Document 6BeajyjjBQ9bED6bK1MVxK93g

PLAINTIFF'S EXHIBIT CAP-1833 HANDBOOK FOR ANALYTICAL QUALITY CONTROL IN WATER AND WASTEWATER LABORATORIES For U.S. ENVIRONMENTAL PROTECTION AGENCY Technology Transfer By ANALYTICAL QUALITY CONTROL LABORATORY National Environmental Research Center Cincinnati, Ohio June 1972 CAPCO JEN 0005351 The mention of trade names or commercial products in this handbook is for illustration purposes, and does not constitute endorsement or recommendation for use by the U. S. En%-iromental Protection Agency, CAPCO JEN 0005352 ABSTRACT One of the fundamental responsibilities of management is the establishment of a continuing ptogram to insure the reliability and validity of analytical laboratory and field data gathered in water treatment and wastewater pollution control activities. This handbook is addressed to laboratory' directors, leaders of field investigations, and other personnel who bear responsibility for water and wastewater data. Subject matter of the handbook is concerned primarily with quality-ewitrol for chemical and physical tests and measurements. Sufficient information is offered to allow the reader to inaugurate, or to reinforce, a program of analytical quality control which will emphasize early recognition, prevention and correction of factors leading to breakdowns in the validity of data. m CAPCO JEN 0005353 ACKNOWLEDGEMENT This handbook was prepared by the Analytical Quality Control Laboratory of the United States Environmental Protection Agency. The contributions of the following individuals in preparing the handbook are gratefully acknowledged. D. G. Ballinger R. L. BootiTM. R. Midgett R. C. Kroner J. F. Kopp J. J. Lichtenberg J. A. Winter R. C. Dressman J. W. Eichelberger J. E. Longbottom inquiries regarding material contained in the handbook should be made to Environmental Protection Agency, National Environmental Research Center, Analytical Quality Control Laboratory, Cincinnati, Ohio 45268. v CAPCO JEN 0005354 TABLE OF CONTENTS Chapter 1 2 ABSTRACT ACKNOWLEDGEMENT TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES ^ IMPORTANCE OF QUALITY CONTROL 1.1 General 1.2 Quality Control Program 1.3 Analytical Methods 1.4 References LABORATORY SERVICES 2.1 General 2.2 Distilled Water 2.3 Ammonia-Free Water 2.4 Carbon Dioxide-Free Water 2.5 Ion-Free Water 2.6 Compressed Air 2.7 Electrical Services 2.8 References 3 INSTRUMENTAL QUALITY CONTROL 3.1 Introduction 3.2 Analytical Balances 3.3 pH Meters 3.4 Conductivity Meters 3.5 Turbidimeters 3.6 Spectrophotometers 3.7 Organic Carbon Analyzer 3.8 Selective Ion Electrodes 3.9 References 4 GLASSWARE 4.1 General 4~2 Types of Glassware 4.3 Volumetric Analyses 4.4 Federal Specifications for Volumetric Glassware 4.5 Cleaning of Glass and Porcelain 4.6 Special Cleaning Requirements 4.7 Dispo.-able Glassware 4.8 Specialized Glassware 4.9 Fritted W'are 4.10 RefereTTces ` iii v vii ix xi 1-1 1-1 1-1 1-2 1-3 2-1 2-1 2-1 2-4 2-4 2-4 2-4 2-5 2-5 3-1 3-1 3-2 3-4 3-7 3-11 3-11 3-23 3-28 3-29 4-1 4-1 4-1 4-2 4-4 4-6 4-7 4-8 4-8 4-9 4-11 vii CAPCO JEN 0005355 5 REAGENTS. SOLVENTS AND GASES 5.1 Introduction 5.2 Reagent Quality 5.3 Elimination of Determinate Errors 5.4 References 5-1 5-1 5-1 5-4 5-7 CONTROL OF ANALYTICAL PERFORMANCE 6.1 Introduction --- 6.2 Precision and Accuracy 6.3 Evaluation of Daily Performance 6.4 Quality Control Charts 6.5 References 6-1 6-1 6-1 6-4 6-5 6-18 DATA HANDLING AND REPORTING 7.1 Introduction 7.2 The Analytical Value 7.3 Report Forms 7.4 STORET-Computerized Storage and Retrieval of Water Quality Data 7.5 SHAVES-A Consolidated Data Reporting and Evaluation System 7.6 References 7-1 7-1 7-1 7-5 7-9 7-11 7-11 SPECIAL REQUIREMENTS FOR TRACE ORGANIC ANALYSIS 3.1 Introduction S.2 Discrete Bottled Samples 8.3 Carbon Adsorption Samples 3.4 Glassware 3.5 Reagents and Chemicals 8.6 Common Analytical Operations 8.7 Gas-Liquid Chromatography 8.8 Qualitative Analysis 8.9 Quantitative Analysis 8.10 Thin-Layer Chromatography 8.1 1 Column Chromatography 8.12 References 8-1 8-1 8-1 8-2 8-2 8-2 8-3 8-4 8-7 8-8 8-9 8-10 8-10 SKILLS AND TRAINING 9.1 General 9.2 Skills 9.3 Training 9-1 9-1 9-2 9-4 VIII CAPCO JEN 0005356 to r j LIST OF FIGURES Figure No. 3-1 Analytical Balance 3-2 pH Meter 3-3 Conductivity Meter 3-4 Turbidimeter 3-5 Spectrophotometer 3-6 Atomic Absorption Unit 3-7 Device for Reproducible Positioning of Burner freight 3-8 Organic Carbon Analyzer 3-9 Selective ion Meter 4-1 Titration Bench 4-2 Example of Markings on Glassware 6-1 Essentials of Control Chart Effects of a and 3 Levels on Standard Control Chart Laboratory Quality Control Charts 7-1 Example of Bench Sheet 7-2 Example of Summary Review Sheet 7-3 Example of STORET Report Form Page 3-3 3-5 3-8 3-12 3-13 3-19 3-22 3-27 3-30 4-3 4-6 6-5 6-8 6-10 7-7 7-8 7-10 IX CAPCO JEN 0005357 LIST OF FIGURES Figure No. 3-1 Analytical Batance 3-2 piI Meter 3-3 Conductivity Meter 3-4 Turbidimeter 3-5 Spectrophotometer 3-6 Atomic Absorption Unit 3-7 Device for Reproducible Positioning of Burner Height 3-8 Organic Carbon Analyzer 3-9 Selective Ion Meter 4-1 Titration Bench 4-2 Example of Markings on Glassware * 6-1 Essentials of Control Chart 6-2 Effects of a and 3 Levels on Standard Control Chart 0-3 Laboratory Quality Control Charts 7-1 Example of Bench Sheet *7^-> Example of Summary Review Sheet 7-3 Example of STORHT Report Form ^ 3-3 3-5 3-8 3-12 3-13 3-19 3-22 3-27 3-30 4-3 4-6 6-5 6-8 6-10 7-7 7-8 7-10 CAPCO JEN 0005358 LIST OF TABLES Table No. Ul W U U IJ IJ -1 Water Purity Comparison of Distillates from Glass and Metal Stills -1 Instruments Commonly Used inWater and^ Wastewater Analysis pH Values of NBS Standards from 0 to 30 C -3 Performance Characteristics of Typical pH Meter -4 Electrical Conductivity of Potassium Chloride Reference Solutions 3-5 Design Features of Some Common Spectrophotometers 3-6 Design Features of Some Common Atomic Absorption Instruments 4-1 Tolerances for Volumetric Glassware --- "T Fritted Ware Porosity 4-3 Cleaning of Filters . 6-1 Precision Data on River Water Samples for Phosphorus AutoAnalyzer Method 6-2 Accuracy Data on River Water Samples for Phosphorus AutoAnalyzer Method 6-3 Factors for Computing Control Chart Lines 9-1 Skill-Time Rating of Standard Analytical Operations Page 2-1 2-2 3-1 3-6 3-7 3-10 3-15 3-25 4-5 4-9 4-10 6-2 6-3 6-1 2 9-3 xi CAPCO JEN 0005359 Chapter 1 IMPORTANCE OF QUALITY CONTROL 1.1 General The role of the analytical laboratory is to provide qualitative and quantitative data to be used in decision making. To be valuable, the data must accurately describe the character istics or the concentration of constituents inj^jg sample submitted to the laboratory. In many cases, an approximate answer or incorrect result is worse than no answer at all, because it will lead to faulty interpretations. Decisions made using water and wastewater data are far-reaching. Water quality standards are set to establish satisfactory conditions for a given water use. The laboratory data define whether that condition is being met, and whether the water can be used for its intended purpose. If the laboratory results indicate a violation of the standard, action is required on the part of pollution control authorities. With the present-emphasis on legal action and social pressures to abate pollution, the analyst should be aware of his,responsibility to provide laboratory results that are a reliable description of the sample. Furthermore, the analyst must be aware that his professional competence, the procedures he has used, and the reported values may be used and challenged in court. To satisfactorily meet this challenge, the laboratory data must be backed up by an adequate program to document the proper control and application of all of the factors which affect the final result. In wastewater analyses, the laboratory data define the treatment plant influent, the status of the steps in the treatment process, and the final load imposed upon the water resources. Decisions on process changes, plant modification, or even the construction of a new facility may be based upon the results of laboratory analyses. The financial implications alone are significant reasons for extreme care in analysis. Research investigations in water pollution control rest upon a firm base of laboratory data. The final result sought can usually be described in numerical terms. The progress of the research and the alternative pathways available are generally evaluated on the basis of laboratory data. The value of the research effort will depend upon the validity of the laboratory results. 1.2 Quality Control Program Because of the importance of laboratory analyses and the resulting actions which they produce, a program to insure the reliability of the data is essential. It is recognized that all analysts practice quality control to varying degrees, depending somewhat upon their train ing, professional pride, and awareness of the importance of the work they are doing. How ever, under tha pressure of daily workload, analytical quality control may be easily neglected. Therefore, an established, routine control program applied to every analytical test is important in assuring the reliability of the final results. The quality control program in the laboratory has two primary functions. First, the program should monitor the reliability (truth) of the results reported. It should continually provide an answer to "How good (true) are the results submitted?" This phase may be termed 'measurement of quality." The second function is the control of quality in order to meet 1-1 CAPCO JEN 0005360 the program requirements for reliabiJity. For example, tlic processing of spiked samples is the measurement of quality, while the use of analytical grade reagents is a control measure. Just as each analytical method has a rigid protocol, so the quality control associated with that test must also involve definite required steps to monitor and assure that the result is correct. The steps in quality control will vary with the type of analysis. For example, in a titration, standardization of the titrant on a frequent basis is an element of quality control. In an instrumental method, the check-out of instrument response and the calibration of the instrument in concentration units is also a quality control function. Ideally, all of the variables which can affect the final answer shotflcTbe considered, evaluated, and controlled. This handbook considers the factors which go into creating an analytical result, and provides recommendations for the control of these factors in order to insure that the best possible answer is obtained. A program based upon these recommendations will give the analyst and his supervisor confidence in the reliability and the representative nature of the sample charactieristics being reported. Without exception, the final responsibility for the reliability*of the analytical results sub mitted rests with the Laboratory Director. 1.3 Analytical Methods In general, the widespread use of an analytical method indicates that it is a reliable means of analysis, and this fact tends to support the validity of the test result reported. Conversely, the use of a little-known technique forces the data user to place faith in the judgement of tiie analyst. When the analyst uses a "private" method, or one not commonly accepted in the field, he must stand alone in defining both his choice of the method and the result obtained. The need for standardization of methods within a single laboratory is readily apparent. Uniform methods between cooperating laboratories are also important in order to remove the methodology as a variable in comparison or joint use of data between laboratories. Uniformity of methods is particularly important when laboratories are providing data to a common data bank, such as STORET*, or when several laboratories are cooperating in joint field surveys. A lack of standardization of methods raises doubts as to the validity of the results reported. If the same constituent is measured by different analytical procedures within a single laboratory, or in several laboratories, the question is raised as to which procedure is superior, and why the superior method is not used throughout. The physical and chemical methods used should be selected by the following criteria: a. The method should measure the desired constituent with precision and accuracy sufficient, to meet the data needs in the presence of the interferences normally encountered in polluted waters. b. The procedure should utilize the equipment and skills normally available in the average water pollution control laboratory. 'STORET is the acronym used lo identify the computer-oriented U.S. Environmental Protection Agency Water Quality Control Information System for STOrago and RKTrieval of data and information. 1-2 CAPCO JEN 0005361 c. The selected methods should he in use in many laboratories or have been sufficiently tested to establish their validity. <1. The method should be sufficiently tapid to permit tontine use for the examination of large numbers of samples. The use of EPA methods in all EPA laboratories provides a common base for combined data between Agency programs. Uniformity throughout EPA lends considerable support to the validity of the results reported by the Agency. Regardless of the analytical method used in the laboratory,the specific methodology should be carefully documented. In some water pollution reports it is customary to state that Standard Methods (1) have been used throughout. Close examination indicates, however, that this is not strictly true. In many laboratories, the standard method has been modified because of recent research or personal preferences of the laboratory staff. In other cases the standard method has been replaced with a better one. Statements concerning the methods used in arriving at laboratory data should be clearly and honestly stated. The methods used should be adequately referenced and the procedures applied exactly as directed. Knowing the specific method which has been used, the reviewer can apply the associated precision anti accuracy of the method when interpreting the laboratory results. If the analytical methodology is in doubt, the data user may honestly inquire as to the reliability of the result he is to interpret. The advantages of strict adherence to accepted methods should not stifle investigations leading to improvements in analytical procedures. In spite of the value of accepted and documented methods, occasions do arise when a procedure must be modified to eliminate unusual interference, or to yield increased sensitivity. When modification is necessary, the :ew>ion should be carefully worked out to accomplish the desired result. It is advisable to assemble data using both the regular and the modified procedure to show the superiority of the latter. This useful information can be brought to the attention of the individuals and groups responsible for methods standardization. For maximum benefit, the modified procedure should be rewritten in the standard format so that the substituted procedure may me used throughout the laboratory for routine examination of samples. Responsibility for the use of a non-standard procedure rests with the analyst and his supervisor, since such use represents a departure from accepted practice. In field operations, the problem of transport of samples to the laboratory, or the need to examine a large number of samples to arrive at gross values will sometimes require the use of rapid field methods yielding approximate answers. Such methods should be used with caution,and with a clear understanding that the results obtained do not compare in relia bility with those obtained using standard laboratory' methods. The fact that "quick and dirty" methods have been used should be noted, and the results should not be reported along with more reliable laboratory-derived analytical information. The data user is entitled to know that approximate values have been obtained for screening purposes only, and that the results do not represent the customary precision and accuracy obtained in the laboratory. 1.4 References standard Method; lor the Examination "I Water ami V.y ican Public Health Association, New York ( l`t"l). .Mil Edition. Amer- 1-3 CAPCO JEN 0005362 CHAPTER 2 LABORATORY SERVICES 2.1 General Quality control of laboratory analyses involves consideration and control of the many variables which affect the production of reliable^ata. The quality of the laboratory services available to the analyst must be included among these variables. An abundant supply of distilled water, free from interferences and other undesirable contaminants, is an absolute necessity. An adequate source of clean, dry, compressed air is needed. Electrical power for routine laboratory use and voltage-regulated sources for delicate electronic instrumentation must be provided. This chapter, therefore, will be devoted to describing methods of maintaining the quality of these services, as used in laboratory operations. 2.2 Distilled Water Distilled or demineralized water is used in the laboratory for dilution, preparation of reagent solutions, and final rinsing of glassware. Ordinary distilled water is usually not pure. It may be contaminated by dissolved gases and by materials leached from the container in which it has been stored. Volatile organics distilled over from the feed water may be present, and non-volatile impurities may occasionally be carried over by the steam, in the form of a spray. The concentration of these contaminants is usually quite small, and distilled water is used for many analyses without further purification. However, it is highly important that the still, storage tank, and any associated piping be carefully selected, installed, and main tained in such a way as to assure minimum contamination. Water purity has been defined in many different ways, but one generally accepted definition states that high purity water is water that has been distilled and/or deionized so that it will have a specific resistance of 500,000 ohms (.2.0 micromhos conductivity) or greater. This definition is satisfactory as a base to work from, but for more critical requirements, the breakdown shown in Table 2-1 has been suggested to express degrees of purity (1). Degree of Purity Pure Very Pure Ultrapure Theoretically Pure Table 2-1 WATER PURITY Maximum Conductivity (micromhos/cm) 10 0.1 0.055 Approximate Concentration of Electrolyte (mg/1) 2-5 0.2-0.5 0.01-0.02 0.00 2-1 CAPCO JEN 0005363 1'roperly designed metal stills from reputable manufacturers offer a convenient and reliable source of distilled water. These stills are usually constructed of copper, brass, and bronze. All surfaces that contact the distillate should be heavily coated with pure tin to prevent metallic contamination. The metal storage tank should be of sturdy construction, have a tight-fitting cover, and a filter in the air vent to remove airborne dust, gases, and fumes. For special purposes, an all-glass distillation unit may be preferable to the metal still. These stills are usually smaller, and of more limited capacity than the metal stills. An actual comparison in which the distillates from an att^lass still and a metal still were analyzed spectrographically for certain trace metal contaminants is shown below in Table 2-2. Table 2-2 COMPARISON OF DISTILLATES FROM GLASS AND METAL STILLS Element and Concentration (Micrograms/1) Source Zn All-Glass Still <1 Metal Still 9 _B_ _Fe 12 1 13 Mn A1 <1 4 <1 <5 Cu 5 11 Ni Pb <1 <2 <2 26 It can be seen that the all-glass still produced a product which had substantially lower contamination from zinc, copper, and lead. Other analyses have indicated the same general relationship, except that a boron concentration of 100 micrograms/liter was found in water from the all-glass still on one occasion. This was probably related to the length of time the distillate had remained in the glass storage reservoir. All stills require periodic cleaning to remove solids which have been deposited from the feed water. Hard water and high dissolved solids content promote scale formation in the evapor ator, and cleaning frequency will thus depend on the quality of the feed water. The boiler of an all-glass still should be drained daily and refilled with clean water. Build-up of scale is easily detected, and the boiler and condenser coils should be cleaned at frequent intervals. Metal stills usually incorporate a constant bleeder device which retards scale formation to some extent. However, these units should still be dismantled and cleaned at regular intervals. Cleaning should always be in accordance with the manufacturer's instructions. Pre-treatment of the incoming feed water will often improve still performance and raise the quality of the distillate. For example, preliminary softening of hard water removes calcium and magnesium prior to distillation. This reduces scale formation in the boiler and conden ser, thereby reducing maintenance service. These softeners employ the ion exchange principle using a sodium chloride cycle, and are relatively inexpensive to operate. A carbon filtration system* installed at the feed water intake, will remove organic materials which might subsequently be carried over in the distillate. If trace concentrations of ions are a major concern, the distillate may be passed through a mixed-bed ion exchanger. At least two commercially manufactured systems are available for production of high purity water by ion exchange. The Millipore Super-Q System (Millipore Corp., Bedford, Mass.), consists essentially of disposable cartridges for prefiltration, organic absorption, deioniza tion and Millipore filtration. The company claims it can produce 10 megohm water, containing no particulate matter larger than 0.45 micron in size, from tap water, at the rate of 20 gallons per hour. Continental Water Conditioning Corp., El Paso, Texas, advertises a 2-2 CAPCO JEN 0005364 system which can be tailored to the needs of the customer. Performance specifications include minimum flow rates of 45 callous per hour and total dissolved solids of less than 0.1 mg/1 when required. Specific conductance is a rapid and simple measurement for determining the inorganic quality of distilled water. Stills of the types previously discussed are capable of producing a distillate with a specific conductance of less than 2.0 micromhos at 250. This is equivalent to 0.5-1.0 mg/1 of ionized material. Frequent checks should be made to determine that optimum performance is being maintained. A jmity meter installed between the still and the storage reservoir will monitor the conductivity of the distillate, in terms of the equiva lent in mg/1 of sodium chloride. If the reading on the meter begins to rise above the present limit of conductivity, effective action should be taken to eliminate the source of contamin ation. Organic quality is more difficult to monitor, but the total organic carbon determin ation is a simple and rapid test for organic contaminants. A piping system for delivering distilled water to the area of use within the laboratory is a convenient and desirable feature. In this ease, special care should be taken that the quality of the water is not degraded between the still and the point of use. Piping may be of tin, tin-lined brass, stainless steel, plastic, or chemically resistant glass, depending on the quality of the water desired and on available funds. Tin is best, but is also very expensive. As a compromise, plastic pipe, or glass pipe with Teflon* O-rings at all connecting joints is satisfactory for most purposes. The glass pipe has an obvious advantage when freedom from trace amounts of organic materials is important. '.VI,en there is no piped-in supply, distilled water will probably be transported to the l.iiuir.iton and 'tored in polyethylene or glass bottles of about 5-gallon capacity. If stored in gi.i.iN containers, distilled water will gradually leach the more soluble materials from the glass and cause an increase in dissolved solids. Therefore, only borosilicate-free glass containers should be used. PoKethylene bottles contain organic plasticizers, and traces of these materials may be leached from the container walls. These are of little consequence, except in some organic analyses. Rubber stoppers often used in storage containers contain teachable materials, including significant quantities of zinc. This is usually no problem, since the water is not in direct contact with the stopper. However, the analyst should be aware of the potential for contamination, especially when the supply is not replenished by frequent use. The delivery tid e may consist of a piece of glass tubing which extends almost to the bottom of the bottle, and which is bent downward above the bottle neck, with a three-to-four-foot piece of flexible tubing attached for mobility. Vinyl tubing is preferable to latex rubber, because it is less leaehable. However, a short piece of latex tubing may be required at the outlet for better control of the pinchcock. The vent tube in the stopper should be protected against the entrance of dust. Ordinary distilled, water is quite adequate for many analyses, including the determination of major cations and anions. Certain needs may require the use of double- or even tripledistilled water. Redistillation from an alkaline permanganate solution can be used to obtain a water with low organic background. When determining trace organics by solvent extrac tion and gas chromatography, distilled water with sufficiently low background may be extremely difficult to obtain. In this case, pre-extraction of the water with the solvent used ' <) / /. ! iir V. 2-3 CAPCO JEN 0005365 Certain analyses require special treatment or conditioning of the distilled water, and these will now be discussed. 2.3 Ammonia-Free Water Removal of ammonia can be accomplished by shaking ordinary distilled water with a strong cation exchanger, or by passing distilled water through a column of such material. For limited volumes of ammonia-free water, use of the Quikpure (Box 254, Chicago, 111.) 500-ml bottle is highly recommended. The ion-free water described below is also suitable for use in the determination of ammonia. 2.4 Carbon-Dioxide-FreeWater Carbon-dioxide-free water may be prepared by boiling distilled water for 15 minutes and cooling to room temperature. As an alternative, distilled water may be vigorously aerated with a stream of inert gas for a period sufficient to achieve saturation and C02 removal. Nitrogen is most frequently used. The final pH of the wateryhould lie between 6.2 and 7.2. It is not advisable to store C02-free water for extended periods. 2.5 Ion-Free Water A multi-purpose high purity water, free from trace amounts of the common ions, may be conveniently prepared by slowly passing distilled water through an ion-exchange column containing one part of a strongly acidic cation-exchange resin in the hydroxyl form. Resins of a quality suitable for analytical work must be used. Ion-exchange cartridges of the research grade, available from scientific supply houses, have been found satisfactory. By using a fresh column and high quality distilled water, a water corresponding to the ASTM designation for referee reagent water (2) (maximum 0.1 mg/1 total matter and maximum conductivity of 0.1 micromho) can be obtained. This water is suitable for use in the determination of ammonia, trace metals, and low concentrations of most cations and anions. It is not suited to some organic analyses, however, because this treatment adds organic contaminants to the water by contact with the ion-exchange materials. 2.6 Compressed Air The quality of compressed air required in the laboratory is usually very high, and special attention should be given to producing and maintaining clean air until it reaches the outlet. Oil, water, and dirt are undesirable contaminants in compressed air, and it is important to install equipment which generates dry, oil-free air. When pressures of less than 50 psi are required, a rotary-type compressor, using a water seal and no oil, eliminates any addition of oil which would subsequently have to be removed from the system. Large, horizontal, water-cooled compressors will usually be used when higher pressures are required. Compression heats air, thus increasing its tendency to retain moisture. An aftercooler is therefore necessary to remove water. Absorption filters should be used at the compressor to prevent moisture from entering the piping system. Galvanized steel pipe with threaded, malleable-iron fittings, or solder-joint copper tubing should be used for piping the air to the laboratory. When the compressed air entering the laboratory' is of low quality, an efficient filter should be installed between the CtTflet and the point of use to trap oil, moisture, and other contaminants. As an alternative, high quality compressed air of the dry grade is commercially available in cylinders when no other source exists. 2-4 CAPCO JEN 0005366 2.7 Electrical Services An adequate electrical system is indispensable to the modern laboratory. This involves having both 115- and 230-volt sources in sufficient capacity for the type of work that must be done. Requirements for satisfactory lighting, proper functioning of sensitive instruments, and operation of high current devices must be considered. Any specialized equipment may present unusual demands on the electrical supply. Due to the special- type of work, requirements for a laboratory lighting system are quite different from those in other areas. Accurate readings of glassware graduations, balance verniers, and other measuring lines must be made/Titration endpoints, sometimes involving subtle changes in color or shading, must be observed. Levels of illumination, brightness, glare, and location of light sources should be controlled to facilitate ease in making these measurements and to provide maximum comfort for the employees. Spectrophotometers, flame photometers, atomic absorption equipment, emission spectro graphs, gas chromatographs, etc. have complicated electronic circuits which require relatively constant voltage to maintain stable, drift-free instrument operation. If the voltage to these circuits varies, there is a resulting change in resistance, temperature, current, efficiency, light output, and component life. These characteristics are interrelated, and one cannot be changed without affecting the others. Voltage regulation is therefore necessary to eliminate these conditions. Many instruments have built-in voltage regulators which perform this function satisfactorily. In the absence of these, a small, portable, constant-voltage transformer should be placed in the circuit between the electrical outlet and the instrument. Such units are available from Sola Basie Industries, Elk Grove Village, 111., and are capable of supplying a constant output of I IS volts from an input which varies between 95 and 130 volts. When requirements are more stringent, special transformer-regulated circuits can be used to supply constant voltage. Only the instrument receiving the regulated voltage should be operated from such a circuit at any given time. These lines are in addition to, and separate from the ordinary circuits used for operation of equipment with less critical requirements. Electrical heating devices provide desirable heat sources, and should offer continuously variable temperature control. Hot plates and muffle furnaces wired for 230-volt current will probably give better service than those which operate on 115 volts, especially if the lower voltage circuit is only marginally adequate. Water baths and laboratory ovens with maximum operating temperatures of about 200C perform well at 115 volts. Care must be taken to ground all equipment which could constitute a shock hazard. The three-pronged plugs which incorporate the ground are best for this purpose. 2.8 References 1. Applebaum, S. B., and Crits, G. J., "Producing High Purity Water." Industrial Water Engineering, Sept./Oct. 1964. 2. I;-6S Book of ASTM Standards, Part 23, Atmospheric Analysis, pp. 225-6, American Society for Testing and Materials, 1916 Race Street, Pliiladelphia, Pa. 19103. 2-5 CAPCO JEN 0005367 CHAPTER 3 INSTRUMENTAL QUALITY CONTROL 3.1 Introduction The modem analytical laboratory depends very heavily upon instrumentation. This statement may be completely obvious, but it should be remembered that the exceptional emphasis on electronic equipment has really begun since the development of the transistor and the computer. Analytical instrumentation to a certain extent, is always in the development stage, with manufacturers continually redesigning and upgrading their products, striving for miniaturization, better durability and sensitivity, and improved automation. The net result to laboratory supervisors and staff members is a bewildering stream of advertising brochures, announcements, and catalogues of newly available equipment. Consequently, the selection and purchase of analytical equipment is, at all times, beset with uncertainty. Table 3-1 lists the instruments most commonly used for water and wastewater analysis. These represent basic equipment used in routine work and should be the subject of careful consideration before purchase. Further, operation and maintenance of these devices ought to be a primary consideration in production of satisfactory data. Obviously, a fundamental understanding of instrument design will assist the analyst in the correct use of the instrument and in some cases aid in detecting instrumental failures. In the pages that follow an attempt is made to discuss basic instrument design and to offer some remarks whenever possible about desirable instrumental features. Table 3-1 INSTRUMENTS COMMONLY USED IN WATER AND WASTEWATER ANALYSIS Analytical Balance Potentiometer (pH meter) Conductivity meter Turbidimeter Spectrophotometers a. Visual b. Ultraviolet c. Infrared d. Atomic absorption Total Carbon Analyzer Gas Chromatographs Miscellaneous a. Temperature devices (ovens, water baths, etc.) b. Recoiders c Selecti\e Ion Electrodes 3-1 CAPCO JEN 0005368 3.2 Analytical Balances The most important piece of equipment in any analytical laboratory is the analytical balance (See Figure 3-1). It bears the same relationship to accuracy of measurements produced by a laboratory as the Greenwich standard clock has to international time-keeping. If the balance is not accurate all data related to weight-prepared standards will contain the same degree of error. The balance, therefore, should be the most protected and cared-for instrument in the laboratory. Unfortunately, care of the balance is frequently overlooked. There are many fine balances on the market designed to meet a variety of needs such as sensitivity, speed weighing, batch weighing, etc. Types of balances include general purpose, micro-, electro-, semi-analytical, analytical and other special purpose instruments. Each type of balance has its own place in the scheme of laboratory operation but the analytical balance is by far the most important in the production of reliable data. Most analytical balances in use today in well equipped laboratories are of the "single pan" variety. Single-pan capacities range from 80 grams to the pbpular 200-gram models with sensitivities from 0.01 to 1 mg. Features of single-pan balances include mechanical lifting and substitution of weights, digital readout of weights, and mechanical zeroing of the empty balance. The advantage of the single-pan balance over the old "two-pan" balance is in greatly increased weighing speed and improved weigliing accuracy because-of mechanical weight handling. With all the design improvements, however, the modern analytical balance is still a fragile instrument, subject to shock, temperature and humidity changes, mishandling and various other insults. Some of the precautions to be observed in maintaining and prolonging the dependable life of a balance are as follows: a. Analytical balances should be mounted on a heavy shock-proof table, preferably one with adequate working surface and a suitable drawer for storage of balance accessories: balance level should be cheeked frequently and adjusted when necessary. b. Balances should be located away from laboratory traffic, protected from sudden drafts and humidity changes. e. Balance temperatures should be equilibrated with room temperature; this is especially important if building heat is shut off or reduced during non-working hours. d. When not in use, the beam should be raised from the knife edges, the weights returned to the beam, objects such as weighing dish removed from the pan, and the slide door closed. e. Special precautions should be taken to avoid spillage of corrosive chemicals on the pan or inside the balance case; the interior of balance housing should be kept scrupulously clean. f. Balance-: should be checked and adjusted periodically by a company service man or balance consultant; if sconce is not available locally, follow the manufacturer's instructions as closeljuas possible. 3-2 CAPCO JEN 0005369 Figure 3-1. ANALYTICAL BALANCE n . ..*73 1 'A * <3 : ?IL S223a28SESS 3-3 CAPCO JEN 0005370 g. The balance should be operated at all times according to the manufacturer's instructions. Standardized weights to be used in checking balance accuracy, and that meet U.S. Bureau of Standards specifications, may be purchased from various supply houses. A very complete set of directions for checking the performance of a balance is contained in Part 30 of ASTM Standards (1). Since all analytical balances of the 200-grain^capacity have about the same specifications with reference to sensitivity, precision, convenience, and price, and since these specifications are suitable for normal weighing requirements in water and wastewater laboratories, it is safe to assume that .there is no clear preference for a certain model, and selection will probably be made on the basis of service availability. 3.3 pH Meters The concept of pH as a means of expressing the degree of effective acidity or alkalinity as contrasted with total acidity or alkalinity was developed by'Sorenson in 1909. Jt was not until about 1940 that commercial instruments were developed for routine laboratory measurement of pH. A basic pH meter (See Figure 3-2) consists of a voltage source, amplifier, and readout device, either scale or digital. Certain additional refinements produce varying performance characteristics between models. Some models incorporate expanded scales for increased readability and solid state circuitry for operating stability and extreme accuracy. All instruments of recent design also include temperature adjustment and slope adjustment to correct for asymmetric potential of glass electrodes. Other features are scales that facilitate u-w of selective ion electrodes, recorder output, and interfacing with complex data handling a -'terns. In routine analytical work, the glass electrode is used as the indicator and the calomel electrode as the reference. Glass electrodes have a very fast response time in highly buffered -olutions. However, accurate readings are obtained slowly in poorly buffered samples, and particularly when changing from buffered to unbuffered samples, as after standardization. Hoctrodes, both glass and calomel, should be well rinsed with distilled water after each reading, and should be rinsed or dipped several times into the next test sample before the final reading is taken. Weakly buffered samples should be stirred during measurement. Glass electrodes should not be allowed to become dry during periods of inactivity. When not in use they should be immersed in distilled water. The first step in standardization of the instrument is done by immersing the glass and calomel electrodes into a buffer of known pH, setting the meter scale or needle to the pH of the buffer and. .adjusting the proper controls to bring the circuit into balance. The temperature compensating dial should be set at the sample temperature. The pH of the standard buffer should be within about two pH units of the sample. For best accuracy, the instrument should be calibrated against two buffers that bracket the pH of the samples. The presence of a faulty electrode is indicated by failure to obtain a reasonably correct .'uv t*>r the pll of the '-econd reference buffer solution after the meter has been .miardi/.ed with, the lirst. A cracked glass electrode will often yield pH readings that are e.'cnti.ilK the same foi both^CTiulards. The response of electrodes may also be impaired by 3-4 CAPCO JEN 0005371 Figure 3-2. pH MUTER 3-5 CAPCO JEN 0005372 failure to maintain the KC1 level in the calomel electrode, or by certain specific materials such as oily substances and precipitates that may coat the surfaces. A faulty condition can be recognized from the check with the two buffer solutions. If either of these conditions should occur the electrode can probably be restored to normal by an appropriate cleaning procedure. Complete and detailed cleaning methods are given in Part 23 of ASTM Standards (2). Because of the asymmetric potential of the glass electrode most pH meters are built with a "slope adjustment" which enables the analyst to correct for slight electrode errors that occur when standardization is performed at two-different pH levels. Exact details of slope adjustment and slope check may vary with different models of instruments. The slope adjustment must be made whenever electrodes are changed, subjected to vigorous cleaning, or refilled with fresh electrolyte. The slope adjustment feature is highly desirable and recommended for consideration when purchasing a new meter. Most pH meters now available are built with transistorized circuits rather than vacuum tubes which greatly reduces warm-up lime and increases stability of the meter. Also, many instruments are designed with a switching circuit so that the conventional 0-14 scale may be used to read a single pH unit. The "expanded-scaie" feature allows for more accurate readings and may be of definite value when the meter is used for potentiometric titrations, it is of dubious value, however, in routine analytical work, since readings more percise than t 0.1 pH are seldom required. Solid state circuitry has led to improved design of compact instruments suitable for field work. Field-type instruments are generally battery-powered, and require more maintenance and more frequent standardization than laboratory instruments. Standard buffer solutions, covering a range of pH, may be purchased from almost any chemical supply house and are completely satisfactory for routine use. Table 3-2 below gives a list of NBS buffers (easily made in the laboratory) and the resulting pH at several different temperatures. Table 3-2 pH VALUES OF NBS STANDARDS FROM 0 - 30C 0.05M Temp. Potassium C Tetroxalate 0 1.67-*10 1.67 15 1.67 20 1.68 25 1.68 30 1.69 Potassium Acid Tar trate (Sat. at 25C) ., -- -- -- 3.56 3.55 0.05 M Potassium Acid Phthalate 4.01 4.00 4.00 4.00 4.01 4.01 0.025 M Potassium Dihydrogen Phosphate + 0.025 M Sodium Dihydrogen Phosphate 6.98 6.92 6.90 6.88 6.86 6.85 0.01 M Sodium Tetra borate 9.46 9.33 9.27 9.22 9.18 9.14 3-6 CAPCO JEN 0005373 Sonic idea of the effect of temperature on pH may be obtained by observing temperature vs pH of various buffers shown in the table. Theoretically, the potential response of the electrode system changes 0.20 mV per pH unit per degree centigrade. Since all pH meters measure potential but read out in pH, a variable compensation is used. A rough rule of thumb is that temperature compensation is about 0.05 pH units per 5 degree increase in temperature. Typical performance data of a conventional expanded scale pH meter is shown in Table 3-3 below. Table 3-3 PERFORMANCE CHARACTERISTICS OF TYPICAL pH METER Range Smallest scale division Accuracy Reproducibility Temp, compensation Input impedance Power requirements Dimensions Normal Scale 0 to 14 pH 1400 mv 0.1 pH 10 mv 0.05 pH 5 mv 0.02 pH 2 mv 0 to 100C (manual or automatic) >I014 115/220V, 50/60 Hz \2'A" w x 11" d x 8'A" h Expanded Scale 1 pH 100 mv 0.005 pH 0.5 mv 0.002 pH 2% of read in 0.002 pH 0.2 mv > 1013 3.4 Conductivity Meters Solutions of electrolytes conduct an electric current by the migration of ions under the influence of an electric field. For a constant applied EMF, the current flowing between opposing electrpdes immersed in the electrolyte will vary inversely with the resistance of the solution. The'reciprocal of the resistance is called conductance and is expressed in reciprocal ohms or mhos. For natural water samples where the resistance is high, the usual reporting unit is in micromhos. Figure 3-3 shows a typical conductivity meter. The passage of direct current through an electrolyte causes changes in the electrolyte, hence, to prevent polarization, it is necessary to use an alternating current or current pulses of short duration when measuring conductivity. Originally, conductivity meters were l uil: using a rapidly alternating current of low intensity in the audio range, and were equipped 3-7 CAPCO JEN 0005374 ) i I i 3-8 CAPCO JEN 0005375 Figure 3-3. CONDUCTIVITY METER with Malice. For obvious reasons, the earpiece lias been replaced by other devices, such as the cathode ray tube commonly known as the "magic eye". Practically all conductivity meters on the present market use some variation of the "magic eye" for indicating solution conductivity, and include a stepping switch for varying resistances in steps of I OX. The instruments are therefore capable of reading conductivities from about 0.) mhos to about -50.000 /* mhos. The sensing element for a conductivity measurement is the conductivity cell, which normally consists of two thin plates of platinized metal, rigidly supported with a very precise parallel spacing. For protection, the plate* are mounted inside a glass tube, with openings in the side walls and submersible end for access of sample. Variations in designs have included use of hard rubber and plastics for protection of the cell plates. Glass may be preferable, in that the plates may be visually observed for cleanliness and possible damage, but the more durable encasements have the advantage of greater protection and reduced cell breakage. One manufacturer offers a cell containing circular carbon plates embedded in an epoxv-type plastic. Reversing the usual procedures, the sample is poured into the cell. The cell is particularly attractive because of its ruggedness and the fact that it can be cleaned without changing the cell factor. V Special precaution is taken with the arrangement of lead wires from the plates. Since stray electroK tic and capacitive current will pass between them, thereby distorting the bridge balance, the wires must be properly separated, usually by use of a non-conducting bead or collar on one of the lead wires. In uiatine use. eells should be frequently examined to insure that (a) platinized coating of plaie' is intact. <b> plates are not coated with suspended matter, tc) plates are not bent, d'-t.'itcd. or misaligned, and id) lead wiles are properly spaced. I emperaiure has a pronounced effect on the conductance of solutions, and must be c<>r!acted, when results are reported. The specified temperature for reporting data used by most analytical groups land all FRA laboratories) is 25''C. Data correction may be accomplished by adjusting sample temperatures to 25"C..or by use of mathematical or electronic adjustment. Adjustment of sample temperature is the preferred system, because ot the empirical nature of the mathematical correction. However, investigative work in the Anah tical Quality Control Laboratory, and at the upper Ohio Basin Office. FRA. has shown that acceptable data is obtained if conductivity readings taken at stream-side are electronically temperature-corrected to 25"C. Instrumental troubles are seldom encountered with conductivity meters because of the design simplicity. When troubles occur they are usually in the cell, and for most accurate work the following procedures should be used: a. Standardize the cell and establish a cell factor by measuring the conductivity of a standard potassium chloride solution (standard conductivity tables may be found in various handbooks). b. Rinse the cell by repeated immersion in distilled water. \g.iin. immcisc the cell in the sample several times before obtaining a reading, d. If the meter is equipped with a "magic eye", determine the maximum width of the 3-9 CAPCO JEN 0005376 shadow at least twice, by approaching the endpoint from a low reading upward, and from a high reading downward. Because the cell constants are subject to slow change, even under ideal conditions, and sometimes to more rapid change under adverse conditions, it is recommended that the cell constant be periodically established. Table 3-4, below, can be used for this operation. Tabled ELECTRICAL CONDUCTIVITY OF POTASSIUM CHLORIDE REFERENCE SOLUTIONS Solution Normality Method of Preparation Temp. (C) Conductivity (/* mhos) . A 0.1 7.4365 g KCf/1 at 20C B 0.01 0.7440 g KC1/1 at 20 C C 0.001 Dilute 100 ml of B to 1 1 at 20 C 0 18 ' 25 0 18 25 25 7,138 11,167 12,856 773 1,220 1,408 147 For instruments reading in mhos, calculate the cell constant as follows: where K, + K, L = ---------- 1--------- 1.000.000 x Kx L = cell constant Kj = Conductivity, in p mhos/cm, of the KC1 solution at the temperature of measurement K-, = conductivity, in mhos, at the same temperature, of the distilled water used to prepare the reference solution K = measured conductance, in mhos Conductivity equipment which has generally been found to be reliable for laboratory work includes the YS1 #31 (Yellow Springs Instrument Co., Box 279, Yellow Springs, Ohio 45387), the Lab-Line Mark IV #1100 and MC3 #1 1025 (Lab-Line Instruments Inc., 15th & Bloomingdale Ayes., Melrose Park, III. 60160), and Industrial Instruments #RC 16B2 and RC-18 (Industrial Instruments Inc., 89 Commerce Road, Cedar Grove, N.J. 07009). The YSI #31 is particularly suitable for routine lab work because of the brilliance of the magic eye. The Lab-Line meters provide the sturdy cell previously mentioned. The Industrial Instruments #RC-1S is designed for extreme accuracy but the number of dial settings required to obtain a single leading do not recommend it for routine analysis. A icceni instrument survey conducted by the Methods and Performance Activity of the Analytical Quality Conirol-fedboraiory showed that a preponderance of EPA Laboratories used the Industrial Instruments RC I6B2. 3-10 CAPCO JEN 0005377 3.5 Turbidimeters Instruments for the measurement of turbidity have traditionally employed principles of design related to transmission or reflectance of light. The luck of a primary standard for turbidity, however, has resulted in a complete absence of uniformity among the available instruments. Further, the Jackson Candle Turbidimeter, which does not depend upon the use of a primary standard, is a primitive instrument, subject to many interferences, and the measurements generally are not reproducible. Recent investigations (3) have resulted in tha-design of an instrument which has been adopted by the Environmental Protection Agency as a standard. The specifications for the instrument are described elsewhere (4). Presently, the Hach Turbidimeter, Model 2100, (Box 907, Ames, Iowa 50010) is the only turbidity measuring device manufactured which meets these specifications. Figure 3-4 shows this instrument. The Hach Turbidimeter Model 2100 employs, for standardization, a suspension of formazin, especially used because of its stability and uniform particle size. For calibration purposes, the formazin is permanently embedded in a cylinder of luciteyfhe cylinder duplicating the size and shape of the sample cuvette. Although the instrument is designed with a series of scales ranging 0-1, 0-10, 0-100 and 0-1000 JCU, it has been recommended that turbidity readings in excess of 40 JCU be rejected (3). Correct use of the meter therefore requires that samples containing turbidities in excess of 40 JCU be diluted to a value below this level and the results multiplied by the proper dilution factor. For production of data with maximum accuracy and precision the following precautions should be observed: a. Protect the lucite standard from nicks, scratches and fingerprints. b. Use a constant orientation of the lucite standard while calibrating the instrument. c. Use a well mixed sample in the sample cuvette; do not take readings until finely dispersed bubbles have disappeared. d. Dilute samples containing excess turbidity to some value below 40 JCU; take reading, and multiply results by correct dilution factor. 3.G Spectrophotometers Since a large portion of routine quantitative measurements are performed colorimetrically, the spectrophotometer (See Figure 3-5) is usually the workhorse of any analytical laboratory. Indeed; the versatility of the instrument, and the number of demands imposed upon it, have resulted in a large variety of designs and price ranges. A systematic listing and detailed discussion of all instrumental types would be beyond the scope of this chapter. As a matter of convenience and practicality, therefore, spectrophotometers are discussed separately as visible, ultraviolet, infrared, and atomic absorption instruments. A spectrophotometer is an instrument for measuring an amount of light or radiant energy transmitted through a solution,.as a function of wave-length. A spectrophotometer differs 3-11 CAPCO JEN 0005378 Figure 3-4. TURBIDIMETER V t i i i 3-12 CAPCO JEN 0005379 \ 3-13 CAPCO JEN 0005380 Figure 3-5. SPECTROPHOTOMETER from a filter photometer in that it uses continuously variable, and more nearly monochromatic, bands of light. Filter photometers are relatively insensitive, and lack the versatility of spectrophotometers. They are used most profitably where a single method can be designed to fit the instrument. The essential parts of a spectrophotometer are: a. A source of radiant energy, usually a tungsten filament bulb, b. The Monochromator, a device for isolating narrow bands of light, c. Cells (cuvettes), for holding the colored solution under investigation, and d. The photodetector, a device to detect and measure the radiant energy passing through the sample solution. v Each of the essential features listed, especially the monochromator and the photodetector system, vary in design principles from one instrument to another. Table 3-5 shows some of the features of the more commonly used spectrophotometers. 3.6.1. Visible Range Desirable features on a visible-range spectrophotometer are determined by the anticipated use of the instrument. Simple, limited programs requiring use of only a few parameters at gross concentrations, can probably be supported by an inexpensive,but reliable instrument, such as the B&L Spectronic 20. (See Table 3-5.) On the other hand, if a laboratory program requires a wide variety of measurements on diverse samples at very' low concentrations, a more versatile instrument may be needed. One of the prime considerations would be adaptability to various cell sizes, at least from 1.0 to 5.0 cm. Many spectrophotometers now available are satisfactory for water quality analyses. As shown in Table 3-5, which lists only a few of the available spectrophotometers, higher priced instruments strive for versatility, including interchangeable sources, detectors, and cells. Complete information on instrument specifications can be found in the publication Industrial Research (Beverly Shores, Ind. 46301), Nov. 20, 1969. 3.6.2. Ultraviolet Range An ultraviolet spectrophotometer is similar in design to a visual range instrument, differences being in the light source and the optics. The light source is a hydrogen or deuterium discharge lamp which emits radiations in the UV portion of the spectrum,* generally from about 200 mp to the low visible region. The optical system, if of the prism type, must be constructed of UV-transparant material, usually quartz. Sample cells must also be constructed of quartz or other UV-transparent material. If a grating system is used in an UV system, the grating may be specially cut (blazed) in the UV region for greater sensitivity. A number of UV spectrophotometers are available: the Beckman DU, the Hitachi, the Bausch and Lomb, the Cary, the Hilger and the Leeds & Northmp. 3 3. inhered Range A number of instrumental changes are required in the construction of spectrophotometers 3-14 CAPCO JEN 0005381 DESIGN FE A TU R E S OF SOME COMMON SPECTROPHOTOMETERS Hydro, Lamp V I Marketed as I 'erkin-CImer $139 o r Coleman $139-001 .sn s 2 3y 3a u C y tcyo tCco/1 C ^E pc oo 2rr: r i to* _n .5 3 e8 Oa ZJ J3 .*72u3 > 3 o . U0> 43y-* _ c E- Q <<c/)2vc5^LQ o o o to CoJ f; P p, lH -u E O t3o '3trt s'a o*0 EC3 to?tZ; C--N JO3 Jy3 o o SI c- ytcrot JrvO> ^rt t&o Q2 a. cl E 4 J2 oCON Jo=>) oS r* 1 -40 t *3^. a5 s"J S 2 2o p y O co d== --u oo o.E -- E o E-* SS u" wCJ co yu y s Q o' u E-^ o o<N O ocoj to' uo. cT t> 5 LrOi to o ,, '/> O y y cy ci ri d o S > coro3 k Zi y s oO cv3 J E j= Z2 O w> s=s 8 C wCc oo ?Sr oo '.vv5) Vi ><. CD ryc 1y> o^ C3 -On u--* ^w *Ouo _02 ooa> 3t-o os *5 yyj .-32-49 oCcJ y 03 2 *9 to* yy to *S- CCJ c o. *'- *6 3 2y -3 3s y yV to r-i E .2 r^y s: p 'w 5^ OE Ey cy SI o *p o -ouy: 25a vo 5 > >* ho: a > o o o o Col - tyo n; yA oo o zz "7ou3. "S 2 o nto J3 ol a O o o to d --* E 1. . d o. p- o 3H 2 *5* yy rT rt *3 G < CL > E) E73 c o . C33 w. O C Vcsi o yo c .c JL1 o ^o to 3 o ~2 -3 roi ci 2 _ 3 y O to o o -J 73 tSl ^ tJr O y u5= ^ y- is ^S 73 C "3a *O- y2 So* t V. P HG O CO CO > Cl y n CO sc: < ! i lo 1 CAPCO JEN 0005382 for measurements in the infrared (IR) region. Modifications are needed,because optical materials, such as glass and quartz, absorb radiant energy in the IR region and ordinary photocells do not respond. Most IR spectrophotometers use front-surfaced mirrors to eliminate the necessity for radiant energy to pass through quartz, glass, or other lens materials. The mirrors are usually parabolic to facilitate gathering the diffuse IR energy. Instruments must be protected from high humidities and water vapor to avoid deterioration of the optical system, and also to avoid extraneous absorption bands in the IR. The energy or light source for an IR instrument may be a Nernst glower or a Globar. Either source has certain characteristics that recommend it for use, but the Globar is more commonly used because it has a more stable emission and it is more rugged. Receiving or detection units may be a thermocouple, a bolometer, thermistor, or a photoconductor cell. Of all the IR instruments available, two are in relatively common use in EPA laboratories. The Perkin-Elmer Infracord is a low cost model designed for routine work with simplicity of operation. The Perkin Elmer #621 (723-G Main Ave., Norwalk^Gonn., 06852) is a far more sophisticated instrument, designed for basic research. 3.6.4 Proper Use of Spectrophotometers The manufacturer's instructions for proper use should be followed in all cases. Several safeguards against misuse of the instruments, however, are mandatory. Instruments should be checked for wavelength alignment. If a particular colored solution is to be used at a closely specified wavelength, considerable loss of sensitivity can be encountered if the wavelength control is misaligned. In visual instruments, an excellent reference point is the maximum absorption for a dilute solution of potassium permanganate, which has a dual peak at 526 mg and 546m/<. On inexpensive grating instruments, which possess less resolution than the prism instruments, the permanganate peak appears at 525 to 550 mt` as a single Hat-topped spike. For both UV and IR instruments, standard absorption curves for many organic materials have been published so that reference material for standard peaks is easily available. Standard films of styrene and other transparent plastics are available for IR wavelength checks. Although most instruments contain built-in transformers for stabilization of the electronic circuits, an exterior, high capacity, constant-voltage transformer is recommended for general laboratory control. A number of controllcd-voltage outlets in the laboratory are especially desirable in industrial areas or in buildings containing heavy, electrically operated equipment where voltage surges on adjacent power lines are apt to be frequent. An unstable voltage is frequently indicated by a nickering needle on the meter. The flickering behavior may be intermittent or it may occur at certain times of the day when heavy machinery in the area may be starting or stopping. Too much emphasis cannot be placed on care of absorption cells. All cells should be kept scrupulously clean, free of scratches, fingerprints, smudges and evaporated film residues. Matched cells should be checked to see that they are equivalent by placing portions of the 3-17 CAPCO JEN 0005383 same solution in both colls and taking several readings of the %T or OD values. If a cell is mismatched it should be discarded or reserved for rough work. (Directions for cleaning cells are detailed in Chapter 4). Generally speaking, trained technicians may operate any of the spectrophotometers successfully. However, interpretation of data from both the UV and IR instruments becomes increasingly complex, and requires more training and specialization. IR interpreta tion requires special training, and because of the special techniques of sample preparation, instrument operation, and interpretation of absorption curves, mere compliance with the operations manual is not sufficient. 3.6.5, Atomic Absorption There are a number of differences in the basic design and accessories for atomic absorption equipment that require consideration before purchase and during subsequent use. These choices concern the light source, nebulizer burners, optical systems, readout devices, and mode conversion. Some of these choices are not readily -pbvious, and require that the purchaser or user be familiar with the types and numbers of samples to be anal;, zed and the specific elements to be measured before a choice is made. For a program analyzing a wize variety of samples for a number of elements at varying concentrations, an instrument of maximum versatility would be required. A typical atomic absorption unit is shown in Figure 3-6. 3.6.5.1 Lamp Mounts A basic design feature of atomic absorption spectrometers is the convenience of the hollow cathode (HC) lamp changeover system. Some instruments provide for as many as six lamps in a rotating turret, all electronically stabilized and ready for use by simply rotating the lamp turret. Other instruments provide for use of only one lamp at a time in the lamp housing, and require manual removal and replacement whenever more than one element is to be measured. A quick changeover system is desirable, especially if a number of lamp changes are needed during a period of operation. Conversely, if lamp changes are infrequent, multi-lamp mounts do not represent a great convenience. For optimal use of the instrument, certain precautions should be observed. After the proper lamp has been selected the hollow cathode curr:nt should be adjusted according to the manufacturer's recommendations and allowed tc electronically stabilize (warm up) before use. This requires approximately 15 minutes. During this period, the monochromator may be positioned at the correct wavelength and the proper slit width selected. For those instruments employing a multi-lamp turret, a warm-up current is provided to those lamps not in use, thereby minimizing the warm-up period when the turret is rotated. In a single-lamp instrument, the instability exhibited during warm-up is minimized by the use of a double-beam'optical system. 3.6.5.1.1 Single Element and Multi-Element Lamps As an adjunct to single-lamp mounts, HC lamps using from two to as many as six elements in combination are available, thereby increasing the versatility of the AA spectrometer. For instance, a single-lamp instrument such as the Perkin-lZlmer may analyze for elements with only one lamp suhstitution .^whereas a six-lamp turret such as the Jarrell-Ash (590 Lincoln Street, Waltham, Mass. 02154), using multi-element lamps could run 18 or more elements. 3-18 CAPCO JEN 0005384 3-19 CAPCO JEN 0005385 Figure 3-6. ATOMIC ABSORPTION UNIT Multi-element lamps, are considerably cheaper per element than single-element lamps, but the savings may not realized if the lamps are not used strategically, because all the elements in the cathode deteriorate when the lamp is used, regardless of which element is measured. The deterioration phenomena results from the different volatilities of metals used in the cathode. One metal volatilizes (sputters) more rapidly than the others and redeposits upon the cathode causing an increase in surface area of that metal, and decreasing the exposed area of the other cathode metals. Thus, with continual use, a drift in signal will be noted with at least one metal increasing and the other (or others) decreasing. If one can ignore the dubious cost savings of multi-element lamps, use of single-element lamps would result in more precise and accurate data. The individual line intensities of an element in a multi-element HC lamp will usually be less than that of a lamp containing a pure cathode of the same element. This is because each clement must now share the discharge energy with all other elements present. However, this reduction should not affect the output by a factor of more than 1/2 to 1/6, depending on the combination and number of elements combined. The output can be even greater in some multi-element lamps because alloying may permit a higher operating current than the pure cathode itself. All HC lamps have life expectancies which are related to the volatility of the cathode metal, and for this reason, the manufacturer's recommendations for amperage at which the lamp is operated should be closely followed. A recent advance in HC lamp design, the high-intensity lamp, promises increased sensitivity for some elements. It is also predicted that the newly designed lamps will be used in atomic fluorescence techniques, with significant gains in sensitivity for metals analysis. Recent improvements in design and manufacture of hollow cathode lamps have resulted in lamps with more constant output and a longer life. Under normal conditions a HC lamp may be expected to operate satisfactorily for several years. At one time, hollow cathode lamps were guaranteed for a minimum amp-hour period. This has been changed, however, to a 90-day warranty. It is good practice to date newly purchased lamps and inspect immediately upon receipt. Operating current and voltage will be indicated on the lamp and should not be exceeded during use. An increase in background noise and/or a loss of sensitivity are signs of lamp deterioration. 3.6.5.2 Burner Types The most difficult and inefficient step in the AA process is converting the metal in the sample from an ion or a molecule to the neutral atomic state. It is the function of the atomizer and the burner to produce the desired neutral atomic condition of the elements. With minor modifications burners are the same as those used for flame photometry. Basically there are two different types of burners. They are the total-consumption or surface-mix bumet, and the laminar-flow or pre-mix burner. There are many variations of these two basic types, such as the Boling, the high-solids, the turbulent-flow, the tri-flamc, nitrous-oxide burner and many others. As one might expect, there are many similarities among the various burners, the different names resulting from the different manufacturers. The element being determined and the type of sample solution dictate the type of burner to be used. .neraily. all types and makes of burners can be adjusted laterally, rotationally and 3-20 CAPCO JEN 0005386 vertically for selection of the most sensitive absorbing area of the flame for the specific I element sought. The vertical adjustment is probably the most important since the position of greatest sensitivity varies from element to element. Burner height is of utmost importance in changing from one element to another. Certain instruments are provided with a vernier adjustment for reproducing burner-height settings, but many are not. Figure 3-7 shows a simply designed device used in the Analytical Quality Control Laboratory for reproducing exact burner height. The gauge is positioned on the burner and the height determined from the lighjj.ieam striking the calibrated scale. The point at which the beam strikes the gauge is recorded for future use. 3.G.5.3 Single-Beam and Double- (Split) Beam Instruments There is a great deal of existing uncertainty among instrument users about the relative merits of single-beam and double-beam instruments. Neither system is the final answer. With a single-beam instrument the light beam from the source^-passes directly through the flame to the detector. In a double-beam system the light from the source is divided by a beam splitter into two paths. One path, the reference beam, goes directly to the detector. The second path, the sample beam, goes through the flame to the detector. The chopper alternately reflects and passes each beam, creating two equal beams falling alternately upon the detector. If the beams are equal they cancel the alternate impulses reaching the detector and no signal is generated. If the beams are different, the resulting imbalance causes the detector to generate an a.c. signal which is amplified and measured. Any difference between the reference and sample beam is measured as a direct function of absorbed light. The advantage of the double-beam design, therefore, is that any variations in the source are of reduced importance, and smaller dependence is placed upon the stability of the power supply. However, stabilization of the power supply can eliminate the apparent need for the split-beam system. Further, a beam splitter requires use of additional mirrors or optical accessories that cause some loss of radiant energy. Neither system, however, compensates for variation in flame intensity. A single-beam system does not monitor source variations but offers certain other advantages. It allows use of low-intensity lamps, smaller slit settings and smaller gain. Asa consequence, the single-beam instrument, properly designed, is capable of operating with lower noise, better signal-to-noise ratio and therefore better precision and improved sensitivity. Because the simplified optical system conserves radiant energy, especially in the shorter wavelengths, it facilitates operation in the low wavelength range. With this advantage, it should be possible to obtain better sensitivity for those elements with strong resonance lines below 350 m^ and even those slightly below 300 mfx. 3.6.5.4. Readout Devices Early models of the AA instrument offered only a meter, calibrated in percentage absorption. In the surge of competitive design, more sophisticated readout devices were built into or offered as accessories to various models. At the present time any desired readout method may be obtained with almost any instrument. Less expensive designs still provide meters with conventional needle indicators. More costly instruments offer any combination of built-in digital scalers, calibrated in concentration, external digital printout ::: concentration, typewriter printon; or typewriter with punch tape. Even ir.cxpenche in-truments are built with recurtftr interfacin'.:. 3-21 CAPCO JEN 0005387 P-HHirc 3-7 Di;\ IC'Er FOrrftEPRODUGIBLE POSITIONING OF BURNER HEIGHT CAPCO JEN 0005388 Choice of a readout system is predicated largely upon laboratory needs and availability of budget. In general, any step toward complete automation is desirable but the degree of automation should be compatible with the hboratory program. 3.6.5.5 Miscellaneous Accessories A number of instruments contain a mode selector, making an instrument usable for either flame absorption or flame emission. The conversion to flame emission is a desirable feature since certain elements are more amenable to analysis by this method. Automatic sample changers are offered for almost all instruments on the market, and as has been previously stated, any automation feature is desirable. However, unless a laboratory program performs a large number of repetitious measurements daily, an automatic sample changer would not be required. As a practical measure, other commonly used sample changing devices not expressly designed for AA use, can easily be interfaced with almost any AA instrument. 3.6.5.B Instrument Choice Table 3-6 summarizes some of the design features of various commercially available instruments. Since many of the features are common to most, models, the basic choice appears to be between double-beam and single-beam instruments. On this basis, at this writing, it is probable that the most widely used instrument in the EPA is the Perkin-Elmer 303. one of two double-beam instruments on the market. The Instrumentation Laboratories Model 153(113 Hartwell Avc.. Lexington. Mass. 02173) has a number of desirable features. In addition to features listed in the table, it includes push-button ignition, push-button wavelength scan, fail-safe solenoids to prevent improper Hame settings, fail-safe flame monitoring, absorption integrator at selected time intervals, visual "peaking" meters, curve correction and several other features. The new Perkin-Elmer Model 403 also offers most of these features in addition to very sophisticated data handling accessories. 3.7 Organic Carbon Analyzer A number of devices designed to measure the organic content of aqueous samples have appeared on the market within the last five years. The oldest, or first of these instruments, is the Dow Beckman Carbonaceous Analyzer (Figure 3-8). The apparatus measures organic carbon as carbon, by oxidizing a very small sample at a temperature of 900, in a stream of oxygen, converting all organically bound carbon to carbon dioxide, which is then measured by a Golay-type thermal detector. The instrument is able to detect about 20 mg. of carbon. A recent modijlcation of the original Dow Beckman Carbonaceous Analyzer employs a dual-combustion-tube system operating at different temperatures to distinguish between total carbon and inorganic carbon. Organic carbon is found by difference. Another such instrument, also developed by Dow Chemical Company, is marketed by the Fisher Scientific Company (Instrument Division, 711 Forbes Ave., Pittsburgh, Pa. 15219) as the "Aquarator". The principle of operation is similar to the Dow Beckman Carbonaceous \-.:h/.er. except that carbon dioxide fed through the combustion tube is reduced to carbon ;!---:o\:dc and measured in wrlnfrared detector. The reaction does not measure carbon, per m\ but reducing materials. The results can be equated with the COD test, but the instrument 3-23 CAPCO JEN 0005389 is best suited for industrial wastes and sewage. It lacks the sensitivity needed for COD measurements on relatively clean waters. The lower limit of measurement is about 40 mg/1 COD. A third instrument developed by Dow researchers is franchised to Ionics Corporation (165 Grove Street, Watertown, Mass. 02 i 72) and sold as the "TOD Analyzer". This instrument is also similar in principle to the carbon analyzer, except that the device measures oxygen in and out of the combustion tube. Depletion of oxygen is correlated with oxygen demand of the sample. However, the reaction of oxygen with sulfur and nitrogen is not stoichiometric, and results for some samples may be questionjWe. A malfunction of the sample injection system appears to be the main problem with the instrument to date. An evaluation study conducted by the Hudson Delaware Basins Office, EPA, indicates a very good potential use for the instrument. Another device, designed by Union Carbide Corporation (Ionics Inc., 65 Grove St., Watertown, Mass. 02172) is similar to the original Dow Beckman Analyzer except that the combustion tube contains a coil of heated palladium wire. An aqueous sample injected into the combustion tube is decomposed to hydrogen and oxygen -through the catalytic action of the palladium. Carbon from organic matter combines with the water-produced oxygen to form carbon dioxide, with a final infrared measurement. Results obtained are theoretically identical to the Dow Beckman Analyzer, but the repeated failure of the sample injection system has prevented evaluation of the apparatus as a laboratory instrument. It is being routinely used, however, as a continuous monitoring system for total carbon. To date the only instrument which has been successfully demonstrated and which meets the needs of EPA laboratories is the Dow Beckman Carbonaceous Analyzer or its successor, the Model 915. Operating instructions foi this device are detailed elsewhere (4). Precautions to be observed in use of the Dow Beckman Carbonaceous Analyzer or the DB Model 915 are detailed below: a. AH inorganic carbon present in the sample as carbonate or bicarbonate must be removed prior to analysis, or accounted for, when using the Model 915. b. Particulate matter in the sample should not be larger than the opening of the hypodermic needle used to take the sample. If necessary the sample should be homogenized by some means designed to reduce the particle size. c. The instrument should be preconditioned with repeated injections of distilled water to obtain stable operating conditions before actual sample injection. d. Furnace temperature and oxygen flow should be maintained at proper settings during period of operation. e. Perform periodic maintenance of instrument as follows: 1. repack combustion tube with fresh asbestos, 2. clean and dry micro filter, 3. clean and dry infrared cell, 4. recharge reference cel! wifii nitrogen. seed for replacement of asbestos packing is indicated by a loss of sensitivity and wide 3-24 CAPCO JEN 0005390 Paper punch t;ipe with typewriter to produce a typed report, push button operation plug-in electronics Ul-SIGN FHATURI-S OF SOMli COMMON A TOM 2 elements siimtl- 5-p;iss system taneously using interference filters. push button operation ai oT . c- f-- 52 73 o wJ O o to 60 3 c/5 O D 7U3. to "3 C 73 o H. 5 Vo5 JO 3 <5 .V 5 or^s K *r, 'A 73 3 9-Sd. 6C0 ur C/J r3 o c 73 Co ~a. co c7O3 jj' .wtnc5or-, cocr:L! to : 5 o zi cr. .ri`-E c. o ti) 6c0 3*3oQ --* tt--o v. iaD/>. C3 c o Wc CL O fc VW 3 s H Q jo 2 .52 56 t--f ^ O <A Oo -c-3 Ca , y tJ If. l5 1 o " .2 i'5 oooc oCN {3 I-- U .a=; c."3 a3 *3 > = J < Hdo c o s: > ^733 *Orj ^O --O S5 5rj o o t>- *- c i2 to 5 o ? o o r- . r- oc .2 6 co ^-- O--' yrvdT. 3<. rC-- ^rs ^o COo --3o. o6c0 c oo CO o 3 5 5 o '5 2Q o-- o cr- C- r; i/1) r > .2 oo oO >t> b--* u *O-- ~ u 'J ^ ^ 'J O1-, cZZ +-> l-- 5 -2 > 8 c oo CO 6CN ? ** c o 22 .3 *U 2 5 C < cO F? R -I ~.X c>/:. *_J "d 2 -- y; < Spcciiil CAPCO JEN 0005391 Figure 3-8. ORGANIC' CARBON ANALYZER I I \ ( \ * .~..JLV',-.* - - - * T-j5L ..* **> 7 :sJv:-6'3^SB r' - ".S^ r- > % i */ >> . --liv'-'-.* . :iHi ; .. . V: --vl; ' * 9n f J > '. V., .'\W-,,. iC:' "..V^ S' ' - ' iy>,-.9r\ v^55^i ; ,' *7y\* ;, i\ \ \*i.;* _jt* -4>rv' - nvi *--<, ---Avt<t,%`J;*.,Vi" '; * ,,.a- 5*- 3-27 CAPCO JEN 0005392 peaks. Unsteady baseline and excessive noise is caused by a dirty or partly clogged micro filter. An accumulation of moisture in the infrared cell is indicated by loss of sensitivity and excessive noise. 3.8 Selective Ion Electrodes In recent years a variety of ion-selective electrodes have been commercially available, and show great promise as fast and efficient tools for jm situ monitoring and for laboratory analysis of all types of samples. A list of ions'susceptible to analysis is indicative of the interest and progress being made in this field. Electrodes, or "probes" as they are popularly called, are available for measurement of monovalent cations, sulfate, nitrate, perchlorate, and a number of others. Dissolved oxygen probes should also be included in this list, although they are not technically selective ion probes. There have been a number of reviews concerning the theory and application of probes, most recently a concise article by Rechnitz (5). Selective ion probes generally measure what they claim to measure-specific ion activity. They do not measure concentrations of un-ionized materials. For example, the probe designed to measure divalent cations promises to measure total hardness as a function of calcium and magnesium. However, since the probe does not respond to un-ionized calcium and magnesium, it does not accurately measure total hardness. As a consequence of this deficiency, much investigation is being carried out to devise means of determining total concentration of the constituent sought. 'Methods for Chemical Analysis of Water and Wastes" (4) outlined procedures for use of two probes, fluoride and DO. The chloride probe is also used on the automatic monitoring devices and will be listed as an approved procedure in future editions of EPA Methods. Evaluation of additional ion selective electrodes for cyanide, ammonium, and sulfide ions is now underway, and will be reported in the near future. Various techniques for use of the probes are reviewed by Riseman (6). Personnel in water pollution laboratories are encouraged to investigate the use of selective ion probes as a means of reducing analytical work and improving data quality. At the present time, however, only the fluoride, DO, and chloride probes are recommended for routine use in data collection. A basic question relating to the use of selective ion probes is the number of standards required to prepare a standard curve. It is generally agreed that the more standards used for the preparation of a colorimetric curve, the more reliable the data resulting from use of the curve. On the other hand, only one or two points are normally used in standardization of a meter for measuring pH, conductivity or DO. In the conversion of a colorimetric procedure such as fluoride to a probe-type measurement, the tendency is to prepare a millivolt vs concentration curve using the usual six to ten standards. Probe manufacturers insist that only one or two points are needed since the linearity of response has been established and only the slope of the line must be known. The alternatives are: a. Take readings in distilled water at 0 and at some concentration approximating concentration of sample; establish slope of line. b. Take readings in distiiled water at 0 and two concentrations bracketing expected concentration of unknowns to establish curve or slope. 3-28 CAPCO JEN 0005393 c. Tiikc readings at 0 and at decade concentrations as I. 10, 100, etc., for standard curve. d. Take reading of sample, add known increment of measured constituent and read again: establish proportionality factor. The method for the use of the fluoride electrode specifies use of multiple standards in the range between 0 to 2 mg/liter, because this system has supplied very precise data when compared to the colorimetric methods using -ttTC same set of standards. The system of incremental addition appears to have considerable merit since the electrode response is established in the presence of possible interferences. At the present time no single procedure for standardizing probe response has been adopted by a majority of users. When a selective ion electrode appears to be malfunctioning, the same check system may be used as for a faulty glass pH electrode. It is unlikely, however, that the electrode will be cracked; it will probably be dry, or insufficiently filled with the necessary solution. The piobe assembly and instructions for refilling customarily accompany the item when shipped in- tiie manufacturer and said instructions should be followed by the user. Selective ion probes are available from several manufacturers including Beekman. Corning, Coleman and Orion. The Orion organization (11 Blackstone St.. Cambridge, Mass. 02139) is the largest producer in the field and offers not only selective ion probes but a sizable complement of electronic equipment for use with the probes. Figure 3-9 illustrates an Orion 'elective ion meter. Diwolved oxygen probes and meters of various designs have been offered by a large number oi manufacturers including Weston & Stack. Beekman. Jarrell-Ash, Union Carbide. Yellow Springs Instrument. Delta Scientific, and others. 3.9 References I 1968 Book of ASTM Standards, Part 30: Testing Single Arm Balances, pp. 1071-84: American Society for Testing and Materials. 1916 Race Street, Philadelphia, Pa. 19103. 2. 1968 Book of ASTM Standards. Part 23: Test for pH of Industrial Waste Water, pp. 292-3: American Society for Testing and Materials. 1916 Race Street, Philadelphia. Pa. '19103. 3. Black, A. P., and Hannah, S. A.. "Measurement of Low Turbidities," JAWWA, 57, 901 (1965). 4 4. Methods for Chemical Analysis of Water and Wastes, EPA, Analytical Quality Control Laboratory, 1971. 5. Rechnitz. G. A., "ton Selective Electrodes." Chemical Engineering News. p. 146, June 12. 1967. *' Riseman. Jean M., "Measurement of Inoiganic Water Pollutants by Specific Ion Electrode." American Laboratory, p. 32. July 1969. 3-29 CAPCO JEN 0005394 V 3-30 CAPCO JEN 0005395 Figure 3-9. SELECTIVE ION METER CHAPTER 4 GLASSWARE 4.1 General The measurement of trace constituents in water demands methods capable of maximum sensitivity. This is especially true for metals and trace organics such as pesticides, as well as for the determination of ammonia and phosgjumis. In addition to sensitive methods, however, there are other areas that require special consideration. One such area is that of the cleanliness of laboratory glassware. Obviously, the very sensitive analytical systems are more sensitive to errors resulting from the improper use or choice of apparatus, as well as to contamination effects due to an improper method of cleaning the apparatus. The purpose of this chapter is to discuss the kinds of glassware available, the use of volumetric ware, and various cleaning requirements. 4.2 Types of Glassware Laboratory vessels serve three functions: storage of reagents, measurement of solution volumes and confinement of reactions. For special purposes, vessels made from materials such as porcelain, nickel, iron, aluminum, platinum, stainless steel, and plastic may be employed to advantage. Glass, however, is the most widely used material of construction. There are many grades and types of glassware from which to choose, ranging from student grade to others possessing specific properties such as resistance to thermal shock, alkali, low boron content, and super strength. Soft-glass containers are usually relatively soluble, and therefore are not recommended for general use. especially for storage of reagents. The mainstay of the modern analytical laboratory' is a highly resistant horosilicute glass, such as that manufactured by Corning Glass Works under the name "Pyrex" or by Kimble Glass Co. as "Kimax". This glassware is satisfactory for all analyses included in "Methods for Chemical Analysis of Water and Wastes (I Depending on the particular manufacturer, various trade names are used for specific brands possessing special properties such as resistance to heat, shock, alkalies, etc. Examples of some of these follow: a. Kimax- or Pyrex-brand glass is a relatively inert all-purpose borosilicate glass. b. Vycor-brand glass is a silica glass (96'?) made to withstand continuous temperatures up to 900C and can be down-shocked in ice water without breakage. c. Coming brand glass is claimed to be 50 times more resistant to alkalies than conventional ware and practically boron free (max. 0.2%). d. Ray-Sorb- or Low-Actinic-brand glass is for use with light-sensitive material. e. Corex-brand labware is harder than conventional borosilicates, better able to resist clouding and scratching. The use of plastic vessels, containeis and other apparatus made of Teflon, polyethylene, polystyrene and pulyproplylcne has increased markedly over recent years. Some of these materials, such as Teflon, quite expensive; however. Teflon stopcock plugs have 4-1 CAPCO JEN 0005396 practically replaced glass plugs in burets, separatory funnels, etc. because lubrication to avoid sticking or "freezing" is not required. Polyprolylene, a methylpentene polymer, is available as laboratory bottles, graduates, beakers and even volumetric flasks. It is crystal clear, shatter-proof, autoclavable and chemically resistant. Some points to consider in choosing glassware and/or plasticware are: a. Generally, the special types of glass listed above are not required to perform the analyses given in "Methods for Chemical^Analysis of Water and Wastes" (1). b. Unless instructed otherwise, borosilicate or polyethylene bottles are to be used for the storage of reagents and standard solutions. c. Certain dilute metal solutions may plate out on glass container walls over long periods of storage. Thus, dilute metal standard solutions are prepared fresh at the time of analysis. d. For some operations, disposable glassware is entirely satisfactory. One example is the use of disposable test tubes as sample containers for use with the Technicon Automatic Sampler. e. Plastic bottles of polyethylene and/or Teflon have been found satisfactory for the shipment of water samples. Strong mineral acids (such as sulfuric acid) and organic solvents will readily attack polyethylene and are to be avoided. f. Borosilicate glassware is not completely inert, particularly to alkalies; therefore, standard solutions of silica, boron and the alkali metals are usually stored in polyethylene bottles. For additional information the reader is referred to the catalogs of the various glass and plastic manufacturers. These catalogs contain a wealth of information as to specific properties, uses, sizes, etc. 4.3 Volumetric Analyses By common usage, accurately calibrated glassware for precise measurements of volume has become known as volumetric glassware. This group includes volumetric flasks, volumetric pipets and accurately calibrated burets. Less accurate types of glassware including graduated cylinders, serological and measuring pipets also have specific uses in the analytical laboratory, when exact volumes are unnecessary. A typical laboratory glassware setup is shown in Figure 4-1. The precision of.volumetric work depends in part upon the accuracy with which volumes of solutions can be measured and there are certain sources of error which must be carefully considered. The volumetric apparatus must be read correctly; that is, the bottom of the meniscus should be tangent to the calibration mark. There are other sources of error, however, such as changes in temperature which result in changes in the actual capacity of glass apparatus and in the volume of the solutions.The capacity of an ordinary glass flask of i ('00 ml volume increases 0.025 ml per degree rise in temperature, but if made of borosilicate glass the increase is much less. One thousand ml of water or of most 0.1 N solutions increases in volumes by approximately 0.20 ml per 1C increase at room 4-2 CAPCO JEN 0005397 I 4-3 CAPCO JEN 0005398 Figure 4-1. TITRATION BENCH temperature. Thus solutions must be measured at the temperature at which the apparatus was calibrated. This temperature (usually 20C) will be indicated on all volumetric ware. There may also be errors of calibration of the apparatus; that is, the volume marked on the apparatus may not be the true volume. Such errors can be eliminated only by recalibrating the apparatus or by replacing it. Volumetric apparatus is calibrated "to contain" or "to deliver" a definite volume of liquid. This will be indicated on the apparatus with the letters "TC" (to contain) or "TD" (to deliver). Volumetric flasks are calibrated to contain a given volume. They are available in various shapes and sizes ranging from 1-to 2000-ml capacity. Volumetric pipets are calibrated to deliver a fixed volume. The usual capacities are 1 thru 100 ml although micro-pipets are also available. In emptying volumetric pipets, they should be held in a vertical position and the outflow should be unrestricted. The tip of the pipet is kept in contact with the wall of the receiving vessel for a second or two after the free flow has stopped. The liquid remaining in the tip is not removed; this is most important. * Measuring and serological pipets should also be held in a vertical position for dispensing liquids; however, the tip of the pipet is only touched to the wet surface of the receiving vessel after the outflow has ceased. For those pipets where the small .amount of liquid remaining in the tip is to be blown out and added, indication is made by a frosted band near the top. The band is usually located far enough down so that it will not touch the technician's lips when liquid is being drawn up or blown out. Burets are used to deliver definite volumes. The more common types are usually of 25- or 50-ml capacity, graduated to tenths of a milliliter, and are provided with stopcocks. For precise analytical methods in microchemistry, micro-burets are also used. Micro-burets generally are of 5- or 10-ml capacity, graduated in hundredths of a milliliter division. Automatic burets with reservoirs are also available ranging in capacity from 10 to 100 ml. Reservoir capacity ranges from 100 to 4000 ml. General rules in regard to the manipulation of a buret are as follows; Do not attempt to dry a buret which has been cleaned for use, but rinse it two or three times with a small volume of the solution with which it is to be filled. Do not allow alkaline solutions to stand in a buret, because the glass will be attacked, and the stopcock, unless made of Teflon, will tend to freeze. A 50-ml buret should not be emptied faster than 0.7 ml per second, otherwise too much liquid will adhere to the walls and as the solution drains down, the meniscus will gradually rise, giving a high false reading. It should be emphasized that improper use of and/or reading of burets can result in serious calculation errors. In the case of all apparatus for delivering liquids, the glass must be absolutely clean so that the film of liquid never breaks at any point. Careful attention must be paid to this fact or the required amou'nt of solution will not be delivered. The various cleaning agents and their use are described later. 4.4 Federal Specifications for Volumetric Glassware Circular 602 of the National Bureau of Standards, "Testing of Glass Volumetric Apparatus", describes the Federal Specifications for volumetric glassware. The National Bureau of Standards no longer accepts stock quantities of volumetric apparatus from manufacturers or dealers for certification and reTBrn for future sale to consumers. This certification service is 4-4 CAPCO JEN 0005399 still available, but apparatus will be tested only when submitted by the ultimate user, and then only after an agreement has been reached with the Bureau concerning the work to be done. Consequently, the various glass manufacturers have discontinued the listing of NBS-certified ware. In its place catalogue listings of volumetric glass apparatus which meet the Federal Specifications are designated as Class A and all such glassware is permanently marked with a large "A". These NBS specifications are listed inJ2ab!e 4-1. The ware in question includes the usual burets, volumetric flasks and volumetric pipets. Table 4-1 TOLERANCES FOR VOLUMETRIC GLASSWARE (Abridged from National Bureau of Standards Data, 1941) Capacity (ml) less than and including Limit of error (ml) Graduated Flasks 25 50 100 200 250 300 500 1.000 2,000 0.03 0.05 0.08 0.10 0.11 0.12 0.15 0.30 0.50 Transfer pipets > 0.006 5 0.01 10 0.02 25 0.025 30 0.03 50 0.05 100 0.08 200 0.10 Burets1 5 0.01 10 0.02 30 0.03 50 0.05 100 0.10 i /.units of error era of total or partial capacity. Customary practice is to test the capacity at 5 intends. 4-5 CAPCO JEN 0005400 In addition to the "A" marking found on calibrated glassware and the temperature at which the calibration was made, other markings also appear. These include the type of glass, such as Pyrex, Corex, Kimax, etc., the stock number of the particular item, and the capacity of the vessel. If the vessel contains a ground-glass connection, this will also be included along with the TD orTC symbol. An example of the markings usually found on volumetric ware is shown in Figure 4-2. PYREX glass co. KIMAX USA *------------------ type -------------------*" USA STANDARD ( )i- - TAPER SIZE A <----------------- . CLASS 500 ml 1 0.20 ml <~ - VO LOME TC 20 C * NO 5680 - TO CONTAIN STOCK NO. -> A 500 ml > TC 20 C NO 28013 Figure 4-2. EXAMPLE OF MARKINGS ON GLASSWARE C'I.iss A glassware need not be recalibrated before use. However, if it should become necessary to calibrate a particular piece of glassware, directions may be found in texts (2) on quantitative analysis. 4.5 Cleaning of Glass and Porcelain The method of cleaning should be adapted to both the substances that are to be removed, and the determination to be performed. Water-soluble substances are simply washed out with hot or cold water, and the vessel is finally rinsed with successive small amounts of distilled water. Other substances more difficult to remove may require the use of a detergent, organic solvent, dichromate cleaning solution, nitric acid or aqua regia (25 percent v/v cone. HN03 in cone. HC1). In all cases it is good practice to rinse a vessel with tap water as soon as possible after use. Material allowed to dry on glassware is much more difficult to remove. Volumetric glassware, especially burets, may be thoroughly cleaned by a mixture containing the following: 30 g sodium hydroxide, 4 g sodium hexametaphosphate (trade name, Caigon), 8 g trisodium phosphate, and 1 liter water. A gram or two of sodium lauryl sulfate or other surfactant will improve its action in some cases. This solution should be used with a buret brush. liivhromate cleaning solution (chromic acid) is a powerful cleaning agent: however, due to iis destructive nature upon clothing and upon laboratory furniture, extreme care must be taken when using this mixture. If any of the solution is spilled, it must be cleaned up 4-6 CAPCO JEN 0005401 immediately. Chromic acid solution may be prepared in the laboratory by adding 1 liter of concentrated sulfuric acid slowly, with stirring, to 35 ml saturated sodium dichromate solution. This mixture must be allowed to stand for approximately 15 minutes in the vessel which is being cleaned and may then be returned to a storage bottle, following the chromic-acid wash, the vessels are rinsed thoroughly with tap water, then with small successive portions of distilled water. Fuming nitric acid acts more rapidly, but is disagreeable to handle. In either case, when the acid becomes dilute, the cleaning mixture is no longer effective. A mixture of concentrated sulfuric and fuming nitric acids is even more efficient but is also hazardous to use. A persistent greasy layer or spot may be removed by acetone or by allowing a warm solution of sodhrflThydroxidc, about 1 g per 50 ml of water, to stand in the vessel for 10-15 minutes; after rinsing with water, dilute hydrochloric acid, and water again, the vessel is usually clean. Alcoholic potassium hydroxide is also effective in removing grease. To dry glass apparatus, rinse with acetone and blow or draw air through it. 4.6 Special Cleaning Requirements Absorption cells, used in spectrophotometers, should be kept scrupulously clean, free of scratches, fingerprints, smudges and evaporated film residues. The ceils may be cleaned with detergent solutions for removal of organic residues, but should not be soaked for prolonged periods in caustic solutions because of the possibility of etching. Organic solvents may be used to rinse cells in which organic materials have been used. Nitric acid rinses are permissible, but dichromate solutions are not recommended because of the adsorptive properties of dichromate on glass. Rinsing and drying of cells with alcohol or acetone before storage is a preferred practice. Matched cells should be checked to see that they are equivalent by placing portions of the same solution in both cells and taking several readings of the transmittance (T'T) or optical density (OD) values. If a cell is mismatched it should be discarded or reserved for rough work. For certain determinations, especially trace metals, the glassware should also be rinsed with a 1:1 nitric acid-water mixture. This operation is followed by thoroughly rinsing with tap water and successive portions of distilled water. This may require as many as 12-15 rinses, especially if chromium is being determined. The nitric acid rinse is also especially important if lead is being determined. Glassware to be used for phosphate determinations should not be washed with detergents containing phosphates. This glassware must be thoroughly rinsed with tap water and distilled water. For ammonia and Kjeldahl nitrogen, the glassware must be rinsed with ammonia-free water (See Chapter 2). Glassware to be used in the determination of trace organic constituents in water, such as chlorinated pesticides, should be as free as possible of organic contaminants. A chromic acid wash of at least (5 minutes is necessary to destroy these organic residues. Rinse thoroughly with tap water,"and finally with distilled water. Glassware may be dried for immediate use by rinsing with redistilled acetone. Otherwise glassware may be oven dried or drip dried. Glassware should be stored immediately after drying to prevent any accumulation of dust. Store inverted or with mouth of glassware covered with foil. Rott'es to be used for the collection of samples for organic analyses should be rinsed successively with chromic acid cleaning solution, tap water, distilled water, and finally .wvcral times with red istilledj^p]vent (e.g., acetone, hexane, petroleum ether, chloroform). 4-7 CAPCO JEN 00054-02 Caps are washed with detergent, rinsed with tap water, distilled water and solvent. Liners are treated in the same way as the bottles and are stored in a sealed container. 4.7 Disposable Glassware When the risk of washing a pipet for reuse becomes too great, as in the case of use with toxic materials, or when the cost of washing glassware becomes prohibitive, disposable pipets may be the answer, provided they meet the necessary specification. Various types are available including bacteriological, serological nTTQ micro-dilution pipets. Disposable glass ware generally is made of soft glass. 4.8 Specialized Glassware The use of vessels and glassware fitted with standard-taper, ground-glass, and ball-and-socket joints has increased because of certain advantages such as less leakage and fewer freezeups. Standard-taper, interchangeable ground joints save time and tro_uble in assembling apparatus. They are precision-ground with tested abrasives to insure an 'accurate fit and freedom from leakage. Ball and socket joints increase flexibility of operation and eliminate the need for exact alignments of apparatus. Symbols and their meaning as applied to standard joints, stoppers and stopcocks are shown below. 4.8.1. Standard Taper (J) | is the symbol used to designate interchangeable joints, stoppers and stopcocks, complying with the requirements of Commercial Standard CS-21, published by the National Bureau of Standards. All mating parts are finished to a 1:10 taper. The size of a particular piece appears after the appropriate symbol. Due primarily to the greater variety of apparatus equipped with ] fittings, a number of different types of identifications are used, as follows: a. For joints-a two-part number, as J 24/40, with 24 being the approximate diameter in mm at the large end of the taper, and 40 the axial length of taper, also in mm. b. For stopcocks--a single number, as J 2, with 2 mmbeing the approximate diameter of the hole or holes through the plug. . c. For bottles-a single number, as J 19, with 19 mm being the appropriate diameter at top of neck. However, there are differences in dimensions between the bottle and flask stoppers. d. For flaslcs, etc. - a single number, as J 19, with 19 mm being the appropriate diameterof the opening at top of neck. 4.8.2 Spherical Joints (J) J is the designation for spherical (semi-ball) joints complying with CS-21. The complete designation of a spherical joint also consists of a two-part number, as 12/2. with 12 being the approximate diameter of the ball and 2 the bore of the ball and the socket, also in mm. 4-8 CAPCO JEN 00054-03 4.8.3 Product Standard () is a new symbol. It will appear in a forthcoming NBS Product Standard for stopcocks with Teflon plugs, with the mating surfaces being finished to a 1:5 taper. As with J stopcocks, a single number is used. Thus. 2 means a Teflon stopcock with a hole of approximately 2-mm diameter in the plug. 4.9 Fritted Ware For certain laboratory operations the use of Fritted Ware for filtration (as jn total dissolved solids and suspended solids determinations), gas dispersion, absorption, and/or extractions may be of an advantage. There are six different porosities of Fritted Ware available, so that precipitates varying in size can be filtered at maximum speed with no sacrifice or retentivity. Porosity is controlled in manufacture, and discs are individually tested and graded into these classifications. The extra-coarse and coarse porosities are held toward the maximum-pore diameter as listed. The medium, fine, very fine, and ultra-fine are held toward the minimum pore diameter as listed' in Table 4-2. Porosity Grade Txtr.i Coarse Coarse Medium 'Fine Very Fine Ultra-Fine Table 4-2 FRITTFD WARt POROSITY Designation 1IC c M F VF UF Pore Size (Microns) 170-220 40-60 10-15 4-5.5 2-2.5 0.9-1.4 Principal Uses Coarse filtration. Gas dispersion, washing, absorption. Coarse filtration. Gas dispersion, washing, absorption. Filtration and extraction. Filtration and extraction. General bacterial filtration. General bacterial filtration. aie de:c; mined h\ the meih".! MVc.fied in . i M I: I2N. md l'vi;ne.il'iiit\ nf Rigid I'oroti- I liters for 1 ,ibu v<ii live." (5) urn Pore 4-0 CAPCO JEN 0005404 4.9.1 Recommended Procedures for Maximum Filter Life a. New Filters. Wash new Filters by suction with hot hydrochloric acid, followed by a water rinse. b. Pressure Limits. The maximum safe differential pressure on a disc is 15 pounds per square inch. c. Thermal Shock. Fritted ware has less resistance to thermal shock than non-porous glassware. Hence, excessive, rapid temperature changes and direct exposure to a flame should be avoided. Heating in a furnace to 500C may be done safely, provided the heating and cooling are gradual. Dry ware may be brought to constant weight by heating at 105-110C. Never subject a damp filter of ultra-fine porosity to a sudden temperature change. Steam produced in the interior may cause cracking. T' 4.9.2 Cleaning of Used Filters In many cases, precipitates can be removed by rinsing with water, passed through from the underside, with the pressure not exceeding 15 pounds per square inch. The suggestions that follow in Table 4-3 will be helpful in dealing with material that will not be removed by the reverse water-wash. The use of strong alkalies, strong hydrofluoric acid and phosphoric acid should be avoided. Also, scratching of the surfaces will weaken the discs. Material Albumen Aluminous and siliceous residues Copper or iron oxides Fatty materials Mercuric sulfide Organic matter Silver chloiide Table 4-3 CLEANING OF FILTERS Removal Agent Hot ammonia or hydrochloric acid 20 Hydrofluoric acid followed by con centrated sulfuric acid. Rinse immediately with water until no trace of acid can be detected. Hot hydrochloric acid plus potassium chlorate. Carbon tetrachloride Hot aqua regia Hot concentrated cleaning solution, or hot concentrated sulfuric acid with a few drops of sodium nitrite. Ammonium or sodium hyposulfite. 4-10 CAPCO JEN 0005405 10 References "Methods foi Chemical Analysis of Water and Wastes," LPA, Analytical Quality Control Laboratory. 19 ~ 1. Wilton*, II. H.. and Furman, N. H., Elementary Quantitative Analysis-Theory and Practice, D. Van Nostrand Co., Inc., New York (1947). 1968 Book of ASTM Standards, Part 30, American Society for Testing and Materials, 1916 Race Street. Philadelphia, Pa. 19103.'**- 4-11 CAPCO JEN 0005406 Chapter 5 REAGENTS, SOLVENTS, AND GASES 5.1 Introduction The objective of this chapter is to provide general information and suggestions that will serve to keep the analyst conscious of his responsibilities in analytical quality control, as they relate to reagents, solvents and gases. While the material presented here will assist the analyst in producing high quality data, it is by no means complete. It is incumbent on the analyst to obtain details of special precautions required to insure proper selection, preparation, and storage of reagents, solvents and gases from the descriptions of individual methods. 5.2 Reagent Quality --- Chemical reagents, solvents, and gases are available in a wide variety of grades of purity, ranging from technical grade to various "ultra pure" grades. The purity of these materials required in analytical chemistry varies with the type of analysis. The parameter being measured and the sensitivity and specificity of the detection system are important factors in determining the purity of the reagents required. For many analyses, e.g., most inorganic analyses, analytical reagent grade is satisfactory. Other analyses, e.g., trace organic and radiological, frequently require special "ultra pure" reagents, solvents, and gases. In methods where the purity of reagents is not specified it is intended that analytical reagent grade be used. Reagents of lesser purity than that specified by the method should not be used. The labels on the container should be checked and the contents examined to verify that the purity of the reagents meets the needs of the particular method involved. The quality of reagents, solvents, and gases required for the various classes of analyses: inorganic, metals, radiological, and organic, are discussed below. Reagents must always be prepared and standardized with the utmost of care and technique, against reliable primary standards. They must be restandardized or prepared fresh as often as required by their stability. Stock and working standard solutions must be checked regularly for signs of deterioration, e.g., discoloration, formation of precipitates, and concentration. Standard solutions should be properly labeled as to compound, concentration, solvent, date, and preparer. Primary standards must be obtained from a reliable source, pretreated, e.g., dried, under specified conditions, accurately prepared in calibrated volumetric glassware, and stored in containers that will not alter the reagent. A large number of primary standards are available from the National Bureau of Standards (NBS). A complete listing of available standards is given in NBS Special Publication 260 (1). Primary standards may also be obtained from many chemical supply companies. Suppliers for special quality reagents, solvents, and gases are noted in later discussions of the various classes of analyses. Reagents and solvents of ail grades are available from many chemical supply houses. There is some confusion among chemists as to the definition of the terms ANALYTICAL RF.AGENT GRADE. REAGENT GRADE, and ACS ANALYTICAL REAGENT GRADE. A review of the literature and jjigmieal supply catalogs indicates that the three terms are synonymous. Hereafter, in this document, the term ANALYTICAL REAGENT GRADE 5-1 CAPCO JEN 0005407 (AR) will be used. It is intended that AR chemicals and solvents shall conform to the current specifications of the Committee on Analytical Reagents of the American Chemical Society (2). The ASTM Manual on Industrial Water and Industrial Waste Water (3), Part 23 of ASTM Standards (4), and "Standard Methods for the Examination of Water and Wastewater" (5) devote separate chapters to problems related to preparation, standardization, and storage of reagents. The information provided therein is particularly appropriate to inorganic determinations. The type of volumetric glasswamjto be used, the effect of certain reagents on glassware, the effect of temperature on volumetric measurements, purity of reagents, absorption of gases and water vapor from the air, standardization of solutions, instability, and need for frequent standardization of certain reagents are among the topics discussed. It is recommended that the analyst become thoroughly familiar with these publications. 5.2.1 General Inorganic Analyses In general, AR-grade reagents and solvents are satisfactory foe inorganic analyses. Primary standard reagents must, of course, be used for standardizing all volumetric solutions. Commercially prepared reagents and standard solutions are very convenient and may be used when it is demonstrated that they meet the method requirements. Ail prepared reagents must be checked for accuracy. The individual methods specify the reagents that require frequent standardization, or other special treatment, and the analyst must follow through with these essential operations. To avoid waste, the analyst should prepare a limited volume of such reagents, depending on the quantity required over a given period of time. Examples and brief discussions of the kind of problems that occur are given under Paragraph 5.3. "Elimination of Determinate Errors". -\s far as possible, distilled water used for preparation of reagent solutions must be free of measurable amounts of the constituent to be determined. Special requirements for distilled v.ater are given in Chapter 3 of this manual and in individual method descriptions. Compressed gases, such as oxygen and nitrogen, used for total organic carbon determination may be of commercial grade. 5.2.2 Metals Analyses All standards used for atomic absorption and emission spectroscopy should be of spectroquality. It is recommended that other reagents and solvents also be of spectroquality, although AR grade is sometimes satisfactory. Standards may be prepared by tlie analyst in the laboratory or prepared, spectrographically standardized materials may be purchased commercially. Standards required for determination of metals in water are not generally available, from the National Bureau of Standards. Analytical reagent grade nitric and hydrochloric acids must be specially prepared by distillation in borosilicate glass and diluted with deionized distilled water. All other reagents and standards are also prepared in deionized water. general. fuel and oxidant gases u>-ed for atomic absorption can be of commercial grade. \u supplied by an ordinary laboratory compressor is quite satisfactory, if adequate pressure i' maintained and nocessary^Tbeautions are taken to filter oil, water, and possible trace 5-2 CAPCO JEN 0005408 metals from the line. For certain determinations, e.g., aluminum, reagent-grade nitrous oxide is required. 5.2.3 Radiological Analyses The great sensitivity of radioactive counting instruments requires that scintillation grade reagents and solvents, or equivalent, be used for all radioactivity determinations. Some of the reagents, for example, strontium carbonate and yttrium oxide carriers used for the determination of strontium 90 and yttrium 90, must be stable, that is free of radioactivity. Barium sulfate, used for coprecipitation of raditim' must be free from all traces of radium. These reagents and solvents are commercially available from chemical supply houses. Calibrated standard sources of specific radioactive materials with known count and date of counting are available from various suppliers. No single company supplies all standards. Gases used for radioactive counting must be of high purity and extra dry. Gases such as helium and air are aged for about 30 days to allow radioactive, background to decay. All gases are checked for background before use. Some cylinders c'ontain inherent radioactivity which is imparted to the gas. When this background is above normal, the gas should not be used for radioactivity determinations. 5.2.4 Organic Chemical Analyses The minimum purity of reagents and solvents that can be used for organic analyses is AR grade. Reference grade standards should be used whenever available. Special note should be taken of the assay of standard materials. Owing to the great sensitivity (nanogram and subnanogram quantities) of gas chromatography (GC), which is often used to quantitate organic results, much greater purity is frequently required (6). The specificity of some GC detectors requires that reagents and solvents be free of certain classes of compounds. For example, analyses by electron capture require that reagents and solvents be free of electronegative materials that would interfere with the determination of specific compounds in the sample. Similarly, use of the flame photometric detector requires that reagents and solvents be free from sulfur and/or phosphorus interference. Pesticide quality solvents are available from several sources. These are often satisfactory for many organic GC determinations. However, the contents of each container must be checked to assure its suitability for the analyses. Similarly, all analytical reagents and other chemicals must also be checked routinely. The. quality of gases required for GC determinations varies somewhat with the type of detector. In general, the compressed gases are a prepurified dry grade. Grade A helium from the U.S. Bureau of Mines has always been satisfactory. The Dohrmann nitrogen-detection system requires the use of ultra-pure hydrogen for satisfactory results. The use of molecular-sieve, carrier-gas filters and drying tubes is required on combustion gases. They are recommended-for use on all other gases. It is recommended that the analyst familiarize himself with an article by Burke (7) on practical aspects of gas chromatography. AH reagents, solvents, and adsorbents used for thin-layer chromatography must be checked to be certain that there are no impurities present that will react with the chromogenic reagent or otherwise interfere with subsequent qualitative or quantitative determinations. Glass-backed layers prepared in the laboratory or precoated layers supplied by a manufacturer may be used. Jipwever, precoated layers are more difficult to scrape. 5-3 CAPCO JEN 0005409 Therefore, it is recommended that layers prepared in the laboratory be used when zones are to be scraped in order to recover isolated compounds. Plastic-backed layers are generally unsatisfactory for this type of analysis. Adsorbents most commonly used for column chromatographic clean-up of sample extracts are Florisil, silica gel, and alumina. These must be pre-activated according to the method specifications and checked for interfering constituents. 5.3 Elimination of Determinate Errors In order to produce high quality analytical data, determinate errors must be eliminated or at least minimized. For purposes of this discussion, we assume that a competent analyst and reliable equipment, in optimum operating condition, are available. Thus, determinate errors that might result from an inexperienced or careless analyst and poor equipment are eliminated. The remaining sources of error are the reagents, solvents, and gases that are used throughout the analyses. The quality of these materials, even though they are AR grade or better, may vary from one source to another, from one lot tp.another, and even within the same lot. Therefore, the analyst must predetermine that all of these materials are free of interfering substances under the conditions of the analyses. To do this he must have a regular check program. Materials that do not meet requirements are replaced or purified so that they can be used. 5.3.1 Reagent Blank The first step the analyst must take is to determine the background or blank of each of the reagents and solvents used in a given method of analysis. .The conditions for determining the blank must be identical to those used throughout the analysis, including the detection system. If the reagents and solvents contain substances that interfere with a particular determination, satisfactory reagents and solvents must be found. Where possible and practical, they should be treated so that they can be used. 5.3.2 Method Blank After determining the individual reagent or solvent blanks, the analyst must determine the method blank to see if the cumulative blank interferes with the analyses. The method blank is determined by following the procedure step by step, including all of the reagents and solvents, in the quantity required by the method. If the cumulative blank interferes with the determination, steps must be taken to eliminate or reduce the interference to a level that will permit this combination of solvents and reagents to be used. If the blank cannot be eliminated, the magnitude of the interference must be considered when calculating the concentration of specific constitutents in the samples being analyzed. A method blank should be determined whenever an analysis is made. The number of blanks to be run is determined by the method of analysis and the number of samples being analyzed at a given time. In some methods, such as the AutoAnalyzer procedures, the method blank is automatically and continuously compensated for since a continuous flow of the reagents passes through the detector. In other procedures, such as the gas chromatographic determination of pesticides, a method blank is run with each series of samples analyzed. Usually this is one blank for every nine samples. 5-4 CAPCO JEN 0005410 5.3.3 Elimination of Interferences and Other Sources of Error Procedures for eliminating or at least minimizing impurities that produce specific interferences or high general background, vary with the reagent and method involved. These procedures may include: recrystallization, precipitation, distillation, washing with an appropriate solvent, or a combination of these. Examples of procedures used for various types of analyses are given below. For complete information, the analyst should consult the individual methods. 5.3.3.1 General Inorganic Analyses Analytical reagent grade chemicals and solvents usually present no interference problems in inorganic analyses. However, some reagents do not always meet methods requirements. An example is potassium persulfate used in phosphorus and nitrogen determinations. This reagent is frequently contaminated with ammonia. Therefore, it is routinely purified by passing air through a heated water solution of the reagent. The purified potassium persulfate is recovered by recrystallization. A problem more commonly encountered in inorganic analyses is the rapid deterioration of the standard reagents and other ingredients. To minimize or eliminate this problem some reagents, for example, ferrous ammonium sulfate, must be standardized daily. Others, such as sodium thiosulfate used for dissolved oxygen determination, may require a substitute reagent, e.g., phenyl arsene oxide. Solid phenol which readily oxidizes and acquires a reddish color can be purified by distillation. Starch indicator used for idiometric titrations may be prepared f<?r each use or preserved by refrigeration, or by addition of zinc chloride or other suitable compounds. 5.3.3.2 Metals Analyses In general, spectrograde chemicals, solvents, and gases present no interference problems in atomic absorption or emission spectrographic determinations. However, standards which do not meet the requirements of the method are sometimes obtained. Ordinarily, no effort is made to purify them. They are simply replaced by new reagents of sufficient purity. Some reagents may form precipitates on standing. Such reagents will reduce the accuracy of quantitative analyses and should not be used. 5.3.3.3 Radiological Analyses In- general, reagents that do not meet the purity requirements for radiological determinations are replaced with reagents that are satisfactory. However, in some instances (for example, barium sulfate used for coprecipitation of radium), it may be necessary to carry out repeated recrystallization to remove all forms of radium, and reduce the background count to a useable level. In some instances, solvents that do not meet requirements may be distilled to produce adequate purity. In some cases, gases having background counts may be useable after aging as described earlier. If not, they should be replaced with gases that are satisfactory. 5.3.3.4 Organic Analyses Many AR-grade chemicals and solvents, and at times pesticide quality solvents, do not meet the specifications required foi*the determination of specific organic compounds. Impurities 5-5 CAPCO JEN 0005411 that are considered trace, or insignificant, for many analytical uses, are often present in greater quantities than the organic constituents being measured. Coupled with the several-hundred-fold concentration of the sample extract that is usually required, such impurities can cause very significant interferences in trace organic analyses. Reagents and solvents found to be unsatisfactory, under the conditions of the analyses, must be replaced or cleaned up so that they are useable. Some useful clean-up procedures are: a. Washing the inorganic reagents with ejich solvent that the reagent contacts during the analysis, b. Washing the adsorbents, such as silica gel G and Florisil, with the solvents that are used for a specified column or thin-layer chromatographic procedure, c. Pre-extracting distilled water with solvents used for the particular analysis involved, d. Pre-extracting aqueous reagent solutions with the solvents involved, e. Redistilling solvents in all-glass systems using an efficient fractionating column, f. Recrystallizing reagents and dyes used in colorimetric or thin-layer determinations, If the reagents and solvents thus produced are not of sufficient purity, they should be replaced. Dirty gases (quality less than specified) are particularly troublesome in gas chromatographic analyses. They may reduce the sensitivity of the detector, and produce a high or noisy baseline. If this occurs, the cylinder should be replaced immediately. Similarly, if cylinders of compressed gases are completely emptied in use, the end volumes of the gas may produce a similar and often more severe effect. Oils and water may get into the system and foul the detector. When this occurs the system must be dismantled and cleaned. Overhaul of the detector may be required. To reduce chances of this, it is recommended that all gas cylinders be replaced when the pressure falls to 100-200 psi. Filter driers are of little help in coping with this type of contamination. 5.3.4 Storing and Maintaining Quality of Reagents and Solvents Having earned out the tasks of selecting, preparing, and verifying the suitability of reagents, solvents, and gases, the analyst must properly store them to prevent contamination and deterioration prior to their use. Borosilicate glass bottles with ground glass stoppers are recommended for most standard solutions and solvents. Plastic containers, e.g., polyethylene, are recommended for alkaline solutions. Plastic containers must not be used for reagents or solvents intended for organic analyses. However, plastic containers may be used for reagents not involved with organic analyses if they maintain a constant volume, and it is demonstrated that they do not produce interferences and do not absorb constitutents of interest. It is important that all containers be properly cleaned and stored prior to use. (Refer to Chapter 4 for details). Standard reagents and solvents must always be stored according to the manufacturer's directions. Reagents or solvents that are sensitive to the light should be stored in dark bottles and/or stored in a coe(, dark place. It is particularly important to store materials used for radiological determinations in dark bottles, since photoluminescence will produce 5-6 CAPCO JEN 0005412 high background if light sensitive detectors are used for counting. Some reagents require refrigeration. Adsorbents for thin-layer and column chromatography are stored in the containers that they are supplied in, or according to the requirements of individual methods. Activated carbon, used for collection of samples for organic analyses, must be stored and processed in areas protected from atmospheric and other sources of contamination (8). The analyst should pay particular attention to the stability of the standard reagents. Standards should not be kept longer than recommended by the manufacturer, or in the method. Some standards are susceptible to changes in normality due to absorption of gases or water vapor from the air. Provisions for minimizing this effect are given in Part 23 of ASTM Standards (4). The concentration of the standards will change as a result of evaporation of solvent. This is especially true of standards prepared in volatile organic solvents. Therefore, the reagent bottles should be kept stoppered, except when actually in use_._The chemical composition of certain standards may change on standing. Certain pesticide!, for instance, will degrade if prepared in acetone that contains small quantities of water. Thus, it is essential that working standards be frequently checked to determine changes in concentration or composition. Stock solutions should be checked before preparing new working standards from them. 5.4 References 1. National Bureau of Standards, Special Publication 260, "Standard Reference Materials", July 1969. 2. "Reagent Chemicals, American Chemical Society Specifications", American Chemical Society, Washington, D.C. 3. ASTM Special Technical Publication No. 148-H, "Manual on Industrial Water and Industrial Waste Water", 2nd Edition, p. 869 (1965). 4. 1968 Book of ASTM Standards, Part 23; p. 897: American Society for Testing and Materials. 1916 Race Street, Philadelphia, Pa. 19103. 5. Standard Methods for the Examination of Water and Wastewater. 13th Edition, American Public Health Association, New York (1971). 6. "FWPCA Method for Chlorinated Hydrocarbon Pesticides in Water and Waste Water", Federal Water Pollution Control Administration, Analytical Quality Control Laboratory-, November 1969. 7. Burke, J,, JAOAC, 48, 1037 (1965). 8. Breidenbach, A. W., et al, "The Identification and Measurement of Chlorinated Hydrocarbon Pesticides in Surface Waters", Publication WP-22, Federal Water Pollution Control Administration, Washington, D.C. (November 1966). 5-7 CAPCO JEN 0005413 Chapter 6 CONTROL OF ANALYTICAL PERFORMANCE 6.1 Introduction This chapter is limited to a discussion of ihe- control of analytical performance in the laboratory. It is assumed that a valid sample has been properly taken, preserved, and delivered to the laboratory for analyses; that the laboratory analyses were done according to currently-recognized methods; and that the recording and reporting of subsequent laboratory results were done in a systematic, uniform, and permanent fashion (See Chapter 7). It must be recognized (and practiced!), however, that quality control begins with the sample collection and does not end until the resulting data are reported. The laboratory control of analytical performance is but one vital link in obtaining valid data. A continuous rapport and conscientious use of quality control betwejen field sampling, laboratory analyses, and management decisions are necessary to insure this validity. Earlier chapters have discussed such key elements as laboratory services, instrumentation, glassware, reagents, solvents, and gases; the reader should refer to these sections to determine the necessary specifications and requirements required for quality control. On the assumption that these variables are under control, that a single method is being used, and that the complete system is initially under control, what should be done in the evaluation of daily performance to document that valid data are being produced? First, valid precision and accuracy data should be available on the method and analyst. Thereafter, systematic daily checks are required to show that reproducible results are being obtained, and that the methodology is actually measuring what is in the sample. These items are discussed in detail in the following sections. 6.2 Precision and Accuracy Precision refers to the reproducibility among replicate observations. In an Analytical Quality Control Program, it is determined, not on reference standards, but by the use of actual water samples which cover a range of concentrations and a variety of interfering materials usually encountered by the analyst. Obviously, such data should not be collected until the analyst is thoroughly familiar with the method, and has obtained a reproducible standard curve. For colorimetric analyses, the initial standard curve should include a blank and a series of at least eight standards encompassing the full concentration range to be used for routine sample analyses. Subsequently, at least two standards (a high and a low) should be analyzed to verify the original standard curve. For other measurements, such as pH, conductivity, turbidity, etc., instruments should be standardized according to manufacturer.'s'instructions (See Chapter 3) and sound, scientific practice. There are a number of different methods available for the determination of precision. One method that has been successfully employed by experienced AutoAnalyzer users, and can be adapted to many other analytical instrumentation and chemical procedures, is described as follows: a. Four separate concentration levels should be studied, including a low concentration near the sensitivitytevel of the method, two intermediate concentrations, and a concentration near the upper limit of application of the method. 6-1 CAPCO JEN 0005414 b. Seven replicate determinations should be made at each of the concentrations tested. c. To allow for changes in instrument conditions, the precision study should cover at least two hours of normal laboratory operation. d. In order to permit the maximum interferences in sequential operation, it is suggested that the samples be run in the following order: high, low, intermediate, intermediate. This series is then repeated seven times to obtain the desired replica tion. e. The precision statement should include a range of standard deviations over the tested range of concentration. Thus, four standard deviations will be obtained over a range of four concentrations, but the statement should contain only the extremes of standard deviations and concentrations studied. An example of data generated from such an approach is shown in Table 6-1. Table 6-1 PRECISION' DATA ON RIVER WATER SAMPLES FOR PHOSPHORUS AUTOANALYZER METHOD Sample 1 2 3 4 5 6 7 Avg. s Kanawha 0.05 0.06 0.06 0.06 0.06 0.06 0.06 0.059 0.004 Cone., (mg P/1) Klamath Arkansas 0.10 0.10 0.10 0.11 0.11 0.11 - 0.48 0.48 0.49 0.48 0.48 0.48 - 0.105 0.482 0.005 0.004 Big Souix 0.62 0.62 0.62 0.63 0.62 0.62 0.62 0.621 0.004 The resulting precision statement would read as follows: "In a single laboratory, using surface water samples at concentrations of 0.06 and 0.62 mg P/1, the standard deviation was 0.004 (Analytical Quality Control Laboratory)." Thus, the statement contams.the number of laboratories involved, the type of samples, the concentrations used, the resulting standard deviation (s) and the reference source. 6-2 CAPCO JEN 0005415 Accuracy refers to a degree of difference between observed and known, or actual, values. Again, accuracy should be determined on actual water samples routinely analyzed, and preferably, on the same series as those used in the precision determinations. The method employed by experienced AutoAnalyzer users consists of the following key steps: a. Known amounts of the particular constituent should be added to actual samples at concentrations where the precision of the method is satisfactory. It is suggested that amounts be added to the low-concentration sample, sufficient to double that concentration, and that an amourtT"be added to one of the intermediate concentrations, sufficient to bring the final concentration in the sample to approximately 75% of the upper limit of application of the method. b. Seven replicate determinations at each concentration should be made. c. Accuracy should be reported as the percent recovery at the final concentration of the spiked sample. Percent recovery at each concentration should be the mean of the seven replicate results. Data were obtained with this approach by using two of the water samples previously used in the precision study reported in Table 6-1 (Kanawha and Arkansas Rivers). They are reported in Table 6-2. Table 6-2 ACCURACY DATA ON RIVER WATER SAMPLES FOR PHOSPHORUS AUTOANALYZER METHOD Cone., (mg P/1) Sample 1 1 3 4 5 6 7 Avg. Kanawha (Added 0.06 mg/1P) 0.105 0.105 0.105 0.110 0.110 0.110 0.105 0.107 Arkansas (Added 0.3 mg/1 P) 0.74 0.75 0.75 0.73 0.74 0.75 0.75 0.74 7r Recovery 0.107 . 0.0J2. + 0.06. x 100 = 90 0.74 . 0.48 + 0.30 6-3 CAPCO JEN 0005416 Again, in order to contain the key elements, the accuracy statement would read as follows: "In a single laboratory, using surface water samples at concentrations of 0.11 and 0.74 mg P/1, recoveries were 90% and 95%, respectively (Analytical Quality Control Laboratory)". Once collected and documented, these precision and accuracy data may be used in a number of ways. Two important examples are: (1) They present clearcut evidence that the analyst in question is indeed capable of analyzing-rttfe water samples for that particular parameter. That is, he has the standard method under control, and is capable of generating valid data; and (2) the data can be used in the evaluation of daily performance in reference to replicate samples, spiked standards and samples, and in the preparation of quality control charts. As observed, the above methods can be adapted to other chemical procedures and analytical instruments. They have been used on manual titration methods for such parameters as alkalinity, chloride, and hardness; on general inorganic instruments such as pH, conductivity, selective ion, and turbidity meters; and on the Beckman Carbonaceous Analyzer. Other instruments, such as atomic absorption and flame emission spectrophotometers, could also be evaluated by these methods; however, radiological instrumentation and gas chromatography systems (See Chapter 8) require special techniques. 6.3 Evaluation of Daily Performance Once valid precision and accuracy data are available on the method and the analyst, systematic daily checks are necessary to insure that valid data are being generated. First of all, verification of the originally-constructed standard curve is mandatory. As previously noted, at least two standards (a high and a low) should be analyzed routinely along with a blank to determine that comparable operating conditions exist. If the data do not substantiate such control, the analyst must systematically trouble-shoot his system until the problem is corrected. In order to document that reproducible results are being obtained (i.e., precision of the method), it is necessary to run replicate samples. Although frequency of such replicate analyses is, by nature, dependent on such factors as the original precision of the method, the reliability of the instrumentation involved, and the experience of the analyst, good laboratory technique is to run duplicate analyses at least ten percent of the time. The resulting data should agree favorably with the known precision of the method. If they do not, the system is not under control, and results are subject to question. Concurrently, quality control should include assurance that the daily system is actually measuring what is in the sample (i.e., accuracy of the method). Although it is far preferable to have obtained values check with known or actual values, it should be recognized that inaccuracy does not destroy the value of data if the degree and precision of the error is known and taken into account. In order to account for background contamination and/or sample interferences, and as a matter of routine practice, spiked samples should be used in addition to standards. As in the case of duplicate sample analyses, good laboratory technique dictates that spiked samples be run at least ten percent of the time. Thus, daily control of analytical performance in the laboratory requires approximately 15-20 percent of the analyst's time. Considering the elapsed time and combined efforts of 6-4 CAPCO JEN 0005417 skilled personnel that are represented in a final laboratory result, this is a comparatively small price to pay for, not a "number", but a valid concentration value. A most convenient way of recording the obtained precision and accuracy data is through the preparation of quality control charts. Plotting of said data systematically answers the question as to whether the laboratory analyses are under control, and is useful in observing developing trends of positive or negative bias. Because of its importance in documenting the quality control being practiced daily in the laboratory, the construction and uses of quality control charts are treated as a separate topic incite next section. A broader and somewhat different form of evaluation of daily performance may be made through routine participation in interlaboratory round-robin studies. Samples analyzed in such a cooperative program should be treated as part of the routine sample load. In so doing, the analyst is able to compare his individual performance against other laboratory personnel, and to have a reliable measure of the particular method's capabilities. In many respects such samples can be regarded as reputable "blind samples"; a necessary' ingredient in the quality control of laboratory results. 6.4 Quality Control Charts Quality control charts were originally developed for the control of production processes where large numbers of items were being manufactured and inspected on an essentially continuous basis. As shown in Figure 6-1, a control chart consists of a graphical chart with the vertical scale plotted in units of the test result and the horizontal scale in units of time or sequence of results. The upper and lower control limits shown on the chart are used as criteria for action, or for judging the significance of variations between duplicate samples. The central line represents the average or the standard value of the statistical measure being plotted. Upper Control Limit Central (Average) Lower Control Limit Time or Order of Results Figure 6-1 ESSENTIALS OF CONTROL CHART As observed in the previous section on the evaluation of daily performance, daily precision and accuracy data can be plotted by means of these quality control charts to determine if valid, questionable, or invalid data are being generated from day to day. There are several techniques available for actually constructing quality control charts and plotting subsequent data. Two currently in use are the Shewhart technique (1,2) and the CuSum technique (3). In both techniques, preeistbn control charts are constructed from duplicate sample 6-5 CAPCO JEN 0005418 analyses, whereas, accuracy control charts are constructed from spiked samples or standards data generated in monitoring recovery efficiencies. At least 15 to 20 sets of duplicate and 15 to 20 sets of spiked sample data from an in-control process are necessary for the initial construction. A system is initially said to be in control when the standard deviation and recovery efficiency data for a given parameter are comparable to those obtained by other experienced laboratories. It is also necessary that the initial and subsequent sets of data be obtained under normal laboratory operation conditions, that the same analyst or group of analysts run the an^djtsis, and that the same analytical method is used. 6.4.1 Cumulative-Summation (CuSum) Quality Control Charts There are various systems currently available for plotting data in the form of cumulative sum charts (4). One system that has been in continuous use within EPA Region VI is that of Harkins and Crowe (3). It has proved most useful in monitoring the validity of data generated by a contracting laboratory and is currently being "used routinely to daily record intra-laboratory performance in technical operations. The following material has been excerpted from their manual (3), in order to accurately describe the construction and use of these charts: 6.4.1.1 Construction of CuSum Quality Control Charts The control charts are derived from three basic calculations: a. Standard deviations (Sd) of the differences between duplicates or, in the case of spiked or standard samples, between the known quantity and the quantity obtained. b. The upper control limit (UL) c. The lower control limit (LL) Prior to these calculations, two decisions must be made: a. The a and 0 levels b. The allowable variability levels Mathematical Equations n, n. (V di)2 V - -di2 ----------- S2 =-------- ---- -------------- = Variance of the differences d N--1 sd = V"sT= Standard deviation of the differences Sq = (-8Sd)2 = .64 S2 (estimates a * ) (a) S2 = (1.2Sd)2 =1.44 Sd (estimates ct j) 6-6 CAPCO JEN 0005419 UL(M) = 2 loge r,1 -f/l i aJ 1i S 2 s2 0 +M rs> i loge Lso J 1i s2 l LL(M) = 11 S5 SJ +M 11 So s 2 1 Where: UL(M) upper limit at M sets of samples LL(M) tii lower limit at M sets of samples the difference between the i1*1 set of duplicates or spiked samples the total number of sets of duplicates or spiked samples used to construct the control charts minimum amount of variation allowed in the system maximum amount of variation allowed in the system percent (decimal fraction) of time you are willing to judge the procedure out of control when it is in control P percent (decimal fraction) of time you are willing to judge the procedure in control when it is out of control M = number of sets of duplicates or spiked samples used in calculating the value to be plotted on the chart By definition, a is the probability of judging the process to be out of control when in fact, it is in control. It is recommended that a be chosen to lie between the boundaries of .05 and .15, that is, the laboratory personnel are willing to stop the laboratory process somewhere between 5 and 15% of the time, judging it to be out of control, when in fact, it is in control. If the cost of examining a process to determine the reason or reasons for being out of control is considerable, then it may he desirable to choose a Iowa. Likewise, if the cost is negligible, it may be desirable to choose a larger a value, and thus stop the process more frequently. (See Figure 6-2) 6-7 CAPCO JEN 0005420 On the other hand, 0 is defined as the probability of judging the process to be in control when it is not. Again, it is recommended that 0 be chosen to lie between the values of .05 and .15; thus, the laboratory personnel are willing to accept out of control data somewhere between 5 and 15% of the time. The economic considerations used for choosing a are also applicable to the choice of B. (See Figure 6-2.) It is also essential to set maximum and minimum allowable variability levels. It is necessary to specify a value for the minimum and maximum amount of variation that will be 2 allowable in the system. These minimum and maximum amounts are referred to as o and j respectively. The values used should be based on a knowledge of the variation in the procedure under consideration. However, if such knowledge is not available, the values may be arbitrarily set at a ^2 = (a - ,20a)2 and a 2^ - (a + .20a ) 2 . Figure 6-2. EFFECT OF a AND 0 LEVELS ON STANDARD CONTROL CHART 6-8 CAPCO JEN 0005421 6.4.1.2 Use of CuSum Control Charts Once the control charts are constructed, and prior to their use, consideration must be given to the number of duplicate analyses to be conducted during a series of samples; likewise, the same decision must be made on spiked or standard samples. In considering the number of duplicate and spiked sample analyses to be conducted in a series of samples, it is necessary to weight the consequences when the data go out of control. The consequences of this situation are reanalyzing a series of samples or discarding the questionable data obtained. The samples tCBe reanalyzed are those lying between the last in-control point and the present out-of-control point. A realistic frequency for running duplicate and spiked samples would be every fifth sample; however, economic consideration and experience may require more or less frequent duplicate and spiked sample analyses. Once the frequency of duplicate and spiked samples has been determined, it is then necessary to prepare spiked or standard samples in concentrations relative to the concentration of the control charts, which should be similar to those of the environmental samples. These spiked or standard samples must be intermitterttly dispersed among the series of samples to be analyzed and without the analyst's knowledge of concentration. Similarly, duplicate samples must be intermittently dispersed throughout the series of samples to be analyzed, and ideally, without the analyst's knowledge; however, this is sometimes very difficult to accomplish. The results of the duplicate and spiked sample analyses should be calculated immediately upon analyzing the samples to allow for early detection of problems that may exist in the laboratory. An example of these calculations follows: Duplicate Sample No. M 1 2 3 Results No. 1 No. 2 5.4 5.2 4.8 4.7 6.1 5.8 Difference (di) 2 .1 .3 di2 2 (di2) .04 .04 .01 .05 .09 .14 Upon plotting the summation or S(di2), one of three possibilities can occur (See Figure 6-3): a. Out of control on the upper limit When "data goes out of control on the upper limit the following steps should be taken: 1.. Stop work immediately 2. Determine problems (a) Precision control chart (1) The analyst (2) Nature of the sample (3) Glassware contamination 6-9 CAPCO JEN 0005422 r SAMPLE SET NO. ANALYSIS IN CONTROL NO PROBLEMS: CONTINUE ANALYSIS SAMPLE SET NO. ANALYSIS OUT OF CONTROL UPPER LIMIT PROCEDURES: 1. STOP ANALYSIS 2. LOCATE PROBLEM 3. CORRECT PROBLEM 4. RERUN SAMPLES 5. START CHART AT SAMPLE SET NO. 1. { I SAMPLE SET NO. ANALYSIS OUT OF CONTROL LOWER LIMIT INCREASED EFFICIENCY OR FALSE RESORTING PROCEDURES: 1. CONTINUE ANALYSIS 2. CONSTRUCT NEW CHART WITH RECENT DATA 3. OBSERVE ANALYST SAMPLE SET NO. ANALYSIS OUT OF CONTROL UPPER LIMIT CONTINUOUS ERROR TREND PROCEDURES: SAME AS ABOVE BUT STOP ANALYSIS WHEN TREND IS DETECTED. Figure 6-3. LABORATORY QUALITY CONTROL CHARTS 6-10 I I CAPCO JEN 0005423 (b) Accuracy control chart (1) The analyst (2) Glassware contamination (3) Contaminated reagents (4) Instrument problems (5) Sample interference with the spiked material 3. Rerun samples represented by that sample set number, including additional duplicate and spiked samples. 4. Begin plotting at sample No. 1 oiTthart. b. In control within the upper and lower limit lines When data continuously fall in between the upper and lower control limits, the analyses should be continued until an out-of-control trend is detected. c. Out of control on the lower limit --- When data fall out of control on the lower limit, the following steps should be taken: 1. Continue analyses unless trend changes 2. Construct new control charts on recent data 3. Check analyst's reporting of data 6.4.2 Shewhart Quality Control Charts Dr. Walter A. Shewhart of Bell Telephone Laboratories developed the basic theory of control charts in the 1920's. His book on statistical quality control (1) grew out of this original work. Since then, industrial acceptance of these control chart concepts and other statistical techniques have refined and quantitated the quest for quality in manufacturing. Although originally developed for control of production processes when large numbers of articles were being manufactured and inspected on an essentially continuous basis, these same concepts have been readily adapted to laboratory operations where the analyst produces comparatively fewer results on an intermittent basis. As-in the CuSum approach, precision control charts are prepared from data resulting from duplicate sample analyses and accuracy control charts from duplicate spiked standards or samples. Once the control charts are constructed, however, data are plotted as individual values rather than cumulative sums. i Certain constants (factors) are also involved in the preparation of Shewhart Charts. Depending upon how the data are grouped, what the size of each grouping is, and what control limit formulation is being calculated, Table 6-3 will serve as a basic reference point: 6-11 CAPCO JEN 0005424 Table 6-3 FACTORS FOR COMPUTING CONTROL CHART LINES (5, 6) Observations in Subgroup (>?) 2 3 4 5 6 7 8 Factor a2 1.88 1.02 0.73 0.58 0.48 0.42 0.37 Factor d4 3.27 2.58 2.28 ' 2.12 2.00 1.92 1.86 Inherent in the Shewhart approach is recognition of the basic assumption that variations exist in every method. That is, no procedure is so perfect, so unaffected by its environment, that it will always give exactly the same assay value or product. Where such situations seem to exist, either the device used to measure the process is not sensitive enough or the person making the measurements is not performing properly. For our purposes, the recorded difference between paired samples should never be less than one-half the minimum detectable limit of the parameter under consideration. In the following outlines for preparing precision and accuracy control charts, nitrate data were used to develop the examples. Therefore the minimum values shown are 0.05 (one-half the. observed minimum detectable limit of 0.1 mg/1 as N). 6.4.3 Precision Control Charts .* These charts are developed by collecting data for many samples, a minimum of 15 to 20, run in duplicate under assumed controlled conditions. Once these data have been generated, preferably over an extended period of laboratory time, the following steps should be followed to construct the control chart: a. List the range (R) for each set of samples. That is, the absolute value of the difference betweeji^sach set of duplicate samples. 6-12 CAPCO JEN 0005425 Note: The following ranges were observed in the nitrate data: 0.1 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.1 0.05 0.05 0.05 0.30 ^"0.10 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 b. Calculate the average range (R) b> summing the list of R values and dividing by the number of sets of duplicates: SR R= * -R = -?=+1=5- = 0.06 35 Calculate the Upper Control Limit (UCL) on the range according to the formula: UCLr = d4r, where D4 is a constant dependent on the number of units in the subgroup. In this case, since two observations are in the subgroup, d4 = 3.27 (see Table 6-3). UCLr = 3.27 R = 0.20 d. Calculate the Upper Warning Limit (UWL) on the range according to the formula: UWLr = 2/3 (D4 R-R)+R, which for duplicate samples reduces to UWLR = 2.51 R UWLr = 0.15 This UWL corresponded the 95% confidence level. 6-13 CAPCO JEN 0005426 e. Now graph R, UWLr, and UCLj^ in the following manner: R Order of Results V' (e.g., duplicate sample sets) f. The above precision control chart for nitrates is now complete, and can be used to plot R values on subsequent duplicate samples to determine, if the system is in control, out of control (plotted R value beyond the UCL), and/or to detect any trends developing within the system. 1. In this example, a trend has developed between duplicate sample sets 4 thru 7. Although the system is not out of control, all variables in the procedure should be checked in an attempt to stop this obvious trend before the UCL is reached. 2. In this example, the system has clearly gone out of control between duplicate samples 3 and 4. At this point, the system can be stopped and all variables in the system checked, or another set of duplicate samples can be run to verify the observed-difference. Once the system has been corrected, all samples between set 3 and 4 should be rerun to insure the validity of the data. 6-14 CAPCO JEN 0005427 0.3 R X UCLr = 0.20 UWLr = 0.15 X X 12 R = 0.06 X -- 3456 Order of Results 7 8 6.4.4 Accuracy Control Charts As in the above system, these charts are developed by collecting-data for many samples, a minimum of 15 to 20, but on spiked samples (preferably) or standards under assumed controlled conditions. Again, these data should be generated over an extended period of laboratory time, and be representative of normal operating conditions (7). The following steps should be followed to construct accuracy control charts: a. List the range (R) and the average (X)* of each subgroup of data. Note: In the following example of nitrate data, subgroups of monthly data involving four observations were used: *X = -d Found - Actual n Month Sept. Nov. Dec. Jan. Actual 1.1 1.1 1.1 1.1 1.1 1.1 1.2 1.2 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 Found 1.1 1.1 1.1 1.1 1.1 1.1 1.2 1.1 1.1 1.0 1.0 1.0 0.9 1.0 1.1 1.0 R 0 0.1 0.1 0.2 X 0 -0.025 +0.025 0 6-15 CAPCO JEN 0005428 b. Calculate the average range (R) by summing tire list of R values and dividing by the number of subgroups: R = *IL 7 r = oa.= o.io c. Calculate the Upper Control Limit (UCL) on the range according to the formula: UCLr = d4r, where D4 is a constant dependent on the number of units in the subgroup. In this case, since four observations are in the subgroup; D4 = 2.28 (see Table 6-3). UCLr = 2.28 R = 0.23 d. Calculate the Upper Warning Limit (UWL) on the range according to the formula: UWLR= 2/3 (D4R-RHR UWLr = 2/3 [2.28 (0.1)-0.11+0.1 =0.19 e. Now graph R, UWLr, and UCLR in the following manner: Order of Results f. Turning now to the X values, calculate the UCL^ by the formula: UCLj- = A2R where A2 is a constant dependent on the number of units in the subgroup. In this case, since four observations are in the subgroup, A2 = 0.73 (see Table 6-3). -TTUCLx = -73^ = -07 6-16 CAPCO JEN 00054.29 Calculate the UWLx by the formula: UWLjJ = 2/3 A2(R) UWLx = 2/3 [ 0.73 (0.1)] =0.05 Note: Lower Warning Limit (LWLy) and Lower Control Limit (LCLj) are simply the negative values of UWLx and UCL^, respectively. h. Now graph the standard Nominal Value (set etjual to zero), UWLx, and UCLj^, LWLx, and LCLj in the following manner: .07 UCLx = -07 .05 UWLx = 0.05 .03 .01 Std Nominal Value order of subgroups -.03 -.05 LWLx = 0.05 -.07 LCL^ = 0.07 i. In order to detail any trends forming within each subgroup, individual differences may be plotted by preparing the following graph: +.2 +.1 0 11 11 I1 . -! -.2 1I 11 11 Individual value 11 i1 11 1! 11 11 subgroup Nominal j. Thus, as in the precision control charts, once all of the above accuracy control charts have been constructed, all future data can be plotted on each set of duplicate spiked sanities or standards, to determine if the system is in control, out of control, and/or to detect any trends developing within the system. 6-17 CAPCO JEN 0005430 6.5 References 1. Shewhart, W. A., Economic Control of Quality of Manufactured Product, 1931. 2. Anon., "Statistical Method-Evaluation and Quality Control for the Laboratory", DHEW Training Course Manual in Computational Analysis, August 1968. 3. Anon., "Laboratory Quality Coflffol Manual", Federal Water Pollution Control Administration, Robert S. Kerr Water Research Center, 1969. 4. Griffin, D. F., "Systems Control by Cumulative Sum Method", Amer. J. Med. Tech., 34. 644(1968). 5. Duncan, A. J., Quality Control and Industrial Statistics, 3rd Ed., R. D. Irwin, Inc., Homewood, 111., Chap. 18 (1965). 6. ASTM Special Technical Publication No. 15-C, "Manual on Quality Control of Materials", pp. 59-64, January 1951. 7. Grant, E. G., Statistical Quality Control, 3rd Ed., McGraw-Hill, New York (1964). 6-18 CAPCO JEN 0005431 CHAPTER 7 DATA HANDLING AND REPORTING 7.1 Introduction To obtain meaningful data on water quality, the laboratory must first collect a representative sample and deliver it unchanged for analysis. The analyst must then complete the proper analysis in the prescribed fashion. Having accomplished these steps, one other important step must be completed before the data are of use. This step includes the permanent recording of the analytical data in meaningful, exact terms, and reporting it in proper form to some storage facility for future interpretation and use. The brief sections that follow discuss the data value itself, recording and reporting the value in the proper way, means of quality control of data, and storaglfand retrieval. 7.2 The Analytical Value 7.2.1 Significant Figures The term significant figure is used rather loosely to describe some judgment of the number of reportable digits in a result. Often the judgment is not soundly based and meaningful digits are lost or meaningless digits are accepted. Proper use of significant figures gives an indication of the reliability of the analytical method used. The following definitions and rules are suggested for retention of significant figures: A number is an expression of quantity. A figure or digit is any of the characters 0, 1,2, 3, 4, 5. 6. 7. 8, 9, which, alone or in combination, serves to express a number. A significant figure is a digit that denotes the amount of the quantity in the place in which it stands. Reported values should contain only significant figures. A value is made up of significant figures when it contains all digits known to be true and one last digit in doubt. For example, if a value is reported as 18.8 mg/1, the "18" must be firm values while the "0.8" is somewhat uncertain and may be "7" or "9". The number zero may or may not be a significant figure: a. Final zeros after a decimal point are always significant figures. For example, 9.8 grams to the nearest mg is reported as 9.800 grams. b. Zeros before a decimal point with other preceding digits are significant. With no other preceding digit, a zero before the decimal point is not significant. c. If there are no digits preceding a decimal point, the zeros after the decimal point but preceding other digits are not significant. These zeros only indicate the position of tile decimal point. 7-1 CAPCO JEN 0005432 d. Final zeros in a whole number may or may not be significant. In a conductivity measurement of 1000 /imhos/cm, there is no implication that the conductivity is ' 1000 1 /imho. Rather, the zeros only indicate the magnitude of the number. A good measure of the significance of one or more zeros before or after another digit is to determine whether the zeros can be dropped by expressing the number in exponential form. If they can, the zeros are not significant. For example, no zeros can be dropped when expressing a weight of 100.08 grams in exponential form; therefore the zeros are significant. However, a weight of 0.0008 grams can be-erpressed in exponential form as 8 x 10'4 grams, and the zeros are not significant. Significant figures reflect the limits of the particular method of analysis. It must be decided beforehand whether this number of significant digits is sufficient for interpretation purposes. If not, there is little that can be done within the limits of normal laboratory operations to improve these values. If more significant Figures are needed, a further improvement in method or selection of another method will be required to produce an increase in significant figures. Once the number of significant figures is established fo^-a type of analysis, data resulting from such analyses are reduced according to set rules for rounding off. 7.2.2 Rounding Off Numbers Rounding off of numbers is a necessary operation in all analytical areas. It is automatically applied by the limits of measurement of every instrument and all glassware. However, it is often applied in chemical calculations incorrectly by blind rule or prematurely, and in these instances, can seriously affect the final results. Rounding off should normally be applied only as follows: 7.2.2.1 Rcunding-Off Rules a. If the figure following those to be retained is less than 5. the figure is dropped, and the retained figures are kept unchanged. As an example: 11.443 is rounded off to 11.44. b. If the figure following those to be retained is greater than 5, the figure is dropped, and the last retained figure is raised by 1. As an example: 11.446 is rounded off to 11.45. c. When the figure following those to be retained is 5, and there are no figures other than zeros beyond the 5, the figure is dropped, and the last place figure retained is increased by 1 if it is an odd number, or it is kept unchanged if an even number. As an example: 11.435 is rounded off to 11.44, while 11.425 is rounded off to 11.42. 7.2.2.2 Rounding Off Single Arithmetic Operations a. Addition: When adding a series of numbers, the sum should be rounded off to the same numbers of decimal places as the addend with the smallest number of places. However, the operation is completed with all decimal places intact and rounding off is done afterward. As an example: 11.1 11.12 11.13 33.35 The sum is rounded off to 33.4. 7-2 CAPCO JEN 0005433 b. Subtraction: When subtracting one number from another, rounding off should be completed before the subtraction operation, to avoid invalidation of the whole operation. c. Multiplication: When two numbers of unequal digits are to be multiplied, all digits are carried through the operation, then the product is rounded off to the number of significant digits of the less accurate number. d. Division: When two numbers of unequal*digits are to be divided, the division is carried out on the two numbers using all digits. Then the quotient is rounded off to the number of digits of the less accurate of the divisor or dividend. e. Powers and Roots: When a number contains n significant digits, its root can be relied on for n digits, but its power can rarely be relied on for n digits. 1,22.2 Rounding Off the Results of a Series of Arithmetic Operations The rules for rounding off are reasonable for simple calculations, however, when dealing with two nearly equal numbers, there is a danger of loss of all significance when applied to a series of computations which rely on a relatively small difference in two values. Examples are calculation of variance and standard deviation. The recommended procedure is to carryseveral extra figures through the calculation and then to round off the final answer to the proper number of significant figures. 7.2.3 Glossary of Terms To clarify the meanings of reports and evaluations of data, the following terms are defined. They are derived in part from American Chemical Society and American Society for Quality Control usage (1, 2). 7.2.3.1 Accuracy Data Measurements which relate to the difference between the average test results and the true result when the latter is known or assumed. The following measures apply: Bias is defined as error in a method which systematically distorts results. The term is used interchangeably with accuracy in that bias is a measure of inaccuracy. Relative error is the mean error of a series of test results as a percentage of the true result. 1.2.22 Average In ordinary usage, the arithmetic mean. The arithmetic mean of a set on ,n_ values is the sum of the values divided bym 7.2.3.3 Characteristic A property that can serve to differentiate between items. The differentiation may lie either quantitative (by variables), or qualitative (by attributes). 7-3 CAPCO JEN 0005434 7.2.3.4 Error The difference between an observed value and its true value. 7.2.3.5 Mean The sum of a_series of test results divided by the number in the series. Arithmetic mean is understood (X). 7.2.3.6 Population Same as Universe. (See subparagraph 7.2.3.13). 7.2.3.7 Precision Degree of mutual agreement among individual measurements. Relative to a method of test, precision is the degree of mutual agreement among individual measurements made under prescribed, like conditions. 7.2.3.8 Precision Data Measurements which relate to the variation among the test results themselves, i.e., the scatter or dispersion of a series of test results, without assumption of any prior information. The following measures apply. a. Standard Deviation (a). The square root of the variance. a b. Standard Deviation, estimate of universe (s). s= c. Coefficient of Variance (V)._ The ratio of the standard deviation (s) of a set of numbers, n, to their average, X, expressed as a percentage: v-* d. Range. The difference between the largest and smallest values in a set. e. 95% Confidence Limits. The interval within which one estimates a given population parameter to Iie795% of the time. 7-4 CAPCO JEN 0005435 7.2.3.9 Sample A group of units, or portion of material, taken from a larger collection of units, or quantity of material, which serves to provide information that can be used as a basis forjudging the quality of the larger quantity as a basis for action on the larger quantity or on the production process. Also used in the sense of a "sample of observations." 7.2.3.10 Series A number of test results which possess common properties that identify them uniquely. 7.2.3.11 Skewness (k) A measure of the lopsidedness or asymmetry of a frequency distribution defined by the expression: (X; - X)3 This measure is a pure signed number. If the data are perfectly symmetrical, the skewness is zero. If k is negative, the long tail of the distribution is to the left. If k is positive, the long tail extends to the right. 7.2.3.12 Unit An object on which a measurement or observation may be made. 7.2.3.13 Universe The totality of the set of items, units, measurements, etc., real or conceptual, that is under consideration. 7.2.3.14 Variable A term used to designate a method of testing, whereby units are measured to determine, and to record for each unit, the numerical magnitude of the characteristic under consideration. This involves reading a scale of some kind. 7.3 Report Forms The analytical information reported should include the parameter, the details of the analysis such as burette readings, absorbance, wavelength, normalities of reagents, correction factors, blanks, and finally, the reported value. To reduce errors in manipulation of numbers, a good general rule is to keep data transposition to an absolute minimum. If this were pursued, the ideal report form would include all preliminary information of the analysis, yet it would be possible to use the same form through to the final reporting of data into a computer or other storage device. However, the ideal report form is not usually in use. Rather, a variety of methods are used to record data. Thev are: 7-5 CAPCO JEN 0005436 7.3.1 Loose Sheets Reporting of data onto loose or ring-binder forms is an older, but much used means of recording data. It does allow easy addition of new sheets, removal of older data, or collection of specific data segments. However, the easy facility for addition or removal also permits easy loss or misplacement of sheets, mix-ups as to date sequence, and questionable status in formal display, or for presentation as evidence. 7.3.2 Bound Books An improvement in data recording is use of bound books which force the sequence of data insertion. Modification beyond a simple lined book improves its effectiveness with little additional effort. Numbering of pages encourages use in sequence and aids also in referencing data, through a table of contents, according to time, type of analysis, kind of sample, analyst, etc. * Validation can be easily accomplished by requiring the analyst to date and sign each analysis on the day completed. This validation can be strengthened further by providing space for the laboratory supervisor to sign off as to the date and acceptability of the analysis. A further development of the bound notebook is the commercially available version designed for research-type work. These note books are preprinted with book and page numbers and spaces for title of project, project number, analyst signature, witness signature and dates. Each report sheet has its detachable duplicate sheet which allows for up-to-date review by management without disruption of the book in the laboratory. The cost is about four times that of ordinary notebooks. Use of bound notebooks is essentially limited to research and development work where an analysis is part of a relatively long project, and where the recording in the notebook is the prime disposition of the data until a status or final report is written. 7.3.3 Pre-Printed Report Forms Most field laboratories or other installations doing repetitive analyses for many parameters day in and day out, develop their own system of recording and tabulating laboratory data. This may include bound notebooks; but a vehicle for forwarding data is also required. In many instances, laboratory units tailor a form to fit a specific group of analyses, or to report a single type of analysis for series of samples, with as much information as possible preprinted to simplify use of the form. With loose-sheet multicopy forms (use of carbon or NCR paper) information can be forwarded daily, weekly, or on whatever schedule is necessary, while allowing retention of all data in the laboratory. Still, the most common record is an internal bench sheet, or bound book, for recording of all data in rough form. The bench sheet or book never leaves the laboratory but serves as the source of information for all subsequent report forms (See Figure 7-1). In most instances the supervisor and analyst wish to look at the data from a sample point in relation to other sample points on the river or lake. This review of data by the supervisor, prior to release, is a very important part of the laboratory's quality control program; however, it is not easily accomplished with bench sheets. For this purpose, a summary sheet can be prepared which comp'ares a related group of analyses from a number of stations. An example is shown in Figure 7-2. Since the form contains all of the information necessary for 7-6 CAPCO JEN 0005437 O o -- rj Figure 7-1. HXAM l L li OF 1JFNC1I SI1L-UT' 3O CO xoc: CO >c0(/o))* rt v rj Q cC_> o cc co ZL Ok. 'f> OSJ fw- cOo. CAPCO JEN 0005438 Susp. Figure 7-2. EXAMPLE OF SUMMARY REVIEW SHEET CO a GO j o 5 CO *ca o H so E oa c o3 <J Z o <_> a> wc-Jo ,ra C-. 2 < CO DS a _oc CD >U o uca ca O >> N u. O a-Earf CUCOOJ < > -J < Z < Q C--\ w!< a: uj z oE CN _ o oCN _ ooo CoN oCO o oCN o r---< 0C0O CCOO o oCN or-* oCCCOOO o co\ o 0CO0 rCoo0COO0 CucoNo <oCGN> CN o0CCO0O -- CN lO oCoN CCG-NN> ooCN rr oCO r^**r rooCO0C0O c--o 0oCO0 CO oo0CO0 co __ rooCOo co CN _4J 3 03 H 12 O P < GHO o oi-t WCJ *oG BCJ Q a. 3 u<v 6 </> u a4C_3 cu 4C-43 *o3 o 2 `> c3 3 CJ 4O- oo *> >ca> 'o ss 4C-3* 4C-J* *O3 O 2 cCGa oG c 4- E .2 s C3 C Go u ca O 2 4CCu-1JJ. CO CGN ca c o 4U<-<. ca is 4J ca O 2 Gca a aGca eckr-a^f s 4c-a* o 2 Q ca 43>4 4J Co *coa 2 4c-a* O r* aC *n ca Eca c41--aJ4 o OCO cc =at) co ! E t) jot--> j g C ccaj c .2 *60 Gocx .Oc4uO3>. ^GO C3 < CO mJ 1* *5 ca C3 ftO E-o E-52 ~~rj -cca S 'r: So d e -cca S IS *-* 'Z2 C5 cl--j fU$- w 03 oCu3. - q o o o CO o o O J O o oooo u &O CAPCO JEN 0005439 reporting data it is used also to complete the data forms forwarded to the storage and retrieval system. The forms used to report data to data storage systems require a clear identification of the sample point, the parameter code, the type of analysis used, and the reporting terminology. Failure to provide the correct information can result in rejection of the data, or insertion in an incorrect parameter. As a group of analyses is completed on one or more samples, the values are reported in floating decimal form, along with the code numbers, for identifying the parameter and the sampling point (station). Figure 7-3 shows an example of a preprinted report form for forwarding data to keypunch. - 7.3.4 Digital Read-out Instrumental analyses, including automated, wet-chemistry instruments, such as Technicon AutoAnalyzer, atomic absorption spectrophotometer, pH meter, selective electrode meter, etc., now can provide direct digital readout of concentration, which can be recorded directly onto report sheets without further calculation. Electronics manufacturers now produce computer-calculators that will construct best-fit curves, integrate curves, and/or perform a pre-set series of calculations required to obtain the final reported value for recording by the analyst. 7.3.5 Key Punch Cards and Paper Tape Since much of the analytical data generated in laboratories is recorded on bench sheets, transferred to data report forms, key-punched, then manipulated on small terminal computers, or manipulated and stored in a larger data storage system, there is a built-in danger of transfer error. This increases with each transposition of data. It is suggested that the analyst can reduce this error by recording data onto punch cards directly from bench sheets. The cards can be retained, or forwarded immediately to the data storage system as desired. IBM now offers a small hand-operated key-punch for this purpose. It is anticipated that in future water quality systems, the intermediate report sheets will be eliminated and the data will be punched automatically by the analytical instrument system onto key-punch cards and/or paper tapes for direct use as computer input. 7.4 STORET-Computerized Storage and Retrieval of Water Quality Data The use of computers with their almost unlimited ability to record, store, retrieve, and manipulate huge amounts of data is a natural outgrowth of demands for meaningful interpretation of the great masses of data generated in almost any technical activity. In August 1961, an informal conference was held in the Basic Data Branch, Division of Water Supply and Pollution Control, U.S. Public Health Service. A number of ideas were brought together-In the basic design of a system for storage and retrieval of data for water pollution control, called STORET. In 1966, the STORET system was transferred, with the Division, into the Federal Water Pollution Control Administration, U.S. Department of the Interior. A refinement of this system is now operated by the Technical Data and Information Branch. Division of Applied Technology. EPA. II properly stored, the d.ita can be retrieved according to the point of sampling, the date. Inc specific parameters stored, etc., oi all data at a sample point or series of points can be 7-9 CAPCO JEN 0005440 Figure 7-3. EXAMPLE OF STORET REPORT FORM WATER QUALITY DATA 1 I 7-10 CAPCO JEN 0005441 extracted as a unit. There is a State/Federal cooperative activity which provides State water pollution control agencies with direct, rapid access into a central computer system for the storage, retrieval, and analysis of water quality control information. Full details on use of the STORET system are given in the STORET handbook recently revised (3). 7.4 SHAVES-A Consolidated Data Reporting and Evaluation System Information systems have been developed to bridge the gap between the analyst and his raw data, and a complex data storage and control system. These systems include preprinted report forms, computerized verification, and evaluation of data and data storage. An example is the SHAVES system. The term, SHAVES, is an acronym for "Sample Handling and Verification System," which originated at the Great Lakes-Illinois River Basin Comprehensive Project Laboratory at Grosse Isle, Michigan. Although the system's original purpose was verification of the calculations following laboratory analyses, it now includes data storage, checks for completeness and consistency of data, procedures for submitting analytical requests, a set of forms for recording sampling and analytical information, and a clerical procedure to account for analyses completed and pending. The primary purposes of SHAVES are the standardization, automation and control of reporting analyses. All samples received at the Pacific Northwest Water Laboratory for routine analysis are processed through the system. Although SHAVES uses a computer to perform its operations, it is not primarily a computer program. It is intended for use as an intra-laboratory quality control tool, and as such compliments the STORET system. It is described in detail elsewhere (4). 7.6 References 1' "Guide for Measure of Precision and Accuracy," Anal. Chem., Vol. 33, p. 480, (1961). p. 480. 2. "Glossary of General Terms Used in Quality Control," Quality Progress, Standard Group of the Standards Committee, ASQC, II, (7), pp. 21-2, (1969). 3. Water Quality Control Information System (STORET), EPA, Washington, D.C. 20460, Nov. 15, 1971. 4. Byram, K. V. and Krawczyk, D. F., "An Evaluation of SHAVES: A Water Quality Sample Handling System," Environmental Protection Agency. Pacific Northwest Water Laboratory, 1969. 7-11 CAPCO JEN 00054-42 I CHAPTER 8 SPECIAL REQUIREMENTS FOR TRACE ORGANIC ANALYSIS 8.1 Introduction The high sensitivity of the instrumentation used in trace organic chemical analysis, and the low concentration of compounds being investigated, dictate that special attention be given this field of analytical endeavor. Contamination of the sample from any possible source must be diligently guarded against, and interferences in the sample must be carefully controlled. Finally, strict attention to method and highly refined technique are required to produce valid quantitative results. 8.2 Discrete Bottled Samples Sample collection should be done with wide-mouth glass bottles, equipped with screw caps fitted with Teflon liners. The use of a screw cap without a Teflon liner may cause contamination of the sample by the liner or adhesive used in sealing the liner to the cap. Plastic bottles (polyethylene) are not used because traces of plasticizer may be leached from the plastic by the water, and can be a source of analytical interference. Moreover, organics from the sample may be adsorbed on the plastic. It has been suggested that high grade Teflon bottles may be satisfactory for this use; however, the cost is prohibitive at present. Many investigators avoid the use of glass sample bottles, because breakage in shipment frequently causes loss of sample. This is overcome by the use of relatively inexpensive, expanded polystyrene foam shipping containers molded to fit the bottle. These shipping containers can be purchased from Preferred Plastics Corp., North Grosvenordale, Connecticut. To insure freedom from organic contaminants, bottles are rinsed successively with chromate cleaning solution, running tap water, distilled water, and finally several times with redistilled solvent (e.g., acetone, hexane, petroleum ether, chloroform). Caps are washed with detergent, rinsed with tap water, distilled water, and solvent. Liners are treated in the same way as the bottles and are stored in a sealed container. Each method designates a recommended sample size for surface water analysis. Duplicate samples are recommended. If analysis by more than one method is to be requested on the same sample, sufficient sample must be simultaneously obtained to supply the needs of each analysis. It is also recommended that when requesting a non-specific analysis, any information that could help direot-`the analytical approach, or aid in interpretation of results, be supplied. Such information could include industrial or agricultural activities in the area from which the sample was obtained, spills, or other accidents that may have occurred in the area. Also mention of similar upstream activity could provide valuable assistance. Samples should be stored in a cool, dark place, and analyzed as soon as possible. If the sample cannot be analyzed immediately, reporting the holding time can help in interpreting results where die-off rates are known. 8-1 CAPCO JEN 0005443 8.3 Carbon Adsorption Samples The affinity of carbon for organic substances requires that supplies of carbon be protected from extraneous sources of contamination. For example, carbon can adsorb organic substances such as paint vehicles and insecticides from the air. Therefore, the carbon is stored and processed in an area adequately protected from such sources of contamination. As an additional precaution, the ventilating, heating, and air-conditioning systems for the laboratories in which carbon adsorption sampjtt-are processed are completely isolated from all other laboratories. All carbon is obtained from the manufacturer in sealed metal drums. Obviously, spraying with pest-control chemicals is not permitted in these areas. Carbon blank determinations supplement these precautions. 8.4 Glassware Proper calibration of volumetric glassware is essential to valid analytical results, because quantitation is performed by comparison to measured amounts of standard compounds, and by accurate measurement of sample volume. Individual concentrator tubes, used to measure final concentrate volumes, must first be calibrated at the working volume. This is particularly important for volumes less than 1 ml. Calibration should be made by noting the number of microliters of solvent required to bring the liquid level (lower miniscus) up to a particular graduation mark. A precision 100 n 1 syringe should be used to measure calibration volume. It is also very important in trace organic analysis that glassware be as free as possible of any organic contaminants. A chromic acid cleaning solution is required for removing all traces of organic material from glassware. 8.5 Reagents and Chemicals The minimum purity of reagents and chemicals should be analytical reagent grade. Analytical standards should be reference grade, when available. The analyst should take special note of the assay of less pure materials (most often pesticides). All reagents and chemicals should be stored according to manufacturer's instructions to prevent degradation. Proper storage is especially important if the chemical is to be used in preparing an analytical standard. Refrigerated chemicals should be allowed to come to room temperature before exposing them to the atmosphere. When preparing stock solutions, it is recommended that at least 0.100 gram of material be used for greater accuracy in weighing. Solutions should be carefully stored so as to preserve their concentration, and to protect them from ultraviolet radiation. Usually storage in ground-glass strippered bottles, either amber-colored, or out of the line of direct lighting, is sufficient. Standard solutions should be prepared using precision syringes, preferably equipped with a Cheney adapter, to measure the volume of stock solution to be diluted. The syringe barrel should be pre-wetted with solvent and air bubbles expelled. Dilution should be done in a Class A volumetric flask to insure accurate measurement. If these solutions are to be used frequently, they are best stored in a screw-cap. septum, sealed vial. These vials allow instant access to the solution antKbfl'er good protection against concentration changes of the standard solution. Evaporation of the solvent caused by repeated removal of the cap is a 8-2 CAPCO JEN 0005444- serious problem with other containers. If septum vials are not available it is advisable to prepare standard solutions in a volumetric container of 100 ml or more and transfer a small portion to a separate container for daily use, then discard that portion at the end of the day. All stock or standard solutions should be carefully watched for signs of changes in concentration or deterioration. As an aid to monitoring these solutions, it is wise to label them as to compound, concentration, solvent used, date, and preparer. Also, in the case of GC solutions, it is necessary to retain some evidence of its chromatographic behavior as a fresh solution for comparison at a later date. Distilled water used as dosed, control samples must be free of organic interferences. A very effective way of removing organic interferences from distilled water is to pre-extract the water with the solvent that is to be used in the analysis, then boil the water to remove the residual solvent. Organic solvents used in pesticide analysis should be pesticide quality, and demonstrated to be free of interferences in a manner compatible with whatever analytical operation is to be performed. Solvents can be checked by analyzing a volume equivalent to that used in the analysis and concentrated to the minimum final volume. Possible interferences are noted in terms of factors such as relative retention times, peak geometry, peak intensity, and width of solvent response. Interferences noted under these conditions can be considered maximum. If necessary, a solvent must be redistilled in glass using a 60-cm column packed with 1/8" glass helices, or an equivalent system. Hexane - ethyl ether and benzene are commonly used in the extraction of water and wastewater in conjunction with analysis by electron-capture gas-liquid chromatography. Because electron-capture detection methods are extremely sensitive to interferences normally found in these sohents, the cleanest possible reagent grade or pesticide-quality solvents must be used. Redistillation in the lab in an all-glass system is usually necessary. Experience in the Analytical Quality Control Laboratory has shown Burdick-Jackson hexane and Baker Chemical Co. ethyl ether to be satisfactory. All solvents vary from lot to lot. Therefore, when a good lot is found, subsequent order should specify that lot number. Solvents in the same lot may also vary and therefore each container should be checked. 8.6 Common Analytical Operations Adequate steps must be taken to eliminate or minimize interferences from solvents and other materials. A blank should be run simultaneously under the same analytical conditions as any block of samples analyzed. A block of samples is defined as any group of one or more samples analyzed using a common batch of analytical supplies. Should any one of the supplies be changed (i.e. solvent, silica gel, Florisil, etc.) a new blank is required. Quantitation of micro amounts of organic materials requires extremely careful technique to avoid loss of sample. Quantitative transfers are essential to obtain accurate and precise results. Practice in these manipulations is recommended for the inexperienced analyst. Concentration of sample extracts to very small volumes for micro analysis requires great care to avoid loss of constituents. A Kuderna-Danish evaporator is a very useful apparatus to accomplish this operation. Instructions in the use of this evaporator must be strictly billowed to avoid loss of desired sample (1). Final concentration in an ampule or calibrated lube is accomplished in a warm water bath with a gentle stream of clean, dry air, if air 8-3 CAPCO JEN 0005445 oxidation is not a problem; otherwise, nitrogen should be used. During the final concentration, the inside walls should be rinsed repeatedly with the working solvent to insure the total sample is contained in the bottom of the tube. Complete evaporation of solvent must be avoided to prevent loss of sample constituents. The step should be accomplished within 10 or 12 minutes for best results. 8.7 Gas-Liquid Chromatography To obtain reproducible results, it is necessary to have very accurate control of the column oven temperature. The temperature should be reproducible withinlC, and have minimum gradients of 2C throughout the oven. For temperature-programmed operation, a low mass oven is required to allow rapid heating and cooling of the column. Most manufacturers produce gas chromatographs which meet these requirements. Many organic compounds decompose when they come in contact with hot stainless steel. For this reason the injection block should be capable of accepting a quartz or glass tube to prevent the compounds from contacting hot metal. Memory peaks observed using -direct, aqueous-injection, gas chromatography can be eliminated or greatly reduced by employing direct on-column injection. A Bio-med injection kit manufactured by Tracor Instrument Co. is very suitable for this purpose. The septum should be changed at the end of each day's use. Changing at the end of each day allows overnight purging of the system of any bleed-off of contaminants from the septum. To avoid most of this bleed-off the septa can be preconditioned by heating at 250C in a vacuum oven for two hours. The gas chromatograph should be equipped with accurate needle-valve, gas-flow controls. If these flow controls are not previously calibrated this can easily be done using a soap-bubble flow meter and a stopwatch. The nature of constituents to be measured dictates the-type of detector to be employed. The electron capture detector is extremely sensitive to electronegative functional groups and substituents, such as: halogens, conjugated carbonyls, nitrites, nitrates, and organometals. It is virtually insensitive to hydrocarbons, alcohols, and ketones. The selective sensitivity to halides makes this detector particularly valuable for the analysis of many pesticides. Electron capture detectors employing two sources of ionization are available: tritium (H3) and radioactive nickel (Ni63). Each has its advantages. Most H3 detectors possess greater sensitivity than Ni63 detectors. However, the H3 detector is limited to 225C because of radioactive leakage. This temperature limitation makes the H3 detector susceptible to a buildup of high-boiling contaminants which reduce its sensitivity. The Ni6 3 detector can be operated up to 400C to prevent this buildup of contamination. ft The microcoulometric detector is specific for halogen-, sulfur-, or nitrogen-containing compounds depending on the conditions used. Although the sensitivity of this detection system is not as great as others available, the extreme specificity makes it a very valuable device in indentification, minimizing the need for cleanup of the extract. Another selective detector, with about the same sensitivity for chlorinated compounds as microcoulometric titration, is the Coulson electrolytic conductivity detector. The specificity of this detector makes it veryniseful in the identification of pesticides, and sample cleanup and pre-treatment are less critical. However, if chlorinated compounds are being detected, a 8-4 CAPCO JEN 0005446 scrubber tube must be used with this instrument to remove any S02 produced by sulfur-containing compounds in the sample. The flame ionization detector (FID) responds to virtually all compounds. Some notable exceptions are air, water, carbon disulfide, and the fixed gases. This makes the FID very useful for direct- aqueous-injection gas chromatography. However, since the detector is sensitive to such a wide range of compounds, it is also subject to interference from extraneous material. This means that the extract must be cleaned up considerably before analysis. The flame photometric detector (FPD)Js used to analyze residues of phosphorusand sulfur-containing pesticides and their metabolities. Little or no cleanup of the sample extract is required, and extraneous material causes no appreciable interference. The alkali flame detector (Thermionic) can also be used in the analysis of organophosphate pesticides. Even though this detector gives an enhanced response to phosphorus, a large amount of extraneous material demands that the extract be cleaned up before the phosphorus-containing constituents are quantitated. An enhanced response to sulfur is not obtained with the alkali flame. v* A one-millivolt, one-second, full-scale-response, strip-chart recorder should be used to maintain a permanent record of the results. The type of detector employed dictates the type of carrier gas that must be used. Nitrogen can be used for the FID, electron-capture, flame-photometric, and microcoulometric titration detectors. The Analytical Quality Control Laboratory has used nitrogen from various suppliers, and has found the J. T. Baker Chemical Company's extra dry grade to be satisfactory'. However, to avoid the risk of system or detector contamination from materials possibly present in the gas cylinder, the cylinder should be replaced when the tank pressure reaches 200 psi. It is also recommended that some type of gas purifier which is composed of a molecular sieve, desiccant, and filter be used on all combustion gases and carrier gases used with electron capture detectors. A precision, gas-tight, microliter syringe, which can be accurately filled, will deliver reproducible injections, and can be easily and thoroughly cleaned, is recommended for gas chromatographic use. A Teflon plunger seal to prevent backwash, and no dead volume are desirable features. These types of syringes are available from various suppliers, but the Glencoe syringes have additional design features to prevent bending plungers. It is not possible to enumerate here the many types and applications of liquid phases and solid supports. The literature (2, 3) will provide helpful guidance in the selection of these materials. The type of column tubing required to perform a specific GC analysis is prescribed by the method for that "analysis. Aluminum has been found suitable for chlorinated organic pesticides and stainless steel for phenols. Generally 1/8" or i/4" OD tubing is used. The 1/4" column is capable of accepting larger injections but the 1/8" column is more efficient at chromatographing small injections. It should be noted that the use of glass columns will prevent the degradation of material often associated with metal columns. Silanized glass wool should be used to minimize degradation or absorption of organic compounds by the glass-wooLpkigs. Glass wool can be silanized by treating with 10% 8-5 CAPCO JEN 00054.47 dimethyldichlorosilane in toluene for 10 minutes, or the prepared material can be purchased from Applied Science Laboratories. Column packings can be prepared by the analyst, or purchased already prepared and conditioned. If prepared in the laboratory a highly refined technique should be developed by the analyst. A well prepared packing is essential to an acceptable chromatographic analysis. Particular care should be taken to^accurately measure loadings, uniformly distribute the liquid phase, and preserve the structure of the fragile solid support. One method used successfully is the beaker slurry method in which the proper amounts of stationary phase and solid support are mixed together in a solvent in a beaker. The solvent can be evaporated by immersing the beaker in a warm water bath or filtering it through a Buchner funnel. The mixture is then dried at 110C to remove residual solvent. A fluidized dryer (4) may also be used to remove the remaining solvent. Details on techniques for preparation of Teflon packings may be obtained from Applied Science Laboratories (Bulletin FF124). An important consideration in packing a column is obtaining a uniform density, not so compact as to restrict gas flow, and not so loose as to create voids during use. Also important is the need to exercise care not to crush particles during packing. It cannot be overemphasized that column conditioning is essential to obtain acceptable GC analysis. Proper conditioning further distributes the liquid phase over available active sites, and removes excess liquid phase that may bleed off the column and impair GC performance. Unstable performance in the form of baseline drift, varying sensitivity, and wide solvent peaks are usually the result of column bleed due to improper conditioning. This also contaminates the detector and makes frequent cleaning necessary. A good general procedure for conditioning columns is as follows: install the column, leaving it disconnected from the detector; heat the column in the GC oven to just above maximum recommended operating temperature for the liquid phase, without gas flow, for 2 hours; follow with a half-hour equilibration period at a temperature at least 40 degrees below the maximum recommended temperature, still without flow; raise the temperature 20 degrees above operating conditions (do not exceed the maximum recommended operating temperature for the liquid phase), adjust gas flow to 50 ml/min, and maintain these conditions 24 to 48 hours. There are many factors that must be considered in any attempt to obtain optimum operating conditions. As indicated, optimum conditions begin with the proper selection of column materials including tubing, glass wool, liquid phase, and solid support. Each method dictates the selection of these items including the percent liquid phase. Liquid load and operating temperature are critical to optimum resolution and elution rate. Uniformly coating the solid support, properly packing the column, and properly conditioning the column all contribute greatly to attaining optimum conditions. Improper attention to these factors contributes to either column bleed, which fouls the detector, or tailing peaks, which prevent resolution of closely eluting compounds. The system must be free of gas leaks since these affect sensitivity and reproducibility. The proper selection of a carrier gas is important. Injection blocks should be of proper flow design, and should be kept efcrrtn. Temperature control of injection port, column oven, and detector must be accurate and constant, and must be at equilibrium to obtain reproducible results. 8-6 CAPCO JEN 0005448 Considerable attention must be given to those factors affecting detector response, sucli as gas flows and regulated current. Also, standard solutions and condition of electrodes for MCT. detectors, condition of radioactive source for EC detectors, and ratio of combustion gases in flame ionization detectors must be carefully attended to. The gas chromatograph must either be operated within the linear range of the detector, or a suitable calibration curve must be employed. The linear range can be determined by chromatographing different amounts of a compound and observing the response. The peak area of the responses should be proportional to the amount of compound injected throughout the full range of the recorder. IfThese conditions are not fulfilled at the attenuation investigated, a more sensitive attenuation should be selected until the linear range is found. Once having obtained optimum operating conditions, the analyst must assure sustained optimum performance through a routine maintenance program, as prescribed in the instrument manual. Such common practices as inspecting for gas leaks, changing the septum after daily use, replacing gas tanks before they run too low,.and keeping a close check on injection block, oven, and detector temperatures must be performed frequently. Column performance can be monitored by observing daily response to a selected group standard, and comparing it to response to that same standard under previously determined optimum conditions. Changes in elution pattern, relative proportions of peaks, and peak geometry are signs of a deteriorating column if the rest of the system has been properly maintained. A column should be replaced as soon as deterioration is observed. A good syringe-handling technique is important when doing GC analysis. Before measuring the volume to be injected into the gas chromatograph, first wet the barrel of the syringe and expel all air bubbles. The volume injected is determined by drawing a selected portion of the extract entirely into the glass, calibrated, syringe barrel, noting the volume, injecting the sample, partially withdrawing the plunger, and measuring the liquid remaining in the syringe barrel. Injections are made rapidly after the needle is in the gas chromatographic injection port. The needle is immediately removed from the system to prevent volatilization of any sample in the needle. This method improves the analytical accuracy since any absolute volume is injected and reproducible technique is not required. It also eliminates the built-in bias of injecting slightly more than expected each time, when direct barrel readings are used exclusively. The standard solutions can be more accurately used if prepared at a concentration that alLows injection volumes similar to those of the sample. 8.8 Qualitative Analysis The retention t(me (Rt) of a component on a given column, under given conditions, is characteristic df that particular component, and is used for qualitative identification. The standard way of reporting retention data is to give the relative retention (RRt) defined as Rt (component) * R( (reference compound). The retention time may be measured as the time elapsed from an unrestrained solute peak elution, to the apogee of the peak of interest. This works well when a flame ionization or electron-capture detector are used. However, when t!ie microccuilometric titration or the flame photometric detector are used (the flame is extinguished by the injection), no unrestrained solute peak is observed. In such, cases, the injection point is manually or^electricallv marked, and used as the point of reference for 8-7 CAPCO JEN 0005449 retention times. Caution must be exercised when manual marking is practiced, so that an accurate reference point is provided. Tentative identification is made by matching the relative retention time of the unknown component witli that obtained from a known compound analyzed under identical conditions, provided that peak geometry is also similar. It must be pointed out that similarity of retention time and peak geometry to a known compound does not unequivocally identify an unknown. Additional gas chromatographic columns of different polarity and other detectors may be used'for confirmation. If retention time and peak geometry match on two or more columns, the identity is corroborated. However, further corroboration using analytical tools such as infrared spectroscopy, mass spectrometry, or tliin-layer chromatography should be used whenever possible. Proof of identity of compounds which produce a multi-peak response should be evidenced, not only by relative retention and peak geometry, but also by the correct number and relative proportion of each peak in the total chromatogram. This is called a "fingerprint" comparison of the known standard chromatogram with that of the unknown constituent. 8.9 Quantitative Analysis The quantitative interpretation of a gas chromatogram is based either on the peak height or the peak area. The area measurement is generally preferred because peak height is extremely sensitive to small changes in the operating conditions, particularly in the column temperature. However, in chromatograms where the peaks are extremely sharp and narrow, the error involved in the area measurement makes height measurements more reliable. Peak area measurement should be carried out using height x width at half height, disc integrator, or electronic digital integrator. These techniques are rapid and simple, and give good results with symmetric peaks of reasonable width. The use of a planimeter, although less precise than other methods, is presently found to be the best method for measuring the area of unsymmetric peaks which do not originate at the baseline. Precision is improved by tracing each peak several times and taking an average value. The electronic integrator is also recommended for those unsymmetric peaks which originate at the original baseline. Concentrations of constituents are determined from standard calibration curves obtained under identical conditions. An absolute calibration curve is obtained from peak areas or peak heights plotted against known weights of a compound chromatographed under identical conditions. These standard injections must be made during the sample run to detect any change in instrument conditions or response which would invalidate the calibration. The use of an internal standard is the most accurate method of quantitating constituents in a sample. A calibration curve can be obtained by simultaneously chromatographing the previously identified sample constituent and a standard, in known weight, ratios, and plotting the weight ratios versus area ratios. An accurately known amount of the standard is then added to the unknown sample and the mixture chromatographed. The area ratios are calculated, and the weight ratio of the sample constituent to the standard is read from the curve. Since the amount^Pstandard added is known, the amount of the sample constituent can be calculate Using this method, injection volumes need not be accurately measured 8-8 CAPCO JEN 0005450 and detector response need not remain constant since any change in response will not alter the ratio. If the following requirements can be met, the internal standard is the preferred method of calibration. The internal standard must be well resolved from other peaks, must elute close to peaks of interest, should approximate the concentration of unknown, and should have structural similarity to unknown. 8.10 Thin-Layer Chromatography Special equipment for thin-layer chromatography as supplied by the Brinkman Instrument Company has been found by the Analytical Quality Control Laboratory to be particularly convenient to use. Adsorbents supplied by Wamer-Chilcott Laboratories, particularly Silica Gel G and Alumina G, are found to be free of impurities. In any case, it is best to check the adsorbent for impurities, and discard it if necessary. Plates precoated by the manufacturer can often be used when quantitation is visual; however, trials-of plastic-backed plates in the Analytical Quality Control Laboratory have shown that they often develop unevenly and very slowly. The use of precoated glass-backed plates is therefore recommended. Prepared plates should be carefully inspected for flaws that might interfere with proper chromatographic development. Plates should be marked in such a way that any grain lines would be perpendicular to the direction of development when used. Prepared plates are best activated by heating at 110C in a convection oven for 30 minutes. Phosphorous pentoxide is recommended as the desiccant for use when storing plates. Two important points to remember before using the developing chambers are to give the chamber atmosphere time to become saturated with solvent vapor, and to keep the chamber in constant temperature surroundings shielded from drafts. Failure to observe these points could result in an erratic chromatographic development, generally off line. When spotting a plate, care must be taken not to overload the spot. Overloading results in poor chromatographic efficiency. Also, when applying a gentle stream of air or other gas to evaporate the solvent, care must be taken not to blow away any of the layer or sample droplets. The analyst should also be aware of whether or not the use of air to evaporate the solvent could cause any oxidation of the sample material, thereby causing erroneous results. If oxidation is a problem an inert gas should be substituted. When spraying a plate with aqueous or other non-volatile sprays, care should be taken not to soak the plate. Soaking with such sprays may cause spots to run as the plate stands vertically. When using any spray it is better to use a stronger solution, then apply increased amounts to develop the spots. Overspraying can produce a background color which impairs the visibility of the developed spot. Quantitation by visual comparison of sample response to a series of standards is semi-quantitative at best. However, by making the spot size the same for samples and standards,by using the same solvent for both, and by using careful development techniques, a fair degree of reliability can be obtained, when comparing the factors of spot size and color intensity. When zone scraping and collating the layer material with the aid of a vacuum, the collection apparatus is usually plugged with glass wool to trap the adsorbent. Since poorly 8-9 CAPCO JEN 0005451 packed glass wool cun cause high loss of material by failure to trap the dust, the scraping of the periphery of the zone should be done first. A clear tygon vacuum line, ora glass section near the end of the tube makes it easy to monitor for untrapped adsorbent. 8.11 Column Chromatography The features of a chromatographic column which define its utility are shape, liquid capacity, and elution rate. The liquid capacity is princjnally a matter of convenience and should be weighed against the increased labor required for cleaning equipment. A flow-control device is critical to an efficient separation because increased flow decreases resolution between emerging components. A Teflon, flow-control stopcock is required because lubricant will produce an analytical interference. Most adsorbents used in column chromatography are preactivated by the manufacturer and shipped in air-tight containers. Storage in the laboratory should also be in an air-tight container because moisture will greatly affect the activatipn state of an adsorbent. If a second activation step is required, the material should be heated at 130C for at least five hours. The activity of the adsorbent can be monitored by eluting a mixture of chromatographically pure dyes (5). The elution rate and degree of separation of the individual dyes is a function of the activation state of the adsorbent. This enables the analyst to accurately attain the same activation for different batches of adsorbent. When the adsorbent is added to the column, gentle tapping or a vibrator should be used to settle the material. This minimizes the space between particles and prevents channeling of the eluting solvent through the adsorbent which reduces separation efficiency. Liquids should be added slowly, down the inside wall of the column, to avoid disturbing the packing surface. .Mixing of the solution above the adsorbent with the fresh eluting solvent can be minimized by introducing the new solvent just as the last of the solution reaches the packing surface. The column top must not go dry or air may be introduced which will lower the separation efficiency of the system. Before addition of the sample, columns should be pre-eluted with 50-75 ml of the solvent prescribed by the procedure. This is done to remove trapped air and to clean the column material of trace contaminants. During this pre-elution it is often necessary to tap the column to free all trapped air, especially if a volatile solvent is used in the pre-elution. 8.12 References 1. Gunther, F. A., et al. Anal. Chem.,.23, No. 2, p. 1835 (1951). 4 2. Lynn, T. R,, et al, Guide to Stationary Phases for Gas Chromatography. Analabs, Inc., 1968. 3. McNair, H. M. and Bonelli, E. J., Basic Gas Chromatography. Varian Aerograph, 1969. 4 Kruppa. K. el a!. Anal. C'hem.._39. 851 (June 1967). 5. Brockman, H. and SchoeWer, lI.,Chem. Ber., 74, p. 73 (1941). 8-10 CAPCO JEN 0005452 CHAPTER 9 SKILLS AND TRAINING 9.1 General Analytical operations in the laboratory can be graded according to the degree of complexity. Some analyses require no sample trtSTment, with the measurement performed in minutes on a simple instrument. Other determinations require extensive sample preparation prior to complex instrumental examination. Consequently, work assignments in the laboratory should be clearly defined. Each analyst should be completely trained and fully understand all the assignments of his job before being given new responsibilities. In this regard, all analysts, sub-professional or professional, should be thoroughly instructed in basic laboratory operations, according to the degree of professional maturity. Some of the basic operations that should be reviewed periodically with laboratory personnel follow: * a. Sample Logging Emphasize the routine procedure for recording of samples entering the laboratory, and assign primary responsibility. Establish what information is required, and how sample is routed to analyst. Discuss stability of samples, and how they should be stored prior to analysis. b. Sample Handling The analyst should understand thoroughly when the sample is to be settled, agitated, poured, pipetted, etc., before removal from the container. c. Measuring The analysts, especially new employees and sub-professionals, should be instructed in the use of volumetric glassware. The correct use of pipettes and graduates should be emphasized as discussed in Chapter 4. d. Weighing Because almost every measuring operation in the analytical laboratory is ultimately related to a weighing operation, the proper use of the analytical balance should be strongly emphasized. Maintenance of the balance, including periodic standardization, should be reiterated to ail personnel. e. Glassware- All glassware should be washed and rinsed following the requirements of the analysis to be performed. Not only must the personnel assigned to this task be instructed, but all lab personnel should know the routine for washing glassware, and also special requirements for particular uses. In addition, the precision tools of the laboratory such as pipets, burets, graduates, Nessler tubes, etc., should be inspected before use for cleanliness, broken delivery tips, and clarity of marking. Defective glassware should be discarded or segregated. 9-1 CAPCO JEN 0005453 In summary, quality control begins with basic laboratory techniques. Individual operator error and laboratory error can be minimized if approved techniques are consistently practiced. To insure the continued use of good technique, laboratory supervisors should periodically review the basic teclmiqucs with each analyst and point out, when necessary, areas of needed improvement. Continuing improvement of technical competence for all laboratory personnel is, of course, the final responsibility of the laboratory supervisor. In a well organized laboratory, however, a big brother attitude of higher ranking to lower grade personnel should be encouraged; each person should be eager to share experience, Jtsicks-of-the-trade, special skills, and special knowledge with subordinates. Obviously, improved efficiency and improved data quality will result. 9.2 Skills The cost of data production in the analytical laboratory, is based largely upon two factors-the pay scale of the analyst, and the number of data units produced per unit of time. However, estimates of the number of measurements that can be made per unit of time are difficult, because of the variety of factors involved. If the analyst is pushed to produce data at a rate beyond his capabilities, unreliable results may be produced. On the other hand, the analyst should be under some compulsion to produce a minimum number of measurements per unit of time, lest the cost of data production become prohibitive. In the following table, estimates are given for the number of determinations that an analyst should be expected to perform on a routine basis. The degree of skill required for reliable performance is also indicated. The arbitrary rating numbers for the degree of skill required are footnoted in the tables, but are explained more fully below: a. Rating 1 -indicates an operation that can be performed by a semi-skilled sub-professional with limited background; comparable to GS-3 through GS-5. b. Rating 2-operation requires an experienced aide (sub-professional) with background in general laboratory technique and some knowledge of chemistry, or a professional with modest training and experience; comparable to GS-4 through GS-7. c. Rating 3-indicates a complex procedure requiring a good background in analytical techniques; comparable to GS-7 through GS-11. d. Rating 4-a highly involved procedure requiring experience on complex instruments; determination requires specialization by analyst who interprets results; comparable to GS-9 through.GS-13. The time limits presented in the table are based on use of EPA methods. A tacit assumption has been made that multiple analytical units are available for measurements requiring special equipment, as for cyanides, phenols, ammonia, nitrogen and COD. For some of the simple instrumental or simple volumetric measurements, it is assumed that other operations such as filtration, dilution or duplicate readings are required; in such cases the number of measurements performed per day may appear to be fewer than one would normally anticipate.-- 9-2 CAPCO JEN 0005454 Table 9-1 SKILL-TIME RATING OF STANDARD ANALYTICAL OPERATIONS Measurement Skill Required (Rating No.) No./Day pH Conductivi ty Turbidity (HACH 2100) Color DO (Probe) Fluoride (Probe) (Simple Instrumental) 1 1 1 1 1,2 1,2 Alkalinity (Potentiometric) Acidity (Potentiometric) Chloride Hardness DO (Winkler) (Simple Volumetric) 1 1 1 1 1,2 * 100-125 100-125 75-100 60-75 100-125 100-125 50-75 50-75 100-125 100-125 75-100 Solids. Suspended Solids, Dissolved Solids, Total Solids, Volatile Nitrite N (Manual) Nitrate N (Manual) Sulfate (Turbidimetric) Silica Arsenic.*. (Simple Gravimetric) 1.2 I, 2 1,2 1,2 (Simple Colorimetric) 2 2 2 2 2,3 20-25 20-25 25-30 25-30 75-100 40-50 100-125 100-125 20-30 SKILL REQUIRED 1 - aide with minimum training, comparable to GS-3 through GS-5 2 - aide with special training or professional with minimum training, comparable to GS-5 through GS-7. 3 - experienced analyst, professional, comparable to GS-9 through GS-12. 9-3 CAPCO JEN 0005455 Table 9-1 (continued) SKILL-TIME RATING OF STANDARD ANALYTICAL OPERATIONS Measurement Skill Required (Rating No.) (Complex, Volumetric or Colorimetric) BOD COD TKN Phosphorus, Total Phenol (Dist'n only) Oil & Grease (Soxhlet) Fluoride (Dist'n) Cyanide 2,3 2,3^" 2,3 2,3 2,3 2, 3 2,3 2,3 (Spec- ial Instrumental) Metals by AA (No preliminary treatment) Metals by AA (With preliminary treatment! Pesticides by GC (Without cleanup) Pesticides by GC (With cleanup) 2.3 2. 3 3.4 3. 4 No./Day 30-40* 25-30 25-30 50-60 20-30 15-20 25-30 10-15 150 60-80 3-5 2-4 SKILL REQUIRED 2 - aide with special training or professional with minimum training, comparable to GS-5 through GS-7. 3 - experienced analyst, professional, comparable to GS-9 through GS-12. 4 - experienced analyst, professional, comparable to GS-11 through GS-13. * - depends on type of sample. 9.3 Training For more experienced, higher grade personnel, formal training in special fields, possibly leading to specialization, should be almost mandatory. Such training can be fostered through local institutions and through the training courses provided by the Environmental Protection Agency. Regional policies on after-hours, government-supported training should be properly publicized. Formalized training for lower grade personnel, comparable to GS-3 to GS-5, is relatively scarce. However, skills can be most efficiently improved at the bench level on a personal, informal basis by more experienced analysts working in the same area. Exposure to pertinent literature should also be a definite program policy. 9-4 * U. S. GOVERNMENT PKINTtNG OKMCL*; 1372 O 456-212 CAPCO JEN 0005456 FRIDAY, APRIL 6, 1973 WASHINGTON, D.C. Volume 38 Number 66 PART I! CAP-1834 ENVIRONMENTAL PROTECTION AGENCY NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR POLLUTANTS Asbestos, Beryllium, and Mercury CAPCO JEN 0005586 S820 ' RULES AND REGULATIONS Title 40--Protection of Environment branes lining the chest and abdomen it is necessary to control emissions from CHAPTER 1--ENVIRONMENTAL PROTECTION AGENCY SUBCHAPTER C--AIR PROGRAMS (30-47). There are reports of mesotheli oma associated with nonoccupatlonal exposures in the neighborhood of as bestos sources (38, 42, 47, 48). An out major man-made sources of asbestos emissions into the atmosphere, but that it is not necessary to prohibit all emissions. PART 61--NATIONAL EMISSION STAND standing feature has been the long In this determination, the Administra ARDS FOR HAZARDOUS AIR POLLUTANTS period, commonly over 30 years, between tor has relied on the National Academy Asbestos, Beryllium, and Mercury the first exposure to asbestos and the ap of Sciences' report on asbestos (53), pearance of a tumor (49, SO). There is which concludes: "Asbestos is too im On March 31, 1971 (36 FR 5931) , pur evidence which indicates that mesothli- portant in our technology and economy suant to section 112 o the Clean Air Act, omas occur after much les exposure to for its essential use to be stopped. But, as amended, the Administrator published asbestos dust than the exposure associ because of the known serious effects of an Initial list ol three hazardous air pol ated with asbestos (51. 52). uncontrolled inhalation of asbestos min lutants which, in his Judgment may It is not practicable, at this time,, to erals in industry and uncertainty as to cause, or contribute to, an increase in establish allowable numerical concentra the shape and character of the dose- mortality or an increase in serious ir tions or mass emission limits for asbestos. response curve in man. it would be highly reversible. or incapacitating reversible, Satisfactory means of measuring ambient imprudent to permit additional contami illness. The pollutants were asbestos, asbestos concentrations have only re nation of the public environment with beryllium, and mercury. On December 7, cently been developed, and satisfactory asbestos. Continued use at minimal risk 1971 (36 FR 23239), the Administrator means of measuring asbestos emissions to the public requires that the major proposed standards for these pollutants. are still unavailable. Even If satisfactory sources of man-made asbestose emission Interested persons participated in the means of measuring asbestos emissions into the atmosphere be defined and con rulemaking by giving testimony at public did exist, the previous unavailability of a trolled." hearings and by sending comments to satisfactory means of measuring ambient The means of control used are limita EPA. Public hearings were held in New levels of asbestos makes it impossible to tions on visible emissions with an option York City on January 18, 1972, and in estimate even roughly the quantitative in some cases to use designated control Los Angeles on February 15 and 16, 1972. relationship between asbestos-caused ill equipment, requirements that certain A third hearing, scheduled to be held ness and the doses which caused those ill procedures be followed, and prohibitions in Kansas City, on February 1. 1972, was nesses. This is a major problem, since on the use of certain materials or of cer canceled because of a lack of requests to some asbestos caused illnesses have a 30- tain operations. These means of control participate. Sixty-eight persons gave year latency period. are required because of the impossibility testimony at the public hearings, and 56 EPA considered the possibility of ban at this time of prescribing and enforc persons sent comments to EPA. Repre ning production, processing, and use of ing allowable numerical concentrations sented were industries, universities, gov asbestos or banning all emissions of as or mas3 emission limitations known to ernmental agencies--Federal. State, and bestos into the atmosphere, but rejected provide an ample margin of safety. The local, and environmental groups. Copies these approaches. The problem of meas alternative of no control of the sources of the public hearing records are avail uring asbestos emissions would make the subject to this standard was rejected able at all EPA Regional Offices and at latter approach impossible to enforce. because of the significant health hazard the Division of Stationary Source En Either approach would result in the pro of unregulated emissions of asbestos into forcement, room 3220. 401 M Street S\V,, hibition of. many activities which are the atmosphere from the designated Washington. D.C. 20460, where copies of extremely important: moreover, the major sources. the comments received are also available. available evidence relating to the health It is the Administrator's judgment The bases for the Administrator's de terminations that asbestos, beryllium, and mercury are hazardous, the deriva tions of the standards now adopted, the Environmental Protection Agency's re sponses to the significant comments received, and the principal revisions to the proposed standards are summarized below. A more detailed statement is available on request from the Emission Standards and Enginering Division, En vironmental Protection Agency, Re search Triangle Park, N.C. 27711. Atten tion: Mr. Don. Goodwin. In addition, the Administrator is issuing Information on control techniques for asbestos, beryl lium, and mercury as directed by section 112(b)(2) of the act. Copies of these documents may be obtained free of charge from EPA Regional Offices. hazards of asbestos does not suggest that such prohibition is necessary to protect public health. For example, demolition of any building containing asbestos fire proofing or insulating materials would have to be prohibited as would the use of materials containing even trace amounts of asbestos which could escape into the atmosphere. Finally, the available evidence suggests a gradient of effects from direct occupa tional, to indirect occupational exposure, to families of workers exposed to asbestos and persons in the neighborhood of as bestos sources--in all of which situa tions asbestos concentrations are un doubtedly high by comparison with most community air. This suggests that there are levels of asbestos exposure that will not be associated with any detectable risk, although these levels are not that the asbestos sources subject to this standard are the major sources of as bestos emissions. In the absence of quan titative emission data, the Administra tor's judgment was based on an national inventory of sources and emissions of asbestos (Si) and other reports (53, 55). The asbestos emissions and emission factors presented in the national inven tory were based on information obtained from production and reprocessing com panies. This information included pro duction figures, estimates of control equipment efficiency and material bal ances; it did not include emission test results. The major sources of asbestos emissions were considered to fall into five categories: (1) Mining and milling; (2) manufacturing; (3) fabrication; (4) de molition; and (5) spraying. In deter mining which of these major sources ASBEST03 known (53). should be covered by the standard pro Asbestos is a hazardous air pollutant within the meaning of section 112. Many persons exposed to asbestos dust de veloped asbestosis when the dust concen tration was high or the duration of ex posure was long (1-7). A large number of studies have shown that there Is an association between occupational ex posure to asbestos and a higher-thancxpected incidence of bronchial cancer (8-30). Asbestos also has been identified as a causal factor in the development of mesotheliomas, cancers of the mem- It is probable that the effects of as mulgated herein, the Administrator con bestos inhalation are cumulative; that is, sidered the effect other Federal regula low-level and/or Intermittent exposure tions will have on the emissions from to asbestos over a long time may be such sources and the proximity of such equally as Important in the etiology of sources to the public. In addition, the asbestotlc disease as high level and/or Administrator considered comments on continuous exposure over a shorter pe-. the proposed standard ar.d additional riod. On the other hand, the available technical data not available before pro evidence does not Indicate that levels posal. The following paragraphs explain of asbestos in most community air cause these considerations and the changes asbestotic disease. Taking both these made to the standard between proposal considerations into account, the Admin and final promulgation. istrator has determined that. In order to The promulgated standard applies to provide an ample margin of safety to asbestos mills, selected manufacturing References at end of article. protect the public health from asbestos. operations, the use of spray-on asbestos FEDERAL REGISTER, VOL. 38, NO. 66--F> DAY. APRIL 6, 1973 CAPCO JEN 0005587 RULES AND REGULATIONS 8821 materials, demolition operations, and the surfacing of roadways with asbestos tail ings. The Administrator will continue to Investigate other existing and new sources of asbestos emission and if any of them a.-e found to be major sources, the standard will be revised to cover them. As applied to mines, the proposed standard would have limited the emis sions from drilling operations and pro hibited visible emissions of particulate matter from mine roads surfaced with asbestos tailings. The Bureau of Mines has prescribed health and safety regula tions 130 CFR 55.5) for the purpose of protecting life, the promotion of health and safety, and the prevention of acci dents in open pit metal and nonmetalitc mines. As related to asbestos mines, these regulations prohibit persons working In a mine from being exposed to asbestos concentrations which exceed the thresh old limit value adopted by the American Conference of Governmental Industrial Hygienists. The regulations specify that respirators shall not be used to prevent persons from being exposed to asbestos where environmental measures are avail able. For drilling operations, the regula tions require that the holes be collared and drilled wet. The regulations recom mend that haulage roads, rock transfer points, crushers, and other points where dust (asbestos) is produced sufficient to cause a health or safety hazard be wetted down as often as necessary unless the dust is controlled adequately by other means. In the judgment of the Admin istrator. Implementation of these regu lations will prevent asbestos mines from being a major source which must be cov ered by the standard promulgated here in. Furthermore, the public Is sufficiently removed from the mine work environ ment that their exposure should be sig nificantly less than that of the workers in the work environment. Accordingly, the promulgated standard does not apply to drilling operations or roadways at mine locations. For asbestos mills, the proposed stand ard would have applied to ore dumps, open storage areas for asbestos materials, tailings dumps, ore dryers, air for proc essing ore, air for exhausting particulate material from work areas, and any mill ing operation which continuously gen erates Inplant visible emissions. The promulgated standard prohibits visible emissions from any part of the mill, but It does not apply to dumps of asbestos tailings or open storage of asbestos ores. The Bureau of Mines' regulations pre viously referenced and regulations Issued by the Occupational Safety and Health Administration (20 CFR 1910.93?.) pro tect from the hazards of air con- t.v.r.in.'.'.v. in the work environment. The r.",;-' Safety and He?l*h .'cinln- : : . . :'a1 .or.s were ;;ro~jl-":ad ; 1. The regulations are in tent;.': to :>i' ,i; c the health of employees fro.--, t-exposure by means of en- cir.ee.- ng r-.r.trols (i.e. isolation, enclo sures.; : 1 'i':-:t collection) rather than by personal protective equipment. It Ls the jii.lm-.ent of the Administrator that measures taken to comply with the Bu The proposed standard would have reau of Mines and Occupational Safety limited emissions from a number of and Health Administration regulations to sources by stipulating that such emis protect the health of persons who work sions could not exceed the amounts which in proximity to dumps and open storage would be emitted from the source if the areas will prevent the dumps and storage source were equipped with a fabric filter; areas from being major sources of asbes or, in some cases, a wet-collection air tos emissions. cleaning device. This would have required The proposed standard would have ap a standardized emission-measuring tech plied to buildings, structures, or facilities nique, which ls not currently available. within which any fabricating or manu The promulgated standard prohibits visi facturing operation b carried on which ble emissions which contain asbestos and involves the use of asbestos materials. provides the option of using specified Comments received on the proposed air-cleaning methods. The existence of standard indicated that the requirements particulate asbestos material In a gas for fabricating and manufacturing oper stream vented to the atmosphere can be ations were confusing. Much of the con determined by collecting a sample on a fusion was created by the use of terms filter and analyzing it by microscopy such as 'any," "continuously," and techniques. The proposed standard stated "forced gas streams." The promulgated that the alr-cleanlng requirement would standard is more definitive as to applica not be met If a number of listed faults, bility of the provisions. The promulgated e.g, broken bags, leaking gases, thread standard prohibits risible emissions from bare bags, existed and It required that the nine manufacturing operations collection hoppers on some baghouses be which. In the judgment of the Adminis emptied without generating visible emis trator, are major sources of asbestos. The sions. Comments received suggested that promulgated standard does not cover this negative approach tended to make fabrication operations. Of all fabrication the quality of alr-cleanlng operations de operations, only those. operalions at new pendent upon the ability of EPA-te an construction sites are considered to be ticipate and to include In the standard major sources of asbestos emissions. The all the factors which would constitute Occupational Safety and Health Admin improper methods. Since the Intent was, istration regulations specify that all and is, to require high quality air-clean hand- or power-operated tools (l.e. saws, ing operations, the promulgated standard scorers, abrasive wheels, and drills) requires proper Installation, use, opera which produce asbestos dust be provided tion, and maintenance without precisely with dust collection systems. In the Judg defining the means to be used. ment of the Administrator, Implementa The proposed standard would have tion of these regulations will prevent prohibited the spraying of any material fabrication operations from being a containing asbestos on any portion of major source which must be covered by a building or structure, prohibited ihe the standard promulgated herein. spraying of any material containing as The proposed standard would have bestos in an area directly open to the prohibited visible emissions of asbestos atmosphere, and limited emissions from particulate material from the repair or all other spraying of any material con demolition of any building or structure taining asbestos to the amount which other than a single-family dwelling. would be emitted If specified air-cleaning Comments Indicated that the no visible equipment were used. Comments re emission requirement would prohibit re ceived pointed out that this standard pair or demolition In many situations, would: (1) Prohibit the use of materials since It would be impracticable, if not containing only the trace amounts of Impossible, to do such work without cre asbestos which occur in numerous nat ating risible emissions. Accordingly, the ural substances, (2) prohibit the use of promulgated standard specifies certain materials to which very small quantities work practices which must be followed of asbestos are added in order to enhance when demolishing certain buildings or their effectiveness, and (3) prohibit the structures. The standard covers institu use of materials In which the asbestos ls tional, Industrial, and commercial build strongly bound and which would not gen ings or structures, including apartment erate particulate asbestos emissions. The houses having more than four dwelling promulgated standard applies to those units, which contain friable asbestos ma uses of spray-on asbestos materials terial. This coverage ls based on the Na which could generate major emissions of tional Academy of Sciences! report (S3) particulate asbestos material. For those which states, "In general, single-family spray-on materials used to Insulate or residential structures contain only small fireproof buildings, structures, pipes, and amounts of asbestos Insulation. Demoli conduits, the standard limits the asbestos tion of Industrial and commercial build content to no more than 1 percent. Ma ings that have been fireproofed with, terials currently used contain from 10- asbestos-containing materials will prove to 80-percent asbestos. The intent of the to be _n emission source In the future, 1-percer.t iimtt is to ben the i::e ofma requiring control measures." Apartment hm:s"s -.l:h four dwelling units or less are considered to be equivalent to single- family residential structures. The stand ard requires that the Administrator be notified at least 20 days prior to the com terials which contain sigrificwn quanti ties of asbestos, but to alio-v ;h? iv,e of materials which would: (i> Contain trace amounts of asbestos which occur in numerous natural substances, and i2) Include very small quantities of asbestos (less than 1 percent) added to enhance mencement of demolition. the material's effectiveness. Although a FcDERAl REGISTER, VOl. 38, HO. 66--FRIDAY-, APRIL 6, 1973 CAPCO JEN 0005588 8822 RULES AND REGULATIONS standardized reference method has not standard will not require disposal. Where 17. Enterline, P. B., and M. A. Kendrick: been developed to quantitatively deter disposal is required, the Occupational Asbestos-dust Exposures it Various Levels mine the content of asbestos In a ma terial. there are acceptable methods available, based on electron microscopy, which Independent laboratories have de veloped. Determining the asbestos con tent of a material with these methods Safety and Health Administration regu lations (29 CFR 1910.93a(h>) require that any asbestos waste, consigned for disposal, be collected and disposed of in sealed Impermeable bags or other closed. Impermeable containers. The contamina and Mortality. Arch. Envir. Health. IS, 181- 160. 1967. % 18. Gloyne, 3. R.: Pneumoconiosis: A His tological Survey of Necropsy Material in 1.205 Cases. Lancet, l, 810-814. 1951. 19. Isselbacher. N. J., M. Klaus, and H. L. Hardy: Asbestosls and Bronchogenic Carci costs approximately 5300, and the results tion of ground water supplies with asbes noma: Report of one autopsted case and re axe accurate within plus or minus 50 percent; these limits on accuracy were taken Into account In establishing the 1-percent limitation. The proposed standard would have prohibited the surfacing of any roadway with asbestos tailings. The promulgated standard applies to all roadways except tos from landfill disposal is not consid ered a potential problem. The substitution of ceramic wool, min eral wool, and fiberglass for asbestos Is not now known to be a problem. There is no evidence that these materials cause health effects in the concentrations found In occupational or ambient environments. view of tbe available literature. Am. J. Med.. 25, 731-732,1953. 20. Jacob. 3.. and M. Anspach: Pulmonary `Neoplasta Among Dresden Asbestos Workers. Ann. N.Y, Acad. Sd., 232,538-548.1965. 21. Klelnfeld, M., J. Messtte, and O. Kobyxnan: Mortality Experience In a Group of As bestos Workers. Arch. Envir. Health, 25, 177180. 1987. those on ore deposits; these roadways are Although the standard was not based 22. Knox, J. P,, R. S. DoU, and X. D. HU1: temporary, and control measures taken to comply with the Bureau of Mines reg ulations prevent them from being ' a major source which must be covered by the standard promulgated herein. At this time, the application of asbestos tailings to public roadways Is not widely prac ticed. but because of the close proximity of roads to the public, a ban on using asbestos tailings on roadways Is Included In the promulgated standard to avoid a future problem and stop the practice where It Is followed. The term "surfac ing" Is defined to Include the deposit of on economic considerations, EPA Is aware of the Impact (55) and considers it to be reasonable. Costs among the various sources covered by the standard are quite variable. Although the standard may ad versely affect some Individual plants or companies which are marginal opera tions, It appears that such effects will be minimal and the Impact to the asbestos Industries as a whole will not be large. Rxfeezxcxs 1. Cooke, W. E.: Flbro3ls of the Lungs due to tbe Inhalation of Asbestos Dust. Brit. Med. J., 2.147.1924. Cohort Analysis of Changes In Incidence of Bronchial Carcinoma In a Textile Asbestos Factory. Ann. N.Y. Acad. Scl., 132, 526-535, 1965. 2J. Knox, J. P.. 3. Holmes, R. Doll, and I. D. Hill: Mortality from Lung Cancer and Other Causes Among Workers in an Asbestos Textile Factory. Brit. J. Ind. Med., 25, 293-303, 1968. 24. Ueben. J.: Malignancies In Asbestos Workers. Arch. Envir. Health. 13. 619-821, 1968. 25. Lynch, K. M., and W. A. Smith: Pul monary Asbestosls, Ed, Carcinoma of Lung In Asbestos-silicosis. Am. J. Cancer, 14, 58-64. 1935. 26. Mancuso, T. P., and A. A. El-Attar: asbestos tailings on roadways covered 2. Cooke, W. E.: Pulmonary Asbestosls. Mortality Pattern In a Cohort of Asbestos with snow or Ice; therefore, this practice is prohibited. Consideration was given to Including provisions In the standard requiring proper disposal of the asbestos material generated during demolition and col Brit. Med. 1,, 2. 1024-1025, 1927. 3. Dreessen. W. C., J. M. Dallavalle, T. XL Edwards. J. W. Miller, and R. R. Sayers: A Study of Asbestoe In tbe Asbestos TextUe In dustry Public Health Bull. 241. Washington. UB. Government Printing Office, 1938,126 pp. 4. McDonald. S.: History of Pulmonary As Worker*. J. Occup. Med., 9, 147-162, 1967. 27. McDonald. J. C.. A. D. McDonald. D. W. Gibbs, J. Slemlatyckl, and C. E. Rossiter: Mortality^ in the ChrysotUe Asbestos Mines and Mills of Quebec. Arch. Envir. Health, 22, 677-888, 1971. 28. Merewether, E. R. A.: Asbestosls and lected in control devices used to comply bestos!s, Brit. Med. J., 2. 1025-1026. 1927. Carcinoma of the Lung. In: Annual report of with the requirements of this standard. 5. Merewether, E. B. A.: Tbe Occurrence of the chief inspector of factories for the year It was decided that this was not neces sary because the Occupational Safety and Health Administration regulations <29 CFR 1910.93a(hl) Include house keeping and waste disposal requirements. Pulmonary Fibrosis and `Other Pulmonary Affections In Asbestos Workers, J. Ind. Hyg.. 12. 198-222. and 12, 233-257. 1930. 6. Mills, R. Q.: Pulmonary Asbestosls: Re port of a case. Minn. Med., 13, 495-499, 1930. 7. Soper, W. B.: Pulmonary Aabehtosls. A 1947. London: M. Y. Stationary Office, 1949, 73 pp. 29. Newhouse, M. r.: A Study of the Mor tality of Workers In an Asbestos Factory. Brit. J. Ind. Med.. 26.394-301,1969. 20. Selikoff. X. J., J. Churg, and E. C. Ham These regulations require that any as report of a caso and a review. Am. Rev. mond: Asbestos Exposure and Neoplasia. bestos waste, consigned for disposal, be Tuberc- 22, 571-584, 1930. JAMA, 188, 22-26.1964. collected and disposed of In sealed im 8. Bonser. O. M., J. S. Faulds, and M. J. 31. Borow, M., A. Conston, L. L. Llvornese. permeable bags or other closed. Imperme able containers. The potential environmental impact of the promulgated standard was evalu ated. and it was concluded that the Stewart: Occupational Cancer of tbe Urinary Bladder In Dyestuffs Operatives and of the Lung In Asbestos Textile Workers and Ironore Miners. Am. J. Clin. Path., 25. 126-134. 1955. 9. Braun. D. C.. and T. D. Truan: An and N. Schalet: Mesothelioma and Its Associ ation with Asbestoe. JAMA, 201, 587-591, 1967. 32. Elmes, P. C. W. T. E. McCaughey. and O. I*. Wade: Diffuse Mesothelioma of the pleura and Asbestos. Brit. Med. J., 2, 350- standard will not cause any adverse ef Epidemiological Study of Lung Cancer In As 353, 1965. fects. The potentially adverse environ bestos Miners. Arch. Znd. Health, 17, 634- 32. Elmes, P. C., and O. L. Wade: Relation mental effects of the standard are: (1) The asbestos-materials which will be collected In control devices and gen erated during demolition will have to be disposed of or recycled. (2) Materials, such as mineral wool, ceramic wool, and fiberglass, will be sub stituted for asbestos presently contained In spray-applied fireproofing and insulat 653. 1958. 10. Buchanan, W. D.: Asbestosls and Pri mary Intratboractc Neoplaama. Ann. N.T. Acad. Scl.. 132, 607-518.1965. 11. Cordova. J. F.. H. Tesluk, and R. P. Enudtson: Asbestosls and Carcinomas of the Lung. Cancer, 15, 1181-1187, 1962. 12. Doll, R.: Mortality from Lung Cancer In Asbestos Workers. Brit. J. Ind. Med., 12, 81-66. 1955. ship Between Exposure to Asbestos and Pleural Malignancy In Belfast. Ann. N.Y. Acad. ScL, 132, 549-557, 1965. 34. Entlcknap, J. B., and W. N. Smither: peritoneal Tumor In Asbestosls. Brit. J. Ind. Med., 22, 20-31,1964. # 35. Fowler. P. B. S., J. C. Sloper, and E. C. Warner: Exposure to Asbestos and Mesotheli oma of the Pleura. Brit. Med. J., 2, 211-213. 1964. ing materials. 13. Dunn, J. E., Jr., and J. M. Weir: A 36. Hammond, E. C,, I. J. Selikoff. and J. In some manufacturing operations, a Prospective Study of Mortality of Several Oc Churg: Neoplasia Among Insulation Workers major portion of the asbestos-material cupational Groups--Special Emphasis on collected by fabric filters is either re cycled to the process or is marketed for Lung Cancer. Arch. Envlr. Health, 17, 71-76. 1968. other u.-e.s. For example, one asbestos text.le recycles large quantities of lon?er-f:ber asbestos for process use and 14. Dunn. J. E.. Jr., and J. M. Weir: Cancer Experience of Several Occupational Greup3 Followed Prospectively. Am. J. Pub. Health. 55, 1367-1375, 1968. sells more than 90 percent of the remain ing collected materials to a brake lining 15. Elwood, P. C., and A. L. Cochrane: A Follow-up Study of Workers from an Asbestos manufacturer. Consequently, a signifi-' Factory. Brit. J. Ind. Med.. 21, 364-307. 1964. in the United States with Special Reference to Intraabdominal Neoplasia. Ann. N.Y. Acad. Scl.. 132, 519-525.1965. 37. Hourihane, D. 0*B.: The Pa`hc-I*^y of Mesothelioma and an Analysis of Ylt.*:r A:sociatlon with Asbestos Exposure. Tho^.ix, 29, 2G8-273. 1964.- 38. Lleben. J.. and H. Plstawka: Mcsc'helloma and Asbestos Exposure. Arch. E..vr. Health, 14, 559-563, 1367. cant portion of the Increased quantities 16. Enterline, P. E.: Mortality Among As 39. Mann. R. H.. J. L. Grcsh. and W. M. of "waste" asbestos materials which will bestos Product Workers In the United States. O'Donnell: Mesothelioma Associated with result from the Implementation of the Ann. :,\Y. Acad. Scl.. 132, 156-165. 1965. Asbestosls. Cancer, 19, 521-526, 1955. FEDERAL REGISTER, VOl. 38, NO. 65---- FRIDAY, APRIL 6, 1973 CAPCO JEN 0005589 RULES AND REGULATIONS 8823 40. McCau?hey, W. T. B., O. L. Wad, and of these were most likely due to exposure tions used to make the dispersion es p. C. ElmeJi Exposure to Asbestoe Dust and prior to the institution of controls, proper timates are given in the Background In Diffuse Pleural Mesotheliomas. Brit. Med. J, assessment of the period of exposure Is formation Report for Asbestos, Beryl Z. 1397. 1963. 41. McDonald. A. D,, A. Harper, O. A. El- Attar, and J. C. McDonald: Epidemiology of Primary Malignant Mesothellal Tumors In Canada. Cancer. 26, 914-919, 1970. 42. ^ewhouse, iL L,, and EL Thompson: Epidemiology of Mesothellal Tumors In the not always possible (f, 2): It Is known, however, that chronic beryllium disease Is associated not only with activities In volving extraction processes, but also that 64 registry cases resulted from exposure during machining operations on beryl lium. and Mercury (APTD-0753), pub lished at the time the standards were proposed. Rocket testing facilities are required to meet the limit of 75 mlcragram-miautes per cubic meter, accumulated dur London Area. Ann. N.Y. Acad. Set., 122, 579- lium materials (J). There are at least 45 ing any period of 2-consecutive weeks. 588, 1965. 42. Owen, W. O.: Mesothellal Tumors and Exposure to Asbestce Dust. Ann. N.Y. Acad. Scl.. 122, 674-679, 1965. 44. SeUkoff, L J, J. Churg. and 2. C. Ham mond: Relation Between Exposure to As bestos and Mesothelioma, New Eng. J. Med, cases of nonoccupattonallyincurred dis eases on file with the registry, of which approximately half have been fatal (3), and retrospective studies of the concen trations of beryllium that resulted in some cases of chronic beryllium disease The limit for rocket testing faculties is the same 33 that developed in 1966 by the Committee on Toxicology of the Na tional Academy of Sciences for protec tion of off-site personnel from intermit tent exposures to soluble beryllium com 272, 560-555, 1965. from nonoccupational exposure have pounds arising from the firing of rocket 45. Wright, O. W.: Asbestos and Health In concluded that the lowest concentration motors <i). 1969. Am, Rer. Reap. Dls, 100, 467-479, 1969. 46. Bellkoff, L J, E. C. Hammond, and J. Churg: Asbestos Exposure, Smoking, and Neoplasia. JAMA. 204, 106-112, 1968, 47. Wagner, J. C,, C. A. Sleggs. and P. Marchand: Diffuse Pleural Mesothelioma and Asbestos Exposure in the North Western which produced disease was greater than 0.01 ftg/m' and probably less than 0.10 ftg/m* (4). In 1949, when It became apparent that beryllium was a toxic material, the Atomic Energy Commission adopted a The proposed standard did not Include a provision on open burning of berylliumcontaining waste. The promulgated standard includes a ban on open burning of beryllium-containing waste. Thi3 change was made because information Cape Province, Brit. J. Tnd. Me<L, 17, 250-271, limit for beryllium concentrations In received after proposal indicated that 1960. 48. Champion. P.: Two cases of Malignant Mesotnelloma After Exposure to Asbestos. Am. Rer. Resp. Dls., 102, 821-326, 1971. 40. Sellkoff, L J,, and E. C. Hammond: En vironmental Epidemiology. DX Community Effects of Nonoccupatlonal Environmental Asbestos Exposure. Am. J. Pub. Health, 58, community air (l.e, 0.01 pg of beryllium per cubic meter of air averaged over a 30day period) (2). Beryllium refining com panies holding contracts with the AEC to operate AEC-owned refinery facilities and expand their own refinery capacity to meet AEC's beryllium requirements, such sources can cause ambient concen trations of beryUium in excess 0/ 0.01 pg/m* and because it 1s not possible to control the emissions from open burning. The promulgated standard does allow disposal of beryllium-containing waste in Incinerators which are controlled so 1653-1666. 1963. were required to observe the community as not to exceed the 10-gram-per-day 59. Wagner, J.C.: Epidemiology of Diffuse Mesothellal Tumors: Evidence of an Associa tion from Studies in South Africa and the United Kingdom. Ann. N.Y. Acad. ScL, 122, 675-578. 1965. 51. National Institute for Occupational Safety and Health: Occupational Exposures air limit. With the termination of these contracts in the 1961-63 period due to a reduction in AEC requirements for beryllium, the refineries were no longer subject to the AEC community air limit. The AEC's health and safety require limit. The disposal of beryllium-contain ing explosive wasto is Included In the standard covering rocket testing. The proposed standard would have covered all machining operations which use alloys containing any amount of be to Abe*toe (Criteria for a Recommended ments, however, have continued to apply ryllium. Comments were received which Standard). Washington, U.3. Department of Ileal: h. Education, and Welfare (FH3, HSMHA), 1972 (ESM 72-10267). 52. SeUkoff, I. J., W. J. Nicholson, and A. M. Langer: Asbestos Air Pollution. Arch. Envtr. Health. 25, 1-13. 1972, 53. National Academy of Sciences: Asbestos (The Need for and Feasibility of Air Pollu tion Controls). Washington, National Acad emy of Sciences, 1971, 40 pp. 54. National Inventory of Sources and Emissions--Cadmium, Nickel, and Asbestos. Report by W. E. Davis U Associates under contract to the Department of Health. Edu cation, and Welfare (Contract No. CPA 2263-131). Feb. 1970. 5$. Research Triangle Institute: Compre hensive Study of Specified Air Pollution Sources to Assess the Economic Impact of Air Quality Standards--Asbestos, Beryllium, Mer cury. Report prepared under contract to the Environmental Protection Agency (Contract No. 63-02-0068). Aug. 1972. Bravnirrac Beryllium la a hazardous air pollutant within the meaning of section 112. The proven effects of airborne beryllium ma terials on human health Include both acute and chronic lethal Inhalation ef to all AEC-owned facilities, some of which fabricate and assemble beryllium parts. In the period since the Implementation of the AEC guideline, no reported cases of chronic beryllium disease have oc curred as a result of community exposure, and the Committee on Toxicology of the National Academy of Sciences concluded that the AEC guideline limit represents a safe level of exposure (I). Accordingly, the Administrator has de termined that In order to provide an ample margin of safety to protect the public health from beryllium, sources of beryllium dust, fume, or mist emissions Into the atmosphere should be controlled to insure that ambient concentrations of beryllium do not exceed 0.01 pg/m'-- 30-day average. The beryllium standard covers extrac tion plants, foundries, ceramic manufac turing plants, machine shops (processing beryllium or beryllium alloys containing In excess of 5 percent beryllium) and disposal of beryllium-containing wastes. claimed that numerous machining opera tions use alloys containing low concen trations of beryUium and do not exceed the 10-gram-per-day emission limita tion. An investigation of these com ments revealed that alloys which include beryllium either contain a large amount (greater than 60 percent) or a small amount (less than 5 percent), and that approximately 8,000 machining opera tions use the low beryllium content al loys. Tests were conducted by the Agency to determine the berylUum emissions from the operations which use the low beryUium content alloys (e.g. stamping, tube drawing, mUling, and sawing). The results indicated that even If the emis sions were vented to the outside air, which they ordinarily are not, they would be significantly below the 10-gram-perday emission limitation. After consider ing these results and the administrative burden If the standard applied to such a large number of sources, the proposed standard was changed to exempt the machining operations which use alloys containing less than 5-percent beryllium. fects (I. 2), as well as skin and conjunc tiva! effects (2). Insufficient data are ail.-.b'e to incriminate beryllium as a ! :..~;r.o-,en but the .ock of o'. .'.-..chamsm for the total elimina- Most affected beryllium sources are lim ited to emissions of not more than 10 grams per d -y. This level was determined through dispersion estimates ns ;ke level The proposed standard would have al lowed all sources of beryllium to choose hetween meeting *'.-.e Iff-::rnm-: cr-rv.- en'.sslcn limit c.vi ccri.wni J" r - t.rn of lyrylLum body burdens, ar.d the which would protect against the occur 0.01 t g.'m* 20-dr.y average is never ex rev.iiv.n~ possibly long residence time mry enhance the opportunity for cancer induction. The Beryllium Registry now c rntains over 820 proven cases of berylilum-re'.ated disease (2), but since many P.-Aorer.ccs at cad of article. rence of 30-day average ambient concen trations exceeding 0.01 pg/m\ Tho sources covered by the standard are the only known ones that could result in am bient beryllium concentrations In excess of 0.01 pg/m\ The assumptions and cqua- ceeded. After reconsidering the proposed standard and the difficulty Inherent in using ambient air quality data, as op posed to emission data, as a regulatory tool. It was decided to limit the use of ambient data as a means of compliance FEDERAL REGISTER, VOl. 38, NO. 66--FRIDAY. APRIL 6, 1973 CAPCO JEN 0005590 SS24 RULES AND REGULATIONS to those sources which have demon strated over a reasonable past period that they can meet and have met the ambient limitation. Therefore, the stand ard being promulgated herein allows the ambient option only to existing sources which have 3 years of current ambient air quality data which demonstrate to the Administrator's satisfaction that the 0.01 rig/m1 level can be met In the vicinity of the source. A minimum of 3 years of data was judged to be necessary to dem onstrate that the ambient guideline of 0.01 r*g/m* (30-day average) can be met because of the possibility of monthly, seasonal, and even annual variations in ambient levels caused by variations in meteorology and production. The exist ing sources which could qualify or this option are four beryllium extraction plants and. possibly, a small number of machine shops. These sources were de signed or modified to facilitate compli ance with the 0.01 pg/m ambient limit. The potential environmental impact of this standard was evaluated and It was concluded that the standard will not cause any adverse effects. Beryllium is a very expensive material, and most gas streams emitting significant quantities of beryllium are controlled with high ef ficiency dry collectors, and the collected material Is recycled or sold back to the primary producers. Wet collectors are rarely used strictly as an air pollution control device, but more often as an ex traction process control device allowing recycle of waste liquids to the process. Absolute filters are often used as final filters and collect small quantities of beryllium from very low concentration gas streams. These filters are usually buried in company owned or segregated dumps or stored In unused mines or buildings. Most of the solid wastes are prepackaged prior to burial to prevent escape of beryllium to the environment. Although the standard Is not based on economic considerations, EPA is aware of the economic Impact (5) of the stand ard. Since most of the sources of beryl lium emissions are already controlled and In compliance with the standard, the economic Impact will be very small. - Reizkznces 1. Committee on Toxicology, National Acad emy of Sciences: Air Quality criteria for Beryllium, and Its Compounds. Report pre pared under contract to the UJ3. Public Health Service (Contract N70HI-291(61)), Washington, March 1.1968. 2. National Institute for Occupational Safety and Health: Occupational Exposure to Beryllium (Criteria for a Reoommeaded Standard). Washington, XJS. Department of Health, Education, and Welfare (PHS, HSMHA), 1972 (HSM 72-10288). 3. Mas-i.ichusetts . Oeneral Hospital, US. Beryllium Case Registry, Boston, Mass. 4. E.ser.bud, M., R. C. Wanta, C. Dustan, L. T. S-e ui:r.an. W. B. Harris, and 3 S. Wolf: Nonoccun.tiiop.ai Berylliosis. J. Ond. il7g. Toxicol., si. .:S2-20+, 1049. 5. Research Triangle Institute: Compre hensive Study of Specified Air Pollution Sources to Ac sess the Economic Impact of Air Quality Standards--Asbestos, Beryllium, Mer cury. Report prepared under contract to the Environmental Protection Agency (Contract No. 68-02-0088). August 1972. Mercury Mercury is a hazardous air pollutant within the meaning of section 112. Ex posure to metallic mercury vapors may cause central nervous system Injury, and renal damage U, 3). Experience with mercury vapor comes almost exclusively from animal experiments and Industrial exposures. Animal (rat) data Indicate a risk of accumulation in critical systems upon proionged exposure, with a poten tial, for example, for selective brain dam age (2, 3). Prolonged exposure to about 100 micrograms mercury per cubic meter of air involves a definite risk of mercury Intoxication (3). To determine the ambient air level of mercury that does not Impair health, the airborne burden must be considered to gether with the water- and food-bome burdens. An expert group concluded, based on Its analysis of several episodes of mercury poisoning In Japan, that 4 micrograms of methyimercury per kilo gram of bodywelght per day would result In the Intoxication of a sensitive adult: application of a safety factor of 10 yielded an acceptable exposure of about 30 micrograms per day for a 70-kllogram man, and this level Is also believed to provide satisfactory protection against genetic lesions, and poisoning of the fetus and of children (4). It should be noted that methyimercury Is considered to be by far the most haz ardous mercury compound, particularly via the ingestion of fish in which it has been concentrated through the food chain, (3, 5). The Environmental Protec tion Agency, in view of the present lim ited knowledge as to the effects of in haled mercury in the general population, and In order to best assure the requisite "ample margin of safety to protect the public health,'* has concluded that It is prudent to consider exposures to methyi mercury (diet) and mercury vapor (air) to be equivalent and additive. It has been estimated that from average diets, over a considerable period, mercury intakes of 10 micrograms per day may be expected (6), so that, in order to restrict total Intake to 30 micrograms per day, the average mercury Intake from air would have to be limited to 20 micrograms per day. Assuming inhalation of 20 cubic meters of air per day, the air could con tain an average daily concentration of no more than 1 microgram of mercury per cubic meter. The standard promulgated herein reg ulates the only two sources, mercury ore processing facilities and mercury cell chlor-alkall plants, which have been found to emit mercury in a manner that could cause the ambient concentration to exceed the Inhalation effects limits of X microgram per cubic meter. The stand ard limits emissions from these facilities to not more than 2,300 crams per day. The emission limit of 2.300 grams per day was derived from dispersion esti mates as the level which would protect against the violation of an average dally ambient concentration of 1 microgram per cubic meter. The assumptions and References at end of article. equations used to make the dispersion estimates are given in the Background Information Report for Asbestos, Beryl lium, and Mercury (APTD-0753), pub lished at the time the standards were proposed. Many mercury cell chlor-alkall plant cell rooms present severe source testing problems due to their design and con struction. Such sources may either recon struct the cell room so that accurate source tests can be made or employ housekeeping and maintenance practices that minimize mercury emissions from the cell room. Source test data and cal culations have Indicated that when such practices are used, 1.300 grams per day Is a reasonable estimate of emissions from the cell room. Therefore, when this option Is chosen, an emission of 1,300 grams per day will be assigned to the cell room. This permits emissions of not more than 1,000 grams per day from the hydro gen and end box ventilation streams c :bined. Compliance with the standard will be determined by the EPA reference method or EPA-approved substitute methods. Where a chlor-alkall plant chooses the housekeeping and maintenance practices option, determination of compliance of the cell room emission will be based on the use of EPA-approved practices. A list of approved practices may be obtained from EPA on request to. regional offices. The only major change In the mercury standard Is the Introduction of the above option of assigning an emission number to the cell room provided certain house keeping and maintenance requirements, are met. When this option is chosen, test ing is not required for emissions from the cell room. This option is offered because comments, testimony, and EPA source testing experience indicated that most existing cell rooms cannot be accurately tested for mercury emissions. Accurate emission tests are unduly complicated and costly because of the cell room configuration. Some of the changes suggested In writ ten comments and public hearing testi mony were considered by EPA but not made. The most significant one Involved the environmental chemistry of mercury, that is. environmental mercury In the at mosphere Is transformed to mercuric oxide by the action of ultraviolet radia tion, and since mercuric oxide Is not os toxic as elemental mercury, the stand ard should be less stringent. This argu ment is based on laboratory experiments under controlled conditions with gener ated radiation. The reaction cited In the testimony occurs when elemental mer cury Is Irradiated with ultraviolet light with a wavelength of 2,537 angstrom (A). Naturally occurring ozone in the upper atmosphere absorbs light in the ultra violet region below 3,000 A: !.T) hence the wavelength of ultraviolet necessary for the reaction is absent in the ambient at mosphere, and the reaction does not pro ceed at as high a rate as implied by the submitted testimony. Field measurements of both mercury vapors and particulate mercury In ambient air Indicate that os much as 96 percent of the mercury de FEDERAL REGIS7ER, VOL. 38, NO. 66--FRIDAY, APRIL i. 1973 CAPCO JEN 0005591 RULES AND REGULATIONS 8825 tected was in an elemental vapor form (data collected b7 EPA at the Federal Budding In Moundsvllle, W. Va.). The Environmental Protection Agency recognizes that mercury and Its com pounds constitute a multimedia contapounds constitute a multimedia contam ination problem, i.e.. strong evidence alter its natural distribution in the en vironment; that such uses may cause or hasten additional deposits Into water or soil over and above those occurring naturally, thereby building up environ mental concentrations; and the mercury levels accumulate In the biota, with the result that potentially dangerous residue levels are reached In foods consumed by man and animals. Current data on the environmental transport of mercury do not permit a clear assessment of the effect of mercury emissions into the atmosphere on the mercury content in the aquatic and ter restrial environments. Results of ongoing research will determine if there is a need for more comprehensive control of mer cury emissions Into the air. The stand ard promulgated herein is intended to protect the public health from the effects of inhaled mercury. The environmental impact of this standard was evaluated and it was con cluded th3t the standard will not cause any adverse effects since the control of mercury emissions to the atmosphere will have only minimal impact on other areas of environmental concern. The simplest control for mercury emissions to the atmosphere is cooling to condense the mercury. This cooling can be indirect or direct, iJy indirect cooling, the mer cury condenses and is retained for re cycle or sale. By direct cooling with a water scrubber, the water is usually re circulated after using centrifugal or gravitational separation to remove the mercury. The water cannot be reused indefinitely and eventually requires addi tional treatment to remove the mercury. In most cases, such treatment facilities are already being utilized to meet water quality standards. A widely used control device for par ticulate mercury emissions Is the mist eliminator. Residues in these devices are removed by gravity and washing with a recycled liquid. Another control method Is chemical scrubbing. In this system, scrubbing liquids are continuously made up while waste materials are usually re cycled to the process feed solutions. Re cycling of these liquids avoids significant contamination of water with mercury residues. The use of adsorption beds Is a highly efficient control method for removing mercury from gas streams. Two primary types arc available: (1) Chemically t.-ti-ared carbon beds. and - - - .`.eves, vfost of ti.e mercury c--..:- . i activated carbon can be rec!:u.r.i.-a by retorting the carbon but this usual;;.- destroys the carbon structure and necessitates disposal. Some small amount if residual mercury will remain with the carbon, but it Is tightly bound and is not easily transferred Into the air or v.aier. Regenerative molecular sieves do not cause a waste disposal problem because the sieves can be regenerated in place without retorting and can be reused many times. Although the standard was not based on economic considerations, EPA is aware of the impact (S) and considers it to be reasonable. Because mercury is an International commodity, world prices determine the fortunes of the domestic mercury mining industry. Historically, mercury prices fluctuate greatly in re sponse to small changes in- demand or supply. Domestic mercury mines are con sidered high-cost producers in relation to foreign producers. Because the average price has dropped from $404 per flask in 1969 to approximately $320 currently, the number of domestic mercury mines in operation has dropped sharply from 109 in 1969 to six or seven In March 1973. As long as the price of mercury remains below marginal costs of production (gen erally about S400), the remaining domes tic mines will be ill equipped to absorb any cost Increases. The total chlor-alkall Industry com prises 68 plants. Approximately 28 are mercury cell plants and account for about 27 percent of the UB. production of chlorine and caustic. The future of the chlorine-caustic in dustry appears healthy. Demand for chlorine is expected to grow at an annual rate of 6 percent projected from 1971. Demand for caustic soda will grow at least at the same rate as chlorine, and perhaps faster. Prices for chlorine and sodium hydroxide have been rising steadily through the sixties into 1971. Based on these trends, the cost of control to comply with the mercury standard will be passed forward to the consumer. Use of these two basic commodities is so di verse that any price increases will be well dispersed through all manufacturing activities. References 1. Report of an International Committee: Maximum AUo*.val>!e Concentrations of mer cury Compounds. Arch. Envlr. Health, 19, B91- 905. December 1959. 2. Clarkson. T. W.: The Pharmacology of Mercury Compounds. Ann. Rev. Pharmacol ogy. 12. 375--10G, 1972. 3. Prlberg. L., and J. Vostat (Eds.): Mer cury In the Environment--A Toxicological and Epidemiological Appraisal. Prepared by the Karolinska Institute Department of En vironmental Hygiene (Stockholm) for the U.S. Environmental Protection Agency (Office of Air Programs), November 1971. 4. Methylmercury in Fish: a Toxicologic- Epidemiologic Evaluation of Risks. Report from an expert group. Nord, Hyg. Tlsdkr. (Stockholm), Supplement 4, 1971 (English translation). 5. Nelson. N.. T. C. Byerly, A. C. Kolbye. Jr.. I,. T. Kurland. It. E. Shapiro, S. I. Slilbko, W. H. Stickle, J. E. Thompson, L. A. Van Den Berg, and A. V/eUsIer: Itararda of Mercury (`jecla! rcp-irc :o the Secretary's IV-.r-ctie .^a-.sory C ^epa.-n-.er.t ,-f lev.:-!:. Education, and ire, NV.--1575). LY.y.r. Res.. !. 1-51. l')7i. G. Westoo, G.: Me-cury In FoocIsnUts--Is There a Great Risk of roAonlnc? VAR FODA, 4. 1-G. 19G5. 7. Leighton. P. A.: Photochemistry of Air Pollution. Academic Press. 1SGI. S. Research Triangle Institute: Compre hensive Study of SpeclQed Air Pollution Sources to Assess the Economic Impact of Air Quality Standards--Asbestos, Beryllium, Mercury. Report prepared under contract to the Environmental Protection Agency (Con tract No. G3-02-0088). August 1972. General Provisions The standards promulgated below are applicable to new, modified, and existing sources. Any new or modified source must comply with the standards upon begin ning operation. Any existing source must comply with the standards within 90 days after promulgation, unless a waiver of compliance is granted. After considering the proposed general provisions and the comments received on them, the Administrator made several changes which are Included in the stand ards promulgated below. A new section was added to specifically require new sta tionary sources to notify the Administra tor before beginnning operation. The requirements for source reporting and request for waiver of compliance were combined into one section. The time for submitting the source report was ex tended from 30 to 90 days to provide sources with more time to complete the information required. Appendix A was added to provide sources' a description and format of the information required. The proposed standards required all sources of mercury and beryllium to test their emissions within 3 months of the effective date and at least once every 3 months thereafter: a provision was in cluded to allow the Administrator to waive the periodic tests for sources in compliance with a standard. Tile stand ards promulgated below require the ini tial test within 90 days of the effective date and include a provision to allow the Administrator to waive this requirement if the source is meeting the standard or has requested a waiver of compliance. Periodic tests are not required unless specifically requested by the Administra tor. The Administrator may cancel a waiver of emission tests and may require a test under the authority of section 114 of the Act at any time. Appendix A speci fies the information which a source must provide the Administrator when applying for a waiver of initial emission testing. The standards promulgated below do not require the owner or operator to request a waiver of compliance before a specific date. However, the owner or op erator should submit the request within 30 days after the effective date of the regulation to be assured that action will be taken on the waiver application prior to the 90th day after the effective date. Continued operation in excess of a stand ard after the 90th day without a waiver is a violation of the act. The Administrator may grant an exist ing source a waiver, permitting a period of up to 2 years for compliance, provided iC'd '}\i I'.vLw'il i**' .-.v.'-s ..i:i he protected ;.iia imminent cnciangerment and provided that .v.icii (.riod is necessary for the installation of controls. To be granted a waiver of com pliance. a source must submit a written request to the Administrator and pro vide certain Information to assist the Administrator in making a judgment. FEDERAL REGISTER, VOl. 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005592 S826 RULES AND REGULATIONS Within 60 days after receiving a request, ferred methods of sampling and analyz Sea the Administrator will notify the owner or operator of approval or intention to deny the waiver. Any waiver of com ing used to determine compliance. The reference methods for beryllium and mercury are Included in appendix B to 61.22 EmlssLon standard. 61.23 Air cleaning. 61.24 Reporting. pliance granted by the Administrator will be in writing and specify conditions the source must meet during the waiver period. If the Administrator Intends to deny a request, the owner or operator will be given a specified time to provide this part. An equivalent method is any Subpart C--National Emission Standard for method of sampling and analyzing which Beryllium has been demonstrated to- the Admin 61.30 Applicability. istrator's satisfaction to have a con sistent and quantitatively known rela 61.31 61.32 61.33 Definitions. Emission standard. Stack sampling. tionship to the reference method under 61.34 Air sampling. additional information or arguments prior to final action on the request. Final action on a request will be in writing by the Administrator, and if denied, will in clude reasons for denial. specified conditions. An alternative method is any method of sampling and analyzing which does not meet all the criteria for equivalency but which can be used In specific cases to determine com Subpart D--National Emission Standard for Beryllium Rocket Motor Firing 61.40 Applicability. 61.41 Definitions. 61.42 Emission standard. The President may exempt any new, pliance. Alternative methods may be ap 61.43 Emission testing--rocket firing Or pro modified, or existing stationary source proved by the Administrator for source pellant disposal. from compliance with the standards for testing; however. In cases where deter 61.44 Stack sampling. a period of up to 2 years, provided the minations of compliance using an alter Subpart E--National Emission Standard for technology is not available to implement native method are disputed, use of the Mercury the standards and the operation of such reference method or its equivalent will 61.50 Applicability. source is required for reasons of national be required by the Administrator. An ap 61.51 Definitions. security. Also, the President may grant exemptions for additional periods of 2 years or less. The construction of a new source or modification of an existing source cov ered by these standards cannot begin without approval of the Administrator. To obtain approval, the owner or opera tor of such sources must apply in writing to the Administrator. Within 60 days, the Administrator will notify the owner or operator of approval or Intention to proved alternative method for beryllium is included in appendix B hereto. All emission data provided to or ob tained by the Administrator hi carrying out these regulations will be available to the public. Records, reports, or informa tion other than trade secrets will be available to the public. Pursuant to section 112(d) (1) of the act. the Environmental Protection Agency intends to delegate the author ity to implement and enforce national 61.52 Emission standard. 61.53 Stack sampling. Appendix A--Compliance Status Information. Appendix B--Test Methods. Method 101--Reference method for determi nation of particulate and gaseous mercury emissions from stationary sources (air streams). Method 102--Reference method for determi nation of particulate and gaseous mercury emissions from stationary sources (hydro gen streams). Method 103--Beryllium screening method. Method 104--Reference method for determi deny approval. If the Administrator in emission standards (except with respect nation of beryllium emissions from sta tends to deny approval, a specified time to stationary sources owned or operated tionary sources. will be given to provide additional infor by the United States) for hazardous air mation or arguments prior to final action pollutants to any State which submits an on the application. The final action on adequate procedure to the Administrator. AnrHORmr: 42 tJJS.C. 1857c-7. Subpart A---General Provisions any application will be in writing by the The requisite procedure for requesting 61.01 Applicability. Administrator, and if denied, will in clude the reasons for denial. Although the demolition of buildings or structures containing asbestos ma terial and the spraying of asbestos ma such delegation will be issued in the future by the Environmental Protection Agency. The regulations for the national emis sion standards for asbestos, beryllium, The provisions of this part apply to the owner or operator of any stationary source for which a standard Is prescribed, under this part. terial will in many cases be modifications and mercury are hereby promulgated ef 61.02 Definitions. of existing stationary sources, the Ad fective upon promulgation (April 6, As used in .this part, all terms not de ministrator's approval is not required be 1973). fined herein shall have the meaning given fore beginning such operations. Section 112(c)(1) of the act specifies that no person may construct any new source or modify any existing source"* unless the Administrator finds that such source Dated; March 30, 1973. Robert W. Far, Acting Administrator, Environmental Protection Agency. them in the act: (a> "Act" means the Clean Air Act (42 UJ3.C. 1857 etseq.). (b) "Administrator'* means the Ad ministrator of the Environmental Pro If properly operated will not cause emis A new Part 61 Is added to Chapter 1, tection Agency or his authorized repre sions in violation of such standard." The Title 40. Code of Federal Regulations, as sentative. demolition and spraying provisions are follows: (c) "Alternative method" means any expressed in terms of procedures to be followed. Therefore, if the source is prop Subpart A--G#n*ral Provisions method of sampling and analyzing for an air pollutant which does not meet all of erly operated, it will be complying with 61.01 Applicability. the criteria for equivalency but which has the standard, and there is no need for 61.02 .Definitions. been demonstrated to the Administra the Administrator to make a finding with respect to each new source subject to these provisions. Each source covered by these stand ards is required to submit to the Admin 61.03 61.04 61.05 61.06 61.07 Abbreviations. Address. Prohibited activities. Determination of constructon or modification. Application for approval of construc tor's satisfaction to, In specific cases, pro duce results adequate for his determina tion of compliance. <d) "Commenced" means that an own er or operator has undertaken a con istrator within 90 days after promulga tion or modification. tinuous program of construction or tion certain Information pertaining to its 61.08 Approval by Administrator. modification or that an owner or operator operation. Changes in the information 61.09 Notification of startup. has entered into a contractual obligation, must be submitted within 30 days after the change, exceot where the change is considered a modification. Then the re quirements for a modified source are applicable. 61.10 01.11 61.12 Ct.l3 61.14 61.15 Source reporting and waiver request. Waiver of compliance. Emission t*s:s and monitoring. Waiver o! emission tests. Source test aud analytical methods. Availability of Information. to undertake and complete, within a rea sonable time, a continuous program of construction or modification. (e) "Compliance schedule" means the date or dates by which a source or cate Three terms are associated with deter 61.16 State authority. gory of sources Is required to comply with mining compliance by means of source testing: (1) Reference method, (2) Subpart B--National Emission Standard for Asbestos the standards of this part and with any steps toward such, compliance which are equivalent method, and (3) alternative 61.20 Applicability. set forth in a waiver of compliance under method. Reference methods are the pre 61 21 Definitions. 61.11. FEDERAL REGISTER, VOL 38. NO. 66--FRIDAY, APRIL 6, 1973' CAPCO JEN 0005593 RULES AND REGULATIONS S827 (O "Construction" means fabrication, yd'--Square yards, of which commenced after the publica erection, or Installation of a stationary w.g.--Water gage. tion date of the standards proposed to source. inHg--Incnes of mercury. tg) "Effective date" Is the date of InH.-O--Inches of water, be applicable to such source, are subject to this prohibition. promulgation la the Federal Register g--Grams, (b) After the effective date of any of an applicable standard or other regu mg--Milligrams. standard prescribed under this part, no lation under this part. N--Normal. (h) "Equivalent method" means any 'R--Degree Rar.kine. owner or operator shall operate any new source in violation of such standard ex method of sampling and analyzing for mm--Minute cept under an exemption granted by the an air pollutant which has been demon sec--Second, strated to the Administrator's satisfac avg.--Average. President under section 112(c) (2) of the act. tion to have a consistent and quantita I.D.--Inside diameter. tively known relationship to the reference O.D.--Outside diameter.. (c) Ninety days after the effective date of any standard prescribed under this method, under specified conditions. iig--Micrograms GO" gram). part, no owner or operator shall operate (1) "Existing source" means any sta %--Percent. any existing stationary source in viola tionary source which is not a new source. Hg--Mercury. (j > -Modification" means any physical Be--Beryllium. tion of such standard, except under a waiver granted by the Administrator in change in. or change in the method of operation of, a stationary source which increases the amount of any hazardous air pollutant emitted by such source or which results in the emission of any hazardous air pollutant not previously emitted, except that: (1) Routine maintenance, repair, and replacement shall not be considered physical changes, and (2) The following shall not be con sidered a change In the method of operation: (1) An increase In the production rate. If such increase does not exceed the op erating design capacity of the stationary source: (ii> An Increase in hours of operation, (k) "New source" means any stationary source, the construction or modification of which is commenced after the publi cation in the Federal Register of pro posed national emission standards for hazardous air pollutants which will be applicable to such source. (Ii 'Owner or operator" means any person who owns, leases, operates, con trols. or supervises a stationary source. (m> "Reference method" means any method of sampling and analyzing for an air pollutant, as described in ap pendix B to this part. fn) "Startup" means the setting in operation of a stationary source for any purpose. i'o> "Standard" means a national emission standard for a hazardous air pollutant proposed or promulgated under tills part. 61.01 AJilresn. All requests, reports, applications, sub mittals. and other communications to the Administrator pursuant to this part shall be submitted In duplicate and ad dressed to the appropriate regional office of the Environmental Protection Agency, to the attention of the Director, Enforce ment Division. The regional offices are as follows: Region I (Connecticut, Maine. Massa- chusetts.New Hampshire. Rhode Island. Vermont). John F. Kennedy Federal Building. Boston. Mass. 02203. Region II (New York. New Jersey. Puerto Rico, Virgin Islands). Federal Office Building. 26 Federal Plaza (Foley Square). New York, N.Y.. 10007. Region III i Delaware, District of Co lumbia. Pennsylvania, Maryland, Vir ginia. West Virginia), Curtis Building. Sixth and Walnut Streets, Philadelphia. Pa. 19106. Region IV ''Alabama, Florida. Georgia. Mississippi. Kentucky. North Carolina. South Carolina. Tennessee). Suite 300. 1421 Peachtree Street, Atlanta, Ga. 30309. Region V (Illinois, Indiana, Minne sota. Michigan. Ohio, Wisconsin), 1 North V,'acker Drive, Chicago, 111. 60606. Region VI (Arkansas, Louisiana, New Mexico, Oklahoma. Texas). 1600 Pater son Street. Dallas. Tex. 75201. Region VII (Iowa. Kansas. Missouri. Nebraska). 1735 Baltimore Street. Kan sas City. Mo. 64108. Region vm (Colorado. Montana. accordance with this subpart or under an exemption granted by the President under section 112(c) (2) of the act. (d) No owner or operator subject to the provisions of this part shall fall to report, revise reports, or report source test results as required under this part. 61.06 Determination of construction or modification. * Upon written application by an owner or operator, the Administrator will make a determination of whether actions taken or intended to be taken by such owner or operator constitute construction or modification or the commencement thereof within the meaning of this part. The Administrator will within 30 days of receipt of sufficient Information to evaluate an application, notify the owner or operator of his determination. 61.07 Application foe approval of construction or mod i (trillion. (a) Tile owner or operator of any new source (o which a standard prescribed under this part is applicable shall, prior to the date on which construction or modification is planned to commence, or within 30 days after the effective date in the case of a new source that already has commenced construction or modifi cation and has not begun operation, sub mit to the Administrator an application for approval of such construction or modification. A separate application shall be submitted for each stationary source. (b) Each application shall include: (1) The name and address of the ap plicant. (p) "Stationary source" means any building, structure, facility, or Installa tion which emits or may emit any air pollutant which has been designated as North Dakota. South Dakota, Utah. Wy oming), 916 Lincoln Towers. 1860 Lin coln Street, Denver, Colo. 80203. Region IX (Arizona, California, (2) Tlie location or proposed location of the source. (3) Technical information describing the proposed nature, size, design, operat hazardous by the Administrator. Hawaii, Nevada, Guam, American ing design capacity, and method of oper 61.03 Abbreviations. The abbreviations used in this part have the following meanings: "C--Degrees Centigrade, cfm--Cubic feet per minute, ft'--Square feet, ft'--Cuoic feet. Samoa). 100 California Street, San Francisco. Calif. 94111. Region X (Washington. Oregon, Idaho, Alaska). 1200 Sixth Avenue, Seattle, Wash. 93101. 61.05 I'mliililleii activities, (a) After the effective date of any ation of the source, including a descrip tion of any equipment to be used for control of emissions. Such technical In formation shall Include calculations of emission estimates in sufficient detail to permit assessment of the validity of such calculations. "F--D-a-ees Fahrenheit. standard prescribed under this part, no 61.08 Approval l>v Adlldnl-lr.itiir. owner or t-p-.v.tor sha'.! construct orinotl- i('i The .'uimir.ii;.-:;'.';.* . ;- - - <,) ify ..ivy - 'a': v.ary source subject to such days i-r I'vveipt .,( sud-cirir. :.:f.,'n si 'iid.ir'l without fir.,; obtaining written to evaluate an application uivier ; vl.07, approval of cite Administrator in accord notify the owner or operator of approval ance with this subpart, except under an or intention to deny approval of con : --Inomeier. r.z-- Oiut'es. v/v--\ -:ne per volume. exemption granted by tile President struction or modification. under section 112(c) >2) of the act. (b) If the Administrator dc;-.-rniines Sources, the construction or modification that a stationary source for which an No (t--Pt.n-------2 FfOESAl REGISTER, VOL. 38. NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005594 SS28 RULES AND REGULATIONS application pursuant to 5 61.07 was sub mitted will. If properly operated, not cause emissions in violation of a stand ard, he will approve the construction or modification of such source. <c> Prior to denying any application for approval of construction or modifica tion pursuant to this section, the Admin istrator will notify the owner or operator making such application of the Admin istrator's intention to issue such denial, together with: (1) Notice of the information and findings on which such Intended denial, is based, and <2> Notice of opportunity for such owner or operator to present, within such time limit as the Administrator shall specify, additional information or argu ments to the Administrator prior to final action on such application. (d) A final determination to deny any application for approval will be In writ ing and will set forth the specific grounds on which such denial Is based. Such final determination will be made within 60 days of presentation of additional infor mation or arguments, or 60 days after the final date specified for presentation, if no presentation is made. (e) Neither the submission of an ap plication for approval nor the Admin istrator's granting of approval to con struct or modify shall: (1) Relieve an owner or operator of legal responsibility for compliance with any applicable provision of this part or of any other applicable Federal. State, or local requirement, or (2) Prevent the Administrator from implementing or enforcing this part or taking any other action under the act. 61.09 Notification of startup. (a) Any owner or operator of a source which has an initial startup after the effective date of a standard prescribed under this part shall furnish the Admin istrator written notification as follows: (1) A notification of the anticipated date of initial startup of the source not more than 60 days nor less than 30 days prior to such date. (2) A notification of the actual date of Initial startup of the source within 15 days after such date. 61.10 Source reporling ami waiver re quest. (a) The owner or operator of any existing source, or any new source to which a standard prescribed under this part is applicable which had an initial startup which preceded the effective date of a standard prescribed under this part shall, within 90 days after the effective date, provide the following information in writing to the Administrator: (1) Name and address of the owner or operator. >2' The '.oration of the source. <3 i 1"-e of hazardous pollutants mifed by the stationary source. <4) A brie: description of the nature, size, design, and method of operation of the stationary source including the op erating design capacity of such source. Identify each point of emission for each hazardous pollutant. (5) The average weight per month of the hazardous materials being processed by the source, over the last 12 months preceding the date of the report. (6) A description of the existing con trol equipment for each emission point. (1) Primary control devicefs) for each hazardous pollutant. (11J Secondary control device(s) for each hazardous pollutant. <iii> Estimated control efficiency (per cent) for each control device. (7) A statement by the owner or oper ator of the source as to whether he can comply with the standards prescribed in this part within 90 days of the effective date. <b) The owner or operator of an exist ing source unable to operate in compli ance with any standard prescribed under this part may request a waiver of com pliance with such standard for a period not exceeding 2 years from the effective date. Any request shall be in writing and shall include the following information: (1)A description of the controls to be installed to comply with the standard. (2) A compliance schedule, including the date each step toward compliance will be reached. Such list shall Include as a minimum the following dates: (I) Date by which contracts for emis sion control systems or process modifica tions will be awarded, or date by which orders will be issued for the purchase of component parts to accomplish emis sion control or process modification: (II) Date of Initiation of onsite con struction or installation of emission con trol equipment or process change; (iii) Date by which onsite construc tion or installation of emission control equipment or process modification Is to be completed; and (iv) Date by which final compliance Is to be achieved. (3) A description of interim emission control steps which will be taken during the waiver period. (c) Changes In the Information pro vided under paragraph (a) of this section shall be provided to the Administrator within 30 days after such change, except that if changes will result from modifica tion of the source, as defined in 9 61.02 (1), the provisions of 5 61.07 and 5 61.08 are applicable. (d) Hie format for reporting under this section Is included as appendix A of this part.' Advice on reporting the status of compliance may be obtained from the Administrator. 61.11 Waiver of compliance. (a) Based on the information provided in any request under 5 61.10, or other in formation, the Administrator may grant a waiver of compliance with a standard for a period r.ot exceeding 2 years from ilie effective date of such itancard. ibi Such waiver will be in writing and will: (1) Identify the stationary source covered. (2) Specify the termination date of the waiver. Tile waiver may be termi nated at an earlier date if the conditions specified under paragraph (b) (3) of this section are not met. (31 Specify dates by which steps to ward compliance are to be taken: and impose such additional conditions as the Administrator determines to be neces sary to assure installation of the neces sary controls within the waiver period, and to assure protection of the health of persons during the waiver period. (c) Prior to denying any request for a waiver pursuant to this section, the Administrator will notify the owner or operator making such request of the Ad ministrator's intention to issue such denial, together with: (1) Notice of the Information and findings on which such intended denial is based, and (2) Notice-of opportunity for such owner or operator to present, within such time limit as the Administrator specifies, additional information or argu ments to the Administrator prior to final action on such request. (d) A final determination to deny any request for a waiver will be in writing and will set forth the specific grounds on which such denial is based. Such final determination will be made within 60 days after presentation of additional In formation or arguments, or 60 days after the final date specified for such presen tation, if no presentation is made. (e) The granting of a waiver under this section shall .not abrogate the Ad ministrator's authority under section 114 of the act. 61.12 Emiuion tests and monitoring. (a) Emission tests and monitoring shall be conducted and reported as set forth in this part and appendix B to this part. (b) The owner or operator of a new source subject to this part, and at the request of the Administrator, the owner or operator of an existing source sub ject to this part, shall provide or cause to be provided, emission testing facili ties as follows: (1) Sampling ports adequate for test methods applicable to such source. (2) Safe sampling platform(s). (3) Safe access to sampling platform(s). (4) Utilities for sampling and testing equipment. 61.13 Waiver of emission tests, (a) Emission tests may be waived upon written application to the Admin istrator if. In his judgment, the source is meeting the standard, or If the source is operating under a waiver of compliance or has requested a waiver of compliance. (b) If application for waiver of the emission test Is made, such application shall accompany the information re quired by 5 61.10. The appropriate form i.> contained ir. appendix A to this part. (c) Approval of any r.-aiv-r granted pursuant to this =ec::cn rhall r.ot abro gate the Administrator's authority under the act or in any way prohibit the Ad ministrator from later canceling such waiver. Such cancellation will be made only after notice Is given to the owner or operator of the source. FEOERAl REGISTER, VOL. 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005595 RULES AND REGULATIONS SS29 61.14 Source test ami analytical mclh- Oils. <a Methods 101, 102, and 104 in ap- Subpart B--National Emission Standard for Asbestos 61.20 Applicability. installation, or portion thereof which contains any boiler, pipe, or load-sup porting structural member that is insu lated or fireproofed with friable asbestos ( pendix B to this part shall be used for The piovisions of this subpart are ap material shall comply with the require all source tests required under this part, plicable to those sources specified in ments set forth in this paragraph. unless an equivalent method or an al 5 61.22. (1) Notice of intention to demolish ternative method has been approved by the Administrator. 61.21 Definitions. shall be provided to the Administrator at least 20 days prior to commencement ib> Method 103 in appendix B to this part is hereby approved by the Admin istrator as an alternative method for sources subject to 61.32<ai and 61.42 (b). <c> The Administrator may, after no tice to the owner or operator, withdraw approval of an alternative method granted under paragraph (a) or (b) of this section. Where the test results using an alternative method do not adequately indicate whether a source is in compli ance with a standard, the Administrator may require the use of the reference method or its equivalent. Terms used in this subpart are defined in the act. in subpart A of this part, or in this section as follows: () "Asbestos" means actinolite. amosite, anthophyllite. chrysotile, crocidolite, tremolite. <b) "Asbestos material" means as bestos or any material containing as bestos. <c) "Particulate asbestos material" means finely divided particles of asbestos material. <d> "Asbestos tailings" means any solid waste product of asbestos mining or milling operations which contains as bestos. of such demolition or anytime prior to commencement of demolition subject to paragraph (d) (4) of this section. Such notice shall include the following information: (i) Name of owner or operator. (ii) Address of owmer or operator. (ill) Description of the building, struc ture. facility, or Installation to be de molished. (iv) Address or location of the build ing. structure, facility or installation. (v) Scheduled starting and completion dates of demolition. (vi) Method of demolition to be em ployed. (>1.13 Availability of information. <e> "Outside air" means the air out ivii) Procedures to be employed to (a) Emission data provided to. or oth erwise obtained by, the Administrator in accordance with the provisions of this part shall be available to the public. ib) Any records, reports, or informa side buildings and structures. if) "Visible emissions" means any emissions which are visually detectable without the aid of instruments and which contain particulate asbestos material. meet the requirements of this paragraph. i2> The following procedures shall be used to prevent emissions of particulate asbestos material to outside air: (i) Friable asbestos materials, used to 'Wry tion. other than emission data, provided to. or otherwise obtained by, the Admin istrator in accordance with the piovisions of this part shall be available to the pub lic, except that upon a showing satisfac tory to the Administrator by any person that such records, reports, or informa tion. or particular part tlieieof iether than emission data), if made public, would divulge methods or processes en titled to protection as trade secrets of such person, the Administrator will con sider such records, reports, or informa tion, or particular part thereof, confi dential in accordance with the purposes of section 1005 of title 18 of the United States Code, except that such records, re ports, or information, or particular part thereof, may be disclosed to other officers, employees, or authorized representatives of the United States concerned with car rying out the provisions of the act or when relevant in any proceeding under tlie act. 61.16 State authority. fa) The-provisions of this part shall not be construed in any manner to pre clude any State or political subdivision thereof from: (1) Adopting and enforcing any emis 61.22 Emission standard. <a> Asbestos mills: There shall be no visible emissions to the outside air from any asbestos mill except as provided in paragraph <f) of this section. Outside storage of asbestos materials is not con sidered a part of an asbestos mill. b i Roadways: The surfacing of road ways v. nil asbestos tailings is prohibited, except for temporary roadways on an area of asbestos ore deposits. The deposi tion of asbestos tailings on roadways cov ered with snow or ice is considered "sur facing." (cj Manufacturing: There shall be no visible emissions to the outside air, ex cept .as provided in paragraph if) of this section, from any building or struc ture in which the following operations are conducted or directly from any of the following operations if they are con ducted outside of buildings or structures. tl) The manufacture of cloth, cord, wicks, tubing, tape, .twine, rope, thread, yam. roving, lap, or other textile ma terials. <2) The manufacture of cement prod ucts. (3) The manufacture of fireproofing and insulating materials. 14) The manufacture of friction products. insulate or fireproof any boiler, pipe, or load-supporting structural member, shall be wetted and removed from any build ing, structure, facility, or installation subject to this paragraph before wreck ing of load-supporting structural mem bers is commenced. The friable asbestos debris shall be wetted adequately to in sure that such debir: remains wet during all stages of demolition and related dling operations. i in No pipe or load-supporting : t tural member that is covered with iriable asbestos insulating or fireproofing material shall be dropped or thrown to the ground from any building, structure, facility, or installation subject to this paragraph, but shall be carefully low ered or taken to ground level. (iii) No friable asbestos debris shall be dropped or thrown to the ground from any building, structure, facility, or in stallation subject to this paragraph or from any floor to any floor below. For buildings, structures, facilities, or in stallations, 50 feet or greater in height, friable asbestos debris shall be trans ported to the ground via dust-tight chutes or containers. (3) Sources subject to this paragraph are exempt from the requirements of S3 61.05(a), 61.07, and 61.09. sion limiting regulation applicable to a (5) The manufacture of paper, miff- (4) Any owner or operator cf a demoli stationary source, provided that such board, and felt. tion operation who intends to demolish a emission limiting regulation is not less () The manufacture of floor tile. building, structure, facility, or installa stringent than the standards prescribed under tills part. ' 2' Requiring the owner or operator (7) The manufacture of paints, coat ings. caulks, adhesives, sealants. 181 The manufacture of plastics and rubber materials. tion to which the provisions of this para graph would be applicable but which has been declared by proper State or lr.cal authority to be structurally un.-o- f a .'.ary .-nurej. o\:er :i a sfa- Oi "be v.ifure of chlorine. vi-.ich :. dri'-'er of im::;:r.c-:v ' e ty .-.u.-ce owned or .-r.vrjted by the '(! L'r-tpy.-.tt.'n: Any owner or opera : .'.'on -.he rsou.ireir.^:;". . . L.Y..ues. to obtain permits, licenses, tor of a d-nioiition operation who intends pa:\u.ra.:.v .-filer fi'.r.n the :- or approvals prior to initiating cor.structi.-n. modification, or operation of such source. to demolish any institutional, commer cial. or industrial building (including apartment buildings having more than four dwelling units), structure, facility. quire:.:en:s specified by f.ar.,grr.i.:i <C.i (1) of this section and tne v.vf.irg of friable asbestos debris as i by paragraph (dii3Mi) of this secta-n. u FEDERAL REGISTER, VOL. 38, NO. 66----FRIDAY, APRIL 6, 1973 CAPCO JEN 0005596 ssao RULES AND REGULATIONS <e) Spraying: There shall be no risible emissions to the outside air from the spray-on application of materials con taining move than 1 percent asbestos, on a dry weight basis, used to insulate or fireproof equipment and machinery, ex cept as provided in paragraph if) of this section. Spray-on materials used to insu late or fireproof buildings, structures, pipes, and conduits shall contain less than 1 percent asbestos on a dry weight basis. (1) Sources subject to this paragraph are exempt from the requirements of 3 61.05(a). 5 61.07,and 3 61.09. (2) Any owner or operator who intends to spray asbestos materials to insulate or fireproof buildings, structures, pipes, con duits. equipment, and machinery shall report such intention to the administra tor at 'cast 20 days prior to the com mencement of the spraying operation. Such report shall include the following infor.m.ai:on: li) Name of owner or operator. (ii> Address of owner or operator. (iii) Location of spraying operation. (iv) Procedures to be followed to meet the requirements of this paragraph. <f> Rather than meet the r.o-visibleemission requirements of paragraphs (a), (c), and (e) of this section, an owner or operator may elect to use the methods specified by 61.23 to clean emissions containing particulate asbestos material before such emissions escape to. or are vented to. the outside air. 01.23 Air-i'lcaning. If air-cleaning is elected, as permit ted by 3 01.22* f ', the requirements of this section must be met. (a) Fabric filter collection devices must be used, except as noted in para graphs (b> and <c> of this section. Such devices must be operated at a pressure drop of no more than 4 inches water gage, as measured across the filter fabric. The airflow permeability, as determined by ASTM method D737-69, must not exceed 30 ft' min ft: for woven fabrics or 35 ft''min, ft' for felted fabrics, except that 40 ftVmin. ft: for woven and 45 ft1/ min,ft: for felted fabrics Is allowed for filtering air from asbestos ore dryers. Each square yard of felted fabric must weigh at least 14 ounces and be at least one-sixteenth inch thick throughout. Synthetic fabrics must not contain fill yam other than that which Is spun. (b) If the use of fabric filters creates a fire or explosion hazard, the adminis trator may authorize the use of wet col lectors designed to operate with a unit contacting energy of at least 40 inches water gave pressure. (ci The administrator may authorize ihe u-<r cf !\!t`*r::'.z equipment other than *.:',/. *'..-. . ! in Paragraphs * a> ar.d tb . '-'.:s u.-in :i :he owner or operator :::mr..-.r ite, to the satisfaction of the administrator that the filtering of part- uiate asbestos material is equivalent to til./, cf the described equipment. * cl i A'.', air-cleaning equipment autiiori d by thL section must be properly installed, i:-ed, operated, and maintained, r.yp.i.-j devices may be used only during upset or emergency conditions and then only for so long as it takes to shut down (h) "Incinerator" means any furnace the operation generating the particulate used in the process of burning waste for asbestos material. the primary purpose of reducing the 61.21 Reporting. volume of the waste by removing com bustible matter. The owner or operator of any existing (i) "Propellant" means a fuel and oxi source to which this subpart is applicable dizer physically or chemically combined shall, within 90 days after the effective which undergoes combustion to provide date, provide the following information rocket propulsion. to the administrator: (J> "Beryllium alloy" means any metal (a) A description of the emission con to which beryllium has been added in trol equipment used for each process; order to increase Its beryllium content (b) If a fabric filter device is used to and which contains more than 0.1 per control' emissions, the pressure drop cent beryllium by weight. across the fabric filter in inches water (k> "Propellant plant" means any gage. facility engaged in the mixing, casting, (1) If the fabric filter device utilizes a or machining of propellant. woven fabric, the airflow'permeability in ft7min/ft:: and. if the fabric is syn 61.32 Emission standard. thetic, indicate whether the fill yam is spun or not spun. (2) If the fabric filter device utilizes a feited fabric, the density in oz/yd', the minimum thickness in inches, ar.d the airflow permeability in ft7min/ft-'. (a) Emissions to the atmosphere from stationary sources subject to the provi sions of this subpart shall not exceed 10 grams of beryllium over a 24-hour period, except as provided in paragraph (b) of' this section. (c) Such information shall accompany (b) Rather than meet the require the information required by 3 61.10. The ment of paragraph (a) of this section, appropriate form is contained in appen an owner or operator may request ap dix A to tills part. proval from the Administrator to meet Subpart C--National Emission Standard for Beryllium 61.30 Applicability. an ambient conentration limit on beryl lium in the *vicinity of the. stationary source of 0.01 /ig/m', averaged over a 30-day period. The provisions of this subpart are ap (1) Approval of such requests may be plicable to the following stationary granted by the Administrator provided sources: that; . (a) Extraction plans, ceramic plants, (1) At least 3 years of data Is avail foundries, incinerators, and propellant able which to the judgment of the Ad plants which process beryllium ore. beryl ministrator demonstrates that the fu lium, beryllium oxide, beryllium alloys, ture ambient concentrations of beryllium or beryllium-containing waste. in the vicinity of the stationary source (b) Machine shops which process will not exceed 0.01 ;ig/m\ averaged over beryllium, beryllium oxides, or any alloy a 30-day period. Such 3-year period shall when such alloy contains more than 5 be the 3 years ending 30 days before the percent beryllium by weight. effective date of this standard. 61.31 Definitions. (it) The owner or operator requests such approval to writing within 30 days Terms used in this subpart are de after the effective date of this standard. fined in the act. in subpart A of this (iii) The owner or operator submits a part, or in this section as follows: report to the Administrator within 45 (a) "Beryllium" means the element days after the effective date of this beryllium. Where weights or concentra standard which report includes the fol tions are specified, such weights or con lowing Information: centrations apply to beryllium only, (a) Description of sampling method excluding the weight or concentration of -including the method and frequency of any associated elements. calibration. (b> "Extraction plant" means a fa cility chemically processing beryllium ore to beryllium metal, alloy, or oxide, or performing any of the intermediate steps in these processes. (c) "Beryllium ore" means any natu rally occurring material mined or gathered for its beryllium content. (d) "Machine shop" means a facility performing cutting, grinding, turning, honing, milling, deburring. lapping, (b) Method of sample analysis. (c) Averaging technique for determin ing 30-day average concentrations. (d) Number, identity, and location (address, coordinates, or distance and heading from plant) of sampling sites. (e) Ground elevations and height above ground of sampling inlets. </) Plant and sampling area plots showing emission points and sampling electrochemical machining, etching, or oilier similar operations. iei "Ceramic plant'- means a manu facturing pinut producing ceramic items. sites. Topographic features significantly affecting dispersion including -plant building heights and !ocai:cns shall be included. ill `Foundry' means a facility en gaged in the melting or casting of beryllium metal or alloy. (g) `'Beryllium-containing waste" means material contaminated with beryllium and/or beryllium compounds used or generated during any process or operation performed by a source subject (tii Information necessary for esti mating dispersion including stack height, inside diameter, exit gas temperature, exit velocity or flow rale, and beryllium concentration. Ill) A description of data, and proce dures (methods or models) used to de sign the air sampling network (i.e.. num to tills subpart. ber and location of sampling sites1. FEDERAL REGISTER, VOt. 3i. NO. 66--FRIOAV. APRIL 6, 1573 CAPCO JEN 0005597 RULES AND REGULATIONS 8831 ti' Air wimplir.gdata indicating beryl lium concentrations in the vicinity of the stationary ,-curce for the 3-year period specified in paragraph (bil> of this section. This data shall be presented chronologically and include the beryl lium concentration and location of each individual sample taken by the network and the corresponding 30-day average beryllium concentrations. cji within 60 days after receiving such report, the Administrator will notify the owner or operator in writing whether approval is granted or denied. Prior to denying approval to comply with the pro visions of paragraph (b) of this section, the Administrator will, consult with representatives of the stationary source for which the demonstration report was submitted. (c> The burning of beryllium and/or beryllium-containing waste, except pro pellants, is prohibited except in incinera tors. emissions from which must comply with the standard. 61.33 .5t.iek sampling. (a) Unless a waiver of emission testing is obtained under 5 61.13, each owner or operator required to comply with 61.32(a) shall test emissions from his source, <1) Within 90 days of the effective date in the case of an existing source or a new source which has an initial startup date preceding the effective date; or (2) Within DO days of startup in the case of a new source which did r.ot have an in.;: >: `.artup date preceding the ef fective- d.te. <b) The Administrator shall be noti fied a* 1-a-t 30 days prior to an emission test so t's.al he may at his option observe the test. (c) Samples shall be taken over such a period or periods as are necessary to ac curately determine the maximum emis sions which will occur in any 21-hour period. Where emissions depend upon the relative frequency of operation of differ ent types of processes, operating hours, operating capacities, or other factors, the calculation of maximum 24-hourperiod emissions will be based on that combination of factors which is likely to occur during the subject period and which result in the maximum emissions. No changes in the operation shall be made, which would potentially increase emissions above that determined by the most recent source test, until a new emis sion level has been estimated by calcula tion and the results reported to the Ad ministrator. (d) All samples shall be analyzed and beryllium emissions shall be determined within 30 days after the source test. All determinations shall be reported to the Administrator by a registered letter dis: .it.. : ''ce 'he r'.o.-e of the next busi- -- such determination. 1 : of emission v.-c rosulrs aivi d needed to determine total cm:"sion> shall be retained at the source and in-he available, for inspection by the Ail::'.::'.:-t:a'.sr. for a minimum of 2 years. (i 1.3 1 \ir -ampling. <.v Stationary sources subject to ? 61 22 o .-hail locate air sampling sites in accordance with a plan approved by the Administrator. Such sites shall be located in .such a manner as is calculated to detect maximum concentrations of beryllium in the ambient air. ib> All monitoring sites shall be op erated continuously except for a reason able time allowance for instrument main tenance and calibration, for changing filters, or for replacemet of equipment needing major repair. <c) Filters shall be analyzed and con centrations calculated within 30 days after filters are collected. Records of concentrations at all sampling sites and other data needed to determine such con centrations shall be retained at the source and made available, for inspection by the Administrator, for a minimum of 2 years. <d> Concentrations measured at all sampling sites shall be reported to the Administrator every 30 days by a regis tered letter. (e) The Administrator may at any time require changes in, or expansion of, the sampling network. Subpart D--National Emission Standard for Beryllium Rocket Motor Firing 61.10 Applicability, The provisions of this subpart are ap plicable to rocket motor test sites. 61.41 Definitions* Terms used in this subpart are defined in the Act, In Subpart A of this part, or in this section as follows: (a> "Rocket motor test site" means any buildir.v. structure, facility, or installa tion where the static test firing of a beryllium recket motor and. or the dis posal of beryllium propellant is conducted. (b) "Ben-Ilium propellant" means any propellant Incorporating beryllium. 61.i2 ICnii-'Mon standard. - (a) Emissions to the atmosphere from rocket-motor test sites shall not cause time-weighted atmospheric concentra tions of beryllium to exceed 75 micro gram minutes per cubic meter of air within the limits of 10 to 60 minutes, accumulated during any 2 consecutive weeks, in any area in which an effect adverse to public health could occur. (b) If combustion products from the firing of beryllium propellant are col lected in a closed tank, emissions from such tank shall not exceed 2 grams per hour and a maximum of 10 grams per day. 61.43 Emission testing--rocket firing or prupoII.uU di-posal. (a) Ambient air concentrations shall be measured during and after firing of a rocket motor or propellant disposal and in such a manner that the effect of these emissions e.-.u be compared with the standard. Sixth sampling techniques shall be approved hy the Administrator. cb) All samples shall be analyzed and results shall be calculated within 30 days after samples are taken and before any subsequent rocket motor firing or pro pellant disposal at the given site. All re sults shall be reported to the Adminis trator by a registered letter dispatched before the close of the next business day following determination of such results. (c) Records of air sampling test results and other data needed to determine in tegrated intermittent concentrations shall be retained at the source and made available, for inspection by the Admin istrator. for a minimum of 2 years. (d) The Administrator shall be noti fied at least 30 days prior to an air sam pling test, so that he may at his option observe the test. 61.44 .Slack sampling, fa) Sources subject to 61.42(b) shall be continuously sampled, during release of combustion products from the tank, in such a manner that compliance with the standards can be determined. The pro visions of $61.14 shall apply. (b) All samples shall be analyzed, and beryllium emissions shall be determined within 30 days after samples are taken and before any subsequent rocket motor firing or propellant disposal at the given site. All determinations shall be reported to the Administrator by a registered let ter dispatched before the close of the next business day following such deter minations. (c) Records of emission test results and other data needed to determine total emissions shall be retained at the source and made available, for inspection by the Administrator, for a minimum of 2 years. <d> The Administrator shall be noti fied at least 30 days prior to an emission test, so that he may at his option observe the test. Subpart E--National Emission Standard for Mercury 61.50 Applicability. The provisions of this subpart are ap plicable to those stationary sources which process mercury ore to recover mercury, and to those which use mercury chloralkali cells to produce chlorine gas and alkali metal hydroxide. 61.51 Definitions. Terms used in this subpart are defined in the act, in subpart A of this part, or in this section as follows: (a) "Mercury" means the element mer cury, excluding any associated elements, and includes mercury in particulates, va pors. aerosols, and compounds. (b) "Mercury ore" means a mineral mined specifically for its mercury con tent. (c) "Mercury ore processing facility" means a facility processing mercury ore to obtain mercury. <d) "Condenser stack gases" mean the gaseous effluent evolved from the stack of processes utilizing heat to extract mer cury metal from mercury ore. (p) "Mercury chlor-olkali cell" means a device which is basic:1 llv < -ed an electrolyzer section v.r.ti a ':-t-:i-r (decomposer) section .ind utni/cs mer cury to produce ciUorir.e gas. hydrogen gas, and alkali metal hydroxide. if) "Mercury chlor-alkali electrolyzer" means an electrolytic device which is part of a mercury chlor-alkali cell and utilizes a flowing mercury cathode to produce chlorine gas and alkali metal amalgam. FEOERAl REGISTER. VOL. 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005598 S$:j2 RULES AND REGULATIONS (r) "Denuder" means a horizontal or vertical container which is part of a mer cury chlor-alkali cell and in which .water and alkali metal amalgam are converted to alkali metal hydroxide, mercury, and hydrogen gas in a short-circuited, elec trolytic reaction. 'hi "Hydrogen gas stream" means a hydrogen stream formed in the chlor- aikali cell denuder. ii) "End box" means a container(s) located on one or both ends of a mercury chlor-alkali electrolyzer which serves as a connection between the electrolyzer and denuder for rich and stripped amalgam. ij) "End box venttlation system" means a ventilation system which col lects mercury emissions from the endboxes. the mercury pump sumps, and their water colection systems. 'k) "Cell room" means a structure's) housing one or more mercury electro lytic chlor-alkali cells. til.52 Emission standard. Emissions to the atmosphere from sta tionary sources subject to the provisions of this su'opart shall not' exceed 2,300 grams of mercury per 24-hour period. 61.53 Stack sampling. (a) Mercury ore processing facility. (1) Unless a waiver of emission testing is obtained under 5 61.13, each owner or operator processing mercury ore shall test emissions from his source, ii) Within 90 days of the effective date in 'he case of an existing source or a new source which has an initial start up date preceding the effective date: or tii) Within 90 days of startup in the case of a new source which did not have an initial startup date preceding the ef fective date. (2) The Administrator shall be noti fied at least 30 days prior to an emission test, so that he may at his option observe the test. (3> .Samples shall be taken over such a period or periods as are necessary to accurately determine the maximum emissions which will occur In a 24-hour period. No changes in the operation shall be made, which would potentially in crease emissions above that determined by the most recent source test, until the new emission level has been estimated by calculation and the results reported to the Administrator. (4) All samples shall be analyzed, and mercury emissions shall be determined within 30 days after the source test. Each determination will be reported to the Ad ministrator by a registered letter disjatched before the close of the next busi ness day following such determination. (51 Records of emission test results and other data needed to determine total emissions shall be retained at the source and made available, for inspection by the 's'.ra;:-r for a minimum c; 2 years. :..vro".ry rhlor-alkah plant--hy'.irvg-.-n and end-oox ventilation gas s'reams. 11 > L'.Uess a waiver of emission test ing i- obtained under 5 61.13. each owner or operator employing mercury chlorjlkali coil's) shall test emissions from his source. (i) Within 90 days of the effective date in the case of an existing source or the Administrator, for a minimum of a new source which has an initial startup 2 years. date preceding the effective date; or (c) Mercury chlor-alkali plants__ rii) Within 90 days of startup in the cell room ventilation system. case of a new source which did not have (1) Stationary sources using mercury an initial startup date preceding the ef chlor-alkali cells may test cell room fective date. emissions in accordance with paragraph (2) The Administrator shall be noti (c)(2) of this section or demonstrate fied at least 30 days .prior to an emission compliance with paragraph (c) (4) of this test, so that he may at his option observe section and assume ventilation emissions the test. of 1,300 gms/day of mercury. (3) Samples shall be taken over such (2) Unless a waiver of emission test a period or periods as are necessary to ing is obtained under 1 61.13, each owner accurately determine the maximum emis or operator shall pass all cell room air sions which will occur in a 24-hour in forced gas streams through stacks period. No changes in the operation shall suitable for testing, be made, which would potentially in (i) Within 90 days of the effective date crease emissions above that determined in the case of an existing source or a new by the most recent source test, until the source which has an initial startup date new emission has been estimated by cal preceding the effective date: or culation and the results reported to the (ii) Within 90 days of startup in the Administrator. case of a new source which did not have, (4) All samples shall be analyzed and an initial startup date preceding the mercury emisions shall be determined effective date. within 30 days after the source test. All (3) The Administrator shall be noti the determinations will be reported to fied at least 30 days prior to an emission the Administrator by a -registered letter test, so that he may at his option observe dispatched before the close of the next the test. business day following such determina (4) An owner or operator may carry tion. out approved design, maintenance, and (5) Records of emission test results housekeeping practices. A list of ap and other data needed to determine total proved design, maintenance, and house emissions shall be retained at the source keeping practices may be obtained from and made available, for inspection by the Administrator. APPENDIX* National Emission Standards for Hazardous Air Pollutants Compliance Status Information 1. S03n.CE REPORT Instructions: Owners or operators of sources of hazardous pollutants subject to the National Emission Standards for Hazardous Air Pollutants are required to submit the information contained In' Section I to the appropriate Environmental Protection Agency Regional Office before (date which is 90 days after the standards are promulgated). is provided in i 61.04. A listing of regional offices A. SOURCE INFORMATION. 1. Identification/Location - Indicate the name and addrtss of each source. A29 -j__ 1__ 1__i_ cohpah't name A4S A49 MlrtBER _l_1__' '_I__ILlftfel'TOfaEis'' L" A6B B19 B33 L_<_ -1--I__1__1_I__I_I CITY L CUz'lTX .634 zip 2, Contact - Indicate the rare and telephone number of the owner or operator or other responsible official iihom ?.1 ray contact con cerning this report. B39 B53. _i__i_1__1__1_i_ j_1__I_1--1--1--LNAME B54 --I--1--l--l lELEPhOJlE B63. B38 J___i FEDERAL REGISTER, VOL. 3D, NO. 66-----FRIDAY, APRIL 6, 1973 CAPCO JEN 0005599 RULES AND REGULATIONS S833 r, APRIL 6, 1973 I CAPCO JEN 0005600 FEDERAL REOISTEI, VOL. 38, NO. 44-- MIDAV, APRIL 4, 1973 8834 >vuj ra O um ou vt f>4J to It V m C*77--<33U J43J 73 3 >a O CPl.--VVSI Cl 4w2 n m r- <*: <u CT 7u-- r- 3 o> *---- cci- -uj nc S- -a- o 0Ou.4N=nl & RULES AND REGULATIONS uc cs oa CUl 7hV3 >ui **<u>u p* o- ra*i ViaI Aa OU COl 4Q*. 4co<1j OuauUo O* x-:* era aat vo 4o4 "O 44 O a-- V* Ua. *O4 Oa. Vau** 41 C (A 41 XI O 3 C= 4 +4 -Oa <4- i t- --au N<uo rt*rt U II VI 73 Cl CCl O VO.V) UuV >v- 4(O-1I uCJ ** o*-.. *> l *OcTW---*ri:| oc CCl Mefci-Of4*- .44S 41 4C4 aCsl Wl~ Ci Is ju;* V3m4S 4vet-=. ` Or: wTJv et- o " VS fc. O aII cIIxvu* EI--^ OA*!O 4CZi I01e3/S14 4> O > 3- 3a3s Cl c > *- *& >* 44Cl 41 3tr *3 aS tcr ci a. t- o ID vr C3 VI O. 0*0 s.O 73 c r- -- f *3 O T- * <a ci VI <e4 3 VI Cl 41 73 vs as C v! U 44 3 0 3 0" : vi >Cl 443 73 OcC oC C4l <r>4l O* >. eCl Cl Orj 7Cc3l O nv E" --'v--r---- -I4io73 I -- fca * w> -- Cl V ii v 4 u iiai- a Ug U a-- c --*- <i.e~*xz 44MV Of U r* 113 L MM4 a x44: o i>i OE4>0a1.Xc3i O -- U. as 4uUmoH-rinii - --<s *,,** sr LuLUv.. 4>1 Cai O C v> 1 vS<e. vai* c<r 3o: ci i c -- 2 <j 44 x: VS 3 I s o > as 0-4- r 44 *44 73i E --Cf- ck, ..Cw IA O U 73 VS 44 Cl Q3J 4C4l ao" 4O4 Ccll 7W3 OCO44 44 cro l <0 u w ^ e J t. 44 U X 73 t. II L.O, CO V<A V C<c kO f>it --C-- >>a- u 7- 44 44 -- g a) as c o ia C o E E4f U- C O >7- C 44 4J -- <,y3SaV^J .--C<7sV3- Sirs ^ 73 c 44 0 E OC bVr*>C<MTo--. +q34 aaO---4X*4 vOU-- O *: e ; e i e5 3 ia u i4/4)J 4: 1. or 4I<- flk__ 7- L. J-- &-- 44 O *3 > CL. O v* d -CO. C. c TS 44 Cl C r- : 4j to -r- o <- U 4 Ca J4 f3c.af-- -Oa- C< Sir- U01 tO. I3A Cl *a!-~34 ra- cr- or 0 O C lA -oaw-l 4U r4 3Locr.{ll- '*- tar ` .= o as 3 ai 7 n._ m v au or< 3s- *o" ou co .c 4:'.5" *4 44 *> 4) U Ca UO. euui* 4OaUCn4j.l. 4CaoUa4jl.z--4OaCz4s 7ztrfS soo<,, 1 74(u-A4 snv>s.>4vco>=i x4an4il4O*-4 nvCUsIITTu33*t<C vc|MvsOo.CCICi<eUci CAPCO JEN 0005601 RULES AND REGULATIONS SS33 (( (( ((. o 1- ft <JM-4>J t4*J UII <a a e -S.i->* c. Qou.uO. aVj----- 2 e.r. ws ++ aa<>o o<j :. ul_ cOJ j .e W. > O--V* -rG3r-- >|W .e <3e ;Vee43i u>J ofc- VI I<I e E_ * o < oU*(.lt,,nuOjcV=iVVIcOSOU Jcu *O- Iro-j *I*- o i e V-1 * *. - 4- JT. -- 4*-Jt JZ-l U OV<1ua.I 4J J i j4i iji I V-1 *<t1 T *o- : *j U > c o o *C OOCl TC^3 I 1>a--t * h- -- Ifo.CS--Pt II---3 Ko*. ec a. < <Q o z o > at " r? s * = o oS5G5 2S = 1*5 g. s *Jss M1 , slgsISsif] x a%p rr c. * <ec. 22.n,, u 2 z 2 g o. - " 5 ^= ~ O J =, n. o ^ g = o"iT>: kJE*jt*i?l*id? s* uc -.. CAPCO JEN 0005602 S$:J6 RULES AND REGULATIONS : t'.tiea of particulate matter. The Alter holder mu*t provide a positive seal agaLnst leakage from ouUtUe or around the filter. A heating v.-i.ein capable of maintaining the filter at a minimum temperature of 250* P. should be u&ed to prevent condensation from occur- 3.3 Analysis--3.3.1 Sodium hydroxide, 10 N.--Dissolve 400 g of sodium hydroxide pellets Ln distilled water and dilute to 1 to 1, 3.3.2 Reducing tyjsnt, 12 percent hydrox- ylamine sulfate, 12 percent sodium chlo ride.--To 60 ml of distilled water, add 12 g of hydroxylsmine sulfate and 12 g of sodium 4.2.2 The sampling site should be at least eight stack or duct diameters downstream and two diameters upstream from any flow disturbance such as a bend, expansion, or contraction. For a rectangular cross section, determine an equivalent diameter from the following equation: 2.1 8 Earometer. To measure atmospheric pressure to 0.1 la Hg. 2.2 'f^s.'iurement of stack conditions (stack pressure, temperature, moisture and irlocti;/)--2 2.1 Pitot tube. Type S, or equivalent. with a coefficient within 5 percent over the working range. 2.2.2 Differential pressure gauge. Inclined manometer, or equivalent, to measure veloc ity held to within 10 percent of the minimum value. Micromanometera should be used Lf warranted. 2 2.3 Temperature gauge. Any tempera ture measuring device to measure stack tem perature to within 1* F. 2 2 4 Pressure gauge. Pitot tube and in clined manometer, or equivalent, to measure stack procure to within 0.1 In Hg. 2.2.5 .Voisiure determmat;07i. Wet and dry bulb thermometers, drying tubes, con densers. or equivalent, to determine stack gas mouture content to within 1 percent. 2.3 recoMery--2.3.1 Leakiest glass samo'e b >r\les. 500 ml and 100 ml with Teflon lined top-5. chloride. Dilute to 100 ml. This quantity Is sufficient for 20 analyses and must be` pre pared daily. 3.3.3 Aeration gas.--Zero grade air. 3.3.4 Hydrochloric acid. 0.3N.--Dilute 25.S ml of concentrated hydrochloric acid to 1 to 1 with distilled water. 3.4 Standard mercury solutions--3.4.X Stock solution.--Add 0.13S4 2 of mercuric chloride to 80 ml of 0.3N hydrochloric acid. After the mercuric chloride has dissolved, add 0.3N hydrochloric acid and adjust the volume to 100 ml. One ml of this solution Is equivalent to 1 mg of free mercury. 3.4.2 Standard solutions.--Prepare cali bration solutions by serially diluting the stock solution (3.4.1) with 0.3N hydrochlo ric acid. Prepare solutions at concentrations in the linear working range for the Instru ment to be used. Soutions of 0.2 Mg/ml, 0.4 HZ ml and 0.6 pg/ml have been found ac ceptable for most instruments. Store all solutions in glass-stoppered, glass bottles. These solutions should be stable for at least 2 months: however, periodic checks should 2LW D"~L + W where: D. = Equlvaleat diameter. L=Length. Width. *1' 101-1 4.2.3 When the above sampling, site cri teria can-be met, the minimum number of traverse points is four (4) for stacks 1 foot in diameter or less, eight (8) for stacks larger than I foot but 2 feet ln diameter or less, and twelve (12) for stacks larger than 2 feet. 4.2.4 Some sampling situations may rea der the. above sampling site criteria Imprac tical. When this Ls the case, choose a con venient sampling location and use figure 101-3 to determine the minimum number of traverse points. However, use figure 101-3 oni-' 'or stacks 1 foot in diameter or larger. 4. To use figure 101-3, first measure the ance from the chosen sampling loca tion v. he nearest upstream and downstream disturbances. Divide this distance by the diameter or equivalent diameter to deter 2 32 Graduated cylinder. 250 ml. be performed to Insure quality,* mine the distance In terms of pipe diameters. 2 3.3 Pitutic far Approximately 300 ml. 2. t A n.alysis--2.4.1 Spectrophotometer, To measure absorbance at 253.7 nm. Perkin Elmer Model 303. with a cylindrical gas cell (approximately 1.5 in. O.D. x 7 in.) with quartz glass windows, and hollow cathode source, or equivalent. / 2 4.2 Gas sampling bubbler. Tudor Scien tific Glass Co.. Smog Bubbler, Catalogue No. TP-1130. or equivalent. 2 4 3. Recorder. To match output of spec trophotometer. 3. It-'cgenn--3,1 Stock reagents--3.1.1 Potc.-'.-tm iodide. Reagent grade. 3 12 D:<t''!cd irater--3.1.3 Potassium. iodide solution. 25 percent. Dissolve 250 g 4. Procedure.--4.1 Guidelines for source testing are detailed In the following sections. These guidelines are generally applicable; however, most sample si?e3 differ to some degree and temporary altera'Ions such as stack extensions or expansions often are re quired to ensure the best possible sample site. Further, since mercury Is hazardous, care should be taken to minimize exposure. Finally, since the total quantity of mercury to be collected generally Is small, the test must be carefully conducted to prevent con tamination or losa of sample. 4.2 Selection of a sampling site and mini mum number of traverse points: of potassium iodide (reagent 3.1.1) in dis 4.2.1 Select a suitable sampling site that tilled water and dilute to 1 to 1. Is as close as Is practicable to the point of 3.1.1 /fydroc/iiorfc acid. Concentrated. atmospheric emission. If possible, stacks 3.1 5. Potassium iodate. Reagent grade. smaller than l foot In diameter should not 3.1.6 Iodine monochloride (ICl) 1.05/. To be sampled. Determine -the corresponding number of traverse points for each distance from fig ure 101-3. Select the higher of the two num bers of traverse points, or a greater value, such that for circular stacks the number Is a multiple of four, and for rectangular stacks the number follows the criteria of section 4.3.2. 4.2.6 If a selected sampling point Is closer than l Inch from the stack wall, adjust the location of that point to ensure that the sample is taken at least i Inch away from the wall. 4.3 Cross sectional layout and location ci traverse points: 4.3.1 For circular stacks locate the trav erse points on at least two diameters accord ing to figure 101^4 and table 101-1. The traverse axes shall divide the stack cross section into equal parts. 800 ml. of 25 potassium iodide solution (reagent 3 13), add 800 ml. of concentrated hydrochloric acid. Cool to room temperature. NUMBER OF 0UCT DIAMETERS UPSTREAM (DISTANCE A) With vigorous stirring, slowly add 135 g. of potassium iodate and continue stirring until 0S 1.0 1.5 2.0 2.S all free Iodine has dissolved to give a clear crange-red solution. Cool to room tempera ture and dilute to 1800 ml. with distilled water. The solution should be kept In amber bottles to prevent degradation. 3.1.7 Sodium hydroxide pellets. Reagent grade. 3.1.8 Nitric acid. Concentrated. 3.1.9 Hydrozylamine sulfate. Reagent grade. 3.1.10 Sodium chloride. Reagent grade. 3.1.11 Mercuric chloride. Reagent grade. 3.2 Sampling--*3.2.1 Absorbing solution, 0.J5/ ICl. Dilute 100 ml. of the 1M ICl stock solution (reagent 3.2.6) to 1 to 1 with distilled water. The solution should be kept In glass bottles to prevent degradation. This reagent should be stable for at least 2 Months: however, periodic checks should be u ri< . *'t to m.S'Te J: I V. V n.:r:a ;-::d-- 3 * , D : ll.d.dcioniied water. 3 2 4 f ! r.z Indicating type, 0 to 16 rv.r-J at 330* F. for 2 hours. 32 5 F* vr (optional). GUoS fiber. Mine S.ifi**;/ Appliances 11063H, or equivalent. A niter ni.iy be necessary In cases where the gas stream to be sampled contains largo quantities of particulate matter. Figure 102-3. Minimum of traverse points. FSDSRAL REGISTER, VOL. 38, NO. 65--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005603 RULES AND REGULATIONS Table 101-1. Location of traverse points- in circular stacks (Percent of stack diameter from inside wall to traverse point) Traverse point number on a diameter 1 2 3 4 5 6 7 8 9 TO 11 12 13 14 15 . 16 17 18 19 2Q 21 22 23 24 2 14.6' 85.4 4 6.7 25.0 75.0 93.3 * Number of traverse points on a diameter 6 8 10 12 14 16 . 18 20 4.4 14.7 29.5 70.5 85.3 95.6 3.3 10.5 19.4 32.3 67.7 80.6 89.5 96.7 2.5 8.2 14.6 22.6 34.2 65.8 77.4 85.4 91.8 97.5. 2.1 6.7 11.8 17.7 25.0 35.5 64.5 75.0 82.3 8a. 2 93.3 97.9 1.8 5.7 9.9 14.6 20.1 26.9 36.6 63.4 73.1 79.9 35.4 90.1 94.3 98.2 1.6 4..9 8.5 12.5 16.9 22.0 28.3 37.5 62.5 71.7 78.0 83.1 87.5 91.5 95.1 98.4 1.4 4.4 7.5 10.9 14.6 18.8 23.6 29.6 38.2 61.8 70.4 76.4 81.2 85.4 89.1 92.5 95.6 98.6 1.3 3.9 6.7 9.7 12.9 16.520.4 25.0 30.S 38.8 61.2 69.4 75.0 79.6 S3.5 87.1 90.3 93.3 96.1 98.7 22 1.1 3.5 6.0 8.7 11.6 14.6 18.0 21.8 26.1 31.5 39.3 60.7 63.5 73.9 73.2 82.0 85.4 83.4 91.3 94.0 96.5 93.9 24 *1.1 3.2 5.5 7.9 10.5 13.2. 16.1 19.4 23.0 27.2 32.3 39.8 60.2 67.7* 72.3 77.0 80.6 83.9 86.8 89.5 92.1 94.5 96.8 98.3 101-4. Co wale* of circular UK* *Ko*if^loc*4.if t/AWM po.nts on p*>pendiculv dUmettfs. flguro 10M. Cfo mcIIm of rtcUnquIar sli:V divided Into tioquat vw, iUt tre(J* point* at cantroid of toc&wo*. 4.3.2 For rectangular stacks divide the eross section Into as many equal rectangular areas as traverse points, such that the ratio of the length to the width of the elemental areas Ls between one and two. Locate the traverse points at the centroid of each equal area according to figure 101-$. 4.4 Measurement of stack conditions: 4.4.1 Sot up the apparatus as shown in figure 101-2. Make sure all connections are tight and leak-free. Measure the velocity head and temperature at the traverse points specified by section 4.2 and 4.3. 4.4.2 Measure the static pressure In the stack. 4.4.3 Determine the stack gas moisture. 4.4.4 Determine the stack gas molecular weight from the measured moisture content and' knowledge of the expected gas stream composition. A standard- Orsat analyzer has been found valuable at combustion sources. In all cases, sound engineering judgment, should be used. mz FEDERAL REGISTER, VOL 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005604 383S RULES AND REGULATIONS 4.5 preparation of sampling train: 4.5.1 Prior to assembly, clean all glassware (probe, impingers, and connectors) by rinsing with wash acid, tap water, O.IM ICl, tap water, and finally distilled water. Place 100 ml of O.IM ICl la each of the first three lmpimrers. and place approximately 200 g of prewel^iied silica gel In the fourth Implnger. Save 60 ml of -he O.IM ICl as a blank in the sampte analysis. Set up the train and the probe as in figure 101-1. 4.5.2 If the gas stream to be sampled Is excessively dirty or moist, the first Implnger may clog or become dilute too rapidly for sufficient testing. A filter can be placed ahead of the impingers to collect the particulates. An initial empty Implnger may also be used to remove excess moisture. If a fifth Implnger Is required, the final Implnger may have to bo carefully taped to the outside of the sample box. 4.5.3 Leak check the sampling train at the sampling site. The leakage rate should not be la excess of 1 percent of the desired sam pling rate. If condensation In the probe or filter is a problem, probe and filter heaters will be required. Adjust the heaters to pro vide a temperature of at least 250* P. Place crushed icc artund the impingers. Add more ice during the test to keep the temperature of the gases leaving the last Implnger at 70' P or less. 4.6Mercury train operation: 4.6.1 For each run, record the data re quired on the example sheet shown In figure 101-6. Take readings at each sampling point at least every minutes and when signifi cant changes in stack conditions necessitate additional adjustments in flow rate. 4.6.2 Sample at a rate or 0.5 to 1.0 cfm. Samples shall be taken over such a period or periods as are necessary** to accurately determine the maximum emissions which would occur in a 24-hour period. In the case of cyclic operations, sufficient tests shall be made so as to allow accurate determination or calculation of the emissions which will occur over the duration of the cycle. A mini mum sample time of 2 hours Is recommended. In some Instances, high mercury concentra tions can prevent sampling In one run for the desired minimum time. This Is indicated by reddening In the first Implnger as free Iodine Is liberated. In this case, a run may be divided into two or more subruns to en sure that the absorbing solutions are not depleted. axit Si^n.1 iOl NO_ Hit) IOI N0_ MlltiH. Si". *t Vi TA.J'.r *? M J * - -- ! | mtwK V - ** JT*C 'S* IAA0KIHC mauac__ ttxnouomm. Miru tot iwwa MUNC1Kk,_ no* HtAtu ttiii'tfJ-- VH0C.TT MAO ,*'4\ mss* oinmiuii AC*75S OWCt 1 * Ml. M.O CAS SAM** VOlWT cas sAxnt rcveounm AfO*T CAS Mil owair . tw(Atwn f fVA.IVGC* *r ; 1 1 -------------------1------------- 4VtACC 1 1 A1. A*tA<. pigur* 101-9. Field data 4.6.3 To begin sampling, position the nozzle at the first traverse point with the tip pointing directly Into the gas stream. Im mediately start the pump and adjust the flow to isokinetic conditions. Sample for at least 5 minutes at each traverse point: samp ling time must be the same for each point. Maintain isokinetic sampling throughout the sampling period. Nomographs which aid in the rapid adjustment of the sampling rate without other computations are In AFTD0576 and arc available from commercial sup pliers. Note the standard nomographs are applicable only for type S pitot tubes and air or a -- aok wirh an equivalent density. Comac: ZIW tne sampling train supplier for n;c\ons when the standard nomo graph is applicable. 4.6.4 Turn o:T the pump at the conclusion of each run and record the final readings. Iinmed.ate'.y remove the probe and iw/zle from the c\ick and handle m accordance with the sampte recovery process described la section 4.7. 4.7 Sample recovery: 4 V.i (All glass storage bottles and the grad- uated cylinder must be precleaned as in sec tion 4.5.1). This operation should be per formed in an area free of possible mercury contamination. Industrial laboratories and ambient air around mercury-using facilities are not normally free of mercury contamina tion. When the sampling train Is moved, care must be exercised to prevent breakage and contamination. 4.7.2 Disconnect the probe from the im plnger train. Place the contents (measured to 1 ml) of the first three impingers Into a 500 ml sample bottle. Rinse the probe and all glassware between it and the back half of the third implnger with two 50 ml portions of O.IM ICl tnlut.on. Add these rinses to the first sample bottle Tor a blank, place 30 ::U of the O.IM ICl in a 100 ml sample bottle. If used, place the filter along with 100 ml of O.IM ICl in another 100 ml sample bottle. Retain a filter blank. Place the silica gel In the plastic jar. Seal and secure ull containers for shipment. If an additional test is desired, the glassware can be carefully double rinsed with distilled water and reassembled. How ever, if the glassware Is to be out of use more than 2 days, the Initial acid wash procedure must be followed. 4.8 Analysis: 4.8.1 Apparatus preparation.--Clean `alt glassware according to the procedure of sec tion 4.5,1. Adjust the instrument settings ac cording to the instrument manual, using an absorption wavelength of 253.7 nm. 4.8.2 Analysis preparation.--Adjust the air delivery pressure and the needle valve to ohtaln a constant airflow of about 1.3 to/ 1/mln. The analysis tube should be bypassed except during aeration. Purge the equipment for 2 minutes. Prepare a sample of mercury standard sotutlon (3.4.2) according to section 4.8.3. Place the analysis tube la the Une, and aerate until a mlxlmum peak height Is reached on the recorder. Remove the analysis tube, flush the lines, and rinse the analysis tube with distilled water. Repeat with an other sample of the same standard solution. This purge and analysis cycle Is to be re peated until peak heights are reproductble. 4.8.3 Sample preparation.--Just prior to analysis, transfer a sample aliquot of up to 50 ml to the cleaned 100 ml analysis tube. Adjust the volume to 50 ml with O.IM ICl If required. Add 5 ml of 10 N sodium hy droxide, cap tube- with a clean glass stopper and shake vigorously. Prolonged, vigorous shaking at this point is* necessary to obtain an accurate analysis. Add S ml of the re ducing agent (reagent 3.3.2), cap tube with a clean glass stopper and shake vigorously and Immediately In sampte line. 4.8.4 Mercury determination.--After the system has been ^stabilized, prepare samples from the sample bottle according to section 4.8.3. Aerate the sample until a maximum peak height is reached on the recorder. The mercury content Is determined by compar ing the peak heights of the samples to the peak heights of the calibration solutions. If collected samples are out of the linear range, the samples should be diluted. Prepare a blank from the 100 ml bottle according to section 4.8.3 and analyze to determine the reagent blank mercury level. 5. Calibration.--5.1 Sampling train.-- 5.1.1 Use standard methods and equipment as detailed In APTO-0576 to calibrate the rate meter, pitot tube, dry gas meter, and probe heater (if used). Recalibrate prior to each test series. 5 2 Analysis.--5.2.1 Prepare a calibra tion curve for the spectrophotometer using the standard mercury solutions. Plot the peak heights read on the recorder versus the concentrations of mercury in the standard solutions. Standards should be interspersed with the samples since the calibration can change slightly with time. A new calibration, curve should be prepared for each new see of samples run. 6. Calculations.--6.1 Average dry gasmeter temperature, stack temperature, stack pressure and average orifice pressure drop. See datasheet (fig. 101-6). 6.2 Dry gas volume.--Correct the sample volume measured by the dry pas meter to stack conditions by using equation 101-2. -ru s hr<.M.-li lit- *ry `lio .Sl-lif'iH;-.., ,;V l", 'i VuMpi- ciila ilry v-0 11: :-l-T T. - Av'-r.'r- t*ni|*-rtiui.* nf -t.v.* ; v-., P. 7\ Av*:.u?* dry tr.elrr icnjwiuire, r Px,\t -- Barometric pressure at the orifice meter, inHg. a//= Average pressure drop across the or. flee meter, inHjO. 13.6=Specific gravity of mercury. p, = Stack pressure, Pb4rstatic pressure, InHg. FEDERAL REGISTER, VOL 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005605 RULES AND REGULATIONS 6 J Volume of water va-por, V*=K.Y.' cq. 101-3 wbtre: Vw Vn'iim- r.f w*br vnporin thegiu <a:np'.< (Hack * tomiKionsJ, (t*. . G.W-67 . when thed* UJ,,US artf uawL Vi -Tola! vl.Uimfl oI bd'iM coltrMed Id iiupmgtra ami jiucii *i;i mo hgur* Wi-*i, ml. T, m A .uack bis t*jupmtui-. * R. pr-i.>iue, /** rfc static pr*a.*urf iif. IIg, m *5 4 Total gas rolunio. Wv?Ii-Total volume of gas sample (stack couditiona), W -m Volume of gas through fas meter (stack conrii* lions), ft*. Vtr -Volume of tenter rapor in gas sample mack * Londitloos), ft*. vcleve or lhjuio tutu ccuzam it#mccx VOLUME. * ml CMC* Mt MIGHT, 9 r.*.u. 1 r.mn. 1 ItCU 3 lOuCCHO 1 ___________ total '.oils* ccitrcito 1 *'l " cofw? ct * rovou.f if l*C*:*s* 4T Ct*.Silt RAtCR. it tsts) >rsis*it a vulLWC :,, ^ PLANT _________________________________ . DATE RUN NO. __________________________ STACK DIAMETER, in.__ BAROMETRIC PRESSURE, in. Hg: STATIC PRESSURE IN STACK (Pg), in. Hg. OPERATORS_______________________________ Traverse point number Velocity head, in. H20 vS7 S839- [--------------------- 1----------- SCHEMATIC OF STACK CROSS SECTION Stack Teirperatuto (%),4F F g/e 101*7. A/ui/Tfcsl data, 6.5 Stack gas velocity. Use equation 101-5 to calculate the stack gas velocity. 'IV... = K>C,(% J?K.e \ LJ;,i cq. 101-5 - Av. r.ure.ta*.k zas velocity. f<.t per rw>nd. AaS'-iJ^ft. (/n,.mIohR.--i-nR.l-f;cu.U,u)V*' ''", CU these units are used. C,-* Pitot tube poefliciciic, dimenionless. (T` Ati'MI-i* stark pas temperature, K. a*.* s.*ce qu irn root of ihe v^ocitv head <if -t u-k ci? tin. IIiOi**** lip. l"l->'. /*. t M.vk pr**ntri\ /*b<t;U:'pr.*sMir>\m. Kp. AT, s M<" ul r weight of stack w\is Mi- elimination of th' product* of the til.tr uiib-lit tf t*,u h iofiijt.i.n.1 :..'.'t,ph1 hy jr* viihp*;. trie propoitnm i.i the mixture, lb.'lb. mole. Figure 101-8 shows a sample recording sheet for velocity, traverse data. Use the averages In the last two columns of figure 101-8 to determine the average stack gas velocity from, equation 101-5. 6.6 Mercury collected. Calculate the total, weight of mercury collected by using equa tion 101-6. W = ViCi~VCi (+7fC-.q. 101-6 where: Wr ~ total weight of mercury collected, *g. AVERAGE: ' : Figure 1Q1-8. Velocity traverse dala. FEDERAL REGISTER, VOL. 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005606 8840 RULES AND REGULATIONS Vi as Total volume of condensed moisture and 1CI in sample bottle, ml. Ci = Concentration of mercury measured la sample bottle, ug/ozl. V=ToUt volume of IC1 used In sampling (implager contents and all wash amounts), ml. C> = Blank concentration of mercury In ICl rolutton, pg/ml. V/ =Total volume of ICt used In Alter bottle (if used), ml. CrsConcerUratton of mercury In Alter bottle (If used), tgfml. 6.7 Total mercury emission. Catcul&te the total amount of mercury emitted from each stack per day by equation 101-7. This equa tion is applicable for continuous operations. For cyclic operations, use only the time per day each stack Is In operation. The total mercury emissions from a source will be the summation of results from all stacks. ,, TVf(tQ,c../4, 86,400 seconds/day 10* pg/g. eq. 101-7 where: It* U it*? of emission. cAlay. of mercury collated, !'** Total volume of sas su:np! (stock conditions/, * ft'. (r)*..aA\vnc< stack gas velocity, feet per second. .-l.ast.wk ar-a, It*. pling Measurements. Paper presented at the Annual Meeting of the Air Pollution Control Association, St. Louis. Mo.. June 14-19, 1970. 11. Smith. W. S., et al.. Stack Gas Sampling Improved and Simplified with New Equip ment, APCA paper No. 67-119.1967. 12. Smith, W. S.. R. T. Shlgehara. and W. F. Todd. A Method of Interpreting Stack Sampling Data, Paper presented at the 63d Annual Meeting of the Air Pollution Control Association, St. Louis, Mo., June 14-19, 1970. 13. Specifications for Incinerator Testing at Federal Facilities PHS. NCAPC. 1967. 14. Standard Method for Sampling Stacks for Particulate Matter. In: 1971 Book of ASTM Standards, part 23, Philadelphia, 1971, ASTM Designation D-2926-71. 15. Vennard. J. K., Elementary Fluid Me chanics, John Wiley and Sons. Inc., New York, 1947. METHOD 102. REFERENCE METHOD FOR DETER MINATION OP PARTICUEATE AND CASEOUS MER CURY EMISSIONS FROM STATIONARY SOURCES (HTOROCEN STREAMS) 1. Principle and applicability--1.1 Princt- pie.--Particulate and gaseous mercury emis sions are lsoklnetlcally sampled from the source and collected in acidic Iodine monochloride solution. The mercury collected (in the mercuric form) is reduced to elemental mercury in basic solution by hydroxylamme sulfate. Mercury is aerated from the solution and analyzed using spectrophotometry. 1.2 Apph'cab(l(ty.--This method ts appli cable for the determination of particulate and gaseous mercury emissions when the carrier gas stream Is principally hydrogen. The method is for use In ducts or stacks at stationary sources. Unless otherwise specified, this method is nob intended to apply to gas streams other than those emitted directly to the atmosphere without further processing. 2. Apparatus--2.1 Sampling train.--A sche matic of the sampling train used by EPA is shown In figure 102-1. Commercial models of this train are available, although complete construction details are described in APTD0581,1 and operating and maintenance pro cedures are described in APTD-0576. The components essential to this sampling train are the following: * acid TRAP 6.8 Isokinetic variation (comparison of velocity of gas in probe tip to stock velocity). , lOOKwi -.*U0(v,h,,. eq. 101-8 where: / Percent of Isokinetic sajnDbnc. T:*tA volume of gas sample v'tock condition*), ft*. .1 I'rahe tip area, ft*, c? *> time, sue. (!'<.*-.Vv-i.u** tjck gjs Velocity, f'-`'t |M`r 7. Dialuation o/ results--7.1 Determina tion of compliance.--7.1.1 Each performance test shall consist of three repetitions of the applicable test method. For the purpose of determining compliance with an applicable national emission standard, the average of results of all repetitions shall apply. 7.2 Acceptable Isokinetic results.-->7.2.1 The following range sets the limit on accept* able Isokinetic sampling results; If 90^^I^1107o, the results are accept able; otherwise, reject the test and repeat. 8. /Je/ere/: '$.--1. Addendum to Specifica. tions for li.j.nerator Testing at FederaL Facilities, PHS. NCAPC, Dec. 6.1967. 2. Determining Dust Concentration In a Gas Stream. ASMS Performance Test Code No. 27, New York. N.Y., 1957. 3. Devorkln, Howard, et a!.. Air Pollution Source Testing Manual, Air Pollution Con trol District, Los Angeles, Calif., Nov. 1963. 4. Hatch, \V. R. and W. L. Ott, "Determina tion of Sub-Microgram Quantities of Mercury by Atomic Absorption Spectrophotometry,** Anal. Chem.. 40:2085-87,1968. 5. Mark. L. S., Mechanical Engineers* Hand book. McGraw-Hill Book Co., Inc., New York, N.Y., 1951. 6. Martin. Robert M,, Construction Details of Isokinetic Source Sampling Equipment. Environmental Protection Agency, APTD- 0581. 7. Me`hvt3 for Determination of Velocity, Wiin'rf. D-: *. and Mist Content of Gases, "z o.t.ivion Division **f J-r Mfc. C*i.: V :->s. Cats'. 3ul. 1363. fi iVrr;\ J. H . Chemical Engineers' K\r.db<v*k. M*Cira.v-H:ll Book Co., Inc,, New York, N.Y.. VW. 9. Tio:'.. J ror.e J.. Maintenance, Calibrar:on. *\;.*1 Operation of Isokinetic Source S.vmp!s* Equipment. Environmental Protection A^esivy. APTD-0576. 10. S'.Uehara. R. T.. \v. F. Todd, and W. S. S.g:u.*icace of Errors In Stack Sam PUMP Figure 102-1. Mercury sampling train 2.1.1 Nozzle. Stainless steel or glass with sharp, tapered leading edge. 2.1.2 Probe. Sheathed Pyrex* glass. 2.1.3 Pifof tube. Type S (figure 102-2), or equivalent, with a coefficient within 5 per cent over the working range, attached to probe to monitor stack gas velocity. 2.1.4 Impingers. Four Greenburg-Sralth Imp Ingers connected in series with glass ball- joint fittings. The first, third, and fourth implngers may be modified by replacing the tip with one-half inch ID glass tube extend ing to one-half inch from the bottom of the flask. 2.1.5 Acid trap. Mine safety appliances air line filter, catalogue No. 81857, with acid ab sorbing cartridge and suitable connections, or equivalent. 2.1 6 M--tz7`.nn system. Vacuum xage. leak- less pump, thermometers capable of measur ing temperature to within 5*F, dry gas meter with 2 percent accuracy, and related equip ment, described In APTD-0581, to maintain an isokinetic sampling rate and to determine sample volume. 2.1.7 Barometer. To measure atmospheric pressure to 0.1 lnhg. 1The*e documents are available for a nomi nal cost from the National Technical In formation Service. U.S. Department of Com merce. 5235 Port Royal Road, Springfield, Va. 22151. Mention of trade names or commercial products does not constitute endorsement by the Environmental Protection Agency, FEDERAL REGISTER, VOL. 33, NO. 66--FRIDAY. APRIL 6, 1973 CAPCO JEN 0005607 RULES AND REGULATIONS 8841 2 2 Measurement of stack conditions 3.3.4 7f;/drochloric acid, 0.3.V. Dilute 25.5 4.2.2 The sampling site should be at least [-.tack pressure, temperature, moisture, and ml of couceutrated hydrochloric acid to 1 l eight stack or duct diameters downstream tvlot'Ujn--2,2.1 Pitot tube. Type S, or equivalent, with a coefficient within 5 per cent over iheworklng range. 2 2.2 Differential pressure gage. Inclined manometer, or equivalent, to measure veloc ity head to within 10 percent of the mini mum value. Micromanometers should be used If warranted. 2 2.3 Temperature gage. Any tempera ture-measuring device to measure stack tem with distilled water. 3.1 standard mercury solutions--3.4.1 Stock solution. Add 0.1354 g of mercuric chloride to 30 ml of 0 3N hydrochloric acid. After the mercuric chloride has dissolved, add 0.3N hydrochloric acid and adjust the volume to 100 ml. One ml of this solution is equivalent to 1 mg of free mercury. 3.4.2 Standard solutions. Prepare cali bration solutions by serially diluting the and two diameters upstream from any flow disturbance such as a bend, expansion or contraction. For rectangular cross section, determine an equivalent diameter from the following equation: D, 2L'.F L+W eq.102-1 where: D#=equivalent diameter. perature to wlthLn i* F. 2.2.4 Pressure gage. Pitot tube and In stock solution (3,4.1) with 0.3N hydrochloric acid. Prepare solutions at concentrations In L = length. W =2 width. clined manometer, or equivalent, to measure the linear working range for the instrument stack pressure to within 0.1 In hg. to be used. Solutions of 0.2 ig/ral, 0.4 /ig/ml 2 2.5 Moisture determination. Drying and 0.6 Ag/ml have been found acceptable tubes, condensers, or equivalent, to deter for most instruments. Store all solutions in mine stack gas moisture content in hydrogen glass-stoppered, glass bottles. These solutions to within 1 percent. should be stable for at least 2 months; how 2.3 Sample recovery--2.3.1 Leakless glass ever, periodic checks should be* performed sample bottles. 500 ml and 200 ml with Tef to Insure quality. lon-lined tops. 4. procedure. 4.1 Guidelines for source 2.3.2 Graduated cylinder. 250 ml. testing are detailed In the following sections. 2.3.3 Plastic jar. Approximately 300 ml. These guidelines are generally applicable; 2.4 Analysis--2.4.1 Spectrophotometer. however, most sample sites differ to some de To measure absorbance at 253.7 nm. Perkin gree and temporary alterations such as stack Elmer model 303, with a cylindrical gas cell extensions or expansions otten are required 4.2.3 When the above sampling site crite ria can be met, the minimum number of traverse points is four (4) for stacks 1 foot In diameter or less, eight (8) for stacks larger than 1 foot but 2 feet In diameter or less, and twelve (12) for stacks larger than 2 feet, 4.2.4 Some sampling situations may ren der the above sampling site criteria imprac tical. When this Is the case, choose a con venient sampling location and use figure 102-3 to determine the minimum number of traverse points. However, use figure 102-3 only for stacks l foot in diameter or larger. (approximately 1.5 In o.d. x 7 In) with quartz to Insure the best possible sample site. Fur 4.2.5 To use figure 102-3, first measure the glass windows, and hollow cathode source, or ther. since mercury U hazardous, care should distance from the chosen campling location equivalent. be taken to minimize exposure. Fnally, since to the nearest upstream and downstream dis 2.4 2 Gas sampling bubbler. Tudor Scien the total quantity of mercury to be collected turbances. Divide this distance by the di tific Co. Smog Bubbler, catalogue No. TP- generally is sm^tl, the test must be care ameter or equivalent diameter to determine 1150. or equivalent. fully conducted to prevent contamination or the distance in terms of pipe diameters. De 2.4 3 Recorder. To match output of loss of sample, termine the corresponding number of trav spectrophotometer. 4.2 Selection of a sampling site and mini el'se points for each distance from figure 3. Reagents.--3.1 Stock reagents.--3.1.1 Potassium Iodide. Reagent grade. 3.1.2 Distilled water. 3.1.3 Pota-satum Iodide solution, 25 per mum number of traverse points. 4 2.1 Select a suitable sampling site that Is as close os is practicable to the point of 102-3. Select the higher of the two numbers of traverse points, or a greater value, such atmospheric emission. If possible, stacks that for circular stacks the number Is a mul cent.--Dissolve 250 g of potassium iodide (re smaller than 1 foot in diameter should not tiple of four, and for rectangular stacks the agent 3.1.1) in distilled water and dilute to bo sampled. number follows the criteria of section 4.3 2. l to l. 3 ! 4 Hydrochloric acid. Concentrated. NUMBER OF OUCT DIAMETERS UPSTREAM' 3.1.5 Potassium todate. Reagent grade. (OISTAWCE A) 3 1.0 Iodine monochloride itCl) I.OM. To 800 ml of 25 percent potassium iodide 05 1.0 T.5 2.0 2 5 solution (reagent 3.1.3), add 800 ml of con centrated hydrochloric acid. Cool to room temperature. With vigorous stirring, slowly add 135 g of potassium lodate and continue stirring until all free Iodine has dissolved to give a clear orange-red solution. Cool to room temperature and dilute to 1.600 ml with dis tilled water. The solution should be kept in amber bottles to prevent degradation. 3.1.7 Sodium hydroxide pellets. Reagent grade. 3.1 8 nitric acid. Concentrated. 3.1 9 Hydroxylamine sulfate. Reagent grade. 3.1.10 Sodium chloride. Reagent grade. 3.1.11 Mercuric chloride. Reagent grade. 3.2 Sampling. 3.2.1 Absorbing solution. 0.1 M ICl. Dilute 100 ml of the 1.0M ICl stock solution (reagent 3.1.6) to 1 1 with dlstsllled water. The solution should be kept In glass bottles to prevent degradation. ThJs reagent should be stable for at least 2 months; how ever. periodic checks should be performed to insure quality. 3.2.2 Wash add. 2:1 V/V nitric acid-water. 3.2.3 Distilled, deionised water. 3.2 4 Silica gel. Indicating type, 6 to 16 mesh, dried at 350*F for 2 hours. 3.3. Analysis--3.3.1 Sodium hydroxide, 10H. Dissolve 400 g of sodium hydroxide pel lets in distilled water and dilute to I 1. 2 Reducing e.gent. I! percent hydror* ........ . 12 percent - d:::sn '.h'oruie. - ; * m- dtstilied-wa^r. :uld 12 i; of hy- ; N.'I.'.r.ur.e sulfate :i:id 12 t c*f i odium chlo- tut: Dilute to 100 ml. This quantity is su:: ::*nt for 20 analyses and must be pre- Fl3?3 104*3. Minimum number of Ittvsrss points. j. ir**d daily. 3 * i Aeration gas Zero grade air. 4.2.6 If a Selected sampling point Is closer tlon of that point to insure that the sample than 1 inch trom stack wall, adjust the loon- Is taken at least l inch away from the wall. FEDERAL REGISTER, VOL, 38, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005608 SS42 RULES AND REGULATIONS 4.3 Cross-sectional layout and location of traverse points. 4 3.1 For circular stacks locate the tra verse points on at least two diameters ac cording to figure 103-4 and table 103-1. The traverse axes shall divide the stack-cross sec tion Into equal parts. 4 3 2 For rectangular stacks divide the cross-section into as many equal rectangular cress as traverse points, such that the ratio of t*ie length to the width of the elemental areas :s between one and two. Locate the traverse prints at the centroid of each equal area ac inp*o4m.tsC*f4p*tflpeicNltloierut loJfrctff'tcMawUUr *nt,k ihowfof loeatfatof cording to figure 102-5. 4.4 Measurement of stack conditions. 4 4 1 Set up the apparatus aa shown in figure 102-2. Make sure all connections are tight and leak free. Measure the velocity head and temperature at the traverse points speci fied by section 4.3 and 4 3. 4.4 2 stack. 4.4 3 Measure the static pressure in the Determine the stack gas moisture. F'furt -5. CM) SMlwn f tuofc 4iv>4*tf Into u uui, *(Ui ua*r* points at ccnus.d ol tseft Table 102-1. Loeatfon of traverse point's In circular stacks (Percent of stack diameter from Inside wall to traverse point) Traverse point number Humber of traverse points on a diameter diameter 2 4 6 8 10 12 14 16 18 20 22 24 T 14.6 6.7 4.4 3.3 2.5 2.1 1.8 1.6 1.4 1.3 T.l 1.1 2 85.4 25.0 14.7 10.5 8.2 6.7 5.7. 4.9 4.4 3.9 3.5 3.2 3 75.0 29.5 19.4 14.6 11.8 9.9 8.5 7.5 6.7 6.0 5-5 4 93.3 70.5 32.3 22.6 17.7 14.6 12.5 10.9 9.7 8.7 7.9 5 85.3 67.7 34.2 25.0 20.1 16.9 14.6 12.9 11.6 10.5 6 95.6 0.6 65.3 35.5 26.9 22.0 18.S 16.5 14.6 13.2 7 89.5 77.4 64.5 36.6 28.3 23.6 20.4 13.0 16.1 8 96.7 85.4 75.0 63.4 37.5 29.6 25.0 21.8 19.4 3 91.8 82.3 73.1 62.5 38.2 30.6 26.1 23.0 10 97.5 88.2 79.9 71.7 61.8 '38.8 31.5 27.2 n 93.3 85.4 78.0 70.4 61.2 39.3 32.3 12 97.9 90.1 83.1 76.4 69.4 60.7 39.8 13 94.3 87.5 81.2 75.0 68.5 60.2 H 93.2 91.5 85.4 79.6 73.9 67.7" 15 95.1 89.1 63.5 78.2 72.8 16 98.4 92.5 87.1 82.0 77.0 17 95.6 90.3 85/4 80.6 18 98.6 93.3 88.4- 83.9 19 96.1 91.3 86.8 20 98.7 94.0 89.5 21 96.5 92.1 22 98.9 94.5 23 96.8 24 98.9 4.4.4 Determine the stack gas molecular weight from the measured'moisture content and knowledge of the expected gas stream composition. Sound engineering judgment should he used. 4.5 Preparation of sampling train. 4-5-1 Prior to assembly, clean all glass ware (probe, implngers. and connectors) by rinsing with wash acid, tap water. 0.1M ICl. tap water* and finally distilled water. Place 10O ml of 0.1M ZC1 In each of the first three implngers. and place approximately 200 g. of preweighed silica gel in the fourth impLager. Save 80 ml of the O.m IC1 as a In the sample analysis. Set up the train and the probe as in Figure 102-1. 4.5.2 Leak' check the sampling train at the sampling cite. The leakage rate should not be in excess of 1 percent of the desired sampling rate. Place crushed lee around the implngers. Add more Ice during the run to keep the temperature of the gases leaving the last implnger at 70* F or less. ,, 4.6 Mercury train operation. > 4.6.1 Safety procedures. It is imperative that the sampler conduct the source test under condition* of utmost safety, since hydrogen and air mixtures are explosive. The sample train essentially is leakless, so that attention to safe operation can be concen trated at the Inlet and outlet. The following specific item* are recommended: 4.6.1.1 Operate only the vacuum pump during the test* The other electrical equip ment, e.g. heaters, fan* and timers, normally are not essential to the success of a hydro gen stream test. 4.6.1.2 Seal the sample port to minimize leakage of hydrogen from the stack. 4.6.lit Vent sampled hydrogen at least 10 feet away from the train. This can be accomplished easily by attaching a /3-tn l.d. Tygon. tube to the exhaust from the orifice meter. 4.0. 2 For each run, record the data re quired on the sample sheet shown in figure 102-6. Take readings at each sampling point at least every S minutes when significant change* in stack conditions.necessitate ad ditional adjustments in flow rate. 4.0. 3 Sample at a rate of 0.5 to 1.0 cfm. Samples shall be taken over such a period or periods os are necessary to accurately determine the maximum emissions which would occur in a 24-hour period. In the case of cyclic operations, sufficient tests shall be made so as to allow accurate determination or calculation of the emissions which will occur over the duration of the cycle. A mini mum sample time of 2 hours Is recommended. In some instances, high mercury concentra- tidna can prevent sampling in one run for the desired minimum time. This is Indicated by reddening in the first implnger as free Iodine is liberated. In this case, a run may be divided into two or more subrtms to insure that the absorbing solutions are not depleted. FEDERAL REGISTER, VOL 33, NO. 60--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005609 RULES AND REGULATIONS 8843' nv.r_____ -- MK_____ r v 'a _____ ll*HHt TP**ATUf. |A*CM(UIC WSSUK. *y.wto vosnjW.Li, fiCtC MttGTtl._ >o._ |AsM>,vUMW)iv| I*t!'U.*C--* icmvahc o mci c*m not tin mITAul'L.CM**. KW1sfjufM.:*lv*M vUcMiA'ojDlo.n eifHM*fcKe*NiUmIltAl L UM(A.UHMRjO. CAttA^VU IV--* * cuAitm0*1iGAi mil MU O#U*U'l*f * t?(WA<*MA$AfIU<. nwMPtVuMiCtMlt ---------- T------- - 1 -------- i----------- amam. *.--- F.jj 102*5- Field dit* 4 6 4 To begin sampling, position the noz gen by dividing by 13. This factor includes zle at the first traverse point with the tip the ratio of the dry molecular weights and a pointing directly Into the gas stream. Imme correction for the different orifice calibration diately start the pump and adjust the flow factors for hydrogen and air. This procedure to uokinetic conditions. Sample for at least Is diagrammed below: 5 minutes at each traverse point: sampling time must be the same for each point. Main tain isokinetic sampling throughout the sam pling period, using the following procedures. 4.6.4.L Nomographs which aid in the rapid adjustment of the sampling rate without othr computations are in APTD-0576 and are available from commercial suppliers. The availaole nomographs, however, are set up !: lit -livil.*l*y la H71 luii m> t.-rbit. for uvi in air streams, and minor chances are required to provide applicability to h>drogen. 4.G 4 2 Calibrate the meter bo* orifice. Use the techniques as described in APTD-0576. l.G 4 6 Operate the sample train at the calculated 1H at each sample point. 4 0 5 Turn off the pump at the conclusion of each run and record the final readings. 4.G 4 3 The correction factor nomograph Immediately remove the probe and nozzle discussed In APTD-0576 and shown on the from the stack and handle in accordance with reverse s*de of commercial nomographs will the sample recovery process described In sec not be used. In Its place, the correction factor tion 4.7. will be calculated using equation 102-2. 4.7Sample recovery. 4.7.1 (All glass storage bottles and the c-.-o.oi (Ct,AM AH<g /\ T..x I'u. K eq. 102-2 where: C = Correction factor. Cr=-Pltot tube coefficient. M< --Mole fraction dry gas. P=Stack pressure. inHg. P = Me:er pressure. InHg. T Meter temperature. *R. M, = Molecular weight of stack gas (from 4.4.4). lb/lb mole. iH# = Meter bos calibration factor, ob tained In step 4.6.4.2. graduated cylinder must be precleaned as in section 4 5.1). This operation should be per formed in an area free of possible mercury contamination. Industrial laboratories and ambient air around mercury-using facilities are not normally free of mercury contamina tion. When the sampling train Is moved, care must be exercised to prevent breakage and contamination. 4.7.2 Disconnect the probe from the 1mplnger train. Place the contents (measured to tfcl ml) of the first three Implngers into a 500 ml sample bottle. Rinse the probe and alt glassware between it and the back half of the third lmplnger with two 50 ml por tions of 0.1M ICl solution. Add these rinses 4.6 4.4 Set the calculated correction factor to the first bottle. For a blank, place 80 ml on the front of the operating nomograph. of the 0.1M ICl In a 100 ml sample bottle. Select the proper nozzle and set the K-factor Place the silica gel in the plastic jar. Seal and on the nomograph as detailed In APTD-0576. secure all containers for shipment. If an ad 4 S 4 5 Read the velocity head in the stack ditional test Is desired, the glassware can be a*, ea-h `ample point from the manometer In carefully double rinsed with distilled water ' !.:: Convert the hydrogen IP to and r^a'*embl?u Ho.vever, if tne `tl^Soware ij a-; ,, .... i.f.ic . aiue for a:r oy multiplying by lv b- o.'t of u:e more than 2 days, the initial a rv.j of the molecular weigh: of air to hy acid wish procedure must Iv allowed. dro tea at the .ttack moisture content. Insert 4 8 Analysis--4.8.1 Apparatus prepara this vdue of aP onto the nomograph and tion.---Clem all glassware according to the read o.f iff, Azatn. convert the iff. which is procedure of section 4.5.1. Adjust the instru an a-.r c-junalent value, to the .*.// for hydro ment settings according to the instrument manual, using an absorption wavelength of 253.7 nm. 4.8.2 Analysts preparation.--Adjust the air delivery pressure and tne needle valve to obtain a constant air flow of about 1.3 l/min. The analysis tube should be bypassed ex cept during aeration. Purge the* equipment for 2 minutes. Prepare a sample of mercury standard solution (3.4.2) according to sec tion 4.8.3. Place the analysis tube in the line, and aerate until a maximum peak height Is reached on the recorder. Remove the analy sis tube, flush the lines, and rinse the analysis tube with distilled water. Repeat with auother sample of the same standard solution. This purge and analysis cycle is to be repeated until peak heights are repro ducible. 4.8.3 Sample preparation.--Just prior to analysis, transfer a sample aliquot of up to 50 ml to the cleaned 100 ml analysts tube. Adjust the volume to 50 ml with 0.1M ICl If required. Add 5 ml of 10 N sodium hydrox ide. cap tube with a clean glass stopper and shake vigorously. Prolonged, vigorous shak ing at this point is necessary to obtain an accurate analysis. Add 5 ml of the reducing agent (reagent 3.3.2), cap tube with a clean glass stopper and shake vigorously and Im mediately place in sample lme. 4.8.4 Mercury defermtnaiion.--After the system has been stabilized, prepare samples from the sample bottle according to section 4.3.3. Aerate the sample until a maximum peak height Is reached on the recorder. The mercury content is determined by comparing the peak heights of the samples to the peak heights of the calibration solutions. If col lected samples are out of the linear range, the samples should be diluted. Prepare a blank from the 100 ml bottle according to section 4.8.3 and analyze to determine the reagent blank mercury level.. 5. Calibration.--5.1 Sampling Train. 5 !.i Use standard methods and equipment as de tailed in APTD-0576 to c.;!jnra:e the rate meter, pitot tube and dry m^'er Recali brate prior to each test series. 5.2 Analyst?.--5.2.1 Prepare a calibra tion curve for the spectrophotometer lining the standard mercury solutions. Plot the peak heights read on the recorder versus the concentration of mercury in the standard solutions. Standards should be interspersed with the samples since the calibration can change slightly with time. A new calibration curve should be prepared for each new set of samples run. C. Calculations--45,l Average dry pas meter temperature, stack temperature, stack pres sure and average orifice pressure drop.--See data sheet (fig. 102-6). $2 Dry gas volume.--Correct the sample volume measured by the dry gas meter to stack conditions by using equation 102-3. y r. (Pb,r+l3~0 cq. 102-3 wh*-rc: Volume o! cas inspl throuch the dry gas ai^ter Otack coiKliuotw, ft.1 7-,= Volume of gas sample through the dry gas meter (meter conditions), ft*. 2*,=: Average temperature of stack gas. SR. T . raAverage dr/ gas meter temperature, St. ?.' r.ir'*:nvtf:c p.-e.vi*.:rj a- * M7 - .V.vr/gc procure drop .\cr_>.*; 'r.? ori fice meter. ir.H.O. 17.6-sSpecirtc gravity of mercury'. P.rrStr.ck pressure. P* rr:Sta::c r.r?w.:rc, inHg. No. 06--Pt. II- FEDERAL REGISTER, VOL 38, NO. 66---FRIDAY, APRIL 6, 1973 CAPCO JEN 0005610 SS44 fl 3 Volume of water vapor. eq. 102-1 b tn: .. r.t vst*r Taper la tb* gas ?api (slack vcr.'v--I;*, K*O.OOJa7 -'7^7, when iheM units *r* used, tcl. --* K Vi.-Total vuitiaw of liquid collected la Unpiagen m.U silica cii (see liiture ici-13. mi. T*-Av^r-iirr c vk i*:j tcmprrTUur^. n. F-St,w.k pressure, /*,, =fc static pressure. In. IT*. 6.4 Total gas volume. Vtot*i=V,+ Fw# cq. 102-5 where: Vt*ti=total volume of gas sample (stack conditions), ft*. V.'WVolum* of zas through dry gas meter (stack lAJidmon*). ltJ, l"w mVuiiuii* of water vapor in gas saispio (stack COKdillvD.'O, It*. VOUjMt 0# LiQU'O waji* coutcuo SI 4 VOUWC. ett JlllCA GtL VtlOll, < flSA*. INlTlAt { vccioccaccrro { TCTAi`.,V-,dCCUtCTW } CONvrcwt'OTc* vauTOvott-<w dividing toul flight ihCUAUiici'cmof iaiu. nu * V01UMC AH*, w* RULES AND REGULATIONS PLANT DATE.__________________________________________ RUNNO._______________________________________ STACK DIAMETER, in. BAROMETRIC PRESSURE, in. Hg;__________________ STATIC PRESSURE IN STACK (Pg), in. Hg. OPERATORS,;_____________________________________ Traverse point number Velocity head, in. HjO v7 [-------------------- --- SCHEMATIC OF STACK CROSS SECTION Stack Temperature Fisur* 102*7# Analytical data. 6.5 Stack gas velocity--Use equation 102-6 to calculate the stack gas velocity. (.)... = K.C.IVaP)..., eq 102-6 where: : Aterser stack gas velocity, feet per second, K, ... ft f lb mile \u\ . . S'"3.in m0le*-K*inH:O ) *b m c. lh*'se amts are used. = Ttrot tube coefficient, dimensionless. Average slack gss temperature, *ft. (VaF) *.* Avomce square root of the velocity head of Stack gas (inlljO)*/' (see figure 102-8). p, K$(I.fut.fk. pressure, Fk4tsuuc pressure, tn ,1/, aiMoleCul.tr weight of stark gas (wet basis), the summation of the products of the molecular weight of each component multiplied by its voftzmetric proportion nr the mixture, Ib,lb*moJe. Figure 102-8 shows a sample recording sheet for velocity traverse data. Use the averages in. the last two columns of figure 102-6 to de termine the average stack gas velocity from equation 102-6. 6.6 Mercury collected. Calculate the total weight of mercury collected hy using eq. 102-7. t AVERAGE: Figure 102-8. Velocity traverse data. FEDERAL U;OISTI, VOl. 33. MO. It--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005611 RULES AND REGULATIONS SS45 Wi = V.Ci-VC.......... eq. 102-7 uiiere: U't = Total ael^ht of mercury collected. MgVi = Total volume of condensed moisture and ICl In sample bottle, ml. Ci = Concentration of mercury measured In rimple bottle, Mg/ml. Vies Total volume of ICl used in samp.ing (implitter contents and all wash amounts), ml. C = Blank concentration of mercury In ICl solution* /tg.'ml. 10. Shistehara. R. T., W. P. Todd, and IV. 3. Smith, Significance of Errors In Stack Sam pling Measurements, Paper presented at the Annual Meeting of the Air Pollution Control Association, St. Lou(3. Mo., June 14-19, 1970. 11. Smith. \V, S., et al.. Stack Gas Sam pling Improved and Simplified with New Equipment. APCA paper No. 67-119, 1967. 12. Smith. W. S.. R. T. Shlgehaca. and W. P. Todd, A Method of Interpreting Stack Sam pling Data, Paper presented at the 63d An nual Meeting of the Air Pollution Control Association, St. Louis, Mo.. June 1+-19, 1970. 2.2.3.Temperature gauge.--Any tempera ture measuring device to measure stack tem perature to within 5% P. 2.2.4 Pressure gauge.--Any device to measure stack pressure to within O.i In. ilg. 2.2.5 Barometer.---'To measure atmos pheric pressure to within 0.1 In. Hg. 1 2.2.6 iVfoisfure determination.--Wet and dry httib thermometers, drying tubes, con densers. or equivalent, to determine stack gas moisture content to within 1 percent. 2.3 Sample recovery.--2.3.1 Probe cieatt- ing equipment.--Probe brush or cleaning rod 6 7 Total mercury emission.--Calculate 13. Specifications for Incinerator Testing at least as long as probe, or equivalent. Clean the totat amount of mercury emitted from at Federal Facilities PHS, NCAPC, 1067. cotton balls, or equivalent, should be used each siack per day by equation 102-8. This equation Is applicable for continuous opera tions. Tor cyclic operations, use only the time 14. Standard Method for Sampling Stacks for Particulate Matter, la: 1971 Book of ASTM Standards, part 23, Philadelphia. 1971, with tho rod., 2.3.2 Leakless glass sample bottles. 2.4 Analysis.--2.4.1 Equipment neces per day each stack U in operation. The total ASTM Designation D-2928-72. sary to perform an atomic absorption, mercury emissions from a source will be the summation of results from all stacks. UVr.Uy* .4, ,3MQ0 sftc^nd- l1.iv* U= !',*. X 107p3/S eq. 102-8 15. Vennard, J. K.. Elementary Fluid Me chanics, John Wiley and Sons, Inc., New York. 1947. MTHO0 103. BERYLLIUM SCREENING METHOD 1. Principle and applicability.--1.1 Principle.--Beryllium emissions are isoklnctically spectrographic, fluorometrlc, chromato graphic, or equivalent analysis. 3. Reagents.--3.1 Sample recovery.--3.1.1 Acetone.--Reagent grade. 3.1.2 Wash acid.--1:1 V/V hydrochloric acid-water. 3.2 Analysis.--3.2.1 Reagents as neces sampled from three points in a duct or stack. sary for the selected analytical procedure. v of -rr.'.ssion. r.day. .'\ \\ \w.-u Ci ::.-f'.ury *!. \\ .,i- r--i.il V'/.ua'.eoibM^jnpltf .>:.!.** .yi.ditw t, tl\ ( .\*,c Av-r.r- *t.iols ?as velocity, t-.-i r^r second. The collected sample Ls analysed for beryl lium uhtr.g an appropriate technique, 1.2 Applicability.--This procedure details guidelines and requirements for methodq acceptable for use In determining beryllium 4. Procedure.--4.1 Guidelines for source testing are detailed In th_` following .vctlon.i. These guidelines are generally applicable; however, most sample sites differ to some de gree and temporary alterations such as stack .1,ci.v.k or.-i, UK r>3 Is^kin'i.c variation (comparison of velocity of pcs to probe tip to stack velocity). _ 1001 lQf*l mI.-.i-' I -V /1.3(r,)..,. eq. 102-0 oli-*VittHu*Mii*.pWc l-T ' .1 c :uo:coff i' s.*i:*j.* *(. k emissions In ducts or stacks at scatiouary sources, as specified under the provisions of 3 61.14 of the regulations. 2. Apparatus--2.1 Sampling trein.---A schematic of the required sampling train configuration Is shown In figure 103-1. The essential components of the train are the following: 2.1.1 Nozzle.--Stainless steel. or equiva extensions or expansions often are required to insure the best possible sample site. Fur ther, since beryllium is hazardous, care should be taken to minimize exposure. Finally, since the total quantity of beryllium to be collected Is quite small, the test must be -carefully conducted to prevent contami nation or loss of sample. 4.2 Selection of a sampling site and num v ; im i.-'M. ; V. *!. >.: . r*r -' ' !. 7. "lion of results.--7 1 Determina tion v: * i.;.J:a:ice.--7.1.1 Each performance teot s'.* .`.t cV/ivdst of three repitr.ions o;' the lent, with sharp, tapered lading edge. ber of runs.--4 2.1 Select a suitable sam 2.1.2 Probe.--Sheathed Pyrex ! gla'.s. pling site that Is as cloie as practicable to the 2.1.3 Filter.--MiUipore AA. or cquiva'ont. point of atmospheric omUsion. If possible, with appropriate niter houi.'r tha* piovides r.tacks smaller than 1 foot in diameter should a positive seal against leakage from outude not be sampled. or around the filter. It is ugi;ested that a 4 2.2 The sampling site should be at least a;pi:.-..`AC test method. For the purpose of determining compliance with an applicable national emission standard, the average of results of all repetitions shall apply. 7 2 Acceptable tsokinct.c results.--7.2.1 The following range sets the limit on ac ceptable isokinetic sampling results: If COH.^I.^UO".. the results are acceptable; otherwl-e. reject the test and repeat. Whatman 41, or equivalent, be placed imme diately against the back side of the MtiUpore filter as a guard against breakage of the MllUpore. Include the Whatman 41 In the analysis. Equivalent fibers must ho at lcat 99 95 percent cificlent (DOP Test) and amenable lo the analytical procedure. mu* clfiht stack or duct diameters downstream and two diameters upstream from any flow disturbance such &3 a bend, expansion or contraction. For rectangular cross-section, determine an equivalent diameter using the following equation: 2LW ----------- -------- --------------------eq. 102-1 8. ft**/vences.--l. Addendum to Specifi cation > for Incinerator Testing at Federal Faci'.i'tes. MIS. NCAPC. Dec. 6, 1967. 2. Determining Dust Concentration in a Gas Stream. ASME Performance Test Code where: D. = equivalent diameter L=lcugth IV=width No. 27, New York. N.Y., 1957. 4.2.3 Some sampling situations may ren 3. Devorkin, Howard, et al.. Air Pollution Source Testing Manual, Air Pollution Con trol District. Los Angeles. Calif.. Nov. 1963. der the above sampling site criteria imprac tical. When, this is the case, an alternate site may be selected but must be no less 4. Hatch, W. R. and W. L. Ott, "Determina tion of Sub-Mlcrogram Quantities of Mer cury by Atomic Absorption Spectrophotom etry,'* Anal. Chem., 40: 2085-87, 1968. 5. Mark, L. S., Mechanical Engineers* Handbook. McGraw-HIU Book Co., Inc., New York. N.Y., 1951. C. Martin. Robert M,, Construction Details of Isokinetic Source Sampling Equipment, Environmental Protection Agency, APTD0.791. 7. f-.*r Determination of Velocity, \ .. uj M:t Cv.r v i'.i'.n Dr l-i .a <*r J .''..mu* Li A:: Cal.Ml jll.'.VP-.70, i H P`.-r". J H. Chemical Engineer.* Handbv-h. J'- Gra*.v-HiU Book Co., Inc . N^.v York, :* v : > - t- f: j. Jerome J.. Maintenance. CaUbra- t*.u. Operation of Isokinetic Source famplu:.' equipment. Environmental Protec tion icy. APTD-0576. Fijwt 103-1. Beryllium Krttningflttthod: tampt* tiifn schemalie. 2.1.4 Meter-pump system.--Any system that will maintain isokinetic sampling rate, determine sample volume, and Is capable of a sampling rate of greater than 0.5 cfm. 2.2 Measurement of stack conditions (v?aefe pressure, temperature, moisture and iclocity).--The following equipment shall he u<ed in the manner specified in section 4.3.1. 2 2.1 P:*'*: f--T-pe N. **r c:.T.`.v'd;`:\r. a is, ...-1c::: u.'rrer.r. :v? working rv.*ge. 2 2.2 Differential procure t;-rnjr.--In clined manometer, or equivalent, to measure velocity head to within 10 percent of the minimum value.1 1 Mention of trade names or specific prod uct-* does not constitute endorsement by the Environmental Protection Agency. than two diameters downstream and onehalf diameter upstream from, any point of disturbance. Additional sample runs are rec ommended at any sample site not meeting the criteria of section 4.2.2. 4.2.4 Three runs shall constitute a test. The runs shall be conducted at three dif ferent poinu. The three points shall pro portionately divide the diameter, l.e. be lo cated at 23. 50 and. 75 percent of the d: irneter from the Inside vMl. For hon.-otrid duc"s. tha d'-unevr rhaU he H *:' dir-'-- 11'J'll ir *..* ; . v *.*h :!:*: ccntrci'i avl i* a .,,i\ if , m?:* .-cs ,% 4 2.7. proportionately divide *.!*- due: commodate the total number of rsn<. ac 4.3 Mvaiurement of stwk c'lV-.-Mm. 4.3.1 Measure the .--lack > pres.c.re. mois ture. and jemuerature, using`the equipment described In J 2.2. Determine the molecular weight of the stack gas. Sound engineering estimates may be made in lieu of direct FEDERAL REGISTER, VOl. 38, NO. 66--FZ10AY, APRIL 6, 1973 CAPCO JEN 0005612 SS4S RULES AND REGULATIONS measurements. The basis for such estimates shall be given, in the test report. 4.4 Preparation o/ sampling train.-- 4.4.1. Assemble the sampling train as shown In flguro 103-i. It Is recommended that all glassware be precleaned by sositing in wash acid for 2 hours. 4.4.2. Leak check the sampling train at the sampling site. The leakage rate should not be In excess at l percent of the desired sample rate. 4.5 Beryllium train operation.--4.5.1 For each run. measure the velocity at the selected sampling point. Determine the isokinetic sampling rate. Record the velocity head and the required sampling rate. 4.52 place the nozzle at the sampling point with the tip pointing directly into the gas stream. Immediately start the pump and adjust the flow to Isokinetic conditions. At the conclusion of the test, record the sam pling rate. Again measure the velocity head at the sampling point. The required isokinetic rate at the end of the period should not have deviated more than 20 percent from that originally calculated. 4.52 Sample at a minimum rate of 02 ft'/mln. Samples shall be. taken over such a period or per:cds as are necessary to deter mine the maximum emissions which would occur In a 24-hour period. In the cue e cyclic operations, sufficient tests shall be made so as to allow determination or calcu lation at the emissions, which would occur over the duration of th* cycle. A sampling time of Z hours Is recommended. 42.4 AIL pertinent data should be in cluded in the test report. 4.6 Sample recovery.-4.6A It is recom mended that all glassware be preeleaned aa In | 4A.1. Sample recovery should also be performed In an area free of possible beryl lium contamination. When the sampling train Is moved, exercise care to prevent breakage and contamination. Set aside a por tion of the acetone used in the sample re covery as a blank for analysis. The total amount of acetone used should be measured for accurate blank correction. Blanks can be. eliminated If prior analysis shows negligible amounts. 4.6.2 Remove the filter and any loose par ticulate matter from filter holder and place In a container. 4.62 Clean the probe with acetone and & brush or long rod and cotton balls. Wash Into the container. Wash out the filter holder with acetone and add to the same container. 4.7 Analysis.--4.7.1 Make the necessary preparation of samples and analyze for beryl lium. Any currently acceptable method such as atomic absorptlon.^spectrographlc, fiuorometrlc, chromatographic, or equivalent may be used* S. Calibration and standards--6.1 Sam pling train.--5.1.1 As a procedural check, sampling rate regulation should be compared with a dry gas meter, spirometer*, rotameter (calibrated for prevailing atmospheric con ditions), or equivalent, attached to nozzleinlet of the complete sampling train. 5.1.2 Data from this test and calculations should be shown in test report. 5.2 Analysis.--5.2.1 Standardization Is made as suggested by the manufacturer of the instrument or the procedures for the analytical method. x 6. Calculations--6.1 Total beryllium emis- .on. Calculate the total amount of beryl lium rmUtecl .'r?m each stack per day by er:ua*".on K'l-2. r.i:s equation is applicable wf continuous * peracions. Eor cyclic opera tions, use only che time per day each stack is in operation. The total beryllium emis sions from a source will be the summation of results fre-m all stacks. .4,,yS6,400seconds/dav V^i 10Vff/ff- wlier*. Raw of*m<sion, pMajr. Ur,m'Total wtjrlit uf btryiliura uf. Vu4*j*Toeai; volume of Kas sampled, ft1. Aver.ice alack g** velocity, feet per second. .U-Stack area, ft1. 7. Test report. 7.1 A test report shall be prepared which shall Include as a minimum..* 7.12 A detailed: description of the sam pling train used and results of the proce dural check with all data, and calculations made. 7.1.2* AIL pertinent data taken during test*.the basis for any estimates- made* cal culations, and results. 7.12 A description of the test site, in cluding a block diagram with a brief de scription of the process, location of the sam ple points in the cross section, dimensions and distances from any point of disturbance. METJfOD- 104. BEFXBXIVCE MTTBOD 70* DETZKStlNATXON OP BEXTIXIUM. EMISSIONS. FROM ST4XX024AXT SOUBCES. 1. Principle and applicability--1.1 Prin ciple.--Beryllium emissions are isokinetically sampled from the source, and the-collected sample Is digested la in. mid solution nd analyzed by atomic absorption spectronho- tometry. * 12 Applicability.--This method, ia appli cable for the. determination ct. beryllium emissions in ducts or stacks at stationary sources. Unless otherwise specified, this method la not Intended to apply to gas streams other than those emitted directly to the atmosphere w tout, further processing* 2. Apparatus--Zl Sampling train.--A schematic of the sampling train used by EPA is shown In figure* 104-1. Commercial models of this train are available, although construction details are described- La- AFTZX0551,* and operating and zoaintenanee pro cedures are described to AFTD-0676. The components essential to this sampling train are the following? 2.1.1 Hozzle.---Stainless steel or glass with sharp* tapered leading edge. 2.1.2 Probe.--Sheathed Pyre** glass. A heating system capable of maintaining a minimum gas temperature in the range of the stack temperature at the probe octree during sampling may be used to prevent condensation, from occurring* PUMP Figure t04-t. Beryllium sampling train 2.1.3 Pitot tube.--Type S (figure 104-2), oc equivalent, with a coefficient within 5 per cent over the working range, attached to probe to monitor stack gas velocity. 2.1.4 Filter AoWer.--Pyrex glass. The filter holder must provide a positive seal against leakage from outside or around the filter. A heating system capable of maintaining the filter at a minimum temperature to therango of the stack, temperature may be' used to prevent condensation from occurring: 2.1.5 Impingers.--Four Cieeub ui g-Smith Impingers connected in series with glass ball Joint fittings. The first, thirds and fourth Impingers may be modified by replacing the tip with a `/--inch l.d. glass cube extending to one-half inch from the bottom of the flask. 2.16 Metering tysfem.--Vacuum gauge, leakless pomp, thermometers capable of measuring temperature to within. 5* P. dry gas meter with 2 percent accuracy, and re lated equipment, described la. APTD-0581, to maintain an Isokinetic sampling rate and to determine sample volume. 2.1.7 Barometer.--To measure atmos pheric pressure to 02 in Hg. 22 Measurement of stack conditions- tstack pressure, temperature, moisture and velocity)--22.1 Pitot tube.--Type S, or equivalent, with a coefficient within 3 percent, over the working range. 222 Differential pressure gauge.--In clined manometer, or equivalent; to measure velocity head* to within 15 percent of theminimum value.1 * * * 5 1 These documents are available for a nom inal cose from the National Technical In formation Service, US. Deuartment of Com merce. 5235 Port Royal Road, Springfield, Va. 2215U 5Mention of trade names on specific prod ucts does not constitute endorsement by the Environmental Protection Agency, FEDERAL REGISTER, VOl. 38, NO. 6--FRIDAY, APRIL 6. 1973 CAPCO JEN 0005613 RULES AND REGULATIONS 8847 however, most sample sites differ to some degree aud temporary alterations such as stack: extensions or expansions often are tequlred to insure the best possible sample site. Further, since beryllium Is hazardous, care should be taken to minimize exposure. Finally, since the total quantity of beryllium to be collected Is quite small, the test must be carefully conducted to prevent* contami nation or. loss of sample. 4.2 Selection of a sampling site and mini mum number of traverse points. 4.3.1 Select a suitable sampling site that Is as ctose as practicable to the point of at mospheric emission. If possible, stacks smaller than l foot In diameter should not be sampled. 4.2.2 The sampling site should toe at least 8 stack or duct diameters downstream and 2 diameters upstream from any flow disturb ance such as a bend, expansion or contrac tion. For a rectangular cross-section, deter mine an equivalent diameter from the following equation: Df=z2LW l+w where: 4?#ssequivalent diameter L=length Ws= width eq. 104-1 NUMBER OF DUCT DIAMETERS UPSTREAM* {DISTANCE A) ffjuv !04*f. Filet lab# rnMowUr 0 & 1.0 1.5 2.0 2.5 2.2.3 Temperature gage.--Any tempera ture measuring device to measure stack tem perature to within 5* F. 2.2.4 Pressure gage.--Pilot tube and In clined manometer, or equivalent, to measure stack pressure to within 0.1 in Hg. 2.2.5 .'v/osture determination,--Wet and dry bulb thermometers, drying tubes, con densers, or equivalent, to determine stack gas moisture content to within l percent. 2.3 Sample recovery--2.3.1 Probe clean- ing rod.--At least as long as probe. 2.3.2 Leakless glass sample bottles.--500 ml. 2.3.3 Graduated cylinder.--250 ml. 2.3.4 Plastic jar.--Approximately 300 ml, 2.4 Analysts--2.4.1 Atomic absorption spectrophotometer.--To measure absorbance at 234.8 nra. Perkin Elmer Model 303, or equivalent, with N^O/acetylene burner. 2.4 2 Hot plate. 2.4.3 Perchloric acid fume hood, 3. F.cagents--3.1 Stock recgcrls.--3.1.1 Hydrochlonc acid.--Concentrated, 3.1.2 Perchloric acid.--Concentrated. 70 percent. 3 1,3 Nitric acid.--Concentrated. 3.1.4 Sulfuric acid.--Concentrated. 3.1.5 Dtstilted and deionized water. 3.1.0 Beryllium powder.--98 percent mini mum purity. 3.2 Sampling--3.2.1 Filter. -- MiUlpore AA, or equivalent. It Is suggested that a Whatman 41 filter be placed Immediately against the back side of the MiUlpore filter as a guard against breaking the MiUlpore filter. In the analysis of the filter, the What Figure i0i-3. Minimum number of traverse points. man 4l filter should be Included with the MiUlpore filter. 3.2.2 Silica gel.--Indicating type, 6 to 10 mesh, dried at 350* F for 2 hours. 3.2.3 Distilled and deionized water. 3.3 Sample recovery--3.3.1 Distilled and deionized water. 3.3.2 Acetone.--Reagent grade. 4.2.4 Some sampling situations may ren der the above sampling site criteria Imprac tical. When this is the case, choose a con venient sampling location and use figure 104-3 to determine the minimum number of traverse points. However, use figure 104-3 only for stacks 1 toot in diameter or larger. 4.2.5 To use figure 104-3, first measure 3.3.3 Wash acid.--1.1 V/V hydrochloric acid-water. 3.4 Analysis.--3.4.1 Sulfuric acid solu tion, 12 N.--Dilute 333 ml of concentrated sulfuric acid to 1 1 with distilled water. Ftaura 104-4. Cress of circular lU;k showing`toctlfat Of traverse paints on ptrptrrticmv dlualtts. the distance from the chosen sampling lo cation to the nearest upstream and down stream disturbances. Divide this distance by the diameter or equivalent diameter to deter mine the distance in terms of pipe diameters. 3.4 2 25 percent V/V hydrochloric acidwater. Determine the corresponding number of traverse points for each distance from fig 3.5 Standard beryllium solution--3.5.1 stock solution.--i ug/tnl beryllium. Dis solve 10 mg of beryllium In 80 ml of 12 N sulfuric acid solution and dilute to a volume of i- -.I . ,:h d.-.rilled v.w.er. Diu"e a 10 ml : -. v ;?* v.::h 25 p**rr*.*it v* V l ydro- ure 104--3. Select the higher of the two num bers of traverse points, or a greater value, such that for circular stacks the number Is a multiple of four, and for rectangular .vacks the number follow* the <*rr*>rla of rc-tti-.M: rh.-nz in d. tvmg a concentration *of 1 -1.3.2. (fA ?:*l. This d.Jure stock solution .hc-uld be prepared fre-h daily. Equivalent strongtn (m bo-yihum) stock solutions may be prepared fro:n bcryllivm salts as BeCl. and 193 percent minimum purity)". 4. Proc:dure. 4.1 Guidelines for source testing are detailed in the following sections. These guidelines are generally applicable; 4.2.3 When the above sampling site cri teria can be met. the minimum number of traverse points Is four (4) for stacks 1 foot lrx diameter or less, eight <8) for stacks larger than 1 foot but 2 feet In diameter or less, and 'twelve (12) for stacks larger than 2 feet. 4.2.0 If a ^elected sampling point ss closer than l Inch from the stack wall, adjust the location of that point to ensure that (he sample Is taken at least 1 Inch away from the wall. 4.3 Cross-sectional layout aud location of traverse points. FEDERAL REGISTER, VOL 30, NO. 65--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005614 ss+s RULES AND REGULATIONS iS? IS a| if* II wu ii r is in item jf5S*-i II! um O B 73 ag-i. * * wO. & "d S4 4^OlMM 5- s S2a5S g 5 > "3 to 2 oE2--O^Sg 5fot-g32'*&tS: g*-D S* S 5? "SIS3 d"2 aS a *2 C2 tf ts c ni &c >tjS; C* g> 7jQ? -**P*?**o d^| w J ac o S s c. S h a 2 >. 5 Sflf?6 s - 11tgf " sSUSSa 5 -Os Es-. <C* 5C 4O5 ; *srS CfSTTS Ill J_L 2..Ho 2JolTt3*"3r o=s. ec&0,.S3ffS5g".---3SgS--St--25* n g2 sf*c-*.$sH*c3g?o,*=S2g^sI"S2Osgc5^^f-eL'j*,Jow2*;*. 5 its a. 5*2=5 a"S5=sp""SfiaSS" ?S5-u": *"siSSs23ea)jSsii R"aaf ,,,a7u 3cC.-,aD w.0eo *s* |* =c. i&-. so 5.5 i 7.9 10.5 23.0 32.3 39.8 60.2 72.8 77.0 80.6 83.9 86.8 89.5 92.1 96.8 98.9 CSM' r- CCMO rCC--MO> VO O0*l CM CM PrVOv iOnrl rf -" U1 o VO *-- VO o CO CM rCM i,C_nO CCcnOO tCvvo.l in vo rOt*o.l CCr*OMv oCM in CO CfmOm o^T* iVnO Ocnl ao> -o0 ao C (CO* Cl r> r- Cl in <vr CM o VO CO CO CO CpVMO Ovorl o |V tOro-l imCnO Ptv- CoO cCoO *VO- revo m cn VO CM cvCnoM CM CO CO oIN vo pCCMO *inr p^ COOi mCM vino vCoO VD Ol. tn m Cl* O CM CO CM in CO mVCOM rppv*. o rs. CCOO fCinvO in ru-i; car I!Oo. 41 "MOC*- (A A s C3 r^> o VO c* CM CvCMnl VtCoOO CvOof *Crr-O* IOOnll tCOnO" rOO--l CO cn CCMO p. r^. _CO r"* iOCnM 'iicnno in vo oin> CCOM * CCCMOO CCOO YOvl. >to ICY CM vo CM CCrOM^ ixanoy rr-.. tnr rC-O* fmN. o .n r> UT cv cCCnOM VNO. VoCOO mOl ft^o ; 1-- !1ii! |j tC/J| 4---* W CM WOV cf r--r mCINl rtons mto VLVOO* VO CainM ipon* cPoO VO r* in CM O *** L0 r> CO Ol o -- CM rCO r- tn rm-- Prv r-- V?- owpCM. <CVMJ CO .* [ <- 'O S2JcCddva0 2oEm-->1i. 5Sgi * eg 5 lcS-^agg2c s-Ss^*Si5 %esfa a eaH"g2 w1 nr -- 5j) tl.sss5? O o'Ss 2g2S is&sd sJSuw,aR!gs*o*-*o4wi -*-nf: 5*s2fl" C B>1 5245S4egt5weo*0<2crdtgla,,,1w ttCsl S~ 5wE-noe---25CdeS-?. !tJ=OVJiSJe5oOSt' IISC g5S-ga s'iligl OQS! AoEa -'-Sga 'r--^ -g-S?,B'c: cOo eJo3 -2 s?l SS d4d5 IVc M > iS 1 , o .o3 fEt ~ A<jS t*S*f<3t^A 3aJ*j -2cOig22tfd3S HzUs u <* . coLIi. ct-=a!l V,,2; *>-o5> 5?ico-a:' "og5S= g gas| AsC rs 2Ji *2* M * *d ** t;3gg u c-2L 3t> ^o52OCaSX" _c.poutiafs2lsOoSsig*c2f>gV-c*..-og5oM5*7^"B*>^*i SeoWi c3oK --152= g"1 in *a .*-3a*ac-3c:*=.&2C4o.*^^o:.,-i-dS.,cu-> | SrV? r2O*-. stco S ^O 5 O??Su5So 5O C O flS C.O ( ( 1 CAPCO JEN 000561S RULES AND REGULATIONS S&J9 from the ,tack and handle In accordance with the sample recovery process desr-:bed m 5 4.7. 4.7 sample recovery,--i.7.1 (All glass storage bottles and the graduated cylinder must be precleaned as In J 4 5.1.) This opera tion should be performed In an area free of possible beryllium contamination. When the sampling train Is moved, care must be exer cised to prevent breakage and contamination. 4.7.2 Disconnect the probe from the lmpinger train. Remove the filter and any loose p.irctculate matter from the filter holder and place In a sample bottle. Place the contents (measured to --1 ml) of the first three lmptngers Into another sample bottle. Rinse the probe and all glassware between it and the back half of the third Implnger with water and acetone, and add this to the latter sam ple bottle. Clean the probe with a brush or a long slender rod and cotton balls. Use acetone while cleaning. Add these to the sample bot tle. Retain a sample of the water and acetone as a blank. The total amount of wash water and acetone used should be measured for ac curate blank correction. Place the silica gel In the plastic Jar. Seal and secure all sample containers for shipment. If an additional test Is desired, the glassware can be carefully dou ble rinsed with distilled water and reassem bled. However, if the glassware U to be out of use more than 2 days, the initial acid wash procedure must be followed. 4.8 Analysis. 4.6.1 Apparatus preparation.--C\ean all glassware according to the procedure of sec tion 4.5.1. Adjust the instrument settings according to the Instrument manual, using an absorption wavelength of 234.8 nm. 4.8.2 Sample preparation.--The digestion of beryllium samples Is accomplished In part in concentrated perchloric acid. Caution: The analyst must insure that the sample Is heated to tight brown fumes after the initial nitric aud addition; otherwise, dangerous perchlorates may result from the subsequent perchloric acid digestion. Perchloric acid also should be used only under a perchloric acid hood. 4 8.2.1 Transfer the filter and any loose particulate matter from the sample container to a ISO ml beaker. Add 35 ml concentrated nitric acid. Heat on a hotplate until light brown fumes are evident to destroy all or ganic matter. Cool to room temperature and add 5 ml concentrated sulfuric acid and 5 ml concentrated perchloric acid. Then pro ceed with step 4.8.2.4. 4.8 2.2 Place & portion of the water and acetone sample Into a 150 ml beaker and put on a hotplate. Add portions of the remainder as evaporation proceeds and evaporate to dry ness. Cool the residue and add 35 ml concen trated nitric acid. Heat on a hotplate until light brown.fumes are evident to destroy any organic matter. Cool to room temperature and add 5 ml concentrated sulfuric acid, and 5 ml concentrated perchloric acid. Then pro 6.3 Volume o/ water vapor. ceed "Mih ilep 4 8 2.4. 4.8 2.3 Weuh the spent .silica gel aud ro- port to the nearest gram. 4 8 2.4 Samples from 1.8 2.1 and 4 8 2.2 nuy be combined here for a.ie of analysis. Replace on a hotplate and evaporate to dry ness In a perchloric acid hood. Cool aud dis solve the residue in 10.0 ml of 25 percent V/V hvdrochlorlc acid. Samples are now ready for the atomic absorption unit. The bervliium concentration pf the sample must be within the calibration range of the unit. Trr __ T- tr 'fjL eq.104-3 where: 1% at u*hr vapor in tin* gas &mipJe {>\ick cuuihri'Mix, ft1. V,tafia(:67 when these units are used. volume of H'juM collected i;i Impin^r* and siiic.i l'i`1 (>* tiaiue lfil-7), ml. Ti.\\vr;U'" >.k eos temperature, *K. i*,**St.n k pi-s,-ure, i'tMrslaUc pressure, lit Ur. 6.4 Total gas volume. If necessary, further dilution of sample with 25 percent V/V hydrochloric acid must be performed to bring the sample within the cf|. 104-4 calibration range. 4.8.3 Ueryllium determination.--Analyze the samples prepared in 4.8.2 at 234.8 mu using a nitrous oxide/acetylene dame. Alumi num, silicon and other elements can Inter fere with this method if present in large quantities. Standard methods are available, however, to effectively eliminate these Inter ferences (see Reference 5). 5. Calibration--5.1 Sampling train.-- where: 1 <rt "Total volume of gas sample (tack conditions), ft*. r*,#"Vuiiunt of gas through dry ga* mj*ter (stack conditions), (V. TVVolum* of water vapor In g.vs sample f>:.ick condition*), ft*. 6.5 Stack gas velocity. Use equation 104-5 to calculate the stack, gas velocity. 5.1.1 Use standard methods a;:d equipment as detailed in APTD-0575 to calibrate the rate teeter. plrot tube, dry gas meter and probe heater (If used). Recalibrate prior to each cq. 104-j test series. 5 2 Analysis__ 5.2.1 Standardization Is made with the procedure as suggested by the manufacturer with staudard beryllium solu tion. Standard solutions will be prepared where: (M*-*."Average stack gas velocity, f*t pec second. Ar-, = 33.J3Mfti;/^lbmoIlbe-_i>n1lTU*iaIj0\j*n. w..n from the stock solution by dilution with 25 percent V/V hydrochloric acid. The linearity these units are used. C, Pitot tube coefficient, d(mensionl**?a. of working range should be established with a series of standard solutions. If collected samples are out of the linear range, the samples should be diluted. Standards should be interspersed with the samples since the calibration can change slightly with time. 6. Csfcn.'af sons--6 1 Arerupc dry gas meter iemprratu'c. stack temperature, stack pretitre and cterage or'.f.ce pressure drop.--See (TV) *." Average stack gas temperature, *R. (y/ap) "Average square root of the velocity h-ad , ofstackgasUnLIiO^Csee figure A "Stack pressure, A./istaUc pressure, la HC. Mok-tjtar Icht of i?k cos 'wet l. ili '<i:nn.atnja u( Lite prflM* t*j idu 'vi.ct.t of *-It . . -it :nn|iipliM i.y its v..ht'n**!r; ; .5. rv> t in the nurture, ih data sheet (figure 104-6). 6.2 Dry gas volume,--Correct the sample volume measured by the dry gas meter to stack conditions by using equation 104-2. VOLUME Of LIQUID VAIO COLUCUD iwiNote VOLUME. Ml 5-UCA CJL cicnr, V FINAL where: IV-Vmmi....-.? sample ti.rou.'li ii.'? dry gas meter t-k condition:*), ft1. V,,Volume of c.1.1 sample thrmich the dry jus meter im-'U'r conditions), ft4. T, ** A v* r.iL.-f' tcMpi'ratur* of stock gas, - .Vyit.un* dry gas meter temperature, *R. Pt>, * Barometric pressure at the orifice meter. In TTft. &//> Average pressure drop across the orifice meter, in HtO. 13.68peciUc gravity of mercury. tack pressure, static pressure, In Hg. INITIAL iicuio ctxucrto TOTAL YOU** COUECTCO f| M C0Vt*T AUGHT Of 1AT TO VOtfWf ir dfvfdicg titil Might U.CttASt sr ocvsitr Of tAitiL (t 1 'All: ricwog lie mi = VOUAX *ATOI at figure 104*7. Analytical data. FEDERAL REGISTER, VOL 38, NO. 68--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005616 SS50 RULES AND REGULATIONS PLANT._______________________________________ DATE RUN NO. - STACK DIAMETER, in.__________________ - BAROMETRIC PRESSURE, in. Hg:__ STATIC PRESSURE IN STACK |Pg), in. Hg. ----------------------------- 6.8 Isokinetic variation (comparison of velocity of gas m probe tip to stack velocity). . x 100V'~' a,,e eq. 104-8 wb*r: I Percent of Isokinetic sampling. VUui- Total volume ofgas sample (stack conditions)-, A*-Probe Op are^ ft*. -Saniplinp time, sec. H*iM,"Average stack gas velocity, feet per second. OPERATORS ___________________________ Traverse point number Velocity head, _,in. H20 v&7 SCHEMATIC OF STACK CROSS SECTION Stack Temperature <y-'F ___ _________ AVERAGE: Figure 104-8. Velocity traverse data. Figure 104-8 shows a sample recording sheet for velocity traverse data. Cse the aver ages in the last two columns of figure 104-8 to determine the average stack gas velocity from equation 104-5. 6.6 Beryllium collected.--Calculate the total weight of beryllium collected by using equation 1C4-6. here: Wt v ViCi-WC*--V-C..cq. 104-6 7.*..*:ghn t.; beryllium collected. l` -7>..U v* fr*':u of hydrochloric acid 4.3 2.4, ml. C ration of beryllium found In r.t: :le. V.--Total volume of water vised la sam- pl*;:g (unplnger contents plus all v.aah amounts ), ml, C*vr-tClar.k concentration of beryllium in vivr, *.,1:11. V*=Total volume of acetone used in sam pling (all wash amounts), ml. C*=B]ank concentration of beryllium In acetone, *ig/ml. 6.7 Total beryllium emissions.--Calculate the total amount of beryllium emitted from each stack per day by equation 104r-7. This equation Is applicable for continuous opera tions. For cyclic operations, use only the time per day each stack Is in operation. The total beryllium enusy.ons from a source will be the summation of results from all stacks. " lu1 p-'l? cq. 104-7 wlar*'1 /.' * of rn:ivitm. c,i w. b'(alot.il w,icliL of l-r>Ilium r*l!r*i.t**i. *4r. Vtvi*j-TTiU velum*' of pas sample otavk tvmJiliorLs), ft*. (r*)#.Avrac* jl.uk pas vivify, f*cc p-r second, .l.-itatk .\r*a, ft-'. 7. Evaluation of resutts--l.l Determina tion of compliance.--'7J.l Each performance test shall consist of three repetitions of the applicable* test method. For the purpoee of determining compliance with an applicable nations!- emission standard, the average of results of all repetitions shall apply. 7.2 Acceptable isokinetic results.--7.2.1 The following range sets the limit on. accept able isokinetic sampling results: U 90 percent percent, the results* are acceptable; otherwise, reject the test and ' repeat, 7. References.--1. Addendum to Specifica tions for Incinerator Testing at Federal Facil ities, PHS, NCAPC. December 6, 1967. 2. Amos. M. D- and Willis, J. B.. "Use of High-Temperature Fre-Mixed Flames- la Atomic Absorption Spectroscopy/* Spectrochtm. Acta, 22: 1325,1966. 3. Determining Dust Concentration in a Gas Stream. ASMET Performance- Test Code No. 27, New York, N.Y,, 1957, . Devorfcttt, Howard et &l,, Air Pollution Source Testing Manual, Air Pollution Control District, Los Angeles, Calif. November 1963. 5. Fleet, B., Liberty. R. V,, and West. T. a.. "A Study of Some Matrix Effects In the Deter mination of Beryllium by Atomic-Absorption. Spectroscopy in the Nitrous Oxide-Acetylene Flame,** Toianta, 17: 203,1970.' 6. Marie, L. S., Mechanical Engineers' Handbook, McGraw-Hill Book Co., Inc., New York, N.Y,, 1951. 7. Martin, Robert M., Construction Details of Isokinetic Source Sampling Equipment. Environmental Protection Agency, APTD056*. 6. Methods for Determination of Telocity, Volume, Dust and Mist Content of Gases. Western Precipitation Division of Joy Manu facturing Co., Los Angeles, Calif. Bulletin WP-50. 1968. 9. Perkin Elmer Standard Conditions (Hey. March 1971). 10. Ferry, J, H., Chemical Engineers' Hand book, McGraw-Hill Book Co.. Inc., New York. N.Y., 1960. 11. Rem, Jerome J., Maintenance. Calibra tion, and Operation of Isokinetic Source Sampling Equipment, Environmental Pro tection Agency, APTD-0576. 12. Shlgehara, B. T*. W. F. Todd, and W. S. Smith, Significance of Errors in Stack Sam pling Measurements, Paper presented at the annual meeting of the Air Pollution Control Association, St. Louis, Mo.. June 14-19, 1970. 13. Smith, W. S. et a}.. Stack Gas Sam pling Improved and Simplified with New Equipment, APCA Paper No. 67-119, 2967. 14. Smith, W. S., R. T. Shlgehara, and W. F. Todd. A Method of Interpreting Stack Sampling Data. Paper presented at the 63d annual meeting of the Air Pollution Control Association. St. Lotus. Mo.. June I4~i0. 1970. 15. Specifications for Z.ic:nera*or Testing at Federal Facilities, PHS, NCAPC, 1567. 16. Standard Method for Sampling Stacks for Particulate Matter, In: 1971 Book of ASTM standards. Part 23. Philadelphia. 1971, ASTM Designation D-2923-71. 17. Vennard, J. K. Elementary Fluid Me chanics. John. Wiley and Sous, Inc., New York, 1947. (FB Doc. 73-6423 Filed 4-5-73:3:45 am] FEDERAL REGISTER. VOL 30, NO. 66--FRIDAY, APRIL 6, 1973 CAPCO JEN 0005617