airborne particulate emissions from a chromic acid

8/12/2019 Airborne Particulate Emissions From a Chromic Acid

1/7

This article was downloaded by: [111.223.252.15]On: 05 March 2014, At: 03:30Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of the Air & Waste Management AssociationPublication details, including instructions for authors and subscription information:

http://www.tandfonline.com/loi/uawm18

Airborne Particulate Emissions from a Chromic Acid

Anodizing Process TankRobert C. Pegnam

a& Michael J. Pilat

a

aDepartment of Civil Engineering University of Washington , Seattle , Washington , USA

Published online: 07 Mar 2012.

To cite this article:Robert C. Pegnam & Michael J. Pilat (1992) Airborne Particulate Emissions from a Chromic

Acid Anodizing Process Tank, Journal of the Air & Waste Management Association, 42:3, 303-308, DOI:10.1080/10473289.1992.10466994

To link to this article: http://dx.doi.org/10.1080/10473289.1992.10466994

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content) containedn the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,

and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otheriabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
http://www.tandfonline.com/page/terms-and-conditionshttp://www.tandfonline.com/page/terms-and-conditionshttp://dx.doi.org/10.1080/10473289.1992.10466994http://www.tandfonline.com/action/showCitFormats?doi=10.1080/10473289.1992.10466994http://www.tandfonline.com/loi/uawm18


2/7

ISSN 1047-3289J. AirWasteManage.Assoc.42: 303-308

Airborne Particulate Emissions from a Chromic AcidAnodizing Process TankRobert C. Pegnam and Micha el J. PilatDepartment of Civil En gineeringUniversity of WashingtonSeattle, Washington

Particulate mass concentration, particle size distribution, andparticle chemical composition measurements have been con-ducted on the gases exhausting from a chromic acid anodizingprocess tank. Particle mass concentrations in the 200 to 2 0,000f tg /m 3 range were measured using open-faced filters (47 mmdiameter) adjacent to the process tank liquid and with closedfilters (90 mm diameter) in the exhaust duct. Particle sizedistributions, measured using University of Washington Mark 3and Mark 20 Cascade Impactors, showed the particle aerody-namic mass median diameter was about 3 microns. Chemicalanalysis of the particle samples obtained by the M odified EPAMethod 5 sampling train, the Mark 20 UW Cascade Impactors,and by the 47 mm and 90 mm diameter filters showed Cr + 6concentrations in the 20 to 1,500 |xg/m 3 range with over 99percent of the chromium in particles larger than 1.0 micronsdiameter. An integrating nephelometer w as used to measure th elight scattering coefficient of the exhaust gases upstream of thewet scrubber. The light scattering coefficient increased by afactor of about 2-3 over the background level during the 40minute time period while a part was being anodized. The bscatvalues ranged from 3 x 1O~5 to 3 x 10~4 meters 1 for theaerosol particles less than about 6 microns aerodynamic diame -ter.

The objectives of this research were to characterize (massconcentration, size distribution, and chemical composition)the airborne particulate emissions from the chromic acidanodizing of aluminum and relate t he emissions to processvariables (electrical current usage in units of amp-hours,

ImplicationsAirborne emissions of hexavalent chromium are facingincreasingly strict air pollution emission stand ards. Mea-surements of the particulate mass concentration, parti-cle size distribution, and particle chemical compositionprovide information of significance to the design ofparticulate control equipment used to reduce the Cr + 6emissions. For the particulate emissions from chromicacid anodizing of aluminum, over 99 percent of thechromium was in particles larger tha n 1 micron diame-ter and accordingly very high control efficiency for thesmaller submicron aerosol particles may not be neces-sary.

square feet of aluminu m part surface area, and tim e duringprocessing of the aluminum parts). This research hassignificance to the control of emissions of aerosol particlesand hexavalent chromium from chromic acid anodizingprocess tanks. Hexavalent chromium is regulated by thePuget Sound Air Pollution C ontrol Agency1with an emis-sion stand ard of 99.0 percent c ontrol efficiency an d 0.03 mgC r+ 6 /amp-hour (if facility-wide Cr + 6em issions are greatertha n 1 kilogram/year after obtaining 95 percent control)and 99.8 percent control efficiency and0.006mg Cr+ 6 / a m p -hour (if facility-wide Cr + 6 emissions are greater than 1kilogram/year after obtaining 99.0 percent control).Literature Review

Chromic acid anodizing of aluminum was developed byBengough and Stuart in England in 1923, as reported byWernick and Pinner (1964).2 The electrolytic process pro-duces an almost insoluble and strongly adheren t protectivefilm on th e alum inum part. A nodizing can also be used fordecorative purposes as well as the forma tion of a porous an dadsorptive base for further coatings applications, such aspainting. The consideration of hexavalent chromium asboth a toxic and carcinogenic air pollutant has caused metalfinishers to conform to more strict air pollutant emissionsstandards.The methodology of mea suring th e air emissions of C r + 6testing has been reported by Bivens and DeWees (1987). 3Similar emission test methods were prepared by PacificEnvironm ental Services, Inc. for th e M etal Finishing Asso-ciation of Southern California.4 In both cases a form of theEPA Method 5 Sampling Train was used for field sourcesampling followed by laboratory chemical analyses of thesamples for Cr + 6 . Chemical analysis methods for hexava-lent chromium concentrations include the diphenylcarba-zide colorimetric and ion chromatography procedures. Ana-lytical methods for hexavalent and total chromiumconcentrations in the airborne particle samples were re-ported by Bhargava et al. (1983),5 Abell and Carlberg(1974),6 and Thomsen and Stern (1976).7 All concur thatthe colorimetric method using diphenylcarbazide is anaccurate means of measuring hexavalent chrom ium concen-tratio ns in aqueous solutions down to about 0.5u,gof C r+ 6 .Total chromium concentrations are measured by atomicabsorption (AA), ion chromatography (IC), and inductivelycoupled plasma (ICP).Koropchak and Roychowdhury (1990)8 reported C r+ 3 /C r + 6 ratios in the 2 to 6 range as a function of particlediameter in the 2.5 to 25 micron diameter range forairborne particles in the room air of a small chromeelectroplating shop. No reports have been found of the sizeCopyright 1992Air Waste Management Association

March 1992 Volume42, No. 3 303


3/7

1Chromic

Anod iz ingTank

1

; < - Port A

>

WetScrubber

P o r t B

Figure 1 . Schematicofair flow from chrom ic acid anodizingtank.

distribution or chemical composition as a function ofparticle size for airborne emissions from chromic acidanodizing process tanks .1 41 6Chromic Acid Anodizing Process Tank

The chromic acid anodizing tank studied is locatedattheAuburn Facilityofth e Boeing Comm ercial Airplane Com-pany. The tank is one of several other (cleaning, alodining,rinse, etc.) tank s usedinthe final processingofaluminumparts priorto painting. The tank dimensionsare115 feetlong, 5 feet wide, and 15 feet deep. The chromic acid liquidlevel is usually maintained at about one foot belowtheupper edgeofthe tank, which correspondsto avolumeofchromic acid of abou t8,050cubic feet (60,200 gallons). Theair from the anodizing tank is exhausted through horizon-tal vents (m ounted flush with the vertical sides of the tan kalong both sides but not on the tan k ends) located about 6-9inches abovethe liquid surface and viaductingto a wetscrubber, asshown inFigure 1,Connectedto theupperedge of the process tank are 38 air exhaust ducts 19exhaust ductsonthe south side connecttoone south sideduct and 19 ductson theno rth side connecttoone n orthside duct).Thenorth sideand the south side ducts jointogether abo ut 30 ft. upstream of the scrubber. The tota l airflow into the exhaust ductsisabout 45,000 acfm (68-80Fand static pressureof-1 .2 H2O )orabou t 1200 acfm perduct. The exhaustairflowsat avelocity of abou t 49 ft/sec.through the 4 ft. by 4 ft. main duct and into the wetscrubber.

Formation of Aerosol Particles During Anodizing ProcessAerosol particles are formed du ring the ano dizing processfrom the bu rsting of gas bubbles generated by the anodizing

Mounting RackPartHanger

^ Gas

C h r o m i c c i d S o l u t i o n

AluminumP ar t(Anode)

TankWall(C a t h o d e )ToScrubber ToScrubber

Figure2. Formationofhydrogen and oxygengasbubblesduring anodizing.

b u b b le f i lm d ro p s jet d r o p sFigure 3. Aerosol formation from bursting of gas bubbles.

process and by the air sparging operation to mix the liquid.The aluminum partstobe anodized ar e suspe nded fromaframe rack platform. This platform andrackareloweredinto thetank ofchromic acid.The platform rests on theprocess tank mounts whichare located just abovethe airexhaust ducts. Once in position, the platform is locked andthe electrical connectionsare set in place (bya hydraulicmechanism).At the start of the anodizing process,thevoltageandcurrent are increased slowlyfor thefirst fiveminutes. For theremaining 35 minutes, the voltage isconstant at about 22.5volts DC. Theelectrical curren t(amperage) may peakashighas7000 ampsin thefirst5minutes but reducesto alower m agnitudeforthe remain-ing 35 minutes. The current magnitude is dependent uponthe total surface area of the pa rtsinthe tank. The c urrentdensity is maintained between5 and 10ampspersquarefoot of pa rt surface area . Air sparging is used for channe lledparts . The air sparging is operated prior to anodizing to mixthe acid solution and to en sure tha t air pockets are removedfrom the channelled parts. Airsparging wassometimesoperated for the d uratio n of the anodizing process (about 40minutes) whereas other times it occurred for the first5minutes.Anodizing isanelectrochemical process. When electricaldirect current isapplied thro ughanaqueous chromic acidsolution, gases are formed a t each of th e electrodes. Oxygenga sisformed on the positively charged (anode) alum inumparts and hydrogen gas on the negatively charged (cathode)tank walls, as is shown schematically in Figure 2 . After theoxygen and hyd rogen gas bubbles rise to the liquid surface,they bu rst a nd em it liquid aerosol droplets (particles) whichare swept awayin the airflowingto theexhaust ducts.Resch (1986)9reported tha t the bu rsting of a gas bubbleatthe surface of a liquid generates two distinct typesofaerosol droplets, liquid film aerosol droplets and liquid jetaerosol droplets, as shown in Figure 3. The liquid filmaerosol droplets are of small diameter (perhap s in the 0.1to1 0 microns diameter) whereas th e liquid jet aerosol dropletsare somewhat larger (about 0.1ofthe diameterofth egasbubble).Experimental Measurements

The aerosol particle emissions were characterized usingfour m easurem ent methods including filters (47 mm a nd 90mm diameter), Greenburg-Smith impingers, cascade impac-tors (Mark 3 and Mark 20 University of WashingtonCascade Impactors), and an integrating nephelometer. Thesampling wasperformed over the liquid surface of thechromic acid anodizing tanks,at theinletto theexhaustducts adjacent to the process tank (port A , and in the mainduct upstrea m of the wet scrubber (port B in Figure 1).

Modified EPA Method 5An EPA Method 5 sam pling train w as modified for use asa sampling system for airborne particulate chromium,similartotha t required by the P uget Sound Air Po llutionControl Agency.10 Theentire sampling section (sampling

nozzle, sample probe, connecting tubes, and impingers)ofthe train wasglass.Theglass sam pling nozzlewas posi-304 J . Air Waste Manage. Assoc.


4/7

tioned about midway inside the 4 ft x 4 ft duct transp ortingth e 45,000 acfm airflowfrom the processing tan k to the w etscrubber (port B in Figure 1). The unh eated glass samplingprobe was connected to the Greenburg-Smith impingerswith ground-glass fittings. The ice-cooled impinger boxcontained four modified and one standard Greenburg-Smith impinger. The first three impingers contained 150mL of 0.1N NaOH, the fourth was dry with a glass woolplug, and th e last co ntained 200 g silica gel and a glass woolplug. The standard Greenburg-Smith impinger was thesecond of the four impingers in the train. The sampledaerosol particles were collected on th e glass nozzle, the glasssampling probe, and the impingers. Trace amo unts of C r + 6were found in the fourth impinger (dry) bu t these con centra-tions were usually lower than the instrument detectionlimit of 0.5 |xg Cr + 6 . Impinger samples, probe wash, andconnector washes were analyzed colorimetrically for Cr + 6and by atomic adsorption for total Cr. The water vaporconcentration was obtained from the amount of watercollected in the impinger silica gel and the ice cooledimpingers (typically the w ater vapor concentration rangedfrom 0.5 to 1.5 perce nt wa ter vapor).

Mark3 UW Cascade impactorThe M ark 3 University of Washington Cascade Impactorswere operated at an air sampling rate of about 1.0 to 1.5acfm for about 3 to 5 hou rs in order to obtain a sufficientlylarge volume of air (about 200 to 400 cubic ft. of air) to beable to weigh the particle samples. The M ark 3 Universityof Washington Cascade Impactors m easure particles in th erange of 0.3 to 20 microns aerodynamic diameter as re-ported by Pilat et al. (1978). n Samples were obtainedduring air sparging with no aluminum parts being pro-cessed (anodized). The two cascade impactors (operatedsimultaneously) were located with the sampling nozzlesabou t 6 to 9 inches above the acid liquid level. Th is distancewas approximately th e same as th e distance from the acidliquid level to the middle of the exhaust duct openings(openings are about 2 to 4 inches high and about 10 incheshorizo ntal). The air sparge (air injection from th e bottom ofthe tank) was then activated and operated for the durationof the tests (about 150 to 300 minutes). Six cascadeimpactor emission source tests were conducted for timeperiods of up to 330 minu tes. During the sampling, it wasobserved tha t th e droplets impacted on both th e inside andoutside of the sampling nozzle. These tests provided data onthe size distribution of aerosol particles formed by the airsparging part of the process. Stainless steel foil substrateswere used to collect the particles sampled on the cascadeimpactor collection plates. The stainless steel substrateswere prepared using ultrasonic cleaning in acetone andfollowed by rinsin g with deionized wa ter. At th e conclusionof the tests an d final weighings, the su bstrates and inside ofthe nozzles were washed and stored in sample bottles with0.IN NaOH.

FiltersOpen faced filter tests were conducted in order to obtainparticle concentrations at the inlet to the air exhaust d uctsfrom the chromic acid anodizing tank. T he 47 mm diameterfilter holders were mounted in two of the exhaust air du cts(port A in Figure 1) such that the entrance to the filterholderwasflush with th e face of th e exh aust d uct. T he filterholders were connected to a 0.5 inch diameter steel sam-pling probe. Tygon tub ingwasused to connect the samplingprobe to a modified Greenburg-Smith impinger containingsilica gel. The impingers, filled with a known amount ofsilica gel, were used for the m easu rem ent of the w ater vapor

concentration. Sampling times were 40 to 50 minuteswhich simulated the anodizing process time of40minutes.

Final weighings of the filters and silica gel in the BoeingQuality Control Lab immediately followed the tests. Theglass fiber filters were stored in a solution of 0.1N NaOHand analyzed colorimetrically and by atomic adsorption atthe U niversity of Washington Dep artme nt of Civil Engineer-ing labs.Two types of 90 mm filter tests were conducted up streamof the scrubber (port B in Figure 1). The first type requiredsampling for a period of40to 60 minutes for the collectionof particulate from a single anodizing process run. Thesecond type involved sampling for approximately 24 hoursfor the collection of the aerosol particles emitted over anentire working day. During this time multiple anodizingruns (typically 12 runs in a 24 hour time period) wereprocessed through the tank. The 90 mm diameter filterholders used a 0.5 inch diameter sam pling nozzle. The filteroutlet was connected via a 0.5 inch diameter samplingprobe and Tygon tubing(3/8 ID thick walled tubing) to thevacuum pump and then to the dry gas meter. Preparationofthe Teflon 90 mm diameter filters consisted of desiccationfor at least 24 hours . Blanks were weighed with each test.

Mark20 UWCascade ImpactorSeven tests were conducted upstream of the scrubber

(port B in Figure 1) during July, August and November1990 using a University of Washington Mark 20 CascadeImp actor. Th e Ma rk 20 cascade impacto r sampled at the 1.8to 2.4 cfm (stack conditions) air flow range for about 24hours, using a 2-stage Leybold-Heraeus vacuum pum p. TheMa rk 20 Impac tor consists of 14 jet stages plus a backup 90mm filter. The inlet to the Mark 20 cascade impactorincludes a samplin g nozzle (tha t is the first jet stage) whichleads directly to the first particle collection plate. Stages 2through 14 consist of multiple round jets. The Mark 20impactor operates at low pressure at the jet stages 10throu gh 14 and hence the air pressure downstream of jetstage 14 was monitored during th e sampling with a Wallaceand Tiernam pressure gauge (the absolute pressure rangedfrom 8-9 Hg). The mylar substrates were washed in dis-tilled water, dried, and then stored in a desiccator. Thebackup Teflon filters were desiccated for 24 hours prior toweighing. Prior to sampling, each substrate, filter, andblank were weighed and the impactor assembled in theBoeing Quality Control lab. One substrate and one filterblank were used for each test. Each b lank was stored in adesiccator. After sampling, all substrates and filters wereweighed and the samples were stored for later inductivelycoupled plasma (ICP) analysis at the U niversity of Washing-ton Department of Chemistry.

Integrating NephelometerThe integ rating nephelom eter was developed at th e Uni-versity of Washington Department of Civil Engineering inabou t 1966 and was first used in optical studies of humidityeffects on hygroscopic aerosols, as reported by Pilat andCharlson (1966).12The UW integrating nephelometer mea-sures the light scattered by air molecules and aerosolparticles, integrated over the 5 to about 175 degree angle(hence the nam e integ rating nephelometer) as reportedby Ahlquist and Charlson (1967).13 The instrument mea-sure s th e light scattering coefficient or b sc a t, most com-monly reported in units of 1/meters. The relationship ofbscat to the light tra nsm ittance is given by the equation:

In = -( bsc a t,meters~1)(L, m eters)Wh ere L is th e distance thr oug h th e volume of air for whichthe light tran smitta nce is being described or measured. Forthis research project the UW Integrating Nephelometer

March 1992 Volume 42, N o. 3 305


5/7

104

is 103

10*

1 I I I 1 1 m i

O

7

Trmr

Iy

pr?

IN

J-

1

/

III

/IJ

r/'

r

Viny

\

Y

Y

i n I I mil

o 4 7 m m F i l t e r s T a n k E x h a u s t]- o -9 0 m m F i lt e rs S c r u b b e r i n l e t ) M a r k 3 O v e r T a n k ) M a r k 2 0 S c r u b b e r In l e t )

mi I I I

: S a m p l T i m e0 . 6 7 h r0 . 6 7 h r3 - 5 h r2 4 hr

11 1

ss S s

1O10.01 0.1 1 5 10 20 304050 6070 80 90 95 99 99.9 99.99

Percent of Samples Less Than Stated ParticleMass ConcentrationFigure 4. Distribution of particle mass concentrations.

was modified in order to protect the instrument from thecorrosive acids in the aerosol emissions from the processtanks. The airwassampled at aflowrate of about to 2 cfmfrom the duct (port B in Figure 1) and was treated toremove the larger (above about 5 microns aerodynamicdiam eter) particles which were more likely to impact on th ewalls in the nephelometer. The air treatment apparatusconsisted of a UW Mark 3 Cascade Impactor inlet sectiontogether with a BCURA cyclone and a 0.375 inch diametersampling nozzle. The UW Mark 3 impactor included thesecond and third jet stages of the normally seven stageMark 3, whereas the jet stages 4 through 7 were omitted.This configuration provided for the removal of the largerparticles and prevented the larger particles from contami-nating the inside surfaces of the neph elometer. Substrate swere used on the particle collection plates downstream ofthe second and third jet stages. No filter was used at theoutlet of the impactor to allow the fine aerosol particles toflow on and into t he integrating nephelometer. An addi-tional air purge pum p and new filters were installed withinthe nephelometer. Each air purge pump provided an airflow of about 200 cc/min. to the phototube section and tothe light trap section of the ins trum ent. The purge air flowprevented th e aerosol particles from entering the se sectionsand contaminating the sensitive components. The averageair sampling flow rate through the nephelometer utilizingthe external vacuum pump was about 2 cfm. The airsampling flow rate was adjusted with a valve locatedupstre am of the external vacuum pum p. The zero and spanchecks of the nephelometer were conducted about everythree days. A zero check was performed by allowing theinstrumen t to befilledwith the pu rge air (requiring about 5minutes). A span check was performed by injecting theinstrument with a standard gas with a knownbs c a t (Freon12 which has a bs c a t of about 3.6 x 10~4 mete r s 1 atatmospheric pressure and 70F). Both the zero and spanmagn itudes were recorded as shown on the recorder ou tputsheet along with the 24 hourbs c a t data.

Particle Sample AnalysisThe particle samples were weighed using a Mettlerelectrobalance (Model AE 240) located at t he on-site Boeing

Quality Control Lab. For the cascade impactor substrates,the outlet filter was weighed first, followed by the lastcollection plate substrate (number 7 collection plate on theUW Mark 3) on up, assembling the impactor with eachstage. The 47 and 90 mm filters were also assembled afterweighing. After sampling, the reverse of the loading weigh-ings were conducted. Visual inspection of the samples wasuseful in providing a qualitative indication of the particlesamp les. The yellowish color of the chromic acid was clearlyvisible on the Mark 20 impactor substrates from the inletjet stage (cut diameter of about 25 microns) to about theeleventh stage (cut diameter of about 0.8 microns), whichqualitatively indicated that there was probably little chro-mic acid below 1 micron diameter. A colorimetric method(spectrometer at a wavelength of 420 nanometers) was usedfor hexavalent chromium concentrations of the ModifiedMethod5samples and the47mm samples. A color indicator(1,5-diphenylcarbazide) was added to a diluted portion ofeach sample. The magnitude of the light transmittancethro ugh a 1.0 cm sample curvett e cell was the n compared toa curve generated from known concentration standards toobtain the C r + 6concentration.Resul t s and Discussion

Particle ConcentrationsThe distribution of the particle m ass concentrations forthe four measurement methods used are presented inFigure 4. The particle mass concentration immediatelyadjacent to the process tank at the inlet to the exhaustducts, as measured by the 47 mm diameter open facedfilters, ranged from 86 to 4200 |xg/m 3. During December1989 to March 1990, 12 (five sets of paired simultaneoussamples and two single samples) open-faced filter sampleswere obtained. The particle concentration differences werepossibly caused by the differences in the distance betweenthe part being anodized and the sampling filter (i.e., thenearer t he part, the higher the particle concentration). Onesampling port allowed for an open-faced filter holder to belocated near the aluminum part being anodized; however,

the 2nd sampling port positioned the 2nd filter holder somedistance from any aluminum part. From this data, itappears tha t the aerosol particles are formed near the partbeing anodized. The UW Mark 3 Cascade Impactor testswere conducted d uring air sparging of the process tank butno parts were being anodized. The cascade impactors werelocated over the process tank liquid (nozzle about 6-9inches over the liquid) and at the inlet to the air exhaustducts, thus sampling the aerosol particles near their placeof generation. The total particulate concentrations fromthese tests ranged from 75 to 290 jxg/m 3 as is shown inFigure 4. Note that the particle mass concentrations sam-pled during air sparging were considerably lower tha n thos eobtained during anodizing. Measurements of the particlemass concentration in the main duct upstream of the wetscrubber were conducted with 90 mm diameter filters andwith the Mark20cascade impactor. Figure 4 shows that theparticle mass concentrations measured by these two meth-ods are in approximate agreemen t (the lower magnitude ofthe Mark 20 concentrations being due to the longer sam-pling times). The particle mass concentrations measuredisokinetically by the 90 mm diameter filters ranged fromabout 200 to 20,000 }ig/m 3 with sampling times rangingfrom 40 to 60 minutes. The Mark 20 Cascade Impactortests conducted over a 24 hour sampling time providedparticle mass concentrations in the 2,360 to 7,613 jig/m 3meter range (4,986 |xg/m 3average concentration).

Particle Size DistributionsThe size distribution results are presented in Figures 5an d6.Th e results of the M ark3Cascade Impactor measure-

6 J . Air Waste M anage . Assoc.


6/7

10 0

M a r k n

N o v e m b e r - D e c e m b e r 1 9 8 9

r ii 4

Jiii

/m]/I

pnParticle Cone.(nn/m3\- o -0

- o - V - -

M3A 131M3B 192M3C 115M3D 86M3E 75M3F 83 -

0.01 0.1 1 5 10 20 3040506070 80 90 95 99PercentMass Less Than Stated Diameter

9 9 . 9 99.99

Figure5. Size distributionofaerosol particles near process tank duringairsparging.

10 4 iii

1 0 3 d

io 1 a

10

4 - i -

i::11'

.| j L.......UJ.1JJ JJjJJ 1 1 1 |

4 7 m m F i lte r T e s t sN o v e m b e r 1 9 8 9 - F e b r u a r y1 9 9 0if.. i L.L .U. iijJ 1 i U.J..J.M] 1 : "Hi

jttti:::::::

44 .

o C r 6 C o n c e n t r a t i o n P r o c e s s C o n c e n t r a t io n :

-i i-Hit 1 i=yH::::::::::::: =:i:::::

jjji [: : | r : : | : : | : S i | : j : : : : : : : i : : : :

; J | J : | E =

illI i i-i-fri-frf

: q z pi

: : : : : : : : : : : : : : :i iri-

iULljlt ::m Mti'f i rft

H ifcjj

- H - i -

tit

I Ili

r

E |

ir

T T T

iH Ii l lit i444T T T

-ill

:iii:= |

|EE r

. . U ij-i-t- i f f

: : : : : : i : : : : t

j . . . . j _

:j::::b104

t 103.o

: i : : i 102

J= 10

10 1 2 1 0 3 1TotalParticulate Concentration(M9/m3)

Figure7 Comparison ofC r+ 6and process emission rate with particle massconcentrations.

ments conducted during air sparging over the process tankand at the inlet to the exhaust ducts at the process tankshow in Figure 5 that the particle mass median diameterranged from 0.5 to 2.5 microns aerodynamic diameter andthe mass concentration from 75 to 192 |ig/m3. The particlesize distributions measured with the Mark 20 in the mainduct upstream of the wet scrubber show in Figure 6 thatthe particle mass median diameters ranged from 2 to 4microns and the mass concentrations from 2,360 to 7,613|xg/m3.ChromiumConcentrationsi ntheAerosolParticles

The distribution of total chromium as a function ofparticle diameter is also shown in Figure 6 for the averageddata from five Mark 20 tests. The particle median diameterof the Cr distribution was 4.0 microns and over 99 percentof the chromium was in particles of diameter greater than

a 1

1

1

M a r k 2 0J u l y - A u g u s t 1 9 9 C

i

P

M

A

dr #l/l\i

m1

f{

ris i -i J TkjO A

- Concentration (/g/m3) Particles

-c-- a

-A

~ -, t

< A 7,613- B 6,411K C 4,680- D 5,482- E 2,362- F 2,360- G 3,596UU . m i

Chromi 42.C59.-44..5 O . f3 2 . 1

DOre22 :

8...I I-.I I L

0.01 0.1 1 5 10 20 3040506070 80 90 95 99 99.9 99.99PercentMass or Cr Less Than Stated Diameter

Figure6. Distribution of mass and chromium in aerosol particlesupstreamof scrubber.

1.0 microns. Note that the smaller aerosol particles lessthan about 1 micron did not contain much chromium 0.5percent of Cr in particles less than 0.8 microns diametercompared to about 10 percent of the particle mass inparticles less than 0.8 microns diameter). Concentrations ofhexavalent chromium measured with 47 mm diameteropen-faced filters in the vent ducts adjacent to the chromicacid anodizing process tank ranged from about 1 to 1300|i,g/m3 of air sampled as is shown in Figure 7. The C r+6concentration was related to the particle mass concentra-tion shown on the abscissa in Figure 7) and the concentra-tion ratio of Cr+6 /mass ranged from 0.15 to 0.24. Figure 7also shows the process concentration mg Cr+6/amp-hour)as a function of total particulate concentration |xg/m3).Process concentrations ranged from 0.018 to 65 mg Cr+6 /amp-hour and the total particle mass concentration was inthe 190 to 4,000 |xg/m3 range. The distribution of Cr+6process concentrations mg Cr+6/amp-hour) measured fromthe Modified Method 5 train are shown in Figure 8. Theconcentrations ranged from 2.5 to 58 mg Cr+/amp-hourwith a median value of about 18 mg Cr+6/amp-hour.

10

1

lodii ie d N lethn d 5

/> f/

/ 4Y

July - September 1989[

0.01 0.1 1 5 10 20 3040506070 80 90 95 99 99.9 99.99Percentof Samples Less Than Stated

ProcessConcentrationFigure8. Distribution of Cr +6process concentrations.

March1992 Volume 4 2 , N o . 3 3 0 7


7/7

0.0003

0.0002

Figure 9. Light scattering coefficient of aerosols in main duct.

Nephelometer Test ResultsThe integrating nephelometer was installed to sample airfrom the main duct upstream of the wet scrubber on July26,19 90. The instru me nt was able to show the variation inthe aerosol particle concentration on a continuous timebasis and reportedbs c a tvalues in the 3 x 10~5to 3 x 10~4m e t e r s 1 range. Figure 9 presents an illustration of thechart recorder trace of bs c a t versus time from midnightSeptember 8 to midnight September 9, 1990 (Saturdaymidnight to Sunday midnight) and shows tha t 13 loads ofparts were anodized during this time period. The lightscatte ring coefficient w hile no part was being anodized wasabout1x 10~4m e t e r s 1and increased rapidly to about 2 x10~ 4 mete r s 1 when the anodizing process was initiated.The bs c a t peak width is about 40 minutes, which corre-sponds exactly with the part anodizing time. This datashows that a continuous monitor such as the integratingnephelometer can be used to indicate the aerosol emissionconcentration for the chromic acid anodizing process ex-haust gases.

ConclusionsThe conclusions of this research project to characterizethe particle emissions from a chromic acid anodizing tankare as follows:

1. The aerosol ma ss conc entration ranged from abou t 40to 20,000 u.g/m 3with the higher concentrations occur-ring during th e anodizing process.2. The aerosol particle size distribu tions were in th e 0.05to 24 micron aerodynamic diameter range. The massmedian diameters were in the2to 4 micron range in t heduct and in the 0.6 to 4 micron range near the processtank.3. The mass median diameter of the total chromiumconcentration of the particles was 4 microns and over99 percent of the chromium was in particles of greatertha n 1.0 microns diameter.4. The hexavalent chrome concentrations were in the 20to 1,500 |Ag/m3 range and approximately correlatedwith the total particle mass concentration.5. The integrating nephelometer was able to detect theaerosol mass concentration increase when the anodiz-ing process was initiated and d emon strated its capabil-

ity for operating as a continuous aerosol emissionmonitor.

6. Measured process conc entration s ranged from 2.5 to 58mg Cr+ 6 /amp-hour with the Modified Method 5 trainand 0.7 to 65 mg Cr + 6 /amp-hour with 47 mm filtertests.Acknowledgments

The autho rs would like to acknowledge the Boeing Com-pany for making this research possible. We greatly appreci-ate the assistance of Hal Alsid, Mary Jane McGarity, DavidMummey, Herb Gasgill, Rich Clasen, Andy Gay, RandyLove, and Kirk Thompson of the Boeing Company. Specialthan ks to the plumb ers stationed in the17-06building (JeffBement, Curt Kinlock, and Kelly Meeds) and also to theprocess line operator, Jimmy Cook. The assistance of JacekAnuszewski, GailDorf, William Dunn, Stacia Dugan, andRon Sletten of the University of Washington D epartme nt ofCivil Engineering with the source testing and chromiumanalysis was very helpful.References

1. Chromic Acid Plating and Anodizing, Reg. Ill of the Pu getSound Air Pollutio n Con trol Agency, Section 3.01.2. Wernick, S.; Pinner, R.The Surface Treatment and Finishingof Aluminum and Its Alloys, 3rd Ed, Robert Draper, Ltd.,Teddington, England, pp. 349,1964.3. Bivens, D.; DeWees,W. Method Development and Testing forMeasurement of Source Levels of Hexavalent and TotalChrom ium, Paper presented at 80th Annual Air PollutionControl Association Mee ting, New York City, NY (June 1987).4. Pacific Environm ental Services, Inc. Efficiency of Ha rshawChemical's MSP-ST for Controlling Chrome Emissions Froma Chromic Acid Anodizing Tank , Conducted for Metal Finish-ing Assoc. of Sou the rn C alifornia, M arch 1989.5. Bhargava, O.; Bumsted, H .; Grunder, F.; Huni, B.; Manning,G.; Riemann, R:j Samuels, J.; Tatone, V.; Waldschmidt, V.;Jeff, S.; Hernande z, P. Study of an analytical method forhexavalent chromium, Am. Ind. Hyg. Assoc. J.4:433 (1983).6. Abell, M.; Carlbe rg, J. A simple reliable method for th edetermination of airborne hexavalent chromium, Am. Ind.Hyg Assoc. J. 35 : 229 (1974).7. Thomsen, E.; Stern, R. A simple analytical technique for thedetermination of hexavalent chromium in welding fumes andother complex matrices, Scand J. Work Environ. Health5 :386 (1979).8. K oropchak, J.; Roychowdhury, S. Evidence for aerosol ionicr ed is t r ib u t io n w i th in ae ro so ls p ro d u ced b y ch ro meelectroplating, Environ. Sci. Technol.24; 1861 (1990).9. Resch, F. J. Liquid Aerosol Formation By Air BubbleBursting, in Aerosols: Formation and Reactivity, Proceed-ings of the 2nd International Aerosol Conference, Berlin,Pergamon Press, pp. 79-82,1986.10. Determ ination of Hexavalent Chromium Emissions fromDecorative and H ard Chrome Electroplating and Chromic AcidAnodizing, Pug et Sound Air Pollutio n Control Agency SourceTest Method, August 9,1990.11 . Pilat, M.; Raemhild, G.; Powell, E.; Fioretti, G.; Meyer, D. Development of a Cascade Impactor for Sampling 0.02 to 20Micron Diameter Pa rticles, EPR I Report No. FP-844, Vol. 1.(1978).12. Pilat, M.; Charlson, R. Theoretical and optical studies ofhumidity effects on the size distribution of a hygroscopicaerosol, J. echerchesAtmospheriques2:165 (1966).13. A hlquist, N.; Charlson, R. A new instrume nt for evaluatingthe visual quality of air, JAPCA 17: 467 (1967).14. U.S. EPA Locating and Estimating Air Emissions FromSources of Chrom ium, EPA-450/4-84-007g, Research Trian-gle Park, NC, 1984.15. U.S. EPA Locating and Estima ting Air Emissions FromSources of Chromium (Supple ment), Research Triangle Park,NC, 1989.16. California Air Resources Board Chrome Pla ting ControlDem onstration Project Staff Report. Sacramento, CA, 1989.

Dr. Pilat is a professor in the Department of CivilEngineering, University of Washington, Seattle,WA98195.R. Pegnam has a BSCE and is a graduate student workingtowards the MSCE degree in the Department of CivilEngineering, University of Washington. This paper wassubmitted for peer review on Janu ary 18,1991.The revisedmanusc ript was received on November 12,1991.

308 J .Air Waste Ma nage. Assoc.

airborne particulate emissions from a chromic acid

Documents