Document KGrGb9BJoRmGN8ay7o4GKyrBo

Heating Piping/Air Conditioning / Engineering data file How to design fume hoods, exhaust systems for research labs A comprehensive guide to safe and satisfactory collection and subsequent dispersal of hazardous or noxious fumes, odors, and dusts encountered in research laboratories By i'/ / a, uEIDER, PE, Federal Housing Administration, HUD, Washington, D.C* T>\v i'i u- of this HPAC En- gitwnny | }at;t File is to describe fumo u.ukIs and their systems and tv' vtvn t,v< Some of the prob* lcm\ lVjjjiiiiiMjt them. It is intend ed m v.iviT tume hoods used in resvwwK IMtnratories, particularly tho-'o universities. A Inmo hood is an extension of an duct, so constructed as to vvni.tiii tm experiment that is privluviun luvardous or noxious fun\\'s% ovltirs, >r dusts that must be u'tnovcd from the laboratory. The i'urptiw of the fume hood, thotvioiu, u jo enclose the experi- mom ,m.l t.> collect and discharge tuiw-\, lar m.niv'i and fine particu Typ*x ol lnu,ds 1<v .m> several types of hovx > ,,( |f,e most common is,, ->v' shown in Fi.Cn I :iuit it consists of an enc'v'>un luiviug three sides and a nV, ,>,, ,,hni vash in front, which ma\ ts- Qf lowered. The !ux\: ,v .-lumped with a baffle '! <........... tilfiliated with the A'i; v,inundation's Archi~ n\ i,,Kineermg Services this HPAC Engineerm- -v t icii.v prepared. near the rear wall, which has two or more adjustable slotted open ings. An exhaust fan is connected to a duct leading to the plenum between the baffle and back wall. The fan draws air from the room through the sash opening, through the slots, through the back ple num, and through the exhaust duct to the outdoors. The slots are adjusted during installation to re move both hot and cold fumes at a suitable uniform sash velocity. A standard hood may be ob tained as a separate unit to be set on top of a laboratory bench, or it may be purchased complete with a cabinet base, which may be used for storage or as a refrigera tor, oven, or ventilated chemical storage space. A is a standard hood whose walls extend to the floor, thus providing sufficient space to accommodate a more elaborate experimental setup re quiring additional height. Such hoods have double or triple hung sashes, which may he raised and lowered to provide access to any part of the setup while the remain ing space is enclosed to contain fumes. The back baffle of such a hood extends over the full height of the hood and is equipped with at least three adjustable slots to regulate the amount of air passing over various parts of the setup. Fig. 3 shows a is constructed to per mit room air to be drawn through the bypass directly into the hood as the sash is lowered. The pur poses of the bypass are to pMM at the sash as it is being lowered and to MMMHMlaHli In large research laboratories, many more hoods are often in stalled than are needed for the purpose of relieving the outdoor air introduced at the air condi tioner for building ventilation. The additional air required for the ex tra hoods must be heated and cooled. Because this is expensive, many mechanical system designers try to reduce the extra condition ing load by installing separate outside air supply systems to se lected hoods. Fig. 4 illustrates three types of are de signed to permit the introduction of outside air directly into the hood itself (the two hoods on the right) or in the front of the sash outside the hood (the hood on the left). If the auxiliary air is intro- Ilciriholii Publishing Corporation. All rights reserved. Printed in U.S.A. 103 { i >vt Fume hood, exhaust system design duced behind the sash (i.e., di rectly into the hood), the {assure* locity through- tbfe-'isnh is dras* ucally reduced, and the hood will ot be able to prevent leakage of fumes back into the room. For this reason, auxiliary air should never be introduced directly into. the hood. Other problems with auxiliary air systems will be dis cussed in greater detail below. Another special type of hood is the perchloric acid hood, shown in Fig. 5. This is a special adaptation of the standard or the balanced air hood for particular use with perchloric acid or other dangerous oxidants. Such hoods are equipped with water sprays along the entire length of the exhaust duct and in the space between the back baffle and the back wall. The sprays are periodically turned on to wash down acid crystals and organic matter that may have accumulated, and drains are provided to receive the wash water. The inside surface of the hood is usually hosed down or washed by hand. If left in the hood and in the exhaust ductwork, accumulations of organic matter and perchloric acid crystals can produce unexpected and violent explosions. fTnrfwnw'w Anffrfu ~u r variations of standard and balanced air hoods, specially constructed to permit easy cleaning. They are usually equipped with HEPA fil ters in the exhaust ducts on top of the hoods to collect radioactive particulate matter. Such hoods are used for isotope work or for other work where the discharge from the hoods must be cleaned of harmful particulates. There are other types of fume hoods also. The most important of these is the exhaust box, which is equipped with small hand openings similar to those provided for glove boxes used for hazardous biologi cal work. Such exhaust boxes re quire much less air to be exhaust ed to the outdoors, but they may not provide adequate space for some experimental setups. Another useful fume collector is the flexible duct or snorkel, which is used to draw heat, moisture, and nonhazardous fumes from a lo calized area. What face velocities? Since the basic purpose of a hood is to prevent leakage of fumes into a laboratory, the velocity of the air through the sash opening must be high enough to capture fumes around the periphery of the sash. A minimum face velocity of 100 fpm is recommended for fume hoods used in chemical research, and 75 fpm for hoods used in un dergraduate teaching laboratories! fume hood operation. Even at 100 fpm and with uniform laminar flow through the sash opening, it is very easy to disturb the flow. This can produce eddy currents that may reverse the air flow at the edges, thus causing the hood to leak. For example, aepHMHHnitoMppMBHttHVat 4 mph creates a wake that will approach a velocity of 350 fpm. Eddy currents from such a wake can very easily dis turb the smooth flow of air through the sash, usually resulting BAL 00004 On the other hand, a velocity much over 100 fpm can interfere with tests being conducted in a hood. For example, an open flame will not burn properly at veloci ties over 100 fpm. Other common causes of face velocity interferences include: lo cation of the hood near an opera ble window; location near a fre quently opened and closed door; or location in the path of any air motion exceeding 30 fpm, such as a nearby air conditioning diffuser or a portable fan. Hood location, design tlBBt to adequate face velocity, the iiHwXiimpn--t M|iri mt ini) Heating/Piping/Air Conditioning, March 1972 factory operation- ar-hood is iUtoortforr.' A hood located near the entrance of a laboratory will be affected by the fanning action of the door, by the inward flow of corridor makeup air, and by the eddy currents of persons passing through the door. Even worse, a hood location behind a door sub jects its user to the possibility of being bumped, which could result in a serious accident. For these reasons, it is impor tant to locate a hood as far into the room as possible and particu larly Safe performance of a hood is far more important than the small increase in friction loss from slightly longer exhaust ducts. A hood should not be positioned so that its sidewall is located direct ly against a partition or window wall, however. This will increase the resistance of the entering air next to the wall, which will reduce the capture velocity near the wall and possibly result in concealed leaks. A hood should be constructed to avoid center posts and sharp edged corner posts. Instead, all liltingi mlpw (where the air en ters the hood at the sash) should be streamlined*1 to prevent eddy currents that may cause the hood to leak. Also, the inside wall sur faces of the hood should be smooth and free of recesses or sharp edges that could disturb the smooth laminar flow through the hood. Of special importance is the leading edge of the hood work surface. In addition to being streamlined, it should have a raised air foil installed (Figs. 2 and 3) to assure a continuous smooth flow of air across the work sur face when the sash is fully closed. Such a flow of air will prevent leakage of vapors that are heavier than air, and the raised surface will defeat the tendency of opera tors to set experimental apparatus too close to the sash opening, thereby disturbing the laminar flow and creating eddy currents. Problems with auxiliary air In an earlier paragraph, auxili ary air hoods were briefly de scribed, and the point was made that auxiliary air should never be introduced into a hood behind the sash. There are other serious prob lems with auxiliary air hoods and auxiliary air supply systems. Although auxiliary air intro duced outside a hood and in front of the sash is flMe, it has the serious disadvan tage of OMffiaaiMHMliMlta through the sash opening. This is particularly true of hoods that have auxiliary air introduced around the periphery of the sash opening immediately in front of the sash and at right angles to the inward air flow. If, for example, the total air volume exhausted from such a hood should become diminished because of a reversed fan motor, a dirty exhaust system, or a faulty damp er, the continued supply of auxil iary air across the reduced laminar flow entering the hood will dis rupt smooth flow, producing se vere eddy currents and leaks. (By way of comparison, if the auxiliary air is introduced inside the sash and the exhaust air volume is re duced, the auxiliary air will almost certainly This is another reason why auxiliary air should never be introduced inside a hood). Even if air inflow interference does not materialize, WHNtoq^jB#elivered overhead arid outside the hood in winter can be extremely to '^e hood user and can cause mtM or fog within the hood. Moreover, unless the auxil iary air is precooled and dehumidi fied in summer, it may add to the room air conditioning load through partial dispersion of the untreated air into the room. Finally, it should be remem bered that an entire auxiliary air supply system is very ^^MMb&nd is especially difficult and gMHil Consequently, for most localities, it is likely to be cheaper and certainly more satisfactory to condition all of the makeup air using the central air conditioning system than to install and main tain a number of auxiliary air sys tems. If an auxiliary air system is be ing considered for a laboratory building, it is especially important to make a complete analysis of Heating/Piping/Air Conditioning. March 1972 tl... 000045373 Fume hood, exhaust system design costs (operating and maintenance costs as well as first costs) to de termine whether the system will really result in economies. More over, if an auxiliary system is de cided on, special attention is needed in the selection of an ac ceptable hood. MosU rfiucernmg designer*- brieve thataturiiwy *ur systems ^should be avoided and that all makeup air should be con ditioned by the central system. How about economy cycle? Economizer systems have fre quently been proposed for use with laboratory buildings in con junction with fume hood exhaust systems. The purpose of these systems is to remove heat from the disotrargioftH^hauauair in win ter and reintroduce the heat back- inte tbe makeufTair for the build ing. The process is reversed in summer. Such a system may in clude a large, slowly rotating wheel filled with porous material of a heat absorbing type; or it may consist of a run-around heat cycle incorporating heat exchange coils, piping, and pumps, or simi lar equipment. Economizer systems are be lieve* .-ha--r exhausniyin>f because the kinds of chemicals and vapors being ex hausted are both unpredictable and potentially damaging to the economizer equipment. Moreover, as will be discussed later in greater detail, the building air intake for a laboratory building should al ways be located as far away from the fume discharge termi nals as possible. If this is done, transfer of heat becomes impracti cal. Also, since there are itkefy to be many individual fume exhaust systems on the roof of the build ing, it may not be economical to recover heat from such a large number of small discharge ducts. weather. So mechanical exhaust fans were installed, usually loca ted overhead above the hoods. It was soon discovered, how ever, that the exhaust ducts read ily became corroded, thus permit ting the fumes (under pressure on the discharge side of the fan) to leak into the laboratory and into other parts of the building along the route of the ducts. Today, all fume hood exhaust fans are in stalled on roofs, usually exposed to weather. Sometimes fume ex haust fans are installed in pent houses. but this practice is dis couraged because of the danger to maintenance personnel. Utility cores In most modern research labo ratories, accessible mechanical service cores are provided for the installation of piping, electrical lines, and ducts. Most such cores are vertical shafts (see Fig. 6) large enough to accommodate a workman; others consist of verti cal service closets next to corri dors, with access doors at each floor. Fume exhaust ducts should always be installed in accessible service shafts so that they may be readily inspected, repaired, or re placed. Locating the exhaust fan Fifty years ago, fume hoods were usually exhausted by gravity or stack effect; that is, fumes were removed by the difference in weight between the warm inside air and the cool outside air. But this did not work well in warm Exhaust system configuration Wherever possible. e^^dEutne ' ' lull ........1...... duff igfrgi <see Fi? 6). If this is done, only isolated hoods will go out of service when equipment fails or a system is cleaned or repaired. Also, individ ual exhaust fans may be turned off when a fume hood is not in use, except, of course, for those needed to remove ventilation air from the building, and when more than one hood are installed in a room or in a contiguous suite of rooms. (The latter problem will be discussed in a subsequent para graph on safety features.) In large chemistry research buildings particularly, there are often more fume hoods installed than are needed for the relief of ventilation makeup air. If hoods are individually controlled, select ed fume hood systems may be shut off to save the cost of heat ing or cooling the makeup air. On the other hand, in high rise research buildings, the large num ber of individual exhaust ducts re quired may occupy more space 106 Heating/Piping/Air Conditioning, March 1972 \ f I than is available in the service cores, especially at the upper floors. By combining some of the hoods, one can reach an optimum compromise between individual hood control and reasonable ser vice core size. When fume exhaust systems are combined, it is important to com bine only those systems handling compatible fumes and managed from a single department. Hoods serving a suite of rooms under a single research program, for ex ample, may be combined. It is very important, however, to isolate all hoods that are likely to be used for hazardous materials such as perchloric acids, radioactive chem icals, or pathogens. In general, such hoods should be operating continuously. Exhaust duct materials Because fume ducts are exposed to a variety of corrosive gases and vapors and because these gases may vary as research programs are changed, it is not usually practical to select a duct ma terial that will satisfy all services. Since most ducts will corrode in time, it is important to insped tilt duets, periodically and to test them regularly for air-tightness. If ex cessive leakage of air into the ex haust duct does occur, the effec tive exhaust volume at the hood may be seriously diminished. Although aluminum, galvanizedsteel, and black steel have been used for fume ducts, they have only limited resistance to common chemicals. Glazed ceramic piping is usually satisfactory, but it is rarely used because of high install ation costs. Materials such as ce ment-asbestos are inclined to ab sorb moisture; also, they break easily and are attacked by some chemicals. Stainless steel and Mon el are satisfactory if they are used only for chemicals that do not at tack them, but there is never any assurance that other chemicals will not be handled at a future date. Reinforced plastics are selected by some designers, but a choice is made more difficult because of the great variety of materials available and their varied reactions to chem icals and especially to attack by fire. Epoxy coated steels were quite popular in recent years, but fabrication of such ducts near building sites, to reduce shipping costs, involved reliability problems. Free stream DRAIN TYPE OFFSET TYPE POWERHOUSE TYPE STACK DRAIN STACK DRAIN STACK 8 Stack drain arrangements that eliminate need for rain caps. In conclusion, it should be re membered that that may be encountered. Consequently, compromises will have to be made, with the understanding that failures may occur. For this rea son, it is important to set up a routine of regular inspections and maintenance if safe and satisfac tory operation is to be achieved. Fume discharge Each individual exhaust fan on the roof should have its own dis charge duct to convey the fumes vertically upward at a high ve locity as far above the topmost roof (usually penthouse) as possi ble. Failure in this will result in potential recirculation of fumes into building air intakes and will be particularly hazardous to per sonnel who use the roof for main tenance, research, or relaxation. As the wind blows over the leading edge of a roof parapet, as shown in Fig. 7, a disturbance is created that sweeps from, the edge of the parapet up over the top of the building. Above the boundary of this disturbance, wind flow is undisturbed. Below the boundary, the influence of the sharp edge of the building creates eddy currents that tend to pocket fumes released at the roof. This is called the wake cavity. Unless where they can be carried away, they will remain relatively undiluted on the roof and in the lee of the build ing, where they can enter building air intakes either on the roof or at the ground. When this happens, all the care taken in designing a good fume exhaust system may be nullified. And with the present concern over air pollution, failure to disperse the fumes may give rise Heating/Piping/Air Conditioning, March 1972 9.&1. 000045375 107 ( Fume hood, exhaust system design Prevailing winri 9 Exhaust tower, architectural in tegrated with building. to legal action against the build ing owner. Fume absorbers such as charcoal have been proposed to relieve the fume disposal problem; so have air washers and catalysts. These de vices have not been used because the kinds and amounts of fumes released are constantly changing in research and are therefore un predictable. Despite the number of warnings in the literature, naasaap#--cone shaped covers or hoods fastened to the tops of vertical stacks--are still being used to prevent rain from entering exhaust stacks. It is vayaiiipmtMfc><lnt' theft" -use be avoided" compiwely. The pres ence of a rain cap or a modern mushroom type roof exhauster will invariably result in the release of fumes laterally across the roof and completely negate the high ver tical velocity needed to carry fumes above the cavity boundary and into the undisturbed air stream above the building. There are several simple stack drain ar rangements that will prevent entry of rain into exhaust stacks when fans are not operating. These are illustrated in Fig. 8. Building air intakes In high rise research buildings, mechanical equipment is frequent ly installed both in the penthouse and in the basement. To reduce air intake resistance, outdoor makeup air is often taken from both the roof and the ground floor. Because of the possibility of recirculating fumes released near the roof, every effort should be made to IJfflUi building iiH'llimi'B b en the' p--wattaf.wfmfwmd Jidu oj-.Uae btritdmg and away from fume ex hausts. A second floor location on the prevailing windward side, shown in Fig. 9, has the added advantage of avoiding most of the dust, odors, and airborne debris 108 Heating/Piping/Air Conditioning, March 1972 that may be present at the ground floor level. Assistance in determining the prevailing wind direction at the building site may be obtained from the local weather bureau. Some universities have meteorol ogy departments that maintain rec ords of wind speeds and direc tions. Such information is usually made available in the form of a wind rose giving the number of days during which the wind is expected to blow from various directions during the year- Architectural appearance The greatest objection to high vertical discharge of fumes usual ly conies from the architect, who dislikes having the profile of his building disfigured with a forest of stacks. To overcome this, some architects specify screens to sur round groups of stacks. Unless a screen permits easy passage of the wind, however, it will simply act as another higher wall with a sharp edge parapet, requiring additional height for the exhaust stacks. To overcome esthetic objections, it is important for the architect to plan for the release of fumes at the outset of his design by in corporating a feature such as an exhaust tower or clusters of ex haust stacks as art architectural element of the building. Fig. 9, for example, illustrates an exhaust tower located on the prevailing windward side of the building. It discharges fumes vertically upward at high velocity, well above the cavity boundary created by the prevailing wind and at the same time high enough to eject the fumes above most cavity bounda ries resulting from winds blowing from other directions. The loca tion of the tower on die prevail ing windward side of the building also provides a convenient makeup air intake for the building at the second floor level, as discussed in an earlier paragraph. Another possible solution to the esthetics problem is to gather the individual fume exhaust stacks over the building service cores as shown in Fig. 10. Bunching of the stacks in this manner also has the advantage of creating a mass of exhaust gases, which is much less readily deflected from upward vertical flow by wind gusts. Some designers combine the dis charges from many fume exhaust fans into one or more totally en closed stacks that discharge the fumes high above the building. If a closed stack is used, it is im portant to oversize the stack so that it offers negligible resistance to the flow of fumes. Otherwise, the backpressure will force the fumes to flow back through an exhaust system that happens to be turned off. Consider adjoining buildings Because fumes are being re leased into the air stream above a research building, the designer should consider the possibility of fumes being carried to nearby buildings. Since the behavior of air in and around a complex of buildings, trees, and other ob stacles is complicated, it is very difficult to predict the degree of contamination at an adjoining building. There are a few simple rules to observe, however. When the wind is blowing in the direction of adjoining build ings, as shown in Fig. 11. fumes released above the roof must be directed vertically upward to a level that will assure release above the combined wake boundary cre ated by the building complex. Thus, if the proposed research building lies entirely within the turbulent cavity behind a taller building, an exhaust stack from the shorter building should be raised to discharge above the boundary created by the taller building to avoid contamination of the lower building. H the lower building is located in front of a higher building, the exhaust stack from the shorter building must be raised nearly to the roof line of the taller building to avoid con tamination of the taller building. The distance between buildings, of course, will influence the amount of dilution occurring be tween buildings. The greater the distance, the greater will be the dilution and the less the need for high stacks. Where the building relationships are particularly complicated or die contaminating effects are uncer tain, it is best to conduct wiod tunnel investigations using simple models of the buildings, trees, and other obstacles involved to deter mine the location and height at which an exhaust stack should re lease its contaminants. There is in formation in the literature that outlines methods for calculating the degree of contamination (the bibliography entry marked with a *, for example); but it may be easier to consult a specialist who is . experienced in such matters. In any event, a predesign study will pay off handsomely in avoiding potential lawsuits and costly cor rections. Safety features Interconnection of hoods-- If two or more hoods independent ly serve a single room or an inter connected suite of rooms, all the hoods in these rooms should be electrically interconnected so that the operation of one will require the operation of all. Otherwise, there is a possibility that fumes will be drawn from a hood that is not in operation by the makeup air demands of those that are in operation. Alarm for hood malfunction --All hoods, especially those han dling highly toxic or dangerous materials, should be equipped with safety devices such as a sail switch, which will warn the hood user and Heating/Piping/Air Conditioning. March 1972 SAL 000045377 Fume hood, exhaust system design the building management that the air volume exhausted from the hood has dropped to a point where it will not provide sufficient cap ture velocity for further safe op eration. Fire dampers--Although most building codes require fire damp ers in all ducts that pass through fireproof walls or floors, it is im portant not to install them in fume exhaust systems. If a fire should occur in a hood, or if heat from a fire near such a damper in a fume exhaust duct should cause the damper to close, the occu pants in the laboratory and ad joining spaces might become cas ualties before anyone realized why the hood had ceased to function, Fire dampers are intended to pre vent transmission of fire from one part of a building to another through a closed circuit of duct work. Fume exhaust systems are not closed systems, however. They deliver air only to the outdoors in an upward direction, and they are therefore essentially chimneys. For this reason, fire dampers are not needed, and their presence is extremely hazardous. 100 percent exhaust from lab oratories--If materials being han dled in a laboratory are hazardous enough to require the installation of a fume hood, the presence of these materials within the labo ratory should also require 100 per cent exhaust of the air from that laboratory. Otherwise, an acciden tal spillage or accidental release of materials at a bench can result in fume recirculation throughout the building. Accidental recircula tion is a serious hazard to other ongoing experiments and to per sonnel elsewhere in the building. In the past, it has required emer gency evacuation in a number of cases. Cost of extra hoods Because research has become more complex and because the presence of a fume hood in a laboratory seems to have become a matter of prestige, there appears to be a tendeocy for research or ganizations to specify more fume hoods than may actually be needed for the research programs involved. Additional fume hoods are very expensive, and it becomes the duty of the designer to warn the owner regarding the high costs of exces sive hoods. In estimating such costs for the owner, the designer should include the following: Bibliography Fume hood design, selection Barrett, James C., "Ventilating Re search Labs," Air Engineering, Janu ary 1962, pp. 31-36. Barrett, James C., and others, comments on "The Design of Ex haust Systems and Discharge Stacks," by John H. Clarke, Healing/Piping/ Air Conditioning, August 1963, pp. 70, 80, 86, 88. Bond, Harry K., "Engineer's Guide to Dust Hoods and Dry Boxes," Electronic Design, March 1961, pp. 72-73. Brief. Richard S.. Church, Frank lin W,, and Hendricks, N. V., "De sign and Selection of Laboratory Hoods," Air Engineering, September, October, and November 1963. Bryant, R. K., "Hood Alarm and Monitoring System Assures Lab Per sonnel Safety," Heating/Piping/Air Conditioning, March 1970, pp. 8384. Burke. J. C., Jr., "Better Labora tory Hood Design," ASHRAE Jour nal, September 1961, pp. 47-50. Clarke, John H., "Planning for Safety-Ventilation in Plant Design," National Safety Congress Transac tions, October 1962. Coleman, H. S., ed., "Hoods and Other Fume Disposal Units," Labo ratory Design, Reinhold Publishing Corp., New York, 1951, pp. 31-34. "Design of Corrosion Resistant Vent System for UCLA Laboratory," Air Engineering, May 1966, pp. 1819. Fconomides. Leander, "F u m e Hoods," Laboratory Management, September-Octobr 1963, pp. 44-45. "Fume Hood Exhaust Systems," Laboratory Design Notes, NIH Div, of Educational and Research Facili ties. Office of Architecture and En gineering, Bethesda, Md., undated, 4 pp. Heider, Scott A., "Fume Hood Re search," ASHRAE Journal, February 1967, pp. 71, 90. Hemeon, W. C. L., "Laboratory Ventilation," Laboratory Design, Reinhold Publishing Corp., New York, 1951, pp. 61-68. Horowitz, Harold, "Planning for Fume Hoods in the Design of Sci ence Buildings," Journal of Chemical Education, May 1967, pp. A439A442. Horowitz, Harold, Heider, S. Af, and Dugan, Caldwell N., "Fume Hoods for Science Laboratories," AlA Journal, July 1965, pp 63-70. Horowitz, Harold, Heider, S.A., and Dugan, Caldwell N., "Hoods for Science Laboratories." Handbook of Laboratory Safety. Ch. 3.3, The Chemical Rubber Co., Cleveland, 1967, 12 pp. Ketcham, N. H., "Testing Labora tory Hoods--Evaluation of Design Changes and Periodic Performance Testing," American Industrial Hy giene Association Journal, August 1958, pp. 324-329. "Laboratories it Engine Test Fa cilities," ASHRAE Guide and Data Book, Applications, 1968, Ch. 10, American Society of Heating, Re frigerating and Air-Conditioning En gineers, Inc., New York. Lind. Arne. "Ventilated Cabinets in a Tuberculosis Laboratory," Bul letin of the World Health Organiza tion, Vol. 16, 1957, pp. 448-453, Nelson. L. N., "Handling Problem Laboratory Atmospheres," Research/ Development, December 1965, pp. 24-29. Papa, Louis J., "A Quantitative Approach to Proper Evaluation of Laboratory Fume Hoods," Air Engi neering, April 1966, pp. 20-22, 25, 30. . Peterson, J. E., "An Approach to a Rational Method of Recommend ing Face Velocities for Laboratory Hoods," American Industrial Hygiene Association Journal, August 1959, pp. 259-265. Peterson, J. E., and Peay, J. A., "Laboratory Fume Hoods and Tbeir Exhaust Systems," Air Conditioning, Heating and Ventilating May 1963, pp. 63-72. Ruddy, John M., "What To Con sider When Designing Fume Hoods for Medium Level Radioaative Con ditions," Heating!Piping!A ir Condi tioning, March 1958, pp. 128-131. Schulte, H. F., and others, "Evalu ation of Laboratory Fume Hoods," American Industrial Hygiene Asso ciation Journal, September 1954. Siefken, Gerald ., Supply-Air Fume Hood Evaluation, Summary Report, NIH Div. of Research Serv ices, Engineering Design Branch, Bethesda. Md., June 1963. 21 pp. Stockdale, W. G., A Discussion of Fume Hood Performance, Labora tory Design for Handling Radioac tive Materials, Conference Report No. 3, BRAB. National Academy of Sciences-National Research Council, Washington. D.C.. 1962. pp. 46-52. Syska & Hennessy. Inc., Labora- \ 4 First costs The total installation cost for each incremental fume hood. The total installation cost for each incremental fume exhaust system required. The pro rata portion of the installation cost for the additional heating and cooling capacity need* ed to condition the makeup air required by each hood. The pro rata portion of the cost for constructing the space oc cupied by the hood. Annual costs-- The annual cost for maintain* ing each additional hood. The annual cost for operating and maintaining each additional fume exhaust system. ; The pro rata share of the an nual cost for operating and main taining the additional heating and cooling capacity needed. The pro rata share of the an nual cost for operating and main taining the space occupied by the hood. Cost of investment-- The annual cost for the total investment for each hood. Before discussing these costs with the owner, the designer should make allowance for the makeup air required for ventilat ing the building. Since some of the hoods obviously can be used to exhaust the ventilation air, the operating costs for such hoods should not be included when to taling the cost of hoods that might otherwise be considered nonessen- tial to the total project. Although costs vary consider ably with locality, quality of con struction, and current rates of in- fiation. an estimate of costs is given in Table 1 as an example. It is assumed that the typical in cremental hood is a 5 ft unit with a 10 sq ft sash opening (4 ft wide by 2'/2 ft high) to be operated at 100 fpm face velocity, requiring 1000 cfm of air exhausted from the laboratory. The tabulated costs are very rough estimates and are not to be used for budgeting costs for a par ticular job ai a particular locality. The costs are intended only to fo cus attention on the importance of giving full consideration to the high costs of owning and operat ing fume hoods, especially those hoods that may not really be needed. Before presenting such an esti mate of costs to the owner, the designer should be prepared to _____ SAL 000045379_ tones and Technical Facilities, Part IT. Handling and Disposal of Fumes, New York, August 1966, 4 pp. Walls. H. L., and Metzner, W. P., "Fume Hoods--Safety vs. Costs," Industrial and Engineering Chemis try, April 1962, pp. 42-45. White, Marshall, Fume Hood Ex haust Duct Work Materials, NIH Div. of Research Studies, Engineer ing Design Branch, Bethesda, Md,, 1965, 10 pp. Fume exhaust and air intake Barry, P. J., Estimation of Down wind Concentration of Airborne Ef fluents Discharged in the Neighbor hood of Buildings, AECL-2043, Atomic Energy of Canada, Ltd., Chalk River, Ontario, Canada, July 1964, 15 pp. Breysse, Peter A., and. Lehman, George, "How To Dismantle Per chlorate Contaminated Lab Exhaust Systems," Heating/Piping/Air Condi tioning, August 1966, pp. 126-127. Changnon, Stanley A., Jr., "Se lected Rain-Wind Relations Applica ble to Stack Design," Heating/Piping/ Air Conditioning, March 1966, pp. 93-98. Clarke, John H., "Air Flow Around Buildings," Heating/Piping/ Air Conditioning, May 1967, pp. 145-154. Clarke, John H., "Design and Lo cation of Building Inlets and Outlets To Minimize Wind Effect and Build ing Re-entry of Exhaust Fumes," American Industrial Hygiene Associ ation Journal, May-June 1965, pp. 242-248. Clarke, John H,, "Effective Stack Design in Air Pollution Control," Heating ! Piping t Air Conditioning, March 1968, pp. 125-133. "What To Do About Fume Re- entry," Laboratory Design Notes, NIH Div. of Educational and Re search Facilities! Office of Architec ture and Engineering, Bethesda, Md., undated, 4 pp. Halitsky, James, "Diffusion of Vented Gas Around Buildings," APCA Journal, February 1962, pp. 74-80. Halitsky, James. "Estimation of Stack Height Required to Limit Con tamination of Building Air Intakes," American Industrial Hygiene Asso ciation Journal, March-April 1965, pp. 106-116. Halitsky, James, Gas Diffusion Near Buildings: Theoretical Concepts and Wind Tunnel Model Experiments with Prismatic Building Shapes, Re search Div., College of Engineering. New York University, New York, February 1963, 122 pp. Halitsky, James. Gas Diffusion Near Buildings, American Society of Heating, Refrigerating and Air-Con ditioning Engineers. Inc., New York, 1963, 18 pp. Hama, George M., and Dowring, Darwin A., "The Characteristics of Weather Caps." Air Engineering, December 1963, pp. 34-37. Hammer, Willie, "What an En gineer Should Know About Micrometeorology," Heating / Piping / Air Conditioning, May 1965. pp. 101-105. Hewson. E. W.. "Stack Heights Required To Minimize Ground Con centrations," ASME Transactions, Oc tober 1955, pp. 1163-1172. Houlihan, Thomas F., "Effect of Relative Wind on Supply Air Sys tems," ASHRAE Journal, July 1965, pp. 28-31. "Laboratories &. Engine Test Fa cilities, "ASHRAE Guide and Data Book, Applications, 1968, Ch. 10, American Society of Heating, Refrig erating and Air-Conditioning Engi neers, Inc., New York. Lord, G. Ross, Baines, W. D., and Leutheusser, Hans J., The Minimum Height of Roof-Mounted Chimneys, Technical Publication Series TP 6409, University of Toronto, Toronto, Canada. October 1964, 15 pp. Martin, James E,, The Correlation of Wind Tunnel and Field Measure ments of Gas Diffusion Using Kryp ton-85 as a Tracer, thesis, Michigan Memorial Phoenix Project, Univer sity of Michigan, Ann Arbor, Mich., June 1965, 130 pp. "New Light Thrown on the Vent- to-Intake Short Circuit," Air Condi tioning, Heating and ' Ventilating, July 1961, p. 81. "Radiation Routs 100 at Hopkins," The Evening Star. Washington, D.C., February 21, 1967, Sherlock, R. H., and Stalker, E. A., A Study of Flow Phenomena in the Wake of Smokestacks, Engi neering Research Bulletin No. 29, Dept, of Engineering Research, Uni versity of Michigan. Ann Arbor, Mich., March 1941, 48 pp, Steere, Norman V., "Ventilation of Laboratory Operations. Part If." Jour nal of Chemical Education, February 1964, pp. A95-A1I0. Steere, Norman V., "Ventilation of Laboratory Operations. Part II," Jour nal of Chemical Education, March 1964, pp. A183-A188. Fume hood, exhaust system design suggest alternatives for the addi tional hoods originally desired. One of these is the exhaust box or glovebox type enclosure described earlier, which requires much less makeup air than a fume hood. An other possibility is to provide double or triple horizontal sliding sashes in place of a single vertical sash, thus limiting the maximum air flow through the sash opening. Another possibility, not recom mended, is to provide auxiliary air for noncrttical hoods as discussed earlier. An economic analysis must first be made, however, to deter mine whether a genuine saving can be realized with such a system (bearing in mind especially that auxiliary air hoods require the in stallation of individual tempered air supplies in addition to the items of costs mentioned above). Since most alternatives are costly and not at all satisfactory, simply not installing unneeded hoods is definitely the best solution. Maintenance all-important Since hood performance may be greatly affected by the cleanliness of the exhaust system and the di rection of rotation of the exhaust fan, it is important to provide an enforced schedule of inspections and performance tests during the year to make sure that the fume hoods are operating safely and satisfactorily. If filters are used to remove particulate matter from the ex haust air, they must be periodical ly inspected and replaced. Also, special attention should be given to possible corrosion of ducts and damper mechanisms and to the collection of debris along the duct As mentioned earlier, excessive corrosion may cause air leakage into the ducts or failure of bal ancing dampers, thereby reducing the already critical capture veloc ity. It is especially important to remember that vitMgnHMM# the llklllj IU UU Hie most commaiE:causg`<rf"'yiiiiJ huuii pertomMefe' Performance tests All hoods shoutd be perodically tested for satisfactory operation whether they are new or have been in use for many years. Two tests are usually performed, one for fume leakage.'and the other for face, velocity. The test for fume leakage consists of releasing 112 Table 1--Investment cost and annual owning and operating cost are tabulated on a per hood basis to illustrate what items should be considered in an economic analysis of an exhaust hood system and give an idea of the magnitude of the costs involved. Costs will vary for particular installa tions depending on such factors as system design, hours of operation, local energy costs, accounting method used to determine annual charges on investment, and applicable insurance premium and/or property taxes. Investment cost (installed) per hood Cost of hood and bench (5 ft hood length at $280 per lin ft) Cost of hood utilities, including 20 ft of piping (water, $125; gas, $100; air, $75; electricity, $80; drain, $70) Cost of hood exhaust system (fan, $400; roof opening, $100; wiring, $300; 75 ft of duct at $10 per ft Pro rata cost of heating and air conditioning equipment (6 tons for 1000 cfm of outside air at $1500 per ton) Pro rata cost of space occupied by hood (5 by 3 ft floor space at $80 per sq ft $ 1,400 450 1,550 9,000 1,200 Total investment cost $13,600 Annual owning and operating cost per hood Cost to maintain and test hood Cost to maintain and test exhaust system Pro rata cost to maintain heating and air conditioning system (6 tons times $4000 per year maintenance contract for 200 tonsystem) Pro rata cost to maintain building (5 by 3 ft floorspace at$3 per sq ft per yr) Cost to operate 1000 cfm exhaust system (1 KW per hp per hr for 1 hp times 24 hr per day times 365 days per yr at $0,015 per KWH) Pro rata cost to operate cooling system (6 tons for 1000 cfm of outside air at 1.25 KW per ton times 1200 hr per yr at $0,015 per KWH) Pro rata cost to operate heating system (110 Btuh per cfm times 1500 hr per yr at $0.10 per therm) Pro rata cost to operate building air supply fan for 1000 cfm makeup air (assume same costs as for exhaust fan) 25 40 120 45 130 135 165 130 Total operating cost 5 percent per yr for $13,600 investment $ 790 680 Annual owning and operating cost $ 1.470 odorous fumes such as ammonia or hydrogen sulfide within the hood. If fumes are detected out side the hood, especially around the face opening, the capture ve locity at the sash opening may be be inadequate, or there may be an interfering air disturbance. Cleaning the exhaust system, ad justing the air flow damper, or increasing the fan speed may im prove the performance if low face velocity seems to be the problem. If, on the other hand, the leakage seems to be caused by in terference from a stream of auxil iary air supply or other air veloc ity near the sash, the nature of the interference may be investi gated by placing liquid titanium tetrachloride on masking tape placed around the periphery of the sash opening and on lengths of string strung as a series of grid traverses over the entire sash opening. Observation can then be made of the path of visible fumes to determine where there is spillage into the room. Smoke bombs have also been used to de termine air flow patterns at sash openings and identify interfer ences. A ally used to measure actual face ' velocity. This is done as a trav erse over the entire sash opening, including especially all edges and comers. The overall average face velocity is obtained by averaging the velocity readings at prescribed positions of the traverse. Since the tests described above depend on subjective observation and on testing procedures that are difficult to standardize, they are considered to be unadaptable and inadequate. For this reason, ASHRAE has set up a research project for developing fume hood per formance criteria and new test procedures. Although ASHRAE has approached fume hood users and manufacturers for support, no funds have been made available so far. Because of national con cern over air pollution, however, and the trend toward suing equip ment manufacturers and building owners because of unsafe equip ment designs or unsafe working conditions, the need for criteria and test procedures is becoming urgent, and special efforts are again being made to find a way to develop the much needed per formance criteria and suitable test procedures. 00 45380 Heating/Ptping/Air Conditioning, March 1972 ^jtM E . L . WALLS W. P. METZNCR FUME HOODS -- SAFETY VS. COSTS Size, air velocity, tctal flow, and economy are compatible The two requirements for fume hood operation-- safety and economy--seem almost incompatible 1650 c.f.m. would severely limit the number of fume hoods that could be located in a laboratory room. in the light of increasing safety requirements and the Building costs are constantly increasing. In addition, ever increasing costs of building and operation. during the last decade advancing technology' and use of Little question can be raised regarding the need for new and unknown materials have raised the commonly fume hoods. They exist for the sole purpose of pro accepted safe face velocity of chemical fume hoods from tecting the user from the toxic or noxious effects of the 50 to 100 f.p.m. (7). This is still going up. Some chemicals or other contaminants with which he works. materials are already considered hazardous enough' to The extent and manner in which it is done, however, is require face velocities in excess of 200 f.p.m. The matter probably the most controversial subject one can choose of face velocity recommendations is well covered in the in the design of today's research, works, and teaching laboratories. Countless articles and books have been written on the subject, but few of these, if any, have really come to grips with the problem of providing a safe fume hood which is economical in both capital and WHY AVAILABLE FUME HOODS DID NOT MEASURE UP operating costs. The conventional fume hood was obviously not accept miomty--air moving from the operator across the contaminating source--determines m mU Regardless of how much total air a fume hood has moving through it, if it doesn't have sufficient face velocity it isn't safe. The source of this air {the working air) is from the room, which in most of today's laboratories is conditioned air. Tkis is the major cost item. Thus, the crux of good fume hood design is providing maximum face velocity at minimum total air movement. able. A very high volume of air lup to 1600 c.f.m.) is re quired to produce safe face velocity when the hood is open far enough to permit freedom of work. The auxiliary-oir hood with internally supplied air was dismissed because it accomplishes absolutely nothing toward reducing the air required from the room to assure a sofe face velocity. In fact, our experience with this type of hood was that it tended to "bounce" air out of the hood. None of these hoods with on auxiliary air fan is used at Monsanto today. The rule-of-thumb correlation between fume-hood The auxiliary-air hood with supplementary air exhaust and air conditioning refrigeration is that each supplied just outside and pulled through the face of the 2fiftmxbic feet per mimiM1 hood, is a satisfactory choice from the safety standpoint. 1 MB of refrigeration, or in terms of cost, equals ap Agoin, large quantities of air must be handled. Also, the proximately S1000 in capital investment. Thus a con auxiliary air, although tempered, passing over the heod of ventional 6-fuot wide fume hood, ranging in cost from the user is not too desirable. S'JOO to 31400, having a face opening of 16.5 square feet Both of the auxiliary air hoods must be provided with o \\ ith a fare velocity of 100 feet per minute (f.p.m.) would bypass to take air from the room when the sash is closed, require an investment in an air conditioning system of or the system wifi be hard to balance. Such a system is also approximately S8200, to say nothing of what it would expensive because it requires three ductwork systems-- cost to operate the system. Also, the requirement of two for supply and one for exhaust. 42 INDUSTRIAL AMD ENGINEERING CHEMISTRY J Gt 2961 1IM6V t 'ON tS TOA ii^OOOO lk;!p, `OOUBuuojJOd. 4Bi;aiis ipiM pooq jBuonuaAuoo b joj papna susuidinbo SuiuouipuoD jib ui ju3ui3S3Aui sip jo %09 J 09 -^!uo SDJinbaj ji 3i3 -}iun jod 00lS Pub 00US uaoMjaq--luomdinbo jeuopuaA -uoo ueq3 sjoui ojhij sjsoo pooq auinj sup `isoo joj sy pooy >i) Jo iW]fpuv tapis >\j] jo saffoq ajffiuis ySnoji/j paanpoxpn si Jty satuij iiv jd pooy yj ySnosyj mojfnv ay) Swyoj.y-- ysvs Fuutj puoijuMuos ay) fo potyui ysvs Svipijt pjuosuoij v fo as/2-- :a/qisso<j aauvuuojiajyons apvm suotpaouvi lolvui ocnj^ pooiiBjEq X|isb3 aq ubo uiajsXs Smuopipuoo jib aqj^ -jian sip ojui j[mq sjsm aaucuasuiBtu Asea puB Aouaioyp uBinnq tonniixBjq 'ajqtssaooB XjisBa sbsjb jjb qjiM `popiAOjd sba\ 2uiuado 300J-9 y 'juauidinba IJ3IAV p3|]y SBA*i B3JB pooq Oqj U3q,\\ U3A3 sA\oy JIB JO ui-j-d 099 -Muo TP!V' "pappojd sba\ -ui'd-j qoi jo .\3p0pA 3DBJ y '3JES SBA\ pouhlSOp pooq H[3 `UJ.Hlj H'liq S)I 11] AJ]BDIUI0U03.-- qjoq OpiAOjd 03 punoj aq 03 psq qoBOjddc a\3U e `uojp qjjvrosaj aip 03 piuouiujap aq p]no.v\ spomj jo joquinu .10 azi* outonp -aj jaqip 33U)S pasiuiojdujoo Dq iou p|noA\ .\3pj13g S3]npoui i(iOjv.ioqv] QS f ^uipjinq ay) uof xua/si's v uSisap o p -- SJS03 Suijviatfo puv ]V)iJvo 3\qvuosv3x uwjuiviu 0 / -- i()i3ojati 330J ySty v jofpau 31/3 aziuSoaai vj^ -- ajnpout jo)Bioqt>i </ov3 0; u3m xnoj ioj apxnonj oj^ -- uviu yava 10fpooy auinj apim }oof-g v aptaojcf oj^ -- SBA\ ` OJ^ `jnaOQ 3A3J3 JB J33U33 lpJB3S3"y A\3U S'OQ JBDIUjaqj OJUBSUOJ^I jo uSisap aqj ui paoBj aq 03 pnq 3Bqj uiaiqojd aq_L t( spooj^ XJOiBJoqB^ joj sanioop^ aobj 5utpuaunuooa,a jo poqjapvj ]euoi;b^j b 03 qoBOiddy uy,, `(`) appjB pooy ySneiyi-yjDm v vitof oj y:vq oj yovq ptjfvjsvi >q uvi spooy omj `suoijvai/ipout tySijs yii^f 'sitfvjvtffo *2joj Jo jtmourn puv uojiDjfDjsut 9)V}t]uuf of *y9VJf ayj Butpnpui `ptoouitJ Xjtj/jJutoj q uva uotws umo] ayj' Tuozpif ysvs stvjS Sutpijs amoj puv aatjcjn Hq tySpy sfi jtnySnojyj f]qtss3M sxjp 'mjo/txfcjo jjta} }oof-gi a\vpommooavpim puv m`f-3 00 of 0S9 tKn pooy ui-yjom >yj, `w/a Qgg Jo aSavyastp uiv jv)0f o fv uot)isotf uxfo spjtyi-ott) fyf ui 'wJ'J $ puv uotjisoJ uJo pjryj-auo ryj ui 'uttf-J QZl f ^lioopn aavj o sppiS pooy ppoui-yauaq *yj. 'jwj uonuntiut id uopvujo >Jvs rptaoit} spooy mvj cnau sayj^ S310NVH 3A1VA Alllim f Figure l. The original prototype hood. Immediately above the sash oftning is a hinged blow-out panel held in place by magnets. The p.mel is designed to relievepressure within the hood in the event of an accident and to reduce the chances of the glass sash blowing com pletely out of the frame. (The glass is Safiex safety plate). The deep well provided at the side of the hood permits the use of tall apparatus. Power is taken into the hood through openings at each of thefront comers Figure 3. Prefabricated piping and electrical systems are seen in place on the hood prior to connection to the main service distribution system. All services except the drain are in the superstructure, so that the hoodis quite easily moved Figure 2. Smoke tests assure air movement awayfrom the hoodface In the final hood, static pressure variation cannot be measured with ordinary instruments with the hood open or closed. This, coupled with the building's double duct high velocity air conditioning system, makes balancing of the total system quite easy. The hood has eye appeal with clean, easily maintained lines. Acknowledgment, The authors wish to acknowledge the contributions of IV. .-1. Cleneot. J. R. Kendall, C. C. fVirts, and the many chemists of Mons'ir.'n Chemical Co. who made the design of these hoods possible. LITERATURE CITED (1) Rarrctc. .Tames C., Safety Maintenance 119, No. 1 (January 1960). (2) Kctcb un. N. H., Am. bid. Hyg. Assoc. Quart. 19 (August 1958). O) IVrt-rson, J. F . Ibid., 20 (August 1939). Figure 4. The second prototype included many changes suggested by testing. The redesigned bottom grill was a major improvement SAL 000045384 VOL 54 NO. 4 APRIL 1962 45 SPECIFICATIONS The fume hood must satisfy the size requirements for conducting experiments in oil branches of chemical research except rodiological chemistry. It must provide safe foce velocities for the work being performed. It should use no more than 550 c.f.m. of room air, due to the room size and occupancy (22 X 30 feet with 4 men). An excess of 2200 c.f.m. of oir movement in the room would result in high discomfort to the occupants. It must be easily balanced os a part of the total air condi tioning system. No air from the laboratory rooms is to be recirculated in order to reduce the danger of experiment contamination. Il must be easily movable. It must be easily maintained and aesthetically acceptable. Although listed and discussed separately, these speci fications are interrelated, and they must be considered simultaneously in the design. Basic Heed Design The first three; namely, size, velocity and total air flow were the key factors which determined the over-all type of fume hood. They seemed so incompatible that the job at first seemed impossible. How could a hood 6 feet wide be used freely while maintaining a face velocity of 100 f.p.m., or better, with only 550 c.f.m. of total air available? Every portion of the hood space should be readily and easily accessible to the user. Yet the hood opening must not exceed 5.5 square feet. The only way these criteria could be satisfied was the use of horizontal sliding sash in lieu of the conventional rising sash. For the 6-foot wide hood, a three-sectioned sash was chosen which, when one third open, yielded 100 f.p.m. velocity with a 550 c.fm. air movement. Thus, three of the requirements were satisfied. The next factor, air balance, was not quite so simple. Air flow formulas were of little value and the design had to be empirical. To reduce the total building space requirement, a decision was made to use a single exhaust fan for the four hoods within the laboratory module. If the air conditioning system was to be balanced it was mandatory that each hood, regardless of sash position, maintain as near constant static pressure drop as possible. The "by-pass'' damper system used on many hoods hav ing a rising sash was not practical with the horizontal sliding sash. The answer to this problem was to make the air flow through the hood at all times. Air was intro duced into the hood at the sides with a simple baffling system sufficient to prevent splash-outs. Severe tur- AUTHORS. E. L. Walls is a Laboratory Design Consultant and a Registered Professional Engineer in St. Louis, Mo. He was Project Architect for Monsanto's Research Center in St. Louis and designed the fume hoods described here. W. P. Meitner is Director of the Monsanto Research Center. 44 INDUSTRIAL AND ENGINEERING CHEMISTRY bulence caused by the side-entering air was e'ffcctivefy eliminated by introducing air at the bottom of the sash through a grill placed level with and between the front edge of the bench top and the sash frame (Figure 1). A stainless steel trough was designed to run under the full length of the grill to catch liquid spills. This trough doubles as a cup sink, for draining off condenser water, etc. Prototype Testing A prototype hood was built and installed in a full size laboratory mock-up at Monsanto's Dayton, Ohio, Laboratory. This approach to testing cannot be over emphasized. Any fume hood is subject to the condi tions of its environment and it is rather common for a hood to perform satisfactorily in one place and fail completely in another. Some of the factors affecting the performance of a fume hood are location within the space relative to traffic patterns, proximity to doors and heat-yielding sources, room air currents, and tempera tures. Smoke tests (Figure 2) were used to establish flow patterns. Tests of the first prototype are a subject within them selves (2). This hood when empty, was safe--i.e., it had no negative flow through a one third open position, with only 300 c.f.m. total air exhaust. Introduction of equip ment into the hood, however, required raising the total air to 410 c.f.m. With the sash closed completely, air entering the bottom opening rose so rapidly across the sash that the Venturi effect caused aspiration of air toward the front. Thus, when the hood was opened, there was a tendency for light vapors to be swept into the room. Although this boil-out was immediately swept back into the hood, it definitely crossed beyond the face of the sash and was highly undesirable. Figure 4 shows the second prototype. The bottom grill was redesigned with smaller openings to eliminate the Venturi effect at the sash opening. Several changes were made for more efficient use. The bottom grill was hinged longitudinally to permit easy cleaning of the trough and to permit easy passage of electric cords into the hood across the entire face. The electric service was incorporated into the frame of the sash and portable variable transformers were placed on a shelf immediately under the trough. __ The side baffles (which were "knuckle knockers") were changed to louvers in the face of the sides with the valve handle extensions passing through the louver. The lights were installed in an inverted drawer accessible by opening the blow-out panel, which permitted use of the space above the hood for storage. Tests of the second prototype resulted in placement of guiding vanes on the lower side of the bottom grill and addition of slots in the back baffle to produce a more laminar flow of air across the hood floor; reduction in the size of the side louvers to reduce the air flow; and a reduction in the height of the sash opening which in creased the face velocity without impairing visibility. These changes were incorporated into the final hood design. Si m Esso Research and Engineering Company (FORMERLY STANDARD OIL DEVELOPMENT COMPANY) P. O. Box 45, Linden. N. J. MEDICAL RESEARCH DIVISION TlLIPHONti WAIUH 5-1600 July 2k, 1962 Mr. F. S. Venable Humble Oil & Refining Company Manufacturing Division Baton Rouge Refinery P. 0. Box 551 Baton Rouge 1, Iouisiana Dear Fred: Attached is a draft of some material vhich ve have put together on laboratory hoods, and you requested this in light of your current in terest in hood installations. Please treat this in a confidential manner as it is being considered for publication and will be subject to scsae changes. Very truly yours. H. V. HENDRICKS NVH:WK Att. - R.D* on "Chemical Type Laboratory Hoods" - 7/1962 NVH c. ^ 1 0 00 0 4 b 3 3 -J CH3CCAL TgS LABORATORY HQ0D3 1. comnsRssp iigJKttistfe a bocds Fume hoods have long bean a standard fixture In chemical labora tories where, until recently, little attention has been given to actual hood performance. The requirements of new laboratories, especially re search facilities have introduced new concepts both In design operation of chemical hoods. Several factors have became important and each has a bearing on the requirements for design and operation. Xn many new laboratories, hood installations are much more exten sive than in older facilities. Some laboratories may treat an entire vail length aa a hood, giving a considerable area of hood face with correspond ing large air flow requirements. Along with this la the acceptance of higher hood face velocities that ars required for adequate control of highly toxic materials. A different problem is seen in many older laboratories where original hood design and related appurtenances have been established. In these situations, the initial design called for lower air movement S ft L 00 0 0 4 fii 3 8 6 is acceptable today, and an often complicating factor ia the addition of other hood* to a fixed system or In son caaea other ventilation systems actually ccarpeting for the available air In the roams. Obviously, this results In levering the efficiency of an already questionable arrangement. Laboratory architects frequently use chectical hoods to exhaust the supply air to the laboratory building. Thus, the total exhaust from the hoods la balanced against air input. This arrangement becomes, critical in terms of building heating and air conditioning vhen total hood opera tion calls for large volumes of air. Depending on actual hood requirements this say even rule out air conditioning for seme laboratories, aa veil aa imposing nnniifti heating problems for others. These nnginnerlng and architectural problems related to hoods certainly Justify serious consideration as to vhat can be done to make maximum use of older hood installations, and further, what can be done with the design of new Installations to minimize these problems. There Is a great need for sound criteria for the design and operation of chemical hoods. 2. K2DICAL RESEARCH DIVISION ggBREST IH HOODS In the course of industrial hygiene studies In a number of loca tions, the Medical Research Division of Esso Research and Engineering Ccm- "pany has made ai-ouuU several hundred chemical hoods under a vide variety of conditlone and at many locations. These have covered velocity and air flow measurements as veil as atmospheric sampling for contaminants around hood' areas. These studies clearly demonstrated need for and value of criteria for uniform hood operations. Da 1958 the Medical Eesearch Division suggested a classification schedule for laboratory hoods In order to take advantage of lover air flov requirements where vork does not involve highly toxic materials, die schedule relates air flov requirements to toxicity of materials handled. These principles proposed at this time are considered valid today. As Interest in hood design and operation increased, further study vas given to this area in i960 and reccooendatlona on laboratory fume hood Installations were prepared. 'Oils material vas broader than the schedule suggested in 1958, and included a number of p features. It vas prepared  after visits to a number of locations vhere studies on hood design and operation had been made, The present discussion brings together the features of these earlier reports and also Includes nav material developed since I960* CHEMICAL TIPS LABORATORY HOOPS 3. HOOP OBJBCTIVBa AJP USAgg When air Is moved through a hood, there should be only one primary objective and this Is the capture and removal of any material vhlch Is emitted Inside of the hood. Air iiorwut for any other purpose in the laboratory can be dose in a wet more economical and efficient manner by other means. The seeded vnlvmm of air to effect capture and removal not only depends on the physical characteristics of the hood and its location, but also on hood objectives sad usage. Depending on the toxicity of the material being handled, the effeetlveaeae of capture and removal can vary over a vide range frcm complete removal for highly trade materials to lover orders of removal efficiency for less toxic substances. For example, -k- 8 A L 0 0 0 0 4 ':;j 3 8 9 I if only vater vapor Is being emitted, the effectiveness needs only to be sufficient to control humidity conditions in the laboratory. If, hcwever, a material such as nickel carbonyl is involved, tbs removal efficiency needs to he complete. Hoods may serve puzposes in addition to removal of emitted materials from either continuous or periodic operations, particularly vhere more than one hood per laboratory is available. Invariably, the hoods are interconnected vith the exhaust system and, often as not, this requirement bee rums the primary function during design stages. Hoods are sometimes used as a safety enclosure of an operation vhich does not emit any material unless an untoward event occurs* Designing for a sufficient volume of air wove--at to contain an explosion is generally uxxvarrentsd. Hoods are also used only for heat removal from hot operations or for hasardcos chemical storage* All of these feetors most be considered vhen deciding cm the volume of air movement needed* k. SCMB FACTORS AT7SCTU0 HOOP OPERATION there are a number of characteristics related to materials used -5- AI... 0 0 0 0 4 5 3 9 0 I In a hood that can Influence hood operation. In addition, there are physi cal factors associated vith activities in the laboratory that may became important. Sobs of the more Important ones are listed belov. A. Characteristics of Materials Handled > Liquids - vapor pressure, temperature, moleeular velght, latent heat vaporization and toxicity. 6 Solids - particle concentration and size distribution in the air toxicity* t Oases - concentration, specific gravity, toxicity. B. Condition of Bood Usage Contaminant emission rate. Contaminant amission velocity. location and direction of emission. C. Disturbing Effects f Traffic past hood. Roca air currents. Heat rate in hood. -6 - SAL 000045391 t Movement of' equipment In hood. Obstruction of air pattern by equipment in .hoed. These Items suggest that careful consideration oust be given to a cumber of factors vhea arriving at the theoretical design of a hood system. Same of these components can be expressed in a quantitative fashion, vhile others, ttongh important, are difficult to describe in a quantitative fashion end are eves sore elusive in terms of Inclusion in actual hood design. Another item is the coat of hood installation and operation. System design beings vlth the hood, and ones these characteristics arm sorted out, It automatically establishes the design of the rest of the system, such as volume of air, duct vork and air movers. It becomes evident that in order to keep these items at tha lamest level cornel stent vith satisfactory operation, considerable effort can be Justified In evaluation of hood feetors and actual design. Two other major factors arm involved, not only in the capital investment, but also In operating expense. These are any air conditioning or heating load that may be occasioned by removal of air from the laboratory. -7- JL Tie extent of these factors are, of course, largely dependent on local climatic conditions* 5. IMPQRTAffT DBSIDB AHD USB CBTCBRIA The following list of recceeaandatiana will provide guide Itnee foe Judgment on hood and air supply location, hood design and hood operation* A. Location* 1* The hood should not he located near aa epee crtnrtrwr, adjoining heavy traffls^adaser doarwaye The presence of roon turbulence caused by these feetore, and poorly located roon air supply outlets, say negate the usefulness of the laboratory hood* 2* Air supply to the laboratory Is hast aoccnpl 1 shed hy use of a perforated celling aa thenxpply pianos* High velocity, supply grills are not desirable because of created air turbulence which nay necessi tate face velocities for adequate hood control* Supply pleauas In the celling, on the other head, haws been shown to have tr--endnus air flow capacity with low turbulence and may be more econcedcal In use for this reason* -8- SAL 0000453? L 3. Spot local exhausters that can be moved to the point of contaminant release, using adjustable blest-gates on dTlaaethle ducts, sen be utilised to advantage. Such units are highly desirable In areas where eeissloa-cresting laboratory work nay be conducted outalde of the hood. 4. Additional exhaust grills in laboratories (other than hooda) can be provided to withdraw supplied conditioned air when the hoods are not operating. 5* Che use of an ea^MWf--*enppl/ to conserve on air-conditioned, sir exhausted by hoods can be Justified only If the auxiliary sir la treated, primarily to prevent eondenaetion probleea. It is reecMeedsd that this air ba brought to near root t--paiatme and outalde mbient timidity, If It la used. 6. Auxiliary air. If used, should ba supplied outside of the hood proper through either m perforated celling or a low velocity grill. \ Auxiliary air supplied Inside the hood disturbs the patterns of rocn sir entry into the hood, and nakes fuae escape highly probable. -9- SAL 000045394 B. resign Feature* 1. The edge* ot the hood should he shaped to give a smooth miat-Soil entrance - Such a shape Is particularly neeessaxy *t the tottaa edgevtoere the optloua design vould be a sueetew allthee. A clear ance should be provided between the bottom air foil and the beach- top to allow air flow under the foil at all times. This action provides a constant flow of air into the hood, clearing It even when the doors are closed* The side edges of the hood are not as critical as the botton edge- sparing ttMddn so they are at a'''b9*fttj0Ne. angle* vlth the hood Is an acceptable substitute for an elliptical section- Cm tapered side* should be at least 6 la. vide and have the edges rounded. 2- The baefc beetle la the hood ahculd have at least three openings, one at the Intersection of the baffle and the bench-top, and the second, at the intersection of the baffle and the hood face ar top, and the third between these two- These openings should, run the full length of the hood, and each should have a open area equal - 10 - to the outlet duct are*. Adjustable slides may be used la conjunc tion with these baffle openings to establish the desired uniformity at the face after installation. 3. Each hood should here its ovn fan exhausting to an individual stack. In critical situations, fan speed and else can then be changed easily to meet changing requirements. Plenums and the use of fans to exhaust many hoods are considered to be undesirable mainly be cause of the Inflexibility of the system and Its v&steful exhaust capacity. Tor an existing building using a plenum esbeuet system, additional hoods might be accommodated even at limited axhanet and supply capacity If the hoods are placed on an on-off operations using electrically controlled deepere. ns dampers would open Into the plen system only vban the hood Is turned on. Hood use factors as lorn as 55 pm cent here bees noted in some laboratories. Advan tage may be taken of this fast in some esses share careful study has bean done and controlled hood use can be effected. She fans should be controllable near the hood face with as on-off 11 - BAL 000045396 switch. Two-speed fan operation. If desired, can be accomplished by a two-spaed motor tor most fans or, at lower cost, by a revers ing switch for forward-curve, centrifugal fans. 5. The exhaust stacks should not have weather protection that requires the air to change Its direction upon discharge* So weather protec tion, or else a butterfly damper cover on the top of the stack eiwiirt he used. Sew style weather exhausts have been developed and good results are reported with than. 6. Where open face veloeitlee exceed 12$ feet per minute, the Install atlon of an atmospheric damper Is suggested to prevent excessive velocities when the open face area la reduced upon closing the sash* Hood Operation 1. Equipment placed within the hood should be so located that the points of contaminant release are s**aiimmt3BMNnm!^mrteJgmsa*lM*iMefe4eee.. Pests have shewn that even at very high face velocity, contaminants located Mar the face of the hood can escape Into the room by "followout" whan an individual walks past the hood face. This point Is of I paramount importance. Laboratory acceptance of the principle of working back in tbs hood can be leaplamented by placing a X/h in. thick edging, 6 In. vide on the bench-top near the hood face. Any similar device to attain a "defense in depth" is warranted. 2. 'Fh* lower hood baffle opening should not be obstructed with large objects or many mall ones. Bottles or chemicals that require gone ventilation should be placed in a ventilated storage area or on separate shelves provided in the hood proper. 3* Concentrated heat loads within the hood presorting about 5,000 watts or 1,000 vatts per foot of hood length create thermal vectors that require higher face velocities for adequate control. 6. CIASamCATIOH SCHBULB for may!mum flexibility in the use of hoods, and also for health protection, tbs ideal situation would be to provide universal hretis with an air flow which would Insure complete capture of materials re leased within the hood. However, a reduction in investment can be effected by on alternate procedure. This would Involve classifying hoods for different - 13 - RAL- 0 0 0 0 4boR servlca. This assumes -that only normal laboratory work would be done and that no unusually large quantities of materials would be involved or re leased. Although the advantage of a reduction in investment has been pointed out, there is the disadvantage of lowering the flexibility of hood use and the further disadvantage of offering no overall modification of ventilation rates once an installation of this type has been made. In establishing categories of hoods, the following classes are suggested. A. Class 8 hoods are reccsnended for handl ing highly toxic materials where complete control la required. Tor work fall ing into this category, hoods should operate at a face velocity of 125 ft./minute. Such hoods should have special design characteristics depending on each Individual situation. Materials handled in these hoods would tetraethyl lead, radioactive materials, beryllium compounds, carbonyls, and materials having a similar high order of toxicity. & Aa a guide, Class S hoods would be used to handle materials SAL 00004539* 14 vlth a maxljBM allowable concentration (MAC^) of <.l pjaa for gas and vapora or <.l mg/cu.m. for dusts, fumes and mists. Class A floods. would operate at a face velocity of 80 ft/mlnute. These could be used for any operation vlth the exception of those offering potentially aevere hazards due to a high order of toxicity of materials to be handled* Basically these hoods vould be used for materials having from a moderate to a high order of toxicity* For reference. Class A hoods could be specified for materials vhose MAC'S are in the range > .1 to < 100 pps for gas vapors, ^ .1 to < 15 mg/cu.m. for dusts, fuses and mists, and - 5 W.P.P.C.F. for mineral dusts* C^e^WWW^BWmvould operate at a face velocity of 50 ft./minute* These could be used for all operations vhere the hazard Is not unusual due to a moderate or lov order of toxicity of materials to be handled. This vould Include essentially all hydrocarbons - 15 - S' A L 00004540 I asd most vould represent a great portion of the work required under a hood In most petroleum laboratories* Class B >**** could be utilized whan the contained materials hare the following MAC'S: - 100 pps for vapors and gases, - 1? mg/cu.a. for dusts, fumes and mists, and > 5 M.P.P.C.P. for mineral dusts* vhere: ppm - Parts of vapor or gas per million parts of air by volume* mg/cu*m* * Milligrams of dust, fume, or mist per cubic aster of air* M.P.P.C.F. - Millions of particles per cubie foot of air* These hood classifications are suggested for installations vhere more than one hood Is to be provided* For situations vhere one hood only is installed, a Class B hood is suggested. It Is suggested that vhere the toxicity of a material la unknown, but because of certain characteristics, it is suspected of possessing a high order of toxicity, the material be handled under a Class S hood* It is further suggested that any other materials not given on the established MAC - 16 - R AI... 0 0 0 0 4 b 4 0 I lists be handled under Class B hoods until toxicity information is obtained. On the basis of such information, -eeaigament to other hood classifications might be possible. In order to insure a reasonable uniformty of air flow across the hood face, it is suggested that w^eeea of'the fWrsindd-lawsa veOocitaf; lese *imms fio of the average rallies Vhen a hood system Is put Into opera-* tlon Its air movement Is at an Initial peak, then vill decrease over a period of ronnt*** or years. Part of this loss is recoverable and part non-reeoverable. Recovery nay be accomplished by malntenanne on the system, fan, duet vork etc. On the other hand, little can be dons about the nan-recoverable portion of the lose, and this may amount to 15-20J&. this point should be recognized and compensation made to cover it in the MNOTWHJVfepamiiM*- While it is pointed out that the greatest flexibility and highest order of protection can be obtained by an installation of universal hoods for normal laboratory use, the schedule outlined above may offer a reason able hood operation with considerable savings. The limitations of this type - 17 - S' A L. () o 00 4 5 4 q 2 I schedule should* however* he recognized. There will he the administrative or supervisory problem of seeing that these hoods are used for the category of materials for which they vere designed. Also* there is the case vhere only one hood Is available for general laboratory vorfc. In this situation ve would suggest a Class A hood. 7. kcod Tsaraao - coMPLiAacg ahd FSRrorouacB A. Instruments Suitable instruments for measuring air velocity include most of the hot wire thnnso sn--fetsi s and the swinging MaMimonster (velooeter). Care oust be exercised In selecting in struments* since many air velocity units do not accurately in dicate velocity below 100 JTM. Instruments found useful by Medical Research Division tnclnds the Alnor Therao-anemcaeter and the Alnor velcneter. Some mention should be made of the need for instrument cali bration. This equipment is quite sensitive and frandUng and shipping will frequently throw an instrument off calibration. - 18 - 000045403 SAl- Periodic calibration la essential. B. Teat Positions In testing the open area of a laboratory hood, XMamqmL tvmmm - afcDI&Id~be used, If possible, for the maximal opes position. With aaahas In the half open position, perhaps, as many as eight areas vould be appropriate. The individual doing the testing can quickly assess tbs number of test points needed to accurately neaeure the linear velocity into the hood account ing for the variations in velocity at the face. As a means of checking both the instruments and the air velocity, two tests are recconendad for checking the average velocity. The first test Is run vith tbs hood open to Its maximum open position} tbs second, run vith the hood half vay open* The average velocities ere obtained simply by averaging the results of the multi point traverse, and in each case, the average velocity is multiplied by Its corresponding open area during the testing* The same value for total volume withdrawn by the hood should be - 19 - SAL 000045404 obtained from the tvo teste. XT not, then retesting will be necessary to be sure that turbulence factors have not changed the readings at any one location. In this regard, it Is important also to test the hood for unusual air flow patterns by using aw/,v` tubes or titanium TiCl^ coated swabs. It is best to run the hood tests under actual operating condi tions, where i y-fvp^at*a stirrers, hot plates, burners, the like, would be used. Such items plan an important part in the effectiveness of the hood operation as measured by the tests outlined here. Tests on sn empty hood without any operation being conducted inside may not give representative results for the same hood being used with equipment such as that described. C. Wbaber of Tests Depending upon hood usage, 1ISSW!Bl9^Wi^SMWPBpiTOK'1hs often ae monthly for hishly changeable toxic experimentation. For other routine operations, yeStiy*f>M>CThg may be perfectly - 20 SAL 000045405 1 satisfactory. Bee hoods should be tasted initially to be sure they neat design-specifleat&oBS* Xf'tba hood Is accept able* tbs data vlll serve as a ccaparison for future testing. Shis is* of course* true for any hood. Data recorded near tbs hood vlll serve as an operating standard far future test ing. She critical point here is not the exact frequency of testing* but the necessity to retest periodically, and in the east Manner. A staple chart as developed by getchaa can be used for per--went record and comparative purposes* S* Evaluation of Results Unlfem flov across tbs face is considered to be important for controlling ceisrlons at all points in tbs hood, for this reason* s criteria of velocity distribution has been established* tbM^uaOeettyetasy gnlni shMH nnd*h^laae tbsn^Ottflft variation greater than this* it is suggested that tbs baffls platss be adjusted to give this dsalred uniformity. Depending upon the operation 21 - SAL. 000045406 being conducted in the hood, the average face velocity at least equal the average value suggested Is these criteria. In comparing teste vlth previous results. It la Important to . recognise that air flow may be decreased vlth time because ot such factors as dirt accumulation In the duets, fan belt slippage, fan wheel corrosion and deterioration. Prcu ex perience, It has been found that the decrease in air flow from initial operation may be as much as after a few years of operation. Thus, In repeated testing. It would be expected that the air flow levels would decreeae slightly between teats. As long as the average velocity and velocity distribution criteria sore Mt, however, no action vlll be necessary. Good alntenairs praotloe* should keep laboratory hood sys tem working properly. - 22 - Research and Development Section National Safety Council eyewash, Safety Shower Continued from preceding page The lever, if used, should be no more than 34 in. above floor height, with an 8-in. handle and require about 10 oz of pressure to activate it. All safety shower valves or control devices throughout the plant should be the same height, the same position, and operate at the same position rela tive to all features of the unit. All controls should be located so that they are directly in front of the employee when he is in a position to be neutralized by the water spray. The unit, once activated, should remain on untii deliberately deacti vated. The control device should operate in the direction of travel. Safely showers and eyewash facili ties should be marked in some manner. It is recommended that a light be located at the shower which turns on whenever the shift is in operation. Alarms should be installed that, when the shower is activated, sound in other areas of (he plant, especially the first aid facilities. The area around the shower or eyewash on the floor should be painted with a 3-ft circle, colored tile, or other markings (preferably high visibility yellow or green). Special facilities should be provided where caustic soda is used based on recommendations by the firm's con sulting medical practitioner. Eyewashes and safety showers should be full-flow tested for pressure and volume at least once a month for one minute and tests recorded. All units should be tested on a daily basis by a bump test. Employees should be trained and tested periodically in the use of eye wash facilities and safety showers. Written directions should be provided at the unit in large letters. --Bill Larson, Chairman R&D Section Training Committee `Adequate' Fume Hood Face Velocities Pose Problems This article "as abstracted by Gordon DeWall. chci-man, R&D Technical Pub lications Committee, .from a paper pre sented at the Research and Development Section program at the National Safety Congress October 19, 1976. WHAT IS an "adequate" face velocity Editor: L. E. Oldendorf Industrial Safety and Fire Protection Engineer U.S. Energy Research & Development Administration Argonne. Hi. 60439 Associate Editors: John M. Hickey Corporate Safety Services Rexnord, Inc. P.O. Box 2022 West Milwaukee, Wls. 53214 Robert G. Nemchin Manager, Office of Occupational Health and Safety Litton Bionetics, Inc. 5516 Nicholson Lane Kensington, Md. 20795 Ralph B. Wainright Supervisor of Industrial Hygiene and Safety American Cyanamid Co. 1937 W. Main St. Stamford, Conn. 06904 The information contained In this newalalt*r na baan obtained from aourcaa ballavad to ba reliable, and the editor* have exercised reasonable care to assure Ha accuracy. ' However, the National Safely Council does not guarantee that contents of this publication are correct, and statements attributed to other sources do not necessarily redact the opinion or position of the Council. 2 for a laboratory fume hood? Much has been written on this subject. Although there are some differences of opinion-, the consensus is that 100 linear feet per minute (fpm) is adequate for most operations involving toxic materials and ISO fpm is required for operations involving highly toxic chemicals such as carcinogens, carbon tetrachloride, beryllium, and others. Occupational Safety and Health Ad ministration (OSHA) standards specify that face velocities for certain car cinogens shall be an average of 150 ; fpm, with a minimum of 125 fpm. * However, face velocities are not spec ified for the balance of the chemical agents for which a threshold limit value (TLV) has been established. The problem is further complicated ' by the fact that what was adequate (of yesterday's research may be totally inadequate today because TLV's may have changed. A hood designed for amaterial with a TLV of 10-100 parts per million of air by volume (ppm) map not be adequate if the TLV is reduced to 1 ppm: Stated as simply as possible, an adequate face velocity for a laboratory fume hood is that face velocity necess ary to maintain the exposure con centration in the breathing zone below the appropriate ceiling values and/or the eight-hour time-weighted averages or TLV's. If exposures can be main tained below the appropriate action levels (normally 50 percent of the TLV), medical surveillance is not re quired, and air and biological monitor ing requirements are greatly reduced. investigative Study Made In order to better define the face velocity necessary to assure, an inves tigative study was undertaken at Oak Ridge National Laboratory. The data in Table 1 represent workroom air sam ples taken in the breathing zone during actual operations being performed in laboratory hoods using the chemicals indicated. The ranges of hood face velocities and concentrations, as well as the total number of samples for each contaminant, are given. The airborne concentrations for each contaminant have been averaged. Concentrations and TLV's are expressed in ppm, except those for mercury, which are in milligrams per cubic meter of air (mg/ mJ). The average concentration of each contaminant sampled was less than 50 percent of the respective TLV. Based on these data, it is apparent that a face velocity of 100 fpm is more than-* adequate to control breathing zone, concentrations of most contaminantsused in laboratory hood operations. As an alternative to performing breathing zone sampling on every em ployee potentially exposed to a toxic chemical, a system was devised using carbon monoxide as the test gas. Pure carbon monoxide gas was metered in at the air foil of approximately 350 hoods at a flow rate sufficient to produce 10 TLV's of carbon monoxide (500 ppm) inside the hoods. The individual per- GAL 000045408 1 Safety Newsletter Research and Development Section Table 1. Air Sampling Data for Operations Conducted in Hoods at ORNL Contaminant Methyl ethyl ketone Toluene Xylene Oioxane Carbon tetrachloride Hydrofluoric acid Trichloro ethylene Tetrach lorethane Tetrabromcethane Methylene iodide Mercury Benzene Average breathing zone concentration (ppm) 17 TLV'(ppm) 200 Range of concen trations (ppm) <1-90 No. of samples 13 Range of hood face Velocities (fpm) 75-170 < 0.1 <0.1 <1 2 100 100 100 10 0-0.1 0-0.1 <0.1 - 1.4 <1 -3 6 3 2 3 70-160 75-140 130 140 <0.3 3 <03-0.5 3 HO <10 <0.5 <0.01 <0.1 <0.005 (mg/m3) 2.6 100 5 1 0.05 10 1 0.2 -13.66 1 1 1 1 1 42 120 140 65 65 110 65-170 a Threshold Limit Values tor Chemical Substances in the Workroom Environment. American Conference of Governmental Industrial Hygienists. Cincinnati (1075). b These operations were not totally confined to the hood. forming the test then went through a series of motions designed to simulate actual work movements. Breathing zone concentrations were determined during these activities. Some 1,000 to 1,200 determinations for carbon mon oxide in the breathing zone were made during this investigation. The highest concentration detected was less than 5 ppm, or less than 10 percent of the TLV. (Note: 5 ppm is considered the lower limit of detection for the MSA carbon monoxide detector tube which was used.) These data confirm that face velocities of 100 fpm protect the employees from exposure to carbon monoxide at the concentra tion tested. Similar studies involving different test contaminants conducted at Oak Ridge also support this conclu sion. Three Observations In the process of generating data from these specific studies, three sig nificant observations were made re garding the practical aspects of hood design. A research laboratory that uses a wide variety of toxic chemicals should design and maintain special use hoods. These hoods would handle materials that either are extremely toxic or have some special property(s) requiring a higher capture velocity and/or a specif ic legal requirement. Such hoods should have a design face velocity of 150 fpm and should not drop below 125 fpm. Insofar as practical, a hazardous duty laboratory hood should be dedi cated to a specific operation using a specific contaminant(s). Examples of such toxic materials are beryllium, certain radionuclides, and chemical carcinogens. These hoods should be specifically labeled, when feasible, for the contaminant they were designed to handle in order to facilitate administra tive control over their use. As OSHA continues to reduce its permissible exposure levels for chemi cal contaminants, the advantages of special use hoods are clear, since such hoods would eliminate the need for installing new ventilation systems for intermittent operations. It should be pointed out that there 3 are certain liabilities associated with face velocities as high as 150 fpm. Probably the major one results from increased air turbulence created inside the hood by high velocities. The use ofa hood air bypass should be discouraged. It has become routine practice to use either a bypass damper or the downward movement of the sash itself to regulate a duct opening for the purpose of allowing room air to bypass the hood face and be exhausted via the hood duct. Although this achieves the seemingly practical goal of a constant air velocity through the hood face, it robs the operator of optimized control, i.e., attaining eiw hanced velocities by merely lowering the sash. The possibility of exaggerated velocities could theoretically be ob tained when the sash is lowered to a minimum opening without a bypass, but in practice this is not a serious problem. The elimination of air by passes permits design of hoods in a manner more consistent with the eco nomics and philosophy of industrial hygiene engineering, i.e., to make every cubic fool of air moved work to reduce exposure. For general laboratory use. a hood with the sash set at the 40 percent open point can still be operated conveniently while improving the shielding and reduc ing the exposure to toxic materials. Hoods should be operated with mini mum sash openings. Despite the general applicability of the 40 percent operating position, it is imperative that the designer know the ultimate use of the hood. There will surely be instances where the normal operating sash opening must be greater than 40 percent. In these cases, the flow should be adjusted to provide 100 fpm at this operating position. The 40 percent figure presented here repre sents a criterion developed from practi cal experience that provides an appro priate design point in the absence of more defmitive data. References Code ofFederal Regulations,Title 29, Section 1910.1003-1016. Committee on Industrial Ventila tion, Industrial Ventilation, A Manual of Recommended Practice, 13th ed., Amer ican Conference of Governmental In dustrial Hygienists, Lansing, Mich. (1974). SAL 000045409 Building Systems Design Reference Section Clearing the Air in Laboratories A discussion of fume hood applications and make-up air systems used to remove contaminants from laboratories and preventing cross-contamination between rooms within a laboratory building. by: JOHN D. CONSTANCE, P.E. Cliffside Park, NJ. SAL 0000454:1.1 An IMPORTANT aspect of environmental engineering is the removal of contaminants from laboratory spaces to prevent cross-contamination between va rious rooms within the labora tory building. This can be most effectively accomplished with ex haust fume hoods and to a lesser degree by exhaust canopies, cor rectly installed at the point of origin or point of emission of the contaminants. Main purpose of this article is to review time hood application and make-up air sys tems. Laboratory work and chemical testing involves procedures that could contaminate the air inside the occupied spaces. The nature of the contaminants varies wide ly: high humidity from steam baths, odors from hydrogen sul phide analyses, corrosion capa bilities of alkalies and acids, solu bility of acetone, explosive prop erties of perchloric acid, health hazards of bacteriological aero sols, and poisonous properties of nickel carbonyl. Ideally, the best procedure is not to emit; but the next best is to remove or exhaust directly and as close to the point of origin for safety of laboratory personnel and possible damage to property. To achieve this end, the most common and accepted method used for containment and re moval of the contaminants is by restricting the contaminant pro ducing procedures to within an enclosure or hood. And simulta neously room air is drawn across the hood face to capture and re 24 move the' contaminants so as to prevent them from escaping into the occupied spaces. In the design of a fume exhaust system utilizing hoods or enclo sures the following factors must be analyzed and evaluated: cap ture velocities (exhaust air rates), fume hood design, fume hood lo cation and arrangement, make up air source, make-up distribu tion, exhaust system, exhaust duct materials, exhaust air treatment, special systems (and considera tions) Capture Velocities The quantities of air required for hood exhaust systems are based on several important fac tors, the most important of which is capture velocities. For most ap plications in the chemical and metallurgical industries, these velocities will range from 50 to 200 fpm. The lower figure is used to control contaminants released at low speed into relatively quiet air (15 to 25 fpm). The higher figure is used to control contami nants released at high rates. Un der certain special conditions face velocities as low as 25 fpm have been used with induction type hoods. The matter of face velocity se lection for laboratory hoods is rather mixed. In the conceptual design of a laboratory facility this consideration is given much ar gument and thought, especially when air conditioning is also be ing included. It is worthwhile to remember at this point that fat ***&*&& ttptmgliil--d--wilm l * to be added to the system capac ity because of make-up air re quirements to "feed" the hoods. At $1,000 per ton of refrigeration the cost of exhausting 1,000 cfm would range from $3,000 to $4,000. This certainly adds to the hood burden and overall costs of capital outlay. SMMaiaivaa state emphat ically that in teaMdOMMMAO dtp** Attempts have been made to relate face velocity to hood ser vice by compromising fume hood usage together with the added re sponsibility of supervision by laboratory personnel to insure that the fume hoods are restricted to the type of contaminants for which face velocities were se lected. To this end, Brief1 offers a method of hood classification as a step toward economy of design and operation. Fume hoods for highly toxic contaminants (threshold limit value of less than 0.1 ppm) are classified as Type "S" hoods and are designed for operating face velocities of 150 to 130 fpm. Type "A" hoods, for moderately toxic contaminants (threshold limit value of less than 100 ppm), can be sized for face velocities of 100 to 80 fpm. Hoods for non-toxic contaminants, Type "B" (thres hold limit above 100 ppm), are sized for a face velocities from 60 to 50 fpm. It should be emphasized that threshold limit values (TLV's) represent airborne concentrations of materials which most workers may be exposed repeatedly dur- BUILDING SYSTEMS DESIGN gineer and others to make any necessary adjustments. This al lows continuous monitoring and control which should predict more exactly the final projected results of tinte required to finish the project, estimated profit, or loss, and similar rtems. An actual project was used in the simulated situation that was tested. This project was set up into its component tasks\utilizing the general model as a panning guide. *T' and "j" nodesNand time durations were assign^ Work assignments were made ii accordance with available man power (in retrospect). In accord ance with this specific arrange ment necessary cost library of task code numbers for personnel, and comparitive data for simula tion purposes were developed and run through computing facilities at the University of Nebraska. Results were then analyzed. The system was shown to be feasible for use in actual design situations with no real penalties in terms of loss of time or money, but rather with an overall gain in the total costs of design. It seemed obvious that all personnel involved would benefit and the profit situation would be improved. Overall Planning Guide This discussion has considered only the HVAC aspects of design./ Extension to all phases of the d sign process is desirable. It should 'not be difficult to set up an over all general model which would program all of the disciplines into units like black boxes' which could then plug into the/total sys tem at various points.yrhis over all planning guide could then be used for all phase/' Civil, me chanical, architectural, and even construction, Aiyowner could ex pect a better picture of his cost in terms of the complete system. All projects in/a particular office could be Arranged in somewhat the same/planning procedure by using thfe black box concept. This should give the management team better information upon which to base long range plan ning, including marketing and accounting, as well as the design divisions. Forecasts of predicted costs, available manpower, etc.. should be possible. It may even be feasible to extend this sort of system planning to selection of mechanical systems in design of the building. The Critical Path Method is no more time-consuming or diffi cult than other management methods. It does, however, clear ly present realistic conditions readily amendable to changes. Factual, intelligent and timely decisions at the onset of the project can be made as well as a critical evaluation in light of. changing conditions. / Summary PHVAC engineering aspects of buildimj design analyzed as a subsystem\of total Project and problems oDplanning, scheduling and time-resobrce/cost control of engineering functions, are sus ceptible to operations research methods. Techniques and\procedures of CPM/PERT (Program Evalu ation Review Techniques^are as applicable to analysis or'engineering design projects as ^tey are to construction and produc tion/projects. Engineering desig) involves definable interrelated'' individual tasks requiring re sources allocation and time as signment in order to develop a workable schedule for execution. The logical plan of action may be shown most effectively and simply by means of a network diagram model. Time-cost limits for compenent tasks may be es tablished through application of CPM/PERT algorithms. The CPM/PERT model of engineering design subsystem provides a dynamic and respon sive management tool from the standpoints of similitude, simula tion, and optimization during planning, scheduling and control phases of project organization and operation. As is the case for most sci entific management systems, CPM/PERT policies and proce dures must be developed within each firm and at all levels of management. Development of a general ized model, similar to the one described, would provide a stan dard yet highly flexible base from which optical models can be derived for ^ach specific ap plication. In fact, the generalized model could be expanded into a system for optimal alternative selection for the best and most economic utilization of available resources at any time referenced to a least time-cost trade-off analysis. Beneficial effects of op timal utilization of personnel up on a company morale cannot be over-emphasized. The CPM/PERT system pre sented must be recognized as nothing more than a manage ment tool. Each individual en gineering firm must adapt tech niques and methods imaginative ly to best fit its structure and operational philosophies. This process requires education and training at all levels of manage ment. Provision must be made to encourage innovation and con tinuing optimization. A policy of total optimization must preclude the usual ever-present problems of sub-optimization of subsystems by departments or individuals. Overall aspects of this kind of systems analysis substantiates the validity of modern trends toward integration of design and construction subsystems into in tegrated "design-and-build" op- rations. in extension of the systems approach which CPM presents seems Vibvious. Investigation of feasible\systems to select the optimum\an become more of a reality because of the availability of more realistic time schedules. When initial estimating is being done to prepare nmes of alterna tive design schemes, various items to be designed are being formulated. It should be possible to include information as to the cost of the various types orequip- ment, since the type of equipment to be designed is known. Up-to- date libraries of costs could be developed. Designers could plug into the system information about alternative designs and output could be comparative costs. This sort of capability would allow' truly optimal design practices. Decisions would then result only after investigating all feasible approaches. AAA I ing a normal work day of 8 hours duration during a working life time. The values should be used with care: they should not be used as sole criteria. Small glove compartments of cabinets are used to confine high ly hazardous radiological and bacteriological procedures. Open hoods should not be used for these purposes. Usually such cab inets have relatively small ex haust requirements due to con finement of the work. Fume hood efficiency depends on the amount of air exhausted. To assure flexibility of operation and maximum safety to labora tory personnel, a fume hood should be designed or selected for exhaust air rates ample for complete removal of all contami nants. This may be a logical step when only one hood or two are involved in a single facility. If a laboratory facility has several hoods, however, generous ex haust through all hoods can im pose a heavy initial and operating cost penalty on the air condition ing and heating system. Where a large facility has only one or two small hoods the room air supply may exceed maximum requirements for hood exhaust air. And no special consideration for satisfactory exhaust hood re quirements is needed. Witwrua* hood* present a veatilatsaMNrt** ofT H cfm pa- sq ft -a fiamMb would be required to maintain uniform conditions. This figure may be used as a design guide. The next factor which must be considered is the total volume of air passing through the fume hood, expressed in cubic feet per minute (cfm). This simple cal culation is made by multiplying the square footage of the fume hood sash opening by the selected face velocity. Allowance must be made for overcoming the effect of static pressure. This is the loss of air or the rate of resistance or the friction incurred by the volume of air as it passes through the hood and connecting ductwork and discharge stack. M--fc hnorii aja ratedLindhir rupeetuBt-Wr intia pressure^ lose* This allowance is also sufficient to compensate for the static resistance of .20 feet .of straight ducting withne 90 el bow. When longer duct lengths or additional elbows are required, the allowance for static pressure must be increased. As we have mentioned pre viously, exhaust requirements are a matter of great concern in tlie design of the overall system. Another way of measuring hood burden on air conditioning load is that it requires more than one third of a ton of cooling capacity per 1,000 cfm of outside air for each degree of wet l?ulb differ ence. In laboratories, make-up air for exhaust accounts for a major portion of the refrigeration cooling load and is also a basic factor in determining the re quired capacity of the air han dling system. It, therefore, be hooves the designer to work out carefully the required exhaust rate. Fume hoods normally add up to be the major source of exhaust, making it desirable from the de sign point of view to standardize on a fume hood that exhausts a minimum quantity of room air in controlling and containing the contaminants. It is safe to say that most chemistry laboratories, no matter what their specific pur pose, contain at least one hood or canopy. From actual experience with laboratory design, it is often difficult to select a one-hood de sign that will satisfy all users and still be adaptable to each need. Fume Hood Designs The function of an exhaust sys tem is to protect the worker from exposure as we have seen. Thus, the heart of the system is the hood or canopy and the design of the system begins with the hood, which is, at best, a compromise between the ideal and the practi cal. Basically, a laboratory hood is a simple box. Fig. 1. Without the necessary indraft shown for the basic ventilated hood. Fig. 2, the material inside the hood can be come airborne and be emitted into the room by any one or combination of the following nor mal laboratory operations: Ther mal action--convection currents within the box. Meehancial agita tion, Aspirating action by cross currents of air outside the box. Material can excape from the basic hood only through the door Sash ___n Fig. 1. Basle laboratory fun* boob design. Fig. 2. Labaraiary hood vftlwet nactuary indraft shewn. SAL 00004 or opening in front. However, in the simple ventilated hood, con taminants are kept inside the hood by the action of the air flowing into the opening. To con tain and keep the material from escaping, sufficient air must be exhausted from the hood to create and maintain an indraft through the face of the hood opening. Exhaust hoods should control contaminated air so that the con taminant does not reach the breathing zone of the laboratory technician in significant quanti ties as pointed out earlier in hood rating. If no emission of contami nant takes place, then the air within the hood should not be contaminated at alL t Hood Location and Arrangement For proper operation of the fume hood, there should be no dis turbance or turbulence at the face of the hood. When planning the laboratory layout it should be done with thought to avoid locating the fume hoods near open windows, open doors, heavy traffic aisles, room supply and air handling units. There may be one exception, however. In the case of Type "S" hoods, thesafety oflaboratory per sonnel may require that the fume hood be near a door for easy and quick exit in case of an accident. It is good practice to locate the ends of the hood at least three feet from any wall of the room, See Fig. 5. This will avoid a one-sided approach of the frontal air stream as it enters the hood. For best op eration room air circulating ve locity should be kept low (25 to 50 fpm) so as to present minimum disturbance to the air pattern as it approaches the face of the hood. Supply air distribution should be carefully designed to this end. E=3 Ceiling Fit. 6. Aspirating typt ceiling diffgam should be used (or supply air distribution. Even at ideal conditions, a per son working at a hood--or just walking by--can cause sufficient dist rrbance in the air flow pattern to spill some hood air from the hood by an aspirating action. For tunately, as the face velocity does go up the chances for spillage go down. From this we can appreciate that the basis for evaluating hood performance is the average face velocity of the air flowing into the hood. Even in a well designed hood the location and setting of the appara tus within the enclosure can make or break the operation. Among the main disrupting factors are the following: Large pieces of equipment lo cated within the hood chamber will interfere with the air flow through the hood. Hot plates, bunsen burners and other concentrated heat loads lo cated at one end of the hood will disturb the air flow pattern. If the fume hood duct stub loca tions are greatly offset to over come building obstacles such as girders, the air flow at the end of the fume hood furthest from the duct stub will be less than at the end where the duct stub is located. (SeeFig. 7) Make-up Air Source Make-up air to balance the air exhausted is the most essential de sign feature of any fume hood ex haust system. When a fume hood is operating poorly closer analysis will most often show inadequate make-up air supply. There is no air for the hood to "breathe" and an improperly sized make-up air supply system will starve the fume hood and restrict its intended op eration. Some designs depend on air drawn from adjoining corridors and office space. Introduction of make-up air by indirect means is a most economical approach. How ever, such a system can lead to bal ancing problems and cross-con tamination between laboratory spaces. Positive introduction of air from corridors and office spaces by use of transfer fans can improve this. Such a positive approach will improve air balance but can still cause cross-contamination. It has been found that the most reliable, flexible, and easily main tained system arrangement is that in which an adequate supply of outside conditioned make-up air is supplied to the laboratory space to balance the air being exhausted through the fume hood. It is good Duct stub ^correct location Fig. 5. Locate ends of hood at least 3 feet from any wall to avoid onesided approach of frontal air stream entering hood. 7Incorrect Incorrect I 1 Ceiling diffusers of the aspirating type should be used for this. See Fig. 6. In other words, laminar flow design is recommended for best results. Air supply terminal units should be of the type de signed for introduction of the sup ply air with minimum turbulence rtf rnnm nir ( ft Fig. 7. ^-Hood BrU- 11 Research and Development Section A. J. Story, Survey ofC-710 Laboralory Fume Hoods (Report No. KYL-714. Union Carbide Corp. Nuclear Division, Paducah Gaseous Diffusion Plant, Paducah, Ky., Prepared for U.S. Atomic Energy Commission under U.S. government contract W-7405 eng. 26 (June 18. 1974). Threshold Limit Values for Chemical Substances in the Workroom nvironment. American Conference of Govern mental Industrial Hygienists, Cincin nati (1975). -- /V. E. Bolton and W. E. Porter Industrial Hygienists Oak Ridge National Laboratory and J. T. Dufour. Industrial Hygienist U.S. Ene'gy R&D Administration Oak Ridge Operations Oak Ridge, Tenn. Newsletter R&D Executive Committee Meets John N. Romine, safety director. Research and Development, Phillips Petroleum Co., addresses the Research and De velopment Executive Committee. Romine, who is first general chairman of the section, retires from Phillips Petroleum Co. in September. Dr. Robert Chanaud, president. Engineering Dynamics, Denver, Colo., addresses the Research and Development Executive Committee during e luncheon meeting. * Fire Prevention Program Aids Off-the-Job Safety THE NEEDLESS loss of life due to home fires can be substantially reduced by means of appropriate training in fire prevention and by the installation of early warning smoke detectors. Fire prevention week (October 9-15, 1977) is a good time to initiate a program which emphasizes home fire safety to employees at work and to neighbors in the community. Two methods of presenting a pro gram have been used successfully at an R&D facility. Both provide employees the opportunity to evaluate their fire safety needs and to order ionizationtype smoke detectors, window escape ladders, and fire extinguishers. Local fire equipment suppliers and distribu tors were most cooperative in provid ing information, demonstration dis plays, films, and general assistance. Appropriate films are also available from the National Fire Protection As sociation (NFPA). Announcements and posters were distributed a week before the presentations. The first method used space in the employee cafeteria where interested employees could stop, look, listen, and ask questions during their lunch hours. A 15-minute fire prevention film was shown every half hour. A 15-minute demonstration on available fire equip ment and smoke detector performance filled the remaining time. Fire escape plans and drills were emphasized, since smoke causes panic and disorientation. A fire equipment display was also set up near the main entrance, with order forms available and where purchases could be made with the supplier or distributor. At least two days are recom mended for this method, since many people like to discuss their needs at home before deciding to purchase such equipment. Literature and order forms were also made available for several months following the program to per mit additional purchases at the special fates. The second method used was di rected towards smaller offices where personnel could meet without upset ting work schedules. This program took up to one hour, since the ABC-TV documentary film on the U.S. fire problem was shown as an introduction. Following are some key points of the presentation, with some of the applica ble statistics that emphasize the home fire problem: Review the U.S. residentialfire experi ence. About 60 percent of the 12,000 annual fire deaths and 40 percent of the three billion dollars in property loss occurs in residences. Another 300,000 people are injured annually in fires. About 18 percent of the fires are 4 electrical, 17 percent involve heating and cooking equipment, and 12 percent are due to smoking. Fire deaths in the home generally are due to carbon monoxide and other toxic gases produced from burning furnishings when the family is asleep. Children are the victims of clothing fires too often, because of playing with matches or near open fires while wear ing flammable clothing. The latest revision to the Federal Fabrics Act requires children's sleepwear (sizes 7-14 X) to be fire retardant. Review the relationship offire preven tion, protection, and planting. Use a visual aid or demonstration (see sample diagram). Discuss evacuation planning and periodic drills. (NFPA visual aids and leaflets may be passed out) Discuss reasonable fire safety precau tions to follow. Minimize fire hazards. Use non combustible furnishings and construc tion materials as much as possible. Inspect for worn electrical cords, flam mable liquids, trash, combustibles near light bulbs, on a frequent basis. Clean and maintain heating appliances in good operating condition periodically. Sleep with bedroom doors closed. Provide a fire warning system. Continue* on page 6 SAL 000045410 Nearly all hood designs pres ently in use attempt to provide protection in three ways: a me chanical shield, direction of air movement, dilution of contami nant by mixing with large vol ume of air inside the hood. The mechanical shield com prises the hood sash. When an ex periment is being set up the sash or door is in the raised position. In many experiments, the sash is lowered two-thirds the way down or even closed off entirely while an unattended experiment is being carried out. Only the oc casional visit by the laboratory worker is needed. Care must be exericised not to lower the sash too much without causing too high an indraft velocity with at tendant overcooling or the snuff ing out of the Bunsen burner flame. Protection is provided by the direction of air movement across the back of the worker (see Fig. 3) and into the hood proper, past t SAL 0000-454 1 4 Fig. 3. Air moving icrats back of worker and into hood preptr, provides protection, diluting contaminated air. the equipment within the hood and thence into the exhaust system. Lastly, because large amounts of air are being moved through th hood, dilution of the contaminated air takes place readily and further reducing the hazard of breathing the air. hacd-.daMffH have been used: 1. GSMiMatipnal-fcood--all ex hausted air taken from the room. This is the simplest, low in initial cost, and effective. However, the amounts of exhaust air place a heavy burden on air conditioning costs and operation. (Fig. 4a) anna*--wwijmwtunFaiiw iu duced face velocity in an attempt to compromise hood effectiveness to reduce the air conditioning load chargeable to the hoods. Al though low in relative cost, it does reduce air conditioning load but its effectiveness in removing fumes generated within the hood is weakened.(Fig. 4b) Exhaust hoods may be needed at random intervals, and it is not likely that they would be simultaneously. As with other types of air conditioning loads. there is a usage or diversity fac tor that is apparent, yet difficult to define. This factor depends up on judgment, experience, and logic. For example, a large num ber of hoods in a laboratory room does not necessarily mean all hoods will be operating at one time since the number of labora tory personnel would be limited and thus reflect on the number of hoods in operation. On the other hand, it is the policy of some lab oratories to keep all hoods in op eration 24 hours a day even though they will be used inter mittently. So much depends on the management of the labora tory facility and it behooves the design engineer to explore the total operation with the ultimate user. (Fig. 4c) L 4f 4h 1 Fig. 43. Convention*! hood. All air taken from room. Fig. 4b. Conventional hoed with reduced face velocity. Fig. 4c. Coirvantienal hoed with use or diversity lactor. Fig. 4d. Internally toppliod flood. Fig. 4e. Externally supplied hood. Fig. 4f. Perforated ceiling supply hoed. Fig. 4g. Horizontal sliding sash door hood. (AU roam air make-up) Fig. 4h. Canopy hood. $ fa 0 00 0 4b 41 -i 4. this type of hood required make up air is fed directly inside the hood without affecting the overall room air conditioning. This air need not be cooled in summer and merely tempered in winter. Al though an additional air handling system is required, the saving on the air conditioning load can off set this. Cost of this hood runs medium to high butmbmaB* fully deai|pie<^and*1IWiitwetffuTitte removal offictivanNl^USTie poor. (Fig. 4d) 5. Externally supplied--hood. Because of the additional duct sys tem required such a system is rela tively more expensive, relatively low cost effect on air conditioning, and because air is being ex hausted across tne face of the hood, AUGUST 1971 fume removal effectiveness is good. (Fig. 4e) 6. Perforated ceiling supply hood. This system allows ample opportunity for the unconditioned air to mix with room air and it be comes often necessary to sensibly cool but not dehumidify this auxil iary supply. Because, again, air is exhausted across the face of the hood into the hood, fume removal effectiveness is good. (Fig. 4f) 7. HmriwMUl niidfffynMiiWin* i "* Compared to the conven tional hood with its vertical slid ing door upward, the horizontal sliding door unit presents much less area to be exhausted and total exhaust is reduced, relative cost of hood is low and since less air is exhausted from the room, air con ditioning costs are low. Air con ditioning effectiveness and fume removal effectiveness are good. (Fig. 4g) 8. ^Rmp94*MdThis is local ex haust ventilation in lieu of an en closed hood. Low are relative cost, air conditioning cost, air condi tioning effectiveness, fume remov al effectiveness (Fig. 4h) In every case where air is drawn across the hood face (a, b, c, e, f, g. and h), it becomes apparent that supply must be from outside the hood to maintain the proper con trol velocity across the face of the hood. It has also been found that high velocity air streams must not be permitted to impinge on the face of the hood. In all of the cases listed, convenience to laboratory personnel was good to fair. 57 i Fig. 8. An adequate supply f outsida condltionad maka-op air it supplied ta lab space to balance air exhausted through fume hood. practice to supply a little less make-up air in this maimer than that being exhausted by the hoods. See Fig. 8. A slight negative pres sure will be maintained, drawing air through door louvers from cor ridors or adjacent offices. Air* nhwtoA.fypBL, *(ftood,j*, never, recirculated and as we know, the continuous loss of con ditioned air increases initial and operating costs of heating and air conditioning systems. Operating costs can be reduced by supplying make-up air from an auxiliary system instead from the air conditioning system. The air so handled is filtered and tempered in winter only. Of the eight basic hood designs previously discussed, numbers d, e, and f, make use of the auxiliary system. The auxiliary air supply is introduced as we have seen and can be either of the cen tral type or unitary type with an outside air inlet for each labora tory. In one air conditioned labora tory it was observed that a unitary system supplied outside air for make-up through the roof of the laboratory room using the face and by-pass principle. Correct selec tion of the type of make-up air system can be made only by an engineering analysis and flow sheet of the fume hood exhaust system. or not, is that at some point the system must end and discharge to atmosphere. Unfortunately, while the exhaust system has ended at this point, the problems associated with that exhaust system may have just begun. If too much air discharged from an exhaust system is recirculated through the supply air distribution not much good has been accomplished. If by poor de sign the exhaust air is- not prop erly located with respect to the in takes of other supply systems po tentially disastrous results can be attained. Such poor designs are often commonplace at many fa cilities. The real cure for this type of problem is not higher exhaust velocities, higher stacks, better weather caps, better separation of discharge and intake openings, or other, although one or more of these can contribute to the cure. The real remedy must start back at the source of the contaminant itself. Because the pattern of natural air flow around buildings is not predictable, contamination con trol by the location of vent efflu ents and air intakes is difficult to put to practice. Halitsky* and Clarke3 have advanced theoretical knowledge and rules of thumb that aid greatly in the solution of such problems. One of the most important char acteristics of an exhaust system, whether it be for laboratory hoods Air Distribution As we have seen, air movement AUftnsT 1971 within each room of a laboratory complex must be such that a defi nite flow pattern will be main tained down through the building along with flow from non-contaminated to potentially contaminated areas. To bring about this differen tial flow pattern, the natural bar riers between the various classes of rooms will assist. The flow pat tern will be assisted by supplying outside clean air to the non-contaminated and semi-contaminated areas and by exhausting air only from the moderately and extreme ly contaminated areas. In general, supply fans should take suction from the upper por tions of the building. In addition, the exhaust fans should discharge to the outdoors through stacks of varying heights depending on ad jacent structures. In order to as sist in the general housekeeping within the area, the building should be maintained at a slight positive pressure with respect to the outdoors. Laboratory rooms should be maintained at a nega tive pressure with respect to the surrounding rooms. In order to have the supply and exhaust in any building in balance, there must be an adequate air sup ply for all exhaust needs. This cer tainly implies that there must be available an excess of air supply over that required for normal ex haust needs. In actual laboratory installa tions, experience has shown that SAL 0000454.17 i when two fans are exhausting from the same space with no provision for make-up air, the stronger fan will take command and air will en ter the room through the weaker fun system. When there are multi ple exhaust hoods and no make-up air, when one hood is turned off, outside air can immediately downdraft through the idle fan. In situa tions when a fan must exhaust out out of a room without an air sup ply, the capacity of the fan will be reduced from the original design flow and will result in less con trol at the hood. See Fig. 9. The exhaust system being under negative pressure, keepingleakage due to poor duct construction or deterioration, will naturally flow into the duct system and contami nation will be confined thereto. This is a relatively simple princi ple, but, nevertheless, many poor installations are found. The best place for an exhaust fan servicing hood is roof mounted. Then all ex haust ductwork on the suction side of the fan will be inside the build ing. If a hood exhaust fan is re quired to be located just above the hood and inside the building, then great attention must be paid as to the tightness of the discharge ductwork and stack. The fan on the roof is still the preferred loca tion. When flammable material is handled, mounting the fan on the roof is a distinct advantage be cause explosion-proof construc tion may not then be required of the motor. This represents a dis tinct savings. In multiple hood exhaust instal lations, duct connections should be streamlined. Branch ducts should and concentration ofcontaminants or chemical reagents, space condi tions, cost, accessibility. Whatever be provided with angular boots material Is selected, the duct joints and not enter the main header at must be leak-tight and the duct an angle of 90. work should have ample supports. Ductwork transporting fumes All longitudinal seams should be may be axed econamicaity'for-- run along the top side of the duct. (Fig. 10) And extensive duct sys tem should have inspection and cleaning facilities. Systems that could develop heavy condensation loading should pitch toward a pocket in the bottom of the duct and provided with a trapped drain. Usually Type 316 passive stain less steel is used for bacteriological, radiological, perchloric acid and other general chemical purposes. 316 stainless is easy to work but is not suitable for chemical hoods handling concentrated hydrochlo ric and sulphuric acids. vaiOTft9^f*9M9fgnni'because ^ 0f_ fers a good compromise between horsepower savings and duct space requirements. Exhaust Duct Materials In the usual newer buildings, ductwork is often concealed in ceilings or inside walls, making duct inspection and replacement a major problem. Where such in accessibility exists it is reasonable to use ductwork with a long life expectancy. For the types ofchem icals used in laboratory work, gal vanized iron and black iron duct work are highly susceptible to cor rosion. Stainless steel, transite, polyvinyl chloride-coated steel or fiberglass-reinforced plastic duct work will not require early re placement for such corrosive ser vice but are costly. Actually, se lection of materials for an exhaust system will depend on the nature Although transite is almost a universal material for its chemical inert properties, it is not commonly used because of its high installed cost and other working disadvan tages. Main disadvantages are the fabrication and the joints. How ever, the joints of round sections are relatively simple to fabricate. Asbestos cement ductwork, be cause of its heavy weight, presents support and space condition prob lems of installation. The material absorbs moisture and could drip when saturated. It also has to be handled carefully during installa tion and maintenance because of its fragile nature. PVC (polyvinyl chloride) and FRP (fiberglass reinforced poly ester resins) areeasytohandleand and fabricate. These materials can be used for general chemical duty except where a highly solvent va por such as acetone is handled. Protected metal offered by a number of suppliers has found some use in this application. Black or galvanized steel coated with acid-resisting materials or coat ings is used for light duty. s- Longitudinal seam ^ Top of duct Fif. 10. All longitudinal seams should run along top side of tho duct. Fans and other moving parts of the exhaust system arc also pro tected by special coatings. There are a number of protective coat ings available but none has uni versal application. Final selection of coatings will depend on the re agents handled in the exhaust sys tem. The fan. bearings should be outside the gas stream separated RimniNA cvcrcikic nrcifiN by the shaft seals. It is good prac tice to provide explosion-proof fan motors if fan is installed below the roof. Exhaust Air Treatment Gases that chemists bubble through reaction mixtures and then discharge to the hood are gen erally, by their nature, reactive enough to be completely elimi nated by a scrubber of some sort. For materials of an acidic nature, a simple caustic scrubber is all that is necessary to assure essentially complete control. Similarly, for materials ofa basic nature, a scrub ber containing sulfuric or another acid can be used to advantage. For those materials that do not react rapidly with either basic or acidic solutions, a column filled with ac tivated charcoal will almost al ways provide the desired control. Some materials such as perchlo ric acid are highly soluble in water and hoods have been developed with a packed section built into the hood superstructure and provided with a water wash header. More on this will be covered under spe cial systems. A fume hood handling radioac tive material should have a high efficiency filter. Exhaust systems for highly hazardous bacteriologi cal experiments can be made safe only by incineration ofthe exhaust air stream, which is heated to about 650 F. to destroy the bac teria. Special Systems Lowered Sash Operation. Ahood exhaust fan maintains proper cap ture velocity at the face of a lab oratory hood when the sash is wide open, but the exhaust hoods* ver tically sliding sashes aresometimes lowered to within a few inches of the work surface when the hood is in operation. It should be recognized that ex haust hoods that would seldom be used with the sash wide open could unnecessarily impose as great a hood burden as those that would often be operated with the sash wide open. It has been found that partially closing a hood sash does very little to reduce the volume of air exhausted by the hood fan in typical installations. A method or system in use to re duce unnecessary wasting of con AUGUST 1971 ditioned air and also to achieve a more constant face velocity over the range of sash positions is the application of a 2-speed fan for each hood. When the sash is pushed up the fan runs at high speed. A micro-switch mounted in the hood is tripped by the sash when it is lowered below a certain predetermined position. The vol ume of air the fan will pull on low speed is adequate to maintain de sired face velocity for the smaller cross-sectional area. The proper placement of the micro-switch can ber 50 to 60 per cent of the vertical face opening, i.e., the fan would go on low speed when the sash is lowered to 50 to 60 per cent of full open ing. This holds for all exhaust hoods despite differences in hood dimensions and other variations in exhaust system configurations. It has been found to apply equal ly to hoods with minimum face velocities of 80, 100, and 125 fpm. The volume of conditioned air that is normally lost is reduced by about one-third in most in stances when, the sash is below the switch position. In a conventional hood with a single speed fan, the excessively high face velocities usually ex perienced at low sash settings and on persons using the hoods. Ex amples of research equipment and materials disturbed by high drafts include, among others, bunsen burners, fine powders, tissue slices, and thin paper. It should be obvious that some ex perimental work could be ad versely affected by the unneces sarily high flow rate of cool room air over the work. High drafts are uncomfortable. Further, still, when the labora tory technician stands in front of an operating exhaust hood his body presents an obstruction to the flow of air into the hood. Thus, a low pressure area de velops in the space between the person and the hood. In extreme cases, this low pressure area can cause fumes to flow from the hood back into the room. The re duction in face velocity using the 2-speed fan reduces the probabil ity of this potential hazard. Type "S" hoods should be pro vided with exhaust fan speeds so that at no point across the hood face should a velocity greater than 250 fpm exists. Another way to control this velocity is to pro vide by-pass dampers in the ex haust ductwork just downstream of the hood itself. By-pass hoods are made to accomplish this ef fect by providing this feature in the hood itself. Bypass Hoods. (Fig. 11) These hoods provide for a constant rate of room exhaust and uniform hood face velocities for any door position. They stabilize the room exhaust and consequently the room supply. The by-pass may be an integral part of the hood itself. As the hood doors begin to close, the damper starts to open. Thus, in any hood door position, the ex haust air quantity and hood face velocity remain constant. The re sulting constant face velocity prevents the disturbance of flames and easily disturbed ob jects when the hood door ap proaches the closed position. One other important aspect and In some installations the by pass effect is provided by the in sertion of a sensitive barometric si damper properly sized to remain closed when a conventional hood is operating with doors open. As the doors are lowered, the change in static pressure inside the duct will open the damper to compen sate for the imbalance and will keep the face opening velocity constant. When the hood door \s completely closed room exhaust will remain fixed. However, the advantage of continuous purging of the hood interior by the integral built-in bypass does not exist. Supply Air Hoods. There are two types of such hoods commer cially available. The first type has auxiliary asc^introduced outsidfe and in front of the sash, normal ly from the overhead position. Fig. 13. Supply air head where auxiliary is fad directly Insjde head. -.... 1_________________ Fig, 12. So#ply sir heeds. TUs type has auxiliary sir totrstert snttMe art ( freet f sash, nemsaHy fraai waifri pasttlaa. (Fig. 12) In this design the auxil iary air supply is drawn into the sash opening as a part of the room air. Balitiim in--t of such hoods compared to the conventional typevisftfaigh. Since the amount of conditioned room air is reduced the relative cost of air condition ing is low. Air conditioning effec tiveness, fume removal effective ness, and convenience to labora tory personnel are good. In the MMBHfcJvpe of supplied air hood the apaaAMPM^a^dSv rectly inside tttfr IMfcff (Fig. 13). Relative cost is rather high, rela tive cost of air conditioning is low for obvious reasons, air condi tioning effectiveness is good, but 32 fuovr ^r-->h'. efi--gfiuim iVi~Kiie ppor. Convenience to laboratory personnel is good. Because ef fective face velocities can drop below the safe value needed to prevent leakage of dangerous fumes, the use of this type is fre quently discouraged by local gov ernment health officials. Induction Venturi. For many fume exhaust applications such as those involving hazardous fumes or vapors, the conventional ex haust method of passing the gases through the fan case could be po tentially dangerous. With exhaust from perchloric acid fume hoods in particular, a build-up of crys tals can occur on duct walls and fans. This crystalline growth is explosive under normal condi tions and special treatment of such a system is mandatory. To overcome this problem, there are commercially available induction venturi systems with water wash facilities. Since per chloric acid crystals are highly soluble such systems are provided with spray rings or nozzles and are washed down internally at regular intervals. Drainage is provided to a trough attached to the back of the hood table. (See Fig. 14) Operation of the system is ac complished by supplying a high velocity air stream jet inside a specially designed venturi. This in turn induces a flow of gas at the inlet to the venturi. This in duced flow can then be used to exhaust hazardous gases or fumes without any of the gas having to pass through the fan. Venturi is of stainless steel (316L). Blower is of mild steel. Such a system used to exhaust 1,200 cfm from a hood against Vfe" w.g. resistance requires a primary flow of clean air of 500 cfm and a % hp fan motor. There are other perchloric acid fume exhaust systems using spe cial centrifugal exhaust fans of PVC construction. Such systems include a fume scrubber just downstream of the hodd and up stream of the fan. The fume scrubber is also of PVC construc tion with access and inspection doors. Other parts of the scrub ber include: raschig rings, water spray header and nozzles, elimi nator blades, and baffles. All Fig. 14. induction venturi, where drainage Is provided to trough attached to back of hood table. I * Main exhaust fan , \ \ Outside makeup All ductwork under negative pressure // \\ From individual hood exhaust fans Outside makeup \ \1 ' Fif. 15. Sinfte exhaust fan system (desiined by autfcar) imwlvas a multi-flood facility- parts coming in contact with the air stream or the wash water must be of Type I PVC. The su perficial velocity of air flow is 400 fpm on the design basis. The centrifugal fan is of nonoverloading characteristics and all PVC construction. Water spray headers run longitudinally inside the ductwork and are also PVC Schedule 80, % in. diameter size. All duct joints and piping are field cemented. Each perchloric acid hood should be provided with its own exhaust fan and duct system. Systems should not be mani folded and exhausted by one fan. Identify hood with warning sign. The washdown system should be turned on daily for at least 20 minutes. All ducts must be pitched back toward the scrub ber. Non-porous materials should be used so that perchloric acid will not penetrate its pores. Inori' ganic flexible duct connectors should be used to isloate different parts of the system. Organic compounds must be avoided in the construction of the system as well as the use of acids and contaminates other than those used normally for testing inside the hood. Each system should be kept under 2,000 cfm and a hood face velocity of 150 ffc>m on the con ventional basis. Multi-hood Single Fan System. In every multiple-hood installa tion the following question fre quently arises: Should each fume hood be provided with its own exhaust fan or should several fume hoods be serviced by a com mon exhaust fan? A common ex haust duct and fan system may be AUGUST 1071 used if the facility handles similar and compatible chemical re agents. Safest practice for per chloric acid hoods is a separate exhaust fan and ductwork for each hood. In the consideration of exhaust systems for a chemical research facility, where the chemical na ture of the reagents to be used cannot be predicted in advance, or cannot be controlled, safest procedure is to use separate and individual exhaust fans and duct work for each hood. In all cases, the ductwork between the hood connection and fan suction would be under negative pressure. For safety's sake the fan should be lo cated on the roof of the complex with fan discharge vertically up. In some cases it is not possible to place the fan on the roof, but must be located high against the under side of the roof construction. Then the fan discharge, ductwork would be under a positive pressure with the possibility of fume leakage into the space. In this instance, ductwork must be carefully fab ricated so as to avoid leakage and the longitudinal duct seam should be run on top of the duct! One such single exhaust fan system designed by the author involved a multi-hood facility. Figure IS depicts the installation. Each hood exhaust fan discharges into the main header which is open-ended to the atmosphere and provided with opposed-blade dampers motor operated. A cen trally operated main exhaust fan located on the roof runs continu ously, taking suction on the head er. With all hood exhaust fans off the dampers are in the preset open position and most of the air fed to the roof fan is outdoor air. As each hood exhaust fan is cut in, the static pressure within the header will build up. This build up is felt by a static pressure sensing head and the dampers will begin to close. With all hood exhaust fans running the outside dampers will be in the closed position as the roof exhauster now takes suction from all hood fans. An alarm and light in the power circuit of the roof exhaust er provides the necessary safety feature; on fan failure light will fail and an alarm will sound. Pejorated Ceiling Supply Hood. Another installation de signed by the author consisted of an auxiliary outside uncondi tioned supply system with tem pered air in winter. Discharge into the laboratory took place through a perforated ceiling at a maximum rate of 3 cfm per sq ft ceiling area. The auxiliary supply duct was provided with opposedblade motorized dampers, one damper interlocked with an ex haust fan, one to an exhaust hood. The main auxiliary supply fan was provided with inlet vanes operable through a static pres sure sensing head in the main dis charge duct. With all hood ex haust fans off, main supply fan is off. As each hood exhauster goes on supply fan operates at con stant speed but static in main supply duct operates to feed hood requirement. As each hood fan goes on, damper motor is actuated to the open position (two posi tion) to permit more auxiliary to enter the perforated ceiling space. With all hood fans on, all inter locked dampers are open wide and supply fan is feeding the total 33 hood exhaust. As the demand for hood operation drops off, duct dampers will shut, static pressure in main header will build up to squeeze down on the inlet vanes. To provide some degree of cool ing (not conditioning) and air conditioning supply is provided and is ducted to individual ceil ing diffusers evenly distributed in the perforated hung ceiling. This provides spot cooling during time hood exhaust fans are in op eration, and full conditioning when all hoods are off. Hoods are conventional with a face velocity of 100 fpm. Expansion of Existing Hood Exhaust System. As we have seen, make-up air must not only be introduced into a building but also must be properly distributed. This problem of distribution was exemplified in an expansion problem for an existing multi hood laboratory facility. The complex was originally provided with 100 per cent outside air sup plied directly to offices and lab oratories for cooling purposes. All the air was exhausted via hood exhaust systems. All lab rooms were kept under negative pres sure. Over a period of time, the addi tion of hoods and the need to up date hood face velocities in creased overall building exhaust rates. In this installation, thus, there built up a need to supply makeup air to two floors and 150 hoods. But there was a problem-- limited space for equipment on the floor and no duct space over head. How could the problem be approached? Here were the steps taken: 1. Investigate existing equip ment for excess capacity. Check coil face velocities, control valve capacity, and duct velocities to as sure them to be within limits of good design practice. 2. Do not arbitrarily increase speed and capacity of a unit based on fan alone to obtain needed makeup air since duct noise, mois ture carryover, poor system tem perature response, and undesir able grille discharge velocities can occur and create an undesirable system. 3. Study use of progressive ventilation technique: reuse air presently being exhausted to sup ply other areas. This approach may be limited by local codes. Supply must be in the direction of increased air contamination. 4. Solution was to install makeup air units outdoors at grade and distribute outside air into spaces above a new perfo rated hung ceiling on each floor. General. Exhaust stacks should be straight and discharge up ward; no weather protection should be used. Fig. 16. Brief suggested that when open-face velocities exceed 125 fpm, the in stallation of an atmospheric damper should be included to prevent excessive indraft veloci ties when the open-face area is reduced upon sash closure. At relative high face velocities, laboratory equipment placed within the hood should be so placed that the points of release of contaminant are at least 6 inches back from the hood face. This matter of technique is of paramount importance. By plac ing a Va inch thick edging 6 inches wide on the bench top near the hood entrance face is Fig. 16. Exhaust stacks should be straight and discharge upward with no weather protection. suggested. Brief found that con centrated heat loads within the hood proper exceeding 1,000 watts per foot of hood width across created thermal vectors that require higher face veloci ties for control of vapor spill into the room. Obstruction of lower hood baf fle openings by large objects is discouraged. Blockage in the path of purging flow causes problems of control. Design Procedure For a most economical design and the use of the various criteria outlined herein the following procedure is suggested: 1. Set inside conditions of dry bulb temperature and relative humidity in the upper range of the comfort zone. Since relative humidity is critical to operating costs, place greater emphasis on this aspect. 2. Select a hood face velocity sufficiently high to control the type hazard, using the recom mendations outlined in reference 1. Review hood operation care fully since not all hoods require same face velocities. 3. In cooperation with labora tory management determine the minimum number of hoods re quiring continuous operation. De termine if a hood or hoods can operate intermittently or a mini mum and estimate if its exhaust flow can be eliminated in so far as its effect on air conditioning load is concerned. 4. Avoid the use of hoods to store material and merely pro vide local exhaust. 5. Determine the acceptability of face screens or shields or hor izontal sliding panels. 6. Locate hoods so that they are set clear of doorways and fre quently traveled aisles. 7. Determine if laboratory management is willing to take a "slip" in room conditions when more air is exhausted than is originally planned. 8. Consider use of perforated ceiling supply hood arrangement with conditioned air supply through ceiling diffusers for spot cooling effect. AAA nimniNc: cystfmc necii'jN FINAL DRAFT Industrial Hygiene Practices Guide LABORATORY HOOD VENTILATION This Guide was prepared by the Committee on Industrial Hygiene Practices from material contributed by the following authors: H. J. Ettinger, M. W. First, and R. N. Mitchell. Supplementary information and comments were obtained from: R. S. Brief, J. H. Clarke, J. E. Peterson, and J. H. Rook. I. INTRODUCTION Dusts, fumes, vapors, and gases are often released in the course of laboratory work. If the amount of exposure is minimal, general ventila tion may be used to dilute the contaminant to a satisfactory level. However, local exhaust ventilation is a more desirable technique when the contamina tion originates in specific areas or when extreme toxicity or other factors make contaminant control by simple dilution impractical. Local exhaust is also recommended for controlling dusts and other particulate contaminants. The laboratory hood is a specialized type of local exhaust hood and is one of the most important of all laboratory safety devices. It requires thoughtful design, careful maintenance, and proper use if it is to be effective in preventing unsatisfactory exposures to laboratory workers. II. LABORATORY HOOD DESIGN A. Face Velocity Under ideal conditions, air flow velocities of 20 fpm may be adequate to control the escape of particles of respirable size or the molecular diffusion of gases and vapors. In actual practice, however, rates of 50-150 fpm are used in laboratory fume hoods for two reasons. 2- - First, room air currents often exceed 50 fpm and these can force or induce the flow of contaminated air out of the hood and into the laboratory. Second, it is good engineering practice when working with hazardous materials to provide a margin of safety to insure that control will always be adequate. Factors which must be considered when selecting the required degree of ventilation control are shown in Table 1. While this listing does not solve the face velocity selection problem, it does delineate the factors of hood use that are important. Procedures are available (Peterson 1959, 1963) for estimating hood face velocity requirements and these are useful for cases where a laboratory hood is to be used with specific chemicals and where the lowest possible safe face velocity must be used. However, when a given hood may be used for a wide variety of materials and operations, an exact estimation of face velocity for type of use becomes impractical. In these instances, it is more realistic just to consider several broad categories of hood use. The type of classi fication shown in Table 2 has been found to be workable and the velocities listed will provide adequate control under most conditions. In those cases where disturbing air currents at the hood face average more than 50 fpm or when the amounts of materials are unusually large, an upward revision of these values may be desirable. Other steps that should also be considered in critical cases include hood relocation, modification of the air supply system, or the use of a glove box. SAL 000045424 -3- B Methods of Obtaining Uniform Face Velocities An arrangement that helps greatly in obtaining uniform air flow both at the face and especially inside the hood consists of the multiple lateral slots shown in Figure 1. Although hoods with only top and bottom slots can provide a uniform velocity at the hood face, there is necessarily a dead spot at the center of the rear baffle. Compared with dual slots, multiple slot designs minimize chances for contaminants to reach the hood face. First, the contaminants are removed closer to their source. Second, air turbulence inside the hood space is considerably reduced. The hood plenum and slots are proportioned so that the major portion of the resistance to air flow occurs in the slot openings and a uniform negative pressure is thus created in the plenum chamber behind the working zone of the laboratory hood. This may be accomplished by main* taining the total slot area at less than two-thirds cross sectional area of the plenum behind the rear baffle. For these conditions, the air velocity through each slot opening will be equal and the amount of air exhausted by each slot will be proportional to its area. Make all slots the same width but construct so that the top and bottom slots can be set to obtain uniform air flow at the time of installation but not so as to be easily adjustable by persons using the hood. Perforated back panels can also be used to provide uniform face velocity, but the holes in the baffle should be at least one-half inch in diameter to minimize their becoming clogged with dirt. Here again, the total hole area should not exceed two-thirds of the plenum cross sectional area. -yiT- 1 r-----w naff- !> -4- Another important design feature of a laboratory fume hood is the bottom airfoil also shown in Figure 1. This arrangement eliminates an eddy zone that otherwise forms just above the hood floor. Thus, vapors evolved from spilled liquids are swept directly to the exhaust system and removed without being circulated to the hood face. Similar airfoils or 45 "picture frames" are often placed at the sides and the tops of the hood openings. While these features do reduce the amount of turbulence around the hood edges, the improvement in hood performance is not sufficiently great to justify lowering the hood face velocity recom mendations of Table 2. The high velocity air stream created when the hood sash is lowered may cause some problems. Above about 150 fpm, for example, bunsen burner flames and fine powders are disturbed and in no case should velocities above 300 fpm be permitted. By-pass arrangements can be used to prevent this by limiting air flow through hood face opening. Examples include: GRAVITY TYH5: A counterbalanced swinging damper that opens progressively as the exhaust system static pressure becomes more negative when the hood sash is lowered. Easy to install but does not give accurate regulation. DIRECT TYPE: As the hood sash is lowered, it uncovers an opening to the exhaust system which allows some of the air to by-pass the hood face. Simple and effective- SAL 0000454 -5- INDIEECT TYPE: As the hood sash is lowered, it actuates the damper through a mechanical linkage. C. Air Conservation Features A positive air supply is needed in each laboratory area to provide general ventilation for the,room and makeup air for that ex hausted through the laboratory hood. If makeup air is not supplied, then the laboratory hood performance will suffer. The cost of heating and conditioning the makeup air is appreciable and various methods of reducing the hood air flow requirements are often considered. Economy in the utilization of conditioned air for laboratory hoods can be achieved most satisfactorily by maintaining the required face velocity but restricting the open area of the hood face. A transparent hori zontal sliding sash arrangement can cut the overall air requirements by 50% if two half-width panels are used. Similarly, if three panels are used, the open area reduction is 33% for a three track set-up and 67%> for a two-track arrangement. This design also has an advantage over the conventional vertical sash because the full height of the hood opening is available. If panels 14-16 inches wide are used, they can also be used as safety shields. are sometimes suggested as a means of conserving conditioned air. Generally, eapBWBMBafcM* for two reasons. First, control of contaminants released inside the hood depends on the velocity of air through the hood opening. For this reason, designs that have un conditioned air delivered inside the hood do not save any room air SAL 0000454' -6 because the auxiliary air merely increases the volume of air exhausted and has no effect on the air velocity through the hood face and thus no effect on control. In some auxiliary air hoods, the-'rmcBBds.tioMd sir enters along the top, sides and bottom of the opening but in front of the sash. This design re^lres ve'ryr'clrerf`uT'halafTdTn^ and extremelycloaa control of supply and exhaust f lox Tfltefy vtrfttr Is"diff Auultu-ec achieve in systems having nuitiple'hQo^r. As might be expected, this arrangement is also quite susceptible to contaminant escape caused by airflow disturbances from pedestrian traffic and cross-drafts. The most satisfactory and safest type of auxiliary air supply is that using a large perforated panel to introduce the auxiliary air near the hodd at about the same velocity as the hood face velocity. This arrangement unfortunately requires the operator to work in a stream of unconditioned air. The second basic reason for the limited application of auxiliary air systems is that savings in refrigeration are over balanced by the cost of a second supply system and by^HWH^ D. Mechanical Design Features 1. Materials of Construction Laboratory hoods are constructed of a wide variety of materials depending upon their anticipated service and allowable cost. Generally, asbestos-cement board has been found to be quite satisfactory for most services. Hoods intended for radioactive work are usually constructed of seamless stainless steel (joints welded, ground, and polished) because of the relative ease of decontaminating this material. Relatively inert SAL 0000454 -/- soapstone such as "Alberene Stone" is recommended for perchloric acid digestion service because of the potential explosive hazard of perchloric acid in contact with any oxidizable material. Here, some additional features are also desirable. The exhaust duct should be sloped so that it drains into a sump at the rear of the hood. Spray nozzles and drains should be installed in the exhaust system to wash the duct walls con tinuously while the hood is in use. This prevents the accumulation of condensed perchloric acid. Also provide sprays and a drain for washing the blower* Access doors are required to permit periodic inspection of the plenum, ductwork, blower and washdown system. Rigid polyvinylchloride plastic ducts are now generally used for perchloric acid service. Although it is an optional feature, most laboratory hoods are equipped with movable transparent sash or panels that in addition to conserving air, provide a safety barrier in case an explosion occurs inside the hood. has generally been found to be the most aCf9fBeiVl^ material for this type service, although quarter- inch or heavier acrylic plastic and tempered glass have also been used. Under used as a safety barrier in the hood sash. 2. Utilities Utilities supplied to hoods vary greatly depending upon the requirements of the specific laboratory. Listed below are those that are generally used along with some additional ones that are often supplied. 00004542? Sftl- GENERAL SERVICE SPECIAL SERVICE Cold water Chilled water Compressed air Large sink Cup sink Hot water Electrical Outlet (110 V/3 prong, grounded) Natural gas Lighting Fixture (vapor proof) Nitrogen Vacuum Voltage Controlled Electrical Outlet The controls for the hood utilities should be located outside the hood enclosure and in the side pieces for reasons of safety. The sink or cup sink in a laboratory hood should be individually trapped and connected to the sewer system. A series of untrapped cup sinks having a common drain line is sometimes installed along a laboratory bench to handle condenser water. When connecting this line to the building sewer system, a separate trap must be used--not the trap serving the hood sink. This precaution is necessary to prevent the escape of odorous or volatile materials that may be poured down the hood sink. Laboratory hood sinks should not be used for the disposal of flammable materials. 3. Storage Space for Chemicals A frequent misuse of a laboratory hood is that of storing chemical containers at the rear so that the bottom slot is obstructed. A vented cabinet for chemicals storage is therefore a valuable adjunct. The cabinet can be fitted into the hood understructure and ventilated by furnishing louvered doors and a 2-3 inch diameter connection to the hood plenum. III. EXHAUST SYSTEM A. Arrangement Individual exhaust blowers for each laboratory hood provide the most flexibility for current or future use and allow changing the hood velocity if the use of the hood-changes. Individual blowers also permit the isolation of chemicals that might result in an explosive mixture if a common duct were used. However, it is usually satisfactory to connect several hoods into a single system if each hood has safety stops to prevent the door from being closed completely. Buildings having relatively few laboratory hoods usually have individual blowers whereas those installations where there are large numbers of hoods are usually manifolded to a few large exhaust fans. Although the usual reason for selecting the manifolded arrangement is lower cost, there are several other advantages. These include (1) greater ease in balancing air supply and exhaust volumes and (2) simplified servicing of the mechanical equipment* (3) greater fan reliability, and (4) permits a more effective discharge stack. B. Duct Requirements The duct material and coating should be selected to withstand any anticipated corrosive materials. The ACGIH Industrial Ventilation Manual gives detailed information on the proper design of a ductwork system. Special attention must be given to ductwork, layout to minimize chances for leakage of contaminants into the building. If possible, place the exhaust fan on the roof to eliminate interior ductwork under positive pressure. Otherwise locate the fan as close as possible to the discharge stack and carefully seal seams and joints on the discharge -10- c. Exhaust Fan The type and size of fan selected depends upon the quantity of air required and the pressure drop through the exhaust system. Consult blower manufacturers for data. Special materials of construc tion and coatings are available where the exhaust fan will be placed in corrosive atmospheres. Where flammable vapors or explosive dusts may be encountered, an aluminum or bronze impeller is recommended to minimize chances of spark generation. Centrifugal blowers are generally used for laboratory hood exhaust systems and a belt drive is preferred so that changes in the laboratory hood exhaust system can be made without replacing the entire fan assembly. Avoid using fans with forward curved blades. A roof location is generally best for the exhaust fans unless a special fan penthouse is provided as is sometimes done for large buildings. This arrangement reduces noise, heat, and chances of contaminant leaks in the laboratory area while still allowing easy access for servicing. To minimize noise in occupied areas, also use vibration isolators to mount the fan and flexible connec tions, such as neoprene coated glass fiber cloth, at the intake and discharge. D. Exhaust Stack To avoid recirculation of laboratory hood fumes, the exhaust stack should be located carefully in relation to the building air in takes. In critical cases, wind tunnel tests with models are recommended to determine the pattern of air flow over the building and the optimum location of the exhaust and intake openings. Generally, stacks extending S A L 0 0 00 A -5 A 3 2 -11- at least 10 ft above roof level for one-story buildings and at least 15 ft above roof level for taller buildings should be employed. Dis charge should be directly upward at a velocity of at least 3,000 fpm to minimize chances of recirculation. One additional advantage of manifolding the exhaust of several hoods into a common discharge stack is the increase obtained in the momentum of the discharged air. Where the system operates continuously, only a simple stack is needed. For intermittent operation use either the low resistance, weatherproof type of stack in Figure 2 or a fan drain. Do not use weathercaps. IV. TESTING AND MAINTENANCE OF LABORATORY HOODS All new laboratory hoods should be checked for compliance with the design values and they also should be tested periodically thereafter with a regularly calibrated instrument. Anemometers of the swinging vane type, such as the ,rVelometerH or the heated thermocouple type are recommended for measuring, hood air flow. SriMHinlMpaMk ingft* in. the -----3---------------------r^P-ry^i firrinnn _QHfir th" hoed A visual check of the air flow pattern at the hood face with a ventilation smoke tube is also help* ful because the smoke test can show where disturbing room air currents cause reverse air flow or where there are obstructions to air flow in the hood. In addition to the ventilation tests, there should be a regular program for checking fan speed and the condition of the ducts, fan belts, and lubrication. 'AL 000045 -12- If the tests show that the total air flow has become inadequate for the intended service, then (a) the hood slots and ducts may need cleaning, (b) the exhaust slots may be closed too much or obstructed by laboratory equipment, (c) the blower wheel may be corroded or dirt laden, (d) the fan belts may be loose' or broken, (e) there may be holes in the ductwork between the hood and the fan, or (f) the fan may be rotating in wrong direction. If any individual reading is less than 80% of the average, then the exhaust slots at the rear and the top of the hood need resetting to provide a uniform air flow. For critical applications a sail switch or an air velocity actuated device may be used in conjunction with an alarm, light or bell to signal the need for attention. A. R. JONES/raar 4-11-67 On t'O TABLE 1 LABORATORY HOOD USE FACTORS AFFECTING _________FACE VELOCITY REQUIREMENTS Toxicity Threshold Limit Value (TLV) Maximum allowable concentration (ceiling) Physical Properties Particle size ) Bulk density ) Ease of dispersion) Dusts Vapor pressure Vapors Pressure Gases Conditions Affecting Generation and Dispersion Amount of material Release rate Release velocity Heat load Chances of leaks and spills Operator participation ic TABLE 2 HOOD CLASSIFICATION Hood Class Suitability Minimum Air Velocity at Face of Fullv-Open Hood, fpm Average for Hood Opening At Any Point X* For radioactive materials, metal carbonyIs, beryllium compounds, volatile carcinogens and other materials of extreme toxicity or hazard. 150 125 II For any operation except those offering 100 severe potential hazards. For moderate to highly toxic materials. Recommended generally for all laboratories. 80 III Only for operations where hazard is not high. For materials of low to moderate toxicity such as acetone, ethanol, hy drocarbons other than benzene, nuisance dusts and fumes. 75 50 Requires by-pass dampers to avoid excessive velocities. Consider glove-box where maximum protection is required. SAL 000045 ft Figure 1 LABORATORY HOOD DESIGN FEATURES Gravity or linkage controlled damper or Figure 2 CONCENTRIC DUCT TYPE OF SELF-DRAINING STACK Discharge Velocity: 3000-4000 ft/min. t v. BIBLIOGRAPHY ASHRAE Guide and Data Book--Applications, Chapter 25, American Society of Heating, Refrigerating, and Air Conditioning Engineers, Inc., New York (1966). Barrett, J. C.: Design Techniques for Ventilating Research Laboratories, Air Eng. 4:131 (January 1962). Brief, R. S., Church, F. W., and Hendricks, N. V.: Design and Selection of Laboratory Hoods, Air Eng. 5:70-22 (September 1963), :34-35 (October 1963), .5:22-24 (November 1963). Burke, J. C-, Jr., and Lanahan, T. B.: Design Factors in Better Laboratory Hood Design, ASHRAE J. 3:47 (September 1961). Clarke, J. H: How to Plan Ventilation Systems, Natl. Safety News 87:28 (January 1963) Clarke, J. H>: The Design and Location of Building Inlets and Outlets to Minimize Wind Effect and Building Re-entry of Exhaust Fumes, Am. Ind. Hyg. Assoc. J. 26:242 (May 1965). First, M. W.: New Techniques in Laboratory Ventilation, Air Eng. 1:27 (No. 5) (August 1959). Halitsky, J.: Estimation of Stack Height Required to Limit Contamination of Building Air Intakes, Am. Ind. Hyg. Assoc. J. 26:106 (March-April 1965). Industrial Ventilation, 9th Edition, American Conference of Governmental Industrial Hygienists, P.0. Box 453, Lansing, Michigan, (1966). Ketcham, N. H.: Testing Laboratory Hoods, Am. Ind. Hyg. Assoc. J. 19:324 (August 1958). Lynch, J. R., Study Shows How to Select Lab Hoods to Cut Cooling Costs. Heating. Piping & Air Cond. 39:133-138 (January 1967). Peterson, J. E-: An Approach to a Rational Method of Recommending Face Velocities for Laboratory Hoods, Am. Ind. Hyg. Assoc. J. 20:259 (August 1959) . Peterson, J.E. and Peay, J. A.: Laboratory Fume Hoods and Their Exhaust Systems, Air Cond., Heat, and Vent. 60:63-72 (May 1965) Schulte, H. F., Hyatt, E. C., Jordan, H. S. and Mitchell, R. W.: Evaluation of Laboratory Fume Hoods, Industrial Hygiene Quarterly, (September 1954). 15:195 Steere, N. V.: Ventilation of Laboratory Operations, J. Chem. Ed. 41:A95 (February 1964) 41_:A183 (March 1964) . Viles, F. J., Jr.: Design and Uses of Laboratory Hood, College and University Business 22:41 (June 1957). t ESSO PRODUCTION RESEARCH COMPANY BUILDING 5s cenfral building in this view. Three story, mubi-wingcd portion at front ond single story portion at rear are previously existing structures. The 65,000 sq ft, two story addition featured in this article lies between the two. Ceiling Plenum Is Air "Sourcs/Sink" in Air Conditioned Research Building ; Two-speed exhaust hood fan operation and central plenum concept reduces air conditioning investment and operating costs while eliminating air balance problems. Cj ^<) 0 0 0 4 5 4 4 1 By WtUIAAt B. MATN5Y Project Engineer General Serried Dciit. Humble Oil & Refining Co. Houston, Tex. Construction of a 65.000 sq ft addition to F.sso Production Re search Companj'i 210,000 sq it three story building in Houston in 1966 presented a number of air conditioning and ventilation prob lems. Because of the wide variety of basic and applied research work carried ort in the building (ex ploration and production re search), a large number of envi ronmental criteria had to he satis fied. Some laboratories needed vibiation-free platform-, closely con trolled tomjeiaiure and himi'ditv, and clindnate'ii of airborne p*>l1-n; many other? needed protec tion against no\b:is oi toxic dust, fume*. vapors, or ga-ev that might he released fnun varum- tvs-i-arch prcfccs-r.s. Environmental court ol reqiiIternmt> win fm ! her nun pli cated hy the fact that the degree of activity involved in the various kinds of research performed changes from time to time. The ventilation and air conditioning systems, then, had to have builtin flexibility so that a broad range of changing environmental de mands could be satisfied. Safety, flexibility, and economy were built into the overall pro gram hy correlating air condition ing design and exhaust hood ven tilation design. Review of design plans and on-the-ground r-study show thut during the two year* that the building addition has been cr>mp!ei.d and in use air condi tioning and ventilation arc being accomplished mere safely and eco nomically than in previous instal lations. Factors Affectlrg Design The effect of the rxhamt hoods on air conditioning ir more -ivnifi- cant than that of any other con struction or utilization factor in the building addition. Other im portant /actors include: hot. hu mid climate; long summer season; high ratio of interior to exterior space; heat resulting from heavy usage of electricity; continuous op eration of some temperature sensi tive research work; and integra tion of the building addition with the dulled water system and heat loads of older parts of the build ing. These factors are interrelated and should he considered both for their effects and for the wav in which they influence the effect of exhaust hood requirements on air conditioning. Although exhaust hood ventila tion volume requirements are seinetirn.'s very huge, the building under consideration is not predom inantly a chemical laboratory. Each ruut.i of tin- building falls in to one of three categories depend ing on u*e. These are: 1) non- 5 luhoralury toi>ms such as genera' offices; 2) physical science labora tories for electronic or optica! work; and 3) chemical l.ihuralnries with exhaust hoods. A room can Le changed from one category to another if necessary. Within chemical laboratories, the chemi cal substances used and the magni tude of operations vary with time and location. The most important design factor for the exhaust sys tem U not the magnitude of die chemical operations hut rather tire variety and changing nature of the work. The design objectives for this air conditioning and ventilation proj ect can he expressed as follows: 1) Design a system that will economically maintain close tem perature and humidity control in a (55,000 sq ft building addition. 2) Include exhaust hoods that arc to be used at random intervals and for various durations, and provide for possible future addi tion or deletion of some exhaust hoods. 3) Provide tempered or de humidified outside air to the air conditioning system in \ drying volumes proportional to the maanitude of exhaust hood operation, and maintain proper air balance throughout the building. Reducing Hood Burden In a hot, humid area such as Houston, high air conditioning costs arc imposed by requirements to replace air exhausted through hoods. These high costs, sometimes referred to as "hood burden,*' are well recognized within the indus try. An informative study and demonstration of these costs by Jeremiah U. Lynch of the U. S. Department of Health. Education, and Welfare was published in the January 1967 issue of Halting. Piping & Air Omditinn'riig. The LPKC building is main tained at 75 F ami 50 percent itn year 'round. Extreme rmf.-ide con dition- anticipated arc 96 k I>H and flit K u'H in snimn-'r and 17 F tut iu winter. .Mo-t nf the hoodcan be classified a "counter top height era I. ..-c<I cabinet cvliau.-t ho-*.!- with ?iriiil- vertical sliding fu-h in>lal!ed iu fu>it and served r \i i j ( I t i i ! tj f ( S -1 i } ki> TYPICAL EXHAUST HOOD in new addition. Supply air diffuser and slotted air troffers to central plenum are seen in ceiling. Service corridor runs behind wall at right, which is rear wall of laboratory. METAL NAMEPLATES are attached to all hoods. Special information is typed in unprinfed aluminum rectangular areas (which appear either block or while here depending on incidence of light). by a ro^f mounted Ain/' The 32 exluiu?' Jiooib vajy in size and configuration, but must of them have a maximum sash operating height of 33 in. and a sash width of 1, 5, 6. or 8 ft. Minimum fare velocities are set at 80, 100. or 125 fpm depending on the material to he exhausted. An exhaust hood fan maintains proper capture velocity at the face of a hood when the sash is wide open, but the exhaust hoods' verti cally sliding sashes are sometimes lowered to w ithin a few inches or the work surface when the lined is in operation. It was recognized that exhaust hoods that would sel dom be used with the sash wide ojk*u could unnecessarily impose as great a hood burden as those that would often be operated with the sash wide open. Partially clos ing a hood sash does very little to reduce the volume of air exhausted by the hood fan in typical installa tions. Some individual hoods can ex haust well in excess of 2000 cfm of air and tints require that an equal volume of outside air be drawn into the building and treated. If future conditions so dic tated. the face velocities of the ex haust hood.- could be changed and more exhaust hoods could be in stalled or some could be removed. The composite of all exhaust air requirements at any given time is the "instantaneous hood burden" for the building addition. The summation of the instantaneous hood burden over a period of tim<\ expressed in cubic feet of outside air introduced, represents an in crease in air conditioning operat ing costs. This is because addi tional energy has to be expended to bring the replacement air to the correct temperature and humidity conditions. The highest expected value for the instantaneous hood burden is a component of the maximum building cooling ami healing loads. The initial co.-l of the air conditioning in>(ullatimi is deter mined by the maximum building cooling and heating loads. \\ liilo the costs of cooling equip ment and cooling opetatiug cu-ts are mme significant in this in stance. the reasoning aba applies to healing costs. The term air randitioning ij, intended heu- in its broadest sense, which includes healing and cleaning of air. To reduce unnecessary wasting of conditioned air and aUo to achieve a more constant fare veloc ity over the range oT tavslv posi tions, a luo-spccd fan is installed for each hood. When the sash is pushed up, creating a large face opening, the fan runs at high speed. A rnicroswilch mounted in tile hood is tripped by the sash when it is lowered below a certain position. The volume of air the fan will pull on low speed is adequate to maintain desired face velocity for the smaller cross section area. Using 1800/1200 rpm fans and motors (1:73), the proper place ment of the microswitch to meet the criteria was found to be 57 to 60 percent of the vertical face opening. The theoretical place ment, taking into account basic fan laws and changes in system characteristics resulting from low ering the sash, would have to be computed separately for each ex haust Stood. T. lio theoretical answer would, in all case's, have to be less than 662y percent of the opening. In this installation, the empirically determined value of approximately 58t.2 percent is valid for all ex haust hoods despite differences in hood dimensions and other varia bles in exhaust hood system con figurations. It applies equally to hoods with minimum face veloci ties of 80, 100, and 125 fpin. The amount of conditioned air that is ordinarily lost to the at mosphere is reduced by approxi mately one-third in all instances when the sash is below' the microswitch position. Two-Speed Fan Operation Is Key The excessively high face ulocitirs usually experienced with verti cal sliding sash exhaust hoods at low sash settings can hmc unde sirable effects on experiments and on persons u.ing e\bau*t hoods. A valuable bypiodiiot of the luospccd exhaust hood operation is flic reduction of excessively high face volfii-itii-? at law sa*h settings. Kxamplts of research equipment and materials disturbed by high UPPER PORTION of front of ex haust hood with one panel re moved to expose microswitch for two-speed fan operation. With sash raised, as in top photo, switch is held in position, causing fan to operate at high speed. A large portion of the available face area below Is open. With sash lowered, as m bottom photo, fan operates at low speed. Switchover point is at approximately 58 '/* percent opening. ShL, 000045443 I drafts include huu^cu burners, fine powders, tissue slices, and thin pa per. Some experimental work could be adversely affected by the unnecessarily high flow of cool air over the material. High drafts are uncomfortable to people. Further, when a man stands in front of an operating ex haust hood, his body presents an obstruction to the flow of air into the hood. This causes the air flow to be more turbulent. A low pres- LABORATORY service corridor runs full length of building addition, with lobs backing up lo both left and right walls. Note access pan els in walls, which provide a mini mum of two routes for piping into each lob. Multiple services are shown stubbed and capped in upper foreground; drain lines and vents at right are of glass. Floor ing in corridor is removable alumi num grating. Corridor is accessi ble only to authorized personnel such as members of maintenance crew. sure area develops in the space be tween the man and the hood, and in extreme cases this low pressure area can cause fumes to flow from the hood to the man. The reduc tion of face velocity by the use of two-speed fans rednees the prob ability of this potential hazard. The effects of excessive air ve locities at exhaust hoods is a com plex subject and has been reported on elsewhere. The intention here is only to acknowledge the nature of the problem and to submit that it is reduced by the two-speed fan operation described. Lnique metal nameplates have been prepared for EPUCs exhaust hoods. Besides data on the twospeed features of the hoods, they display other information germane to proper use and maintenance of individual hoods. SAL 000 Air Distribution, Makeup The exhaust hoods under con sideration arc needed at random Secot\d floor, tome os first Central plenum Hood exhaust Fans, typical Mixing box, Diffuser one per fob r--------------- Mixing box, ne servos several offices I_____I Diffuser. rt P.ovr of offices at positive pressure; air flows through troffer* to cenVrol plenum above -1- t I .t T Row of chemical lobs with pressures varying X from below to obovc atmospheric; flow through troffers may be up or down 1 il \fxjxj Row oF lobs Row oF office! Pedestrian corridor maintains near ofmoshpcrlc or slight positive pressure Exhaust hoods, typical Lob service corridor \ (no plenum above) at negative pressure; separately ventilated Pedestrian corridor SECTION through new addition illustrates arrangement of offices, pedestrian corridors, labs, service corridor. and central plenum. Air conditioning fan Heating coil Cooling coil Outside air, tempering coils (oversized) 1 Hill 1 Cold >TTi7lTl 111 H Hot I u I \ Variable flow cf outside air to mechonicol room Variable return from central plenum to mechanical room; openings In woll permit flow but maintain pressure differ ential TEMPERED OUT5IDE AIR and oir from central plenum ore supplied t inlet of air conditioning fan. Proportion depends on number of exhays hoods operating. SAL 000045445 all rooms arc provided with slotted air handling troffers tliat serve as return air grilles opening into a large central plenum overhead. Be cause of their number and si2e, the openings afford considerably more total cross sectional area than would l>c required to permit the amount of air supplied to a room to flow from the occupied space to the plenum. As a result, there is very little pressure drop across the slots. The central plenum opens to the inlet of the high pressure air conditioning fan through a series of manually set dampers, which impose a pressure drop. Tl\us, the central plenum is one source of air for the fan. Recall, however, that the fan inlet is also connected di rectly to an oversized outside air tempering unit. For this' reason, the pressure maintained at the high pressure fan inlet is virtually atmospheric, and the fan does not tend to impose a negative pressure on the central plenum. Instead, the fan receives from the- central ple num volumes of air that are sur plus to the plenum as controlled by positive displacement of air within the plenum. At those times when few exhaust horn] fans are operating, a considerable amount of return air js pushed from the central plenum to the inlet of the air conditioning fan. When many exhaust hood fans are operating, correspondingly less air returns from the central plenum to the air conditioning fan. The fan rect-ives propoitiouately more of its re- It intervals, ami It is not likely that all would In* used simultaneously. As villi other types of air condi tioning loads, there is a usage fac tor or diversity factor that is ap parent, yet difficult to define. The two-speed fan system which allows hoods to exhaust less than maxi mum air volume increases the di versity factor. Having decided on a high pres sure dual duct air distribution sys tem early in the design of the building, engineers gave consider ation to possible later additions of exhaust hoods. With all factors studied, a judicious selection for the maximum makeup air was made. The outside air tempering units, complete with hot and cold coils, were sized for this flow vol ume. The tempering units do not normally handle this large volume. Each laboratory has its own mixing box and thermostat. Some offices and other rooms, served from the same high pressure air system as the laboratories, share thermostats and mixing boxes in groups of three. The supply air volume for each space was designed to meet its cooling and heating re quirements without specific regard to the need for the rootns exhaust ho*>d makeup air. \\ here a large laboratory has only one small hood, the supply air may exceed maxi mum requirements for exhaust hood air. For example, in the case of a large laboratory with a heat load requiring 1200 cfm of air for cooling and with only one hood ca pable of pulling a maximum of 900 cfm, no special consideration for satisfying the exhaust hood air re quirement is needed. The method of providing for the more common situation where maximum exhaust hood air requirements exceed room supply air will be developed in this discussion. The Design Philosophy The addition to the building is so arranged that all laboratories are toward the core or center of the addition. Each laboratory backs up to and shares a common wall with a M-rvie- corridor that runs tin* h-ngtli of the addition. Thi* configmatioii adds U> flexibil ity by making till laboratory serv ices readily available to each lab oratory. It also facilitates technical maintenance. In addition to increasing initial equipment co?t< and operating costs for air conditioning, exhaust lmods usually pose certain control problems. These include excess face velocity problems, already discussed, which occur when hood sashes are lowered. In addition, exhaust hoods often make it diffi cult to introduce the proper amount of outside air at ail times and to achieve mid maintain good air balance throughout the build ing. To seek a solution to these dif ficulties, attention should be di rected to air distribution concepts. The selection, sizing, and place ment of appropriate supply air dif fusers and the determination of proper air velocities and volumes arc important aspects of air condi tioning design. Often there is more than one "correct" solution for a room. Some rooms could he served with two or three air supply dif fusers, assuming all other features were compatible with the number or diffusers selected, and cither sidewall or coiling diffusers might suffice. Other variations are pos sible. It is often possible, therefore, to consider esthetics, cost, and fu ture flexibility in selecting from among the several correct solu tions. Supply air design is not a simple matter of making things big enough. While satisfactory re sults do not demand extreme accu racy, providing diffuser sizes or air volumes appreciably larger than needed can defeat the effort to achieve good comfort condi tions as surely as errors in the op posite direction (undersizing). SERVICE CORRIDOR view shows verticol section of galvanized ex haust hood duct at left ond hori zontal runs at top center. These are short and have only two bends per duct. They are of stainless steel since experience showed that galvanized duel might not withstand corrosive liquids that could condense out on surfaces. Corridor has ample space for more ducts. It is sealed off from occupied spaces end has forced ventilation pulling a:r out at the roof. A somewhat similar aspect of air conditioning is the work of se lecting proper sizes and locatiuus for reLum air grilles. Doth aspects deal with the science of air flow, although return air design is some what less critical and time consum ing than supply air design. This can be appieciated when one con siders that while Loth the supply and return air systems have to transport the proper quantity of air, the supply system must also provide fur correct throw, velocity, induced air, and directfun. In many instances there would be esthetic and economic objec tions to making the return air sys tem, and more specifically the re turn air grilles, larger or more numerous than necessary. But the return air system can he designed and built larger than the minimum requirements without negating good air conditioning results. If there were accruing benefits from oversizing the return air system-- say if it aided in solving air bal ance problems--and if the benefits exceeded any esthetic or economic disadvantages, then the oversizing would be justified. At the EPKC building addition, L quircmenls through the outside air unit as a result. In effect, the rooms, Including the laboratories, are open at the top to a large pool o[ slowly mov ing air. which is conditioned and only slightly warmer than room air. (Of course, the visual impres sion is that of a normal, attractive ceiling.) Consider an individual chemical laboratory room that has several large exhaust hoods, all in operation. The air required l*y the hoods exceeds the air supplied to the room by the fan, and no air flows from the room to the central plenum through the slotted air handling troffers, there being no positive displacement of air toward the plenum. In fact, air flows down into the room from the cen tral plenum as a result- of the slight pressure in the plenum coupled with the slight negative pressure the hood fans impose on the room. The pressure differen tials involved are minute, but the cross sectional area for flow is great. When large quantities of air are exhausted by the hoods in a par ticular room, air from the central plenum enters across the full ex panse of ceiling at luw velocity. If the air flowing from the plenum to the chemical laboratory is a de gree or two above the thermostat setting, as it usually is, the warm ing effect is sensed by the thermo stat, which tends to reset the sup ply air temperature to compensate and maintain comfort conditions in the room. The flow of air from the central plenum into the room directly follows any changes in ex haust hood operation, such as turn ing off a hoocl fan or changing fan speeds. Thu main air conditioning fans are in no way deprived of air when laboratories draw from the pool of return air because the fans have unimpeded access to outside air^ The source of air to the main air conditioning fans automatically divides between tempered outride air and centrjJ plenum air so that tiie system satisfies all makeup air needs. This capability is a function of the design and occurs without any sophisticated et>ntu>l system or special sensrus or dampers be yond is typical fur this geit- Thc design ha? not affected the cost of conditioned air that needs to he exhausted from laboratory hoods. Each cubic foot of air ex hausted through a hood has been conditioned at a price. Hut by re ducing the amount of air that is sometimes exhausted through hoods, the design eliminates the wasteful portion of the costs. The design has also eliminated the problem^ of air balance and other air conditioning problems frequently encountered in air con ditioned lahoialory buildings that have exhaust hoods. Evaluating the Results Test data obtained under a va riety of operating conditions have verified that air flow cannot become unbalanced in this building with the current distribution of exhaust hoods. It is felt that the features of the system will be adequate, without modification, to accommo date such future additional ex haust requirements as may reason ably he expected at this facility. Like any mechanical system, this or a similar system has limitations and must not lie overtaxed. Oth erwise, safely and proper opera tion could be compromised. Prop er sizing ami selection of system components must be based on knowledge of the facility's initial and future exhaust requirements and an evaluation of the com ponents* characteristics. Some readers, particularly tliose used to dealing with processes that arc continuous and more exclusive ly of a toxic chemical nature, mayobject to this design on the grounds that laboratory air should not lie recirculated. Like all brief admonitions, the statement "labo ratory air should not be recircu lated" must be studied and ..under stood. not simply followed blindly. To discuss the caution' properly, we must define each term. ('What constitutes "lahoialory air"?) We must km<w the intent of the cau tion ami the tvpes of situations to which it applies. In this building addition, all chemical work lakes place in prop er cxfuut hoods, and all undesir able air. ai*niN. and the liko are ly is recirculated. Excellent safety practice* and special safely precautions are in effect throughout the laltoratory building, which includes older por tions without two-speed exhaust hood funs and central plenums. Al so, there is a regular schedule for special medical examinations for those employees who work with materials that have health implica tion*. On rare occasions, chemical odors have been detected in old areas of the building, indicating some failure of earlier exhaust ventilation systems to capture all .chemicat materials. This has not occurred in the new addition with its special ventilation features. There have been no injurious ven tilation shortcomings anywhere, but nevertheless all old exhaust hoods were recently rclesflsd, and many were adjusted or Modified in an effort to insure optinnun pro tection aud see that everyone is af forded a work environment com parable with that in the new' addi tion. A 5500 sq ft laboratory* building at the same site, completed in Julv of this year, has ventilation fea tures very similar to those de scribed herein. Conclusions. SAL <>0004544? It is hoped that those who con sider this design in detail will con cur in the conclusion that for this application tlie design with twospeed exhaust hood fans and a cen tral plenum has achieved the fol lowing: 1) Reduced air conditioning in vestment. 2) Reduced operating costs. 3) Eliminated air balance and associated problems. 4) Most important, provided an environment with additional com fort, convenience, and safely for scientists aud laboratory tcclmi* cia ns. The objectives of this air con ditioning aud ventilation project were satisfied. It was primarily the two-speed exhaust hood configura tion. used in conjunction with the special features of the air condi tioning distribution system, that fc. / * AMERICAN INDUSTRIAL HYGIENE ASSOCIATION MEETS: Speakers Decry "Mysterious Air Treatment Devices"; Discuss Laboratory Hood Design; Defend Unvented Makeup Heaters Over 1400 engineers, physi cians, chemists, nurses, physicists, (2) air ionization for thera peutic purposes and other professional personnel (3) air sterilization by dis attended the annual American In dustrial Hygiene Conference, held in Washington. D.C., from May 14 through 17. The meeting, spon sored by the American Industrial Hygiene Association and the American Conference of Govern infectant-coated fibrous fibers, and (4) control of dust soiling by "self-charging electro static" filters and electrical devices which "shatter" particles. mental Industrial Hygienists, was" described by the association as Laboratory Hood Design (Part one of the most successful ever, both in attendance and number of 1). Melvin W. First, Sc.D., con sulting and reserach engineer, technical papers presented. Ab stracts from a few of the 130 papers conclude this report. Newton Highlands, Mass. Rapid increases in the number of fully air conditioned labora The Donald E. Cummings Me tories has focused attention on morial award for outstanding con- the fresh air requirements of tributions to the knowledge and X fume ho..o.d..s. S_ in_c__e_, __f_o_r___s_a_f_e_t_y_'s_ practice of the profession of in- > sake, no part of the exhaust air dustriai hygiene was presented to ^ may be recirculated, the heating Herbert T. Walworth. Mr. Wal- / and (especially) the cooling load worth, Director of Industrial J requirements for exhaust air Health and Hygiene for the Kem- j often represent a major part of per Insurance Group, was the ^ the total. 16th person to receive the award, first given in 1944 and named for a pioneer in industrial hygiene. Some measure of economy may be achieved by using the lowest hood face velocities consistent Mysterious Air Treatment De vices. Melvin XV. First, Sc.D.t consulting and research engi- \ with safety. Most laboratory pro cedures can be conducted safely in J a well designed hood with a min r jmum face'velocity of 70-80 fpm neer, Newton Highlands, Mass.^ when the fume cabinet is located Large numbers of air treatment devices of no discernable merit are being purchased by the public at large and by commercial and in an area free from draughts and out of traffic flow. Essential hood design features for minimum flow include a smooth, converging en industrial institutions. Some have been promoted in vigorous nation wide advertising campaigns for the relief of respiratory ailments in spite of the fact that toxic irri tants such as ozone are emitted. These practices should be a mat ter of grave concern to profes sional specialists in this field lest sound practices be condemned in discriminately with the worthless by a disillusioned public. try shape and flow distribution baffles at the rear. These mini mize air turbulence and differ ences in flow velocity at different locations in the hood face. For especially hazardous operations (isotopes, beryllium dust), a min imum face velocity of 100 fpm is generally specified when optimum conditions exist; and even higher velocities may be preferred to provide additional safety factor. Four classes of mysteifious de vices based on fanciful theories derived from ill-understood scien tific fragments are: ll'i odor control by "counter action" or "destruction with ozone" Because minimum exhaust air volume, even when employing ex cellent hood designs and place ment, usually amounts to 800-1200 cfm per fume hood, further air conditioning economies have sometimes been attained by intro ducing a part of the hood exhaust volume inside the hood enclosure in the form of untempered outside air. This arrangement produces ^ turbulence currents inside the hood enclosure and when hoodU^ face velocity is reduced to 25-30/V'fc'j^~- fpm, safety is sacrificed. Air flow ^ measurements and smoke observations indicate clearly that air supplied directly to the interior of a fume hood has no influence on the volume of room air required for safety and, in fact, produces no discernable beneficial effect. The only factor of importance is air flow which makes a contribu tion to the maintenance of hood face velocity. Laboratory Hood Design (Part 2). Richard Chamberlin and F. J. Vile*, Jr., Massachusetts Institute of Technology, Occu pational Medical Service, Cam bridge, Mass. Proper design of supply air hoods requires that the supply air be introduced outside of the hood face and in such a manner that the operator is always breathing clean air. Studies have been conducted to determine the advantages of sev eral supply plenum designs, and their positions. Included were the use of lateral supply plenums, lat eral supply* plus overhead, over- head with and without full and tapered sides, and overhead sapply with supplemental lateral sup ply at the base. The lateral supply system was found to work well on small hood openings provided the tempera ture of the incoming air did not vary considerably* from, the room air temperature. When the incoming air was 10 deg F higher than the room air, particularly for hoods greater than 4 feet wide, an additional overhead plenum was re quired to overcome the rising effeet of the lateral supply due to density difference of the air. In- (Continned on page 1-261 ^ o X' o :<r AIR CONDITIONING. HEATING AND VENTILATING, JULY, 1942 75 AMERICAN INDUSTRIAL HYGIENE ASSOCIATION MEETS: (Continued from -page 75) coming air at temperatures of 20 deg or more below room tempera tures caused the reverse effect and indicated that for cold weather operations the incoming air would need to be tempered. The maxi mum discharge velocity for lateral supply plenums, without causing turbulence, was found to be 100 LFM. With overhead plenums, the discharge velocity was increased to 140 LFM and the air density problem particularly under warm weather conditions was mini mized. Provision of side enclo sures with the overhead plenums increased the percent of supply air that could be provided with out effecting hood control and loss to the room environment. The maximum percent supply achieved was 60% for overhead plenum types, 70% for overhead with side enclosures and 80% for combination enclosures with both lateral and overhead supply plen ums. Markedly improved face vel ocity profiles in front of the hood operator is obtained particularly with designs involving lateral sup ply air. This study showed that where the cost is warranted, supply air laboratory hoods can be designed that do not interfere with the principles of good hood control. Health Hazards and Control of Direct Fired, Unvented, Make up Air Heaters. George M. Hama, Bureau of Industrial Hy giene, Detroit Department of Health, Detroit, Mich. A study was made of one type of direct-fired makeup heater (also known as a line type gas burner, fiue-less makeup air heater or unvented makeup air heater). The burner investigated takes from the passing air stream at least -/% of the total air re quired for complete combustion at the maximum firing rate and is supplied from an external blower all the air required for combus tion at minimum firing rate. Air determinations were made for carbon monoxide, carbon di oxide, oxides of nitrogen, alde hydes, sulfur dioxide water, vapor and unburned hydrocarbons with the burner operation under nor mal and adverse poor maintenance conditions. Studies were made on one un vented makeup air heater under laboratory conditions and several direct fired makeup heaters in ac tual industrial operations. Test data indicates that toxic and nui sance contaminants were below the maximum allowable concentra tions. Classified and Resale Section CLASSIFIED ADVERTISING RATES: $12 per single column inch ELECTRICAL AND MECHANICAL ENGINEERS Send!* Corporation has immediate openings for gradual* Electrical and Mechanical Engineers to act a* dasign and consulting enginaars in the design of buildings and facilities. Sandia Corporation, located in Al buquerque, N.M., is engaged in re search and development of nuclear weapons and other projects for the AEC. Albuquerque is a modern city of about 225,000; has an axcallent cli mate and many cultural and racraetlanai attractions. Winters are mild, summer nights are cool, and there's plenty of year-round sunshin*. Sandia Corporation offers liberal employee benefits including vacation, retirement and insurance plans, an educational assistance program, and paid reloca tion allowance. Send resume to: Professional Employment, Section 537. S/\rs] ED IAK CORPORATION AsauQurnour. new Mexico Art equal opportunity employer. PRODUCT DEVELOPMENT ENGINEER Recent graduate engineers, with 2-3 years experience primarily in the air filtration field to develop filter media, new types of filters and new applica tions for existing products. Some ex perience in designing testing and customer contact worlt desirable. All re plies confidential. Sand resume to: Mr. A. J. Pearson Manager Product Development Laboratory Owens-Coming Pibarglas Corporation Newark Ohio WANTED ENGINEERS Healing Accessory Manufacturer has positions available in Engineering Quality Control, aad Manufacturing Departments for graduate Me chanical Enginaars. Experience in hydronic Industry preferred. Opportunity for growfti In a small independent company. Send complete resume of education and exparianee including salary requirements and availability to: lox 777, Air Conditioning. Heating A Ventilating. 73 Worth Street. New York 13, N. Y. WANTED: Manufacturers ataats by a areasineet manufacturer ef unit heater*, faresfare, baatine an MeliM fils, aad air eondidenl** unite. Terri fariee ape* in neural seeliens ef the United State* inetadfa* Budsle. Wasblneto* D.C.. Co. hmtbu*. New Orleans. Little Neek. San Aateal*. Albuquerque, El Rase. Hocbetter, Oklahoma City, Or* Molnee, and Rerttand. Neely te fas l(MH. Air Ceaditlenine. Heating and Ventllatlac. S3 Werlb Street. Mew Yerk 13, New Verb, fl-riae inforwatioa toneernlne yew orqtatocllea and etlfar Knee bandied. MANUPACTURERS AGENTS As eetaMfahed waaefweturar with a yean male ease eredawae venillauaq egmawial fa faeklee *w menuteaterer't baeets is etna* sendee* M the eeuetry te tell Ikdr new line ef Ininetrtel fane. The erednate tnnlude anial dew - aed tabeanlal fane, beta dlreet aad belt driven, and newer reef rentilarar*. WANTED MIGIKATION ENGINEER Tltrae to fire ytart experience in dtsfen and da. nlopMMt ot rfrierati(M equipment. Must ban MX. or EX degree er eqanaient. Peminent position with Midwest manufacturer at ice aikinf equipment Excelteat opportunity. Send reman, including salary requirements end availability to: Box 996, Air Conditioning, Hcstinq A Ventiiatlap, 9> Worth Street, New York U, N. Y. AVAILABLE FOR EUROPE Hexline. Air Coeditienine. Retrlferatinj, Plumb. Ine. A Sprinkler Engineer, 83 in MI, 20 year* exeerienee in au ahum ineiudtne Mia*, available in June fer Eareeean esaienaient reeretentinq AneHiaa manufacturer er as Seaiee teetien bead e* lonq farm AC ereject- Rcoistcred In 4 state*. Teaching axetrifee at eolfaee level. Muter etcamfitter. Widewar. eMIdleu. willies te travel. Plea** reply fa Box tfiooo. Air Cendltfanin*. Hftins A Ventilaline, 3 Wertb Street, Hew Yerk 13. N. V. mt JULY. 1962. AIR CONDITIONING, HEATING AND VENTILATING FINAL DRAFT Industrial Hygiene Practices Guide LABORATORY HOOD VENTILATION This Guide was\ prepared by the Committee on Industrial Hygiene Practices from\material contributed by the following authors: H- J- Ett'inger, M. W. First, and R. N. Mitchell. Supplementary information and comments were obtained from: R. S. Brief, J. H. Clarke, J. E. Peterson, and J. H. Rook. \/ \I. INTRODUCTION / Dusts, fumes, valors, and/gases are often released in the course of laboratory work. tion may be used to \/ If the a^iouni of exposure is minimal, general ventila- V dilute the Contaminant to a satisfactory level. However, local exhaust ventilation is morse desirable technique when the contamina tion originates in specific^ireas oiS^/hen extreme toxicity or other factors make contaminant control by simple dilution impractical. Local exhaust is also recommended for controlling dusts arwl other particulate contaminants. kThe laboratory hood is specialized type of local exhaust hood and Is one of the most important .^f all laboratory safety devices. It requires thoughtful design, careful maintenance, and proper use if it is to be effective in preventing unsatisfactory exposures^ laboratory workers j II. LABORATORY HOOD DESIGN A. Face Velocity Under ideal conditions, air flow velocities 20 fpra may-be adequate to control the escape of particles of respirabl^ size or the molecular diffusion of gases and vapors. In actual practixe, however, races of 50-150 fpra are used in laboratory fume hoods for t^o reasons. b'AL 000045450 Evaluation of Laboratory Fume Hoods H. F. SCHULTE, E. C. HYATT, H. S. JORDAN and R. N. MITCHELL Industrial Hygiene Group, Los Alamos Scientific Laboratory Los Alamos, New Mexico n evaluation of laboratory fume hoods under normal conditions of use. Normal use A was originally undertaken with the in this sense would include average room aim of obtaining data on the numerousair currents, activity of personnel in front types of existing hoods at Los Alamos and of and partially inside- the hood, and the for developing design specifications for new generation of fumes in almost any part of hoods. In no instance in the field of hood the hood. The desirable air pattern for a application is the designer asked to recon* hood, therefore, is one with a stable, con ciie as many contradictory requirements as tinuous flow of air from the face to the ex in the design of laboratory fume hoods. haust slots. Any feature which causes the For this reason, there are many custom* air to become unstable and change direc built hoods offered for sale and fume hoods tion would be undesirable. appear to serve as favored subjects for ad* Smoke for the study of flow patterns was vertising copy. The word "fume" in this generated by means of five-minute smoke paper is meant to include vapors, gases, candles and also by means of the unit shown smokes, mists and other airborne materials. in Fig. 1 which burns a mixture of sawdust Design Data and Test Methods and motor oil. An attempt was made to de velop an odor test to give quantitative re A survey of the literature showed that the sults but this was unsuccessful. The test * only design criterion specified in vari procedure used was to observe the flow pat ous handbooks was that of face velocity and terns revealed by smoke generated in vari the recommended value of this varied from ous parts of the hood under varying oper 50 to 150 feet per minute.1'2*3 Even the ating conditions. These conditions included manual on industrial ventilation of the different face velocities, cross draft velo American Conference of Governmental In cities, thermal loadings, slot openings, en- dustrial Hygienists which has become a standard reference work contains little but contradictory information on this equip ment.1 Stockdale and Turner studied sev eral hood types by means of smoke tests and developed a hood which they state op erates satisfactorily at 50 feet per minute face velocity.4'* Others also have used com binations of smoke and face velocity tests.3'*'7'* An inquiry addressed to hood manufacturers regarding their methods of testing yielded absolutely no information. Proposed Method of Testing 'T'HE principal criterion of hood performA ance is the effectiveness with which fumes are captured, retained and exhausted Thb document la based on work performed at the Loa Alamos Scientific Laboratory under the auspices of the Atomic Energy Commission. V Fig. 2. Effect of cross draft velocity in excess of the face velocity. trance shapes and others where such fac tors could be varied on specific hoods. Results Orum uti0n&,rmttm -- Studies indicated that cross currents and other r: Horn air movements are fjwtor Mifectiiig flew wiihir'tteHwwi. It was found that whenever the velocity of the cross currents equals or exceeds the face velocity, serious disturbances of the hood air flow pattern results. The degree of this disturbance is practically independent of hood design. To insure proper performance of a hood under conditions of normal use, the face velocity thus must exceed the expected cross draft which nullifies attempts to provide an in dependent source of unconditioned or un tempered air to a hood. Tests were made on one hood where air was brought in along the sides and top of the hood face to sup ply about one half of the total amount of air exhausted by the hood. Smoke tests on this hood are shown in Fig. 3. Correspondence with other engineers and observations made in other laboratories have convinced us that the independent air supply is not a solution to the problem of saving heated or cooled air. --Careful con sideration should be given to the location of the hood in the laboratory* although here many compromises have to be made. It is desirable to minimize the effects of thermal air currents from heating units and cur rents from doors, windows and supplied air inlets. A perforated ceiling as a supplied air source to a room was found to be almost ideal in its noninterference with hood op erations.5'9 Tests on one recessed radiator showed air currents with velccities as high as 100 feet per minute which would seriously interfere with the hood operation. Also, heat from a radiator located near a hood is largely wasted through the hood exhaust. Since most hoods will be used for operations that potentially may result in a fire or explosion, the hood should bo located so that the exit from the laboratory will not be blocked in case of such an accident in the hood. --It has been reported of f**t fr This is based on actual measurements of cross drafts and air movements produced by op erating personnel. than this value dMMtapBsvariaadMiaateMaitoeei under normal conditions of use. Conservation of heated or cooled air at the expense of an adequate control velocity simply is not justified. In this investigation, controlled cross drafts were induced with a variable speed fan and were measured parallel to the face of the hood with a velometer. The effect of cross draft velocity in excess of face velocity is illustrated in Fig. 2. It is this detrimental effect of cross drafts ft*. 3. Smoke tat on hood having untamparad air supplied through grills on side and top of hood face. 0000 that heavy thermal loads inside a hood cause large variations in hood performance.10-11 To test this, a hood with a work surface 2'x3', a face area 2V'x3' and an average face velocity of 80 feet per minute was equipped with three 1200 W hot plates and four Meeker burners. With only the hot plates in use, the air temperature in the upper part of the hood was 8 C. above room temperature and there was no effect on flow pattern or face velocity distribution. There was no leakage of smoke from the upper part of* the hood. With the four Meeker burners operating as well as the hot plates, the upper air temperature reached 38* C. above room temperature. Again, there was no adverse effect on flow pattern or face velocity. This is illustrated in Fig. 4. How ever, there was an appreciable leakage of smoke around a poorly fitted by-pass damp er in the upper part of the front of the hood. This is due to the thermal head cre ated by the difference in density between the heated air inside the hood and the air outside. This thermal head is larger than the small static suction required to pull air into the hood. Repeating these same tests at a face ve locity of 50 feet per minute, again, showed no effect of heat on the face velocity but an unfavorable effect on the air flow pattern. The leakage of smoke was considerably in creased and some even leaked out of the upper part of the hood face. Obviously, this is an extremely heavy thermal loading and is much more severe than is likely to be encountered in actual practice. In the test Fig. 4. Smofca Hst shewing lack of effaet of high thermal loading on air flow pattern. fig. 5. Air currants moving toward the hood face along tha side walls and bottom of a hood with eornar posts and depressad basa. procedure, all hoods were tested with a thermal loading of one Meeker burner per foot of hood length and in no case was there a serious disturbance of the air flow pattern. study revealed that any plane surface intersecting the air stream would cause serious disturbance of the air flow pattern within the hood and, in many cases, cause air to flow toward the face of the hood. If such a disturbing sur face is at or near the hood face these air currents may be quite serious. For this rea son, hoods without comer posts, center posts or excessively depressed bottoms were found to give the best pattern of air flow. Hoods that have such obstructions have currents of air moving along the side walls and bottom toward the face of the hood as illustrated in Fig. 5. Fumes captured in these currents will move toward the face and be recaptured by the main air stream only at the very edge of the hood opening. Under these conditions, the slightest ex ternal disturbance by cross drafts will ac tually aspirate such fumes out into the Ial* oratory. The inside of the front of the hood above the opening is another surface that will cause air to move toward the face. Fumes captured in this air flow will run down this A. WLAN VIEW OF HOOO WITH CONNER POSTS . PLAN VIEW OF HOOO WITH PLAIN ENTRANCE Rg. 6. Smoke moving down the inside surface of the hood and being recaptured at the face opening. surface and form a sag of variable depth before being recaptured by the incoming air again, at the face of the hood. This is illustrated in Fig. 6. Tests on one hood in dicated that an air foil along the top of the hood opening or streamlining the top of the hood had no practical value in improving this condition, at least within the limits im posed by the operations carried on inside the hood. Sinks were found to create unfavor able air patterns and it is felt that tiMiant Fwfrfcwf nnjr iinliahiiniri be iinlu., Inwh fium Hoods, therefore, should have a smooth and as continuous a surface as possible. Air foils along the side and bottom of the hood entrance improve the air flow some what by eliminating the vena contracta. With face velocities of 80 to 100 feet per minute, an ordinary unobstructed entrance will create a vena contracta about six inches inside the hood face and extending to with in one inch of the side walls. This will cause a slight disturbance to any fumes generated right at the edge of the hood and thU condi tion could be corrected by air foils. It is doubtful, however, that the installation of air foils could be justified if the cost was appreciably in excess of that required for BASE Rg. 7. Air flew pattern! in hood. an ordinary unobstructed entrance. Fig. 7 illustrates the air flow pattern in these vari ous hood types. doors on a laboratory fume hood is largely one of personal preference. There is no doubt that they represent a convenient safety shield that may be put into place quickly. However, as far as the action of the hood itself is concerned, it is doubtful whether doors aid in any manner. It is felt that hoods equipped with doors also should be equipped with some sort of by-pass to minimize fluctuations in the face velocity of the hood and in the amount of air sup plied to the laboratory. Our tests verify the generally accepted fact that sdartlivdNM the hood. By-passes must be carefully designed and constructed. The type of by-pass actuated by means of a mechanical linkage has prov en superior to- gravity-operated or electron ically controlled by-passes. The gravity-type by-pass similar to a furnace draft damper has the disadvantage of not forming a fume-tight seal in the top of the hood and does not give completely satisfactory con trol over the face velocity. The differential pressure produced by the low rate of. air flow into the hood is not sufficient to make the by-pass a positive acting one. Extreme precision is needed in the mounting and bearing assembly of such a by-pass if it is to function satisfactorily. A system of hoods :> 0004 54 54 each equipped with an electronically con trolled by-pass was included in this study. The operation of the electronic controls was highly unstable and the mechanism tended to "hunt" continuously. It became necessary to disconnect all controls and servo motors and lock the control damper in a fixed posi tion. Slots and Plenums--An ideal laboratory fume hood would be exhausted through the entire back of the hood and then directly outside. Normally, however, the space behind the hood is limited and the exhausted air must be carried by ducts upward or down ward from the hood. Thus, it is common practice to exhaust through slots in the rear usually located at the top and bottom of the rear wall of the hood. The slots may have in dividual ducts or may be open into a plenum behind the hood. The relative amounts of air flowing through each of the two slots are usually made adjustable to provide for an even distribution of face velocity. This may be done either by making the slot width ad justable or by dampers in individual ducts leading to each slot. The former method is least desirable since the adjustment is read ily and frequently made by the user who has no air flow instruments to check on the adjustment. As a result, such hoods are usually found out of balance. Also, in many cases, the depth of the plenum behind the hood may be smaller than the slot width and, hence, little is accomplished by small adjustments of the slot width. The most satisfactory arrangement appears to be a separate duct of proper design for each slot with a control damper in each duct. Location of Blowers--Normally, a hood will be utilized to control a toxic or irritant substance and, therefore, it will be unde sirable for such a substance to escape from a hood or from a conveying duct before reaching the discharge point outside the building. Hoods should have the blower located at the discharge end in order to maintain a negative pressure in the con veying ducts.12 Generally, this is not a prob lem when hoods are connected together in an exhaust system. Unfortunately, it seems to be a prevailing practice to install indi vidual blowers in a housing immediately above the hood proper. Exhaust Ducts--Experience at Los Ala mos would seem to indicate that down draft hoods with ducts underneath the floor give little trouble from condensation. It is also an easier task to maintain washdown sprays in such ducts. Hoods connected to over head ducts have given trouble due to con densation and scale particles falling back into the hood. Therefore, if hoods are to be F!g. 9. Los AUmos Hood 8. SAL OOOOQ4J54 Fig. 10. Lot Alamo* Hood C. connected to overhead ducts, the location and design of the top slot should insure that dropping particles or droplets will not fall onto the work surfaces of the hood. While the matter of individual hood filters and washdown sprays behind the hoods was not investigated in this study, the need for such equipment should be considered carefully when the system is being designed.1,1*'14 Utility Connections--The number and variety of utility outlets within the hoods apparently are increasing. The latest Los Alamos hood has a total of 20 outlets for 12 different services. The placement of this number of outlets may create an obstruc tion sufficient to disrupt the flow pattern. This is especially true if the outlets are placed on the side wait close to the hood face. The mounting of utility connections should be such as to insure ease of access for repair and maintenance.**15 Characteristics of Specific Hoods I_TOOD MANUFACTURERS have designed and x installed many more types of hoods than can be described in this paper. Therefore, this review will be limited to a number of hood types installed at Los Alamos which illustrate specific hood features. Los Alamos Hood A (Fig. 8)--This par- U **. Fig. It. Lo* Alamos Hood D. ticular hood is characterized by a sharp edged entrance and a mechanically linked by-pass which exhausts air near the floor. In actual use, these hoods have an average face velocity of 125 to 150 feet per minute. In spite of this high face velocity, the sharp edged entrance visible in the above figure causes back flow and cross drafts will often pull material out from the face of the hood. There is an air flow adjacent to the front half of the side walls and inside the comer post with a velocity of 100 feet per minute toward the hood face. Los. Alamos Hood B (Fig. 9)'--This hood is characterized by streamlined entrances and is commercially advertised as the Oak Ridge type hood. This hood is very satisfac tory under normal operating conditions as long as an average face velocity of 100 feet per minute is maintained. Los Alamos Hood C (Fig. 10)--This hood has a blower located above the hood haa a by-pass damper actuated mechanical ly by the door. It also has a lower slot raised six inches above the hood base. Teste on this hood demonstrated the adverse effect of corner posts and the fact that the ordi nary unobstructed base was comparable in effectiveness to a base equipped with an air foil. b A L 0 0 00 4 5A Los Alamos Hood D (Fig. 11)--This hood is one installed in our own laboratories. It is characterized by streamlined entrances and individual ducts with dampers for both top and bottom slots. It has a gravity-type by-pass at the top of the front of the hood. This hood operates at an average face ve locity of 100 feet per minute and has given very satisfactory performance under nor mal operating conditions. The by-pass does nut operate quite as well as anticipated and smoke will escape through the by-pass under heavy thermal loading. Los Alamos Hood E (Fig. 12)---This hood has a streamlined entrance, no doors and is twelve feet long. This hood was specifically designed for the digestion of urine samples with nitric acid and operates satisfactorily for this purpose. Los Alamos Hood F (Fig. 13)--This hood is included by reason of the fact that it is one of the most common types found in in dustry and universities throughout the country. It has no by-pass and showed the same adverse effects of corner posts. How ever, at an average face velocity of 100 feet per minute, it gives satisfactory perform ance. Los Alamos Hood G (Fig. 14)--This hood has a nearly plain entrance and was origin- Pig. 13. Lo> AUmtt Hood P. ally installed with an electronically con trolled by-pass. The control for this by-pass consisted of a balanced hot wire located in a small opening on the side of the hood, the air flow through this opening increasing as the hood door was pulled down. As previous ly mentioned, this control was unstable and was finally disconnected and the control damper locked in position. The hood oper ates satisfactorily under normal conditions with the doors open. 12. Let AUmot Hood E. Conclusions 1 ---- '----------per minute and Ail mn rniiii* i .. .............I i 1 H face is necessary for hoods handling sub stances of moderate toxicity. 2. The hood should be located in the lab oratory at such a point aa to.minimize haz ards to personnel in the event of fires or ex plosions and in so far as possible should be away from the cross-drafts, high velocity supplied air inlets and other sources of air disturbance. 3. Normal heat loadings will not adverse ly elfect the performance of a hood with an average face velocity of 100 feet per minute. SAL 00004S4b maintain a negative pressure in the con veying duct. 8. Utility connections for hoods should be installed in such a manner to insure maxi mum access for repair and maintenance and minimum interference with air flow near the face of the hood. 9. Conservation of heated or conditioned air can only be obtained by minimizing the size of the hood opening rather than at tempting to introduce unheated air at the hood face or reducing the face velocity. Hoods with very high heat loads should have the major portion of the air exhausted through the top slot. 4. The hood entrance should be kept free of obstructions and irregularities such as corner posts, sinks and excessively de pressed bases. A slight improvement in. the air pattern is obtained by use of air foils. 5. A hood door does not contribute to the performance of a hood but may be desirable for protection in case of a spill or an ex plosion. When doors are used, a mechanical ly operated by-pass of proper design is de sirable to keep the face velocity below 300 feet per minute when the sash is closed to a six inch opening and to provide a con stant room exhaust rate. 6. Provision should be made for adjust ment of the relative quantities of air drawn through the top and bottom slots. This ad justment is best done by means of dampers located outside the hood and not readily accessible to the user. 7. Hood blowers should be located at the discharge end of the exhaust system to Dibliography 1. Industrial Ventilation. pp. 5-55, Lithoprinted Edwards Brothers. Inc., Ann Arbor. Michigan. 1051. 2. Heating, Ventilating and Air Conditioning Guide. p. 995, American Society of Heating and Ventilating Engineers. 1953. 3. Wkat We Make, Catalog No. BOO-1, 2nd ed.. p. 174, B. F. Sturtevant Company. Division of Weiting- house Electric, Hyde Park, Boston, Uaas. 4. J. F. Turner: New Laboratory Fume Hoods Cut Air Conditioning Load. Heating. Piping and Air Coni.. 23: US-6. 1951. 5. Laboratory Design for Handling Radioactive Ma terials. Research Conference Report No. 3. conducted by Building Research Advisory Board, Washington, D.C., 1952. 6. C. G. Dgrwn.BR: A Performance Study of Laboratory Fume Hoods. Knolls Atomic Power Laboratory Re port No. ft64 (January) 1954. T. S. H. Webster. . J. Liljegret* and C. C. Poweu,: Hood for Radioactivity Work. Nucleonics, 10: 65. 1952. 8. A. D. Mackintosh: The Radiochemical labora tory--An Architectural Approach to its Deaim. Nucleon ics. 5: 48 (November) 1949. 9. C. Haines: Planning the Scientific Laboratory. Arch. Record, 108: 107-25, 1950. IQ. H. S. CouiUK, editor: Lotentem Dings, pp. 61-58, Beiohold Pub. Corp.. New York. 1961* 11. H. B. Clay: Controlling Fume Hood Exhaust in Atomic Energy Laboratories. Heating, Piping and Air Cond.. 22: 78. 1955. 12. J. B. Bmbnsd: Fadioieotopea in Jntfnifrp. Rein hold Publishing Co^ New York. 1953. 13. W. B. SWIM: Perchloric Aeid--It Can Be Used Safely. Foundry, 8 : 95, 1962. 14. L. E. PeEVSS and J. H. WATSON: Deaign and Construction of a Small Radioactivity Laboratory. Nneloonie*. 6: 11, 1950. 16. C- N. Sjogren and J. X. Sau.: Oaln and Con struction of Petroleum Refining Laboratory, /ad. and Eng. Cheftk. 41: 1657. 1949. 15. A. K. Solomon and C A. Foster: a Hood f<v Work with Radioactive Isotopes. Ansi CAem., it: 1949. 17. R. E. Holms: Ventilation of an Atomic Energy Laboratory. Refrig. Eng,, 59: 755-758. 1851. . Reprinted from American Industrial Hygiene Association Quarterly, 15:3, September, 1954 (Copyright, 1954, Industrial Medicine Publishing Company) SAL 000045458 Research and Development Section STANDARD PROCEDURES FOR TESTING FUME HOODS IN PLACE --AN UNFULFILLED NEED By ROYAL S. BUCHANAN Dir., Research and Technical Services, American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. For many years, the American Society of Heating, Refrigerating, and Air Condition ing Engineers, Inc., Guide and Data Book, Applications Volume, has contained a chap ter on Laboratories and a section on Fume Hoods1 but the information is confined to a general description of various types of laboratory fume hoods and comments con cerning good and bad practice in their in stallation and use. General requirements for safe operation are described, performance tests are recommended, and methods of test ing are suggested, but such methods have not been standardized nor have minimum standards of performance been agreed upon. Four years ago, ASHRAE's Technical Committee 4.3 on Industrial Environment became concerned with the frequent misuse or poor performance of fume hoods and the lack of any rationally accepted design cri teria and test procedures to assure the safety of laboratory personnel. Although there was already a great deal of public concern and technical interest in connection with urban air pollution, building designers and research administrators seemed to be unaware of air pollution problems within some of their own laboratories, often caused by inadequate performance of fume hoods. Fume hoods of one kind or other have been used for more than fifty years and, as often happens with such familiar equip ment, their fHPMtfHMiNiRBBipMphMMK flHH. Architects, mechanical designers, and laboratory administrators have often assumed that the installation of a fume hood manufactured by a commercially-responsible company (a name brand, if you will) will insure a safe and satisfactory atmosphere for the laboratory worker. However, after re viewing several hundred science buildings, the Architectural Sendees Staff of the Na tional Science Foundation reported in 1965 that much work needs to be done to develop design criteria and performance standards for fume hoods and to achieve wide dissemi nation and understanding of the basic prin cipals of fume hood design, installation, and use.* Improved product design alone is not enough, for variations in placement, use, and balance of the laboratory air conditioning system may override the product design fea tures and result in unsatisfactory and unsafe operation. For example, raaptti nrartMpHBba. Most authorities agree that this may be as low as 50 fpm when the fume hazard is low and should be 150 fpm when fumes may be highly toxic. But a person walking past the front of the hood at only one mile per hour may cause an air movement of 88 fpm, 33 fpm greater than the minimum face velocity for hood safety. A more normal indoor walking speed of two miles per hour might cause an air movement of 176 fpm, 26 fpm greater than the maximum face velocity recommended for a hood. a in or through the laboratory is obviously undesirable and may make safe operation almost impossible regardless of the hood design. Any air cur rents external to the hood may disturb the air pattern of the hood and cause fume outfall. In addition to foot traffic, these may be caused by thermal convection from a nearby heat source, movement of condi tioned air supplied to the room, or rapid op eration of doors or windows in the lab oratory. Any`air motion at the hood face in excess of the hood face velocity disturbs face air flow and may cause outfall. Although safe operation is of prime im portance because the comfort, health, and even the lives of people are involved, the advent of year-round air conditioning for laboratory buildings has greatly intensified the economic problems associated with the selection and utilization of fume hoods. The 'br'iXfcaiiStaJ by fume hoods acids substan- 53 SAL 0000 4 S 4' i Research and Dcvefufment Section enced engineer and made by a reliable manu facturer. Fume hood manufacturer.-; contend that their designs arc adequate and they have no control over the problem applica tions caused by location, air conditioning system balance, etc. Research proposed would include only the most commonly used hood types: standard air-foil laboratory hoods; and air-foil hoods with auxiliary air supply. The scope of the work would include (I) a thorough literature search to determine what research has been done and reported in this area; (2) determination of the mini mum face velocity that must be maintained for control of hazardous operations per formed within the hood when such hood is properly located (i.e., when ambient air movement and outside interferences are mini mal) ; (3) determination of the limits of practical outside interferences such as foot traffic and general air currents that can be tolerated without adversely affecting the control provided by the laboratory hood when the minimum face velocity is main tained (this may require field observations of the ambient air movement and outside interferences likely to be encountered in actual hood installations) ; (4) development of standard tests to be used for evaluating the factors mentioned in (1) and (2) above; (5) establishment of basic design criteria for auxiliary air supply type hoods to main tain the hazard control features provided by a standard air-foil laboratory hood; (6) de velopment of a standard test procedure for evaluating performance of the hood with auxiliary air supply, such test procedures to be applicable under imbalance conditions (i.e., where supply air quantity is the same as the exhaust quantity) ; and (7) develop ment of a standard test procedure for deter mining the percentage of supply that is en trained and exhausted. It is intended that test procedures de veloped as a part of this research will take into account interferences typically encount ered in laboratories and will thus provide a standard method of generating the per formance data published by the hood manu facturer. The tests would usually be made and performance characteristics certified by the manufacturer. Such standardized per formance requirements could also be used in the specifications issued by the purchaser. A somewhat less complicated standard pro cedure would be developed for testing and evaluating the hood performance in-place to assure that no unforeseen conditions exist (beyond those anticipated by the standard) which would interfere with the safe and uniform functioning of the hood. There seems to be general agreement that standard test procedures are needed to as sure safe performance of laboratory fume hoods as they are actually used. ASHRAE is willing and anxious to develop such pro cedures in an impartial way and is even willing to pay part of the cost of research needed. Considering the number of laboratory fume hoods installed annually and the poten tial hazards involved, there is surely enough interest among users and manufacturers to assure financial support for the remaining part of the cost. I am sure that we are not many years away from air quality standards for the air inside buildings. When such standards are set, there will be no allowance for the malfunction of fume hoods. The crash pro gram required at that time to improve fume hood design and application will cost many times more than the proposed research. I hope that we will anticipate the need and be ready before there are legal require ments in this field. Reference* 1. ASSRAS Guide and Data Booh, 1968 Ap plications Volume. Chapter 10. 2. Horowitz. Harold; Gelder, 3.A.; Dugan, Caldwell N. r "Fume Hoods for Science Laboratories." AIA Journal, July 1965. ss SAU )00A'o460