Full text of "DTIC ADA491483: A Residual Life Indicator (RLI) for Physical Adsorption Capacity of Nuclear, Biological, and Chemical Filters. Part 3. A Novel RLI Design for Collective Protection Demonstrated Using Breakthough and Chemical Pulse Data"

See other formats

EDGEWOOD--

CHEMICAL BIOLOGICAL CENTER

U S ARMY RESEARCH, DEVELOPMENT AND ENGINEERING COMMAND

ECBC-TR-658

A RESIDUAL LIFE INDICATOR (RLI)

FOR PHYSICAL ADSORPTION CAPACITY
OF NUCLEAR, BIOLOGICAL, AND CHEMICAL FILTERS

PART III

A NOVEL RLI DESIGN FOR COLLECTIVE PROTECTION
DEMONSTRATED USING BREAKTHOUGH AND
CHEMICAL PULSE DATA

I ^

Manufacturing Company

A buunacs yvt of Huntor r*otvx»»og**«

HUNTER APPLIED RESEARCH CENTER
Edgewood, MD 21040

Gregory W. Peterson

RESEARCH AND TECHNOLOGY DIRECTORATE

David Friday
Marc Shrewsbury
Scott Deibert

November 2008

Approved for public release;
distribution is unlimit'' J

20081230007

ABERDEEN PROVING GROUND, MD 21010-5424

DISCLAIMER

The findings in this report are not to be construed as an official Department of the Army position
unless so designated by other authorizing documents.

REPORT DOCUMENTATION PAGE

Form Approved
OMB No. 0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the
data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing
this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suita 1204, Arlington, VA 22202-
4302 Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently
valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS

1. REPORT DATE (DD-MM-YYYY)

XX-11-2008

2. REPORT TYPE

Final

3. DATES COVERED (From - To)

Jun 2006 - Nov 2007

4. TITLE AND SUBTITLE

A Residual Life Indicator (RL1) for Physical Adsorption Capacity of Nuclear,
Biological, and Chemical Filters, Part 111, A Novel RL1 Design for Collective
Protection Demonstrated Using Breakthrough and Chemical Pulse Data

5a. CONTRACT NUMBER

5E22A

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S)

Friday, David; Shrewsbury, Marc; Diebert, Scott (HARC); and Peterson, Gregory
W. (ECBC)

5d. PROJECT NUMBER

BA05PRQ102

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

DIR, ECBC, ATTN: AMSRD-ECB-RT-PF, APG, MD 21010-5424
Hunter Defense Technologies, Applied Research Center
2109 Emmorton Park Road, Suite 120, Edgewood, MD 21040

8. PERFORMING ORGANIZATION REPORT
NUMBER

ECBC-TR-658

9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES)

Defense Threat Reduction Agency
8725 John J. Kingman Road, Room 3226
Fort Bel voir, VA 22060-6201

10. SPONSOR/MONITOR’S ACRONYM(S)

DTRA

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION / AVAILABILITY STATEMENT

Approved for public release; distribution is unlimited.

13. SUPPLEMENTARY NOTES

14. ABSTRACT

Wc start by simulating the condition of a fielded filter by selecting representative threat vapors and representative contaminant
vapors. We assume that only contaminants that are moderately and strongly adsorbed can affect the residual life. A simulated
contaminated filter is configured using contaminated carbon at the bed inlet and fresh carbon at the bed outlet. Breakthrough
experiments are then completed using an organic to simulate the threat vapor. These data define “residual life” based on
relative humidity (RH), threat vapor, contaminant chemical, and the extent of bed contamination. Using a novel residual life
indicator approach that employs dual satellite beds, one sampling the filter air inlet and one sampling the filter outlet, chemical
pulse tests are conducted. Simulated contaminated beds are tested at different RHs. The difference between the reference bed
effluent concentration profile and the contaminated bed effluent concentration profile is correlated to the residual life using a

15. SUBJECT TERMS

Acetone

Organic contamination

Pulse testing

Physical adsorption capacity

Residual life indicator (RLI)

16. SECURITY CLASSIFICATION OF:

17. LIMITATION OF
ABSTRACT

18. NUMBER OF
PAGES

19a. NAME OF RESPONSIBLE PERSON

Sandra J. Johnson

a. REPORT

b ABSTRACT

c. THIS PAGE

19b. TELEPHONE NUMBER (include area code)

410-436-2914

Standard Form 298 (Rev. 8-98)

Prescribed by ANSI Std. Z39.18

Blank

PREFACE

The work described in this report was authorized under Project
No BA05PR0102. This work was started in June 2006 and completed in November 2007.

The use of either trade or manufacturers’ names in this report does not constitute
an official endorsement of any commercial products. This report may not be cited for the
purposes of advertisement.

This report has been approved for public release. Registered users should request
additional copies from the Defense Technical Information Center; unregistered users should
direct such requests to the National Technical Information Service.

Blank

CONTENTS

1. INTRODUCTION

1.1 Background.1

1.2 Classification of Threat Vapors for RLI Purposes.2

1.3 Approach to Estimating Residual Life of a Filter.2

2. BREAKTHROUGH TESTING.3

2.1

2.2

2.3

2.3.1

2.3.1.1

2.3.1.2

2.3.1.3
2.3.2

2.3.2.1

2.3.2.2

2.3.2.3

2.3.3

2.3.4

2.3.4.1

2.3.4.2

2.3.4.3

2.4

Simulated Contaminated Filter.

Breakthrough Apparatus and Procedure.

Breakthrough Test Results.

Hexane Challenge/Octane Contaminant.

Effect of RH.

Effect of Contaminant Loading.

Effect of Contaminated Bed Fraction...
Hexane Challenge/Dodecane Contaminant

Effect of RH.

Effect of Contaminant Loading.

Effect of Contaminated Bed Fraction ..
Nonane Challenge/Octane Contaminant....
Nonane Challenge/Dodecane Contaminant

Effect of RH.

Effect of Contaminant Loading.

Effect of Contaminated Bed Fraction ...
Summary of Breakthrough Results.

..3
..4
..5
..7
..7
..8
..8
10
10
10
11

14
14

PULSE TESTING

3.1 Pulse Testing Concept and Approach. 18

3.2 Pulse Testing System Apparatus and Procedure.19

3.2.1 Apparatus.19

3.2.2 Bed Preparation.20

3.2.3 Test Procedure.21

3.3 Pulse Test Results.21

3.3.1 Testing Overview and Experiment Summary.22

3.3.2 Octane Contaminant Results.22

3.3.2.1 15% RH.23

3.3.2.2 50% RH.25

3.3.2.3 80% RH.26

3.3.3 Dodecane Contaminant Results.28

3.3.3.1 15% RH.28

3.3.3.2 50% RH.29

3.3.3.3 65% RH.31

3.3.3A 80% RH.31

3.4 Pulse Testing Summary.33

4. RLI CORRELATION.33

5. SUMMARY.38

LITERATURE CITED.41

FIGURES

2.1 Bed Configuration for Simulated Containment Filter.4

2.2 Test System Schematic.5

2.3 Hexane Challenge with Octane Contaminant at 15% RH. 9

2.4 Hexane Challenge with Octane Contaminant at 50% RH.9

2.5 Hexane Challenge with Octane Contaminant at 80% RH. 10

2.6 Hexane Challenge with Dodeeane Contaminant at 15% RH.11

2.7 Hexane Challenge with Dodeeane Contaminant at 50% RH.12

2.8 Hexane Challenge with Dodeeane Contaminant at 80% RH.12

2.9 Nonane Challenge with Octane Contaminant at 15% RH.13

2.10 Nonane Challenge with Octane Contaminant at 80% RH.14

2.1 1 Nonane Challenge with Dodeeane Contaminant at 15% RH.15

2.12 Nonane Challenge with Dodeeane Contaminant at 80% RH.16

3.1 RL1 Pulse Apparatus P&1D.20

3.2 Results for 0.1 g/g Octane Loading at 15% RH Using PFCB.24

3.3 Results for 0.2 g/g Octane Loading at 15% RH Using PFCB.24

3.4 Results for 0.1 g/g Octane Loading at 50% RH Using PFCB.25

3.5 Results for 0.2 g/g Octane Loading at 50% RH Using PFCB..26

3.6 Results for 0.1 g/g Octane Loading at 80% RH Using R123.27

3.7 Results for 0.2 g/g Octane Loading at 80% RH Using R123.27

3.8 Results for 0.1 g/g Dodeeane Loading at 15% RH Using PFCB.28

3.9 Results for 0.2 g/g Dodeeane Loading at 15% RH Using PFCB.29

3.10 Results for 0.1 g/g Dodeeane Loading at 50% RH Using PFCB.30

vii

3.11 Results for 0.2 g/g Dodecane Loading at 50% RH Using PFCB.30

3.12 Results for 0.2 g/g Dodecane Loading at 65% RH Using PFCB.31

3.13 Results for 0.1 g/g Dodecane Loading at 80% RH Using R123.32

3.14 Results for 0.2 g/g Dodecane Loading at 80% RH Using R123.32

TABLES

2.1 Simulant and Agent Properties.6

2.2 Contaminant Properties.6

2.3 Complete List of Breakthrough Experiments.7

2.4 Estimated Breakthrough Times Using MTZ that is 35% of Bed.18

3.1 Pulse Gas Properties.22

3.2 Complete List of Pulse Experiments.22

4.1 Summary of PFCB with Octane as Contaminant.35

4.2 Summary of PFCB with Dodecane as Contaminant.35

4.3 Summary of 80% RH Experiments Using R123.36

4.4 Break Time, Mass Ratio Comparison with Octane as Contaminant.37

4.5 Break Time, Mass Ratio Comparison with Dodecane as Contaminant.37

viii

A RESIDUAL LIFE INDICATOR (RL1) FOR PHYSICAL ADSORPTION CAPACITY
OF NUCLEAR, BIOLOGICAL, AND CHEMICAL FILTERS

PART III

A NOVEL RLI DESIGN FOR COLLECTIVE PROTECTION DEMONSTRATED
USING BREAKTHROUGH AND CHEMICAL PULSE DATA

I. INTRODUCTION

I. I Background

Estimating the remaining filtration capacity of a filter has been investigated for
more than 20 years. Typically, these efforts have been divided into two types of indicators,
namely, (1) cnd-of-service life and (2) residual filter life end. An end-of-service life indicator
(ESLI) refers to a method of detecting a toxic chemical eluting through a filter during a chemical
event. In collective protection (ColPro) applications, this may not be as useful as determining
the residual life prior to a chemical event since filter change-out during that event is not practical
(or possibly even feasible). A typical example of an ESLI is a device that changes color when
exposed to a given chemical or group of chemicals. We will discuss the problems with this in the
next paragraphs. Is an ESLI that responds specifically to cyanogen chloride (CK) or Sarin (GB)
really useful? At what concentration does the sensor respond? What is the protocol following a
sensor response? How many potential threats do NOT respond to the detector? Given all of these
questions, what is the real value of an ESLI that responds to one or two chemicals?

The filter “residual life” is the remaining filtration capacity of filter and whether it
should be subjected to a chemical event is the focus of this report. Specifically, wc will
demonstrate how ambient contaminants can degrade filter capacity during peacetime operations.
Many previous efforts have been conducted to estimate the residual life of filters. 1 g While most
of them produced little or no useful results, even those efforts that did provide some valuable
insight into a residual life indicator (RLI) failed to answer the fundamental question: what is the
residual life of the filter with respect to specific challenge gas or gases? This work attempts to
address this question in its entirety by first addressing only those gases that are affected by
adsorbed contaminants.

Assessing the residual life of collective protection filters is particularly important,
since unlike individual protection filters, it is much more difficult and expensive to change them.
In fact, the current military requirement for individual protection is to have the capability to
change filters in a chemically contaminated environment. This is clearly not possible for
collective protection filters. Current change-out doctrine for collective protection filters is based
solely on time-in-service. Using this approach, it is very likely that good filters are changed too
early (increasing life cycle cost) and poor filters remain in service too long (potential protection
risk).

1.2

Classification of Threat Vapors for RL1 Purposes

Generally speaking, one can divide threat vapors into two categories in terms of
removal mechanisms by ASZM-TEDA: (1) those vapors removed primarily by physical
adsorption [e.g. GB and Distilled Mustard (HD)] and (2) those vapors that require subsequent
chemical reaction to prevent elution into the product [e.g. CK, hydrogen cyanide (AC), and
phosgene (CG)]. Toxic Industrial Chemical (TIC) vapors can also be classified in this manner.
Gases from both categories can be adversely affected by adsorbed contaminants as shown by
Peterson et al. 0 Gases such as CK and AC that require chemical reaction can also be adversely
affected by a reduction in impregnant activity. Both high water loadings (caused by exposure to
high RHs) as well as reactive ambient contaminant gases such as NOx, SOx, etc., can reduce
impregnant reactivity.

1.3 Approach to Estimating Residual Life of a Filter

Based on the two-category concept, we have divided our RLI effort into two
parts: Part 1, determine the effect of adsorbed contaminants on filter protection performance,
and Part 2, determine the effect of loss of impregnant activity on the filter performance. This
report details the efforts undertaken to complete Part 1. The loss of filter life due to impregnant
degradation will be the focus of a follow-on effort.

A critical component of the RLI approach is that the remaining filter life can only
be estimated knowing the current state of the filter. That means that all of the adsorbed
constituents (contaminants) that can adversely affect the ability of the threat vapor to adsorb
must be characterized. The only method to accomplish this task is to sample the adsorption
capacity of all the ASZM-TEDA carbon in the entire filter. The best method to accomplish this
task is to use a weakly adsorbed (non-destructive) probe gas that does not react with the
impregnants. That is the approach used in this effort.

The RLI correlation is developed in two parts. We start by simulating the
condition of a fielded filter (Section 2). We assume that only contaminants that are moderately
and strongly adsorbed can affect the residual life since weakly adsorbed gases elute quickly
through the filter and can easily be displaced by water and most threat gases. We configure a
simulated contaminated filter using contaminated carbon at the bed inlet and fresh carbon at the
bed outlet. Breakthrough experiments are then completed using an organic to simulate the threat
vapor.

In Section 3, we configure the same simulated contaminated beds used in the
breakthrough study described in the previous paragraph. This time, these beds are challenged
with pulses of perfluorocyclobutane (PFCB) or l,l-dichloro-2,2,2-trifluoroethane (R123). Our
approach uses two beds, a contaminated bed and a reference bed made up of only fresh carbon.
Both beds sample the same RH, assuming that the filter effluent RH is nearly the same as the
filter influent RH, or in other words, the ambient RH air does not change rapidly. Both beds will
have almost the same moisture content, resulting in a method that allows us to “subtract out” the
effect of adsorbed water on the pulse gas result.

In Section 4, we correlate the results form the breakthrough studies of Section 2
with the pulse studies performed in Section 3. A relationship is developed that relates the
difference between the contaminated bed pulse and the reference bed pulse.

2. BREAKTHROUGH TESTING

In this section, breakthrough experiments are conducted to simulate the behavior
of a contaminated filter when exposed to a chemical threat. With this approach, we can then
establish a true filter residual life based upon the remaining filter capacity as determined by the
breakthrough data. Several critical parameters that affect filter residual life are explored. These
include the following:

• The fraction (%) of the filter that is contaminated - 25 and 75% of the total
bed depth

• The strength of adsorption of the contaminant - octane and dodecane

• The loading (adsorbed phase concentration) of the contaminant - 0.1 g/g
and 0.2 g/g

The strength of adsorption of contaminants needs further explanation. Simply put,
the stronger the chemical is adsorbed, the more difficult it is to displace. This means that a very
strongly adsorbed vapor will not be displaced by even strongly adsorbed threat vapors such as
Sarin (GB) or Distilled Mustard (HD). Octane is a strongly adsorbed chemical, but is not nearly
as strongly adsorbed as dodecane. Generally speaking of carbon adsorbents, chemicals with
higher boiling points are more strongly adsorbed than those with lower boiling points.

2.1 Simulated Contaminated Filter

Shown below in Figure 2.1 is a schematic of the “contaminated” filter concept.
This concept is critical to the correct interpretation of residual life. This concept assumes that the
contaminants that affect residual life will generally be confined to the inlet of the filter. This is
consistent with the fact that chemicals that can effectively prevent the threat vapors from
adsorbing will themselves be strongly adsorbed; thus, they will not easily elute through the filter.
Therefore, these contaminant vapors will tend to adsorb primarily at the filter inlet.

I i

25%

Contaminated

75%

Contaminated

75%

Fresh ASZM-T

25%

Fresh ASZM-T

i I

Figure 2.1 Bed Configuration for Simulated Contaminated Filter

2.2 Breakthrough Apparatus and Procedure

The beds used in these experiments are 4.1 cm in diameter. The volumetric flow
rate is set at 18.6 L/min at ambient temperature to simulate the air velocity through a M98 filter.
The bed depth is 5.5 cm slightly greater than a M98. The challenge concentration used in all
experiments is 4,000 mg/m\ Figure 2.2 illustrates the system used in the experiments.

DMMP GAS LIFE TESTING SCHEMATIC

Aif Supply

Miller INdafrn

KoiaMrtm

Figure 2.2 Test System Schematic

Breakthrough testing was eondueted on a push-pull-vented test apparatus at a
temperature of approximately 25 °C. Glass test tubes with an inside diameter of approximately
4.1 em were packed using a snowstorm method through a 30 in. drop tube. The challenge was
delivered by Bowing air through saturated vapor in a temperature controlled glass eell and
diluting with elean humidified air. The feed and effluent streams were monitored with a Hewlett
Paekard 5890 Series 11 gas ehromatograph equipped with a flame ionization deteetor (GC/FID).

2.3 Breakthrough Test Results

Breakthrough testing is performed to demonstrate the effect of RH, the adsorption
strength of the adsorbed eontaminant, and the adsorption strength of the threat vapor. Two
different threat vapor simulants are used, each with a different adsorption strength. We consider
the effeet of adsorbed water on the breakthrough behavior as well. This will be important in the
pulse vapor analysis described in the next section. So, the breakthrough testing consists of the
following parameters:

• Two threat vapor simulants - hexane and nonane

• Two contaminants - octane and dodeeane

• Two eontaminant loadings - 0.1 g/g and 0.2 g/g by weight

• Three RHs - 15, 50, and 80%

Given below in Tables 2.1 and 2.2 are several important physical properties of the
simulants and contaminants used in the breakthrough experiments. A critical parameter that is
related to adsorption strength is the boiling point. Note that the stronger adsorbed simulant
(nonane) has a higher boiling point than the weaker adsorbed contaminant (octane). One would
expect that nonane would be able to compete favorably for adsorption space.

Table 2.1 Simulant and Agent Properties

Property

n-Hexane

Nonane

Values for GB

CAS

110-54-3

111-84-2

107-44-8

Molecular Weight

86.2 g/mol

128.3 g/mol

140.08 g/mol

Boiling Point

68.9 °C

150.6 °C

150 °C

Vapor Pressure @ 20 °C

124 mmHg

3 mmHg

2.48 mmHg

Water Solubility @ 20 °C

0.002 wt.%

Insoluble

Miscible

Liquid Density

0.66 g/mL

0.72 g/mL

1.09 g/mL

8-hr TWA

50 ppm

None

0.00003 mg/m 1

IDLH

1100 ppm

Not determined

0.1 mg/m’

*Calculated at 25 °C

Table 2.2 Contaminant Properties

Property

Octane

Dodecane

CAS

11-65-9

112-40-3

Molecular Weight

114.2 g/mol

170.34 g/mol

Boiling Point

125.6°C

216°C

Vapor Pressure @ 20°C

10 mmHg

0.3 mmHg

Water Solubility @ 20°C

0.00007 wt.%

Insoluble

Liquid Density @ 20°C

0.70 g/mL

0.75 g/mL

8-hr TWA

500 ppm

Unknown

IDLH

1000 ppm

Unknown

Given below in Table 2.3 is the complete list of breakthrough experiments. The
far right column refers to the figure number where that result is plotted. We have decided to
organize all the breakthrough experiments by the threat vapor simulant-contaminant pair.

Table 2.3 Complete List of Breakthrough Experiments

Challenge

Vapor

Contaminant

Contaminant
Bed Fraction

i%L

Contaminant
Loading
(g/g)

RH (%)

Figure #

Hexane

Octane

0, 25,75

0.1

2.2

Hexane

Octane

25,75

0.2

2.2

Hexane

Octane

0, 25,75

0.1

2.3

Hexane

Octane

25,75

0.2

2.3

Hexane

Octane

0, 25,75

0.1

2.4

Hexane

Octane

25,75

0.2

2.4

Hexane

Dodecane

0, 25,75

0.1

2.5

Hexane

Dodecane

25,75

0.2

2.5

Hexane

Dodecane

0, 25,75

0.1

2.6

Hexane

Dodecane

1 25,75(1

0.2

2.6

Hexane

Dodecane

0, 25,75

0.1

2.7

Hexane

Dodecane

25,75

0.2

2.7

Nonane

Octane

0, 25,75

0.1

2.8

Nonane

Octane

25,75

0.2

2.8

Nonane

Octane

0, 25,75

0.1

2.9

Nonane

Octane

25,75

0.2

2.9

Nonane

Dodecane

0, 25,75

0.1

2.10

Nonane

Dodecane

25,75

0.2

2.10

Nonane

Dodecane

0, 25,75

0.1

2.1 1

Nonane

Dodecane

25,75

0.2

2.11

2.3.1 Hexane Challenge/Octane Contaminant

This system would represent a less strongly adsorbed threat vapor (hexane) with a
moderately adsorbed contaminant vapor (octane). Shown in Figures 2.3 through 2.5 are the
results for each contaminated bed configuration at the three different RHs: 15, 50 and 80%. In
every experiment, both beds were allowed to equilibrate at the test RH prior to introducing the
challenge chemical.

The following discussion compares the results of Figures 2.3 through 2.5.

2.3.1.1 Effect of RH

Clean Bed (Baseline) - The breakthrough times (based on approximately half of
the feed concentration) at 15 and 50% RH (Figures 2.3 and 2.4) are about the same - about
105 min. However, the breakthrough time and the shape of the breakthrough curve for 80% RH,
as shown in Figure 2.4, is entirely different. This is because hexane is adversely affected by
adsorbed water at 80% RH. The hexane begins to breakthrough at about 30 min but rises only to
about half of the challenge concentration at 40 min For the next 60 min or so, it rises slowly
towards the feed concentration.

Contaminated Beds - Visual observation shows that the results at 15 and
50% RH are almost the same. The only difference appears to be a slight effect of octane loading
at 50% RH where the breakthrough of the 0.2 g/g loaded bed occurs at about 75 min and the
0.1 g/g loaded bed occurs at about 90 min. The results at 80% RH are very different. The trend
observed for the 15 and 50% RH runs is the same, but the times are compressed. The shapes of
all of the breakthrough curves at 80% RH are also different. For example, the baseline
experiment appears to rise up to around 2000 mg/m 1 2 3 4 5 at around 40 min and then gradually
increase towards the challenge concentration of 4000 mg/m 3 . This is caused by a multi-
component adsorption effect between adsorbed water and hexane.

2.3.1.2 Effect of Contaminant Loading

Visual observation shows that the results at 15 and 50% RH are almost the same.
The only difference appears to be a slight effect of octane loading at 50% RH where the
breakthrough of the 0.2 g/g loaded bed occurs at about 75 min and the 0.1 g/g loaded bed occurs
at about 90 min. The results at 80% RH are very different. The trend observed for the 50% RH
run is the same, but the times are compressed. The shape of all the curves at 80% RH matches
the shape for the baseline run. These results are consistent since adsorbed water is itself a
“contaminant” for hexane.

2.3.1.3 Effect of Contaminated Bed Fraction

The difference in breakthrough times between the 25 and 75% contaminated beds
at 15% RH are different depending upon the loading. At 0.1 g/g octane loading, the difference
between the 25 and 75% contaminated beds is about 20 min. At 0.2 g/g loading, the difference
between the 25 and 75% contaminated beds is greater, about 35 min. The same trend is observed
at 50% RH (25 min and 40 min, respectively), although there is not as large of a difference
observed in the 0.1 g/g loaded beds. At 80% RH, it is difficult to determine a difference in
results as a function of loading, although the general trend from baseline as the longest break
time is still the same. The following is the order from longest to shortest breakthrough time:

1. Baseline

2. 0.1 g/g loading, 25% contaminated bed

3. 0.2 g/g loading, 25% contaminated bed

4. 0.1 g/g loading, 75% contaminated bed

5. 0.2 g/g loading, 75% contaminated bed

Concentration (mg/m3) Concentration (mg/m3)

5000

4500

Figure 2.3 Hexane Challenge with Octane Contaminant at 15% RH

60 80 100

Time (min.)

Baseline

0.1 g/g Octane 25% Bed
[—0.1 g/g Octane 75% Bed
0.2 g/g Octane 25% Bed
0.2 g/g Octane 75% Bed

140

160

180

Figure 2.4 Hexane Challenge with Octane Contaminant at 50% RH

4000

Figure 2.5 Hexane Challenge with Octane Contaminant at 80% RH

2.3.2 Hexane Challenge/Dodecane Contaminant

In these tests, we will examine the effect of a more strongly adsorbed contaminant
(dodecane) on the less strongly adsorbed threat vapor simulant (hexane). Figures 2.6 through 2.8
show the results for the hexane-dodecane system at 15, 50, and 80% RH.

2.3.2.1 Effect of RH

Clean Bed (Baseline) - The breakthrough times at all RH used in this study are
the same data shown in Figures 2.3 through 2.5.

Contaminated Beds - These results are almost identical to those measured for the
hexane-octane system. Here is a slight difference, however. The 0.1 g/g dodecane experiments
breakthrough later than their 0.1 g/g octane counterparts. This is almost assuredly due to the fact
that the octane loading is actually higher than 0.1 g/g or the dodecane loading is lower than
0.1 g/g. From an adsorption strength point of view, this would not make sense.

2.3.2.2 Effect of Contaminant Loading

Once again the results and trends are very similar to those observed for the
hexane-octane system.

2 . 3 . 2.3

Effect of Contaminated Bed Fraction

The difference in breakthrough times between the 25 and 75% contaminated beds
at 15% RH are different depending upon the loading. At 0.1 g/g dodecane loading, the difference
between the 25 and 75% contaminated beds is about 25 min. At 0.2 g/g loading, the difference
between 25 and 75% contaminated beds is greater, about 50 min. The same trend and virtually
the same differences are observed at 50% RH. At 80%RH, as with the hexane-octane system, it
is difficult to determine a difference in results as a function of loading, although the general trend
from baseline as the longest break time is still the same. From longest to shortest breakthrough
time the order remains

1. Baseline

2. 0.1 g/g loading, 25% contaminated bed

3. 0.2 g/g loading, 25% contaminated bed

4. 0.1 g/g loading, 75% contaminated bed

5. 0.2 g/g loading, 75% contaminated bed

5000

4500

4000

Figure 2.6 Hexane Challenge with Dodecane Contaminant at 15% RH

Concentration (mg/m3)

5000

Figure 2.7 Hexane Challenge with Dodecane Contaminant at 50% RH

4000

3500 -I

3000

2500

2000

1500

1000

500

Baseline

—O—0.1 g/g Dodecane 25% Bed
—X—0.1 g/g Dodecane 75% Bed
—A—0.2 g/g Dodecane 25% Bed H
—■—0.2 g/g Dodecane 75% Bed

Time (min.)

120

Figure 2.8 Hexane Challenge with Dodecane Contaminant at 80% RH

2.3.3

Nonane Challenge/Octane Contaminant

This system represents a more strongly adsorbed threat vapor with a less strongly
adsorbed contaminant vapor. Shown in Figures 2.9 and 2.10 are the results at 15 and 80% RH.
The 50% RH experiments were not conducted since the 80% RH experiment (Figure 2.10)
shows very little effect of moisture.

These results can be summarized simply. Whereas the more weakly adsorbed
simulant (hexane) is greatly affected by moisture, the more strongly adsorbed simulant (nonane)
is able to effectively displace moisture due to preferential adsorption. Therefore, nonane
breakthrough is affected more by the more strongly adsorbed contaminant (dodecane) loading
than moisture content and RH.

4000

3500

3000

f, 2500

•2 2000
ITS

S 1500

1000

500

. Baseline

—O—0.1 g/g Octane 25% Bed
—X—0.1 g/g Octane 75% Bed
—*—0.2 g/g Octane 25% Bed
—*—0.2 g/g Octane 75% Bed

0 20 40 60 80 100 120 140 160 180

Time (min.)

200

Figure 2.9 Nonane Challenge with Octane Contaminant at 15% RH

3500

3000

Figure 2.10 Nonane Challenge with Octane Contaminant at 80% RH

2.3.4 Nonane Challenge/Dodecane Contaminant

This system represents the more strongly adsorbed threat vapor (nonane) with the
more strongly adsorbed contaminant (dodecane). The results for this pair are shown in
Figures 2.11 and 2.12.

2.3.4.1 Effect of RH

Clean Bed (Baseline) - As in Section 2.3.3, there is a slight effect of adsorbed
water. The breakthrough time at 15% RH is about 160 min, and the breakthrough time at
80% RH is about 140 min.

Contaminated Beds - Visual observation shows that the results at 15 and 80% RH
are quite different. At 15% RH and 0.1 g/g dodecane loading, the difference in the break times
for the 25 and 75% contaminated beds is about 50 min. While at 80% RH and 0.1 g/g dodecane
loading, the difference in the break times for the 25 and the 75% contaminated beds is about
30 min.

2.3.4.2 Effect of Contaminant Loading

The effect of contaminant loading at 15% RH is not large. At 0.1 g/g and 0.2 g/g
dodecane loading, the difference is only about 8 min. Another interesting observation is that the
100% contaminated bed of 0.2 g/g loaded dodecane has a breakthrough time of about

25 rmn. This means that the remaining adsorption space, even where dodecane is adsorbed, is
being utilized, at least to some extent, by the nonane.

2.3.4.3 Effect of Contaminated Bed Fraction

The difference in breakthrough times between the 25 and 75% contaminated beds
at 15% RH, for both the 0.1 and the 0.2 g/g dodecane loadings is almost identical, about
50 min. At 80% RH, the difference for the 0.2 g/g dodecane loading is about 20 min greater -
30 min for the 0.1 g/g dodecane loading and 50 min for the 0.2 g/g dodecane loading.

4000

3500

4000

3500

Baseline

0.1 g/g Dodecane 25% Bed
—X— 0.1 gig Dodecane 75% Bed
— A — 0.2 gig Dodecane 25% Bed
—•—0.2 gig Dodecane 75% Bed
— 0.2 gig Dodecane 100% Bed

100 150

Time (min.)

250

Figure 2.12 Nonane Challenge with Dodecane Contaminant at 80% RH

2.4 Summary of Breakthrough Results

The breakthrough results confirm the complexity of a true RLI. It is shown that
the breakthrough time (filter life) is dependent on a number of parameters such as RH,
adsorption strength of the contaminant, and the loading of the contaminant. The sensitivity that
each of these parameters has on breakthrough behavior is critically dependent on the adsorption
behavior of the threat chemical.

If hexane is representative of the threat vapor:

• A contaminant that adsorbs as strongly as octane can prevent hexane from
adsorbing (although not completely).

• A contaminant that is very strongly adsorbed such as dodecane can almost
completely exclude hexane at 0.2 g/g loading.

• Adsorbed water can greatly affect residual life.

If nonane is representative of the threat vapor:

• Octane has a small effect on breakthrough time. Nonane can displace
and/or adsorb the remaining adsorption volume, even at a 0.2 g/g octane
loading.

• Adsorbed water has only a small negative impact.

• Dodecane is a representative contaminant and can adversely affect
breakthrough times.

Therefore, an RL1 must consider both the contaminant and the threat vapor. The
implication is that less strongly adsorbed vapors such as CK will be affected to an even greater
extent than hexane. Therefore, the following conclusions must be taken into consideration:

• The threat vapor adsorption strength must be considered in the RLI test.

• The contaminant loading and adsorption strength relative to the threat
vapor can drastically change the true filter residual life.

• A safe-sided approach is to assume the worst situation, that is hexane
threat simulant and dodecane adsorbed contaminant

In order to convert the data measured here in Section 2.0 to filter life, it is
necessary to consider the effects of mass transfer. The breakthrough time in these experiments is
determined at half of the challenge concentration, about 2,000 mg/m\ This is close to the
stoichiometric center of a single transition wave, and the adsorption capacity can easily be
calculated from the value. The calculated capacity cannot be directly correlated to the residual
life since the mass transfer zone (MTZ) has not been taken into account. The MTZ is the length
of bed where the adsorption wave transitions from the challenge concentration down to the
breakthrough concentration. For an M98, that constitutes about 35% of the bed. For the
following breakthrough experiments, one can estimate the bed “residual life” by subtracting
about 35% of time off the stoichiometric center breakthrough time of a clean bed. Thus, a
breakthrough time of 100 min w'ould imply a residual life of about 65 min. A breakthrough time
of 35 min from a contaminated bed would imply instant break, thus a residual life of zero.

Given below in Tabic 2.4 arc the stoichiometric times (time where 2,000 mg/m 3 is
measured in the effluent) and estimated breakthrough times using an MTZ that is 35% of the
bed. The clean bed stoichiometric time breakthrough is multiplied by 0.65 to get the estimated
breakthrough time for a clean bed. If wc assume that the MTZ is not affected by RH, then every
clean bed result can be multiplied by 0.35 to get the “offset” time. The “offset” time is the
amount of protection “consumed” the MTZ, and it is subtracted from the stoichiometric time to
get the estimated breakthrough time.

For example, for hexane to beds with 0.1 g/g octane loading, the clean bed time is
105 min. The estimated breakthrough is 0.65 x 105 = 68 min, while the “offset” time at 15% RH
is 0.35 x 105 = 37 min. We round to the nearest 5 min to get 70 and 35 min, respectively. So, the
estimated breakthrough time for the 25 contaminated bed is 80 - 35 = 45 min, and the estimated
breakthrough time for the 75% contaminated bed is 55 - 45 = 10 min.

Table 2.4 Estimated Breakthrough Times Using MTZ that is 35% of Bed

Challenge

Vapor

Contaminant

Bed

Fraction

(%)

(g/g)

(%)

Stoichiometric
Center Times
(min)

Breaktime
Estimates
Using MTZ
(min)

Hexane

Octane

0, 25,75

0.1

105, 80,55

70,45,20

Hexane

Octane

25,75

0.2

80,30

45, instant

Hexane

Octane

0, 25,75

0.1

105,90,60

70,55,25

Hexane

Octane

25,75

0.2

75,35

40,instant

Hexane

Octane

0, 25,75

0.1

40,35,20

25,20,5

Hexane

Octane

25,75

0.2

35,15

C 20, instant ]

Hexane

Dodecane

0,25,75

0.1

105,95,70

70,60,35

Hexane

Dodecane

25,75

0.2

85,35

50, instant

Hexane

Dodecane

0, 25,75

0.1

105,95,60

70,60,25

Hexane

Dodecane

25,75

0.2

85,35

50,instant

Hexane

Dodecane

0, 25,75

0.1

35,30,30

20,15,15

Hexane

Dodecane

25,75

0.2

35,10

20, instant

Nonane

Octane

0, 25,75

0.1

140, 125,120

90,75,70

Nonane

Octane

25,75

0.2

130,145

80,95

Nonane

Octane

0, 25,75

0.1

135,140,140

85,90,90

Nonane

Octane

25,75

0.2

140,135

90,85

Nonane

Dodecane

0, 25,75

0.1

140,110,60

90,60,10

Nonane

Dodecane

25,75

0.2

100,30

50, instant

Nonane

Dodecane

0, 25,75

0.1

140,130,90

90,80,45

Nonane

Dodecane

25,75

0.2

110,50

60, instant

For every system where 75% of the bed is contaminated with a 0.2 g/g loading,
there is instant break, except for the nonane feed to octane contaminated beds system where
nonane is virtually unaffected by adsorbed octane.

3. PULSE TESTING

In Section 2, data were measured that correlate directly to the remaining fdter life.
Several different parameters were investigated to better understand the true meaning of residual
life. In this section, we will probe each of the bed configurations tested in Section 2 using a non¬
destructive chemical vapor pulse.

3.1 Pulse Testing Concept and Approach

The satellite beds are probed using a non-destructive gas, typically referred to as a
chemical pulse method. The basic requirements of the probe gas are that it is weakly adsorbed,
non-toxic, non reactive, and easily detectable. Even though the probe gas is weakly adsorbed, the
carbon will still retain it for a measurable period of time. As the bed becomes filled with

contaminants, the retention time of the chemical pulse will be reduced. In addition, based on
basic fundamentals of adsorption/chromatography science, the peak (maximum) chemical
concentration will rise as the level of bed contamination increases.

The two major problems with a chemical pulse method to determine the filter
residual are listed below:

1. A large amount of test chemical may be required, especially for large filter
installations that employ a large number of M98’s.

2. Any non-destructive pulse chemical will almost certainly be adversely
affected by adsorbed water. Thus, adsorbed water will be interpreted as a
potential contaminant (the estimated filer residual life w'ill be lower) when
we already know that most agents arc not adversely affected by adsorbed
water.

The RL1 approach overcomes these two shortcomings by using (1) a satellite
cartridge that samples the inlet air to mimic the environmental history of the filter or filter system
and (2) a second satellite filter at the filter outlet to serve as a reference for RH.

3.2 Pulse Testing System Apparatus and Procedure

The system is designed to make sure a precise mass of pulse chemical is
introduced into each bed for each test. This is critical in order to accurately assess the pulse tests
results since even a 0.1 g/g increase in mass delivered could produce a result that is interpreted as
a leak.

3.2.1 Apparatus

An infrared analyzer (MlRAN 1A) is used to detect the bed probe chemical. It is
set to a wavelength of 10.3 microns, a path length of 20.25 m, and an absorbance sensitivity of
.IX to detect the probe chemical. The air mixing chamber supplies humidified air to the test
beds, and has a volume of 4.5 L. The dew point hygrometer samples the humidified air from the
mixing chamber to determine RH.

Rotameters are used to control the flow rate through the sample beds. Their range
is 0 to 23 L/min, and for this test, they are set at 7.8 L/min to meter the correct velocity
(25.4 cm/s) through the bed.

The beds are 1-in. diameter and are made from stainless steel. A spring and a
porous top plate are used to keep the carbon particles from fluidizing during the test. Bed depth
should be sized to mimic the filters in the filter bank. For this test, the beds used are 5.6 cm in
depth and tested at an air flow velocity of approximately 25.5 cm/s.

Solenoid valves are used to control the flow of probe chemical through the test
system. The valves are 14 in., 2-way valves. Bed selection valves are used to select the bed to be
tested. These valves are % in., 3-way valves.

Type T thermocouples are used to measure the air temperature in the test system.

A mass flow controller is used to control the flow rate through the probe chemical
reservoir (volume = 20 mL) during the test. It is set at 40 mL/min. It is rated at 0 to 100 mL/min.

Figure 3.1 shows a piping and instrumentation diagram (P&ID) of the RLI pulse

apparatus.

Dew Point Hygrometer

HumkfrOed Air

Vacuum

- 0 -

Probe Chemical
Reservoir

Vent

== ^$ 1

ir^Kd

Probe

Chemical

Mass Flow
Controller

To Vacuum

Key

— Solenoid Valve

0 — Bed Selection Valve for Chemical
-HD — Thermocouple

Figure 3.1 RLI Pulse Apparatus P&ID

3.2.2 Bed Preparation

1. Fill the reference bed to 5.6 cm with ASZM-T carbon. Measure bed depth
with a small metric ruler.

2. For the 25% contaminated bed, fill to 4.1 cm with ASZM-T carbon. For
the 75% contaminated bed, fill to 4.1 cm with ASZM-T carbon. Use fresh
ASZM-T to make up the rest of the bed for a total bed depth of 5.6 cm.

3. Before conducting the test, make sure beds are fully equilibrated to the
desired RH. This will take at least 2 hr for the 80% RH tests.

3.2.3

Test Procedure

1. Open valve 1 to vacuum down the probe chemical reservoir to 28 in. of
Hg (vacuum) to purge the reservoir, Once the gauge reads 28 in., close
valve 1. (Valves 2, 3, 4, and 5 should remain closed.)

2. Open valve 3 to fill the reservoir with the probe chemical until the gauge
reads 0 in. (ambient pressure). Once the gauge reads 0 in., close valve 3.
Open valve 1 to vacuum the reservoir back down to 28 in., purging any air
out of the reservoir, and then close valve 1. (Valves 2, 4, and 5 should
remain closed.)

3. Open valve 3 once again to fill the reservoir with the probe chemical until
the gauge reads 12 in. (vacuum) for the 15 and 50% RH tests (PFCB) or
until the gauge reads 6 in. (R123) for the 80% RH tests. Then close
valve 3. (Valves 1, 2, 4, and 5 should remain closed.)

4. Use the bed selection valves A and B to configure the correct flow path for
the first bed to be tested. Simultaneously open valves 2 and 4 and set the
mass flow controller to 40 mL/min. to start the test.

5. The response from the Miran 1A is sent to an SCX1 data acquisition
system. This, in turn, is being accessed and controlled using a PC. Data
are recorded continuously using Labview software. Monitor the 1R
analyzer for a response on the meter to see that the probe chemical is
being detected. The reading on the meter should peak and then fall back
to zero. When the reading reaches zero, the test is complete. When the
test is completed, set the mass flow controller back to 0 and close valves 2
and 4. (Valves 1, 3, and 5 should remain closed.)

6. Repeat test procedure steps 1 through 4 for the remaining bed.

3.3 Pulse Test Results

Pulse testing is performed using the same contaminated bed configurations tested
in Section 2. Two pulse gases are used depending upon the RH. PFCB is used for the 15 and
50% tests. R123 must be used at 80% RH, because the PFCB is almost totally displaced by
water, i.e., very little adsorption occurs, even in the reference bed. An additional set of tests is
performed at 65% RH using dodecane as the contaminant to demonstrate the likely maximum
RH where PFCB can be used as the probe chemical. Properties of the test gases are presented in
Tabic 3.1, and the experiments performed are summarized in Table 3.2.

Table 3.1 Pulse Gas Properties

Property

PFCB

R123

Molecular Weight

200.04 g/mol

152.93 g/mol

Boiling Point

-5.6 °C

27.8 °C

Water Solubility @ 20 °C

Insoluble

0.21 wt%

Liquid Density @ 20 °C

1.5 g/mL

1.48 g/mL

3.3.1 Testing Overview and Experiment Summary

Table 3.2 Complete List of Pulse Experiments

Challenge

Vapor

Contaminant

Contaminant
Bed Fraction
(%)

Contaminant

(g/g)

RH (%)

Figure #

PFCB

Octane

25,75

0.1

3.2

PFCB

Octane

25,75

0.2

3.3

PFCB

Octane

25,75

0.1

3.4

PFCB

Octane

25,75

0.2

3.5

R123

Octane

25,75

0.1

3.6

R123

Octane

25,75

0.2

3.7

PFCB

Dodecane

25,75

0.1

3.8

PFCB

Dodecane

25,75

0.2

3.9

PFCB

Dodecane

25,75

0.1

3.10

PFCB

Dodecane

25,75

0.2

3.11

PFCB

Dodecane

25,75

0.1

3.12

R123

Dodecane

25,75

0.2

3.13

R123

Dodecane

25,75

0.1

3.14

3.3.2 Octane Contaminant Results

The results of pulse tests conducted at 15, 50, and 80% RH using octane as the
contaminant are summarized in this section. There are two plots for each RH; one plot displays
the results for the 0.1 g/g octane loading and the other plot displays the 0.2 g/g octane loading
results. Every plot shown consists of two pairs of experiments (four in total): the reference bed
and 25 and 75%, respectively, of the contaminated bed. Every reference bed experiment is
different, since even at the same target RH, there may be a slight difference in the actual RH.

Each of these pulse experiments can be compared to a corresponding
breakthrough experiment. In the following discussion, the breakthrough reference plots will be
noted and used for comparison.

3.3.2.1

15% RH

Shown below in Figures 3.2 and 3.3 are the pulse results at 15% RH. In each
figure, there are two test “pairs”: a 25% contaminated bed with its corresponding reference run
and a 75% contaminated bed with its reference run. A reference test is performed with every
contaminated bed test to account for any deviation in RH and temperature that might occur. The
difference between the reference and contaminated beds should be proportional or related to loss
of adsorption capacity.

The peak concentration and peak time are used to analyze the test results. An
increase in peak concentration will usually occur with a decrease in peak time. This is consistent
with the well-characterized behavior of chemical vapors on chromatographic columns. As the
adsorption capacity increases, the peak time increases. With the increase in adsorption capacity
typically one will see a decrease in the peak concentration. This behavior is directly related to the
favorable shape of the adsorption isotherm (concave down). Therefore, the shorter peak times
and higher the peak concentrations relative to the reference bed correlate to less adsorption
capacity, and therefore a shorter residual life.

Figure 3.2 shows the results for the 0.1 g/g octane loading. For the 25%
contaminated bed, there is a measurable difference from the reference (pink and red lines). The
peak concentration increases from about 10 to about 15 ppm. For the 75% bed, the peak
concentration is about 55 ppm, an increase of about 45 ppm, and the peak time has shifted back
almost 40 min from the reference. Clearly this is indicative of a large amount of capacity loss
when going from a 25 to a 75% contaminated bed. Generally speaking, this is the same relative
trend that is observed in the breakthrough testing.

When comparing Figures 3.2 and 3.3, notice that the peak concentrations for the
contaminated beds at 0.2 g/g octane loading are considerably higher than those at 0.1 g/g. This is
especially noticeable for the 75% contaminated bed where the increase in peak concentration is
from 55 (0.1 g/g loading) to 250 ppm (0.2 g/g loading). This is clearly indicative of PFCB co¬
adsorbing in the available adsorption sites. For example, at lower loadings (0.1 g/g), more PFCB
adsorbs due to more available pore volume, resulting in lower peak elution concentrations.

Concentration (ppm) Concentration (ppm)

Figure 3.2 Results for 0.1 g/g Octane Loading at 15% RH Using PFCB

Figure 3.3 Results for 0.2 g/g Octane Loading at 15% RH Using PFCB

3.3.2.2

50% RH

Shown below in Figures 3.4 and 3.5 are the pulse results at 50% RH. There are
several key features to highlight. First, the same trend observed at 15% RH is seen at
50% RH. There is a large difference between the 25 and 75% bed results. Second, the effect of
adsorbed water is seen in the reference results as well as in the contaminated bed results. For
example, the reference at 15% RH from Figure 3.2 has a peak concentration of about 10 ppm at
about 40 min, while the reference bed at 50% RH has a peak time of about 35 ppm at about
17 min. In addition, the peak concentrations of the contaminated beds are much higher than their
15% RH counterparts. For example, the 75% contaminated bed with 0.1 g/g octane loading at
50% RH has a peak concentration of 170 ppm compared to 55 ppm at 15% RH.

As with the 15% RH tests, the peak concentrations also increase as the octane
loading is increased.

Figure 3.5 Results for 0.2 g/g Octane Loading at 50% RH Using PFCB

3.3.2.3 80% RH

At 80% RH, PFCB is not viable, since very little PFCB adsorbs on clean carbon.
Therefore, a more strongly adsorbed pulse chemical, R123, is required.

Shown below in Figures 3.6 and 3.7 are the pulse results at 80% RH. These
results are very different from the results for PFCB at the lower RH. First, the peak elution times
are almost the same for all three beds: reference, 25 and 75% contaminated. Second, the peak
times occur very early, all within 5 min or so. However, the trend still holds. The largest peak
concentration is seen with the 75% contaminated bed for both octane loadings.

3.3.3

Dodecane Contaminant Results

3.3.3.1 15% RH

Shown in Figures 3.8 and 3.9 are the results using dodecane as the contaminant at
15% RH. These results are similar to those observed for octane indicating that octane and
dodecane have about the same effect on PFCB. In other words, any vapor equal to or greater than
octane in adsorption strength will generate the same RLI result with PFCB..

■0.1 gig Dodecane 25% Bed Effluent Concentration
•Reference 0.1 gig Dodecane 25% Bed Effluent Concentration
•0.1 gig Dodecane 75% Bed Effluent Concentration
•Reference 0.1 g/g Dodecane 75% Bed Effluent Concentration

Time (min.)

Figure 3.8 Results for 0.1 g/g Dodecane Loading at 15% RH Using PFCB

3.33.2 50% RH

Shown in Figures 3.10 and 3.11 are the results using dodecane as the contaminant
at 50% RH. These results are similar to those observed for octane showing that the change in RH
does not change the fact that any vapor more strongly adsorbed than octane will produce the
same result.

Concentration (ppm) Concentration (ppm)

Figure 3.10 Results for 0.1 g/g Dodecane Loading at 50% RH Using PFCB

Figure 3.11 Results for 0.2 g/g Dodecane Loading at 50% RH Using PFCB

3.3.3.3

65% RH

Shown in Figure 3.12 are the results using dodecane with a 0.2 g/g loading at
65% RH. These experiments were conducted to identify the RH limit where PFCB can be
effectively used as an indicator.

While these results do show the same trend as described previously and can likely
be used to estimate the residual life, PFCB elutes through the bed quickly (short peak times) and
at relatively high concentrations. At 65% RH, we are very close to the usable limit of PFCB,
since a reliable indicator must be able to differentiate between the reference and the
contaminated bed. If the reference bed has a very low adsorption capacity, then quantifying the
difference between reference and contaminated beds is more difficult.

Figure 3.12 Results for 0.2 g/g Dodecane Loading at 65% RH Using PFCB

3.3.3.4 80% RH

Shown in Figures 3.13 and 3.14 are the results using dodecane as the contaminant
at 80% RH. The pulse chemical used was R123. These results are similar those observed for
octane.

Concentration (ppm) Concentration (ppm)

160

140

120

-0,2 g/g Dodecane 25% Bed Effluent Concentration
-Reference 0.2 g/g Dodecane 25% Bed Effluent Concentration
-0,2 g/g Dodecane 75% Bed Effluent Concentration
-Reference 0.2 g/g Dodecane 75% Bed Effluent Concentration

15 20 25

Time (min.)

Figure 3.14 Results for 0.2 g/g Dodecane Loading at 80% RH Using R123

3.4 Pulse Testing Summary

The following arc some key conclusions drawn from these pulse tests:

• The effect of RH is dramatic when using PFCB. Compare reference beds,
for example, at 15 and 50% RH where the peak time goes from 40 to
about 17 min. This is a major problem that has confronted attempts to
develop an RL1 in the past. The reference bed approach allows one to
“subtract” the effect of RH.

• PFCB pulse tests are similar between dodecane and octane. Therefore,
PFCB will give us a weakly adsorbed vapor response (safe sided).

• The limit where PFCB is effective is close to 65% RH.

• The relative response of RI23 to contaminated bed changes is much
smaller than that of PFCB, but it appears to be great enough to provide a
measurable difference.

4. RL1 CORRELATION

As we have shown from the breakthrough results and the pulse test results, RH
(water adsorption) can dramatically change the relationships between variables. In addition, we
have seen that the adsorption behavior of the threat vapor and the contaminant can also have a
measurable effect on the breakthrough time.

A qualitative review of the pulse gas results at each RH show that the general
trend from 25 to 75% contaminated remains consistent. At the higher RHs, the reference bed
allows much more PFCB to penetrate earlier so that the difference between the reference and
contaminated beds is much less. This in turn diminishes the resolution in the test, e.g., it becomes
much more difficult to determine if the residual life is 70 or 30% of the original filter life. The
resolution issue is especially true for the R123 results.

In terms of the procedure to be followed in operational situations for the dual¬
probe chemical system, the reference bed would be tested first using PFCB. If the peak
concentration is below a preset value and the peak time is greater than a preset value, then PFCB
can be used to probe the contaminated bed. Otherwise, R123 must be used for both beds. For
example, in this work, a maximum PFCB concentration for the reference might be set at
200 ppm and a minimum peak retention time would be set at 3 min (see Figure 3.12). Clearly
another alternative could be to keep the RH for both the reference and contaminated beds below
60% or so. This could be accomplished by heating the air entering the satellite beds to
approximately 35 °C.

RLI Estimation Algorithm

An algorithm that provides a reasonable correlation for residual life is as follows:

1. Find the reference peak time.

2. Integrate the reference effluent concentration from time = 0 to the peak
time.

3. Integrate the corresponding contaminated effluent concentration from
time = 0 to the peak time.

4. Compute the ratio of the two numbers (contaminated mass/reference
mass).

5. Calculate the residual life by applying the MTZ effect.

As an example, let us examine how this would work for a 0.1 g/g bed
contaminated with octane at 15% RH. Refer to Figure 3.2. The peak time for the reference is
about 40 min. If we integrate under the effluent concentration curve for the 25% contaminated
bed up to 40 min, we get 24 mg. Performing the same integration for the reference concentration
curve up to 40 min results integrated mass of 8 mg.

A qualitative observation of the 15 and 50% RH reference bed results for all of
the experiments shown in the plots in Section 3.0 is the large difference in reference bed results.
We should be able to use the large change to scale our results based on the implied RH of the
reference bed result. For example, if the mass and peak time for 15% RH results were kept as a
baseline then the mass ratio could be scaled using this information. In this case, if we multiply
the 50% RH results by 1.5 the mass ratios between 15 and 50%RH, the experiments are much
closer.

Tables 4.1 and 4.2 summarize the correlated results for PFCB using octane and
dodecane as the contaminant.

The next to last column is generated by integrating the effluent breakthrough
curve up to the peak time. The last column is the ratio of the effluent mass of chemical between
the contaminated and the reference bed. It is that value we propose to use to correlate to residual
life.

From inspection, integrated ratio numbers below about 2.5 correspond to about
25% contaminated beds. Numbers above 6 or so correspond to the 75% contaminated beds. From
a residual life perspective, if one assumes that about a quarter of the bed is mass transfer zone,
then 25% contaminated bed will result in a 33% loss of protection time and a 75% contaminated
bed will result in an almost immediate breakthrough.

Table 4.1 Summary of PFCB with Octane as Contaminant

Contaminant

Bed

(g/g)

Bed

(%)

Peak

Time

(min)

Peak

Cone

(ppm)

Integrated
Mass at Ref.
Peak Time

(mg)

Integrated

Ratio

Contam/Ref.

Octane

0.1

3.0

Reference

Octane

0.1

5.6

9.2

Reference

Octane

0.1

9.6

2.0

Reference

Octane

0.1

3.7

171

5.7

Reference

Octane

0.1

3.5

Reference

Octane

0.2

2.2

249

13.2

Reference

Octane

0.2

6.6

2.8

Reference

Octane

0.2

305

8.5

Reference

Table 4.2 Summary of PFCB with Dodecane as Contaminant

Contaminant

(%)

Bed

(%)

Peak

Time

(min)

Peak

Cone

(ppm)

Integrated
Mass at Ref.
Peak Time
(mg)

Integrated

Ratio

Contam/Ref.

Dodecane

0.1

2.78

Reference

Dodecane

0.1

7.88

Reference

Dodecane

0.1

9.2

1.59

Reference

Dodecane

0.1

3.8

3.93

Reference

Dodecane

0.2

2.82

Reference

Dodecane

0.2

10.31

Reference

Dodecane

0.2

2.25

Reference

Table 4.2 Summary of PFCB with Dodecane as Contaminant (Continued)

Contaminant

(%)

Bed

(%)

Peak

Time

(min)

Peak

Cone

(ppm)

Integrated
Mass at Ref.
Peak Time
(mg) .

Integrated

Ratio

Contam/Ref.

Dodecane

0.2

2.7

174

5.56

Reference

Dodecane

0.2

1.6

562

2.09

Reference

2.5

323

Dodecane

0.2

2.5

286

1.91

Reference

3.5

189

Shown in Table 4.3 are the results for all of the 80% experiments using R123.
These results are not as consistent as the results for the PFCB. For example, consider the
integrated mass ratio values

Table 4.3 Summary of 80% RH Experiments Using R123

Contaminant

(%)

Bed

(%)

Peak

Time

(min)

Peak

Cone

(ppm)

Integrated
Mass at Ref.
Peak Time
(mg)

Integrated

Ratio

Contam/Ref.

Octane

0.1

3.9

103

5.8

2.3

Reference

4.4

2.5

Octane

0.1

3.9

137

11.9

2.5

Reference

4.1

4.7

Octane

0.2

4.4

4.3

1.2

Reference

4.4

3.6

Octane

0.2

4.1

113

2.7

Reference

4.4

3.3

Dodecane

0.1

4.2

1.7

Reference

4.2

Dodecane

0.1

3.1

184

3.0

Reference

3.8

123

Dodecane

0.2

3.4

124

3.8

Reference

4.8

Dodecane

0.2

3.9

137

2.6

Reference

4.2

Given in Tables 4.4 and 4.5 below are integrated mass ratios along the estimated
breakthrough times from Table 2.4.

Table 4.4 Break Time, Mass Ratio Comparison with Octane as Contaminant

Octane

(%)

Bed

(%)

Rll

(%)

Integrated
Mass Ratio
Contam/Ref.

Estimated

Hexane

Break

Time

(min)

Hexane

Residual

Life

(%)

Estimated
Nonane
Break Time
(min)

Nonane

Residual

Life

(%)

0.1

3.0

0.1

9.2

0.1

2.0

0.1

5.7

0.1

2.3

0.1

2.5

0.2

3.5

0.2

13.2

Instant

None

0.2

2.8

0.2

8.5

Instant

None

0.2

1.2

0.2

2.7

Instant

None

Table 4.5 Break Time, Mass Ratio Comparison with Dodecane as Contaminant

Dodecane

(%)

Bed

(%)

Integrated
Mass Ratio
Contam/Ref.

Estimated

Hexane

Break

Time

(min)

Hexane

Residual

Life

(%)

Estimated

Nonane

Break

Time

(min)

Nonane

Residual

Life

(%)

0.1

2.8

0.1

7.9

0.1

1.6

n/a

0.1

3.9

n/a

0.1

1.7

0.1

3.0

0.2

2.8

0.2

10.3

Instant

None

Instant

None

0.2

2.3

n/a

0.2

5.6

Instant

None

n/a

0.2

3.8

0.2

2.6

Instant

None

Instant

None

SUMMARY

In this report, we have quantified the effect of heavy ambient contaminants on
agent simulant breakthrough and developed a chemical vapor pulse method to determine the
residual capacity, or residual life, of in-service ColPro filters. Many conclusions have been
drawn from the data collected in this report.

In quantifying the effects of heavy contaminants on agent simulant breakthrough,
we reached the following conclusions:

1. If hexane is a representative of the threat vapor (a moderate-to-high
vapor pressure TIC), then

• A contaminant that adsorbs as strongly as octane can prevent hexane
from adsorbing, although not completely.

• A contaminant that is very strongly adsorbed, such as dodecane, can
almost completely exclude hexane at 0.2 g/g loading.

• Adsorbed water can greatly affect residual life.

2. If nonane is a representative of the threat vapor (a low vapor pressure
nerve agent), then

• Octane has a small effect on the breakthrough time. Nonane can
displace and/or adsorb in the remaining adsorption volume, even at a
0.2 g/g octane loading.

• Adsorbed water has only a small negative impact.

• Dodecane as a representative contaminant can adversely affect
breakthrough times.

Some key conclusions from the pulse tests are

• The effect of RH is dramatic when using PFCB. Compare reference
beds, for example, at 15 and 50% RH where the peak time goes from
40 to about 17 min. High RH has been the major hindrance to developing
an RLI in the past. The reference bed approach allows one to subtract the
effect of RH.

• PFCB pulse tests are similar between dodecane and octane. Therefore,
PFCB will give us a weakly adsorbed vapor response (safe-sided).

• The limit where PFCB is effective at probing the carbon bed for
contaminants is approximately 65% RH.

• The relative response of R123 to a contaminated bed is much smaller than
that of PFCB, but appears to be great enough to provide a measurable
difference.

A method has been developed to correlate vapor pulse response to the remaining
life of in-service filters. Major conclusions are listed below:

• The dual satellite bed approach to “subtract out" the effects of adsorbed
water has been shown to be effective

• The pulse results using the mass ratio calculation correlate well to the
estimate breakthrough times for most of the test conditions investigated.

• Results for R123 at 80% RH are not as reliable as the results for PFCB at
15 and 50% RH. A reasonable approach in future development would
include keeping the satellite beds at an elevated temperature to maintain a
RH below 60% or so.

• The RLI concept proposed here has been fully evaluated and it is ready for
technology transition to an engineering prototype.

Blank

LITERATURE CITED

1. Favas, George. End of Service Life Indicator (ESL1) for Respirator Cartridges. Part I.
Literature Review, DSTO-TN-0657; Defence Science and Technology Organisation: Canberra,
Australia, 2005.

2. Lawhon, S.J.; Richardson, A.W.; Hofacre, K.C.; Gardner, P. Development and Design of a
Colorimetric End-oJ-Service-Life Indicator, ECBC-CR-065; U.S. Army Edgewood Chemical
Biological Center: Aberdeen Proving Ground, MD, 2003; UNCLASSIFIED Report

(AD B298 053).

3. Mix, T.W.; McDonald, T.C. A Vapor Challenge Method of Measuring the Residual Life of
Gas Filters; CRDEC-CR-086; U.S. Army Chemical Research Development and Engineering
Center: Aberdeen Proving Ground, MD, 1990; UNCLASSIFIED Report (AD A227 711).

4. Smith, M.E. The Adsorption of Pulses of Vapour through Charcoal Beds, Master of
Philosophy Thesis, Chemical Defence Establishment: Porton Down, Salisbury, Wiltshire,
England, 1983.

5. Hammarstrom, J.; Bac, N.; Sacco, A., Jr. Residual Life Method for Determining Gas
Protection of A SC Whetlerite Carbon Beds; ARCSL-CR-83046; U.S. Army Chemical Systems
Laboratory: Aberdeen Proving Ground, MD, 1983; UNCLASSIFIED Report (AD B079 938).

6. Kladnig, W.F.; Weiss, A.H. Residual Protective Life of Carbon Beds; Quarterly Progress
Report #9; Worcester Polytechnic Institute: New England, 1978.

7. Van Dongen, R.H. Non-Destructive Test of Charcoal Filters. Part III. Characteristics of the
elution peaks of ethane and propane on partially spent charcoal filters of various moisture
content. TNO: Brussels, 1975.

8. Wheat, J.A.; Hyde, J.C. Estimation of the Residual Adsorption Capacity> of Charcoal Filters;
Report No. 663; Defence Research Establishment: Ottawa, 1972.

9. Karwacki, C.J.; Morrison, R.W. Adsorptive Retention of Volatile Vapors for Nondestructive
Filter Leak Testing. Ind. Eng. Chem. Res. 1998 , 37, 3470-3480.

10. Peterson, G. W.; Jones, P.; Keller, J.; Weller, E. Contaminant-Induced Degradation of
ASZM-TEDA Reactivity. ECBC-TR-629; U.S. Army Edgewood Chemical Biological Center:
Aberdeen Proving Ground, MD, 2008; UNCLASSIFIED Report (AD B341 218).

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "DTIC ADA491483: A Residual Life Indicator (RLI) for Physical Adsorption Capacity of Nuclear, Biological, and Chemical Filters. Part 3. A Novel RLI Design for Collective Protection Demonstrated Using Breakthough and Chemical Pulse Data"

See other formats