TECHNICAL REPORT

ANSI/ISATR92.06.031999

Feasibility of Chlorine
Detection Instrument Testing

Approved 1 January 2000

ANSI/ISATR92.06.031999
Feasibility of Chlorine Detection Instrument Testing

ISBN: 978-1-55617-701-9

Copyright 1999 by the Instrument Society of America. All rights reserved. Printed in the United
States of America. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means (electronic, mechanical, photocopying, recording, or
otherwise), without the prior written permission of the Publisher.

ISA
67 Alexander Drive
P.O. Box 12277
Research Triangle Park, North Carolina 27709
 3 ANSI/ISATR92.06.031999

PREFACE

This preface, as well as all footnotes and annexes, is included for information purposes and is not part of
ANSI/ISA-TR92.06.03-1999.

This Standard has been prepared as part of the service of ISA, the international society for measurement
and control, toward a goal of uniformity in the field of instrumentation. To be of real value, this document
should not be static but should be subject to periodic review. Toward this end, the Society welcomes all
comments and criticisms and asks that they be addressed to the Secretary, Standards and Practices
Board; ISA; 67 Alexander Drive; P. O. Box 12277; Research Triangle Park, NC 27709; Telephone (919)
549-8411; Fax (919) 549-8288; E-mail: standards@isa.org.

The ISA Standards and Practices Department is aware of the growing need for attention to the metric
system of units in general, and the International System of Units (SI) in particular, in the preparation of
instrumentation standards. The Department is further aware of the benefits to USA users of ISA standards
of incorporating suitable references to the SI (and the metric system) in their business and professional
dealings with other countries. Toward this end, this Department will endeavor to introduce SI-acceptable
metric units in all new and revised standards, recommended practices, and technical reports to the
greatest extent possible. Standard for Use of the International System of Units (SI): The Modern Metric
System, published by the American Society for Testing & Materials as IEEE/ASTM SI 10-97, and future
revisions, will be the reference guide for definitions, symbols, abbreviations, and conversion factors.

It is the policy of ISA to encourage and welcome the participation of all concerned individuals and interests
in the development of ISA standards, recommended practices, and technical reports. Participation in the
ISA standards-making process by an individual in no way constitutes endorsement by the employer of that
individual, of ISA, or of any of the standards, recommended practices, and technical reports that ISA
develops.

CAUTIONISA ADHERES TO THE POLICY OF THE AMERICAN NATIONAL STANDARDS

INSTITUTE WITH REGARD TO PATENTS. IF ISA IS INFORMED OF AN EXISTING PATENT THAT IS
REQUIRED FOR USE OF THE STANDARD, IT WILL REQUIRE THE OWNER OF THE PATENT TO
EITHER GRANT A ROYALTY-FREE LICENSE FOR USE OF THE PATENT BY USERS COMPLYING
WITH THE STANDARD OR A LICENSE ON REASONABLE TERMS AND CONDITIONS THAT ARE
FREE FROM UNFAIR DISCRIMINATION.

EVEN IF ISA IS UNAWARE OF ANY PATENT COVERING THIS STANDARD, THE USER IS
CAUTIONED THAT IMPLEMENTATION OF THE STANDARD MAY REQUIRE USE OF TECHNIQUES,
PROCESSES, OR MATERIALS COVERED BY PATENT RIGHTS. ISA TAKES NO POSITION ON THE
EXISTENCE OR VALIDITY OF ANY PATENT RIGHTS THAT MAY BE INVOLVED IN IMPLEMENTING
THE STANDARD. ISA IS NOT RESPONSIBLE FOR IDENTIFYING ALL PATENTS THAT MAY
REQUIRE A LICENSE BEFORE IMPLEMENTATION OF THE STANDARD OR FOR INVESTIGATING
THE VALIDITY OR SCOPE OF ANY PATENTS BROUGHT TO ITS ATTENTION. THE USER SHOULD
CAREFULLY INVESTIGATE RELEVANT PATENTS BEFORE USING THE STANDARD FOR THE
USERS INTENDED APPLICATION.

HOWEVER, ISA ASKS THAT ANYONE REVIEWING THIS STANDARD WHO IS AWARE OF ANY
PATENTS THAT MAY IMPACT IMPLEMENTATION OF THE STANDARD NOTIFY THE ISA
STANDARDS AND PRACTICES DEPARTMENT OF THE PATENT AND ITS OWNER.
ANSI/ISATR92.06.031999 4

ADDITIONALLY, THE USE OF THIS STANDARD MAY INVOLVE HAZARDOUS MATERIALS,

OPERATIONS OR EQUIPMENT. THE STANDARD CANNOT ANTICIPATE ALL POSSIBLE
APPLICATIONS OR ADDRESS ALL POSSIBLE SAFETY ISSUES ASSOCIATED WITH USE IN
HAZARDOUS CONDITIONS. THE USER OF THIS STANDARD MUST EXERCISE SOUND
PROFESSIONAL JUDGMENT CONCERNING ITS USE AND APPLICABILITY UNDER THE USERS
PARTICULAR CIRCUMSTANCES. THE USER MUST ALSO CONSIDER THE APPLICABILITY OF
ANY GOVERNMENTAL REGULATORY LIMITATIONS AND ESTABLISHED SAFETY AND HEALTH
PRACTICES BEFORE IMPLEMENTING THIS STANDARD.
The following people served as members of ISA Subcommittee SP92.06:

NAME COMPANY

R. Warburton, Chairman Industrial Scientific Corporation

D. Bishop, Managing Director Chevron Production Technology Company
W. Alexander Mine Safety Appliances Company
D. Alpha Detcon, Inc.
G. Black* Canadian Standards Association
L. Cerovich Exidyne Instrumentation Technology
J. Chilton NIOSH PRC
M. Coppler Ametek, Inc.
E. Croft Occidental Chemical Corporation
T. Donkin Enmet Corporation
S. Hildebrand Exxon Company USA
B. Holcom Gas Tech, Inc.
K. Johnson KWJ Engineering, Inc.
D. Li* Canadian Standards Association
A. Maynard Gas Measurement Inst. Ltd.
R. Menot Factory Mutual Research Corporation
J. Miller Detector Electronics Corporation
B. Northam Munro Electronics
R. Novack Ametek
R. Pellisier RKI Instrument, Inc.
K. Rollick MSA Instruments
M. Schaeffer Control Instruments Corporation
B. Shao CSA
M. Stryker NCASI
J. Thomason Omni Industrial Systems, Inc.
D. Wagner Industrial Scientific Corporation
D. Wechsler Union Carbide Corporation

The following people served as members of ISA Committee SP92:

NAME COMPANY

J. Thomason, Chairman Omni Industrial Systems, Inc.

D. Bishop, Managing Director Chevron Production Technology Company
W. Alexander Mine Safety Appliances Company
D. Alpha Detcon, Inc.
G. Black* Canadian Standards Association

______
* One vote per company.

K. Burden Sensidyne Inc.

 5 ANSI/ISATR92.06.031999

M. Coppler* Ametek, Inc.

T. Donkin Enmet Corporation
B. Henderson Chevron USA Inc.
B. Holcom Gas Tech, Inc.
K. Johnson KWJ Engineering, Inc.
A Maynard Gas Measurement Inst. Ltd.
R. Menot* Factory Mutual Research Corporation
F. McGowan* Factory Mutual Research Corporation
J. Miller Detector Electronics Corporation
D. Mohla Union Carbide Chemicals & Plastics
G. Naujoks* Keithley Instruments, Inc.
B. Northam Munro Electronics
R. Novack* Ametek, Inc.
R. Pellissier RKI Inst. Inc.
M. Schaeffer Control Instruments Corporation
W. Shao* Canadian Standards Association
A. Spataru The Adept Group Inc.
M. Stryker NCASI
P. Stupay* Keithley Instruments
D. Wagner* Industrial Scientific Corporation
R. Warburton* Industrial Scientific Corporation

This technical report was approved for publication by the ISA Standards and Practices Board on
15 April 1999.

NAME COMPANY

H. Dammeyer The Ohio State University

H. Baumann H. D. Baumann, Inc.
D. Bishop Chevron Petroleum Technology Co.
P. Brett Honeywell, Inc.
M. Cohen Senior Flexonics, Inc.
M. Coppler Ametek, Inc.
W. Holland Southern Company
A. Iverson Ivy Optiks
R. Jones Dow Chemical Co.
V. Maggioli Feltronics Corp.
T. McAvinew Instrumentation & Control Engineering LLC
A. McCauley, Jr. Chagrin Valley Controls, Inc.
R. McFarland Honeywell, Inc.
R. Reimer Rockwell Automation
J. Rennie Factory Mutual Research Corp.
R. Webb Altran Corp.
W. Weidman Parsons Energy & Chemicals Group
J. Weiss EPRI
J. Whetstone National Institute of Standards & Technology
M. Widmeyer Consultant
R. Wiegle CANUS Corp.
C. Williams Eastman Kodak Co.
G. Wood Graeme Wood Consulting
M. Zielinski Fisher-Rosemount Systems, Inc.

______
* One vote per company.
This page intentionally left blank.
 7 ANSI/ISATR92.06.031999

Foreword

Workplace safety

Elemental chlorine is a strong oxidizing agent1, and in water it disproportionates to form hypochlorous and
hydrochloric acids. This high chemical reactivity is reflected in the toxicity of chlorine. The workplace
exposure limit, TLV-TWA, is only 0.5ppm2. Consequently safety requirements of chlorine are a high
priority, and there is a considerable need for chlorine gas detection instruments for workplace safety.
These instruments may be either fixed or portable; and there are many companies that manufacture
instruments for chlorine gas detection in the workplace. In order to promote a safe work environment, the
ISA SP92 Committee is developing performance standards for gas detection instruments for workplace
safety. One of the subcommittees, SP92.06, has the responsibility of developing a standard for chlorine
gas detection.

Chlorine detection instrument performance tests

Several questions have been posed by the Committee about testing chlorine gas detection instruments.
Since the TWA-TLV is low (0.5ppm), it is desirable for the tests to be performed near this chlorine
concentration level. For the test to be valid, the concentration of the gas applied to the instrument under
test has to be known to a significantly higher precision and accuracy3 than the allowed tolerance of the
instrument response. As currently written, many of the instrument performance tests require the
instrument to give a reading or activate an alarm based on a difference in gas concentration of 0.1ppm.4

Effect of humidity5

The standard for chlorine gas detection instruments (ISA-S92.06.01-1998) includes a performance
requirement that the instrument detect a chlorine gas test mixture of known concentration within set
tolerances over a non-condensing humidity range between around 15 and 90+ percent. There has been a
lot of debate within the Committee about whether this test is valid. There are two principal arguments in
this discussion.

The first argument is that humidity is a very important parameter that should be tested since many of the
locations where chlorine detection instruments are used are very humid. Many paper mills, for example,
operate with the atmosphere almost constantly saturated with water vapor. Therefore, a performance
standard for chlorine detection instruments should include a test of the effects of humidity on the chlorine
sensor response.

______
1
Cl2(g) + 2e <==> 2Cl; E = 1.358V vs. SHE; CRC Handbook of Chemistry & Physics, 70th Edition, 1989-1990, R.C. Weast et al.
(Eds.), Publ. CRC Press, Boca Raton, FL.

2
1994-1995 Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices, ACGIH.

3
For this discussion, accuracy is a measure of how close the nominal chlorine concentration is to the actual chlorine concentration.
The precision is a measure of the variance of the chlorine concentration from the mean concentration upon repeating the same
measurement many times.

4 This report assumes that the atmosphere and gas pressure is one standard atmosphere (1.01325 Bars). At constant pressure,
and assuming ideal gas behavior, gas concentrations can be expressed as volume ratios, i.e. the volume of component divided by the
total gas volume, written as, for example, percent (%) or parts per million (ppm).

5 Relative Humidity - The ratio of the quantity of water vapor present in the atmosphere to the quantity that would saturate at the
existing temperature.
ANSI/ISATR92.06.031999 8

The second argument is that the test is not reasonable since chlorine may react with water in the vapor
phase, forming other reactive species. If species such as hypochlorous acid (HOCl) are being formed,
they may not diffuse into the detector in the same way as chlorine gas. The most common technology for
detecting chlorine in the workplace uses electrochemical sensors. Typically these sensors contain a
diffusion barrier consisting of a porous Teflon membrane, which the gas has to diffuse across to reach the
electrode/electrolyte where the detection reaction occurs. If hypochlorous acid is formed in the gas phase,
it is likely that it will absorb onto the first surface it meets, such as the porous membrane. The instrument
would therefore read low under high humidity conditions because the real chlorine concentration had been
decreased by the high humidity. Since this standard is for chlorine (Cl2) detectors and not active chlorine
species detectors; the humidity test may not be a valid performance criterion since the true chlorine (Cl2)
concentration would not be known at high humidities if these vapor-phase reactions occurred to a
significant extent.

The Committee agreed to try to find out as much as possible about the reactions of chlorine with water
vapor and combine these with experimental results to produce recommendations for ISA-S92.06.01-1998.
This report is a summary of the results of this investigation.

Abstract

This report has been written to answer questions about the feasibility of performing the tests described in
the ISA standard for chlorine gas detection instruments (ISA-S92.06.01-1998). The report discusses two
questions; the first is the availability of a chlorine gas supply of sufficient accuracy and precision for
performing the tests, and the second is the viability of testing a gas detection instrument with chlorine
under a variety of humidity conditions. The report surveys the available chemical literature, presents
experimental and calculated results of humidity testing, and summarizes information on test gas
generation and analysis.

Key Words

Accuracy, Chlorine, Cl2, Gas, HCl, HOCl, Hypochlorous acid, Humidity, Hydrochloric acid, Instrument,
Precision, Sensor, Standard
 9 ANSI/ISATR92.06.031999

CONTENTS

1 Introduction..................................................................................................................................... 11
2 Review of chlorine chemistry 6,7,8,9.............................................................................................. 11
2.1 Chlorine................................................................................................................................. 11
2.2 Chlorine oxides ..................................................................................................................... 12
2.3 Oxy acids and anions............................................................................................................ 13
2.4 Water solubility...................................................................................................................... 13
2.5 Vapor phase reactions .......................................................................................................... 13
3 Sensor manufacturers specifications............................................................................................. 14
3.1 Draeger16 ............................................................................................................................. 14
3.2 City Technology17 ................................................................................................................ 14
3.3 Transient humidity effects ..................................................................................................... 14
4 Experimental results ....................................................................................................................... 14
4.1 Overall approach................................................................................................................... 14
4.2 Chlorine instrument tests: Bacharach19 ............................................................................... 15
4.3 Chlorine instrument tests: Mine Safety Appliances20........................................................... 15
4.4 Chlorine instrument tests: NCASI21 ..................................................................................... 17
4.5 Chlorine instrument tests: National Draeger22 ..................................................................... 18
4.6 Equilibrium calculations 23................................................................................................... 19
4.7 Titration of humidified chlorine28 .......................................................................................... 21
5 Discussion of results....................................................................................................................... 22
5.1 Summary results from instrument/sensor tests..................................................................... 23
5.2 Specification in Standard ISA-S92.06.01-1998, Section 7.14.1............................................ 24
6 Sources of chlorine test gas ........................................................................................................... 24
6.1 Gas tanks .............................................................................................................................. 24
6.2 Electrochemical generators................................................................................................... 25
6.3 Permeation Tubes................................................................................................................ 25
7 Analysis of chlorine gas concentration ........................................................................................... 25
7.1 Literature methods for chlorine test gas analysis.................................................................. 25
8 Recommendations for testing with chlorine.................................................................................... 25
8.1 Experimental considerations ................................................................................................. 25
8.2 Correction for water vapor..................................................................................................... 26
9 Other gas performance standards.................................................................................................. 27
10 Acknowledgments ........................................................................................................................ 27
This page intentionally left blank.
 11 ANSI/ISATR92.06.031999

1 Introduction
This report has been written to provide support to the work of the ISA SP92.06 Subcommittee responsible
for drafting a performance standard for chlorine detection instruments. The scope of the standard is
limited to those instruments intended for the determination of chlorine gas in air to enhance workplace
safety. In Committee discussions two technical questions were raised; the first was the feasibility of
generating chlorine gas concentrations with sufficient accuracy and precision necessary for the
performance tests. The second question was whether there was a significant reaction between chlorine
and water in the vapor phase since the extent of this reaction would impact the validity of the proposed
humidity performance test.

The first part of the report discusses issues related to the question of the possible interaction of chlorine
with water vapor in the gas phase. The latter part of the report discusses the generation of chlorine gas.

This report is divided into several sections. The first section provides a brief overview of the chemistry of
chlorine and related compounds. This section is intended to provide an outline of the chemical basis for
later discussions and to summarize the results obtained from literature searches. Since a standard is by
necessity based on the current technology, the available information from sensor manufacturers is
summarized so that a comparison between the proposed specifications and manufacturers specifications
can be made. Several instrument tests have been performed by both the NCASI and several
manufacturers using both Draeger and City Technology chlorine sensors. These tests were aimed at
finding the effect that humidity has on the response of the instrument to a nominal chlorine concentration.
Lastly, a theoretical approach using free energy calculations was used to estimate the thermodynamic
feasibility of the proposed vapor-phase reactions between chlorine and water.

The second part of the report addresses sources of chlorine test gas and the feasibility of generating
chlorine test gas mixtures with sufficient accuracy and precision necessary to do the performance testing
called for by the standard (ISA-S92.06.01-1998). This section also provides information about the analysis
of chlorine test gas mixtures and a short section with some practical experimental recommendations for
working with chlorine.

2 Review of chlorine chemistry 6,7,8,9

The chemistry of chlorine is briefly reviewed as relevant to this report. Particular emphasis is placed on the
reactions of chlorine with water and the products and equilibrium formed. Further information may be
found in the references cited.

2.1 Chlorine

Elemental chlorine is a greenish yellow diatomic gas, Cl2, with a molecular weight of 70.9 g/mol; atomic
number 17; mp -101.0C, bp -34.05C; and its density is approximately two and a half times that of air.
Chlorine occurs most commonly in nature as the chloride ion in sea water and in salt deposits. It was first
isolated in 1774 by Karl Scheele, who did not recognize it as an element; the elemental nature of chlorine

______
6 Merck Index, 11th Edition, Publ. Merck & Co. Inc.,1989.

7 Kirk-Othmer Encyclopedia of Chemical Technology, Third Edition, Volume 1, John Wiley, New York, (1978).

8
F. A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, Fourth Ed., Publ. John Wiley, New York, 1980.

9 G. C. White, The Handbook of Chlorination and Alternate Disinfectants, Third Ed., Publ. Van Nostrand Reinhold, New York (1992).
ANSI/ISATR92.06.031999 12

was recognized some 30 years later in 1810 by Sir Humphry Davy.10 Chlorine is a very widely used
material, and chlorine manufacture ranked seventh by volume at 8.2 million metric tons for chemicals
produced in the USA (1975)11 and is currently around 13 million metric tons. The main source of chlorine
is the electrolysis of brine; and for the world, the amount of chlorine produced is estimated at 47.3 million
tons (1994). Chlorine and chlorine derivatives are used in a huge array of products ranging from polymers,
elastomers, solvents, lubricating oils, pharmaceuticals, paper chemicals, and so forth. Further details may
be found in a recent review of the chlorine industry and its environmental impact.12

2.2 Chlorine oxides

Chlorine forms a number of oxides. The oxides tend to be strong oxidizing agents and are very reactive.
They are briefly included in the report since they are related to the oxo anions discussed below.

Dichlorine monoxide (Cl2O)13 is a well characterized yellow-red gas, boiling point 2C that forms a yellow
solution in water, containing hypochlorous acid (HOCl) of which dichlorine monoxide is formally the
anhydride. The major use of Cl2O is as a bleaching agent for wood pulp. Cl2O is often prepared by
passing chlorine into sodium carbonate with only a small amount of water. This method produces aqueous
sodium chloride and hypochlorous acid (HOCl); but the solubility of hypochlorous acid is soon exceeded
and chlorine monoxide is evolved.

Cl2 + H2O <==> HOCl + HCl

2Cl2 + 2Na2CO3 + H2O ==> Cl2O + 2NaHCO3 + 2NaCl

and 2Cl2 + 2NaHCO3 ==> Cl2O + 2CO2 + 2NaCl

Dissolving dichlorine monoxide in water gives the reverse reaction and forms a solution of hypochlorous
acid.

Cl2O(g) + H2O(l) <==> 2HOCl(aq)

Studies of the vapor above an aqueous solution of hypochlorous acid have shown that both Cl2O(g) and
HOCl(g) are present, the former predominating at high concentrations and the latter at low concentrations.

Chlorine dioxide (ClO2)14,15 is a yellowish, very reactive gas that is made on a large scale, usually where it
is required. Water solutions decompose in the light to form hydrochloric acid (HCl) and chloric acid
(HClO3). It is commonly prepared from reaction of chlorine and sodium chlorite or from sodium
hypochlorite and sodium chlorite. Chlorine dioxide is typically used for bleaching purposes such as for
wood pulp, textiles, and disinfecting drinking water.

______
10 I. Asimov, Asimovs Biographical Encyclopedia of Science and Technology, Publ. Pan Books Ltd., London, 1975.

11
Kirk Othmer Encyclopedia of Chemical Technology, Third Edition, Volume 1, p799, Publ. John Wiley, New York, 1978.

12 B. Hileman, J.R. Long, E.M. Kirschner, C&EN, November 21, 1994, p12.

13 Further details of the chemistry of dichlorine monoxide (chlorine monoxide) can be found in J.J. Renard, H.I. Bolker, Chemical
Reviews (1976), 76, 487.

14
Further details can be found in W.J. Masschelein, R.G. Rice, Chlorine Dioxide, Publ. Ann Arbor Science Publishers Inc., Ann
Arbor MI (1979).

15 Ozone and Chlorine Dioxide Technology for Disinfection of Drinking Water, Ed. J. Katz, Publ. Noyes Data Corp., Park Ridge, NJ
(1980).
 13 ANSI/ISATR92.06.031999

Dichlorine heptoxide (Cl2O7) is the most stable chlorine oxide. It reacts with water and hydroxide to form
perchlorate ion (ClO4-) and has application in organic synthesis of alkyl perchlorates.

Other chlorine oxides exist, such as dichlorine tetroxide (ClOClO3) and dichlorine hexoxide (Cl2O6);
however, they are generally unstable.

2.3 Oxy acids and anions

There are several chlorine oxy acids and anions in which the formal oxidation state of chlorine occurs as
+1, +3, +5 and +7. The principal oxy acids are HOCl, HOClO, HOClO2, and HOClO3 respectively. The
interaction between all the species possible is beyond the scope of this report.

2.4 Water solubility

Chlorine is somewhat soluble in water, in which it reversibly reacts to form hydrochloric and hypochlorous
acids:

Cl2 + H2O <==> HCl + HOCl

The solubility in alkalies is much greater because the above equilibrium is pulled to the right-hand side.
Dissolution of chlorine into alkali and alkali-earth hydroxides is used to make the hypochlorites, which are
also powerful oxidizing agents, often used for bleaching.

2NaOH + Cl2 ==> NaOCl + NaCl + H2O

Chlorine dissolves in water (25C) with the formation of aqueous Cl2 (0.062M), HOCl (0.030M) and Cl-
(0.030M), giving a total solubility of 0.092M. Hypochlorous acid (HOCl) is a weak acid, with a pKa (HOCl)
around 7.5. The reaction between chlorine and water is fast, and the increase in rate at high pH has been
explained as a reaction between chlorine and hydroxide anion. At temperatures below 9.6C, yellow
chlorine hydrate crystals may form in solutions of chlorine water.

2.5 Vapor phase reactions

In the vapor phase, HOCl can be formed by the equilibrium reaction of chlorine dioxide (ClO2) with water:

H2O(g) + ClO2 (g) <==> 2HOCl(g)

The hypochlorite anion can disproportionate in aqueous solution over time to form the chlorate anion:

3ClO- <==> 2Cl- + ClO3-

The reaction is thermodynamically very favorable, with an equilibrium of 1027, but the reaction rate is slow
at room temperature but becomes fairly rapid at higher temperatures (~75C), and good yields of ClO3-
can be obtained. The chlorate anions produced can, in turn, slowly disproportionate further to give the
perchlorate anion (ClO4-). In summary, chlorine has a well-known chemistry in water in which it readily
dissolves, forming hypochlorous acid (HOCl) and hydrochloric acid (HCl). Hypochlorous acid is in
equilibrium with the anhydride ClO2 both in solution and in the gas phase, and hypochlorous acid may
undergo disproportionation to form chlorates and perchlorates under some circumstances. There is,
however, very little literature describing reactions of chlorine with water in the gas phase.
ANSI/ISATR92.06.031999 14

3 Sensor manufacturers specifications

If the humidity test is to be implemented in ISA-S92.06.01-1998, suitable specifications will be required.
Since standards are written around current technology, the sensor manufacturers specifications and
information are a good initial starting point for determining values for the specifications.

3.1 Draeger16

Operating range 10 to 95 percent r.h. non-condensing

Performance specification for the effect of humidity at 25C

Zero: < +/- 0.001 ppm/percent r.h.

Sensitivity: < +/- 0.1 percent of measured value / percent r.h.

These specifications are equivalent to a humidity response of +/- four percent of measured value,
assuming testing at 15 and 90 percent relative humidity, compared to the response at 50 percent RH.

3.2 City Technology17

Operating range

Continuous 15 to 90 percent non-condensing

Intermittent 0 to 99 percent non-condensing

3.3 Transient humidity effects

Most manufacturers specifications for humidity give the steady state behavior of the sensor over a range
of humidities. However, electrochemical sensors from many manufacturers exhibit transient humidity
effects after being exposed to a change in relative humidity. Typically, steady state changes in humidity
will not affect the sensor output. Step change in humidity, however, can produce a transient signal that
dies away after a few seconds.18 These effects would show up on tests that involve following the
instrument behavior upon exposing the instrument to the humidified gas stream, but would not show up if
the instrument were allowed to equilibrate with the new gas mixture prior to making the measurement.

4 Experimental results
4.1 Overall approach

A series of tests were performed involving testing either instruments or sensors with dry and humidified
chlorine test gas mixtures. The chlorine concentration was typically in the range 1 to 5 ppm. These tests
were run by four organizations: Bacharach Inc., Mine Safety Appliances Inc., NCASI and National Draeger
Inc., representing three manufacturers and a user.

______
16
Data sheet for Draeger Polytron II chlorine sensor, 6808865, 1st Edition, September 1994.

17
Specifications for Chlorine CiTicels, City Technology Ltd., Document Reference: PDH.pm4, April 1, 1993.

18 Information provided by Max Carcas, City Technology Ltd.

 15 ANSI/ISATR92.06.031999

These test results are intended to give the actual behavior of the sensor to humidified chlorine gas
mixtures. An increase or decrease in the output signal with humidity may be due to many factors, including
possible vapor-phase reactions between chlorine and water, increase in the adsorption of chlorine onto
instrument/sensor surfaces with increased water content, and so forth.

Whatever the fundamental cause, these test results will provide the main basis for determining the
specifications since a specification that cannot be met by the current technology is not useful. The
intention, therefore, has been to broaden the testing as far as possible, involving several organizations and
sensors from more than one manufacturer (Draeger and City Technology). Apart from the general goal, the
test methods and equipment were selected independently by each laboratory.

4.2 Chlorine instrument tests: Bacharach19

Tests were performed using a City Technology chlorine cell in a Sentinel 44T calibrated with 50 ppm
chlorine in nitrogen; the following results were obtained with a test gas concentration of 1.5 +/- 0.2 ppm
chlorine.

Effect of Humidity on Bacharach Chlorine Instrument Response

Relative Humidity (%) Sentinel Reading (ppm)

15 0.5

50 1.4

90 1.4

These results appear to indicate that high relative humidity does not have a large effect on the instrument
chlorine response, though the low chlorine response at 15 percent RH is unexpected.

4.3 Chlorine instrument tests: Mine Safety Appliances20

The results shown in the table below were obtained using a City Technology chlorine sensor in a Passport
Instrument exposed to between 1 and 5 ppm chlorine.

______
19
Test results supplied by Chuck Gardner and Debi Beyer of Bacharach Inc.

20 Results provided by Joseph Jolson of MSA Instrument Division.

ANSI/ISATR92.06.031999 16

Effect of Humidity on MSA Chlorine Instrument Response

1 ppm Cl2 2 ppm Cl2

% RH Passport Reading % Deviation from % RH Passport Reading % Deviation from

(ppm) Normal Gas Concn. (ppm) Normal Gas
Concn.

1.9 0.9 0 2.1 2.2 0

14.5 0.9 0 11.8 2.2 0

29 0.9 0 24.8 2.2 0

39 0.9 0 36.3 2.2 0

48.1 0.9 0 46.8 2.2 0

58.7 0.8 11 58.7 2.15 2

73.3 0.8 11 69.3 1.9 14

88.1 0.7 22 77.2 1.75 20

3 ppm Cl2 5 ppm Cl2

% RH Passport Reading % Deviation from % RH Passport Reading % Deviation from

(ppm) Normal Gas Concn. (ppm) Normal Gas
Concn.

2.2 3 0 1.9 4.8 0

3.8 3 0 2.7 4.8 0

11.6 3 0 3.5 4.8 0

26.3 3 0 14 4.8 0

37.2 3 0 27.6 4.6 4

48.8 2.8 7 39.3 4.5 6

58.2 2.6 13 49.3 4.2 13

69.2 2.4 20 ~58 4 17

These results indicate that there is little effect on the response to chlorine at low humidities, but as the
humidity increases the response to chlorine decreases.
 17 ANSI/ISATR92.06.031999

4.4 Chlorine instrument tests: NCASI21

The effect of humidity on the response of two chlorine detection instruments was measured. One
instrument was a PAC III instrument from National Draeger and the other was a portable instrument from
Bacharach. Chlorine was generated from permeation vials (Thermedics Inc.) housed in a U-shaped tube,
which was set in a constant temperature bath. A low flow of nitrogen was passed through the U-shaped
tube to sweep out the permeated gas, with the nitrogen flow rate controlled by a critical orifice. This
chlorine gas flow was Teed into cylinder air. The cylinder air was either dry or was humidified by passing it
through a fritted impinger filled with water. All apparatus that came into contact with the test gas was made
of either Teflon or glass. Test gases were analyzed using the spectrophotometric method described in
NCASI Technical Bulletin No. 521. The total flow rate for each test gas was set at 500 - 600 mL/min, the
rate specified for calibrating both monitors. Humidity was checked with an EdgeTech Model 911 Digital
Humidity Analyzer, with the chlorine vials taken out of line to protect the instrument, but the same flow of
nitrogen incorporated as if they were in line. Concentrations were changed by using one or three chlorine
permeation vials.

For all testing, the calibration adaptors made for the respective instruments were used when applying the
test gases. Initial response to a five-minute exposure to 1.0 ppm Cl2 in dry air was Draeger 1.10 and
Bacharach 0.3 ppm. The Draeger monitor was not recalibrated because it was close to being accurate.
Both monitors display readings tended to drift over time after being exposed to chlorine, so an arbitrary
three-minute exposure was applied for the subsequent tests. The Bacharach monitor was recalibrated to a
5.0 ppm chlorine dry-test gas (after a five-minute exposure, then zeroed, then the chlorine was reapplied
and spanned after three minutes). The Draeger instrument zeroed quickly after exposure when the
chlorine stream and the calibration cup were removed. The Bacharach unit was slower, so clean air was
applied through its calibration apparatus to hasten its return to zero. The tests were started the following
day, and the results are summarized in the following table.

All tests were performed at ambient laboratory temperature, about 22C, in an exhaust hood. After a three-
minute exposure to the test gas, a monitors reading was recorded. Exposure to a given test gas was
repeated three times for each monitor, with a 3- to 4-minute recovery time in room air (Draeger) or with
cylinder air, then room air (Bacharach) as the other monitor was exposed. Averages of three readings to a
given test are included in the table.

It was found that the Draeger monitor response was diminished with the humidified test gas (compared to
the dry test gas), and the Bacharach was enhanced.

______
21 Test method and results from Margaret N. Stryker, NCASI.
ANSI/ISATR92.06.031999 18

Results of NCASI Testing with Bacharach and National Draeger Instruments

Date Analyzed Chlorine Relative Humidity Draeger Instrument Bacharach

(ppm) Reading (%RH) Response Instrument
Response

4/2/96 5.00 5.4 6.35 5.2

Cl2 6.45 5.8
6.70 8.1
Average Average
6.50 6.37

4.98 86.5 3.50 6.8

Cl2 4.15 6.6
4.35 6.9
Average Average
4.00 6.77

4.94 6.6 6.35 8.0

Cl2 6.15 9.7
6.65 8.1
Average Average
6.38 8.6

4/4/96 1.45 6.1 1.80 2.2

Cl2 1.60 1.9
2.10 1.8
Average Average
1.83 1.97

1.45 88.0 0.90 2.4

Cl2 1.20 3.2
1.15 3.1
Average Average
1.08 2.90

4.5 Chlorine instrument tests: National Draeger22

Tests were conducted to measure the effect of humidity on the response of sensors to chlorine. The
chlorine test gas was generated by blending 300 ppm chlorine in nitrogen from a compressed gas cylinder,
with dry nitrogen with Mattheson mass-flow controllers. All tubing was Teflon, and the flow system had a
slow continual purge of chlorine gas mixture for several months to ensure that all surfaces had saturated.
The chlorine test gas was blended with humidified nitrogen at a T. The humidified nitrogen was produced
by blending dry nitrogen (boiled off from liquid nitrogen tank) with nitrogen saturated with water (passed
through two water bubblers in series fitted with gas dispersion tubes). The humidity was varied by
modifying the ratio of the dry and humidified nitrogen streams while keeping the total nitrogen flow rate
constant. The total flow to the sensor was 200 mL/min, using a Teflon calibration adaptor.

A test was performed with a Draeger PAC sensor in a Multi.pac instrument exposed to nominal 5 ppm
chlorine in which the relative humidity was varied between zero and 75 percent RH. The relative
humidities were only approximate. The nitrogen supply was assumed to be zero relative humidity, and the
humidified nitrogen coming out of the two water bubblers in series was assumed to be 100 percent. The
other values were calculated based on dilution. The results are as follows:

______
22 Results provided by Richard Warburton, National Draeger Inc.
 19 ANSI/ISATR92.06.031999

Effect of Humidity on National Draeger Chlorine Instrument Response

% Relative Humidity Instrument Reading (ppm) % Deviation from Nominal Gas

Concentration

0 5.0 0

12.5 5.0 0

25 4.6 8

37.5 4.6 8

50 4.3 14

75 4.4 12

The chlorine sensors respond to chlorine over the full range of the humidities up to the maximum tested
(75 percent RH); however, the response to chlorine appears to decrease at high humidities. Unfortunately,
with the experimental set-up available, it was not possible to increase the relative humidity beyond 75
percent RH. There is good agreement between these results and the corresponding 5 ppm chlorine
results obtained with the MSA tests above.

4.6 Equilibrium calculations 23

The proposed reaction occurring between chlorine and water in the gas phase is formation of HCl and
HOCl, by analogy to the well-known aqueous chemistry of chlorine. Unfortunately, no information was
found in the chemical literature documenting this reaction and, therefore, estimates were made about the
thermodynamics of the process. Since energy and mass are conserved in all reactions, one can use the
energy balance to estimate whether a reaction is probable or not; however, thermodynamics gives no
information about how fast a reaction will occur.24

Estimates were made to determine whether a reaction was likely to occur in the gas phase between water
and chlorine from thermodynamic calculations based on changes in the Gibbs free energy for the
reaction.25

______
23
Further information about the theory of free energy calculations may be found in many Physical Chemistry text books, such as
W.J. Moore, Physical Chemistry, 4th Edition, Publ. Prentice-Hall Inc., Englewood Cliffs, NJ (1972).

24 The simplest analogy for thermodynamics is financial accounting. If one calculates a balance sheet, one can estimate if it is
possible to make a purchase; but the balance sheet will not determine if or how quickly the purchase will be made.

25
Free energy values from CRC Handbook of Chemistry & Physics, 76th Edition, 1995-1996, D.R. Lide, H.P.R. Frederikse (Eds.),
Publ. CRC Press, Boca Raton, FL.
ANSI/ISATR92.06.031999 20

Two reactions were considered:

The first reaction is between chlorine and water to form hydrogen chloride and hypochlorous acid in the
gas phase.

Cl2 (g) + H2O(g) ==> HCl(g) + HOCl(g)

Gf 0 + -237.1 ==> -95.3 + -66.1

(kJ/mol)

The free energy change G = +75.7 kJ/mol.

The free energy change for this reaction was calculated from the difference of the free energies for the
products and the sum of the reactants. Since this value is positive, the reaction is not thermodynamically
favorable and so is unlikely to occur in the gas phase.

A quick check was made on this number using various equilibrium data from the chemical literature. A
thermodynamic cycle can be drawn up for the above reaction, after converting all concentrations to mol/L
and assuming ideal gas behavior for all vapor phase species.26

Proposed reaction, equilibrium constant K

Cl2 (g) + H2O <===> HCl(g) + HOCl(g)

Solubility of chlorine in water

K1 = [Cl2 (aq)]/[Cl2 (g)] = 1.52 (no units)

Ionization of chlorine in water

K2 = [HOCl(aq)][H+ (aq)][Cl-(aq)]/[Cl2 (aq)] = 4.36 x 10-4 mol2L-2

Solubility of HCl in water

K3 = [H+ (aq)][Cl-(aq)]/[HCl(g)] = 4040 mol L-1

Solubility of HOCl in water

K4 = [HOCl(aq)]/[HOCl(g)] = 0.7 (no units)

By rearranging these relationships, the value for K may be calculated

K = K1K2/(K3K4) = [HCl(g)][HOCl(g)]/[Cl2 (g)] = 3.2 x 10-11 mol L-1

______
26 Values calculated from the data in the Merck Index, 11th Edition, Publ. Merck & Co. Inc, Rahway, NJ (1989), and J.J. Renard, H.I.
Bolker, Chemical Reviews (1976), 77, 487.
 21 ANSI/ISATR92.06.031999

This equilibrium constant is very small and indicates that this gas-phase reaction between chlorine and
water is unlikely, in agreement with the free energy calculation. Presumably, this low value for the
equilibrium constant in the gas phase compared to aqueous solution is a consequence of the lack of
energy of hydration, especially of the HCl, which is a strong acid in water. It should be noted however that
these two calculations are not independent, since both represent thermodynamic calculations, and Gibbs
free energies are calculated from equilibrium values. The above equilibrium value corresponds to a free
energy of approximately +60 kJ/mol.27 This agreement is reasonable in view of the errors in the numbers
used. Some of the numbers were for 25C and some for zero. Some of the other assumptions necessary
to make these calculations may also have added to the errors.

The second reaction considered is between chlorine and water in the gas phase to produce hydrogen
chloride and dichlorine monoxide.

2Cl2 (g) + H2O(g) ==> 2HCl(g) + Cl2O(g)

Gf 0 + -237.1 ==> 2(-95.37) + +97.9

(kJ/mol)

The free energy for the second reaction can be calculated similarly: G = +144.3 kJ/mol.

Since this value is also positive, the reaction is not thermodynamically favorable and so is unlikely to occur
in the gas phase.

From these simple thermodynamic calculations, it would appear unlikely that these reactions will occur in
the gas phase. It should be noted that these calculations are only rough estimates and that the
assumptions made behind them, such as not forming gas-phase clusters of water around reaction
products, may not be valid. There is little information available about gas-phase hydration of active
chlorine species.

4.7 Titration of humidified chlorine28

A test was performed to provide some more experimental data about the suggested reaction between
chlorine and water in the gas phase. In the experiment, dry chlorine gas was generated and analyzed and
the test repeated with humidified chlorine test gas. If there was a reaction between the chlorine and water
that resulted in the formation of either aerosols, or water clusters, etc., that would adhere to the tubing and
other apparatus in the gas path more than simple chlorine, a difference would be expected.

Humidified chlorine was generated by passing zero cylinder air (300 mL/min) over several Thermedics
chlorine permeation vials at constant temperature. This gas stream was mixed with humidified zero air (by
passing through 500 mL water-filled impinger), to give approximately 80 to 90 percent and zero relative
humidity, while keeping the nominal chlorine concentration constant.

The humidified chlorine was analyzed for chlorine by two methods: The first involved collecting the gas
stream in an impinger containing 2 percent KI, buffered to pH and analyzing spectrophotometrically the
absorbance of iodine at two pH values (to differentiate chlorine from chlorine dioxide in other applications).

______
27 Calculated using relationship G = -RT.ln(K), where K is the equilibrium constant, R is the gas constant (8.314 JK-1 mol -1), T is
temperature (K) and K is the equilibrium constant.

28 Test method and results from Margaret N. Stryker, NCASI.

ANSI/ISATR92.06.031999 22

In the second method, the chlorine-containing air was captured as chloramine in an impinger containing
0.1 percent sulfamic acid, which upon later acidification forms iodine with KI, which is quantified.

No statistically significant difference was seen in the results between potassium iodide and sulfamic acid
capture methods, nor in the results between the humidified and non-humidified gas streams. Therefore, if
HOCl/HCl is being formed in the humidified chlorine-containing gas, it appears to be acting as chlorine.
The formation of hydrochloric acid and hypochlorous acid is a reversible equilibrium in aqueous solution,
and is expected to be similar in the gas phase. If the HCl and HOCl produced were remaining as gas-
phase species and not forming aerosols and large water clusters, etc., then both species would be present
in the final collecting solution if they were not adsorbed along the way. If this scenario occurred, then the
result would be the same as if chlorine were present. This data supports the argument that there is no
significant effect of water vapor on the chlorine (as Cl2) concentration in the gas phase.

5 Discussion of results
The chemical literature was searched by three primary means, an online search of the chemical abstracts
database on STN,29 by the information service NERAC,30 and by consulting several textbooks on chlorine.
In summary, there was a great deal of information about the reactions of chlorine with water in the liquid
state, but there was very little information available about the reactions of chlorine with water in the
gaseous state. As far as is known, no experimental results identifying gas-phase chemical species in
humidified chlorine gas mixtures have been performed that are applicable to this report. From the
experimental data, no evidence was found for water vapor interacting with chlorine in the gas phase. The
titration of dry and humidified gas streams produced similar results indicating that either the water vapor is
not reacting with the chlorine, or the species produced have the same adsorption behavior as dry chlorine
test gas.

Simple thermodynamic calculations were performed, which indicated that the reactions were very unlikely
to occur. Therefore the chlorine concentration is probably not significantly affected by gas-phase reactions
with water vapor below 100 percent relative humidity.

Experiment results from instrument testing show a variety of responses. Draeger sensors tend to show a
decrease in response to chlorine as the humidity is increased. City Technology sensors tested by MSA
showed a similar behavior, even though the sensors are of different design; one sensor has an acid
electrolyte and the other a neutral electrolyte. For both sensor designs, the gas has to pass through
several diffusion barriers, such as the porous Teflon membranes, used in the manufacture of
electrochemical sensor electrodes. Adsorption of chlorine on the humidified surface may result in a lower
chlorine concentration until the surface has saturated. This behavior is expected to give a slower response
time under conditions of high humidity than low humidity and may be the cause of the dependence on
humidity observed in the experiments described above.

It is interesting that the tests performed on Bacharach Instruments showed an increase in response with
higher humidity, even though the Bacharach instrument uses a City Technology chlorine sensor. However,
this effect is probably the result of cumulative exposure to chlorine, rather than the effect of humidity.

______
29 STN International is operated in North America by CAS, a division of the American Chemical Society. CAS, 2540
Olentangy River Rd., Columbus, OH 43210. Tel (800) 848-6533.
30 NERAC, One Technology Drive, Tolland, CT 06084. Tel. (860) 872-7000.
 23 ANSI/ISATR92.06.031999

The instrument tests represent the current technology and may provide a basis on which to select
specifications for the performance test. These tests do not, however, resolve the question about the
possible reaction between chlorine and water in the gas phase.

There are, however, likely to be other effects of humidity on the performance of most electrochemical
sensors, even without gas-phase reactions. At high humidity, there will most likely be a significant amount
of water adsorbed onto the various surfaces of the sensor in the gas-diffusion path.

5.1 Summary results from instrument/sensor tests

The specifications of ISA-S92.06.01-1998 are in the form of a percent reading or a ppm chlorine deviation
from the response to chlorine test gas at 50 percent relative humidity. The results discussed above are
presented in this format in the table below.

Summary of Instrument Response Deviations from the Response to Chlorine

at 50 Percent Relative Humidity

Test Performed, Test % Change from PPM Change % Change from PPM Change
(Nominal Cl2 Test No. Value at 50% from Value at Value at 50% from Value at
gas Concentration) RH (% RH) 50% RH RH (% RH) 50% RH (%RH)
(% RH)

Bacharach 4.2 -64, -0.9, 0, 0,

(1.5 ppm) (15 %RH) (15% RH) (90 %RH) (90 % RH)

MSA 4.3 0, 0, -22, -0.2,

(1 ppm) (1.9 %RH) (1.9% RH) (88.1 %RH) (88.1 %RH)

MSA 4.3 0, 0, -20, -0.45,

(2 ppm) (2.1 %RH) (2.1 %RH) (77.2 %RH) (77.2 %RH)

MSA 4.3 7, 0.2, -14, -0.4,

(3 ppm) (2.2 %RH) (2.2 %RH) (69.2 %RH) (69.2 %RH)

MSA 4.3 10, 0.6, -5, -0.2,

(5 ppm) (1.9%RH) (1.9 %RH) (58 %RH) (58 %RH)

NCASI,31 4.4 -3, -0.18, 3, 0.22,

Bacharach32 (5.4 %RH) (5.4 %RH) (86.5 %RH) (86.5 %RH)
(5 ppm) _______ _______
31, 2.05,
(6.6 %RH) (6.6 %RH)

______
31 Relative humidity readings at 50% RH are not available. Deviations from 50% RH assumed linear change in chlorine response
with humidity, based on instrument response for the first two humidity values. Thus the response for 50% RH was estimated to be
6.55 ppm and 2.40 ppm for the Bacharach Instrument, and 5.37 ppm and 1.48 ppm for the National Draeger Instrument for testing at
nominal 5.0 and 1.45 ppm chlorine test-gas mixtures respectively.

32 Second exposure to humidified gas, i.e. the test-gas relative humidity profile was 5.4 % RH, 86.5 %RH, and 6.6 %RH.
ANSI/ISATR92.06.031999 24

NCASI, Draeger 4.4 21, 1.1, -26, -1.4,

(5 ppm) (5.4 %RH) (5.4 %RH) (86.5 %RH) (86.5 %RH)
_______ _______
19, 1.0,
(6.6 %RH) (5.4 %RH)

Test Performed, Test % Change from PPM Change % Change from PPM Change
(Nominal Cl2 Test No. Value at 50% from Value at Value at 50% from Value at
gas Concentration) RH (% RH) 50% RH RH (% RH) 50% RH (%RH)
(% RH)

NCASI, Bacharach 4.4 -18, -0.43, 21, 0.5,

(1.45 ppm) (6.1 % RH) (6.1 %RH) (88.0 %RH) (88.0 %RH)

NCASI, Draeger 4.4 23, 0.35, -27, -0.40,

(1.45 ppm) (6.1 %RH) (6.1 %RH) (88.0 %RH) (88.0 %RH)

Draeger 4.5 16, 0.7, 2, 0.1,

(5 ppm) (~0 %RH) (~0 %RH) (~75 %RH) (~75 %RH)

5.2 Specification in Standard ISA-S92.06.01-1998, Section 7.14.1

The current specification is as follows: For instruments having meters or other output signals, the meter
and output indications at each humidity extreme shall not vary from the 50 percent RH exposure indication
by more than 0.1 ppm or 30 percent, whichever is greater after correcting for displacement by water vapor.

Comparison of the standard with the experimental results discussed above indicates that the allowed
deviation of 30 percent would encompass most of the test results obtained above.

6 Sources of chlorine test gas

In this section, several common commercial methods for producing chlorine test gas are briefly described
along with the published errors. All the methods are widely used for calibrating chlorine detection
instruments; however, not all of them have the accuracy and precision required for the ISA-S92.06.01-
1998 performance standard testing.

6.1 Gas tanks

Chlorine gas mixtures are available in compressed gas tanks. However, the content of chlorine gas
cylinders may change with time; and suppliers may have difficulty guaranteeing concentration accuracy at
low concentrations. For example, two major gas-cylinder suppliers were asked for the lowest chlorine in
nitrogen gas concentration that can be supplied. Mattheson can supply 5ppm chlorine in nitrogen with an
analytical uncertainty of 20 percent33 and Scott Specialty Gases can supply 10ppm (uncertified).34

Many chlorine gas detectors that are used for industrial safety are periodically calibrated using chlorine
from a calibration gas cylinder. These calibration gas cylinders typically have the concentration accuracy
similar to above. Therefore chlorine from gas cylinders is unlikely to have sufficient accuracy for
performance testing instruments for ISA-S92.06.01-1998.

______
33 Telephone call to Mattheson, Pittsburgh Office, (412) 261-2782, December 6, 1995.

34
Telephone call to Scott Specialty Gases, Philadelphia Office, (215) 766-8861, December 6, 1995.
 25 ANSI/ISATR92.06.031999

6.2 Electrochemical generators

Electrochemical chlorine gas generators are also available and have the required gas accuracy.

Advanced Calibration Designs35 have a chlorine gas-generating instrument that operates on the 0.2 to 2, 2
to 20 and 20 to 200ppm ranges. The accuracy of the chlorine gas generated is approximately +/- 10
percent.

6.3 Permeation Tubes

Permeation tubes offer a convenient method for producing chlorine gas at low concentration with good
accuracy. For testing purposes, chlorine test gas generation by permeation tube for performance testing
instruments for ISA-S92.06.01-1998 is a reliable method. A permeation tube, such as one available from
Kintek36 or VICI-Metronics,37 should produce chlorine test gas with an accuracy of better than two percent.

Permeation tube methods can be used to generate chlorine with sufficient accuracy and precision
necessary for ISA-S92.06.01-1998 performance standard testing.

7 Analysis of chlorine gas concentration

7.1 Literature methods for chlorine test gas analysis

Chlorine gas concentrations at the low fractional ppm concentration can be determined using an impinger
capture method developed by Margaret Stryker and Robert Fisher of the NCASI.38 Alternative methods
also exist, including a sample followed by ion chromatography,39 and a spectrophotometric procedure
following the bleaching of methyl orange by chlorine.40

8 Recommendations for testing with chlorine

8.1 Experimental considerations

Working with low concentrations of chlorine can be difficult due to the reactivity of the chlorine. It is
important to ensure that all the components of the gas delivery system are chemically compatible with
chlorine. With the exception of perfluorinated elastomers such as Viton (Dupont), most rubbers should be
avoided since they will often efficiently remove ppm levels of chlorine from a test gas.

______
35 Advanced Calibration Designs, 1664 S. Research Loop, Ste. 206, Tucson, AZ 85710, (602) 290-2855.

36 Kin-Tek Inc, 504 Laurel St., La Marque, TX 77568, (409) 938-3627.

37 VICI Metronics, 2991 Corvin Drive, Santa Clara, CA 95051, (408) 737-0550.

38
M. Stryker, R.P. Fisher, NCASI Technical Bulletin, Number 521, April 1987. NCASI is the National Council of the Paper Industry
for Air and Stream Improvement Inc., P.O. Box 13318, Research Triangle Park, NC 27709, (919) 558-1999.

39 NIOSH Manual of Analytical Methods, 4th Edition, Ed. P.M. Eller, M.E. Cassinelli; Publ. US. Dept. of Health and Human Services,
CDC, NIOSH (1994).

40
NIOSH Manual of Analytic Methods, Second Edition, Part I, NIOSH Monitoring Methods, D.G. Taylor (Manual Coordinator),
Volume 1. Publ. US. Dept. of Health, Education and Welfare, CDC, NIOSH (1977).
ANSI/ISATR92.06.031999 26

Chlorine will adsorb to some extent on any surface. Porous materials should be avoided since they usually
have a high surface area. It is also advisable to keep the length of tubing used to a minimum.

Even on a surface chemically inert to chlorine, there may be loss of chlorine by reaction with water
absorbed on the surface. It is therefore best to keep all components of the system both clean and dry (with
exception of humidity test). Before running a test, it may be necessary to have the chlorine purge the
system until it is saturated.

Chemically resistant, hydrophobic materials such as Teflon often work well. If the chlorine test gas has to
pass through a flow controller, there may be a significant loss of chlorine within the controller, especially if
it has not had a lot of exposure to chlorine. Devices such as flow controllers should be kept to a minimum
in the chlorine test gas stream, and verification of the chlorine concentration at the point of delivery is
strongly recommended.

8.2 Correction for water vapor

If the humidity is high, the concentration of other components may be diluted accordingly. This dilution
may need to be taken into account when conducting humidity tests, depending on the design of the
experiment.

The vapor pressure of water at 25C (298K) is 23.756 mm Hg,41 which corresponds to 100 percent
relative humidity. One atmosphere is ~760 mm mercury, so the concentrations of gases in a humid
atmosphere will be reduced by about 3.0 percent (= 23.756/760 ). For example the composition of dry and
humid air as percent by volume is shown in the following table.

Composition of Dry and Humid Air (Percent Volume)

Nitrogen Oxygen Water Scaling Factor

Dry 78.08 20.95 0 1

15% RH 77.71 20.85 0.47 0.995

50% RH 76.86 20.62 1.56 0.984

90% RH 75.64 20.30 2.81 0.972

The applicability of this correction depends on how a gas mixture is prepared. If the component gas is
mixed with balance gas and the resulting mixture is humidified, the above correction should be applied.42
If, however, the component gas is blended with humidified balance gas, then no correction is necessary
since the humidity will dilute the concentration of the balance gas but not the component gas.

______
41 Data and definitions from Handbook of Chemistry and Physics, 70th Edition, 1989-1990, CRC Press.

42 For a typical test gas composition, e.g. 10ppm chlorine in nitrogen, the chlorine is the component gas and the nitrogen is the
balance gas.
 27 ANSI/ISATR92.06.031999

9 Other gas performance standards

It is one of those unfortunate realities that the more reactive gases tend to be more toxic and so need to be
detected at lower concentrations. However, the same reactivity makes generation of stable gases, gas
delivery and thus reliable detection more difficult. This report has discussed chlorine and chlorine
detection, but similar considerations need to be made for other gases as well. Gases such as ammonia,
hydrogen chloride, sulfur dioxide etc. may be especially prone to problems with humidity. The effect of
humidity on each gas will have to be assessed separately.

10 Acknowledgments
The following people are thanked for their help in providing material and for helpful discussions toward the
writing of this report.

Max Carcas City Technology Ltd.

Jerry Connell Capital Controls Inc.
Chuck Gardner Bacharach Inc.
Leigh Greenham Crowcon Detection Instruments Ltd.
Gunter Huber National Draeger Inc.
Joseph D. Jolson MSA Inc.
R.G. Landolt Texas Weslyan University
Davey Li CSA
Dale Merriman MerTech Inc.
Margaret Stryker NCASI
Bill Thorn NIST
David Wagner Industrial Scientific Corporation
This page intentionally left blank.
Developing and promulgating sound consensus standards, recommended practices, and technical reports
is one of ISAs primary goals. To achieve this goal the Standards and Practices Department relies on the
technical expertise and efforts of volunteer committee members, chairmen and reviewers.

ISA is an American National Standards Institute (ANSI) accredited organization. ISA administers United
States Technical Advisory Groups (USTAGs) and provides secretariat support for International
Electrotechnical Commission (IEC) and International Organization for Standardization (ISO) committees
that develop process measurement and control standards. To obtain additional information on the
Societys standards program, please write:

ISA
Attn: Standards Department
67 Alexander Drive
P.O. Box 12277
Research Triangle Park, NC 27709

ISBN: 978-1-55617-701-9