ML20153C497

Mine Safety Appliances GMR-1 Canister for Use Against Radio Iodine & Organic Iodides

ML20153C497
Person / Time
Site:	Vogtle
Issue date:	04/25/1984
From:	Mckee E MINE SAFETY APPLIANCES CO.
To:
Shared Package
ML20153C307	List: ML20153C305 ML20153C497
References
NUDOCS 8809010163
Download: ML20153C497 (21)

Text

,

.a ENCLOSURE 2 TO VL-41 THE HSA GMR-! CANISTER FOR USE AGAINST RADIO ICOINE Afl0 CRGANIC lC0!CES f

i

\\

1 l

Note:

Presented by Dr. E. 5. McKee, Mine Safety Acoltances Company, Pittsburgh, Pennsylvania for Alabama Power Company to Nuclear Regulatory Commission staff on April 25, 1984 at Beth6sda, Maryland 1

i 1

l

-e 88090101 $3 000826 PDR ADOCK 05000424 P

PNU s

...,.,,--.-n..,

_,--_. ~, - _,,. -_,___.- - _ -. -.--,,

TEST CanolTIous Canllenge Conc.:

5 - 10 ppe Cil 1 Ihaddity: 60 + 35 and 90 + 31 3(minianas of six cans at eacE humidtty) -

Temperature:

1108F Cyclic Flou:

192 ti40 for 0.82 sec.; O LPf6 for 1.64 sec.,

repeating tiils cycle throughout the test.

This gives a minute volume of 64 L.

Breakthrough Conc.:

11 of the challenge concentration t

e r

9 i

e

_t_

e i

+

5

.=.

Table 1.

. Results 601 Cl

.o #

Mfg. Date Service Time Comument

- Can i

• I**

b'S-4 Mf. Date Service Time

- Comment 9

5*

11/30/83 720 12 0.251**

min.

krs.

6*

37 4/14/83 1410 23.5 0.07 "

7*

38 1/9/84 1896 31.5 0.33 "

29 2/2/84 2160 36.0 39 4/14/83 10110 18.0 i

10 2520 42.0 40 1/9/84 2220 37.0 11 2670 44.5 57 3/28/84 2490 41.5 l

12 4/14/83 1200 20.0 58 i

13 2280 38.0 1500 25.0 59 I

14 2610 43.5 1410 23.5 60 15 2/2/84 1680 28.0 61 2460 41.0 2250 37.5 16 4/14/83 1530 25.5 62 2460 41.0 4

90% Rif 3*

11/30/83 1215 20.3 0.30 "

4*

26 2/2/84 1560 26.0 1215 20.3 0.15 "

j 1t*

10/21/83 990 16.5 0.45 "

2070 34.5 27 l

9*

0.25 "

! 10*

28

~

1

0. 4 3 * * -

2220 37.0 41 4/l&/83 1230 20.5 I 11*

11/30/83

. 720 12 0.67 **

1320 22.0 42 i 12*

10/21/83 43 0.04 "

1650 27.5

. 13*

11/30/83 44 8.47 "

'1320 22.0 1 14*

1/9/84 795 13.3 ~

0.83 "

1500 25.0 45 15*

46 3

0.34 **

1260 21.0

! 16*

47 Test Invalid i

0.35 a j 17 1/9/84 1890 31.5 ConsLTFlow 49 48 1350 22.5

iff 3180 53'.0 1290 21.5 i

50*

? 19 840 14.0 0.62 "

2530 42.2 in 9/13/83 2390 39.3 51 3/28/84 1650 27.5 I*

52 1800 30.0 1530 25.5 0.44 "

53 1620 27.0 1 *2*

2280 38.0 0.09 "

54

! ~3 10/21/83 2490 41.5 1530 25.5 i.4 55 y

1740 29.0

=

f5 2910 48.5 56 2490 41.5 1520 27.0 i

lesL stopped before 1% breakthrough.

"I.cakgle when les t 5 topped.

1

\\

Tahia.

SLctistical Analysis cf Lct 4/14/83

~

X (1 lut)

Y (Svc. Time)

Log X Log Y 60 1200 min.

1.77815 3.07918 Ave. Y60 = 1355 min. (22.6 hrs.)

~

60 1500 3.1760n 60 14 10 3.14922 Ave. Y90 = 1365 min. (22.7 hrs.)

60 1530 3.18469' 60 1410 3.14922 60 ISM 3.03342 90 1650 1.95424 3.21748 90 1230 3.08991 1

90 1320 3.12057 i

90 1500 3.17609 90 1260 3.I8037 i

90 1350 3.I3033 90 1290 3.11059 90 1320 3.12057 991 Prediction Interval for tog Y. Given tog X = 2 (1001 Ril) 991 Interval = Logy *(t 1.99/2),,7 S

(Equivalent to the ra==an expression of X 135).

gg y Where tog Y = bo t bi Log X and bo = 3.08231. by = 0.02606, where b is the intercept and b.

the slope of the plot of Log Y vs. Log X.

o tog Y = 3.13443 (1362-min., 22.7 hrs.), when tog X - 2 or X a 100 S2{g,1/n*(Log _I 5 tog Y =

2]=.05543 E

L 991 Interval of Log Y when Lo9 X = 2 or X = 100 = 3.134431 (3.055)

(.05543) = 3.13443 *.16934

= 3.30377 to 2.%509

(

Y = 33.5 krs to 15.4 hrs.

\\

l

)

i.

! I i

6 E M ~GMR-I_, _ _ _ _ _ _ _ _ - - - - -

~ ~ ~ ~ ' ~ ~ i. T -.

'~

a-CAIIS LST 4/14/83 og Relativa numielity Linear A.H.

- (so & Sol llo*F Cyclic flow 192 LFM witti Log Servica L Ifo for o.82 Sec: o LPM I.64 Sec.

'~

Log y

4

~

(

)

Y-(X i)-(Y V)

(y-y)2 ~

Col Muse.)

1.ife (Min.)

6e 1200 1 77815 3.07318

.10062

.o1012

.05209

+.00524 3.12865

.0024g 60 1500 a

3.1760s

+.o4482

.00451

.00225 60 1410 3.14322

+.01735

.00181 a

.00042 6e 1530 3.18 % S

+.05342

.00538

.00314 t

a 60 1410 3.14922

+.01795

.00181

.00o42 60 late 3.03342

.09785

+.00985

.00907 30 1650 1.95424 3.21748

+.o7547

.0057a

+.0862:

+.00651 3.13324

.007:o 30 1230 3.08331

.04:36

.00312

.00188 So 1320 3.12057

.0I070

.00081

.00016 90 1500 3.37609

+.e4482

+.00338

.00184 a

90 1260 3.10037

.03090

.00233

.00:08

~

r 90 1350 3.13033 l

.00094

.00007

.00001 a

So 1290

' 3.11059

.o2068

.00156

.00058 a

So 1320 3.12057

.01070

.00001

.00086 881 Y - 1.87877 Y - 3.13427 I

.30623 I-+.00277 I

.03049 I

I s

. I

Page 2 cf 2

~

2 (X- ) (Y-Y)

+.o0277 n

b Y - b, + bg X

i

• a.02&o' f_(X-Y)2 i

.io623 Y - b Y - 3.13:27

.o26o6 (i.sys77) - +3.os23:

b i

o f(Y-)/sa.-2 s

.ojens/s2

.oo254 t

s, 2

.oo2544.lo623

.o2331

@ # *~D 2

st ff,2 l

s,i j.o23si ensul 4

4

]

392 coo r. u.i n s o slope G,

b, t((i.ss/,)l= -2 s,I

.o26e6

• 3.055 (.1541) a

+.4ss4o.

.4u2s j

t

)

sst Prediction intervals for Logy (Life) Clwen tog X - 2 7

• c ((s.ss/ )] at -2 sy ; i - b,+ b,, - 3.os231

+.o2606(2)

[

3 g,, if,, g, _ y)2]

.05541

[

2 y-3.13443 (22.7 nrs);

s

-f s y

g r2 l

5 3.134's3 2 3.055(.05543) - 3.3o377 (u);

2.36503 (L) -

e

n 10 90 G:

. _ - =.. _..

t Y

m_

7

=. -.

Th;oratica1 alope

- gy.ghg ?f-

7. "._g 1_--l--

of tervice life ata

.=

w"7.3..!.L.

..*T l' m M ; :.I.." -r m :- m ~

l..6',!"

14..Il

.l-i 90%

RH = 22.7 hrs.r _

--==_.C:

e and cervice time gh.

M B=.-.iW El@ M W.Q ih toi4lj # Y 1."

at 100% RH =

=_ =. F.h.._.,_ s.,.;:.1 0 0 '.,F... i : 0,: hrs. '

r 8.0 hrs.

..... t..

4 i, -

i e

e l..

.i

,,L 6

l.

,e A_

3.n. : :.:

.- --.. 4. {2.: :.:... -

.i

.c e

..... - - -=_.

.m-

k.. -..n

-:: =.....:

m.

1

==

. - -. =.

1 :.:

5 - =..= - :.- = :. : n ::.......v.I~

.v. ~ t.-

....... ~......,..

t

......t_.. _... _..,.....

..:,,,. e.. o f v 1 g hgMeq ( f'y {

.g

.s

..a.

g u

ee4n.3-g

..e j.,.

t

==

p i..

j.

.p i

i s-.

..l

. i

..j

.i

.)

s g

m l....

l l

101:

g l.

CJ

...O. '.

• y i

l' ?

l' t

.l 9

g

e.........

3 e

..i.....

6 m.......,

. l,.......1..

.y, g

g A

l-

.....i......M......

.. o.

p..

=,

C.

7:

........ l.

Oe g'

t

% g

.1 9

g.

t i

I i

4 3

y e.

e t

I t

e

.l I'

6 I

f I

~

l

.l e

4 i

l i,e I

6 w

p..

,1

... +..

, n g,

l f

n t

-"..**.4*114.. 8 :-.

-"J...

v-*

_. :.. 2 : :..-

.~

{.

.u-m j

. ;. :. ].-* 1-l

. - '.... j *-

4.J -

. -*.."..: -. t. -

4.. '.. * : ~. '

.j.

[,

l

-.t

.4 m

.......,..~.l...

.).

-.o...

g 4

~..

\\

..l e

_m.g........

a...

b i

..L.

9..

.. -.J

.l

...t.

.i i

i...

..p...

..l.....

..g...

...6

... j.... e...

.......{.

j

.,..~..-

. g. -

y

REQUIREMENTS FCR NIOSH APPROVAL FOR AN ORGANIC VAPOR CHIN CANISTER PER 30 CFR 11 Int _ Condi tiens Challenge conc. 5000 ppm CCl 4 Test Hunidity 5015% RH Test Temperature 2512.56 C Flew 64 LPM for as received canisters 32 LPM for equilibrated canisters Breakthrough conc. 5 ppm Ecu111bration Conditiens j

3 Canisters as received.

2 Canisters equillbrated for 6 hrs., 64 LPM, 255 RH, Room Temp.

2 Canisters equilibrated for 6 hrs., 64 LPM, 85'. RM, Room Temp.

Total 7 canistars.

Service Time Recuirement 12 minutes.

No statistical requirements.

If all seven canisters have service times of' 12 minutes or more, the canister is approved.

e

\\

\\

j

\\

\\

\\

EXA.'tPLE OF LOT EVALUATION PIR MIL,-STD-414 SIMGLE $?ICITICATION LIMIT - TOP.M 1 i

VARIABILITY UNC0W - STNi3AC CEVIATION 'tEIHOD (PJT. PACE 37)

LEVEL II AQL = 1.0%

SPEC. LIMIT 1.0%

LOT SIZE - 500 CANS SAJ2LE SIZE (TA3LE A. 3-1) = 7 (n) 3 i

TIIT KISULTSs S HOUR 3REAKTRROUGH CAN #

_ CONCENTRATION (t) 41

.036 42

.028 4

43

.019 44

.064 45

.027 46

.035

~~

49

.170 t

SAMPLE MEAN =.0612P (2)

PATIMATE OF LOT STANDARD DEVIATION 9.05354 (s) l

. THE QUANTITT (U-5)/s a 1.00

.06129, g7,33 I

.05356 1

ACCEPTA34.ITY CONSTANT (k) = t.62 (TA3LE l-L)

LOT.

ACCEPTABILITY CRI IK:!" SINCE U-i/s> k

, $'r l

..N

~

i PROPOSED LOT AECE?IANCE PLAN 4

4.1.1 MIL-STD 414,tevel II, AQt. 1% would be used to (i) select the payer number of cans to test, depending on lot size, and (2) to interpret the results regarding lot acceptance or failure.

,4 >*

g-4.1.2

• The cans umuld be tested under the conditions of section 1; however, all tests would be conducted at 905 20.

Tests would be stopped at eight hours and the procent leakage recorded at this time. From evidence presented in the preceding sections. results at $sg are not significantly different from those at 1001.

~'

~.

4.1.3 The percent Yeakage values would te conpared to the spec. Ilmit of 1.01 using the s!.31c spec. Ils!t. variablistpneun, standard deviation method of MIL-STO 414. Acceptance soaid be based on this analysis.

LOT $IZE TEST SAMPLE 300-500 7

501-800 10

~

801-1,300 15 e

1.301-3,800 20

~ ~ ~

3,200-8,000 25

~ 4

- - -... ~

~

g j

SAFETY FEATURES-EUILT IllTO THE PLAal l

1.

Flow Rate:

64 LI9t ----- a person could not.possibly breath at this rate for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.

P least twice the average rate.

i 2.

8 deurs Service Tlas ------

\\

this is probably double the actual use time required.

i 3.

Actual Service Times ------ minimum of 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> ----- 2-1/2 times the required l

a bears.

i i

~

t

==

Conclusion:==

l niculd need a catastrophic failure for a can to not give proper protection -----

up such a failure.lio destructive test samp11ag plan will pick

)

i l

1

~

{

1 i

l 3

l' I

l r

- 10,-

l

'l a

n m

PtANNED fHIURE WORK The following parameters will be further investigated to give additi onal styport to the for y conclusions and proposals, and to develop a better lot acceptance plan 1.

challen9e concentration:

1. 10, 100. 250, 500 ppe 2.

Nt/T og Ni as/1 5

15 25 34 43 4.5 66 35 19.5 12 7.5 l

9 70 39 24 15 18 79 49 30 36 54 97 60 90 Numbers in the table are the

~

relative humidity percentages torresponding to the absolute humidi ty/ tempera ture condi tions.

3.

Rate of Flow: 16, 32, 64 LPM 4.

Cyclic vs. Constant flow.

^

i 11 -

r

(

i

~

4 Sl9 MARY i

1.

Data supports approval of the GMR-I can for its intended use.

I!

~

2.

The p6 acceptance plan will assure quality of future lots.

3.

Further leers'will be done to:

3.1.

Seppert the conclusions drawn in 1 and 2.

i 3.2.

W th lot acceptance plan by:

~ '

1 3.2."1 Wing the time rpquired for testing and running the canisters to a 11 breakthrough service time.

l 3.2.2 Simplifying the test proceddre.

1

. a I

{*

]

i y

j-O q

e i

r 12 -

3

@ '?

L s e s

A;

5.

=-

=

x k%..X v y_ n y

,.,m z m; l

T J

O @-

@ g

~

I. FINE DUST FILTER

5. SPRING

~

2. COARSE DUST FILTER

6. HI EFF FILTER ASSY.

3. SORBENT

7. BOTTOM PLATE

4. SEPARATOR

8. INHALATION CHECK VALVE GMR-I CANISTER m

m e

s

\\

\\

s l

Tatc1 Radictica Dostge--

l Comparison of Data Showing Time Involved and Workers Wearing Air-Line Respirators or SCBAs vs. CHR-I Respirators Corresponding No. o f Persons Task Time Total Dose Dose Rate Required to Work Hours (Total Task (MR/hr)

Perform Task Required to Time x Dose for Each Wearing:

Perform Task Rate)

Task with Workers with Workers Wearing:

Wearing:

A-LR A-LR A-LR or WR-I or WR-I or QiR-!

SCBA Can SCBA Can SCRA Can Pressure 26 6

6 7

5 182 130 Safety Valve Testing containment 11 6

6 57 37 627 407 Sump Work RCS Seal 55 12 9

180 133 9,900 7.425 Inspect &

Replace Reactor 22 18 18 314 235 6,908 5.170 cavity

,Decon RER Check 37 5

4 52 39 1,924 1,443 Valve l

Repair i

Accumulator 45 8

6 77 58 3,465 2,610 i

check Valve Repair RNR Heat 59 13 to i44 100 8.496 5,900 Exchanger Casket Replacement Spent Fuel 12 4

4 11 16 252 192 Pool Transfer Canal Work Containment 75 9

6 7.7 12 2,025 900 Entry at 100% Pot er Incore 17 16

!2 90 68 1,530 1,156 Thimble Cleaning l

Totals 97 81 969 705 35,309 25,333

~

Mine Safety Apol'ances Company

• 6C0 Penn Center Soulevard. Pittsourgh, Pennsvivania t!2:5 sit,an seco April 13, 1984 412-273-5140 i

PERFORMANCE DATA FOR MSA'S GM-I CANISTER SUBMITTED TO NRC In accordance with our egreement, the following reort is submitted for ycur approval.

1.

General g

It was agreed with Comoany on March 8,'1984, that MSA would test GMR-! cans to comoiegion in order to be able to statistic-ally project performance at 110 F and 100% RH.

In addition, other tests had been run prior to the March 8th agreement and the data are shewn in Table !.

The tests were conducted under the following con-ditions:

Challenge Conc.: 5 - 10 com CH Humidi ty:

60 + 31 and so r 3u 3 inimum of six cans at eacn humidity)

Temnera ture! [110*fl i

ICyclic Flew:1 192 LPM for 0.82 sec.; O LPM for 1.64 sec.,

repeating this evele throuchout the test.

This Breakthrough Conc.:gives alminute volume of 64 L.1 It of the cnarienge concentration 2.

Test Results Ouring tNis program, 48 GMR-! cans have been tested (47 valid tests).

These cans came from six production lots made over the period Acril 14, 1983, to February 2,1984.

Sixteen cans were tested from lot April 14, 1983,10 at 90% RH and 6 at 60; RH.

Only eight results at 90% RH were used in the statistical analysis given below, as one test was invalid (No. 47) and another was stopped befort completion.

Only a few cans t

4 were available from the other lots, so they could not be statistically analyzedt however, all cans run to ecmpletion had a service time of i

i

}*'

__.____,,---.___.___.._._-,-..____.__.,_____.,__,,_--m__.

l j

I 1

)

20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> or greater.

The cresults are shewn in Table 1.

The original 14 cans En run to ccmolation had service times well in excess of 12

{

hours - much in excess of the eight hours desired.

3.

Statistical Analysis of Lot 4/14/83.

Table 2 shows the data used and the statistical analysis to give the 99% prediction interval for individual values of Log Y (log service time), when X (relative l

humidity) is 1005.

The lower limit of this interval is calculated i

to be 15.8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.

This predicts that over 99% of the individual GMR-! can service times would. be greater than.15.8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> at 100% RH and the other test parameters used in this program.

This gives a i

considerable safety margin over the eight hours desired.

i One other interesting point to note from the data in Table 2, as i

well as all of the test data on the GMR-! cans, is that humidity has little or no effect en the service time over the humidity range studied, 60 to 901.

close to those at 90% on a log service time--log RH plot, u I

j slope were extremely steep--which is not the case.

t 4

Proposed Acceptance Plan.

The extremely long service times experienced j

in this program for the GMR-1 cans run to comoletion, an average of over 29 hours3.356481e-4 days <br />0.00806 hours <br />4.794974e-5 weeks <br />1.10345e-5 months <br />, makes testing to completion for routine lot accootance i

impractical; therefore, the following clan is proposed.

4.1 Interim Plan.

On an interim basis, until more data can be 4

gathered as explained in section 4.2, the proposed lot accootance

)

j would be as follows:

1 4.1.1 MIL-STO 414, Level II. AQt.15 would be used to (1) select Jl the proper number of cans to test, depending on lot size' and o interpret the results regarding lot acces'.ance 4.1.2 The cans would be tested under the conditions of section 1; i

however, all tests would be conducted at 90% RH.

Tests would be stoceed at eight hours and the percent leakage i

recorded at snis time.

From evidence' presented in the preceding sections, results at 901 are not significant1v j

different from those at 1005.

i 4.1.3 The percent leakage values would be comoared to the soec.

limit of 1.04, using the single spec. Ilmit, variables unknown, standard deviation method of MIL STO 414.

Acceptance would be based on this analysis.

)-

4.2 Future.

Because the tests in section 4.1 are very time consuming i

and somewhat difficult to run for regular cuality assurance lot acceptance testing, we plan to do further testing on the GMR-!

can in an attempt to reduce the time required for testing and also to simplify the test.

Parameters that will be investigated are:

l

l 4.2.)

Increasing the challenge concentration of CH,I in an effort to reduce the time to test.

Under current cdnditions, a test to completion might run 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> t we would like to reduce this to about two hours.

If there were a simple, straight-line relationship between service time to a It breakthrough and challenge concentration. It would indi-cate that a challenge concentration of accroximately 200 com would be required to do this. We wish to firmly establish the service time---challenge concentration re-lationship over a range of challenge concentrations from 1 ppm to 500 ppm.

4.2.2 Constant Flow vs. Cyclic Flow.

Constant flow tests are much simpler to conduct than cyclic flow tests.

From some preliminary information. it apoears that constant flow gives similar service times as cyclic flow.

If, by further tests this can be verified. constant flow would be used in lot acceptance _ tests.

4.2.3 Temperature and Humidity Effects.

Further tests will be run to study the effects of temperature and humidity on j

the perfonnance of the GMR-! can.

It to test cans for lot acceptance at 25,would be preferable C and 85% RH (standard N!0SH conditions t

tt can be proven that these conditions are as severe as),43{C and 90% RH. or if a g correlation between these two conditions can be established.

5.

Conclusion.

5.1 Forty-seven GMR-! cans have been validly tested undsr the conditions specified in section 1.

All of these cans had cervice times well in excess of 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

This compares to a desired service time of eight hours.

5.2 There were 14 valid tests run on lot 4/ 4/83.

of this data, projected to 100% RH.110}F. Indict.te that over 995 Statl of the GMR-! cans in this lot have service times well over eight hours (15.8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />).

Incidentally, from the data af Table 1. this lot appears to have the shortest average service time of the lots testAd.

5.3 In licht of sections 5.1 and 5.2. lthe GMR-I can should be consicerecl laua' ified lo oive service timen over eicht hourst under the con-i ditions:

% breakghrougn. cyc' ic flow (peak 192 LPM. average 64 LPM). 110,F (43 C) and 100 RH.

0 5.4 Lot Acceptance will be determined by using MIL STD-414. Lsvel II.

AQL 1%.

The percent leakage at eight hours service time will be comoared to the scoc. limit of 1.05, using the single scoc

limit, variables unknown, standard deviation method of MIL-STO-414, i

5.5 Further tests will be run studying the effects of challenge ccncentration, constant flow rate, temperature and humidity on the service time of GMR-1 cans.

This program is intended to shorten the required test time and. simplify the test crocedure.

5.6 Frca data in this investigation, it appears that relative humioity between 60 to 905 has little effect on service time of the GMR-1 canister.

plot, suggests that the service times at 90% and 1001 R not significantly different.

If you have any further 4t estions, please do not hesitate to contact me Very t,ruly yours,

(1 Wayde Miller, Jr.

Diret: tor of Product & Sales planning

/jw j

Attachments / Table I and 2 e

I O

s

__,y

-,---+w^

d Tablo 1.

Service Time of GR-I Canisters Test Condittoas: As given in section I

^

601 RH Can 8 Nfg. Date Serviceim7 Comment Can #

Mfg. Date Service line Comment stia.

hrs.

min.

hrs.

Leak 9 12 hrs.%

5 11/30/83

>>720

>>12 0.25 g

34

14/83 1410 23.5 6

0.07 gg 35 2/84 1680 28.0 7

0.33 g 0; 36 14/83 1530 25.5 29 2/2/84 2160 36.0 r

37 30 1410 23.5 2520 42.0 38

,/84 1890 31.5 31 2670 44.5 39 4/14/83 1980 18.0 32 4/14/83 1200 20.0 40 1/9/84 2220 37.0 l

33 1500 25.0 90% Ril

• Leak' age 3

11/30/83

>>I215 >>20.3 0.30 23 10/21/83 2490 41.5 2

4

>> C15 >>20.3 0.15 24 2910 48.5

~

8 10/21/83

>> 990 >>16.5-0.45 25 2490 41.5 9

0.25 26 2/2/84 1560 26.0 10 0.43 E

27 2070 34.5

)

Leak 9 12 hrs.3 11 11/30/83

>>720

>>l2 0.67^

28 2220 37.0

~

}

12 10/21/83 0.04 41 4/14/83 1230 20.5 13 11/30/83 0.47 e

42 1320 22.0 l

14 1/9/84

>>795 >>13.3 0.64 j

43 1650 27.5 i

15 0.34 44 1320 22.0 i

16 0.75 45 1500 25.6 l

17 1/9/84 1890 31.5 Const. flom, 46 1260 21.0 18 3180 53.0 47 Test invalid 19 2530 42.2

., 48 1350 22.5

}

20 9/13/83 2390_ 39.8 49 1290 21.5 i

21 1530* 25.5*

• Test Stopped 50 840*

il4.0*

Test Stopped i

22 22110 *i38.0*

alo dreakthrouch Ne 9re4 throush I

h

...... =

l

c Table 2 l

~

Statistical ~ Analysis of Lot 4/14/83

)

X (1 RH)

Y (Sve. Time)

~

Log X Log Y j

}

60 1200 min.

1.77815 3.07918 Ave. Y60 = 13. min. (22.6 kn.

i 60 1500 3.17669-aie 1410

"~

3.14922 Ave. Y90 = 1365 min. (22.7 hrs.

W 1530 3.18469 60 1410 3.14922 60 1000 i

3.03342 90 1650 1.95424 3.21748

)

90 1230 3.00931 90 -

1320 3.12C57 90 1500 3.17609 90 1260 3.18037 90 1350 3.13033 90 1290 3.11059 1

90 1320 3.12057 s'

i 991 Prediction Interval for tag Y given Log X = 2 (loot MI) 99% Interval-Y+t[1.99/2)n-2 N l

A 10nere Y = bo + b X and bo = 3.08231, by = 0.02606 y

j

=.3.13443(1362 min.,22.7 hrs.)

Sp=[S [1+1/n+(I-X)2]=05543 I

E S

• '( ~

E g,2 n-2

r. = K-K 991 Interval = 3.13443 + (3.055)

(.05543) = 3.13443 +.16934

= 3.30377 to 2.96509 ff = 33.5 krs. to 15.4 hrs.1 I

Revision 4/26/84 i

Attachmenc Two NUREQ/C3 3403 O

LA.9427.P R Prestess Report RH Criteria and Test Methods for Certifying Air-Purifying Respirator Cartridges and Canisters Against Radiciodine October 1,1978-Septe:nber 30,1982 Gerry O. Wood i

Frank O. Valdez Vincent Gutschic'4 Manuscriot suomitted: June 1943 Date Duchsned: August 1983 Preoaroe for Occusanonas Aacianon protecnon Stanen Civision of Wacdity Coeranues Omco of Nwc: ear s gwiatory Researca e

US Nucear Requiatory Cornemisa.on Wasm.agton. CC 20555 N AC FIN No A7041

'O-0@

d W

O e@ Los Alamos,New Mexico 87 Los Alamos NationalLaboratory G

e

& \\2.0400 0

40fP'

t

.t

\\Q CONTESTS A B STRA CT............................................. I

!. I b7R O D U CTI O N...................................... 2

11. eLEMEb7AL IODINE GENERATION r.ND ADSORPTION ON ACTIVATED CHARCOAL 2

A. O bj ect iv e s......................................... 2 B. Generation 3

C. Retention 3

111. RADIO!ODINE STUDIES-IXPERIMENTAL 4

A. Flow Sysvtms 4

B. Generation Methods 4

C. Detection Methods 6

D. Reagents 6

0, E. TesiBeds 7

IV. RADIOlODINE STUDIES-RESULTS AND CONCLUSIONS 7

A. Compatisens of Yapor Species 7

B. Methyllodide Versus Methyl Radiciodide 9

C. EiTects of Bed Depth and Contact Time

........................11 D. Effects of Cha!!enge Concentrations

..........................12 E. Cartridge Comparisons.................................

14 V. EFFECTS OF USE C ONDITIONS............................

15 A. Relative Humidity

.................................15 B. Temperature

......................................18 C. Flowrste

........................................21

.I D. Reproducibilities of Senice Liff Measurtments

....................21 l

VI. EFFECTS OF CYCLIC FLOW BREATHING PATTERNS

..............22 A.

Background.......................................22 B. Computer Modeling Study

...............................22 C. Experimentaj Study...................................26 D. Conclusions.

.....................................25 i

i b

I

i

)

l

\\

c

/

l t

V11. DESORPTION OF TEDA FROM IMPREGNATED CH ARCOALS.......... 29 A. B a c k g round....................................... 2 9 B. Apparatus and Procedures

...............................29 C. Results and Conclusions

...............................30 VI!!. Ten Apparatus Development

...............................32 IX. DEVELOPMENT OF APPROVAL CRITERIA FOR RADIOf0 DINE CANISTERS 33 A. History

.........................................33 B. Cunent Recommendations

...............................35 X. ASSISTANCE TO NIOSH IN ESTABLISHING A TESTING AND CERTIFICATION PROGRAM

.............................36 REFERENCES

..........................................37 g

A P P E N D ! x............... '.............................

3 8 i,

i l

i i

()')

J l

(>}

\\

CRITERIA AND TEST METHODS FOR CERTIFYING AIR. PURIFYING RESPIRATOR CARTRIDGES AND CANISTERS AGAINST RADIOIODINE October I,1978 - September 30,1982 by Gerry O. Wood Frank O. Valdez Vincent Gutschick l

Prepared for Oflice of Neefear Regulatory Research U.S. Nuclear Regulatory Commission Washington. DC 20555 k

l NBC FIN No. A7041 ABSTRACT A project has been completed which provides experimental data and recommends, tions for establishing a standard test procedure and acceptance criteria for air purifying respirator cartridges and cants.ers used for altbome radiolodine. Previous experimental work with methyllodide vapor was extended to generate elementallodine and measure its removal by charcoals. A special apparatus was constructed and used to simu!une, ously rnessure penetrations of radiciodine and normal lodine vapor species through beds of verbus charcoals. Normal methyl lodide (1,127) was selected as th.t mest representative vapor species for testing and its limitations were identined. Effects of testing and use conditions (bed depth, contact time, concentration, relative humidity, temperature, flowrate, and flow cycling) were studied to identify testing requirements.

Temperature and almulated breathing flow cycling were showTi to have much more significance than was previously realized. Recommendations for testir.g and approval Include considering the effects of all these parameters. An apparatus designed and built for testing has been delhered to the National Institute for Occupational Safety and Health, in one related study the desorpden of triethuenedismine (TEDA), a charcoal,

trnpregnant for organic lodide removal, was found to be insignincant at normal canister use conditions.

r

,-,~.n~.----

,n.,.

,.,n._-...

n-

)

(0 L INTRODUCTION (11) Development of final acceptance tests, sp.

pustus, and crite:ia to be recommended to NRC for The main goal of this project has been to preside the ' approval of respirator cartridges against radiciodine.

I Nuclear Regulatory Commission (NRC), the National (12) Publication of results of this project and transfer Institute for Occupational Safety and Health (NIOSH) of the test procedures and techtuques developed to the Tesdag and Certification Brusch (TCB), respirator NIOSH TCB and assistance to them in the develcpment manufacturers, and respirator users with data, recom-of an approval schedule.

mendations, and proven test methods for certifying air.

Items 8,9, and 10 have been added to the original plan purifying respirators against radiciodire Since facepie:c to address concerns which have arisen as the project fit is being determined at Los Alamos and elsewhere in deve!oped. In addition, a complete, ready to-use test other studies, the main concern in this project was with apparatus has been buut for the NIOSH TCB to use for the air purifying canister or cartridge used with certification testing.

facepieces.

A preceding progress report' covered the first three of Steps which have been taken to accomplish this goaj the above steps and included the background for this are:

project. This report includes and organizes work re.

(!) Survey and anaj sis of the literature relating to ported since September 1978 (a quarterly letter reports, y

air. purifying respiraters, vapor adsorption, and radio-presentations at professional meetings, and publications.

iodine air cleaning. Contacts with professionals ex-Wita the exceptions of some journal publications to perienced in these fields.

follow, this is the fina] report for this project.

(2) Design and construction of an experimental apparatus for sorbent testing, including ge::eration and

'etection systems for nonradioactive 8871 vapor species.

!!. E!.EMENTAL IODINE GENERATION AND (3) Experirnental study of the adsorption of methyl ADSORPTION ON ACTIVATED CHARCOAL iodide on a variety of potential respirator sorbents and examination of the effects of endronmental and cartridge A. Objectives design puameters on this adsorpdon.

(4) Experimentaj study of the adsorption of elemen.

Testing of a selected adsorbent, an unimpregnated taj icdine vapors under limited conditions, activated charcoal, for adsorption of elementaliodine at (5) Experimental study of the adsorption of ppm chauenge concentratient was done to examir. th:

hypoiodous acid (HOI) vapors.

usefulness of1 generation and detection methods and to 3

(6) Design and construction of facDities for the use demonstrate the kinds of results that might be expected of rad 5 iodine for sorbent testing and development of in a respirator cartridge test.

radiciodine generators and detectors.

G) Experimental study of the adsorption of lodine vapor species tagged with "'I for comparisons of resujts B. Generadon with those obtained using stable d'l species.

(8) Studies of the efTects of re!adve humidity, tem.

One I generation technique used a flow of a.r (10 perature, Dowrate, and concentration on cartridge per.

L/rc.in) to pick up is and H 0 evaporating from an 1

f;rmance and senice life.

aqueous solution, (s 10-8 moles /L) Relative humidity.

(9) Measurements of desorption rates of charcoal resulting from H 0 evaporation was about 50%. The 3

Impregnants used to enhance methyliodide removal, cha!!enge and test bed breakthrough concentrations were (10) Evalua en of effects of eyeiic 0ow on etliciencies measured using calibrated oxidant meters (Mast Model d

and service lives of poten:ial radiciodine canisters.

724 5). The challenge concentration (C,) ofla generated b

i i

(g 1

in air was directly proportional to In concentratiori in 3

'l

'E solution [C, (mpm3) = 26400 [1 ) (moles /L)]. Bo.h E

3 concentrations decreased linearly with time as 1 3

w evaporated faster than H 0.

3 O

Another generation technique involved the sublima-tion of In cryst:Js at controlled temperatures into a z

0 4

flowing air stream. Challenge concentrations of 8 30 10 mg/m'(1-4 ppm), determined by weight losses and air o

i E

0.5 '

flow rates (101/ min), were relatively steady for up to a O

o.1

~

week.

5

.02-sc C. Retention 8

,,,,,,1 g3 1

10 100 The activated charcoal used for these studies was 6/16. mesh from Union Carbid CH ALLENGE CONCENTR ATION (ppm)

FractionaJ bed breakthrough C,/C, from I: genera.

Fig.1. lodine breakthrough times for a 6/16-mesh Union tion from solutions increased from zero to a constant Carbide medveted charcoal bed,1.25<m deep. 2.4<m value in a time interval (10-120 min) dependent on bed darn,10 M & b. @n symM are kr m condition and chajiense concentration. This limit value relative humdity and solid symbols are for 90% relative humidity experiments. Fractional bed penetradons: C for of C,/C, was constant over a wide range (X 300) of C, 0.02, a for 0.1, o for o.J.

and decreased exponentia))y with bed depth, D,le., C =

C,e-*D. All of these observations suggest that this initial bed penetration is contro!!ed by kinetic adsorption radiciodine environments. However, due to the long processes rather than by adsorbent capacity. It should, experimentaj times involved, the determination of therefore, be equal tc, adsorption emeiency at much cartridge lifetimes may not be a practical way of lower challenge concentrations.

measuring and comparing cartridge performances. The Retention studies of la generated from crystals used observed initial breakthrough may be a more useful beds of 6,'16 mes' tharcoaj,2.4.cm diameter and usually indicator of cartridge performance.

1.25.cm deep, which corresponds at 10 L/ min to a linear Five types of potentia] radiciodine adsorbents were flow velocity of 22 m/ min. Relative humidities of 50%

compared for 1 adsorption emeiency at the following 3

and 90% were used. In these longer term experiments at condiuens:

constant C., after the inidal constant bed breakthrough 2.4.cm diam x 1.25-cm deep beds was subtracted out, a subsequent increase in penetration 50% relative humidity developed more slowly, requiring up to 7 d to reach a.1 0.0127 g 1:in 100 rn! H:0 generator solution.

additionai 10% bed penetration. This wu due to loss of 10 L/ min (22.1 m/ min) air flow rate.

capacity as active sites were being used up.

The measured penetration fractions were:

This subsequent breakthrough curve wss best de-0.27% Westvaco WY H, coal charcoal, not im.

scribed by the equations of the Statistica] Moments pregnated.

Theory, as was previously found for tr. ethyl iodide.8 An 0.14% Sutcliffe Speakman 207A, 1.5% K1 im-la hallenge concentration effect (Fig.1) was observed:

pregnated.

c t, = kC7s (t, = breakthrough time for a selected 0.!!% Coast Engineering Silver Zeolite, AgZ.

frs:0onal penetration C,/C.), similar to what was ob.

0.072% Sutcliffe Speakman 20EC, 5% TEDA im.

served for CH:1. Again, this implies that such cartridge pregnated.

((

lifetimes, determined at ppm levels using normaliodine 0.032% Witco 337, petroleum charconj, not im-would be conservedve for much lower levels expected in pregnated.

s f

4 DI. RADIOIODINE STUDIES-EXPERIMENTAL the headspace over a heated water reservoir. A hunddity monitor / controller (Phys Chemical Research Corp.)

A. Flow System which regulated water temperatt:re was calibrated with a dew point hygrometer (EG&G 911) at the test bed The apparatus used to measure the penetrations of location. Water level was maintained automatically by a '

volatile lodine and radioiodine compounds through test conductive liquid level control (turneh Electronic beds, canisters, und cartridges is diagrammed in Fig. 2 Co.).

and shown in Fig. 3. It was built inside a fume hoed to A "Standard Operating Procedure for Use of n'! in exhaust any toxic vapors which might have been re-the Testing of Respirator Components"8 was prepared leased. Radiciodine solutions and contaminated sorbents and approved by internaj review. It describes the ex.

were contained for further safety within a glove box with perimental spparatus, procedures, and precautions tc be charcoal and HEPA exhaust Illters. Vapor generation used with this radionuclide.

and test bed exposures were done within the glovebox.

Compressed air was filtered. regulated for paper flow -

rate, and humidified before entering the glove box. An B. Generation Methods electronic mass flow meter (Datametrics 8001.) which monitored airlflow was periodically checked using a dry Yapers were generated in two ways shown in Fig. 2.

test meter (Singer DTM.325) at the test bed location.

Liquid methyl iodide and methyl radiciodide sealed in a Humidi.qcation was accomplished by passing air through Tcilon permeation tube were released at a steady rate by

~

g TEMPERATURE TEST BED HUMIDITY AfiD WATER LEVEL SEtiSOR C0f1 TROLLED BATH f ?

VEtiT TO t

~

GLOVEB0X

!d I

\\%s J

e SAMPL! fig

'SAMPLit?G LOOP 2 LOOP 1 g

P. ASS FLOW SEfiSOR A /.CH4 VALVES GAS I

F CHROMAT0 GRAPH CARRIE PARTICULATE FILTER WITH ECD rF%

,pggp' CHARC0AL FILTERS J

tt PRESSURE REGULATOR RAD 1010Dif1E CHARC0AL TRAPS

,L PARTICULATE FILTER c

NA[ CRYSIALS O

PHOTOMULTIPLIER PERMEAT10ft SOLUT!0:1 TUBES GEtiERATOR GEfiERATOR p

Fig. 2. Espenmenal appustus (ce resung air pun /Hng respastcv carmdass and canJaters using ts6oioine and riormaj iodans espor species.

m a,_

h 4

[45#.l1,tv-9 n;

7 77-i

.x

,p ;

p.s a.r t

e o

t s.

.m

.c

. i p, i

m 4

s e.g

y g;. u i-

.5'.

bg

} $..

' _ / i, Du

/

N.

f.

l' c.a. -

'V.~.,*

. &. t

.)

dl E1"7 8

m 1

e.

-~

4 a

(h n-1[$ '/4 e1 h

)

Si74W"F M fjC

'I.P

$~.

jl l

1]

ifgl:

.}l.s

/

nel 1

I if g' G.,f ?- Q h 2,

si.

!W, 7

g iq-;o!

=

5, y

w.

-1 Y[

g J

n g,

g,3g n

g

i

. c. --

.i

[ ;.g.i.

e F-M

.e.1

.4

.r h a t' f

3 b

9:k si.MGMkh <

ff

[

'7 1.s : 7 Q74'/.;f,JE.lg*'el*t 9Hl o/

,, : m +>

);

1 ;

k

)"I I I.

M d 4)>

~

i)

Ll-W$ & % &j j

]h 7--

\\

h pe"m m:

a#$h,e. m '

.m q\\ qg ea.,s

.wm m

~O H.~Y., E k.N<.; M. a m a *__ R' ) w, E if[ :.; :

\\

t Wn e m-s

.\\.

L

_,,.m,..--,.--e-

~ " - ' -

/

/

l

/

e

[

penr~ ting into a 500 cm / min airCow. Temperature digital valve sequence programmer (Vajco Instrument 8

csntru (25 70'C

• 0.l'C) of this permention tube was Co.) to alternately inject the upstream and downstream by'he Calibration System (Analytical!nstrument Devel-air at 5 minute inte vals. The chromatographic column opment, Inc., Model 303). Alternately, methyl radio-was 1.8 m x 4 mm.id. 31sss packed with 15% OV 7 on iodide (CHj881), elert mtal radiciodine (8881 ), and 100/120 mesh Chromosorb G. Operating conditions 3

hyporadiciodous acid (HO'8'l) were generated from were 100'C and 20 cm / min 19:1 Ar:CH. canier gas.

8 aqueous solutions. A tyringe was used to inject 10 mL of An electronic peak inteststor (Spectra Physics Mini-solution into 100 mL of distilled water or other reagent grator) quantitated the methyliodide peaks and recorm solution in the glass container in the lower center of Fig.

elapsed times. Calibrations of this analytical system were

2. The volatDe iodine compounds in this stirred mixture made using weighed permention tubes to generate known 8

entered the head space and were swept by 500 cm / min methyliodide concentrations in air.

L of air through TeDon and ; ass tubing into the main The radiometric detectors continuously coUeeted and airflow. Water vapor was also generated. Output of measured 88'l from the 0.3 L/ min air samples passing volatues from solution dropped exponentiaUy from the through the gas chromatograph sampling valve. Fig. 2 time ofinjection. Generator output and main airstream shows the charcoal trap and 7.6 cm diam x 7.6-cm tnick passed through sumcient length of 2.4 mm.i.d. glass Nal (TI) well type (52 mm deep x 29.mm-diam) scintula-tubint. and two c! bows to mix thoroughly before entering tion crystal with intestal photomultiplier tube (Harshav the test bed. Sections of the glass now system and the Chemical Co.). High-emeiency charcoals were used: 5%

test bed were connected with O ring seals and clamps.

TEDA impregnated (Barnebey Cheney CN 2762) for Chauense air and test bed emuent air were sampled CHj881 and activated charcoal (Union Carbide ACC) continuously through Tenon tubes connected to the glass for 88'!: and HO'88L The majority of radiciodine was system and into the gas chromatograph and charcoal couected at the bottom of the wc!L resulting in good Q beds.

detection emeiencies (-0.5) for the 0.364 MeV gamma.

T:le technique of generating volatUe iodine s,pecies ray. Each detector for upstream and downstream air had from aqueous solutions for the testing of serbent beds or its own preampliner, amplifier, single channel analyzer, respirator cartridges has proved to be quite useful and counter (all from Ortec). They shared the power bin t

Concentrations in water (and in air) decrease with time, (Ortec), high voltage power supply (Canberra), timer approximately exponent!ajly, dependi.g on species vola-(Ortec), and printer (Ortec). Linear log rate meters tility and,in some cases (CH 1, the rate of stirring of the (Mech Tronics) were used for count rate monitoring.

3 solution. One advantage of this generation method is that Detector counts were taken from 5 minute intervals and a range of challenge concentrations is produced in a printed together. Each detector trap and crystal was single experiment. This can give information about the shielded by 5 cm of! cad to reduce background counts.

adsorption isotherm of the test bed. Another advantage Fresh charcoal was placed in the detector traps for for inerganic species, particularly, is that generation is background counts before each new bed was tested. The from a scurce similar to field soarecs, such as reactor detectors were compared almost daily for relative coolant waters or spent fuel cooling pools. Jt is also sensitivities by sampling the same radiciodine containing possible that experimental generator solutions can be s'r..

matheed (pH, additives, etc.) to actual aqueous field sources.

D. Reagents C. Detection Methods The source for radioiodine 131 was ICN Chemical and Radioisotope Division, Irvine, CA. Mett.yl radio-The detector for methyliodide was a gas chromato-iodide was creeted as 5 mCl 8881 in 3 mL of total graph (Varian 1520) w;th a linearized electron capture methyliodide. Stated purity was a least 99%. Two

(( detector '(Tracor instrEments). Air from upstream and miUiliters were used to fdl a permestion tube and 1 mL downstream.of thef est bed was drawn (0.8 L/ min) was dissobed in 1 L of double dist0!cd water. This t

through matched. Tenon sampling loops attached to a aqueous solution (2.3 g/L or 0.016 meUL) was used 10 10 port vaM('Vaic6 friitt'ume3t Co.) of H astalloy C (for mL at a time for generating as described above. Radio-inertness). This valve}s0pneamatically actuated by a iodine in the form of Nal in 0.05 N NaOH was

d j

?

purchased in a 5 mci amount. Stated purity was at least against lodine for radiciodine vapors. Each type con.

99% with an '"I/88'I ratio less than 10. This material of tained a particulate filter followed by a sorbent bed cbout I mL volume was added to 1 L H 0 contairdng containing a coarse grained charcoal. Some of the 3

0.127 g of dissolved I, (0.127 g/L or 5 x 10-* mol/L).

charcoal sorbents were reponedly impregnated with Isotope exchange occurred to form 88'1. This solution reactive chemicals for radioiodine removal, such as 3

was used to generate both '881 and Hol'88. For HO 881, triethylenediamine (TEDA) and K!3 (K! + 1 ). The 3

3 10 rnL of this latter solution were injected into 100 mL of distinction which is made in this paper between canisters 4 x 10-8 mol/L NalO: at pH = 2 to cause the reactions:'

and cantidges is that the latter are used in pairs and are physically smaller. For some experiments beds of 2.4.cm 21 + 10; + 6H* + 2H 0 = $ H Ol*

diam were prepared from charcoals taken from canisters.

3 3

3 The term "test bed" will be used in this repon to refer to H Ol* + H 0 - HO! + H O+

a canister, a cantidge, or an experimental bed. Table I 2

3 lists characteristics of the canisters and cartridges and No anempt was made to determine the extent of HO!

their charcoaj contents, production, since no analytical method is knowTi which distinguishes this unstable species from I.

IV. R ADIO!ODINE STUDIES-RESULTS AND CONCLUSIONS E. Test Beds A. Comparison of Vapor Species gj Air. purifying respirator cardsters and cartridges were obtained from three U.S. corntnercia] sources: Mine Penetration test results at high humidity (97 :e 3% for Safety Appliances Company (MSA), Pittsburgh, PA; the three radiciodine vapor species are tabulated in Table Nonen Company, Safety Products Division, Cranston,

!! for five canisters (64 L/ min) and in Table !!! for four RI: and Scon, He:Jth/ Safety Produ:ts, South Haven, cartridges (32 L/ min). Pulses of challenge vaper were M1. They ca:h claimed by labeting or personal manufac.

generated from solution at 2. hour intervajs. Two. hour turer information, to be of some use for protection average penetrations and standard deviations (given in TABLE I. Characteristics of Canisters and Cartridges Tested Charcoal Bed Geometry' Cross Charcoal Impregnants' Source Tne Designation Section (em') Depth (em) Volurne (em')

(Weight Percent)

MSA Canister G M R.!

IlW 3.2 350 3% K1 3 Caruster OMR !(TEDA)*

110' 3.2 350

$% Kl,2% TEDA' i

Canister G M R.S 110' 3.2 350 Metal and Ammonium Salts' Scott Canister 600232 75 87 3.8 330

$% TEDA Cardster 282 OAP.R 87 3.8 330 Metal and Ammonium Salts' Canridge 604$50 73 48 1.3 52 5% TEDA Cantidge 604403 75 48 1.3 62 5% TEDA Nonen Cartriose Type l' 44 2.4 106

$% TEDA Canridge Type !!"

44 2.4 106

$% TED A g.

' Measured from opened canisters.

b'

'Best Information from manufacturers.

' Oval cross section.

'The GMR.!(TEDA) designatlon is used for GMR.! canLiers manufacturtd after July,1979 through at least Apri?,1980.

'TEDA = triethylenedismine.

'Whetlerized charecal.

8 Granule sine 316 mesh.

~~.

/

TABLE 11. Radiciodine Test Results for Canisters' 1

Average Percent Instantaneous I'"'t'*li "S(""d S'*"d"'d D' u 'I "'I' Canister Test Type Vapor 02h 24h 4-6 h 68h 810 h 0.24 Scott CH "'I -

3 (0.02) 600252 75

!.07 0

0.6 I (0.06)

(0.05)

"'!, 0-0-

4 0.10 0.08 0-HO"'I (0.03). (0.03)

GMR1 CH,"'I 0.24 4.43 6.09 (TEDA)

(0.08) (0.16)

(0.17) 0.99 2.46 7.54 (0.41) (0.73)

(1.06) 8"!,

0.71 0.10 0.10 (0.04) (0.02)

(0.02) 0.08 4

0-0-

HO"81 (0.04)

Q GMR1 CH 8"I, 3J4 8.40 21.29 3

(0.52) (0J3)

(0.65) 0.17 0.07 0

(0.02) (0.01) 0-0.07 0.17 0.11 HO"'I (0.03) (0.07) (0.03)

Scott CH "'I 19 98 100 3

282 OAP.R (2)

(4)

(7) 8"1 0-0.07 0.18 3

(0.01)

(0.02) 0-4 0.27 0.75 HO"'I (0.04) (0.04)

GMRS 8"!:

4 4

4 HO"81 0-4 4

'641/ min 97 i 3% RH.

'Zero value (0-) means not significantly greater than aero at the 95% cordidence level. Dash (-) means not messated.

parentheses) were determined by linest regression analy-cardster or cartridge was tested.

Methyllodide wa: the vapor form of radiciodine that j

sis of 5 minute counts in the donstream detector versus most readily ynetrated the respirator canisters and j

the upstream detector. Relative sensithity of the two cartridges which were tested. Penetrations of 8"I, and radiciodine detectors determined by daily calibrations HO"81 at high humidity were low (s0.15%) and, with g

was taken into account. Any gnetrations calculated to one exception, did not increase significantly with ex.

be within 95% confidenec 1cvels of teto were considered posure and loading. Since methyl iodide is the most as zero and listed as 0. In only two cases were more than one canister or cartridge of a type tested for a given volatile organic iodide compound, other crganic lodidas should be retained on these canisters er cartridges witn radiciodine vapor. Therefore, these results cannot reflect the sarne or higher emeiencies. Therefore, methyllocide

-F d-at eh5 s iven me. At lesst three of each N

J

[

/

was rn n ce ng neuly c nuant L5 TABLE 111. Radiolodine Test Results for Cartridges, 0.5% penetration of both methyliodide and radiciodine Average Percent throughout the experiment. The GMR 1 ($% K1, 2%

3 Instantaneous Penetrations TEDA) was less emeient at about 10 2% methyl Cuddge Ten Type Vapor 02h 24h 46h lodide penetration and 5

• 1% methyl radiciodide penetration after an initial equilibration period. The Norton Type !

CH "'I 0.03 1.94 3.34 3

GMR.! (5% K! ) charcoal was most efficient at the 3

(0.01)'

(0.06) (0.60) beginning c'f the experiments, but rapidly and steadily

"'Is 4

4

-0 deteriorated to give a 60% cumulative fractional methyl Norton Type !! " 81 0-

+

0*

3 iodide penetration and a 17% cumulative fractional Scott CH "'I 1.18 9.27 11.50 3

methyl radiciodide penetration by 100 minutes.

604403 75 (0.33), (0.29)

(0.34)

Results from seventeen experiments with iodiced

"'Is 0.04 0

4 charcoals are compued in Fig. 6, which shows CH "'I 3

@01) cumulative percent penetration versus CH ! cumulative 3

Scott C H "81 1.98 10.71 12.87 3

percent penetration. The data points all fall below the 604550 75 (0.17)

(0.99)

(0.90) equality (dashed)line, Lc., CH 1 penetration greater than 3

'J2 Umm,97

• 3% RH.

CH "81 penetration. Also, in the region of practical 3

'Zeio value ( 0-) means not signincantly greater than aero at interest (less than 10% penetration) the difference is an 8 Standard ti ns i

i Cases.

A more extensive comparison of fractional penetra.

LO.

tions for Scoi (5% TEDA) bees is summuited in Fig. 7.

should be used as the test species to determine the upper These results are from 14 experiments at two humidities limit penetration of vapors containing iodine.

for two generation methods, and for three bed depths Milo Kabat and coworkers at Ontario Hydro have (1.25 3.75 cm). Each graphed point represents the aver.

challenged four cartridges and canisters with CH !, HO!,

age of 20 to 30 measurements for a given experiment.

3 and 1 forms of radiciodine.8 The results shown in Table The penetration values all fall close to the theoredcal 3

IV, conArm that HO! and 13 removal and retention (dashed) equality line. Therefore, for this type of serbent emciencies are greater than or essentia]!y equaJ to those (TEDA only) measurements of mele:ular CH 1 penetra.

3 i

for CH 1.

tions ne direct measurements of the "'I penetration 3

when the radiciodine chaljenge is in the fcrm of CHj"1.

A fourth type of charconj. from an MSA GMR S B. Methyllodide Versus Methyl RadiciodJde canister, was tested to compare methyl iodide and j

radiciodide penetrations. This Whetlerized charcoal is Cumulative percent penetrations through three twes impregnated with metal and ammonium sajts, but con.

of impregnated charcoajs are compared for methyl tains no impregnants that react with methyl radiciodide.

i iodide (Fig. 4) and for methyl radiciodide (Fig. 5). The Therefore, removal of "81 in CH "81 can occur only by 3

test beds,3.75 cm deep by 2A-cm diameter consisted of physical adsorption of the molecule. Cumulative frac.

charcoals taken from MS GMR4, GMR !(TEDA) and tional penetrations of methyl lodide and methyl radio-1 Scott 60025215 canisters. Each bed was preconditioned lodide are compared in Fig. 8 for duplicate experiments.

for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> at the test condidens of 3 t/ min airlfew and The data points closely fit the equivalence line until the 86 3% relative humidity before being challenged with amount desorbing from the test bed equ Js that entering 8

1.5 ppm (7 mg/m ) methyliodide tagged with 8"!.

it. Then there was a slight deviation in the direction of Cumulative fractionaj methyl radiciodide penetrations greater radiciodine penetration than methyliodide pene.

'(

were calculated directly from 5 minute in'erval counts of tration. This deviation is expleined as the result of

/

radioiodine trapped in the detectors. Cumulative frac.

forming volati!e lodides other than methyl iodide. The tional penetrations of methyl iodide were cajculatted by radioiodine detector is not compound spec 4c. but the integrating instantaneous upstream and downstream gas chromatograph is and would not measu.v the other concentrations deterrrined by su chromatography. The iodides. Instantaneous fractional penettstions (eLent 5% TEDA.irnpregnated chueca] from the Scott canister concentratien/ challenge con:entrauen) er methyl iodide i

/

/

/

/

/

=

TABLE IV. Comparisons of Radioiodine Species Removal Emelencies by Kabat*

Adsorber Radioiodine Altflow Range of Percen 8881 Type Species 1/ min Adsorption (3h)

Desorydon (2L)

GMA.H CH1 20 98.13 99.29 3.! !

3 Cartridge HO!

99.24 99.90 0.21.5 I:

99.92 99.96

<0.10. I I GMIH CH1 20 99.95 99.96

0. < 0. I 3

Cartridge HO!

99.87 99.92

< 0.1 1,

99.92 99.96 0.19-0.33 Canadian C1 CH1 40 52-47 41 46 3

Canister H01 99.2 99.6 0.34.3 1:

98.3 1.1 MSA Twe N CH!

40 99.91 99.93 0.14 0.59 3

Canister HO!

99.21 99.93 0.1 1

99.87 99.96

<0.10.15 3

'From Reference 3. 95 percent RH. 25'C.

ao g

(O a00 1

wg M

FIs.4. Methyllodide cornulsdve percent penetradons as fum tions of 4g g

time for charconJs fro.s three respirsto canisters: C. Scott p

600232 75 (3% TIDA); o. MS A GMR I(5% KJ,. 2% TIDA);

A. MS A GMR 1 (3 % KJ,).

" m N. W. W m " T.~ m g

0 20 40 40 so 10 0 TIPE (MIN)

G 20 c'

E h15 W

t 1

Fig. 5. Met.hyl redoMdr cumWsdvs percent peneradoes as futc.

10 tions of dine for ther:4als from three respirator cardeurs: C. Scoct l

600232 13 (5 % TIDA); o.MSA GMR 1(3% KJ, e 2% TIDA);

3 l

A. MS A GMR I(5% K!,).

g

(, bd v # D " U'

((

5 0

20 40 ao s0 10 0 TIPI.(miw) n

/

p 8 20 80 i

/

f g

~

g

,I h 60 15

~

z

/

5 A

/

go[

d t

h 100 %

/

d INSTANTA'iE003 e

h8 jo D

/

y 40

- PENET UTION C

4 1

g

-l#

y 5

5

_ 2O

&g m

j Q

l 5

g'^ o i

O 20 40 60 80 0'

l C

20 40 60 80 CH 1 CUMul.ATIVE PENETRATION C)

CH 1 CUM.ATIVE PENETMTICN (!)

3 3

Fg. 6. Comparisons of rumWadve percent penceadons of Fig. 8. Compedions nfcumdsdie percent penevadons of mecil methyl ra6oiciide and methyllo6da for two $% K!, charcoals:

redoiodide and methyllo6de for Whederized charcoal (GM A f)

A. GMR 1; O. GMR 1 (TIDA).

for two separate expenmentA (4 4Ad o ).

increased with time and even exeteded 100% as the '

100;

,]

vepor adsorbed at the beginning was displaced in the air.

Breakthrough times of methyl iodide averaged 33 $ 3 g

g minutes at 0.1%. 49 6 minutes at 1%. and 68 :: 8 g

,e minutes at 10% Instantaneous penetration.

5 k,

Nermaj methyl iodide can be used to determine the

= 10 r upper limit of penetration to be expected for methyl y

,4' radiciodide. Isotope equiv Jent efTi:lencies have been O,

l demonstrated for sorbents not impregnated with nerma]

@g" iodine or iodide. Normal methyl iodide tests car. net 0,

5 I

,e

]

measure the removal of 8881 by isotope exchange on

~

,e lodized charcoajs and, therefore, give a high (cen.

tJ h

servative) estmate of methyl radiciodide penetration.

e' However. there are currently no commercial radioicdine

/

canisters or cartridges which have charconfs im-O.1 pregnated with iodide only.The GMR 1 canister with 3%

0.1

)

to 100 XI, packing is no longer available.

AVEW.E CH 1 PEXITMTION (1) 3 l

Fig f. Cornpadsen of everste pertem penc4 ens of methyl C. Effects of Bed Depth :nd Contact Tirne esdye&de and enrhyl M4e for a $b TIDA imprignated shascosi (Scon 600212 73). Pcmesdon tube graersuont o.

Another series of experiments with the TED A and K}3 91% RH and e.Il% RH;agveou saludon genersdon:fi.91%

(rnpregnated respirator canister chsico415 was done to (g

RH.

estaolish the rate orders of methyliodide and radioicdide

\\

w

/

\\

/

/

\\

l

/

/

a

[

removal. The ranges of test conditions were:

yltadiciodide concentration. The range of airdow rates Bed depth: 1.25 3.75 cm was r.ot large enough to notice the velocity effects found Bed diameter: 2.4 cm later.

/

Airflow rate: 1.8-4.2 IJmin or 6.715.3 crrt s Ten such experiments using 5% TEDA charcoal from Bed residence time: 0.16 0.75 s Scott 600252 75 canisters were also done at similar Relative humidity: 86 3%

conditions. Semilog plots for penetration percents (both Concentrations: 61200 nCi/m' 8881 and 0.19 72 methyliodide and radiciodide) versus bed contact times 3

mg/m CH 1 (Fig. 9) showed that the chemisorption reaction is also 3

Conditioning Period: 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> described by a simple first order rate. Both meth) iodide Seventeen tests with mehtyl radiodide generated from a and radiciodide are removed from air at the same rate.

permention tube were done using lodized charcoal from Four experiments were also done at different bed deptns an MSA GMR 1 canister. In each test the meth-(1.25 5.0 cm) using charcoal from MSA GMR !

yltadiciodide instantaneous penetration remained nearly (TEDA) canisters, penetrations of methyl lodide and constant, while methyl iodide instantaneous penetration radiciodine during each run were both constant, but not i

inercased steadily with time until it exceeded 100%.

equal. The difTerer.cc for this mixed impregnant (2%

When the logarithm of methyttadiciodide penetradon TEDA and 5% K! ) serbent is due to isotop: exchange 3

percents were plotted against bed contact times a straigt which removes the 88'I from the methyttadiciodide but line with an intercept of 1.0 resulted (Fig. 9). This

! caves a molceule of methyl iodide. Average drst order indicates that the methyttadiciodine removal reaction rate cer!icients calculated from the slopes of plots such (isc. ope exchange) is simple first order in meth-as Fig. 9 are listed in Table V.

The reaction of TEDA impregnant with methyliodide vapor is by first order kinetics. The isotope exchange of (C'

100 ioejde impr,,nant to remove 13e radjoiodine from meth-y!radiciodide is also by first order kinetics. Emuent vapor concentrations decreased exponentially with bed O

depth.These results indicate that removal emciency was o

independent of vapor concentrations within the bed. This is an important conclusica, since the radiciodine concen-3O o

w a

untions to be encountered in nuclear environments are

~

many orders of magnitude lower than the ppm concen-N trations required for a nonradiometric test. The first order kinetics also implied that contact time of vapor g 10 :

o within the serbent bed is critical. Centact time is g

drtermined by canister geometry and airdow rate (i.e.,

l l

workload). A high flow rate should be chosen for a g

canister test to approach the upper limit of average vapor g

pencuation. The arbitrary test standard is 64 IJmin for g

4 l

canisters and32 IJmin for individual cartridges used in pairs.

Canister charconjs containing 5% TEDA impregnant were more effective for methyliodide removsj than those A

containing 5% K! impregnants are more emcient for 3

methyl radioiodide removal than those without. except j

i i

i i

e 0

0.2 0.4 0.6 0.8 I 'sh it Periods with fresh canisters.

((

$0EENT BED CONTACT TIME (s)

D. Effects of Challenge ConcentratJons Ms.

b. Amsse Wtutaneovi percent penevadoes a,

logarismic betions of t=4 sentact timent o enemyl ts6o lo&de for a $% K!, charcoal (GMR lh t inedyl ladde and Five tests were made with 5% K!. impregnated 3

meGyl rs&oiocde for a 5% TID A charces)(se*"k charcoal under these conditions:

• e O

f f

[

,s

/

./

/

TABLE V. First Order Rate CoeiT.clents for Methyl lodide and Radlolod!de Removal' Rate Coefficient (s-')

Charcoal Charcoal Total Total Isotope Impregnant Source CH,"81 CH,I Exchange' 5% TEDA Scott 6.9 i 0.5 7.1 t 0.3 None 600252 75 5% KI G MR.1 3.6

• 0.3 3.6

• 0.3 3

5% K! +

GMR1 4.9

• 0.1 3.0

• 0.1 1.9

• 0.2 3

2% TEDA (TEDA)

• 2.h preconditierting at 86% RH.

• Difference of preceeding two columns.

3.15.cm depth x 2.5.cm-diam bed The Wheeler adserption equation predicts the 8

1.81/ min airnow: 6.6 cm/st 0.57s bed contact time logarithm of penetration as a linear function of time for 86% RH; 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> equilibration before testing 0.19 3.8 low penetrations (<!!%9, and such plots have been 8

reported for CH 1. The penetration curves for the mg/m CH l i

3 8

0.0061-0.125 pCl/m

"'t.

experitnents reported here with the iodized charcoa!

Chauenge concentrations varied over a factor of 20.

ensistently fit the Statistical Moments Theory (SMT)

Breakthrough times (t,) of CH 1 for 1%,10%, and 50%

equations' and an empirical exponential C/C, = at' 3

instantaneous penetrations were nearly the rame for all equation better than the Wheeler equation. For example, challenge concentratins (Table VI.) Individual break.

four data sets at penetrations less than 15% yielded the through times were used to calculate de breakthrough correlations in Tab!: VII.

capacities plotted versus chajlenge concentrations in Fig.

The consistent failure of the Wheeler equation to give

10. The linearity of thve plots indicated that CH 1 the best rit of penetration resuhs frem many experiments 3

adsorption and desorption occurred according to a brings to question its use in extrapolating to define initial j

simpte linerar isotherm (Henry's law). Other charcoals penetration at initial exposure. However, it will ajways 1

which have been tested with CH 1 have not indicated give a cm rvative (higher)initia] value relative to the 3

linear isotherms.8 true one due to the curvature of the breakthrough curve.

TABLE VI. Effects of Challenge Concentrat.as 1

CHti CH, 8"!

Breakthrough Times (min)

Conc.

Cone.

Percent (mg/m )

'l%

'l0%

'50%

(nCL/m )

Penetration' 8

8 O.19 6.8 18.0 39.6 6.1 12.3 0.41 2.4 14.7 41.5 13.6 16.0 1.29 3.8 15.4 41.8 42.6 9.7 2.19 7.4 20.5 45.8 72.3 13.4 (g

3.78 3.4 15.I 38.7 124.7 11.5 Average 4.8 16.7 41.5 Average 13.0 Std. Dev.

2.2 2.5 2.7 Std. Dev.

2.7

• Agerage instantaneous penetration aAer the inidal period in =tdeh phpical adsorption was s'gnificant.

~

4 i'

,/,

l Four cartridges, all containing 5% TEDA.im.

0,3 pregnated chercoals, were tested with methyltadiciodide at 0,2, and 4. hour exposure times to 32 L/rnin,97

• 3%

gg RH air. Methyliodide penetrations again increased with exposure times. Maximum penetrations (humidity equi.

.!0.2 a

E librations) were reached in about 3-4 hours. Average penetrations measured during 46 hours5.324074e-4 days <br />0.0128 hours <br />7.60582e-5 weeks <br />1.7503e-5 months <br /> by gas 10 g chromatography and by radiometric coundng are listed g 0.1 in Table VIII. The v Jues from the two methods are in good agreement. Cartridges with larger sorbent volumes 1g (Table 1) of similar sized and impregnated charcoals gave lower penetrations,. That this can be attributed to in.

h h

creaded bed contact time is shown in Fig.11. This O

y semilog plot also includes data from Table 11 for the g

O(31 MMT!@ M Scott 600252 75 censiter. The average Erst order rate i

coencient is 17.6s-H (standard deviation = 1.3 s-"). This Fig.10. Dyumic adsorpuon foodms at thm lastan.

correlation should be useful for improving emeiencies by taneove penetrodon frecuens for ewthyt ht;de and MsA GMR 1(2% XI,) chanos!.

redesigning canisters and cartridges.

The larger carusters (used alone) were more efective For the above expcirments, the Wheeler equation gave for methyl iodide removal than cartridges (used in pairs) even though the flow rate through each cartridge was

( g an initiaj penetration vajue of 0.33% (std. dev = 0.18%) ha!f as much. Also, the cartridges deteriorated in elTi.

to be compared with the SMT initial value of 0.094%.

ciency more rapidly due to high humidity. Magnitudes of One of the best nts of the breakthrough curves was for a emeiencies can be correisted with volumes of charcoal C/C, = at' empirical equation which has three dif.

Y and bed contact times.

ficulties:(1) it has no theoretical basis: (2)it does not Insumcient data are available to rate cartridges and allow extrapolation of penetration to irUtial exposure canisters for radiciodine removal. Variations within time; ani (3) extrapolated vajues very rapidly at short brands and types ahve not been established Also, their times, for this examp!e,0.020 at 0.5 rninutes,0.073% at contents are subject to change by the manufactureres.

1 minute, and 0.265% at 2 minutes.

These unknowns emphasize the need for an ongoing cert"ication program. Such a program to be carried out in the NIOSH Testing and Certincaiton Branch, is an E Cartridge Comparions uJtimate product of this project.

TABLE VII. Fits of Penettstion Data to Equatio'ns Linent Correlation Cocmelent, r Equation Wheeler:

In(C/C.) = In a + b t, 0.9256 Exponential In(C/C.) = In a + b in t, 0.9999 t, - m, + m/6m m

- - frE + g, X, 0.9999 SMT:

(

/

/

/

f TABLE Vill. Cartridge Penetration Freetions of Methyl Iodide and Methyl Radiolodide after EquDibration at High Humidity' Average Percent Instantaneous Penetradons Bed Charcoal Cartridge (and Standard Dedadons)

Contact Granule Type CH1 CH,88'I Time (s)

Mesh 3

Norton Type 1 3.75 3.34 0.20 8 16 (0.76)

(0.06)

Norton Type !!

1.50 0.20 12 20 (0.15)

Scott 604403 75 13.24 11.50 0.12 8 16 (0.59)

(0.34)

Scott 604550 75 16.52 12.87 0.12 8 16 (0.6 l}

(0.90)

'J21./ min. 97 2 J% RH. 4h.

100 whue continuously exposed to very high humidity (95 3% RH) air, increasing penetrations were usually ob. (

served. This is illustrated by the results in Tables 11 and Ill. Such an effect could be due to (1):en6ng the test bed p

with methyliodide in previous tests and/or (2) leading it E

I with water vapor by exposure to large volumes of high 3

humidity air. Studies were done to sort out these effects Q

using Scott 600252 75 canisters (5% TEDA charcoal) at E

64.L/ min airflow rate. Methyliodide was generated from l1

• ueous solutions (0.23 g/100 mL) at selected times r

i while a canister was exposed to high (97 2 3% or y

i l

medium (50 t 3%) relative humiity air. Penetration a

l esults versus exposure times are shown in Fig.12. Bot 0.1 10 0

0.1 0.2 0.3 0.4 SOG ENT BED CONTACT TIE O)

E

~

Fig. II. Dependence of sverage lastantaneous p"cent penetracons j

g

/

en bed contact times for etnisters and cartridges omtaining 3%

g

~

TEDA4mpregmazed charcoaja C, medyl lo6 den A. methyl ro60 W

ladde.

i I.

n.s.

N y ci I - l=

l 1

1 E

G

~

V. EFFECTS OF USE CONDITIONS

!l o.ul 0

.4 a

12 to 20 24 2s A. Relative Humidhy af talit tre:tw fit t.)

1. Cemparisons of Water Yapor and Methyl lodide Fts.12. Aversse initwaneous preens Penetratic'i' *f t"'*!'i*M' Loading. Wher' cardsters or cartridges were tested more throvsh seen 600:f:13 (1% TIDAL etsiters ** feccens of

.s..

.....s.s

....,..r.:.a.

r........

-e

h-f pe.t ranges of data obtained. In the first (open rectangles) a fresh canister was tested 2

1. 6, and 24. hour exposures to 97% RH sir.

o ps increased by over two orders of magrutude.

g b 'M experiment an ther canister was exposed i

p% Joo %o.

g gity for 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> before being tested at 16,

[

h g M boura. These data (shown as rectangles with

,,g fg ces the same curve as those from the first 5

j In the third experiment (solid rectangles) g g e:. canister was tested at the 50% RH and 0,2,4 3

g p g 3. hour aposures. Even for the longest time and g

g,en bed loading the penetration at 50% RH was not gp;nezntJy changed from the beginning.

A gonon Tne !! cartridge (12 20 mesh 5% TEDA) c.ol i

u a e

i so :

sN h io A H

,,3 gy,anenged with 1.7 ppm (7.6 mg/m ) methyliodide TIME S) 8 ct 3; L/ min air and 90% RH (Fig.13). During the first 3 ry.13. meet or esposun time on meest iede innameeous hours, the Pencadon fracdon increased nearly 2 orders penetntion fee a fnt'. Nonon Type Il cartridst temd at 90% RH.

of mapitude to 1% where it remained Constant for at M Umin adow, and 1.7 ppm p.6 mg/m') medyllMde.

least 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />. Since the bed was being loaded constantly with methyl iodide and there was no change in penetra-con fraction after 1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />, the initial change must be not exhibit heating due to additional water vapor adsorp-P attributed to something other than sorbent exhaustion by tien (see Se: tion V. A.4.). Also, the penetradon is often methyl iodide loading. Apparently,3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> was required maximized at such equilibrium. A Scott cartridge (642-for the charcosJ to be:ome equuibrated with water vapor TEDA H) containing 5% TEDA impregnated charcoal

{fi equilibrium with the 90% RH air.The larger canisters was exposed at 32 L/ min airflow 0.57 mg/m' (0.13 require more time (Fig.12). The adsorbed wster vapor ppm) CH 1,27 0.4'C, and 50,71,and 91% RH. After 3

gther blocks se:ess ermethyllodide vapor to the TEDA equilibrium was rea:hed at es:h humidity penetradon or removes TEDA effectiveness by measurements were made at seven or 10.more minute impregnant hydrolysis:

intervals and averaged.

These values of the equilibrium CH 1 penetration 3

N + H 0 = N(CH CH )3NH* + OH-fra: tion P were related to water vapor concentradon in N(CH CH )2 s

3 3

3 2

air [H 0) by:

3 V,' hen the challenge of a vspor to a test bed is at a high enough concentration and continuous, the bed wDl P = exp (--47/[H 0]),

3 become loaded and will decrease in efliciency of vapor removal. The resulting increase in penetration with tirr:e Th's is consistent with the simple competitive rne:ha-

!s called a breakthrough curve. B 4:1through times for nism:

selected penetration fractions are often dependent on the cha]Jenge concentinuon. At high relative CH 1 + N(CH CH ),N - N(CH CH )3NCH3 + I-vapor 3

3 3

3 3

humidities charcoal beds become loaded with water vapor, also increasing penetration of test vapor with H O + N(CH CH ),N m: N(CH CH )3NH' + OH-3 3

3 3

3 time. Tre above experirnents have shown that for efficient sorbents at low cha]!cnge concentrations or where water vapor reacts with TEDA, making it un-1:adings, the relative humidity efTe:t may be the most available for CH 1 removal.

3 significant. Therefore, the time of esposure of a canister Larger MSA canisters, also containing 5% TEDA g

i cr cartridge to high humidity air is an important charcoal, were also measured for methyliodide penetra-paf arneter in a test procedure or a usage protocol.

tion at several hurridities. Flow rate was 64 L/ min and relative humidities ranged from 50 to 45%. Penetrations

2. Equilibrium Penetradons. A charcos] bed at equi-at water vapor equaibrium, P,, wue less sensiuve to librium with the water vapor in the air ente ing it does (H:Ol changes than in the case of the Scott cartridges

1

~

./

f

/

7l mth one fifth as much charcoal. Times required for fresh cartridges decrease with increasing relative humidity of MSA canisters to reach water vapor equilibrium varied the air passing through them. This is illustrated by the from 9.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> at 75% RH to 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> at 50% RH.

results in Fig.14. If 1% is chosen as the muimum At relative humidities above 75% the CH ! penetra.

penetration to be a!! owed, the service lives t, for fresh N

3 tion at water vapor equ2ibrium was not the highest MSA 5% TEDA canisters decrease from 635 minutes at

/

penetration value. This is illustrated in Fig.14 with 50% RH to 200 minutes at 85% RH. Another selected penetration fraction versus time curves for MSA penetration value would give another set of service lives.

canisters containing 5% TEDA charcoal. At 85% RH a For examp!c, tests of three fresh GMR 1 (TEDA) muimum penetration of 7.6% was reached at 450 carusters at 64 L/ min yicided the results in Table IX.

minutes as compared with an equDibrium penetration of An empirical relationship was found which described 4.1% (std dev - 0.2%). This muimun is attributed to the effects of relative humidity on service lives (t,) of t!'e displacement by water of CH 1 previously physically fresh canisters. Log t, versus log [H 0} (or los percent 3

adsorbed. Such a muimum is commonly seen at au RH) plots were found to be linear with slopes between 2 humidides with various charcoal beds, and 3. Fig.15 shows such plots for MSA 5% TEDA Conclusions reached from studying humidity effects canisters at 64 L/ min airflow, Scott 5% TEDA canisters for equ0ibrated carusters are the fo!!owing:(1) We now (600252 75) at 641/ min airflow, and Scott 5% TEDA understand how water vapor reduces the e!Ticiencies of cartridges (642 TEDA.H) at 32 L/ min airflow. The two TEDA impregnated charcoal beds. It is by tying up the brands orcanisters, which have nearly the same volumts impregnant and making it unavaDable for reaction with of charconj, had equivajent service lives (Fig.15). Even methyliedide. (2) Long times. up to 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />, required to at half the airflow rate the cartridge with much smaller p

reach humidity equUibrium limit the practicality of using charce:J volume had much shoner sersice lives which penetration at humidity equilibration as a measurement were more serious!y alTected by humidity. Similar data of canister performance. (3) Since the penetration at high with MSA GMA canisters containing urumpregnated humidity equilibrium is not the highest value which activated charconj were also !!near on a log t, versus log occurs,its usefulness for canister or carvidge perform.

[H 0) plot.

3 ance specification is questionable.

The usefulness of this relationship for a cenification L

program is for extrapolating from one humidty to 9

3. Senice Lives. Measurement of senice life, the another. It may allow the selection of two humidit time required to reach a selected penetration fraction,is conditions for evaluating canisters and cartridges. Also, an alternadve to measurement of penetration at humidity the selected test humidides could be high (70100%)

equilibrium. Senice lives of air purifying canisters and where senice lives are shorter and where, therefore, test times would be shorter. This is desirable for mujmum efficiency of a certificadon test program.

TABLE IX. Service Life ( 5 min) at Three get g Relative Humid' ties E

75: M g

gg M Penetradon Percent Relative Humidity Q)

Fraction 60 80 100 W

i 53 M 3

0.0002 145 85 30 f

[

0.0005 185 105 35 0.001 225 135 45 g

0.002 275 155 55 q'

0.005 375 195 65 0,1

' b

' ;c '

0 70 9h 01 455 20 M

gg 0.02 555 245 95 0.05 705 305 135 Fig 14. Hwutty efree on eneGyliode bnaharogh survu at 64 0.1 855 475 235 t/ min for MS A 3 % TED A canned

}

/

/

RELATIVE HU:1!DITY AT 25'C Measureable heating (4.2'C) continued for periods up l'J 20 40 60 80 to 340 minutes for this series. Other cartridges and

[

3 canisters (64 L/ min) showed similar heating effects.

'E Dew points were measured for air leaving test 5

cartridges as we!! as for air entaring them. This allowed g

determination of the rates of' water vapor adsorption at times throughout an experiment. Temperature increases were proportional to water vapor concentration de-100 :-

~:

creases. This relationship was used to calculate heats of h

i adsorption ranging from 4 to 9 kcal/mo!c.

g Humidity heating effects are important to note since (1) they can make air purifying respirators less com-fortable to wear and (2) they complicate the descriptions of how canisters work. The comfon effect is more

....,,,,i relevant to the user than t the certi0 cation test.

10 100 WATER VAPOR PRESSURI (tons)

Fig.13. Comf adons of re!adve humidity and eartice lift (1%

B. Temperature bewahhrovsh) for cardsters (::. Scon and 4 M5A) and a cartn4 ( 0. scoa) contawns s% TEDA charcoals.

3. Equilibrium Penetrations. Ambient air temperatures for applications of air purifying respirators (4

can vary. In addition, as mentioned above, air drawn

4. Hun.ldity Heating. The adsorption of water' vapor through a canister can be heated by water vapor from air passing through charcott packed cardsters and adsorption on the adsorbent. Temperature effect2 can be

cartridges heated the air to significant extents for long complex since higher temperatures enhance chemical [

periods. Temperature rises for Scett canisters (642-reactions (chemisorption) with impregnants, but reduce TEDA H) at 32 Umin airDow and three humidities are physicaj adsorption of vapors.

shown in Fig.16. The maximum increase of 10'C In a series of experiments with Scott eartridges (642 (18'F) was observed for the highest (85%) relative TEDA H) at 32 Umin airflow, temperature of enterins humidity at about 5 minutes from initial exposure, air was varied from 26.4 to 38.0'C. Test cartridges were equilibrated at dew points from 15.1 to 25.7'C before methyl lodide penetrations were measured at the equj-libration humidities (50-75% RH). Fig.17 shows a plot 12 4.6 r-U10 0

w q

E,q

@3 w

85! M h6

{

+-

(4,2 m

~

WM

~

q,0 2

301 M i(,

pe"' w _.

3,8 0

0 10 23 30 0

50 60 3.20 3.25 3.30 3.35 3.0 11Pt (min) 3 10 /T (*K~I)

F416. Humidity bestirs ofrecu for a Scott 3% TIDA carm4e at Fig.17. Ciamron plot to servetste equn Wm meht)I 31 !./enin aircow.

bias Fnetr:Uon Mth ee TWnt and air um;writures.

/

,/

/

7 O

3 c( In (-[H O] In Pg) versus 1/T, which turns out to be Company (MS A) canister containing charcoal with two

,g knear with a slope of 3020'k.

impregnants,2% TEDA and $% K1.The breakthrough 3

s, The usefulness of this data la in sorting out the curves show increased penetrations for incrossed interrelated efTects of relative humidity, ambient temper-temperatures at all experimental times. When logarithms ature, and dew point temperature. The foUowing semiem-of percents penettation were plotted asains pirical equations were derived from data for this temperatures, ws obtained apparently straight lines. Fig.

cartridge:

19 shows such plots for the MSA canister (2% TEDA,

$% Kl ) and a Scott 642 cartridge ($% TEDA). Simuar 3

$300 3020 results were obtained with an MSA GMA canister in P = -0.0096 t, exp T*** "" ~T""

Containing Unimpregnated activated charcoal. In su three cases, the temperature effects corresponded to approx.

Imately doubling the penetration for each 5'C (9'F) 2280 In P = (0.96 t,%RH) **P T* 8 8 t".

The params'ter cf air temperature (T), relative ine ease in tempersture.

8 I

humidity t'RH), and dew point temperature (DP) are The exponent $300/Toi, poi, comes from the de*

interreisted and cannot independently affect senice lives pendence of saturation vapor pressure of water on (breakthrough times at selected penetrations). Therefore, temperature. At constant dew point, i.e., at constant we have studied temperature effects first with RH water vapor con:entration, equ0ibrium penetration of constant and then with DP constant. As before, methyl methyllodide decreases with increasing ambient temper

• iodide is the test vapor, since it is the most penetrating sture. However, when relative humidity is held constant, vapor form ofiodine we have found.

O

• • "'i"*"*"'""*"'""'"*"''''h"'m At constant RH,inereasing temperatures shift break-bient temperature. This is due to the higher water vapor through curves to higher penetrations and, consequently, concentration for fixed humidity at higher temperatures, result in shorter seni:e lives (Fig. 20). Canisters and j

cartridges containing three types of charcoal ure t

2. Penetration and Senice 1.lves. We have done studied at constant RH. The result (Table X) show experiments to determine how much effect the tempera-gj,,tificant service life decreases with increasing ture of inspired air has on the efficiencies and senice

^

temperatures, up to 13% de:rease per 'C (8% per 'F).

lises of fresh (unequilibrated) canisters anc cartridges for methyl lodide. First, we observed thai su:h an effe:t does exist, even over a limited expe:ted range of use 0.4 temperatures (!!.33'C or $9 9$'F). This is iUustrated in i

e i

$c[c $ng 2m Fig.18 by data obtained with a Mine Safety Appliances 32 Uml=

10 1 {

$;; p

]

i i

i i

i e

i i

g

),

s g

~

C 35 *C W 0.2 g

31 'C t

2 4

1 2

5: X1 )

]

E

{

]

3 O

tr a

o g

64 Umin g

50: M 26*C g

2M

\\

t t

t t

i i

, g,y

(

0 1

2 3

4 5

6 7

g 9

10 23 25 33 35 40

~

M ("I TU tt M M.E ('C)

Fle. II. Air temperature efreet on methyl Iculide bredthrough surves at 64 t/rmn and 30% RH for an MS A 2% TEDA. 3% KI, Fig.19. ENects of air temstroture on methyllodide Penetrsoon at 84 miser.

selected times after irutiaung 11ows of 30% RH ad.

l

e

/

/

/

/

/

400

'.2 5,300 h

O td 0

3:

A 200

~.a g

100 OM (E

to C/J 05 0

20 30 40 AI,R TEMPERATURE (OC)

Fig. 20. EEect of alt tem;catur,y en service Ives at connut 30%

RH for sa MSA cuiser G% TEDA.3% KJ,) at 64 Urnin..

'4 TABLE X. Temperature Effecu on Service Lives at Constant Humidity Charcoal Percent T Range Percent CH 1 Senice Life Decrease 3

Type RH

('C)

Penetration (Percent Per'C) 2% TEDA, $% K1

$2 30 35 1

4 3

50 31 35 1

4 26 35 0.5 7

26 31 0.2 15 Activated 30 25 34 1

7 70 25 30

$0 25 30 10 25 30 1

3% TEDA

$0 29 38 I

9 in one case (2% TEDA, $% K1 ), the temperature effect turn out to be the case. Average increases in senice lives 3

varied with peneustion fraction selected to denne senice life.

at 1% penetration were 12% per 'C at 19'C DP and 3%

per 'C at 24*C DP.

At constant DP (1.c., constant water vapor concenua-The :en:!usica of these studies h that temperature can tion in air), senice lives increased signincantly with sITect service lives much more than the 110% reduction increasing temperatures (Fig. 21). This la due to the per 10'C increase reported in the literature.' Therefore, combined efTects ofless water adsorption (air /charcoa]

temperatures at which cartridges and canisters are tested j

equ0ibrium shift) and enhanced reaction ormethyliodide must be more closely conuo!1cd than t2.$'C speciSed with triethylenediamine (TEDA) impregnant. The main in CFR 30. Part II.* Also. the units must be tested at reasen for doing constant DP studies was our hope that maximum T and RH of expected use or tested silower temperature efTects would be less signincant. This did not values with exuspelations of service lives to the uorst \\

i

/

/

/

/

/

/

/

7 n

eax

\\

^

.5 E

,a.

set 100 -

g

s E!

g w

O M

NO 40 AIR TDdPERmmE (* C)

Fis. 21. Errect of air temperatun on seence Ilns at constam 24'C dewpoint for ea MSA c4:uster (2% TIDA. 3% K!,) at 64 Umm.

case conditions. Users must be made aware of the potendal for reduced senice life when these units arp D. Reproducibilldes of Senice Ufe hiessurements used at even more elevated temperatures.

The question was raised as to what reproducibilities can be expected for senice life determinations, consider.

C. Flowtate ing variabilities due to manufacturing and testing. Re.

sults oflimited studies are shown in Table X11. Precisica A canister with 2% TEDA, 5% Kl, impregnated was worst for high protection facters and short times charcoal was tested at 30'C and two RH's, DP's, and where bed deterioration due to humidity was most rapid.

airflow rates. The results shown in Table XI clearly These results and other experiences indicate that at 1%

indicate that senice life is inversely prcportional to penetration for one bat:h of cartridges or canisters a \\

airflow rate. This confirmation of a well established reproducibility of 10% relative standard deviations of L senice life is reasonable. Reprcducibility between re!:tionship was necessary, since in this case senice life is determined by water vapor loading rather than the batches can only be determined with more extensive testing.

contaminant (methyl iodide) vapor loading.

TABLE XI. EfTects of Altflow Rate and Humidity on Senice 17fe' t

Service 1.Ife (25 min)

Flowrote Dew Polnt Relative Humidity at Pennradens (L

f

, / min)

('C)

Percent i Percent 10 Percent 32 19 32 173 300 64 19 31 80 32 24 71 113 195 64 24 67

$5 105

'_J0'Ct 2% TIDA, $% K1, caniners.

6 L-h a

,l j

l

/.l i

/

/

[

TA LI XII. Re,mfucibilices or Ser*e tXs Menswvments Semme LE W

[

Charvoel Temprrstwo % RelasJve Flyerste Nwntier of Pwan:

g g,i,g, Trp

('C)

Med4,ry (1/edn1 maturemreu Penetradon A, wise suL Dev.

But. Dn.

S% TEDA 3e 70 32 6

3 46 6J 14 3

(4 4.6 7

10 M

4.2 4

2e 130 12.2 s

l 2% TEDA.

30 10 64 3

8 124 10.4 a

ss 10, s% TEDA 23 90 32 3

0.02 23 24 H

tes as is 3s o.:

ss e

0 0.2 80s l0 10 0.3 132 7

s l

VI. EFFECTS OF CYC1.!C FLOW (BREATHING breathing cycle (work load) were measured for cajeu.

PATTERNS) lated and related to instaataneous emeiencies throughout the cyc!c.

A. Background

]

The effects of variable now rates and now (breathing)

B. Computer Modeling Study

(g patterns on the average ef0ciency needed to be.de.

termined also. Evidence is available (Section IV. C.' that A computer program was *Titten which could cajeu.

the ef0ciency of a sorbent bed for removing vaport from late canister penetrations of methyl iodide based on air decreases with increasing airdow rate. There are data assumed airnow rate and reaction kinetics of removal.

which demonstrate, among other things, that peak Breathing patterns of airdow were taken frort the work inspiration now rates can be very h!sh (200 Umin) at of Sijverman, et aj..' who mmured and characterized moderate work loads." Therefore, at the peak of inhalation and exhajation cur.'es at ten work rates for inspiration in the breathing cycle, the ef6ciency may be res' stances approximating those of gas masks and other very poor and certainly will be very different from that at breathing apparatus. Their Table 4 provided four the standard 64 Umin test airdow.

pr imeters used to simulate the varying Dows:

The work of Gary Nelson"is widely misinterpreted as R = respiration rate (per minute) showing no such effect. Actually, he demonstrated only A = maximum inspiratory DowTate (Umin) that, in a limited number of cases, the cartridge capacity I

= fraction of total cycle that is inspiration (lifetime) was unaffected by airduw rate or cycling. In the F,,, = minute volume, mean inspirstory Dowtate over f'

cases of highly toxie vapors (radiciodine, other radio, an entire breathing cycle (Umin) nuc!! des, carcinogens, etc.) the serbent bed ef0ciency, OrJy the inspiration now was considered since exhala.

rather than its capacity,is the limiting factor in determin.

tion is usus!!y through an exhajation valve, rather than ing usefulness. This is because only low levels in air are through the canister. During expiration, now through the expecteu' to be encountered, resulting in low bed loading.

canister was set at zero.

Experimental measurements and theoretjeal computa.

The equation which best Ot the experimental breathing tions were done to identify and quantitate the effects of curves was a linear combination of sinusoidal and cyclic Dow patterns. Average efficiencies for a si$en el!!psoidal functions:

1 22

?

Y

/

~

nt f 2t i

8" l

's I -l - Il F = A sin

+A where i

\\

)

c Ostst P = penetration fraction of methyllodide i

k = first order rate coefficient for remova)(per see) where t, = bed contact time (sec) = 60V/F V = charcoal bed volume (L)

F.F.

= instantaneous and average volumetric Howrates (Umin)

These equations were combined to c Jeulate instan.

t,

= 1/R, the average time for inspiration (min) taneous and integra] Dowtates and penetration fractions A,, A = constants scJected for each work rate to match for steady and cyclic Dows for selected values of k and V the experiment:J vajues of maximum nowrite' in the ranges of experimental values. The constant A, such that flow 7ste required to give a penetration equal to that of the nclic flow was also calculated.

The simplest case, where k is a velocity independent A=A E*

8 + A and F" =

A 8

8.* I 8 1

A (n

j Table XIV for k = 17.6 s'i and V = 330 cm, both 8

experimental vajues (Secuen !Y. E.). Average cyclic now Table XIII lists the input parameters and calculated pene rations were much higher than those for equal values of A,, and A for ten workrates.The(trapezoida))

steac'y Dowvates. Higher steady Cowrates (2.0 to 3.6 integrated flowTates for the best fit curves are given in the timeA) were required to give penetrations equal to the last column.

cyche flow penetrations. The efTects of varying the A second assumption was that the canister was product k V on the cyclic penetration (P ), steady

.g equilibrated at a set of temperature and relative humidity

,- etration (P.), and ratio were given by conditions, where the removal of methyl lodide was described by first order kinetics (Section IV. C.):

In P, = -< (k V).

P = exp [-kt,)

In P, = -s (k V).

P/P, = exp ((s - c)(k V)),

TABLE XI!!. Input and Cajeujated Parameters for Fitdng Experimental Breathing Curves Aversge MaxJmum Respiration Best Fit Constants e

Integrated Tate Flowrate Rate Inspiradon A

A, Flowrate i

j h)

(Umin)

(per r'Un)'

Fractfon (Umin)

(Umin)

(Umin)

L1 37 14.8 0.382 33.20 1,79 9.0 83.2 44 17.5 0.431 26.42 17.38 13.1 19.8 60 18.7 0.4 64 29.92 30.08 19.6 27.0 78 20.1 0.479 32.89 4 $.11 26.7 28.2 79 22.0 0.487 27.83

$ 1.17 27.9 36.2 101 22.5 0.490 36.62 64.38 35.8 A8.9 128 27.4 0.312 33.76 94.24 48.4 64.4 160 32.3 0.319 10.61 149.39 63.7 81.3 192 34.2 0.539

-0.26 192.26 80.4 90.3 240 42.0 0.514 86.13 153.87 89.3 v

~

J

,1

/

.f

/

/

[/

ThBLE XIV. Cyclic.and *;teady Flow Resu:ts Calculated for a Constant Rate Coemeient' Average Penetration Fracdon Equivajent Steady l'1owtate

. Integrated Flowrate (Umln) Cyclic Bow Steady Flow Rau RowTate (1/ min)

Ratio 9.0 2 x 10-'

2 x 10-"

l x 10" 32.2 3.58 13.1 1 : 10-*

3 x 10-"

3 x 10' 38.2 2.92 19.6 0.M ' l 2 x 10-8 5 x 10' 50,9 2.60 26.7 0.0 W 7 2 x 10-*

2 x 10' 64.5 2.43 27.9 0.0051 4 x 10-*

1280 66.0 2.36 35.8 0.0145 6 x 10-8 244 82J 2JO 48.4 0.0340 0.0007 46 102.0 2.13 63.*

0.0675 0.0042 16 129.3 2.03 80.4 0.l N 6 0.013:

S 15 A.4 1.92 89.3 0.1423 0.0202 7

178.7 2.00 8

_= 11.6t*'. V - 330 cm

• k where for each we krate c. and s are average valuts of phenomenon."'" Including the selocity, v (err /sec).

60/F for cyclic and steady flows, respectively. The dependence, penettation ratia is a function of k V and, therefore, a function of the per.etration fraction (P, or P,). At the in P = -k v"V/F penetration fraction of most irterest for dstermining Gartridge service life, P, = 0.01, and at a total breathing where k is a "true" constant. Whee!cr has shown rate of 64 Um'. (32 Umin through each of two theoretically that the s aJte of n should be 0.5 for the case cartridges of volume 165 cm'1, k V = 4.962, P, -

of a mass transfer. limited rate.8 Dietz, et al," obtained n 0.000087, and P,/P, = ll5.

values of 0.45 t t, 0.58 for their hex.

We experiments were showi'ig much smaller cyclic at..ethylnetetramine/ iodine / sodium hyJ.txide.im.

flow c' Tecti see belov), the nmplest model was modified pregnated charcoals. The data of May and Polson" has to ine!ude velocity dependence of the rate cocmcient.

been used to calculate the n vah.es for a 5 per cent Dietz, Blech!y, and fonas observed a continear increase TEDA impregnated charconj shown in Table XV. Th.se of the first order rate coemclent with vieresses in linear results show a relatise humidity dependence for n,which flow velocity for methyl iodide removal by impregnated ranged frr.m 0.23 to 0.42.

charcoals." Others have also observed this TABLE XV.

? ("~%-!cnt Velocity Dependence Calculated from the ay ant,.') wn' bg.L g Lust Squaru Fit l

Percent CoefL Relati:

dumber Correla.

s

.bn Humidity tt

<f Points n

(r')

50 11 c'

4 0.42 1.000 e

17.7 101.1 4

0.38 03.9

' ((

b.

17.3 103.2

.3 8

0.29 0.970 37.5. 01.5 21e31.8 4

0.23 0.994 f

3 l

' L..'..tnce 13.

• k v'.

J i

f Tht computer program was modified tolac!ude linear Cyclic / steady penesation ratios for P, = 0.01 varied veJocity dependence in the above penetration equatkn.

from 2.49 to 5.13. Steady / cyclic lbwrate rat.ios for P, =

Raults shown in Table XVI were computed from n = 0 P, = 0.01 ranged from 1.90 to 2.76, with a!! but the to 0.75, average F = 64 Umin (32 Umin through each of lowest two workrates in the range 2.0

• 0.2. The two cartridges), and k V selected so that P, = 0.01, flowrate rat'os for the extreme case of n = 0 were also in Computed squivalent (P, = P = 0.01) steady flowtates this range for cyclic breathing rates of 48 Umin and are listed in the last column.

above (Table XIV). The cyclic / steady flownte ratio :

The value of a = 0.7 was used with the computer less variable :han the penetration ratio, and is a possible program and the parameters of Table X111 to calculate way to take into account cyclic flow efTects.

the results in Table XYll for tan workrates.

TABLE XVI. Computed Values for' Selected Yelocity Parameters

• _ _ __

Steady Average Penetradon Fraction Equlvajent Steady n

Cye :e Flow Steady Flow Ratio Flowtate (Umin) 0.00 0.01 8.7x!0.s 64.6 0.50 0.01 0.0019 5.3 38.9 0.67 0.01 0.0036 2.8 58.1 0.70 0.01 0.0040 2.5

$ 8.6 0.73 0.01 0.0046 2.2 59.3

' Assumed 32 Umin through u.d of two carvidges and breat.ing curst correspondirig to 64 Umin total flowrote.

+-

TABLE XVII. Results Computed a for Yelocity. Dependent Rate Coefficient

• and One Percent Cyclic Flow Penetration Integrated Aurage Penetradon Fraction Equivalent Steady Flowrate Flowrote (Umin)

Steady Flow Ratio F'owtate (Umin)

Ratio 9.0 0.0019.5 5.13 24.8 2.76 13.1 0.00254 3.94 31.2 2.38 19.6 0.00296 3.38 42.8 2.18 26.7 0.00318 3.34 56.0 1.10 27.9 0.00328 3.05

$ 7.5 2.tr6 35.8 0.00335 2.99 72.9 2.04 48.4 0.00364 2.75 93.6 1.93 63.7 0.00375 2.67 121.2 1,90 80.4 0.00402 2.49

46.8 1.83 89.3 0.00378 2.65 169.2 1.89 c

6.. i vv. v. 33o es.

/

)

f

/

[/

The computer program was modified to include linear velocity dependenet b the above penetration equation.Cyclic / steady penetration ratios for P, = 0.01 varied Results shown in Table XVI were computed from n = 0 from 2.49 to $.13. Steady / cyclic Dowrate ratios for to 0.73, average F = 641/ min (32 Umin through each of P, = 0.01 ranstd from 1.90 to 2.76, with all but the two cartridges), and k V selected so that P = 0.01.

towen two worktates in the range 2.0 + 0.2. The Computed equivalent (P = P, = 0.01) steady flowratesOcwrate ratios for the extreme case of n = 0 we are listed in the last column.

this range for cyclic breathing rates of 48 Umin and The value of a = 0.7 was used with the computerabove (Table XIV). The cyclic / steady flowrate ratio is program and the para;neters of Table XI!! to calculate less variable than the penetratien ratio, and is a possible way to take into account cyclic flow c!Tects, the results in Table XVil for ten workrates.

TABl.E XVI. Computed Values for Selected Yelocity Parameters *

=

Steady Aserage Penetration Fraction Equivalent Smdy n

Cyclic Flow Staady Flow Ratio Flowrate (t./mi]n 0.00 0.01 8.7 x 10-8 113 64.6 0.30 0.01 0.0019

$.3

$ 8.9 0.67 0.01 0.0036 2.8

$ 8.1 g

0.70 0.01 0.0040 2.5 38.6 0.75 0.01

• 0.0046 2.2 39.3

' Assumed 32 t/ min through each of two cartridges and ',reathing curse corresponding to 64 L/ min total Courate.

=

TABl.E XVll. Results Computed a for Velocity. Dependent Rete Coeflicient* an Percent Cyclic Flow Per'ttration Integrated Aserage Penetra:Jon Fraction, y,,j,,, steady Flowrate Flou rate (l ' il Steady Flow Ratlo Flourate (Umin)

Ratio 9.0 0,00195 5.13 24.8 2.76 13.1 0.00234 3.94 31.2 2.38 19.6 0.00296 3.38 42.8 2.18 26.7 0.00318 3.I4

$ 6.0 2.10 27,9 0.00328 3.05 37.5 2.06 35.8 0.00335 2.99 72.9 2.04 48.4 0.00364 2.75 93.6 1.93 63.7 030375 2.67 121.2 1.90 80.4 0.00402 2.49 146.8 1.83 89.3 0.00378 2.6 $

I69.2 1.89

'k. 4 k V." V = 330 cm'.

{

I

i

,/.

/

/

/

Additional experiments were done with other C. Ezperimental Study

/

charconJs to explore the generality of the cyclic now A Scott Breathing Simulator (Scott Asiation Corpor.

effect. Compatisens of pnetrations at cyclic and steady g

ation, Lancuter, NY, Part No. 800116) was used to tow conditions (32 Umin) were done with several produce cyclic flow partems, it la a dual.pisten pump cartridges, canisters, and beds packed with 32 53g of opratad by a motorized cam to simulate a breeGng charcoal (7.5 cm-diam). The measured cyclichteady panern. The cam used in these experiments was des!g.

peneustion ratios and experimental conditions are given mated 622 XOM. Total volumetric Dowrate of air was in Table XIX. Figure 22 shows breakthrough cunes adjusted with the pump speed control and measured obta!ned for two 5% TEDA charcoals, one which downstream of the test bed with a dry test meter (Singer showed a definite cyclic flow effect (1.6 times higher Model DTM.325) oser at leut 20 cycles. A series of pnetraJens) and one which did not. Likewise for the check vahes and a filtered air supply was used so that other charcoals, seme showed an effect itad others did during half of the cyc!: backnow through the cartridge not. There was ra obvious way to predict when the I

was prevented. This was to simulate one.way (in pira.

cyclic flow would give a higher methyllodide penetration tion) flow through a cartridge on a typical air purifying or how much higher it would be, respirator. A 20 L polyethylene bottle was used to nix The significance of the cyctic flow efrect is seen in Fig.

enethyl iodide from the permention tube with the main 22 for the Scott charcoal. There appear to be two aircow to smoeth out possibic variatiens in cha!!enge separate breakthrough curves differing by the factor of concentrations due to cycling aircow. Total 0ow volume 1.6.The end of ser ice life, denned as pneustion reach.

over a time corresponding to a full number of cycles ing 1%,is 300 min for standy airdow, but orJy 180 for 20-25) was meuured with a dry test meter downstream cyclic airnew at 32 Umin. At 1.3% the differences are g

(of the test bed to determine average flowrate. Other than nuch greater (>>!00 min vs. 260 min).

these mcdificatione, the experimental apparatus was the A maximum penetration at about 120 min was same as that described in Sections !!! and Vill.

observed for the Wil! serv!nco J% TEDA charcoal Test beds of 7.!-cm iam and 0.5 to 1.5-cm. depth breakthrough curat (Fig. 22). This has been seen before were packed with varying amounts of 5 preent TEDA.

for this and other ch.ttcos!s and is attributed to bed impregnated charcoaj (Bunebey Cheney CU 2762) to heating by water vaper adsorption (see Secuen V.A.d.),

obtain a range of per.etraion fractions at sa!ceted Both the cyclic Ocw ud steady flow penetration fol.

relative humiditits (25 95 per cent) and ambient temper.

Iow ed the sarne breakthrough curve.

sture (23 : l'C) Air 0cw was maintained at 32 Unin for both steady and cyclic (20.5 cycles / min) situations.

After the test bed wu equilibrated at the selected relatise D. Conclusions hurnidity, penetration fracticas of methyl iodide were determined at 10. min intervals, ajternating steady and Both the computatienal and experimental approaches cyclic flow for about 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> periods. The resulting to determining cyclic / steady penettstion rstics led to the meuurements were aseraged and a standard desittion same conclusion: significant differences between break.

was cajeulated. Altemating Cow petterns for the same through fractions at a w.ceted bed conditiens (tirm of test bed e!iminated the sit.nificant between. bed variations use, humidity, temp"sture, averags Cowtate, etc.) can esist for a variety of :harcoals. The limited experimenta) expenenced in earlier experiments.

Results sur. mar.ced in Table XYll! show definite and computed data aegired so far does net reves) the -

pencuation difTerences for the cyclic and steady flow factors determining the esistence or magnitude of this cases. Cyclic Dow pnetration was greater by factors errect for any sisen charcoal. It is of great interest for a frem 1.2 to 4.2. However, there was no consistent variety of respirator ar;1isst'onso net merely tr& iodide removal, to identify these critical parameters. The ec. '

cortelation of rstio with penettation fraction, contrary to

(

the computer csJculations. This suggests that the rea) is,ence of a cyclic flow effectimpacts on manufactu* ins.

situation is complicated by unknown factors (e.g rela-testing, certi0 cation, appresa!, and use of chede*J air tise hurnidity, granu!e sire, packing density, bed depth, cleaning respirator cartridges and canisters, etc.) having unknown effects.

1

l l

TABLE XVill. Experimental Results of Cyc!!c Flow Study at 321/ min J

for one 5% TEDA charecal(BC CU.2762)

Relative Cyclic (C) or Average Standard Ratlo Humidity %

Steady (S) Flow Penetration Deviation (C/S) 25 C

0.023 0.003 4.2 3

0.006 0.001 C

0.145 0.019 1.3 3

0.112 0.010

$0 C

0.045 0.00$

1.9 5

0.023 0.002 C

0.076

• 0.013 1.7 S

0.045 0.005 C

0.086 0.007 1.6 5

0.0$$

0.012 75 C

0.304 0.027 1.3 0.238 0.022 86 C

0.081 0.003 2.3 0.032 0.003 C'

O.067 0.008 3.5 S

0.020 0.001 95 C

OJ12 0.012 1.2 5

0.443 0.028 C

0.177 0.008 1.7 S

0.101 0.007

==r -

1 An immediate concern for this proje:t 9 ho w to take The se:end option. doubling the steady a.

w,is into ac:ount cyclic flow effects on methyllodMe penetra.

based on the computed resdt in T'.bles XVI and, t'll, s, t!on.Three options have been considered:

which show that this approximately compensates for

!) Derme end-of senice life of a canister er cyclie flow at a vanety of conditiens. Unfortunately, the gJ certirtse at a lower penetration fraction (e.g.,

experimental resu!ts reveal a more complicated situation.

0.5% instead of 1%).

This option also has the same obje:tions as the first.

2) Deuble the testing airCow rate.

Humidity effects wodd be me than doublei by

3) Use a breathing simulator pump for testing, douFing the air 0cw rate. And modif1:stions would be The first option +cdd mquire identifying a constut required for the existing testing apparatus. The third cr average cycli:/tt:acy dew penetration factor. Tables option also would require modifying the esist'ns XVill and XIX show a signincant range for this ratio.

apparatus by 1) inserting a breathing simdator i

Furthstrere, this option wedd penalize those manufac.

pump and one way valves between the futered air supply turers asing charcoah %ich have no such e! Test and and the humidity chamber,2) inserting a 20 L buffer (Q

may hinder the development nf more effe:tive sorbents.

volume Wtween the methyl fodide generator and the test For examp!,the WCtioMnce en tridge in Fig. 22 w ould bed and 3) changing the method of everage flow rate s

have failed (> l%) entirely ir: stear of having s senice life mortitoring (volume: tic average instead of an instan.

of about 45 min.

taneous Dowmeter readout). The question remains as to A

m 14J m.

' ~.

/

/

l j

i

?

I TABLE XIX. Cyclic / Steady Flow Penetration Ratlos for a Variety of Charcoals.

Charcoal Original Packed Preconditioned? Run Pelative Number of Average Cyclic / Steady Source and Type' Urdt Bed'

(% RH)

Hurnidity (%)

Comparisons Penetration Ratio Norton OV Cartridge.

X No 75 2

1.0 0.1 0$001)

X No 85 2

1.0 t 0.1 MSA OY Canridge, X

35 30 3

1.3

• 0.1 GMA (44.33) 3 1.0 a 0.1 MSA OV Canister, X

$0 83 GMAC MS A Cardster.

X No 85 4

1.0 t 0.1 GMR 1, $% KI, MSA Cardster,'

X No 85 4

1.7 2 0.2

$% K1,2% TEDA 3

MSA Cardster' X

95 95 4

1.0 0.1

$% TED A Sc:n Canister, X

No 83 7

1.6 s 0.1

$% TEDA (600232 75)

X 85 95 2

1.7 2 0.2 Barnebey Cheney X

No 85 1

1.9 e 0.2

$% TEDA (CU.2762)

X 85 85 4

1.8 s 0.2

%"t!!sorv1nco Cartridge,'

X No 85 2

1.320.1

$% TEDA (Lot B)

X 95 93 1

1.3 0.1 Willson/Inco Canridge,'

X No 85 4

1.0 t 0.1

$% TEDA (Lot V)

X No 85 1

0.0 20.2 X

No 83 2

1.20.2

'impregnants and amounts based on manufacturers'information.

'Protot>ps suppted by manufaeruters.

'$$ g in 1.5 cm4iam bed.

I I

(

f 7

/

.h

/

This compound has a normal boiling point of 174'C, r

sm.

j 34

'y but is known to sublime readDy at room temperatures.

P.

4 * **

A The volatility of the pure crystals has brought up the g

A question of the volatility of TEDA trnpregnated in C

acdvated chuecals. The reason for this concem is the E ** * -

possible release of significant amounts of this amine of unestablished toxicity from serbents, especially in air

,e

,/

purifying canister applications.

em.

pe There are no toxitological data avaI!able for TEDA; however. TEDA belongs to the class of organic aliphade

' " e (' a amines, many of which have been shown to be toxic.

m m m au soo su.ao.ao a no Threshold I.imit Values (1982)" for similar emines are:

to m g/m' ppm s ois.

i Ethylamine 18 10 g o eie.

e Diethylamine 30 10 y

,. ;,.c Triethylamine 40 10 8

g Etylenediamine 15 10 Diethylenetriamine 4

1 By structural and functional simi!rrides. TEDA can be am e

u m m m ano =

3" considered moderately toxic with a concern level of 1 O

ex>o vac 1mc t.e.

p,m o, greaier. vapor,rc,su,es measured over the rig. 22. Breshhrough curves s!terristifts steady ftow ( o ) and eretic range 50110'C hase been extrapolated to give 0.58 tort flow (t.) for r=o 3% TID A charcoals. UFPer curvt: scon ($!g. 7J.

at 25'C. However, there was rio informadon avaDable em sam). l.c=ct curse: Whordaco (3:g. 7.0<m 6am),13 % RH.

on the volatDity of desorption rate of TEDA imptegnattJ 32 m on activate chuecal.

70 supply data to answer these concerns we have measured TEDA desorydon from commerical im.

which breathing curve parameters to use and how critical pregnated charcoals.

are they to the final penetration results.

The third option appears to be the most desirabic, at least until more data is deseloped. The selection of B. Apparatus and Procedures parameters would be no more arbitrary than the ulec.

tion of an avezage breathing rate, usu Dy 641/ min, for The apparatus used for measuring TEDA desorption testing. The data of Suverman. et al,'is avaDable to make is diagrammed in Fig. 23. The detector for TEDA in air these ulections less arbitran, was a pho r> ionization detector through which air sam.

pies are drawn. Detector response wss ampLfied and attenuated with the electrometer component and re.

Yll. DESORPTION OF TEDA FROM IM-corded on a strip chart. The detector was calibrated by PREGNATED RESPIRATOR CHARCOALS sublimation of TEDA crystals at 30.0*C into flowing air. Weight loss rate and dDuent air flow rates were A. B ackground measured and used to calculate calibration coecentra-tions.

Data reported ear!!ct (Section IV.) have been shown A gas chromategraph oven was used to control

(

that TEDA [N(C H.),N]i an effective charcoalimpreg.

temperatures (70-120'C) of test beJs. the air entering cant for the trapping of orgartic forms of radiciodine them, and the sampling lines. Charcoal samples of I 4 from air. Four canister manufacturers have plars to cin' volume were packed into stainless steel tubes and continue or begin packing their radiciodine canisters held in place by glass wool. This resulted in bed depths of

• ith 5% by weigth TEDA impregnated charcoals, 1.4 3.6 cm.

f was also tested under the same conditions to provide a

/

p.,,,,,,,,,

/

n Photoionization reference r.nd to identify any iodine release upon heating.

Pump.

t j oelector I\\ l Onn OO C. Results and Conclusions o

1 1

y o

Vent Since breathing through a respirator cutridst is net at Blonk Test a fixed, constant now rate, we first studied the effect of Bed sed airnow velocity. The results showm in Fig. 24 for one of the 5% TEDA charcoals show the absen:e of effect of now rate over the range 1.6-6.8 err /s. This implies that the air was TEDA saturated and the volatilization rate Recorder eered Air was itpid.

Since bed depths also vary for different designs of Fig. 23. The apparatus maad for eru4es or TEDA dewrydon from cartridges and canisters, we also vuicd this puameter.

et.arcents.

Again, no effect was obser,cd (Fig. 23). This result, combined with no velocity effect, implies that the air Compressed air from cylinders was passed through a passed through the impregnated chuecal was saturated fi!ter of activated charcoal before use. It was quite dry with TEDA vaport in other words, at equilibrium.

initially. For higher humidity studies a fraction or all of Humidity was also varied over a range from 5% to the air 0:w was passed through the headspace of a water 99% RH at 25'C. Dew points of-18 2 4 ' C, 15.1

() reservoir. Resulting relative humidites were determined 1.4'C, and 24.8 0.7'C were measured at test using a dew point hygrometer, conditions of 70'C,90'C, and !!O'C. At 25'C these Two chuecal beds were placed in the oven in such a dew po:ati correspond to relathe humidities of 4%,54%,

way that the airnow could bc swit:hed by a valve to and 99%. Increasing water va'>cr concentrations de-dther. One bed contained unimpregnated activated crtased the response of the photoionization detector, charcoaj and the other the test charcoal. Air Cowed When this response change was taken into accetmt, no through the former to the detecter during even equilibra.

dete: table changes in TEDA desorption rates wert tion. Upon rea:hing a steady dete: tor base!!ne signal the air 0cw was switched to the test bed. An upscale signal shift occurred, no a

a te c Such signal shift measurements were repeated at the same conditions, often using a fresh bed. At least three y

tmperatures were used for each charcoal. Signal shifts

}t-re:ctded on the strip were measured with h ruler, j

multiplied by attent.,ation factors, and compared with g

8 calibration curves to get TEDA concentrations (mg/m ).

g a.

Three kinds of TEDA. impregnated charconjs from 5 commercial sources have been rtudied for TEDA de.

terption. Four of these contained s 5% by weight g,,

7 e*c leading of TED A.

Another had a mixed im.

g pregnated-2% TEDA and 5% XI. And one charcoaj 3

ens impregnated 5% with a new compound called "C.

O Alkyl TEDA" or "Heavy TEDA," which has an alkyl

'e i

a a

i a

a F

group, such as ethyl, added to ene or more of the Annow mocrty t..M

{ ethylene bridges. The main obje:tive is to make a higher g

Fig. :4. tract or airfb. niocty on TEDA user dmeoa

[

mole:ular weight compound with lower volatility. The couenuaaon er i.o sempentum.

added a!kyl group should not alTe:t the reactive nitro.

gens. Another chuecal, impregnated with 5% XI, enj,

y o

)

1

,f

.[

..e

)

i r

/

characteristics (actisity, surface ares, pore structure, pore size, etc.).

Fig. 27 shows a comparison of desorption concentra-i};

^

g" tions of TEDA ar.d Heavy TEDA. Both charcoa.ls were from the same manufacturer who said the same base g

chuecal was used. Note that the Heavy TEDA desorp.

l tion was about 10 times lower than that for TEDA.This is what was expected. EfDeiencies for trapping methyl vW, lodide have been found to be similu for both impreg.

I nuu.

a O

70 %

As we have seen from Fiss. 26 and 27, Clapeyron g<

equation plots (log C versus 1/T) are Encar. This was 9

expected from analogy with evaporation and sublimition i

s processes. The slopes of these plots are directly prepor.

e i

a 4

' ' ' '""'**8 tiond to heats of deserption. The range of meas 2 red FO. 23. Emet of bed dep6 on TIDA espor doorpdon sonaatts.

heats of desorption are shown in T& ole XIX. The noa St ar" tempersruris, average is 25 kcal/mol, much higher than the 14 kcal/mo! heat of TEDA sublimation from pure crystab.

The difTerence is due to TEDA.chueca) interactions.

observed over these ranges of experimental puameters.

The 25 kcal/mol average corresponds to a doubling of

]'

Only dry air was used in other experiments.

desorption concentration with every 5'C rise in tempera.

For the ordinary $% TEDA chuecals desorption tare.

csncentrations varied widely (Fig. 26). For example, at Another use of the Clapeyron equation plots is 8

90'C the range was 4 to 48 mg/m. The mixed extrapolation to lower temperatures where TEDA de.

8 impregnant charcoa) gave a value of 6 mg/m. at the sorption is too small to measure direct!y. Such extrapola.

locer end of this range. No iodine or other desorbing tions to 25'C yieldea the TEDA vapor concentrations vapors were detected from the $% K13 (only) im.

showtiin Table XX.

pregnated charcoal up to 120'C.

The most important conclusion from these studies is The differences in desorptien rates for the four $%

shown in this table: The ma.ximum desorbed TEDA TEDA charcoal are signi%:st. They may be he to vapor concentration at 2$'C was calculated to be 0.12 impregnation methods or due to the charcoa) base TImPee Afuse *c) ttareeAtuet(N) m me se m

g,

1. 08 :

14 30 0 to.

se 7e l

e F

SE TEDA C

ss MoA 881 A Coat g =':

I!

N i

$1H TEDA 8

i=,

i W W1 e

f!

O l

l 4

0 t

H to t'7 7e N

1 37 to M

30 atcreoC At feare1 Atutt 34 (*ES M

96 stCPeOC AA TsmPflafvet svet*tS i

I. 26. C!speren plots for TID A eav semed frem 3% TID A rg _27. Cla;wpon piou for ispor de.orted from a "normaI* TID A 1

4

.......s..s.....--a s......

L

. A

P

/

/

,/

p. No toxicological data is available for TED4, but of methyl iodide pea'o to the electron capture detector O a well below the Threshold 1.fmit Values for simi:ar (Valco Instrument Co., Model 140B). Each sample

/ s min e s, which range fr.om 4 mg/m8 for passes through its own column (1.0 or 0.6 m long) with j-cethylenetri4mine to 40 mg/m for triethylandne."

its individually controlled carrier gas flow rate to ac.

8 Theref;te, thert should be no toxic hazard from using complish this. A third vajve momentarily ents the I.

TEDA. impregnated charcoals up to the 5% by weight emuent from both columns to keep air from pusing l

level, through the detector. A downstream sarnpling loop 10 i

times larger than the upstream one gives methyllodide VIII TEST APPARATUS DEVELOPMENT peaks closer together in size during the earlier stages of bed penetration. These improvements, the substitution of During the course of this project several experimental a dew point hygrometer (EGhG. Model 911) for a apparatuses base been butt and used for chauenging and resistance type hygrometer, and high output generation tsting serbent beds, canisters, and cartridges. The two of methyllodide from permention tubes were major steps 8

earliest apparatuses have been described :sewhere and to the final test apparatus design.

In Section !!! and Fig. 2 of this report.

The final test apparatus, pictured in Fig. 28, contains After experiments with radiciodine were completed the in one unit on wheels: (1) air flow, humidity and analytical instruments !91 for sampling and meuuring temperature control, (2) methyl iodide generator, (3) methyl iodide in air was redesigned and rebuDt. Goals sampling pumps and automatic samplers, (4) dual col.

were compactness, simplicity, autoristion, and low cost.

umn gas chromatograph with sampling vajves and Sampling vahes and loops were mounted in a heat valve eJeetron capture detecter, and (5) data integrator with cven (Carle Instrument Inc., Model 4300) to overcome chart recorder. It requires for operation:(1) compressed the pr:blem of water condensation during high humidity air (2) distuled water (3) argon / methane carrier gas, and

't7,sts. A more emeient gas chromatograph column pack.

(4) elect ic power. Two Respirator Cartridge / Canister g.'Porapak Q.S) for sepuating methyliodide from air Test Systems has e been buut. one for our use and one for ou found. This rnade possible smaner co!umns that a]so

• N10SH. The capabihties are:

could be meunted in the valve oven and eliminated the Temperature: Ambient--40'C need f;r a separate large gas chromatograph. Two Dew Point: s25'C sampling vahes wer: ganged by gears for simultaneous-Airilow Rate: s100 IJmin (Constant) samp!!ng upstream and dowmstream of a test canister.

Penetration Fraction: 20.001 This eliminated the need for interpolating between peak Charenge Concentration: 20.1 ppra CH t i

arcu of alternate samp'es when making comparisons.

Advantages of having two units include the capsbuities Simultane% sampling requires different times of anival to confirm testing results, to help NIOSH with I

I TABLE XX. Triethylentdiamine Dssorption Charcoal Heat of Desorption Vapor Concentration Impregnants (kcaVmol) at 25'C (mg/m')

5% TEDA 19.6 0.12 23.2 0.032 31.6 0.0003 26.6 0.0035 2% TEDA 28.5 0.0011

+5% KJ 3

!% H TEDA 19.0 0.016

(

, 32

t 9

troub! shooting, and to preside backup in case of major In November 1981, meetings were held in Rochille, breakdowns.

MD, with NRC and NIOSH personnel to refine some of A detaDed and descriptive operations manual for this these proposals. We identified probable maximum use test system has been written" and will not be repeated conditions (90'F or 32*C: 100% RH). The proposal of bert. The table of r,ontents is given in the Appendix of user discretion in setting senice life bar:d on data to be this report to illustrate tne information provided to the provided on work rates and breathing volumes was NRC and NIOSH. It includra diagrams, photographs, rejected, since it was felt that user knowledge was often specifications, instructions, precautions, and componem inadequate and the radiciodine has no warning

manuals, properties in case of overuse. We identified some addi.

The nrst draft was subjected to an evaluation sus-tional use restrictions (interferences, storage, maximum gested by Donald Campbell of NIOSH. Five technicians concentration, facepiece performances etc.) that must be and staff members not fam0iar with the apparatus were part of the approval. Another revised set of testing given the instructions and apparatus and asked to conditions (30*C and 25'C at 50% RH and 85% RH) perform a cartridge test. These evaluations revealed was proposed. Steps necessary fer foUow up of this some unc! car and out of sequence instructions and meeting were agreed upon.

presided useful suggestions for improvements. The final The ANSI Ad Hoc Respirator Testing and Approval draft was once again evajusted in this way to make sure Subcommittee meeting in Los Ala.nos in December 1981 the changes had been effective.

was another good opportunity to discuss re!cvant sub-In the light of the discovery of significant cyc!!c flow jects with representatives from many industry and gov.

,efTects, this apparatus and manual will need to be ernment organizations.

I modified to include a breathing simulator pump and Approval requirements were modified to allow several '

associated parts (Section VI.D.)-

classes of approvals iy humidity range (high and l

moderate) and minimum service life for 1% penetration:

IX. DEVELOPMENT OF APPROVAL CRITERIA FOR RADIOlODINE CANISTERS Hah Humidity, Jo r.dnutes at 30'c and i

Hhlf Hour 100% RH t

A. History High Humidity.

60 minutes at 30'C and Preliminary proposals for approval (acceptance) criteria were prtsented and discussed at NRC and Moderate Humidity, 6C minutes at 30'C and NIOSH in Febn ary 1981. Testing conditions proposed One. Hour 75% RH were: 0.3 ppm CH81 challenge at 64 L/ min,23'C, and rws hurrudities, ;0% and 85% RH for freshly opened Moderau Humidiu, D0 nm u 30'C ud Tw o Hour 75% RH l

(n:t equilibrated) canisters. Acceptable senice lives j

proposed were 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> at 50% RH and 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> at 85%

The reasons for more than one class cf approv4s are:(1) l RH, which extrapolares to 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> minimum at 100%

to aJlow some currem canisters to be appresed and (2) to RH.

pre ide incertive for manufseturers to deselop improved Further discussions and additional experiments led to canisters for higher classes (i.e., High Humimy, Ught a revised set of proposals in April 1981 Testing at a Hour)of approval, higher temperature (30'C) was added. Close control of The approval schedule should s!so include peaiodic Ril (*2%) and T (tl'C) was required. A sesting 'o venfy shelf hfe claims of manufacturers, reproducibDity requirement of *10% on.enice life Additional use restrictions that must be put into the

{

measurements was proposed, as was identifying service regulations for use and on approvallabels inclrde:

life in terms of total breathed volume,instead ofin terms

1. Not to be used in the presence of organic solvent L

cf time of use. A knowledgeable industrial hygienist or vapors.

/,

supeniser would then he able to calculate a service life

2. To bc stored in scaled, humidi.y barrier pa:Laging i

based on T. RH, and work lese!. These ideas were in cool, dry environments.

discussed over the next several months with various

3. Senice life in to be cajeulated frem the tirne of inttrested parties.

unsealing including periods of non exposure.

\\

1 l

l to w

m 0

il I

_w

., w i

w w e

1 i

n 6 19i i

9 #3 i

i

. G, i

g. _,..

. i...

....~.. g i

I i

3 ni,.

~,-

g g

a.

w W

Mi!2

.4 l

4 g,

/'

' I M s.

3 i

e' 1.

1 r

1558 J

4

.e..-e j,

d 4 1... _*_

n (b ~

\\

N ~ ?..

l7-;..K M

C y

7

+p.w

\\

.~

/S. A._

3

.~

%.b'.thl,7GM@$

b.$ $,Q_

m

.... d 4

ke M_'

i m

M,

~~ Z _

e..

s Awse 1

1 E J, R.~3 i

J m2g i

( A~ m.-- '

M[.[M J ap.,

3 4

,. g -, p =. D[ s u g' N E

$*a.d.

,j w i1 7<

l E.

a

)

,e

..$.e.a.

,e.

....A 1

m

- [-tt.9 9.,,;ig ;?

1

=

h.~ -.,N % 'I;,'.' :

1 e..

e m/ --met. m-.

E' A. e i....

w w.: v : N i:;4. w. w. in. 3 2 U. w 3:efr _hw,;)s:.gQ,

n-

. w z.v wnw{yv._h,..g u.

s.

1 y:2 ' '..bs.......

j al.* l... .. '.., h. : 43,?..;;_7.,.y.

i a 4.m.'.-:

-u d. ~.

.,,; q 'f g g.Q; Qr - g y ih f.:)...:

v-

a. v -

g. w..c,ey..,.. Lg e.

.s (L

. u...,. *,. ~

a _j 3;,,. 4 w.

1

.4 rs 2_.

• _w.

4 28. Apparatus dettloped for tesus radiciodl.ne cartridges a.nd san.' sten uLM aut.hyllad;h.

L 1

'D

4. To be use0 with a facepiece capable of providing dition and provides a safety factor for use at less severe protection fectors greater than 100, as detemuned by conditions. The maximum testing humidity was redu.ed esting with a HEpA fi!!ct and acroso!.

from 85% to 75%, sin 6e at 30'C the latter corresponds S. Not to be used in chaIJense concentrations of total to a dew point of 25'C the maximum practically orgarde lodide, including nonradiometric lodide, greater suainable without placing the testing apparatus in a than 1 ppm.

warm (> 25'C) room or environtnental chamber. Linear Also in December 1981. NIOSH laitiated by internal, extrapolation of resujts to 100% st 30'C using los memo procedures for establishing an approvs1 schedule ' service life versus los RH plots is recommended, on the following conditions: (1) NRC will first estsblish Triplicate instead of dup!!cate service life determinations administrative controls, (2) Los Alarnos will provide wi!! better define reproducibuity and the need for addi.

NIOSH with the testing equipment,(3) approval will be tional testing, for methyllod;de, the testing agent, only. NRC then can a!!ow use for other iodine vapor species based on Los Alamos and other data.

X. AS$1 STANCE TO NIOSH IN ESTABLISHING A TESTINO AND CERTIFICATION PROGRAM B. Current Recommendations All data, conclusion, and propeujs generated from this project have been shared *ith the NIOSH TCB from The current recommendations for radiciodine the beginning. This has been accomplished by visits to cartridge and canister testing :enditions and acceptance one another's laboratories,in. person and telephone con.

criteria are summarized in Table XXI. Also listed are the versations, trip reports, progress reports, public presenta.

current criteria fmm the 1.'.S. Code of Federal Regula.'

tions, and publications.

O tions' for organic vapor canisters for comparison. The A final test apparatus described in Section Y111 was latter are cal:ed current reccmmendations, rather than bu0t and shipped to the NIOSH TCB for thnt use in final ones, since discussions wi!! con'inue in the regula, certincation testing. An extensive eperation manual was to n process.

prepared and also given to N!OSH. Fo!!owup visits to Testing should be done at eso relative hur-Jdities and the Morgantown, WV, laboratories are planned to help l

j at 64 Umin cyclic airnow for canisters and 32 Umin NIOSH in setting up vid using this equipment. Los cyclic airSow l'or cartridges used in pairs. Cha!!enge Alamos will also be avaUsble for telephone consultations, concentration should be I ppm methyllodida., although as needed. The duplicate apparatus at Los Alamos wCl this is not a critical parameter. IJnits are to be tened as be useful for identifying and correcting problems N!OSH l

received and freshly opened. Tests at 258C were may enceunter, n well as for performms interlaboratory i

sliminated since 30'C represents a more severe con, comparisons of test results.

l i

l l

k n

j

O TABLE XXI. Testing Conditions and Acceptance Criteria for Organic Vapor Cidn. Style Gas Mask Canliters Test CFR Title 30, Radiotodfrte Parameter Fart !!,102 Proposal Vapor CCl.

CH1 3

Concentration 5000 ppm I ppm Relative Humidity 30 t $%

$0 75% (t2%)

Temperature 23

• 2.!'C 30

• l'C Total Airflow As Re7Ived 64 UmIn 64 Umin Cyclic Flow 4 Equi 1Ruted 32 UMin EquDibration LAs Received All As Raceived (6 H st 64 Umin) 2 At 25% RH 3 At 50% RH 2.At 85% RH 3.At 75% RH Manlmum Penetration 0.1% ($ ppm) 1% (0.01 ppm)

Minimum Service Life 12 min 30 min at 100% RH*

60 min at 75% RH

• Extrapolsted from 50% and 15% RH.

(

REFERENCES

$. A. Wheeler, in Catalysis, Yo!. II, p. !!O, P. H.

Emmet, ed., Reinhold Publishing Co., New York,

1. G. O. Wood, G. J. Vogt. D. C. Gray, and C. A.

NY (19 3).

Kasuni:,"Criteria and Test Methods for Certifying Air.Pu'ifying Respirators Against Radiciodine "

6. O. Grubner and D. W. Undertal, "Caj:ulation of Los Altmos Scientific Laboratery Progress Report Bad Capacit) by the Theory of Statistical Mo.

i NUREG/CR 10$$, LA 8029.PR (September rnents," Separati:n Scienee 5, $ *$ (1970).

1979) Ava2able from the National TechnicalInfor-tr.Ation Senice, Spnnstield, VA 22151,

7. G. O. Nelson, A. N. Cr:reia, and C. A. Harde',

"Respirator Cartridge Effirier.cy Studies:Vll. Effe:t

2. G. O. Wood "SOP for Use of 88'lin the Testing er of Relative Humidity and Temperature," Am. Ind.

Respirater Components," Unpublished document.

Hys. Assoc. J. 37,281 (1976)

Los Alamos National Laboratory,' indatrial Hygiene Group l.as Alamos.NM 87544 (February

8. R C. Lee and L. Suverman,"An Apparatus for 1979)

Measuring Air Flow During Inspiration," Rev.

Scient. lutruments 14, 174 (1943).

3. M. J. Xabat private communication of unpublished data, Ontario Hydre, Health a.nd Safety Division,

9. L SDvermen, et al, "Air Flow Measurements on Torento, Canada (1982)

Human Subjects with and withou Respiratory Re.

sistance at Several Work Rates" A. M. A. Arch.

4. Department of the Interior Bureau of Mines, Ind. Health 13,(1956).

"Respiratory Protective Devices: Tests for Fe missibaity: Fees," Titje 30, Code of Federal

10. G. O. Nason and C. A. Harder, "Respirator Regulations, Part !!, Fed. Res. 31, No. $9 (March Caruidge Efficiency Studies IV. Efre:ts of Steady-23, 1972).

State and Pulsating Flow" Am. Ind. H)g. Assoc J.

33s 797 (1974).

~..

'~

r O

11. V. R. Deitz, C. H. Blachly, and L. A. Jonas, Avausble from the NationaJ 1echnical informtdon "Dependence of Gas Penetration of Charcoal Beds Service, Springfield, VA 22161.

~

on Residence Tirne and Linear Velocity" Proceed.

ings of the 14th ERDA Air Cleaning Conferenrn,

!$. R. D. Ackley,"Removal of Radon.220 for HTOR CONF 760822, Vol 1, p. 233, M. W. First, ed.

Fuel Reproetssing and Refabrication Off.Ges (1977). AvaDable from the Nadonal Technical Infor.

Strcams by Adsorption. DRNL.TM 4883 (Apri!

1973) AvaDable from the Nadonal Technical Infot.

Cation Seni;e, Springfield, VA 22161.

mation Service, Sprir4 ic1d, VA 22i61.

f

12. L. A. Jonas and J. A. Rehrmann, carbon 12,93

16. Threshold Limit Values for Chemical Substances (1974).

and Physical Agents in the Workrocra Environment with Intended Changes for 1982 American Con.

13. F. G. May and H. J. Polson, "Methyl lodide Penetration of Charcoal BW: Variation with Rela.

ference of GovernmentalIndustrial Hygienists, Ch.

tise Humidity and Face Velocity" Australian cinnati, Ohio (1982).

Atomic Energy Corr.tnission report AAEC.E322

17. G. O. Wood, V. Gutschick, and F. O. VsJdez.

(1974).

"Operating Manual Respirator Cartridge / Canister

14. D. W. Underhill. "Mass Transfer of Krypton.83 in Test System Using Methyl lodide."Industrial i

i Charcoal Adsorbers" Proceedings of the 9th AEC Hygiene Group, Los Alamos Nations! Labora.'ory, Air Cleaning Conference, AEC.660904, Vol. 2, p.

Los Alamos NM 87343 (1982).

824. J. M. Morgan and M. W. First, eds, (1967).

'C 4

II

.Y%.

9 APPENDIX Table of Contents from Refemnce 17.

OPERATING MANUAL RESPIRATOR CARTRIDGE / CANISTER TEST SYSTEM USING METHYL IODIDE 1

Contenu Page

!. GENERAL PRINCIPLES OF OPERATION A. Introductica I

B. System Descriptica.................................... I

1. Block t: Main Air $upply 2

2. Block 2: Methyl lodide Cha!!enge Gem ator..................... 3 J

3. Block 3: Humidifying and Hags Section. Majn Air Flow 4

4. Block 4: Measurement Section 4

3. Bloct 5: Gas Chromatograph and Act.essories................,... 7 C. Component Identities and Specifications

........................10

1. Commercial Components..............................10

2. Components Built at Los Alames National La!, oratory

. 14

(

IL INITIAL SET.UP TO STANDLY CONDITION A.

Input Requirements: Power. Air. Water. Ce trier Gas

. 15 l

B. Connections and Adjustments to Reach Standby Condition

. 16

)

!!!. CHOOSING AIR FLOW RATE, TEMPERATURE, AND RELATIVE HUMIDITY A.

Flow Ra:e

........................................18 B. Temperature end Relative Humidity

..........................18

!Y. STAR 1UP FROM STANDBY 10 RUN CONDITION

. 19 V. CALIBRATING AIR FLOW RATE AND OAS CHROMATOGRAPH SENSITIVITY A.

Calibrate the Main Air Fluwmeter

...........................24 B. Check Gas Chromategraph's Peak Resolution

. 25 C. Measure the GC Sensitivity R atio (Upswea:n:Dovmstream)

. 26 VI. INITIATING A TEST RUN A. Preparation

.......................................27 B. Periodic Checks During Automated Run........................ 30 VII. RESETTING TFE SYSTEM FOR A NEW RUN A. New Air Flow R ate................................... 31 B. New Main Air Temperature

..............................32 C. New R elativ H umidity................................. 3 2

(

r v

ee >

e

])

Yll!. TREATMENT OF DATA A. For Each Individual Rt.n

................................32 i

B. Extrapolating a Set of Run to Reference Conditions

............,.....33 IX. SHUTDOWN FROM RUN TO STANDBY CONDITION

...............34 X. SHUTDOWH FROM STANDBY TO *' FULL OFF" CONDITION

..........35 XI. MAINTENANCE AND REPAIRS............................37 A. Air Supply FDiers (Both)

B. Carrier Oas Puriner C. Permentior Tube D. GC Packed Columns E. Electron Capture Detector F. Humidifying Water Bath G. Maht Air Flowmeter APPENDIX !.

CHECKLIST OF CONTROL SETTINGS FOR FULL OFF. STANDBY. AND RUN CONDITIONS

.......................................39 APPENDIX II. INSTRUCTION MANUALS FOR COMMERCIAL COMPONENTS

...43 A. Main Air Digital Flow Meter: Datametries Model 810L Flowineter C,

B. Liquid Level Contro!!ct Relay: Pope ScientiSc Company. Lab Mo. titer III.

C. Digital Humidity Ar.alyzer/ Controller: EO/G Model 911 Dew.At! Digital Humidity Analyzer.

D. Dual Ten. Port Multi Functional Sampling Valves: Valco Instruments Ca. Model AH.V.10 HPa which includes Valve Actuators (Model A60), Valve Oven (Model HVE.2), and Sampling Loops for Application 32.

E.

Digital Valve Programmer: Valco Instruments Co.

F. Temperature Contrems: Valco Instruments Co.

G. Carrier Gas Puriner: Supelco H. Electron Capture Detector: Ya!co Instruments Co. Model 140.

I. ECD Chart Recorder: Cole Parmer Instrument Co. Model 8377 10.

J. Ar'omatic Peak Integrator: Spectra Fhysics "Minigrator" i

\\

l

--