Hazardous Gas Monitors

Chapter 1 Introduction
Chapter 1
Introduction
T
he detection of hazardous gases has always been
a complex subject and makes choosing an appro-
priate gas monitoring instrument a difficult task.
To address this problem, this book aims to provide
the following essential tools:
• A simple guide to the various sensor technolo-

gies available
• Information to help you intelligently select the
proper instruments for specific applications
• Information engineers can use to design a com-
plete monitoring system
• Technical data and practical procedures that
technicians can use to check and maintain a gas
monitoring system
The main emphasis of the book is on gas detec-

tion technology that is used in the field of area air
quality and safety. This field primarily involves the
protection of personnel and property against toxic
and combustible gases. The discussion includes the
types of sensors used, the various instruments avail-
able, and the applications that incorporate these in-
struments.
1
Hazardous Gas Monitors
Analytical Instruments and Monitoring Systems

To date, no gas sensors exist that are 100% selec-
tive to a single gas. Achieving such selectivity requires
the use of instruments that employ analytical tech-
niques to identify gases.
Examples of such instruments include Fourier trans-
form infrared (FTIR) instruments that use the infrared
spectral characteristics of gases, gas chromatographs that
use analytical columns, and mass spectrometers that iden-
tify molecules through characteristic variable deflec-
tions from a magnetic field.
These instruments provide fairly accurate and se-
lective gas readings. Some typical applications for
these kinds of instruments include airport bomb
detection, drug abuse screening, and analyzing air
pollutants. However, these analytical instruments re-
quire skilled and knowledgeable operators, and are
generally very expensive and designed for laboratory
tabletops or specific on-line applications for in-plant
An Analytical Instrument. The
example shown above automates
installations.
gas chromatography with the help In addition, many suffer from limitations such as
of its built-in robotic technology. high maintenance, slow response time, and large size,
(Courtesy of CE Instruments) making them impractical monitors for area air qual-
ity and safety. Thus, they are typically used only as a
last resort for applications in which a suitable sensor
is not available.
For work area air quality and safety applications,
monitoring systems must meet a number of practical
criteria. These monitoring systems must be:
• rugged and corrosion-resistant
• weather- and dust-proof
• capable of being installed in hazardous areas
• durable and long-term
A Gas Monitor. IST’s MP-204
is a wall-mounted unit housed • operationally stable
in weatherproof enclosure with • easy to maintain
four sensor channels.
• operated by a minimally skilled person
2
• suitable for multisensor systems that, for ex-

ample, can be used for an entire chemical plant
• low cost
This book deals with gas monitors for work area
air quality and safety applications. For practical pur-
poses, we will not delve into the realm of the much
more complex analytical instruments which, for the
most part, do not meet these criteria.
Gas Sensors
A gas sensor is a transducer1 that detects gas mol- 1
Any device that converts input en-
ecules and which produces an electrical signal with a ergy of one form to output energy
magnitude proportional to the concentration of the gas. of another.
Unlike other types of measurement, types that are Gas Molecules
relatively straightforward and deal with voltage, tem-

perature, and humidity, the measurement of gases is GAS TRANSDUCER
Electrical
Signal
much more complicated. Because there are literally
hundreds of different gases, and there is a wide ar-
ray of diverse applications in which these gases are 2
present, each application must implement a unique
set of requirements. For example, some applications
may require the detection of one specific gas, while 1
eliminating readings from other background gases.
Conversely, other applications may require a quan-
titative value of the concentration of every gas present
in the area.
Types of Gas Sensors. There are many different
technologies currently available for the detection of 4
gases, each with certain advantages and disadvantages. 3
The following sensing methods are the focus of our
discussion and are the five types most suitable and
widely used as gas monitors for area air quality and
safety applications:
• electrochemical • infrared 5
• catalytic bead • photoionization Some Examples of Gas Sensors

1. Catalytic Bead. 2. Infrared.
• solid state 3. Solid State. 4. Electrochemical.
5. Photoionization.
3
All of these sensors are commonly used for detection of

toxic and combustible gases in the work area for hu-
man and property protection, or for process control.
One common characteristic of these sensors,
despite what is often claimed or implied, is that they
are not specialized to detect any one specific gas.
Each sensor is sensitive to a group or a family of
gases. In other words, the sensor is non-specific and
is subject to interference by other gases much like
a smoke detector in a house cannot distinguish be-
tween the smoke caused by a furniture fire and the
smoke caused by food burning in the stove or oven.
In limited cases, a chemical filter can be installed
to filter out interference chemicals while permitting
the target gas to pass through to the sensor. Alterna-
tively, an analytical column can be installed to iden-
tify chemicals qualitatively and quantitatively.
For gas monitoring applications, a proper sensor
is usually selected to match the specific application
requirements and circumstances, with the user inter-
preting the readings based on an awareness of the
sensor’s limitations.
Terms, Definitions, and Abbreviations

Units of Measure for Gas Concentration
ppm: parts per million by volume (see Table 1, oppo-
site page)
ppb: parts per billion by volume
Fuel mg/m3: milligrams per cubic meter
mg/cc: milligrams per cubic centimeter
g/m3: grams per cubic meter
g/cc: grams per cubic centimeter
For Combustible Gases
Oxygen Ignition Source
Flash Point (Fl.P): The temperature at which a com-
The Combustion Triangle bustible liquid gives off enough vapor to form
4
Table 1. Equations for Deriving Units of Gas Concentration

Gas concentration is commonly expressed as percent (%), ppm, or ppb. Mathematically, these are
unitless expressions since they do not carry a unit for volume or weight but simply express the ratio of
gases in relation to background air.
For instance, one ppm of CH4 simply means one part of methane amongst 999,999 parts of back-
ground air. It is expressed as
Vg Vg
=
Va + Vg VT
Vg = volume of gas; Va = volume of air; VT = total volume of air and gas
Multiply the fraction derived from the formula above:
a) by 102 % to obtain the percentage
b) by 106 ppm to obtain the ppm
c) by 109 ppb to obtain the ppb
For example, if you mix 1 cc of gas with 99 cc of air, the calculation is as follows:
1 cc
= 0.01
1 cc + 99 cc
Thus, 0.01 x 102 % = 1%
0.01 x 106 ppm = 10,000 ppm
0.01 x 109 ppb = 10,000,000 ppb = 107 ppb
In cases like this, one would normally not cumbersomely express the units as ten million parts per
billion. Instead, the simpler expression, 1%, is preferred.
This volumetric expression of concentration is straightforward. Additionally, the volume ratio is equal
to the pressure ratio according to Dalton’s law of partial pressures. It is expressed as
Pg Pg Vg
= =
Pa + Pg PT VT
Pg is the partial pressure of the gas within the total pressure PT and Pa is the partial pressure of air.
As an example, 1 psi of gas within 99 psi of air with total pressure of 100 psi has a concentration of 1%
as in the volume expression.
Another unit, commonly used in the medical and metallurgical industries, is mg/m3 or milligrams per
cubic meter, in situations wherein the chemical is either in liquid or solid state at room temperature. To
convert mg/m3 to percent or ppm, the ideal gas law must be used. Chemical conversion factors are
included in the Gas Data section in Appendix II, page 199, and the conversion formula is discussed on
page 173 in Chapter 11, Gas Sensor Calibration.
5
an ignitable and combustible mixture when air

is present near the liquid’s surface.
That is, it is the temperature at which a com-
2
The law of partial pressures, first bustible liquid chemical has sufficient partial
formulated by James Dalton in pressure2 in the air to be ignited. The partial
1802, states that the pressure of pressure curves of chemicals are available from
a mixture of gases, P, which do not
react chemically, is the sum of the
chemical libraries or manufacturers of the
independent pressures (partial chemicals.
pressures) that each gas exerts: Lower Explosive Limit (LEL) or Lower Flammable
P = P1 + P2 + . . . + Pn Limit (LFL): The minimum concentration of gas or
H2O+ = 9%
vapor mixed with air (percent by volume, at room
= 1.323 psi
temperature) that will cause the propagation of
flames when it comes in contact with a source of
O2 = 21%
ignition. In common terminology, mixtures be-
= 3.087 psi
low the LEL or LFL are too lean to ignite.
N2 = 70%
= 10.29 psi
Upper Explosive Limit (UEL) or Upper Flammable
Limit (UFL): The maximum concentration of gas
or vapor mixed with air (percent by volume, at
P = PO2 + PN2 + PH2O+ room temperature) that will cause the propaga-
P = 3.087 psi + 10.29 psi + 1.323 psi tion of flames when it comes in contact with
P = 14.7 psi an ignition source. In common terminology,
mixtures above the UEL or UFL are too rich to
support combustion. The combustible range is,
therefore, between the LEL and the UEL (see
Figure 1 below).
100%
TOO RICH FOR COMBUSTION

GAS CONCENTRATION
(MIXTURE IN AIR)
UEL
COMBUSTIBLE MIXTURE
LEL
TOO LEAN FOR COMBUSTION
0%
Fig. 1. A Window of Combustibility
6
Specific Gravity (Sp.Gr.): The ratio of the weight

per unit volume or mass of a substance at 68°F
(20°C) to the mass of an equal volume of dis-
tilled water at 39.2°F (4°C).3 3
Water at 4°C has the LOWEST
Vapor Density: The weight per volume of gas or volume per gram. Ice expands
when its temperature goes below
vapor compared to dry air; both components
0°C and will crack a rigid encase-
having the same temperature and pressure. For ment of it even if it was made of
example, air has a vapor density of 1; carbon cement.
dioxide, 1.52; hydrogen, 0.07; methane, 0.55;
and propane, 1.52.
For Toxic Gases

The National Institute for Occupational Safety and
Health (NIOSH) is a branch of the U.S. Department
of Health and Human Services, Public Health Service,
Centers for Disease Control and Prevention. Acting
under the authority of the Occupational Safety and
Health Act of 1970 and the Federal Mine Safety and
Health Act of 1977, it publishes recommended exposure
limits (RELs) for hazardous substances or conditions
in the work place.
To formulate these recommendations, NIOSH col-
lects and evaluates data from the fields of industrial
hygiene, toxicology, occupational medicine, and ana-
lytical chemistry. These recommendations are then
published and transmitted to the Occupational Safety
and Health Administration (OSHA) and the Mine
Safety and Health Administration (MSHA) for use in
promulgating legal standards.
OSHA published the permissible exposure limits
(PELs) that are known as the General Industry Air
Contaminants Standard.
The American Conference of Governmental In-
dustrial Hygienists (ACGIH) is a professional society,
not an official government agency. Membership is lim-
ited to professionals in governmental agencies or edu-
cational institutions engaged in occupational safety
and health programs in the United States and around
7
the world. ACGIH publishes the exposure standard,

threshold limit values (TLV).
These three different standards—NIOSH’s RELs,
OSHA’s PELs and ACGIH’s TLVs—are similar to each
other, yet in some instances there are variations among
them. All are based on time-weighted-average (TWA),
short-term exposure limit (STEL), ceiling (C) and immedi-
ately dangerous to life or health (IDLH) concepts. The
following definition of these terms are for informa-
tional purposes only. It is beyond the scope of this book
to include detailed discussions of each.
Time-Weighted Average (TWA) is the average con-
centration of contaminants over a specified time
period. Mathematically,
During the first hour, the TWA is: TWA is the integrated area
(50 ppm x 1 hr.) / 1 hr. = 50 ppm under the concentration
Up to the end of the second hour, the TWA is: curve over time divided by
(25 ppm x 1 hr.) time period. To illustrate
(50 ppm x 1 hr.) + (50 ppm x 1 hr.) +
2
2
= 56.25 ppm the concept of TWA (Figure
2), let us assume that dur-
Up to the end of the third hour, the TWA is:
(50 ppm x 2 hrs.) ing a three-hour period, CO
(50 ppm x 3 hrs.) +
2
= 66.7 ppm
concentration is constant at
3 50 ppm during the first
The numbers within the parentheses represent areas under the curve. hour; then the CO concen-
tration increases steadily to
100 100 ppm 100 ppm at the end of the
third hour. The TWA at any
GAS CONCENTRATION
(% MIXTURE IN AIR)
given moment is repre-

66.7 ppm
sented by the red line.
50 56.25 ppm TWA Many microprocessor-
based instruments and data
logging programs are ca-
pable of performing the
0
0 1 2 3 TWA calculation, and the
HOURS time interval used by the
instruments to calculate the
Fig. 2 Simplified Three-Hour TWA. PELs are based on an eight-
hour TWA. This is a simplified version to illustrate the TWA concept.
TWA is much shorter than
one hour.
8
Recommended Exposure Limit (REL) is TWA con-

centration permissible for up to ten-hour work-
days during a forty-hour work week.
Permissible Exposure Limit (PEL) and Thresh-
old Limit Value (TLV) are TWA permissible
concentrations, to which workers may be ex-
posed continuously, day after day, without ad-
verse effects, for a normal eight-hour workday
and a forty-hour work week.
Short-Term Exposure Limit (STEL) is defined as a
fifteen-minute TWA exposure which should not
be exceeded at any time during a work day even
if the eight-hour TWA is within
CEILING
limits. Exposures at the STEL
should not be longer than fif- GAS CONCENTRATION
(MIXTURE IN AIR) STEL
teen minutes and should not be
repeated more than four times TWA (red line)
per day. There should be at least PEL
60 minutes between successive

exposures at the STEL. The con-
0 5 10 15
cept of STEL is illustrated in Fig- MINUTES
ure 3. Fig. 3. A graphical example of permitted TWA excursions
Ceiling Limit: The concentration above STEL, provided the 15-minute TWA does not exceed
which should not be exceeded the STEL.
at any time.
Immediately Dangerous to Life or Health (IDLH)
concentration is the maximum concentration
above which only a highly reliable breathing
apparatus providing maximum protection for
workers is permitted. The IDLH value is based
on the ability of a worker to escape without loss
of life, irreversible health effects, or other
health effects such as disorientation or incoor-
dination that could prevent escape.
The preceding information is a simplified version
interpreted from the NIOSH Pocket Guide to Chemical
Hazards. Appendix II at the end of this book lists data
9
IDLH extracted from the same pocket guide

GAS CONCENTRATION
that includes REL, PEL and IDLH

(MIXTURE IN AIR)
CEILING
concentrations.
STEL Figure 4 illustrates the overall sum-
mary of the toxic gas safety concept.
PEL
TWA Combinations of Substances. When
0 1 2 3 4 5 6 7 8 9
two or more hazardous substances
HOURS are present in the area, the additive
Fig. 4. Basic Rules for Compliance: TWA for eight
effects should be considered. That
working hours does not exceed PEL/TLV. Observe the is, if the sum of the fractions of haz-
definition of STEL and never exceed the ceiling limit. ardous substance concentrations di-
vided by the respective PEL exceeds
unity, then the threshold limit of the mixture must
be considered excessive. This case is illustrated in
the following formula:
C1 C2 C C
+ + 3 + ... + n < 1
PEL1 PEL2 PEL3 PELn
If it is greater than one, then the PEL is exceeded.
C1,C2 and C3 represent the TWA of various hazardous
substances.
For example, if the ambient air contains 35 ppm
of carbon monoxide (PEL 50 ppm) and 350 ppm of
carbon dioxide (PEL 5000 ppm), the calculation is as
follows:
35 350
+ = 0.77
50 5000
The threshold limit is not exceeded.
This discussion is for conceptual purposes only. It does
not take into account the effect of the combining of
chemicals that can react with each other, resulting in
a final mixture that can be more toxic to humans than
the individual toxicities of each gas.
Performance Specifications
Accuracy: Webster’s dictionary defines accuracy as
10
“the quality or state of being accurate or exact;

precision; exactness.” Accurate is defined as “free
from mistake or error; precise; adhering closely
to a standard.” Precise is defined as “strictly de-
fined; minutely exact; low tolerance; etc.” Accord-
ingly, a measurement can be precise but not nec-
essarily accurate. The accuracy can only be deter-
mined when compared to a standard.
Accuracy is the most important definition of Measured Standard
Weight Weight
the quality of performance for most of the ob-
ACCURACY
jects we deal with every day; for example, your
watch, the weather thermometer, the bathroom
scale, and the measuring tape, to name a few. The
definition of these measurement standards are
well-defined.
At the National Institute of Science and Tech- The International Vocabulary of Basic and
nology (NIST; formerly National Bureau of Stan- General Terms in Metrology defines the
dards), standards for weight, length, temperature, terms for result of the measurement as
etc. are kept. Internationally, there is a total agree- follows:
ment about the “absoluteness” of those standards. 1. Accuracy: Closeness of the agreement
In real life, the most accurate instrument may between the result of a measurement
not necessarily be the best. For example, the mea- and a true value of the measurand.*
suring tape which is used by the tailor is not very 2. Repeatability: Closeness of the agree-
ment between the results of successive
accurate, but it is practical for the task. Calipers
measurements of the same measurand
used in a machine shop are more accurate than carried out under the same conditions
the tailor’s tape but would not be suitable for use of measurement.
by the tailor. Thus, each instrument serves a dif- 3. Reproducibility: Closeness of the agree-
ferent objective. The gas monitors serve more like ment between the results of measure-
a tailor’s tape than a machine shop’s caliper. ments of the same measurand carried
The Challenge of Accuracy. With gas moni- out under changed conditions of mea-
surement.
toring systems, there is no standard by which to
compare accuracy. There are hundreds of differ- 4. Linear scale: Scale in which each spac-
ing is related to the corresponding scale
ent chemicals, each having its own unique chemi-
interval by a coefficient of proportional-
cal and physical properties. ity that is constant throughout the scale.
As an example, what is 100 ppm of carbon mon-
* measurand–a particular quantity subject
oxide (CO) in air? Mathematically, this equates to
to measurement; e.g., vapor pressure
0.01% of CO and 99.99% air. After the mixture is of a given sample of water at 20°C.
made, its accuracy is difficult to determine since
11
there is no standard 100 ppm of CO to compare

with and there is no common agreement that de-
fines a 100 ppm CO mixture.
Because calibration standards are difficult to
define in practice, accuracy is the most misunder-
stood and abused term in gas monitoring. There
are few agreements amongst manufacturers of in-
struments, and there is no common understand-
ing in general.
Realistically, it is best to establish a calibration
method that can yield consistent and precise calibra-
tion data. The accuracy of this calibration method
can be compared to an accepted standard when
challenged.
As long as the calibrations are done with high
precision, the accuracy of your calibration can be
established when an accepted standard is available.
Repeatability: Repeatability is the ability of sensors to
repeat the measurements of gas concentrations when
the sensors are subjected to precisely calibrated gas
O2 samples.
Zero Air: “Zero air” is available on the market in the
N2 form of a mixture of oxygen and nitrogen in a
high pressure tank. In normal applications, how-
ever, gas monitoring sensors are put to work in
A. Artificial Air “non-ideal” environments and, consequently,
or there are many other components in the ambient
air besides O2 and N2 , such as water vapor, car-
H2O bon dioxide, carbon monoxide, and other trace
O2 gases.
Therefore, it is not practical to zero a sensor
CO2 &
N2 others to a simple mixture of oxygen and nitrogen. Some
sensors can be zeroed with dry air or N2 but some
cannot. For instance, most infrared (IR) detectors
B. Environmental Air can be zeroed with dry air or N2 as long as the
wavelength being detected has a minimum water
Zero Air vs. Ambient Air vapor effect. However, with solid-state sensors or
photoionization detectors (PIDs), very different
12
readings may result from dry air compared with

wet air.
In many applications, sensors may be satisfac-
torily zeroed by exposing the sensor to a bag of
air collected from a location where the air is “nor-
mal.” In fact, this is the easiest way to verify
whether a sensor is giving a real alarm Sensor Output
or a false alarm. in volts
Linearity: Quantitatively, linearity refers
to the output signal in relation to the 10
gas concentration: If 1 volt equals 10
ppm and a full scale 5 volts represent
50 ppm, then the output will be lin- Near Saturation
Linear Nonlinear (Poor Resolution)
ear. With most sensors, the initial 5
output of the sensor is linear or close
to it, but as the gas concentration in-
creases, the output signal is gradually
reduced. Figure 5 shows a typical sen- 0
Gas Concentration
sor response curve.
Fig. 5. Typical Sensor Output Curve.
Specificity or Selectivity: This is the ability of an in- As the gas concentration increases,
strument to detect a target gas without being af- the output signal becomes smaller in
fected by the presence of other interfering gases. relation to the increase in gas
Most sensors are sensitive to a family of gases, concentration, resulting in poor
resolution. Most sensors provide
and there are no sensors specific to only one gas. better accuracy at lower concen-
Among the more specific sensors is an electro- trations than at very high concentra-
chemical sensor for the detection of oxygen. tions. Thus, instruments on the
Different techniques are employed in order market today commonly have output
signals that are digitally linearized.
to achieve some degree of selectivity to suit prac-
tical applications. For example, a charcoal filter
is used to filter out most hydrocarbons while let-
ting only CO, H2, and CH4 pass through.
In solid-state sensors, the surface temperature
of the sensor can be set differently in order to
make it more sensitive to one gas and less sensi-
tive to other gases.
The most common practice is to use an ana-
lytical column, in which the gas sample stream is
introduced into the column and the chemical
13
components are separated and come out at the

end of the column at different times, where they
are detected by a sensor.
This method works well for laboratory use but
has proven to be impractical for gas monitoring
because it requires a high degree of user knowl-
edge and a high degree of maintenance. Addi-
tionally, since the sample must be drawn through
the column, the time necessary to take readings
can be relatively excessive. Sample times of 15 to
30 minutes are not unusual.
For ambient air monitoring, it is much more
practical to use a sensor directly installed at the
location being monitored and compensated for
the different gases which may be present.
Interference Ratio: As mentioned earlier, sensors are
not selective to a single gas and will read other
gases as well. Thus, a common practice for manu-
facturers of gas monitoring equipment is to pro-
9 ppm 30 ppm vide data indicating the ratios that different gases
H2 CO
will read on the sensor. For instance, on a 100
ppm carbon monoxide sensor, hydrogen may
33 ppm read at a 3-to-1 ratio. This means that 3 ppm of
CO Sensor CO Signal H2 will read the same as 1 ppm of CO.
In many cases, even though it is stated that a
9 ppm H2 shows as 3 ppm CO certain gas will not interfere, if the concentra-
tion of this gas is high enough, it may in fact ac-
tually interfere. For example, while a CO sensor
with a charcoal filter has little interference from
certain solvents at 100 ppm, when the concen-
tration is increased to 1,000 ppm, they may in-
terfere drastically.
Because there are so many gases, it is not pos-
sible for manufacturers to present data on cross
sensitivity ratios for all gases. Therefore, if inter-
ference data is not provided for a gas that is of
interest to you, you should inquire with the
manufacturer if the sensor is selective to a spe-
14
100%
Scale
50%
0% Time
T80 T100
Fig.6. A Typical Sensor Response
cific gas being targeted; in which case, the manu-

facturer could provide the data needed.
Response/Recovery Time: This is typically defined
as the time it takes for a sensor to read a certain
percentage of full-scale reading after being ex-
posed to a full-scale concentration of a given gas.
For example, T80 = 30 seconds means that the
sensor takes 30 seconds to reach 80% of the full-
scale reading after being exposed to a full-scale
gas concentration.
Temperature and Humidity: Specifications for these
parameters are easy to understand, but beware of
the humidity specification. Relative humidity is
an indication of the amount of water vapor in air
as a percentage of the total amount possible at a
given temperature.
Quantitatively, the amount of actual water vapor
in air is a function of temperature. For instance, at
80% relative humidity and a temperature of 25°C,
water vapor is present at a level of 3%. However, the
same 80% relative humidity at a temperature of 48°C
produces a water vapor level of 10%. In the pres-
ence of chemicals, combined with the changing of
temperature between day and night, the possible
water condensation and resultant corrosive mixtures
can compromise the life expectancy of a sensor.
15
Hysteresis: The difference in response of the sensor when

calibrating from a zero level to mid-scale compared
to the response when calibrating from full scale to
mid-scale, is known as hysteresis. This quantity is nor-
mally expressed as a percentage of full scale. For ex-
ample, a 100 ppm instrument, when calibrated from
0 to 50 ppm and exposed to a 50 ppm calibration gas,
will indicate 50 ppm. However, when the sensor is cali-
brated to 100 ppm gas but is exposed to 50 ppm the
sensor may indicate 55 ppm. This variation of 5 ppm
is 5% full-scale hysteresis. Most infrared and photo-
ionization instruments do not exhibit hysteresis, but
many other sensors, including electrochemical, solid
state, and catalytic sensors do exhibit hyteresis.
In alarm setting, the difference between the on
point and off point alarm is also referred to as hyster-
esis. For instance, if the alarm comes on at 100 ppm,
the alarm will not turn off until the gas is below 90
ppm. This hysteresis is needed; otherwise, an alarm
can be chattering at the set point of 100 ppm.
Zero and Span Drift: While there is no specific definition
for these two terms, common understanding holds that
this drift is the percentage change of the zero or span
calibration over a specified period of time, typically
30 days or more.
Hazardous Locations
Gas monitoring instruments are often installed in in-
dustrial process and production areas. These areas are
often classified as hazardous locations. Industrial fa-
cilities in which potentially explosive gas atmospheres
exist or may exist must utilize proper explosion proof
protection methods when using these types of instru-
ments. It is beyond the scope of this book to provide
full details; however, the following information may
be helpful. The reader is advised to consult with the
manufacturer for specific needs.
North America and other parts of the world that
16
have been influenced by North American practices

have traditionally used the National Electric
Code(NEC®) articles 500-503. They employ a class
and division system: Classes identify the type of haz-
ard present as gases or vapors, combustible dusts,
and flammable fibers. Divisions define the condi-
tion under which the hazardous material
may be present. The devices designed and North American Certification Agencies
manufactured for these hazardous loca- Underwriters Laboratories Factory Mutual
tions should be tested and approved for
use by a nationally recognized laboratory
such as Under writer’s Laboratories
(UL®), Factory Mutual (FM), or the Ca- Canadian Standards Association
®
nadian Standards Association (CSA ).
The NEC 500 hazard classifications are
as follows:
Class I: Flammable gases or vapors.
Class II: Combustible dusts
Class III: Easily ignitable fibers and flyings
Groups are based on flame propagation characteris-
tics, ignition temperature, and pressure generated during
explosion of various gases and vapors. There are four dif-
ferent groups in Class I. These groups are as follows:
Group A: Acetylene
Group B: Acrolein, butadiene, ethylene oxide, form-
aldehyde, hydrogen, propylene oxide, and propyl
nitrate
Groups C and D: All other combustible gases belong
to Groups C and D
Division 1: Where ignitable concentrations of gases,
vapors, dusts, and fibers can exist all the time or some
of the time under normal operating conditions.
Division 2: Where ignitable concentrations of gases,
vapors, dusts, and fibers do not exist under normal
operating conditions. Hazardous conditions only ex-
ist in the event of abnormal conditions, such as acci-
17
dental rupture or breakdown of a container, stor-

age tank, etc.
Gas monitoring instruments are typically designed
and certified for use in Class I, Division 1, Group B, C
and D hazardous locations for use in North American
markets.
Zones. European countries, as well as a majority of
other nations of the world, have been influenced by
the International Electrotechnical Commission’s (IEC)
three-tiered zone approach. The IEC separates the
potentially explosive atmosphere into Zones 0, 1, and
2 based on the probability of occurrence and length of
time a potential explosive mixture may be present.
CENELEC Marking Apparatuses designed for use in these areas are usually
Example: Ex EEx d IIb T3
tested and approved for use by the European Commit-
tee for Electrotechnical Standardization (CENELEC)
Approved mark for
apparatus certified test authorities using Euronorm (EN) standards. The
by an EC test division of these three zones are:
authority
Zone 0: An area in which an explosive gas atmosphere
Symbol for apparatus
built in accordance is continuously present for long periods.
with a European
standard Zone 1: An area in which an explosive gas atmosphere
is likely to occur in normal operation.
Flameproof enclosure
(type of protection) Zone 2: An area in which an explosive gas atmosphere
Explosion group is not likely to occur in normal operation, and if
Temperature class
it does, it will exist for a short period only.
Zones 20, 21, and 22 are a subset of Codes 0, 1, and 2
that refer to ignitable dust clouds.
Definition Comparisons. In accordance with their
explosive properties, the combustible gases and vapors
are divided into temperature classes and explosion pro-
tection subgroups. There are no direct comparisons
between the current NEC and IEC standards. National
Fire Protection Agency (NFPA) in America adopted
article NEC 505 which is comparable to IEC standards.
A brief comparison of IEC (world), CENELEC (Eu-
rope) and NEC (USA) are as follows:
1. Condition: Hazardous conditions exist continuously
18
or for long periods of time.

NEC 505: Class 1, Zone 0
NEC 500: Class 1, Division 1
IEC: Zone 0.
CENELEC: Zone 0
2. Condition: Hazardous condition is likely to occur in
normal operation.
NEC 500: Class 1, Division 1
IEC: Zone 1
CENELEC: Zone 1
3. Condition: Hazardous condition is not likely to oc-
cur in normal operation and if it does, only infre-
quently and for a short period.
NEC 500: Class 1, Division 2.
IEC: Zone 2
CENELEC: Zone 2
4
The lower case letter in a CENELEC-
Types of Protection approved marking designates the
type of protection offered by an
There are several acceptable types of protection enclosure.
for electrical equipment in hazardous locations. The
more common types are the following:
A. Flameproof Enclosure (d)4 : The enclosure,
such as the one shown in Figure 6, will with-
stand an internal explosion, without causing
ignition of an external explosive atmosphere.
The enclosure joints and structure covers are
designed and manufactured for such purposes.
This type of protection is most commonly used
for gas monitoring applications and can meet
the requirements of:
NEC 500 - Class 1, Division 1 & 2.
NEC 505-Class 1, Zone 1 & 2, AExd.
IEC-Exd. Fig. 6. An Explosion-Proof
CENELEC-EExd. Enclosure
19
Types of Flamepaths in Flameproof Enclosures
Joint Threaded Spigot
Materials used: Aluminum, Iron (Courtesy of Cullen Associates)
B. Intrinsic Safety (i): The electrical energy in

an intrinsically safe circuit which enters into the
enclosure is not sufficient to generate a spark
and ignite a combustible mixture in the haz-
ardous area, in any worst case scenario. To meet
this requirement, safety barriers or other devices
limiting the electrical energy
HAZARDOUS SAFE AREA
AREA are placed on the wires to
Process limit the electrical energy al-
Single Channel Controller
Resistor Fuse
(+) 24 v lowed to flow through the cir-
Zener cuit before the wire enters
Diodes
Sensor
Module
into the hazardous location.
M R Safety barriers are a combi-
nation of zener diodes, power
Power supply must
not be grounded resistors, and fuses which are
Intrinsically designed to limit the amount
Safe Ground
of electrical energy allowed to
(Courtesy of Cullen Associates) flow through the wires. Various approved and
certified safety barriers are available as standard
electrical components. These are limited to low
power device applications only. This method of
protection can meet the requirements of:
NEC 500 - Class 1, Division 1 & 2.
NEC-505 - Class 1, Zones 0, 1 & 2.
AExi(a).
IEC - Exi(a).
CENELEC - EExi(a).
20
C. Purged and Pressurized (p): This is the pro-

Pressure
cess of supplying sealed electrical enclo- Optional Pressure Indicator
Relief Device
sures with a protective gas to prevent the
entrance of flam- Protective
mable gases while Gas Supply Optimal
maintaining a positive Alarm or
Power
enclosure pressure. Pressure Cut-Off
Regulator Protective Switch
This type of protec- Enclosure
tion can meet the requirements of:
NEC 500 - Type X, Y & Z.
NEC 505 - Type X, Y & Z.
IEC - Exp.
CENELEC - EExp.
Type X pressurizing: Reduces the classification
within the protected enclosure from Class 1, Di-
vision 1 or Class 1, Zone 1 to unclassified.
Type Y pressurizing: Reduces the classification
within the protected enclosure from Division 1
to Division 2 or Zone 1 to Zone 2.
Type Z pressurizing: Reduces the classification
within the protected enclosure from Class 1,
Division 2 or Class 1, Zone 2 to unclassified.
D. Increased Safety (e): This is a type of explosion
protection applied to electrical apparatus that
does not produce arcs or sparks in normal ser-
vice, in which additional measures are applied
so as to give increased security against the pos-
sibility of excessive temperature and of the oc-
currence of arcs and sparks. This method of
protection can meet the requirements of:
NEC 500 - No Standard.
NEC 505 - Class 1, Zone 1 & 2, AExe.
IEC - Exe.
CENELEC - EExe.
21
E. Other Protection Methods: Oil immersion(o),

powder filling(q) and moulding(m).
Enclosure Classifications For Nonhazardous Areas

In North America, the National Electrical Manu-
facturer’s Association (NEMA), as a way of standard-
izing enclosure performance, classified the enclosures
in different ratings which are intended to provide in-
formation for users to make proper product choices.
This rating system identifies the ability of the enclo-
sure to resist various possible conditions. The classifi-
cations are as follows:
NEMA Type 1: For general-purpose indoor use, pro-
vides protection against incidental contact with
the enclosed equipment.
NEMA Type 2: In addition to NEMA Type 1, pro-
vides protection against a limited amount of
falling water and dirt.
NEMA Type 3: For outdoor use. Provides protec-
tion against windblown dust, rain and sleet, as
well as formation of ice on the enclosure. Pro-
vides rust resistance.
NEMA Type 3R: Same as NEMA Type 3, but does
not provide dust protection.
NEMA Type 4: Same as NEMA Type 3, except, it is
for indoors or outdoors, provides protection
against direct water hose down.
NEMA Type 4X: Same as NEMA Type 4, but it also
provides corrosion resistance for indoor use.
NEMA Type 6: Same as NEMA Type 4X, but pro-
vides protection against water during tempo-
rary submersion at a limited depth.
NEMA Type 7: For indoor use in hazardous loca-
tions, Class 1, Groups A, B, C, and D.
NEMA Type 9: For indoor use in dust applications,
Class II, Groups E, F, and G.
22
NEMA Type 12: For indoor use, provides protec-

tion against dust, falling dirt, and dripping non-
corrosive liquids.
NEMA Type 13: For indoor use, provides protec-
tion against dust, spraying water, oil, and non-
corrosive coolant.
Internationally, the protections for electrical appa-
ratuses are designated with ingress protection (IP) followed
by a two-digit number which defines the degree of pro-
tection. The first digit (0-6) defines the protection
against contact and entry of foreign objects while the
second digit (0-8) defines the protection against water.
Protection First Digit Second Digit

0 No protection No protection
1 Large objects of more Vertically falling water
than 50 mm diameter
2 Medium-sized objects of Falling water at up to
more than 12 mm diameter 15 degrees from vertical
3 Small objects of more Falling water up to
than 2.5 mm 60 degrees from vertical
4 Granular objects of more Water splashes from
than 1 mm diameter any direction
5 Dust protected, Water from a nozzle
not completely tight from any direction
6 Dust tight Powerful water jet
7 Short-term immersion
8 Continuous immersion
The IP classification is set by the International

Electrotechnical Commission. The definitions for pro-
tections are different from that of NEMA. Therefore,
the IEC enclosure classification designations cannot
23
be exactly equated with NEMA enclosures.

As examples, IP 66 is approximately equivalent to
NEMA 4 or 4X , IP 67 is equivalent to NEMA 6, and
IP 55 is equivalent to NEMA 12.
Summary
There is no clear definition delineating a gas moni-
tor from an analytical analyzer. The distinction be-
tween the two is based largely upon their usage in the
actual application. A gas monitor is most frequently
used to monitor gases in toxic and combustible ranges
for area air quality and safety applications. For this
type of safety application, the concept of the units used
in the measurements and the definition of terminolo-
gies is somewhat unique. It is important to understand
the terms used. However, the examples of chemical
toxicity presented in this book do not consider the
complexity of actual toxicology and interreactions be-
tween contaminant chemicals. In critical applications,
a specialist on the subject needs to be consulted.
24
Chapter 2 Electrochemical Sensors
Chapter 2
Electrochemical
Sensors
T
he oldest electrochemical sensors date back to
the 1950s and were used for oxygen monitoring.
More recently, as the Occupational Safety and
Health Administration (OSHA) began requiring the
monitoring of toxic and combustible gases in confined
space applications, new and better electrochemical
sensors have been developed.
By the mid-1980s, miniaturized electrochemical
sensors became available for detection of many dif- Fig. 1. Electrochemical Sensors
ferent toxic gases in PEL ranges, with the sensors ex-
hibiting good sensitivity and selectivity. Currently, a
variety of electrochemical sensors are being used ex-
tensively in many stationary and portable applications
for personal safety. Figure 1 shows a small collection
of such electrochemical sensors.
The physical size, geometry, selection of various
components, and the construction of an electrochemi-
cal sensor usually depends on its intended use. Quite
often, the final design results in a compromise between
various performance parameters of the electrochemi-
cal sensor. The most common misconception about
electrochemical sensors is that they are all the same.
In fact, the appearance of the electrochemical sen-
sors used to detect various gases may be similar, but
their functions are markedly different. Consequently,
27
Gas Molecules Micro Ammeter

one can expect varying performance from each of
these sensors, in terms of sensitivity, selectivity, re-
Anode Cathode sponse time, and operating life.
+ –
For example, a low concentration gas sensor with
very high sensitivity uses a coarse-porosity hydrophobic
Electrolyte
membrane and less restricted capillary to allow more
gas molecules to pass through to produce enough sig-
Fig. 2 Basic Sensor nal for better sensitivity. However, this design also al-
lows more of the electrolyte’s water molecules to es-
cape out to the environment. In other words, an elec-
trochemical sensor with high sensitivity would have a
relatively short operating life due to evaporation of
moisture through the porous membrane.
Similarly, the electrolyte composition and the sens-
ing electrode material is selected based on the chemi-
cal reactivity of the target gas. By careful selection of
the electrolyte and/or the sensing electrode, one can
achieve the selectivity towards the target gas, but the
sensitivity may be reduced.
In summary, different electrochemical sensors may
appear very similar, but are constructed with differ-
ent materials including such critical elements as sens-
ing electrodes, electrolyte composition, and porosity
of hydrophobic barriers. Additionally, some electro-
chemical sensors use external electrical energy to
make them reactive to the target gas. All components
of the sensors play a crucial role in determining the
overall characteristics of the sensors.
Principle of Operation
Capillary
Diffusion Electrochemical sensors operate by reacting with
Barrier Hydrophobic the gas of interest and producing an elec-
Membrane
trical signal proportional to the gas concen-
Sensing Electrode
tration. A typical electrochemical sensor con-
Reference Electrode
sists of a sensing electrode (or working elec-
Counter Electrode
trode), and a counter electrode separated by a
Electrolyte thin layer of electrolyte, Figure 3.
Fig. 3 Typical Electrochemical Sensor Setup Gas that comes in contact with the sensor
28
first passes through a small capillary-type opening and

then diffuses through a hydrophobic barrier, and even- Water and Gas
tually reaches the electrode surface. This approach is
adopted to allow the proper amount of gas to react at
the sensing electrode to produce a sufficient electri-
cal signal while preventing the electrolyte from leak- Hydrophobic
ing out of the sensor, Figure 4. Membrane
The gas that diffuses through the barrier reacts at
Gas
the surface of the sensing electrode involving either an
oxidation or reduction mechanism. These reactions are
Fig. 4 HydrophobicElectrolyte
Membrane:
catalyzed by the electrode materials specifically devel- prevents liquid electrolyte from
oped for the gas of interest. leaking out.
With a resistor connected across the electrodes, a
current proportional to the gas concentration flows be-
tween the anode and the cathode. The current can be
measured to determine the gas concentration. Because
a current is generated in the process, the electrochemi-
cal sensor is often described as an amperometric gas sensor
or a micro fuel cell.
Importance of a Reference Electrode. For a sen-
sor requiring an external driving voltage, it is impor-
tant to have a stable and constant potential at the sens-
ing electrode. In reality, the sensing electrode potential
does not remain constant due to the continuous elec-
trochemical reaction taking place on the surface of the
electrode. It causes deterioration of the performance
of the sensor over extended periods of time. To improve
the performance of the sensor, a reference electrode is
introduced.
This reference electrode is placed within the elec-
trolyte in close proximity to the sensing electrode. A
fixed stable constant potential is applied to the sens-
ing electrode. The reference electrode maintains the
value of this fixed voltage at the sensing electrode. No
current flows to or from the reference electrode. The
gas molecules react at the sensing electrode and the
current flow between the sensing and the counter elec-
trode is measured and is typically related directly to
29
the gas concentration. The value of the voltage applied

to the sensing electrode makes the sensor specific to the
target gas.
The micro fuel cell-type electrochemical sensors do not
require an external driving voltage. For example, an elec-
trochemical sensor specific to oxygen has an anode, ei-
ther Pb or Cd, that supplies electrons for the reduction of
oxygen at the cathode. During the oxidation of the an-
ode, the electrons are released which then travel via an
external circuit to the cathode where oxygen molecules
consume the electrons as follows:
In acidic electrolyte
Oxidation at the anode: 2Pb + 2H2O → 2PbO + 4H+ + 4e-

Reduction at the cathode: O2 + 4H+ + 4e- → 2H2O
In basic electrolyte
Oxidation at the anode: 2Pb + 4OH- → 2PbO + 2H2O + 4e-
Reduction at the cathode: O2 + 2H2O + 4e- → 4OH-
The overall reaction in both cases is: 2Pb + O2 → 2PbO.

These types of sensors do not require a reference electrode.
Major Components
An electrochemical sensor consists of the following ma-
jor components:
A. Gas Permeable Membrane (also called hydrophobic
membrane): This is used to cover the sensor’s sens-
ing (catalyst) electrode and, in some instances, to
control the amount of gas molecules reaching the
electrode surface. Such barriers are typically made
of thin, low-porosity Teflon membranes. Such sen-
sors are called membrane clad sensors. Alternatively,
the sensing electrode is covered with a high-poros-
ity Teflon and the amount of gas molecules reach-
ing the electrode surface is controlled by a capil-
30
lary. Such sensors are referred to as capillary-type

sensors. Besides offering a mechanical protection to
the sensor, the membrane performs the additional
function of filtering out unwanted particulates. Se-
lecting the correct pore size of the membrane and
capillary is necessary to transfer the proper amount
of gas molecules. The pore size should be such as
to allow enough gas molecules to reach the sensing
electrode. The pore size should also prevent liquid
electrolyte from leaking out or drying out the sen-
sor too quickly.
B. Electrode: The selection of the electrode mate-
rial is very important. It is a catalyzed material
which performs the half cell reaction over a long
period of time. Typically, the electrode is made
from a noble metal, such as platinum or gold, and
catalyzed for an effective reaction with gas mol-
ecules. Depending on the design of the sensor, all
three electrodes can be made of different materi-
als to complete the cell reaction.
C. Electrolyte: The electrolyte must facilitate the cell
reaction and carry the ionic charge across the elec-
trodes efficiently. It must also form a stable refer-
ence potential with the reference electrode and be
compatible with materials used within the sensor.
If the electrolyte evaporates too quickly, the sensor’s
signal will deteriorate.
D. Filter: Sometimes a scrubber filter is installed in
front of the sensor to filter out un-
wanted gases. There is a limited selec- Capillary
Diffusion
tion of filters, each with different de- Barrier Charcoal Filter
grees of effectiveness. The most com-
monly used filter medium is activated Sensing Electrode
charcoal, as shown in Figure 5. The ac- Reference Electrode
tivated charcoal filters out most chemi- Counter Electrode
cals with the exception of carbon mon- Electrolyte
oxide and hydrogen gases. By properly
Fig. 5 Filtering with Activated Charcoal
selecting the filter medium, an elec-
31
trochemical sensor can be made more selec-

Typical Gases and the
Range of Measurement of
tive to its target gases.
Electrochemical Sensors Choosing the suitable materials for the above
components, and arranging the geometry of all these
GAS NAME PPM RANGE
components to determine the optimum operating per-
Ammonia, NH3 10 formance presents a challenge to scientists. Minor
Arsenic Hydride, AsH3 1 variations in the details of the sensor design can have
Bromine, Br2 30 a profound influence on the sensor’s accuracy, re-
sponse time, sensitivity, selectivity, and life expectancy.
Carbon Monoxide, CO 300
Chlorine, Cl2 5
Importance of Oxygen. The reactions at the sens-
ing electrode(anode) for some gases are as follows:
Chlorine Dioxide, ClO2 5
CO + H2O → CO2 + 2H+ + 2e-
Diborane, B2H6 1
Fluorine, F2 10 H2S + 4H2O → H2SO4 + 8H+ + 8e-
Germane, GeH4 2 NO + 2H2O → HNO3 + 3H+ + 3e-
Hydrogen, H2 2000
H2 → 2H+ + 2e-
Hydrogen Chloride, HCl 30
2HCN + Au → HAu(CN)2 + H+ + e-
Hydrogen Cyanide, HCN 30
Hydrogen Fluoride, HF 10 Simultaneously, the reactions at the counter elec-
Hydrogen Sulfide, H2S 30
trode (cathode) need oxygen molecules to complete
the process:
Nitric Oxide, NO 100
O2 + 4H+ + 4e- → 2H2O
Nitrogen Dioxide, NO2 50
Oxygen ppm levels to
An inadequate supply of oxygen to complete the
100% by vol. reaction will shorten the life of the sensors, hence the
Ozone, O3 3
sensors will not operate properly.
Sensors involving a reduction reaction of the tar-
Phosphine, PH3 1
get gas—such as the reduction of nitrogen dioxide,
Silane, SiH4 50 chlorine, and ozone—at the cathode produce water
Sulfur Dioxide, SO2 100 as a byproduct. At the anode, water is simultaneously
oxidized. Such sensors do not require the presence
of oxygen to function properly, as shown by the fol-
lowing:
NO2 + 2H+ + 2e- → NO + H2O
Cl2 + 2H+ + 2e- → 2HCl
O3 + 2H+ + 2e- → O2 + H2O
32
Characteristics
There are many different ways that electrochemi-
cal sensors are constructed, depending both on the
gas to be detected as well as the manufacturer. How-
ever, the main characteristics of the sensors are es-
sentially very similar. Following are some of the com-
mon characteristics of electrochemical sensors:
1. With a three-electrode sensor, there is nor-
mally a jumper which connects the working
and reference electrodes. If it is removed dur-
ing storage, it will take a long time for the sen-
sor to stabilize and be ready to be used. Some
sensors require a bias voltage between the elec-
trodes, and in such cases, the sensors are
shipped from the factory with a nine-volt bat-
tery powered electronic circuit. It takes any-
where from thirty minutes to twenty-four hours
for the sensor to stabilize, and it will continue
to stabilize over a three-week period.
When installed in a portable or stationary
instrument, the sensor cannot be removed
from power for an appreciable amount of time.
It is wise to double-check the instrument be-
fore use if batteries or power were removed at
some point. The portable instrument’s cir-
cuitry provides a small current needed to
maintain the sensor in the ready-to-use condi-
tion, even if the instrument is turned off.
Two-electrode sensors do not require any
bias voltage. For example, oxygen sensors do
not require a bias voltage.
2. Most of the toxic gas sensors require a small
amount of oxygen to function properly. There
is a vent hole on the side or back of the sensor
for this purpose. It is wise to double-check with
the manufacturer in applications that use non-
oxygen background gas.
33
3. Electrolyte within the sensor cell is an aqueous

solution separated by a hydrophobic barrier
which will not allow the aqueous solution to
leak out. However, water vapor can pass
through, just as other gas molecules can. In
high humidity conditions, prolonged exposure
can cause excessive water to build up and cre-
ate leakage. In low humidity conditions, the
sensor can dry out. Sensors that are designed
to monitor high gas concentrations have less
porous barriers to limit the amount of gas
molecules that pass through, and therefore are
not affected by the humidity as much as sen-
sors that are used to monitor low gas concen-
trations, which have more porous barriers and
allow a more free exchange of water molecules.
Pressure and Temperature

Electrochemical sensors are minimally affected by
List of Typical Gas pressure changes. However, it is important to keep the
Interference Ratios entire sensor within the same pressure since differen-
for CO Sensors tial pressure within the sensor can cause sensor dam-
age. Electrochemical sensors are also quite sensitive
Gas Without Filter With Filter to temperature and, therefore, the sensors are typi-
cally internally temperature-compensated. However,
H2S 0.3:1 10:1 it is better to keep the sample temperature as stable
as possible.
SO2 2:1 20:1 In general, when the temperature is above 25°C,
the sensor will read higher; when it is below 25°C, it
NO 3.3:1 10:1 will read lower. The temperature effect is typically 0.5%
to 1.0% per degree centigrade, depending on the
NO2 1.6:1 10:1 manufacturer and type of sensor.
H2 2:1 2:1 Selectivity
Electrochemical sensors are generally fairly selec-
The higher the interference ratio,
the less effect an interference tive to the target gas they are designed for. The de-
gas has on the sensor. gree of selectivity depends on the type of sensor, the
target gas, and the concentration of gas the sensor is
34
designed to detect. The best electrochemical sensor Typical Toxic Gas

is for the detection of O2, which has good selectivity, Electrochemical
is very reliable, and has a long life expectancy. Other Sensor Specification
electrochemical sensors are subject to interference
from other gases. A typical list of interference ratios Sensor Type: 2 or 3 electrodes; mostly 3
for CO sensors is shown on page 34 as an example. electrodes
The higher the ratio, the less the effect of inter- Range: 2-10 times permissible exposure limit
ference gas on the sensor. The interference data are Life Expectancy: 12 to 24 months normal;
taken using relatively low gas concentrations. In ac- depends on manufacturer and sensor
tual applications, interference concentrations can be Temperature Range: –40oC to +45oC

quite high, causing false readings and/or alarms. Relative Humidity: 15-95% noncondensing
Response Time T80 : < 50 sec.
Life Expectancy
Long Term Drift: drift down 2% per month
The life expectancy of an electrochemical sensor
depends on several factors, including the gas to be
detected and the environmental conditions in which
the sensor is used.
Generally, a one- to three-year life expectancy is
specified. In reality, the life expectancy will be highly
dependent on the total amount of gas exposed to
the sensor during its life, as well as other environ-
mental conditions, such as temperature, pressure and
humidity.
Summary
Electrochemical sensors require very little power
to operate. In fact, their power consumption is the
lowest among all sensor types available for gas moni-
toring. For this reason, the sensors are widely used in
portable instruments that contain multiple sensors.
They are the most popular sensors in confined space
applications.
A sensor’s life expectancy is predicted by its manu-
facturer under conditions that are considered normal.
However, the life expectancy of the sensor is highly
dependent on the environmental contaminants, tem-
perature, and humidity to which it is exposed.
35
Chapter 3 Catalytic Combustible Gas Sensors
Chapter 3
Catalytic Combustible
Gas Sensors
C
atalytic bead sensors are used primarily to de-
tect combustible gases. They have been in use
for more than 50 years. Initially, these sensors
were used for monitoring gas in coal mines, where
they replaced canaries that had been used for a long
period of time.
The sensor itself is quite simple in design and is
easy to manufacture. In its simplest form, as used in
the original design, it was comprised of a single plati-
num wire. Catalytic bead sensors were produced all
over the world by a large number of different manu-
facturers, but the performance and reliability of these Fig. 1 A Catalytic Bead Sensor
sensors varied widely among these various manufac-
turers. A catalytic bead sensor is shown in Figure 1.
Combustible gas mixtures will not burn until they
reach an ignition temperature. However, in the pres-
ence of certain chemical media, the gas will start to
burn or ignite at lower temperatures. This phenom-
enon is known as a catalytic combustion. Most metal
oxides and their compounds have these catalytic prop-
erties. For instance, volcanic rock, which is comprised
of various metal oxides, is often placed in gas burning
fireplaces. This is not only decorative, but it also helps
37
the combustion process and results in cleaner and

more efficient burning in the fireplace. Platinum,
palladium, and thoria compounds are also excellent
catalysts for combustion. This explains why the auto-
mobile exhaust system is treated with platinum com-
pounds and is called a catalytic converter. This kind
of gas sensor is made on the basis of the catalytic prin-
ciple, and therefore is called the catalytic gas sensor.
A gas molecule oxidizes on the catalyzed surface
of the sensor at a much lower temperature than its
normal ignition temperature. All electrically conduc-
tive materials change their conductivity as tempera-
ture changes. This is called the coefficient of temperature
resistance (Ct). It is expressed as the percentage of
change per degree change in temperature.
Platinum has a large Ct in comparison to other
metals. In addition, its Ct is linear between 500°C to
1000°C, which is the temperature range at which the
sensor needs to operate. Because the signal from the
sensor is linear, this means that the concentration of
gas readings are in direct proportion to the electrical
signal. This improves the accuracy and simplifies the
electronic circuitry. Also, platinum possesses excel-
lent mechanical properties. It is physically strong and
OUTPUT can be transformed into a fine wire which can be pro-
cessed into small sensor beads.
Furthermore, platinum has excellent chemical
properties. It is corrosion resistant and can be oper-
RB ated at elevated temperatures for a long period of time
d
ea
D.C. POWER
without changing its physical properties. It is capable

eB
tiv
Ac
R1 of producing a constant reliable signal over an ex-

Re
tended period of time.

fe
re
RB
nc
The electrical circuit used to measure the output

eB
ea
of catalytic sensors is called a Wheatstone bridge, in honor

d
Fig. 2 A catalytic bead sensor of English physicist and inventor Sir Charles Wheat-
Wheatstone bridge–a circuit for
measuring an unknown resist- stone (1802-75). Wheatstone bridges are commonly
ance by comparing it with known used in many electrical measurement circuits. As
resistances. shown in Figure 2, four circuit branches are arranged
38
in a square. The source of the electrical current is con-

nected, and between the other pair of opposite cor-
ners, the output measurement circuit is connected.
In operation, R1 is the trim resistor that keeps the
bridge balanced. A balanced bridge has no output sig-
nal. Resistor value RB and trim pot R1 are selected with
relatively large resistance values to ensure proper func-
tion of the circuit. When the gas burns on the active
sensor surface, the heat of combustion causes the tem-
perature to rise, which in turn changes the resistance
of the sensor. As the bridge is unbalanced, the offset
voltage is measured as the signal. It is important that
the reference sensor or bead maintains a constant re-
sistance during the exposure to the combustible gas;
otherwise, the measured signal will be inaccurate.
Evolution of the sensor. The original catalytic sen-
sor was a coil-shaped platinum wire. The coiled shape,
illustrated in Figure 3, was used to obtain a compact Fig. 3 Hot Wire Sensor
geometry for efficient heating and to produce a strong
enough signal to function as a gas sensor. Unfortu-
nately, despite the excellent physical and chemical
properties of platinum, it is a poor catalyst for com-
bustion of hydrocarbon gases.
For the proper detection of hydrocarbon gases,
the sensor requires a heated surface temperature be-
tween 900°C and 1000°C so that the sensor can prop-
erly react with gases at a sufficiently high and stable
rate. At this temperature, however, the platinum starts
to evaporate. The evaporation rate increases as the
gas molecules start to react with the sensor and as the
sensor temperature increases. This causes a reduc-
tion in the cross-section of the platinum wire, and, as
a result, the resistance increases. This affects the
sensor’s operating temperature, which shows up as zero
and span drifts.
The reference wire ideally should be the same as
the active wire, with the same geometry and operat-
ing temperature, but should be nonreactive with the
39
gas. This is not practically possible, however. A com-

promise is made by operating the reference wire at a
temperature that is substantially lower so that no oxi-
dation takes place in the presence of hydrocarbons.
In addition, the reference wire is chemically treated
to reduce the catalytic property of the platinum. This
may also be achieved by coating platinum wire with a
non-catalytic metal, such as gold.
Another problem with hot platinum wire is that it
becomes very soft at a temperature of 1000°C. There-
fore, it is difficult to maintain its coil shape. Also, the
coefficient of thermal resistance becomes less linear
as the temperature increases. This situation also re-
sults in poor zero and span quality of the sensor, as
well as a relatively short operating life.
One way to improve stability of the sensor is to
coat the platinum wire with suitable metal oxides.
Thus, the final step is to treat the finished sensor or
bead with a catalyst, such as platinum, palladium or
thoria compounds. Figure 4, shows the sensor bead.
PLATINUM WIRE
CATALYTICALLY-TREATED
METAL OXIDE
Fig. 4 A Catalytic Bead Sensor
The construction of the catalytic sensor bead is

analogous to constructing a building by using rein-
forced concrete. The coating makes the sensor physi-
cally very rugged. The sensor becomes a very small
40
mass which helps make it resistant to shock and vibra-

tions. Most importantly, the catalyst coating reduces
the temperature needed to achieve a stable signal for
hydrocarbons between 400°C and 600oC.
The use of fine diameter wire not only reduces
the size of the sensor, but it also increases the signal,
because finer wire has a higher magnitude of resistive
value and the signal output is the percentage change
of total wire resistance. This also reduces power con-
sumption.
The reference sensor can be constructed in the
same way as the active sensor, with the exception that
the catalyst chemical is eliminated. The bead can be
further treated with chemicals, such as potassium, to
prevent the reference bead from reacting with the gas.
A near perfectly compensated pair of sensors is now
possible. The sensor is called a “catalytic” sensor be-
cause the use of the catalyst is the main ingredient
involved in the proper functioning of the sensor. The
catalytic sensor is stable, reliable, accurate, and rug-
ged, and has a long operating life. The output is lin-
ear because the platinum wire has a good linear coef-
ficient of thermal resistance.
Characteristics
The sensor’s output is directly in proportion to
the rate of oxidation. The maximum output of the
signal occurs at about the stoichiometric1 mixture of 1
Pertaining to substances that are in the
the gas, or it is based on the theoretical combustion exact proportions required for a given re-
reaction formula. Methane, for example: action.
CH4 + 2O2 + 8N2 → CO2 + 2H2O + 8N2

It takes 10 moles of air for one mole of methane
to complete the reaction, assuming there is one part
of oxygen and four parts of nitrogen in air.
Therefore, for a theoretical combustion to take
place, one part of methane will require 10 parts of air
41
to complete the combustion, or theoretically 9.09%

of methane in a mixture of air.
For a sensor to detect methane, the signal out-
put will respond linearly from 0–5% of methane
(which is 100%LEL). As the concentration reaches
close to the stoichiometric value of 9%, the signal
increases very rapidly and peaks at around 10%. The
signal starts to drop slowly as the concentration of
gas passes approximately 20%; after 20% it drops
straight down to a level that reflects no output as the
concentration of gas reaches 100%. Figure 5, illus-
trates this effect.
0-3%: NEARLINEAR; 3-5%: SLIGHTLY LESS LINEAR
LEL UEL
10
8
Sensor Output
6
in volts
0
5 10 15 20 40 60 80 100
% Methane Gas Concentration
Fig. 5 Sensor Output vs. Gas Concentration
Consider another example, propane. The reaction

formula for propane is:
C3H8 + 5O2 + 20 N2 → 3CO2 +4H2O + 20N2
or one part of propane per 25 parts of air for theoreti-
cal combustion of propane. The actual theoretical
combustion concentration for propane is 3.85%.
The LEL for methane is 5% and for propane is
2.1%. This value is near half of the theoretical com-
bustion value. There is a safety factor of 2 added to
ensure safety.
42
Sensor Operation Factors

There are several factors affecting the operation
of the catalytic sensor.
1. Catalyst Poisoning: There are chemicals which
will deactivate the sensor and cause the sensor to lose
sensitivity and eventually become totally nonrespon-
sive to gases. The most common chemicals that can
poison catalytic sensors are those that contain silicon,
such as the common oil and lubricants with silicon
compounds used as additives in machinery. Sulfur
compounds, which are often released with gases, chlo-
rine, and heavy metals also cause the poisoning of the
sensor.
The exact cause of this poisoning is very difficult
to identify. Some chemicals, with very small concen-
trations, will totally destroy the sensor. There have been
instances in which the silicon contained in simple
hand lotions has caused problems with catalytic sen-
sors.
2. Sensor Inhibitors: Chemicals such as halogen
compounds, which are used in fire extinguishers and
Freon used in refrigerants, will inhibit the catalytic
sensor and cause it to temporarily lose the ability to
function.
Normally, after 24 or 48 hours of exposure to am-
bient air, the sensor starts to function normally. These
are just a few typical chemicals that inhibit the sensor
performance and are by no means to be considered
as the sole possible inhibitors.
3. Sensor Cracking: The sensor, when exposed
to excessive concentration of gases, excessive heat,
and the various oxidation processes that take place
on the sensor surface, may eventually deteriorate.
Sometimes this will change the zero and span set-
ting of the sensor.
4. Correction Factors: Catalytic sensors are most
43
Relative Sensitivity commonly calibrated to methane for 0-100% LEL full

scale range.
As an example for a typical sensor The manufacturers generally provide a set of cor-
calibrated for 100% LEL methane
gas, the relative sensitivity to rection factors that allow the user to measure different
other gases is as follows: hydrocarbons by simply multiplying the reading by
the appropriate correction factor to obtain the read-
Gas Reading
ing of a different gas. The reason for using methane
Methane 100%
as the primary calibration gas is that methane has a
Propane 60% saturated single bond that requires the sensor to op-
n-Butane 60% erate at the highest temperature in comparison to
n-Pentane 50% other hydrocarbons. For instance, a typical catalytic
n-Hexane 45% sensor for methane gas may require a 2.5-volt bridge
Methanol 100% voltage to obtain a good signal, while the same sensor
will only need 2.3 volts for butane gas. Therefore, if
Ethanol 70%
the sensor is set to read butane, it will not read meth-
iso-Propyl Alcohol 60% ane properly.
Acetone 60% In addition, methane gas is a very common gas
Methyl Ethyl Ketone 50% and is often encountered in many applications. Fur-
Toluene 45% thermore, it is also easy to handle and has the ability
to be mixed into different concentrations easily.
However, it should be noted that the correction fac-
tors are a set of numbers that should be used with
great care. The correction factors can vary from sen-
sor to sensor, and they can even change on the same
sensor as the sensor ages. Therefore, the best way to
obtain precise readings for a specific gas is to actually
calibrate the sensor to the gas of interest directly.
5. Percent LEL for Mixtures of Hydrocarbons:
For combustion to take place, the following require-
ments must be present:
a. Combustible mixture
b. Oxygen
c. Ignition source
This is sometimes referred to as the combustion
triangle. But in real life, the process of igniting a com-
bustible mixture is much more complicated. The en-
44
vironmental conditions, such as pressure, tempera-

ture, temperature of the ignition source, and even
humidity can have an affect on the combustible mix-
ture concentration.
If two or more chemicals are involved, it is not even
possible to calculate and determine the combustion
range of the mixture. Therefore, it is best to consider
the worst-case scenario and calibrate the sensor ac-
cordingly. Furthermore, a sensor calibrated at a per-
centage LEL for one gas cannot necessarily be used
for other gases. Many instruments on the market to-
day have a scale unit as a percentage of LEL without
indicating that the unit is calibrated on methane.
Therefore, if the unit is used for some other gas or
mixture of gases, the data can be totally meaningless.
For example, a catalytic sensor calibrated on meth-
ane produces lower readings when exposed to hydro-
carbons of higher carbon content, while infrared in-
struments will produce much higher readings if ex-
posed to a higher carbon content gas. This is a very Typical Specifications
common mistake made by many users of gas detec- for Catalytic Sensors
tion equipment.
Sensor Type: Diffusion catalytic bead
Summary Temperature Range: –40oC to +60oC
A catalytic sensor is relatively easy to manufacture. Response Time: 10 to 15 sec. to 90% of
However, the quality of the sensor varies quite drasti- reading
cally from one manufacturer to another. Accuracy: ±5%
The overall technology of making a sensor for
Repeatability: 2%
the market is more of an art than a predictable sci-
entific event. This is particularly true in selecting, Drift: 5–10% per year
preparing and processing all the chemicals needed Life Expectancy: Up to 3 years; depending
to make the final sensor. There are too many vari- on application
ables in the process that inhibit the making of a pre- Sensors can be remotely mounted up to
dictable final product. Therefore, most users of cata- 2,000-3,000 meters, depending on the
manufacturer and cable size used to wire
lytic sensors select their sensors based on the reputa- the sensor.
tion of the manufacturer.
45
Chapter 4 Solid-State Gas Sensors
Chapter 4
Solid-State Gas Sensors
W
hen scientists were doing research work re-
lated to semiconductor p-n junctions,1 they
discovered that these junctions were sensitive
to environmental background gases.
At that time, such a behavior was considered a
problem. This problem, however, was solved by en-
capsulating the semiconductor chip so that it was no
longer exposed to the outside environment. Subse-
quently, unsuccessful attempts were made to utilize
the sensitivity of the semiconductor junction as a gas
detection device.
It wasn’t until 1968 that Mr. N. Taguchi marketed Fig. 1 A solid-state sensor
a simple semiconductor, or a solid-state sensor, for the used for detecting more than a
detection of hydrocarbons in LEL combustible ranges. hundred toxic gases.
The intention was to provide an alternative to the
1
popular catalytic bead sensor, which suffered from The positive and negative junctions
several problems, including loss of sensitivity with time in a semiconductor.
due to poisoning and burning out when exposed to
high gas concentrations.
In 1972, International Sensor Technology (IST) in
Irvine, California introduced a solid-state sensor for
the detection of hydrogen sulfide in a range of 0-10
ppm. A few years later, IST developed solid-state sen-
sors for the detection of more than 100 different haz-
ardous gases at low ppm levels. This was a significant
development, since OSHA was being formed at about
47
the same time and began to regulate acceptable gas

concentration levels for safety at the workplace.
Today, solid-state sensors are available for the de-
tection of more than 150 different gases, including
sensors for the gases which could otherwise only be
detected using expensive analytical instruments.
There are now several manufacturers of solid-state
sensors, but each sensor has different characteristics
and different manufacturers offer different levels of
performance and quality.
Properly manufactured, solid-state sensors offer a
very long life expectancy. It is not unusual to find fully
functional sensors that were installed 30 years ago.
The Rise in Popular Use of Solid-State Sensors. In
the early 1980s, Japan passed a law that required gas
detectors to be installed in residential apartments
where gas bottles were being used. For this huge mar-
ket, the competition was between solid-state and cata-
lytic bead sensors.
While there were some initial complaints about
solid-state sensors that produced false alarms, this was
far outweighed by the long life the sensor provided.
Because catalytic sensors burn the gas being detected,
sensor material is consumed and changed in the pro-
cess and the sensor eventually burns out.
With solid-state sensors, on the other hand, gas
simply “adsorbs” onto the sensor surface, changing
the resistance of the sensor material. When the gas
disappears, the sensor returns to its original condi-
tion. No sensor material is consumed in the process,
and hence the solid-state sensors offer a long life ex-
pectancy.
After a few years of usage, the catalytic sensors had
become less popular for such applications due to the
need for frequent sensor replacement.
A solid-state sensor consists of one or more metal
48
oxides from the transition metals, such as tin oxide,

aluminum oxide, etc. These metal oxides are prepared
and processed into a paste which is used to form a
bead-type sensor. Alternatively, thick or thin film-chip
sensors are made when the metal oxides are vacuum
deposited onto a silica chip, in a fashion similar to
making semiconductors.
A heating element is used to regulate the sen-
sor temperature, since the finished sensors exhibit
different gas response characteristics at different
temperature ranges. This heating element can be a
platinum or platinum alloy wire, a resistive metal
oxide, or a thin layer of deposited platinum. The
sensor is then processed at a specific high tempera-
ture which determines the specific characteristics
of the finished sensor.
In the presence of gas, the metal oxide causes the
gas to dissociate into charged ions or complexes
which results in the transfer of electrons. The built-
in heater, which heats the metal oxide material to an
operational temperature range that is optimal for the
gas to be detected, is regulated and controlled by a
specific circuit.
A pair of biased electrodes are imbedded into the
metal oxide to measure its conductivity change. The
changes in the conductivity of the sensor resulting
from the interaction with the gas mol-
ecules is measured as a signal. Typically, Platinum
a solid-state sensor produces a ver y Coil Heater Collector
strong signal, especially at high gas con- VOUT
centrations.
There are different ways of making
solid-state sensors, each arrangement
making the sensor’s performance
Heater
characteristics different. Two typical Control
styles are the following:
1. Bead-type sensor (Figure 2)
2. Chip-type sensor (Figure 3, next page) Fig. 2. Schematic Diagram of a Bead-type Sensor
49
Most solid-state sensors have three

or four pins, depending on how the
Collector heater and bias electrodes are con-
Termination
nected.
Termination
Characteristics
Silicon Substrate Solid-state sensors are among the
Heater Resistor most versatile of all sensors, as they de-
tect a wide variety of gases, and can be
Fig. 3 Chip-type Sensor
used in many different applications. Dif-
ferent response characteristics are achieved by vary-
ing the semiconductor materials, processing tech-
niques, and sensor operating temperature.
Among the unique attributes of the solid-state sen-
sor are the abilities of the sensor to detect both low
ppm levels of gases, as well as high combustible levels.
Longevity. The main strength of the solid-state
sensor is its long life expectancy, as the sensor typi-
cally lasts 10 years or more in clean applications. This
is a major advantage compared to other sensor types,
such as catalytic bead or electrochemical sensors,
which typically last only one to two years.
However, while solid-state sensors have a longer
life expectancy, they are also more susceptible to in-
terference gases than the other types of sensors. Thus,
in applications where other background gases are
present, solid-state sensors may trigger false alarms.
In certain instances, the interferences from other
gases are minimized by using appropriate filtering ma-
terials that absorb all other gases except the gas to be
detected.
For example, a solid-state sensor for monitoring
carbon monoxide and hydrogen can be equipped with
a charcoal filter which eliminates the majority of in-
terfering gases. This way the sensor performs very well
and becomes very selective for those two gases.
Versatility. The versatility of the solid-state sensor
50
is one of its main advantages. For example, in chemi-

cal plants, a gas monitoring system may involve the
monitoring of many different gases and ranges, or
even the same gas with multiple ranges.
Often, the lower ranges need to be monitored for
certain gases for toxic concentrations while simulta-
neously, the same gas needs to be monitored in the
combustible range for explosive concentrations.
The solid-state sensor is capable of detecting gas Typical Specifications
in both ranges. This greatly simplifies the system de- for Solid-State Sensors*
sign and maintenance required because it eliminates Accuracy: ±3% to 10% of full scale
or minimizes the use of multiple sensor technologies Response Time: T 80 ranges from 20
which must be designed and maintained differently. seconds to 90 seconds
A list of gases that are detectable by using IST’s Temperature Range: –20oC to +50oC
solid-state sensors is shown in Table 1 (pages 52-53).
Humidity: 5–9%
The typical ppm ranges which are chosen are
three to five times the permissible exposure limits Life Expectancy: 10+ years
for an eight-hour work day or ranges that can be Power Consumption: Approx. 300mW
based on the sensitivity as well as the interference
*Actual specifications will vary depending on
characteristic of the sensor, whichever is the most gas and range.
practical in an application.
51
Table 1: IST’s Solid-State Sensors Gas List

FULL-SCALE RANGE FULL-SCALE RANGE
GAS Low ppm GAS Low ppm
or higher LEL or higher LEL
Acetic Acid 100 —- Deuterium yes
Acetone 100 yes Diborane 10 —-
Acetonitrile 100 —- Dibromoethane 50 —-
Acetylene 50 yes Dibutylamine —- yes
Acrolein(Acrylaldehyde) 50 —- Dichloroethane (EDC) 50 yes
Acrylic Acid 100 —- Dichlorofluoroethane 100 —-
Acrylonitrile 50 yes Dichloropentadiene 50 —-
Allyl Alcohol —- yes Dichlorosilane 50 —-
Allyl Chloride 200 —- Diesel Fuel 50 yes
Ammonia x50 yes Diethyl Benzene —- yes
Anisole 100 —- Diethyl Sulfide 10 —-
Arsenic Pentafluoride 5 —- Difluorochloroethane —- yes
Arsine 1 —- Difluoroethane (152A) —- yes
Benzene 50 yes Dimethyl Ether —- yes
Bophenyl 50 yes Dimethylamine (DMA) 20 —-
Boron Trichloride 500 —- Epichlorohydrin 50 —-
Boron Triflluoride 500 —- Ethane 1000 —-
Bromine 20 —- Ethanol 200 yes
Butadiene 50 yes Ethyl Acetate 200 yes
Butane 400 yes Ethyl Benzene 200 yes
Butanol 1000 yes Ethyl Chloride 100 yes
Butene —- yes Ethyl Ether 100 yes
Butyl Acetate 100 yes Ethylene 100 yes
Carbon Disulfide 50 —- Ethylene Oxide 5 yes
Carbon Monoxide 50 yes Fluorine 20 —-
Carbon Tetrachloride 50 —- Formaldehyde 15 —-
Cellosolve acetate 100 —- Freon-11 1000 —-
Chlorine 10 —- Freon-12 100 —-
Chlorine Dioxide 10 —- Freon-22 100 —-
Chlorobutadiene —- yes Freon-113 100 —-
Chloroethanol 200 —- Freon-114 1000 —-
Chloroform 50 —- Freon-123 1000 —
Chlorotrifluoroethylene —- yes Fuel Oil or Kerosene —- yes
Cumene —- yes Gasoline 100 yes
Cyanogen Chloride 20 —- Germane 10 —-
Cychlohexane 100 yes Heptane 1000 yes
Cyclopentane 50 —- Hexane 50 yes
52
Table 1 (continued) : IST ’s Solid-State Sensors Gas List

FULL-SCALE RANGE FULL-SCALE RANGE
GAS Low ppm GAS Low ppm
or higher LEL or higher LEL
Hexane 100 yes Nitric Oxide 20 —-

Hydrazine 5 —- Nitrogen Dioxide 20 —-
Hydrogen 50 yes Nitrogen Trifluoride 50 —-
Hydrogen Bromide 50 —- Nonane 2000 —-
Hydrogen Chloride 50 —- Pentane 200 yes
Hydrogen Cyanide 20 —- Perchloroethylene 200 —-
Hydrogen Fluoride 20 —- Phenol 100 —-
Hydrogen Sulfide 5 yes Phosgene 50 —-
Isobutane 1000 yes Phosphine 3 —-
Isobutylene —- yes Phosphorus Oxychloride 200 —-
Isopentane 1000 —- Picoline —- yes
Isoprene —- yes Propane 100 yes
Isopropanol 200 yes Propylene 100 yes
JP4 1000 yes Propylene Oxide 100 yes
JP5 1000 yes Silane 10 —-
Methane 100 yes Silicon Tetrachloride 1000 —-
Methanol 200 yes Silicon Tetrafluoride 1000 —-
Methyl Acetate 30 —- Styrene 200 yes
Methyl Acrylate 60 —- Sulfur Dioxide 50 —-
Methyl Bromide 20 —- Tetrahydrofuran 200 yes
Methyl Butanol —- yes Tetraline 100 —-
Methyl Cellosolve —- yes Toluene 50 yes
Methyl Chloride 100 yes Toluene Diisocyanate 15 —-
Methyl Ethyl Ketone 100 yes Trichloroethane 50 —-
Methyl Hydrazine 5 —- Trichloroethylene 50 yes
Methyl Isobutyl Ketone 200 yes Triethylamine (TEA) 100 —-
Methyl Mercaptan 30 —- Trifluoroethanol 25 —-
Methyl Methacrylate 100 yes Trimethylamine (TMA) 50 —-
Methyl-Tert Butyl Ether —- yes Tungsten Hexafluoride 50 —-
Methylene Chloride 20 yes Turpentine —- yes
Mineral Spirits 200 yes Vinyl Acetate 1000 yes
Monochlorobenzene —- yes Vinyl Chloride 20 yes
Monoethylamine 30 —- Vinylidene Chloride 50 —-
Morpholine 500 —- Xylene 100 —-
Naptha 1000 yes
Natural Gas 1000 yes
53
Chapter 5 Infrared Gas Sensors
Chapter 5
Infrared Gas Sensors
I
nfrared (IR) gas detection is a well-developed
measurement technology. Infrared gas analyzers
have a reputation for being complicated, cumber-
some, and expensive. However, recent technical ad-
vancements, including the availability of powerful am-
plifiers and associated electronic components, have
opened a new frontier for infrared gas analysis. These
advancements have resulted from an increase in de-
mand in the commercial sector, and these demands
will likely continue to nourish the advancement of this Gas Cell
technology.
Gases to be detected are often corrosive and reac-
Fig. 1 An example of an IR
tive. With most sensor types, the sensor itself is directly gas monitor with the gas cell
exposed to the gas, often causing the sensor to drift assembly exposed.
or die prematurely.
The main advantage of IR instruments is that the
detector does not directly interact with the gas (or
gases) to be detected. The major functional compo-
nents of the analyzer are protected with optical parts.
In other words, gas molecules interact only with a light
beam. Only the sample cell and related components
are directly exposed to the gas sample stream. These
components can be treated, making them resistant to
corrosion, and can be designed such that they are eas-
ily removable for maintenance or replacement.
Today, many IR instruments are available for a wide
variety of applications. Many of them offer simple,
55
rugged, and reliable designs. In general, for toxic and

combustible gas monitoring applications, IR instruments
are among the most user friendly and require the least
amount of maintenance. There is virtually an unlimited
number of applications for which IR technology can be
used. Gases whose molecules consist of two or more dis-
similar atoms absorb infrared radiation in a unique man-
ner and are detectable using infrared techniques. Infra-
red sensors are highly selective and offer a wide range
of sensitivities, from parts per million levels to 100 per-
cent concentrations. This chapter provides general in-
formation, with a special emphasis on instruments used
for area air quality and safety applications.
The infrared detection principle incorporates only a
small portion of the very wide electromagnetic spectrum.
The portion used is that which we can feel as heat. This
is the region close to the visible region of the spectrum
to which our eyes are sensitive. Electromagnetic radia-
tion travels at close to 3 x 108 m/sec and has a wave-like
profile. Let’s review the basic physics of electromagnetic
radiation by defining the terminology involved with it.
Wave: Similar to a wave in the ocean, the electro-
magnetic radiation waves oscillate, one wave followed
by another. There are both electromagnetic and me-
chanical waves, with mechanical waves having a much
longer wavelength. Figure 2 illustrates a mechanical wave.
1 cm 2 cm
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 sec 2 sec
Frequency = 10 Hz. Wavelength = 0.1 cm, Wave Number = 10 cm -1
Fig. 2 A simple mechanical wave showing 10 waves per centimeter to

illustrate the concept of the wave.
56
Frequency: Number of waves per second passing

through a point. An electromagnetic wave travels at
the speed of light which is 300 million meters per sec-
ond, or 3 x 108 m/sec. Therefore, the frequency is the
speed of light divided by the wavelength, and is ex-
pressed as the number of waves per second, or hertz
(Hz).
Wavelength: The distance between two peaks of
the wave, or the spacing between two waves. It is com-
monly expressed in microns. It is a very popular term
used in representing gas molecular absorption bands
as well as optical component characteristics.
Wave number: The number of waves in one centi-
meter. It is the reciprocal of wavelength. Since 1 mi-
cron = 10-6 m = 10-4 cm, the reciprocal of one micron is
1/10-4 (10,000 wave numbers per cm), and 2 microns =
5000 wave numbers per cm. The formula is:
Wave number = 1/wavelength
Micron: A common unit used to express wave-
length in the infrared region. It is one millionth of a
meter(10-6m) or a micrometer, and is abbreviated as µ.
Transmittance: The ratio of transmitted radiation
energy to the incident energy. The energy not trans-
mitted is absorbed and reflected. It is used to specify Reflected Energy
optical parts. 10
Absorbance: Opposite of transmittance. Used to

describe the amount of energy absorbed by gas mol- Absorbed
Incident Energy Energy Transmitted Energy
ecules. Both percent absorption and percent transmittance 100 25 65
are used as the y-axis versus wave number or wavelength
as the x-axis in the infrared spectra.
Wave number and wavelength are common terms
65
used by scientists to describe the infrared region for Transmittance =
100
= 65%
gas analysis because they provide a convenient method
to express radiation frequency and the mechanisms of
interaction between infrared radiation and gas mol-
ecules. Mathematically, they are the reciprocals of each
57
Ultra-
Radio Television Microwave Infrared
violet
X-rays Gamma rays
6 9 12 15 18 20
10 10 10 10 10 10
4 x 1014 8 x 1014
Frequency (cycles/sec.)
Fig. 3 Location of Infrared in the Electromagnetic Spectrum
other. For example, methane gas has the absorption wave-

length of 3.4 microns, or a wave number of 2941 cm-1.
(Figure 4, on page 59, shows a spectroscopic description
of methane gas which illustrates that methane has a strong
absorption peak at 3.4µ, or a wave number of 2941 cm-1.)
Electromagnetic waves propagate through space or
matter by oscillating electric and magnetic fields. In a
vacuum, they travel at the speed of light. The complete
range of frequencies of these waves is called the electro-
magnetic spectrum.
These frequencies range from gamma rays of 1020 Hz
to radio waves of 106 Hz. They are classified from higher
to lower frequencies as gamma rays, x-rays, ultraviolet
light, visible light, infrared light, microwaves, and radio
waves. Figure 3 shows the electromagnetic spectrum.
Visible light, at about 4 x 1014 Hz (or 0.4 to 0.7 mi-
crons), is actually only a very narrow portion of the spec-
trum. Infrared is just below visible light, and this explains
why we feel, but do not see, temperature. The infrared
region is most useful for gas analysis because absorption
by gas molecules is unique and selective in this region.
Unique Gas Absorption “Fingerprints.” The com-
plexity of the gas molecules determines the number of
absorption peaks. The more atoms that form a molecule,
the more absorption bands that will occur. The region
in which this absorption occurs, the amount of absorp-
tion, and the specific character of the absorption curve
is unique to each gas. Gas molecules can be fingerprinted
using their absorption characteristics and archived for
58
METHANE CH4
CH4
Fig. 4 A Spectroscopy Scan of Methane Gas
gas analysis and identification purposes. A library of these

curves can then be stored in the memory inside an in-
strument. When a given gas is scanned by the instrument,
the graph is then compared with the stored curves to
identify the gas molecules. This method of gas analysis is
the most popular in analytical chemistry.
In gas monitoring applications, only one specific ab-
sorption region is used to quantitatively determine the
gas concentration. The wavelengths in this region are
between 2 and 15 microns or wave numbers of 5,000 to
670 cm-1. A typical spectroscopy scan of methane is
shown in Figure 4.
The scan shows that methane gas has a strong absorp-
tion peak at 3.4 microns, which is the wavelength used
for methane detection. As a matter of fact, most com-
mon hydrocarbon gases have strong absorption in the
3.4 micron region. On the other hand, carbon dioxide is
absorbed strongly at 4.26 microns, while carbon monox-
ide absorbs strongly at 4.7 microns.
Natural Frequencies of Gas Molecules. Gas molecules
are made up of a number of atoms bonded to one an-
other. These interatomic bonds are similar to Natural
Frequency
springs, connecting atoms of various masses
together. Figure 5 illustrates water molecules
Oxygen
with one oxygen and two hydrogen atoms. This Hydrogen
bonding vibrates with a fixed frequency called

Hydrogen
the natural frequency. Fig. 5 Water Molecule
59
All matter has a natural frequency. The Golden

Gate bridge in San Francisco has a natural frequency,
and the balcony in a theater has a natural frequency.
Even though these are mechanical waves, in theory
they are similar to electromagnetic waves. If the wind
or an earthquake shakes a bridge at the same fre-
quency as the bridge’s natural or resonant frequency,
1
0
this can cause a much more violent vibration of very
1 high amplitude, as shown in Figure 6, resulting in
Natural Frequency severe damage. People moving around a balcony gen-
2 erate vibrations of certain frequencies, which can
1 cause a similar effect. Thus, the natural frequency of
0
a structure is a concern for structural engineers.
1
2 Gas molecules have a number of natural frequen-
External Resonant Wave cies. The larger the molecules, the more modes of
3 natural frequencies they have. Natural frequencies are
2 also determined by the molecular structure of the
1
chemicals. They are always the same for a given mol-
0
1
ecule and bonding structure. The particular proper-
2 ties exhibited by the chemical become its signature
3
and offer clues to identify the molecular structure of
Resultant Wave
the given chemical compounds.
Fig. 6 Energy Absorption by
Gas Molecules
Two Ways of Detection. Infrared radiation con-
tains a wide spectral content. When this radiation in-
teracts with gas molecules, part of the energy has the
same frequency as the gas molecule’s natural fre-
quency and it is absorbed while the rest of the radia-
tion is transmitted. As the gas molecules absorb this
radiation, the molecules gain energy and vibrate
more vigorously.
This vibration results in a rise in the temperature
of the gas molecules. The temperature increases in
proportion to gas concentration, and is detected by
the detector. On the other hand, the radiation ab-
sorbed by the gas molecules at the particular wave-
length will cause a decrease in the original source
strength. This radiation energy decrease can be de-
tected as a signal also.
60
Key Components for Analysis

To get a useful signal for gas analysis, there are
many different components and parts used in the vari-
ous designs found today. However, there are no spe- THERMOCOUPLE
cific rules regarding the selection of each of the com-
ponents. The configuration of these instruments de- Metal 1
pends on what needs to be accomplished and the dis- – +
cretion of the designer. Following is a description of
the major components. Dissimilar
1. Detector: Infrared detectors convert electro- Metal 2 Metal
Junction
magnetic radiation energy or temperature changes T1 T2
into electrical signals. There are many types of infra-
red detectors and each detector type offers a wide
range of performance characteristics. Some of the
T1 = Reference or normally ice point
types are described briefly below:
T2 = Measuring probe
a. Thermoelectric: A detector that converts tem-
perature into an electrical signal is commonly
known as a thermocouple. The junction of dissimi-
lar metals generates a voltage potential, which is
directly proportional to the temperature. This
junction can be made into multiple junctions to
improve sensitivity. Such a configuration is called
a thermopile.
Using techniques used in the semiconductor
industry, many junctions are connected in a series
to multiply the output of the device, yet the detec-
tor package is miniaturized and compact. The size
and mass of the device are important in determin-
ing its response time and other characteristics of
the device.
This detector has a relatively slow response
time, but offers the advantages of DC stability, re-
quiring no bias, and responding to all wavelengths.
It is the simplest way of converting light energy
into an electrical signal.
b. Thermistor Bolometer: A bolometer changes re-
sistance when incident infrared radiation interacts
61
with the detector. This thermally sensitive semi-

conductor is made of a sintered metal oxide ma-
terial. It has a high temperature coefficient of re-
sistance.
PYROELECTRIC c. Pyroelectric Detector: Pyroelectric materials are
crystals, such as lithium tantalate, which exhibit
spontaneous polarization, or a concentrated elec-
Infrared tric charge that is temperature dependent.
As infrared radiation strikes the detector sur-
face, the change in temperature causes a current
Pyroelectric Crystal
to flow. This current is proportional to the inten-
sity of the radiation. This detector exhibits good
sensitivity and good response to a wide range of
wavelengths, and does not require cooling of the
detector. It is the most commonly used detector
for gas monitors.
d. Photon Detector: Photons possess energy based
on their wavelength and intensity. A photon de-
tector detects the quantum interaction between
incident photons and semiconductor material. A
photon that strikes an electron with sufficient en-
ergy can raise the electron from a nonconducting
state into a conducting state. The presence of elec-
trons in the conduction band will increase the con-
ductivity of the chip, and a bias voltage registers
this change as a signal.
The excitation of electrons requires photons
to have a certain amount of energy. Shorter wave-
lengths have higher frequencies, and therefore
more energy. This detector functions in a limited
spectral region, which depends on the detector
material being used. Typically, the detector must
be cooled with a thermoelectric cooler or even
liquid nitrogen for it to function properly. A typi-
cal example of such detectors are lead sulfide
(PbS) detectors used in the 1-3 micron region, and
lead selenide (PbSe) detectors used in the 1-5
micron region.
62
e. Luft Detector: The word “luft” is a German LUFT

word meaning “air,” and the original luft detec-
tors were designed in Germany. A luft detector con-
sists of two chambers, either linked by a micro flow IR Source
sensor or divided by a diaphragm. The chambers Chopper

are sealed with a target gas at a low pressure. IR Sample Motor
Out
transparent windows are fitted to seal the cham-
bers and the same intensity of pulsed infrared ra-
diation is received by both chambers when no tar-
get gas is present. SAMPLE REFERENCE
CELL CELL
When the sample containing target gas flows
through the sample cell, a reduction in radiation Sample IR-Transparent
energy is received by the detector chamber, which In Window
causes the temperature and pressure to drop in
the detector chamber. The amount of temperature Reference
Chamber
or pressure drop is in direct proportion to the gas
Detector
concentration. In the case of linked chambers, the Chamber Micro-Flow Sensor
or Diaphragm
pressure difference between the two chambers
causes a detectable flow, which is measured as a
signal. In the case where a diaphragm separates
the two chambers, a movement of the diaphragm
causes a measurable change in capacitance.
This detector finds many useful applications
as an analyzer, and has good potential for further
development.
f. Photoacoustic Detector: This detector is similar
to the luft detector except that the pressure change
is measured by a condenser microphone. The PHOTOACOUSTIC
sample gas is passed through a chamber at a pre- Infrared
set time interval and the chamber is sealed with a
fixed volume of sample gas trapped inside. A spe-
cific wavelength of infrared radiation is pulsed into
the chamber via an infrared transparent window.
The pulsating pressure change is measured by the Microphone
microphone as a frequency change which pro-
duces the signal.
2. Infrared Source: A regular incandescent light
bulb is a good infrared source. A heated wire filament,
63
similar to that in a pen flashlight, radiates sufficient

energy in the 1-5 micron range for the detection of
most hydrocarbons, carbon dioxide, and carbon mon-
oxide. This simple and inexpensive light source of-
fers long life and long-term stability.
Any source that can generate enough radiation at
the wavelength of interest for the purpose of detect-
ing the specific target gas may be used. There are many
light sources available, ranging from specially designed
heating filaments to electronically generated sources.
Modulating the light. Depending on the type of
detector used, it may be necessary to modulate the
light source, turning it on and off at a specific fre-
quency, in order for the detector and corresponding
circuitry to function properly. Typically, this is done
by passing the light through a chopper blade, which
resembles a fan blade. The blade interrupts the light,
creating a pulsing frequency. The frequency is deter-
mined by the speed of the motor and arrangement of
the chopper blade.
The advantage of the chopper is that it is simple
and can provide a high chopping frequency, one which
would otherwise be unattainable by a pulsing filament
source. This is due to the fact that the filament is a
heated wire and is limited to how fast it can be heated
up and cooled down. Although choppers do have ad-
vantages, the large size of the chopper and motor as-
sembly can be a liability for applications that require
a simple, rugged instrument for use in harsh locations.
For this reason, pulsing filament sources are used
whenever possible. The pulsating filament source typi-
cally provides a wavelength in the range of 2-5 microns,
depending on the filament temperature and bulb en-
closure material. There are many different light
sources and modulating techniques available.
3. Optical filter: There are two basic types of gas
analyzers, namely dispersive and nondispersive. The dif-
ference between the two is the way in which the spe-
64
cific wavelength of interest is extracted from the infra-

red source.
Dispersive types utilize an optical device such as a
grating or prism to spread the light spectrum over an
area containing the wavelength of interest.
Nondispersive types use discrete optical bandpass
filters, similar to sunglasses that are used for eye
protection to filter out unwanted UV radiation. This
type of configuration is commonly referred to as
nondispersive infrared (NDIR). Almost all commercial
IR instruments are of the nondispersive type. The dis-
persive type instrument is typically used only for spe-
cial requirements. The bandpass filter is one of the
most important components in designing for the type
of target gas and selectivity of an analyzer. The filters
are generally produced by a specialty optical manu-
facturer. Typical specifications for a methane filter are
shown in Figure 7.
Percent transmittance specifies the ratio of transmit-
100%
Peak Wavelength
80% Peak Transmission

Transmittance
60% Transmission Region
40% Bandwidth
Center Wavelength
20%
0%
3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.0
Wavelength (µm)
Fig. 7 Methane Filter Specifications
65
ted radiation to the incident radiation. Bandwidth is

defined as the range of wavelengths that pass through
the filter at half of the peak transmission point. Band-
width determines the selectivity of the filter, and hence
the selectivity of the instrument. Center wavelength de-
termines the gas that will be detected. There are cer-
tain designs where filters of different wavelengths are
mounted in a carousal arrangement. By exposing the
gas mixture to different filters, different gases in the
gas mixture can be identified.
4. Gas Cell/Light Path: Gas cells are often designed
in such a way as to allow the light path to interact with
the sample gas. This is normally done by using a
SELF-SAMPLING GAS CELL tube that allows light to enter from one end and
exit the other, where it meets the detector. There
Sample Gas Entrance/Exit
are “inlet” and “outlet” ports that allow the sample
gas to circulate through the tube.
The path length, or the distance in which the
IR Source light passes through the gas, is called the length of
Detector
the gas cell. This length is in direct proportion to
the amount of radiation absorbed; that is, the longer
the path length, the more radiation will be absorbed.
A longer path length results in a greater signal using
the same amount of radiation. Since the gas cells
are in direct contact with the gas sample, it is desir-
able to construct them using chemically inert mate-
rials to ensure the long-term stability and reliability
of the instrument.
Configuration
There are a number of ways by which various IR
components can be arranged to produce a gas ana-
lyzer. The design may be relatively simple, or very
complicated, using many different optical components
depending on the type of analyzers for the applica-
tions. For applications that require high sensitivity, se-
lectivity, and stability, the design of the analyzer is more
complex. For applications where the selectivity and
66
sensitivity can be compromised for reli- Sample In Sample Out

ability in a hostile, industrial environ-
ment, a simpler design can be imple-
mented.
Figures 8, 9, and 10 illustrate some of
the basic features of an IR analyzer. IR Source Filter
Figure 8 shows a basic layout of: (1) Detector
an IR source, (2) bandpass filter, and (3) Fig. 8 A Basic Infrared Gas Detector Layout
the interaction with the gas sample and
detector. Depending on the detector used, the
bandpass filter could be placed in front of the light
source, instead of placing it in front of the detector.
Figure 9 shows a similar
layout except that two de- Active
Filter/Detector
tectors are used. Modulated
MIRROR
flashing IR sources are re- IR Source
flected back to the detec-
Reference
tors. The active detector Filter/Detector
has a filter for the target
Fig. 9 A Two-Detector Layout
gas, while the reference de-
tector has a filter with a different wavelength. In other
words, the active detector is used to detect the target
gas and the reference detector is used to ignore the
target gas. In actual operation, the reference detec-
tor provides a base point value or zero point while the
active detector is used to provide the signal; with the
differential between the two detectors providing the
actual span value of the instrument.
This arrangement offers the advantage of compen-
sating for the changes that occur in the detector’s sen-
sitivity with time. For instance, the intensity of the light
source can change with time due to contamination,
which will create a zero drift. The two-detector arrange-
ment minimizes this type of drift. Also, in this arrange-
ment, the path length is doubled which leads to higher
signal strength.
Figure 10 illustrates another popular design. This
design uses two tubes or cells. One is a reference cell
67
Chopper Motor Sample In Sample Out Filter/Detector
MIRROR
SAMPLE CELL
MIRROR
REFERENCE CELL
IR Source
Fig. 10 Double Beams with Chopper Layout

that is filled with a pure target or reference gas, while
the other is a sampling cell in which the sample gas
passes through. A chopper is used in this configura-
tion, which is basically a disc with a number of slots
in it. As the chopper rotates, it alternately allows the
light beam to pass through the sample and reference
cells. The single detector gets its base reading from
the reference cell, similar to the reference detector
in Figure 9. The gas signal is acquired from the
sample cell.
Characteristics
1. Temperature: An IR detector is essentially a
temperature sensor and is, therefore, potentially very
sensitive to changes in the ambient temperature. How-
ever, a properly designed detector can be operated
between -40°C to 60°C without being susceptible to
ambient temperature fluctuations. Most detectors do
not react well to sudden temperature variations. The
instrument typically requires 10 to 20 minutes to
achieve temperature equilibrium. For outdoor appli-
cations, this usually is not a serious problem because
the ambient temperature changes only slowly. Gener-
ally, the detector unit is operated at a temperature
slightly above the surrounding temperature to prevent
condensation. Water vapor condensation at the op-
tics and/or on the detector can seriously affect the
68
performance of the analyzer.

2. Humidity: Normal environmental humidity has
very little effect. However, high humidity could promote
corrosion and contamination that lead to the failure of
the analyzer. High humidity poses an even more serious
problem in the presence of corrosive gases. The wave
path (sample chamber) can be constructed of any mate-
rial that does not absorb the IR light. The most common
materials used are stainless steel, aluminum, or copper,
plated with a corrosion-resistant coating. For some ex-
tremely “wet” applications, such as a confined space, wet
containers, or drainage sumps, the wet sample should
be “dried” before exposing it to the detector.
3. Sensitivity: The IR energy absorption is directly
proportional to the molecular structure of the hydrocar-
bon (in addition to the concentration of the hydrocar-
bon present). For example, the detector is least sensitive
to methane (CH4) with its simple, single bond. With pro-
pane (C3H8) and butane (C4H10), however, the sensitivity
increases dramatically. An example of the drastic differ-
ence between the sensitivity among varying hydrocarbons
can be seen by the fact that an IR detector can be cali-
brated to 100% pure methane, but only a few percent by
volume of propane or butane will saturate the system. In
the case of percent lower explosive limit (%LEL) or lower
flammable limit (LFL) applications, the volume concen-
tration of each gas to reach 100% LFL or LEL varies,
and the response curves are nonlinear; hence, each hy-
drocarbon must have its own curve programmed into
the system. Unlike the catalytic sensor, which has near
linear response to gases at the LFL range, the IR unit
requires some means to linearize the output.
4. Life Expectancy: The IR detector is a solid-state
device that is sealed inside a standard electronic pack-
age with a sapphire window filter. It has a long life ex-
pectancy, similar to most electronic devices. The IR light
source typically has a life expectancy on the order of 3 to
69
Signal 5 years. This lifetime can be dramatically prolonged by

operating the source at a lower-than-designed energy.
Zero Alternatively, the IR source can be designed so that it
can be easily replaced when needed.
Application
As shown in Figure 11, the IR detector responds to
Span
radiation by generating a constant signal, which is con-
sidered the “zero” point for the source. Once the zero
0 point is established and maintained, the span calibra-
Gas Concentration
tion is automatically taken care of. This is due to the
Fig. 11 Advantage of an IR
Detector. Zero point has the
fact that the absorption of radiation by the gas is always
highest output making it easy to in the same proportion, regardless of its initial source
detect any abnormal condition. intensity. Therefore, as long as the zero point is main-
tained, the accuracy of the detector remains intact. This
is one of the biggest advantages of IR technology. How-
ever, a routine calibration check is an invaluable safety
check and should not be eliminated from any periodic
maintenance.
For gas monitoring applications, the design of the
unit should be relatively compact. Sampling should be
done by diffusion. The extractive type methods that
require a pump to move the sample into the detector
are more troublesome due to the limited electrical
motor life expectancy and the maintenance required
on the pump.
The IR instruments used for this monitoring appli-
cation are typically limited to the detection of higher
concentrations (1% and above) of hydrocarbons and
carbon monoxide. Carbon dioxide absorbs infrared ra-
diation very strongly, and many monitors are available
that can detect carbon dioxide in concentration ranges
of 0.1% and higher.
With the filter used for the %LEL combustible gases
application, the center wavelength is typically at 3.4
microns. This is the wavelength of most hydrocarbons
and it is also where most of the hydrocarbon derivative
gases have a strong absorption.
70
Following is a list of common gases that are detected by

this detector:
1. Alkanes or saturated hydrocarbons such as methane,
ethane, propane, butane, pentane, hexane, and
heptane, etc.
2. Cycloalkanes such as cyclopropane, cyclohexane,
methyl cyclohexane, etc.
3. Alkenes or unsaturated hydrocarbons such as ethylene,
propylene, butene, pentene, hexene, octene, etc.
Acetylene has absorption at 3.1 microns which is not
detectable.
4. Cycloalkenes such as cyclohexene and pinene.
5. Aromatics such as benzene, toluene, and xylene.
6. Alcohols such as methanol, ethanol, propanol, and
allyl alcohol.
7. Amines such as dimethyl amine, trimethyl amine,
butanamine, cyclopropanamine, and pyridines.
8. Ethers such as dimethyl ether, ethyl ether, n-propyl
ether, methylvinyl ether, vinyl ether, ethylene oxide,
tetrahydrofuran, furan, and 1,4-dioxane.
9. Ketones such as acetone, methyl ethyl ketone,
pentanone, methyl isobutyl ketone and heptanone.
10. Aldehydes that have a central wavelength mostly at
the 3.55 micron region and generally have a weak
detection signal at 3.4 microns.
Carbon dioxide at 4.3 microns and carbon monoxide
at 4.6 microns have very little interference by other gases.
Summary
Detection of hydrocarbons using IR has been available
for a few years. However, because of the many design op-
tions, the features and implementation of the technology
vary greatly from one manufacturer to another. Neverthe-
less, IR detection has been well received by many indus-
tries, including the petrochemical industry.
For the purpose of area air quality and safety applica-
71
tions, some comparisons between IR, solid-state, and cata-

lytic bead sensors, used for the detection of combustible
gases are given below.
1. Poisoning: This is the main problem with catalytic
sensors. Various chemical compounds, such as hydrogen
sulfide, silicon compounds, and chlorinated or fluorinated
compounds, among others, can poison the catalyst in the
sensor and cause the sensor to lose sensitivity. IR detec-
tors do not suffer from this problem.
2. Burn Out: Catalytic senors will burn out if exposed
to high gas concentrations. Again, IR detectors do not
have this problem.
3. Life Expectancy: Catalytic sensors have a life expect-
ancy of about 1-2 years, while solid-state sensors typically
last more than 10 years. A well-designed IR unit also has a
life expectancy of more than 10 years.
4. Calibration: Periodic calibration must be done on
all sensor types. However, on IR units, as long as the zero
is maintained, the IR unit is assured a good response and
good span accuracy. Because of this characteristic, ab-
normal functioning of an IR unit can be easily determined.
5. Continuous Exposure to Gas: In applications that re-
quire the detector to be constantly exposed to a gas stream
to monitor for hydrocarbons, catalytic and solid-state sen-
sors will experience a shortened life span. Continuous ex-
posure to gas ultimately changes the characteristic of the
sensor and leads to permanent damage. However, with
IR instruments, the functional components are protected
by the optical parts, which are basically inert to most
chemicals. Only the IR radiation interacts with the gas;
therefore, as long as gas samples are dry and noncorro-
sive, IR instruments can be used to monitor a gas stream
continuously over a long period of time.
By selecting one of the three available sensor tech-
nologies (IR, solid-state, and catalytic), one can tailor a
solution to almost any hydrocarbon detection application.
72
Chapter 6 Photoionization Detectors
Chapter 6
Photoionization
Detectors
T
he photoionization detector (PID) utilizes ultra-
violet light to ionize gas molecules, and is com-
monly employed in the detection of volatile or-
ganic compounds (VOCs). This technique originally
found use in bench top laboratory instruments, but
its complexity limited its use elsewhere. The heart of
the photoionization detector is an ultraviolet source, Fig. 1 A Pocket PID Monitor
which is essentially a lamp. Early versions of this lamp
used electrodes inside the lamp similar to those used
in the early days of the vacuum tube and were quite
costly to manufacture.
The lamps used today do not contain electrodes
and are both less costly and have a longer life expect-
ancy. In the 1980s, with the advent of integrated cir-
cuit (IC) technology, the electronics that were de-
signed were better able to process the small signal from
PID sensors into useful and reliable data.
Additionally, requirements for the monitoring
of underground storage tanks to help prevent
ground water contamination began to emerge,
which required monitoring of VOCs.
These events led to the design of small, portable
PIDs (see Figure 1) that have proven to be both prac-
tical and reliable, and which offer fast response and
the ability to detect low gas concentrations. To this
73
day, PID sensors are the preferred choice for the de-
tection of VOCs.
Ultraviolet (UV) light is the radiation group that
has frequencies that are higher and are directly above
visible light in the electromagnetic radiation spectrum.
UV wavelengths, being in the nanometer (ηm = 10-9
meters) region, are much shorter than IR wavelengths.
The wavelengths of infrared light used for gas analy-
sis are in the micron region, 10- 6 meters.
Because shorter wavelengths have higher frequen-
cies, and hence more energy, UV radiation has more
energy than IR radiation. UV radiation energy levels
are commonly described in terms of electron volts, or
1
eV = energy gained when a particle of eV.1 Wavelength is related to eV through Planck’s Con-
one electronic charge is accelerated stant2 which is 4.135 x 10-5 eV. [An electron volt equals
through a potential of one volt. 1.2395 x 10-6/wavelength (ηm). The term eV is used
for convenience, to give a simple numerical expres-
2
Planck’s Constant is a fundamental con-
stant in quantum physics. The radiant
sion of the radiant strength.]
energy only changes in “quanta” or The PID Lamp. The heart of the detector is the
discrete steps. PID lamp which has an exterior shape resembling a
medical ampule and comes in different sizes and di-
mensions, depending on the manufacturer. The lamp
is filled with a low-pressure inert gas. When this gas is
energized with energy in resonance with the natural
frequency of the gas molecules, an ultraviolet spec-
tral radiation is produced.
The wavelength of the UV light emitted depends
on the type of gas in the lamp. For instance, krypton,
when excited, will emit 123.9 ηm and 116.9 ηm radia-
tion, or the equivalent of 10 eV and 10.6 eV. This 10.6
eV lamp happens to be the most popular lamp cur-
rently used in PID instruments. Various other gases
are also used. Figure 2 on the following page shows
wavelengths emitted by argon, krypton, and xenon.
Many different techniques are employed to power
the PID lamp. A high voltage is required for the lamp
74
XENON KRYPTON ARGON

147.6ηm 123.9ηm
105.9ηm
116.9ηm
129.1ηm
8.4eV 9.6eV 10.0eV 10.6eV 11.7eV
Fig. 2 The wavelengths (ηm) emitted by argon, krypton, and xenon.
to function. The source needs to remain stable over

time and also be efficient, so as not to consume too
much power. A typical design would place electrodes
inside the lamp. These types of lamps look similar to
the vacuum tubes used in the early days of radio and
TV. There are electrical terminations on the tube, and
the gas discharge is confined to a small capillary. While
the voltage used is in excess of 1000 volts, the current
is very low. Although this type of lamp is used in many
laboratories, it is quite expensive. As a result, a more
popular lamp in use today is the electrodeless lamp.
The Electrodeless Lamp. This lamp is only filled
with a low-pressure inert gas and there are no elec-
trodes inside the lamp. The simplicity of this design
allows the lamp to be miniaturized, making the de-
sign of the portable instrumentation very compact.
In the electrodeless lamp, the low-pressure inert
gas is separated by the lamp wall from the outside
world. Because there are no electrodes that provide
power directly to the gas, the only way to excite the
75
gas molecules is to use radiation energy that can pen-

etrate the lamp wall.
There are several different ways this can be accom-
plished. One method is to use electromagnetic radia-
tion. A more popular technique, however, is to place
a pair of electrodes on the exterior wall of the lamp. A
high-voltage, low-current charge is applied to the elec-
trodes, and the energy applied is sufficient to excite
the low-pressure gas inside the lamp. The design is
similar to that of a fluorescent light used in a home or
office, except on a much smaller scale.
Special Window Materials. At the wavelength of
radiation, meaning the wave frequency of UV light,
most materials will absorb the radiation and prevent
the transmission of such radiation.
Therefore, special window materials are placed at
the discharge end of the lamp which allow spectral
emissions to pass through. This window material is a
crystal that allows good transmission of the targeted
UV wavelength. For instance, 10.6 eV lamps commonly
use krypton gas and magnesium fluoride windows,
while 11.7 eV lamps, which have more energy, use ar-
gon gas and lithium fluoride windows. These windows
are soft glass and are very fragile. They are quite ex-
pensive, and require special care and handling. Fig-
ure 3 illustrates typical PID design.
Gas Molecules
Signal i
+
Insulation V
-
Electrode
Window
Electrode Plate
for Lamp Illumination
Lamp
Fig. 3 A Typical Photoionization Detector Configuration
76
A pair of electrode plates are placed in close prox-

imity to the lamp window where the light is emitted.
The electrodes are biased with a stable DC voltage
which will generate signals in case any small changes
in the electrical field occur. As gas molecules move
into the radiated field in the space between the elec-
trodes, they are ionized and the free electrons are
collected at the electrodes resulting in a current flow
whose magnitude is directly proportional to the gas
concentration.
Distinct Ionization Potentials. Each gas has its
own unique ionization potential (IP). Gases with IP val-
ues below the eV output of the lamp will be detected.
For most portable instruments, the 10.6 eV lamp is
most widely used because it detects most volatile or-
ganic compounds, and the lamp is easy to clean. How-
ever, other lamps are also available, including 10 eV,
9.5 eV, 8.4 eV, and 11.7 eV lamps. Most manufactur-
ers design instruments with a lamp chamber that al-
lows easy interchange of lamps. The ionization po-
tentials for many gases are included in Appendix II at
the end of this book.
Following are examples of gases detectable using
different lamps:
1. Gases detected by 9.5 eV lamp such as benzene,
aromatic compounds, amines.
2. Gases detected between 9.5 eV and 10.6 eV such as
ammonia, ethanol, acetone.
3. Gases detected between 10.6 eV and 11.7 eV such as
acetylene, formaldehyde, methanol.
A 9.5 eV lamp will ionize gases with ionization
potentials below 9.5 eV and will not ionize gases with
higher ionization potentials. On the other hand, if
one uses an 11.7 eV lamp, gases with ionization po-
tentials up to 11.7 eV will be ionized. Interchanging
lamps in an unknown air sample helps to separate it
into groups of gases.
77
Characteristics
PID sensors offer fast response for detection of
many volatile organic compounds. They will respond
to all gases that have ionization potentials equal to or
less than the eV output of their lamps.
Correction Factor. PIDs are typically calibrated
with isobutylene, a stable gas with a slightly pungent
odor. This gas is easy to handle and can be stored at
high pressure, allowing calibration bottles to provide
many calibrations. PID instruments typically detect
gases at low concentrations, and most of these gases
are normally liquid solvents or other gases that are
not easy to calibrate. Thus, it is much easier to cali-
brate these instruments using isobutylene as the cali-
bration gas. Readings for other gases are obtained by
multiplying the reading by a correction factor. For ex-
ample, benzene has a correction factor of 0.5, which
means that a reading of 100 ppm isobutylene on the
sensor indicates a concentration of 50 ppm benzene.
For ammonia, which has a response factor of 10, a
reading of 100 ppm on the sensor would indicate an
ammonia concentration of 1000 ppm. Instrument
manufacturers typically supply a list of correction fac-
tors with their product.
An important point to remember is that correc-
tion factors are not absolute, and data can be slightly
different from manufacturer to manufacturer. In fact,
these factors can even vary somewhat from lamp to
lamp, and the results may vary depending on the qual-
ity of the UV output of the lamp at the time of mea-
surement. Thus, for the most accurate readings on
specific gases, it is necessary to individually calibrate
the gas of interest.
PID instruments are very sensitive to the composi-
tion of gases in the sample. The results will differ vastly
if one uses gas with dry nitrogen as the balance to
calibrate, from those results obtained when one uses
air as the balance. Therefore, it is important to use
78
isobutylene in air as the calibration gas when the instru-

ment is used for area air quality and safety applications.
Still, there will be some error because the calibration
mixture is under high pressure and is dry; therefore, it
does not exactly duplicate environmental air.
Zeroing Instruments for Best Results. Generally,
high humidity decreases the response by up to 30%,
when compared to dry air. In dry nitrogen, readings
are typically 10-30% higher than in dry air.
If the sample gas constituents include gas compo-
nents that are ionizable but the ionization potentials
are above the detecting lamp output potential, and al-
though the lamp radiant energy will not ionize these
gas molecules, the UV rays can be scattered and ab-
sorbed, resulting in a lower output reading. This is
known as the “quenching effect.” For example, in envi-
ronmental air, such “quenching” gases are water vapor,
carbon dioxide, methane, carbon monoxide, etc. This
is another reason why a representative air is needed to
zero the sensor and prepare the calibration gas mix-
ture. The common practice of calibrating the instru-
ments with isobutylene in an air mixture, without mak-
ing corrections, results in an inaccurate reading.
PID sensors will exhibit different zero readings
when exposed to nitrogen, dry, clean air, and environ-
mentally clean air. Thus, it is best to zero the sensor in the
same type of conditions that are present in the application in
which it will be used. Clean, ambient air is, therefore,
the best choice for zero air
5
as well as for mixing cali-
bration gas. 4
Output Voltage
PID Response Charac- 3

teristics. The output of the 2
PID sensor is relatively lin-
1
ear below 200 ppm, and the
0
detector output becomes 0 50 100 150 200
saturated above 2000 ppm. Concentration Isobutylene
Figures 4, 5, and 6 illustrate Fig. 4 Zero-to-200 ppm Curve
79
5 the response curves for 200

4 ppm, 2000 ppm, and 4000
Output Voltage
3 ppm isobutylene in room

air.
2
You will notice that the
1
200 ppm curve is relatively
0 linear, while on the 2000
0 500 1000 1500 2000
ppm curve, 1000 ppm pro-
Concentration Isobutylene
Fig. 5 Zero-to-2000 ppm Curve
duces an output of 3.6 volts
and 2000 ppm produces an
output of 4.8 volts. The additional 1000
500
ppm produces only 1.2 volts, or one third
400 of the initial 1000 ppm. From the 4000 ppm
Output Voltage (mV)
curve, you can see that the detector be-

300 comes saturated.
200
Thus, this output region does not pro-
vide a reading with a good resolution.
100
Applications
0
0 1000 2000 3000 4000 PID instruments offer ver y fast re-
Isobutylene Concentration (PPM) sponse, high accuracy, and good sensitivity
Fig. 6 Zero-to-4000 ppm Curve for detection of low ppm volatile organic
compounds (VOCs). The major shortcom-
ing of these instruments is that the PID lamp requires
frequent cleaning. Because the lamp window is directly
exposed to the sample stream, the condition of the
window is very critical to accurate readings, and a dirty
window will produce much different results than a
clean window. The frequency of cleaning needed for
the window depends on the sample stream conditions.
Because PID sensors require periodic cleaning and
have limited life expectancies, they are not practical
choices for use in stationary monitors, which sample
continuously. Their use is limited to portable models
in which only periodic readings are required.
Lamp Life Expectancy. The life expectancy of the
lamp depends on the type of lamp. Generally, 10.6 eV
lamps have the longest life expectancy, approximately
80
6000 hours.
Typical Specifications
The 11.7 eV lamps have lithium fluoride win-
for PID Instruments
dows that transmit the shorter UV wavelength light
at 11.7 eV, which allows for the detection of many Detector: 10.6 eV electrodeless discharge
lamp. 11.7 eV lamp optional.
more gases when compared to lower eV lamps.
However, the intensity of the light they emit is Response Time: 3 secs. to 90% reading.
weaker than the light emitted from lower eV Ranges: 100 ppm, 1000 ppm, and 2000
lamps. The weaker energy emitted means less sig- ppm isobutylene equivalent, auto-ranging.
nal, which results in less stable and more tempera- Sensitivity: 0.1 ppm isobutylene.
ture-sensitive instruments. Thus, 11.7 eV lamps are Accuracy: 1-10%, depending on range.
not typically used for general applications. Further-
Temperature Range: –20oC to +50oC.
more, lithium fluoride is hygroscopic, which means
that lithium fluoride crystals attract and absorb Humidity: 0–95% relative humidity,
moisture from the air, causing the window to de- noncondensing.
grade. They also cannot be exposed to cleaning sol-
vents because most solvents contain small amounts
of water. Therefore, a special, fine solid aluminum
oxide powder is needed to clean the lamp.
81
Sensor Selection Overview

Application Requirement
• Safety Monitoring (LEL) • Exposure Assessment (TWA)
• Toxic Limit Detection (PEL) • Ambient Air Quality Compliance Monitoring
• Leak Detection
• Personal Safety
LIMIT DETECTION ALARM QUALITATIVE & QUANTITATIVE ANALYSIS
Monitors Analyzers
Monitor Type: Target Gas: Analyzer Type:
Electrochemical O3 UV Photometers
Toxic
Gases
Solid-State SO2 UV Fluorescence
NOx Chemiluminescence
Catalytic Air
Quality THC*
Monitor Flame Ionization
Combustible Solid-State
Gases
Single CO Infrared
Infrared Gas
(Online)
CO2
Continuous
Volatile Emission SO2
Organic Photoionization Monitor
Compounds
O2 Paramagnetic;
Zirconium Oxide;
Electrochemical
Total
Hydro- Flame Ionization Most UV/IR
Carbons UV/IR Spectrometer
Absorbing Gases
Multi- Most IR
Gas Absorbing Gases Fourier Transform IR
Thermal (Laboratory)
Other Conductivity
Gases Most Gases Mass Spectrometer
Colorimetric
Gas Chromatograph
*Total hydrocarbon
Chapter 8 Sensor Selection Guide
Chapter 8
Sensor Selection Guide
E
ach of the following sensors—electrochemical,
catalytic bead, solid state, infrared and photo-
ionization detectors—must meet certain criteria
to be practical for use in area air quality and safety
applications. Some of the basic requirements are:
1. The sensor should be designed for a housing
that is small and rugged. The sensor should be suit-
able for use in hazardous locations and harsh envi-
ronments, and, it should also be explosion-proof. The
sensor should be cost-effective, designed for installa-
tion and use in industrial production areas, and
installable at a reasonable cost.
2. For portable applications, the instruments
should have reasonable energy consumption and the op-
tion of powering the instruments with batteries should
be easily available. The instruments should be small
and portable so they can be carried easily. They should
be safe for use in industrial environments. Preferably,
the instruments should be certified as intrinsically safe
for use in a hazardous area.
3. The operation and maintenance of the instru-
ments should be easily performed by regular plant
personnel with minimal special training require-
ments.
4. In stationary installations, the sensors should
be able to function continuously and reliably for a period
of time, preferably longer than 30 days. The sensor
103
should be able to function in an industrial environ-

ment for at least two years or longer and should be
replaceable or renewable at a reasonable cost. It
should be easy to install into a multi-point system and
be managed by a controller or a computer controlled
distribution system.
5. The cost of the instruments should be reason-
able so that multiple sensors can be installed to effec-
tively protect the area.
Four of the five sensors discussed in this book all
meet the above criteria. The exception is the photo-
ionization detector. The PID is a good detector for
portable applications but is limited by the lamp be-
cause it has a relatively short life expectancy and the
frequency of maintenance required may not be prac-
tical for stationary applications. However, there are
PID stationary instruments available that can be use-
ful as long as users are aware of the limitation.
There are other types of sensors which meet the
above criteria, but most have limitations. For ex-
ample, thermal conductivity sensors are mostly used
for high concentration applications and are not widely
used as gas monitors.
Factors to Consider When Selecting Sensors

One of the most frequently asked questions regard-
ing sensors is: “Which sensor is the best?” Of course,
there is no simple answer to this question. Each sen-
sor has certain capabilities and limitations, and thus
the suitability of a given sensor depends largely on
the application in which it is to be used. Thus, to
choose the correct sensor, one must first properly de-
fine the application. The illustration on page 102
shows an overview of various application requirements
and their detection technologies. It is common for
manufacturers to exaggerate the capabilities of the
sensors that they offer and downplay sensors that they
do not offer.
104
In determining which sensor to use for a given ap-

plication, the following factors/observations should
be considered:
A. Realistically define what objective one is trying to ac-
complish and define an instrument specification
that meets the minimum requirements. The
specifications should define the gases and
ranges of the sensors. The ranges or the con-
centration of the gases to be measured should
be 3 to 5 times the actual monitoring concen-
tration. As with a voltmeter, one should always
select a range higher than the actual voltage
to be measured. For example, select a 50-volt
range to measure a 12-volt battery.
B. Determine the background gases in the moni-
toring area. In cases where the background
gases cannot be determined, a representative
sample should be analyzed. A major cause of
sensor failure is the presence of background
gases that the instrument’s manufacturers did
not take into consideration. The selectivity or
specificity of the sensor must be acceptable for
the application.
C. The temperature ranges in which the sensor is
to be installed should be within the sensor
specifications and should be suitable for the
gases to be monitored. For example, jet en-
gine fuels have very low vapor pressure. It is
useless to install a sensor to measure the com-
bustible range in a hangar if the temperature
will never exceed 100° F because the vapor
concentration cannot reach combustible lev-
els. In this example, it is more appropriate to
measure in ppm ranges.
The temperature changes between day and
night, and during summer and winter, should
also be considered. A wide temperature change
105
can cause moisture condensation. This is par-

ticularly important in a confined space such as
a closed container where air circulation is poor.
D. A typical specification for humidity is 95% non-

condensating. The occurrence of condensation
is a function of temperature change, as seen
on wet windows and car windshields in the
morning. Normally, there is no problem in
normal industrial background environments,
even during the hot summer months in coastal
areas such as the Gulf of Mexico, as long as the
air circulation is normal. Areas with poor air
circulation can often be the cause of conden-
sation.
Both solid-state and catalytic sensors have
heated elements. In addition, their transmit-
ters are designed to operate at 14-24 VDC,
which generates heat. Therefore, the sensor
transmitters are always a few degrees warmer
than the environment in order to minimize
the possibility of condensation. Electrochemi-
cal sensors normally require relatively much
less power; therefore, the temperature of their
transmitters is similar to the surrounding tem-
perature. In this case, it is easier for conden-
sation to occur.
E. In applications requiring the sensors to be con-

stantly exposed to gas, special considerations are
required and the sensor specifications and the
suppliers of the sensors may need to be con-
sulted. A properly designed sampling system
may make difficult or otherwise impossible ap-
plications possible to handle.
It is difficult to mention every consideration
needed, but a carefully evaluated and studied appli-
cation can yield savings in both time and money.
106
With the specifications of the sensors, one can

decide which sensor best meets one’s requirements.
There is no general consensus that establishes which
sensor is the best for a given application. Hence, the
information that follows contains some guidelines
that may be helpful in making the proper selection
of a sensor.
Toxic versus Combustible Gas Monitoring

Gas monitoring applications are generally classi-
fied into toxic or combustible range monitoring. Toxic
gas monitors are generally used for human health
protection and the ranges of the monitors are 3 to 5
times higher than the permissible exposure limits. For
most gases, ranges are in ppm concentrations.
For combustible applications, ranges are typically
100% lower flammable limits or a fraction of these
ranges, such as 50% LFL. The gas concentrations are
high and are generally in the range of several percent.
In other words, for toxic gas applications, a sensor
must be able to measure gases at low concentrations
while, for combustible gas monitoring, a sensor must
measure high gas concentrations.
Summary
Electrochemical Sensors. Except for oxygen appli-
cations, electrochemical cell sensors are designed to
be used as toxic gas monitors. These sensors are only
suitable for low concentration ppm ranges. For por-
table applications, the electrochemical sensor has
many advantages: it has very low power consumption,
responds quickly to gas, and is not affected by humid-
ity. Also, the sensors are only exposed to gas periodi-
cally, which maximizes the sensor life.
Electrochemical sensors are therefore a good
choice for portable instruments. Electrochemical sen-
sor life expectancy is two years; however, depending
on the application, it can be much shorter. The cost
107
of replacement sensors is high, especially when the

number of instruments in use is large. The annual
budget and labor to keep the instruments function-
ing need to be considered.
There are approximately 20 gases that can be
monitored by electrochemical sensors. For the rest
of the gases (for ppm ranges), solid-state sensors or
PIDs will need to be used.
Catalytic Sensors. In portable combustible gases
in the LEL ranges, a catalytic sensor is good for nor-
mal, simple applications. The sensors can last for a
long time because they are used sporadically in por-
table applications. Catalytic sensors are relatively in-
expensive but one has to make sure they are made by
reputable suppliers.
Infrared and Solid-State Sensors. For gases that
can poison the catalytic sensors and make them im-
possible to use, the choice is between infrared sensors
and solid-state sensors. Depending on the gases to be
detected, infrared sensors have the better perfor-
mance but detection of gases is limited. On the other
hand, solid-state sensors can detect most chemicals in
the LEL ranges.
In stationary applications, the sensors are con-
stantly exposed to environmental background gases.
For toxic gas applications, it is generally favorable to
use solid-state sensors, especially when the number of
sensors is sizable. In applications where interference
can be a problem, it is best to study the sensor specifi-
cations and to consult with the supplier.
For stationary combustible gas applications, the
choice is among catalytic sensors, solid-states sensors
and infrared sensors, which are fully described in
Chapters 4, 5, and 6 respectively.
There are no standardized specifications for a gas
monitor. The following table entitled “Typical Speci-
fications for Gas Monitoring Instruments” is a sum-
mary of several “requests for quotation” from cus-
108
tomers and is presented here as a reference.
Typical Specifications for Gas Monitoring Instruments
Remote Sensor Transmitter Specifications

Gas to Be Monitored: Methyl Bromide, 0–500 ppm.
Sensor: Solid-state sensor
Temperature: –10o C to 45o C.
Humidity: 95% RH, noncondensing.
Accuracy: 5%
Response Time: 50 seconds to 90% full scale
Background: Normal environmental conditions
Transmitter shall be certified for use in Class 1, Div. 1, Groups B, C, D areas. Provides
functional adjustments without the need to open the transmitter’s cover, and pro-
vides protection for outdoor applications.
Power: 14 VDC to 24 VDC
Output: 4–20 mA linear output.
Controller Specifications
1. Four-channel control unit, microprocessor-based.
2. Independent digital display of reading for each channel. Provides calibration func-
tions, three alarm settings, and diagnosis functions.
3. Three-level alarm set points (low, mid, high) with LED indicator for each alarm.
4. Relays: SPDT, 5 amp. 220 VAC max. resistive
Common relays for each alarm level (3 relays per channel)
Options: Individual relays for each alarm level ( 3 relays per channel)
“Fault” relays for each channel, indicating electrical fault of the sensor
5. 4-20 mA analog output signal for each channel.
6. Wall-mounted weatherproof NEMA 4X enclosure.
7. Provide 24 VDC, 500 mA for each sensor module, accepts and processes 4-20 mA
transmitter output signal.
109
Chapter 11 Gas Sensor Calibration
Chapter 11
Gas Sensor Calibration
G
as sensors need to be calibrated and periodi-
cally checked to ensure sensor accuracy and
system integrity. It is important to install station-
ary sensors in locations where the calibration can be
performed easily. The intervals between calibration
can be different from sensor to sensor. Generally, the
manufacturer of the sensor will recommend a time
interval between calibration. However, it is good gen-
eral practice to check the sensor more closely during
the first 30 days after installation. During this period,
it is possible to observe how well the sensor is adapt-
ing to its new environment.
Also, factors that were not accounted for in the
design of the system might surface and can affect the
sensor’s performance. If the sensor functions prop-
erly for 30 continuous days, this provides a good de-
gree of confidence about the installation. Any possible
problems can be identified and corrected during this
time. Experience indicates that a sensor surviving 30
days after the initial installation will have a good
chance of performing its function for the duration
expected. Most problems—such as an inappropriate
sensor location, interference from other gases, or the
loss of sensitivity—will surface during this time.
During the first 30 days, the sensor should be
checked weekly. Afterward, a maintenance schedule,
161
including calibration intervals, should be established.

Normally, a monthly calibration is adequate to ensure
the effectiveness and sensibility of each sensor; this
monthly check will also afford you the opportunity to
maintain the system’s accuracy.
The method and procedure for calibrating the
sensors should be established immediately. The cali-
bration procedure should be simple, straightforward,
and easily executed by regular personnel. Calibra-
tion here is simply a safety check, unlike laboratory
analyzers that require a high degree of accuracy. For
area air quality and safety gas monitors, the require-
ments need to be simple, repeatable, and economi-
cal. The procedure should be consistent and trace-
able. The calibration will be performed in the field
where sensors are installed so it can occur in any type
environment.
Calibration of the gas sensor involves two steps.
First the “zero” must be set and then the “span” must
be calibrated.
Step One: Setting the “Zero” Reading

There is no established standard that defines zero
air. Many analytical procedures, including some spe-
cific analyzer procedures such as EPA methods, use
pure nitrogen or pure synthetic air to establish the
zero point. The reason for this is that bottled nitro-
gen and pure synthetic air are readily available. As a
result, it is popularly believed that using bottled nitro-
gen or synthetic air is a good method to zero a sensor.
Unfortunately, this is not correct. Normal air con-
tains traces of different gases besides nitrogen and
oxygen. Also, ambient air normally contains a small
percentage of water vapor. Therefore, it is much more
realistic and practical to zero the sensor using the
air surrounding the sensor when the area is consid-
ered to be clean. This reference point can be diffi-
cult to establish. Therefore, a good reference point
162
can be in the area where air is always considered

clean, such as in an office area. This will give a more
realistic representation of the zero point because it
will be representative of the local ambient air condi-
tion. The lack of water vapor can cause the zero point
setting to read lower than in ambient air making the
sensor zero appear to drift. This is most noticeable
in solid-state sensors and PIDs.
Calibration Methods. Taking all factors such as the
type of sensor and the conditions of the application
into consideration, the following are some proposed
methods of calibration: H2O
A. In applications where the ambient air is nor-
mally clean, and, based on the operator’s judgment O2
that no abnormal condition exists and the instrument CO2 &

is indicating a close to zero reading, the procedure to N2 others
zero the sensor can be skipped. When in doubt, use a
plastic bag to get a sample of what is considered to be
“clean air” in the facility and expose it to the sensor
for a few minutes. This is a very quick and easy proce- Ambient air is the best zero air.
dure. It is also a very effective way to differentiate a
real alarm from a false alarm.
B. Compressed air has the advantage that it is easy
to regulate and can be carried around in a bottle. Also,
in many facilities, shop air is available throughout the
plant, making it very accessible and convenient. How-
ever, most shop air contains small concentrations of
hydrocarbons, carbon monoxide, carbon dioxide, and
possibly other interference gases. Also, the air is typi-
cally very low in humidity. A solution to this is that the
air can be filtered through activated charcoal to re-
move most of the unwanted gases and water vapor can
be added into the air using a humidifier in the sam-
pling system. After this conditioning, the air can be
used to calibrate most types of sensors. However, it is
important to note that carbon monoxide is not re-
moved by charcoal filters.
It is therefore imperative to make sure that the
163
CO concentration in the shop air is the same as in the

ambient air. Furthermore, a soda ash filter should be
used to remove carbon dioxide. This is also a very good
way to zero carbon dioxide sensors since placing a soda
ash filter in-line with the sampling system will remove
all carbon dioxide, thus providing an easily obtain-
able zero baseline.
Although synthetic air is usually very pure, it can-
not be used with solid-state sensors or PID sensors
because these sensors require some water vapor in the
sample stream. A simple solution to this problem is to
add a wet tissue paper in the sample line. This acts as
a humidifier in the sample stream and provides
enough water vapor for the sensor to read properly.
Another option is to use a Nafion tube, which is de-
scribed thoroughly in Chapter 10, “Sampling Systems
and Designs.” Figure 1 illustrates this concept.
Calibration Gas
Flow Nafion
Meter Dryer Tube or To Sensor
Wetting Material
Regulator Fig. 1 Adding Moisture to Calibration Gases
Step Two: Span Calibration

The span calibration can be quite easy or it can be
very complicated and expensive, depending on the gas
type and concentration range. In principle, to achieve
the best accuracy, a mixture of the target gas balanced in the
background environmental air is the best calibration gas.
However, although this can be done, it usually requires
that the operators be more skilled than would usually
be required. In practice, most calibration gases are pur-
chased from commercial suppliers. The following sec-
tion describes a few methods of span calibration.
A. Premixed Calibration Gas
164
This is the preferred and most popular way to cali-

brate gas sensors. Premixed gas mixtures are com-
pressed and stored under pressure in a gas bottle. The
bottles are available in many sizes but most field cali-
brators employ smaller, lightweight bottles. These
small portable bottles come in two different catego-
ries: a low-pressure and a high-pressure version.
The low-pressure bottles are thin-walled, light-
weight bottles that are usually nonreturnable and dis-
posable. High-pressure bottles are designed to bottle
pure hazardous chemicals. For calibration gases, these
bottles are normally made of thick-walled aluminum
which has a service pressure of 2000 psi.
To get this highly pressurized gas out of the bottle
in order to calibrate the sensor, a regulator assembly
is needed. This assembly consists of a pressure regu-
lator, a pressure gauge, and an orifice flow restrictor.
The orifice flow restrictor is a fitting with a hairline
hole that allows a constant air flow at a given pres-
sure difference. In operation, the high pressure from
the bottle is reduced to a lower pressure of only a few
psi, which provides a constant air flow through the
orifice. Flow rates between 600-1000 cc/
min are most common. Models can be
fitted with an adjustable pressure regu-
lator so that the flow rate can be adjusted
accordingly. Figure 2 illustrates typical
models of high- and low-pressure bottle
assembly.
Many gases can be premixed with
air and stored under pressure, but some
gases can only be mixed in inert gas
backgrounds, such as nitrogen. Some
mixtures can only be stored in bottles
that are specially treated or condi-
tioned. Each type of mixture will have
a different amount of time before it ex- Low-Pressure Assembly High-Pressure Assembly
pires or before it can no longer be used. Fig. 2 Calibration Gas Bottles
165
Detailed information about storage and shelf life can

be obtained from the manufacturer. Generally, high
vapor pressure gases with low reactivity, such as meth-
ane, carbon monoxide and carbon dioxide, can be
mixed with air and stored under high pressure. Low
vapor pressure gases, such as liquid hydrocarbon sol-
vents, can only be mixed with air and stored under
low pressure. Most highly reactive chemicals are mixed
with a nitrogen background. With certain sensors, such
as solid-state sensors, whether the mixture of the gas
is in the air or in the nitrogen background will dra-
matically affect the sensor reading.
During calibration, some sensors may need mois-
ture to get a proper reading. Moisture can be added by
following the same procedure described in Step 1 for
zeroing the sensor.
To estimate the volume of a pressurized gas in a
cylinder, take the total pressure (P) divided by the at-
mospheric pressure (Pa) and multiply this ratio by the
volume of the cylinder:
Vmix = V . (P/Pa)
where
Vmix = the volume of the gas mixture
V = the volume of the cylinder
P = the pressure in the cylinder
Pa = the atmospheric pressure
For example, let’s say a given lecture bottle has a 440

cc volume (V). Assume the bottle has a 1200 psi pres-
sure. The estimated volume of the premixed gas at atmo-
spheric pressure is: (440 cc) x (1200/14.7) = 35,918 cc.
If the flow rate of the calibration gas is 1000 cc per
minute and it takes approximately one minute per sen-
sor to calibrate, a single cylinder can be used to cali-
brate approximately 30 times.
166
B. Permeation Devices
A permeation device is a sealed container that con-
tains chemicals in liquid and vapor phase equilib-
rium. The gas molecules are either permeated
through the permeable container wall or through
the end cap. The rate in which the gas molecules
permeate depends on the permeability of the mate-
rial and temperature. The rate of permeation is con-
stant over long periods of time. At a known rate of
permeation at a given temperature, a constant flow
rate of air mixed with the permeated chemicals forms
a constant stream of calibration gas. A calibrator with Examples of Permeation Tubes
constant temperature and flow regulation is needed.
However, the permeation tube continuously emits
chemicals at a constant rate thus creating a storage
and safety problem. Also, the rate of permeation for
a given gas of interest can be too high or too low for
a given application. For example, high vapor pres-
sure gases permeate too quickly while very low vapor
pressure chemicals have a permeation rate that is too
low to be of any use.
Permeation devices find most of their use in labo-
ratories and in applications using analytical analyzers.
For gas monitoring applications, the concentrations
needed to calibrate the sensor are typically too high
for the permeation device. Therefore, they have been
found to be of limited use.
C. Cross Calibration
Cross calibration takes advantage of the fact that
every sensor is subject to interference by other gases.
For example, for a sensor calibrated to 100% LEL hex-
ane, it is usually much easier to use 50% LEL meth-
ane gas to calibrate the sensor instead of using an ac-
tual hexane mixture. This is because hexane is a liq-
uid at room temperature and it has a low vapor pres-
sure. Therefore, it is more difficult to make an accu-
rate mixture and to keep it under high pressure.
167
On the other hand, methane has a very high va-

por pressure and is very stable. Furthermore, it can
be mixed with air and still be kept under high pres-
sure. It can be used for many more calibrations than a
hexane mixture in the same size bottle and it has a
long shelf life. A 50% LEL methane mixture is also
readily available. Therefore, it is common practice for
manufacturers of combustible gas instruments to rec-
ommend the use of methane as a substitute to cali-
brate for other gases.
There are two ways to accomplished this task. The
first method is to calibrate the instrument to meth-
ane while other gas readings are obtained by multi-
plying the methane reading by response factors that
are included in the manual. This is commonly done
with catalytic sensors. Catalytic sensors have a linear
output and therefore the use of this response factor is
applicable to the full-scale range. For example, pen-
tane has an output of only half that of methane gas
when the sensor is calibrated to methane. Therefore,
it has a response factor of 0.5. So, if the instrument is
calibrated to methane but is used to measure pentane,
the reading is multiplied by 0.5 to obtain the pentane
reading.
The second method is to still use methane as the
calibration gas, but double the value of the reading of
the calibration. For instance, use 50% LEL methane
calibration gas and calibrate with this as 100% LEL
pentane. After the calibration, the instrument directly
indicates the pentane gas concentration although it
was calibrated using methane gas.
Many low-range toxic gas sensors can be calibrated
using cross gas calibration. Also, with infrared instru-
ments, any gas within the same wavelength of absorp-
tion can be used for cross calibration. The advantage
of cross calibration is that it allows the sensor to be
calibrated with a gas and range that is easier to obtain
and handle.
168
However, there are some problems with using cross

calibration. One is that the response factors for each
sensor can be different as it is generally impossible to
make most sensors exactly alike. For example, in cata-
lytic sensors, the heater voltage has to be as specified
in the manual; otherwise, the response factor will not
be applicable. The response characteristics will vary
with different heater voltage settings. Therefore, it is
a good practice to periodically check the calibration
of the sensor with the actual target gas.
Mixtures of stable noncombustible and nontoxic
gases with various concentrations are available from
many supply sources. Check with the instrument
manufacturer for more detailed information.
D. Gas Mixing
Not all calibration gases are available. Even if they
are available, it is very possible that they would not be
available in the right concentration or in the proper
background mixture. However, many mixtures are
available for some process uses which can be diluted
to use in calibration of gas monitors in lower concen-
tration ranges. For example, 50% LEL methane has
a concentration of 2.5% or 25,000 ppm. To make a
20% LEL mixture having a volume of 2000 cc, the
following formula can be used:
C . C – Cb
Vb = V, Va = . V
Cb C
and
Va = V – Vb
where
Cb = concentration in the bottle, 50% in this
case
C = new concentration, 20% in this case
V = total final volume, 2000 cc in this case
Vb = volume of mixture
Va = volume of air or other dilutant
169
Vb = 20/50 . 2000 = 800 cc

Va = 2000 – 800 = 1200 cc
The final mixture would be made by taking 800 cc
of the calibration gas and mixing it with 1200 cc of air
to make the mixture equal to 20% LEL.
Another example is to dilute this 25000 ppm of
methane calibration gas to make a 100 ppm of mix-
ture.
Vb = 100/25000 . 2000 = 8 cc
therefore
Va = 2000 – 8 = 1992 cc
By mixing 8 cc of calibration gas into 1992 cc of
air, 2000 cubic centimeters of 100 ppm gas mixture is
obtained.
Some Calibration Tools

To perform the above procedure, the following
tools are needed:
1. Syringe and Needle: This is the
most inexpensive way to accurately
measure the amount of gas. A dis-
posable medical syringe with a large
A 2”-Diameter, 1000 cc Syringe gage needle is most practical but
there are few syringes with more than one hundred
cubic centimeter volume. Hence, large volume mea-
surements can be troublesome. However, it is easy to
make a syringe using any standard size pipe having
about a 2-inch diameter. It provides an easy and con-
venient means to make a mixture on a regular basis.
For very small volume measurements, there are mi-
cro syringes that are readily available in chemical
supply catalogues.
2. Calibration Bag: Most of the materials used in
food packaging or storage are quite inert; otherwise,
Standard Medical Syringes: food would be contaminated with odor. Therefore,
1 cc, 10 cc, and 50 cc
170
food storage bags can be used to hold most chemicals

as long as they are used for relatively short durations.
This is an important point to keep in mind since gas
molecules will eventually diffuse through the many
thin layers of a plastic bag. For example, potato chips
can stay fresh in their original bag for long periods of
time because the bag material is less permeable by
gas molecules than normal food storage bags. This is A 5-liter Sampling Bag
demonstrated by the fact that when the potato chips
are transferred into a tightly sealed food container
bag, they will lose their crispness in a very short time.
There are also many commercially available sampling
bags on the market. One common example is a Tedlar
bag. It is made from polyvinyl fluoride and has low
absorption of gas molecules. However, this type of bag
is still permeable, so a heavy gauge material will be
needed if permeability is a major concern. Sampling
bags normally come with a valve and a septum that is
used as an injection port.
Pressure Formula
Earlier, we described preparing a mixture based
on a volume relationship. Based on the ideal gas
law, the same volume formula can be used as a pres-
sure formula. As an illustration, take an 800 psi mix-
ture of 50%LEL methane with a 1200 psi mixture
of air. This will result in a 2000 psi mixture of
20%LEL methane.
Preparing gas mixtures can be a very difficult task. 1000 cc Calibration Cans
It is best to consult with the instrument manufac-
turer regarding the best method of calibration and
availability of gas mixtures.
Following are some examples of how gas mix-
tures can be made:
For ppm gas mixtures:
Cppm = Vc /(Vc + Vd) . 106 ppm
where Vc is target gas volume and Vd is the Mixing Calibration Gas
171
dilutant volume. For example, what is the concentra-

tion in parts per million when 1 cc of CO is added to
a 1000 cc container?
C = 1/1000 . 106 = 1000 ppm
ppm
Many containers are sized by gallons. For % range gas mixtures:

Therefore, it is useful to know that C% = Vc /(Vc + Vd) . 102 %
one gallon is equal to 3785 cc.
The Vc term in the denominator can be insignifi-
cant in low ppm mixtures.
Calibrating Liquid Chemical Mixtures. To make a
calibration mixture for liquid chemicals, a known vol-
ume of liquid is vaporized in a known volume of
dilutant air. The ideal gas law states that one gram
mole of molecules will occupy 24,500 cc of volume at
25 degree centigrade and at 760 mm of mercury or
sea level atmospheric pressure. This temperature and
pressure is also called the standard condition. At stan-
dard conditions, the equation is:
C = 24.5 . 109 . (V x D)/(V . M)
ppm a
where V = volume of liquid, D = density of the liq-

uid, which is the same as the specific gravity, Va = the
volume of the dilutant air, and M = the molecular
weight of the liquid.
Since it is easier to measure the liquid using a mi-
cro syringe, the equation then becomes
V = C . V . M/(24.5 . 109 . D)
ppm a
where all units are in milliliters, cubic centimeters,

and grams.
For example, benzene has M = 78.1 g and D =
0.88 g/cc. What is the amount of benzene needed to
make a 1000 ppm mixture in a 2000 cc bottle?
V = 1000 . 2000 . 78.1/(24.5 . 109 . 0.88)
which yields
V = 7.2 . 10-3 = 0.0072 cc = 7.2 microliters
172
Ventilation Hood for Calibration Procedures
In air pollution, industrial hygiene, and medical

toxicology work, the commonly used unit of concen-
tration is milligram per cubic meter. The following
equation expresses this relation, again assuming stan-
dard conditions.
Cppm = C . 24.5/M
where
C = mg/m3
M = molecular weight
In conclusion, for the calibration of gas monitors,

accuracy is not extremely important because these are
not analytical devices or systems. However, it is most im-
portant to keep the calibration methods standardized and
easily traceable. If procedures are standardized, data
can be normalized at a later date if necessary.
173
Appendix I PID Correction Factors
Appendix I
Compound Name Formula 9.8 C 10.6 C 11.7 C IP (eV)
PID Correction Factors

for Various Gases and Lamps with Instrument Calibrated to Isobutylene

Acetaldehyde C2H4O 5.5 + 10.23 Notes:
Acetic Acid C2H4O2 NR + 22 + 2.6 + 10.66
1. The values indicated by a plus (+) sign in
Acetic Anhydride C4H6O3 NR + 6.1 + 2.0 + 10.14 the “C” column are confirmed values; all
Acetone C3H6O 1.2 + 1.1 + 1.4 + 9.71 others are preliminary and subject to change.
Acetonitrile C2H3N 100 12.19 2. The correction factors in this table were mea-
Acetylene C2H2 2 11.40 sured in dry air. Actual readings may vary
with age and cleanliness of the lamp, rela-
Acrolein C3H4O 42 + 3.9 + 1.4 + 10.10 tive humidity, components in the sample,
Acrylic Acid C3H4O2 12 + 2.0 + 10.60 and other factors. For accurate work, the
Acrylonitrile C3H3N NR + 1.2 + 10.91 instrument should be calibrated regularly
under the operational conditions in which
Allyl alcohol C3H6O 2.4 + 1.7 9.67 the instrument is used.
Allyl chloride C3H5Cl 4.3 0.7 9.9
3. IP (eV) data was taken from the CRC Hand-
Ammonia H3N NR + 9.7 + 5.7 + 10.16 book of Chemistry and Physics, 73rd Edition,
Amyl alcohol C5H12O 5 10.00 and NIST Standard Ref. Database 19A.
Aniline C7H7N 0.50 + 0.5 + 0.5 + 7.72
Anisole C7H8O 0.8 8.21 (Courtesy of RAE Systems, Inc.)
Benzaldehyde C7H6O 1 9.49
Benzene C6H6 0.55 + 0.5 + 0.6 + 9.25
Benzonitrile C7H5N 1.6 9.62
Benzyl chloride C7H7Cl 2 0.7
Bromobenzene C6H5Br 0.6 0.5 8.98
Bromoform CHBr3 NR + 2.5 + 0.5 + 10.48
Bromopropane,1- C3H7Br 150 + 1.5 + 0.6 + 10.18
Butadiene C4H6 1.0 + 1.1 9.07
Butadiene diepoxide,1,3- C4H6O2 25 + 3.5 + 1.2 ~10
Butane C4H10 1.2 10.53
Butanol, 1- C4H10O 70 + 4.7 + 1.4 + 9.99
Butene, 1- C4H8 0.9 9.58
Butoxyethanol, 2- C6H14O2 1.8 + 1.2 + 0.6 + <10
191

Butyl acetate, n- C6H12O2 2.6 + 10
Butyl acrylate, n- C7H12O2 1.6 + 0.6 +
Butylamine C4H11N 7 8.71
Butyl mercaptan C4H10S 0.5 9.14
Carbon disulfide CS2 1.2 + 0.3 10.07
Carbon tetrachloride CCl4 NR + NR + 1.7 + 11.47
Chlorine Cl2 1.0 + 11.48
Chloro-1,3-butadiene, 2- C4H5Cl 3
Chlorobenzene C6H5Cl 0.44 + 0.40 + 0.39 + 9.06
Chloro-1,1-difluoroethane,
1- (R-142B) C2H3ClF2 NR NR 12.0
Chlorodifluoromethane CHClF2 NR NR NR 12.2
Chloroethane C2H5Cl NR + NR + 1.1 + 10.97
Chloroethanol C2H5ClO 10.52
Chloroethyl methyl ether,2- C3H7ClO 3
Chloroform CHCl3 NR + NR + 3.5 + 11.37
Chlorotoluene, o- C7H7Cl 0.5 0.6 8.83
Chlorotoluene, p- C7H7Cl 0.6 8.69
Crotonaldehyde C4H6O 1.5 + 1.1 + 1.0 + 9.73
Cumene C9H12 0.58 + 0.5 + 0.4 + 8.73
Cyanogen bromide CNBr NR NR NR 11.84
Cyanogen chloride CNCl NR NR NR 12.34
Cyclohexane C6H12 1.4 + 9.86
Cyclohexanol C6H12O 1.1 9.75
Cyclohexanone C6H10O 1.0 + 0.9 + 0.7 + 9.14
Cyclohexene C6H10 0.8 + 8.95
Cyclohexylamine C6H13N 1.2 8.62
Cyclopentane C5H10 0.6 10.51
Decane C10H22 4.0 + 1.4 + 0.4 + 9.65
Diacetone alcohol C6H12O2 0.7
Dibromoethane, 1,2- C2H4Br2 NR + 1.7 + 0.6 + 10.37
Dichlorobenzene, o C6H4Cl2 0.54 + 0.47 + 0.38 + 9.08
192

Dichlorodifluoromethane CCl2F2 NR + NR + 11.75
Dichloroethane, 1,1- C2H4Cl2 11.06
Dichloroethane, 1,2- C2H4Cl2 NR + 0.6 + 11.04
Dichloroethene, 1,1- C2H2Cl2 0.9 9.79
Dichloroethene, c-1,2- C2H2Cl2 0.8 9.66
Dichloroethene, t-1,2- C2H2Cl2 0.5 + 0.3 + 9.65
Dichloro-1-fluoroethane,
1,1- (R-141B) C2H3Cl2F NR + NR + 2.0 +
Dichloropropane, 1,2 C3H6Cl2 0.7 10.87
Dichloro-1-propene, 2,3- C3H4Cl2 1.9 + 1.3 + 0.7 + <10
Dichloro-1,1,1-trifluoro-
ethane, 2,2- (R-123) C2HCl2F3 NR + NR + 10.1 + 11.5
Diesel Fuel #1 m.w. 226 0.9 +
Diesel Fuel #2 m.w. 216 0.7 + 0.4 +
Diethylamine C4H11N 1 + 8.01
Diethylaminopropylamine,
3- C7H18N2 1.3
Diethylmaleate C8H12O4 4
Dimethylacetamide, N,N- C4H9NO 0.87 + 0.8 + 0.8 + 8.81
Dimethylamine C2H7N 1.5 8.23
Dimethyl disulfide C2H6S2 0.2 + 0.20 + 0.2 + 7.4
Dimethylformamide, N,N- C3H7NO 0.8 9.13
Dimethylhydrazine, 1,1- C2H8N2 0.8 + 0.8 + 7.28
Dimethyl sulfate C2H6O4S ~23 ~20 + 2.3 +
Dioxane, 1,4- C4H8O2 1.1 9.19
Epichlorohydrin C2H5ClO ~200 + 8.5 + 1.4 + 10.2
Ethane C2H6 NR + 15 + 11.52
Ethanol C2H6O 12 + 8 10.47
Ethanolamine
(Not Recommended) C2H7NO ~4 + ~3 + 8.96
Ethene C2H4 10 + 3 10.51
Ethoxyethanol, 2- C4H10O2 3.5 9.6
Ethyl acetate C4H8O2 4.6 + 10.01
193

Ethyl acrylate C5H8O2 2.4 + 1.0 + (<10.3)
Ethylamine C2H7N 0.8 8.86
Ethylbenzene C8H10 0.52 + 0.5 + 0.5 + 8.77
Ethylene glycol C2H6O2 16 + 6 + 10.16
Ethylene oxide C2H4O 19 + 3 + 10.57
Ethyl ether C4H10O 1.1 + 9.51
Ethyl formate C3H6O2 1.9 10.61
Ethyl hexyl acrylate, 2- C11H20O2 1.1 + 0.5 +
Ethyl (S)-(-)-lactate C5H10O3 13 + 3.2 + 1.6 + ~10
Ethyl mercaptan C2H6S 0.6 9.29
Ethyl sulfide C4H10S 0.5 + 8.43
Formaldehyde CH2O 0.6 10.87
Furfural C5H4O2 0.9 + 0.8 + 9.21
Gasoline #1 m.w. 72 0.9 +
Gasoline #2, 92 octane m.w. 93 1.3 + 1.0 + 0.5 +
Glutaraldehyde C5H8O2 1.1 + 0.8 + 0.6 +
Halothane C2HBrClF3 0.6 11.0
HCFC-123 (see 2,2-Dichloro-1,1,1-trifluoroethane, R-123)
HCFC-141B (see 1,1-Dichloro-1-fluoroethane)
HCFC-142B (see 1-Chloro-1,1-difluoroethane)
Heptane, n- C7H16 2.6 + 0.5 9.92
Hexamethyldisilazane,1,1,
1,3,3,3- HMDS C6H19NSi2 0.2 + 0.2 + ~8.6
Hexane, n C6H14 300 4.3 + 0.5 + 10.13
Hexene, 1- C6H12 0.8 9.44
Hydrazine H4N2 2.6 + 2.1 + 8.1
Hydrogen H2 NR + NR + NR + 15.43
Hydrogen peroxide H2O2 NR + NR + NR + 10.54
Hydrogen sulfide H2S NR + 3.3 + 1.5 + 10.45
Iodine I2 0.1 + 0.1 + 0.1 + 9.40
Isobutane C4H10 100 + 1.2 + 10.57
Isobutanol C4H10O 19 + 3.8 + 1.5 10.02
194

Isobutene C4H8 1.00 + 1.00 + 1.00 + 9.24
Isobutyl acrylate C7H12O2 1.5 + 0.60 +
Isoflurane C3H2ClF5O
Isooctane C8H18 1.4 9.86
Isopar G Solvent m.w. 148 0.8 +
Isopar M Solvent m.w. 191 0.7 + 0.4 +
Isophorone C9H14O 3 9.07
Isoprene C5H8 0.69 + 0.6 + 0.60 + 8.85
Isopropanol C3H8O 500 + 6.0 + 2.7 10.12
Isopropyl acetate C5H10O2 2.5 9.99
Isopropyl ether C6H14O 0.8 9.20
Jet fuel JP-4 m.w. 115 1 + 0.4 +
Jet fuel JP-5 m.w. 167 0.6 + 0.5 +
Jet fuel JP-8 m.w. 165 0.6 + 0.3 +
Kerosene (C10-C16 petro.distillate - see Jet Fuels)
Mesitylene C9H12 0.36 + 0.35 + 0.3 + 8.41
Methane CH4 NR + NR + NR + 12.51
Methanol CH4O NR + NR + 2.5 + 10.85
Methoxyethanol, 2- C3H8O2 4.8 + 2.4 + 1.4 + 10.1
Methoxyethoxyethanol, 2- C7H16O3 2.3 + 1.2 + 0.9 + <10
Methyl acetate C3H6O2 1.6 10.27
Methyl acrylate C4H6O2 3.7 + 1.2 + (9.9)
Methylamine CH5N 1.0 8.97
Methyl bromide CH3Br 110 + 1.7 + 1.3 + 10.54
Methyl t-butyl ether C5H12O 0.9 + 9.24
Methyl cellosolve (see 2-Methoxyethanol)
Methyl chloride CH3Cl NR + NR + 0.7 + 11.22
Methylcyclohexane C7H14 1.1 9.64
Methylene chloride CH2Cl2 NR + NR + 0.89 + 11.32
Methyl ethyl ketone C4H8O 0.86 + 0.9 + 1.1 + 9.51
Methylhydrazine C2H6N2 1.4 + 1.2 + 1.3 + 7.7
Methyl isobutyl ketone C6H12O 1.2 + 0.9 9.30
195

Methyl isocyanate C2H3NO NR + 4.6 + 1.5 10.67
Methyl mercaptan CH4S 0.6 9.44
Methyl methacrylate C5H8O2 1.4 + 1.4 9.7
Methyl propyl ketone C5H12O 0.9 + 0.8 + 9.38
Methyl-2-pyrrolidinone, N- C5H9NO 1.0 + 0.8 + 0.9 + 9.17
Methyl salicylate C8H8O3 2
Methylstyrene, a- C9H10 0.5 8.18
Mineral spirits
(Stoddard Solvent,
see also Viscor 120B) 0.7 + 0.39 +
Mineral Spirits
Viscor 120B
Calibration Fluid 1.0 + 0.7 + 0.3 +
Naphthalene (Mothballs) C10H8 0.45 + 0.4 + 0.40 + 8.13
Nitric oxide NO 5.2 + 2.8 + 9.26
Nitrobenzene C6H5NO2 2.6 + 1.9 + 1.6 + 9.81
Nitroethane C2H5NO2 3 10.88
Nitrogen dioxide NO2 NR + NR + 9.75
Nitromethane CH3NO2 4 11.02
Nitropropane, 2- C3H7NO2 2.6 10.71
Nonane C9H20 2 9.72
Octane, n- C8H18 13.2 + 1.8 + 9.82
Pentane C5H12 80 + 8.4 + 0.7 + 10.35
Peracetic acid C2H4O3 NR + NR + 2.3 +
Peracetic/Acetic acid mix C2H4O3/C2H4O2 50 + 2.5 +
Perchloroethene C2Cl4 0.69 + 0.57 + 0.31 + 9.32
PGME C6H12O3 2.4 + 1.5 + 1.1 +
PGMEA C6H12O3 1.65 + 1.0 + 0.8 +
Phenol C6H6O 1.0 + 1.0 + 0.9 + 8.51
Phosphine in N2 PH3 2 + 1.4 9.87
Photocopier Toner 0.5 + 0.3 +
Picoline, 3- C6H7N 0.9 9.04
Pinene, a- C10H16 0.3 + 0.5 8.07
196

Pinene, b C10H16 0.38 + 0.4 + 0.4 + ~8
Piperylene, isomer mix C5H8 0.76 + 0.7 + 0.6 + 8.6
Propane C3H8 NR + 1.8 + 10.95
Propanol, n- C3H8O 6 1.7 10.22
Propene C3H6 1.7 + 9.73
Propionaldehyde C3H6O 1.9 9.95
Propyl acetate, n- C5H10O2 3.5 10.04
Propylene oxide C3H6O 6.5 2 10.22
Propyleneimine C3H7N 1.5 + 1.3 + 1.0 + 9.0
Pyridine C5H5N 0.78 + 0.7 + 0.7 + 9.25
RR7300
C4H10O2/
(70:30 PGME/PGMEA) 1.4 + 1.0 +
C6H12O3
Stoddard Solvent - see Mineral Spirits
Styrene C8H8 0.45 + 0.40 + 0.4 + 8.43
Sulfur dioxide SO2 NR + NR + 12.32
Tetrachloroethane, 1,1,1,2- C2H2Cl4 1.3 ~11.1
Tetrachloroethane, 1,1,2,2- C2H2Cl4 NR + NR + 0.60 + ~11.1
Tetraethyllead C8H20Pb 0.4 0.3 0.2 ~11.1
Tetraethyl orthosilicate C8H20O4Si 0.7 + 0.2 + ~9.8
Tetrafluoroethane, 1,1,1,2- C2H2F4 NR NR
Tetrafluoromethane CF4 NR + NR + >15.3
Tetrahydrofuran C4H8O 1.9 + 1.7 + 1.0 + 9.41
Therminol 0.90 + 0.7 +
Toluene C7H8 0.54 + 0.50 + 0.51 + 8.82
Tolylene-2,4-diisocyanate C9H6N2O2 1.4 + 1.4 + 2.0 +
Trichloroethane, 1,1,1- C2H3Cl3 NR + 1 + 11
Trichloroethane, 1,1,2- C2H3Cl3 NR + NR + 0.9 + 11.0
Trichloroethene C2HCl3 0.62 + 0.5 + 0.4 + 9.47
Trichlorotrifluoroethane,
1,1,2- CFC-113 C2Cl3F3 NR NR 11.99
Triethylamine C6H15N 1.3 7.50
Trifluoroethane, 1,1,2- C2H3F3 34 12.9
197

Trimethylamine C3H9N 0.9 7.82
Trimethylbenzene, 1,3,5- - (see Mesitylene)
Turpentine C10H16 0.4
Undecane C11H24 2 9.56
Vinyl actetate C4H6O2 1.2 9.19
Vinyl bromide C2H3Br 0.4 9.80
Vinyl chloride in N2 C2H3Cl 2.0 + 0.6 + 9.99
Vinyl-2-pyrrolidinone, 1- C6H9NO 1.0 + 0.8 + 0.9 +
Viscor 120B - see Mineral Spirits - Viscor 120B Calibration Fluid
Xylene, m- C8H10 0.50 + 0.4 + 0.40 + 8.56
Xylene, o- C8H10 0.57 + 0.6 + 0.7 8.56
Xylene, p- C8H10 0.5 + 0.6 + 8.44
198
Thermocatalitic sensor in dynamic mode
1. General principle
Schematic diagram of sensor for methane concentration in dinamic mode is shown on the figure below:
C0
1. Cover made of strained metal

1 chippings or porous ceramic
2. Calibrated hole between prechamber
and combustion chamber
2 3. Combustion chamber
Thermocatalitic element, platinum
coil with small temperature inertia
Co - concentration of the gas outside
3 the sensor
C
C- concentration of the gas inside the
combustion chamber
St —overall signal (current or voltage)
4 which is result of heathing of the coil
from curent flow and of gas combustion
onto the coil.
St
Figure 1
When the coil is powered, it will start to heat up, and gas into chamber will start to combust. At the
beguinning inflow of the gas through calibrated hole is smaller then the quantity of the combusted gas
onto the coil, concentration of gas C will be reduced. When the quantity of combusting gas becomes
equal to quantity of inflow gas, concentration stabilize with new value Ci (see diagram 3). It is easy to
show theoreticaly, and experiments certify, that gas concentration into chamber reduce with exponential
trend.
At the moment when concentration almost reach Ci, power is turned off and the chamber starts to fill
again, it fills up for about the same time. The power is turned on again and so on alternatively.
Transitive process in which we measure the signal in two consecutively points is repeated, and on that
base we determine concentration of the target gas, as we will see later. It is validated truth that
transitial process is much more reliable and much informational than the stationar.
Ss
Si
I II III IV
Figure 3
One short time period after energizing the coil (t=0) there is no respond from system and that is
inertial time of the system, a signal remains null (period I), after that time, coil starts to heat up
rapidly caused from current and also from gas combustion until the signal reach its maximal value Ss
(period II). In the period III, which is of the greatest interest for metrology, flux of outer gas inflowing
through the hole is smaller than the quantity of the combusted gas and the concentration decrease ‘till
reach equilibrium point when the signal becomes Si. Signal then stays constant on minimal value Si
(period 4).
The following trend of exponencial decreasing of the concentration in the period III is changing by
following formula:
C = Cs [ r + ( 1 - r ) e-t/T]
Where are : r = Ci/Cs factor of concentration reduction, and T — is time constant. The both of this
factors are characteristics specific for each. Time t is calculated from the end of the period II (t=0).
Because the signal is proportional with concentration S = k x C, it folows:
S = k Cs [ r + (1 - r ) e —t/T ]
This is, of course, only a part fo the signal and its origin is from the gas combustion. Other part of
the signal, parasitic signal, is the result in the first place of resistor heating caused by current flow but
its also caused by ambient temperature. It also consists of other relevant factors as is the air pressure,
humidity, resistive changes with time, etc.
St= S + Spar
Signal total St
Ss
L = St(t1)-St(t2)
Si
Spar
t=0 t1 t2 Temps (Unités arbitraires)
Figure 4
On this curve coordinate beguinning (t=0) is shifted at the end of period II. Total signal is measured in
two points t1 and t2 on transition curve.
The difference of the signals in those two points L (t1 , t2 ) = S (t1) - S (t2) is proportional with gas
concentration.
From theory stated above it can be seen that the number of parameters are before the constructor to
adapt them in the best way to his specific applications. We will comment only briefly some of them.
The diameter of the calibrated hole and the capacity of combustion of coil, are obviously in relation to
each other. It’s stated before that it is beter to have coil with as smaller thermal inertia as it is
possibile, but when it is achieved by minimizing the vire dimesions and coil itself, thermocatalitic
surface layer should achieve very intensive combustion and in that way increase sensitivity of sensor
itself. The diameter of calibrated hole considerably influence on the form of exponential curve,
determining its gradient, level of the Ci (Si) and the time for reaching that value (time constant).
Selection of the points t1 and t2 in which we measure the signal is of great importance. Point t1
should be chosen far enought from maximal value, to avoid shifting of that point to left side from the
maximum, because during time coil characteristic gets deformed and maximum point is sliding slowly
towards right side.
The distance between points t1 and t2 is from great importance. It won’t be same for stationary
devices in mines which are constantly exposed to target gas and devices in industry (stationary or
portable) where the presence of target gas is mostly incidently.
The proper choose of this two parameters enables construction of so called “universal” explosimeter.
As a matter of fact, classical explosimeters shows accurately only concentration of gas for which they
are calibrated and indicate it in percents of lower explosive limit (LEL). Any other explosive gas they
only indicate, and measuring of explosivity itself can deflect up to 300% in extreme cases. This makes
those exposimeters practically useless, in areas where we don’t know which gas is present, or when
there is a mixture of gases.
Dinamic mode enables to, with proper chose of parameters t1 and t2, realize explosimeter wich will
equaly acurate measure explosivity of mostly combustile gases, and for extreme cases deflection is
below few tenth of percent (referred to curently 300%).
2. Instrument realized on basis with thermocatalitic sensor in dinamic mode

The stated principle is able to realize only with use of micro-controller, which is done in cooperation
of French company Oldham and Russian laboratory NPC-ATB, and it is realized and specific
measuring instrument is called DMS-01. Measuring range of this device is from 0-100% methane, it
has beside catalitic sensor in dinamic mode, which works in range of 0-5% methane, installed in the
same measuring chamber another catharometric sensor wich cover range of 5-100% methane.
Design of measuring cell (sensor) is given on following figure:
gas
sinter
0.5 - 0.7mm
HC LC
Principle of opertion is as follows: first of all, micro-controller turns on thermoconductive sensor for
high methane concentrations HC with short current impulse which lasts for 0.35 sec., during that time
thermocatalitic sensor is turned off (no methane combustion).
As the concentration of methane is higher, the slope is under lower angle because the wire start to
cool more intensive (element is constantly energized) this is previously stated distinction of the
catarometric sensors.
If a methane concentration is above level of DGE (5%CH4), sensor for lower concentrations stays
turned off, otherwise it turns on, but only if the concentration is less than DGE. Time line of one
measuring cycle, in case that after initial turning on the DMS-01 has measured concentration below
DGE, is shown on following figure:
1 — short current impulse for thermoconductive sensor, lasting for 0,35 sec
2 — sensor response HC
3 — current impulse for heating the LC sensor, lasting for 2,5 sec (turns on at t=1s and turns off at
t=3.5s ), when proper temperature is reached (about 450°C) and combustion of methane starts.
4 — sensor response LC in lack of methane
5 — sensor response in presence of methane
Briefly, LC sensor is a tipical thermocatalitic sensor, it is installed in a chamber with calibrated hole
0.5 — 0.7mm. The gas pass thru this calibrated hole, sensor (LC) turns on and when wire reach proper
temperature the methane combustion starts. Quantity of the gas that combust is in proportion with
concentration of the gas that’s inside the chamber, and after some particular time, because the power
supply of the sensor is turned off, the balance renew. Transition process from the moment of turning
on ‘till stabilization is used for measuring.
Measuring is conducted in two points, first point is at 1.3 sec. and the second one at 2.3 sec. (after
turning on the sensor). Concentration is equal to difference of these two signals.
1.3s 2.3s
1s
After impulse of 2.5 sec., there is a pause of 5 sec. to fill the chamber again, and the whole process is
cyclicly repeating. Total cycle of measurment lasts for 7,5 sec. and is cyclicly repeating, untill to
moment when sensor of lower concentration exceed the DGE (5% methane) when the sensor of higher
concentration HC turns on. Farther work of this device is on the same principle as after turning on
(when the concentration falls below DGE, sensor HC is turned off, and sensor LC turns on).
3. Analysis of Ex protection for thermocatalitic sensor in dynamic mode

The next results are achieved in establishing a special conditions for Ex protection:
- Power supply of the sensor via electronic circuitry with intrinsically safety with maximal
values of electrical parameters: Um=1.41V, Iib=116 mA.
- Limited heating of the thermoelement to safety limit (≤ 450°C).
- Thanks to exceptional low power consumption enough for accurate sensor operation (low wire
diameter), a high temperature coefficient of resistive change ( K=3.9 x 10-3), and rather low
thermal inertia (measuring cycle lasts for 7.5 sec., 2.5 sec for heating, and 5 sec. for cooling
down), its created the effect of developing a rather low amount of thermal energy, which leads
to low operating temperature of the sensor (Tmax=73°C), and which is a far lower from
maximal safety temperature tbez=295°C, derived with a safety factor of 1.5. Measurements in
stated example are achieved with power supply conditions U=1.41V, I=116 mA, in presence
of mixture of propane 5,3% and ambient air, at the temperature of 50°C (mixture is in comply
with group of gases IIA).
- Inlet of the target gas inside the sensor is thru sintered mass which eliminate possibility of
intruding of coal dust inside the sensor, which lead to higher permited temperature limit inside
the sensor from 1500 to 450°C.
Design of the sensor, that means the operational principle of dynamic mode, have allowed to designers
to construct the measuring intrument, with intrinsically safety conditions in compy with limit curves
(IEC 79-11) witout any problems, i.e. far from the limit curve.
4. Summary
Thermocatalitic sensors in dynamic mode of operation has following advantages compared to
“classical” thermocatalitic sensor:
1. With only one element, without reference, we determine the gas concentration, avoiding in that way
any inconvenience which reference element can cause (need for selection — complying of elements,
operative with the comparative, different aging of this two elements, which is inevitable, and which
lead to zero drift, etc.)
2. With dynamic mode, actually we compare element with itself, it is referent to itself, comparative, if
its characteristics change over time, those changes are the same in points t1 and t2. (figure No. 4). On
that way problem of zero drift is solved which is one of the biggest advantages of this sensor.
3. Power consumption of sensor itself is manifold smaller in compare to classical thermocatalitic

sensors, because there is no reference element, thus only one element is heating (consumption is 50%
reduced). Sensor isn’t working all the time (only 2.5sec) and 5 sec is the time for new charging of the
chamber (additional consumption reduction). And finally, thin wire of 10 μm heats and cools down
very quickly.
4. Unlike classical thermocatalitic sensors which has to be installed in flameproof housing, i.e. realised
in Ex d, Ex protection of this sensor is Ex ia, thus eases in a great way construction of detector itself.
5. Proper choose of parameters of thermocatalitic sensor in dynamic mode allows construction of so

called “universal” explosimeter. Particulary, classical explosimeters shows acurate readings only for
concentrations of gas for which they are calibrated for and indicate it in percents of lower explosive
limit (LEL). Any other explosive gas, they only indicate, and readings itself of the explosivity could
deflect up to 300% in extreme cases. This make them practically useless, in areas where we don’t
know which gas is present, or when there is a mixture of gases.
Dynamic mode allows us to realize explosimeter, with proper selection of parameters (t1 and t2),
which will equaly acurate measure explosivity of most of common combustive gases, and for those
extreme cases deflection is around few tenth of percents (referred to curently 300%).
General conclusion is, that the despite of use of IR sensors, changed thermocatalitic sensor in dynamic
mode of operation has its distinctive position as well in mining as so in industry.

Hazardous Gas Monitors

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hazardous Gas Monitors

Uploaded by

Copyright:

Available Formats

Chapter 1 Introduction

• A simple guide to the various sensor technolo-

The main emphasis of the book is on gas detec-

Analytical Instruments and Monitoring Systems

• suitable for multisensor systems that, for ex-

relatively straightforward and deal with voltage, tem-

• catalytic bead • photoionization Some Examples of Gas Sensors

All of these sensors are commonly used for detection of

Terms, Definitions, and Abbreviations

Table 1. Equations for Deriving Units of Gas Concentration

an ignitable and combustible mixture when air

TOO RICH FOR COMBUSTION

TOO LEAN FOR COMBUSTION

Specific Gravity (Sp.Gr.): The ratio of the weight

For Toxic Gases

the world. ACGIH publishes the exposure standard,

given moment is repre-

Recommended Exposure Limit (REL) is TWA con-

60 minutes between successive

IDLH extracted from the same pocket guide

that includes REL, PEL and IDLH

“the quality or state of being accurate or exact;

there is no standard 100 ppm of CO to compare

readings may result from dry air compared with

components are separated and come out at the

cific gas being targeted; in which case, the manu-

Hysteresis: The difference in response of the sensor when

have been influenced by North American practices

dental rupture or breakdown of a container, stor-

or for long periods of time.

Types of Flamepaths in Flameproof Enclosures

Joint Threaded Spigot

Materials used: Aluminum, Iron (Courtesy of Cullen Associates)

B. Intrinsic Safety (i): The electrical energy in

C. Purged and Pressurized (p): This is the pro-

E. Other Protection Methods: Oil immersion(o),

Enclosure Classifications For Nonhazardous Areas

NEMA Type 12: For indoor use, provides protec-

Protection First Digit Second Digit

The IP classification is set by the International

be exactly equated with NEMA enclosures.

Gas Molecules Micro Ammeter

first passes through a small capillary-type opening and

the gas concentration. The value of the voltage applied

Oxidation at the anode: 2Pb + 2H2O → 2PbO + 4H+ + 4e-

The overall reaction in both cases is: 2Pb + O2 → 2PbO.

lary. Such sensors are referred to as capillary-type

trochemical sensor can be made more selec-

3. Electrolyte within the sensor cell is an aqueous

Pressure and Temperature

designed to detect. The best electrochemical sensor Typical Toxic Gas

tual applications, interference concentrations can be Temperature Range: –40oC to +45oC

the combustion process and results in cleaner and

without changing its physical properties. It is capable

R1 of producing a constant reliable signal over an ex-

tended period of time.

The electrical circuit used to measure the output

of catalytic sensors is called a Wheatstone bridge, in honor

in a square. The source of the electrical current is con-

gas. This is not practically possible, however. A com-

Fig. 4 A Catalytic Bead Sensor

The construction of the catalytic sensor bead is

mass which helps make it resistant to shock and vibra-