edited by MALCOLM M. RENFREW University of Idaho MOSCOW.
I&
83843
Laboratory Air Quality: Part II. Measurements of Ventilation Rates Samuel S. Butcher, Dana W. hlayo, and Sandra M. Heberl Bowdoln College. Brunswick, ME 0401 1
Ronald M. Pke Menlmack College. North Andover, MA 01845
Part I of this paper ( 1 ) described a simple model for estimating laboratory concentrations of gas phase pollutants. In this part the measurement of ventilation rates and applications of the model will be discussed.
Direct Measurement of Air Flow
Measurement of Clearance Time
Where either the air inlet or the exhaust for the laboratory occurs at a well-defined mint.. the volume rate of air flow may be ohtained as the oroduct of an air vel6eitv and an area. For the general c u e of air flow measurements the reader ia referred to The Committee on Industrial Ventilation (6). In the case of many undergraduate laboratories, all air exhausted is taken out through hoods. Flow rates through these hood faces may be conveniently measured with any of several air velocity instruments. For these purposes an accuracy of 20-30% is adequate. Velocities may not be the same everywhere, and it is important to measure at several points. The hood opening may be set ta a height which will give an air velocity in the readable range of the instrument being used. Then if the velocityreadings represent equal areas, the area averaged velocity is simply the numerical average of these readings. As an example, if one wishes to take four readings across the face of the hood, the hood opening is divided into four equal areas and a velocity reading is caken at the center of each of these are=, at ',s,?s. %. and 'a of t he drstance across the hood face. For a wider hood opening one might use eight readines as two rows of four readings. The number of points measured dependson the degree to which the velocity varies in the h w d opening. The volume flow through the hood is then the product of the area-averaged velocity and the area of the hood opening. Many velocity measuring instruments and equipment design specifications are expressed in feetlminute or cubic feetiminute. It may be convenient first to get the flow rate in cubic feetlminute and then make the conversion to euhic meterslminute, since the engineering staff is mare likely to be familiar with cubic feet notation. After the volume flow rate for each hood of exhaust port has been measured, the sum of these rates is the room ventilation rate, Q.
The use of the clearance time in cbaracterizing the ventilation rate is discussed by the Committee on Indoor Pollutants (7). The concept of clearance time may be examined by going back to eqn. (1) (in part I). In this method a tracer is injected into the room a t t = 0 and the concentration of the tracer is measured as a function of time as the ventilation equipment clears the room. The concentration of the tracer in the room will be given by the following equation if the room is well mixed.
.
~~
Measurement 01 Ventllatlon Parameters Emission factors for selected processes The other factor have been reported (MI. required in the concentration model is the ventilation rate, Q. Bayer (2) discussed one method for measuring the ventilation rate in laboratories, and an overview of several methods has been presented by the Committee on HazardousSuhstances in the Laboratory (5).A survey of methods appropriate for instructional laboratories is presented here. The results of applying these methods to sin instructional laboratories are summarized in the following section. It is very important in the application of any of these methods that all factors that affect the ventilation rate he controlled. Doors, windows, and unit ventilators all may have significant side effects on the main system ventilation rate.
Same1 S. Butcher ard Dana W. received their PhD degrees from Haward University and Indiana university, respe*ively, and are prcfessors of chmsnry at owd do in College. Ronald M. Pike received his PhD from Massachusetts lnsliiute of Technology and is pmfes~wot dlemistry at Menimack College. Pike has also spent mree semesters at Bowdoin College as Visiting Pickard Professor of Chemlsby. Sandra M. H M is e I984 chemisby graduate of Bowdoin College. Mayo
~
~~
~
.
Volume 62
In this case co is the initial concentration of tracer (at t = 0); t is the time at which the measurements of e are taken; and Q and V retain their previous meanings. The clearance time, or well-mixed ventilation time, is also expressed by t, (= VIQ). tu is the time required for the concentration to fall to l/e of its initial value. The reciprocal of t, is the air exchange rate (usually expressed in ACH. or air changes oer hour). One may ohtain a value fur t, by plotting In (c) as a function o f t . The dope of this curve is then -l/r,. It is important that the air in the room remain well mixed while these measurements are being made. (An 18-inch diameter fan worked well in our studies.) Bayer (2) has measured t, with this method by using a nontoxic hydrocarbon tracer and determining the eoncentrationby direct injection of air samples into a gas chromatograph. Sulfur hemfluoride and other gases have been used in several studies (7). We have used an integrating nephelometer to measure the light scattering from MgO smoke produced by a flash powder as a function of time. Any method employed should have the capability of acquiring a t least three or four measurements of concentra-
.. .
(Continued on page A262)
Number 10 October 1985
A261
Table 1.
Room/Vent:
tiou in time t,. Poorly ventilated lahoratories will have a t, in excess of 20 minutes, whereas t, will be less than 10 minutes for well-ventilated spaces. Once t , is known, Q is obtained as Vlt, in eqn. (6). The air erchange rate and t, provide a measure of the performance of ventilation equipment which may he useful for laboratory to lahoratory comparative purposes. The Committee on Hazardous Substances in the Lahoratory (5) suggests a range of 6 1 2 air changes per hour as adequate for general laboratory ventilation. Ventiiation Rate from Equipment Design
In some cases, physical plant personnel are able to provide an estimate of the ventilation rate from knowledge of the design characteristics of the air-handling equipment. Such an estimate may provide a useful approximate value for Q, hut care should be taken in using these figures. Flow rates depend on the availability of make-up air, the condition of filters, and the state of maintenance of the air moving equipment. Thus, if a comparison of the design ventilation rate with the value measured hv method A or R above leads to a large dLparity, that obgervation may indirate a major pwh. lem with the ventilatmn equipment. Measurements in Selected Laboratorlea
We have measured ventilation rates for several lahoratories using the methods described above. The air velocities have been measured at hoods or a t other well-defined openings with an electronic instrument using a heated ceramic probe. This instrument (Kurz Mode1 440) can measure air velocities in three ranges up to a range of 0-2500 feetImin and offers readability of about 5 ftlmin at 100 ftlmin. The instrument was calihrated against a Pitot tube at air velocities of 550-700 ftlmin and the same relative correction (approximately 15%)was applied a t the lower velocities encountered in these studies. The clearance times have been measured by the MgO smoke technique and the results of several trials are collected in Table 1.
It may be seen in Tahle 1 that there is a cmsrdcrablr range in the observed air ex~ h a n g erates. Other things heing equal, the concentration will he inversely proportional to the air exchange rates in Tahle 1. This means that there is also a large range in the expected concentrations if the emissions are the same or, conversely, a large range in operations that could he allowed .-withnut ------- sen---erating excessive concentrations. The results of Table 1and our experience with measuring ventilation rates give some idea of the types of errors that can be expected. They also indicate some of the factors which may reduce the effectiveness of any ventilation system. Many of these prohlems have also been discussed by others (5).
A282
Journal of Chemical Education
Illy Values in Air Changes per Hour
Room Volume fcubic meters1
Measured by Measured by Method A. Method 8, o/!P dincldP
Bowdoin 101 Bowdoin 3
1
565
2.8
2.8
2A 28
754
754
2.2 2.8
1.6 2.0
3A
519 519 519
14 26 10.2
23 46 7.8
(10) (4)
Bates 206
546 546
13 9.6
(3)
Colby 303
38 3C
(0.7)
~otes
(0.1)
48 4C
546
N.M.
12 5.0 20
5A 50
323 323
10.8 4.1
9.0 7.2
(1.2)
Menimack 302
6
345
N.M.
10.2
(3.6)
Menimack 301
4A
(2)
'The numbers in me len-hand column refer to laboratories and me leners define dinerem ventilation configurations mr the same laboramry. ON.M, indicates not measured. (The numbers in me parenmses are me ranges of cbrsrvatlonsfwhsremere is mwe than onedeterminatlon).me ventilation rate was avsilabletromvemllationspeciflcationa foronly one case. The rate was 12 air changes pw hour fw case 3c.
.
1) Errors from the uncertnintv . in mca.. . ... surement of low-velocity flows. Air flows less than 100 ftlmin are difficult to measure because of the noise in the velocity signal generated by turbulence in the room. This error becomes more significant when the measurements are made in very large openings, such as doors. The velocity measurements for 3B and 4B are especially suhject to this sort of uncertainty. ~
~~
2) I t may he difficult to achieve the wellmixed condition necessary for the clearance rate measurement at higher ventilation rates. In 3A and 3B the air was exhausted from the room by 32 individual hoods and three large hoods. T h e make-up a i r entered through a n open door a t 150-300 ftlmin and may have impacted on the nephelometer, giving a higher than expected ventilation rate.
3) Theconfiguration of exhaust and inlet may result in a "short circuit" in which the make-up air is poorly mixed with pollutants before heing exhausted from the room. In 4A and 4C the same exhaust hoods were used. In 4A, the make-up air was supplied mainly through an open door which was not far from one hank of hoods. In 4C the door was closed and the make-up air was supplied from loose fitting doors, partly open windows, and heating vents.
Some applications of the concentration model and ventilation measurements are presented to give a general idea of conclusions which may he drawn from a few ohservations. As one example, the application of ventilation parameten and emission factors to an estimate of concentration are outlined. The emissions of solvent incetone) per student are observed to he 5600 mg when the rccryitallization uf napththaleoe is carried out a t the scale of the conventional organic laboratory: for the equivalent microscale proeess the emissions are 1200 mg per student. The recrystallization in this case uses ordinary filtration not assisted by vacuum. A value of 0.3 is assumed for the mixing factor (k) with a 20-student laboratory section lasting three hours. In the ease of lahoratory configuration 1with 2.8 air changes per hour (see Tahle I), the dilution fador kQtoln, is 71 cubic meters per student. For laboratory configuration 3B with 28 air changes per hour there are 654 cubic meters per student available. The average concentrations expected to result from the recrystallization process alone are collected in Table 2. Thus, in Laboratory 1, macroscale recrystallizations should not be conducted in the open lehoratory usine solvents with PEL'S less than 80 mgjm3 (assuming avolatility similar to that for acetone). Of course, the final evaluation of risks from vapors must also include other operations with this solvent and exposures from other substances. A procedure is de~
from a large number of sources which mav he difficult to measure. This was the ease fgrr rooms 4 and 6. In these cases, the clearance rare method may be theonly way to measure theventilation rate.
~
~
~
Table 2. Average Concentration Expectedto Result lrom Recryrtalllzation (Acetone Solvent)
Laboratory Configuration L) Some laboratories receive make-up air
~
Concentration, mg/rn3
Macroscale
Labwatwy I Laboratory 3 8 Laboratory 1 Labaratwy 3 0
17 2
from more than one substance (8). In a related stud".. . we have made a brief evnluatiun of vrlority measuring inscrumentswhich presrntly lieat theeatremeaof priceand flexibiliry.Oneof there isa heated ceramic probe electronic instrument which casts about $8W and has three sensitivity ranges going up to 2500 ftlmin (Kurz Model 440). I t has averyreadahle scale and aprohe on a long cord for use in messuringvelocities in ducts. The other instrument is a suspended vane meter which costs about $30 (Dwyer "Vaneometer"). The vane meter is designed for measurement of hood face velocities and has an upper limit of about 400 ftlmin. The instruments were used "as is" with factory settings. The differences between averages of seven measurements for the two instruments were only 15-2490 over the range of velocities studied (50-125 ftlmin). Thus, it would appear that in a number of cases, ventilation rates may be successfully monitored with very inexpensive devices. The measurement of higher velocities and the measurement of velocities in duets will require a probe type instrument, of which there are many available in the $2W-$8W price range. ~~~~
~~~~~~
~
Dlscusslon The simplified model described for estimating vapor-phase pollutant concentrations can provide a useful starting point in planning for safer instructional laboratories. The a ~ d i e a t i o nof such a model in an accurate piedietive sense will have to wait until there is better understanding of mixing factors ( k )and emissim factors ( m j . The values derived from this a p p n m h may he calibrated with measurementsof concentration to provide a means of estimating impacts of selected changes in lahoratories. As an example, if the measured concentration of a certain vapor is two times the desired limit, the model suggests a number of possible courses of action. 1) Emiasions per student may be cut in half. The microscale laboratory (3,9) approaches the prohlem through this measure and brings about a large reduction in emissions by reducing the amount of material used. Reduction could also he obtained hy changing to less volatile solvents.
2) Operations may he restricted to hoods. 3) The main system ventilation rate can be doubled. Since the mixing fadar may be affected by s change in the ventilation rate, the change in the concentration may be less than that predicted hy the model.
-
carried out in the men. While knowing the venrilotion rate will highlrght rhr optiun~i avnilahlr for imprwing lahorntory air quality, periodic measurement uf the vrntilation rate can also provide an indication of the state of maintenance of ventilation equipment. Measurement of the rate by the clearance method does take more time and care hut this mav be the oulv aooro~riatemethod in eases where air enters and leaves the - ~ -..-~. laboratory at many points. In those situations where air is exhausted through a few locations which can be easily measured, the ventilation rate for a laboratory can he obtained in a few minutes. In some cases these rates may be obtained with the same inexpensive instruments used to monitor hood face velocities. Use of this model i~ not mtended to substitute completely for the actual memurement of concentrations. In those cases where modeled concentrations are much less than Permissible Exposure Levels, no further work may he necessary. Where modelled results approach or exceed PEL'S, personal exposures should be evaluated by monitoring. ~~~~
. .. .
~
~
~
~~~~~~~
~~~~~~
Acknowledgment We thank James Boyles, Phil Meldrum, and Larry Shoer of Batea College, Gary Mahbott of Colby College, and J. David Davis and Rita Fragala of Merrimack College for their assistance in making the ventilation measurements. The authors also acknowledge the generous support of the Surdna Foundation and the ARC0 Foundation for the development of this laboratory program. Literature Clted (1) Buteher,S.S.,PikeR., Mayo, D. W.,andHelwt,S.M., J. CHEM.EDUC.,in press. 12) Bwer, R.. J. C m . EDUC., 69, A385 (1982). (3) Buteher, S. S., Mayo, D. W.. Pike, R. M..Fmte, C. M.. Hotham. J. R., and Page, D. S., J. C u m . Eouc.,62. 147 (1985). (4) Bayer. R E., J. C m . EDUC.,57, A287 (1980). (5) Committee on Hazardous Substanew in the Labmato-
w, National Reaeaueh Council. "Prudent Praetrrrs for
Handling Hazardous Chemicals in Laborak&s." National Academy PI-. Washin@on,DC. 1981. ( 6 ) Committee on InduaVial Ventilation, '"lndu3trialVentihtion,"AmericmCounciiofGovernmentalIndusVial Hygieniah, Lanaing. MI, 1982. ( I ) Commitfee on lndoar Paliutants," National Academy Press, Washington, DC. 1982. (8) "Code of Federal Regulations Title 29," Part 1910.IWO. U. S. Go*. Printing Office, Weshiomton, DC, 1984. (9)May0.D. W., Buteher, S. S., Pike.R. M.. Fwte, C. M., Hotham, J. R.. m d Page, D. S., J. C m . Eouc.,62, 149 (1985).
4 ) The numher of students in the labora-
tury semon mn be reduced by 50%. 5 ) A substance can be substituted which
has a safe limit two times that of the original substance. Theventilation rate plays a central role in the concentration model; therefore, values of Q must be obtained with reasonable seeuracy. There are also other benefits to he derived from measuring the ventilation rates for laboratories in which experiments are Volume 62
Number 10
October 1985
A283