-------
the process under consideration and the types and compositions of hazardous
wastes that may be burned in each incinerator. Where possible, published test
data were used. In certain cases, such as the pre-air pollution control
device zone, there are no published test results and only engineering
estimates can be given. In other cases, such as combustion zone monitoring,
the few test results that are available have been obtained during the
combustion of hard-to-incinerate wastes such as PCBs and herbicide orange.
The operating parameters recorded during these tests may be more severe than
(and not necessarily representative of) typical hazardous waste incineration
efforts. Nevertheless, where data were available, they were used to develop
the presented ranges.
The reported monitoring parameter ranges are intended to provide a
minimum and maximum for expected gas conditions. They should include most,
but not necessarily all, conditions that may be experienced in a hazardous
waste burn. More detailed breakdowns are provided in Volume II. Individual
incinerator operators may decide to operate at higher temperatures or greater
oxygen concentrations (excess air rates) than presented here to ensure
complete destruction of the waste being incinerated. These situations were
not prejudged in the establishment of a range for each parameter. In like
fashion, the range of pollutant concentrations may be exceeded if a relatively
pure stream of a single pollutant compound; i.e., hydrogen sulfide, is
incinerated. Again it was judged that such an event is unlikely, and
therefore should not be used to define an unreasonably high maximum
concentration.
MONITORING APPLICATIONS
The choice of a conditioning system and continuous gas analyzers is
influenced by the concentration ranges of the species of interest in the
incinerator system. The conditioning system must be designed and constructed
to withstand the high temperature corrosive environment of the combustion zone
as well as the wet, cool environment after a scrubber. At present, federal
regulations specify temperature, pressure, CO, and 02 ranges as indicators
of destruction efficiency. The performance specifications for these
continuous analyzers are defined and should be used as a guide in selecting
instruments for the other parameters of interest.
Three monitoring zones are considered throughout this report to address
four basic goals of hazardous waste incinerator emission monitoring. The most
important goal is the documentation of pollutant destruction efficiency. With
monitored waste feed rates and compositions, continuous monitors in the
combustion zone would enable destruction efficiencies to be determined. A
second goal is the continuous documentation of control device removal
efficiency. Comparing pollutant concentrations measured before and after the
pollution control device would allow real time monitoring of efficiency. The
third goal is the continuous documentation of the emissions to the
environment. The compliance status of the source, with respect to applicable
air pollution control regulations, would be more fully documented through
continuous monitoring in addition to source performance testing conducted
periodically. The fourth goal is to provide the source operator with process
information to aid in determinations of maintenance needs of the incinerator
and pollution control equipment.
11
-------
SECTION 4
CONCEPTS OF SAMPLE CONDITIONING
Systems used to measure the performance of a hazardous waste incinerator
must be equipped to measure criteria pollutants (NOX, S02, CO, etc.) and
selected organic species under a wide range of sampling conditions. Depending
on the incineration system and the sampling point location, temperature
extremes of up to 1700°C (3000°F), moisture levels approaching 65 percent, and
acid gas levels that simultaneously reach 17 percent by volume per acid
component could be encountered. The GEM conditioning system must be
sufficiently versatile to accommodate the hot, corrosive atmosphere of the
combustion zone as well as the cooler, wet sampling conditions in the exhaust
stack.
The objective of sample conditioning is to provide an unadulterated
sample that is compatible with the analyzers in use. Maintaining the
integrity of the inorganic and organic analytes while removing compounds that
are interferences or are harmful to the instrument (i.e., HC1 gas) is
difficult when large changes in the physical state of the sample are
required. The three sample zones represent a wide cross section of
temperatures, pressures, moisture contents, acid gas contents, and particulate
loadings. Ideally, one inorganic and one organic sample conditioning system
could be used for all three locations through proper system design and choice
of analyzers. Even more desirable is the ability of this same conditioning
system to be used for both inorganic and organic species of interest to
minimize the capital investment required by the incinerator owner.
Sample conditioning equipment is defined as all flow related equipment
preceeding the continuous gas analyzers. The following subsystems are
integral parts of the sample conditioning equipment:
• Probe/preconditioning module/heat traced transport lines
• Flow control/pumping system
• Distribution control system
Flue gas is typically passed through a heated probe and into a
conditioning module where large particulate, acid mist, high boiling
compounds, and moisture removal is accomplished. From this point the sample
gas is transported to the continuous gab analyzers through transport tubing,
flow control systems, and various valving systems. Both calibration gases and
sample gases are passed through a distribution control panel. It is desirable
12
-------
to route calibration gases through the probe inlet for complete system
calibration. The operating requirements of a particular analyzer play an
important part in the design of the sample interface system. Requirements
that must be considered are as follows:
• Sample temperature, pressure, and flow rate
• Sample particulate and moisture levels
• Interfering species in the gas stream
• Chemical corrosion of interface surfaces.
Besides maintaining sample integrity, other important considerations are
maintainability and effects on response time. For effective long term
operation, the conditioning systems should only require minimal routine weekly
maintenance. The conditioning should be relatively quick, ideally less than
1 minute, in order to provide a sample as close to "real-time" as possible.
Conditioning systems have been designed and manufactured on a case-by-case
basis where the range of wastes to be incinerated was well defined. Some
degree of adaptation of any system must be done by a user. Specific
conditioning requirements of each continuous monitor, if known, will be
referenced in the appropriate instrument sections. Conditioning methods and
the advantages and limitations of each method as applied to measurement of
inorganic and organic species will be discussed in the following sections.
PARTICULATE REMOVAL
Most extractive gas analyzers and organic species analyzers require the
removal of particulate matter prior to sample analysis. In addition,
particulate matter can clog sample lines and shorten the life of the sample
pump. Particles can be removed by coarse and fine filters or inertial
separators. There are numerous types of filter bodies and filter elements
constructed of both organic and inorganic materials. Selection is based on
particle size and stream loading factors, as well as the nature of the analyte
gases. Particles are usually removed in two stages: an initial stage (coarse
filtration) and a final stage (fine filtration).
Coarse Filtration
Larger particles can be removed by passive filtration, inertial
separation, or inertial filtration. Passive filtration is accomplished by
placing a sintered metal filter of large surface area, typically constructed
of sintered 316 stainless steel, at or near the sample stream extraction
point. Sintered filters normally remove all particles that are larger than 10
to 50 micrometers. Glass, ceramic, quartz, and other metals (Carpenter 20,
Hastelloy C) may also be used in sintered filters. The porous medium should
be protected by a baffle to prevent excessive particulate buildup on the
leading surfaces when mounted in the flue. These filters should be maintained
at or above the flue gas temperature to minimize the incidence of plugging and
corros ion.
13
-------
Larger particles can also be removed by inertial separation. This type
of coarse particulate removal can be accomplished by using a flow-through
cyclone. In the cyclone, the sample gas is introduced tangentially and
exhausted through the bottom. The cyclonic flow causes the particles to be
accelerated toward the inner walls of the device. The sample stream is
extracted from the vortex present inside the cyclone. Cyclones have a
distinct particle size upper cut point dependent upon geometry, flow rate, and
gas viscosity. Gases containing water droplets and/or high concentrations of
large particles, would be more effectively "coarse" cleaned with a cyclone
than a filter. Cyclones do not plug easily and can be cleaned using automatic
blow-out systems. A disadvantage is higher cost; however, the failure rate is
low. The cyclone should be maintained at or above the flue gas temperature.
Many materials of construction are available.
The flow-through tube filter utilized in inertial filtration combines the
techniques of passive filtration and inertial separation. The large ratio of
axial to radial sample gas velocity in the tube filter prevents larger
particles from impinging on the filter pore structure. Smaller particles
establish a dynamic membrane on or within the porous wall, and in equilibrium
with very low drag forces, effectively prevent transmission of particulate
contaminants much smaller than the filter pore size. The turbulent nature of
particulate laden gas flowing through the filter tends to keep the filter
clean by abrasion. The flow-through tube filter can be mounted either
internal or external to the stack or duct. This element would require the
same high temperature materials as passive filtration and cyclone inertial
separation devices.
Fine Filtration
The majority of extractive gas analyzers and organic species analyzers
require almost complete removal of all particles larger than 1 micron from the
sample gas stream. In order to reduce particles to this level, a low
resistance, high efficiency filter is needed. Fine filtration is usually
accomplished by adding a second fine filter near the analyzer inlet. Fine
filters are divided into two categories: surface filters and depth filters.
Surface filters remove particles from the gas stream using a porous matrix.
These filters can remove particulates smaller than' the actual filter pore size
as a result of particulate cake buildup and electrostatic forces acting to
trap smaller particles without excessive resistance. Depth filters collect
particulate matter within the bulk of the material. A depth filter may
consist of loosely packed fibers or relatively large diameter granules. The
spun glass filter is a depth filter which when maintained at elevated
temperatures (e.g., 220°F (104°C)), reliably and efficiently removes
particulates as small as 0.5 micron. Spun glass, when packed to a density of
0.1 gm/cm3 and a bed depth of at least 2 inches, can act as an inexpensive
depth filter for normal gas flow rates.
Application Requirements
The corrosive constituents encountered in monitoring hazardous waste
incinerators are: nitrogen oxides, sulfur dioxide, dilute to concentrated
sulfuric acid (wet 803 or acid mist), and halogen acids (HC1, HBR, HF). An
14
-------
instack filter should be fabricated of sintered metal, refractory, metallic
screen, or high temperature fabrics. Ordinary steel is limited to stack
temperatures below 1400°F (760°C) while 316SS is usable to 1600°F (870°C). At
higher temperatures, a water-cooled probe/filter may be used or stainless
steels with higher nickel compositions such as Hastelloy or Inconel.
The chemical resistance of various materials to some expected gas
constituents has been collected from various sources and summarized in
Table 7. Of the metals, titanium is the most resistant (and also the most
expensive) followed by the nickel alloys and stainless steel alloys. Finally,
the inert materials such as quartz and glass are quite resistant but the most
fragile.
Studies have shown that some organic compounds adsorb onto the filtered
particulate matter which result in a low bias at the instrument. The relative
amounts of adsorbed material is a function of filtration temperature among
other factors. Stack tests at the Marquardt liquid injection incinerator^
were performed where combustion zone gases were cooled from 1000°C to 100°C
prior to particulate filtration. Analysis of the filtered particulate was
performed by extracting with methylene chloride and using instrumental IR and
low resolution mass specrometry (LRMS) whereby alkyl hydrocarbons and alkyl
esters of phthalic acid were found. Higher boiling point compounds not
extracted by methylene chloride may also have been present. Particulate
removal by filtration at elevated temperatures may prevent condensation of
organic compounds, however, this poses a problem in that only coarse filters
are routinely used at high temperature combustion zone sampling locations.
Fine filters that remove particles smaller than one micron need to be found
that will withstand 1000 to 1500°C temperatures. Filter materials must be
chosen that will not catalyze additional chemical reactions. The fine filters
should also be self-cleaning and only require weekly maintenance or changing
since maintenance performed near the hot zones of an incinerator are dangerous
procedures.
For the zones before and after the control device, distinguishing between
particulate matter and condensed organic material is the major problem. As
the effluent gas cools, higher boiling point compounds may condense and form
aerosols. Filtering and removing these organic aerosols could only be
prevented through revaporization of the sample. Analysis of these higher
boiling compounds is of prime importance when determining the overall
destruction efficiency. In all three zones, the affect of the filter
material, binders and filtered particulate matter on passing organic compounds
needs to be determined. Acid catalyzed reactions, fine particles acting as
reaction sites, and numerous other possibilities exist for altering or
trapping the organic compounds of interest.
The compatability of filter materials to the waste gas stream is also
important. High concentrations of various organic compounds have been tested
for compatability with many types of filters. Table 8 provides an indication
of the chemical resistance of three general types of filters to some common
organic compounds. Obviously teflon is the most inert, however, other
considerations such as pressure drop, pore sizes, costs, temperature
limitations and replacement frequency need to be considered when designing a
15
-------
'
CO
EH
Fd
H
M
H
CO
O
cj
CO
o
0
H
EH
CJ
Pd
p-<
X!
Pd
o
H
CO
hJ CO
<3 pd
M O
W o
S o o
0) -Q O
H fl -<
en
O
en
4J CM
* en
>, CM
h 0
^
-3 S
Z
x cn
U CM
a M
ri o^
^ 3
g j — ^
o -• cn
to a ^CM
S
§ «s
u
-2 S
n x •-<
c ax
o
•H
CO J ^
2<-i O
a x
O
O 4J CM
CU r-t
3 U
M CM
£9
a 8
Cu
CO
t5 SM
3G
,
cn
V
a z
O
z
u
a
03
1-1
a)
S
o o
0 00
en m
,~« CM
CL, i,
s r
CM r-<
N
fM ml r-l
-« CM CM
CM —1 ^H
CM -J r-l
CM ^1 CM
(M r-l ^
CO CM CM
CN CN ^H
CM CM CM
CM fM ^
N " "
N - -
"^
§
f-l
•"
2
-•
X
e o
3 ^
» 'c *«
!t U 0)
3 H 5
* Cn c a c i -i
,fl o O O O
-------
o. > o
BOO
a) -a o
E-» fl —•
a) o
3 M
r-» CM CM 04
*H CM CM CM CM
CM CM
0)
4J
§
o
9
-------
TABLE 8. ORGANIC SPECIES/FILTER MATERIAL COMPATABILITY
Depth
nitrocellulose
ACIDS
Concentrated
6 Normal and weaker
ALCOHOLS
n-Amyl, butyl, glycerol
ethyl, methyl, isopropyl
BASES
CRYOGENIC GASES
ESTERS
ETHER ALCOHOLS
FUELS
HALOGENATED HYDROCARBONS
cci4
Chloroform
Methylene Chloride
Perchloroethylene, trichloroethylene
HYDROCARBONS
KE TONES
OILS
PHOTO RESISTS
SOLVENTS
WATER
D
G
G
D
D
G
D
D
G
G
G
F
G
G
D
G
D
G
G
Screen
po lye arbonat e
G
G
G
G
D
G
G
F
G
G
D
D
F to G
G
G
G
G
G
G
Porous
teflon
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G - Good
F - Fair
D - Dissolves
18
-------
specific filtration system. Very low concentrations of the organic species
are expected to be present, therefore, minimizing gross compatability
problems; however, substantial acid gas concentrations are expected to greatly
influence the choice of filter material.
ACID MIST AND HIGH BOILING COMPOUND REMOVAL
Incinerator effluent gases can contain significant quantities of acid
mists and vapors. Analyzer temperatures must be kept above the acid dewpoint
to prevent condensation within the analyzer. An acid mist coalescer, which is
a knockout bowl of nonreactive glass or plexiglass, can be used to remove acid
mist. The trap should have a drain line made of nonreactive material such as
teflon, which leads to a collection bottle, treatment system, or drain. The
line can have an automatic or manual drain valve. The automatic valve should
dump the filter bowl when a back purge is activated. Analysis of the water or
condensed sample could be performed to verify no loss of sample. The knockout
bowl is maintained at a lower temperature than the gas stream. This drop in
temperature through the bowl may effectively remove most of the acid mist. If
cyclones or tube filters are used for the initial particulate removal, these
will also remove acid mist droplets. The filter elements must be inert and
should not include any binders that undergo chemical changes. One of the
better materials for this purpose is polypropylene. Polypropylene is
hydrophobic (repels liquid water) and is available as a yarn, fabric or a
fused, porous tube. As a yarn it can be wound on a core to provide a larger
surface area for coalescing acid vapors. As a fabric it should be pleated to
give a larger surface area. Fiberglass yarn can be wound on a stainless steel
core and, if the winding lubricant is extracted, a pure glass element can be
produced. The glass element can be used at high temperature and is chemically
inert. Many filter housings designed to be chemically inert, withstand high
pressures, and provide visibility of the filter element can be assembled and
disassembled without tools. Filter housings are designed to maintain good
heat transfer. The chemical resistance of the filters and materials of
construction are presented in Table 9.
Sulfuric acid mist can be removed by controlling the filter temperature.
The acid dewpoint temperature, shown in Figure 1, is a function of ^SO^
concentration and temperature. At temperatures below 212°F (100°C), the
concentration of acid as a vapor is below 0.1 ppm. Oxides of sulfur (S02
and SO^) and water vapor will pass through a filter element; however
sulfuric acid will condense at the dewpoint temperature.
When high boiling organic compounds exist in the gas stream, it is
important that removal is accomplished to prevent condensed tar-like materials
from plugging the instrument. Removal can be accomplished by condensing the
compounds on either an inert substance, such as glass beads, or a disposable
substance such as fiberglass. The adsorbing medium should be in an inert
container. To ensure that the water in the sample is maintained in the vapor
state while the high boiling compounds are removed, condensor temperatures
greater than 212°F (100°C) should be used. The inert condensing surface is
then cleaned or discarded periodically, which requires a bypass to an unused
system. High boiling compound removal can be affected between the initial and
final particulate removers.
19
-------
TABLE 9. CHEMICAL RESISTANCE OF FILTERS TO EXPECTED GAS CONSTITUENTS
IN HAZARDOUS WASTE INCINERATORS
Chemical
Acid
HC1
HC1
HN02
H2S04
H2S04
Gases
NH3
N02
S02
H2S
C12
C12
03
Cone.
(%)
1
io
5
1
5
10
1
5
10
1
100
1
Max
temp.
°C.
50
50
80
80
80
80
80
80
80
80
80
80
Filter
SS
NR
NR
R
NR
NR
R
R
F
R
NR
NR
R
AN
NR
NR
R
F
F
R
R
R
R
NR
NR
R
housing
AF
R
R
R
R
R
R
R
R
R
R
F
F
PP
R
R
R
R
R
R
R
R
R
R
R
R
Filter
elements
SS
NR
NR
R
NR
NR
R
R
F
R
NR
NR
R
PP
R
R
R
R
R
R
R
R
R
R
R
R
SG
NR
NR
R
NR
NR
R
R
R
R
NR
NR
R
EP
E
E
R
R
R
R
R
R
R
R
F
F
Seals
W
R
R
R
R
R
R
R
R
R
R
R
R
TFE
E
E
E
E
E
E
E
E
E
E
E
E
SS - Stainless Steel
AN - Aluminum - Nickel
AF - Aluminum - Teflon
PP - Polypropylene
SG - Stainless Steel and Fiberglass
EP - Ethylene - Propylene Rubber
VV - Viton - Rubber
TFE - Teflon
NR - Not Recommended
P - Poor
F - Fair
R - Recommended
E - Excellent
20
-------
TEMPERATURE, °F
— M i\* l\» f\> IV u '
0 2! w w >4 tf = i
o o o o o o o <
,X
1 III
X
X
1 111
X
s
\ III
/
/
1 III
o.oi ai i.o 10
S03 (H2S04)IN GAS.ppm
Figure 1. Plot of H2SO, vapor versus temperature.
21
-------
Acid vapors and mists adversely affect the lifetime of the majority
presently available organic species instrumentation. Corrosion, cracking,
etching, salt formation and deposition can all lead to premature instrument
failure. The anticipated high levels of inorganic acid gases should be
reduced in the conditioning system without removal of organic acids and other
organic compounds of interest. Removing acid gases is possible by bubbling
the gas through a caustic solution, by lowering the temperature to well below
the acid gas dew point or by using a selective membrane or a dry alkaline
absorbent. The effects of these treatments on the integrity of the organic
species of interest will need to be fully documented. Determining which of
the available methods is suitable for organic species monitoring is dependent
in part on the organic species of interest. ^^-04 hydrocarbons are not
appreciably soluble in water or basic solutions and do not condense at ambient
temperatures. A simple water bubbler containing sodium hydroxide could be
used to neutralize acid gases when analyzing for these hydrocarbons. Organic
acids, aromatics and other soluble or acidic organic species would be trapped
in the caustic solution, therefore, this method would not be applicable if
these compounds were to be analyzed.
Ion exchange resins which are specific for inorganic acids have been
developed. The use of these resins for continuous monitoring applications has
not been well documented and the quantitative recovery of organic species
should be determined prior to use. No commercially available conditioning
systems for organic instrumentation using ion exchange columns were found. A
variation of this, the use of precolumns in a chromatograph, has been in wide
usage in hydrocarbon continuous analyzers where a timed backflush valve is
used. After a sample is injected into the GC, a timer is used to activate a
backflush of the precolumn after elution of the species of interest. This
conditioning method of sorts prevents other species from entering the detector
and unnecessarily lengthening the analysis time. This same theory may be used
to remove unwanted inorganic gases and acids if the appropriate materials are
selected.
A semi-continuous conditioning method for organics using Tenax, XAD or
similar adsorbant resins has been developed primarily to concentrate a sample
for GC/MS analysis. By automating a typical manual sampling method, a
conditioning method has been developed wherein the sample is passed through
one to two or more adsorbant tubes for a preset length of time; a fresh trap
is then cycled into place while desorbing the first trap of organic material.
The resin trap organic materials while allowing inorganic gases (acids) to
pass through. This conditioning method mimics several EPA methods involving
condensation traps with adsorbants, most notably Level 1 SASS measurements.
This method has not been used in continuous monitoring applications, but is a
promising area for future research. The effect of HC1 concentration on the
adsorption efficiency of the resin should be investigated since the
acidification of the resin may prevent collection of neutral or alkaline
organic species. Cycle times are available that allow 2 minute to 2 hour
adsorption times with thermal desorption times on the order of 1 minute.
Solvent desorption systems are available. This purge and trap procedure would
be applicable for parts per trillion sensitivity siuce no continuous
monitoring instrumentation is presently available for use in this
concentration range. Analysis of the concentrated sample requires a trade-off
22
-------
with time, the greater adsorption times providing a higher concentration
sample with decreasing levels of "real time."
MOISTURE REMOVAL
Incinerator stack gases can contain significant quantities of water
vapor. For some analyzers which are insensitive to water vapor and can
operate at higher temperatures (e.g. the UV instruments) the problems of
moisture removal can be avoided by keeping the temperature of the sample above
the dewpoint. This requires heated sample lines, heated filters, and possibly
a heated sample pump. Use of a totally heated system appears to be
inconsistent with the criteria for a minimum design.
For analyzers which cannot operate at very high temperatures, moisture
removal is a necessity. There are a variety of methods for moisture handling
and removal. The methods are condensation, permeation distillation, membrane
separation, and air dilution.
Condensation
Moisture removal is most readily accomplished by allowing the sample to
cool. Refrigerated condensers are available for cooling sample gas streams.
A typical condensation system design is shown in Figure 2. This design
requires a primary and secondary condenser to decrease the moisture level to
less than 1 percent (v/v). In the primary condenser, the moisture level is
reduced to approximately 2 to 3 percent. Because this condenser is at an
absolute pressure that is less than the ambient pressure, the condensation
process is not as efficient as it would if the condenser pressure were greater
than ambient pressure. Final moisture removal is accomplished by passing the
pump effluent under pressure through the secondary condenser. The increased
pressure enhances the condensation process by raising the vapor pressure of
moisture and increasing the residence time in the coil. Sample gas existing
the secondary condenser typically contains moisture levels approaching
0 percent.
Permeation Distillation
Permeation dryers have been used with success in drying combustion stack
gases.^ The dryer uses a semipermeable membrane as a water removal system,
and a dry gas for continuous regeneration. Specifically, the dryer is a
bundle of membrane tubes with a common header in a shell and tube
configuration. Openings adjacent to the sample inlet and product outlet
(Figure 3) are provided for the dry gas. The wet gas sample flows through the
tubes, and the dry gas flows countercurrent to the shell which permits water
vapor to be purged and transported through the tubing. Water diffuses to the
dry side of the bundle by a simple diffusion mechanism where the natural
tendency is to equalize concentrations on both sides of the membrane. The
efficiency of the dryer at constant temperature and humidity is based on the
amount of tubing in the shell and the sample and purge flow pressures.
23
-------
s
3
to
UJ
en
tc
UJ
N
o
•H
Ift
IO
2 *
Z UJ
or
o
§
Eo
0.0
CO
0)
T3
O
a
CO
3
1
0)
0)
t-l
3
O
2
0)
S-i
3
60
o r to
K d >
U. UL. tO
-------
UJ
x
(O
UJ
oc
0.0.
tltltl
t
t
4
4
4
t
t
4
4
4
t
t
I
f
4
4
Is
(O Ul
(O (9
UJ K
o: 2
Q. o.
UJ
-I
UJ O
_|Z
o) =
(8
CU
0)
o.
(3
U
cx
>.
H
OJ
S-l
60
25
-------
Membrane Separation
Membrane separation by reverse osmosis is a newly applied gas/liquid
separation technology. A membrane that is selective to certain components of
a gas or liquid stream will allow passage through while one or more other
components cannot do so. The semipermeable membrane is sometimes used with
pressure to reverse the natural tendency of osmosis to equalize
concentrations. The flow of material from the least concentrated to the more
concentrated phase or the tendency to dilute concentration is reverse
osmosis. Research is underway to separate CC^ from methane by this
method.5 other gases could be selectively permeated.
Air Dilution
The sample can be diluted in the probe with a clean dry non-reactive gas
in order to overcome instrument range problems or to minimize corrosion. This
will generally be dry instrument air which contains virtually none of the
contaminants to be measured. The dilution ratio selected at the probe is
generally greater than 70:1 to reduce the dewpoint of the sample for transport
in unheated tubing to the ambient air analyzers. With ambient air as the
diluent, SC>2, CO, NO, N02, Hydrocarbons, and C02 can be measured.
Calibrations are generally made using dilution air as zero air.
Dilution ratios must be selected by calculation using the dilution range
desired and the measurement range of the analyzer. Typically, a dilution
system has dilution ratios ranging from 12:1 to 22,500:1 or larger with a
double dilution technique. The analyzers, however, should have at least 20:1
rangeability or larger. A vacuum is drawn on a stack gas sample continuously
by an ejector pump, sometimes mounted inside the stack probe. The instrument
air stream (pressurized air) with an adjustable flow creates a vacuum in the
space between the primary and secondary nozzle of the ejector pump. The
vacuum is used to transport the stack gas through a critical orifice also
sometimes mounted inside the stack probe. The critical orifice defines the
sample flow at all pressures below the critical pressure. Control flow will
be reached when the vacuum in the ejector pump ranges from 0 to 0.34 bar (0 to
35 cm Hg). Within certain limits the flow through the critical orifice
depends on the pressure in the stack. Flow also depends on the viscosity of
gas which will vary with process conditions and moisture content. A sample
pressure variation of 10 mbar typically changes the flow of diluted sample by
1 percent. As the flow of dilution air must be adjusted by the operator, the
dilution ratio can be varied between certain limits. The diluted sample
stream is directed to an emission monitor through a tube in the umbilical
cable.
Wet Measurements
Not removing the water contained in the gas is desirable when analyzing
for easily condensed or water soluble components. There are two methods for
keeping water in the vapor phase in a gas stream: heating and pressure
reduction. Heating a stream above 212° (100°C) will keep water in the vapor
phases if the total pressure is near atmospheric. By dropping the pressure of
the gas stream, a lower condensation temperature can be realized. Figure 4
26
-------
1000
100
X
E
E
^
UJ
CO
cn
UJ
tr
0.
UJ
CO
CO
10
I
0 20 40 60 80 100
MINIMUM TEMPERATURE TO KEEP WATER IN VAPOR PHASE , °C
Figure 4. Condensation pressure.
27
-------
shows che relationship for temperature required to keep water in the vapor
phase as a function of pressure. The application of this method is to heat
the gaseous sample above the dewpoint until the pressure can be reduced to a
level where condensation will not occur at room temperature. Once the
appropriate pressure is achieved, no further heating would be required.
Applications and Problems
Some commercially available systems maintain the sample above the
dewpoint. This is done primarily to avoid solubility losses of the species
being measured. Some instruments in particular are analytically sensitive to
water vapor. This sensitivity can be circumvented by the use of optical
filters within the analyzer, by removal of water vapor, or by the use of
constant water concentration in both the sample and calibration gases. It is
important to determine the amount of water vapor that must be removed in order
to operate within the accuracy limits of the analyzers. The effect of water
vapor will increase as lower concentrations of S02 or NOX are measured by
an NDIR instrument. A typical water sensitivity for an NDIR instrument is on
the order of 20,000 to 1. That is, 20,000 ppra water registers as 1 ppm
pollutant on the analyzer. The use of a constant moisture level in both the
sample and calibration gases is a method of minimizing the effects of water
vapor. This is done as simply by condensation or the use of desiccants.
There is a possibility of interactions with constituents of the sample stream
other than water. Some loss of S02 on dessicant materials has been
observed.6 Dessicants are not considered as a viable water removal
mechanism for most analyzers.
Permeation distillation has been used with some success as mentioned
previously. Some advantages for permeation dryers are listed:
• materials of construction do not come into contact with liquid
condensate and therefore are less prone to corrosion;
• a condensate trap is not required for removal of liquid;
• there is little possibility of sample loss by solution in liquid
condensate.
Some disadvantages are:
• particulate plugging;
• a vacuum pump and controls must be maintained on the shell side of
the dryer or
• a regulated supply of dry purge air must be used.
If the dryer is located in the instrument house, the entire sample line must
be kept above the sample dewpoint and if it is located in the vicinity of the
stack, the heat tracing problem is reduced but the accessories and controls
(vacuum pump, pressure gauge, flow meter and valve) would be in a less
accessible location. Long sample lines could be run to and from the dryer but
these are inconvenient.
28
-------
Overriding advantages for condensation systems include the simplicity of
operation and minimal need for heat traced sample lines. The disadvantages of
the condensation method include:
• corrosion on condenser parts;
• solubility losses of constituents from gas stream; and
• response time lags due to the large volume of the condensate trap;
• refrigeration compressors are expensive to purchase, operate and
maintain.
The air dilution sampling system is a particularly effective method of
sampling hot, dusty, and corrosive stack gases. This systems advantages
include:
• condensation in the sample line is avoided;
• quick cooling and dilution of the sample avoids reactions of sample
components;
• no moisture is removed by refrigeration or semipermeable membranes;
• the system has no moving parts,
• no additional heating is required;
• if required, a second diluter unit can be incorporated in the sample
stream for greater dilution flexibility.
Disadvantages of air dilution include capillary plugging problems, even after
fine filtration.
Several methods of removing excess moisture have been discussed. A wide
range of boiling points of organic species could be present in the sample gas
stream, therefore, moisture removal mechanisms that lower the gas stream
temperature may also remove organic species. If only lower boiling,
relatively insoluble organic compounds (0^-04) are to be monitored,
moisture removal by any of the methods presented earlier would not seriously
affect the integrity of the sample. If easily condensed organic compounds are
to be analyzed, dilution methods and heating the gas sample above the dew
point are methods of choice. The sample temperature must always be above the
boiling point of the compound of interest. Table 10 provides example
hazardous compounds classified by boiling point and Table 11 provides a
listing of compound classified by solubility in water. From these two tables,
one can predict some of the compounds that may be prematurely removed through
condensation of moisture or temperature changes. Methods that involve no
removal of water hence no inadvertent loss of condensed organic material would
be preferred. No organic species analyzers are sensitive enough to allow for
substantial dilution of the sample, however, several are insensitive or can be
desensitized to substantial quantities of water vapor. Conditioning systems
29
-------
TABLE 10. BOILING POINT RANGES OF SELECTED HAZARDOUS ORGANIC COMPOUNDS
Boiling point range
Low boilers
(60-212°F)
Medium boilers
(212-400°F)
High boilers
(400-600°F)
Acetone
Acetyl chloride
Chloroform
C -C, alkanes
1,2-dimetyhlhydrazine
2-nitropropane
Tetrachloromethane
Octane
Ethyl benzene
Toluene
1,1,2-trichloroethane
Chlorobenzene
Trichlorohydrin
Phenol
Dodecane
Napthalene
Fluorene
Trichlorobenzene
Benzothiophene
Biphenyl
Diphenyl sulfide
TABLE 11. WATER SOLUBILITIES OF SELECTED HAZARDOUS ORGANIC COMPOUNDS
Insoluble
Soluble
Highly soluble
Miscible
Acetaldehyde
Analine
Chlorobenzene
Acetophenone
Benzyl chloride
Dieldrin
Ethylmethacrylate
2-chloronaphthalene Lead acetate
di-n-butyl pthalate Paraldehyde
Acrylamide
n-butyl alcohol
Chloroform
Acetone
Dimethylamine
1,2 dimethyl hydrazine
Cyclohexanone Chresylic acid Malononitrile
1,2 dichloropropane Pyridine
Ethylenediamine Acrilic acid
30
-------
consisting of no moisture removal mechanisms and only heat traced lines (to
400°F) have been used routinely with flame ionization detectors. Conditioners
using a membrane drying system have also been successfully used since once an
equilibrium level has been reached, normal alkanes are not removed by the
membrane dryer either through filtration or surface attractions. Moisture
condensation, which activates the corrosiveness of HC1 gases, must be
prevented in order to maintain the analyzer in good order and to prevent loss
of soluble compounds.
Connective Line
Gas transport or connective tubing requires certain considerations.
These are as follows:
1. Tube interior-exterior diameter
2. Heat resistance
3. Chemical resistance to gases being sampled
4. Cost
Connective tubing must be sized to ensure an adequate gas flow rate and
linear velocity with a reasonable pressure drop to provide good system
response times. A flow rate of 2 slpra (enough to supply two gas analyzers)
through 1/4 inch tubing exhibits a pressure drop between 1 and 3 mmHg per
30 meter length (depending on wall thickness). This pressure drop is quite
acceptable for most sampling pumps.
Teflon and stainless steel tubing exhibit corrosion and heat resistance
in addition to being chemically resistant to stack gases and acid mist. The
corrosion resistance of stainless steel is enhanced by keeping gases above the
dewpoint.7 Heat traced stainless steel and teflon tubing is commercially
available from several suppliers. Copper or brass could be specified for
temperatures up to slightly over 400°F, providing the gas is not highly
corrosive and will not be significantly affected by potential corrosion
products. Aluminum has a temperature limit of about 500°F. Carbon steel
tubing, generally annealed for ready bending, can go to 700°F or even 800°F
when corrosion is not a problem. For higher temperatures, either type 304SS
or 316SS is suitable. With carbon steel, for temperatures above 800°F,
carburization can occur. At temperatures above 900°F carburization can occur
in stainless steels. For highly corrosive, but moderate temperature
conditions, plastic tubings and fittings are effective. Teflon, polyethylene,
plasticized PVC, and nylon are available with either corrosion-resistant-metal
or plastic fittings and are useful up to about 350°F. Connections for pipe
may be threaded with either ordinary or modified threads to resolve leakage.
Brazing and welding can be used, but crevices can form causing flow
difficulties and erosion from abrasive particulate laden gases. Several
designs of all-metal tubing connectors are available and in common use. These
fittings utilize ferrules which are driven tightly against the tube wall to
seal the connection. Fittings are available in 316 ss and brass.
31
-------
As summarized in Tables 7 and 9, stainless steel is a good material for
in-stack tubing. The middle grade, type 304, is used for a wide range of
slightly corrosive materials. Type 316, molybdenum containing steel, is used
for process chemicals, acids, bleaches and more corrosive materials.
Carpenter 20 is the better of the stainless steels used for boiling 112804
and the most severely corrosive materials. Other alloys that could be used in
the pre-filtration area include Hastalloy C, Incoloy 825, and Inconel 625.
The stainless steels and alloys are used primarily before the water removal
and particulate removal systems.
Teflon is recommended for out-of-stack heat traced lines. When
moisture-laden gases are below the dewpoint, teflon, Carpenter 20 or
Hastalloy C, should be used.
Dry sample gas is generally compatible with 316 stainless steel because
the corrosivity of the dry gas is low. (See Tables 7 and 9.) Polyethylene
and polypropylene are economical choices for dry sample gas maintained above
the freezing point without heat tracing. These plastics are not compatible
with nitric acid over the long term.
Materials of construction can interact with the sample by catalytic
reaction, bulk absorption or surface adsorption. These factors are considered
in detail in the section entitled Sample Integrity. No losses by reaction are
anticipated for any of the materals listed in Table 7. Absorption and
adsorption by the walls of the system must necessarily be transient phenomena
since saturation will eventually be reached and, at steady state, the correct
concentrations will be measured. Sorption (both absorption and adsorption) by
the surface walls will manifest in a slower system response to changes in
concentration at the probe tip. Experimental measurements (Figure 5) of
system response using sample lines of various materials indicate that
adsorption and absorption are negligible for 316 SS, Teflon and polypropylene;
are moderate for polyethylene; and are large for Tygon.
Alloys and metals are resistant to abrasion by particulate laden gas
streams. Valves, air movers, and probes are subject to abrasion. If heating
components fail, water vapor may condense to form an acidic or basic slurry.
Components located between the initial and final particulate removers and the
primary water vapor removers would be more subject to abrasive effects. High
velocity particulate streams tend to remove metallic oxide coatings thereby
increasing corrosion potential of some metals.
The heat resistance of various materials is illustrated in Table 12.
Typical post control device stack temperatures (Tables 4, 5 and 6) are below
the upper temperature limit for Teflon and plastics. Less heat resistant
plastics cannot be used in the vicinity of the probe when sampling in the
combustion zone.
Sample Integrity
Regardless of the exact design of the sampling interface, it is essential
that the sample be transported from stack to analyzer with minimal losses and
interactions. There are several mechanisms as mentioned by which interaction
32
-------
cfl
•H
t-l
OJ
a)
C
•H
(U
iH
a.
CO
o
(8
•H
CO
C
O
U
Cfl
CU
4J
c
a.
o
CO
to
0)
3
•H
33
-------
TABLE 12. MAXIMUM CONTINUOUS OPERATING TEMPERATURES
FOR GEM MATERIALS OF CONSTRUCTION
Material
Maximum Temp. (°C)
Plastics
Teflon
Viton
Polyethylene*
Polypropylene
PVC
Tygon*
Stainless Steels
Carpenter 20
316 SS
304 SS
Nickel Alloys
Hastelloy C-276
Inconel 625
Incoloy 800
Incoloy 825
Non Metallics
Aluminum silicate
Quartz glass (fused silica)
Ceramic
Zirconium oxide
Pure Element
Titanium
250
150
80-125.6
110
110
60-82.2
871
870
788
1,038
980
760
704
1,540
900-1200*
1094-1538*
2, 204
800-1000
*Depending on type used.
-------
can occur, including reaction, absorption, adsorption, dilution. Unquantified
dilution can occur by air leakage into portions of the system which are under
vacuum. Gas phase species can be lost both by homogeneous gas phase reaction
and by heterogeneous catalytic reaction on system components or collected
particulates.
35
-------
SECTION 5
MEASUREMENT TECHNIQUES
INORGANIC COMPOUNDS
Inorganic gaseous analyzers can be characterized by detection principles
and grouped into the following categories: absorption spectrophotometers,
emission analyzers, electroanalytical methods, conductimetrie methods, and
coulometric methods. Nondispersive infrared (NDIR) techniques can be used for
S02, NO, CO, and C02. Chemiluminescent techniques can measure NOX.
Flame photometric techniques can be used for S02. Electrochemical
techniques can be used for S02, NOX, HC1, 02, and CO. Ultraviolet
techniques can be used for S02 and NOX. Other techniques such as gas
filter correlation and combinations of the above techniques can be used for
HC1 and CO. Still other techniques such as Fourier transform infrared
spectroscopy can measure HC1 and various other species.
Nondispersive Infrared Analyzers
Nondispersive Infrared (NDIR) analyzers have been developed to monitor
S02, NOX, CO, C02, and other gases that absorb in the infrared
wavelength region. An NDIR analyzer is basically an instrument that does not
disperse the light that is emitted from an infrared source. NDIR instruments
utilize a broad band of light that is centered at an absorption peak of the
pollutant molecule of interest. This broad band is usually selected from all
the light frequencies emitted by the infrared source by using a bandpass
filter. Table 13 gives the band centers for several of the gases of interest.
In a typical NDIR analyzer, infrared light from a lamp or glower passes
through two gas cells—a reference cell and a sample cell. The reference cell
generally contains dry nitrogen gas, which does not absorb light at the
wavelength used in the instrument. As the light passes through the sample
cell, pollutant molecules absorb a portion of the incident infrared light. As
a result, the light emerging from the end of the sample cell will have less
energy than when it entered. It also will have less energy than the light
emerging from the reference cell. The energy difference is then sensed by a
detector, such as a thermistor, a photodetector, a thermocouple, or
capacitance arrangement.
A common problem with analyzers that use infrared adsorption detectors is
that other gases that absorb in the same spectral region as the pollutant
molecule will cause a positive interference in the measurement. For example,
water vapor and C02 interfere in the measurement of CO. These gases must be
removed before the sample gas enters the analyzer. Maintaining a constant
36
-------
TABLE 13. INFARED BAND CENTERS OF SOME COMMON GASES
Location of
band centers Wave number
Gas (urn) (cm~l)
NO 5.0-5.5 1800-2000
N02 5.5-20 500-1800
S02 8-14 700-1250
H20 3.1 1000-1400
5.0-5.5 1800-2000
7.1-10 3200
CO 2.3 2200
4.6 4300
C02 2.7 850-1250
5.2 1900
8-12 3700
NH3 10.5 950
37
-------
moisture level in the sample gas and calibration gas will sometimes calibrate
out this effect. A unique solution to this problem is to put the detector
cells in series instead of in parallel. The front chamber of the detector
absorbs the infrared radiation primarily at the frequencies in the center of
an absorption band. Since the front cell takes away energy from the light
beam at the center frequencies, the rear measuring chamber will absorb more of
the energy in the outer edges of the band than from the center. The
geometries and gas concentrations of each measuring chamber are chosen so that
the pressure in each remain the same as no pollutant molecules in the sample
cell. Once pollutant molecules are introduced into the sample cell, the
amount of energy reaching the detector is reduced; however, most of this
reduction arises from absorption at the band center, and the front chamber of
the detector is affected less by the incoming radiation. The front chamber
therefore, is cooler than the rear chamber, causing a pressure difference and
a distention of the thin-metal diaphragm. This method is often called
negative filtering.
Interfering species generally do not have an absorption band that
coincides exactly with that of the species of interest. In such a case,
absorption occurs evenly over the region, and the interference is minimized.
Several monitors have been constructed utilizing this principle and need less
conditioning to remove such species as water and CC^.
The advantages of the NDIR-type analyzers are relatively low cost and the
ability to apply the method to many types of gases. Generally, a separate
instrument is required for each gas, although several instruments have
interchangeable cells (benches) and filters to provide more versatility.
The disadvantages are that NDIR instruments require more optical
maintenance, since large changes in window condition are not automatically
compensated for. Also since NDIR instruments are pressure and temperature
sensitive, care must be taken to control these parameters. The microphone
type detectors are sensitive to vibration and often require both electronic
and mechanical damping.
Nondispersive Ultraviolet Analyzers (NDUV) - Differential Absorption
Several available nondispersive systems are based on absorptions in the
ultraviolet and visible regions in the spectrum rather than in the infrared.
To analyze for SC^, these instruments utilize one of the narrow absorption
bands in the ultraviolet absorption spectrum.
NC>2 may be determined by analysis of its absorption spectrum in the
visible region. The instruments designed to operate within these regions are
somewhat different from the NDIR methods discussed previously. Essentially,
the analyzers measure the degree of absorption at a wavelength in the
absorption band of the molecule of interest, 280 nm for 862 and 436 nm for
NC>2, for example. The major difference is that a reference cell is not
used. Instead, a reference wavelength, in a region where S02 or N02 has
minimal absorption, is utilized for signal stabilization and processing. This
method of analysis is often termed differential absorption, since measurements
are performed at two different wavelengths. This method is not limited to
38
-------
extractive monitoring systems. It is also used in both in-situ analyzers and
remote sensors.
Extractive analyzers using differential absorption have proven to be
reliable in monitoring source emissions. Several of the instrument models
currently available were designed for in-plant environments. The differential
absorption SC>2 analyzers are somewhat more sensitive than the NDIR
counterparts. The sequential nature of the NOX analysis may limit the
utility of the method to cases where response times are less important. As
with some extractive monitoring systems, particulate matter should be removed
before entering the analyzer. It is not necessary, however, to remove water
vapor. A heated sample line and heated cell should be used to prevent
condensation in the analyzer. Since water does not absorb light in this
region of the ultraviolet spectrum, no spectral interferences occur.
Luminescence Analyzers
Luminescence is the emission of light which results from a molecule
returning from an elevated energy state to the ground state.
Photoluminescence is the release of light after a molecule has been excited by
ultraviolet, visible, or infrared radiation. The emission of light from a
molecule excited in a chemical reaction is known as chemiluminescence. The
atoms of a molecule can be excited to luminescence in a hydrogen flame. These
three types of luminescent processes are used in source monitoring
applications. Monitors utilizing the principle of luminescence can be
specific for given pollutant species and can have greater sensitivity than
some of the absorption or electrochemical methods in certain matrices.
Monitors that use each of these luminescent processes are discussed in this
section.
Fluorescence—
Fluorescence is a photoluminescent process in which light energy of a
given wavelength is absorbed and light energy of a different wavelength is
emitted. In this process, the molecule that is excited by the light energy
typically remains excited for about 10~"8 to 10~^ second. This period of
time is sufficient for the molecule to dissipate some of this energy in the
form of vibrational and rotational motions. When the remaining energy is
reemitted as light, the energy of the light will be lower and light of a
longer wavelength (shorter frequency) will be observed. The fluorescence
spectrum for SC>2, shown in Figure 6, illustrates this point. Commercially
available instruments contain either a continuous or a pulsed ultraviolet
light source. The light from the source is filtered to a narrow region that
is centered near 210 nm in the near ultraviolet range where the SC>2 molecule
will be excited. The fluorescent radiation is measured at right angles to the
sample chamber with a photomultiplier tube. Another band pass filter is used
to select only a portion of the fluorescent radiation for measurement, since
interferences can occur over the range of the fluorescence emission spectrum.
Fluorescence monitors have been successfully applied to monitoring source
gases as well as ambient air. Using these instruments in source monitoring
requires attention to the problem of signal quenching. In the flourescence
process,the excited SC^ molecule may collide with another molecule before it
can release its extra energy as light. The energy, instead, will be lost in
39
-------
0.3
0.2
5
TE
0.0
I I I
S02 ABSORPTION SPECTRUM
SO2 FLUORESCENCE EMISSION SPECTRUM
_ ^ ABSORPTION
BANDPASS. FLUORESCENCE
FILTER EMISSION
I \
200
250 300 350
WAVELENGTH
-------
the collision and will make the molecules move faster after the collision.
Water, C02, 02, N2, or any other molecule, for that matter, can quench
the emission of radiation. Each of these molecules has a different quenching
efficiency. If the composition of the background gases in a sample changes by
5 percent 02 and 10 percent CC>2 in a combustion gas, the S02 reading
obtained would be different from that obtained if the background gas were
ambient air containing 21 percent oxygen. The quenching effect of CC>2 is
approximately 50 percent that of oxygen. A decrease of oxygen in a flue gas
generally means a relative increase in CO2- The errors due to the
differences range from 5 to 10 percent of the S02 concentration. The S02
values can be corrected by knowing the C02 and oxygen percentages.
Fluorescence monitors, outside of the quenching problem, have few other
significant interference problems. Particulates and water must be completely
removed from the sampling stream before entering the sampling chamber or the
instrument will be easily fouled. Permeation dryers generally are used in
these instruments to eliminate any remaining water vapor that is not removed
by the extractive system.
Chemiluminescence—
Chemiluminescence is the emission of light energy that results from a
chemical reaction. It was found in the late 1960's that the reaction of NO
and ozone, 03, will produce infrared radiation from about 500 to 3000 nm.
Figure 7 shows the emission spectrum observed in this reaction. Monitors that
measure NO concentrations by observing the chemiluminescent radiation select
only a narrow region of the total emission. A filter is used to select light
in the region from about 600 to 900 nm. Nitrogen dioxide (N02) does not
chetniluminesce and must be reduced to NO before it can be measured by this
method. Most commercial analyzers contain a coverter that catalytically
reduces N02 to NO. The NO produced is then reacted with the ozone and the
cherailuminescence is measured to give a total NO + N02 (NOX) reading.
Ozone is generated by the ultraviolet irradiation of oxygen in a quartz
tube or by spark discharge. The ozone is provided in excess to the reaction
chamber to ensure complete reaction and to avoid quenching effects. Since the
photomultiplier signal is proportional to the number of NO molecules, not to
the NO concentration, the sample flow rate must be carefully controlled. The
N02 to NO converter chamber is generally made of stainless steel or
molybdenum to effect the catalytic decomposition. A few monitors on the
market will allow switching the sample gas automatically in and out of the
converter to give almost continuous readings for both NO and N02-
The Chemiluminescence method has been proven reliable and is the most
widely used method for source NO/NOX analysis. Molecules, such as 02,
N2, and C02, quench the light radiation of this cherailuminescent reaction
as in the fluorescence measurement technique. The quenching problem has been
minimized by (l) choosing a flow rate of ozone into the sample chamber much
greater than that of the sample flow rate and (2) using high vacuums to
minimize molecular collisions. The ozone dilution gives a relatively constant
background gas composition and the effects caused by different quenching
efficiencies of different molecules are minimized. The only serious
interference is ammonia, which will oxidize to NO in stainless steel converter
41
-------
CO
LU
BANDPASS FILTER
100-
co
CO
LU
50-
400
1200 2000 2800
WAVELENGTH nm
Figure 7. The chemiluminescent emission spectrum of N0«.
42
-------
chambers. This is not usually a problem when the monitor is used on a
traditional combustion source, however, many hazardous wastes could hydrolyze
to form ammonia. Molybdenum converters operated at lower temperatures will
not oxidize ammonia.
Photoluminescence (Flame Photometric)—
Another luminescence technique used to detect gaseous pollutants is that
of flame photometry. Flame photometric analyzers are primarily used in
ambient air sampling, but have been applied to stationary source sampling by
using sample dilution techniques.
Flame photometry is a branch of spectrochemical analysis in which a
sample is excited to luminescence by introduction into a flame. Instead of
using an ultraviolet or visible light source to excite the S02 molecule, as
in photoluminescence, a hydrogen flame is used to excite the sulfur atom. The
excited atom will in turn, emit light in a band of wavelengths centered at
about 394 nm, which is then detected by a photomultiplier tube. The method is
specific to elemental sulfur, not to sulfur dioxide. Compounds, such as
H2S, 863, and mercaptans, will contribute to the ultraviolet emission to
give a measure of the total sulfur content of the sample stream. With the use
of scrubbers or chromatographic techniques, selective determinations are made
for each of these sulfur compounds.
A disadvantage of flame photometric analyzers is the narrow concentration
ranges available and the requirement of hydrogen for the flame. Facilities
that have strict regulations concerning the use of hydrogen and hydrogen
cylinders may find it inconvenient to utilize this method. There are
currently only a few manufacturers of source-level flame photometric analyzers
capable of detecting sulfur compounds over the expected concentration ranges.
Electronanalytical Methods
Most of the instruments discussed in this report rely on spectroscopic or
electro-optical techniques to monitor gases. Another class of instruments
based upon electroanalytical methods has found great utility in source
monitoring applications. A number of monitors based on polarographic and
electrocatalytic methods are available for source monitoring applications.
Polarographic analyzers have been developed for a number of gases and can be
inexpensive and portable, therefore ideal for inspection work. Complete
continuous source-monitoring systems also are available from manufacturers of
these instruments. The electrocatalytic or high temperature fuel-cell method,
as it is often called, is only used to monitor oxygen. Both extractive and
in-stack monitors are available using this technique. The method of
conductivity is less widely used and is subject to a number of interferences.
Descriptions of this method are given here, since a few instruments employing
this principle are still marketed.
The principles behind the polarographic and electrocatalytic methods are
somewhat more difficult to understand than the spectroscopic principles
discussed earlier. A combination of classical electrochemistry and modern
fuel-cell technology has provided the theoretical basis for these
developments. An understanding of the underlying operational principles,
however, is important for their evaluation.
43
-------
Polarographic Analyzers—
Polarographic analyzers have been called voltametrie analyzers or
electrochemical transducers. With the proper choice of electrodes and
electrolytes, instruments have been developed utilizing the principles of
polargraphy to monitor S02, N02» CO, 02, l^S, and other gases.
The transducer in these instruments is generally a self-contained
electrochemical cell in which a chemical reaction takes place involving the
pollutant molecule* Two basic techniques are used in the transducer: (1) the
utilization of a selective semipermeable membrane that allows the pollutant
molecule to diffuse to an electrolytic solution, and (2) the measurement of
the current change produced at an electrode by the oxidation or reduction of
the dissolved gas at the electrode.
The generation of electrons at the sensing electrode produces an electric
current that can be measured. There are two reasons why this type of system
may be termed polarographic or voltametric. In typical polarographic analyzers
used in chemical laboratories, the electric current in the system is related
to the rate of diffusion of the reacting species to the sensing electrode. It
turns out that if the rate of which the reactant reaches the sensing electrode
is diffusion controlled, the current is directly proportional to the
concentration of reactant.
The cells come in a number of configurations, depending upon the
manufacturer. Various claims are made about the response and selectivity of
the instrument related to the cell design. These systems are small and
portable and compared to practically all other source monitoring instruments,
are the least expensive. These two factors make them ideal for source
inspection, as warning detectors
The polargraphic analyzers in the earlier development stages were
temperature sensitive, but temperature compensation devices are now provided
to avoid this problem. The cell electrolytes generally are used up in three
to six months of continuous use. The cells can be sent back to the company
and recharged or new ones can be purchased. It is extremely important that
the sample gas be conditioned before entering these analyzers. The stack gas
should be at ambient temperature and the particulate matter and water vapor
removed to avoid fouling of the cell membrane.
Electrocatalytic Analyzers For Oxygen—
A new method for the determination of oxygen has been developed over the
past several years as an outgrowth of fuel-cell technology. These so-called
fuel-cell oxygen analyzers are not actually fuel cells, but simply
electrolytic concentration cells that use a special solid catalytic
electrolyte to aid the flow of electrons. These analyzers are available in
both extractive and in-situ (in-stack) configurations. This versatility of
design is making them popular for monitoring dilute oxygen concentrations in
combustion sources.
In basic electrochemistry, one of the common phenomena studied is the
flow of electrons that can result when two solutions of different
concentrations are connected together. The electron flow results from the
-------
fact chat the chemical potential is different on each side and that
equilibrium needs to be reached. The electromotive force (EMF), or output
voltage, that results from a concentration cell is described by the Nernst
equation.
The instruments designed to continuously monitor oxygen concentrations
utilize different concentrations of oxygen gas expressed in terms of partial
pressures. A special porous material, zirconium oxide, serves both as an
electrolyte and as a high temperature catalyst to produce oxygen ions. A
schematic of the electrocatalytic sensing system is shown in Figure 8.
When sampling combustion gases, the partial pressure of the oxygen in the
sample side will be lower than the partial pressure of oxygen in the reference
side, which is generally that of air. When such a cell is kept at a
temperature of about 850°C, oxygen molecules on the reference side will pick
up electrons at the electrode-electrolyte interface. The porous ceramic
material of ZrC>2 has the special property of high conductivity for oxygen
ions. This occurs because the metal ions form a perfect crystal lattice in
the material, whereas the oxygen ions do not, resulting in vacancies. Heating
the zirconium oxide causes the vacancies and oxygen ions to move about. The
oxygen ions migrate to the electrode on the sample side of the cell, release
electrons to the electrode, and emerge as oxygen molecules.
A number of manufacturers are presently marketing oxygen analyzers. Both
extractive and in-situ type systems have been developed, providing the source
operator with versatility in application. A supply of clean air for the
reference side of the cell is required. Calibration gases can be injected
into the measuring cavity contained within the thimble to check instrument
operation.
Paramagnetic Analyzers For Oxygen
Molecules behave in different ways when placed in a magnetic field. This
magnetic exhibited behavior is either diamagnetic or paramagnetic. Most
materials are paramagnetic; they are attracted by a magnetic field.
Paramagnetism arises when a molecule has one or more electrons spinning in the
same direction. Most materials will have paired electrons; the same number of
electrons spinning counterclockwise as spinning clockwise. Oxygen, however,
has two unpaired electrons that spin in the same direction. These two
electrons give the oxygen molecule a permanent magnetic moment. When an
oxygen molecule is placed near a magnetic field, the molecule is drawn to the
field and the magnetic moments of the electrons become aligned with it.
There are three methods of applying the paramagnetic properties of oxygen
as a measurement technique in commercial analyzers. These are the magnetic
wind or thermomagnetic methods, the magnetodynamic and the magnetopneumatic
methods.
Magnetic Wind Instruments (Thermomagnetic)—
The magnetic wind instruments are based on the principle that
paramagnetic attraction of the oxygen molecule decreases as the temperature
increases. A typical analyzer utilizes a cross-tube wound with filament wire
heated to 200°C.
45
-------
POROUS
ELECTRODE
Zr02 POROUS ELECTROLYTE | ELECTRON CURRENT
PREF(02) > PSAMPLE (02)
©
Figure 8. Schematic of an electrocatalytic sampling system.
46
-------
A strong magnetic field covers one half of the coil. Oxygen contained in
the sample gas is attracted to the applied field and enters the cross-tube.
The oxygen then heats up and its paramagnetic susceptibility is reduced. This
heated oxygen then is pushed out by the colder gas just entering the
cross-tube. A wind or flow of gas, therefore, continuously passes through the
cross-tube. This gas flow, however, effectively cools the heated filament
coil and changes its resistance. The change in resistance detected in the
Wheatstone bridge circuit is proportional to the oxygen concentration.
Several problems can arise in the thermomagnetic method. The cross-tube
filament temperature can be affected by changes in the thermal conductivity of
the carrier gas. The gas composition should be relatively stable if
consistent results are desired. Also, unburned hydrocarbons or other
combustible materials may react or combust within the heated filaments and
change their resistance.
Magneto-Dynamic Instruments—
The magneto-dynamic method utilizes the paramagnetic property of the
oxygen molecule by suspending a specially constructed torsion balance in a
magnetic field. Here, a dumbbell will be slightly repelled from the magnetic
field. When a sample containing oxygen is added, the magnet attracts the
oxygen and field lines surrounding the dumbbell are changed. The dumbbell
swings to realign itself with the new field. Light reflected from a small
mirror placed on the dumbbell is used to indicate the degree of swing of the
dumbbell, and hence, the oxygen concentration.
Magnetopneumatic Instruments—
The magnetopneumatic method utilizes the paramagnetic property of the
oxygen molecule by introducing oxygen into a dis proportional magnetic field.
The paramagnetic gaseous substance is attracted to the magnetic field to cause
the pressure in that field to rise. The pressure elevation is picked out of
the magnetic field by using a nonpararaagnetic gas (nitrogen). The
electromagnet is alternately magnetized and the pressure change is converted
into an electrical signal by a condenser microphone.
All of the commercial paramagnetic analyzers are extractive systems.
Water and particulate matter have to be removed before the sample enters the
monitoring system. It should be noted that NO and N02 are also paramagnetic
and may cause some interferences in the monitoring method if high
concentrations are present.
Gas Filter Correlation Analyzers
The technique of gas filter correlation (GFC) spectroscopy offers
improved specificity over conventional non-dispersive infrared (NDIR)
techniques. GFC spectroscopy is based upon comparison of the detailed
structure of the infrared absorption spectrum of the measured gas to that of
other gases also present in the sample being analyzed. The technique is
implemented by using the measured gas itself, in high concentration, as a
filter for the infrared radiation transmitted through the analyzer.
47
-------
The basic components of a GFC spectrometer are shown in Figure 9.
Radiation from an IR source is chopped and then passed through a gas filter
which alternates between the reference gas and N2 due to rotation of the
filter wheel. The radiation then enters a multiple optical pass sample cell
where absorption by the sample gas occurs. The IR radiation exits the sample
cell and passes through a narrow band pass interference filter to an IR
detector. Gas filter correlation techniques are presently used for continuous
measurement of CO and HC1, but could be used for continuous measurement of NO,
and S02.
Second-Derivative Absorption Analyzers
The second-derivative measurement technique provides a very specific
measurement of compounds with narrow band spectral absorption. Light with a
modulated wavelength, is projected through the same gas, or measurement
cavity. If the gas to be detected is present, the intensity of the detected
light varies at twice the modulation frequency. The resulting signal is
processed to determine the second derivatives (d2) value which is directly
proportional to the concentration of the gas. Since the technique is
sensitive to the curvature of the spectra and not the intensity of the
spectra, it provides enhanced sensitivity and permits measurement of a single
gas in complex mixtures.
Fourier Transform Infrared Spectroscopy Analyzers (FTS-IR)
Although relatively new on the market for process gas streams, fourier
transform infrared spectroscopy can be used for pollution measurements. The
FTS-IR uses a rapid scanning interferometer consisting of two mirrors, one
movable and one fixed, and a beam splitter. A FTS-IR schematic is shown in
Figure 10.
There are no slits, no gratings and no conventional chopper. The mirror
is the only moving part in the spectrometer. The beam splitter reflects back
50 percent and transmits 50 percent of the incoming infrared radiation. These
two beams are then reflected back towards the beamsplitter by the two
mirrors. Depending on the position of the moving mirror, these two beams
recombine at the beam splitter with a specific path difference between them.
This produces the interferogram. The interferogram is generated by the
interferometer modulating the infrared beam as the moving mirror is
translated. The modulation frequencies depend on the wavelength of the
incident radiation and the velocity of the moving mirror. The interferogram
is produced after absorption by a sample, the modulated radiation reaches the
detector where intensity is recorded as a function of path difference. The
FTS-IR is capable of measurement ranges from the near visible to less than
10 cm~l. FTS-IR systems are compatible with gas chromatograph systems and
analysis can be performed in close to real time. Other applications besides
quantitative analysis in pollution measurements are spectral subtraction,
spectral separation, far infrared analysis, emission spectroscopy and time
resolved spectroscopy.
48
-------
4J
1)
CJ
0)
o-
CD
o
•H
4J
«
VJ
SJ
O
O
0)
4J
r-i
•rl
CO
(fl
ao
49
-------
DETECTOR
V////tf////A
• Ji
p
MIRROR | i
(MOVABLE)
BEAM SPLITTER
SOURCE OF
RADIATION
Figure 10. FTS-IR Schematic.
50
-------
ORGANIC COMPOUNDS
Many varieties of organic species analyzers are available. Gas
chromatographic systems include all devices that separate organic species
through use of a packed or capillary column and measure the organic
concentration using a detector at the end of the column. Several analyzers
are hybrid chromatographs, in that, organic/inorganic separation is performed
through chemical or thermal techniques and analysis is performed by a typical
chromatographic detector, i.e., flame ionization. A third category includes
spectrometric systems which bombard compounds with light in the UV, visible or
infrared region and measure characteristic absorbtivities and emissivities.
Chromatographic Techniques
Gas chrotnatography is a common technique used for separating and
analyzing mixtures of gases and vapors. A gas mixture is percolated through a
column of porous solids or liquid coated solids which selectively retard
sample components. A carrier gas is used to bring the discreet bands to a
detector and through analysis of the detector response and the component
retention time, the sample can be identified and quantified. Gas
chromatography has been in use in the laboratory since 1905; however, it has
only recently been used in continuous monitoring applications.
Columns—
The column is the heart of the chromatograph in that the choice of the
proper column and packing materials is crucial to efficient separation and
elution of the components of interest. In some cases, only a capillary tube
with a thin liquid coating on the inside wall is used. The most common type
of GC column is a liquid coated on an inert solid support contained in a small
bore (1/8 in.) stainless steel or glass tube. The choice of the proper liquid
coating is crucial and is usually similar in chemical structure to the sample
components of interest. Organic compounds can be divided into five classes of
solute polarity ranging from most polar (Class I) to nonpolar (Class V).
Table 14 lists solute classifications for organic compounds. Compounds having
similar boiling points can be separated by choosing column materials of the
appropriate polarity. With the right choice, the order of elution of the
sample components can be manipulated. Through selection of the best solid
support/liquid phase combination, and through optimizing sample size, flow
rates, column temperatures and length, and the sensitivity of the detector, a
gas chromatograph can be adapted for almost any sample analysis situation.
Since the purpose of a hydrocarbon monitor is to measure concentration of
incomplete combustion products of the waste feed material, a chromatographic
column and detector must be chosen that is appropriately sensitive to the
types of hazardous waste being incinerated and the anticipated partial
combustion products. Eighty percent of all hazardous wastes listed in
Appendix VIII contain carbon and hydrogen and 28 percent are in the
organochlorine compound category. Of the 380 wastes listed, two thirds can be
categorized as shown in Table 15. Choosing the correct column components
(solid and liquid phases) for an incinerator that burns otily one class of
compounds is a fairly routine matter and most instrument manufacturers have
compiled recommendations based on their experience and their equipment
51
-------
TABLE 14. SOLUTE CLASSIFICATION
CLASS I (Most Polar)
Water
Glycol, glycerol, etc.
Amino alcohols
Hydroxy acids
Polyphenols
Dibasic acids
CLASS III (Intermediate)
Ethers
Ketones
Aldehydes
Esters
Tertiary amines
Nitro compounds with
no -H atoms
Nitriles with no -H atoms
CLASS II (Polar)
Alcohols
Fatty acids
Phenols
Primary and secondary amines
Oximes
Nitro compounds with -H atoms
Nitriles with -H atoms
NH3, HF, N2H4, HCN
CLASS IV (Low Polarity)
CHC13
CG2C12
CH3CHC12
CH2CLCH2C1
CH2C1CHC12 etc.
Aromatic hydrocarbons
Olefinic hydrocarbons
CLASS V (Non-Polar)
Saturated hydrocarbons
CS2
Mercaptans
Sulfides
Halocarbons not in Class
IV such as CC14
Source: Ewell, R. N., Harrison, J. M., and Berg, L., Ind.
Eng. Chera. 36, 871 (1944).
52
-------
TABLE 15. APPENDIX VIII CONSTITUENTS ACCORDING
TO ORGANIC FAMILIES
Appendix VIII
Organic family (percent)3
Aromatic 18
Straight chain paraffins 16
Polynuclear aromatics 8
Nitroso compounds 7
Straight chain olefins 5
Biphenyls 4
Hydrazines 2
Thioureas 2
Ring N aromatic 1
63
aPercent of 300 organic Appendix VIII constituents.
Source: Cudahy, J. J., Sroka, L., and Troxler, W.,
Incineration Characterization of RCRA
listed Hazardous Wastes, July 1981, EPA
Report.
53
-------
restrictions. A commercial hazardous waste incinerator which burns a wide
variety of compounds may need to have several columns available. Column
switching systems are available with two to four columns installed in a single
instrument. Table 16 provides a listing of the types of columns that have
been successfully used to analyze certain organic compound classes.
Chromatograph columns are available for separation of most types of organic
hazardous waste compounds.
An important parameter to consider is response time. A column should be
chosen that elutes the compounds of interest first and rather quickly. A
shorter column would be preferable over a longer column in order to decrease
response time. In commercially available monitors, column lengths vary from 4
inches to several feet (coiled). Shorter columns do not separate compounds as
efficiently or as completely as a longer column of the same material. Several
short columns of different materials could be arranged to give rapid detection
of several compound classes. Backflushing the columns after a designated time
can be automated and prevents unwanted compounds from reaching the detectors.
The backflush feature is important for compounds like HC1 that destroy a
sensitive flame ionzation detector.
Detectors—
Once the compounds of interest have been appropriately separated in the
column, they are swept into a detector. The following sections describe the
available detector types with any limitiations for use with the anticipated
hazardous waste streams.
Flame lonization (FID)—Most organic compounds are pyrolyzed in a
hydrogen-air flame. In the FID, the ionic intermediates produced migrate to a
detector plate which is appropriately charged. This migration is essentially
an electric current which is measurable with sensitive electronic equipment.
A chromatographic column is used to separate organic species followed by flame
ionization to detect the quantities of carbon and hydrogen in the separated
species. The electronic current produced by ionization is proportional to the
concentration of the species.
An FID does not respond to the following inorganic compounds:
He CS2 NH3
Ar COS CO
Kr H2S C02
Ne S02 H2o
Xe NO Sicl4
02 N20 SiHCl3
N2 N02 SiF4
Straight chain hydrocarbons produce the greatest FID response- An FID
responds differently to different compound classes. Equal amounts of
hydrocarbon, esters and ethers do not produce equal FID responses. The
sensitivity of an FID is also dependent on the relative flow rates of carrier
54
-------
TABLE 16. RECOMMENDED COLUMN LIQUID PHASES BY SAMPLE TYPE
Classification
of compounds
Stationary phase
ALCOHOLS
crc5
C1"C18
Di-Poly
ALDEHYDES
crc5
C5~C18
ALKALOIDS
AMINO ACID DERIVATIVES
N-Butyl trifluoroacetyl
esters
BORANES
ESSENTIAL OILS
General
ESTERS
Mixed
ETHERS
GLYCOLS
HALOGEN COMPOUNDS
Freons
Hallcoraid M-18 OL
Carbowax 600 or 1540
FFAP
Carbowax 20M
FFAP
QF-1
Ethofat
Carbowax 20M
QF-1
SE-30
DEGS/EGSS-X
Apiezon L
FFAP
Carbowax 20M
Dinonyl phthalate
Porapak Q
Carbowax 20M
Proapak Q
Carbowax 20M
QF-1 (FS-1265)
FFAP
Dibutyl Tetrachlorophthalate
UCON Polar 2000
(continued)
55
-------
TABLE 16 (continued)
Classification
of compounds
Stationary phase
HYDROCARBONS
Aliphatic
C1~C5
C5~C10
Aromatics
Hydroxy
Olefins
C1-C6
C6-up
Polynuclear
KETONES
NITROGEN COMPOUNDS
Amines
Amides
Ammonia
NITRILES
ORGANO METALLIC
PESTICIDES
Propylene Carbonate
Carbowax 400
Tributyl phosphate
Didecylphthalate
SE-30
Tetracyanoethylated
Pentaerythritol
Dibutyl tetrachlorophthalate
2,4 Xylenyl Phosphate
AgNO3/Benzyl cyanide
DimethyIsulfolane
Propylene Carbonate
Carbowax 20M
SE-30 on DMCS treated support
FFAP on DMCS treated support
PMPE (5 ring)
Lexan
FFAP
Dowfax 9N9/KOH
Versamid 900
Ethofat or Carbowax 600 on
Chromosorb T
Te tracyanoe thylated
pentaerythritol
FFAP
XF-1150
FFAP
SE-30
Dow 11
QF-1 (FS-1265)
SE-30
OV-1 OV-17
(continued)
56
-------
TABLE 16 (continued)
Classification
of compounds
Stationary phase
PHOSPHORUS
SE-30
STAP
SILANES
SF-96
FFAP
STEROIDS
SUGAR DERIVATIVES
TrimethyIsilylathers
SULFUR COMPOUNDS
STAP
XE-60
QF-1 (FS-1265)
SE-30
OV-1 OV-17
QF-1
SE-52
Carbowax 20M
FFAP
Dinonylphthalate
Porapak Q
Source: McNair, H. M., and Bonelli, E. J.
Chromatography 45, 5th Ed., 1969.
Basic Gas
57
-------
gas, hydrogen and air and upon electrode and linear jet geometry. The FID has
the widest linear range of any chromatographic detector in common use, on the
order of 107, and is sensitive to as little as 2 x 10~H gm of alkanes.'
The major problem with applying this measurement technique is corrosion
of the detector plates. HC1 gas attacks the stainless steel plates used as
grids in the detector and very quickly renders them insensitive. Several
manufacturers were contacted to determine if FID detectors were available that
would withstand attack by HC1. No detectors were found that were built to
withstand HC1 gas corrosion, however several manufacturers indicated that
detector plates could be built of Hastelloy C-276 or Inconel. The increased
cost of highly resistant detector plates should be weighed against the
replacement cost of stainless steel detector plates and the costs of
additional sample conditioning to remove HC1. At present, an FID will last
for several years with routine maintenance at HC1 gas levels less than
1000 ppm. Experience on the incinerator ship Vulcanus has shown an FID
lifetime of about two weeks at elevated HC1 concentrations (8 to 17 percent
v/v).9 Acid resistant detector plates would require only minimal
conditioning for HC1 since all other GC components are readily available with
acid resistant materials such as glass and Teflon.
Electron Capture (BCD)—The electron capture detector measures the loss
of electrical signal rather than the produced electrical current. Carrier gas
molecules are excited by a radioactive source to produce a steady background
current under a fixed applied voltage. When a sample is introduced that
absorbs electrons, the current is reduced and this reduction is indicated by
an amplifying electrometer. The source of the electron current in an ECD is
usually the radioactive decay of nickel 63 or tritium. The high energy
electrons emitted interact with the carrier gas which in turn interact with
sample constituents. A problem exists when the ionized sample molecules react
with other sample constituents to produce an electron which results in
confusing detector responses. Usually, a small amount of methane is added to
the carrier gas (argon) to deactivate further reactions by excited molecules.
However, this dilution of sample gas results in reduced sensitivity. Pure
nitrogen carrier gases have been used in laboratory studies with mixed
success. The ECD is well adapted for use in the analysis of 10~10 to
10~12 levels of pollutants incinerator effluents. The detector suffers from
the same potential for corrosion as the FID. The ECD may be operated above
350°C if the Ni63 source is used. However it is highly sensitive to
temperature changes. Baseline drifts of 50 percent have been reported for a
2°C room temperature change.H This detector is extremely sensitive to sub
picogram quantities of pesticides, alkyl halides, conjugated carbonyls,
nitrides and organoraetallics, and is quite insensitive to hydrocarbons,
alcohols and ketones. Table 17 lists the relative ECD response factors for
various organic compound classes.In summary, the ECD is highly sensitive, not
very linear, and is very sensitive to surrounding conditions, therefore
probably not well suited to the rigorous demands of continuous monitoring.
Photoionization (PIP)—The photoionization detector is similar in
principle to an FID and ECD except that an energy source in the UV-visible
wavelength region is used. Many inorganic species form ions upon exposure to
UV radiation, some organic species requiring higher UV energy levels to ionize
58
-------
TABLE 17. ELECTRON CAPTURE DETECTOR RELATIVE RESPONSE FACTORS
(BENZENE = 1)
Factor Compound or class
<1 Aliphatic hydrocarbons
1-10 Aromatic hydrocarbons
Alcohols, ketones, ethers
Monofluoro compounds
10-100 Amines, esters, aldehydes, nitriles
Monochloro compounds
Trifluoro compounds
100-10^ Dichloro aliphatic compounds (some)
Stilbenes*
Oxalates
10-^-10^ Dichloro aliphatic compounds (some)
Monobromo compounds
Hexafluoro compounds
10^-10" Dichloro aromatic compounds
Mononitro compounds
10" Monoido compounds
Dibromo compounds
Trichloro (or greater) compounds
Source: American Laboratory, Electron Capture Detectors
and their applications to Toxicology.
59
-------
than others. Formaldehyde, acetonitrile, chloroform, and carbon tetrachloride
are some of the organic species that are difficult to photoionize. Some
inorganic gases that are difficult to photoionize include hydrogen sulfide,
chlorine, and oxygen. Using the appropriate column with the appropriate
energy level UV lamp allows characterization of compound classes. Available
lamp energies range from 8 to 12 eV. The lower lamp energies provide
increased selectivity for polyaromatic compounds while the higher energy lamp
is more sensitive to aliphatic and olefinic organics. In order to identify a
variety of organic species in a given sample, three to five separate lamps
would be required along with a computer to quickly interpret the relative
responses. PID detectors are also available that do not use chromatographic
columns. Even in combination with a column, a single PID instrument would
have limited uses at a facility burning a wide variety of compounds. These
detectors are typically specified for applications where the sample matrix is
well characterized since the PID responds to both inorganic and organic
species.
Advantages of the PID include sensitivities equal to or greater than an
FID (10 to 50 times for certain organic compounds) and the fact that the
sample is not destroyed. The PID is available with all parts contacting the
sample, except the light window (quartz), coated with Teflon. Disadvantages
include reduced sensitivity from dust or coatings on the sample cell windows
and the need for several instruments or differing lamp energies to identify
and detect a complex mixture of organic species.
Other Detectors—The previously mentioned detectors - FID, ECD and PID,
are the most commonly used chromatographic detectors for air pollutant
analysis. The thermal conductivity detector (TCD) is a common detector that
is sensitive to all compounds. Since pollutants at trace levels are of
interest in hazardous waste incinerators, the other components of the sample
would pose an insurmountable selectivity problem with this analyzer. Highly
selective detectors have been developed where sensitivity to certain compounds
or elements in compounds has been enhanced. Alkali Flame lonization Detectors
(AFID) were developed to analyze pesticides. With the AFID, a small pellet of
Cs, Br, or Rb2S04 is placed at the tip of a typical FID burner which is
operated in a starved oxygen mode. The sensitivity of the detector to
phosphorous containing compounds is enhanced 5000 to 1 over normal
hydrocarbons. Different salt pellets have been used to enhance sensitivity to
different elements, the next most common application being nitrogen
compounds. The Hall Electrolytic Conductivity Detector (HECD) is particularly
sensitive to nitrogen, sulfur and halogen containing compounds and is more
discriminating than the ECD. The Thermionic Specific Detector (TSD) is also a
highly sensitive detector that is specific for nitrogen containing compounds.
These detectors usually achieve selectivity and sensitivity at the expense of
durability and reliability. Many of the more recent noncontinuous EPA
sampling and analysis methods recommend the use of these detectors. The
durability and reliability of these detectors may be improved to the point
where continuous monitoring applications would be feasible.
Several of the EPA methods also recommend using mass spectrometry (MS) in
combination with a GC. A mass spectrometer is a powerful instrument capable
of identifying compounds and structures. GC/MS and other
60
-------
instrument/instrument combinations such as GC/IR and FTIR/GC are presently in
use in the laboratory environment. GC/MS has been used in mobile monitoring
efforts, however, no continuous monitors per se were found to be in current
use. Mass spectrometers are finding some use in process monitoring in certain
applications, however not enough data is available to reasonably assess the
adaptability of these instruments and combinations of instruments to
continuous monitoring of incinerator effluents.
Dual Detectors—Combining the high sensitivity of the FID detector to
easily pyrolized, low boiling organic compounds with the high sensitivity of
ECD detectors to higher boiling aromatics, etc., is one method for more
complete quantification of organic emissions from an incinerator. One
instrument could be chosen with two columns, each specified to enhance the
separation of and response time associated with the appropriate organic
compounds. One or both detector(s) could be used at a given time, depending
on the waste being burned. The nondestructive nature of the PID detector
allows its use in series with other GC detectors. PID/FID, PID/ECD and
PID/NPD combinations have all been used for specific applications. A typical
PID/FID application is the identification of hydrocarbon classes. The PID
response increases with increasing degrees of unsaturation and the FID
response is mostly unaffected by double bonds. By comparison of the relative
response of these two detectors, alkanes, olefins and aromatic compound
classes can be identified in complex sample matrices.
Hybrid Chroma to graphs
A basic gas chromatograph consists of a series of sample separation and
detection equipment as discussed previously. Many combinations of these basic
components are possible which would allow virtually any organic compound to be
identified. Numerous columns can be used in parallel or in series to tailor
the sample separation and retention time to a particular need. Various
detectors could be used in parallel or in series to enhance sensitivity and
selectivity to compounds of interest. Since gas chromatography involves both
a separation and detection step, in some cases only one of these steps may be
needed. A variety of other instrumental methods; GC/MS, GC/IR and GC/FTS-IR
combines a chromatographic separation with an instrumental identification of
the components of the eluted materials. The combination of a GC and other
instrumentation results in more of a research oriented monitor and the capital
costs are orders of magnitude higher than the simple GC.
Many commercially available organic species analyzers are comprised of a
sample alteration system and an FID detector. Organic species may be reduced
to methane, water, or C02 which is then analyzed. Many other instruments
consist of only an FID detector with no chromatograph or other sample
treatment. Combining these two instruments provides a method for determining
methane and nonmethane hydrocarbons simultaneously. One sample is passed
directly into the flame to determine total hydrocarbons while a second sample
is passed through a stripper column to remove everything except methane and by
difference, the nonmethane hydrocarbons can be quantified. Using a selective
combustor instead of a stripper column provides an indication of total
reactive hydrocarbons. Depending on the operating temperature of the
combustor, one can convert olefinic (250 to 500°C) or paraffinic (to 1000°C)
61
-------
compounds to water and/or methane and by difference between total and reacted
detector responses, one can determine the methane concentration. With proper
sample conditioning, these dual response instruments may be well suited to
order of magnitude analysis of incinerator effluents.
GC/IR--
Combining the separation and detection power of a GC with the
identification capabilities of an IR absorption spectrometer produces a more
specific instrumental technique for identification of organic compounds. To
date, the major problem with combining these techniques has been the
development of suitable equipment to interface the two instruments. The
volume of sample from the GC is traditionally only a tiny fraction of the
volume of sample usually required by an IR instrument. Reducing the cell
volume of the IR decreases sensitivity unless additional passes through the
cell are provided. IR cells (light pipes) have been developed that are
compatible with capillary GC instruments and through the development of
computer software, GC/IR and GC/FTS-IR are gaining popularity as the "poor
mans" GC/MS. Within the next few years, this technology will be developed to
the point where it will be applicable to continuous air pollutant monitoring.
At present, only 55 toxic substances have been detected in the laboratory with
the GC/FTS-IR technique with minimum detectable quantities on the order of
1 to 10 micrograms.12
GC/MS—
A compound in a gas matrix can be more fully identified through analysis
of retention time in a GC and the mass spectrum. Identity can be established
by comparing the total ion current profile of an eluted compound to a published
standard spectrum. GC/MS techniques are particularly suited for analysis of
organics in water through a concentration step. GC/MS has been used to
identify organic ambient air contaminants. The concentration step involves
passing the air sample through an absorber column that traps the organic
material followed by thermal or solvent desorption of that material into the
GC. This technique is semicontinuous and overall response times of a GC/MS
are typically greater than 3 minutes. This powerful research tool could be
adapted to identify and quantitate organic compounds in incinerator effluents
in close to real time. At present, no GC/MS instrumentation is in routine use
as a continuous monitor. Double mass spectrometry (MS/MS) and laser
multiphoton ionization mass spectrometry have been identified as potential
on-line or real time instruments for the identification of polycyclic aromatic
hydrocarbons.13 These instruments do not use the GC for separation of
components and therefore do not involve the same delays in response time.
A disadvantage of GC/MS and MS/MS techniques is the complexity and cost
of the instrumentation. Investments costing more than $75,000 are usually
required. The mass spectra produced is complex and close to real time results
can only be provided through a computer with extensive library searching
capabilities. The MS can scan for certain compounds within seconds; however,
full spectrum scans usually take greater than 3 minutes. These disadvantages
should be weighed against the high sensitivity and resolution capabilities of
the GC/MS system.
62
-------
Radiation Emission/Absorption Instrumentation
All organic compounds absorb electromagnetic radiation because all
contain valence electrons that can be excited to higher energy levels.
Absorption measurements in the visible and ultraviolet wavelength regions are
useful for detecting the presence of certain functional groups that contain
valence electrons with relatively low excitation energies. Absorption
measurements in the infrared wavelength regions are useful for identifying
specific compounds since no two organic compounds (except for optical isomers)
have identical infrared absorption curves. The applicability of absorption
measurements to characterize incinerator effluents involves the choice of the
appropriate energies and wavelength.
A given compound normally exhibits one or two maximum absorption peaks in
the UV and visible region and may have numerous maximum and minimum
absorptivities in the IR region. Aromatic hydrocarbons absorb strongly in the
near ultraviolet region, however, aliphatic hydrocarbons absorb in the region
of 3 microns. Ideally, measurements are made at several wavelengths where the
substances of interest absorb while ofher components do not. Since all
compounds absorb radiation, these methods suffer from a lack of specificity.
Through chemical alterations and substitutions, one can identify several
compounds in a sample; however, this laboratory procedure has not been
demonstrated on a real time basis for more than one compound (phenol).
Absorption or emission in the X-ray region of the spectrum will not be
considered here since the instrumentation has not yet been applied in a
continuous emissions monitoring situation.
Radiation absorption instrumentation consists of a radiant source, a
wavelength control system, the sample chamber, a signal detector, and an
indicator. Various types of lamps are used; from a common tungsten filament
light bulb to a CC^ laser depending on the wavelength region of interest and
power requirements. Wavelength selection is accomplished most cheaply with
absorption and interference filters where a short wavelength range is
transmitted. A monochrometer can resolve radiation into its component
wavelength through use of a system of slits and lenses, prisms and/or
diffraction gratings. Sample compartments or cells must be constructed of
materials that pass radiation in the spectral region of interest with minimal
interference. Radiation detection and indication is usually accomplished
through use of photovoltaic cells, photomultiplier tubes, or photoconductive
cells. Digital, dial, and other indicators are used to provide the
quantitative results.
The following sections describe the principles of various absorption/
emission measurements related to organic compounds, the compounds for which
the method is selective and any advantages and/or limitations to using the
instruments at the three incinerator sampling locations of interest.
Infrared Absorption—
The infrared portion (IR) of the spectrum ranges from wavelengths of 2.4
to 14 microns and the near infrared (NIR) ranges from 1 to 2.5 microns. Most
organic species absorb radiation in the IR region and the concentration of
species can be determined by the nature of the transmitted radiation. The
63
-------
concentration of an absorbing compound is inversely proportional to the sample
cell path length and proportional to the energy transmitted through the sample
divided by the logarithm of the energy entering the sample (Beers law).
Sensitivity can be increased through increasing the cell path length and by
including multiple radiation passes through the sample. High radiation energy
sources such as CC>2 lasers are finding increased
The near infrared region has found increasing application to process and
pollution monitoring for several reasons. Sample cell window and lens
materials have been developed which are physically more suitable for
extra-laboratory use. Radiation source energies and detector sensitivities
are also correspondingly higher. The most common radiation source is an inert
heated solid material, commonly known as a black body generator. The Nernst
glower, composed of rare earth oxides or the Globar, composed of a silicon
carbide rod, are typical commercial IR sources. A typical NIR source consists
of a quartz diode lamp.
Determining the appropriate wavelength to monitor will depend on the
organic species of interest. General IR absorption characteristics of classes
of compounds are given in Table 18. Scanning the 6 to 9 micron (1665 to
1110 cm~l) wavelength region would allow for detection of most types of
organic compounds. To identify individual compounds in a complex mixture with
IR is complicated by overlapping absorption bands and shifts in characteristic
wavelengths. Computerized compound identification, where large numbers of IR
spectra are stored and rapidly compared with the spectra at hand, may allow
almost real time identification of simple sample matrixes. The use of
computers with IR instrumentation is primarily a laboratory function and not
yet useful as a continuous monitoring tool.
Presently, most continuous monitoring instruments measure absorption at
only one or two wavelengths. In-situ IR instruments are available but are not
considered applicable to monitoring a hazardous waste incinerator. Extractive
IR instrumentation have advantages and disadvantages depending on the system
design. Instruments that utilize one and two IR sources and a separated
reference and sample cell suffer from drift caused by particulate in the
sample. Instruments using a single cell use mirrors to increase the path
length which require maintenance and in many cases, careful adjustments.
Ultraviolet —
Aromatic hydrocarbons absorb radiation in the near ultraviolet wavelength
region. Aliphatic hydrocarbons typically do not absorb in this region to any
great degree. Compounds can absorb over both wide and narrow wavelength bands
and continuously absorb over several discrete bands. By determining the
absorption maxima, wavelength, and absorption intensity one can identify an
organic chromophore (e.g., a carbonyl group). Wavelength shifts can be
attributed to the presence of double and triple bonded species and rules have
been developed to aid in the identification of these species. There are many
variables and interferences involved with UV absorption measurements since
inorganic species also absorb at these wavelengths. Absorption is affected by
pH and nonabsorbing species that can react with solvents and numerous other
reagents to form absorbing species. Differentiating organic from inorganic
species and separating overlapping spectral bands can be done through use of
64
-------
TABLE 18. IR CHARACTERISTIC WAVELENGTHS
Compound class
Band of interest Wavelengths
(stretch, wag, rotation) m (cm~l)
Aromatic Hydrocarbons
Aromatic Halogenated
Hydrocarbons
Aromatic Ether
Aromatic Ketones
Aromatic Alcohols and Phenols
Aromatic Amines
Nonaromatic Amines
Nonaromatic Aldehydes
Nonaromatic Ketones
C-H stretch
C-C stretch
C-Fa stretch
CH2-C1 wag
C-0 vibration
C-0 stretch
0-H stretch
C-0 stretch
N-H stretch
C-C stretch
N-H stretch
N-H deformation
C-O stretch
C-0, C-C stretch
3.2-3.3
(3125-3039)
6-7
(1665-1430)
8
(1250)
8
(1250)
8.5-9
(1175-1110)
6-6.2
(1665-1615)
2.7-3.2
(3705-3125)
7.5-8.5
(1335-1165)
2.9
(3450)
6.2
(1665)
3
(3335)
6.2
(1615)
5.9-6
(1695-1665)
5-6.7
(2000-1495)
(continued)
65
-------
TABLE 18 (continued)
Compound class
Band of interest
(stretch, wag, rotation)
Wavelengths
m (cm"*)
Nonaromatic Alcohols
Nonaromatic Hydrocarbons
Nonaromatic Halogenated
Hydrocarbons
C-0 stretch
G-H stretch
C-H bend
C-F stretch
CH2 deformation
8-10
(1250-1000)
3.4-3.5
(2940-2855)
6.8
(1470)
15
(665)
8
(1250)
aF =» halogen, Cl, Br, etc.
Source: The Aldrich Library of Infrared Spectra, Second Edition, 1975 by
Charles J. Pouchart
66
-------
derivative UV absorption spectroscopy. Here, selectivity is improved by
modulating the incident wavelengths and detecting the second harmonic of the
emitted modulation frequency. This method is preferred over simple UV
spectroscopy since the results are not influenced significantly by sample
opacity, light intensity fluctuations or the presence of particulate matter in
the sample. Instrumentation is presently available that can scan several
wavelengths in a few seconds and in effect provide a real time analysis.
Aromatic compounds all have a very strong absorption peak at 184 nm.
Identification of each aromatic compound usually requires analysis of the
weaker absorption bands. Table 19 provides a listing of the next two
wavelengths where weak absorption occurs for several aromatic compounds. An
instrument capable of scanning 180 to 320 nm may allow identification of these
compounds.
TABLE 19. UV ABSORPTION WAVELENGTHS OF AROMATIC COMPOUNDS
Weak band Weaker band
Compound
Benzene
Toluene
m-Xylene
Chloro benzene
Phenol
Phenolate ion
Aniline
Anilinium ion
Thiophenol
Naphthalene
Styrene
C6*6
4H5CH3
C6H4(CH3)2
C6H5C1
C6H5OH
C6H50-
C6N5NH2
C6H5NH3
C6H5SH
C10H8
C6H5CH=CH2
max
(nm)
204
207
-
210
211
235
239
203
236
286
244
max
(nm)
256
261
263
265
270
287
279
254
269
312
282
Other Radiation Methods—
The intrumental measurement of absorbtivities in the UV and IR wavelength
regions discussed previously was based on first order absorption. Derivative
or wavelength modulated derivative spectroscopy is a technique where the
incident wavelengths are modulated by a few nanometers and the absorbed second
haroraonic of this frequency is detected. Basically, instead of determining
peak absorption wavelengths, the curvature of the absorption peak is also
measured. Measurement of this curvature provides increased selectivity. A
prototype derivative UV absorption spectrometer (DUVAS) has recently been
field tested and appears to be a promising continuous montioring tool. These
instruments require computer controllers and can scan a 100 nm wavelength
range within minutes. As additional derivative absorption spectrums are
developed and stored in computer memories, more and more complex samples can
be quickly identified.
67
-------
Polycyclic organic materials are inherently fluorescent and the use of
laser-induced fluorescence has been determined to be a potential real time
monitoring tool.15 This technique requires a separation step to be
selective and through use of a tuneable laser, problems with band overlapping
and broadening could be overcome. There is little to no data on the
fluorescent spectra of aromatic compounds and no commercial continuous
monitoring instruments were found.
Total luminescence spectroscopy is a technique where excitation and
emission wavelengths are compared to identify a compound. This technique is
applicable to identification of a single compound in a noninterfering
background and is not applicable to the determination of components of a
mixture. Polynuclear hydrocarbons in hexane have been identified at ppm
levels with this technique however no continuous monitoring instrumentation
appear to be available.
DATA ACQUISITION AND CONTROLS
The development of a continuous emissions monitoring system goes beyond
the gas analyzers and conditioning system. The analyzers must measure
emissions within specified time periods. The measurements then must be
recorded in some manner. After the data are recorded, these must be converted
into units of the emissions standard.
Presently, after a permit is issued, incinerator operators must
continuously monitor temperature, waste feed rate, air flow rate and stack gas
carbon monoxide levels. Incinerators operating in interim status are required
to monitor any existing instruments which relate to combustion and emission
control at 15 minute intervals.
A complete emission monitoring system, therefore requires some means of
recording the analyzer data. Strip-chart recorders have been used most often,
but data loggers and computer systems are beginning to become more popular.
Data processers have been developed specifically to reduce the time necessary
to evaluate and report emissions.
A data reporting system may encompass anything from manual resolution of
raw strip chart data and compilation of associated data to the near fully
automatic preparation of complete emission reports. The choice of the data
reduction and reporting system may be the most important factor in the overall
continuous emission monitoring system.
The control/data systems can be tailored to yield a complete printed
report. The user has the advantage of implementing digital processing
automation to whatever extent desired contingent upon economic and technical
constraints. Multiple stacks or duct locations can be accomodated with
hourly, daily and weekly reports with standard data processing hardware and
software. The addition of reason codes generally requires the addition of a
hard disc. While the standard software supports reports on a weekly basis, a
larger storage capability allows custom software to support data for a
calendar quarter, or longer. The cost of such a system is largely dependent
upon the features requested.
68
-------
SECTION 6
EVALUATION OF COMMERCIALLY AVAILABLE EQUIPMENT
Presently available continuous emission monitors utilize a wide range of
techniques for determination of pollutant emission rates. Evaluating the
usefulness or the adaptability of these instruments to monitoring in three
zones of an incinerator involves numerous considerations. The monitor and
conditioning system must operate with minimum maintenance requirements and be
reliable enough to be linked to automatic shutdown systems. The conditioning
system must provide a sample in close to real time and the analyzer should
respond within seconds. In the following sections, the available equipment
will be summarized by outlining advantages and disadvantages based on the
above considerations.
INORGANIC MONITORS
Specifically this program involved an assessment of continuous source
monitors. Manufacturers were contacted to obtain information on their
respective analyzers and commercial products required in their sampling
systems. This information was then used as a basis for the sampling system
requirements for each analytical technique.
The design of an adequate inorganic extractive interface for continuous
monitoring of SOX, NOX, CO, C02, 02, and HC1 from hazardous waste
incinerator sources involves a number of engineering choices and tradeoffs.
An "adequate" or a "minimum" sampling interface is the simplest and least
expensive system that will permit continuous monitoring within certain
specified tolerance limits.
The design of a sampling interface must be predicated on the specific
source/analyzer combination for which it is intended. Typical source
characteristics were given for hazardous waste incinerators. Sample
conditioning requirements are given for nine types of commercially available
analyzers including NDIR, UV (differential absorption), luminescence,
electroanalytical, conductimetrie, coulometric, colorimetric, thermal
conductivity, and specialized IR. Although it is difficult, in the absence of
field test data, to recommend a specific system which will achieve the
aforementioned criteria, a minimum assessment is presented.
Most analyzers are sensitive to pressure and must be controlled at a
fixed pressure value by venting to atmosphere. Sample flow rate is not a
critical parameter and the need for flow control devices or back pressure
regulators is not anticipated. A flow rate on the order of 1 SLPM is typical
of analyzers surveyed. Sample temperature is not a critical parameter as long
69
-------
as it remains within fairly wide limits. All analyzers which are sensitive to
sample temperatures provide internal control. The sample pump must be
positioned upstream of the analyzer if pressure is controlled by venting to
atmosphere.
The proper choice of materials of construction is an important part of
the interface design. Information has been presented on chemical resistance
and heat resistance. The stainless steel and nickel alloys are somewhat
versatile but limited in corrosive atmospheres and the chemically inert
materials such as glass, ceramics, and refractory metals are extremely heat
resistant but fragile. Plastics and fluoroplastics are extremely good in
corrosive atmospheres but have limited heat resistance.
For incinerator sources, coarse filtration is required before the sample
is withdrawn from the stack. A few types of filters have been mentioned:
(1) the "external" or exposed filter in which the filter element is supported
within the duct or stack, and (2) the internal or in line filter which is also
supported within the stack or duct but is exposed only to the withdrawn
sample. The internal arrangement is recommended when calibration gases are to
be injected at the probe tip. Calibration of the analyzers can be
accomplished by using simple timing devices to actuate solenoid valves.
Sample line materials can be plastics or fluoroplastics except for the
first few meters in the vicinity of the stack or duct when sampling in the
combustion zone. Heated sample lines can be used to keep the sample above the
dewpoint. The tolerance limit or system response time can be easily met
without extracting excess sample and bypassing the excess to vent. Dead end
volumes have little effect on system response time but large mixing volumes
can have a significant effect. When pressure is controlled by venting to
atmosphere, the most suitable type of sampling pump is the diaphragm pump.
Other types of air movers have been discussed, such as eductors and ejectors,
but should be used when dilution is necessary and moisture is not removed from
the system.
Moisture removal is an important part of the interface design. Only NDIR
instruments require the maintenance of a constant moisture level. Moisture
can be removed by condensation, permeation distillation and membrane
separation or not removed at all by maintaining the sample above its dewpoint
or by sampling using air dilution.
Fine filtration is accomplished either in the conditioning system or
internally in the analyzer itself. Useful control measurements for the
conditioning system include the sample flow rate, the pump suction and
discharge pressures, and sample line temperatures.
The exact nature of hazardous waste incinerator exhaust gases depends to
some extent on the type of emission control process that is used. If a
dry-absorption process is used to reduce emissions, moisture removal
components are less necessary, however, acid mist must be removed. If a wet
scrubber is used to control emissions, acid mist removal is accomplished but
water vapor must be removed or controlled at a fixed level for NDIR
instruments. The basic conditioning system components for sampling dry stack
70
-------
gases are probe tip filter, acid mist remover, sample pump, and fine filter.
The basic components for sampling wet stack gases are probe tip filter if
required, sample pump, moisture removal or dropout system, and fine filter.
If large volumes of NC>2 are present heat tracing should be used to keep the
sample above its dewpoint or a permeation dryer can be used to selectively
remove water vapor while retaining other gases. Heat tracing as the only
control measure for high moisture contents is not recommended for NDIR
instruments which are sensitive to water vapor. Heat tracing may be
unacceptable for chemiluminescent and electroanalytical analyses if the
dewpoint is above their maximum allowable sample temperature. Heat tracing as
the only control is a viable alternative for all other instruments.
Permeation drying can be used with all instruments and can represent a cost
savings over heat tracing. The same additional interface components are
required.
The distinguishing disadvantage of NDIR analyzers is their sensitivity to
water vapor. This requires either moisture removal to below 3 percent or
maintenance of a constant moisture level in both the sample and calibration
gases. Many flue gas constituents have overlapping absorption wavelength
bands and it is difficult to isolate an absorption wavelength band for a
specific constituent that is not influenced by other constituents of the stack
gas. The main advantage of NDIR analyzers in continuous measurement is that
it is less expensive than comparable UV analyzers. It also is the best way to
analyze combustion control parameters (CO and
An advantage of UV analyzers is high operating temperature capabilities.
Most manufacturers recommend that the sample be kept above the dewpoint. UV
analyzers are tolerant to limited amounts of particulates. If sufficient
particulate removal is obtained in coarse filtration, fine filtration is not
required.
There are no particularly noteworthy aspects of electroanalytical
analyzers except for removal of interferences and low price.
The chemiluminescent analyzers must have some means, internal to the
instrument itself, to control the flowrate of the sample to the reaction
chamber. Some manufacturers control internal flow and pressure with
regulators. Some analyzers draw sample through the analyzer using a pump on
the downstream side. Moisture can be removed in chemiluminescent analyzers by
a room temperature trap provided the analyzer is kept above room temperature.
The main advantage of the chemiluminescent technique is sensitivity and broad
concentration ranges measurable with one instrument.
Other luminescent techniques, such as flame photometric detection and
fluorescence have advantages in that they are specific to sulfur compounds. A
disadvantage is that, when applied to flue gas measurement, dilution is
required from 1,000-10,000 to 1 so that the detector can operate in its
optimum range.
Other analytical techniques include gas filter correlation and Fourier
transform infrared which could be applied to continuous measurement of HC1 and
a combination colorimetric/photometr ic analyzer for 803 measurement. These
71
-------
techniques cannot be assessed on a continuous basis on hazardous waste
incinerators without a proper field evaluation.
One of two problems faced with implementing an interface design is
locating manufacturers of suitable components. A number of components are
fabricated cheaply, but in many cases the cost of labor for fabrication makes
it less expensive to buy commercially available components.
In conclusion, sample conditioning development for source sampling is in
a state of flux. Each manufacturer seems to have his own idea as to what
configuration a sample conditioning unit should take. Well designed, reliable
sample conditioning systems can cost as much or more than the analyzers
attached to it. It must be remembered that no matter how good the analyzer
is, no analyzer will work properly with poor sample conditioning.
Sample conditioning systems have been and are being presented to industry
in a variety of forms. One such system consists of a probe, heated sample
line, water trap, and aspirator. There are pressure-type systems which
utilize a probe, heated sample line, heated pump, heated filters, and
refrigerated water trap. There are systems that only utilize a filter at the
probe, and others that use heated probes with external filters.
In short, presently available commercial conditioning systems are not
fabricated to handle typical incinerator conditions but specialized fabricated
systems can be developed by the vendors to fit the particular application.
ORGANIC SPECIES MONITORS
Presently available organic species continuous monitors can be divided
into two basic types, instruments that can detect a large number of compounds
and instruments that detect only one or two compounds. A commercial hazardous
waste incinerator burning a wide variety of wastes would likely need an
instrument capable of detecting at least four organic compounds or compound
classes. The capital costs of these two general types of equipment are not
appreciably different. If further regulatory demands include continuous
compliance monitoring of more than one organic component, instruments that
convert organic compounds to C02 or CH,/,., followed by detection of these
species, may not be applicable. Single or double wavelength spectrometers and
methane monitors would be of little use compared to spectrometers that scan a
wavelength range or chromatographs that also detect a variety of compounds.
Most presently available monitors offer ppm levels of sensitivity whereas
an incinerator effluent typically contains parts per billion or parts per
trillion levels assuming that 99.998 percent or greater ORE is maintained.
Instruments that offer sensitivity at these low ranges are presently
laboratory type instruments that are not yet field proven for continuous
monitoring applications. ORE field tests are performed by concentrating the
sample prior to analysis in order to provide sufficient sensitivity. The
dynamic ranges of typical organic laboratory instrumentation is provided in
Figure 11. The ranges of continuous emission monitots using these detection
techniques would be similar. Data on the long-term use of continuous sample
concentrating equipment, such as that sometimes used as an interface between a
72
-------
>> H M
2 W M pi O
O OS 2 H H < ,-J
>— 1 3 O U MCL<2 *^
CK Cd CM H CK 3 S
O
01
CO
(3
O
0)
n.
CO
ts
cfl
oo
O
U-l
a)
60
C
04
u
•r-l
6
rt
60
73
-------
GC and a MS spectrometer, is not yet available. It is expected that if
continuous sample concentration techniques are more frequently used, reliable
resins and sample extraction systems will be developed. Also likely is the
development of more sensitive continuous monitors.
Adaptability is also very dependent on sample conditioning requirements.
With present sensitivities, dilution methods of sample conditioning are not
appropriate. Particulate removal is required for all organic analyzers.
Mechanisms for particulate removal are well developed, however the effect of
particulate removal on the integrity of the organic sample is not well
documented. HC1 gas concentrations of up to 17 percent must be reduced below
1,000 ppm for most presently available instrumentation to ensure a reasonable
instrument lifetime. Instruments that can withstand high acid gas
concentration are available and instruments presently sensitive to HC1 can be
adapted for higher concentrations by specifying different materials of
construction of key components. Moisture removal mechanisms are also well
developed and moisture does not pose an insurmountable problem for most
analyzers. The loss of soluble organic species in the removal of moisture is
of concern. Sample conditioning methods that maintain sample gas temperatures
above the boiling point of water and above the boiling points of all compounds
of interest are preferred to condensation or permeation drying mechanisms. It
is anticipated that a conditioning system consisting only of several
particulate filters and heated lines would be required for organic analyzers
equipped to handle significant HC1 gas concentration.
Gas Chromatography
The most widely used organic species monitor is the GC/FID combination.
Advantages of this combination include:
(a) linear response over 6 decades
(b) insensitive to air, H20, or inorganic gases
(c) sensitive to all organic compounds
(d) relatively simple and inexpensive instruments are available.
Gas chromatographs are highly adaptable since many variables can be
involved in the selection and operation of an instrument. Column lengths,
column packing materials, operating temperatures and flow rates and sample
handling equipment can all be manipulated to provide the desired results.
Instruments are avilable with multiple columns, multiple detectors and
multiple sample handling capabilities such that several classes of compounds
can be detected rapidly and simultaneously.
Disadvantages of the GC/FID combination include:
(a) high degree of particulate removal required
(b) not real time analysis
74
-------
(c) requires pure hydrogen, air and inert carrier gases
(d) detector susceptible to corrosion by HC1.
Particulate matter must be removed to lower and lower levels as the
sensitivity of the FID is pushed higher. Particulate causes a noisy baseline
that could overshadow a response to an organic compound. Some laboratory
GC/FID instruments require 1 hour to complete an analyses of one complete
sample. Recent advances in high speed microparticulate packed columns has
provided response times of 30 seconds where several minutes would normally be
required. 16 A. continuous monitoring GC/FID could be tailored to provide
response times on the order of 1 minute, however separation efficiency and
sensitivity may decrease as the response time is decreased. Response times of
less than 3 minutes would allow separation and identification of most
hazardous compounds expected to be present, however, two or more columns would
probably be required.
An FID that is resistant to attack by HC1 could be built through use of
hastelloy C or Inconel detector plates. Presently, no "off the shelf"
detectors, constructed of these or other acid resistant materials (besides
stainless steel which is not considered resistant to high concentrations of
HC1) were found, however several manufacturers indicated a willingness to
construct special detector plates.
A gas chromatograph equipped with an electron capture detector (ECD) has
been successfully used in ambient air monitoring applications where
sensitivities of low part per trillion for chlorofluorocarbons have been
demonstrated over the long term (1975-1981).I7 The Ni-63 ECD is preferred
over the tritium ECD for long-term stability and greater precision. This
highly sensitive detector is suitable for trace level determination of a
variety of hazardous organic species such as PNAs, PAHs, and PCBs. In
combination with an FID, a dual range instrument that is sensitive to
12 decades of concentration of organics would result. This dual detector
instrument would be ideally suited to continuous monitoring applications where
large swings in concentration of organics are expected. The sensitivity of
the ECD requires longer restabilization times between large upsets, whereas
the FID recovers more quickly. Operating the incinerator within the range of
the ECD would document 99.99 percent ORE in most cases, assuming low parts per
billion emission rates. For long-term use with high acid gas concentrations,
both detectors should be constructed to minimize corrosion problems.
A gas chromatograph with photoionization detector (PID) is presently the
most resistant to corrosion of the chromatographic methods. The PID is a
nondestructive detector that can be purchased with teflon-coated internals
with no decrease in sensitivity. Advantages of the PID include 10 times
greater sensitivity than the FID to certain compounds, a linear range of 107
(10X greater than FID) and its nondestructive nature. Disadvantages include
lower selectivity (inorganic compounds are also detected), drastic reductions
in sensitivity due to coatings on the lamp windows and the need for
interchanging several lamps of various energies for identification of
compounds. A PID in series with an FID is a recently available laboratory
75
-------
instrument with potential for continuous monitoring applications. The ratio
of the PID to FID response is a tool for further identification of a compound
not available with the individual detectors.
Radiation Techniques
All organic compounds and most inorganic compounds absorb radiation.
Using UV, VIS, or IR analyzers to continuously detect and identify compounds
in a complex effluent stream has, to date, been limited. Only certain
compounds are monitored, their characteristic wavelengths are scanned, and
absorption at these wavelengths is measured. Since no sample separation steps
are involved, interfering species may cause erroneous results. Overlapping
spectral bands makes identification of compounds difficult in a continuous
monitoring situation. Spectral shifts occur from variations of pH and other
factors. The adaptability of these instruments to the variable
characteristics of incinerator effluents is limited to detection of one or two
compounds per instrument. These techniques are widely accepted as useful in
the determination of many inorganic gases. Lack of specificity for organic
compounds makes these type of instruments less suitable for organic species
monitoring.
Other Techniques
Combining instruments is an ideal way to overcome disadvantages.
Specificity problems associated with radiation instrumentation are overcome in
GC/IB. and GC/FTS-IR combinations. GC/MS is widely recognized as a very
powerful tool for compound identification. Combination instruments are
gaining wider acceptance as continuous monitoring tools, however, at present
the high cost of such systems has limited their use to mostly research
application. These instrument combinations are not considered as presently
available continuous monitors, however as the technology develops these
instruments may be available as monitors in the near future.
76
-------
REFERENCES
1. U.S. Environmental Protection Agency, Regulations for Hazardous Waste
Management, 40 CFR Parts 260-265, 267.
2. Bonner, T. A., et al. Engineering Handbook for Hazardous Waste
Incineration, SW889, U.S. Environmental Protection Agency, Cincinnati,
Ohio. June 1981.
3. Destroying Chemical Wastes in Commercial Scale Incinerators, NTIS Final
Report SW-122c.l, U.S. EPA. 1977.
4. Personal Communication with Mr. Jack Kertzman, Perma Pure Products, Inc.,
Oceanport, NJ (September 1982).
5. Telecon. Charles Iltis, Osmonics Inc., Hopkins, Minnesota, with F.
Abell, GCA/Technology Division, October 26, 1982.
6. Houser, E. A. "Gorkman Analysis Systems for SC>2/NOX in power plant
Stack Gases." From Beckman Representatives' Memorandum (November 1971).
7. Cheremisinoff, P. N., et al. Corrosion Resistance of Piping and
Construction Materials. Pollution Engineering. August 1973. pp. 23-26.
8. McNulty, et al. Investigation of Extractive Sampling Interface
Parameters. U.S. Environmental Protection Agency, Washington, D.C.,
October 1974.
9. Handbook Continuous Air Pollution Source Monitoring Systems. U.S.
Environmental Protection Agency, Cincinnati, Chio, June 1979.
10. Personal Communication with Mr. Anthony Eggleston, Kilkelley
Environmental Associates, Inc.
11. Bazer, F. L. "An Overview of Chroma to graphic Instrumentation: Problems
and Solutions" Journal of Chromato graphic Science, Volume 20.
September 1982.
12. Gurka, D. F. , and P. R. Laskan. The Capability of GC/FTIR to Identify
Toxic Substances in Environmental Extracts. Journal of Chroma to graphic
Science, Volume 20. April 1982.
77
-------
13. Johnson, I. et al. "Sampling and Analysis of Hydrocarbons in Combustion
gases" ANL/CEN/FE-80-22 Quarterly Report. October to December 1980.
Argonne National Laboratory.
14. Jones, H. R. and P. A. Wilks, Jr. The C02 Laser and Infrared
Spectroscopy American Laboratory. March 1982. p. 87-91.
15. Vo-Dink, T. New Luminescence Techniques Simplify Air Analysis. Intech.
May 1981.
16. Jonker, R. J., Poppe, H. and J. F. Huber. "Improvement of Speed of
Separation in Packed Column Gas Chromatography. Anal Chem. 1982, 54,
2447-2456.
17. Brice, K. A. et al. "Measurement of CC^F and CC14 at Harwell over
the period January 1975 to June 1981 and the Atmospheric Lifetime of
CC13F". Atmospheric Environment, Volume 16, No. 11. pp. 2543-2554.
1982.
18. Cudahy, J. J., L. Sroka, and W. Troxler. Incineration Characteristics of
RCRA Listed Hazardous Wastes. U.S. Environmental Protection Agency,
Office of Research and Development, Cincinnati, Ohio. July 1981.
19. Redmond, J. D., and K. H. Miska. The Basics of Stainless Steels.
Chemical Engineering. October 18, 1982. pp. 79-93.
20. Corrosion Resistance of Nickel-Containing Alloys in Caustic Soda and
Other Alkalies, Corrosion Engineering Bulletin, Inc., New York, NY. 1973.
21. Corrosion Resistance on Nickel-Containing Alloys in Hydrofluoric Acid,
Hydrogen Fluoride and Fluorine. Corrosion Engineering Bulletin, Inc.,
New York, NY. 1968.
22. Schillmoller, C. M. Alloys to Resist Chlorine, Hydrogen Chloride and
Hydrochloric Acid. Chemical Engineering. March 10, 1980.
23. Clausen, J. F., et al. At-Sea Incineration of Organocchlorine Wastes
Onboard the M/T Vulcanus. EPA-600/2-77-196. U.S. Environmental
Protection Agency, Research Triangle Park, N.C., September 1977.
24. Personal Communication with Mr. Pat Conner, Columbia Scientific
Industries Corporation, Austin, TX (October 1982).
25. Atmospheric Emissions from Sulfuric Acid Manufacturing Processes. PHS
Pub. No. 999-AP-13, Publiv Health Service, Durham, NC. 1965.
26. Perry, R. H., Editor, Chemical Engineers' Handbook, 5th Edition, New
York; McGrawHill (1973).
78
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
I. REPORT NO.
EPA-600/8-84-011a
2.
4. TITLE AND SUBTITLE Feasibility Study for Adapting Present
Combustion Source Continuous Monitoring Systems to
Hazardous Waste Incinerators; Vol. 1. Adaptability
Study and Guidelines Document
3. RECIPIENT'S ACCESS!Ol*NO.
5. REPORT DATE
March 1984
6. PERFORMING ORGANIZATION COOE
7. AUTHOR(S)
John Podlenski, Edward F. Peduto, Robert Mclnnes,
Frank Abell, and Stephen Gronberg
8. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-82-60-G(l)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA/Technology Division
213 Burlington Road
Bedford, Massachusetts 01730
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3168, Task 55
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 10/81 - 9/82
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES IERL-RTP project officer is Merrill D. Jackson, Mail Drop 62,
919/541-2559. Volume 2 is a review and estimate of incineration test conditions.
16. ABSTRACT
repOrt gives results of an adaptability study of commercially available
sample conditioning and measurement systems, in the form of a guidelines document
to be used by EPA and industry personnel. As part of EPA- sponsored research pro-
grams to investigate sampling and analysis methods for hazardous waste inciner-
ation (focused on adapting existing methods for identifying and quantifying constit-
uents listed in 40 CFR 261), the adaptability of existing continuous emission moni-
toring systems (CEMS) involves such measurement categories as SO2, SOS, NOx,
CO, CO2, O2, HC1, and organic materials. Study results indicate that commercially
available extractive continuous monitors can be adapted to incinerators through pro-
per sample conditioning. Available CEMS provide the ranges and sensitivities
needed to accurately measure concentrations of the organic and inorganic compo-
nents of interest.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Pollution
Monitors
Incinerators
Wastes
Toxicity
Measurement
Samples
Treatment
Pollution Control
Stationary Sources
Continuous Monitors
Hazardous Waste
Sample Conditioning
13B
14G
06T
14B
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
85
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
79
-------