Document BRkpQ0eGxMvjpw0KL5DQVdN1j

If PB-253 110 AN APPRAISAL OF TESTS AND STANDARDS FOR THE EVALUATION OF ELECTRICAL INSULATING FLUIDS National Bureau of Standards PREPARED FOR Energy Research and Development Administration 14 May 1976 HONS 204750 f KEEP UP TO DATE :\ Between the time you ordered this report-- which is only one of the hundreds of thou sands in the NTIS information collection avail able to you--and the time you are reading this message, several new reports relevant to your interests probably have entered the col lection Subscribe to the Weekly Government Abstracts series that will bring you sum maries of new reports as soon as they are received by NTtS from the originators of the research. The WGA's are an NTIS weekly newsletter service covering the most recent research findings in 25 areas of industrial, technological, and sociological Interest-- invaluable information for executives and professionals who must keep up to date. The executive and professional informa tion service provided by NTIS in the Weekly Qovemment Abstracts newsletters will give you thorough and comprehensive coverage of government-conducted or sponsored re search activities. And you'll get this impor tant information within two weeks of the time it's released by originating agencies. WGA newsletters are computer produced and electronically photoeompoeed to slash the time gap between the release of a report and Its availability. You can learn about technical innovations immediately--and use them In the most meaningful and productive ways possible for your organization. Pfease request NTIS-PR-206/PCW for more infor mation. The weekly newsletter series will keep you current But /earn whet you have mieeed in the peet by ordering a computer NTISearch of all the research reports in your area of Interest dating as far beck as 1964, if you wish. Pisses requsst NTIS-PR-iae/PCN for more inlormstion. WRITE; Managing Editor 5266 Port Royal Road Springfield, VA 22161 - Keep Up To Date With SRIM SRIM (Selected Research in Microfiche) mlcroflched report Your SRIM service begins provides you with regular, automatic distri as soon as your order Is received and proc bution of the complete texts of NTIS research essed and you will receive biweekly ship reports only in the subject areas you select. ments thereafter, if you wish, your service SRIM covers almost all Government re will be backdated to furnish you microfiche search reports by subject area and/or the of reports issued earlier. originating Federal or local govemmant agency. You may subscribe by any category Because of contractual arrangements with or subcategory of our WGA (Weekly Govern several Special Technology Groups, not all ment Abstracts) or Qovemment Reports NTIS reports are distributed in the SRIM Announcements and Mas categories, or to program. You will receive a notice in your the reports issued by a particular agency microfiche shipments Identifying the excep such as the Department of Defense, Federal tionally priced reports not available through Energy Administration, or Environmental SRIM. Protection Agency. Other options that will give you greater selectivity are available on A deposit account with NTIS is required 1 request before this service can be Initiated. If you have specific questions concerning this serv The cost of SRIM service is only 45# ice, please call (703) 461-1566, or writs NTIS, domestic (60# foreign) for each complete attention SRIM Product Manager. This information product distributed by ||||C m^^ U.S. DEPARTMENT OF COMMERCE National Technical information Service 5295 Port Royal Road Springfield, Virginia 22161 MONS 20A751 NBSIR 76-1054 PB 253 110 An Appraisal of Tests and Standards for the Evaluation of Electrical Insulating Fluids Electricity Division institute for Bask Standards National Bureau of Standards Washington, D. C. 20234 May 14, 1976 Final Report Pr ireaea^eMrwed for Energy Re--arch and Development Administration Electric Energy Systems Division_____________ Contract No. 349-18) - 2062 | NATIONAL TECHNICAL ; MOHMATION SOtVICS HONS 20^752 Mfli.t U* Ifltv T Jjt 1)1 0( OFCOM aisuocitAphie data SMfET 4. TH 1 1 'I'llTHI.I l-l 111 u A1ION OK HI MOH| NO. ., HBSIR 76-1051* 2. Go.'l Am ,hi,.ii An Appraisal of Tests and Standards for the Evaluation of Electrical Insulating Fluids 1 Al 1 IIOKIM t. PtKHiRNlNl. ORGANIZATION NaMF AND ADDRESS National bureau op jtahoardj department op commerce WASHINGTON. O.C. 30214 12. Spon*ufin Orjuniz.nun N*ir and Cooplti* Addrra, (IlrHl, City, Stata, ZtP) Energy Research and Developmant Administration 20 Massachusetts Avenue Washington, O.C. 205^5 IS. SUPPLEMENTARY NOTES 'PB253TI0 S* I'abl it-Date May 1A, 1976 * Performing Otgitnixaiiun l Hertorming Organ. Re port No to. TfBSiI/HT7k6/-1a0.i?54U,.. Sc 2110506 11. Caninet/Grant No 13. Tjrp* si Report M Prnod Covered Final 14 Sponsoring Agency Code. '* ABSTRACT (A ]0O>m< or iui ficauf nHM7 at tnotl tl0Ullttnl Hifafanfian. /Moouioil tneludti a ilfilflenl ' RiAflogropfcr or lUonlwi aurrar, Mnlioi II hara.) Based on a literature study and interviews vith representatives of the electric utility industry, manufacturers of insulating fluids sad electrical apparatus, government regulatory agencies, organisations preparing standards and codes, trade associations and independent testing laboratories, the status of existing standards and test procedures for insulating fluids is reviewed. Askarel-type transformers and capacitors are described and the characteristics of several currently-used as veil as nev candidate insulating fluids are given. The possible impact of codes and government regulations on the introduction of nev fluids into use is reviewed. Reeds for nev or revised test procedures snd standards are noted and recommendations made for research and development efforts as veil as administrative actions to facilitate the qualifying of nev insulating fluids as acceptable replacement! for the askarels currently in use in a certain class of transformers and capacitors. 17, KEY WORDS (Mi fa twalra mi trim*; mtjtiatatieal Old**, capitaItaa an ty die tint fatter at * lltat kay *ord untamm a prapt' naira.- *paralad by tmmKalomm ) Asfcarel; capacitors; insulating fluids; liquid insulators; polychlorinated biphenyls; |i0S 9MKT TO CHANGE Is. A\AILADILITY IX Unliauiid C Kar Official Distribution, Do Not ftelcaae to NTlS Q Order F/om Sup. of Dot,t U.S. Govinom Printing Oftlco ' fuhcnitM. OX. 20*102. SQ C, Na ClI 5p~" Order Front National Technical Information Se/eice (NTIS) Springfield, Virginia 22151 * | It. SECURITY CLASS (THIS REPORT) mwftrtP PACES UNCLASSIFIED 30. SECURITY CLASS (THIS PAGE) UNCLASSIFIED t I uicowmfcoc ** MONS 204753 N8SIR 76-1064 AN APPRAISAL OF TESTS AND STANDARDS FOR THE EVALUATION OF ELECTRICAL INSULATING FLUIDS Electricity Division Institute for Basic Standards National Bureau of Standards Washington, D. C. 20234 May 14, 1976 Final Report Prepared for Energy Research and Development Administration Electric Energy Systems Division Contract No. E(49-18) - 2062 U.b. DEPARTMENT OF COMMERCE. Ettot L Richardson. Iiowarr James A. Bator. III. UwdSr Seorecary Dr. Betsy Anotor Johnson. 4MManf fawcev Nr Mmm md TtotaeNr? NATIONAL SUftfAU OF CTANOAIIOa. Imaet AmWsr. AmNf OSeeSar HONS 204754 AM APPRAISAL OF TESTS AND STANDARDS FOR THE EVALUATION OF ELECTRICAL INSULATING FLUIDS David B. Miller, Principal Investigator, Purdue University, Consultant to Electricity Division, NB8; Vincent 3. Bover, F. Ralph Kotter and Oskars Petersons, Electricity Division, MBS and Merritt M. Blrky, Clayton M. Huggett and AndreJ Macek, Fire Science Division, NB8 it KONS 204755 Table of Content* 1. Summary 1.1 Introduction 1.2 Appraisal and Recommendations 1.3 Acknowledgments 2, Status of Test Techniques and Standards for New Insulating Fluids -- A Sumary 2.1 "Askarel-class" Transformers 2.2 "Askarel-claea" Capacitors 2.3 Available Insulating Fluids 2.3.1 Fluid Characteristics 2.3.2 Test Techniques and Standards 2.3.3 Flammability 2.3.4 Toxicity 2.3.5 Degradability 2.4 The Scope and Influence of GovernmentRegulations 2.5 The Scope and Influence of Insuranceand Fire Code Regulations 3< Appraisal and Recommendations 3.1 Electrical Tests on Fluids 3.2 Flammability Tests for New Fluids 3.2.1 Flash and Fire Points 3.2.2 Oxygen Index 3.2.3 Autoignition Tests 3.2.4 Spray Flammability Tests 3.2.5 Arcing Tsste 3.3 Toxicity Tests on Fluids 3.3.1 Appraisal 3.3.2 RecoBtendatlons 3.4 Recommendation on Degradability andBiodegradability Testa 3.5 Insulation System Tests and Standards --Transformers 3.5.1 Insulation System Tests 3.5.2 Standards -- Transformers P*e 1 ^ g g g g 10 ^.o ^ ^ 12 13 14 15 17 17 19 20 21 21 21 22 22 22 24 25 26 26 28 MOWS E04756 Section 1 Summary 1,1 Introduction Askarels, the principal constituents of which are polychlorinated biphenyls (PCBs), have been used for over forty years aa insulating fluids in power capacitor and flame proof transformer applications. However, recently they have come under close scrutiny and criticism because of suspected toxicity and long persistence in the environment. Large scale use of PCBs has now been restricted to electrical equipment applications under stringent controls to prevent accidental spills. As a result of these concerns and the possibility of further restrictions being placed on the use of PCBs, alternative fluids are being considered; however, so far none has gained general acceptance. Among the factors impeding the acceptance of new fluids are the cost considerations and the excellent electrical properties of PCBs. Besides these major factors, it was felt that lack of adequate test procedures and test-related performance standards for the evaluation of new fluids did present an impediment to their acceptance. For example, the question was asked: Are the existing flassaablllty performance requirements and the associated teats realistic in terms of fire risks? Of equal concern were tests and requirements regarding toxicity and degradability of the material In the environment. Could one arrive at concisely defined yet meaningful requirements for these properties coupled with a comprehensive set of tests and measurements? - In the early part of 197j the National Bureau of Standards was approached by the Energy Research and Development Administration concerning standards for testing nev nonflammable insulating fluids for transformers and capacitors. The objective was to detail in a report or position paper the test methods and standards and the test-related research and development required to facilitate the evaluation and qualifications of new insulating fluids for distribution transformers and power capacitors. In the study leading to this report the authors attempted to find answers to the type of questions mentioned above and to Identify test-related roadblocks to the acceptance of nev insulating fluids. The material for this report was gathered in literature reviews and interviews with representatives of the electric utility industry, manufacturers of insulating fluids and electrical apparatus. Government regulatory agencies, organizations preparing standards and codes, trade associations, and Independent testing laboratories. The resources and the time frame of the project were such that not every teat for the properties of insulating fluids could be examined in detail--American Society for Testing and Materials (AflTO) standards (see Appendix A.3.2) list about ^0 properties--nor do we believe that such an examination would have been desirable and cost effective. Instead we took the position of responding to problem areas as brought to our attention by numerous personal contacts. The status portions. Section 2 and the Appendix, review the existing test procedures and standards for insulating fluids. They also describe the askarel-type transformers and capacitors and stress the Interdependence between the insulating fluid and the rest of the device. 1 HONS 20*757 To illustrate the properties of insulating fluids and trends in the fluid development, several existing and proposed insulating fluids are described. Finally, since a new material introduced in commerce must meet the requirements of numerous cades and Government regulations, an outline of these is provided at the end of Section 2, with details in the Appendix. The appraisal and recommendations are the authors' considered opinions. In Section 3 we comment on the entire field of tests and stsuidards without firm suggestions for a division of the recoomended future work among the industrial, trade organization, or public sectors. In 1.2, we emphasize those areas which would benefit from public support. Admittedly, the public and industrial responsibilities cannot be always clearly delineated. During this investigation we were frequently close to the PCB controversy. However, our scope and objectives did not include the evaluation of merits and deficiencies of PCBs; ve do not pass judgment in this area. Similarly, although several potential replacement fluids are mentioned in the report, we offer no firm conclusions regarding the availability and suitability of replacements. The identification of commercial materials and their sources in this report is made for illustrative purposes only. In no case does this identification imply recommendation by the National Bureau of Standards, nor does it imply that the materials are the best available. . 1.2 Appraisal and Recosmendations In this part of the report we sunnarize our most Important conclusions, our appraisal, and our reconaendatlons for future action. Not all our findings are abstracted here; for a complete appraisal the reader Is referred to Section 3 of the report. The flavor of some of the complex aspects of the problem, such as toxicity and Government regulations, can only be obtained by reading the pertinent parts of the entire report Including the appendix. The summarized recomendations here are mostly those which we feel warrant the consideration of public support they are emphasized for the convenience of the sponsor of this study. We are now convinced that the problem of impedimenta to the acceptance of new fluids is, with only minor exceptions, not associated with formalized test procedures as prescribed by standards documents and the associated measurement technology. The U.S. system of voluntary standards is sufficiently flexible to provide for ad hoc exceptions if such exceptions are warranted. On the other hand, compulsory codes and regulations based on the characteristics of specific existing products rather than on required performance present serious handicaps to progress. While simple laboratory tests on the insulating fluids themselves can be performed using standard procedures, these alone do not suffice for evaluation of the fluid. Essential tests on models and prototypes in advanced stages of development are not readily standardized. HONS E04753 2 The latter, however, are essential since ultimately the fluid *ust be qualified in the envirotent imposed by the equipment. Adverse inter actions among various opponents of the equipment, rather than an isolated fluid property, say be the real barrier to the acceptance of a fluid. The problems with such system tests are threefold; (1) the test technology and procedures may not be available; (2) the tests may have to be tailored to individual ineulation systems; and (3) the tests are very costly. Finally in certain areas such as toxicity, uniform test protocols may not be desirable because of the complexity of the system and required responses. One must accept a lengthy set of experiments in the process of evaluation. Because of the above concerns, we did not restrict the Investigation to standardised tests and the related measurements, but considered a broader spectrum of experimental procedures snd regulatory constraints aasociated with the Introduction of new fluids. At its conclusion vt see this study as having confirmed the existence of a need for the development of additional tests and standards. While for meeting seme of these needs the course of action appears clear, for others it does not. ,, Within the scope of our investigation the most signifiesnt roadblocks appear to be in the fleaaubility arsa. The problem lies with the building codes and firs regulations and the lack of a simple procedure for evaluating the risk in actual fire eituations. Current codee end regulatione dealing with the flammability specifications for insulating fluids refer only to specific chemical fsmilits--mineral oils and aekarels and thus tend to exclude new developments. Furthermore, it appears that in some cases flaamability performance evaluations are being made with greater consideration given to the properties of the liquids currently in use then to established criteria of operating sad environmental conditions snd the fire risks involved. It is highly probable that replacements, dictated by environmental considerations, will have poorer fire resistance properties than the aekarels. The problem is to detsrmine how much relaxation of the flanability requirements may be possible without Introducing undue risk. It iM Kttowmtndtd that 'ii&MAch be. conducted -in <ut attempt to develop {to--tiitffy pcAjoksunce cfiiti/ua joe new frbudA. This work should consider the analysis of past accidents, the environment In vblch the equipment operates and the failure modes. This research activity must have a strong input from the insurance industry, Underwriters Laboratories, and peraonnel from the National Electrical Code Committee. The results, if to be of benefit, must be usable In preparation of code revisions. HONS 204759 3 The laboratory tests for initial screening of materials--flash and fire point and autoignition testa--appear generally satisfactoryHowever, they are tests of relative flannability sad not true indicators of fire safety involving large scale equipment failures. In the past, the definitive tests for fire safety have been the arc tests involving actual prototypes. This is not a satisfactory state for the introduction of new fluids since the entire fluid and equipment development procedure must be carried out before the test can be conducted. Besides the development costs, the test itself is very expensive. It would be highly desirable to arrive at a realistic fire-safety teat procedure using small models of the insulation systems, not necessarily scaled down actual devices. Since actual failures follow electrical arcs and energy releases, arc tests appear promising. They are little explored, but whatever little has been done in this area has attracted vide interest. Thai ouA, ncond Azcommtndation in thi onto, oi itaimubitity tuting -u ioA. iht development oi a tut which Mould be a AzatUtic. indicator oi iiAZ hazard. The test involving an arc in a simulated insulation system appears promising. Since any new test has to be implemented in practice through code revisions, interaction with the previously mentioned groups is essential. Questions have been raised about the lack of a quantifiable definition of flammability; indeed the term is largely used in s qualitative and relative sense. Ve believe any improved definition of flssassblllty of insulating materials, although still relative,, must follow the development of an Improved test procedure. Whether or not the insulating fluid passes the test will qualify it as nonflammable or flaauble. Potential toxicity is of great concern in the introduction of new fluids. The underlying characteristic of this area revealed in our study is its extreme complexity. Relatively short testa with well definsd measurements such as those oaployed to determine physical properties do not apply here. Instead, lengthy and numerous experiments must be performed on animals. As examples of the complexity of the subject, a reference work, cited in this report, mentions 27 differsat organs in which responses can be specifically identified; 22 different routes of administration; and finally 23 levels of animal prefsrence. There are several definitions of toxicity but relatively few standard experimental procedures for determining the toxicity of a new chemical. The attempt to develop a standard set of protocols for routine assessment of the toxicological haxard of a new insulating fluid or any chemical hv several dangers that mist be recognised: (l) it may lead to unnecessary tests for certain substances; (2) it may fall to aak questions which are of overriding significance; (3) in view of the multiplicity of possible biological systems. It may be wasteful; (k) in view of the multiplicity of responses, sobs of which may require a long time before the effect can be observed, such testing cannot guarantee the complete safety of a new chemical. k HONS 204760 Tvio type* oi tati, Amti tut <wd cMonic itudiu, <vu. iuqyuttd tkU xzpo-tt jo* Aoatinz laxty icAZUiing oi mtexiati, We believe the early toxicological screening belongs with the development of the insulating fluid. The fluid manufacturers are either aware of or are studying toxicity of their materials. As the development of the fluid approaches the time when there is reasonable probability of actual use, it would he desirable to conduct impartial, critical and in-depth * appraisals of the manufacturer's toxicity data. Cl/e Atcosmtend that uidzpzndejit appAOUali oi thz toxlzAty data, be ipcnioAzd. We favor them since such appraisals will minimize the chances of discovering adverse effects years after the insulating fluid has been established in commercial use. The degradability of insulating materials in the environment is also a complicated hut even a less explored subject than toxicity. The process by which a substance degrades to naturally occurring compounds may be biological, chemical, or physical. The subjects of toxicity and degradability are often interdependent; for exsaple, rapid degradability implies that the substance is highly reactive and therefore perhaps toxic. In dealing vith the question of degradability, the following aspects are important: the rate of degradation, toxicity of the degradation products, snd accumulation of ths original substance and intermediate degradation products in living organisms. - There are no standard degradability tests for the electrical insulating fluids, although individual manufacturers have performed experiments on thsir products. One industry group which has been successful in developing a standardised test for biodegrsdability, including a definition of what is meant by biodegradable, is ths Soap and Detergent Association. Thsir success has lead us to believe that an attempt to develop a similar procedure for insulating fluids vould be a worthy endeavor. Thai, tot AZCOtmtnd KtAtaKch to ZXptOKt thz dzvztopmznt oi itandcuuiiztd dzgMuJability zvatuation pAoczduAti. In the electrical and other areas we see no major deficiencies in the laboratory type of tests on insulating fluids. There are several minor problems mentioned in Section 3* These concern the types of cells used for breakdown tests; possibly inadequate tests for moisture content of new fluids; specifications snd slgaificence of corona Inception snd extinction levels. We feel the situation is different for the tests in the more advanced stages of equipment development during which insulation models and proto types are evaluated. Of concern is the stability of the Insulation system, and how it is affected by interactions among various components. Such interactions may degrade the insulating properties of the fluid snd reduce the life expectancy of ths apparatus. An Important cause of failure is corona, but there are other factors contributing to Insulation deterioration. It could be argued that tests at this stage are ths sols responsibility of the developer of a particular apparatus. W# believe, however, that guidelines to uniform system testing vould accelerate ths introduction of new fluids. Having well developed snd verified guidelines vould 5 MQNS 204761 reduce the cost of insulation system evaluation and would facilitate the comparison of results of competitive systems, and the availability of carefully-performed and well-documented results of a validated test will mitigate user reluctance in accepting a new product. Tests in accordance with such guidelines would he expected to uncover possible deficiencies which could cause expensive, unexpected failures at a future date and possible withdrawal of the product. CofUCOUCii&y, Mt tecormend xaiaAch leading to the. development orf journal guideline* joe tAaniioKmen and capaccton. imutation by* til* tiAtb. The long-term performance of an insulation aystea is often determined through accelerated aging procedure*. To arrive at a valid accelerated test, the basic mechanisms of failures must be understood. A study in the basics of insulation failures, as they affect tests, would be desirable. Finally, we consider the effect of published voluntary performance standards, compulsory building codes, and Government agency regulations. These do slow down the process of introduction of new products, if for no other reason than because of multiplicity of requirements which must he considered. Sometimes these requirements are ambiguous and conflicting. The impediment is very severe if the codes and regulation! recognise only certain specific products. The process of preparing voluntary standard* could be accelerated by encouraging greater uaer and third party participation in comlttees. This objective could be partly achieved through allocation of a special fund for these purposes. - A close liaison should b maintained between the Energy Research and Development Administration and Federal and State regulatory agencies through flow of information in both directions, participation in meetings, and assis tance in project reviews and steering. Wt Mzoimtnd inCJUObtd inttAdtpaxtmyjUat rooJuLLruUian oj tegutatony aatiuitif. A teak force consisting of members of concerned agencies charged with a minimum duty of keeping each other Informed would alleviate problems in this area. If the task force could then become the focal point for coordinating the specific actions of the agencies which impinge on the electrical powmr industry, and thus become a center to which the industry could look for guidance, it wuld be of even greater value. 1.3 Acknowledgment s This report ia based on information gathered from a large maber of sources, and in particular from meetings with many individuals and groups who are involved in one way or another with insulating fluids. We wish to express our appreciation to these groups, listed below, who very willingly supplied their time, information, and suggestions in order to best meet the stated objectives of this study: Government Regulatory Agencies Environmental Protection Agency (Water and Hazardous Materials Group, Office of Toxic Substance#, Environmental Monitoring Laboratory, Environmental Research Laboratory) Department of Health, Education and 6 MONS 204762 Welfare (Food and Drug Administration, National Institute for Occupational Safety and Health, Public Health Service); Department of Labor (Occupational Safety and Health Administration); Department of Transportation (Federal Railroad Administration, Urban Maas Transit Authority); Department of Natural Resources, State of Michigan; and Department of Natural Resources, State of Wisconsin. Manufacturers Dov Chemical U.S.A.; Dov Corning Corporation; Electric Utilities Company; Exxon Research and Development Company, Corporate Research Laboratories; General Electric Company (Transformer Laboratory Operation, Capacitor Products Dept., Research and Development Center, Environmental Protection Operation); McGrav-Edison Company (Pover Systems Division); Monsanto Industrial Chemicals Company (Specialty and Process Chemicals Division); RXE Corporation; and Westinghouse Electric Corporation (Transformers Divisions, Research and Development Center). Electric Utilities Commonvealth Edison Company; Consolidated Edison Company; Kansas City Pover and Light Company; Potomac Electric Pover Company. Standards Organisations. Trade Associations and Testing Laboratories American Society for Testing and Materials; Edison Electric Institute; Electronic Industries Association; Factory Mutual Engineering and Research; International Association of Electrical Inspectors; Manufacturing Chemists Association; National Electric Manufacturers Association; and Underwriters Laboratories, Inc. The authors thank J. Z. Cunningham for typing the manuscript, including numerous drafts, under tight deadline conditions. We also acknowledge vith thanks the contributions of M. G. Broadhurst and C. G. Malmberg of the Polymers Division, NBS; F. L. Hermach and H. S. Turgel of tha Electricity Division, NBS; W. H. Kirchhoff of the Office of Air and Water Measurement, NBS; and G. Carlson, Department of Toxicology, Purdue University, in critically reviewing sections of the report. 7 HONS 204763 Section 2 Statu* of Test Technique* and Standard* for lev Insulating Fluid* -- A Summary "Askarel" 1* a generic term for a class of fire-resistant, synthetic, chlorinated hydrocarbon fluids used by the electric Industry for insulation in capacitors and a certain class of transformers. Their "fire resistance," good heat-transfer and non sludging characteristics and their high dielectric constants make these fluids particularly attractive for these applications. The industry has grown accustomed to the particular features of the aakarels over the UO-odd years they have been available and many standards and operating guidelines have been developed in the light of their specific characteristics. It is the purpose of Section 2 to review in a summary form the. status of existing standards and test procedures for Insulating fluids as they relate to the nev fluids being proposed as replacements for the askarels (the Appendix contains detailed background material). In reviewing the available procedures, the reader should keep in mind not only the applicability of given procedures to nev fluids, but also the adequacy in the face of today's much more complex and stringent requirements on electrical materials and apparatus. Reliability, fireresistance, nontoxicity and minimum environmental impact a--Kr( in the way of materials and apparatus testing than was required when askarels' were Introduced. It is an appropriate time, therefore, that the adequacy of insulating fluids test techniques and standards be reviewed. In this section, background on the requirements of aakarel-insulated transformers and capacitors is surveyed in order to identify the operational requirements which new fluids oust fulfill within the apparatus environment. Fluids characteristics themselves are then Investigated. Finally, existing constraints due to governmental regulation and fire codes are explored to reveal possible additions and changes which might be required in this area to accommodate new fluids. 2.1 "Askarel-class" Transformers Specifications for transformer aakarels (proprietary blends of polychlorinated biphenyls (PCBs) and other ingredients) are given in ASTM Standard D 2283-73- The fluid is enclosed in a metal tank, and the transformer windings are completely immersed in the fluid. Askarels are used as the inaulants in transformers where the fire resistant characteristic is particularly important. These uses Include utility and industrial distribution service, where the transformer serves a heavily concentrated load in an urban, commercial or industrial area and where the transformer must be in close proximity to populated buildings. Electric railroads snd many electrostatic precipitators also use fireresistant transformers. 8 MONS 204764 The fire resistant quality of the askarels ia due, in part, to the nonflunable gas, mostly hydrogen chloride, which it produced when the fluid la aubjected to arcing or high temperature. This gaa is also corrosive, and scavengers must be used to absorb small araunts of hydrogen chloride produced in normal operation. Transformer rooms or vaults must be adequately vented to prevent build up of HC1 vapor, and any leading or spilled aakarel must be contained because of its toxic and environmental problems. The transformer fluid could be spilled during manufacture, transport, processing, filling, normal operation, service failure, repair or disposal of the transformer. Guidelines have been issued for proper handling (ANSI1 ClOT.l-lSTU; IEEE2 76-1971*). Spilling during service most often results from an internal arc causing the transformer tank, to rupture. The arc generally starts as a small discharge, and pressure builds up slowly; a pressure relief device, if present, prevents failure from this type of malfunction. Small discharges, however, may grow rapidly to high current arcs. The resulting pressure vave then builds too fast for a pressure relief device to operate. Although a current-limiting fuse can be used to reduce the magnitude of the pressure wave, the tank, must he made exceptionally rugged to withstand this condition. Because of Intensive preventive programs, users report very low (0.01 to 0.02$) yearly tank rupturing failures. In spite of the low failure rate, indoor mineral-oil Insulated transformers are Installed in fire-proof vaults in accordance with national Electrical Code recommendations. Vault requirements apply to all mineral-oil insulated transformers and to aaksrel-insulsted transformers rated over 35,000 volts. . The fire-resistance of a transformer fluid must ultimately he determined by testing a complete transformer. The askarels were originally qualified by such tests conducted by Underwriters Laboratory. More recently, HTE Corporation has run a set of demonstration tests comparing mineral oil, askarel, silicone fluid, and an cTE proprietary material "RTemp." In this test, the three newer materials were shown to be less fluBBmble than mineral oil; this test could form the foundation for a more rigorously controlled and Instrumented standard test of transformer insulating fluids. When considering alternatives to mineral oils or askarels in transformers, in addition to the obvious electrical and thermal properties the following special characteristics will have to he considered before a nev fluid can be considered to qualify*, material compatibility, ability to impregnate the solid insulation used, hydrolytic stability, fire safety, toxicity and environmental Impact. Standards for axternal protection of transformers with nev fluids should also be considered; this includss fusing, circuit breakers, and vault requirements. A large number of standard* exist which guide the selection and <ng of mineral oil and askarel in transformers. These are listed in A.1.5 presumably similar standards for new fluids would eventually be required. ^American Rational Standards Institute institute of Electrical and Electronics Engineers 9 MONS 204765 2.2 "Aakarel-clasa" Capacitors A large percentage of liquid-filled electrical capacltora uae askareis as the dielectric fluids, primarily because of their relatively high dielectric constants and the fact that they are approximately equal to the dielectric constant of readily available paper Insulation. Their good electrical insulating and stability qualities also Bake than attractive for capacitors. Although not a prime requirement for large power capacitors, the fireresistant feature of askareis is also a significant advantage. The fire resistance appears to be important in certain types of snaller f"Insistrial"? capacitors, such as those used in conjunction with fluorescent lights and appliance motors. There is an interdependence of the transformer and capacitor askarel manufacture which should not be overlooked in considering the economics of substitutes. Capacitors generally fail by an Internal arc destroying the electrical insulation. The arc will also often rupture the capacitor tank. Capacitor reliability Is generally quite good, with yearly failure rates being less than one-percent. In addition to the special characteristics noted as Important In the case of candidate replacement fluids for the askareis used In transformers, the fluids for use In capacitors must possess high dielectric constants. The esse with which the solid insulation can he fully impregnated with the liquid Is even more important in the capacitor application. Existing standards for capacitors are described In A.2.5. These deal with the fluids themselves (specifications, acceptance, handling, disposal) and with the requirements of the total capacitor. _ 2.3 Available Insulating fluids 2.3.1 fluid Characteristics We have described several available new fluids in A,3-1 of the Appendix, for the sake of completeness, descriptions of mineral oils and askareis are also included. A listing of these substances including their principal properties follow. The askareis are mixture* of polychloroblphenyls and chlorobenzenes with additives. These liquids have excellent electrical properties and are highly stable,but evidence of their toxicity has developed. Their excellent stability has given them high reliability on the one hand, but also given them the disadvantage of poor degradability in environment. Silicone liquids (mixtures of dimethyl slloxane polymers) have good electrical properties (although the dielectric constant Is not as high as In the case of the askareis). They are useful over a wide temperature range, but axe less fire resistant than the askarela and form sollde during electrical discharge. They are high molecular weight liquids which appear to be relatively nontoxic. Mineral oils are mixtures of paraffinic (or naphenic) and aromatic compounds of petroleum origin with aromatic additives to prevent oxidation and 10 HONS 204766 sludging. The electrics!, properties are satisfactory sad the cost is lav. These materials are flammable (see table p. 72). Their toxicity Is low and results largely from the toxicity of additives. However, the additives are satisfactorily degraaable. Hydrogenated paraffin oils exhibit characteristic* similar to those of mineral oils except for higher flash points and consequently lover flamnabllity. Mixtures containing diaryl sulfones. particularly formulated for capacitors, possess high dielectric constants. The liquids are somewhat flammable (see table p. 72). Their toxicity and electrical properties are now being thoroughly investigated. They appear to be satisfactorily biodegradable. The phthalate esters are readily available in large quantities since they are now used extensively in the plastics industry. Their high dielectric constants (about 3) and othsr electrical properties make them more attractive for capacitor than for transformer applications. They exhibit low toxicity and satisfactory degradability although their flamabllity characteristics (see table p. 72) represent some disadvantage. Butylated monochlorodlphenyl oxides have good slectrical properties (dielectric constant about 5), and~are~being considered for capacitor ,, use. They are flanaable (see tabls p. 72). The slight toxicity resulting from a minor component has been investigated. The fact that the offending component is biodegradable and that its degree of bioaccumulation is also low permits these to remain as candidate liquids. 2.3.2 Test Techniques and Standards The standards relating to testing of insulating fluids are summarized in A.3.2. These refer to procedures for specific fluids (askarels, silicone fluids, insulating oils) and to specific test techniques Independent of the particular fluid being tested. 2.3.3 Flexibility Fire and explosion are potential hasards in the operation of transformers and to a leaser degree capacitors and since one outstanding feature of the askarels is their lav flsxability, the flanabillty or "fire resistance" of potential replacement fluids must be evaluated. Flash and fire points are defined as the temperatures at vhich an ignition source vlll cause a transient end a continuing flame in the vapors of the fluid. These are the most vldely used measures of flammability. Closed sad open cup methods for measurement of these parameters are described end discussed more thoroughly In A.3.1*. HONS 204767 11 In the Section A.3-1* on flammability testing, several other tests are also described. The Oxygen Index (0.1.) Test is a frequently used technique in which the lowest combust ion- supporting percent of o^gen in an oxygen-nitrogen mixture is determined. Some limitations on the use of the 0.1. test are discussed. A heated atomized mist of the substance is sprayed into an open flame in the Spray Flammability Test; this is a very severe test and has not been applied to any extent in evaluating the flammability of insulating fluids. Another flamability test is the Auto Ignition (A.I.) Test, in which the temperature for spontaneous ignition is determined. This has been used for insulating fluids; the A.I. temperatures are generally quite high, veil above normal operating temperatures. The energy release during combustion of an insulating fluid is a very pertinent parameter since physical damage to the apparatus and its surroundings is the thing vhich one hopes to eliminate if combustion occurs. Energy release measurements have not been applied to insulating fluids, but In A. 3.** suggestions are made toward this end. A basic problsm related to the flamnability issue is ths general lack of a rigorous, quantised definition of "flanability.1* We have used this term, and also such terms as "fire resistant" without really having an operationally ussful meaning for ham. Standardized, interpretable tests such as described above will provide the data on vhich quantitatively meaningful definitions can be baaed. Tables in A.3.1 and A.3.1* aumarize existing flammability data for some of the available insulating fluids. 2.3.** Toxicity The toxicity of the askarels has recently became a significant issue, leading to severe restrictions on their present day use, and creating the intensified search for substitute fluids. The toxicity of each candidate replacement must therefore be an important parameter to evaluate before the liquid is put into widespread use in electrical equipment. Some reference is given to the toxicity of the various insulating fluids in A.3.1* In A.3.5, toxicity determination techniques are explored. Perhaps the most important conclusion is that the toxicity issue Is extremely complex. A variety of definitions of toxicity are presented in A.3.5, emphasising that a fluid cannot merely be labeled "toxic" but that its specific toxic actions must be identified. Questions raised include: At what dose level does any toxic response occur, and what is the lethal doaaT Is the chronic response to long-term, continuous exposure different from the acute response to a single large dose? A large variety of responses occur. Table C p. 88 indicates the variety of organs and systems in vhich responses can be specifically identified, 12 HONS 204768 including such things as skin, blood, nervous system sod liver. Likewise, the response csa depend on the route of administration, e.g., orally or by inhalation; 22 different routes ere identified in Table B p. 66 Finally, the toxicity of a given substance may differ from one species to another; 23 levels of specie preference are listed in Table A p. 85 . The importance of continually improving chemical analysis techniques must be recognised in any attempt to establish definitions or protocols for toxicity. Gas chromatography sad mass spectroscopy have, for instance, greatly advanced analysis, both qualitatively sad quantitatively, in recent years. The o{ Toxic. Chemical Sub&tancti, published by the National Institute of Occupational Safety and Health, is identified in A.3.5 as listing all chemicals which are documented as having shown any toxic effect in any species. The effect, dose, administration route, species and reference source are also given. A.3.5 also contains further governmental legislative and regulatory enactments which give operational definitions of toxicity. These include: The Federal Hazardous Substance Act, Department of Transportation Regulations, Occupational Safety end Health Administration Regulations, Food and Drug Administration Regulations, and Environmental Protection Agency Regulations. 2.3.5 Degradability Because of the highly stable nature of the polychlorohiphenyl molecule, the aakarels which accidently or upon disposal reach the environment persist for long times; they are not biologically degraded In the manner or at the rate of most organic compounds. This persistence has amplified the toxicity problem and has brought out the need to Investigate the degradability as well as the toxicity of replacement fluids. Degradation to more simple, harmless, environmentally compatible compounds may occur dus to natural physical processes or, as in the case of most organics, by biological processes. New compounds require testing for both types of processes. Silicones, for instance, do not biodegrade, but are shown to decompose to basic silica and water in the presence of solar (ultraviolet) radiation, air and water. Tests for biodegradability of potential pollutants include the Fiver-Die Away Test, continuous activated sludge tests and batch activated sludge teste. The SuhccMlttee on Biodegradation Test Methods of the Soap and Detergent Association have promulgated a standard sulfonate biodegradability test. The Biochemical Oxygen Demand (BOD) and the Oxygen Electrode tests are identified as useful biodegradation tests in the EPA Manual of Methods for Chemical Analysis. All of the above tests are described and referenced in A.3.6* 13 HONS 204769 2.U The Scope end Influence of Government Regulations In 2.3 above. Federal legislation and regulations were Identified as one source for definitions of toxicity. We return to these sources now to examine the authority and procedures by vhich governmental bodies vlll influence the acceptance of nev insulating fluids. Not* detailed examinations and referencing of these various lavs and regulations vlll be found In A. **. The Environmental Protection Agency (EPA) has the following authorities from the Federal Water Pollution Control Act; 1. Prepare a list of toxic pollutants; criteria for selection, standards for allowed levels. , 2. Assign oil and hazardous substance liability; regulations for reporting and clean-up of spills. 3. Establish pretreatment standards; standards for pretreatment of pollutants which could enter public treatment works. 1*. Establish effluent standards; standards for allowed levels of discharge of each identified pollutant. 5. Issue pollution discharge permits; EPA may issue permits limiting the amount of discharge of specific pollutants from identified sources into navigable waters. Various versions of a Toxic Substance Control Act are also being con sidered In Congress. If this Act is passed it is expected that SPA will be given direct control of the use, as well ee the discharge, of toxic substances. The Federal Food, Drug and Cosmetic Act. empowers the Food end Drug Administration (FDA) to maintain safe levels of food quality in the United States. Under this authority, the FDA can regulate the use of poisonous Ingredients in agricultural and food processing industries. In the specific case of the PCSe, the FDA has issued regulations where PCB-containlng materials can and cannot be used. Thus, nev insulating fluids may fall under FDA Jurisdiction if used when food contamination is possible. Currently, PCBs ere allowed In seeled electrical transformers and capacitors; presumably these regulations could be stiffened if and when replacement fluids come into generally accepted, commercial use. The Occupational Safety and Health Act directs the Occupational Safety end Health Administration (08HA) to assure working persons safe end healthful working conditions. Hew insulating fluids could fall within the Jurisdiction of this Act both during manufacture (of the fluid or the apparatus) end during use of the fluid-filled apparatus in a working area. The Secretary of Labor (through OSHA) is directed to issue standards for toxic substances and to inspect working pieces in order to enforce these standards. The Secretary of Health, Education and Welfare (through the national Institute of Occupational Safety and Health, NIOSH) is to carry out research and develop the criteria for establishing standards. lU HONS 204770 NIOSH regularly updates its "Registry of Toxic Effects of Chemical Substances" (see Ref. &5 of A.3). This list, for instance, specifically cites several "Aroclors" (a manufacturer's tradename for mixtures of PCBs), The list should be consulted for toxicity information on other candidate fluids. OSHA regulations have incorporated the whole of the Rational Electrical Code, making this code, and specifically its provisions on transformers and capacitor fluids, part of Federal lav. OSHA has also nov issued draft technical standards on chlorobiphenyls. Department of Transportation (DOT) regulations vould have an impact on nev insulating fluids in two vays: 1. Control of the insulating fluid in transformers of electric railroad cars; the availability of a sufficiently nonflammable fluid must be recognized vhen considering a restriction on askarels in this application. 2. Control of the transportation of fluids', the transportation of hazardous materials falls vlthin DOT's authority. Class 3 poisons are defined in U9CFB 173.3U3* Finally, ve mention that the Department of Health, Education and Welfare (HEW) has regulatory control of solid waste disposal through provisions of the Solid Waste Disposal Act. HEW is, for instance, to publish guidelines on solid waste recovery, collection, separation and disposal. Federally controlled disposal facilities are obligated to follow these guidelines. 2.5 The Scope and Influence of Insurance and Fire Code Regulations The National Electrical Code (NEC) is a guide to the safe installation of elactrical equipment. It has been adopted, as is, or with modifications, as a part of the legal building code by many state and local authorities. It has also been incorporated into the Federal Occupational Safsty and Haalth Regulations. The current edition of the NEC mentions only mineral oils and askarels for use as insulation fluids in transformers and capacitors. Independent testing laboratories such as Underwriters Laboratories (UL) and Factory Mutual Engineering end Research (TV) will carry out testing of apparatus and issue an approval, Indicating that the item meets NEC standards. UL has also published flamabllity ratings of materials. FM will issue sa "acceptance" rating. Indicating that a particular installation meets safsty standards. These ratings are important in aiding the insurance companies to decide whether a given installation should be Insured. 15 HONS 204771 Fire safety has always been a major concern associated vlth electrical apparatus, sad It will he particularly Important In equipment using new fluids in place of the ''fire-resistant" asharels. For building code and fire Insurance protection, therefore, new fluids oust he acceptable by fire code standards' Further details regarding Insurance and fire code regulations will be found in the Appendix (A.5) HONS 204772 16 Section 3 Appraisal and Recommendations 3.1 Electrical Tests on Fluids It appears that existing specifications, standards and test procedures are adequate for dealing with the mineral oils and asltarela currently in use. While It does not seem feasible to propose a single specification which could be applied to old as veil as new candidate fluids for all applications, it does seem quite feasible to use at lease most of the same electrical test procedures for all. So far as ve have been able to determine none of the current standard electrical test procedures called out for insulating fluids actually act to impede the acceptance of new fluids except perhaps Indirectly. An example of this Indirect impact Is the situation with respect to dielectric breakdown tests dealt with in two ABTM Standards: D 077 "Standard Method of Test for Dielectric Breakdown Voltage of Insulating Liquids Using Disk Electrodes" and D l8l6 "Standard Method of Test for Dielectric Breakdown of Insulating Oils of Petroleum Origin Using VD2 Electrodes." These standards differ in that D 877 calls for a test cell having a small volume and an electrode geometry which provides a near-uniform field, whereas D 1816 requires a substantially larger cell with electrodes which establish a uniformly decreasing field about a maximum field point. The liquid in this larger cell is maintained in gentle forced flow between the electrodes. The D l8l6 procedure provides a test which is more sensitive to the presence of moisture in the liquid being tested and Is less influenced by the contaminating by-products of previous breakdowns; however, it requires a significantly larger volume of the test liquid. Either method can be used for testing the candidate silicone fluids; because of breakdown by-products, however, the specified alternative D 877 procedure which calls for refilling the test cell with new fluid prior to each breakdown test must be followed in order to obtain consistent results. The circulation of the liquid avoids this problem if D l8l6 is used. Thus with either procedure the volume of liquid required for the test is several times that required in a mineral OH or askarel test and the higher cost of the silicone fluid then appears as an indirect deterrent to its use. It should be noted that poor repeat breakdown strength of a fluid alone does not necessarily maan poor repeat breakdown atrsngth of the composite system when the fluid is used as the lmpregnant of a porous solid. ASTM Standard D 2413 applisa to such composite systems. Another indirect deterrent ia the obvious requirement that the test fluid not be contaminated by chemical interaction with the materials used in construction of the test cells. The commercially available cells which are inert to the mineral oils and askarela (and not e-ii are inert to both) are not inert to all new candidate liquid*. Still another indirect deterrent may be the fact that current standard test procedures for moisture content are not as satisfactory for silicone fluids or perhaps for other liquids as for mineral oils and aakarels. These can only be considered as temporary impediments to the adoption of new fluids since prospective widespread use of any new fluid would toon lead to the development of test procedures suitably tailored to its peculiarities. m0NS 204773 17 A characteristic of askarels which adds to their attractiveness as electrical insulating fluids is the behavior when partial discharges (corona) occur. With mineral oil and some other insulating fluids, the partial discharge inception voltage is significantly higher than the extinction voltage. Thus a temporary voltage excursion which initiates the partial discharge may lead to a prolonged discharge period with a resultant degradation in the properties of the fluid. With askarels, the inception and extinction voltages differ very little so that temporary voltage excursions which lead to partial discharges are inconsequential. This characteristic of an askarel tends to encourage its being specified as the insulating liquid. Recommended test procedures relating to this characteristic exist.^ However, since acceptable limits may vary from one application to another and from liquid to liquid, no universally accepted basis for Judging the relative merits of new versus old candidate liquids in this respect is available. The lack of such a standard may be considered another "indirect" deterent to the adoption of new fluids. The stability of an insulating fluid describes the permanence of its important physical and electrical characteristics in the operating transformer or capacitor. This is an extremely important characteristic since these fluids are expected to operate reliably and essentially unattended in long-service-life apparatus. Although bench tests on the fluids themselves to screen out obviously unsatisfactory candidates are essential, fluid must finally be qualified in the apparatus environment. Such tests are described in Sections 3.5 and 3*6. Within these tests are also included suggestions for developing cleaning and maintenance procedures for fluids in service. Finally, guidelines and standards must be developed for proper disposal of new fluids as they come into use. ANSI C107.1, guidelines for proper handling of askarels, would serve as a useful example. These guidelines would have to await quantitative determination of the toxicity and degradability of each new fluid, and they would have to be cognizant of changing federal and local waste disposal regulations. ^Institute of Electrical fc Electronics Engineers (IEEE) Standard 5U--1973 "Reconended Practice for the Detection and Measurement of Partial Discharges (Corona) During Dielectric Tecta" National Electrical Manufacturers Association (NB4A) Standard ICP-P2-1973 Official Standards Proposal "Teat Procedure for Measure ment of Internal Partial Discharges in Capacitors" American Society for Testing and Materials (A3TM) Standard D 1868-73 "Detection and Measurement of Discharge (Corona) Pulses in Evaluation of Insulation Syatema" HONS 204774 18 3.2 Flammability Tests for Kew Fluids Fire safety tests have been classified in a variety of ways [1.2].1 For our purpose, it is convenient to classify them as: (a) properties of materials testa (b) system tests (c) apparatus prototype tests. Tests of the first class are small scale laboratory tests designed to measure an intrinsic property of a material which is related to its fire performance, e.g., flash point or heat of combustion. Such tests are useful in developing a qualitative picture of the hazards associated with the use of a material and for purposes of specification and quality control. They are relatively inexpensive and would be the first tests to be run on new candidate fluids. They do not necessarily predict the hazard potential of a material under actual use conditions. System tests involve the interaction of the material with the environment and with other elements of the system in which it is intended to he used. An example is the RTE Arc Tests (A.1.3). System teats attempt to simulate essential features of the device, and thus are less susceptible to standardization than property teres. They are usually larger in scale, ' require more careful planning and are relatively more costly than testa of component materials. Thus, they would find use in the latter stages of the development end qualification of new insulating fluids. Prototype tests involve the testing of the complete device under conditions simulating actual use, for example, the Underwriters Laboratories Tests (A.1.3). Such tests can be very expensive, not only because of the cost of the teat units and test operation but because most of the development costa of the device will have been incurred before the first unit is ready for testing. They should he undertaken only after a careful analysis of the conditions under which the device is intended to function, the possible failure modes, and the type of information that is to he sought in the test. The purpoee of flammability testa and standards for insulating fluids Is to reduce the probability of loss from fire incident to the use of these materials. In order to set realistic performance standards and select appropriate test methods, it is necessary to know the propertlee of the environments in which these fluids will he used, the failure modes that occur, and the consequences of these failures. We have been unable to find any systematic compilation of such hazard information in the course of the present study. Consequently, the establishment of suitable flammability criteria for new candidate insulating fluids will at this time be based on inadequate knowledge of the nature of the hazards which may be encountered. ^References shown in brackets, are found in 3-3. 19 MONS 204775 Recommendation; It Is reconmended that a systematic survey and analysis be made of the environmental conditions in which insulating fluids may be used and incidents involving the use of such fluids which result In fires or related loss. Current flammability specifications for insulating fluids apply to specific chemical families of fluids and vary from one type to another. This presents a barrier to the introduction of new chemical types since they do not fit under existing specifications and there are no clear guidelines for acceptability. It appears that seme present performance specifications have been established with greater consideration given to performance levels which the fluid in question can readily meet than to the actual requirements of the application. Setting too high a requirement may result in the elimination of promising candidate fluids whose properties are entirely adequate for the intended application. The excellent flammability performance of the askarels has led to a very low incidence of firea involving these materials. It is highly probable that replacements for these materials, dictated by environmental considerations, will have poorer fire resistance properties than the askarels. The problem is to determine how much relaxation of flaaaability requirements may be possible without introducing undue riak. This question cannot be answered by reference to the historical record of the performance of devices using askarels. Reco"""dation: It is recommended that flammability performance requirements for new candidate insulating liquids be based on a consideration of the environment in which they will be used and on analysis of likely failure modes, rather than on the properties of the fluids they are designed to replace. 3-2.1 Flash and Firs Points The measurement of flash and fire points gives important information on tne flaaHSbiiity hasard characteristics of a liquid. The Cleveland Open Cup method (ASTM D 92) is used for this purpose. A number of other methods are available for the measurement of flash points. Closed cup methods generally give somewhat lower values for ths flash point than the open cup method, but the latter baa ths advantage of permitting the concurrent determination of the fire point. Since the flash point is a relative rather than absolute measure of flaMSbility hazard, the small differencss in results between the two methods are of littls significance In the present esse. An ASTM ad hoe Coamlttee on Flash Point Methodology and Government Response Is currently considering new test methods with the major objective of resolving conflicting flash point definitions and test methods used in different government regulations t31 The cossaittee appears to favor a closed cup, equilibrium temperature method. 20 HONS 204776 Recoanendatloni The measurement of fluh and fire points using the Cleveland Open Cup method (ASTM D 92) should he retained, at least for the present. Specifications should he set to provide an adequate murgin of safety above the highest anticipated operating temperature for the planned application. If a nev flash point test Is adopted for regulatory purposes its use for determining the flash point of Insulating fluids should he considered, but the determination of the fire point should he retained. 3.2.2 Oxygen Index No standard method for determining the oxygen Index of liquids is available. The oxygen Index Is related to the fire point [4,5], Since Insulating fluids vlll have high fire points, heat loss to the apparatus Is an important factor and the measurements are extremely apparatus sensitive. No meaningful Interpretation of the oxygen Index of liquids In terms of fire hatard characteristics is available at the present time. Racoamendation: Since oxygen index measurements, as presently made and Interpreted, do not provide useful Information on the fire hazard characteristics of insulating fluids, they should not he Incorporated into electrical Insulating fluid specifications. 3.2.3 Autoignition Tests R<ertWMHidation: The autoignition test, ASTM D 2133 sad similar tests do not simulate any hazard mode identified with the use of devices incorporating insulating fluids and thus are not appropriate as fire safety tests. They do provide useful Information, in conjunction vlth other material property tests, on the general performance properties of materials, and thus may he useful in the preliminary characterization of nev candidate insulating fluids. 3.2.4 spray FI--ability Tests Rmeo^egdatios: Inasmuch as our limited survey of the uses and misuses of insulating fluids has revealed no circumstances vfaich would he closely simulated by these droplet combustion testa, there is no reason vhy they should be Included in the specifications for candidate replacement fluids. A further survey of accidents and potential accident circumstances may provide grounds for the further consideration of this type of test in the future. MONS 204777 21 3-2,5 Arcing Teats Arcing appears, at this tine, to be the nost cosmon failure of devices incorporating insulating fluids, failure may result froa a gradual growth of low level discharges during nonal operation or froa sudden overvoltage conditions. Tests which alaulate such failures provide essential information an the fire safety properties of insulating fluids. Such tests have been conducted on sn ad hoc basis [61. There is need for a standard procedure to permit evaluation and comparison of the fire safety properties of new fluids under simulated use failure conditions. Separate procedures for transformers and capacitors aay be needed. Both gradual and catastrophic failures should be simulated. The energy release characteristics of the failure should be measured. Recomaendation: A standard teat method or methods should be developed to simulate the arcing failure of devices incorporating insulating fluids under application conditions. The test method should be based on a careful analysis of the observed or plausible failure modes of such devices. Further reeosmendatlona regarding such a test will be found in 3-5 and 3,6. 3.3 Toxicity Testa on Fluids 3.3.1 Appraisal The determination of the toxicological hazard or riak of a chemical in the environment, whether insulating fluid or otherwise, depends on s large number of factors. Ths exposure to cbeaicsl toxicants can be divided into two rather broad categories: (1) direct exposure of humans during the manufacture and handling of fluids including accidental exposure due to spills, and (2) contaainatlon of the environment that may ultlmatsly lsad to sxposurs of animals and biaans. Tbsss two categories includm both the acute (accidental spill) and the chronic exposures. Both categories, the work environment and environmental contamination, usually produce chronic exposure situations, that is,low level rub-lethal exposures that aay continue for periods of months and years. This is particularly true in the industrial environment. Therefore this type of study should be emphasised. Both categories require an assessment of the effects of the fluid on a complex biological system that may respond in a multitude of weye. A partial list of the biologic responses that the toxicologist must observe in studies to determine the toxicity of a cbmsical is gives in Table C, page 68. For example the production of tumors (neoplastigenesis), benign or malignant (carcinogenesis); the production of changes in the offspring, whether traaaadssible (mutagenesis) or not (teratogenesis); and the production of death are Just a few of the observations that are used as qualifying toxic effects. MONS 204778 22 Tbe observed or uuurtd toxicity will depend on a number of parameters such as mode of administration (see Table B, page 86), length of exposure, the animal species used for testing, the period of observation, etc. Many of these factors have been reviewed for PCSs by Kimbrough (see A.3.7, Ref. 11) and further illustrate the diversity of responses to a chemical using a variety of animal species. This review illustrates the difficulty if not the impossibility of specifying a comprehensive set of test methods. Exposure of humans (industrial toxicolo): While a number of agencies have defined toxicity (see A.3.5), there are relatively few standard tests for determining the toxicity of a new chemical. In the first category above, protection of the work environment is generally achieved through the use of a "threshold" limit value (TLV) formerly known as maximum allowable concentration (MAC). In the United States the threshold limit values are set by the American Conference of Governmental Industrial Hygienists and published annually. The TLV may or may not have been arrived at through animal testing to determine "safe" levels. The levels are constantly undergoing revisions as new information and experience becomes available. Perhaps the most recent example of this is the lowering of the threshold limit value for vinyl chloride monomer due to the finding of liver carcinoma (angiosarcoma) in industrial workers who were exposed to low sub-lethal levels of vinyl chloride over a period of 20 to 30 years. While there is no "one" test method used for determining the toxicity of a chemical to humans, the most frequently used chronic end acute tests are: (1) oral LD50; (2) dermal toxicity (skin absorption)} (3) irritation to determine the effect on skin, eyes or mucous membranes; (It) inhalation; (5) chronic or ingestion studies. These determinations are generally made on animals and the relationship between human and animal data continues to raise numerous questions that cannot be resolTed here. Details on performance of most of these tests are given in the text of the Federal Hazardous Substances Labeling Act (Public Law 86-613, see A.3.5). It should be emphasized that these tests ars recownrtationa and not meant to be all-inclusive. Protection of the environment: The second category referred to earlier as the basard to the environment is certainly nore complex and generally represents a chronic exposure. The hasard that a particular chemical poses to the environment not only depends on its inherent toxicity but also the chamlcal and physical properties of the chemical For example, the chemical stability, water and lipid solubility, biodegradability, vapor pressure, etc., will determine the extent to which the chemical will survive and be distributed throughout the environment. These factors will determine to a large degree the ability of the chemical to he involved in biocoocentration as it proceeds along the food chain. Biodegradation or hiotransformation, may itself produce a nore toxic product, more easily assimilated into biological systems than the original chemical. Biological accumulation requires high lipid solubility versus water solubility, high persistence in the biological syrtern, and a low 23 MONS 20*779 tendency for complex formation vith organic ligands. The degree to which a chemical affects the environment may lead to chronic exposure of man by the vay of the food chain. Thus the foregoing discussions regarding the contamination of the environment may ultimately lead bach to human exposure. These factors must be considered in the evaluation of a new Insulating fluid. 3.3.2 Recommendations 1. Avoid rigid protocols: Any attempt to develop a standard set of protocols for routine assessment of the toxicological hazard of an insulating fluid,or any chemical, has several dangers that must be recognized. Such a course of action may lead to the use of tests that are unnecessary. In addition, such standard protocols may fail to ash questions that may be of overriding significance. Given the multiplicity of the possible biological systems, on all inclusive standardizing test protocol, which considers all possible reactions, would be wasteful of resources if not impossible. Furthermore, such testing would not guarantee the "complete11 safety of a new chemical. 2. Multidisciplinary assessment of toxicity requirsmenta; The extent and nature~of~tests should reflect (1) the possible level of exposure and route of administration to be expected in usage, (2) the consequences that can be expected, and (3) a margin of safety between expected environmental - levels and the dose that produces some adverse effect. In terms of insulating fluids it is important to examine the pattern of usage disposal, the chemical and physical properties and the biological effects of the compound. An understanding of these factors demands that chronic toxicity evaluations be assessed by a sophisticated multidisciplinary group which involves chemists, toxicologists, microbiologists, and biochemists, etc., worhing together. Anything short of this approach is likely to lead to a repetition of the present PCB situation. 3. Ames test: It is recomended that all new fluids he screened for mutagenesis la the Salmonella/microsods test otherwise known as the Ames test [?]. The mutagenicity of the combustion products of new fluids should also be screened using this procedure. Since this test is relatively simple and a high correlation has been obtained between carcinogenicity and mutagenicity with large number of chemicals,it should be run prior to any other toxicological studies to determine the mutagenic potency of a new fluid. Vith data available on the mutagenic potency of a new fluid, a decision to do more extensive testing or to terminate the effort can be made at a relatively early stage. k, Chrtnic tests for initial screening: While no set of routine animal investigations will be universally applicable for the definition of the toxicological properties of all insulating fluids, a screening procedure is valuable. Since the problem is mainly one of low level chronic exposure, this type of study should be emphasized but acute exposure should not be neglected. In fact many toxicological studies start with acute work even though the exposure is chronic. Since chronic exposure in the work environment is likely to occur through the respiratory tract, this mode of administration should be amphasized. Dua to the fact that environmental probls frequently occur by way of water contamination, avian and fish studies are extremely valuable and should be used. 2k MONS 204780 Acuta toxicity studies are frequently used to: (l) determine relative toxicity in order to determine what type of further studies should be made (chronic and acute), and (2) delineate the specific toxic effect and mechanism of action. Such studies aid in identifying target organ effects which are valuable in the design of chronic exposure studies which is the primary concern in the evaluation of insulating fluids. In addition to LD50 data, acute toxicity tests provide information on species variability, sex differences, and signs of intoxication which include the relationship of the response or primary interest and the dose response curve. While a specific protocol is inappropriate, general but detailed guidelines, as suggested by the National Academy of Science in "Principles for Evaluating Chemicals in the Environment" (A.3-7, Hef. hU) are appropriate It 13 highly recommended that appendix B and D of that document be followed closely in evaluating the toxicological effects of a new insulating fluid prior to its introduction into commerce. 5. Combustion product toxicology: The toxicological hazards aseociated with the combustion products of an insulating fluid are likely to repreeent an acute exposure situation with a single nigh concentration of products for a relatively short period of time. The major route of exposure is by inhalation. The assessment of this hazard is complicated by the large number of products that are likely to result during accidental fire. On the other hand, the situation is simplified by the fact that environmental contamination with toxic combustion products is not likely to create a problem aince accidental fires art expected to be relatively infrequent. At the present time there is no accepted or standardized procedure for this type of hazard assessment. However, a recommended protocol for determining the relative toxicological hazard of combustion products for various synthetic materials is under sctive development in a number of laboratories in the world. This is one of the major objectives of the Program for Toxicology of Combustion Products at the National Bureau of Standards. The protocol under development would appear to be applicable with minor modifications to insulating fluids. 3.1* ReCQM"*ndatlon on Degradability and Biodegradability Tests Our knowledge of the degradability of the materiala of Interest as PCB replacsmsnts coats mainly from several sxpsrimenta which were dictated by the isediate needs of individual organizations. There has been no coordinated, industry-wide effort to study the degradability of the electrical insulating fluids currently used or proposed for use in transformers and capacitors. A standard degradability teet, developed by the Soap and Detergent Asaociation, has been successfully employed in that industry's self policing program. The same test procedure has also proven significant 25 HONS 20*781 In evaluating the biodegradability of the pfathalate eaters bj that Industry, A need exists for a similar standard test procedure, or set of procedures, for evaluating the degradability--biological, chemical, or physical--of electrical insulating fluids. In view of the urgency and of the almost universal impact of tha decisions to be made it seems appropriate to reccmiend involvement of governmental agencies--by initial sponsorship and subsidy--in fostering the development of such standard test procedures. Initial phasls should be on tests applicable to those liquids which appear on other grounds to be the most promising candidate replacements for the askarels. Much stress has been placed upon the toxicity of askarels and other widely used fluids. It Is clear, however, that the toxicity of the Intermediate substances in the breakdown processes of the proposed askarel replacement fluids as well as the fluids themselves should he scrutinized. The Ideal solution may be elusive. Since an increase in degradability means an increase in reactivity, as new insulating materials for use in power transformers and capacitors are sought, repeated confrontation with dilemaa such as the following must be expected: 1. Increased degradability may bring with It Intensified toxicity. 2. Increase in degradability may increase fire risk. 3. Increased degradability may cause a lowering of reliability of performance of the insulating liquid. . A consideration of all these points will necessitate a comprehensive examination of all properties of a fluid along with an examination of all properties of its breakdown products, an litnae task. It is evident that until knowledge accumulates care will have to be exercised and Judgment rendered and compromises accepted for each new class of insulating liquids. 3-5 Rec""1--'ded Insulation System Teats and Standards -- Transformers 3.5.1 Insulation System Tests Teats of the separate properties of new transformer fluids, as described in 3.1, 3.2, 3.3, and 3.1*, are adequate for screening the potential usefulness of the fluids, but they must ultimately be qualified in the true transformer environment. The influences of winding, tank, and core geometries, and of possible reactions between the fluids and core, tank and solid insulating materials cannot be evaluated outside the actual transformer configuration or its realistic model. 26 HONS 204782 Each transformer manufacturer will, of course, test new fluids in bis own designs, hut it would he additionally helpful to hare a standard test configuration developed so that different fluids and different fluid/solid combinations can he put on a cooperative basis. This would he useful to fluid and transformer manufacturers, transformer users, and to fire insurance, code and regulatory personnel. Each of these interests should therefore he represented In planning and guiding the detailed design, initial testing, and preparation ox descriptions and instructions for this standard transformer test. As an example, we recommend for consideration the test guidelines described in the remaining portion of this subsection. We must caution that this is only our Initial concept of such a test based on a limited exposure to the problem. Firm guidelines for testing can only be established and validated after considerable experimentation. We recommend research which could lead to such guidelines. The following features should be considered for a transformer test: 1. Tank: fixed site, shape (tank material might be considered to be a variable) 2. Bushings: fixed 3. Sealing and filling techniques: fixed. A provision should be made for regular sampling of the fluid. ' . U. Winding; winding Insulation; core size, shape, and materials; fixed 5. Internal supporting structure and electrical connections: fixed 6. The size of the standard model should be large enough for realistic instrumentation, but not so large that cost and laboratory size are excessive. A unit in the range of a few kVA and a few kV might be the right size. Three types of testing should be carried out for an evaluation of a new transformer fluid: stability, corona, and arcing testing. Instrumentation and measurements which might be considered are described below.1 1. Stability teats: The objective of these tests Is to measure the permanence of the fluid, and the degree of deterioration of solid surfaces in contact with the fluid. Tsmperature cycling should occur. The fluid should be sampled at regular intervals and tested for changes in electrical (dielectric constant, dialactric strength, resistivity, loss factor) and physical (density, viscosity, spacific heat, flsnmiability) proparties. Chemical analysis should be carried on to identify contaminants. The tank walls, core and wire insulation surfaces should be inspected regularly for detection of surface reaction. Standardized procedures should be employed wherever available. 27 HONS 204783 2. Corona tssts: A standard corona source should be designed into the standard model transformer, possibly associated with one of the windings. Corona inception and extinction levels should be measured, using standard procedures. Corona should persist at a prescribed level for a specified duration. The fluid should be regularly sampled during this tine to measure changes in electrical and physical properties, and to detect contaminants. Tank walls, core and conductor insulation should be regularly inspected for surface reactions. An attempt should be made to detect and identify any tendency to convert from corona to a higher current arc. Procedures for removing contaminants should be considered. 3- Arcing tests: A standard method should be developed for igniting an arc within the model transformer. Possibly the standard test could include different arc modes, between adjacent layers of a winding, between windings and from one winding to the tank. The following parameters should be controllable variables: source voltage and impedance or arc current, arc duration and initial temperature. The following phenomena might be meaaured or observed: a. Growth of arc current and arc length and position. b. Pressure wave within the tank and on the tank wall. c. Conditions to cause tank rupture. d. Energy release from the fluid. e. External fire; fire threat to neighboring structures. f. External blast wave and damage to neighboring structures. g. Escaping combustion/explosion products. After an acceptable Insulation model has been developed and thoroughly teated, the design and testing procedures should be documented for consideration by standards, codes, lnsursnce groups and regulstory agencies. 3.5.2 Standards -- Transformers Several of the AOT C 57 series of standards dtaling with transformers make reference in the title or elsewhere to the liquid filling, l.e., "mineral-oil imersed," "mineral-oil filled," stc. These are probably only intended as restrictions to eliminate dry-type transformers since AI8X C 57.80 (Terminology) states "the term oil Includes synthetic liquids as well as mineral transformer oil." However it would seem appropriate that the whole series be studied and necesssmy modifications made in future revision! to remove the restriction* where they are not essential. In those cases where the restrictions are necesssmy* corresponding documents which apply to transformers with other insulating liquids.should be prepared. lASTM D 1868-73 "Detection and Measurement of Discharge (Corona) Pulses in Evaluation of Insulation Systems" 28 MQNS 204784 Son* standards and formal guideline* which might be needed for transformer* with new fluids are: 1. Standard sizes and ratings 2. Guidelines for acceptance testing. 3. Installation and operation codes. k. Guidelines for maintenance of new fluids In transformers. 5. Guidelines for storage and disposal.1 6. Guidelines for fusing. The insulation system tests on transformer models described above should be designed to provide the information needed for these new standards and guidelines. While fusing is strictly outside the scope of this study, more flammable fluids could be used if higher levels of fusing were present; the model system tests could place quantitative requirements for fuse protection of new, more flanmable fluids. 3.6 Recommended Insulation System Tests and Standards -- Capacitors 3.6.1 Insulation System Tests Tests of the separate properties of new capacitor fluids, as described in 3*1* 3.2, 3.3* nd 3.U, are adequate for screening the potential usefulness of the fluids, but they must ultimately be qualified in the true capacitor environment. The influence of electrode and solid dielectric geometry and possible reactions with these surfaces are essentially impossible to evaluate outside the actual capacitor configuration. Each manufacturer will, of course, test new fluids in his own capacitor designs, but it would be additionally helpful to have a standard test configuration so that diffsrent fluids and different fluid/solid combinations can be put on a comparative basis. This would be useful to fluids tnd capacitor manufacturers, to capacitor users, and to fire inaurmaee, code and regulatory personnel. All of these interests should, therefore, participate in guiding the detailed design and use of such s standard capacitor test. The suggested model tests on capscitor insulation systems are identical in approach to those recommended for transformers in 3.5.1. Many factors responsible for insulation failures in transformers are present also in capacitors. There are several obvious practical differences in the components of trinsformers snd capacitors. Because there are ^ contributor has suggested that guidelines and standards will also he needed for the proper substitution of new fluids for askarel in existing transformers. 29 HONS 204785 rtifferences, the entire suggested guideline for capacitor nvrt+1 test* is described. As for transformers, we again emphasise that the following guidelines are our initial recommendations for a model test; they point out a number of details which should be Investigated in any rodel test. However, they are not firm and verified test procedures. These can only be established after considerable experimentation. The following features should be considered for a standard capacitor model test* 1, Container: fixed size, shape, material. 2, Bushings: fixed, 3, Sealing and filling techniques: fixed; provision should be made for regular sampling of the fluid, possibly by dismantling one out of a group of test units at regular intervals. **. Electrode foil; Insulation film size, thicknesses and roll configuration: fixed. 5. Foil and film materials: variable, so that the reaction of the test fluid with different foil and film materials can be measured. 6. Foil/film roll supporting structure and electrical connections: fixed. ,, Three types of tests should he carried out for a complete evaluation of a given fluid/film/foil combination: for stability, for corona effects, and for arcing effects. Instrumentation and measurements which might be made In each test are described below. 1. Stability testa: The objective of these teats is to measure the permanence of the fluia, and the degree of deterioration of solid surfaces in contact with the fluid. Temperature cycling should occur. The fluid should be sampled at regular intervals and tested for changes in electrical (dielectric constant, dielectric strength, resistivity, loss factor) and physical (density, viscosity, specific heat, flinability) properties. Chemical analysis should be carried on to identify- contaminants. The container vails, foil and film should be inspected regularly for detection of surface reaction. Standardized procedures should be employed wherever available. 2. Corona teats; A standard corona source should be designed into the standard capacitor configuration. This might be the foil itself, or an auxiliary electrode. Corona inception and extinction levels should be measured, using standard procedures. Corona1 1ASTM D 1868-73 "Detection and Measurement of Discharge (Corona) Pulses In Evaluation of Insulation 3ystema" 30 HONS 204786 should persist at a prescribed level for a specified duration. The fluid should be regularly sampled during this time to measure changes In electrical and physical properties sad to detect contaminants. Foil, film and container vails should be regularly Inspected for surface reactions. An attempt should be made to detect and Identify any tendency to convert from a corona to a higher current arc. Procedures for ren^vlng contaminants should be considered. 3- Arcing tests: A standard method should be developed for Igniting an arc vithln the container. The source voltage and Impedance or arc current, arc duration and Initial temperature should be controllable variables. The following parameters might be measured or observed: a. Arc voltage. b. Pressure wave (intensity and time parameter) vithln the container. c. Pressure on the container vail. d. Energy release from the fluid. e. Conditions to cause container rupture. f. External fire; fire threat to adjacent capacitors. g. External blast vave and damage to neighboring capacitors snd structures. h. Escaping arc products. _ . After a suitable model has been developed snd thoroughly tested, the design snd testing procedures should he documented for consideration by standards, codes, insurance and regulatory agencies. 3.6.2 Standards -- Capacitors Several guidelines and standards vlll eventually have to he developed for capacitors vitta new fluids; these can he grouped in several categories*. 1. Standard sixes snd ratings. 2. Guidelines for acceptance testing. 3. Installation and operation codea, U. Guidelines for maintenance of new fluids in capacitors. 5. Guidelines for storage and disposal. 6. Guidelines for fusing. HONS 20^787 31 These would serve capacitors with new fluids much as AISI C?5.2, NEMA CPI and BIA RS-392 now serve askarel-filled units. The model system tests described above should be designed to provide the information needed for these new standards and guidelines. We should note that BIA is moving in this direction already with its RS-392-I standard (alto AISI C83.679) but that the combination of nonaskarel fluid and nonpaper dieleotric has not yet been covered by EIA standards. Some brief explanation of the above fusing recommendation is in order. While fusing is strictly outside the purview of this study, it does possess a close relationship to the flammability issue and therefore deserves some mention. A more highly flammable fluid can possibly be tolerated in a capacitor if the fusing protection is sufficiently good. The system model tests described above should be designed to quantize this trade off between flaannability and fusing. 3.7 Standards, Regulations, Codes In previous sections of this report, frequent references have been made to standards, regulations, and codes, but mostly in the context of specific test techniques or specific new fluids. In this present section we will briefly review this same material, not in terms of the techniques themselves but rather to investigate mechanisms for formal adoption of ' the techniques. - It is true that existing codes and standards are such that locations can os found for initial service testing of capacitors sad transformers with new fluids; however, new or revised standards will he required as new fluids, and new test procedures for these fluids are "proved out" in field testing and as they approach full-seal* coamrcial use. Because of the long time required in the preparation of such standards, the existing standardizing committees in IEEE, A3TM, IXKA, and EIA should be encouraged to accelerate current efforts and to initiate new programs where necessary to permit the development of standards to proceed as nearly as possible in parallel with the field testing phase. Of concern in a discussion of the need for new, or the revision of existing, standards documents is the problem of achieving adequate "user" participation in the technical society committees which prepare them. It has been called to our attention, for exaa^le, that of the near 100 members of the A3TM D 27 comlttee which ia responsible for insulating fluids standards, only the order of 10$ are from utilities, and participation of federal Government agencies is at a very much lower level although the Government through GSA, DoD, DoI, etc., is perhaps the largest single purchaser of power transformers. It seems evident that greater participation by the "user" group might significantly expedite the preparation of needed standards. 32 MONS 204788 Concern with the coeta of staff participation in standardizing activities probably is the principal deterrent to greater involvement both of this "user" group and of the academic community. This concern no doubt also sets a limit to the level or effort in these activities by the manufacturing community vhich currently appears to be making the greatest contributions. Recommendation: Establish a fund to pay travel costs incident to staff participation in the preparation of the most urgently needed standards, in order to broaden the base of industry and academic participation and greatly reduce the time required in the preparation and adoption of nev and revised standards. The National Electrical Code (NEC) plays a key role in commercial acceptance and usage of nev fluids, as baa already been pointed out in earlier sections. It is used by fire insurance companies in Judging the acceptability of installations for insurance coverage, it forms the basis of many local building codes and ordinances, and it has been incorporated into federal regulations. It is important, therefore, that the Electrical Coordinating Committee of NEC he cognizant of, and give counsel on development of nev flammahility tests such as discussed in Sections 3-2, 3-5, and 3.6. Independent testing laboratories such as Underwriters Laboratory and Factory Mutual Engineering and Research must also he given the opportunity to contribute to formulation of these nev flammability ' test procedures. . Each of the pertinent Government regulatory agencies has its ovn particular area of concern regarding nev insulating fluids, and can contribute from its ovn special strength in developing nev teets, and in interpreting regulations as they apply to nev fluids. In particular, the folloving specific comments can he made: 1. OSKA: Consideration should be given to modifying regulations to allov insulating fluids not yet recognized by the National Electrical Code. OSBA might also encourage development of airborne toxicity tests and flammability teats vhich are particularly adapted to the vork place. 2. FDA and State Departments of Public Health: Early evaluation of the nev candidate fluids Is needed, regarding not only toxicity but also likely avenues of entry into food production and distribution. 3. DOT*. The hazards of nev fluids particularly associated vltb transportation should he explored; also a determination should be made whether any nev candidate fluids fall within DOT'S existing hazardous or toxic substance classifications. U, HEW: Test procedures will need to be developed for determining toxicity and degradability of nev flulde. MQNS 204789 33 5- EPA. State Departments of Natural Resources and Public Health: Analytic procedural vill be needed for measuring pretence, toxicity end degradability of nev fluids in the natural environment; present background levels of the subetances should also be determined for future comparison. An evaluation sbould be made of the Impact of proposed toxic substance legislation on the Industry If It turns to the nev candidate fluids. Hecommendation; Form an Interagency task force vlth representation from each of the federal (and perhaps frca a fev of the more active and involved state) agencies, charged with the task of developing a consistent governmental approach to solution of the problems imposed by the safety oncern vlth electrical insulating fluids. If such a task assignment .s not feasible, even the minimal assigned responsibility for keeping *ach other and the electrical pover industry at large veil informed as o planned actions would veil Justify its formation. The task force could become the focal point for the industry's contacts vlth the Government In this area of concern. HONS 3U 3.8 References for Section 3 1. A. F, Robertson, "Test Method Categorization sad Fire Hazard Standards." Standardization Neva. VqL. 3 (1975), pp. 18-20. 2. H. L. Malhotra, "The Philosophy and Design of Fire Tests." Proceedings of the International Symposium on Fire 3afety of Combustible Materials. University of Edinburgh. Oct. 1975, p. 1U9, 3. H. A. Wray, "New Flash Point Tester for the Paint Industry." Journal of Paint Technology, Vol. U5 (1973), pp. UU-5U. 1. A. F. Roberts, "Extinction Phenomena in Liquids." Proceedings of the Fifteenth Symposium (International) on Combustion, The Combustion Institute, 197 k, pp. 305-313. ~ 5. D. J. Rasbash, "Relevance of Fire Point Theory to the Assessment of Fire Behavior of Combustible Materials." Proceeding of the International Symposium on Fire Safety of Combustible Materials, University of Edinburgh, Oct. 1975, p. 169 6, D. A. Duckett, "Catastrophic Explosion Teste in Insulating Fluids." Internal Report, RTS Corp., Waukesha, Wisconsin 53186, Hov. 9, 1971*. 7. J. McCann, E. Choi, S. Tanasski and S, Ames, "Detection of Carcinogens, as Mutagens in the Salmonella/Micro some Test*. Assay of 300 Chemicals." Proc. Natl. Acad. Sci., Vol. 72, Dee. 1975, pp- 5235-5239. ` HONS 204791 35 appendix Status A.l "Askarel-class" Transformers A.1.1 Description of the "asharel class" of transformers Electrical transformers are used to change voltage, current and impedance levels in a vide variety of applications in electrical apparatus. These transformers range from the very smallest types in solid-state electronic circuitry to very large 60-Hz power transformers. This study is concerned vith the following restricted but important, set of transformers: 1. Utility and industrial distribution transformers; these include a wide variety of applications: uses in utility "network." systems, in secondary unit substations and factory load centers; they may be under ground, within buildings or adjacent to buildings or other occupied areas. This group is therefore Identified by industrial and cosmercial use in areas where the load and population densities are high- Typical ratings are 500-2500 kVA, 4-34 kV. 2. High voltage transformers, for electrostatic prmclpltators which remove particulate pollutants from furnace exhaust gases. . 3. Transformers for multiple-unit type electric railroad cars; these art located under the flat-bed of the car. While many precipitator and most railroad transformers are askarel'illed, only 10-15$ of utility and industrial distribui-ion transformers identified above a*e the askarel-type. Most power transformers use a liquid insulant which serves two purposes: electrical insulation and cooling. Compared with air insulation, the liquid allows much closer spacing (smaller size) for a given voltage, and compared with air or solid insulation (as used in "dry-type" transformers), the circulating liquid coolant leads to a much smaller size for a given power rating. Mineral oil was traditionally used for this purpose and is still used for most applications. Estimates are that as much as 96$ of all liquid-cooled power transformers use mineral oil, on an KVA or gallon-of-coolant basis. However, the particular set of applications described above have critical nonflaaability requirements since they are generally located in heavily populated areas or are used in high-temperature or high firerisk conditions. "Askarel" transformer liquid was developed, using blends of the polychlorinated biphenyls (PCBs), in the early 1930's to replace mineral oil in applications where nonflwabillty was an important requirement. The PCBs for these transformer askarels are manufactured by Monsanto Industrial Chemicals Company and are sold by Monsanto under the trade name "Aroclor." Two grades are presently used (shoving percent chlorination and 1974 domestic sales in thousands of pounds) Aroclor 1242 Aroclor 1254 (42$) (54$) 6207 6l85 36 HONS 204792 Ten to twenty percent of this production joes into electrostatic precipitator and railroad transformers*; most of the remainder Is used in the utility and industrial transformers. Approximately 5000 askarelfilled transformers are manufactured per year in i-h* U.S., at a value of about $U5 million. The term "askarel" which describes a broad class of fire-resistant [l], synthetic, chlorinated, hydrocarbon insulating liquids is defined more rigorously in the following standards and guidelines. IEEE Std. 76-197**', Tkk Guide for Acceptance and Maintenance^ of Transformer Askarel in Equifwnt ASTM D-2203-73; Chlorinated Aromaf'r Hydrocarbons (Askarels) for Transformers (i su see D-2233-70) National Electrics] Code, 1975; Art. 100 ("Definitions1'), 1*50 (Transformers and Transformer Vaults), 1*60 (Capacitors). In addition to the fire-resistant characteristica of tne askarels, their high dielectric strength, low electrical loss, adequate viscosity, specific heat, and thermal stability, and their nonsludging qualities . have combined to make them Increasingly attractive transformer lnsulants since their introduction in 1932 [2], Their high cost relative to mineral oil has, however, restricted their use to transformers for the abovedescribed fairly well defined set of applications--the "aakarel-class1' of transformers.^ It la for this set of uses, then, that alternative transformer fluids are being offered, and test methods and standards for these alternative fluids should therefore be responsive to these specific applications. The askarels provide fire-resistance by generating Inert hydrogen chloride upon decomposition. Their flash and fire points are also very high, making them difficult to ignite. Hydrogen chloride is a highly corrosive material, and so extensive special designing and installation must be observed. Only particular, 'orrosion-resistant winding and tank coating materials can be used, "scavenger" materials must be present in the fluid to take up any small amounts of hydrogen chloride which are generatea in normal operation, and in addition the transformer room or vault must be adequately ventilated to prevent buildup of HC1 vapor. It should also be noted that askarels are very efficient solvents and that consequently all solid insulating materials, including gaskets and paints, must be carefully selected. ^These two special uses are growing as mass transportation and air pollution control programs expand. O Askarel transformers are specifically Identified in the National Electrical Code and in many local fire codes and ordinances. 37 HONS 204793 Because of the toxicity and environmental problems of the PCBa, drainage from askarel transformer installations must also be adequately collected so that small or large scale spilling of fluid, during filling, testing, operation or catastrophic rupture, will be contained and not reach the ground water or navigable water systems. These special requirements of existing askarel-transformer designs and installations are being Identified in order to have a clearer picture of the processes which will be involved in replacing the PCB fluid with an alternative. Either the new fluid will have to be compatible with all of the characteristics of the existing askarel installation, or new Installations will have to be designed to accommodate the new fluid. These constraints will shape the requirements of testing procedures to evaluate and qualify new fluids and transformers filled with new fluids. A. 1.2 Sources and consequences of transformer tank rupture In the construction of "asksrel-class" transformers, the windings are completely immersed in the fluid, inside a tank which ia sealed to prevent the fluid from escaping.^ The tank muat be heavy vailed to with stand the sudden pressure of a high-current electrical fault within the ' tank. The tank la generally finned for convective heat transfer to the surrounding air, and It often has a pressure relief device to bleed off any small but steady rise of gas pressure within the tank (as from occasional lov-level arcing or corona). If mineral oil or s new liquid inaulant such as a silicone fluid is to be used as a replacement for askarel In transformers, one can be sure that containment of this fluid will be a crime objective, for safety, environmental and economic reasons. In anticipation of require ments for containment tests and standards, it will be thm purpose of this section to review the present state of stsadards, testa and understanding In the area of transformer fluid containment. Although the fluid is held in substantial drums or in the sealed transformer tank, loss of the fluid can still occasionally occur under several different circumstances: a. during transport of the fluid or of the filled transformer, b. during processing and handling of the fluid by the manufacturer or user, ^The amount of aaksrel fluid in such a transformer ranges from UO to 1500 gallons, depending on the size and rating of the transformer. 38 HONS 204794 c. during in-service operation of the transformer, and d. upon disposal at the end of the unit's life. Instructions for proper transport of the fluid and of the transformer are included within the general procedures of the Department of Transportation [3], (see also Section A.4). In addition, the following are cited as examples of the types of guidelines and standards which could he developed to prevent leakage during acceptance, maintenance, handling and disposal of transformers and fluids; a. "Guidelines for handling and disposal of capacitor and transformer grade askarels containing polychlorinated biphenyls," ANSI C10T.1 - 1974; American National Standards Institute, 1430 Broadway, New York, N.Y. 10018. b. "IEEE guide for acceptance and maintenance of transformer askarel in equipmentIEEE Std. 76 - 1974, Institute of Electrical and Electronics Engineers, 345 E. 47 Street, New York, N.Y. 10017. These standards are discussed more fully in Section A.1.5. The oil spill regulations developed and enforced by the Environmental Protection Agency [4] also can serve as a model far Judging the restrictions on other fluids leaking into the environment. These are discussed in Section A.4. A very rugged design is used for transformer tanks in order to maintain containment under most external or internal stresses likely to be encountered In service. These tank designs are apecifled by Industry standards published by ANSI [5]. Tank dimensions, static pressure withstand capability and corrosion resistance are specified. These standards should be reviewed for applicability with new fluids. Rupture of a transformer tank while in service is most often due to arcing within the tank. This is a very complicated phenomenon which has been studied for many years. The following paragraphs describe the phenomenon, its effects, and possible controls as presented in recent articles by Barkan et. al. [6], Goodman and Zupon [71 * Nettleton [8], and Ristuccia and Benton [9] a. The arc move often starts as a low-current, high-impedance insulation breakdown between adjacent turns somewhere within the transformer windings. 1Revision of this document ia currently under way. 39 HONS 204795 b- A "high proportion" [T] of transformer failures result fro* the gradual buildup of pressure in the tank due to the gas evolved by the low current arc. The preasure-relief valve present on most network transformers could avoid this as a cause of rupture in this important application of "askarelclass" transformers. Some users claim, however, that the pressure relief valve is more trouble than asaet, and therefore avoid its use. A fuse could also be used to clear these lowlevel faults, although network transformers are generally operated without primary fusing. c. The high-impedance arc will grow with time to a high-current fault as the arc attacks the insulation on more turns. Because the use of pressure relief valves generally prevents the type of tank rupture described in part b. above, the high current arc may be the most prevalent cause of tank failure in the network transformers. We presume that the final growth of current is rapid, creating a bubble of high pressure gas within the fluid. According to Barkan [6], this bubble, as it expands, drives a "piston" of oil against the tank wall, momentarily overloading the wall and causing rupture. d. Techniques for calculating the peak tank wall pressure under arcing faults are presented in both the Nettleton [6] and the Barkan [6] articles, but both also point out the difficulty of this for actual tank geometries. The etatic pressure relief valve Is of no use here because of the rapid rate of pressure rise [9l. and gas-flame venting techniques are also generally Inapplicable for the same reason [8]. There is apparently a proposed ''vision to the ANSI transformer standard (C5T.12.201971) for a standard transformer arc test to determine the necessary tank arc pressure withstand capability [9]. e. Further complicating the pressure calculation are the effects of energy release by burning of the fluid and of fluid products. This energy will build up the pressure, and the pressure wave will have characteristics due to being flame- as well as arcdriven [8]. f. One should not rely purely on tank withstand strength to prevent tank rupture under arcing, but rather this should be coordinated with a properly specified current-limiting fuse [6, 9] The peculiar impedance characteristics of the arc prevent use of the familiar I*t type of fuse specifications; rather, the fuse manufacturers should recomend the proper fuse, based on the user's known voltage and fault-current levels.A ^One user points out that network transformers are often in subways and must therefore be submersible, and fuses are therefore avoided. 40 204796 MOWS Protection against the high current ere la of greet Importance, not only beceuee It nay be the most prevalent source of "askarel-class" transformer failurea, but elao becauae the ere la e high-temperature source which can Ignite the fluid or the decomposition producta due to the ere. Thua, protective teata end ateoderda must coordinate the available temperature and energy (aa for lnatance detemlned by a current limiting fuse) with the Ignition tempernture and energy of the fluid and of flammable componenta created by the arc. The "tranaformer system testing" discuased In A.1.3 will explore thie type of testing in depth. It la worth noting at this point that aa arc in an insulation fluid can cause a vide range of decomposition products, depending on the chemical nature of the fluid, the availability of reactant gases such aa oxygen, and the Intensity (temperature) of the arc. Products can be flammable, explosive, corrosive and toxic. Aa examples, arcs (or low level discharges such aa corona) can decoaipose mineral oil Into hydrogen and hydrocarbon gases (and probably water if air la pressnt), while aakarels are known to break down to hydrogen chloride, and some hydro carbon gases. Ritrogea, carbon monoxide and carbon dioxide are also aald to result [10] but again probably only in the presence of air.1 Hot only can theae products be directly harmful, but they can also weaken the dielectric strength of the insulating fluid and thereby set the stage for subsequent high current flaxhover and tank rupture. One should, therefore, seek standard procedures by which the fluid can be cleaned up at regular intervals, or after known arcing, in order to avoid future equipment failure. As guides In preparing such standards, we should note the following procedures which are now regularly oployed. a. "Scavenger1* substances are mixed with aakarels to take up the corrosive hydrogen chloride which is produced by occasional low level electrical discharges within the fluid. b. In spite of the presence of scavenger material, users are instructed to bubble nitrogen through the fluid imaediately after known arcing in order to pick up hydrogen chloride before it can do corrosion damage. . c. The IXSK Standard 76-1971* suggests that moisture content and dielectric strength of askarel be regularly checked if electrical discharge is known to be occurring. Because of the combination of preventive measures outlined above, manufacturers sad users report that present transformers of the network type experience about a 0.01 to 0.02% yearly rate of tank rupturing 1A recent report (Sational Conference on PCBa, Chicago, ov. 1975) Indicates that chlorinated dibensofurans are formed in PCB heat exchanger fluids. 1*1 HONS 204797 failures. Tests, standards, and guidelines must be prepared for new fluids in order that low failure rates will also be obtained for new transformer fluids. Even though a low rupture rate is experienced with transformers, it is still industry practice, generally bached by fire or environmental code enforcements, in indoor installations to place fluid-insulated transformers in vaults or special rooms. Mineral oil units must be in fire-proof vaults 111 ]. Low-voltage askarel transformers need not be in fire-proof vaults but must be installed so that ventilation is adequate to prevent accumulation of toxic vapors [12] - Drainage facilities should be such that spilled fluid , contaired A.1*3 Transformer system testing In the early days of askarel usage a short-circuit test and an arcing test were performed by Underwriters' Laboratories, Inc., on actual transformers filled with an askarel (General Electric "Fyranol"). Details of these tests are reported in the UUffi No. 2581 Report, dated _ September 29, 193U. More recently, catastrophic explosion (arc) tests with several insulating fluids were performed by RTE Corporation [131 1. Underwriters' Laboratories Teats The transformers were rated 60 Hz, 5 kVA, 220(^110 volts. One of these was short-circuited externally. The other was subjected to an arcing test by being connected to a 12,000 kVA generator so as to apply 7,500 volts between the ungrounded 122-volt lead and ground. In both cases, pressure-tight covers were provided with pressure-relief diaphragms designed to rupture at not less than 1(^12 psig (69-83 kPa gauge). The results of the two tests were rather similar: no appreciable change in pressure was noted for approximately 2 minutes after current was supplied to the transformer. Thereafter, tbs pressure Increased rather rapidly until the diaphragms ruptured. Large volumes of gases were liberated during the tests, hut the gases did not ignite upon application of a test flame. 2. RTt Arc Tests Pour Insulating fluids were tested: mineral oil, DC-200 Silicone fluid (Dow Corning), hydrogenated paraffin oil "RTemp" (RTE Corporation), and an askarel (Westinghouse Inerteen 70-30). Approximately four gallons of each fluid were used, preheated to 150 C. The containers were covered before the test, but the rupture pressure (presumably roughly the same in all four cases) was not specified. In each test the liquids were subjected to a massive arc discharge: 1*660 bo 1*820 A, the arc being initiated with U800 volt open circuit. Each of the four samples showed a violent flash initially and wide spraying of the liquid as the cover was blown free, but only mineral oil continued 1*2 WONS 204798 to bum. It is important to not* that mineral oils used in transformers have flaah points in the vicinity of 150 C and fire points about 15 c higher. The highly-chlorinated askarels, on the other hand, have no fire point up to the boiling point. They have no true flash points but socalled pseudo flash points may be observed under conditions of the test. Silicone fluids have very high flash points at about 300 C (see the section on flammability tests). RTemp, betn* a hi*n hydrocarbon freet . presumably alsr> has a high flash point. Thus the tests were conducted with initial fluid temperature near the flush and fire points of mineral oil, but substantially below the tlash points of the other fluids. 3. React Ion from Industry The consensus appear* to be that both sets of tests (UL and RTE) show askarels to have very low flammability. The UL rating of 2-3 on the scale of 100 for flammable fluids reflects this consensus. Furthermore, the Factory Mutual representativea appear satisfied that the RTE tests show also the silicone fluids and RTemp to have flammabilities in the same, generally acceptable, range as askarels. UL representatives have not comnented on the RTE tests. On the other hand, the nature of the initial flash in RTE tests and of the observed pseudo flashes in flash point testing .of askarels does not appear to he understood; some concern therefore remains. Intense heating of the decomposition products by the arc may lead to some reaction of these products with air. Contributors have also suggested that in arc explosion tests of askarels the fluid is electrically heated to decomposition temperatures, and that decomposition Itself la exothermic. As a result, several representatives frcm Industry and insurance buslneaa have commented that a carefully standardized arcing test should he developed. Manufacturers, users, and testing laboratory personnel have indicated a willingness to cooperate in developing such a teat. A.1.4 Alternatives to askarels in "askarel-class" transformers; Influence on total transformer tests end standards We must emphasize again that the purpose of this study is not to evaluate alternatives to askarels hut rather to survey the status of teats and standards by which such alternatives may he Judged. The characteristics of the available alternatives must he explored, however, to the extent that unique characteristics will influence the definitions of tests and standards. This section will emphasize the effect of alternatives on the total unit, while the Section A. 4 will look at the fluids themselves. Many concepts originally brought up in A.1.2 ("Sources and consequences of transformer tank rupture") will he revisited in terms of the goals of this section. The characteristics of available alternative fluids are presented in detail in Section A.3 end therefore need not be repoated here. It is important to note, however, that these represent a wide range of density, 43 HONS 204799 viscosity, operating temperature, corrosiveness, volatility, hydrolytic stability, flammability, etc. Because of the diversity of these values, we will see that universal fluid performance specifications will he difficult to write without taking into account the unique characteristics of each particular insulant. This will be equally true in establishing performance standards for the total transformer assembly. The important fluid characteristics which must be accounted for in any future transformer standards and specifications are listed below: a. Material compatibility. The chemical activities of the available insulation fluids differ in kind and degree. Tests must, therefore, be devised to measure adequacy of the materials used for protective coating on inside of the tank valla, for insulation on the wire used for the transformer windings, and for miscellaneous other components such as disconnects and fuses which might be in contact with the flqid. b. Temperature. If the particular insulating fluid requires an unusually high operating temperature, all materials used to construct the transformer must be compatible with this temperature, and in addition the location of the operating transformer must be selected to be acceptable to this higher operating temperature. Existing standards such as: "General Principles for Temperature Limiti in the Rating of Electric Equipment," IEEE 3td. 1-1969. are probably adequate for this purpose. Standard sampling techniques may also be desired to insure the stability of the fluids at high operating temperatures. ' . c. Cooling. The fluid in a transformer is used as both an insulating and cooling medium. An economically competitive transformer must have its cooling system deaign closeiy matched to the characteristics of the coolant fluid; new fluids will undoubtedly require some redesign of the transformer. This may necessitate new standards in transformer tanks, cooling surfaces, circulating equipment, etc., and the user may demand more thorough cooling specifications on transformers with new coolant fluids. d. Hydrolytic stability. Water is generally a problem in transformer fluid, but the degree of difficulty will probably change from one fluid to another. Sealing against moisture entry, and processes for removal of water should, therefore, be specified for each new insulant. e. Containment. The different fluids will have different densities and required volumes for any given transformer power and voltage ratings, and space and supporting structure specifications must be made up for each new fluid. The types of seals will also have to be correlated with each fluid, since different fluids have different leakage characteristics. Since rigorous contain ment of any fluid may be the goal of future regulations, standards and tests for evaluating containment of each tank/fluid combination will be of Increasing importance. UU MGNS 204800 Standards for external protection hould be coordinated with the particular class of transformers, the fluid and the level of tank containment security* By external protection, ve secondary protection provisions, beyond the tank Itself. This Includes both vaults, for extra containment in case of tank failure and protective circuit elements (such as circuit breakers and fuse*) to prevent tank rupturing arcs from occurring. Much as an electrical system has its various pover and protective elements coordinated with one another, so also might the industry consider more quantitative coordination of the levels of transformer containment. One might, for instance, consider that increased speed and redundancy of circuit breaker and fuse protections could allow a less cumbersome and expensive vault structure. Extensive probability analyses of actual operating experience would be required before such a coordination plan could he acceptable for fire codes and insurance purposes. Although no new standards or tests may he required, we Include, for completeness, a mention of two other available alternatives, namely drytype and gas-cooled transformers. These transformers depend on ambient air or special gas Insulation. The dry types are available now and are being used extensively in place of askarel transformers. Relative reliability and overload capability have been questioned by some users but are quite acceptable to others, which may mean that standardized tests on these characteristics should he available to resolve such disagreements. Gas cooled transformers are not yet as readily available for ' substitution. Tssts may he required to establish the harmlessness or the need for containing products of arcing within the insulation gas. Development of these tests should be guided by the extensive existing - work on entire gas-insulated substations. A.1.5 Test techniques and standards used by manufacturers and users to evaluate fluid-filled transformers The transformer manufacturer and user have the following standards documents available for guidance in selecting and handling the Insulating fluid. 1. ASTM^ D 2283-T1* "Standard Specification for Chlorinated Aromati. Hydrocarbons (Akkarels) for Transformers." This document gives detailed specifications of the physical, chemical and electrical properties of six askarels which have been, or are being used in transformers and identifies specific test methods. 2. ASTM D 3146-75 "Standard Specification for Oxidation - Inhibited Mineral Insulating Oil for Use in Trsnaformers tnd Circuit Breakers." ^-American Society for Testing and Materials, 1916 Race 3t., Philadelphia, Pa. 19103* 45 HONS 204801 This document gives detailed specifications of the physical, chemical and electrical properties of the oil together with approved test methods. 3- ASTM D lQl*0-73 "Standard Specification for Uninhibited Mineral Insulating Oil for Use in Transformers and in Oil Circuit Breakers." This document gives detailed specifications of the physical, chemical and electrical properties of the oil together with approved test methods. **. ANSI^- C 59.131-1971 "Guide for Acceptance and Maintenance of Insulating Oil in Equipment" [Also issued as IEEE2 Std. 6U-1969]. The purpose of this guide is to assist in evaluating the serviceability of oil received in equipment', oil as received from the refiner for filling new equipment at the installation site, and as processed into such equipment. It also is intended to assist the operator in maintaining his oil in serviceable condition. The guide reconmends standard oil tests and evaluation procedures, methods of reconditioning and reclaiming, the levels at which these become necessary, and the routinea for restoring oxidation resistance where required by the addition of inhibitors. 5. IEEE Std. 76-1971* "Guide fur Acceptance and Maintenance of Transformer Askarel in Equipment." This guide asslata in evaluating aakarels as received in transformers, reactors, and accessory equipment operated at power frequencies and in efforts to maintain aakarels in serviceable condition. It recommends standardized tests and evaluation procedures. Methods are outlined for reconditioning and reclaiming aakarels whenever necessary. 6. ANSI C 107.1-1971*. "Guidelines for Handling and Disposal of Capacitor- and Transformer-Grade Aakarels Containing Polychlorinated Biphenyls." This document gives typical physical, chemical and electrical properties of aakarels used in insulating fluids in transformer and capacitor construction, standard methods by which these properties are measured, details methods of shipping and handling, spells out necessary safety precautions, recoanends labeling practices and details approved disposal techniques. It also lists organizations having facilities for analysis and for disposal and it describes analytical procedures for the determination of PCBs in air, water and sediment. 1Aaericen National Standards Institute, 1930 Broadway, New York, H.Y. 10018. institute of Electrical and Electronics Engineers, 3**5 2- **7 Street, New York, N.Y. 10017. HONS 204802 1*6 7. IEEE Drart "Specification for New, Unused Transformer-type Askarels." This specification covers bulk shipment in tank cars and drums of new, synthetic, nonflammable electrical insulating liquids of the chlorinated aromatic type known as askarels which are used as insulating and cooling media in liquid-filled transformers. While this specification is intended to cover the composition and characteristics of askarels for purchase only and does not apply to liquids in apparatus, a number of discontinued types are included for purposes of identification and historical significance. 3. IEEE Std. 283-1968 "Guide for Installation of Oil-Inmersed Trans formers." This document suggests procedures to be followed in filling with oil those large transformers vhlch are customarily shipped to the point of installation without the oil. 9. A NEMA1 Standards Document titled "Proposal for American National Standard on Mineral Insulating Oil for Use in Electrical Apparatus" is in the final stages of ^reparation. It is intended "to provide a functional Industry standard that will assure a continuing supply of a single mineral insulating oil that adequately meets the needs of equipment manufacturers and oil refiners." It sets functional limits for the essential physical, electrical, and chemical properties and specifies the ASTM test methods to be used. Presumably it will ultimately replace items 2 and 3 above. 10. FSS-GSA2 Federal Specification W-l-530a "insulating Oil, Electrical (For Transformers, Switches and Circuit Breakers)." This standard specifies the physical properties required, establishes quality assurance provisions, and gives Instructions for testing. It identifies several ASTM documents as forming a part of the specification. 11. FSS-OSA Federal Specification W-l-1219 "Insulating Fluid, Electrical (Noncombustible)." This standard applies to "liquid of the chlorinated aromatic hydrocarbon type (askarels)" with coverage similar to that of W-l-530a. ^National Electrical Manufacturers Association, 155 E. *** Street, New York, New York 10017. ^Federal Supply Service, General Services Administration, Washington, D.C. 20^*05 U7 HONS 204803 The following standards relate to the specifications and performance of the transformer itself: 1. ANSI C 5T.12,00-1973 (also IEEE #U62) "General Requirements for Distribution, Fower, and Regulating Transformers." This standard specifies (l) service conditions; (2) preferred ratings; (3) insulation classes and dielectric tests; (**) tolerances on losses, impedance, ratio, regulation and temperature rise; (5) tests; (6) such construction details as bushing class, marking of terminals, nameplate data, etc.; and (7) short circuit characteristics. 2. ANSI C 57.12.90 (also IEEE #262) "Test Code for Distribution, Fower, and Regulating Transformers." This standard prescribes methods for performing the tests specified in C 57.12.00-1973 (above) and in the separate transformer standards of the C 57.12 series. The test methods covered are: (l) resistance measurements; (2) electrical insulation; (3) losses and impedance; (U) ratio and regulation; (5) temperature rise; (6) insulation power factor; (7) polarity and phase relation; and (8) short circuit calculations. C 57*98, an Appendix to C 57.12.90, covers impulse tests. 3. C 57.12.90S-1971* (also IEEE #262A) (Draft Standard) "Distribution and Power Transformer Short-Circuit Test Code." - This code defines a procedure by which the mechanical capability of a transformer to withstand short-circuit stresses siay be demonstrated. Other standards and guides in the ANSI C 57 series (there are about 30) include such documents as: C 57.12.20-1971* "Requirements for Overbead-Type Distribution Transformers 67,000 Volts and Below" C 57.12.21-1969 "Requirements for Fad-mounted Compartmental-Type Single-Phase Distribution Trsnsformers" C 57.12. **0-1967 "Secondary Network Transformers, Subway and Vault Type (Liquid Immersed)" C 57.100-19T1* "Thermal Evaluation of Oil-Immersed Distribution Trsnsformers" C 57*93 (slso ITO(A TR-5-1956) "Guide for the Installation and Maintenance of 011-Immersed Trsnsformers" C 57.92 "Guide for Loading Oil-lamersed Distribution and Power Transformers" !*. NEMA TR1-197U "Transformers, Regulators and Reactors" This document identifies the ANSI transformer standards which have U8 HONS 204804 been approved an NEMA standards and includes sure detailed specifications regarding auch characteristics aa audible sound levels, external clearances between live parta, accessories, and some test procedures. 5- NEMA TRll-1967 "Small Power Transformers with 65C Average Winding Rise and Distribution Characteristics," This covers electrical characteristics and mechanical features of this class of transformers. The above are representative of about ten ITEMA standards dealing with various classes of transformers. 6. UL (Underwriters Laboratories) 506 "Specialty 'Transformers" This standard covers requirements for air-cooled transformers and reactors for general use but does not cover oil-filled units. 7. AAR (Association of American Railroads) SM262 "Specifications for Impulse Transformer for Coded Systems Control." 8. AAR SM165 "Specifications for Transformer, 011-Immersed, Self- Cooled." - 9. FM (Factory Mutual) 5-27 "Fire Prevention Transformers." 10. FM 5-275 "Arc-Furnace Transformers." 11. FM lb-8 "Inspection and Maintenance of Transformers." (Note: Items 7-11 above are not In hand and thus no further description Is available). Section A.3.2 should be consulted for designation of the standards and techniques used for specifying and measuring the various fluid material properties Identified In the above transformer standards. A.1.6 Rtferencts, Section A.l Much of the material In this section was obtalnsd through personal conversations with the manufacturers and users cited In the Introduction. In addition to these personal communications, the following sources were found to he helpful. 1. In Re: Proposed Toxic Pollutent Effluent Standards, (SPA) FR1 Dec. 27, 1973; testimony on behalf of: I Federal Register !*9 HONS 204805 a. General Electric Co., Mr. George B. Farnsvorth, March lU, 1971* b. Westinghouae Electric Carp*, Dr. James H, Wright, March 15, 197U. ii. "The Role of Polychlorinated Biphenyls in Electrical Equipewnt," General Electric Co., Schenectady, N.Y., Feb. U, 1972, (internal publication). iii. "Study of the Potential Impacts of the Proposed Toxic Substances Control Act, as Illustrated by Senate Bill S. 776," for Manufacturing Chemists Association, Washington, D*C., by Foster D. Snell, Inc., June 26, 1975- It. National Conference on Polychlorinated Biphenyls, sponsored by Environmental Protection Agency, Chicago, Illinois, Nov. 19-21, 1975. Tne foil wing specific references have been cited in the text: 1. "Tran-,former Askarel Inspection and Maintenance Guide," Monsanto Company Bulletin No. IC/FF-38R, revised March 1975. 2. R. N. Sillars, Electrical Insulating Materials. IBS Monograph 1U, - 1973; Section 10.2.1. . 3. 49CFR1 100-199, revised October 1, 197**. U. "Oil Pollution Prevention," Environmental Protection Agency, 38 FR 3U16U, December 11, 1973. 5* C 57.12.U0; American National Standards Institute. 6. P. Barkan, et. al., "Overpressure Phenomena in Distribution Transformers with Low Impedance Faults: Experiment snd Theory," F 75 U6U-8, ieks PES Sumer Meeting, San Francisco, Calif., July 1975. 7. E. A. Goodman, L. 2upon, discussion to Barkan paper (7 75 U6U-8). 8. M. A. Nettleton, "Explosions Due to Faults in Electrical Equipment," Electrical Rerlev. July 25, 1975, pp. 11^119* 9. D. J. Rietuccia, R. E. Benton, "Ten Most-Asked Questions on Violent Transformer Failure," Transmisalon and Distribution, January 1975, pp. 30-31. 1Code of Federal Regulations 50 MOWS 204806 10. A. E. Knovlton, editor-in-chief, Standard Handbook for Electrical Engineer!. Ninth Edition, McGraw-Hill Book Co., Hew York, 1957, 4-561, 4-570. 11. National Electrical Code, 1975 j Sec. 450-24, "Oil-Inaulated Transformers Installed Indoors," NFFA, Boston, (also see Ref. 10, Sections 15-102, 4-572, 14-300). 12. National Electrical Code, 1975 j Sec. 450-23, "Askarel-Insulated Transformers Installed Indoors," NFFA, Boston. 13* D. A. Duckett, "Catastrophic Explosion Tests in Insulating Fluids," RTB Corporation, Waukesha, Wisconsin 53186, November 9, 1974. M0NS 204807 51 A. 2 "Askarel-class" Capacitors A.2.1 Description of the "askarel class" of capacitors Aa in the case of transformers, there is also a vide ranee in types and sizes of capacitors vhich are used in electrical apparatus. These are used in electronic circuits, for electric energy storage, and for pover factor correction purposes. Their ratings can range from a few oicofarads to many microfarads, with voltages from a few volts to over one-hundred kilovolts. Although some high-frequency capacitors use air insulation, most capacitors are made up of multiple, thin layers of metal and solid insulation. To eliminate any weakening by the presence of air, these Layer assemblies are then totally Immersed in a liquid dielectric1; often paper is used as one part of the solid insulator, and this paper is impregnated by the liquid. In addition to being a good electrical insulator, the liquid should have a high relative permittivity (high dielectr constant, e/cQ). The PCBs possess both of these features, and so grades of capacitor askarels have been used in capacitors since the introduction of the askarels in the 1930's. Whereas only a small percentage of transformers contain askarels, nearly one-hundred percent of pover and industrial liquid- filled capacitors use this type of liquid insulant. . Because of the widespread use of PCBs in capacitors, Monsanto in 1971 introduced a more environmentally compatible formulation, Aroclor 1016, especially for capacitors. In 197k, Monsanto sold 21,9?? thousand pounds of this fluid; the capacitor market for PCBs is considerably greater than the transformer PCB demand. The market value for PCB-insulated capacitors is in excess of $100 million per year. Askarel capacitors are often divided into two categories, "power" and "industrial." The power capacitors are high voltage units which generally contain several gallons of fluid. They are used for power factor correction to increase the efficiency of the electric power system. Additional uses in the "power" category include dc filters, energy storage, and Induction heating capacitors. "Industrial" capacitors are generally ftmalhwr-units which serve a large number of purposes including motor start and run, and fluorescent light starting. They are also used in arc welders and pover supply filters. The industrial class of capacitors is estimated to use one-half to two-thirds of the demand for capacitor askarels. ^Except for a few capacitor types, e.g., mica capacitors, which use only solid insulation. HONS 204808 ?2 Capacitors s..i both categories are t,-inert. I j.y constructed by winding alternate sheets cf metal foil and paper and/or film dielectric in round or flattened rolls. These rolls are connected together inside a closed metal housing vhlch is filled vlth an askarei. The paper dielectric material has a high permittivity, but it is dried before assembly into the capacitor, and it contains a large volume percentage of open air space. The askarei impregnates the paper, filling the air-space voids, Askarels are superior to mineral oil for this dielectric application because the high permittivity (e &t9) reasonably matches the paper, and therefore causes less non uni fomuty in electrical stress. The net high permittivity and the high dielectric strength of the askarels also allow for a smaller capacitor for a given power or energy rating. This combination of high permittivity and high dielectric strength alone would suffice to make askarels favorable impregnates for capacitors. Our survey of manufacturers and users indicates that they generally have also come to depend on the stability and nonflammability performance of askarel capacitor fluids. Since power capacitors are often out-of-doors or in vaults, however, it appears that more flammable fluids may in the future be more readily accepted. This may not be true with the industrial capacitors which are often installed in potentially flamnable environments, for example, fluorescent light fixtures and appliance motors in buildings. This flammability issue is further discussed in A.2.3- In considering test techniques for alternate capacitor fluids, then, we can at this time identify the following factor? as significant: dielectric constant (relative permittivity, c/c Q ), dielectric strength, impregnability, flammability, and stability. Toxicity, environmental degradability and compatibility with other capacitor materials are added to the list in subsequent sections. ' There is sn important economic interdependence between the capacitor and transformer fluids markets which should not be overlooked in deciding on alternatives to PCBs. The capacitor and transformer askarels have different percentsge contents of various PCBs, but they are both formulated from the same basic stock. They, therefore, aid each other in achieving the cost advantsges of large scale production. Any shift of either market away from askerek would influence the price of the askarels for the other use. This factor should be considered in evaluating the relative cost advantsges of substitutes for PCBs. A.2.2 Probability and consequences of rupture The rate at which the small industrial capacitors experience a tank-rupturing failure la very low; end when one of these small capacitors fails, there Is little loss of fluid.1 The total amount of fluid in all small industrial capacitors is, however, large, as was mentioned previously. ^In a capacitor containing paper insulation, 80S of the fluid is absorbed and so will not spill out even if the can does rupture. 53 HONS 204809 In the case of the larger, power-type capacitor, the failure rate has been historically low. Generally, a capacitor will fail by an internal arc short-circuiting the unit. Each power capacitor should be individually protected by a current-limiting fuae^ which will blow when the capacitor fails, leaving the capacitor-fuse unit open-circuited. If the industrial capacitors contain more than 3 gallons of "flammable" liquid, the National Electrical Code also directs that these be enclosed in a vault if indoors or within a fenced enclosure if outdoors (NEC-1975, Art. 460-2), Often, power capacitors are used in parallel-series connected groups, or banks, and upon failure of one capacitor in the bank, all parallel capacitors will try to discharge through the failed unit. A current-limiting fuse on each unit limits this "in rush" current in order to protect the other capacitors in the bank.2 NEMA standards^ exist for identifying the proper fuse to use to protect any given capacitor. These are somewhat out of date, however, so a manufacturer will generally provide the necessary information for his capacitors. There is need, therefore, for a newly created set of industry-wide standards for protection of new capacitor designs with new types of lmpregnant dielectric fluids. A.2.3 Full capacitor arc testing We have established in Section A.l that nonflaivability is sn important characteristic of "askarel-class" transformers and that the insulating fluid has been in the past and should continue to be qualified by full transformer arc testing. We have also recognised the importance uf nonflammability of capacitors, but it must be reported that there is little documented experience or interest in arc or flame testing of capacitors. The apparent Justifications for this lack of testing are the low rate of capacitor failure and the small amount of fluid in any one capacitor. Fuse protection as described in A.2.2 also reduces concern for rupture and fire. ^The National Electrical Code calls for "overcurrent protection" on any capacitor operating at potentials greater than 600 V (NEC-1975; Art. 460-25). 2 Power capacitors on distribution feeders are often group fused. 3CP 1-19T1, "Shunt Capacitors," National Electrical Manufacturers Association, New York. See also ANSI C55.1-19^6 (IEEE No. 18), "Shunt Power Capacitors." 54 HONS 204810 In evaluating the need for full unit flsnaability testing of capacitors, one should also keep in nind the folloving factors: s.. New candidate capacitor dielectric fluids generally are more flammable than askarels. b. The paper and film, which make up a large percentage of the bulk material within a capacitor, are generally flamble. c. Power factor correction capacitora are generally used not singly but rather in large multiple-unit banks. A study should therefore be made of the probability of fire spreading from one faulted unit to other capacitors in the bank. d. Use of many capacitors in large, closely spaced banks also has another risk. Unless each capacitor is properly fused, all parallel branches of a bank can discharge through a single unit which develops an internal, short-circuit fault. Fusing requirements of capacitors with new fluids should therefore not be overlooked. Manufacturers of fluids and capacitors have privately run occasional ire tests on capacitors to study flamiability effects after a tankrupturing fault. No definitely recomenaed format for such testing has ,, been brought to our attention, however, and the indication is that considerable design and testing would be required before a quantitatively interpretable and repeatable test could be devised. Several parties involved in the manufacture and use of capacitors did, however, feel rather strongly that such a test should be developed, particularly in light of the increased flammability of the new fluids. A.2.U Effect of alternative fluids on design and testing of the full capacitor unit The dielectric fluid oust be compatible with all other materials with which the fluid makes contact inside the capacitor casing. As in the transformer situation, the inside coating of the natal capacitor tank must he noareactive with the fluid. Since the fluid permeates directly to the metal foil electrodes, either the metal and fluid must he compatible or the metal must he treated with a protective coating. In capacitors, the most serious material coordination problem is between the fluid and the Insulating sheets which separate the foil electrodes. If this is a paper, then the fluid must be able to completely Impregnate the paper, leaving no voids in which corona discharges can be initiated. Although certain fluid characteristics such as molecular weight and viscosity will help to screen out obviously Improper fluids, adequacy of impregnation is best evaluated finally by testing the impregnent fluid in a true fils^electrode capacitor geometry. 55 MQNS 204811 One should be on the vetch not only for immediate corona from inadequate impregnation, but also for weakening of the dielectric strength of the fluid, which can result from the fluid dissolving impurities out of the solid insulation material. Incompatibility of a new fluid with the existing film iua paper may be solved by selecting new paper and film rather than by rejecting the new fluid. Thus, tests must be devised so that combinations of fluid and solid films can be evaluated on a controllable and comparative basis, and criteria for interpreting these tests must be devised in order that new fluids are not rejected merely because of inadequacies in the paper or film. The best practical combinations of fluid and solid must be compared. In the previous paragraph, we sought the best combination of impregnant and paper from a materials compatibility standpoint. It is important, however, to also coordinate dielectric characteristics. Severely mismatched dielectric constants will cause very uneven electric stress which can lead to corona and flashover in the more highly stressed member. Thus, askarel is a good fluid to use with paper since both have relatively high dielectric constants. The plastic films which are now in use in capacitors, such as polypropylene, have lover dielectric constants, and are therefor* more compatible with lover dielectric constant fluids. Using lower dielectric constant materials means that tna capacitance per unit area of electrode 1> less for a given electrode separation. Larger voltage or site are therefore needed for the same energy storage. Regardless of what material it selected, its dielectric constant should be accurately known under actual operating conditions, and its dielectric constant should be measured with the material in the true operating electric field. This may suggest that an extension be made in the ASTM standard test for dielectric constant (D 92b) The size of the capacitor is influenced not only by the dielectric constant, but also by the dielectric strength. Dcreaaed capacitance and energy storage becauae of decreased dielectric constant can potentially be regained by higher dielectric strength materials which allow higher fields and voltagea. The design of the capacitor ia, therefore, influenced by tb dielectric strength (breakdown strength) of the fluid. The breakdown strength of the fluid itself may be measured by standard teat procedures, as described in Section A.3.2. final capacitor designs, and teat capacitor geometries can be evaluated by methods outlined in Section A.2.5. The operating voltage of a capacitor is generally set by its corona inception and extinction levels, rather than by the flashover voltage. The low current, but continuous corona discharge causes damage to solid and liquid dielectric material and therefore leads to short life and failure of capacitors even though the operating voltage may be well below the breakdown level. Detection and measurement of corona and particularly of its inception and extinction points are therefore 56 HONS 204812 extrsmely important in ordsr to specify the maximum safe voltage for long life. Theee corona levele are not characteristic of Juat the insulating fluid, itself, however, but rather of the fluid in the actual electrode configuration. Including the electrode material, the microscopic surface conditions, and purity of the electrode/fluid interface. An absolute measure of corona inception for a pure, isolated fluid is therefore of little value, and one must instead determine the inception level for the whole electrode/insulator configuration, assembled according to routines which would be followed in manufacture. Corona measurement techniques are specified, for example, in ASTM D 1068-73. In addition to corona, losses in capacitors result from resistive and dielectric properties of the fluid. These can generally be evaluated for the pure, isolated liquid as described in Section A.3.2. The measurement should duplicate operating conditions (frequency, temperature, electric field) as closely as possible. Generally, higher dielectric constant materials exhibit higher loss, because of the large and rapid molecular polarization changes required by high permittivity materials in ac fields. A final important characteristic of the capacitor dielectric fluid is its stability. It must show little change in dielectric, chemical or physical characteristics under actual electrical and thermal operating conditions over the many years expected life time of the - capacitors. Frequent comments from fluids manufacturers, capacitor . manufacturers and users brought out (l) the importance of this characteristic, (2) the --ifficulty in defining and measuring stability, and (3) the importance at tne measurement being in ss realistic an environment as possible, l.e., in a teat capacitor configuration. In conclusion, this section has shown that the particular dielectric fluid chosen for a capacitor has a major effect on the total capacitor design. Consequently, and of more major importance to this study, we can further conclude that all aspects of new capacitor dielectric fluids finally need to be evaluated In an environment which simulates realistic operating conditions. A.2.5 Test techniques sad standards employed by manufacturers end users to evaluate fluid-filled capacitors The capacitor manufacturers end users have the following standards documents available for guidance in selecting sod handling the insulating fluid. 1. AHSI C 59ill (also ASTM D 2283-71*) "Standard specification for Chlorinated Aromatic Hydrocarbons (Asfcarels) for Capacitors." This document gives detailed specifications of the physical, chemical and electrical properties of four askarela which have been used as capacitor impre^iants and details test methods. 57 HONS 204813 2. ANSI C 59.123-1970 (also ASTM D 2297-68) "Standard Specification for Continuity of Quality of Electrical Insulating Mineral Oil for Capacitors and Cable Accessories." This document provides detailed specifications of the physical, chemical and electrical properties of the oil. 3- ANSI C 59-122-1970 (also ASTM D 2296) "Standard Specification for Continuity of Quality of Electrical Insulating Polybutene Oil for Capacitors." This document provides detailed specifications of the physical, chemical and electrical properties of the oil and details test methods. 1*. ANSI C 107.1-197^ "Guidelines for Handling and Disposal of Capacitor- and Transformer-Grade Askaralt Containing Polychlorinated Biphenyls." This dociawnt gives the typical physical, chemical and electrical properties of askarels used as capacitor lmpregaaats, details methods of shipping and handling, spells out necessary safety precautions, recommends labeling practices and details approved disposal techniques. It also lists organizations having facilities for analysis and for disposal and details an analytical procedure for the determination of airborne PCBs. ' - The following standards relate to the specification and performance of the capacitor itself: 1. ANSI C 55.1-1968 (IEKI No. 18) "Shunt Power Capacitors" Definition of terms, ratings, tolerance, operation, production and design testing, fusing. Such charseteristics as stability, corona start, radio influence voltage, overvoltage, capecltance, leakage, loss and dielectric withstand are identified, but the specific detailed test techniques art not described, nor ere appropriate ASTM procedures cited. This standard is for shunt power capacitors without reference to the dielectric fluid; with nonasksrel fluids, certain specified values and limits might possibly have to change, and specific test procedures may not be the same as have been traditionally used for aakarel capacitors. 58 HONS 20481^ 2. ANSI C 55*2-1973 "Series Capacitors for Transmission and Distribution-line Compensation" This standard includes requirements for safety, rating, and gap settings; functional requirements for protective equipment, alarm devices, servicing, type, and routine tests and a guide for operation. 3- NEMA CPI-1971 "Shunt Capacitors" This standard is similar in coverage to ANSI C 55.1* 1*. EIA1 RS-392 "Fixed Paper Dielectric Capacitors for Alternating Current Applications" This standard covers the requirements for oil-paper dielectric capacitors hermetically sealed in metal cases for general purpose application on ac voltages. It specifies (l) standard designations, (2) standard test conditions, (3) marking, (U) quality assurance test programs, (5) tests, and (6) applications suggestions- Although the term "oil" is used, this standard is meant to apply to askarel-filled capacitors. 5. EIA RS-392-1 (AH3I C83-67a-197>*) "Fixed Paper Dielectric Capacitors with Kon-PCB Impregnant for Alternating Current Application" ' - This standard appllts to capacitors with paper dielectric impregnated with non-PCB fluid. Bote that thert is no existing EIA standard covering non-PCB capacitor fluids with plsstic film dielectric. 6. EIA RS-bOl "Paper, Paper/Film, Film Dielectric Capacitors for Power Semiconductor Applications" The coverage of this standard is similar to RS-392 above. 7. UL 8l0 "Capacitors" (power factor correction capacitors up to 15 kVar or 600 V) This standard doss not cover capacitors intended for use as parts of applltncss, motors, transformers or slsctrlc-discharge-lsap ballasts. It does not specify that the liquid used ha nonflamable, hut does require that the capacitor ha marked to "Indicate whether the liquid Is combustible or nonflwable. If the liquid Is combustible, the amount of the liquid in gallons shall ha indicated" ^Electronic Induetrits Association, 2001 Eya Street, Washington, D.C- 20006. ^Underwriters Laboratories, 333 Pfingsten Road, Northbrook, Illinois 60062- HONS 20^815 59 A.2.6 References, Section A.2 Much of the materiel In this section was obtained, through personal conversations with the manufacturers and users cited in the Introduction. In addition to these personal connnunications, the following publications and reports were found to be helpful: 1. R. H. Munch, "New Capacitor lapregnantsConference paper presented at IEEE Power Engineering Society Meeting, New York, New York, Jan. 30, 1975. 2. B. H. Goldy, V. 0. Solberg, "A New Liquid Dielectric for Capacitors," Inaulatlon/Clrculta. Jan. 1975. 3- John Lapp, "Concepts in Systems Testing of Dielectrics in Capacitors," IEEE Trana on Power Apparatus and Systems, Vol. PAS 91*, Jan/Feb. 1975, PP- 63-71. 1*. j. h. Wright, "Consents of Vestlnghouse Electric Corporation." Testimony at hearings regarding "Proposed Toxic Pollutant Effluent Standards" (ZPA) (FR Dec. 27, 19731 given March 15, 1971*- 5. L. L. Jackson, "Industrial Organic Chemicals as Alternative Dielectric Fluids," Conference paper presented at IEEE Power Engineering Society Meeting, New York, New York, Jen. 1971*. ' 6. Statement of Electronics Industries Association, to EPA, concerning Proposed Toxic Pollutant Effluent Standards, in a letter to the Hoaorefele Bussell Train, Administrator, June 25, 1975- HONS 204816 60 A.3 Dielectric and Insulating Fluids A.3-1 Available Fluids In this section ve shall treat the major materials available at the present time or potentially available to the electric power industry as transformer and capacitor fluids. The power industry in general has viewed with reluctance any change from the use of askarels in certain transformer and capacitor applications because the askarels possessed outstanding electrical characteristics. Turtherswre, the askarels have low flammability, and, while not inexpensive, their cost has been at a level the industry could menage. It must be emphasised that there is no other fluid now available which hoe quite so broad a range of application and usefulness in the electric power Industry. The major fluids proposed as substitutes for the askarels in power transformer and capacitor applications are for transformers: (a) silicone liquids, (b) mineral oils, (c) highly saturated paraffin oils. For capacitors, the fluids ore (d) diaryl sulfones and mixtures, (e) long chain esters of phtbollc acid sad (f) alkylated monochlorodiphenyl oxides. The Japanese are reported to be working an a diarylalkane substitute, and the French on "ChloraJJQrlvie," an isopropyl chlorlnsttd biphenyl. There are other fluids which have high dielectric strength and excellent electrical properties such as the fluorocarbons snd the perfluoroalkyl furans. These are marketed for use In small devices. Their cost ($25 per liter, snd up) probably prohibits their use at the present time in large Installations such ss those that Interest us hers. A tsble listing ths properties of intsrsst of the fluids described here is appended st the end of this station (page 72). Before proceeding to the consideration of the nonaskarel fluids a few remarks on the oekmrels themselves may not he amiss. ^Reported at the Rational Conference on Polychlorinated Upheayls, Environmental Protection Agency, Chicago, Illinois, Kovenber 19-21, 1975- 61 HONS 204017 The Askarels (PoLyehlorobiphenyls) The askarels are proprietary mixtures of chlorinated biphenyls and chlorinated benzenes, tie empirical formulas of which are C H Cl 12 10-n n (vhere n_ nay run from 1 to 10) and Cg Hg_n C1R (where n may run from 1 to 6) respectively, These compounds may be represented by the general structural formula' a chiorobiphenyl chlorobenzene The proprietary mixtures mentioned, above usually are designated as having an effective percentage of chlorine and this percentage usually appears somewhere In their trade numerical designation. The singly chlorinated biphenyls are light oils snd the completely chlorinated (n * 10) biphenyls usually assume the character of a resin or wax. Although first synthesized nearly a century ago [l]1, it was not until the third and fourth decades of this century that they were produced in considerable quantities [2]. It was during these decades that the superior electrical properties of these materials became apparent. Within a few years, lsrge scale manufacture of the askarels was undertaken [3]. Besides the excellent electrical properties snd the low flammability of the askarels, it was observed that with the somewhat higher dielectric constant over, say, transformer oils, the capacitor dimensions could be reduced [4]. The askarel fluids have been used in a vide variety of applications outside the electric power industry, mostly as heat transfer fluids, additives to petroleum oils, in textile coatings and in surface coatings and paints [5]. Applications to sealing compounds, printing inks and papers, and casting waxes are all well attested [?] The variety snd successfulness of application of the askarel mixtures has led to widespread use and unfortunately to their widespread occurrence in the environment. Most of the areas vhere the polycblorobiphenyls occur in the environment are, of course, associated with eminently intense industrial activity. However, the high chemical stability of the askarels Insures their persistence wherever they are transported. ^References for Section A.3, shown in brackets, will be found in A.3.7. 62 MONS 204818 The result is that they are found at some polluting level in a great many places around the world. The polychlorobiphenyls have been documented as existing in North and South America in both marine and freshwater environments [6-10], A detailed discussion of these findings may be found in Kimbrough's comprehensive review paper [ll]. There would seem to be little point in expatiating upon the toxicity of the polychlorobiphenyls in tms brief resume since this report concerns the adequacy of tests of prospective replacements for these very fluids. Suffice it to say that the polychlorohiphenyls accumulate in, persist in, and are destructive to animal tissues. There is In fact a ' large body of literature supporting the toxicity of PCBs [see for example 11-15]. More recent reevaluation of poisoning incidents seem to indicate, however, that another substance, polychlorinated dihenzofuraa (PCDF, formed by oxidation of PCBs during thermal aging) was also present and may have significantly contributed to the toxic action [l6]. Silicone Liquids Probably the foremost contenders among the silicone liquids as s substitute for the askarels are mixtures of the dimethyl siloxsne polymers. These mixtures can he made up in a variety of viscosities. One of the most important properties of the dimethylsilicone liquids is their low temperature coefficient of viscosity [17]. They are also mechanically resistant to high shear rates [l8, 19]. These compounds are generally inert and are resistant to oxidative and thermal degradation [20, 21 ]. The dimethylsllicose liquids may be compounded to have viscosities running from 1 to well beyond 30,000 cent 1stoltes The dielectric constants for all such mixtures remain in the neighborhood of 2.7 over many orders of magnitude in frequency [22] and over a bread range of voltages [25], The flaBehlllty of the various mixture* exhibiting this viscosity range may be muted by the range of the open cup flash points from about 37 C for a 1-centlatolts fluid to about 320 C for a 200-centiatoke fluid [22, 23]. The solid-liquid transitions for the mathylalloxane liquids that interest us here, occur for the moet part in tha neighborhood of -SO to -U0 C [22]. Tha dielectric strength of these fluids Is in the neighborhood of Uo kV/0.25 cm [22]. The dissipation factor is shout 0.0003 at 100 Bi. All in all, these fluids are nearly ideal as power transformer fluids. They do not sludge or oxidise easily and with proper design one obtains good heat transfer. They offer good resistance to acid and alkaline contaminants in s transformer. Offsetting these virtues somewhat is the fact that when these fluids are subjected to arcs their dielectric strength may be reduced by the formation of solid SiOg filamentary bridging. In addition there seems to he a possibility of silicon carbide, SiC, being formed in an arc in these 63 HONS 204819 liquids. Silicon** nrbide is conductive and refractory, and provides another impediment to the self-healing of the dielectric liquid [2U], However, it is claimed that self-healing does occur if the liquid is impregnated into kraft paper. These drawbacks are much more detrimental to applications in capacitor than in transformers and explain why the liquids are proposed mostly for transformer use. Since the breakdown products of the silicones resemble the products from which they are synthesized, a few words should perhaps be added here about the chemistry of silicone liquids. What is given below is not meant to represent current proprietary methods of silicone liquid production hut to give elementary background material for discussions elsewhere in the report. Early expectations that silicon would possess a structural chemistry analogous to carbon organic chemistry were never fulfilled. Saturated silicon chains analogous to the paraffins can Indeed be formed but the silicon-silicon bond is relatively weak. At the beginning of the century Alfred Stock prepared the first few members of the silicon hydride series, by dropping [see 26] acid on magnesium silicide: MggSi H Cl SiH^ Si2 Hg + Si^ Hg + . . . + MgClg . He found that the hydrides were unstable and he was not able to prepare any compound in pure state higher than 31,3.. [see 26]. lbs modern silicon compounds which are of interest here^ao not rely on the Si-Si bond which Is subject to oxidation but upon the stable Si-O-Si bond. From this bond true polymers may he constructed as was evidenced by the thorough investigation of compounds of this class by Kipping in the first half < f the century [see for Instance, 2T]- We might pursue one possible par.h to synthesis of a dimethyl silicone. If we treat sand (SiOg) at about 1000 C with chlorine gas and s reducing agent (carbon) an important product will be silicon tetrachloride, a corrosive liquid at room temperature (SiCl],). Use of a methyl Grlgnard reagent (CH~MgCl) will allow attachment of methyl group# to the silicon atom: Stri^ 2 CH3KgCT * (CH^ SiCl2 + 2 MgClg . This last step Is performed commercially by causing silicon to react with methyl chloride at 250-300C with a metal catalyst Si + 2(CH3C1) 2e?tC ' (CH3)2 - Si - Cl2 MOMS 204820 The dimethyl silicon dichloride is asst useful 'because vs simply allow it to react with water: CH 3 CH 3 Cl - Si - Cl + 2H20 * HO - Si - OH + 2 RC1 ii CH 3 CH 3 This product is called dimethyl disilanol and will undergo a dehydrative condensation, thus: ?> ? C - OH + HO Si - OH * HO - Si - 0 - Si 11 J CH, CH, 33 CH, 3 CH3, giving us a primitive dimethyl silicone. A glance at the equation will suggest to the reader that longer chains and cyclic compounds can be built up from our simple silicone, and such is the case. Of course the last reaction may be reversed by altering the conditions, i.e., may yield silanols from silicones, a point of interest in the study of the manner in which these compounds may degrade In the environment. These compounds are considered combustible but they have no true combustion points. Combustion can be sustained if the heat supplied to the compound is sufficient to cause depolymerisation or decomposition. Thus, only the short chain polymers really burn [25, 28]. The low toxicity of the silicone liquids is claimed as an advantage of their use in the applications which concern us here. Indeed there is little in the literature to refute the contention that the material has extremely low toxicity, and numerous toxicity studies have been conducted at the behest of the asaufacturers of the silicones. Mineral Oils Most of the oils used as transformer fluids are complex mixtures of paraffinic and aromatic compounds (as inhibitors) of mineral origin. Many these paraffinic compounds are both saturated and cyclic. Such cyclic compounds are designated as "naphthenic" by the petroleum industry. 65 HONS 204821 The oils are nearly always the result of a process where the refining method is tailored to the final use of the product. The aromatic content of the oil must be controlled in order to prevent sludging or oxidation of the paraffin components; but some aromatics are added aa oxidation inhibitors. A typical additive which is compounded with the oil as an antioxidant is ditertiary butyl paracresol (DTBP or DBPC) which frequently is added to the extent of a few tenths of a percent [2U]. Such antioxidants are largely confined to distribution transformer applications. The use of such compounds is largely avoided in power transformers where the fluid is hermetically sealed into the transformer under a blanket of nitrogen gas. The addition of aromatic compounds is also designed to inhibit gassing in an oil under the operating conditiona of a transformer. The low cost and self-healing properties of mineral oils coupled with their satisfactory electrical properties have contributed to their widespread use. These oils have a dielectric constant of about 2. The dissipation factor is about 0.001 at 60 Hz. The dielectric strength of mineral oils is generally greater than 30 kV/0.25 cm, and 1012 ohm-cm is a typical value for the resistivity of a fresh oil. The toxicity of oils composed entirely of saturated paraffins is very low. Highly saturated oils find perennial use in cosmetic products. Some of the highly refined oils are pharmacopoeal. The presence or addition of aromatic inhibitors increases tht toxicity of the oils. The degradation of a pure oil in a normal atmosphere probably follows the course of oxidation of alkanes to, ultimately, carbon dioxide and water. Such a process would require high-energy radiation (ultraviolet) for initiation of the reaction. The mechanism of this process is not yet clearly understood because of the paucity of knowledge of slow reactions [31]. All paraffins are to some slight extent soluble in water and can therefore in time react with compounds present in natural waters. Furthermore, volatile fractions of an oil may escape into the atmosphere where if they are excited by energetic radiation they may form free radicals which readily react with atmoepherlc components. Whatever the details of the case. It Is believed that tba degradation products whether produced by physical or biological means are not highly damaging to the environment. The flaanability of the mineral oils is considered high since they are hydrocarbons with no quenching atoms (such as halides). If the oil is to be suitable for transformer use it must be both insulating and non-viscous at operating temperaturee since it will be used as a convective heat transfer fluid. However In general, the less viscous an oil, the closer It is to its flash point so that a compromise between hasard of fire and efficiency as a beat exchanger is required. These considerations have led power companies to tht use of oil in transformers almost exclusively is exterior installations. If a transformer operates with the insulating liq^iid at a temperature far below its flash point the hazard from fire would seem to be minimal. HONS 204822 66 There is some advocacy (see for example Duckett RTE Corp., [30]) of using high-purity high-flash-point paraffin oils in inside installations, much as askarels are used at present. The test used to suggest the suitability of oils is discussed in the next section. Hydrogenated Paraffin Oils One suggested fluid for transformer use is a product which the RTE Corporation of Waukesha, Wisconsin is proposing as a substitute for askarels in transformers in locations proximate to or inside buildings. The firm maintains that the high flash point and consequent low flammability of this oil so much reduces hazard from fire in interior installations that the fluid should qualify for service in such installations. This fluid is not produced by the RTE Corporation itself. The supplier is a large oil company [2$] and the fluid was designed, or expected to be used, in transformers which operate at temperatures high enough to reduce its viscosity to an acceptable level. The product is marketed as "RTemp.'' The composition does not appear to be anything unusual in the way of petroleum oils. It is a high-viscosity (800 cs at 25 C), high-flash-point (296 C) oil with a dielectric strength of about 37 kV/0.25 cm, (ASTM D87T), and high resistivity, l(j oha-cm. The oil is a petroleum product twice subjected to a hydrogenation process, thereby acquiring its saturated paraffinic - properties. RTemp is therefore a highly stable mineral oil which is claimed . to be of food grade [30]. It is presumed that the environmental characteristics of this highly saturated oil would be similar to the phamacopoeal mineral oils [30]. The general physical and electrical properties of this oil resemble the ordinary transformer oils. The flinmeblllty properties of RTemp trsnsforaer oil are better than most transformer oils as would be indicated by its high flash point. The expected lower flammability is confirmed in a general way by the surge-cell tests conducted at RTE Corporation by Duckett [30.]. The surge of electric energy into the oil caused vaporization of the oil and dispersal of the oil outside the test apparatus. Ho fire resulted, however, because the temperature of the liquid oil was well below the combustion point. This behavior was in marked contrast to an ordinary transformer oil subjected to the seme test which ignited immediately and did not extinguish, possibly because the transformer oil was initially closer to its flash point than the RTemp end considerably lesa energy was required to bring the transformer oil to combustion temperature. Oils of the highly saturated type are used in cable and underground applications. In spite of the high viscosity of such oils they are acceptable at the canon operating teaperatures of transformers In temperate, or even moderately cold, weather. But in the severest cold thsss oils require the protection of enclosure or burial to maintain their fluidity. Although the RTE surge test* show, and the combustion data predict, that these highly saturated oils are probably safer than ordinary transformer oils, the national Electrical Code Committee will probably need independent testing before explicit acceptance of this fluid. 6t HONS 204823 Mixture! Containing Diaryl Sulfonea These compounds are proposed by one of the major chemical companies of the country as substitutes for the aakarels in capacitor use. The basic chemical structure is: It n A SO 2 ItA where the It is an alkyl group with probably 1-8 carbon atoms, runs from 1 to 3 and A is a phenyl, naphthyl or indan group. All these compounds are halogen free [35]. The actual product is presumably a proprietary mixture of such compounds and other ingredients such as non-chlorinated biphenyls. Cited as in example in German sod U.S. Patent disclosures [32] is a mixturs of tolyl xylyl sulfeme, isopropylbiphenyl and minor ingredients. It is presumed that in use further ingredients will be sdded to these compounds as antioxidants The mixtures as proposed have a dielectric constant of about 6. No dielectric strength figures are available at present, but life-tests of capacitors are continually being run by the manufacturer of the fluid [33]. The fluid is tested in ballast-capacitor use att 1000 volts at 90 C working temperature. The flash point of the mixtares nos advocated by the manufacturer is about 150 C, combustion sustained at about 165 C and the fluid will auto-ignite in the region of I*50 C. The termlnal-to-case resistance is listed as '3 x 10? ohm, the dissipation factor as about 0.003 [31*]* No pourpoint figure is available but the capacitors function veil at -50 C [33]. Some toxicity studies have been conducted by the manufacturer. These studies indicate that irritation can be induced by intense or prolonged contact of the mixture with eye or akin of rodents. The manufacturer recommends that normal industrial hygisnic precautions be taken (attainable, in the main, with soap and vatsr) In order to assure safe handling. These materials when fed to rodents broks down Into excretshle metabolites according to studies conducted by, or on order of, the manufacturer. Polychlorobiphenyls fed under similar conditions persist in the *"~lbody after the aulfons mixtures in the tissue have become undetectable. A toxicity statement is available from the medical department of the manufacturer [ 35 ]. HONS 204824 68 The Phthalate Eaten The pbthalates can be offered immediately as products for use in the capacitor industry because they are already supplied by the petroleum industry as plasticizers for polyvinyl chloride plastics in tonnage quantities. In 1972, about 1 billion pounds of some 20 different phthalate esters were manufactured [36]. The generally excellent electrical characteristics of the esters of phthalic acid make them natural candidates as substitutes for the askarels in selected capacitor use should the substitution become necessary. They must be limited to low voltage applications, however, because of the problem of corona extinction in the phthalate esters. The phthalic anhydride is typically manufactured from naphthalene which is oxidized in the presence of vanadium oxide to phthalic anhydride or alternatively from o-xylene by a similar oxidation process to the same end product The phthalate esters supplied as insulating liquids are mostly diisononyl phthalate (DU?) and dioctyl phthalate (HOP) [37]. One major manufacturer prefers DIB? over DOP because of the expected resistance to hydrolysis bestowed by steric hindrance. The chemical and petroleum companies use various methods for producing the alcohols from which DOP and DIOP are formed; all of which start from petroleum refinery products. They may be typified by the one reaction for 2 ethyl-hexacol which begins with propylene to form s aldehyde which in turn, by aldol condensation, yields a Cg aldehyde. The Cg aldehyde is hydrogenated to 2 ethyl-hexanol. 69 HONS 204825 CO., H- , H, C3H6 -- 1 \ "** Sau.ttW =8 ""W. -J- 2 thylheiajiol. The 2-ethylhexanol then esterifies the phthalic acid C*0 0 + (2 R-OH) / C*o producing the veil-known plasticizer 2-diethylhexyl pbthalate. The octanols and isononanols used in the manufacture of insulating fluids will use the same or similar reactions -n their preparation. The aromatic dibasic pbthalate esters typified by diiaononyl and dioctyl pbthalate have specific resistances in the neighborhood of 10* obm--cja. In applications as capacitor fluids they offer a dielectric constant of about 5- The voltage breakdown point occurs at-around 30 kV/0.25 cm. The boiling point at reduced pressure (5 am Hg) is about 250 C, and yet the poufpoint is in the neighborhood of -50 C. The flash point as determined by the Cleveland Open Cup Method for these fluids is in the neighborhood of 220 C; fire point is determined by shoving that combustion is sustained at 250 C or thereabout [38]. The dibasic pbthalate esters are available on the market, apparently in any quantity desired and are marketed ty at least two major chemical companies. At least one equipment manufacturer has modified the design of Mi capacitors to accommodate the pbthalatea as capacitor fluids [3T]> Engineers from another capacitor manufacturer report that thair commercial use of tha pbthalate estera has been in lov voltage electronic unite for the export market. The report by Rutkovski and Forster [38] presents a sundry of toxicity atudias indicating lov toxicity and there is work supportive of the barmlesaneas of the pbthalate* [39, Uo]. For detailed studies of the toxicity of the pbthalatea see [1*1]. 70 HONS 204826 Butylated Honochlorodinhenyl Oxide One of the largest chemical companies in the country has for some years been engaged in a Joint effort with an electrical pover equipment manufacturer in the electrical and environmental evaluation of butylated monochlorodiphenyl oxides. This effort has resulted in the offering of a product to the electrical industry which may be used in high voltage power capacitors [k2]. The butylated monoehlorodipbenyl oxides are offered as direct replace ments for PCBs, i.e., no change in capacitor dimensions is rsqulred. The electric properties are similar to other dielectric fluids offered. The dielectric constant is about 5 and the dielectric strength about 35 kV/0.25 cm. The resistivity is about lO^1 ohm-cm; the vapor pressure is low. At room temperature it is about 10"4 torr. The viscosity is about 10 centistokea at room temperature. The pourpoint ia about -J*5 C. The flash and fire points are considerably (about 100 C) above the operating temperature, and the flash point is above that of some of the lower-chlorinated askarels. The values of flash and fire point are listed as 17U C sad 199 C respectively by the msaufacturer. Prolonged reliability tests have been performed, and the manufacturer has indicated a villlniesi to produce the material at the million poundper-annum level [U2], Safety and performance tests vsre, as of December 2, 1975, still being assessed. The manufacturer states that the major constituent of the fluid is known to be nontoxic to fish. The minor constituent is slightly toxic to fish but is also the K>re biodegradable. Solubility Is low ms is bioaccumulation. An environmental impact statement may be obtained from the manufacturer (1*3] 71 HONS 204827 Nnuid Propartes,. dielectric COQftlBl dialactrl* nrinu 8T/0.25 a llialpailai fitter ctoluc, t*cpxiat fin nlat coc. *c Him niIiIItUi Ukinli 5.8 35 0.00. 180* *0 ~ao J x 101* Mineral 011a 2 Suatrr of typical Propartlaa of A*fctrtla 011a and Propoaad S*plaeant fluid* Silicon* Liquid* Anri Sulfonate Mi arturea Phthalat* Eatar DllP Butylittd Mroochlorodlpbenyl Ox Ida fifthly aaturatad paraffin oil 2.5 i k.66 t t-k.T <100 *C) 22 Ox 001 '190 MTO 44 h0 .0003 (100 to) an --b -60 --b 0.09 150 168 ' -80 1*" 10* K>" 30 0.05 `30 895 -a 10* >55* *g.oi (100 la) 1T8 199 -85 1011 <100 Xx) 37 <0 09 (60 la) 296 321 -21 1013 an -wriM ttm qmliWUn * Ural i -- --a *r 1b MHllar i Dili Vila lx tkx taUx, uii>> ftr I tra MiImop nlM wt *180 flaUi mar la tfcx 9xUa M M> 80* -- * n> 290-hl0 --* b--rat atctad br raafactunr tm fluk pstn* 1 in tfpUal or IWi Ilua. Ifeot of too nail* tloo. n--0>m*f Mi ftp* ppwptpA pmmppp tar ppoolfU HONS 204828 A.3.2 Standards Relating to tha Testing of Available Fluids Three compilations of ASTM test procedures relating to the performance of electrical Insulating liquids have been Issued: 1- ANSI c 39*2 (also ASTM D 117) "Standard Methods of Testing Electrical Insulating Oils" 2. ANSI C 59-62 (also ASTM D 901) "Standard Methods of Testing Askarels" 3- ANSI C 59-118 (also ASTM D 2225) "Standard Methods of Testing Silicone Fluids for Electrical Insulation." ASTM test procedures called out In these documents are listed on the next page. The above standards deal vlth the fluids currently In coaaon useInitial testing of a nev candidate liquid would logically be done using the procedures already proven appropriate for existing liquids; however, it Is to be expected that some procedures would not be satisfactory and that nev procedures perhaps unique to a particular fluid would have to be developed. As illustrations of this situation note (1) that a different test procedure is requl**d for the detection of inorganic chlorides In askarels than in oils and (2) that the test procedure for determining the water content of askarels and oils is not acceptable for siliconesIn fact, the tabulation indicates that agreement within ASTM on standard test procedures for several Important characteristics of silicones has not yet been reached. 73 MONS 204829 Acidity, Approximate Acidity, Approximate, and Polar Contamination Coefficient of Thermal Expan sion Color Chlorides, Inorganic Chlorine Compounds, Hydrolyzable Dielectric Constant Dielectric Strength Using Metal Disk Electrodes Dielectric Strength Using VDE Electrodes 2,6 Ditertiary-Butyl Par a- Cresol Content Examination, Visual Plash Point Fire Point Gas Content Gas Content (Nonacidic) Inorganic Chlorides and Sulfates Interfacial Tension Neutralization Value (Neutralization Number) Oxidation Stability Peroxide Number Pourpoint Power Factor Refractive Index (and Specific Optical Dispersion) Resistivity Sampling Saponification Number Scavenger Content Sediment and Soluble Sludge Sludge Specific Gravity Sulfur, Corrosive Thermal Stability Viscosity Visual Examination Volume of Oil Water Content ASKARELS (D901) D9TU D1903 D2129 D1821 D102O, D2UU1 D92U D877 D92 D97l*. D66U D97 D92h D1807 D1169 D923 D1701 D1810 D1936 D08, Di*U5 D1702 D1808 D1533 ELECT. INS. OILS (OUT) D153^ D1902 D1903 D1500 D878 D877 PI816 DIU73 D152U D92 D831 D182T D878 D971 D661t, D97U D2UUo* D1563 D97 D92h D1807 DII69 D923 D9U D1698 D1313, D131** D1298 D12T5 D88, DUU5, D2161 D1315, D1533 SILICONE FLUIDS (02225) D2129 V92k D92 D9T> D97 D92U D1807 DH69 D1298 DUU5, D2l6l D9l*3, D1931*, ud D2112 alto deal with oxidation stability. D31U6 is called out in "Standard Specification for Oxidation -- Inhibited Mineral Oil for Use in Transformers and Circuit Breakers." 7U MGNS 204830 The following related ASTM Standards have recently appeared: D 3300 "Standard Method of Test for Dielectric Breakdown Voltage of Insulating Oils of Petroleum Origin Under Impulse Conditions" D 3303 "Standard Method for Rapid Gas Chromatographic Estimation of Higher Boiling Homologues of Chlorinated Biphenyls for Capacitor Askarels" D 330U "Standard Method for Analysis of Environmental Materials for Polychlorinated Biphenyls" (This document is a modified version of Appendix B of AMSI C 107.1-197L). A.3* 3 Test Techniques By Which Electrical Properties Are Evaluated a. Dielectric constant and power factor. Techniques for measurement of the dielectric constant of insulating fluids are well documented and pose no problems if applied with care- ASTM Standard D 92k "Standard Method of Test for Power Factor and Dielectric Constant of Electrical Insulating Liquids" outlines test procedures and specifies suitable test cells and ASTM Standard D 130 "Standard Methods of Test for A-C Loss Characteristics and Dielectric Constant (Permittivity) of Solid Electric Insulating Materials" outlines the theory and describes typical measuring circuits. b. Dielectric Strength. While not suitable for determinations of intrinsic dielectric strength, two designs of test cells have gained acceptance for ac measurements of the dielectric breakdown characteristics of insulating fluids. ASTM Standard D S7T "Standard Method of Teat for Dielectric Breakdown Voltage of Insulating Liquids Using Disc Electrodes" specifies a cell using 25-** am flat disc electrodes (with square edges) spaced 2-5 mm and outlines acceptable test procedures. ASTM Standard D lSl6 "Standard Method of Test for Dielectric Breakdown Voltag* of Insulating Oila of Petroleum Origin Uaing VDE Electrodes" specifies a call with a propeller for circulating the liquid and having brasa electrodes, the opposing surfaces of which are spherical In ahape and separated 2.0U mm (with an optional spacing of 1.02 os if the voltage limitation of the teat transformar requires It). A third cell end test procedure is specified in ASTM Standard D 3300 "Standard Method of Test for Dielectric Breakdown Voltage of Insulating Oils of Petroletn Origin Under Impulse Conditions." Alternate slectrode configurations are specified as follows: (l) sphere-to-sphere, using 12.7 sh brass or steel spheres and (2) point-to-sphere, using a 12.7 nm brass or steel sphere and a steel point, with a 0.06 am radiui of curvature. While concern haa been expressed by seme users regarding the desirability of placing greater emphasis on cleaning the test cell, on care in sample preparation, and on the fact that the electrode material 75 MONS 204831 may slightly influence the results, these standards are generally considered satisfactory for the purpose- c. Resistivity. ASTM Standard D II69 "Standard Method of Teat for Specific Resistance (Resistivity) of Electrical Insulating Liquids" specifies the test cells and procedures for measurement of resistivity using direct voltage. The theory and measuring equipment no he used is specified in ASTM Standard D 25T "Test for DC Resistance or Conductance of Insulating Materials." These standards appear adequate for the purpose. d. Partial Discharges (Corona). ASTM Standard D 1868 "Detection and Measurement of Discharge (Corona) Pulses in Evaluation of Insulation Systems"; IEEE Standard W5L "Recommended Practice for the Detection and Measurement of Partial Discharges (Corona) During Dielectric Tests"; and NEMA CP-P2-19T3 "Teet Procedure for Measurement of Internal Partial Discharges in Capacitors" are similar documents. They deal vlth a phenomenon vhlch relates to the total system rather then to the fluid alone; however, the corona characteristics of different fluids are often compared independent of their operating environment in particular apparatus. A.3-U Test Techniquss By Which Fluid Flsnability Is Evaluated Most would agree that the concepts of combustion and flamabillty of a material are qualitatively clear and easily understood. The quantitative definition of flammability of the material, that is, whether it will ignite and than continue to burn under specified temperature, pressure and ambient atmosphere conditions is not, however, as easily achieved. What is required Is a flaawblllty test in vhlch all relevant parameters are rigorously controlled and In vhlch the results are unequivocally Interpretable, and this has been difficult to develop to everyone's satisfaction. The various tests which are recognized and in coaaon use are described below, along with coMenta on their applicability and shortcomings: a. Flash and Firs Points The only test specified by the standards-setting agencies (ASTM, ANSI, IFPA1, IEEE) for Insulating fluids, mineral oils and askarels, is tha Cleveland Open Cup teat for flash and firs points (ASTM D 92) The specifications are: Mineral oil: flash point of 1*6 C (in some areas 130 C); Askarel: no fire point up to boiling point. Beside the D 92 test, ASTM recognizes sevsrsl other variants for determining the flash point, vhlch, however, are not prescribed for Insulating fluids: the Tag Open Cup tester (D 1310), the Tag Closed ^Rational Fire Protection Association, 60 Batterymarch St., Boston, Mass. 02110 76 MGNS 204832 Tester (D 56), and the Pensky-Mnrtena Closed Tester (D 93). In all experimental arrangements the flash point is determined ty slowly increasing the temperature of the cup containing the liquid sas^le. An ignition source, placed above the surface of the liquid, is actuated at specified intervals- The flash point, by definition, is the lowest temperature of the cup at which the vapors above the liquid surface flash. The flashes are transient, because only the vapor phase burns, not the liquid. In addition to the determination of the flash point, the D 92 test specifies the procedure of increasing the temperature beyond the flash point up to the temperature at which transient application of the ignition source causes continued burning of the liquid. The lowest temperature at which this happens is the fire point. Determinations of flash and fire points give valuable indications of flaamability hazards, and should continue to be used. However, it must be recognized that developments in recent years have progressed significantly beyond the stage at which the standards for Insulating fluids were set, and also beyond the state-of-the-art prescribed by officially adopted ASTM standards. The following items should be considered before flash snd fire point stsndards are reccusnded for replacement fluids: (1) There has been a general trend toward closed-cup methods (e.g., D 93). In particular. Underwrite*s* Laboratories have favored closed-cup tests for soat time now. These generally give lover flash points than the open-cup variants. For example, the flash points ' of the Dow Corning 50CS (silicone) fluid are 30^ C sad 277 C as determined by D 92 end 2 93 tests respectively. (2) The ASTM Flash Point Comittee is moving toward recceending International adoption of an "equilibrium" method. Equilibrium methods specify Flash-Ho Flash determination at a series of constsnt (rather than continually increasing) temperatures. Since in equilibrium methods the entire system, sample and container, is at the ssme temperature, complications associated with beat transfer between the sample and the container are eliminated. Therefore ssmple site is arbitrary and much smaller staples csn he used. In addition, these methods do not depend on viscosity because convective currents within the ssmple are also eliminated. (3) Scientific analyses of fire-point determination are poorly developed. Therefore, the significance of the test Is not clearly understood. This point will be further discussed In connection with the Oxygen Index test.* b. The Oxygen Index Test In the Oxygen Index (01) test the test sample is placed inside a tube containing a mixture of oxygen and nitrogen gases. The gas mixture, normally at room temperature, flows gently upward. The ssmple is Ignited and the percentage of oxygen in the gas mixture Is decreased. 01, by definition, is the uercentmge of oxygen in the oxygen/nitrogen mixture at which extinction occurs.1 1A reviewer points out that In any event fire point is of less importance If the more conservative flash ooint is used In specifications. 77 HONS 204833 The specified ASTM 01 technique applies to 3olid samples only (D 2863). Various laboratories have made their own modification* of the test to apply it also to liquids, and have reported the resulting data, hut there i3 no uniformity of procedure. Sharply divided opinions exist as to the value of the 01 test toward a definition of flammability. For example, a limited polling of the industrial opinions in this survey show* that some manufacturers use the test for practical purposes, while others consider it "worthless." The reasons for this division are not difficult to see. On one hand, it i3 recognized that the burning of a small laboratory sample (a few grams) in room-temperature environments in no way simulates the behavior of large amounts of combustible materials in a well-developed fire. On the other hand, there are reasons why the 01 test should not be discounted in the context of this survey. Fundamentally, the test is the prototype of extinction of a diffusion flame, and it bears roughly the same relationship to practical fire-point determinations that the fundamental flammability-limit data do to practical flash-point determinations. As was discussed earlier, both flash and fire points are specified standards for definitions of flaanability of insulating fluids. The problem is, however, that fundamentally meaningful fire point, or 01, tests are more difficult to design than the meaningful flash point tests. To obtain consistent extinction data one would have to develop an 01 test for liquids and a fir* point tester in such a way that an 01 value of 21 (i.e., air) is obtained at the experimental fir* point. Recent work at the DBS Center for Fire Research shows that such development would require very careful design of the test geometry. Therefore, an 01 test for Insulating fluids should be recoeSMnded only after some additional study, allowing specifications of the testing arrangement. _ ' c. Spray Flanability Tests The ASTO Mist Spray Flaamability of Hydraulic Fluids test (D 3119-72T) could be applied also to electric insulating fluids, but we have no evidence that anyone has done that. The test is a very severe one, entailing the spraying of a preheated atomized mist into an open flame. Underwriters1 Laboratories consider it too severe and use on* of their own variants in its stead (UL File MB 01L66, p. Tl-6). The UL test consists of applying drops of the liquid to the surface of s heated steel plate, and observing the "behavior." Factory Mutual Research Corporation haa its own spray flamability test, akin to D 3119- It is also a very sever* one. Most hydraulic fluids do not pass it. Factory Mutual is currently using th* results of their spray flasBahility test to establish acceptable operating temperatures for aakarel substitutes. d. Autoignition Temp*ratuf* (AIT) Several techniques exist for the heating of liquids in air until they ignite. ASTM D 2155 i* * specified procedure for liquid petrol*TM products. Underwriters' Laboratories us* s similar t#at of their own. 78 HONS 204834 As one should expect, autoignition temperature* of electric insulating fluid* are quite high. Some reported value* are: 332 6C for a mineral oil and 393 C for Dow Corning 50CS fluid, both ty the D 2155 teat; 662-670 C for askarels hy the UL teat. As in the case of the spray flammability testa, operational conditions, normal or abnormal, are unlikely to give rise to these temperatures in the case of insulating fluids. The AIT test, therefore, does not appear particularly pertinent. (At this point,, it should be mentioned that Underwriters' Laboratories does not evaluate liquids only on the basis of tests which simulate some aspects of actual operations, but also on the basis of the total information about stability and flammability of the liquid. If this philosophy is adopted, then of course the AIT test and any other reproducible test which can be tied to some property of the liquid, becomes pertinent. In fact, UL did use the AIT test among others, to arrive at a numerical flammability rating of askarels). e. Energy-Release Tests Hazards tests designed to evaluate the amount of explosive energy released in an accident apparently have not been used ih connection with insulating fluids, but there are good reasons why they should be. Such tests almost Invariably give a measure of energy release by the effect on the surroundings damage to a target, displacement of a target , or generation of a blast wave in air. Since the hazard of an arc-initiated fire in a transformer or capacitor tank may be due to the blast wave damage as well as the subsequent fire, a measure of the explosive energy release capability of an insulating fluid would appear to be useful. Such a test would be particularly important if it could be shown to be a useful screening mechanism before going to the more expensive, full-unit tests discussed in Sections A.2.3 and A.3.3 (see also Sections 3-5 and 3-6). Direct measurement of the Dlast wave intensity may be the most valid energy release test, as described more fully In Section 3. The table on page 80 sumariies presently accepted flammability data on several Insulating fluids. Fluid flamiability measurement, by whatever test procedure is adopted, should be carried out under conditions as closely approaching operation as reasonably possible. In particular, pressure and temperature conditions are known and should be duplicated. Thus, power transformers of the "aakarel class" have a temperature increase of about 65 C under full load, so that bulk fluid temperatures will probably not exceed 100 C even on hot days. Local temperatures near the transformer winding may be ten to twenty degrees higher. Most utilities operate their aakarel transformers with a pressure relief device as prescribed by the National Electrical Code (Section 1*50-23); the relief pressure is usually only a few psi (20-1*0 kPa). Power capacitor tanks rupture at low pressures but small industrial type capacitors generally are sealed to 1*0-50 psi (280-3^0 kPa). The fluid temperatures within capacitors would not be much above the ambient, which itself might vary widely depending on location. 79 HONS 204835 un im * aM Hr Cunts* im nn-n , lit I Him PspST JOM 6 296] 01 tofart- maK 'ini' l*g) HI ** .*--H- I. uPyIKi IUU wf--r HI !**! am *TS 091M uh aaMj (sUfiM) 1U (130) IK us imnu--< pa Bli 1016 12U o1o36W*l 1) IT* [llMl *) m in Tiro Mlal (tl tf* U MiatMitita JttK. m UK U.l Mo Carols* i*r* db-m in mu k Ina iT Mm MM1 K Ct (fttiKtHuri) IT-l-lTH. IMrMUA (U) NMfl <61 TlM.) Mm liar noo. JO* 36* Jtl m U) S13J* m ut 13* JL BiaitM U ilr nr Ti Mbr wUi MKMiMli^rmlUTtM1l*i MtoitmHIku*iMetom. IlHf aiKM*6M(tAam( olin* iKMKIOflua* 1M iMMIUlUr' MM IM M> JO* Boo *e. Caomral*s*MJO1CKtITKT lom M Mo Carols*. all I^H MM mu Mmi KKUn IoiIm rota (m/mo. 10 M ! I inm) O.QKflf yt lift 0,60(0 6mm q 1. 1. . toita at 6, 9. CM, *TpUli aoO naoOUKr Tutoit* ( Larlwu/ lOUOHHUfll^ M' f <1**3I 2. 0. r. MO aat . MOIhKK Nor 6. Tima. MKortO OHoUrioMt*. !MnoaTlMrroa, M. oMtant.oJlaawOiTils3.6,.,41(MT*.lwm Mo CT* 111 0, 1*T* Mattr *. 1. F. MrM CMAimi, rMo6slTso. .MOMiio, lIilmlllMl. Iiffilloll*HT3i.ll*I6TTT]l,Mlfnoal. MNlllll Mm. 1*T3 DoNLs T-*1--------1 CUorl 80 HQ NS 204836 For a new fluid, vith different loss and heat transfer characteristics, these typicel values say have to be revised. A.3.5 Some Test Techniques For The Evaluation Of The Toxicity Of Dielectric Fluids In the past, toxicologists have not only made, the Caboxatoxy assess ments o& toxicity, but in many instances they have, modi the iinat judgment m to the. social couAbt to be. taken on tilt bail* oi a paxticulax set oi finding*. Instead, the technical experts should be changed with staining an objective independent determination oi the extent, nature, and ixequency oi adverse eiiects. They should be asked to explain the xetative gravity oj these ejects {on the target system, be it humans on wildtiie. Similarly, other qualified technical expects should be xequested to make an objective assessment oi the benefits oi use and oi alternative materials ox processes. However, the iinat judgment as to a txade-oH between an advexse health eiiect and a desired benefit is a social decision and should be made utith the paxticipation oi those utko axe afiected. This is not to say that technical expects using theix technical expertise will not participate, but it does state that they should not be the sole judges of determining the balance between the benefit and the xibk. "Teams such as 'toxicological insignificance,1 'saie,' 'ttco tolecance,' 'no eiiect level,' and 'negligible xibk' have been in rather common ate. Kit oi these contain in one way ox anothex value judgments ox technical implications which have no place in an objective assessment oi xisk. Thexe is no substance which, under cextain ciccumstances, could not be dangexous and unsa{e. Thexe is no battexy oi tests, howevex elaboxate, which can prove beyond challenge the complete saiety oi a chemical, fox the toxicologist to apply the texms 'toxicotogically insigniiicant' ox 'negligible xibk' to a set oi obsexvations makes a premature judgment in the wxong axena by the wxong ptxson as to insignificance ox acceptability. Attempt has been made to eliminate such texms ixom this xepoxt. Anothex common texm oi wide usage is the 'no-eiiect level.' This is statistically meaningless and thexeioxt oi limited value since it merely means that no effect was obsexved in studies using a gxoup oi animals orf paxticulax size. Such an observation is completely compatible with the presence o{ an advexse eiiect, which in iuxthex studies with taxgex sample sixes ox with diiiexent types oi observation might lead to a positive outcome. Ve peejee the usage oi the texm 'no obsexved efiect,' which should always caxxy with it a qualiiying statement as to size oi the gxoup in which no advexse eiiect was obstxved." We have opened this section with quotations froa tha report of the National Academy of Sciences, "Principles for the Evaluation of Chemicals in the Environment"- [Wt], as a forewarning of the complexities and difficulties ve may anticipate in looking at the field of toxicology. 81 HONS 20483? Our knowledge of the toxicity of insulating fluids results, in the main, from two sources. The first source is clinical observation. In this instance the observer, generally a physician, is confronted with a victim of a toxic condition with which he may have had no prior experience, and even if be is successful in treating the condition, the one event may well remain an isolated one to him. A more informative kind of clinical event (and more costly in human terms) oceurs when there is a mass poisoning such as the now famous Yusho outbreak of 1968 in Fukuoka, Japan. Here there were a sufficient number of casts for a local medical school to mount a full scale epidemiological study. The study group was successful in determining the origin of the poisoning, polychlorobipheryl heat transfer licuid which leaked into a food preparation vat*[ll]. It is evident that the clinician or epidemiologist must seize his opportunity end extract as much Information as possible from the case at hand before the opportunity is lost. The second source of our knowledge of toxicity comes from tnimal experimentation. In contrast to clinical studies, animal experimentation provides the kind of systematic data which may be acquired by independent observers and confirmed at will. This kind of experiment constitutes the main source of information that a developer of a dielectric fluid submits to government regulatory agencies. It will be this type of experiment which will be dealt with in what follows. _ It is possible to develop a definition of toxicity through the use of populations of animals a is done in the Federal Hazardous Substances Act [PL 86-613 and amendments, 15 U.S.C., CH. 30] where the definitions under Section 2 paragraphs (g) through (k) read: ' ` "(g) The tenm 'toxic' shalt apply to any substance 10then, than a nadioactive substance) which has the capacity to produce personal injury ok illness to man thnougk ingestion, inhalation, ok absorption thnough any body suKiact. (h) (1) The tenm 'highly toxic' means any substance which tails within any o{ the iollowinq categories (a) produced death within iounteen days in kali ok ihokc than kali i a group ten ok more tabonatony white eats each weighing between two hundred and thnee hiindeed gnome, at a single dose oi iHM miltianams ok less pee kitogsam oi body weight, when oKatly administered ok (b) produced death within iounteen fay* in halt ok moKt than hal{ oi a group o{ ten ok moee laboKatony white eats each weighing between two handled and thKee hundred gnome, when inhaled continuously ion a peniad ai one hour ok less at an atmospheric concentration oi two hundred pants pen million by volume ok less oi gas ok vapon oi two miltignams pen liter by volume ok less oi mist and dust, provided such concentnation is Likely to be encountened by nan when the substance is used in any seasonably ionesetable manner; ok (c) produced death within iounteen days in kali or more than kali i a group oi ten or more nabbits tested in a dosage oi two hundned mUlignams ok less per kiloqnan oi body weight, idien administened by continuous contact with the bare skin \ok twenty-io\m house ok less, (2) i{ the Secretary iinds that available data on human experience with any1 1A more recent evaluation [l6] suggests that a contaminant of the PCB rather than the PCB itself may have caused the poisoning. 82 HONS iO1*838 tubt-tonce indecatt rtAult* dl{{erent ^rom tho*t obtained on animal* An the above-named doAcgt* ox concentration*, the human Ahall take precedence. U) The term 'corroAive' mea.ru any AubAtance which in contact with living tl**ue will came de*truction o{ tiAAut by chemical nation; but Ahull not refer to action on inanimate Aux{aceA. [jl The term 1 wujtant' meant any tubt.fcutce not coxxOAlvt withen the meaning o{ Aubparagxaph U) which on iimetUate, prolonged, ox repeated contact with normal living tlAAue well induce a local xn{lamatory reaction, (k) The term 'Atrong AenAitlzex' meanA a AubAtance which will came on normal living tiAAut through an allergic ox photodynamic proceAA a hypeuenAitlvity which become* evident on xeapplication o{ the Acme AubAtance and which Ia deAigmted at Auch by the Secretary. Before deAignatinq any AubAtance at a Atrong AtnAitlzer, the Secretary, upon coruidexation o{ the frequency o{ occurrence and Aevtxify oi the reaction, a kali bind that the AubAtance hat a Aignifrcant potential {ox earning hyper* enAitlvity." Pros the operational definitions provided above one can by means of a few modifications, define other operational definitions such as are found in the Registry of Toxic Effects of Chemical Substances [1*5] of the Rational Institute for Occupational Safsty and Health: "TVLo-Toxic Vo*i Low - the loweit doAe o{ a AubAtance, at publUhed ox made available to pubti*h, introduced by any route other than inhalation over any givtn period o{ time and reported to produce any toxic e{{ect in man or to produce carcinogenic, teratogenic, mutagenic, or neoploAtigenic e{{ecl* in human* or animat. TCLo-Toxic Concentration. Low - any concentration o{ a AubAtance in air to which man or animal* have been expoeed {or any given period o{ time and that hat been reported to produce any toxic effect in man, or to produce a carcinogenic, teratogenic, mutagenic, m neoploAtigenic toxic e{{ect in animat* or human*. LVLo-Lethal VoAt Low - the lowtAt doAt o{ a AubAtance other than LV50 introduced by any route other than inhalation over any given period o{ time and reported to hone earned death in man or the lowest single doAt introduetd in one or more divided portion* and reported to have corned death in animal*. Entree* {or lethal doAt* tuhniniAtered to animal* {or the qualifying toxic doit, TXDS, were made in accordance with -the dote limit* o{ Table 1.1 See Ke{. 45, p. xviii 83 HONS 204839 LV50-Lethal Vose Fifty ~ a calculated don oi a chemical substance whxch <s expected to cause the death oi 50% oi an entire population ol an experimental animat species, as determined pujm the exposure to the substance, by any route other than inhalation, o{, a signiftcant timber ftom that population." Other lethal dose percentages, such as LD1, LD10, LD30, LD99, as well as others, may he published in the scientific literature for the specific purposes of the author. Such data would be published in the list if these figures, in the absence of an LD50, were the lowest published in the article. A substance may qualify by an LDLo published value even though a higher LD50 value, which may also exceed the limits of Table I1 was reported in the same publication. " LCLo-Lethal Concentration Low - the lowest concentration oi a substance, other than an LC50, in air which has been reported to have caused death in man or to have caused death in animats when they have been exposed iar 24 hours or less. Entries ior lethal concentrations ior animals ior the qualiiying toxic dose, TXDS, were made in accordance with the dose limits o< Table I.' LC50-Lethal Concentration Fiity - a calculated concentration oi a substance xn axr, exposure to which ior 14 hours or less would cause the death oi SOt oi an entire population oi an experimental animat species, as determined irom the exposure to the substance oi a signiiicant number irom that population. These entires quatifted in accordance with the dose limits set iorth in Table I.' See LV50 above ior other parameters which may similarly apply to entries ior the list." It has also coma to out attention that a scale developed by Hodge is frequently used by toxicologists [561. The complexity of the whole environmental system requires the examination of the effect of pollutants on more than one type of animal. It is generally considered desirable to Include experiments on birds, amphibians, and fish as well as mainals, in order to obtain an idea of the total impact of a substance upon the environment. The more responsible the concern for the environment, the more detailed will he the experiments, for ws are trying to extract from a few tailing experiments information applicable to virtually tbe whole of nature. To acquaint himself with Just the first notions of the complexity of the subject before us here, the resder might refer to a list of experimental animals used in toxicity experiments. Such a list is given herewith as Table A. It is taken from the Registry of Toxic Effects of Chemical Substances [1*5 Table III pxx]. The list runs the gamut from amphibians to man and will doubtless be expanded still further as toxicological studies proceed through the years. In addition to tha range of animals to be considered, there is also a variety of methods of administration as given in Table B, taken from [45 Table II p, xix]. This list may also be expanded as experimentation continues, "Intraocular" for example which does not appear on the list ^See Ref. 1*5, p. xviii HONS 204840 84 TABU A Specie* t>ird -- toy domestic or laboratory bird reported but not otherwise *--*< bird -- wild bird tpeda cat chicken - child dog donuatic animals: goat, iheep, bom duck (domestic) tog gwbil goiaMpic hiflism bonus infant -- speck* unspecified In lafitreao* mas monkey mOBM P*i ptgHS quail (laboratory) rabbit rat squirrel toad turkey womarn Order of Preference for Data Acquisition 22 23 7 11 17 1 4 12 19 20 13 9 10 1 l 13 1 6 3 8 16 IS 3 2 14 20 21 1 Designation brd bdw cat ctl ckn chd dog dors dek bt pb 8PI bam hmn inf WilH mat mfcy mu* P'8 pgn qal rbt rat sql tod trk wmn HONS 204841 85 TABLE B Route Intraarterial Intraaura1 Intracerebral Intracervical Intradermal Intraduodenal Inhalation Implant Intramuscular IntrapiaccmaJ Intrapleural Intnpentoncel Intrarenal Intratracheal Intravaginal Intravenous Ocular Oral Parenteral Rectal Skin Subcutaneous Unreported routes of administration to, or exposure of, ANIMAL SPECIES TO TOXIC SUBSTANCES Abbreviation tat ial tee 1CV idr idu ihl imp ims ipc tP* ipr irn itr ivf ivn ocu ml P rec tka sen unit Definition administration into the artery administration into the ear administration into the cerebrum administration into the cervix administration within the dermis by hypodermic needle administration into the duodenum inhalation in chamber, by canmilauan, or through mask placed surgically within the body -- location described m reference administration of dose into the muscle by hypodermic needle administration into the placenta administration erf dose into the pleural cavity by hypodermic needle administration into the peritoneal cavity administration into the kidney administration into the trachea administration into the vagina administration of dose directly into the vein by hypodermic needle administration directly onto the surface of the eye or into the conjunctival sac per os, intragastric, feeding introduction with drinking water administration into the body through the skin. Reference cited is not ipacific concerning the route used. Could be Ipr, scu, ivn, ipi, ims, in, or ice administration of dote by way of rectum to the rectum or colon in form of *-- suppontory application to the intact skin, dermal, cutaneous administration under the skin dose, but not route, is spedfted in the lefsrance 66 MQNS 204842 in this year's edition, but probably will soon as there already are such studies for silicones [1*6]. Another li3t of the effects observed (extracted from the same document) given here as Table C [1*5 Table VI p. xxiv] shows the almost endless variety of observations to be met with in the field of toxicology. In addition to the foregoing variables and constraints there are to be considered, age, sex, and periods of exposure and observation. The iaporatance of considering these last two parameters may be Judged from the National Acadeny Report [1*1*]. "Acute toxicity can be. demoted as exposu*e to a test ngent jo* 14 houns ok less. This debinition will pnovide jo* the 14-koux exposune pentad. often used in denmal pnocedunes and the 1 to t-koun pentads often used in the inhalation pKoc.tduA.ti. The ternn 'acute toxicity' has often been used to dtftnt the. immediate (usually 24 houns ok leu) ejjects oh an agent on the animat; it ii moKt appKOpKiate to use longen pentode oh obsenvation a.ftex. the single exposure xn ondtK to detect any delayed ejjects axising oven an extended pentad oh time. The 24-kouK pentad, oh obsenvation may detect pnimanity phanmeo dynamic activity, although otheK biological mechanisms might be so aliened as to count death ok otheK eftects Wcthin a longen pentad oj obsenvation. A pnolonged obsenvation pentad mould alio penmit detection oh ejjects that might be nelated to anatomical, biochemical, ok hematological changes that mould not cause immediate death. Acute toxicity studies one needed in safety evaluation jo* two distinctly diftenent neasonst one is as a seneening pnocedune to aid in identifying compounds oh such lout toxicity that, when considtKtd in Kelation to a paepostd use oh lam exposone, extensive investigations to mate a judgment oh safety one not justifted. The otheK is as a type oh acute toxicity pnocedune designed to delineate the speciftc toxic eftect, and mechanism theneoh, which might be associated with a massive exposun.t ok a nonmal use associated with a high level oh uposuKt to a cowpound. This latten type oh study is designed to assist in the development oh am appnopntate clinical management pnognam jo* individuals involved, in a misuse ok accident and to assist in appnatsal oh the sahety oh a paxticulan use involving significant human viposune, even though that exposure might be chncnic in natuu. Such studies can aid in identifying tanget oKgan ejjects which may be valuable in the pnopen design oh chnonic studies. Tn the seneening type oj acute toxicity pKocedune, the amount oj matexial to which an animal Is exposed is so massive that it genenaily beans no pnacticat Ktlatlonskip to the expected human exposunt. It is neasottabte to establish maximum exposone levels (e.g., 5000 mg/kg anally, 5000 mg/kg denmatly, etc.) jo* each acute pnocedune, which can be used to pnedict that the agent is essentially acutely nontoxic ij no advenst ejjects one detected. 87 HONS 204843 Abbreviation* ALR BCM BLD BPR CAR CNS COR CUM CVS DDP EYE GIT GLN IRR MMI MSK MUT NEO PNS PUL RBC SJCN SYS TER TFX UNS WBC TABLE C NOTATIONS DESCRIPTIVE OF THE TOXICOLOGY Definition* (not limited to effect* luted) Allergenic-systemic reaction tucb u might be experienced by individual* let--itifnj to penicillin Blood clotting mechanism -- any effect which increases or decreases elretiwg Blood effects -- effect on *11 blood elements, electrolyte*, pH, premia, oxygcncarrying or releasing capacity. Blood pressure effects -- any effect which change* any aspect of Not'd pressure away from normal -- increased or decreased. Caronogenic -- producing cancer -- a cellular tumor, the nature of which is fatal or ts associated with the formanon of secondary tumors (meustasa). Central nervous system -- includes effects such u headaches, tremor, drowsiness, convulsions, hypnosis, anesthesia. Corrosive effects -- buns, desquamation. Cumulative effect -- where lubstanca is retained by the body in greater quantities than is excreted, or the toxic effect is increaeed in seventy by repeated boduy insult. Cardiovascular effects -- such si when heart activity is increased or decreased through an effect on veatnek or auricle; fibrillation; or when the arterial or venous system is dilated or constricted. - Drug dapeadenee -- any indication of addiction or dependence. Eye effects -- irritation, diplopUa, cataracts, aye ground, NIndian* by eye or the optic aem. the Gastroiaiestiaal tract effects------ diarrhea, constipation, ulceration. Glandular affects -- any effect on the endocrine glandular system. Irritant affect -- any irritant effect on the skin, aye, or mucous membrane, Mucous membrane effects -- irritation, hyperplasia, tiliaxy activity r hinged Muscalo-skeietal affects -- inch u osteoporosis, muscular riagwntinn Mutation or mutagwuc -- tiansmiasibia *--g-- produced in the offspring, Neoplastic effect -- the production of tumors. Peripheral nervous system effects. Pulmonary tymm effects-- effects on respiialien and respiratory pathology. Rad Mood ceil affects -- indude* the severe! anemias. Skin affects-- inch as erythema, rash, senirtixatinw of skin, pisschitl hemorrhage. Systemic affects -- affects on the metebodc and excretory function at the fiver or Termognic affects -- noutnnsmisaibie changes produced in the offspring. Toxic effects -- used to introduce the pnndpel organ system affected m reported or the pathology. Unspecified effects -- the toxic effects were unspecified in source. White Mood ceil effects -- effects on any of the cellular units other dun erythro cytes. including aay change ia number or form. Editorial note: Death is, of course, another definable response. 88 HONS 204844 A decision process band on both parameters o j txpvuiiKxtaJUy derived toxic, tiiects and estimated exposure levels should be developed through subchronic studies in an eifart to identxiy then compounds and uses uthick merit chronic studies. In other words, as the knowledge oi exposure level change* and toxicity infarmation become* available, iurther testing way not only be desirable bat necessary. However, that (hint way bt circumstances other than those which warrant chronic studies is recognized. Structural characteristics may indicate the need oi a camnogeruc bxoaiiay. No short-term ox subchronic tests axe known that can give assurance oi the patience of possible carcinogenic activity, When a new chemical ii presented. fax evaluation oi toxicologic hazard, the chemical and physical characteristics oi the material must be considered so that appropriate steps can be taken to assure adequate exposure oi the animal to the chemical in the various procedures. The chemical stability oi the agent, vapor pressure, solubility, particle size, etc,, can be oj importance. Befare any judgment regarding health hazard can be made, it is necessary to know the intended use and the anticipated route, faequency, and level oi exposure to man. The type oi exposure being considered *s generally that likely to result iron the normal usage oi chemicals rather than faiom an abnormal situation Such as would result faom a spill- - [UU, p. 991. - A further complication la the requirement that adequate analytical chemical techniques should exist (for every conceivable compound and metabolite) to determine the administered quantity of a toxic substance, to Identify the type and quantity of any contaminants in the substance, and also to determine the amount and presence in the biological system. Analytical techniques, such as gas chromatography combined vith mass spectrometry, which can measure trace materials in water down to levels of 1-10 parts per trillion are available [l7l. Finally, the overall complexity of the subject can be Inferred from the criteria fbr inclusion of a substance in the Registry of Toxic Effects of Chemical Substances [U5, p. xvl]i "QpaliiyinQ Toxic Dost (TX5S). All toxic dosu appearing in the Registry toeu derived iron reports 04 toxic eiitcls produced by individual substances, A toxic eifacl *s defaced as any bodily injury -- reversible or irreversible; any tumor - benign or malignant; any mutagenic or teratogenic efaecl; or death which has been reported to have resulted faom exposure to a chemical substance vie the respiratory tract, skin, eye, mouth, or any other route. for humans the toxic tUect is any toxic eHect that was reported in the source rtiertnee. There is no qualiiying limitation to the duration oi exposure nor to the quantity oi concentration oi the substance, nor is there a qualiiying limitation oi the circumstances that resulted in the exposure. Regardless 0 j the absurdity oi the circumstances that were involved in a toxic exposure, it is assumed that the same circumstances could recur. HONS 204845 09 For anilnot* the toxic dose must be limited in order that the Registry wiIt have practical useiulness as a guide, to hazardous ncu to humans. The quatiiying symptomatology far animals cannot bt elucidated with any practicality' therefore, `mane, objective signs o& toxic eiiects muit be relied upon. The production o& tuition [neoplastigenescs I, benign or malignant (carcinogenesis); the production oi change* in the oifapning, whether transmissible (mutagenesis) or not (teratogenesis)} and the production o& death are the entente that are used a* the quatiiying toxic eiiects fan animat data. For a substance to be included, there is no limitation* a* to the time o& expo*ure, nor the quantity or concentration oi the do*e oi the *ub*tance reported to have caused mutagenic, teratogenic, neoptastigenic, or carcinogenic eiiects in animat*. Thi* i* became, the do*e-e&&ect relationship ha* not been correlated far animat* and man. It i* presumed that there is some potential far the occurrence in man oi eiiect* simitar to those noted in animats, ii man is exposed to these substances.'' We have now come to the point where we have seen the possibility of defining an experimental protocol in terms of populations of animals and we have seen a few of the effects which are looked for In the tables. It might be well for the interested reader to consult a review such as that of Kimbrough [ll], where an overview of the toxicity of a large number of polychlorinated polycyclic compounds la given. We see from Kimbrough [ll]. and from the Interdepartmental Task Force on PCBs [18] the following general results for the particular case of the PCBs. ' - 1. Response depends on route of application: inhalation, skin and oral techniques have been usual in PCB studies. 2. Response depends on species; a large number of animals have been used, and data also exist on human response. 3. A wide variety of specific reactions occur including liver: size, action, enzyme production; skin: chloracne, edema; reproduction; embryonic effects, gastrointentinal lesions, chemical porphyrogenic effects, neurotoxicity, iomuce response, death, with sufficiently large dose. . Response time; because of the stability of the PCBs, they tend to accumulate in specific body tissues and their effects are therefore generally slow to appear and are long lasting. 5. Analytic techniques; sources indicate the problem of identifying contaminants which probably accompanied the PCBs in controlled exposure tests. Because of the accumulation phenomenon, toxic level concentrations may be very low. In addition, regulatory personnel have suggested PCB standards as low aa 1 ppt (l part in 10;. Analyses at this level are routinely available in special situations [Vf]. A citation here of three paragraphs of Kimbrough's review should Indicate something of the methods, the range, and the vaatness of the 90 HONS 204846 difficulty of this type of study. From page U55 of Kimbrough [ll] we see: "lit a pheasant reproduction itudy with Aroclar 7254,*^'*^ a iignificantly greater difference in the number, of e<j<ji tiutX. weAe pepped but not hatched wa* found in the gKoup of ken* that received 50 mg of PCSa weekly. Egg production and hatchaluJUty wcu lower in thi* gnoqp, the *urvivat 0 j duckena that did hatch at 4-ot week* of age usu reduced, and the one* that did iurvive were iiqnrficantly tighter than the controls. In another not reproduction itudy94 the lowest doie oi Aroctor 1254 that affected reproduction tuu 20 ppm US mg/kg/day] which manifested itself in a decreased number oj Litter* and Its* pupa per litter. Neither 1 ppm nor 5 ppm oi Aroctor 1254 had cut e^ect on reproduction in the Sherman strain rat; however, an exposure ojj the darn to the Aroctor* at 5 ppm or higher increased the liver-to-body weight ratioa in weanling rati of both iexe*. At 1 ppm the increase in lover weight war observed only in Fla and Flb weanling male rat*. The dietary Level* 0(5 700, 20, and 5 ppm oi Aroctor 1260 did not affect reproduction; however, the Liver-to-body weight ratio of 21-day-old pup* wa* increased at alt exposure Level*. Hud1 are highly 41aceptibtc to the toxic effect* of PCS*. A daily dietary intake of 30 ppm resulted in death in about 6 month*. ExpoAure to 5 ppm Aroctor 1254 in tht daily diet Acvtnely affected reproduction. Unfortunately, it i* often very difficult to compare the variou* dietary level* that affect different iptcin, *ince the food consumption may vary decisively from one animal iptcin to another and the daily food intake in g/kg body weight i* often not included in tht icientific report. Some mice strain*, for Instance will be expoitd to a* much of a given iubstance that i* present in the diet at the concentration 0 j 300 ppm a* rat4 on a diet containing 600 ppm or more. With *omt iptcle* it i* of count difficult or almost impoiiiblt to meaiure the food conAumptuon, but a greater effort in establishing there very basic fact* would be helpful for the interpretation of toxic effect* in different Aptcie*." A number of additional U.3. Government Regulations ax1st giving furtbar operational definitions of toxicity. Among them are: 1. Department of Transportation - From CFR Titla 7*9 "Transportation" Part 173* 3*3. Claaa B nolson ("lass dangerous") sec. 173.37*3 liquid or solid __^ substance "bum* to be *o toxic to man a* to afford a hazard to health 91 HONS 204847 during transport, on which, in the absence o{ data, on human toxicity... {alt within any one o{ the {oilowing catcgoAces1: J| Oral toxicity- product death within 4t hours in hat{ on mom o{ a group o{ TO on mom white tabonatony rats weighing 200 to 300 grms at a tingle dote o{ SO milligrams on let* pen kitognam o{ body weight when administered onatty. Z) Toxmity ok inhalation: those which pnoduce death within 4t hours in hat{ on mone than hal{ o{ a group o{ TO on none white tabonatony nats weighing 200 to 300 grams, when inhaled continuously {on a period o{ one houn on less at a concentration o{ 2 milligrams on less pen Liten o{ vapor, mist, on dust, provided such concentration is likely to be encountered by man when the chemical pnoduct is used in any seasonable {oresteablc manner. 3) Toxicity by skin absorption: those which product death within 4t hours in hal{ on mone than hal{ o{ a gnoup o{ TO an mone rabbits tested at a dosage o{ 200 mulignams on less pen kilogram body weight, when administered by continuous contact with the bane skin {on 24 hours on less. The {onegoing categories shall not apply i{ the physical characteristics o{ the probable hazards to humans as shown by experience indicate that ' the substances will not cause serious sickness on death, neither the display o{ danger on wanning labels pertaining to use non the toxicity tests set {onth above shall prejudice on prohibit the exemption o{ any substances {nom the provisions o{ Pants 170-119 o{ this chapter. 129 FR 11753, Vec. 29, T96i, as amended by Arndt. 173-3. 33 Fit T49tt, Oct. 4, 1931; 33 FR J9I23, Pec. 27, 1931)." 2. Occupational Safety and Health Administration: Air contamination standards (CFR 29 1910.93) cites: "Threshold Limit Values of Airborne Contaminants for 1968," Aaer. Conf. on Goto. Industrial Hygienists. 3* Food and Drug Administration (FDA): FDA relies on EPA to establish tolerances. FDA also will act on any legitimately verifiable indication of toxicity of a substance; rather than having its own specific regulatory definitions of toxicity or lists of toxic substances. Tolerance levels for specific substances may, however, be established, as was done for PCBa on July 6, 1973 (38 FR 10095; also see 21 CFR 122.10) milk manufactured dairyproducts poultry eggs animal feed animal feed components fish infant foods 2.3 ppm 2.5 5 0.5 0.2 2 5 0.2 The reader will note the similarity between these definitions and those used in the Federal Hazardous Substance Act (15 U.S.C., Ch 30) discussed earlier. 92 HONS 204848 U. Environmental Protection Agency (EPA): A Federal Water Pollution Control Act, Amendments of 1972 (Public Lav 92-500 16 Stmt. 8l6); the term "toxic pollutant" is defined in section 506(13): 502(13) "The term 'toxic pollutant' mean* thou pollutants, ok combination* of pollutant*, including din cant-canning agent*, which after dinchange and upon expo*ure, ingention, inhalation m an*imilation into any organinm, either dietetic/ fKcm tht environment ok uidmectly by ingention through food chain*, wilt, on tht of information available to tht Adnunintrator, count death, dimane, bthavoKial abnanmaJUtien, canctK, genetic mutation*, phyniological malfunction* (including malfunction* in nepnoduction} ok phynical deformation*, in Audi OKganinmn ok theix offnpKing." Section 30h(g) directs the Administrator of EPA to publish test procedure guidelines: ' "304(g) The Adminintrator nhall within one hundred and eighty day* fxom the date of enactment of thin title, promulgate guideline* entablinhing tent procedure* for die analynin of pollutant* that nhall include the factor* which mint be provided in any certification puriuant to nection 40J of thin Act ok permit application punuant to nection 40t of thin Act." Test procedures for 30b(g) vere published in 38FS 28758 Oct. 16, 1973; for "chlorinated organic compounds (except pesticides)," gas chromatography is the approved tsst procedure. Section 307 directs the Administration to establish a list of toxic pollutants: "307(e) (a) (I) The Adminintrator nhall, within ninety day* after tht date of enactment of thin title, publink (and from time to time thereafter revinel a lint which include* any toxic pollutant ok combination of Audi pollutant* foK which an. effluent ntandand (which may include a prohibition of die dinchange oL Audi pollutant* ok combination of Audi pollutant*} will be titahlinned under thin nection. The AdminintratoK in publinhing nuck a lint nhall take into account the toxicity of die pollutant, it* pernintence, degradability, die unual or potential pKtntnce of die affected oKganinm* in any water*, die importance of die affected oKganiim* and die nature and extent of die effect of the toxic pollutant on Audi organinm*. (2) bithin one hundred and eighty day* after die date of publication of any H*t, or Ktvinion thereof, containing toxic pollutant* or combination of pollutant* under paragraph [ > I of thin nub*ection, the Adminintrator, in accordance with nection 553 oj title. 5 of die United Staten Code, nhall publink a proponed effluent ntandand (or a prohibitionI for Audi pollutant or combination of pollutant* which nhall taka into account die toxicity of die pollutant, it* pernintence, degradability, die uauo or potential pnnenct of the 93 HONS 204849 affected ongfuvuma in any waters, the xnportance of the attend organisms and tht nature and extent of the tffaU of the toxic pollutant on iuch organisms, and he shall publish a notice for a public heading on such proposed standard to be held within thirty days. As won as portable after such hearing, but not latex than six montiu after publication of the proposed effluent standard ton prohibition), unless tht Administrator finds, on the record, that a modification of iuck proposed standard (or prohibition) is justified based upon a preponderance of evidence adduced at Such hearings, such standard (or prohibition] shall be promulgated. (3) If after a public hearing the Administrator fads that a modification of such proposed standard prohibition] is justified, a revised effluent standard (or prohibition) for such pollutant or combination of pollutants shall be promulgated immediately. Such standard (or prohibition] shall be reviewed and, if appropriate, revised at least every three years. (4) Any effluent standard promulgated under this section shall be at that level which the Administrator determines provides an ample margin of safety,. (5) athen proposing or promulgating any effluent standard [on prohibition] under this section, the Administrator shall designate the category or categories of sources to which the effluent standard (or prohibition) shall apply. Any disposal of dredged material may ~ be included in such a category of sources' after consultation with the Secretary of the Army. (6) Any effluent standard (or prohibition] established pursuant to this section shall take effect on such date or dates as specified in the order promulgating such standard, but in no case more than one year from the date of such promulgation. (7) Prior to publishing any regulations pursuant to this section the Administrator shall, to the maximum extent practicable within the time provided, consult with appropriate advisory committees. States, independent experts, and federal departments and agencies." The "List" ni published according to proposed criteria in 38TO l80hl*, July 6, 1973 "Proposed Criteria...." "Pollutants were chosen for inclusion on the initial list in light of the following criteria: t. Vote from laboratory or field studies indicate that the pollutant could, if discharged into water, constitute a serious environmental threat. The types of data examined in making this judgment are set forth in Part 7 below. Z. The pollutant is discharged, or has the potential to be discharged, fwm point iouncti and constitutes a very serious environmental threat. 9U HONS 204850 3. Dalit axt available to establish tUluent ameeting the. Kequinement* oi the Act. 4. Standard *etting unden. Section 307 (a) is appnopniatt, and ii neces*any become the pno*pectxve timing and eHectcvenes* o(j abatement action* unden othen provision* oj the Act one not cormen*unate utcth the natane and *eniou*ne*i oi the pnoblem* identiced by the above criteria. " 1. Explanation oi Criteria "The iinst criterion concenn* the actual on potential damage that a oaten discharge oi the*e materials nay emote by virtue oi certain toxicological pnopertie*. Detailed toxicological data one examined to detenmine whether one on morn oi the&e type* oi eiiect* one fenouwi bioaccumulation, cancinogenic, mutagenic and tenatogenic eiiect*, and high acute toxicity. Speciiioally, the iollowing type* oi data one nevlevied: T. Available data concenning uihethen a *ubstance ok it* chemical, on biological iron* ioAmotion pnoduct i* bioaccuimlated thnough ' an aquatic mechanism to the extent that they ne&ult in: (a) Reproductive impairment in any important. specie*, on (b| Concentration* in iood *ouKces in exceii o{ applicable Fedenally established tolenance level*. 2. Data concenning carcinogenic, mutagenic, and tenatogenic e^ectA. Maternal* likely to be cancinogenic, mutagenic, on tenatogenic to man one iunthen evaluated a* to thein chemical, biological, and physical itability in water. 3. Data concenning high onden oi toxicity a* measured by ikont-tenm lethality tJUt*. In thi* connection a cla**i{ication a *ub*tance a* *highly toxic' to man accondinq to ctassijicatian* e*tabti*ked by the Department oi Tnan*poMotion and the Environmental Protection Agency (W""flT7934f Ftbnuany It, 1971, 40 CFR 16t.il and the Environmental Pnotection Agency and the Food and Drug Administration testing procedures (40 CFR 16t.t and tl CFR I9I.T0) ii particularly relevant. According to the neitnence* cited above, *ub*tance* which have an onat LP5C oi 50 mgjkg oj body weight on lei* on have a denmal LP50 oi 200 mg!kg o{ body weight ok le*& one deiined a* highly toxic to metmals and aKt examined iurther. The term 'onat LP50* mean* that iingle anal dcMe which is lethal to one-holi oi the test population within 14 dayi. The tenm `denmal LP5C mean* that do*e denmally abionbed in 24 koun* which i* lethal to one-holi oi the test population within 14 day*. HONS 204851 95 For the protection of aquatic. Hft, substances proposed to be classified as `highly toxic.' according to thi national Academy oj Sciences water pollution ranking system may oe considered fox inclusion in thi Hit. Sacn substances are thou which are toxic to aquatic life ai measured by a 96 hour LC50 orf 10 mg/1 o>i lea. Thi term `LC50` means that concentration of a Substance in water uhich ii lethal to one-half orf thi tat population in the specified tint periods. Teateng methods such ai thou found in `Standard Methodi fan. thi Examination o ^ Water and Wastewater, ' 13th Edition, pant 231, 1971, on. thesr equivalent utill be considered adequate." "Thi niond criterion conclAM thi seriousness o ^ discharges on potential diichangii of the pollutant fnom point iouncU. Relevant here one inch fadorn ai thi nature and extent of toxic iffecti aiiadated with the pollutant, the extent to which diichangii of the pollutant haul bun identified and thi production, distribution, and uii pattern of the pollutant. Vote mint be available fnom well documented field itudiu ihoming damage to important organiimi from diichangii of the compound into wateni, or technically sufficient to show the material hoi the potential to be environmentally harmful after being diichanged from point iouncei to the water." "The third criterion concern iitting effluent itandandi for pollutanti on the Hit. The following are representeteve of the categories oj date useful in itandand iitting for each material. . 1. Toxicity to man and other organiimi; 2. Carcinogenicity, mutagenicity, and teratogenicity date; 3. Tramport pathi of the material in the environment; 4. Hoaccumulation and bioconcentration; 5. Chemical, physical, and biological tram formation in the environment; 6. Reliability and accuracy of analytical procedures; 7. Chemical ckanacteriitia; S. Production and industrial or commercial utilization; 9. Sources of the pollutant to water; 10. Enviromnental incidents attributed to the material [fish hills, etc. I; 11. Pretence of the material in the environment, residue levels in. various organises, ambient concentratiom in riven, tabes, etc. It. State and Federal regulatory requirements concerning the material; 13. Classes and characteristics of waters into which the material is discharged." HONS 204852 96 "The iouxth cxitexion conctxns the ovexalt environmental eUeot i the contxol measures available, xnc/.uding the use oi substitute pxoducts, and possible eiiects upon gxound mter ox othex environmental media." Section 311(b) deals with regulating the discharge of oil and hazardous substances: "3/Mb) (I) The Congxess hexeby deataxes that it is the polity oi the United Statu that thexe should be no dischaxges oi oil ox hazaxdous substances into ox upon the navigable waters oi the Untied State* adjoining shorelines, ox into ox upon the waters oi the contiguous zone." There is at the writing of this report no Toxic Substances Act administered by the authority of the Environmental Protection Agency which defines the tests which should be adhered to by manufacturers of polychlorylbiphenyl replacements. The Agency is now developing guidelines for such tests and is consulting with industry on the matter. A.3.6 Test Techniques By Which Degradability Is Evaluated In this section we are concerned with the standards and tests which might apply to degradability of a compound in the absence of living organisms (physical or chemical degradation) or in their presence (biodegradsbility). We are not aware of standard tests for the evaluation of the purely physical or chemical degradation of transformer and capaeitor fluids. One must seek such information on a piecemeal basis and either design one's own experiment or make inferences from experiments in the physico chemical literature. Inference is the source of much of our estimate of the safety of certain products. An example of the kind of experiment which is relied upon to provide information on nonblological degradation is the study by Siegel and Stewart on the degradation products of silicones whan exposed to high energy radiation (1*9), The technique used by the authors was the exposure of a dimethyl silicone polymer to vacuum-ultraviolet radiation in a photolysis tube. At the conclusion of the photolysis the products were drawn off into a mass spectrometer where analysis was made, and thence the quantum yields in ethane and methane and the energy transfer were calculated. These results when coupled with knowledge of the chemistry of formation of the silicones give a clue to the effect of sunlight upon silicones, namely that the silicones yield degradable products. In a similar fundamental study of photolysis in the near ultraviolet of dimethysilicones, Delman,Landy and Simms [50] found from the use of two different activating lamps that if the ultraviolet radiation is energetic enough and if oxygen is present, silanes are formed. These findings were made under carefully controlled physical end chemical conditions. The conclusion of these results is that in sunshine and air the silicones will degrade to ailanols (S1-0H structures) and thence to silica and water. In the environment there is also water present, and one would expect all the more the formation of silanols. MOHS 20^853 97 Attempt* are under way in industry to provide store direct environmental testing of the dimethyl silicones to determine with certainty the degradation of these products but the studies will be long and difficult. There are, however, several standard methods of biodegradability measurement. Most of these depend upon the degradative activity of sewage sludge or selected bacteria. Processes range from the simple so-called River Die-Away test to the difficult continuous-activatedsludge test (51]- This last la specified by the lav of at least one European nation for detergent degradability tests [52]. The river die-away test consists of mixing a quantity of natural river water with a known amount of the substance under study. Periodic sampling allows the Investigator to follow the rate at which the original substance decays. There is also a well known, often used, batch activated-sludge test. It is similar to the river die-away test except that a prescribed amount of activated sludge is added. The sludge or culture used may be made highly uniform and is commercially obtainable (531 - Semicontinuous methods of sludge addition have also been devised [51]- These are more likely to involve pilot-plant scale equipment than the Mason Jar* called for in the River Die-Away and batch tests, - So far as we are aware the only test for biodegradahility which has the stomp of approval of an industry-wide coosittee, is that for sulfonate biodegradahility promulgated by the Subcooslttee on Biodegradation Test Methods of the Soap and Detergent Association [53]. This test, which defines what is meant by "biodegradable" for the purposes of the soap and detergent industry, is a detailed procedure for tasting sulfones by a modified semicontlnuous activated-sludge method. Defined in the test are solutions and vessels to be used, design of aerator, source of uniform bacterial culture, and directions for sampling and rate of aeration. The Soap and Detergent Association test has been applied to the phthalate esters with results that were gratifying to producers of the phthalates [55]. Appended to this section are two tests selected from the SPA Manual of Methods for Chemical Analysis of Waters snd Wastes 19T1* [51*]These tests were culled to illustrate common tests in the field. HONS 20*85* 98 BIOCHEMICAL OXYGEN DEMAND (5 Days. 20*C) STORE! NO. 0031 a 1 Scope and Application I 1 The biochemical oxygen demand test (BOD) is used for determining the relative oxygen requirements of municipal and industrial wastewaters. Application of the test to organic waste discharges allows calculation of the effect of the discharges on the oxygen resources of the receiving water. Data from BOD tests are used for the development of engineering criteria for the design of wastewater treatment plants. 1.2 The BOD test is an empirical bioassay-type procedure which measures the dissolved oxygen consumed by microbial life while assimilating and oxidizing the organic matter present. The standard teat conditions include dark incubation at 20*C for a specified time period (often S days). The actual environmental conditions of temperature, biological population, water movement, sunlight, and oxygen concentration cannot be accurately reproduced in the laboratory. Reaults obtained must take into account the above factors when relating BOD results to stream oxygen demands. Z Summary of Method 2.1 The sample of waste, or an appropriate dilution, is incubeted for 3 days at 20*C in the dark. The reduction in dissolved oxygen concentration during the incubation period yields a measure of the biochemical oxygen demand. 3. Comments 3.1 Determination of dissolved oxygen in die BOD teat may be made by use of either the Modified Winkler with Full-Bottle Technique or the Probe Method in this manual. 3.2 Additional information relating to oxygen demanding characteristics of waste waters can be gained by applying the Total Organic Carbon and Chemical Oxygen Demand tests (also found in this manual). 4. Precision and Accuracy 4.1 Eighty-six analysts in fifty-eight laboratories analyzed natural water samples plus an exact increment of biodegradable organic compounds. At a mean value of Z1 and 175 mg/I BOD, the standard deviation was 0.7 and 26 mg/I, respectively. (EPA Method Research Study 3). 4.2 There is no acceptable procedure for determining the accuracy of the BOD teat 99 HONS 204855 5 References 5.1 The procedure to be used for this determination is found in. Standard Methods for the Examination of Water and Wastewater, 13th Edition, p 489, Method 219 (1971). 20^856 HONS 100 DISSOLVED OXYGEN (Electrode) STORET NO. 00299 1. Scope and Application 1.1 The probe method for dissolved oxygen is recommended for those samples containing materials which interfere with the modified Winder procedure such as sulfite, thiosulfate, polythionate, mercaptans, free chlorine or hypochlorite, organic substances readily hydrolyzed ih alkaline solutions, free iodine, intense color or turbidity, biological floes, etc. 1.2 The probe method is recommended as a substitute for the modified Winkler procedure in monitoring of streams, lakes, outfalls, etc., where it is desired to obtain a continuous record of the dissolved oxygen content of the water under observation. 1.3 The probe method may be used as S substitute for the modified Winkler procedure in BOD determinations when It Is desired to perform nondestructive DO measurements on a sample. 1.4 The probe method may be used under any dreumstandts as a substitute for the modified Wmkler procedure provided th*> the probe itself fa standardised against the Winkler method on samples free of Interfering materials. 1.5 The electronic readout meter for the output from dissolved oxygen probes normally calibrated in convenient scale (0 to 10, 0 to 15, 0 to 20 mg/1 for exampit) with a sensitivity of approximately Oil5 mg/litsr. 2. Summary of Method 2.1 The most common instrumental probes for determination of dissolved oxygen in water am dependant upon electrochemical reactions. Undti steady-state condi tions, the current or potential can be correlated with DO concentrations. Intarfadal dynamics at the probe-sample Interface arc a factor in probe response and a significant degree of interfacial turbulence is necesmry. For precision performance, turbulence should be constant 3. Sample Handling and Preservation 3.1 See 4.1,4.2, 4.3,4.4 ustder Modified Winkler Method. 4. Interferences 4.1 Dfcsotved organic materials are not known to interfere in the output from dissolved oxygen probes. HONS 204857 101 4.2 Dissolved inor(Janie salts ire a factor in the performance of dissolved oxygen probe 4.2.1 Probe* with membranes respond to partial pressure of oxygen which in turn is a function of dissolved inorganic salts. Conversion factors for seawater and braclush waters may be calculated from dissolved oxygen saturation versus salinity data. Conversion factors for specific inorganic salts may be developed ex penmen tally. Broad variations m the kinds and concentrations of salts in samples can make the use of a membrane probe difficult. 4.2.2 The thallium probe requires the presence of salts in concentrations which provide a minimum conductivity of approximately 200 mkromhos. 4 3 Reactive compounds can interfere with the output or the performance of dissolved oxygen probes. 4 3.1 Reactive gases which paae through the membrane probes may interfere. For example, chlorine will depolarize the cathode and cause a high probe-output Long-term exposures to chlorine will coat the anode with the chloride of the anode metal and eventually desensitize the probe. Alkaline samples in which free chlorine does not exist will not interfere. Hydrogen sulfide will interfere with membrane probes if the applied potential is greater than the half-wave potential of the sulfide ion. If the applied potential is less than the half-wave potential, an interfering reaction will not occur, but coating of the anode with the sulfide of the anode metal can take place. 4.3.2 Sulfur compounds (hydrogen sulfide, sulfur dioxide and mercaptans, for example) cause interfering outputs from the thsllium probe. Halogens do not interfere with the thallium probe. 4.4 At low dissolved oxygen concentrations, pH variation below pH 5 and above pH 9 interfere with the performance of the thallium probe (approximately 0.05 mg/1 DO per pH unit). The performance of membranes is not affected by pH changes. 4.5 Diaolved oxygen probes are temperature sensitive, and temperature compensa tion is normally provided by the manufacturer. The thallium probe ha* a temperature coefficient of 1.0 mv/"C. Membrane probes have a temperature coefficient of 4 to 6 percent/*C dependent upon the membrane employed. Apparatus 5.1 No specific probe or accessory is especially recommended as superior. However, probes which have been evaluated or are in use and found to be reliable are the 1M HONS Z1*858 Weston & Stuck DO Analyzer Model 30, the Yellow Sprmp Instrument (YSI) Model 54. und the Beckman Fieldlab Oxygen Analyzer. 6. Calibration Follow manufacturer instructions 7. Procedure Follow manufacturer instructions. 8. Calculation Follow manufacturer instructions. 9. Precision and Accuracy Manufacturer's specification claim 0.1 mg/1 repeatability with *1% accuracy. MOWS 204859 ft 03 A.3-7 References for Section A.3 1. Schmidt, H., and Schultze, G., (Liebig's) Annalen. 1881, p 207, p. 238. ' 2. Penning, C. H.,"Physical Characteristics and Commerical Possibilities of Chlorinated DiphenylInd. Eng. Cheat. 1930, Vol. 22, p. 1180. 3. Clark, P. M.,"Nonflammable Dielectric Organic Compounds," Ind. Eng Chem. 1937, Vol. 29, p. 698. -------- 4. Clark, F. M., Trans. Electrochem. Soc 1934, Vol. 65, p. 59. 5- Broadhurst, M. G.,"Use and Replaceability of Polychlorinated Biphenyls." Environ. Health Perspect 19T2. Vol. 2, pp. 81-102. 6. Holden, A. V.,'Source of Polychlorinated Biphenyl Contamination in the Marine EnvironmentNature 19T0, Vol. 228, pp. 1220-1221. 7. Koeman, J. H., Ten Soever De Brauv, M. C., and Devos, R. H., "Chlorinated Biphenyls in Fish, Mussels, Birds from the River Rhine and the Netherlands Coastal Area." Nature 1969, Vol. 221, pp. 1126-1128 8. Jensen, S., Johnels, A. G., Olsson, M., and Otterllnd, G-., "DDT and PCB in Marine Animals from Swedish Waters." Nature 1969, Vol. 224, pp. 247-250. 9- Zltko, V., Bull. Environ. Contain- Toxicol. 1970. Vol. 6, p. 464. 10. Veith, G. D.,"Recent Fluctuations of Chlorobiphenyls (PCBs) in Green Bay, Wisconsin Region." Environ. Health Perspect. 1972. Vol. 1, pp. 51-62. 11. Kimbrough, R., CRC Reviews of Toxicology. Jan. 1974, p. 448, Chemical Rubber Company, Cleveland. 12. Schweitzer, G. E., statement at the Wisconsin State Hearings on PCBs, Madison, Wisconsin, August 19T5* 13. Stallings, D. L., and Msyer, F. L.."Toxicitiea of FCBa to Fish and Environmental ResiduesEnviron. Health Perspent. 1972. Vol. 1, pp. 159-164. 14. Hansea, 0. J., Parrish, P. R., Lowe, J. J., Wilson, A. J., Jr., and Wilson, P. D.,*Mhronic Toxicity, Uptake and Retention of Aroclor 1254 in Two Estuarine Flshee." Bull. Environ. Contam. Toxicol. 19T1, Vol. 6, pp. 113-119. 15. "PCBs and the Environment," Interdepartmental Task Force on PCBs, Com-72-104l9, National Technical Information Service, U.S. Department of Coaaerce, March 1972. 16. Kurataune, M., "Yusho; Characteristics and Long-tsrm Effects," National Conference on Polychlorinated Biphenyls, Environmental Protection Agency, Chicago, November 19-21, 1975. 104 HONS ZO*860 IT. Dauppi, T. A., and Currie, C. C., Product Engineering. 19^*9, Vol. 20, p. 108, 18. Currie, c. C., and Smith, B. F.,'*?low Characteristics of Organopolyailoxane Fluids and Greases,"Ind. Ena. ChMi. 1950. Vol. k2, p. 2U5T- 19- Fitzsimmons, V. G., Pickett, D. L., Militi, R. 0., and Zisman, W. A., Trans. A.S.M.E. (19^6), Vol. 68, p. 365. 20. Scott, D. W., J. Am. Chen. Soc. 19b6. Vol. 68, 356. 21. Patnode, W., and Wilcock, D. F., ibid., p. 358. 22. Fordham, S., The Silicones. The Philosophical Library, New York, 1961, also Noll, W., chemistry and Technology of the Silicones, Academic Press, New York, 1968. 23. ASTM Standard Method D 92, Annual Book of Standard Methods Part 1*0 American Society for Testing and Materials, Philadelphia, Pa., 19T5. 2b. Clark, F. M., Insulating Materials for Design and Practice. John Wiley li Sons, New York, 1962, pp. 136 and 320. 25. See for example "Dielectric Properties of Dow Corning Silicone Liquids," Issued by Dow Corning Corporation, Midland, Michigan. 26. Stock, A., and Saaleski, C., 19l6, Vol. **9, p. 111. 27. Kipping, F. S., Proc. Roy. Soc. 1937, Vol. 159A, p. 131. 28. McGregor, R. R., Silicones and Their Uses. McGraw-Hill Book Company, New York, 195**. 29. Link, E., RTE Corporation -- personal commlcation. 30. Duckett, D. A., Paper before the General Meeting of the Edison Electric Institute, Transmission and Distribution Conittee, Minneapolis, Minnesota, May 8, 1975. 31. Morrison, R. T., and Boyd, R, R., Organic Chealetrr. Allyn k Bacon, New York, 1966. 32. Munch, R. H., German Patent Disclosure No. 2,310,807,13, Sept. 1973, U.S. Patent No. 3,796,93b. 33. Munch, R> H., Monsanto Chemical Company, St. Louis, Mo. -- personal communication. 3b. Munch, R. H., "New Capacitor Impregnants," Conference Paper presented at IEEE Power Eng. Soc, 1975 Winter Meeting Session, 35. Report on MCS 1238, Obtainable from Corporate Medical Dept., Monsanto Chemical Company, St. Louis, Mo. 63166. 36. Graham, P. R.,"fhthalate Ester Plasticisers - Why and How The Are UsedEnviron. Health Perapect. 1973, Vol, 3, pp. 3-15, 105 MOUS 204861 37. Forster, S. 0. -- personal communication. 38. Rutkovski, A. J., and Forster, E. 0., Presented at American Power Society (IEEE) Winter Power Meeting, Sew York, 19T5, also Proc. VII Int. Conf. on Electrical Insulating Materials, Inst. Elec. Eng., Tokyo, Japan, 197**. 39- Moffitt, A. E., Jr., Clary, J. J., Levis, T. R., Blank, K. D., Perone, V. B., J. Am. Ind. Hyg. Assoc-. 1975, Vol. 36, p. 633. 1*0. Tanaka, A., Adachi, T-, Takahashi, T., Yamaha, T., IsULtfifiiagx 197 5, Vol. 1*, pp. 253-26U. ' 1*1. Peters, J. W. , Cook, R. M., Environ. Health Persnect. 1973. Vol. 3, pp. 9I-9L. See also Rubin, R. J., and Jaeger, R. J., ibid., 1973, Vol. 3, pp. 53-59, and other papers from S.I.E.H. Sciences Conference on Phthalic Acid Esters, Sept. 6-7, 1972 in Environ. Health Perspect. 1973, Vol. 3, pasalm. 1*2. Capacitor Fluid XFS-U169L statement on development of fluid by Dow Chemical Company and McGraw-Edlson Company, Dec. 2, 1975, Dow Chemical Company, Midland, Michigan. 1*3. Environmental Impact Statement on XFS-i*l69L Capacitor Fluid. Dow Chemical Company, Midland, Michigan. - 1*1*. Rational Academy of Sciences report on the Principles for Evaluating Chemicals In the Environment. Rational Academy of Sciences, Washington, D.C., 1975* 1*5. Registry of Toxic Effects of Chemical Substances, 1975 ed., U.S. Dept, of Health, Education and Welfare, Rational Institute for Occupational Safety and Health, Rockville, Maryland 20852. 1*6. Mukal, H., Lee, P. F-, Oguri, M., Schepena, C. L., Can. J. Ophthalmol 1975, Vol. 10 (3), p. 391* 1*7. Perkin-Elmer Chromatography newsletter, 197^, Vol. 3, P- 15. 1*8. "Polychlorobiphenyls and the Environment." Interdepartmental Task Force on FCBs. Cam.-72-10l*l9 Rational Technical Information Service, U.S. Dept, of Commerce, Springfield, Virginia 22151. 1*9* Siegel, S., and Stewart, T.,"Vacuum Ultraviolet Photolysis of Polydlmethylailoxane Gas Yields and Energy Transfer." J. Fhya. Cham. 1969, Vol. 73, p. 823. 50. Delmn, A. D., Lendy, M., and Sima, B. B., J. Polymer Scl. 1969, Vol. 7, p. 3375. 51. Weaver, P. J., and Coughlin, F. J.,^teasurement of Biodegradability," Amer. Oil Chem. 3oc. 196k. Vol. 1*1, p. 738* 52. Publication 253/62 of the Federal Council of the German Federal Republic, 1962. 106 HQNS 53. Soap and Detergent Association, Aar. Oil Chen. 3oc. 1965, Vol. U2, p. 986. " 5I*. Published by U.S. Environmental Protection Agency, Office of Technology Transfer, Washington, D.C. 20460. 55. S&eger, V. W., Tucker, E. 3. Ill,'Phthalate Eaters Undergo Ready BiodegradationPlastics Eng. 1973. Vol. 29 (8), pp. l*6-i*8. 56. Gleason, M. H., ed. Clinical Toxicology of Commercial Products. l*th Ed., Williams and Wilkins, Baltimore, 1975. ViOHS 20^3^^ 107 A.U The Scope end Influence of Government Regulation* The purpose of thi* *ection 1* to identify the several government agencies which have potential regulatory pover over new insulating fluids, to describe the particular area of concern over which each exercises regulatory power, and to define what criteria and procedures are used to determine whether a new fluid falls within the purview of each particular agency or regulation. The emphasis will, therefore, be on the regulations themselves; criteria and testing procedures which support or are cited in these regulations will be briefly identified. In Section A.3.2, these test techniques are more thoroughly explored. A.lt.l Environmental Regulations^ Environmental protection regulations are developed and enforced by the Environmental Protection Agency (EPA) and by the various state departments of natural resources. The state agencies generally follow the Federal regulations, and they may In fact have operational responsibility for enforcing the Federal environmental regulations. Being more closely in touch with the local Issues and being under less complex operating constraints, the states csn often act mora rapidly than the Federal enforcement agencies. It Is important, therefore, that the state and Federal environment regulatory bodies act under rigorous, quantitatively definable criteria in establishing and enforcing their regulations. We will review the existing and pending Federal regulations, recognizing that these will also probably be reflected in any state acts. Insulating and dielectric fluids are typically very stable compounds which have low vapor pressures and which, therefore, do not represent a significant air pollution problem. The primary environmental effect is In water, and so the primary source of environmental regulation of such fluids is: Public Law 92-500 (86 Stat. 816) "Federal Water Pollution Control Act Amendments of 1972" Three sections in this act are of particular importance for evaluating the Impact of regulations: 1. Sec. 307a "Toxic effluents." A proposed list and a final list of toxic substances were published by EPA in the Federal Register on July 6, 1973 (38FRl8oUU)3 and on Sept*b*r 7, 1973 (38FR2b3b2), respectively. Criteria for selecting these substances were included, thereby indicating the standards by ^Section A.3.It also contains discussions of Federal environmental regulations in terms of definitions of test techniques and criteria. 2See Section A.U.3 "Occupational health," for air quality criteria in the working place. ^Federal Register (FR), vol. 38, page 180UI*. 108 HONS 204864 which new lubstsnce* can be evaluated, the toxic substance list is not enforceable, however, until standards fbr allowable levels of these toxic substances are finally promulgated. Although standards were proposed on December 27, 1973 these have not yet reached final legal status. ' 2. Sec. 307b "Pretreatment standards." The Administrator of SPA is authorized and directed to publish and promulgate standards for pretreatment of pollutants which could be introduced into publically owned treatment works. 3. Sec. 311 "Oil and Hazardous Substance Liability." (spills, leakage, pouring hazardous substances into navigable waters of the U.3.). On August 22, 197**, an advanced notice was issued: "Vuigttation and dettMiunation oi Ktnovabitity oi hazardouA iubitancti ptcm u)attKn (39?R30<*66). In response to this advance notice, written coments were received and were incorporated into a proposed set of rules issued an December 30, 1975 (1*07R5996o). When promulgated, these rules will provide the basis of the hazardous chemical (including PCBa) spill control program. Key elements activated by these regulations will probably include mandatory reporting of discharges, clean-up and damage mitigation, and civil penalties which may be assessed by both the Environmental Protection Agency and the Coast Guard. U. Sec. **02 "Pollution discharge permits." This section empowers the regional offices of SPA to issue permits limiting the allowable levels of pollutants in effluent discharges. Limitations on polychlorinated biphenyls have already been included in several permits. "Pollutant" is defined (in Sec. 302) as effluents which "(Wold iattKiVU. mitk tkt attainmtnt ox mainttnanct oi that wttK quality in a Aptcx^tc poKtion o{ tkt navigablt wttK* mhick thatt auu/it pnottctZon oi public, wttK iuppliti, agKiciittunat and induatriat usee, and tkt prottetian and propagation oi a batanctd population oi iheltii&h, tiih and uiitdliit, and attorn xtcKtntionat activititi in and on tkt wttK,...." In edditlon to the existing legislation cited ebove. Congress is in the process of working on a Toxic Substances Control Act, which if passed would give XPA more direct control on production and use of identified toxic substeaces. A.fc.2 Poisonous Ingredients in Poods The Food and Drug Administration (FDA), an agency of the Federal Department of Health, Education and Welfare, has the responsibility for administering the consumer protection standards established by the Federal Food, Drug, and Cosmetics Act of 1936 with amendments (FFDCA) (U.S. Code, Title 21). The various state departments of public health cooperate in 109 HONS 204865 maintaining proper food standards, under authority of state public health legislation. Whereas FDA enforcement authority concerns only foods involved in interstate commerce (Sec, 703 and 70U of FFDCA), the state may, of course, take local action. "Poisonous substances" include all substances documented to be harmful to human health. Thus, they may be specifically identified in FDA regulations (as in the case of PCBs; see 38FR18096, July 6, 1973) or known to be poisonous from other reliable sources such as "The Toxic Substances List" published by the national Institute for Occupational Safety and Health (see Section A.U.3 below). Likewise, tolerance levels are not specifically identified by FDA for each known poison, but rather this is judged on a case-by-case basis, as specified in See. Uo6 of FFDCA.1 Any new Insulating or dielectric fluid which is identified by any Federal agency to be poisonous would thus corns within FDA Jurisdiction. In relation to this, it should be noted that the FDA regulations on PCBs (38FR18096) specifically exesept PCB transformers and capacitors; their being in sealed containers was an influential factor in this decision. In enforcing its regulations, TOA can request a manufacturer to recall an adulterated product, under threat of prosecution. It can, under court authority, seise a shipment of adulterated food. And it can legally prosecute manufacturers of illegally adulterated foods. Actions are reported in; "FDA Weekly report of seizures, prosecutions, injunctions, field corrections and recalls," published weekly uy the FDA ''rese Office, Rockville, Ksrylsnd 20852. A.U.3 Occupational Health The Occupational Safety and Health Administration (06HA), a part of the Department of Labor, hat the responsibility "to oAAmu. Ac jo*. a* ponibit tv&uf Milking man and vxtman in the. Nation ait and kaath{ut uxvdung condition*...." as stipulated in; Public Lsnr 91-596 (8U Stat. 1590) "Occupational Safety and Health Act of 1970" Dielectric and insulating fluids could cgsm within the scope of 08HA regulations in two ways; 1. During processing of the material by the capacitor and transformer manufacturers. 1Although levtls were set for PCBs, in the July 6, 1973 regulations. 110 HONS 204866 2. By use of the capacitor* and transformer* in on cloaely adjacent to work areas. In the Act cited shove, the following responsibilities are defined: 1. The Secretary of Labor (i.e., OSSA) is to promulgate standards for toxic substances (Sec. 6). 2. The Secretary of Labor also is empowered to inspect sorting places (Sec. 8) and to enforce the standards (Sec. 10, IT). 3. The Secretary of Health, Education and Welfare, through its agency. The national Institute of Occupational Safety and Health (NIOSH) is to carry out research, develop criteria, and carry on training concerning health and safety in the working place (Sec. 20, 21, 22), Thus, NIOSH establishes standards, while OSSA promulgates and enforces these standards. The following publication* and standards have resulted regarding toxic substances: "Registry of Toxic Effects of Chemical Substances," 1975 Edition, NIOSH, Rockville, Ml. 20852, June 1975 In this publication, a large number of toxic substances are listed, along with toxic dose levels, end the reference sources for these toxic doses. Four PCBs are included. "Occupational Safety and Health Standards," 39FR23502, June 27, 1971*. Table 01 lists contaminants, including chiorodiphenyl (l*2j Cl); skin 8-hour time weighted average exposure not to exceed 1.0 ag/m^, and cblorodlphenyl (5U| Cl), 8-hour time weighted average skin exposure not to exceed 0.5 mg/sr. (These are currently listed in 29C7R 1910.1000). OSSA has developed draft technical standards establishing require ments for monitoring employee exposure, medical surveillance, handling end use, training, record keeping, sanitation and housekeeping. These are now available for comnant (*0015^033, November 20, 1975) Tbs source of these standards is UlCFR 50-20^.50 which cites: "Threshold Limit Values of Airborne Contaminants for 1968," Anar. Conf. of Goto. Industr. Hygienists. The National Electrical Code (NSC) - 1971 (ANSI Cl-1971) has been incorporated within OSSA regulations (29C7R 1910.308; aee 39FR23502, June 27, 1974). Dielectric fluids for transformers and capacitors 111 HONS 204867 therefore fall within 06HA regulations Insofar as use of these fluids is affected by the NEC. (See Section A.5 of this report for furthsr discussion of the NEC). A.!*.!* Transportat ion The Department of Transportation (DOT) is involved in both the use and the control of insulating fluids in these two ways: 1. Use. High speed mass transit systesu employ electric traction. The electric transformer is a key item. It must be small because of the site limitations on the transit cars, and fireproof in the event of an accident. Askarel transformers are therefore universally used in present ground electric transit systems.1 Thus, DOT may become a significant participant in the move toward both safe and reliable techniques for continued use of PCBs and for substitutional use of new insulating fluids; however, no documentation regarding efforts by this agency to encourage development of improved transformers has yvt corns to our attention. 2. Control. The transportation of basardous materials falls within DOT'S regulatory power, aa identified in: U9CFB Parts 100-199 Transportation; revised of October 1, 1975- The following sections are particularly pertinent to the insulating fluids question: 1. List of Hfesydnna Materials; see. 172.5 a. lone of the PCBs, by any of their chemical or trade disks, is on the list. b. Any substance which qualifies for the list under definitions given in other sections must he treated as if on tbe list. 2. Dielectric and insulating fluids would be Identified as Class B poisons if they were found to be toxic. (See See. 173.3^3, quoted on pp. 91, 92). ^The Japanese retained aakarel in their electric rail cars even after their otherwise complete ban on PCBs; they apparently are now experimenting with a silicone substitute. 112 HONS 20^868 3. Effect of list. Proper procedures for the transport of hazardous materials are staged in the following sections: Pt. 173 Shippers "173.34* PadUng {or class 8 poisom, liquid a) Closing and cushioning. HI contouners must bt tightly and securely cLoad. Inside containers mat be cushioned 06 prescribed, ok in any case when necessary to prevent breakage ok leakage. b) Outage. Outage {ok containers o{ Liquid, potion {ok tKanApoKtation by carriers by Kail freight, Kail express, highway, ok water muAt be oa {otlows: T) Container must not be entiKtly {Hied. Su{{icient interior Apace mat be lt{t vacant to prevent leakage ok diAtontion o{ containem due to the expansion o{ the content* {nom increase o{ temperature dating Hamit. Z) The proper vacant Apace (outage) in a tank cat ok other shipping container depends on the coefficient o{ expansion o{ the liquid and the maximum Increase oj tempenatwie to which it will be subjected in Hamit. Outage must be calculated to the total capacity o{ the containtA. 3) Liquid poison must not be loaded into domes of tank cam. 4) Ik tank cam, outage must be calculated to percentage o{ the total capacity of the tank, i.t., shell and dome capacity combined. 1{ the dome o{ the tank can dots not provide su{{icient outage, then vacant space must be left in the shell to make up the required outage. 5) The outage {ok tank cam mat not be less than 1 percent. 4) Mo cargo tank or compartment thereo{ used {or the tramportation o{ any liquid poison shall be completely {Hied} su{{icient space shall be left vacant in every case to prevent leakage {ram or distortion o{ any such cargo tank by expansion o{ the contents due to rise in temperature in transit, and Such free space {outage) shall be su{{icient in every case so that such cargo tank shall not become entirely {Hied with the liquid at 130 *F. HONS 204869 113 173.345 Exuptiaru foK palionoui Icquxdi, Clan 8. a| Poi&onoui liquid*, cla** 8, a* defined in 173.343, except tho*e foK which no exwptionA eua pnovided a* indicated by the "No exemption" statement in 172.5 oj thu> chapteK, ok a* pnovided (ok in 173.355(c), in tightly cloitd -mAidt container, itcuntly cat hioned when necenaKy to pntvtnt bneahage and packed a* follow, one exempt foam iptcifocation packaging, manking, and labeling KtquiKement*, except that manking name o( content* on out*ide containen La nequined foK ihipment* via caimitK by waten. ShipmentA foK tnampontation by highway caKnitA* one exempt alia foam Pant 177 o( thi* chapteK, except 177.117, art Pant 357 o( thi* title. 1) In gla*4 ok eaKthemoane containen* not oven 1 quant capacity each, ok in metal containen* ok polyethylene bottle* not oven 1 gallon capacity each, packed in itxong outaide wooden boxe* ok bonnet*,, 2} In gla** ok eantherwaKt container not ovck 1 pint capacity each, ok in metal ok polyethylene container not ovck 1 quant capacity each, packed in itnong outiide (ibtnbcand boxei." The following sections are alto of uat relevances 173.4l4 Poisons Material labels Pt. 17U 174.532 17**-538 17U.586 Rail freight Loading other hazardous aaterials Rail freight ; loading and storage chart of hazardous Material Handling hasardous Materials, by carriers by rail freight Pt. 175 Carriers by Rail Bxpress Pt. 176 Rail Carriers in Baggage Service 176.702 Hazardous Materials Pt. 177 177.821 177.841 177.848 177.860 Public highway carriers Hazardous Materials forbidden or limited for transportation Loading; poisons Loading and storage chart of hasardous materials Accidents or leakage; poisons Pt. 178 Shipping container specifications Pt. 179 specifications for Tank Cars 114 HONS 204870 A.U.5 Solid waste disposal The Department of Health, Education and Welfare (HIV) has certain regulatory control on aolid vaate dlapoaal through the authority of the Solid Waste Disposal Act (PL 89-272 and amendments; see 1*2 H.S.C., Ch. 39). In particular, HEW is to recoamend and publish guidelines and information on solidwaste recovery, collection, separation and disposal. Facilities owned or controlled by the Federal Government oust comply vith these published guidelines. The disposal of the PCBs and of solids contaminated by *skarels is becoming increasingly difficult.. Problems associated with disposal of new fluids should be anticipated at an early stage in their development; the Federal solid waste disposal regulations font one set of criteria by which disposability of new fluids can be Judged. 115 HONS 204871 A.? The Scope and Influence of Insurance and Fire Code Requirements In addition to the various federal and state governmental regulations which night restrict the use of fluids In electrical equipment, one must also sake sure that such equipment does not conflict with insurance and fire code regulations. The two are closely related and therefore can be most easily discussed together. The National Electrical Code (NIC) is one part (volume 6) of the 1975 National Fire Codes, Issued by the National Fire Protection Association (NFPA). This code is developed and continually updated by a set of conittees which is sponsored by NFPA and is composed of representatives from aanufacturers, users, fire insurance companies. Independent testing laboratories, and municipal and state fire safety and regulatory agenclee. As such, it is an authoritative sort universally recognized as a standard statement of safe electrical fire practice. Although the NEC is not in itself lav, it is usually adopted with various levels of modification as the legal electrical building code. Equipment to be legally installed within the Jurisdiction of such a code must therefore conform to NIC. A further note an the NIC is that it has been adopted in whole by OSHA ("Occupational Safety and Health Standards," 29CPR 1910.300 - See 39FR23502, June 27, 197*0 and it therefore has a very broad legal Importance in Industrial facilities. OSHA regulations are discussed more extensively In Section A.U.3. As pressntly written, the ISC specifically refers only to alnsral oils and askarels as insulating fluids. Sines no more general descriptions or definitions are given, one is lsd to infer that other fluids are excluded unless specifically identified in future editions. To get a new material listed in the NIC, the asnufseturer must propose the desired Inclusion to the Electrical Coordinating Committee of NIC, along with evidence subetantiating the fire safety of the material. Before approval, however, the committee would probably require testing tnd safsty verification by an independent laboratory such as Underwriters Laboratory (UL) or Factory Mutual Nbginaering and Research (IMR). UL Is an independent not-for-profit testing lsborstory which carries out tests on products for manufacturers and makes "listing" or "classification" judjwnt based on the** tests: "Listing" - the product is deemsd to meet tbs requirements of the NSC. "Classification" - a material will he given a flanmability clas sification, based on a scale between 0 (water) and 100 (diethyl ether); the procedure is outlined In: "Tests tor Comparative Flesmablllty of Liquids," UL Standard No. 3**0. HONS 20^872 116 Whereas SIC Is tn installation guide, local ordinances vill often look to the UL listing to verify that the product meets 1XC standards and therefore can he Installed as directed hy the code. FMDt Is a service arm of the factory Mutual System of Insurance companies. It carries out testing of equijment and Issues "approvals" or "acceptances" based on these tests: "Approval" - the product Is deemed to be safe and suitable for general applications. "Acceptance" - a particular installation of equipment and materials meets fire safsty standards. Such a safety test could be Initiated upon request by a manufacturer or by one of the 7M Insurance companies. An Installation of electrical equipment must not only meet fire ordinances but also must be covered by fire insurance. The insurance rates are generally set by the State Insurance Commission, but an insurance company must have verification of the safety of the Installation before agreeing to vrite insurance coverage. The insurance inspector vill therefore generally look for UL or M acceptance, or to tome other indication that the installation meets IXC requirements. The inspector might also base his decisions on his co^mny's rating of the requesting company and possibly on the users or manufacturer's own tests. ' An experimental unit vith a nev insulating fluid might for instance be accepted for further field testing under this latter situation. If one unit vithin a larger installation does not meet standards, temporary approval may be granted vith the understanding that the substandard unit be upgraded to an acceptable level vithin a specified period. It should be noted that utilities are often self Insured and are granted exclusion from the locel codes based on BIC. It is tbs general practice of utilities to comply vith BIC, however, sad they usually vork closely vith local fire protection people to make electrical installations safe. When utility owed equipment is Installed in users buildings, often the local code or insurance requirements vould necessitate that BIC practice be followed. USCOMM-HBS-DC 117 HONS 204873