xvEPA
United States Industrial Environmental Research
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
EPA-600 7-78-134
July 1978
Guidelines for
Preparing
Environmental
Test Plans for Coal
Gasification Plants
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide-range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsemeht or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-78-134
July 1978
Guidelines for Preparing
Environmental Test Plans
for Coal Gasification Plants
by
G.C. Page, W.E. Corbett, and W.C. Thomas
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
Contract No. 68-02-2147
Exhibit A
Program Element No. EHE623A
EPA Project Officer: William J. Rhodes
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The report outlines a philosophy and strategy for
preparing environmental assessment sampling and analysis (test)
plans. Five major points of test plan development are addressed:
(a) defining the test objectives; (b) performing an engineering
analysis of the test site; (c) developing a sampling strategy;
(d) selecting analytical methods; and (e) defining data manage-
ment procedures. The important considerations involved in each
of these areas are discussed in relation to three types of envi-
ronmental tests: (a) waste stream (Levels 1, 2 and 3); (b) con-
trol equipment; and (c) process stream characterization. Some
specific sampling and analytical methods are presented, with
numerous references cited for more detailed information.
-------
TABLE OF CONTENTS
Number Page
1. 0 INTRODUCTION 1
2.0 TEST PLAN OBJECTIVES 4
2. 1 WASTE STREAM CHARACTERISTICS 5
2.1.1 Screening Waste Streams (Level 1). 6
2.1.2 Detailed Waste Stream Charac-
terization (Level 2) 7
2.1.3 Continuous Monitoring of Waste
Streams (Level 3) 7
2 . 2 CONTROL EQUIPMENT CHARACTERIZATION 7
2.3 PROCESS STREAM CHARACTERIZATION 8
3.0 ENGINEERING ANALYSIS 10
3 . 1 GENERAL PRINCIPLES 10
3.1.1 Process Flow Sheet Development.... 11
3.1.2 Process and Waste Stream
Characterization 14
3.1.3 Definition of Process Variable
Effects 16
3.2 SPECIFIC EXAMPLE - ENGINEERING ANALYSIS
OF A CHAPMAN GASIFICATION PLANT 17
3.2.1 Process Description 18
3.2.2 Waste Stream Prioritization 24
4. 0 SAMPLING STRATEGY 29
4. 1 SAMPLE POINT SELECTION 30
4.1.1 Sampling Point Selection for Gases 31
4.1.2 Sampling Point Selection for
Liquids 34
111
-------
TABLE OF CONTENTS (Continued)
Number
4.1.3 Sampling Point Selection for
Solids 34
4.1.4 Other Sampling Point Considera-
tions 34
4.2 SAMPLING METHOD SELECTION 35
4.3 SAMPLING FREQUENCY AND TIMING 37
4.3.1 Sampling Frequency 37
4.3.2 Sampling Timing 39
4.4 SAMPLING QUALITY CONTROL PROGRAM 40
5.0 ANALYTICAL STRATEGY 45
5.1 SAMPLE HANDLING AND PRESERVATION 52
5.2 ANALYTICAL TECHNIQUES 52
5.2.1 Analytical Screening Techniques ... 53
5.2.2 Quantitative Analytical Proce-
dures 61
6.0 DATA EVALUATION AND MANAGEMENT 88
6.1 PLANNING EXPERIMENTS 88
6.1.1 Statistical Experimental Design ... 89
6.1.2 Quality Control Program 91
6.2 DATA VALIDATION 93
6.2.1 Material Balance Calculations 93
6.2.2 Statistical Data Validation 94
6. 3 DATA EVALUATION 94
6.3.1 Waste Stream Characterization Test. 95
-------
TABLE OF CONTENTS (Continued)
Number Page
6.3.2 Control Equipment Characterization
Tests 101
6.3.3 Process Stream Characterization
Test 103
6.4 DATA HANDLING 103
6.4.1 Manual Data Base Organization 104
6.4.2 Computer Data Base 106
7.0 REFERENCES 10 7
APPENDIX SAMPLING METHODS 113
A-l GAS SAMPLING 113
A-2 LIQUID SAMPLING 151
A- 3 SOLID SAMPLING 168
A-4 REFERENCES 176
v
-------
FIGURES
Number Page
1-1 INTERRELATIONSHIPS AMONG GENERAL AREAS
INVOLVED IN PREPARING AN ENVIRONMENTAL
TEST PLAN 3
3-1 SIMPLIFIED PROCESS FLOW DIAGRAM - BURNHAM
COAL GASIFICATION COMPLEX (REF. 2) 12
3-2 EL PASO BURNHAM GASIFICATION PLANT
SECTION (REF. 3) 13
3-3 MODULAR APPROACH TO PROCESS ANALYSIS 17
3-4 SIMPLIFIED PROCESS FLOW DIAGRAM FOR THE
CHAPMAN GASIFICATION PLANT 20
4-1 NUMBER AND LOCATION OF TRAVERSE POINTS
DEFINED BY EPA METHOD 1 GUIDELINES 32
4-2 SAMPLE PORT ARRANGEMENT FOR "HEAD-ON"
SAMPLING 33
4-3 EXAMPLE OF A QUALITY CONTROL CHART FOR
SPIKED SAMPLES (REF. 5) 42
5-1 OUTLINE OF LEVEL 1 ELEMENTAL ANALYSES USING
SPARK SOURCE MASS SPECTROMETRY (SSMS) 55
5-2 LEVEL 1 ANALYSES FOR MERCURY, ANTIMONY
AND ARSENIC 56
5-3 OUTLINE OF LEVEL 1 ORGANIC ANALYSES 57
5-4 EXAMPLE INORGANIC ANALYTICAL SCHEME FOR
LIQUID SAMPLES 74
5-5a INTEGRATED SCHEME FOR SEPARATION AND ANALYSIS
OF ORGANIC CONSTITUENTS FROM COAL GASIFICATION
PROCESSES 86
5-5b INTEGRATED SCHEME FOR SEPARATION AND ANALYSIS
OF ORGANIC CONSTITUENTS FROM COAL GASIFICATION
PROCESSES 87
-------
FIGURES (Continued)
Number Page
6-1 A FULL FACTORIAL DESIGN INVOLVING TWO
INDEPENDENT VARIABLES AT THREE LEVELS 90
6-2 EXAMPLE OF A QUALITY CONTROL CHART FOR
SPIKED SAMPLES (REF. 5) 92
Vll
-------
TABLES
Number Page
3-1 PROCESS DATA TO BE COLLECTED DURING
SAMPLING 2 8
5-1 ANALYTICAL PARAMETERS OF INTEREST FOR
ENVIRONMENTAL TESTING OF COAL GASIFICATION
FACILITIES 46
5-2 ENVIRONMENTALLY HAZARDOUS ORGANIC SPECIES
POTENTIALLY PRESENT IN COAL GASIFICATION
FACILITY PROCESS AND WASTE STREAMS 47
5-3 ORGANIC SPECIES FROM EPA EFFLUENT
GUIDELINES (REF. 6) 50
5-4 BIOASSAY PROTOCOL FOR SOLID, LIQUID
AND GAS SAMPLES (REF. 1) , 51
5-5 ANALYTICAL PARAMETERS FOR SCREENING
INORGANIC CONSTITUENTS IN SOLID, LIQUID
AND GAS SAMPLES 54
5-6 BIOASSAY PROTOCOL FOR LEVEL 1 WASTE
STREAM CHARACTERIZATION TESTS 59
5-7 INORGANIC ANALYTICAL METHODS FOR COAL
GASIFICATION PROCESSES (GASEOUS) 64
5-8 INORGANIC ANALYTICAL METHODS FOR COAL
GASIFICATION PROCESSES (LIQUIDS) 66
5-9 INORGANIC ANALYTICAL METHODS FOR COAL
GASIFICATION PROCESSES (SOLIDS) .. . 70
6-1 'LEVEL 1 ORGANIC ANALYSIS RESULTS COMPARED
TO MATE VALUES 97
6-2 BIOASSAY TEST RESULTS FOR A COAL
GASIFICATION FACILITY 98
6-3 KEY WORD CATEGORIES FOR ENVIRONMENTAL
ASSESSMENT DATA BASE 105
VI11
-------
ACKNOWLEDGEMENTS
The authors wish to express their thanks to R. V,
Collins and J. M. Harless for their contributions to the
sampling and analysis sections and to R. A. Magee and M. P.
Kilpatrick for their review comments.
IX
-------
SECTION 1.0
INTRODUCTION
This report outlines a philosophy and strategy for
preparing environmental assessment sampling and analysis plans
(referred to as "test plans" throughout this document) for coal
gasification plants. Its primary purpose is to provide general
guidelines for the development of conceptually sound site spe-
cific test plans. While an environmental test plan may include
both source evaluation and ambient monitoring, this report only
addresses source evaluation. It is not intended to be a source
of the actual procedures required, although many of the appli-
cable sampling and analytical techniques are either referenced
or discussed in some detail.
The five general areas which must be addressed in
developing an environmental test plan are:
defining the test objectives,
performing an engineering analysis of the
test site,
developing a sampling strategy (including
selecting sampling points and sample
handling methods) ,
selecting analytical methods, and
defining data management procedures.
The important considerations involved in each of these areas are
briefly discussed below.
Any test plan development effort should start with a
definitive statement of the anticipated test objectives. This
definition is important because it will influence many of the
decisions that must be made in the other four areas of test plan
design.
-------
An engineering analysis should be performed to iden-
tify and characterize the process steps, control equipment, and
waste streams associated with a particular site. Ultimately,
this analysis should lead to an understanding of the functions
and principles of operation of all key pieces of plant hardware,
as well as the materials and chemicals involved.
Developing a sound sampling strategy involves address-
ing such considerations as sample point selection and timing.
Obviously, sampling points should be selected so that represen-
tative samples of the streams being studied are obtained. The
timing and frequency of sampling are usually dictated by the
objectives of the test and characteristics of plant operation
such as cyclic operations and material residence times.
Sampling methods should be chosen to allow samples
from the specified streams to be collected and preserved so that
subsequent chemical analysis data are representative of those
streams. The physical conditions of the stream, the accuracy
requirements of the test, the parameters of interest, and the
amount of sample required are important considerations here.
The analytical methods chosen must allow the chemical
species of interest to be quantified to acceptable levels of
accuracy. Factors affecting analytical method selection include
the composition of the stream, the type of sample taken, the
sampling method used, and accuracy requirements.
Aspects of the data management function which should
be considered in the test planning phase include data validation,
data reduction techniques, and data evaluation.
Although each of these areas is distinct, as indicated
in Figure 1-1, the decisions which must be made within each are
dependent upon limitations inherent within all of the other areas.
It is this interdependency in the decision-making process that
makes test plan construction difficult. In subsequent sections
of this report, each of the five areas just mentioned is dis-
cussed in detail. While the individual areas are discussed
separately, the need to consider the interrelationships among
these areas is stressed.
-2-
-------
Informationj
Needed !"
Engineering
Analysis
Definition of Test
Plan Objectives
Analytical
Strategy
Development
Data Evaluation,
Techniques
MAJOR AREAS IN TEST PLAN PREPARATION
Completed
Test Plan
Figure 1-1. INTERRELATIONSHIPS AMONG GENERAL AREAS INVOLVED IN PREPARING AN
ENVIRONMENTAL TEST PLAN
-------
SECTION 2.0
TEST PLAN OBJECTIVES
In this section, the various types of tests which can
be used to characterize the environmental aspects of coal gas-
ification technology are described. The primary goal of this
discussion is to summarize these tests and the results obtained
from each and to emphasize that defining the test objectives is
one of the most critical steps in developing sound environ-
mental test plans. Typical types of environmental tests and
their objectives are summarized below.
Test Type
Waste Stream
Characterization
Control Equipment
Characterization
Process Stream
Characterization
General Statement of Test Objectives
To identify and quantify the
pollutants found in a facility's
multimedia (gaseous, liquid, and
solid) waste streams and to eval-
uate their health and ecological
effects.
To determine the effectiveness of
existing or developing control
equipment for removing pollutants
from waste streams.
To determine the origins and fates
of pollutants as they pass through
selected processes and to evaluate
the effects that process operating
parameters have on pollutant types
and concentrations.
The above tests may be executed during one or more of
the following process operating conditions:
normal operation,
start-up,
shutdown, and
emergency upsets.
-4-
-------
It is necessary to specify the process operating conditions
required since these conditions will affect the physical and
chemical characteristics of process and waste streams along
with the performance characteristics of control equipment. The
desired operating conditions should be defined by the engineering
analysis of the plant (or process) to be tested, as discussed
further in Section 3.0. The objectives section of an environ-
mental test plan, therefore, should address not only the specific
test objectives, but also define the plant operating conditions
required.
The following text summarizes the various types of
environmental tests (waste stream, control equipment and process
stream characterization) and the results that can be obtained
from each. Where appropriate, examples are given to help
clarify this discussion.
2.1 WASTE STREAM CHARACTERIZATION
A waste stream characterization test is designed to
identify and/or quantify the pollutants emitted from a plant or
process. The scope of this type of test depends upon the
specific test objectives. The Environmental Protection Agency
(EPA) has established guidelines for waste stream characteriza-
tion tests (Ref. 1) which define the following three levels of
testing:
a semiquantitative overview or screening
(Level 1),
an accurate quantification of selected
pollutants in selected waste streams
(Level 2), or
continuous monitoring of pollutants in
specific waste streams (Level 3) .
Typically, Level 1 testing along with an engineering analysis
of the plant (as discussed in Section 3.0) can be used to
identify environmental problem areas which require more quanti-
tative characterization (Level 2 testing). The results of Level
2 testing can then be used to determine the specific pollutants
and streams that require continuous monitoring (Level 3).
-5-
-------
A waste stream characterization test may be performed
during any of the four plant operation modes (normal conditions,
start-up, shutdown, or emergency upsets), although generally it
would be performed during normal plant operation. An exception
to this would occur when the results of the engineering analysis
indicated that there may be significant and/or unique environ-
mental problems associated with the other plant operating modes.
In these cases, waste stream characterization tests may be
desired during start-up, shutdown, and/or emergency upsets.
The objectives and scope of the three levels of waste
stream characterization tests are discussed in the following
text. A more detailed discussion of these tests is presented in
the IERL-RTP Procedures Manual: Level 1 Environmental Assess-
ment, (EPA-600/2-76-160a and b) (Ref. 1).
2.1.1 Screening Waste Streams (Level 1)
The purpose of a Level 1 waste stream characterization
test is to identify gaseous, liquid, and solid waste streams
which require further characterization. A Level 1 test, there-
fore, usually involves screening all of the multimedia waste
streams emitted from a plant or process to insure that all
possible environmental problem areas are covered.
Although the primary emphasis of a Level 1 test is on
waste streams, some process streams such as feedstocks should
also be considered because of their effects on the waste stream
characteristics and control equipment performance. In a Level
1 test, samples of gaseous, liquid, and solid waste streams are
analyzed for organic and inorganic pollutants, biological and
ecological effects and particle morphology. Process stream
data are also required to determine whether the Level 1 testing
occurred during the operating conditions specified by the test
plan objectives.
In summary, a Level 1 waste stream characterization
test will provide a general overview of the environmental pro-
blem areas associated with a facility. The chemical, biological
and ecological results from the Level 1 test can then be used to
establish priorities for more quantitative characterization
testing (Level 2).
-6-
-------
2.1.2 Detailed Waste Stream Characterization (Level 2)
The purpose of Level 2 testing is to provide quantita-
tive data for specific pollutants in specific waste streams. As
in Level 1 testing, certain process streams should be tested in
order to provide information concerning process operating con-
ditions during the test effort.
Level 2 testing is usually too specific and too costly
to be performed on all of the waste streams from a plant. There-
fore, the results of Level 1 tests and/or an engineering analysis
of the plant should be used to identify specific streams for
Level 2 testing. If an engineering analysis indicates that an
environmental hazard may exist in a particular stream, Level 2
testing can be performed simultaneously with, or in some
cases, in lieu of, Level 1 testing. Timing and cost benefits
are the important considerations in this decision.
2.1.3 Continuous Monitoring of Waste Streams (Level 3)
The purpose of Level 3 testing is to quantitatively
monitor the amounts of specific pollutants found in selected
waste streams on a continuous basis. These results should pro-
vide data on the relationships between process operation (normal
operation, start-up, shutdown and emergency upsets) and the
emission rates of specific pollutants. The results from Level
3 testing may also provide guidelines on the specific pollutants
and waste streams that require continuous monitoring during
plant operation.
2.2 CONTROL EQUIPMENT CHARACTERIZATION
The purpose of a control equipment characterization
test is to assess the performance of a control device in
treating a specific waste stream and/or to obtain the informa-
tion required to design the control equipment required to treat
a specific waste stream. As for waste stream characterization
tests, a control equipment characterization test should be per-
formed during defined plant operating conditions (normal opera-
tion, start-up, shutdown, and/or emergency upsets).
-7-
-------
In order to assess the performance of a control pro-
cess, the distributions of specific pollutants or elements of
interest in the inlet and outlet streams of the process must be
determined. This may also involve performing a material balance
test for specified compounds or elements around the control equip-
ment. Material balance tests are usually restricted to selected
elements especially if the pollutant in the inlet stream under-
goes chemical changes. Therefore, the performance of the control
equipment should be assessed both by performing material balances
on certain elements and by analyzing specific pollutants in
selected inlet and outlet streams.
For example, a Stretford process can be used to con-
trol the emissions of gaseous sulfur species from low-Btu gas-
ification plants. The inlet stream to a Stretford process may
contain a variety of gaseous sulfur species (H2S, COS, CS2, etc.)
The outlet sulfur-containing streams consist of by-product
sulfur, sulfur species in the blowdown Stretford solution, and
gaseous sulfur species in the treated gas and oxidizer vent
gas. Therefore, to assess the performance of the Stretford
process in controlling sulfur species, an elemental sulfur
material balance should be attempted along with determining the
concentrations of specific gaseous sulfur species in the inlet
and treated gas streams.
The pollutants to be measured in control equipment
characterization tests are defined by either the results of
waste stream characterization tests or from the results of the
engineering analysis. The data collected from these tests
should provide guidelines for selecting and/or developing
adequate pollution control processes.
2.3 PROCESS STREAM CHARACTERIZATION
The purpose of a process stream characterization test
is to provide information on the relationship between process
operating parameters (inlet stream composition, temperature,
pressure, etc.) and the characteristics of the process waste
streams. This type of test is similar to control equipment
characterization tests in that performing material balances
around the process and/or measuring specific compounds in the
process inlet and outlet streams may be required.
An example of a process stream characterization test
would be to determine the effect of coal sulfur content on the
-8-
-------
amount of gaseous sulfur species in the raw low-Btu gas from an
atmospheric, fixed-bed gasification process. To accomplish
this, an elemental sulfur material balance would be performed
around the gasification process along with a measurement of the
concentrations of sulfur species in the coal and in the raw
product gas. The results of the material balance could be use-
ful in defining an adequate process for removing the gaseous
sulfur species (H2S, COS, CS2, etc.) found in a typical raw
product gas stream.
In summary, there are various types of environmental
tests that can be performed to accomplish specified environ-
mental objectives. Waste stream characterization can be asso-
ciated with a plant (or a process) or used to define needs for
specific control equipment. Control equipment characterization
is useful in evaluating the effectiveness of equipment used for
treating process waste streams. Process stream characterization
is useful in determining the relationships between process
operating parameters and process waste stream characteristics.
It must be emphasized that while performing any of
these environmental tests, the plant operating conditions
(normal operation, start-up, steady-state, and emergency up-
sets) must be defined. These desired plant conditions should
be determined by performing an engineering analysis of the
plant.
-9-
-------
SECTION 3.0
ENGINEERING ANALYSIS
In this section, the general principles involved in
performing an engineering analysis of a candidate gasification
plant test site are discussed. Application of the principles
discussed is illustrated by a specific example.
3.1 GENERAL PRINCIPLES
An engineering analysis should provide a sound basis
for translating a set of general test objectives into a set of
specific process data needs. The element which is fundamental
to any engineering analysis is the development of an accurate
understanding of the functions of all environmentally signifi-
cant process operations within a plant. The specific steps
which would usually be involved in developing this understand-
ing include:
constructing an up-to-date plant flow sheet
showing all process and emission streams which
are relevant to the objectives of the test
program,
characterizing the process and waste streams
of interest with respect to their flow rates,
compositions, and physical characteristics,
and
identifying (and defining the normal operating
levels of) process variables which affect
the characteristics of the process or waste
streams being considered.
The important factors involved in each of these areas are
addressed in the following sections.
-10-
-------
3.1.1 Process Flow Sheet Development
The development of an up-to-date process flow sheet is
a necessary first step in any engineering analysis because it
provides a convenient mechanism for identifying:
the major processing steps and equipment
items found at a given facility,
the input and output (including waste
streams) associated with each equipment
item, and
the process characteristics which might
be expected to have an impact on waste
stream properties.
Depending on the complexity of the facility in question, it may
be possible to represent the whole plant on a single flow sheet,
or it may be necessary to use several sheets. An example of a
relatively simple facility is the gasification plant
which is described later in this section. Extremely complex
facilities such as the conceptual El Paso Burnham gasification
plant are best handled by: 1) breaking the overall plant into
sections (see for example, Figure 3-1), and 2) presenting indi-
vidual flow diagrams for the various sections. Figure 3-2, which
shows the processing units associated with the Burnham plant
gasification section, is indicative of the level of detail which
is desired on a test plan process flow sheet.
In most cases, existing plant flow diagrams (e.g.,
piping and instrumentation diagrams) can be used as a basis for
constructing suitable flow diagrams for test planning purposes.
In cases where this type of resource is not available, a flow
diagram should be constructed from information gathered during
a preliminary site screening-visit.
The importance of a site visit in terms of its poten-
tial value to a test planning effort cannot be emphasized
enough. In order to derive the maximum possible benefit from
a site visit, it should be recognized that a number of informa-
tion needs should be addressed during a site visit.
-11-
-------
TO ATMOS. McHE
TO INCINERATOR
N>
I
AIR
ELEMENTAL SULFUR
PIPELINE GAS
— —— RECLAIMED WATER TO
IN-PLANT USERS
AMMON4A SOLUTION
TOSoLFUR PLANT POWER TO
PLANT USERS
Figure 3-1. SIMPLIFIED PROCESS FLOW DIAGRAM - BURNHAM COAL GASIFICATION COMPLEX
(REF. 2)
-------
LO
o
o
iiSC£__iOG._
s'i.Cb
\^\ "VN.
i^L:
}::>
__l}*i
•...j'^ajjgjo1
,oii_4j3fl.*S'
t 5196
5S60'
"sTTr
."ii
36 i
j[aw¥T
•3.^ 38«
s.*5;,<;i
j.^T_4[_
~S3t!~
; --9.,*iii ioo~
Zl.iQ
I.O33.5O1
4t it
^989"
.,•'02.837^
Figure 3-2. EL PASO BURNHAM GASIFICATION PLANT SECTION (Ref. 3)
-------
First of all, as discussed above, an accurate, up-to-
date plant flow sheet should be developed to serve as an aid in
establishing test program priorities (e.g., objectives, scope,
sampling strategy, etc.). A plant visit can play an important
role in the development of such a flow sheet.
The physical characteristics of the plant should
receive specific attention during a site visit. In particular
potential sampling point locations should be carefully screened,
taking into consideration:
Suitability - can the samples needed to
fulfill the test objectives be obtained?
Important considerations here include
existing vs. new sampling port, port
size, port location (e.g., relative to
upstream/downstream obstructions), and
characteristics of the stream to be
sampled (e.g., pressure, temperature,
flow velocity).
Accessibility - availability of the sampling
platforms/scaffolding, electrical connections,
etc. needed to safely accommodate the work
space requirements of the sampling team.
The sizes and physical layout of the various plant
components should be noted during the site visit so that
appropriate residence time considerations can be taken into
account in planning the sampling effort. The physical dimen-
sions of lines and ducts should be determined in order that
average flow velocities can be estimated from known stream
volumetric flow rates .
In addition to the information needs which were just
discussed, a considerable amount of process data should be
gathered during a site visit. Some of these data needs are
addressed in the following two sections.
3.1.2 Process and Waste Stream Characterization
Once the basic plant flow scheme has been defined, all
environmentally significant process and waste streams shown on
the flow sheet should be characterized with respect to their
-14-
-------
flow rates, compositions and physical characteristics. Obviously,
an environmental test program will tend to focus on waste streams
rather than process streams. However, in many cases, certain
process stream characteristics will be of interest because of
their effects upon the waste streams produced in downstream
processes.
It is very important to consider the need for gathering
accurate process characterization data during a preliminary site
screening visit. Some of the reasons for this are as follows:
Quite frequently, published process data
such as that shown on process flow sheets
(see Figure 3-2, for example) will reflect
design or average, rather than actual
process conditions.
An appreciation for the variability of the
process can be gained by observing its
operation, by reviewing plant operating
logs and by discussing the magnitudes of
and the driving forces for those variations
with the plant operators.
Another issue which should be addressed during the
site visit is the availability of desired process monitoring
instrumentation. Generally, an operating plant will not be
equipped with all of the instrumentation needed to fully char-
acterize the operation of the process. Particularly in a
commercial plant it would be more typical to find that the
minimum number of process monitors needed to safely control
the process would be used. A developmental or demonstration
unit would tend to be better instrumented.
Available data on stream characteristics should be
considered in an engineering analysis to provide a basis for
establishing test priorities. If time and/or budget con-
straints exist, it will generally be reasonable to concentrate
test program efforts on:
the most environmentally significant
streams in terms of their mass emission
rates of specific hazardous compounds
or classes of compounds,
-15-
-------
the streams which are the most applicable
to broad rather than specialized areas of
a technology, and
the streams about which the least is known.
Some of the important considerations involved in each of these
areas are addressed in the example which is discussed in
Section 3.2.
3.1.3 Definition of Process Variable Effects
This task addresses one of the more difficult areas
in test plan construction and in some cases may be beyond the
test plan scope. The proper execution of this task requires a
thorough understanding of the plant and its component processes,
Once this understanding is gained, however, specifying the vari-
ables which would be expected to affect the characteristics of
potential waste streams is reasonably straightforward. Ulti-
mately, this effort should also provide a basis for defining
what process data should be collected in conjunction with the
execution of the task so that subsequent analyses of the test
results will yield useful correlations.
The recommended approach to addressing the problems
inherent in this task is a multifaceted one. First of all,
available data from the candidate site should be evaluated.
Applicable information from other sites or from related tech-
nology areas should also be considered to determine which pro-
cess variables would be expected to have the major impacts on
the facility's waste stream characteristics.
A modular type of analysis (symbolically represented
in Figure 3-3) is a useful way to approach this problem. In
simple terms this approach involves defining the expected char-
acteristics of the waste stream from a process or group of pro-
cesses in terms of their input stream characteristics and oper-
ating parameters. In theory, an entire plant could be analyzed
using this approach. However, it is probably more reasonable to
start with an application of this technique to the specific
equipment items from which the various plant waste streams
originate.
-16-
-------
Waste
Streams
11
input
S treams
7
>>
f
Process (es)
' Output
^ streams
Operating Parameters &
Equipment Characteristics
(Temperatures, pressures,
residence times, etc.)
Figure 3-3. MODULAR APPROACH TO PROCESS ANALYSIS
The discussion which has been presented to this point
has described in general terms the major tasks involved in an
engineering analysis. The following section illustrates how
these techniques can be applied to a specific gasification
facility.
3.2
SPECIFIC EXAMPLE - ENGINEERING ANALYSIS OF A CHAPMAN
GASIFICATION PLANT
The Chapman gasification plant selected for this exam-
ple is an actual facility which was tested as part of Radian's
low-Btu gasification technology data acquisition program. This
facility was selected as a test site for the low-Btu program
for the following reasons:
It is one of only a few commercial-scale
gasification units which is operating in
this country.
The facility has a well-defined operating history,
It uses fixed-bed, atmospheric pressure,
single-stage gasifiers which are repre-
sentative of the gasifier type which is
-17-
-------
currently being considered for widespread
commercial use in this country.
The plant uses bituminous coal, a widely
available feedstock.
The plant includes a gas quenching-
scrubbing system that provides a means
of evaluating the tar and oil by-products
associated with a raw gas cooling operation.
For purposes of this discussion, it will be assumed
that the objective of the test program being developed here will
be to characterize the emission streams leaving the Chapman
facility. The first step to be followed in the pursuit of this
objective involves the development of an accurate description
of the facility, including:
a process flow sheet showing all significant
process and waste streams, and
a discussion of the significant operating
characteristics of the plant, with emphasis
on the factors which affect the characteristics
of the plant's waste streams.
It should be noted that the following section was gen-
erated from information gathered during a preliminary visit to
the Chapman site. The discussion presented here is representa-
tive of the level of understanding needed to make reasonable
test program decisions.
3.2.1 Process Description
The coal gasification facility discussed in this
example produces low-Btu gas which is used as a process furnace
combustion fuel. While twelve Chapman gasifiers are operational
at this facility, only two gasifiers are operated at any one
time to meet current fuel demands.
Four processing operations are used in the plant:
(a) coal handling; (b) gasification; (c) gas purification,
-18-
-------
which includes particulate removal and gas quenching and
scrubbing; and d) water (process condensate) treatment. A pro-
cess flow sheet for the Chapman gasification unit is presented
in Figure 3-4. This diagram also shows the air, water, and
solid waste streams associated with each process operation. In
the following text, each of these operations and their respec-
tive multimedia waste streams are discussed in more detail.
Coal Handling -
The coal handling operation at the Chapman facility
consists of: a) delivery/storage of presized Virginia bituminous
coal in hopper cars, b) conveying, and c) storing this coal in
the gasifier feed hoppers. No coal grinding, crushing, sizing,
or drying operations are used at the plant site.
The major waste streams associated with the coal
handling operation are particulates from coal conveying and
coal storage. The compositions of these emitted particulates
should be similar to that of the coal feedstock.
Gasification -
The gas producers are single-stage, atmospheric,
fixed-bed, air-blown Chapman gasifiers. The coal feedstock
enters the top of the gasifiers through a barrel valve and is
spread across the bed by a distribution arm. Steam and air
introduced into the bottom of the gasifier pass through a grate
which evenly distributed these gases and also supports the coal
bed. Ash from the gasifier is collected in a water sealed ash
pan and removed from the unit using an ash plow. The hot raw
gas exits the top of the gasifier at 840-940°K (1050-1250°F)
and enters a cyclone to remove entrained particulates (ash and
coal dust). Pokeholes located on top of the gasifier are
positioned so that rods can be periodically inserted to break
up any coal agglomerates which form.
The discharge streams associated with the gas produc-
tion operation are: a) gaseous emissions from the barrel valve,
pokeholes, and leaks around the gasifier seals; and b) moist
ash exiting the bottom of the gasifier. These gaseous emissions
are released to the atmosphere.
-19-
-------
O
I
Coal Dust
Barrel Valve Pokehole Liquor Trap
Vent Gases Gases Vapors
Fugitive
Separator
Vapors
Gaslfier Collected
Ash Particulars
By-Product Tars
and Oils to
Utility Boilers
Evaporator
Gases
Low-Btu Gas to
Process Furnaces
Figure 3-4. SIMPLIFIED PROCESS FLOW DIAGRAM FOR THE CHAPMAN GASIFICATION PLANT
-------
Gas Purification -
The gas purification operation at the Chapman facility
consists of the following steps:
particulate removal, and
quenching and scrubbing.
These steps are described in the following sections.
Particulate Removal -
Particulate removal is accomplished in a hot, refrac-
tory lined cyclone that operates at a temperature slightly lower
than the gasifier overhead temperature. Each gasifier is
equipped with its own cyclone. The particulates removed by the
cyclones consist of coal dust, ash and tar entrained in the raw
gas. These particulates collect at the bottom of each cyclone.
Pokeholes are located in the top of each cyclone and in the hot
gas ducts so that steam lances can be periodically inserted to
break up agglomerated particulates.
Atmospheric emissions from the cyclones consist of
pokehole vent gases and leaks. The collected particulates,
which constitute a solid waste stream, are combined with the
gasifier ash for disposal.
Quenching and Scrubbing -
The hot gas leaving the cyclones is quenched by a
series of sprays located inside the exit lines from each cyclone.
Excess quench water is collected in a series of liquor traps
(one trap to each gasifier/cyclone). The gas from all operating
liquor traps enters a collecting main. Water sprays located
inside this main cool the gas to approximately 340°K (150°F).
Excess tar and quench liquor from both the liquor traps and the
collecting main are directed to the liquor separator. Pitch
which accumulates in the liquor traps is periodically collected
and disposed of off-site.
-21-
-------
After the initial quenching step, the gas is scrubbed
and further cooled in two tray scrubbers which are operated in
parallel. In this operation most of the tars, oils, and parti-
culates are removed as the gas is cooled to approximately 330°K
(135°F).
The gas exiting these tray scrubbers is combined and
compressed before entering the final spray scrubber. In this
spray tower, some residual tars, oils, and particulates are
removed as the gas is further cooled to 320°K (120°F). The
effluent liquor from both the spray and tray scrubbers is
returned to the liquor separator.
The liquor separator at the plant is a large concrete
tank (approximately 5 x 13 x 2 meters). Process condensate and
any condensed tars and oils from the quenching/scrubbing steps
described above enter at one end of the tank. A series of
baffles is used to minimize the turbulence caused by the incom-
ing liquor and to keep oils and tars which have settled to the
bottom of the separator from entering the clean liquor uptake
line. The tars and oils which accumulate in the separator are
burned as an auxiliary fuel in a coal fired boiler. The water
is recirculated to the quenching and scrubbing operations.
The.discharge streams from this operation include
fugitive emissions from the liquor traps, and vent gases from
the liquor separator. A steam ejector is used to vent the
vapor space above the liquor in the separator.
Water Treatment -
Water treatment problems are minimized at the Chapman
facility by operating the gasification process such that there
is no net accumulation of water. If excess water accumulates,
it is evaporated. Air emissions from the evaporator should
contain volatile materials found in the condensate.
Waste Stream Summary -
Potential waste streams identified in the Chapman
gasification facility include:
-22-
-------
Gaseous Emissions
Barrel valve vent gases
Fugitive emissions from the gasifier
and cyclone pokeholes
Gas/liquor trap vapors
Liquor separator vapors
Evaporator vapors
Stack gases from the process heaters
Particulates from coal conveying and
storage operations
Liquid Effluents
By-product tars and oils
Solids Wastes
Gasifier ash
Collected particulates (cyclone)
Pitch from liquor traps
Sludge from liquor separator
These discharge streams and their expected compositions can be
summarized in general terms as follows.
The compositions of the pokehole, and barrel valve
vent gases should be similar to that of the raw gas exiting the
gasifier. The gaseous emissions from the liquor trap, liquor
separator and evaporator should contain a wide range of volatile
inorganic and organic compounds most of which should be found in
the condensate or tar/oil fraction as well. The only liquid
effluent from the Chapman gasification plant is the by-product
tar and oil stream. The four types of solid wastes identified
above would all be fairly unique materials.
-23-
-------
3.2.2
Waste Stream Prioritization
Gas Streams -
On a gross emission rate basis, the process heater
stack gas is the most significant gaseous emission stream at
the Chapman facility. The concentrations of specific hazard-
ous components present in the stack gas should be small, however.
Stack
Gas
Low-Btu
PrnHnrf N-
Gas
Process
Heaters
Combustion
Air
The fates of environmentally hazardous low-Btu pro-
duct gas components in a combustion process is of interest to
the program. Certainly, most of the organic materials present
in the product gas should be oxidized to less harmful flue gas
components in a combustion process. However, it would probably
be reasonable to screen the stack gas using waste stream char-
acterization (Level 1) sampling and analysis procedures. Then,
if positive indications of harmful levels of specific compon-
ents were obtained, subsequent studies to more fully character-
ize the stack gas would be justified. These studies might
entail defining the emission rates of specific flue gas compon-
ents as functions of the input fuel gas composition and the
operating characteristics (e.g., excess air firing rate) of
the furnaces.
The characteristics of the barrel valve vent gas
should be determined by direct sampling. This should be done
to determine the extent to which raw gas components such as tars
are sorbed by the coal feed.
-24-
-------
The characteristics of the various fugitive emis-
sions could be determined most efficiently by sampling the raw
gasifier product gas. If emission rate estimates indicated that
those streams were a significant emission source, an attempt
should be made to measure the rates and compositions of those
emissions directly.
The liquor separator vent is another significant
gaseous emission stream. Simultaneous sampling and subsequent
analyses of both the vent stream and the separator liquor are
desirable here since this should provide a basis for relating
the composition of the vent gas to levels of volatile materials
found in the separator liquor.
Vent Gas
Ambient Air
Inleakage
Inlet Quench
Liquor &
Tars/Oils
Steam to Ejector
Quench Liquor
Tar/Oil
-> Quench Liquor Out
To Boiler
It would not be recommended that any significant time
or effort be devoted to the characterization of particulate
emissions from the coal handling operations at the Chapman
plant. This is because:
these emissions are not unique to a
gasification plant, and
adequate characterization data for
this type of emission source already
exist in the literature.
Evaporator vent gases would not need to be sampled
for the following reasons.
-25-
-------
This is an intermittent stream which is
generated only when the evaporator is
operated (which, in theory, is
infrequently).
The characteristics of this stream
should be similar to those of the
separator vent.
Likewise, gas/liquor trap vapors would not be recommended for
sampling because the components in this stream should be similar
to those encountered in the liquor separator vent stream. A
method of estimating the flow rate of this stream, however,
should be investigated.
Liquid Streams -
While the tar/oil stream, literally, is not an
effluent stream at the Chapman facility (it is stored and sub-
sequently burned), it should probably be sampled. The analysis
data gathered for the tar should consist of the following:
a broad organic screening analysis to
identify classes of hazardous compounds
present in the tar, and
a detailed characterization of the
light ends fraction since this would
help to define the expected com-
position of potential storage tank
vent gases.
Analysis of the tar/oil stream would also provide a
basis for defining the distributions of key pollutants (NH3,
sulfur species, cyanides, trace metals, etc.) between the
aqueous and the organic layers in the liquor separator.
-26-
-------
Solid Waste Streams -
Coal .. . >
(Evaporated
in the air
stream)
Gasifier
Gasifier
Ash (Wet)
Raw
"" froduct -*
Gas
I
Cyclone
Collected
'articulates
"Particulate-
Free" Product
Gas
The gasifier ash and the collected particulates
exiting the bottom of the gasifier and the cyclone, respec-
tively, should clearly be sampled and characterized using Level
1 procedures because these are the two major solid waste streams
produced at the Chapman facility. The flow rates of these
streams should also be determined.
Samples of the liquor trap pitch and the liquor sepa-
rator sludge should also probably be analyzed to determine the
significance of these materials as potential sinks for certain
pollutants (heavy metals, for example). Net production rates
for these materials would have to be estimated based on informa-
tion obtained from the plant operating staff, i.e., the fre-
quency of cleanup and the quantity of material removed.
The operating data that should be collected during
these tests are presented in Table 3-1. These data will be used
in conjunction with the results of the sampling task to per-
form an engineering analysis of the plant's performance and to
insure that the plant is operating at desired conditions when
the samples are collected.
-27-
-------
Table 3-1. PROCESS DATA TO BE COLLECTED DURING SAMPLING
Gasifier
Coal feed rate
Pressure
Air flow rate, temperature and moisture content
Outlet gas temperature and flow rate
Ash flow rate and temperature
Cyclone
Overhead temperature
Accumulation rate of collected particulates
Tray Scrubbers
Inlet gas temperature and flow rate
Outlet gas temperature and flow rate
Inlet liquor temperature and flow rate
Outlet liquor temperature and flow rate
Separator
Tars/oils - net production rate
Spray Scrubber
Inlet gas temperature and flow rate
Outlet gas temperature and flow rate
Inlet liquor temperature and flow rate
Outlet liquor temperature and flow rate
Quench Water
Inlet temperature and flow rate
Outlet temperature and flow rate
Product Gas
Temperature and flow rate
-28-
-------
SECTION 4.0
SAMPLING STRATEGY
In this section, the selection and execution of sam-
pling procedures to be used in producing reliable waste stream,
control equipment, and process stream characterization data
are discussed. In developing a sampling strategy for an envi-
ronmental test, the following items should be addressed:
sampling point selection, including location,
sampling methods selection,
sampling frequency and timing, and
sampling quality control program.
The selection of sampling points and methods should
be based upon the accuracy requirements of the test, the phys-
ical condition of the stream to be sampled (e.g., temperature
and pressure), analytical requirements with respect to sample
size or sample pretreatment requirements, the expected chemical
composition of the stream, potential component reactivities
(sample stability and safety considerations), and the physical
arrangement of piping and ducting containing the stream to be
sampled. Other considerations may include sampling practices
of plant personnel, effects of sampling on plant operation,
and safety and work area constraints.
Sampling frequency and timing involve decisions on
how often to sample and when to sample, respectively. Sampling
frequency constraints include the sampling method itself, plant
operational variations, quality control requirements, and data
evaluation needs. Sampling timing is process dependent and is
defined by the desired plant operating conditions (normal oper-
ation, start-up, shutdown, and upsets) and the type of environ-
mental test (waste stream, control equipment, and process stream
characterization) being performed.
-29-
-------
A sampling quality control program should be used to
insure that the data collected during the environmental test
are both accurate and precise. This program should include the
following elements:
calibrating sampling and analytical
equipment,
taking replicate samples,
performing replicate analyses,
using alternative (if available) sampling
and analytical methods,
comparing with plant personnel obtained values,
and
establishing a clearly defined chain of respon-
sibility for sample collection, sample analysis,
and evaluation of the results.
In the following text, each of the four major areas
involved in developing a sampling strategy is discussed. Spe-
cific sampling methods are presented in the Appendix. Examples
are also given to illustrate how and why these items are incor-
porated into an environmental test plan.
4.1 SAMPLE POINT SELECTION
The process considerations for sampling point selec-
tion (including process stream selection and accessibility) are
discussed in the Engineering Analysis Section (Section 3.0).
This section is concerned primarily with selecting the actual
sampling point locations and focuses upon problems related to
stream flow characteristics.
In the following text, methods for selecting sampling
points for gases, liquids and solids are presented.
-30-
-------
4.1.1 Sampling Point Selection for Gases
The flow characteristics of the gas stream to be
sampled are a basic consideration that affects the ease with
which a representative sample of the total stream can be col-
lected. EPA's Method 1 guideline document (Ref. 4) presents
specific criteria for selecting a gas stream sampling port.
The EPA approach is based upon selecting a sampling location
having minimum turbulence and a capability for traversing a
representative cross section of the stream to provide a statis-
tically valid sample. The number and locations of the traverse
points are dependent upon the distance from the sampling point
to the nearest disturbance in the stream (e.g., elbow or tee)
and the specific configuration of the pipe or duct containing
the gas streams. Guidelines for sample point selection for
stacks and ducts are shown in Figure 4-1.
Small pipes and ducts present a special problem in
the selection of a sampling point. The traverse approach of
Method 1 requires that the sampling port (or ports) be con-
structed at a right angle to the direction of flow. This allows
traverses across the stream's cross-sectional area to be made.
For small pipes, the sampling port itself may be a source of
turbulence. As a general rule, if the cross-sectional area of
the required sample port is greater than one-fifth of the
cross-sectional area of the pipe to be sampled, the port should
not, if possible, be constructed at the proposed sampling
point. The alternative recommended is to find a section of
pipe having a minimum length of 10 pipe diameters and terminating
downstream in an elbow. The sampling port should then be con-
structed in the elbow such that a probe can be inserted into
the center of the stream at a distance of two pipe diameters
upstream of the elbow (see Figure 4-2). The probe's cross-
section should be less than one-fifth of the pipe's cross-
sectional area to avoid significant turbulence caused by the
probe.
The previous discussion presents the criteria which
apply to ideal sampling point locations. However, for certain
environmental tests, such as a Level 1 waste stream character-
ization test, the accuracy requirements (within a factor of ±2)
are such that installing new sampling ports may not be justified.
This is particularly true when there are existing sampling ports
in the pipe or duct. A judgment must then be made whether the
existing sample port locations will meet the accuracy require-
ments of the environmental test.
-31-
-------
0.5
*NUMBER OF DUCT DIAMETERS UPSTREAM
DISTANCE A
1.0 1.5 2.0
2.5
*FROM POINT OF ANY TYPE OF
DISTURBANCE (BEND,EXPANSION.CONTRACTION,ETC.)
i I I I I I
345678
*NUMBER OF DUCT DIAMETERS DOWNSTREAM
DISTANCE B
Minimum number of traverse points.
o
_____
0
o
o
,
o
0
1
1
o i o
_j.
1
0 1 0
1
1 i__
1
0 1 0
1
1
Example showing rectangular stack cross section divided into
12 equal areas, with traverse points at centroid of each area.
TRAVERSE DISTANCE
POINT * of diameter
l<*. t
20.5
TO.5
35.3
95.6
X/ fTT
y\ 11!
T
* t
Example showing circular stack cross section divided into 12
equal areas, with location of traverse points at centroid of
each area.
Figure 4-1. NUMBER AND LOCATION OF TRAVERSE POINTS DEFINED BY
EPA METHOD 1 GUIDELINES. (Ref. 4)
-32-
-------
uo
I
Sampling
/Port
/
tr 2d Jf ^fl(l ^1
- K - *« ' °" -1
* ^ \ A
., /
1 \ iv
f — v^
r \
Probe
Gas
* Flow
Figure 4-2. SAMPLE PORT ARRANGEMENT FOR "HEAD-ON" SAMPLING
-------
4.1.2 Sampling Point Selection for Liquids
The preferred sampling points for enclosed liquid
streams are existing valves, either in-line or on a side stream.
These valves provide a ready sample source from the stream and
should be used when compatible with the objectives of the test
program. Another point of easy access is at outflow orifices
where the liquid streams flow into ponds, tanks or other open
vessels. Open or noncontained streams may be sampled at any
point compatible with the accuracy requirements of the test.
A major restriction in selecting sampling points is
stream homogeneity, particularly for liquids composed of mixed
phases (e.g.., aqueous and organic). To ensure a well-mixed
sample, sampling should be done downstream from points of
turbulence, such as elbows or pump-discharge lines. It may be
necessary to have sampling valves installed at points where
none exist.
4.1.3 Sampling Point Selection for Solids
Samples of solid streams may be collected from stor-
age piles or bins, transport containers (e.g.,, trucks and rail-
road cars), or conveyors. Often it is desirable to composite
solid samples obtained over a specified time period. Generally,
samples collected from storage piles, bins and transport con-
tainers are already composite samples. However, samples
obtained from conveyors represent the solids being used or pro-
duced during a specific period of plant operation. Therefore,
those streams may need to be sampled at a specified frequency
in order to obtain a representative composite sample.
4.1.4 Other Sampling Point Considerations
In selecting sampling points for an environmental
test, special considerations which may be involved in the final
decision include:
sampling practices of plant personnel,
effect of sampling on process operation, and
safety and work area requirements.
-34-
-------
Most facilities have established sampling and analysis
methods for routine process monitoring requirements. The sam-
pling points and methods used by plant personnel should be eval-
uated with respect to the requirements of the particular test
program. If these sampling points are acceptable, they should
be used. Many times the sampling team can arrange to have plant
personnel catch extra samples at their established sampling
points, thus reducing the sampling effort.
Another consideration for sample point selection is
the effect of the sampling effort upon the operation of the
process itself. Sampling points must be selected so that entry
into a stream will not adversely affect the process. This
potential problem should be thoroughly discussed with plant
personnel and if any danger of plant upset exists, operator
approval should be obtained.
Safety and work area requirements are very important
criteria for sampling point selection. The noise level, temper-
ature, and atmospheric conditions to which the sampling team
will be subjected must be considered. Additional considerations
should include access to the sampling point, the availability
of safe scaffolding or ladders and the availability of suitable
means for transporting equipment to the sampling site. Problems
associated with worker exposure to potentially hazardous mater-
ials must also be addressed.
4.2 SAMPLING METHOD SELECTION
The careful selection and execution of sampling pro-
cedures is the most critical step in producing reliable environ-
mental test data. Samples must accurately represent the compo-
sition of the stream sampled and must be compatible with the
analytical techniques to be applied. Factors which must be
considered in order to maintain sample integrity and provide a
representative sample include:
stream composition as well as spatial and temporal
variations in composition,
potential for sample contamination and for changes
in sample composition following its removal from
the stream,
-35-
-------
sampling equipment reliability and size,
possible limitations of the analytical techniques
to be used, and
the accuracy requirements of the test.
In addition, budget constraints will always be an important
consideration.
Obtaining a small sample which is statistically repre-
sentative of a much larger quantity of material (composite sam-
ple) is a common sampling problem. Spatial variations in compo-
sition can be averaged by compositing aliquots collected over
the cross section of a flowing stream or throughout the volume
of a static storage vessel or pile. However, it is best to
avoid these variations by selecting a sampling location where
the material is well mixed. Temporal variations can occur for
a variety of reasons, ranging from stratification in a storage
vessel to process operating fluctuations. They can be averaged
by compositing aliquots collected over a period covering sev-
eral process cycles or characterized in detail by analysis of
each aliquot.
Samples may undergo changes in composition during
sample collection (for example, vaporization of light organics
from liquid samples or leaching of contaminants from the sam-
pling equipment), periods of storage, or transport. Such
changes must be minimized. For example, the loss of dissolved
gases in a liquid sample may be prevented by collection and
storage under pressure. Changes resulting from chemical reac-
tions may be inhibited by storage at low temperatures or by
designing chemical treatments specifically for the preservation
of particular components. Unless a proven technique is avail-
able, time-stability studies to develop suitable preservation
procedures or immediate on-site analysis should be considered.
The reliability of sampling equipment can impact the
choice of sampling method. Generally, the more complex the
equipment, the harder it is to maintain reliable operation for
the duration of the sample collection period. In addition, as
the sampling equipment increases in complexity, it generally
increases in size. This may be a critical consideration if
space is at a premium at the sampling location.
-36-
-------
Sampling procedures may be grouped into two basic
categories - manual (non-continueus) and automatic (continuous)
sampling methods. Manual methods (e.g., sampling trains) are
generally more flexible, more easily executed, and more labor
intensive than continous methods. Manual methods are often
used as precursors to the installation of continuous monitors
(on-line instrumentation) or for calibration of continuous
sampling devices. Continuous techniques tend to be less flex-
ible and more capital intensive than manual methods. Therefore,
continuous techniques are usually preferred for long-term appli-
cations, whereas manual ones are often more appropriate for the
short-term.
Presented in the Appendix is a detailed discussion of
proven techniques for collecting representative samples from
the three major stream types: gas, liquid, and solid. Sampling
procedures for mixtures of these phases are addressed under the
section dealing with each major phase. For example, sampling
gases containing entrained particulate matter is discussed in
the gas sampling section. Sampling liquid-solid slurries is
discussed in the liquid sampling section.
4.3 SAMPLING FREQUENCY AND TIMING
Sampling frequency and timing are concerned with how
often and when a sample should be collected. Sampling frequency
(how often) is determined by the sampling method itself, the
type of environmental test being performed, plant operational
characteristics, quality control requirements, and data evalua-
tion needs. Sampling timing (when) is concerned with insuring
that the samples are collected under the desired plant operating
conditions as defined by the test plan objectives.
4.3.1 Sampling Frequency
Sampling frequency may be method or process limited.
Some sampling methods require several hours for collecting a
suitable sample, such as obtaining a large sample of particulate
matter in a gas stream having a low particulate loading. In
such a case, sampling frequency may be limited to one sample
per day. Other methods require only short collection times,
such as collecting grab samples of a solid stream. These meth-
ods usually entail higher sampling frequencies and compositing
techniques (see Appendix) if valid average samples are to be
obtained.
-37-
-------
There are two general types of processes (continuous
and cyclic) that need to be discussed with respect to sampling
frequency. For continous processes (e.g., acid gas removal
processes), the sampling time should be sufficiently long to
average out normal process variations. Sampling cyclic or
intermittent processes (e.g., gasifier coal feed lock hoppers)
can be approached in either of two ways:
If the steps in the cycle are relatively long
and well-defined, individual samples can be
taken for each step.
If the steps are short and undefined, sampling
periods which span one or more cycles of oper-
ation can be used.
In some cyclic processes, there will be no input or output
flows except at the beginning or end of the cycle. This would
be the case for coal feed lock hoppers. In those cases, sample
collection should take place only during flow conditions.
As previously indicated, the coal feed lock hopper is
a good example of a cyclic process. In a Lurgi gasification
plant, coal feed enters the gasifier through a lock hopper
which cycles approximately every 12 minutes. Product gas is
often used to pressurize the hopper before coal is fed into
the gasifier. After the hopper is emptied, it is sealed and
depressurized. The gas resulting from the depressurization
step is the hopper outlet gas. Sampling the lock hopper might
involve measuring key species in the pressurizing gas at the
beginning of the cycle and measuring those same species in the
outlet gas at the end of the cycle.
The sampling frequency can also be affected by the
stipulations of the sampling quality control program (see
Section 4.4) and by the data evaluation requirements (see
Section 6.0). Sampling method limitations or plant operating
characteristics may prevent full adherence to the sampling
frequency specified in the test plan. If this occurs, devia-
tions should be made with care such that the samples collected
will still be representative of the streams and meet the test
objectives.
-38-
-------
4.3.2 Sampling Timing
The sampling timing (when to collect samples) is
established primarily by the plant operation and the type of
environmental test being performed. The plant operating condi-
tions (normal operation, start-up, shutdown, emergency upsets)
for the test will be specified by the test plan objectives.
The following factors are important in defining the timing for
sample collecting.
For sampling at normal plant operating
conditions, the plant should be given
sufficient time to stabilize. Process
data should be monitored to determine
that these conditions are present. The
engineering analysis of the plant should
be used to define normal plant operation.
For sampling during start-up, shutdown,
and/or emergency upsets, sampling timing
is critical and demands continuous or
frequent sampling methods. The types of
methods employed will depend upon the
time required for start-up, shutdown,
or the duration of the upset.
Sampling timing is most critical when determining the
distribution of species or elements in process input and out-
put streams (material balance tests). In material balance
tests, the results of the engineering analysis of the process
should be used to help define sampling timing. The engineering
considerations include:
residence time of the material in the process,
and
chemical reactions occurring in the process.
The residence time of the material in the process is dependent
upon the flow characteristics of the material through the pro-
cess. Knowledge of these flow characteristics allows estimates
to be made of when the material or products of reactions invol-
ving the material will be present in the various process output
streams.
-39-
-------
A brief discussion of a sulfur balance around a coal
gasifier will illustrate this concept. The residence time of
a coal particle in a coal gasifier may be many hours. This in-
cludes the time required for the particle to enter the gasifier,
react with steam and oxygen, and exit the gasifier (as ash).
During most of this time interval, the coal particle will be
continually undergoing devolatilization or gasification reac-
tions. Therefore, sulfur species from the coal particle can
contribute to the raw outlet gas sulfur species during the
entire particle residence time.
In order to obtain a sulfur balance around the gas-
ifier, several grab samples of inlet coal should be collected.
The sulfur content in the outlet gas (including particulates,
tars, and gaseous sulfur species) should be collected during
the entire residence time interval. Several grab samples of
ash should be collected before and after the calculated coal
residence time in the gasifier. This time interval for col-
lecting ash samples is necessary because of the possibility of
channeling or recycle of coal particles inside the gasifier.
In summary, the engineering analysis of the plant
should be used to define when to collect a sample in order to
accomplish the test objectives. In many cases, judgments must
be made on the time it takes a material to pass through a pro-
cess (residence time) and when products of reactions will be
present in the various process output streams.
4.4 SAMPLING QUALITY CONTROL PROGRAM
The purpose of a sampling quality control program is
to prevent propagation of determinative errors (bias) through
the sampling/analysis/evaluation phases of the test program
The specific items which should be addressed in a quality con-
trol program include:
facilities and equipment inventory,
training program,
document control,
quality control charts,
-40-
-------
supervision,
materials inventory and procurement,
reliability and maintenance,
data validation,
equipment calibration, and
correlation tests.
Reasonable inventories of sampling and analytical accessories,
spare parts, etc. should be maintained for the duration of the
test. Periodic checks of the on- and off-site laboratories,
sampling platforms, and data logging and processing areas
should be performed to insure that they are in good condition.
Prior to the initiation of field work, all personnel
involved in the test should be thoroughly briefed on the goals
of the sampling and analysis program and the procedures to be
followed including the use of special sample acquisition systems.
All personnel should also be made aware of the need for instru-
ment calibration, replicate samples, and other elements of
quality control.
Provisions should be made to maintain complete secur-
ity of sample log books, sampling and analytical data recording
forms, and operating procedure documents to insure that data
are not lost or mishandled. Duplicate record keeping is recom-
mended along with specific procedures for distributing and
making revision to documents affecting the quality control
program.
Quality control charts should be used to track the
day-to-day results of sampling and analytical efforts. The
Shewhart Control Chart (Figure 4-3) is generally used for
achieving this goal. Such a chart should be kept for duplicate
analysis results, calibration constants, "spiked" sampling
results (percent recovery), isokinetic sampling rates, and other
factors which have a direct bearing on quality control. Nor-
mally, 3a control limits are used, with 2a being used as a
warning limit. Sigma, a, is the standard deviation of a set of
two or more duplicate sample results.
-41-
-------
UPPER
CONTROL LIMIT
LOWER
CONTROL LIMIT
CONSECUTIVE DAYS
Figure 4-3.
EXAMPLES OF A QUALITY CONTROL CHART FOR
SAMPLES (REF. 5).
-42-
-------
Sampling personnel should routinely check the confor-
mance of test team members to standard procedures for sampling,
calibration, analysis, and data handling. Any deviations should
be corrected immediately and an assessment of the effect of the
deviation on results should be made. If necessary, tests should
be repeated to ensure data validity.
Routine inventories of all reagents and supplies which
are needed for quality control maintenance should be maintained.
Purchasing guidelines should be established and documented so
that delays in delivery will not affect data quality.
All measurement system components should be routinely
serviced in accordance with manufacturer or developer recom-
mendations. Reviewing equipment maintenance records prior to
the time that the equipment is packed for shipment to the field
is strongly recommended. The reliability of the equipment
should routinely be checked to ensure that the maintenance pro-
cedures are adequate, and revisions should be made as needed to
maintain data quality. Maintenance and reliability data should
be recorded in log books.
Criteria for data validation should be stated in the
test plan and revised as needed to maintain quality control.
In the ideal case, data validation consists of two components:
a routine set of both manual and computerized
checks of measurement system performance at
various points in the sampling/analysis/data
processing sequence, and
random audits of quality control performance by
non-test team members. The quality control audit
should check the test team adherence to standard
procedures (a system review ) and make indepen-
dent quantitative checks as needed.
No quality control program can succeed unless all
equipment is routinely calibrated by the use of appropriate
standards. In addition, recovery tests for "spiked" samples
should be a regular part of laboratory procedures, in order to
check for chemical interferences (matrix effects). A quality
control chart should be retained for recording recovery results.
-43-
-------
Where possible, alternative sampling and analysis
methods should be used to provide added assurance of data
accuracy. Statistical correlation techniques can be used to
ensure that the two methods are producing identical results,
within experimental error.
The sampling strategy for an environmental test pro-
gram needs to be thoroughly developed before the test is init-
iated. The sampling strategy should specify the actual loca-
tions of sample points, the methods for sample acquisition, the
frequency and timing for sample collection, and the quality
control program for obtaining representative samples such that
with proper analysis techniques, the data will meet the objec-
tives of the test.
-44-
-------
SECTION 5.0
ANALYTICAL STRATEGY
An analytical strategy for an environmental test plan
should define sample handling and preservation techniques along
with specific analytical procedures required to accomplish the
objectives of the test. Factors to be considered in selecting
an analytical strategy include:
compatibility with sampling procedures
(amount of sample required and/or available),
expected concentration levels and required
detection limits,
presence of interfering species,
accuracy and precision requirements,
requirements established by the quality
control program, and
time, equipment and cost limitations.
The selection and validation of analytical methods
should be the responsibility of the analyst involved with the
program. A number of alternate methodologies may be applied
depending on available facilities and personnel. A selected
list of analytical parameters of probable interest for environ-
mental tests (waste stream, control equipment and process stream
characterization) of a coal gasification facility is given in
Table 5-1. Tables 5-2 and 5-3 list some specific organic
species of potential concern. Table 5-4 outlines the protocol
for the bioassay tests referred to in Table 5-1.
In the following sections, a discussion is presented
of handling and preservation techniques for gas, liquid and
solid samples followed by a discussion of the analytical methods
for organic and inorganic analyses of those samples. Because
-45-
-------
Table 5-1. ANALYTICAL PARAMETERS OF INTEREST FOR ENVIRONMENTAL
TESTING OF COAL GASIFICATION FACILITIES
Solida Liquids Gases
Morphology Physical Factors e Particulate Matter8
teachability Study3 Anlons Major Components
Ultimate Analysis Ammonia Minor Components
(wt Z of C, H2, N2, ,
S, 02) Oil & Grease Metal CarbonylsJ
Proximate Analysis Phenols Trace Elements
(wt Z of moisture, ash . c
volatile matter, fixed trace Elements Organic Species
carbon) .
. Organic Species0 Bloassay Tests
Trace Elements d
Bloassay Tests
Organic Species
Bloassay Tests
teachability studies: same analytical parameters as for liquids.
Trace Elements: S, Al, Ca, Fe, Mg, K, Si, Na, P, Ti, Sb, As, Be, B, Cd,
Cu, Cl, Cr, F, Pb, Li, Hg, Mn, Mo, Ni, Se, Ag, Tl, Sn,
U, V, Zn
Actual trace elements analyzed will depend upon preliminary
results from Sparks Source Mass Spectrometry Analysis.
Organic Species: POM, PNA, BAP, Benzene, Nitrosamines, Phenols, Carboxyllc
Acids, and GC/MS analysis of nonpolar compounds, moderately
polar compounds, polar neutral compounds, ethers of phenols,
methyl esters of carboxylic acids, basic compounds, polar
water soluble compounds, and very polar water soluble com-
pounds. Tables 5-2 and 5-3 list specific organic species
of potential concern.
The extent of organic characterization will depend upon
preliminary organic screening tests and the amount of organics
present in a sample.
4
Bioassay Tests: Table 5-4 gives the bioassay tests for solid, liquid, and gas
samples.
Physical Factors: pH, temperature, specific conductance, TSS, TDS, hardness,
acidity, alkalinity, COD, BOD, TOC
Anions: Chloride, Fluoride, Sulfate, Sulflte, Sulfide, Nitrate, Phosphate, Cyanide.
8 Particulate Matter: Same analytical parameters as for solids.
Major Components: CO, C02, N2, H2, 02.
Minor Components: HCN, H2S, COS, CS2, S02, Mercaptans, Thiophenes, NHj, NO
* Metal Carbonyls: Fe(CO)5, NKCOK.
-46-
-------
Table 5-2. ENVIRONMENTALLY HAZARDOUS ORGANIC SPECIES POTENTIALLY
PRESENT IN COAL GASIFICATION FACILITY PROCESS AND
WASTE STREAMS
Compound
Haxard
Compound
Hazard
POLYCYLCIC AROMATICS
Naphthalenes
Nephthalene
1-Methyl naphthalene
Acenaphthenea
Acenaphthene
Anthracenes
Anthracene
9-ttethylanthracene
Phenanthrenes
Suspected carcinogen
Toxic
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Dlbenzanthracenes
Dlbenz(a,H)anthracene
Dlbenz(a,c)anthracene
Dibenz(a,J)anthracene
Dlbenzopyrenes
Dlbenzo(a,i)pyrene
Dlbenzo(a,d)pyrene
Dlbenzo(a,e)pyrene
5-MethyIdibenzo(a,1)pyrene
Suspected carcinogen
Carcinogen
Suspected carcinogen
Suipected carcinogen
Carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Penanthrene
Pyrenes
Pyrene
Fluoranthenes
3-Hethylfluoranthene
Fluoranthene
Indenopyrenes
Indeno (l,2,3-c,d)pyrene
Cholanthrenes
Cholanthrene
3-Methylcholanthrene
Chrysenes
Chrysene
5 , 6-Demethylchrysene
5-Methylchrysene
Benzanthracenes
Benz (a) anthracene
6 , 8-Dine thylbenz (a) anthracene
7 , 12-Dimethybenz (a) anthracene
8 , 12-Dime thy Ibenz (a) anthracene
1-Hethylbenz (a) anthracene
2-Methylbenz (a) anthracene
Benzophenanchrenes
Benzo(c)phenanthrene
2-Me thylbenzo(c) phenanthrene
Benzo fluoranthenes
Benzo ( j ) f luoranthene
Benzo(b)fluoranthene
Benzo (k) f luoranthene
Benzopyrenes
Benzo (a) pyrene
Benzu(e) pyrene
3-Me thylbenzo (a) pyrene
Toxic
Carcinogen
Suspected carcinogen
Toxic
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Carcinogen
Suspected carcinogen
Carcinogen
Carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Carcinogen
Suspected carcinogen
Carcinogen
Carcinogen
Suspected carcinogen
Carcinogen
Carcinogen
Suspected carcinogen
Benzoperylenes
Benzo (ghi)perylene
Dlbenzopentaphenea
Dibenzo (h,rst)pentaphene
Dibenzoperylenes
Peropyrene
Carcinogen
Suspected carcinogen
Suspected carcinogen
NITROGEN HETEROCYCLICS
Pyrroles
Pyrrole
Horphollnes
Morpholine
N-Ethylmorpholine
B ismo rpho 1 inome thane
Pyridlnes
Pyridine
a-Picoline
3,4-Dihydroxypyridine
n ° e
Indole
Carbazoles
Carbazole
Benzocarbazoles
Benzo (a) carbazole
Dlbenzocarbazoles
7H-Dibenzo (c , g) carbazole
7H-Dibenzo (a , 1) carbazole
7H- Dibenzo (a ,g) carbazole
Ouinollnes
Qulnoline
Isoquinoline
Toxic
Irritant
Irritant
Carcinogen
Irritant
Irritant
Carcinogen
Suspected carcinogen
Toxic
Suspected carcinogen
Carcinogen
Suspected carcinogen
Suspected carcinogen
Toxic
Toxic
f1 r\** i--i i-tii £*i3
-47-
-------
Table 5-2.
(Continued) ENVIRONMENTALLY HAZARDOUS ORGANIC SPECIES
POTENTIALLY PRESENT IN COAL GASIFICATION FACILITY PRO-
CESS AND WASTE STREAMS
Page 2
Compound
Hazard Compound Hazard
Benzacrlndines .. .
Benz(c)acrldine
7,9-Dimethylbenz(c)acridine
7-Methylbenz(c)acridine
Dibenzacridinea
Dibenz(a,J)acrldlne
Dlben(a,h)acrldine
Phenazinea
PhenazlDe
SuBpectad carcinogen
Suspected carcinogen
Carcinogen
Carcinogen
Carcinogen
Suapected carcinogen
NON-HETEROCTCLIC NITROGEN COMPOUNDS
Nltrllea
Acetonitrile
Acrylonltrlle
Aliphatic Amines
Flperidine
Methylamlne
Ethylamine
n-Propylamine
n_Butylamlne
Triethylamlne
Ethylenediamlne
Cyclohexylamine
Dicyclohexylamine
Dime thylamine
Allylamine
Amides
Acrylamide
Acetamide
N,N-Diethylacetamide
Acetylaminofluoranthene
Thioacetamide
Acetanilide
Aromatic Amines
o-Toluidine
m-Toluidine
B-Naphthylamine
Diphenylamine
fienzidine
4-Amino-biphenyl
Aniline
4,4-Methylenedianillne
Aminoazobenzene
Benzylamine
p-Phenylenediamine
Imlnes
Ethylenimine
Hydroxylamines
N-Hydroxyanillne
N-2-Naphthylhydroxylamine
Hydroxy lamine
Toxic
Toxic
Toxic
Irritant
Irritant
Irritant
Irritant
Irritant
Irritant
Irritant
Irritant
Mutagenic
Irritant
Toxic
Carcinogen
Suspected carcinogen
Suspected carcinogen
Mutagenic
Mutagenic
Suspected carcinogen
Suspected carcinogen
Carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Toxic
Suspected carcinogen
Suspected carcinogen
Irritant
Irritant
Toxic
Mutagenic
Suspected carcinogen
Mutagenic
Hydrazlne
1,3-Dlethylhydrazine
• Hethylhydrazine
Hydroazobenzene
Sealcarbazldea
Semlcarbazide
Azo Compounds
Azobcntene
Aioethene
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Carcinogen
Suspected carcinogen
Suspected carcinogen
SULFUR COMPOUNDS
Thloureas
Thiourea
Mercaptans, Aliphatic
Methylmercaptan
Ethylmercaptan
Isopropylmercaptan
n-PropyImercaptan
2-Pentylraercaptan
Isoamylmercaptan
n-Araylercaptan
n-Hexylmereaptan
3-Mercaptoethanol
Mercaptans. Aromatic
Thiophenol
BenzyImercaptan
p-Thiocresol
Thioraphthol
Sulfonic Acids
Benzenesulfonlc acid
Methanesulfonic acid
Sulfuric Acid Esters
Dimethylsulfate
n-Propylmethanesulfonate
Sulfoxldes
Dimethylsulfoxide (DMSO)
Sulfldes
Carbon disulflde
Thlophenes
Thiophene
Suspected carcinogen
Disagreeable
Disagreeable
Disagreeable
Disagreeable
Disagreeable
Disagreeable
Disagreeable
Disagreeable
Mutagenic
odor
odor
odor
odor
odor
odor
odor
odor
Disagreeable odor
Disagreeable odor
Disagreeable odor
Disagreeable odor
Irritant
Irritant
Suspected carcinogen
Mutagenic
Suspected carcinogen
BENZENE DERIVATIVES
Polyaryla
Blphenyl
Toxic
Continued
-48-
-------
Table 5-2. (Continued) ENVIRONMENTALLY HAZARDOUS ORGANIC SPECIES
POTENTIALLY PRESENT IN COAL GASIFICATION FACILITY
PROCESS AND WASTE STREAMS
Page 3
Compound
Double Bond Conjugated Benzene*
Styrene
Alkylbenzenea*
Benzene
PHENOLS
Phenol
b-Chlorophenol
2,4-Xylenol
2,5-Xylenol
2,6-Xylenol
3.4-Xylenol
3,5-Xylenol
o-Cresol
B-Cresol
p-Creaol
1-Naphthol
2-Naphthol
Pyrogallol
Hydroqulnone
CARBOXYLIC ACIDS
Aliphatic Acids
Formic Acid
Acetic Acid
Propionic Acid
Butyric Acid
Valeric Acid
Caproic Acid
Acrylic Acids
Acrylic Acid
Methacrylic Acid
Carbamates
Methylcarbamate
Ethylcarbamate
n-Butylcarbamate
n-Propylcarbamate
Lactones
B-Propiolactone
v-Butyrolactone
Amino Benzoic Acida
Hazard
Toxic
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Toxic
Toxic
Toxic
Toxic
Toxic
Irritant
Irritant
Caustic
Caustic
Disagreeable odor
Disagreeable odor
Disagreeable odor
Disagreeable odor
Irritant
Irritant
Mutagenlc
Mutagenic
Mutagenic
Mutagenlc
Mutagenic
Mutagenlc
Compound
Aliphatic Ketonea
Acetone
Methylethylkecone
Cyclohexanone
a. 3 Unsaturated Carbonyls
Acrolein
Crotonaldehyde
Aromatic Carbonyls
Acetophenone
Furfural
Qulnonea
Benzoquinone
Anthraquinone
2-Me thylbenzoquinone
OTHER SPECIES
Oxygen Heterocyclics
p-Dioxane
Coumarin
Alcohols
Methanol
Ethanol
Allyl alcohol
Ethylene Glycol
Cyclohexanol Carcinogenic
Gaseous Speciea
Hydrogen Cyanide
Carbonyl Sulfide
Carbon Monoxide
1,3 Butadiene
Hazard
Irritant
Irritant
Care inogen
Irritant
Irritant
Irritant
Irritant
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Suspected carcinogen
Toxic
Suspected carcinogen
Irritant
Suspected carcinogen
Toxic
Toxic
Toxic
Irritant
Anthranilic Acid
Aliphatic Aldehydea
Formaldehyde
Acetaldehyde
Propionaldehyde
Butyraldehyde
Hyceraldehyde
Suspected carcinogen
Irritant
Suspected carcinogen
Irritant
Irritant
Mutagenlc
*0ther major alkylbenzenes present as major constituents of their class are
also significant.
-49-
-------
Table 5-3.
ORGANIC SPECIES FROM EPA EFFLUENT GUIDELINES
(Ref. 6)
1,3-Dichlorobenzene
1,4-Dlchlorobenzene
Hexachloroethane
1,2-Dichlorobenzene
bis(2-Chloroisopropyl)ether
Hexachlorobutadiene
1,2,4-Trichlorobenzene
Naphthalene
Nitrobenzene
bis(2-Chloroethyl)ether
bis(2-Chloroethyoxy)methane
Hexachlorocyclopentadiene
2-Chloronaphthalene
Acenaphthylene
Acenaphthene
Isophorone
Fluorene
2,6-Dinitrotoluene
2,4-Dinitrotoluene
4-Chlorophenyl phenyl ether
Diphenylhydrazine
Dimethyl phthalate
N-nitrosodiphenylamine
Hexachlorobenzene
4-Bromophenyl phenyl ether
Diethylphthalate
Phenanthrene/Anthracene
dj o-Anthracene
Di-n-butylphthalate
Fluoranthene
Pyrene
Benzidine
Butyl benzyl phthalate
Chrysene/Benzo(a)anthracene
bis(2-Ethylhexyl)phthalate
Benzo(b)fluoranthene/
Benzo(a)pyrene
Indeno(l,2,3-c,d)pyrene
Dibenzo(a,h)pyrene
Benzo(g,h,i)perylene
N-nitrosodimethylamine
N-nitrosodi-n-propylamine
Endrin aldehyde
Endosulfan sulfate
3,3'-Dichlorobenzidine
2,3,7,8-TCDD
Chloromethane
Dichlorodifluoromethane
Bromomethane
Vinyl chloride
Chloroethane
Methylene chloride
Trichlorofluoromethane
1,1-Dichloroethylene
Bromochloromethane (IS)
1,1-Dichloroethane
Trans-l,2-dichloroethylene
Chloroform
1,2-Dichloroethane
1,1,1-Trichloroethane
Carbon tetrachloride
Bromodichloromethane
bis(chloromethyl)ether
1,2-Dichloropropane
Trans-1,3-dichloropropene
Trichloroethylene
Dibromochloromethane
1,1,2-Trichloroethane
cis-1,3-Dichloropropene
Benzene
2-Chloroethyl vinyl ether
2-Bromo-l-chloropropane (IS)
Bromoform
1,1,2,2-Tetrachloroethane
1,4-Dichlorobutane (IS)
1,1,2,2-Tetrachloroethene
Toluene
Chlorobenzene
Ethylbenzene
Phenol
2-Chloropheno1
2-4-Dimethylpheno1
2-Nitrophenol
2,4-Dichlorophenol
p-Chloro-m-cresol
2,4,4-Trichlorophenol
2,4-Dinitrophenol
4-Nitrophenol
4,6-Dinitro-o-cresol
Pentachlorophenol
dt0-Anthracene (IS)
-50-
-------
Table 5-4. BIOASSAY PROTOCOL FOR SOLID, LIQUID AND GAS SAMPLES
(Ref. 1)
Sample
Health Effects
Ecology Effects
Water/Liquids
Microbial Rodent Acute
Mutagenesis Toxicity
Algal Static
Bioassay Bioassay
Soil
Microcosm
Solids
Microbial Rodent Acute
Mutagenesis Toxicity
Algal Static
Bioassay Bioassay
Soil
Microcosm
Gases
Plant
Stress
Ethylene
Particulates Microbial Rodent Acute
Mutagenesis Toxicity
(Cytotoxicity)
Soil
Microcosm
Sorbent (XAD-2)
Microbial
Mutagenesis Cytotoxicity
-51-
-------
of the large number of analytical methods discussed in this
section, they are presented according to the parameter to be
analyzed. A brief summary of the specific methods and a list
of references where detailed discussions of the methods can be
found are presented.
5.1 SAMPLE HANDLING AND PRESERVATION
After samples have been obtained from various streams
in the process, they should be prepared for analysis and/or
storage. The bulk samples from each sampling point should be
taken to some central location equipped to perform the preserva-
tion and treatment procedures, probably a stationary or mobile
laboratory facility. The handling and preservation requirements
for the analytical parameters of interest in each stream should
be integrated into a total sampling/preservation/analytical
scheme for greatest efficiency. The requirements of such a
scheme will determine the number of individual aliquots into which
the bulk sample is divided, Some level of sample division may
have already occurred as part of the sampling techniques. For
many of the parameters of interest, preservation will allow the
sample to be transported to off-site facilities for analysis.
Others will require immediate, on-site analysis due to their
instability.
5.2 ANALYTICAL TECHNIQUES
In this section the analytical techniques for analyz-
ing samples collected from coal gasification facilities are
summarized. A detailed description of each technique is not
included here. Instead, references are given so that the reader
may readily obtain detailed descriptions of each technique.
Again, it should be emphasized that the selection of analytical
techniques depends upon the following factors:
compatibility with the sample procedures,
expected concentrations and detection limits,
presence of interfering species,
accuracy and precision requirements,
-52-
-------
requirements established by the quality
control program, and
time, equipment and cost limitations.
In the following sections analytical methods are
divided into screening (qualitative) and quantitative tech-
niques. In each of these divisions, analytical methods for
inorganic and organic analyses are presented separately. Bio-
logical tests are discussed only in the section dealing with
screening techniques.
5.2.1 Analytical Screening Techniques
In this section, analytical screening techniques for
environmental tests are discussed. These techniques, which are
used for Level 1 waste stream characterization tests, are
specified in the lERL-RTP Procedures jfanual: Level 1 Environ-
ment Assessment (Ref. 1).
Inorganic Analysis -
The Level 1 (screening) procedures for inorganic
analyses in solid, liquid, and gas streams are given in Table
5-5. For most trace element analyses, spark source mass
spectrometry (SSMS) is used. Exceptions to this are mercury,
antimony, and arsenic. The sample type, preparation and analysis
procedures for SSMS and for mercury, antimony, and arsenic
analyses are outlined in Figure 5-1 and 5-2, respectively.
Reagent test kits are used for the other inorganic
components shown in Table 5-5. These analyses are primarily for
liquid samples (including leachates). The analytical techniques
used for each of these components are given in Section 5.2.2.
Organic Analysis -
Level 1 screening procedures for analyzing organics in
gaseous, liquid, and solid waste streams are shown in Figure 5-3,
Gases and organic liquids (including extracts) are analyzed by
gas chromatography for Ci-C5 hydrocarbons. Liquids and extracts
are further examined by infrared (IR) spectrometry for the
-53-
-------
Table 5-5. ANALYTICAL PARAMETERS FOR SCREENING INORGANIC
CONSTITUENTS IN SOLID, LIQUID, AND GAS SAMPLES
Sample Type
Solid
Gas
Analytical Parameters
Trace Elements
Trace Elements
Analytical Technique
SSMS
SSMS
Liquid (including
leachates)
Particulate Matter
Trace Elements ,
Water Quality Parameters
Trace Elements
Morphology
SSMS
(See Table 5.8 in
Section 5.2.2)
SSMS
Ref. 1
Detailed discussion of the analytical techniques for these
analytical parameters are given in the Level 1 environmental
assessment manual (Ref. 1).
Water quality parameters:
acidity, alkalinity, pH, COD, BOD,
TOG, conductivity, TDS, TSS, dissolved
oxygen, ammonia, cyanide, sulfate, sulfite,
sulfide, nitrate, nitrite, carbonate, and
thiocyanate.
-54-
-------
SAMPLE FOR
ELEMENTAL
ANALYSIS
SAMPLE
WATER AND
NON-ORGANIC
SOLUTIONS
XAD-2
SORBENT
ORGANICS:
LIQUID OR
SOLID
PARTICULATE
MATTER. ASH OR
NON-ORGANIC
SOLIDS
HOMOGENIZE
AND DIVIDE
PARR BOMB
COMBUSTION
OVER HNOs
EXTRACT
FOR
ORGANICS
PARR BOMB
COMBUSTION
OVER HN03
PARR BOMB
COMBUSTION
OVER HNOs
PREPARATION
SSMS
SSMS
SSMS
SSMS
ANALYSIS
Figure 5-1.
OUTLINE OF LEVEL L ELEMENTAL ANALYSES USING SPARK SOURCE MASS
SPECTROMETRY (SSMS)
-------
SAMPLE TYPE
SAMPLE PREPARATION
PARTICULATE MATTER
ASH, AND
NON-ORGANIC SOLIDS
Ul
C^
I
SAMPLE ANALYSIS
Figure 5-2. LEVEL I ANALYSES FOR MERCURY, ANTIMONY AND ARSENIC
-------
,
ALIQUOT FOR
GRAVIMETRIC
S TCO
r
INFRARED
ANALYSIS
i
F
LOW RESOLUTION
MASS SPECTRA
ANALYSIS
*Recoimn«ided procedure, but not officially part of Level 1 analysis.
Figure 5-3. OUTLINE OF LEVEL 1 ORGANIC ANALYSES
-57-
-------
presence of organic functional groups. The liquids are then
subjected to liquid chromatography to fractionate the sample
according to compound class. Each of these fractions are sub-
sequently analyzed by IR spectrometry and low resolution mass
spectrometry (LRMS).
Bioassay Analysis -
Bioassay tests are used in order to provide informa-
tion concerning the biological effects (synergism vs. antagonism)
of waste streams containing complex mixtures of inorganic and
organic compounds. In most cases those effects cannot be pro-
jected from the results of physical and chemical tests alone.
Bioassay tests may be performed on the whole sample or fractions
of the sample depending upon the waste stream characteristics.
For example, if a gaseous waste stream contains particulate mat-
ter, bioassay tests may be needed for both the gas and particu-
lates.
At present, specific bioassay analytical procedures
are not completely defined. However, types of bioassay tests
are being studied to assess their applicability for evaluating
the health and ecological effects of waste streams from coal gas-
ification facilities. These tests, their expected results and
the amount of sample required are given in Table 5-6. A brief
discussion of these bioassay tests is presented below:
• Health Effects
Ames Test: is used to measure the potential
mutagenicity (carcinogenicity) of the sample.
Three histidine deficient Sallmonella
Typhimurim strains (TA-1535, TA-1537 and TA-
1538) are used. The reversion of the strains
to prototrophy indicates mutation (Ref. 7, 8,
and 9).
- Cytotoxicity Test: is used to estimate the
acute cellular toxicity of a sample from an
in-vitro cell mortality test using rabbit
lung aveolar macrophages. The results of
this test are extrapolated to obtain an
ECso value for the sample.
-58-
-------
Table 5-6.
BIOASSAY PROTOCOL FOR LEVEL
CHARACTERIZATION TESTS
1 WASTE STREAM
Bioassay Tests (Classification of
Sample to be Tested: Solid, Liquid,
Gas)
Results of
the Test
Amount of
Sample Tested
Health Tests (Solid, Liquid)
Ames
Cytotoxicity
WI-38
Rodent Acute Toxicity
+1-
+1-, EC so
+1-, EC so
+1-, LDso
1 8
1/2 g
1/2 g
50 g
Ecological Tests (Solid, Liquid)
Freshwater
• Algal
Daphnia
• Fish
Saltwater
• Algal
• Shrimp
• Fish
Terrestrial
Soil microcosm
• Plant stress ethylene
(gas sample only)
+1-, EC so
+1-, LCso
+1-, LCso
+1-, EC so
+1-, LCso
+1-, LCso
+1-, Ranking
+1-
50 £
50 £
50 I
50 H
50 £
50 £
10 g
1360 £
The exact procedures for each of these bioassay tests are currently
being developed by the EPA.
+1-: toxic/nontoxic; EC so and LCso: concentration required to kill
50% of the biological species tested; LDso: dose required to kill
50% of the species tested.
-59-
-------
- WI-38 Test: is used to estimate the acute
cellular toxicity of a sample from an
in-vitro WI-38 cell mortality test. As in
the cytotoxicity test, the results are
extrapolated to obtain an ECso for the sample.
- Rodent Acute Toxicity Test: is used to
measure the acute toxicity of a sample in a
whole animal. The test is performed by
administering known levels of the sample to
a small population of mice and extrapolating
the mortality rate to obtain an LC50 value.
Other effects on the mice, such as loss of
hair, bleeding at the nose, etc., should also
be recorded during the test.
Ecological Effects
- Freshwater Tests: are used to estimate the
acute toxicity of a sample by exposing selected
aquatic species (algae, daphnia and fathead
minnows) to several levels of sample concentra-
tions. An ECso, for algae, and an LCso, for
daphnia and minnows, are obtained by
extrapolation.
- Saltwater Tests: are used to estimate the acute
toxicity of a sample by exposing selected species
(algae, mysid shrimp and sheepshead minnows) to
various concentrations of sample. An ECs0 and
LCso's are estimated from the algal test and the
shrimp and minnow tests, respectively.
Plant Stress Ethylene Test: is used to estimate
the effects of gaseous waste streams on five-
week-old soybean plants. Variations in the
amounts of ethylene produced by the plants are
measured for various exposure times and gas
volumes. The results of this test indicate
whether the sample was toxic (increase in
ethylene production) or nontoxic (no change in
ethylene production).
Soil Microcosm Test: is used to measure the
variations in oxygen consumption and carbon
dioxide production of soil samples containing
-60-
-------
microorganisms. The results from this test
give a toxic/nontoxic result along with a
severity ranking of each sample tested.
In summary, screening analytical techniques are effec-
tive in determining whether a stream or portions of the stream
need further chemical or biological characterization. Pro-
cedures for quantitative chemical analysis of environmentally
significant streams are presented in the following section.
Quantitative biological tests have not been defined by the EPA.
5.2.2 Quantitative Analytical Procedures
Quantitative analytical techniques for characterizing
chemical species in waste and process streams are discussed in
this section. These techniques are used for Level 2 and 3 waste
stream characterization tests and for control equipment and pro-
cess stream characterization tests. In the following text, a
summary of quantitative techniques for analyzing inorganic and
organic species in gas, liquid, and solid samples is presented.
A detailed description of each of these procedures is not
included; however, references are given to indicate where a
detailed discussion can be found.
Inorganic Analysis -
Inorganic species of interest in environmental tests
for coal gasification facilities are given in Table 5-1.
Inorganic analysis generally consists of two steps: 1) sample
preparation, and 2) species identification and/or quantification.
Preparation techniques for gas, liquid, and solid samples are
discussed below. Analytical methods for identifying and quanti-
fying inorganics in those samples are then presented.
Sample Preparation - In most cases gaseous, liquid
and solid samples require preparation prior to analysis for in-
organic species. These preparation techniques are summarized in
the following text. A more detailed discussion is contained in
a report by Richard Luthy, "Manual of Methods: Preservation and
Analysis of Coal Gasification Wastewaters" (Ref. 10).
-61-
-------
The gas samples collected in impinger solutions require
preparation procedures similar to those for liquid samples being
analyzed for analogous species (e.g., NH3 and HCN are analogous
to NH* and CN~). Sulfur species in impinger solutions (COS, H2S,
CS2 and S02) are oxidized with hydrogen peroxide for analysis as
sulfate. Condensable organics collected in impingers may require
extraction or dissolution of an immiscible organic phase prior
to analysis.
Vapor phase trace elements collected by wet electro-
static precipitator (WEP) sampling (described in the Appendix)
require that particulates suspended in the electrolyte solution
be filtered, digested, and the digested solution added back to
the electrolyte solution prior to analysis.
Gas grab samples may require pretreatment (e.g., dry-
ing and filtering) prior to analysis by gas chromatography or
other instrumental techniques. Great caution must be exercised
to insure that the gas species of interest are not lost or
changed in pretreatment. Another area of precaution involves
the probability of certain gaseous species sorbing on or reacting
with the walls of the gas collection container (glass bomb or
bag). Pretreatment of the collection container with a gas con-
taining the gas species of interest may be required. However,
the possibility of desorption of the pretreatment species from
the container walls into the sample gas must also be avoided.
Some liquid samples may require distillation or acid
digestion prior to analysis to exclude interfering species or
to stabilize the compound of interest. Examples of inorganic
species that require distillation pretreatment include cyanide,
fluoride and ammonia. Analyses for trace elements usually
require acid digestion to oxidize the elements present to a
stable valence state (Ref. 11).
Organic liquids, such as tars and oils, may require
dilution with a solvent to facilitate sample handling. Trace
element analysis requires a high temperature and pressure acid
digestion technique (Ref. 12) using a sealed metal bomb with
a Teflon insert.
The first step in preparing solid samples is to insure
homogeneity. This can be accomplished by mixing and grinding the
sample to less than 200 mesh. Drying the sample is often
-62-
-------
required with concurrent gravimetric determination of moisture
content. The drying temperature must be selected to avoid loss
of volatile compounds. For samples with a high moisture content
drying may be necessary to allow mixing, grinding and sizing.
Identification and Quantification Procedures - Once the
samples are in a form compatible with the analytical technique,
the species of interest can be identified and/or quantified. The
extent of this determination can range from the identification of
classes or groups of species to the accurate quanitification of
specific compounds .
Analytical procedures which have been proven effective
in analyzing for the inorganic species listed in Table 5-1 are
summarized in Tables 5-7 through 5-9.
These tables list the species to be analyzed, the
amount of sample required, the sample container, how the sample
is collected, the maximum holding time allowable before analy-
sis, the sample preparation required and a general description
of the analytical technique along with references. General
reference sources for inorganic analysis include:
Standard Methods, 14th edition (Ref. 13),
1977 Annual Book of ASTM Standards - Part
31 - Water (Ref. 15^
Manual of Methods for Chemical Analysis of
Water and Wastes, 1974, EPA (Ref.
Assessment, Selection, and Development of
Procedures for Determining the Environmental
Acceptability of Synthetic Fuels Plants Based
On Coal, (Ref. 15),
Detection and Determination of Trace Elements,
6th Edition, Maurice Pinta (Ref. 16), and
Quantitative Analysis of Gaseous Pollutants,
Ruch, (Ref. 17).
An example of an integrated scheme for inorganic
analysis of liquid samples from a coal gasification facility is
-63-
-------
Table 5-7. INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION PROCESSES (GASEOUS)
Sample Quantity
Gases Requirements Container
Grab Samples:
CO, CO;, Kz, 1 liter each Scotchpack
Hz, Oa
H2S, CS2, 1 liter each Glass Bomb
COS, S02
NO 1 liter Tedlar bag
X
Ammonia 250 mis P, G
Cyanide 250 mis P, G
Sulfur Species:
H2S 250 mis P, G
S02 250 mis P, G
Total Sulfur 250 mis P, G
(H2S, COS,
CS2, S02)
Collection
System
Grab Sample
Grab Sample
Grab Sample
Impinger
Collection
in 10X H; SO.,
Inpinger
Collection
in 101 NaOH
Impinger
Collection in
basic Zn
acetate
Hz02 implnger
Impinger
Collection In:
1st: Iodine
2nd: QaCl/NH,OH
3rd: EtOH, KOH
Ref. AS
Holding
Time Preparation Analytical Method, Reference
(on-site Drying Gas Chroma tography,
analysis) Ref. 1, 18, 19, 20
(on-slte Drying Gas Chroma tography,
analysis) Ref. 1, 18, 19, 20
(on-site Drying Chemiluminescence, commercial
analysis) instrumentation, Ref. 21
Cool, 4°C, Distillation Titrimetric, Ref. 11, 13, 14
24 hours
Cool, 4°C, Distillation Spectrophotometric, Ref. 11, 13, 14
24 hours
Cool, 4*C, Precipitation, Titrimetric, Ref. 9, 13
24 hours filtration,
dissolution
Cool, 4°C, None Turbidimetric, S0» . Ref. 11,
7 days 13, 14
Cool, 4°C, H202 oxidation Turbidimetric, SO,, Ref. 11,
24 hours 13, 14
Continued
-------
Table 5-7. (Continued) INORGANIC ANALYSIS METHODS FOR COAL GASIFICATION
PROCESSES (GASEOUS)
Sample Quantity
Elemental Analysis:
Impingers: 250 mis P
F P
Cl P
Collection Holding
System Time
HN03 l> NaOH 6 months
impingers
NaOH impingers 7 days
Ref. 49
NaOH Implnger 7 days
Preparation Analytical Method, Reference
None See Table 5-8 for
liquid analyses
None Specific ion electrode,
Kef. 22
None Specific ion electrode,
standard additions,
Ref. 23
Metal Carbonyls Depends on con-
(Fe, Ni) centration
I2/HC1 Implnger 6 months
Atomic Absorption,
Kef. 24, AA Graphite Furnace
0.01 g
Participates
collected on filters
Acid digestion
See Table 5-9 for solids
analyses
P - polyethylene
T - teflon
G - glass
-------
Table 5-8. INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION PROCESSES (LIQUIDS)
Sample
Liquids Requirement
Acidity, total 100 mis
Alkalinity, total 100 mis
Biochemical 1000 mis
oxygen demand
(BOD)
Chemical oxygen 50 mis
demand (COD)
Chloride 100 mis
Cyanide 500 mis
Fluoride 500 mis
Hardness, total 100 mis
Nitrogen, 500 mis
ammonia
Nitrogen, 100 mis
nitrate
Oil & Grease 1000 mis
Organic Carbon, 100 mis
total
pH 25 mis
Container1 Preservation
P, G Cool, 4°C
P, G Cool, 4'C
P, G Cool, 4°C
P, G HzSOi,, pH <2
P, G
P, G NaOH, pH >12
PbNOs to remove
sulfide, Cool,
4°C
P Cool, 4°C
P, G Cool, 4°C
P, G H2SOi,, pH <2,
Cool, 4°C
P, G HjSOi., pH <2,
Cool, 4*C
G, T H2SOi,, pH <2,
Cool, 4°C
G, T H2SOn, pH <2,
(brovn) Cool, 4'C
P, G Determine on-
site
Holding
Time Preparation Analytical Method, Reference Remarks
24 hrs Ticrimetric, Ref. 11, 13, 14
24 hrs Titrimetric, Ref. 11, 13, 14
6 hrs Bloassay, Ref. 11, 13
7 days Dichromate Titrimetric, Ref. 11, 13, 14
reflux
removal, tlon using specific moval necessary
Ref. 13 ion electrode, Ref. 13 due to organics
6 sulfur usually
found in samples
24 hrs Distillation, Spectrophotometric,
Ref. 13 Ref. 11, 13, 14
7 days Distillation, Specific ion electrode,
Ref. 13 Ref. 11, 13, 14
7 days Titrimetric, Ref. 11, 13, 14
24 hrs Distillation, Titrimetric, Ref. 11, 13, 14
Ref. 11, 13, 14
24 hrs Spectrophotometric, Ref. 13
24 hrs Soxhlet extraction
gravimetric,
Ref. 11, 13
24 hrs Carbonaceous analyzer,
Ref. 11
6 hrs pH meter, Ref. 11, 13, 14
-------
Table 5-8. (Continued) INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION
p 2 PROCESSES (LIQUIDS)
Liquids Requirement
Phenol 500 mis
Phosphate, 100 mis
total
filterable ,
& volatile
Specific 100 mis
conductance
Sulfate 50 mis
Sulfide 100 mis
Sulfite 100 mis
Temperature 1000 mis
Container l Preservat ion
G, T H3SO%. pH <4
Ig CuSO,.'5H20/l
Cool, 4"C
G (acid Cool, 4"C
washed with
1:1 HS03)
G P Coo 1 4 ° C
t, G Cool , 4°C
P. G Cool, 4°C
P, C Zn .icefate
P, G Cool, 4"C
P, G Determine
on-site
Holding
Tine Preparation Analytical Method, Reference Remarks
24 hrs Distillation, Spectrophotometric ,
Ref. 11, 13, 14 Ref. 11. 13, 14
24 hrs Acid digestion, Spectrophotometric, A«cotbic acid
Ref. 11, 13, 14 Ref. 11, 13, 14 method
7 davs Gravimetric Ref- 11 13 14
24 hrs Conductivity meter,
Ref. 11. 13. 14
7 days Tnrbtdimetric,
Ref. 11, 13, 14
24 hrs Tltrimetric,
Ref. 11, 13
24 hrs Titrimetric,
Ref. 11, 13, 14
No holding Ref. 13
Elemental Analysis 1000 mis
(Elemental Survey) 200 mis
(SSMS)
Antimony
HNOj to pH <2
Spark source mass
spectroraetry, Ref. 25
Atomic absorption, Ref. 24, 25
Organic extrrac- AA - graphite furnace,
tion, Ref. 38, 50 Ref. II, 26, 27
Berylliu
Organic extrac-
tion, Ref. 39
AsHi generation,
Ag-DEDC pyridlne,
Ref. 11, 14
Atomic absorption, graphite
furnace, Ref. il, 27
Continued
-------
I
cr>
00
Table 5-8. (Continued) INORGANIC ANALYTICAL METHODS
p 3 PROCESSES (LIQUIDS)
Holding
Liquids Requirement Container' Preservation Time Preparation
Boron Ion exchange
Separation
Ref. 40
Cadmium Ref. 56
Calcium
Chroaium Ref . 6
Copper Ref. 6
Iron
tion, Ref. 29
Lithium
Magnesium
Manganese
Mercury
Molybdenum Solvent extrac-
tion, Ref. 46
Nickel
Potassium
Selenium
Silver Ref. 6
Silica
Sodium
Thallium
FOR COAL GASIFICATION
Analytical Method, Reference
Spectrophotometric, Ref. 28
AA - graphite furnace, Ref. 11,
27 Solvent Extraction - AA,
Ref. 29
Atomic absorption, Ref. 11, 14,
Atomic absorption, Ref. 11, 14
AA - graphite furnace, Ref. 27
Atomic absorption, Ref. 11, 14
AA - graphite furnace, Ref. 11,
Atomic absorption, Ref. 11, 14,
Atomic absorption, Ref. 24
AA - graphite furnace, Ref. 11,
27
Atomic absorption, Ref. 14
Atomic absorption, Ref. 11,
14, 24
Atomic absorption, Ref. 14
AA - graphite furnace, Ref. 27
Cold vapor, flameless atomic
absorption, Ref. 11, 14, 30
AA - graphite furnace, Ref. 27
Atomic absorption, Ref. 14
AA - graphite furnace, Ref. 27
Atomic absorption, Ref. 11, 24
Fluorometry, Ref. 31, 32, 33
AA - graphite furnace standard
additions, Ref. 11, 27
Atonic absorption, Ref. 24
Atomic absorption, Ref. 11, 24
Atonic absorption, Ref. 11, 24
AA - graphite furnace, Ref. 27
Remarks
24
0.1-5 mg/1
0.1-5 mg/1
27
24
<0.3 ppm
1-5 mg/1
0.04-5 mg/1
<0.1 mg/1
0.2-10 rag/1
<0.2 mg/1
0.2-5 mg/1
•0.2 mg/1
Continued
-------
VO
Table 5-8. (Continued) INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION
p e 4 PROCESSES (LIQUIDS)
Holding
Liquids Requirement Container' Preservation Time Preparation Analytical Method, Reference Remarks
Tin Atonic absorption, Ref. 11, 24 4-350 mg/1
AA - graphite furnace, Ref. 27 <4 mg/1
Titanium Spectrophotometric, Ref. 34
Uranium Fluorometry, Ref. 11. 35
Vanadium AA - graphite furnace standard <1 mg/1
additions, Ref. 27, 36
Zinc Atomic absorption, Ref. 11, 24 0.1-1 mg/1
AA - graphite furnace, Ref. 27 cO.l mg/1
*P • Polyethylene G - Glass T - Teflon
-------
Table 5-9. INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION PROCESSES (SOLIDS)
Solids
Sample
Requirements Container
Preservation
Holding
Time
inorganic Analysis:
Morphology
1 teg
Ultimate Analysis; 10 g
(Ct H, 0, N, S,
Ash)
Proximate Analysts: 10 g
(H20, ash, volatile
matter, fixed
carbon, Btu content)
Leachability Study 500 g
G,T
Cool, <4°C
Preparation
Ref. 37
Extract 24 hours
with HjO and/or
dilute HC1
Analytical Method
Dry, crush, sieve,
blend
Photometric photography,
scanning electron micro-
scope
Ref. 37
Ref. 37
Elemental, anion,
organic analysis
o
I
Elemental Analysis:
Elemental Survey 5 g
(SSMS)
Elements: 50 g
(Al, Ca, Fe, Mg,
K, Si, Na)
L1B02 fusion,
Ref. 51
Perchloric acid
digestion
Spark source mass
spectrometry, Ref. 25
Atomic absorption,
Ref. 24
Colorimetry, Ref. 13
Ascorbic acid
method
Perchloric acid
digestion
Colorimetry, Ref. 34
Perchloric acid
digestion
Extraction, atomic
absorption, Ref. 38
Perchloric acid
digestion
AsHa generation,
AgDEDC-pyridine
collection, Ref. 14
Perchloric acid
digestion
Organic extraction,
atomic absorption,
Ref. 39
NazCO] fusion,
H2SOi, dissolution
Ref. Ill
Ion exchange separa-
tion , Ref. 40; color-
imetry, Ref. 28
Cd
Perchloric acid
digestion
AA graphite furnace,
Ref. 27; solvent extraction
AA. Ref. 29
Continued
-------
Table 5-9
Page 2
(Continued) INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION
PROCESSES (SOLIDS)
Solids
Sample
guiremen,ts
Container
Preaervation
Holding
Time
Preparation
Analytical Method
Remarks
Cu
Cl
Perchloric acid
digestion.
Ref. 42
Perchloric acid
digestion
NazCOj fusion
Ref. 42
Perchloric acid
digestion
Perchloric acid
digestion
AA, Ref. 14
AA - graphite furnace
Ref. 27
Specific ion electrode,
standard addition
Ref. 41
AA, Ref. 14,
AA, graphite furnace,
Ref. 27
Specific ion electrode,
standard addition
Ref. 42
Solvent extraction - AA,
Ref. 29, AA - graphite
furnace, Ref. 27
Atomic absorption,
Ref. 14
Hg
Gold amalgamation,
coid vapor, flameless AA,
Ref. 43, 44, 45
Perchloric acid
digestion
Atomic absorption,Ref. 14
AA - graphite furnace,
Ref. 27
Perchloric acid
digestion
Perchloric acid
digestion
Solvent extract ion,
atomic absorption.
Ref. 46
AA, Ref. 14; AA - graphite
furnace, Ref. 27
Perchloric acid
digestion
Organic ext ract ion,
fluorescence, Ref, 32, 47
Perchloric acid
digestion
AA - graphite furnace,
standard additions,
Kei. t/
Continued
-------
Table 5-9,
Page 3
INORGANIC ANALYTICAL METHODS FOR COAL GASIFICATION PROCESSES (SOLIDS)
Continued
Solids
Sample Holding
Requirements Container Preservation Time Preparation
Analytical Method
rks
Tl
Sn
Zn
Perchloric acid
digestion
Perchloric acid
digestion
Fusion, Ref. 35
Perchloric acid
digestion
Perchloric acid
digestion
Atonic absorption, Ref. 24
AA - graphite furnace,
Ref. 27
Atomic absorption, Ref. 24
AA - graphite furnace,
Ref. 27
Fluorescence, Ref. 35
AA - graphite furnace,
standard additions,
Ref. 27
Atomic absorption, Ref. 24
AA - graphite furnace,
Ref. 27
P - Polyethylene
C - Class
T - Teflon
-------
shown in Figure 5-4. This scheme is based upon the information
given in Table 5-8.
Organic Analysis -
By comparing the inorganic species in Table 5-1 with
the organic species listed in Tables 5-2 and 5-3, the complexity
of quantitative organic analysis is readily apparent. The
following discussion will provide the reader with descriptions
of the tools available to the analyst and a general approach to
guide the analyst in formulating a strategy for characterizing
the organic species emitted from a specific coal gasification
plant. A more detailed discussion of this subject is available
in Assessment, Selection and Development of Procedures for
Determining the Environmental Acceptability of Synthetic Fuel
Plants Based on Coal,(Ref.15).The general strategy outlined
begins at the point where a sample is presented to the analyst.
This strategy is presented in terms of:
separation techniques,
identification techniques,
quantitative techniques, and
integrated analytical scheme.
Samples from coal gasification processes that contain
highly complex mixtures of organic species will require at least
some degree of simplification by fractionation prior to identi-
fication and/or quantification. The examination of these frac-
tions can then range in depth from the identification of func-
tional groups which will indicate the presence or absence of
classes of species to the quantification of individual species.
Separation Techniques - The separation techniques
described here perform three primary functions. These functions
are:
removal of the species of interest from the
sample matrix (solid, liquid, or sorbent) ,
concentration of these species in a solvent
-73-
-------
Liquid _
Sample
Preparation Analytical
Method Technique
h
p p y
Standard Specific Ion
Additions Electrode
pec p y
Specie*
Thiocyanace
Phosphate
Nitrate
Chloride
Fluoride
Sulfite
Ion Tritrlmetry
Exchange ' (Spectropnotometry;
_J r\ f < 1 1 I n.l
Sulfate
Ammonia
Sulfide
Cyanide
Phenol
Mercury
m
LU
Organic
Extraction
AJ>
1
Standard
Addition
1 -[ HGA-AA [
Inorganic [ Hr v ~n
Complex 1 '^ i
Beryllium
Molybdenum
Manganese
Antimony
Cadmium
Lead
Nickel
Thallium
Aluminium
Calcium
Chromium
Iron
Lithium
Magnesium
Potassium
Silicon
Sodium
Copper
Zinc
Vanadium
Arsenic
organic „,
Extration
Detection
(ppa In orl*i
1.
.01
.1
1
.01
.02
.2
.04
.1
.008
.001
.001
.008
.008
.06
.004
.004
.004
.004
.002
2.0
.09
.1
.2
.04
.008
.005
2.
.02
.1
.01
.05
.003
.01
Hmub DlgesLlou ^t^^y
Ion hxcnange J
Sulfur
-I
. ~l
Figure 5-4. EXAMPLE INORGANIC ANALYTICAL SCHEME FOR LIQUID SAMPLES
-74-
-------
suitable for introduction into the remainder
of the analytical scheme, and
• simplification of the sample by fractionation
into groups containing fewer individual species.
The separation techniques which are considered appli-
cable to samples from coal gasification processes include:
a) extractions from solid samples and sorbent
materials,
b) liquid-liquid partitioning,
c) distillation,
d) inert gas stripping,
e) ion exchange,
f) gel permeation,
g) column chromatography,
h) gas chromatography, and
i) derivatization.
a) Extractions - Extractions from solid samples or
sorbent materials should be performed in a Soxhlet extractor
(Ref. 52 and 53) or similar device. The extraction solvent must
be of high purity. Commercial solvents may require purification
prior to use or characterization to quantify a reagent blank.
The solvent should be volatile. Extractions should be conducted
at as low a temperature as possible to avoid sample loss by vol-
atilization or thermal decomposition. Poor extraction efficien-
cies should be rectified, if possible, by selecting alternate
solvents rather than by increasing temperature.
Extraction of an aqueous sample can be performed batch-
wise, as in a separatory funnel, or in a continuous extractor.
Saponification (Ref. 54) can provide a more thorough extraction
by treating a dispersion of the solvent in the sample. However,
emulsions must be avoided during extraction . In some situations,
-75-
-------
this may require the use of an alternate solvent. As with extra-
tions from solids, sequential extractions with additional sol-
vents may be required for better extraction efficiency.
b) Liquid-liquid partitioning - Liquid-liquid parti-
tioning involves the transfer of selected species from one
liquid phase to another. Species are separated on the basis of
acid/base or polar/nonpolar character. Several aspects of
liquid-liquid partitioning must be considered when designing a
separation scheme using this technique.
Distribution coefficients are a measure of
the degree of partition achievable between two
phases. Sequential partitioning with small
volumes of the extractant which are then
combined will result in a greater total
efficiency than a single extraction with the
same total solvent volume.
Solvent volumes may become large enough for
evaporative concentration to be required.
Several techniques are available for con-
centration including a reduced pressure rotary
evaporator, a Kaderna-Danish evaporator, freeze-
drying and evaporation with an inert gas (Ref.
55, 56). The prime consideration here is to
accomplish the solvent re'moval without loss
of sample through volatilization or thermal
decomposition.
Solvent purity must be assured since solvent
impurities can become concentrated in the
sample. Continuous solvent purity checks
should be an integral part of the analytical
quality control program.
Acid or base catalyzed reactions may result
in the loss of specific sample components.
Standards of reactive species should be
carried through each step to identify reaction
losses as part of the analytical quality
control program.
c) Distillation - Steam distillation provides a means
of separating species on the basis of volatility. Addition of
an acid or base to the sample solution will add an additional
-76-
-------
dimension of separation by acid/base character. The problems
discussed above for liquid-liquid partitioning apply in general
to distillation. The potential for loss by reaction or thermal
decomposition is even greater due to the elevated temperatures
This may be partially alleviated by vacuum distillation but the
use of standards to check the procedure is still required.
f .V1 d) !nert gas stripping - Inert gas stripping provides
a feasible means of separating volatile species which would be
lost in the evaporative concentration steps of other techniques
Non-polar species with boiling points below 100°C can be recov-
ered. The species are stripped from the sample with a low flow
of inert gas (<30 ml/min) and sorbed on Tenax or other sorbents
Recovery by thermal desorption allows direct introduction to a
gas-chromatographic analyzer.
e) Ion-exchange - A separation scheme in which acids
are removed on anion-exchange resins, bases on cation-exchange
resins and neutral nitrogen compounds on FeCl3/attapulgus clay
has been applied extensively to petroleum analysis (Ref. 57,58).
Potential problems with ion exchange procedures include repro-
ducibility, resin swelling, acid resin catalysis of reactions,
extremely long separation times, and losses of trace components.
f) Gel-permeation chromatography (GPC) - Gel permeation
chromatography is an exclusion technique in which retention is
based on molecular size. Molecules too large to penetrate the
support media pores are eluted with the solvent while smaller
molecules are selectively retained. GPC support media are avail-
able in a variety of pore sizes ranging from <50°A to >3000°A.
The mobile phase (solvent) is chosen on the basis of low viscosity,
dissolution properties and compatibility with the support media.
g) Column chromatography - The types of liquid chrom-
atography generally applicable for separating coal conversion
process samples are: 1) nonlinear elution chromatography, 2)
linear-elution chromatography, and 3) reversed-phase chromato-
graphy. Nonlinear-elution chromatography uses an active support
media which tightly binds selected species until they are eluted
with a different solvent. The sequential use of several solvents
or mixtures will result in an equal number of discrete fractions.
Linear-elution chromatography uses a less active support_media
which binds species in varying degrees resulting in continually
eluting bands in a single solvent or a series of solvent mixtures
of gradually increasing polarity. Reversed-phase chromatography
is performed on a non-polar support with polar solvents.
-77-
-------
h) Gas chromatography - Gas chromatography is a tech-
nique widely applied in qualitative and quantitative analysis
but its value as a preparative separation procedure is often
ignored. Preparative scale gas chromatography (PSGC) is in
essence a scale-up of systems more generally used in qualitative
and quantitative analysis. The inlet to the column must be
designed to ensure flash evaporation of the larger sample
quantities involved or severe tailing and incomplete separation
will result. Fractions eluting from the column may be collected
by condensation or sorption on a suitable support medium.
i) Derivatization - Derivatization is a technique by
which a component of interest is, through selective reaction,
converted to a compound which is compatible with a separation.
identification or quantification technique. The most generally
applied derivatization technique is methylation. Many methyla-
tion agents are available and a general discussion of their
applications has been provided by Webb, et. al. (Ref. 56).
Identification Techniques - Many of the techniques
utilized in organic analysis provide both identification and
quantification. This section will deal with those techniques
which are primarily applied only to identification and for which
quantification is difficult or only feasible in specialized
situations. Some of these techniques can provide semi-quantita-
tive data without extensive calibration or development; however,
the accuracy and precision expected of quantitative techniques
is impossible or extremely difficult to attain. These identifi-
cation techniques include:
a) high-performance liquid chromatography,
b) thin-layer or paper chromatography,
c) functional group analysis,
d) mass spectrometry,
e) nuclear-magnetic resonance spectrometry,
and
f) infrared spectrometry.
An analytical separation (as opposed to gross or class prepara-
tive separations) is an integral part of some of these techniques
-78-
-------
a) High-performance liquid chromatographv (HPLC) -
The major advantage of HPLC is its ability to separate very polar
compounds without derivatization, highly reactive compounds, and
compounds which have very low vapor pressure or are thermally
labile. HPLC is a relatively new technique still undergoing a
great deal of development but of great utility in the analysis
of very polar water-soluble species. Column packing alternatives
include porous silica spheres, liquid phase coated support,
bonded phases and ion exchange materials. Elution solvents are
selected to be compatible with both the sample and the column
packing material, and may range from aqueous solutions to non-
polar organic solvents. The use of sequential elution with
varying strength solvents or solvent mixtures is particularly
useful for complex mixtures.
The detectors for HPLC are typically either a differ-
ential refractometer or a fixed wavelength ultraviolet (U.V.) cell
Differential refractometers measure differences in refractive
index between pure solvent and solvent containing solute. Al-
though responsive to all compounds, they are inherently insensi-
tive. U.V. cells are the most widely used HPLC detector and,
depending on the U.V. absorption characteristics of the compound,
is much more sensitive than the refractometer. Recent develop-
ments have provided U.V. cell detectors with variable wavelength
capabilities.
Identification of compounds is achieved in HPLC
through comparison of retention time with standards and, to a
limited extent, examination of U.V. absorption and refractive
index.
b) Thin-layer chromatography (TLC) - This chromato-
graphic technique has limited utility analysis, but its analy-
tical separation capability combined with specific visualization
reagents, fluorescene, and fluorescence quenching, make it use-
ful for general detection. Standards which must be run with each
unknown TLC may also serve as a useful analytical preparative
technique since the spots may be removed and subjected to analy-
sis by other techniques.
c) Functional group analysis - Functional group_detec-
tion by spot test is not useful for the identification of indivi-
dual compounds but provides an inexpensive, "quick and dirty"
check for a given functional group. Spot tests have been
developed for almost every known functional group or combination
of functional groups (Ref. 59). Spot tests can be used in the
-79-
-------
field to provide a quick indication of the presence of a class
of compounds or in a complex analytical scheme to check the
efficiency of separations or derivatizations.
d) Mass spectrometrv - There are two general categor-
ies of mass spectrometry which are decidedly different in their
applications:
High-resolution (HRMS), and
Low-resolution (LRMS) .
The primary value of HRMS is the determination of pos-
sible presence or certain absence of preselected compounds. The
accurate mass determination of HRMS will often provide definitive
identification if mixtures are not overly complex. HRMS is not
capable of distinguishing between isomers and often cannot offer
data for differentiating between compounds of isoatomic structures.
LRMS cannot give the precision of mass that HRMS supplies
since it only measures integral mass numbers. It does, however,
offer several advantages including fast scans over a wide mass
range and chemical ionization. It can be combined with gas
chromotagrphy (GC-MS). The fast scan capability of LRMS pro-
vides data over a large range of mass with a limited sample
quantity. Chemical ionization can help indicate the molecular
weight of a compound. GC-MS combines the separation capabilities
of gas chromatography with the identification capability of mass
spectrometry to provide the most powerful analytical tool avail-
able today. More detailed discussions of GC-MS will be provided
in the section on quantitative analysis.
•
e) Nuclear magnetic resonance spectrometry (NMR) -
NMR has very limited use in an identification scheme for trace
organic compounds contained in complex mixtures. For pure or
relatively pure samples in sufficient quantity NMR may provide
adequate information for complete identification. Some of the
sensitivity limitations can be overcome by Fourier Transform
NMR (Ref. 60), but the specificity limitations remain.
f) Infrared spectrometry (IR) - IR suffers from the
same sensitivity and specificity limitations as NMR. The
greatest application of IR is its power to indicate the presence
of various functional groups through correlation with reference
spectra. The sensitivity of IR can be enhanced through the use
-80-
-------
of attenuated total reflection (ATR). Advantages of ATR include
ease of sample preparation, extreme sensitivity, and non-destructive
analysis. Another recently developed IR technique is Fourier
Transform Infrared (FTIR) which provides enhanced sensitivity
over conventional IR. FTIR can be used to scan peaks from a gas
chromatograph as they emerge.
Quantification Techniques - Three organic quantitative
analytical techniques are discussed here:
a) Gas Chromatography (GC),
b) Gas Chromatography - Mass Spectrometry
(GC-MS), and
c) Colorimetry.
Gas chromatography is the most widely used technique for both
identification and quantification of organic compounds avail-
able today. Gas Chromatography-Mass Spectrometry is treated
separately because of the unique power of this combination.
Colorimetric techniques were widely used in the past but due to
their limitations are currently secondary to GC and GC-MS.
a) Gas Chromatography (GC) - A wide variety of separa-
tion capabilities can be accomplished using various column con-
figurations and a number of detection systems. Identification is
based on the comparison of retention times with those of stand-
ards and the use of selective detectors. The instrument must be
extensively calibrated for retention time measurements. Quanti-
tation is accomplished by comparison of detector response for an
unknown with a calibration curve prepared from standard solutions
of the same compound. Detector response will vary with instru-
ment conditions (flow rates, oven temperatures, and maintenance
of the instrument), compound identity and column type. The use
of literature "response factors" that relate the response of a
detector for various compounds to a common basis is not adequate
for quantification unless the response factors were generated on
the instrument in use at the same operating conditions.
Several detectors are available which lend a great deal
of flexibility and specificity to GC analysis.
-81-
-------
Flame lonization Detectors (FID) - The FID
is a useful detector for quantification
because of its wide linear range, sensitivity,
good response to high concentrations, insen-
sitivity to temperature and pressure changes,
and response to all organic compounds of
interest except formic acid and carbon dis-
sulfide. FID response is in general pro-
portional to sample weight rather than molar
concentration.
Electron Capture Detector (ECD) - The ECD
is much more responsive to electronegative
compounds than to other types. It is suit-
able for polynuclear aromatics, nitrated,
chlorinated and perfluorinated compounds. The
ECD is highly sensitive to these groups of
compounds but is limited by a small linear
dynamic range. Sensitivity and response
vary widely from compound to compound requiring
calibration for each compound over the full
range of expected concentration.
Thermal Conductivity Detectors (TC) - TC
detectors are relatively insensitive by
comparison to FID or ECD and are, therefore,
most useful for higher concentration char-
acterization for inorganic gases or volatile
organics.
Conductivity Detectors - The electrolytic con-
ductivity detector is commonly called the
Coulson or Hall detector (a much improved ver-
sion) . This detector combines catalytic com-
bustion in a controlled atmosphere with
dissolution of the combustion products in an
electrolyte. The resulting variations of
electrolytic conductivity are measured. The
detector is sensitive to sulfur-containing
and halogenated compounds when the combustor
operates in the oxidative mode. The detector
is sensitive to nitrogen-containing compounds
when the combustor operates in the reductive
mode.
Flame-Photometric Detector (FPD) - The FPD
can be made specific to sulfur or phosphorus-
containing compounds by the appropriate choice
of interference filters which pass either the
sulfur or phosphorus emission bands. The
-82-
-------
detector is highly sensitive (0.001 yg,
or less) but is totally nonlinear.
b) Gas Chromatography-Mass Spectrometry (GC-MS) - Gas
chromatography-mass spectrometry (GC-MS) combines separation
identification, and quantification. Both GC retention time and
mass spectra are used for identification. Quantification can be
achieved by three techniques: auxiliary GC detector, total ion
current monitoring, and selected ion monitoring. The use of an
auxiliary GC detector is identical to quantification by GC alone,
while total ion monitoring is similar with the mass spectrometer
serving as the detector. Selected ion monitoring (SIM) also
uses the mass spectrometer as a GC detector but greatly increases
sensitivity and specificity. Either the electron impact or chem-
ical ionization mode may be used for quantification.
A GC-MS system normally includes four major components:
the gas chromatograph, the mass spectrometer, an interface between
the two and a computerized data system. Gas chromatography has
been previously discussed. Both magnetic sector and quadrapole
mass spectrometers have been used. Quadrapole instruments have
rapid scan time and can perform multiple SIM scans simultaneously.
The interface between the GC and MS consists of a molecular sep-
arator of which there are three categories: 1) enrichment by
effusion, 2) enrichment by preferential effusion through a semi-
permeable membrane, and 3) fractionation in an expanding jet
stream. All of these are designed to enrich the concentration of
the solute in the carrier gas. The rapidity with which a GC-MS
system generates data makes the availability of a dedicated com-
puterized data system a virtual necessity.
c) Colorimetry (Ultraviolet-Visible Spectrometry) -
The greatest utility of UV-visible techniques is in quantisation.
A limited amount of qualitative information may be obtained from
evaluation of spectral scans, but the information is limited in
sensitivity and specificity. Quantification depends on the pre-
sence of a unique absorption band either for the compound of
interest or some derivative. It is based on the Beer-Lambert Law
and concentrations are determined from a calibration curve gen-
erated from standard solutions. The technique of standard add-
itions may be useful for eliminating matrix interferences.
Integrated Scheme for Organic Analysis - Now that the
techniques available to the analyst have been briefly described,
it is necessary to consider how these can be applied in some
logical manner to the characterization of a sample from a coal
-83-
-------
conversion process. Figure 5-5 provides an example of such an
integrated scheme which is designed to apply to solid, liquid,
or gaseous samples. This scheme relies heavily on the use^of
gas chromatography-mass spectrometry for the final identification
and quantification of organic compounds and includes some
inorganic components to indicate the beginnings of the total
integration of the analytical function which is ultimately
required.
The operations in the upper part of Figure 5-5 provide
examples of gaseous species, both organic and inorganic, which
are determined on-site or with minimal preparation prior to off-
site analysis. Immediate analysis of the reactive gases is
required because of their instability. These include carbonyl
sulfide, carbon disulfide, hydrogen sulfide, methyl mercaptan,
methyl amine, hydrogen cyanide and other low molecular weight
sulfur and nitrogen species. The nonreactive gases include low
molecular weight hydrocarbons and "fixed" inorganic gases such as
oxygen, nitrogen, carbon dioxide. Acid gases trapped in a KOH
impinger include inorganics and organics such as carboxylic
acids, phenols, sulfonic acids and aryl thiols. Basic gases,
such as low molecular weight amines and ammonia, will be col-
lected in an acidic impinger. Aldehydes are easily oxidized and
thus should be collected as the more stable aldoximes, hydrolyzed
with HCL, and then analyzed as aldehydes by GC.
General reference sources which should prove useful in
the analysis of organic compounds include:
Identification and Analysis of Organic
Pollutants in Water, L.H. Keith, 1976
(Ref. 61);
Current Practices in GC-MS Analysis of
Organics in Water, Webb,Garrison,Keith
and McGuire (Ref. 56);
Sampling and Analysis Procedures for
Screening of Industrial Effluents for
Priority Pollutants,Revised, April 1977,
EPA (Ref.6);
Techniques of Combined Gas Chromatography/
Mass Spectrometry, McFadden, 1973 (Ref. 62) ;
and
-84-
-------
Assessment, Selection and Development of
Procedures for Determining the Environmental
Acceptability of Synthetic Fuel Plants Based
On Coal, (Ref. 15) .
-85-
-------
Figure 5-5a. INTEGRATED SCHEME FOR SEPARATION AND ANALYSIS
OF ORGANIC CONSTITUENTS FROM COAL GASIFICATION
PROCESSES
-86-
-------
LEGEND:
LC - Liquid chrooatography
ORC - Organic solvent phase
AQ - Aqueous phase
ALC - Alcoholic (MeOH or EtOH)
GRAV - Gravimetric analysis
GC-HS - Combined gas chromatography -
mass spectronetry analysis
TCO - Total chronatographable organlce
analysis by GC-FID (CT-Cj7)
TOC - Total organic carbon analyals
HPLC - High pressure liquid chronatography
GC/FID - Gas chromatography using flame
lonlzatlon detector
GC/FFD - Gas chromatography ualng flame
photometric detector
Figure 5-5b.
INTEGRATED SCHEME FOR SEPARATION AND ANALYSIS OF
ORGANIC CONSTITUENTS FROM COAL GASIFICATION
PROCESSES
-87-
-------
SECTION 6.0
DATA MANAGEMENT
The previous sections of this document have addressed
the general areas of test objectives, engineering analysis and
sampling/analysis. Another important aspect of an environmental
assessment test program is data management, which is assumed here
to consist of:
experiment planning,
data validation,
data evaluation, and
data handling.
From a data management point of view, experiment plan-
ning involves the establishment of a quality control program and,
in some instances, the use of statistical experimental design
procedures. Consideration of how the data will be validated,
evaluated, and reported is also important since this may iden-
tify potential problems in using the data to meet the objectives
of the test.
6.1 PLANNING EXPERIMENTS
A properly executed experiment planning effort should
accomplish two very important functions. These are:
identifying the data necessary to satisfy the
test objectives (portions of this point have
been discussed in Sections 2.0 and 3.0), and
providing a method for ensuring that the data
obtained are accurate.
To accomplish these objectives, sound experimental and quality
control programs must be developed. The details of an
-88-
-------
experimental program need to be established early in the test
plan since its design affects and is affected by the sampling
and analysis methods to be used. A quality control program is
needed to minimize the propagation of determinative errors (bias)
through the samp ling/analysis/evaluation chains of the test
program.
6.1.1 Statistical Experimental Design
From a statistical point of view, experimental design
is concerned with the planning of an experiment or set of exper-
iments so that the resulting data can be used to answer the ques-
tions of the experimenter through the use of statistical analysis,
It also provides a set of rules for collecting samples. Although
data from normal operating logs can sometimes be used for regres-
sion analysis or analysis of variance, an experimental design
provides assurance that all data needed to satisfy the experimen-
tal goals will be collected. The following discussion serves
basically to introduce the subject. The classic work by Cochran
and Cox (Ref. 63) is strongly recommended for a detailed approach
Perry's Chemical Engineer's Handbook (Ref. 64) also provides a
good introduction with specific applications to process experi-
mentation.
Three types of experimental designs which have appli-
cability to quantitative waste stream and control equipment
testing are:
full factorial,
fractional factorial, and
Box-Wilson.
In order to illustrate the significant characteristics of these
techniques, the following examples have been prepared. Consider
an experiment where the effects of volume flow rate (V) and
temperature (T) on S02 emissions from a control process are to
be assessed. Assume 3 levels of flow rate and 3 levels of temp-
erature will be studied in the experiment. A full factorial
experimental design would appear as in Figure 6-1, and would
consist of 9 different tests.
-89-
-------
Vi
V3
T2
T3
VjTi V2T! V3T!
VjT2 V2T2
ViT3 V2T3
V3T2
V3T3
Figure 6-1 A FULL FACTORIAL DESIGN INVOLVING TWO INDEPENDENT
VARIABLES AT THREE LEVELS
The number of tests required for a full factorial
design rapidly becomes excessive as the number of variables
and/or the number of test levels increase. Fractional factorial
and Box-Wilson designs are procedures for reducing the number of
tests required from that of a full factorial design. The frac-
tional design uses an integer fraction (a multiple of the number
of test levels) of the number of tests required by the full fac-
torial design. For example, a 1/3 fractional design for an
experiment involving 4 variables at 3 test levels (31* = 81'pos-
sible tests) would consist of 27 tests. The combinations of
levels to be tested are chosen randomly. Two National Bureau of
Standards (NBS) documents (Ref. 65 and 66) provide tabulations
of factorial designs for variables tested at three levels and
two levels, respectively.
Box-Wilson designs approach the problem from a process
modeling viewpoint, but the end result is the same, -i.e., the
number of required tests is reduced in a random fashion. A
general set of experiments has been designed to derive the math-
ematical model of the process. The experimental design includes
three types of independent variable combinations: axial, fac-
torial and center. Axial points consist of each variable at its
extreme points (maximum and minimum). while all other variables
are at their mid-points. The factorial points include all com-
binations of intermediate levels (i.e., extreme points are
omitted). The center point consists of all variables at their
mid-points, and is usually repeated several times for purposes
of estimating error.
-90-
-------
6.1.2 Quality Control Program
A quality control program needs to be developed during
the planning stage of a sampling and analysis program in order
to ensure that the data obtained are accurate. Developing a
sound quality control program involves:
estimating the expected variability of the data,
estimating the error propagation through the
sampling/analysis chains of the program, and
selecting methods for maintaining sampling and
analytical accuracy.
Estimates of data variability and error propagation involve
interplay with the engineering analysis and with the selection
of sampling and analysis techniques. Because data variability
and error propagation are also important concerns in data valida-
tion, these areas will be discussed in Section 6.2.
The methods which can be used to maintain a quality
control program are relatively simple and generally should be
performed in the field to ensure rapid feedback of information
when a loss of control is indicated. The two most common types
of analysis are the use of correlation tests and the maintenance
of control charts.
One of the most important techniques for maintaining
accuracy in a sampling/analysis chain is the regular use of
alternative pairs of results determined by two different methods.
Correlation analysis of data pairs will indicate the degree to
which the results are similar. (They do not have to be identi-
cal, but they must be linearly related.) To ascertain whether
apparent differences in the results are real, several common
statistical methods may be used, such as confidence intervals,
t-tests, or F-tests. (Refer to any statistical methods book,
such as Ref. 67 or 68.)
For daily maintenance of quality control, an excellent
technique is the use of a quality control chart, such as the
Shewhart Control Chart (Figure 6-2). These charts can be used
to display such quality control data as: replicate sample results,
-91-
-------
110-
> 105
QC
UJ
>
O 100-
111
oc
* 95-
90-
UPPER
CONTROL LIMIT
LOWER
CONTROL LIMIT
I
5
10
I
15
I
20
CONSECUTIVE DAYS
Figure 6- 2. EXAMPLE OF A QUALITY CONTROL CHART FOR SPIKED
SAMPLES (REF. 5).
-92-
-------
isokineticfisampling rates, EPA sampling train calibration fac-
tors, and spiked sample recovery results. For most applica-
tions, the control limits are set at three standard deviations
(3a) for replicate results. The initial estimates of 0 can be
based on previous results or on laboratory results. Later a
running estimate of the pooled standard deviation should be main-
tained for setting the control limits. If any variations that
are outside of the control limits occur, the source of the pro-
blem should be identified and corrective measures should be taken
If a detailed evaluation of the sources of error in the
sampling and analysis chain is needed, analysis of variance can
be used. However, this procedure requires replication at every
stage of sampling and analysis at which variations can occur.
In most field testing, this is impractical.
6.2 DATA VALIDATION
The importance of data validation to a successful sam-
pling and analysis program cannot be overemphasized. In order
to draw meaningful conclusions and recommendations, test results
must be validated. After the analytical results have been
transformed into stream concentrations (i.e., data reduction),
useful data validation techniques include material balance and
statistical calculations. Test results should also be compared
with expected results, i.e., estimates based on engineering
calculations or the results of previous tests of a similar nature
6.2.1 Material Balance Calculations^
As stated previously, material balance tests around a
process are usually performed on an element (e.g., sulfur). If
the results of the element balance indicate that the amount of
the element entering a process does not equal (within experi-
mental error) the amount exiting the process, the methods used
to measure stream flow rates, collect the samples and analyze
the element need to be checked and corrective procedures
initiated.
It should be emphasized that, closure of a material
balance does not mean that there are no errors in the sampling,
analytical or stream flow rate determination techniques. For
example, errors in flow rate measurement may offset errors in
-93-
-------
sample collection and/or analysis. To minimize this problem,
instrument maintenance calibrations should be performed fre-
quently during the test.
6.2.2 Statistical Data Validation
Methods for estimating data variability should be con-
sidered in all test plans. Several statistical quantities use-
ful in measuring data variability are:
mean value - the arithmetic average of a set of
data,
sample standard deviation - a measure of the
distribution of a set of data about the
calculated mean value,
confidence interval - a range about the calculated
mean wherein a certain percentage of the measured
data should lie if they are normally distributed
(e.g., the probability for a measurement to fall
within one sample standard deviation of the
mean is about 68%; two sample standard devia-
tions is about 95%) , and
error propagation - a measure of the cumulative
error in a value due to error in the methods used
to obtain the value (for the case of test data,
the error is introduced by the sampling and analy-
tical techniques).
It is not the intent of this document to present a detailed dis-
cussion of statistics. Instead, reference is made to the excel-
lent works by Snedecor and Cochran (Ref. 67) or Bowker and
Lieberman (Ref. 68).
6.3 DATA EVALUATION
The types of data evaluation techniques required are
dependent upon the objectives of the test. In the following
text, data evaluation techniques for typical waste stream, con-
trol equipment, and process stream characterization tests are
discussed. Specific examples are also given in order to help
explain the evaluation procedures.
-94-
-------
6.3.1 Waste Stream Characterization Test
As discussed in Section 2.0, waste stream characteri-
zation tests are grouped into three test types: screening tests
(Level 1), quantitative tests (Level 2), and continuous monitor-
ing (Level 3) . Data evaluation procedures for each of these
tests are discussed below.
Level 1 Screening Test -
The primary objective of a Level 1 waste stream char-
acterization test is to identify the potentially hazardous waste
streams that need to be further characterized. Identification
of these streams is achieved by performing both chemical analysis
and bioassay tests. A detailed discussion of the procedures for
reducing and evaluating the raw Level 1 chemical data is pre-
sented in two draft reports: Approach to Level 2 Analysis Based
on Level 1 Results , MEG Categories and Compounds and Decision
Criteria (Ref.69), and Suggested Report Format for Level 1
Organic Analysis Data (Ref 70). A brief discussion of the pro-
cedures used to reduce Level 1 chemical and biological data and
to identify potentially hazardous waste streams is discussed in
the following text.
Level 1 Chemical Dat^a - In order to reduce the raw chem-
ical data obtained from a Leve^l 1 test, organic data from liquid
chromatography (LC), gas chromatography (GC), infrared spectro-
scopy (IR), low resolution mass spectrometry (LRMS), and spark
source mass spectrometry (SSMS) need to be converted to concen-
trations (mass loadings) of compounds or compound classes (e.g.,
phenols) in the waste stream sample. These concentration values
may be reported in yg/m3 for gases, yg/g for solids and yg/£ for
liquids. These concentrations are then compared with appropriate
decision criteria values in making a judgment on whether further
characterization is necessary. It is currently recommended that
minimum acute toxicity effluent (MATE) values be used as these
decision criteria. These MATE values are given in Multimedia
Environmental Goals for Environmental Assessment (Ret!71).
MATE values are specific for individual chemical spe-
cies while the results obtained from Level 1 tests give _concen-
trations primarily for compound classes with only a limited
number of specific compound concentrations. Therefore, for each
category that is found in the sample, the worst chemical that
-95-
-------
could be present is used as a comparison value. In the cases
where LRMS is used, specific compounds will be identified and
their concentrations should be compared to their respective
MATE values.
The following is a general procedure for analyzing the
results of a Level 1 waste stream characterization test .
1. List the categories of compounds found in the
sample by MEG category number and estimated
concentration .
2. For each MEG category, identify the compound
with the lowest MATE value (Ref. 69).
3. If the estimated concentration in the stream
is lower than this MATE value, go to the
next category .
4. Review the data to see if there is any
evidence that the compound with the lowest
MATE value cannot be present in the sample.
This may be indicated from LRMS data on
molecular weight ranges of the compounds
present in the samples .
5. If the compound with the lowest MATE value
can be ruled out, identify the compound with
the next lowest MATE value and reiterate
steps 3, 4, and 5 until MATE values have
been selected for each category
6. List the compounds and their MATE values
selected for comparison between the estimated
concentrations calculated from the Level 1
rest
7. If the estimated concentration values are
greater than the MATE values, the stream
or stream fraction may need further
characterization.
An example of a Level 1 organic analysis results reporting for-
mat is given in Table 6-1. This table shows the MEG category
and number, estimated concentration from the organic analysis
of the sample, the MATE value of the most hazardous compound in
-96-
-------
Table 6-1. LEVEL 1 ORGANIC ANALYSIS RESULTS COMPARED TO MATE VALUES
MEG Concentration Found Lowest MATE in Category
Category Found Number mg/m3 (compound) *
Fused Aromatics 21, 22
>216 MW
Fused Aromatics 21, 22
<216 MW
Heterocyclic N 23
Compounds
Carboxylic Acids 8A, B
Sulfur Compounds 53
Heterocyclic S 25
Compounds
Heterocyclic 0 24
Compounds
Phenols 18
Esters 80
Aromatics Benzenes 15
Aliphatic HC's 1
c
-/
5
1 *
1
0.6
0.5
0.3
0.1
0.08
0.06
0.06
2.0 x 10
30 mg/m3
16 mg/m3
6 mg/m3
1 mg/m3
1 mg/m3
590 mg/m3
2 mg/m3
5 mg/m3
1 mg/m3
200 mg/m3
5mg/m3 (benzo [ajpyrene)
(me thylphenan threnes )
(quinoline)
(phthalic acid)
(sulfuric acid)
(benzonapthothiophene)
( tetrahydrof uran)
(1,4-dihydroxybenzene)
(phthalates)
(biphenyl)
(cyclopentadienes)
Ratio
Found /MATE
2.5 x 10s
0.2
0.06
0.2
0.6
0.5
0.0005
0.05
0.02
0.06
0.0003
* eliminating compounds that cannot possibly be present,
-------
Table 6-2.
BIOASSAY TEST RESULTS FOR A COAL GASIFICATION
FACILITY
Coal
HEALTH TESTS
1. AMES SP
2. Cytotoxlctty
WI-38, EC-50 (cell count, ^60(s)
Ug solid, m3 gas/ml
culture)
RAM, EC-50 (cell count, lig >1000(s)
solid and liquid, m! gas/ml
culture)
3. Rodent Acute Toxlcity M
l.D-50 (g sample/kg rat) >10
ECOLOGICAL TESTS
Fresh Waterb
Algal, EC-50 (15 days
Daphnia, LC-50 (48 hours)
Fathead minnow, LC-50 (96 hours)
Salt Waterb
Algal, EC-50 (12 days) Filtered/
unfiltered
Shrlnp, LC-50 (96 hours)
Barrel
Barrel Valve Separator
Valve Vent Cas Vent Gas
Vent XAD-2 Gaslfier Cyclone XAE-2 Separator
P N N P SP N
4 x 10~"(g) - - 7 x 10~6(g)
>2 x lo"3(g) >300(s) >1000(s) >1000(s) >1 x I0"!(g) >600(1)
- - L M H L
>10 >10 >10 - >10
1.0 to 0.1Z
0.1U
0.02Z
0.53/0.41Z
0.25Z
Sheepshead minnows, LC-50
(96 hours)
Terrestial
Soil microcosm
Plant stress ethylene
0.16Z
* Indicates a plant waste scream.
Test not performed.
SP: Slightly Positive
P: Positive
N: Negative
H: High toxicity
M: Medium toxicity (i.e., rats showed hair loss, eye discoloration, etc.)
L: Low toxicity (i.e., no significant effects noted)
EC-50'o were calculated on the XAD-2 extract for the harrel valve vent and separator vent gases by:
Nm1 vent gas/ml culture
EC -50 reported
EC-50 ] in ui Of extract
per ml culture
(g) : gas, (s) : solid, (1) : liquid
mg of organics
extracted per ml
of extract
mg o f organics
per Nm3 of
vent gas
EC-5CTs and LC-5Cfs for fresh and salt water tests are presented in wt % of the sample in water.
CSoil microcosm tests are ranked for No. 1 the most toxic to No, 5 the least toxic.
Source: Environmental SaiBp_Ii_n_g. and Analysis Program in a Commercial l.ow-Btn Gasification
Facility - Preliminary Results ~
-98-
-------
the MEG category that could be present in the sample and the
ratio of estimated concentration to MATE value.
From the example data given in Table 6-1, the follow-
ing conclusions can be drawn.
The fused aromatics category (>216 Molecular
Weight) had a concentration/MATE ratio much
greater than unity; therefore, further char-
acterization of this sample is required.
• Fused aromatics (<216 MW) , carboxylic acids,
sulfur and heterocyclic sulfur categories
have ratios close to unity and may require
further characterization.
For the remaining compound categories, the
ratios are much less than one; therefore,
further characterization may not be required.
From the above conclusions, further characterization of LC frac-
tions 2 and 3 (containing fused aromatics) is required. This
characterization, however, may range between just looking for
benzo-a-pyrene to performing a comprehensive GC/MS character-
ization of all the organic species present in these fractions.
A decision to go back to the test site and perform a Level 2
quantitative waste stream characterization test may also be
appropriate to characterize the emission rates of specific fused
aromatic compounds in the waste stream sampled.
Level 1 Bioassay Tests - Bioassay tests for Level 1
studies can be performed on the total waste stream sample or
on certain fractions of the sample (e.g., the extract from the
XAD-2 resin used to capture organic species). As discussed in
Section 5.0, Analytical Procedures, Level 1 bioassay tests in-
clude screening tests for both health and ecological effects.
The exact procedures and interpretations of the results from
these bioassay tests are currently under development by the EPA.
Bioassay tests have been performed on certain samples
collected from a low-Btu gasification facility (as described in
Section 3.C). The results of these tests are shown in Table 6-2.
The conclusions that can be drawn from these results are listed
below.
-99-
-------
The separator liquor sample was very toxic to
aquatic species; however, it was least toxic in
the soil microcosm test, had negative results
for mutagenicity (carcinogenicity), cytotoxicity,
and rodent acute toxicity. Because of the toxic
effects on aquatic species, this liquor would
need to be treated before being discharged into
a receiving water body.
The cyclone dust and tar were the most toxic
samples analyzed by the soil microcosm tests
and would require treatment if these streams
were sent to a landfill.
The separator tar had a positive Ames test
result and was the most toxic sample tested
in the rodent acute toxicity tests. Although
this stream is not a waste stream, fugitive
tar emissions from pumps, valves, etc. could
present a. hazard to workers in the plant.
Positive results were obtained from the Ames test
on the barrel valve vent gas XAD-2 extract. There-
fore, the barrel valve vent gas should be treated
before being discharged into the atmosphere.
The gasifier ash was lowest in soil microcosm
toxicity and showed low toxicity signs in the
rodent acute toxicity tests. Leachate from
the ash should be tested independently; however,
these data indicate that disposing of the ash
in landfills may be an acceptable control.
From the bioassay test results shown in Table 6-2, it is obvious
that not all of the tests give all positive (high toxicity) or
negative (low toxicity) results (e.g., the separator liquor is
highly toxic to aquatic species but showed negative or low
toxicity in the other tests). Therefore, all of these tests are
necessary to ensure that a complete biological screening is
performed on each sample.
Level 2 Quantitative Test -
The purpose of a Level 2 waste stream characterization
test is to provide quantitative data for pollutants in waste
streams. These data may be either compound classes or specific
-100-
-------
compounds, depending upon the objectives of the Level 2 test-
For example, quantitative data for classes of compounds may be
specified for denned input streams to a needed control device
Quantitative data for specific compounds may be required to
determine whether a control device is needed or whether a spec-
ific compound needs to be continuously monitored (Level 3).
In order to satisfy the objectives of a Level 2 test
raw chemical data need to be reduced such that the emission rates
of pollutants in a waste stream can be determined. The results
of the Level 2 characterization program can then be used to
answer the following questions.
• Are the pollutant (e.g., NO SO ) emission
rates within environmental standards?
• Will specific pollutants present a potential
health and environmental hazard?
What type of control equipment will be required
to treat the waste stream?
Is continuous monitoring of specific pollutants
required?
It should be emphasized that because the objective of Level 2
tests is to obtain quantitative results, data validation tech-
niques, as discussed in Section 6.2, should be used before eval-
uating those results.
Level 3 Continuous Monitoring Tests -
The purpose of a Level 3 waste stream characterization
program is to provide continuous data on the emission rates of
specific pollutants in a waste stream. Data reduction and eval-
uation techniques are similar to those required for Level 2 data
to obtain pollutant emission rate information.
6.3.2 Control Equipment Characterization Tests
The purpose of a control equipment characterization
test is to obtain information on the effectiveness of a control
process in treating a waste stream or to collect data required
-101-
-------
to design a control process to treat a waste stream. The data
obtained from this type of environmental test require evaluation
techniques as simple as calculating compound flow rates or as
complex as developing an equation for the design of a control
process.
An example of the complexity of data evaluation proce-
dures for a control equipment characterization test is to obtain
design information for a hot acid gas removal process. The type
of design data required include:
mass flow rates of streams, major compounds,
and acid gases to be treated;
mass flow rates of compounds that may interfere
with the removal of acid gases;
physical conditions (temperature and pressure)
of the stream to be treated; and
kinetic expression for the rate of removal of
acid gases as a function of reactant species
concentration, temperature, and pressure.
Statistical analysis techniques can also be applied to
the results of control equipment characterization tests. Of these
techniques, two are particularly important - correlation analysis
and regression analysis. Detailed discussions of these statis-
tical techniques are available in the literature (Ref. 67 and 68)
The features of correlation analysis are:
gives a quantitative assessment of the extent
to which two parameters vary together,
requires only that data for the two parameters
be collected in pairs and that they come from
a bivariate normal distribution,
is an excellent means for comparing the
results of alternative sampling/analytical
procedures, and
cannot be used to study cause-and-effeet
relationships between variables.
-102-
-------
In comparison, regression analysis:
provides an estimate of error in the data
generation chain,
provides an estimate of the dependence of
one variable on one or more other variables,
and
can be used to develop a simulation model
of a process or process module.
6.3.3 Process Stream Characterization Test
The purpose of a process stream characterization test
is to provide information on the relationship between process
operating parameters and the characteristics of the process
waste streams. The data reduction and evaluation procedures
required for this test are similar to those for control equipment
characterization tests. Mass flow rates and concentrations of
pollutants need to be calculated from the raw sampling and anal-
ytical data. Process operating parameters (stream flow rates,
temperature, pressure) also need to be accurately measured if
data are to be obtained for checking or developing a process
design equation. As for control equipment characterization tests,
statistical analyses often are useful data evaluation techniques
for process stream characterization tests.
6.4 DATA HANDLING
The results of an environmental sampling and analysis
program will usually comprise an extensive data base. Each piece
of analytical data must be correlatable or identifiable with the
plant operating conditions at the time of sampling. These data
may either be recorded in tabular form and stored under process
key words or coded and stored in a computer data base.
-103-
-------
6 . 4.1 Manual Data Base Organization
The key feature of any data storage system is its
ability to provide efficient and logical accession of its con-
tents. If the data are organized in a manual retrieval data
base, all bits of data should be related to the specific test
run in which they were generated and organized under key word
subject categories.
Each test run should be given a unique name and ref-
erence number. Test run references should then be listed under
key word subject categories so that personnel desiring informa-
tion under one of these topics can quickly identify which test
runs are of interest. One way to handle this is to list ref-
erences to test runs on 3 x 5 library cards.
Of course, referencing test run information through
key word categories only provides a list of data sources (unique
test run numbers). The actual data must still be available in
some convenient and readily accessible form. This could be the
final reports on the results of each characterization test run.
Pertinent information from each characterization test run can
also be summarized on tables and each table given a unique code
number. In this case, it would be these numbers which would be
listed under the key word categories to guide personnel in
accessing characterization data.
Key data subject categories should consist of topics
for which it is anticipated that retrieval of data will be
desirable. As an example, it would be desirable to access data
by process, by control equipment, by streams and their sources,
and by pollutants. Suggestions for organizing the data in each
of these key word categories are shown in Table 6-3.
Information under the process and control equipment
key word categories should include information unique to that
process that might separate it from a similar process in another
plant, such as feedstock process capacity, operating parameters,
input and output streams, and product gas end-use. Information'
under stream and effluent key words should define the process
and site to which the stream data are related: flow rates, pres-
sure, temperature, chemical composition, stream source and
-104-
-------
Table 6-3.
KEY WORD CATEGORIES FOR ENVIRONMENTAL ASSESSMENT
DATA BASE
Process Key Words
Coal Pretreatment
Drying
* Partial Oxidation
• Storage
Handling
Crushing/sizing
Briquetting
Pulverizing
Coal Gasification
Fixed-bed, pressurized, dry ash
Entrained-bed. pressurized, slagging
Fixed-bed, pressurized, slagging
Fixed-bed, atmospheric, dry ash
Fluid-bed, atmospherid, dry ash
Entrained-bed, atmospheric, slagging
Gas Purification
Particulate removal
Tar removal
Quenching and cooling
• Acid gas removal
Control Equipment Key Words
Air Pollution Control
Particulate control
Sulfur control
Hydrocarbon control
Nitrogen oxide control
Others (noise, heat, etc.)
Water Pollution Control
Oil/water separation
Suspended solids removal
Dissolved organics removal
Dissolved inorganics removal
Solid Waste Control
Chemical fixation
Sludge reduction
Ultimate Disposal
• Landfill
Evaporation ponds
Streams and Their Sources
Process Streams (Source)
• Raw product gas (Wellman-Galusha Gasifier)
- Waste Stream (Source)
• Coal feeder vent gas (coal feed lock hopper)
Pollutants (Stream/Source)
- Sulfur Oxides (flue gas/low-Btu gas boiler)
-105-
-------
stream disposition, and any basic plant information which would
impact stream flow or composition (coal feed rate and charact-
eristics, unusual plant operating conditions, etc.)- Data
available under chemical species should be similar to that listed
under stream or source names. These data should include the
amount or rate of discharge of the chemical species, physical
and chemical characteristics of the stream, and information on
the effluent source and plant conditions at the time the data
were obtained.
All bits of information should reference the environ-
mental test during which the data were gathered. This would
allow data to be considered in terms of the time, plant condi-
tions , and any extraneous problems that occurred during that
particular characterization effort.
6.4.2 Computer Data Base
A computer data base would be organized similarly to
a manual data base and would contain the same information. All
bits of characterization data would have to be coded and entered
in the computer. Information could then be retrieved by com-
puter output and search commands. The primary advantage of the
computer data base is that it could more easily handle the large
amounts of information which would be generated in a number of
environmental test efforts. Also a computer data base could be
adapted for search and comparison routines (e.g., identification
of all SOX sources that exceed a specified level, or identifica-
tion of sources that contain a certain combination of pollutants
and exceed a specified flow rate).
As in the manual data base, all information would have
to be linked to the unique environmental test (process, site,
and time) and to process conditions and operating mode that
provide the basis for the test.
-106-
-------
SECTION 7.0
REFERENCES
1. Hamersma J. W. , S. L. Reynolds, and R. F. Maddalone , IERL
RTP Procedures Manual: Level 1 Environmental Assessment
EPA 600/2- 76-106a. Redondo Beach, CA, TRW Systems Group',
j uii G , i y / o .
2. U.S. Department of the Interior, Bureau of Reclamation,
Upper Colorado Region, El Paso Gasification Project, San
Juan County, NM, draft environmental statement DFS-7A-77
1974. ' "•
3. El Paso Natural Gas Company, Application of El Paso Natural
Gas Company for a Certificate of Public Convenience and
Necessity. Docket No. CP73-131. El Paso, TX, 1973.
4. Environmental Protection Agency, "Sample and Velocity
Traverses for Stationary Sources, 40 CFR 60, App . A,
Reference Method I," Fed. Reg. 41(111), 23061-63, (1976).
5. National Institute for Occupational Safety and Health,
Public Health Service, The Industrial Environment - Its
Evaluation and Control. Contract No. HSM-99- 71-45 ,
Washington, DC, GPO , 1973.
6. Environmental Protection Agency, Environmental Monitoring
and Support Laboratory, Sampling and Analysis Procedures
for Screening of Industrial Effluents for" Priority
Pollutants, Revised. Cincinnati, OH, April 1977.
7. Ames, Bruce N., et . al . , "Carcinogens as Frameshift Muta-
gens : Metabolites and Derivatives of 2-Acetylaminof luorene
and Other Aromatic Amine Carcinogens," Proc. Nat. Acad. Sci .
691(11), 3128-32 (1972).
8. Ames, Bruce N., Frank D. Lee, and William E. Durston, "An
Improved Bacterial Test System for the Detection and Class-
ification of Mutagens and Carcinogens," Proc. Nat. Acad. Sci
70(3), 782-86 (1973).
9. Ames, Bruce N., at al., "Carcinogens are Mutagens: A Simple
Test System Combining Liver Homogenates for Activation and
Bacteria for Detection," Proc. Nat. Acad. Sci. 70_(8) , 2281-
85 (1973).
-107-
-------
10. Luthy, Richard G., Manual of Methods: Preservation and Anal-
ysis of Coal Gasification Wastewaters. FE-2496-8.Pittsburgh,
PA, Carnegie-Melion University Environmental Studies Institute,
July 1977.
11. Environmental Protection Agency, Methods for Chemical Anal-
ysis of Water and Wastes. EPA 625 /6-74-003.Washington,
DC, Environmental Protection Agency, Office of Technology
Transfer, 1974.
12. Harstein, A. M., R. W. Freedman, and D. W. Platter, "Novel
Wet-Digestion Procedures for Trace-Metal Analysis of Coal
by Atomic Absorption," Anal. Chem. 45_(3), 611 (1973).
13. American Public Health Association, American Water Works
Association, and Water Pollution Control Federation, Standard
Methods for the Examination of Water and Waste Water, 14th
ed.Washington, DC, American Public Health Association, 1975.
14. American Society for Testing and Materials, 1977 Annual Book
of ATSM Standards, Part 31. Water. Philadelphia, PA, 1977.
15. Oldham, Ronald G., and Robert G. Wetherold, Assessment,
Selection, and Development of Procedures for Determining
the Environmental Acceptability of Synthetic Fuel Plants
Based on Coal~ Revised Report, 4 vols . FE-1795- 3. Austin,
TX, Radian Corp., May 1977.
16. Pinta, Maurice, Detection and Determination of Trace Elements.
Absorption Spectrophotometry, Emission Spectroscopy and
Polarography,Trans.by Miriam Bivas.London, Ann Arbor, 1972.
17. Ruch, Walter E., Quantitative Analysis of Gaseous Pollutants.
Ann Arbor, MI, Interscience,1970 .
18. Bailey, J. B. W., N. E. Brown, and C. V. Phillips, "A Method
for the Determination of Carbon Monoxide, Carbon Dioxide,
Carbonyl Sulphide, Oxygen, and Nitrogen in Furnace Gas
Atmospheres by Gas Chromatography," Analyst (London) 96, 477
(1971). —
19. Jones, Carolyn N., "Gas Chromatographic Determination of
Hydrogen, Oxygen, Nitrogen, Carbon Monoxide, Carbon Dioxide,
Hydrogen Sulfide, Ammonia, Water, and C\ Through C2 Saturated
Hydrocarbons in Refinery Gases," Anal. Chem. 39(14), 1858-60
(1967).
20. Kim, Ann G., and Louise J. Douglas, Gas Chromatographic
Method for Analyzing Gases Associated with Coal. R.I. 79~03.
Pittsburgh,PA, Pittsburgh Mining and Safety Research Center
1974.
-108-
-------
S1!?;, 5ranS8 MV^d^hn E" Si§sby. "Chemiluminescent
Method for NO and NO (NO + N02) Analysis," Environ Sci
Tech. 8(2), 149 (197$).
22. Baker, R.L., "Determination of Fluoride in Vegetation Using
the Specific Ion Electrode," Anal. Chem. 44(7), 1326 (1972).
23. Orion Research Inc., Analytical Methods Guide 3rd ed
Cambridge, MA, 1972. '
24. Perkin-Elmer, Analytical Methods for Atomic Absorption
Spectrophotome try. Norwalk, CT, 1973.
25. Jacobs, M. L., Sandra L. Sweeney, and Clyde L. Webster,
"Determination of Trace Elements in Coal by Spark Source
Mass Spectrography," Reprint. Keystone Coal Industry
Manual. New York, NY, McGraw-Hill, 1975.
26. Ediger, Richard D., "Atomic Absorption Analysis with the
Graphite Furnace Using Matrix Modification," Atomic
Absorption Newsletter 14(5), 127 (1975).
27. Instrumentation Laboratory, Inc., Analytical Instrument
Division, Atomic Absorption Methods Manual, Vol. 2,
Flameless Operations"! December 1976 .
28. Mair, James W. , Jr. and Harry G. Day, "Curcumin Method for
Spectrophotometric Determination of Boron Extracted From
Radio-Frequency Ashed Animal Tissues Using 2-Ethyl-l,
3-Hexanediol," Anal. Chem. 44(12), 3015 (1972).
29. Kinrade, John D., and Jon C. Van Loon, "Solvent Extraction
for Use with Flame Atomic Absorption Spectrometry," Anal.
Chem.46(13). 1894-8 (1974).
30. Hoog, Han, "Future Trends in Refining," World Petrol.
1968 (June).
31. Levesque, M. , andE. D. Vendette, "Selenium Determination
in Soil and Plant Materials," Can. J. Soil Sci. 51, 85-93
(1971).
32. Rankin, John M. , "Fluorometric Determination of Selenium
in Water w/2 , 3-Diaminonaphthalene , " Environ. Sci. Tech.
7(9), 823 (1973).
33 (G K ) Turner Associates, "The Fluorometric Determination
of'Selenium- A Literature Review," Fluorometry Reviews,
Ace No. 11863. Palo Alto, CA, September 1972.
-109-
-------
34. Rader, L. F., and F. S. Grimaldi, Chemical Analysis for
Selected Minor Elements in Pierre Shale.U.S.G.S. Profes-
sional Paper 391-A.Washington, DC, GPO, 1961.
35. (G. K.) Turner Associates, "Uranium," Fluorometry Reviews
Series. Palo Alto, CA, February 1968.
36. Cioni, R., F. Innocenti, and R. Mazzuoli, "The Determination
of Vanadium in Silicate Rocks with the HGA-70 Graphite
Furnace," Atomic Absorption Newsletter LI (5), 102 (1972).
37. American Society for Testing and Materials, 1974 Annual
Book of ASTM Standards. Pt. 26, Gaseous Fuels; Coal and
Coke; Atmospheric Analysis. Philadelphia, PA, 1974.
38. Burke, Keith E., "Determination of Microgram Amounts of
Antimony, Bismuth, Lead and Tin in Aluminum, Iron and
Nickel-Base Alloys by Non-Aqueous Atomic-Absorption Spec-
tres copy," Analyst 9^7, 19-28 (1972).
39. Bokowski, D. L., "Rapid Determination of Beryllium by a
Direct-Reading Atomic Absorption Spectrophotometer," Amer.
Ind. Hyg. Assoc. J. 29_(5), 474-81 (1968).
40. Carlson, R. M., and J. L. Paul, "Potentiometric Determination
of Boron as Tetrafluoroborate," Anal. Chem. 40(8), 1292 (1968)
41. Orion Research Inc., Analytical Methods Guide, 3rd ed.
Cambridge, MA, 1972.
42. Thomas, Josephus, Jr., and Harold J. Gluoskoter, "Determin-
ation of Fluoride in Coal With the Fluoride Ion-Selective
Electrode," Anal. Chem. 46_ (9), 1321 (1974).
43. Kalb, G., and Charles Baldeck, The Development of the Gold
Amalgamation Sampling and Analytical Procedure for Inves-
tigation of Mercury in Stack Gases. PB-210 817. Columbus,
OH, TraDet, Inc., 1972.
44. Diehl, R. C. et al., Fate of Trace Mercury in the Combustion
of Coal. TPR-54. Pittsburgh, PA, Pittsburgh Energy
Research Center, 1972.
45. O'Gorman, J. V., N. H. Suhr, and P. L. Walker, Jr., "The
Determination of Mercury in Some American Coals," Applied
Spectroscopy 26(1), 44 (1972).
-110-
-------
46. Kim C. H. C M. Owens, andL. E. Smyth, "Determination
of Traces of Mo in Soils and Geological Materials by Solvent
Extraction of the Molybdenum-Thiocyanate Complex and Atomic
Absorption," Talanta 21. 445-54 (1974).
47. Levesque, M., andE. D. Vendette, "Selenium Determination in
Soil and Plant Materials," Can. J. Soil Sci. 5_1, 85-93
48. Jacobs, Morris B., The Analytical Toxicology of Industrial
Inorganic Poisons. New York, NY, Interscience, 1967.
49. Monteriolo, S. Cerquiglini, and A. Pepe, "Comparative Study
of Methods for the Determination of Airborne Fluorides "
Pure Appl. Chem. 24(4), 707-14 (1970).
50. Headridge, J. B. , and D, Risson Smith, "Determination of
Trace Amounts of Antimony in Mild Steels by Solvent Extrac-
tion Followed by Atomic Absorption Spectrophotometry "
Lab. Practice 20(4), 312 (1971).
51. Boar, P. L., and L. K. Ingram, "The Comprehensive Analysis
of Coal Ash and Silicate Rocks by Atomic-Absorption Spec-
trophotometry by a Fusion Technique," Analyst 95, 124-30
(1970). ~
52. Spindt, R. S., First Annual Report on Polynuclear Aromatic
Content of Heavy Duty Diesel Engine Exhaust Gases, For Period
Ending 1 July 1974. PB-238 688. Pittsburgh, PA, Gulf
Research and Development Company, 1974.
53. Kleopfer, Robert D., and Billy J. Fairless, "Characterization
of Organic Components in a Municipal Water Supply," Environ.
Sci. Tech. 6(12), 1036-37 (1972).
54. Hughes, D. R. , R. S. Belcher, andE. J. O'Brien, "A Modified
Extraction Method for Determination of Mineral Oil in Sea
Water," Bull. Environ. Contam. Toxicol. 10.(3) , 170 (1973).
55. Pierce, Ronald C., and Morris Katz, "Dependency of Poly-
nuclear Aromatic Hydrocarbon Content on Size Distribution
of Atmospheric Aerosols," Environ. Sci. Tech. 9_(4) , 347
(1975).
56. Webb, Ronald G., et al., Current Practice in GC-MS Analysis
of Organics in Water. EPA-R2-73-277.Athens, GA, EPA,
Southeast Environmental Research Lab., 1973.
57. Happel, John, Robert E. Lief, and Reiji Mezaki, "A Theore-
tical Model Suitable for Higher Conversion Data of Sulfur
Dioxide Oxidation," J. Catalysis 31. 90-95 (1973).
-Ill-
-------
58. Jewell, D. M., et al., "Ion-Exchange, Coordination, and
Absorption Chromatographic Separation of Heavy End Petro-
leum Distillates and the Analysis of Hydrocarbon Fractions,"
ACS, Div. Pet. Chem., Prepr. 16_(4) , C13-20 (1971).
59. Fiegel, F., Spot Tests in Organic Analysis, 7th ed. New York,
NY, Elsevier, 1966.
60. Farrar, T. C., and E. D. Becker, Pulse and Fourier Transform
NMR. New York, NY, Academic Press, 1971.
61. Keith, Lawrence H., ed., Identification and Analysis of
Organic Pollutants in WaterTAnn Arbor, MI, Ann Arbor
Science, 1976.
62. McFadden, William, Techniques of Combined Gas Chromatography/
Mass Spectrometry. New York, NY, Wiley,1973.
63. Cochran, William G., and Gertrude M. Cox, Experimental
Designs, 2nd ed. New York, NY, Wiley, 1957.
64. Perry, John H., Chemical Engineers Handbook, 5th ed. New York,
NY, McGraw-Hill, 1973.
65. Conner, W. S., and Marvin Zelen, Fractional Factorial Exper-
imental Designs for Factors at Three Levels. Applied Math-
ematic Series 54.Washington, DC, NBS (GPO), 1959.
66. Deming, Lois S., Fractional Factorial Experimental Design
for Experiments at~Two Levels. Applied MathematicsSeries
41TWashington, DC, NBS (GPO), 1957.
67. Snedecor, G. W., and W. C. Cochran, Statistical Methods,
6th ed. Ames, IA, Iowa State University Press,1967.
68. Bowker, A. H., and G. J. Lieberman, Engineering Statistics.
Englewood Cliffs, N J, Prentice-Hall, 196TT
69. Ryan, L. E., R. G. Beimer, and R. F. Maddalone, Approach
to Level 2 Analysis Based on Level 1 Results, MEG Categories
and Compounds and Decision Criteria. Redondo Beach, CA,
TRW, 12 January 1978.
70. Harris, Judith C., Suggested Report Format for Level 1
Organic Analysis Data. Cambridge, MA, Arthur D. Little, Inc.,
21 October 1977.
71. Cleland, J. G. and G. L. Kingsbury, Multimedia Environmental
Goals for Environmental Assessment, Vol. 1, Final Report.
EPA-600/7-77-136A, Research Triangle Park, NC, Research
Triangle Institute, November 1977.
-112-
-------
APPENDIX
SAMPLING METHODS
In this appendix the approaches and techniques for
collecting samples from three major stream types (gas, liquid
and solid) are presented. Sampling procedures for collecting
samples containing mixtures of these phases are discussed under
the section dealing with the major stream phase. For example,
particulate matter entrained in a gas stream is discussed in the
gas sampling section. Liquid-solid slurries are discussed in
the liquid sampling section.
A-l GAS SAMPLING
The selection of sampling techniques for collecting gas
samples requires a knowledge of the following:
stream physical conditions (temperature,
pressure, particulate and condensable
content, etc.),
major chemical components in the stream,
reactivity of stream components relating to
both sample stability and safety considerations,
and
physical arrangement of the piping or
ducting containing the stream.
This information is available from the analysis of the test site.
The primary emphasis of the following sections is on
manual methods for collecting gas samples, since the decision
to employ continuous monitoring is often based on the results
of these methods. Execution of either approach requires:
-113-
-------
a carefully selected sampling point, and
access to the stream and withdrawal of a
portion of that stream with immediate or
delayed analysis. (Some continuous techniques
avoid the withdrawal of gas from the stream
by analysis "in situ.")
Selection of the proper sampling point is discussed in Section
4.0 The following text contains a discussion for sampling
train selection (sample withdrawal system) and continuous moni-
toring techniques. Also discussed are techniques for collecting
fugitive gas samples from valves, flanges, etc. in a coal
gasification facility.
A-1.1 Sampling Train
The sampling train, or system of collection devices,
performs several functions, each of which must be considered in
selecting a suitable train. These functions are to:
transport a portion of the gas stream to the
collection (and conditioning) sections,
provide a controlled interface between the
gas stream and the plant work environment,
condition the sampled gas as required for
the collection sections,
collect the components of interest,
measure the volume of gas sampled, and
measure stream conditions (temperature,
pressure, flow rate).
The sampling train designed to fulfill these functions
consists of four sequential sample-processing units. These are:
probe assembly,
particulate collection unit,
-114-
-------
vapor collection unit, and
metering unit.
Through various combinations of its individual components, this
general-purpose sampling train is used to characterize a wide
range of analytical parameters. The sampling involves removing
the gas sample from the stream over a specified period of time.
The train concept presented here is based upon the EPA's Method
5 for characterizing particulate matter in gas streams (Ref. A-l) .
This concept can also be applied to the Source Assessment Sam-
pling System (SASS) developed for Level 1 waste stream character-
ization tests. The Method 5 and SASS trains are illustrated in
Figures A-l and A-2.
Systems of this general type are available commercially
and are suitable for a variety of applications. Gas stream tem-
perature and pressure and the specific requirements of the param-
eters analyzed may dictate modifications. The following sections
discuss each unit of the train, its limitations, and many of the
possible options for modification.
A-1.1.1 Probe Assembly
The probe assembly performs several discrete functions,
each of which must be considered in its design or selection.
These functions are to:
provide sample removal from the stream at
a point or series of points (traversing)
to give a representative sample,
provide a controlled interface between
stream and environment (particularly
important for pressurized or toxic
streams),
• transport the sampled gas to the remainder
of the train without contamination or
alteration of composition, and
• provide measurement of stream velocity
(flow rate), temperature and pressure,
as required.
-115-
-------
Vapor Collection
Probe Assembly
PROOE
REVERSE-TYPE
PITOT TUOE
PHOT MANOMEItill
Particulate
Collection
•IEATEO
AREA
THERMOMETER
PILTEH
HOLDER
(MPiNGER TRAIN OPTIONAL. MAV HE REPLACED
av AN EQUIVALENT CONDENSER
THERMOMETER
CHECK
VALVE
I '
12 iiir sr 1
'
:z
—
1
rii t-
r
i
i
i
i
L-Xx?-_V— J
y \
Mr,rns !CE OAT"
ORIFICE
THERMOMETERS
DRV GAS MEIER
PUMP
Metering
VACUUM
LINE
Figure A-1. EPA 5 SAMPLING TRAIN
-------
FILTER
STACK T.C.
GAS COOLER
5ENTRAUZEO TEMPEBATUREI
AND PHESSUne 3EADQUT
CONTROL MODULE
10 CFM VACUUM PUMP
SOURCE ASSESSMENT SAMPLING SCHEMATIC
HOT QA3
FROM OV6N
LIQUID PASSAGE
GAS PASSAGE
GAS COOLER
x 3-WAY SOLONOIO VALVE
•-V^^XJ"~J}C__J—- TO COOLING BATH
FROM COOLING BATH
rrL
Lgr~f
XAD-! CARTRIDGE
CONOEN3ATE _J^.---———
REGEHVOIH r—^,\ L_>—
COOL
COOLING FLUID
RESERVOIR
| [-• IMMERSION HEATER
—-LICUIO PUMP
TEMPERATURE
CONTHOLLER
c
XAD-2 SORBENT TRAP MODULE
Figure A-2. COLLECTION OF ADSORBABLES WITH SASS TRAIN
-117-
-------
A probe assembly consists of the probe itself, the
outer jacket, and selected elements for measuring stream con-
ditions. Examples for right-angle and head-on approaches to
the stream are shown in Figure A-3.
Sample Removal -
To provide access to points within the stream, the
probe assembly must protrude through the duct or pipe wall either
"head on", as shown in Figure A-4, or at a right angle to it.
To allow traversing of points within the cross-sectional of the
gas line, the probe must be mobile during operation, and more
than one sampling port should be available. Sample port loca-
tions are discussed in detail in Section 4.0 of this report.
Permanently mounted probes may be used if sampling at
several points is required. For streams containing high levels
of particulates or condensable tars, the use of permanently
mounted probes is not advisable due to potential probe plugging
problems. Removable probes should be used to allow cleaning
or replacement.
Probe Interface -
For nontoxic gas streams at pressures close to atmos-
pheric, the design of the stream-environment interface is not
particularly important. A loose fitting cap, or stuffing around
the probe assembly which allows it to move freely, will suffice.
The interface becomes critical, however, for toxic or pressurized
streams demanding a leak-tight seal. A lubricated packing gland
similar to valve-stem packing is the most generally applicable
option. Figure A-5 illustrates such an interface. The packing
gland is mounted on a fully opened gate valve or ball valve of
adequate internal diameter (3-inch is generally adequate) to
allow insertion of the probe assembly. The valve provides clo-
sure of the sampling port when not in use.
Gas Transport -
The probe's primary function is collecting and trans-
porting the sample without contamination or alteration. The
nozzle and sample-transport tube must be constructed of inert
-118-
-------
PROBE JACKET
\
GOOSE-NECK
NOZZLE
SAMPLE TRANSPORT
TUBE
PROBE DESIGN FOR "RIGHT ANGLE" SAMPLING
\
PITOT-STATIC TUBE
^THERMOCOUPLE
^STRAIGHT NOZZLE
,PROBE JACKET
\
SAMPLE TRANSPORT
TUBE
PROSE DESIGN FOR 'HEAD-ON' SAMPLING
Figure A-3. SAMPLING PROBE ASSEMBLIES
-119-
-------
N>
O
Sampling
/Port
^
7
^ — v^ _
L 2d j -*FH i
I /I
\ 1 (
Probe
Gas
X Flow
Figure A-4. SAMPLE PORT ARRANGEMENT FOR "HEAD-ON" PROBES SAMPLING
-------
MOUNTING FLANGE
PACKING CAP
TENSION ADJUSTMENT
BOLTS
GLAND BODY
Figure A-5. PACKING GLAND ASSEMBLY CUT-AWAY DIAGRAM
-121-
-------
material to avoid contamination or reaction with the tube walls.
The EPA Method 5 train mentioned earlier accomplishes this with
a pyrex glass probe liner. This serves very well for applica-
tions where temperature and pressure limitations of the glass
are not exceeded.
Table A-l summarizes some of the possible materials
for transport-tube construction and their limitations. Extreme
conditions (temperatures, pressures, or corrosive gases) require
expert design and material selection to avoid contamination from
scaling and to ensure safe operation. In these cases, the equip-
ment must meet the same piping code specifications for construc-
tion as the appropriate process piping. When materials are
selected for corrosive applications, compromises may be made for
short-term use, as long as potential sample contamination is
kept at a minimum. The sample tube may need heat tracing to
avoid condensation of gas stream components especially when
sampling the tar-laden gas streams found in certain coal gasifi-
cation facilities.
Table A-l. MATERIALS OF CONSTRUCTION FOR SAMPLE TRANSPORT TUBES
Maximum Maximum
Pressure Temperature
Pyrex Glass 2 psig 770°K (930°F)
Quartz 2 psig 1370°K (2010°F)
Teflon 50 psig* 500°K (440°F)
Stainless Steel (304 or
316 Series) * 820°K (1020°F)
Inconel 600 Series * 1170°K (1650°F)
* Dependent on wall thickness and construction
The nozzle at the collecting end of the sample-
transport tube should be constructed of the same material as the
transport tube. This is not always a rigid requirement, since
the sampled gas is exposed to the nozzle for only a small frac-
tion of the transport time. Interchangeable nozzles of varying
internal diameters are necessary to meet the requirements of
-122-
-------
isokinetic sampling and remain within the flow-rate limitations
of the sampling train. Isokinetic sampling (when the sample flow
velocity equals that of the stream) is discussed in the section
on particulate collection. To minimize turbulence, the nozzle
edges must be tapered as sharply as possible.
Sample Measurement -
The final function of the probe assembly is to measure
the stream's temperature, pressure, and flow rate. The designs
shown in Figure A-3 include thermocouples and pitot tubes for
doing this. Thermocouples with sheathing material selected for
stream conditions are recommended for their increased durability
and reliability. Two types of pitots are in common use, S-type
(reversed) and pitot-static. The function of each is to deter-
mine the velocity of the gas by measuring the pressure differen-
tial between the velocity head and' the stream static head (S-type
differential includes a negative pressure effect produced by the
eddy effect on the downstream side). For each type, the follow-
ing equation relates the differential pressure to the gas
velocity:
V = 85.48 C
s p
T (P -P )
sv p w
P M
s
(A-l)
where
V =
P =
w
P =
P -P
p w
f~i _
gas velocity (feet per second),
measured static pressure (inches of water),
measured total pressure (inches of water),
measured differential pressure (inches of water),
pitot coefficient (1.00 for pitot-static,
^0.85 for S-type; but it should be determined
by calibration),
T = stream temperature (°R),
-123-
-------
P = absolute stream pressure (inches of Hg), and
s
M = gas molecular weight (wet basis).
The static pressure of the gas stream is measured on the static
leg of the pitot-static or, if an S-type pitot is used, by
rotating the probe so that the pitot is oriented at a right angle
to the flow and measuring the pressure on either leg, with the
other leg disconnected.
A-l.1.2 Particulate Collection Unit
Particulate matter entrained in a gas stream is col-
lected for one or more of four possible objectives. Aside from
the obvious one of providing particulate-free gas for the vapor-
collection unit, three common functions for providing informa-
tion about the nature of the particulate matter in a gas stream
are:
measuring particulate loading in the stream,
determining particulate size distribution, and
determining particulate composition.
For all three functions, compositing and isokinetic sampling
techniques are required.
Compositing -
Particulate material is not necessarily distributed
evenly throughout the gas stream. Such maldistribution requires
traversing of the sample points within the sampling plane, as
discussed in Section 4.0, to collect a representative sample.
The traversing may consist of individual samples (or determina-
tions) for each point, or else a composite sample created by
particulate collection at each point for an equal time period.
-124-
-------
Isokinetic Sampling -
Isokinetic sampling requires that the sampled gas
stream enter the probe nozzle at the same velocity as the gas
stream velocity. Deviation from this requirement will result in
a size separation of particulates at the nozzle tip due to the
momentum of the larger particles. The gas flow patterns for
isokinetic rates and deviations from those rates are illustrated
in exaggerated fashion in Figure A-6. Small particulates will
follow the gas flow patterns shown while larger particulates
will follow a straight path relatively independent of gas flow
patterns. These two extremes of behavior will be exhibited over
the entire size distribution of the particulates in varying
degrees, depending on the gas velocities involved. Higher gas-
stream velocities will increase the tendency of smaller parti-
culates to follow a straight path, while lower velocities will
reduce this tendency.
The obvious result of nonisokinetic sampling will be
that measured size distributions will be biased: toward smaller
size ranges during super-isokinetic sampling and toward larger
size ranges during sub-isokinetic. Not so obvious is the effect
on determinations of particulate loading. Super-isokinetic
sampling will result in lighter-than-actual loadings, because
some of the particulates in the sampled gas will bypass the
nozzle. Sub-isokinetic sampling will result in heavier-than-
actual loadings, because particulates not already in the gas
sampled will enter the nozzle.
Even less obvious is the effect of nonisokinetic
sampling on particulate composition. The composition of parti-
culates is rarely independent of particle size, so a shift in
the representation of size distribution is likely to result in
a compositionally unrepresentative sample. Deviations of the
sampling rate within 10 percent on either side of isokinetic
will not significantly affect the results.
The Method 5 train (Figure A-l) collects entrained
particulates with the passage of the gas sampled from the probe
to a filter mounted in an oven maintained at 380°K (225°F).
The EPA specification of a collection temperature in this pro-
cedure effectively defines what material is considered to be
particulate matter. In this case, materials which condense
above 380°K (225°F) are included and those which remain vaporous
at 380°K (225°F) are excluded.
-125-
-------
GAS FLOW
GAS FLOW
GAS FLOW
SUPERHSOKINET1C
ISOKINET1C
SUB-ISOKINET1C
Figure A-6.
NOZZLE GAS FLOW PATTERNS ILLUSTRATING ISOKINETIC
SAMPLING
-126-
-------
The selection of the collection temperature must,
therefore, be made in light of test objectives (e.g., regulatory
requirements require collection at a specified temperature) and
stream composition. If the primary test objective is determin-
ing the actual stream particulate content, then collection at
stream conditions is recommended. One approach which inherently
requires collection at stream conditions includes mounting the
particulate collection device on the end of the probe extended
physically into the gas stream. This has the added advantage of
avoiding losses of material to the transport-tube walls.
If collection outside the stream is chosen, the sample-
transport tube must be designed to minimize losses. This is
accomplished primarily by providing smooth flow contours that
avoid protrusions or sharp directional changes in flow and by
heating the transport tube to avoid condensation. Particulates
entrained in pressurized gas streams should be collected at
stream pressure to avoid losses of material during passage
through a pressure-reduction step. Since some probe losses are
inevitable, the material must be recovered by washing the trans-
port tube with a suitable solvent (Method 5 specifies acetone)
following completion of any collection.
The specific collecting devices selected for the
particulate-collection unit depend upon the techniques found
effective for the primary purpose of the particulate collection -
whether for loading, size distribution, or composition.
Particulate Loading -
The common basis for all particulate-loading measure-
ments is the particulate weight per unit volume of gas. The
determination of particulate loading in a gas stream involves
collecting all particulate material from a portion of the stream,
determining the weight of collected material, and dividing the
particulate weight by the. total volume of gas sampled.
The particulace-loading procedures differ only in
their collection technique. EPA's Method 5 specifies an all-
glass system with a glass-fiber filter and a glass cyclone^(for
heavy loadings), both mounted in an oven maintained at 380°K
(225°F). The ASME Power Test Code specifies an alundum thimble-
type filter located in the main gas stream (Ref. A-2). A modifi-
cation of this ASME procedure uses a 47 mm glass-fiber filter,
also in the gas stream. The particulate collection technique
-127-
-------
for the SASS train is a series of three cyclones followed by
a glass fiber filter.
Modifications of these procedures for special appli-
cations, or for accommodating gas stream conditions, are per-
fectly valid if designed with the restrictions of the above
discussion taken into consideration. For example, a particulate
collection unit designed to determine the collection efficiency
of a high-temperature cyclone operated under pressure, might
include these four elements: a probe assembly with packing
gland and gate valve at the interface, a sample-transport tube
constructed of Type 316 stainless steel and heat traced to main-
tain process temperature, a filter holder containing an alundum
filter maintained at process temperature and pressure, and a
pressure-reducing and flow control valve at the filter exit.
This system would provide a measure of the actual stream parti-
culate loading, excluding any material not condensable at process
temperature.
Special Considerations -
There are certain types of streams that are unique to
coal conversion processes, and therefore, require special
sampling techniques. An example of these types of streams
is hot product gas (810-1480°K, 100-2200°F) containing
high concentrations of entrained particulates, tars and oils.
To obtain a representative gas sample from a "hot" stream having
high levels of entrained particulates, tars and oils, a sample
pretreatment train must be used. This train should remove the
particulates, tars and oils in a manner such that the con-
centrations of the gaseous species are not changed. An example
of such a pretreatment train consists of an in-line filter, a
knock-out pot, an inert filter, a permeation drier, a pump,
and a flowmeter. Bags or sample bombs can then be used to
collect the gases for analysis. A schematic of this train is
shown in Figure A-7.
Particule Size Distribution -
The most commonly used sampling device for the deter-
mination of particulate size distribution is the cascade
impactor, shown schematically in Figure A-8. Several vendors
manufacture devices which, although different in design, operate
-128-
-------
Gas In
I
1-J
t-0
Alundum Filter Holder
Used at the Cyclone (OPTIONAL)
Stainless Steel Probe
Teflon
Filter
Rotometer/
Flow Controller
Teflon Bag
(Total Hydrocarbon
Teflon Bag (NO )
Perma Pure
L
U
Drier
E=
U
4
=J
rn /
ol
^_ Glass Bomb
(Sulfur Species)
— ». Scutchpak Bag
(Fixed Gases)
^ Miran IA Analyzer
(HCN. NH,)
Humid
Air
Stainless Steel
Water Trap Used
at the Separator
Vent (OPTIONAL)
I Teflon-Lined
Dry Vacuum Sample
Alr Pump
Figure A-7. GRAB SAMPLE COLLECTION AND PREPARATION SYSTEM
-------
Plates
Gas Flow
Figure A-8. CASCADE IMPACTOR
-130-
-------
on the same basic principle. The sample gas stream passes
through a series of stages, each of which accelerates the
stream to a higher velocity. The increased velocities impart
increased momentums to the entrained participates, which results
in successively smaller particulates being collected by impaction
on the collection surface.
Some designs make provision for the use of removable
substrates of metal foil, polycarbonate, or glass filter media
for the collection surface. Others simply use a clean metal
surface^or plate, which may be coated with a film of grease to
reduce "bounce-off" and reentrainment. For each stage, the
sizing specifications are dependent on sampling rate, temperature,
particle density and gas density according to relationships
supplied by the device's manufacturer. A cascade impactor sizes
not as a function of physical size but as a function of effec-
tive aerodynamic size, which is dependent on physical size,
density, and shape. Due to material losses within this device,
the sum of the material collected on the individual stages can-
not be used to calculate an accurate measure of the total
particulate loading.
Particulate Composition -
The objective of particulate collection when deter-
mining compositon is to provide the analyst with material
representing the composition of the particulates entrained in
the gas stream. Three alternative devices deserve considera-
tion, depending on the problem at hand: filters, wet electro-
static precipitators (WEP), and cyclones.
Filtration is the most generally applicable technique
for this purpose. The filter medium must be selected to con-
tribute the minimum background for the parameters of interest
and to be unreactive with the gas stream or the collected
particulate sample. To minimize the influence of the filter on
particular analytical parameters, pretreatment of the filter
substrates is often required. Gaseous components of the stream
samples may react with, or sorb on, the filter medium to^produce
interferences in subsequent analyses. For example, studies have
shown that sorption of sulfur dioxide on filter media, followed
by oxidation, is a significant interference in determining sul-
fates Many filter configurations are limited in sampling
capacity The alundum thimble filter assembly specified by
ASME can provide increased sample quantities over other arrange-
ments available.
-131-
-------
Using the wet electrostatic precipitator (WEP) system,
shown in Figure A-9, can avoid many drawbacks present in filtra-
tion collection. This system is particularly useful in collect-
ing material for elemental analysis. In entering the WEP, the
sample gas bubbles through the electrolyte reservoir. The
electrolyte is a 5% solution of nitric acid. The gas then
passes through the WEP body, where electrostatic collection is
induced by a d-c potential of 12 kv between the suspended
platinum electrode and the electrolyte-wetted wall. Collected
material is washed from the wall by the electrolyte, which is
circulated by a peristaltic pump. This system overcomes the
sample quantity and interference limitations of filtration
collection; however, it simultaneously collects both particulates
and vapor-phase elements, which may be a disadvantage in some
situations. The WEP is almost as efficient as filtration col-
lection, having at least 99% of the latter's collection
efficiency.
For determination of compositon by size fractions,
cyclones may be selected for collection. Specially designed
sets of cyclones provide controlled fractionation. Each set
consists of a series of cyclones designed with progressively
smaller effective cutoff diameters so that the material collected
in each provides a sample sized between the effective cutoff of
the collection cyclone and of the preceding one. The fine
particulates escaping the last cyclone may be collected by either
filtration or the WEP method. The High Volume Source Assess-
ment Sampling System proposed by EPA for general environmental
assessment includes a set of cyclones with effective cutoffs
of 14y, 3y , and ly. To ensure the collection of adequate
material in each size range, the approximate particle-size
distribution in the gas stream should be known so that a set
of cyclones with appropriate cut-offs may be used. The actual
size distribution of each collected fraction should be deter-
mined by microscopy following collection, since many effects -
such as particle agglomeration, wall losses, and electrostatic
collection - may cause non-ideal operation of the cyclones. The
mass of material collected in the cyclones and its distribution
among the various size fractions should be considered mainly
for composition; it is only a rough indication of particulate
loading or of particle size distribution.
A-1.1.3 Vapor-Collection Unit
The particulate-collection section provides a particu-
late-free gas stream which is, generally, suitable for the deter-
mination of vaporous components and fixed gases. Two precautions
-132-
-------
SAMPLE
OUTLET
PERISTALTIC
PUMP
HIGH VOLTAGE
POWER SUPPLY
PLATINUM
ELECTRODE
WEP BODY
SAMPLE
INLET
CIRCULATING
ELECTROLYTE
RESERVOIR
Figure A-9. WET ELECTROSTATIC PRECIPITATOR
-133-
-------
apply, however, because maldistribution can occur from stream
non-uniformity and from temperature differentials during
collection.
Traversing with the probe is recommended when gaseous
components may not be uniformly distributed in the gas stream.
It has been shown that either laminar or turbulent flow condi-
tions can create this problem.
Controlling the temperature of the particulate-
collection unit may be important when this unit is the source
of the sample gas. Some vaporous components of interest may
condense or sorb on the particulates if the temperature is
lower than that of the gas stream.
There are four general techniques for the collection
of vapor components for analysis: sorption in liquids, sorp-
tion on solids, condensation, and grab sampling.
Sorption in Liquids -
Impingers or bubblers are used to collect and concen-
trate vapor-phase components of the gas stream by dissolution,
reaction, or both. The sampled gas is passed through a specific
reagent solution. Since the solubility of gases and vapors
generally increases with decreasing temperature, this collection
solution should be maintained at 273°K (32°F) in an ice bath.
The collection efficiency of liquid sorption tech-
niques is dependent on the driving force making the component
enter into solution and the degree of contact between gas and
liquid. The driving force is dependent on the temperature, the
mass-transfer rate across the gas-liquid boundary layer, and
the vapor pressure of the component or its reaction products.
The degree of contact between gas and liquid is dependent on
the impinger's or bubbler's design and gas flow rate (residence
time in solution).
Commonly used liquid impinger devices are illustrated
in Figure A-10. These are sized to accommodate various ranges
of solution volume and gas flow rate. Good practice dictates
that a minimum of two impingers be used in series to allow the
-134-
-------
SMITH-
GREENBURG
MODIFIED
SMITH-
GREENBURG
FRITTED
BUBBLER
Figure A-10. VAPOR COLLECTION DEVICES
-135-
-------
collection efficiency to be calculated from separate analyses
of the solutions.
Table A-2 lists several vapor or gaseous components
of probable interest and sorption solutions for their collec-
tion. Analytical techniques for these components or their
reaction products are presented in Section 5.0. The solutions
presented are not intended to include all possible options but
one example for each component. The selection of a sorption
solution and the analytical technique should take into consider-
ation the gas stream compositon, the concentration range of the
component of interest, the detection limit required by the pro-
gram objectives, and the potential analytical interferences.
Table A-2. SORPTION SOLUTIONS FOR VAPOR COMPONENTS*
Component
NH3
HCN
H2S
COS
CS2
S02
S03
NO
X
Trace Elements
Sorption
Solution
5% H2S04
10% NaOH
2% Zn(C2H302) 2
7.5% CaCl2/l% NH^OH
Alcoholic KOH
6% H202
80% Isopropanol
5% HaSO^/HzOz
0.2M (NH1|)2 S208/
Reference
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
A-3
A-4
A-5
A-4
A-4
A- 6
A- 7
A-8
0.02M AgN03
* Grab sample drawn into evacuated glass vessel containing
sorption solution.
Sorption on Solids -
Collection of components by sorption on solids,
followed by their recovery through thermal desorption or solvent
extraction is a rapidly developing area of sampling techniques.
-136-
-------
These techniques provide for the simultaneous collection and
concentration of components, analogous to the liquid sorption
techniques discussed above. The collection and determination
of organic components using solid substrates - particularly
porous polymer resins (as used in gas chromatography), activated
charcoal, silica gel, XAD-2 and Tennax - is the primary applica-
tion of_solid sorption techniques. Most applications are pre-
sently in the earily stages of development. Therefore, a defini-
tive characterization of their use for the wide range of organic
components of interest is needed. Before using a solid sorbent
for a particular organic compound, the following factors must
be considered and quantified:
collection efficiency as a function of flow
rate and temperature,
breakthrough capacity,
component stability, and
recovery efficiency.
These factors must be determined empirically for each component,
or combination of components, before a solid sorption technique
can be considered reliable for quantitative determinations. Col-
lection efficiencies can often be enhanced by collection at low
temperature. If complete characterization of a technique is not
feasible, the use of two or more collection devices in series can
give an indication of collection efficiency and breakthrough
capacity. Component stability can be enhanced by sample storage
at low temperatures under an inert gas. Results not generated
under these circumstances must be handled with considerable
skepticism and can be considered semiquantitative at best.
Solid sorption techniques have received limited appli-
cation for inorganic gaseous components. Silica gel has been
used as a support for direct reading colorimetric tubes (Ref.
A-9). These devices have been developed for a variety of com-
ponents and are useful for semiquantitative surveys. Silica gel
has long been accepted as a standard approach to the determina-
tion of moisture content of gas streams (Ref. A-10). The deter-
mination of mercury vapor by collection through amalgamation of
gold or silver, followed by thermal desorption, has been exten-
sively documented (A-ll).
-137-
-------
Condensation -
The viability of condensation, including cryogenic
techniques, for a particular component in a gas stream can be
roughly evaluated by comparison of the component's vapor pres-
sure (at the condensation temperature) with its partial pres-
sure (directly related to the concentration on a volume/volume
basis) in the gas stream. The vapor pressure must be at least
an order of magnitude below the partial pressure in the gas
stream for a condensation technique to have any potential for
application. In general, condensation techniques are more use-
ful for organic components than for inorganic ones, because most
inorganic gases require extremely low temperatures. Condensation
or freezing at cryogenic temperatures (usually with liquid
nitrogen) is very cumbersome and requires that the sample be
analyzed immediately or stored at the collection temperature to
avoid losses. Condensation and freezing of gas-stream moisture
can cause severe plugging and dilution problems when cryogenic
techniques are used. Determination of moisture content by a
combination of condensation and silica-gel sorption is a widely
practiced technique (Ref. A-10). This is an example of an often
used combination of condensation backed up by solid sorption.
Condensation is often used as a clean-up procedure for
removing condensables from a sample gas stream to minimize
interferences when determining a noncondensable component. This
is particularly useful for gas streams containing high levels of
condensable organics. Although this approach is often useful,
extreme caution must be exercised to avoid loss of the component
by sorption in the condensed materials. Here, analysis of the
condensate for the component of interest may be used as a check
or correction procedure.-
Grab Sampling -
The use of rigid or nonrigid containers for the col-
lection of gas samples in the gas phase is termed grab sampling.
There are two major sources of error which restrict the use of
this technique: adsorption on (or reduction by) the container
walls, and reaction of gas components with each other.
These effects can be minimized through the use of
inert materials such as glass or teflon in container construc-
tion and in the preconditioning of the container prior to
-138-
-------
sampling. Analysis of grab samples should be conducted as soon
after collection as possible, and sample stability as a function
of time should be determined. The most difficult source of error
to avoid is the loss of components by sorption in condensate
(usually water vapor) produced during cooling of the sample.
Maintaining the container at a temperature high enough to prevent
condensation is not usually practical, and reheating rarely is
effective. Removal of moisture by sorption on silica gel or
by a permeation dryer prior to the container may be a solution
if the component of interest is not sorbed on the desiccant or
passed through the permeation dryer tubes.
As this discussion indicates, severe problems may
arise with the reliability of grab sampling techniques and they
can only be used with confidence for nonreactive fixed gases or
following extensive verification. The sorption technique for
NOX listed in Table A-2 illustrates a modification of grab
sampling to include a sorption solution in the collection vessel.
A-1.1.4 Metering Unit
With the exception of grab sampling (which often
requires a vessel of known volume), all of the techniques de-
scribed for particulate and vapor collection require measuring
the sampled gas volume before calculating concentration in
the gas stream from the analytical results. In addition, the
sample-gas flow rate must be known in order to maintain iso-
kinetic sampling rates.
The most commonly employed metering technique to pro-
vide a cummulative volumetric measurement is a dry gas meter,
the type used for household natural-gas metering. These meters
are available in a wide range of maximum flow-rate capacities
and are useful down to a minimum flow rate of 0.1 cfm. Dry gas
meters must be calibrated frequently against a standardized
meter or a wet-test meter and then handled carefully to avoid
inverting or jolting. The sample gas must be conditioned prior
to metering by the drying and removing of corrosive components.
To convert the meter readings to standard conditions, the inlet
and exit meter temperatures and meter pressures must be recorded
during sample collection. The use of wet-test meters in field
situations is not recommended.
An alternative approach is to measure flow rate and
total sampling time to provide the total gas volume. Rotameters
-139- .
-------
and orifice meters are well suited for this function. A com-
prehensive discussion of these meters is available in Perry's
Chemical Engineers' Handbook; equations are available for
correcting rotameters and orifice meters for variations in gas
temperature, pressure, density, and viscosity (Ref. A-12).
A-1.2 Continuous Monitoring
The results of an extensive sampling and analytical
program may call for the continuous monitoring of some species
in a gas stream. The basis for deciding which stream and
analytical parameter to monitor must be a part of the overall
test plan. The decision could be based on many factors,
including the following:
compliance with governmental regulations,
reduction of undesirable chemical effects
(including toxicity),
process control, and
research and development investigation.
Due to the repetitious and long-term nature of continuous
monitoring, automated or instrumented systems are inherently
more suitable than the manual techniques previously described.
Physical conditions - including temperature, pressure, flow
rate, and density - are conveniently monitored using commer-
cially available instrumentation, such as normally is installed
for process control. While not directly a part of the test
plan, such instruments supply process information related to
the testing.
The analytical parameters frequently of interest for
continuous monitoring are discussed in the remainder of this
section. While not intended to be all-inclusive, this discus-
sion is a guide providing examples of continuous monitoring
applications.
Continuous monitoring equipment should be selected
only after all costs for initial purchase, installation, and
-140-
-------
maintenance are considered. An inexpensive instrument with fre
quent repair and calibration problems can be more expensive fnd
less reliable than manual methods. The consequence? of ins t?u-
of the%^ T^ °?eration an<* for the successful execu
of the test plan are important considerations.
Planning for automated monitoring should take into
account three areas of choice: sampling approaches, analytical
determinations, and instrument calibration methods.
A- 1 . 2 . 1 Sampling Approaches
There are two basic approaches for instrumentally
monitoring a gas stream: extracting the sample from the stream
for analysis and analyzing the stream in situ.
Extraction Systems -
The extractive approach is the oldest one with more
information available on applications. After a probe is
installed in the gas stream, the gas is drawn through the probe
and then through a sample conditioning system. This system
removes particulates and adjusts the sample to the appropriate
temperature, pressure, and moisture content. The sample then
enters the analyzer. Detailed attention to the sample condi-
tioning system is required for maintaining sample integrity.
There are many advantages to extractive systems. They
frequently are available at a lower cost and are more flexible
in their application and location, because the analyzers are no
longer subject to the temperature, pressure, and contamination
problems associated with in-situ analysis. These problems are
eliminated by the conditioning unit in an extraction system.
Such a unit protects the analytical instruments by removing
impurities from the stream and adjusting stream conditions to
conditions for which the instrument has been calibrated. An
extractive system can also be provided with a more representa-
tive sample by using a multiple-orifice probe or a multiple
acquisition system. A combined sample-acquisition and condi-
tioning system can supply a sample stream to several analyzers;
also a single analyzer can handle samples from several streams.
-141-
-------
One obvious problem with extractive systems is that if
the species of interest is very reactive, the extractive and
conditioning procedure can alter sample composition from actual
stream composition. In such cases an in-situ system would be
preferred.
In Situ Systems -
In situ continuous monitors of gas streams may in
general be classified as optical or non-optical systems. In situ
optical systems provide an average value of the light path over
one diameter of the stack (possibly two diameters if the optical
path is folded by a reflector). In situ analyzers have a limi-
tation imposed by the physical parameters of the stream analyzed.
They are, however, not subject to the limitations and possible
loss of sample integrity present in a conditioning system. This
makes them ideal for the more reactive species. The installa-
tion of an in situ system, however, is often more costly and
requires more frequent maintenance. An in situ optical system
requires, for example, periodic cleaning of the optical windows.
Some non-optical techniques are available for in situ
monitoring. Also, some of the recently developed analyzers
employ features of both in situ and extractive techniques. In
one arrangement, the analyzer is contained in a porous housing
extended into the gas stream. Because it is located in~situ, the
analyzed gas needs no conditioning to maintain stream condi-
tions. Yet, by maintaining the analyzer slightly evacuated, the
inflowing sample is filtered through the porous housing.
Calibration is performed by pressurizing the porous container
with known gas mixtures. Thus, the benefits of in situ condi-
tioning and extractive filtration are combined.
A new in situ technique for optically analyzing stack
gas just after: it enters the atmosphere is based on laser optics
that allow remote detection. While not at present commercially
available, this technique shows great promise for ease of
installation and flexibility in location.
A-1.2.2 Analytical Determinations
Sample analysis systems can use either the extractive
or in situ techniques. These techniques can best be reviewed in
-142-
-------
terms of the analytical parameters to be monitored. In this
text these parameters are considered under three categories:
particulates, criteria pollutants in general, and other gaseous
components.
Particulates -
The instrumental monitoring of particulates is normally
limited to the determination of particulate loading (including
particulate size distribution) - the concentration of mass
within a given gas volume (as discussed in Section A-l.1.2). The
guidelines for selecting manual sampling points and sample
handling apply also to continuous monitoring. Particulate con-
centration is normally monitored by measuring collected part-
iculate matter from a known volume of gas.
Devices for directly determining particulate concen-
tration, including particulate concentration as a function of
particle size, are currently marketed. These usually operate in
a cyclic or batch mode and are complicated and costly. To main-
tain an isokinetic sampling rate, the flow rate must be automat-
ically controlled, based on the output of a velocity-measuring
device.
A more commonly used approach is the indirect deter-
mination of particulate concentration with an optical method.
The optical density of the gas stream is monitored by passing
the gas stream through a light beam. The relative particulate
concentration is indicated by changes in optical density. The
relationship between this optical density and particulate con-
centration is determined empirically, with the absolute part-
iculate concentration estimated from the established relationship
The particulate size distribution can be monitored by
measuring the degree of light scattering caused by particulates
in the gas stream. Unlike particulate concentration measure-
ment, particulate size distribution is normally monitored by
using an extractive technique outside the confines of the gas
stream duct. Also, a large amount of clean, particulate-free
air is normally required for sample dilution. These factors
require extra attention to details of the sample acquisition
and handling systems.
-143-
-------
Criteria Pollutants -
Effluents from a coal gasification plant, especially
the combustion products, will be monitored for "criteria
pollutants" to meet requirements of government regulations.
Besides particulates, these pollutants include two frequently
monitored gases - sulfur oxides and nitrogen oxides. If the
monitoring of these emissions is a direct result of government
regulation, care should be taken to ensure that the installed
systems are acceptable to the regulatory agency, in addition to
providing the data required for the test program. The EPA has
promulgated specifications and standards for instrumentation in
over two dozen industries or processing areas (Ref. A-13).
Both in situ and extractive techniques are available
to monitor these criteria species. Currently accepted instru-
mentation employs optical, wet chemical, electrochemical, or
chemiluminescent methods for oxides of sulfur and nitrogen; the
flame method is available for detecting oxides of sulfur. The
remote (laser) in situ method mentioned earlier looks most pro-
mising for the measurement of S0x and NOX.
Other Gaseous Components -
Going beyond these criteria pollutants are the major
gas and vapor components of the process and effluent gas streams
Important ones are illustrated in Table A-3. In this table,
many gases are included in more than one category. There are
continuous analytical methods for analyzing these species.
Table A-3. GAS AND VAPOR COMPONENTS
Fixed
Gases
H2
CO
N2
02
Ar
NH3
CH,,
Acid
Gases
H2S
CO 2
S0x
NO
X
HF
aci
HCN
Sulfur
Species
COS
S0x
CS2
H2S
RSH
RSR
Non-Condensable
Hydrocarbons
ctu
C2H5
C2H6
C3 to C6
Condensables
H20
SO 3
>CS Hydro-
carbons
-144-
-------
Although continuous-reading instruments are marketed
for measuring moisture content, molecular weight, gas density
and total heat content, this discussion focuses on the two major
approaches for the quantitative determination of such species-
chromatographic and optical.
Process gas chromatography (GC) is an extractive tech-
nique and is probably the most widely applicable method. Ail of
the gases listed can be analyzed with a process GC. In most
cases, multiple analyses can be performed.
A typical process GC installation consists of a sample
acquisition and conditioning system, a column for species separa-
tion and a detector. Each of these three components must be
tailored to the particular task performed. In some cases,
multiple columns and detectors can be used for one collecting
instrument. The sample handling and conditioning considerations
discussed earlier apply to GC analysis. Separation columns for
process GC are available in many types. While a discussion of
these is beyond the scope of this effort, detailed information
is available from GC instrument manufacturers.
The GC detectors most often used in process instru-
mentation include three types: thermal conductivity (TC), flame
ionization (FID) and flame photometric (FPD).
The TC detector will analyze all the species and is
most commonly used. Its major drawback is its lack of sensi-
tivity. Also, it cannot analyze high concentrations (several
percent) of one species and low concentrations (< 100 ppm) of
other species in the same sample. As a result, the TC is
usually used for the analysis of "major" species in a gas stream.
The FID has a very large sensitivity range, almost
four orders of magnitude. Flame ionization is also selective
for hydrocarbons, so is widely applied to the analysis of organic
species. Other advantages expand the usefulness of the FID. A
sample may be introduced to the detector without prior separa-
tion and still obtain a total hydrocarbon value. Other tech-
niques, such as the separation and catalytic hydrogenation of
CO to form methane, extend the applications.
-145-
-------
The FPD is selective for sulfur-containing species
and is a very sensitive detector. These two characteristics
make flame photometry very useful in the analyses of sulfur
species, although the separation techniques for these species
are often difficult. The major limitation of the FPD is its
relatively narrow range of linearity (usually less than two
orders of magnitude) with respect to concentration.
The other main continuous approach, optical analysis,
is used in numerous commercially available instruments. The
ability of many species to absorb a specific wavelength of light
is the principle utilized. These analyzers detect light from
the ultraviolet, visible, and infrared regions, employing either
in situ or extractive techniques.
In nondispersive infrared (NDIR) analysis, the extra-
tive mode is more common. This technique can be used for most of
the species listed in Table A-3 (except for those having diatomic
molecular structure), if interfering species can be masked.
NDIR instruments are also unique in being convertible at minimal
cost for monitoring an entirely different species - thus
increasing their flexibility.
A-l.2.3 Calibration Methods
Quality control for instrumental monitoring is pri-
marily a matter of frequent and careful calibration. These
requirements are often specified in detail by the manufacturer,
and sometimes by a governmental regulatory agency.
The basic approach is to check the analyses against
known samples. Precision is ensured by the consistency of
replication when a particular sample is retested repeatedly.
Accuracy is ensured by periodic comparison with a standard
having a known content. In addition, the instrumental results
may be checked at any time by comparing them against the results
of manual sampling and analysis.
A-1.3 Fugitive Gas Emissions
Fugitive gas emissions are those which are expelled
directly into the environment without passing through any
-146-
-------
transport device such as a pipe, duct, or stack. They may be
generated in enclosed areas such as buildings and transmitted to
the environment through structural openings and vents, or they
may result from open area sources such as leaking equipment, open
storage piles, and effluent disposal areas.
For environmental assessment testing, fugitive emis-
sions are assigned to one of two categories: specific source
emissions and site source emissions. Examples of specific
sources are leaking valves, coal piles, and grinding mill
building vent. Examples of site sources are the coal gasifica-
tion unit and the plant as a whole.
A-1.3.1 Sampling Point Selection
The selection of sampling points for fugitive emis-
sions are based on the source type and a good deal of subjective
judgment on the part of the sampling team. Sources must first
be identified and categorized according to type. They must be
significant enough to warrant sampling within test guidelines.
Specific sources may be sampled from downwind of the source to
determine atmospheric distribution or they may be sampled
directly by enclosing the source. Site sources are usually
sampled from both upwind and downwind locations to determine
area distributions of pollutants. Specific sampling point selec-
tion is highly site-specific and depends on source size and
location and homogeneity of emissions. These factors will have
to be determined by pre-test site surveys.
A-1.3.2 Sampling Techniques
The analytical parameters measured by fugitive emis-
sion sampling are similar to those discussed for manual gas
sampling (Section A-l.l). Basically, the parameters of inte-
rest include gaseous compounds, particulates, trace organic
species, and trace elements.
The basic types of sampling methods for collecting
area and source fugitive emissions are listed below.
-147-
-------
Area Emissions
High volume sampling
SASS train sampling
Grab sampling
Continuous sampling
Source Emissions
Source enclosure followed by a sample or
collection or pretreatment train similar
to those discussed in Section A-l.l
High Volume Sampling -
High volume sampling involves drawing large volumes of
air through a filter to trap particulates. A split stream from
the sampler is passed through an adsorbent canister or impingers
to trap organic or inorganic species. Grab samples may also be
taken from this stream. This technique is used for area
sampling. Figure A-11 shows a typical high volume sampling
assembly.
SASS Train Sampling -
SASS train sampling has been previously discussed and
is much like high volume sampling. The major difference is that
the SASS train will collect particulate matter fractions, organic
species, and gaseous species (by impingers).
Grab Sampling -
Grab samples are taken of the fugitive emissions for
gas component analysis. These are usually taken in three liter
evacuated vessels and are obtained either directly from the
atmosphere or from the high volume or SASS train sample streams.
The gas compounds analyzed would include those not collected by
solid sorbents or impingers.
-148-
-------
\
1/4 INCH
SWAGE LOK
BULKHEAD
COPPER
TUBING
WALL
Figure A-11.
TO 5 efm PUMP
EXPANDED VIEW OF CONNECTIONS OF XAD-2 CARTRIDGE
TO HIGH VOLUME SAMPLER
-149-
-------
Continuous Sampling -
Continuous sampling is used to monitor area fugitive
emissions. Currently continuous CO area monitors are being
recommended for coal gasification facilities.
Source Enclosure Sampling -
Source enclosure sampling is the least defined sampling
technique in the literature, but provides the most precise and
representative data. This method involves bagging, tenting, or
otherwise enclosing the emission source. Once the source is
enclosed, samples of the trapped gases are taken and analyzed.
If the rate of change in concentration of one or more of the
species being leaked into the enclosure is monitored, a measure
of the leak rate can be computed by knowing the enclosure volume
during the sampling period. The source may be enclosed using a
bagging or tenting technique. Mylar or teflon are a suggested
bag material because it is easily shaped, relatively inert, and
used for gas grab sample collection bags. Decisions on whether
to enclose a source to measure its fugitive emissions depends
upon the following:
source size (can the source be enclosed),
source temperature (bagging material may
be temperature limited),
expected composition of the fugitive emissions
(will the gas component react with the bag
material), and
source location (can the source be accessed for
enclosure).
Data which should be recorded by the sampling team
during fugitive emissions sampling include:
area data (temperature and air flow patterns),
sampling location and method,
-150-
-------
• emission source and characteristics
(temperature), and
• sampling equipment operational data.
A-2 LIQUID SAMPLING
-------
For pressurized single or multiphase liquid streams
which contain dissolved gases it must be resolved whether the
escaping gas needs analysis. Two sampling techniques may be
used. In one method, the sample is taken at atmospheric pres-
sure and the dissolved gases are allowed to escape; only the
liquid portion of the sample is then analyzed. In the other,
the sample is caught under process pressure in a suitable
sampling bomb. Later the pressure is reduced, and the gases
emitted are analyzed separately.
High-temperature streams may require the application
of cooling methods to reduce the sample temperature to a level
that is both below its boiling point and compatible with the
sampling techniques used. Molten solid streams must be sampled
by methods and equipment which allow for the fact that the
sample will solidify on cooling.
A- 2 .1 Sample Point. Selection
The preferred sampling points for liquid streams are
existing valves, either in-line or on a side stream. These
valves provide a ready source from the stream and should be used
when compatible with the objectives of the test program.
Other points of easy access are outflow orifices where the
liquid streams flow into ponds, tanks, or other open surfaces.
Open or noncontained streams may be sampled at any point com-
patible with accuracy requirements.
The major restriction in selecting sampling points is
stream homogeneity. To ensure a well-mixed sample, sampling
should be done just downstream from points of turbulence, such
as elbows or pump-discharge lines. It may be necessary to have
sampling valves installed at points where none exist.
A-2.2 Grab Sampling Methods
The selection of sampling method, the size of the
sample, the frequency of collection, and the method of preser-
vation must be based on the goals of the test program.
In selecting sampling methods for liquids the analyt-
ical techniques planned must also be considered. Analytical
-152-
-------
parameters for liquid slurry streams consist of four types:
physical characteristics,
inorganic species,
organic species, and
water quality parameters.
Table A-4 lists typical parameters for each of the
first three categories. Overlap with water quality parameters
is_indicated in the table. The factors involved in deciding
which parameters to analyze are defined by the test plan objec-
tives. The sampling methods selected should be scrutinized for
compatibility with the proposed analytical methods. In general,
sample size will be determined by the number of analyses to be
made and the specific requirements of each technique.
Samples may be taken at regular intervals over the
duration of the test and then either analyzed individually or
combined to give an averaged sample. If possible, the test
duration should be long enough to cover normal process variations
For some cyclic processes, it may be best to sample at each step
of the operation. In cases where sample requirements are less
stringent, a single sample may suffice.
Grab sampling techniques may be adapted for liquid
sampling according to test plan requirements. Depending on
analytical requirements and stream characteristics in any given
situation (as discussed below), either of two general methods
may be appropriate: tap sampling and dipper sampling.
A-2.2.1 Tap Sampling
Tap sampling is used to collect liquid samples from
enclosed pipes or storage tanks. Two tap systems are shown in
Figure A-12 The simplest, shown in the bottom illustration,
consists of a valve attached to the wall of the vessel or line
from which the sample is drawn. To sample nonhomogeneous streams
more accurately, the method shown in the top illustration of
Figure A-12 is required. In this system, a probe is inserted
-153-
-------
Table A-4. ANALYTICAL PARAMETERS FOR LIQUID AND SLURRY SAMPLES
Inorganic Species
Dissolved Oxygen*
Ammonia
Chloride
Fluoride
Sulfate
Sulfite
Sulfide
Nitrate
Nitrite
Phosphate
Cyanide
Carbonate
Iodide
Orthophosphate
Total Phosphate
Major Elements:
S Mg
Al P
Physical Characteristics
pH *
Solids Content (IDS and TSS)*
Water Hardness*
Specific Electrical Conductance*
Acidity*
Alkalinity*
Turbidity*
Temperature
Organic Species
Biological Oxygen Demand (BOD)*
Chemical Oxygen Demand (COD)*
Oil and Grease*
Phenol
Total Organic Carbon
Speciation •
Polynuclear Aromatics
Heterocyclics
Halogenated Organics
Aromatics
Other Hydrocarbons
Ca
Fe
Na
Trace
Sb
As
Be
B
Cl
Cd
K
Si
Ti
Elements '
Cr
Cu
F
Pb
Li
Mn
Hg
Mo
Ni
Se
Tl
U
V
Zn
* Water-quality parameters
-154-
-------
Proba
Packing
Gland
Valv»
Flow
Ball Valva
Sanpling Configuration using Probe through
Exiacing Valve for Liquid Streams
LINE
OR
TANK
WALL
7 MM
(1/4 IN.)
Assembly for Tap Sampling
Figure A-12. TAP SAMPLING TECHNIQUES
* Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-155-
-------
through a valve into the center of the stream. This method
minimizes the problems caused by reduced flow along the walls
of the pipe. It does, however, create design problems, since
the seals and valves must allow probe insertion without leaks
from the stream being sampled.
Some alternate sample tap configurations may be used_
for accurate testing when a representative sample can more easily
be obtained, such as with static samples of homogeneous liquids.
These provide representative samples, but are more difficult to
install (see Figure A-13).
Tap sampling lines should be as short as possible to
facilitate flushing. The samples are collected by inserting the
sample line, usually Teflon, into the sample bottle so that it
extends nearly to the bottom to minimize air entrainment in the
sample. The sample line and bottle should be flushed thoroughly
with sample material before a portion is retained. Specific
examples of tap sampling can be found in the references cited
in Table A-5.
Table A-5. LIQUID SAMPLING METHOD REFERENCES
Manual Reference No.
1973 Annual Book of ASTM Standards, Pt. 23. "Water; Ref. A-14
Atmospheric Analysis."
ASTM, "Standard Recommended Practice for Sampling Ref. A-15
Industrial Chemicals, " 1973.
EPA, Handbook for Monitoring Industrial Wastewater, 1973 Ref. A-16
EPA, IERL-RTP Procedures Manual, Level 1; Ref. A-17
Environmental Assessment, 1976.
EPA, Tentative Procedures for Sampling and Analysis Ref• A-18
of Coal Gasification Processes, 1975.
ERDA, Assessment, Selection and Development of Ref. A-19
Procedures for Determining the Environmental
Acceptability of Synthetic Fuel Plants Based
on Coal, Revised Report, A vols.
Radian Corp., Sampling Plan, Characterization Ref. A-20
of the Effluents from the C02 Acceptor Process,
1976.
-156-
-------
CLBOW
TIP IN
mot
FiRure A-13 VARIOUS SAMPLE TAP INSTALLATIONS
(Ref. A-15)*
* Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-157-
-------
The temperature and pressure of the stream being sam-
pled are not as critical in sampling liquid and slurry streams
as they are with gas sampling. In general, high- and low-
pressure liquid and slurry lines may be sampled by the standard
tap techniques described. Care should be taken that valves on
high-pressure lines are opened slowly to prevent injury to per-
sonnel. Sampling personnel should also be aware of the possibil-
ity of the sudden surge in the flow of sample systems due to the
release of plugs in the probe or tap. Plant management should
be consulted when any high-pressure lines are to be sampled, and
plant safety regulations followed.
Open tap sampling of high-pressure liquid streams con-
taining dissolved gases will result in a loss of dissolved gases
after the reduction in pressure. If samples containing dissolved
gases are desired, high-pressure bomb-sampling techniques must
be used. A typical sampling apparatus is shown in Figure A-14.
The bomb can either be evacuated before sampling or it may be
filled initially with an inert gas, such as nitrogen or helium.
When the latter method is used, the initial gas pressure must
be known in order to calculate the dilution of gases released
from the sample on depressurization.
The tap sampling of liquids flowing at subatmospheric
pressure requires pump assistance to remove the sample. Figure
A-15 shows the necessary equipment.
The tap sampling of liquid or slurry streams at high
temperature (i.e., above the liquid boiling point) requires
special procedures. The liquid sample must be cooled to a tem-
perature below its boiling point before its entry into the
sample container or it will flash vaporize. The sample is
cooled by passage through an air or water jacket system before
collection (see Figure A-16). Again, care must be taken to
thoroughly flush all parts of the system before material is
retained for analysis.
In most cases, the use of a cooling system will cause
some loss of sample integrity. A loss of material through
deposition on the cooled wall is unavoidable. In some instances,
plugging of the sample line will result. These problems can
be minimized if the sample is cooled just enough to handle, but
not below the temperature at which sample integrity is shown to
be lost.
-158-
-------
PROBE
PACKING
GLAND
FLOW
BALL VALVE
STAINLESS STEEL
SAMPLE BOMB
Figure A-14. TAP SAMPLING FOR HIGH PRESSURE STREAMS
-159-
-------
PROBE
: FLOWMETER
PUMP
Figure A-15. PUMP ASSISTED TAP SAMPLING OF SUBATMOSPHERIC LINES
-160-
-------
PROBE
VALVE
HIGH PRESSURE
PROBE
FLOWMETER
PUMP
SUBATMOSPHERIC PRESSURE
Figure A-16. SAMPLING METHODS FOR HIGH TEMPERATURE STREAMS
-161-
-------
A-2.2.2 Dipper and Thief Sampling
Dipper sampling is for collection at liquid surfaces,
whereas thief sampling collects material below the surface of a
contained liquid.
The dipper sampling method is useful for sampling open
ditches or sluices, storage tanks, and outflows from open pipe
ends. The dipper is made with a flared, Teflon-coated bowl and
a handle long enough to allow access to the areas being sampled.
Multiple samples, if necessary, may be collected to obtain a
sufficient sample volume. If possible, streams should be
sampled at points where they are stagnant or not flowing.
A variation on the dipper method that provides a much
more representative sample is collection where the stream flows
out the end of a roughly horizontal pipe or flume. An inert
container with sharp parallel sides is passed at an even rate
through the falling stream, obtaining a full cross section of
the stream (Figure A-17). The container should not be allowed
to overflow if the effects of stream stratification or a repre-
sentative proportion of solids in the liquid stream are desired.
If the sampling of a large tank or other stagnant
vessel is unavoidable, a thief sampling procedure should be
used. Thief sampling equipment is shown in Figures A-18 and
A-19. A sample thief (or bailer) should be used to obtain
samples from several depths in the tank (Figure A-20). Some
thiefs are available which will collect a sample at an even rate
while it is being pulled up from the bottom of the tank.
A-2.3 Continuous Sampling
Samples obtained from a liquid stream on a continuous
basis will be much more representative than those obtained on a
grab basis. If this higher level of accuracy is required, a
sample probe of a design similar to one of the three shown in
Figure A-21 should be used, unless the stream is known to be a
homogeneous liquid.
The sample can be withdrawn and transferred to a con-
tainer in different ways. The sample may be withdrawn periodically
at a uniform rate. This can be done in increments with a smaller
-162-
-------
Figure A-17. STREAM SAMPLING CUP
(Ref. A-15)*
*Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-163-
-------
ALTIRNATC RIG
I LITER (I QUART)
WEIGHTED BOTTLE SAMPLER
(CAN it FAINICITEO TO FIT »Wt
UK «LMS- ffOPPtHfO
Figure A-18. ASSEMBLY FOR BOTTLE SAMPLING*
* Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-164-
-------
CHAIN FOR
LOWERING
— 4 LUGS
15 J" (40.0 cm)
»L-3 l/2*(a39on) -\
CIA.
(a) Bomb-Typ« Sampling ThicJ
(*) Cort Thief. T4p-T>p.
Figure A-19. SAMPLING THIEFS *
* Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-165-
-------
HATCH
ST:.JUOWER TMIRO
OUT1_£T
"BOTTOM SAMPLE
Figure A-20. SAMPLING DEPTHS FOR THIEF SAMPLES
(Ref. A-15)*
* Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-166-
-------
»<*»«« ro >•«.*» a*
COM «
-------
time interval between each unit volume fraction of the sample,
thus providing a representative sample. The continuous sampling
of a slurry in the most representative manner requires isokinetic
sampling (discussed in Section A-l), with the velocity of the
stream equaling the velocity at which the sample enters the sam-
ple probe. This assures that.suspended solids are included in
their proper proportion.
During continuous sampling, it is also possible to
remove solids if desired by filtering or centrifuging, and to
add the required preservatives as the sample is collected.
A-3. SOLID SAMPLING
Solids commonly found in coal gasification facilities
are fuels, process additives, by-products, and waste products.
They can range in size from fine power to lumps, and they can
vary in consistency from dry solids to thick, non-flowing pastes.
To accomodate all possible situations, a variety of sampling
techniques are required.
The types of analyses performed on solids samples are
highly dependent on the material sampled. For solid piles or
streams, the nature of the analytical method rarely affects the
sampling method used, although the handling and preservation
methods are affected.
The sampling plan designer may need to arrange for
solid samples from either of two general kinds of solids source:
unit aggregations and process conveyors.
The unit aggregations are either large and stationary
or they are transport containers. Solids used or produced in
the gasification processes in large quantities are usually found
stored in large open piles or enclosed silos. Both storage and
transport units yield samples which may have already been com-
posited, but with differing compositing periods. The stratifi-
cation of solids in such static locations can make obtaining a
representative sample difficult.
Samples from process conveyors represent solid mater-
ials being used continually during defined periods and under
defined operating conditions. These streams, therefore, are
-168-
-------
the important ones for material K^i ^^^^ j -. j i
response tests. ^nenal balances and control-module
Sampling methods for solids use two general techniques,
grab sampling and grab-and-composite sampling. Grab sampling
is the general sampling technique used where low precision is
required Grab-and-composite sampling is the more precise tech-
nique. In most cases, the difference between the two is only a
matter of degree; the sample collecting methods are identical.
In the second, grab samples are collected periocally over the
duration of the test, then composited to form a single sample.
The following sections present more detailed sampling procedures
for the various sources of solid samples. References for
specific methods are cited in Table A-7. Although most of these
methods are for coal, they are readily adaptable to other solid
streams.
Table A-7. REFERENCES FOR SOLID SAMPLE COLLECTION AND HANDLING
Reference No. Comments
Ref. A-21 General considerations for the collection of samples
to measure trace components
Ref. A-22 ASTM Method D 2013-72: sample handling
Ref. A-22 ASTM Method D 2234-74: gross sample collection
Ref. A-15 ASTM Method E 300-73: sample handling and collection
A-3 .1 Sampling Methods for Storage Facilities and Transport
Containers
Solids encountered in coal gasification facilities are
usually stored in piles or contained in enclosed bins with rela-
tively large circumference and depth. The selection of sampling
techniques for these materials will depend upon the test require-
ments, the sample pile form, and its accessibility. There are
three basic methods: the shovel technique, the pipeborer tech-
nique, and the auger technique.
-169-
-------
The shovel technique is used to remove a sample from
the perimeter of a pile. Although this is the simplest solids
sampling technique, it may well be inadequate. Unless there is
reason to believe that the pile is homogenous, it may result in
a highly nonrepresentative sample.
Somewhat more representative samples of materials
stored in piles, silos, or bins can be obtained using boring
techniques. These techniques involve inserting a pipe into the
pile from top to bottom; the sample in the pipe represents a
vertical composite of the pile. This technique cannot be used,
however, with wet, coarse-grained, or lump materials. Small
borers can be used to sample solids in sacks if the borers can
reach to the bottom of the container.
There are two basic designs used for pipe borer sam-
pling. The first is a straight length of pipe tapered at the
insertion end. This is plunged through the pile, and the sample-
filled pipe is withdrawn. The second type, a solids thief, is
used when a more vertically representative sample is desired.
It is made of two close-fitting concentric pipes, each with a
sealed conical base and longitudinal slots in the walls (see
Figure A-22). The thief is inserted with the slots non-aligned
(turned away from each other). When it is in position, the
pipes are rotated (relative to each other) so that the slots
are aligned and provide access to the center of the inner pipe.
The openings through which a sample flows (either the end of
the simple borer or the thief slots) should be larger than the
largest particle size to be included in the sample. Thief
design variations are shown in Figure A-22.
The auger sampler is particularly suitable for sam-
pling materials that are packed too tightly for pipe or thief
techniques. An auger (Figure A-22) is like a large drill bit
which is turned into the pile from the top. When the auger
is withdrawn, the sample is packed in the helical grooves. If
necessary to prevent sample spilling, it can be enclosed in a
casing.
A-3.2 Sampling Methods for Process Conveyors
Process conveyors are the preferred sources for col-
lecting representative samples of solid materials, because
there is less segregation according to particle size, and the
material obtained is often a sample of solids actually being
-170-
-------
n
li
Split Tube Thief
Multi-Slot Tube Thief
Single-Slot Tube Thief
u MM • • •
Grain Probe
ga fcl B a BTEL>
Missouri Trier
Auger Sampler
Figure A-22. SOLIDS SAMPLING EQUIPMENT
(Ref. A-15)*
* Reprinted by permission of the American Society for Testing and Materials,
copyright 1973.
-171-
-------
used in the process. Generally, the best samples from conveyors
are obtained at a terminus where the exiting solid material falls
vertically from the conveyor. An inert container having sharp
parallel sides is passed laterally through the falling stream at
a uniform rate, thus collecting a representative sample. This
container should not be allowed to overflow. Other methods for
collection depend on which type of conveyor is used: belt,
ladder-tray, screw, or duct.
Samples may be taken from belt conveyors while they are
moving or, preferably, while stopped momentarily. In both cases,
a shovel is used to remove a full cross-section cut (having par-
allel sides). The cut should be at least three times as wide
as the largest piece of material encountered.
Ladder-tray conveyors are sampled by periodically
removing a shovelful from one of the trays.
Screw conveyors, which usually transport sludge-type
materials, are enclosed systems. These must be sampled at the
inlet or exit to the system by passing a sampling container
laterally through the stream.
Duct conveyors are either gravity-feed systems or
employ chain-drive scrapers. These conveyors can be either
open top or completely enclosed. Open ducts are sampled from
above the duct top by taking a shovelful from the duct-conveyed
solids. Closed ducts must be sampled at the exit point.
Sample reduction is required for making the sample
manageable for analysis. The quantity of material collected in
solid-stream sampling is usually much larger than that needed
for analysis. Therefore, the samples must be reduced in size
without affecting the distributions of components and particle
sizes. Two methods are generally accepted for accomplishing
this: the coning and quartering method and the riffling method.
It is frequently necessary to reduce the sample's
particle size before this sample reduction can begin. Above a
certain particle size, coning and quartering a sample into
smaller quantities is not a valid technique. Particle size
reduction is also necessary for most rifflers.
-172-
-------
The coning and quartering technique is illustrated in
Figure A-23. The sample is shaped into a conical pile which is
then sharply divided into quarters. Two opposite portions are
combined and the composite is then further reduced by again
coning and quartering. This is repeated until a sample of the
desired size is obtained.
The riffling technique is a mechanical way of sub-
dividing the sample systematically. The mechanical devices
known as rifflers are illustrated in Figure A-24. When the
sample is poured evenly over the top of the riffler, it is
divided into two equal fractions. Further sample size reduc-
tion is accomplished by recycling one of the fractions.
Continuous samplers for solid streams are commer-
cially available. In most cases, they are an automated version
of the falling-stream sampler described earlier, followed by
an automated sample-reduction system such as a riffler (or a
whistle pipe), with the excess sample returned to the solid
stream.
-173-
-------
Cone
Quarter
1+4
2+3
Combine
Figure A-23. CONING AND QUARTERING
-174-
-------
FEED CHUTE
AT LEAST
FOURTEEN
13 MMM/2")
TO 25 MM (1")
OPENINGS
RIFFLE SAMPLER
A. LARGE RIFFLE SAMPLERS
RIFFLE BUCKET AND
SEPARATE FEED CHUTE STAND
AT LEAST TWENTY-FOUR
6 MM (1/4") OR 25 MM (1")
OPENINGS
NOTE:
MAY BE CONSTRUCTED AS EITHER
CLOSED OR OPEN TYPE. CLOSED
TYPE PREFERRED.
B. SMALL RIFFLE SAMPLER
Figure A-24. RIFFLES FOR SAMPLE SUBDIVISION
(Ref. A-18)
-175-
-------
A-4 REFERENCES
A-l. Environmental Protection Agency, "Determination of Parti-
culate Emissions From Stationary Sources, 40 CFR 60, App.
A, Reference Method 5," Fed. Reg. 41(111), 23076-83 (1976)
A-2. American Society of Mechanical Engineers, Determining the
Properties of Fine Particulate Matter, Power Test Codes.
ASME PTC 28-1965, ANSI PTC 28-1974.New York, NY, 1965.
A-3. Texas Air Control Board, Laboratory Methods for Determina-
tion of Air Pollutants, Revised ed.Austin,TX~June 1976.
A-4. Jacobs, Morris B., The Analytical Toxicology of Industrial
Inorganic Poisons. New York, NY, Interscience, 19^67.
A-5. Texas, State of, Water Commission, Groundwater and Elec-
tronic Data Processing Divisions and Texas Water Pollution
Control Board, A Statistical Analysis of Data on Oil Field
Lys
Brine Production and Disposal in Texas for 1961 From an
Inventory Conducted by the Texas Railroad Commission.
:ory L
., TX,
Austin, TX, February
A-6. Environmental Protection Agency, "Determination of Sulfur
Dioxide Emissions From Stationary Sources, 40 CFR 60, App.
A, Reference Method 6," Fed. Reg. 41(111), 23083-85 (1976).
A-7 Environmental Protection Agency, "Determination of Sulfu-
ric Acid Mist and Sulfur Dioxide Emissions from Stationary
Sources, 40 CFR 60, App. A, Reference Method 8," Fed. Reg.
41(111), 23087-90 (1976).
A-8. Environmental Protection Agency, "Determination of Nitrogen
Oxide Emissions From Stationary Sources, 40 CFR 60, App.
A, Reference Method 7," Fed. Reg. 41(111), 23085-87 (1976).
A-9. Leichnity, J., Detector Tube Handbook, 2nd ed. Dragerwerk,
A. G. Lubeck, October 1973.
A-10. Environmental Protection Agency, "Determination of Moisture
in Stack Gases, 40 CFR 60, App. A, Reference Method 4 "
Fed. Reg. 41(111), 23072-76 (1976).
-176-
-------
A-ll. Weissburg, B. G., "Determination of Mercury in Soils by
Flameless Atomic Absorption Spectrometry , " Econ. Geol.
£6(7), 1042 (1971). -
A- 12. Perry, John H. , Chemical Engineers ' Handbook . 5th ed.
New York, NY, McGraw-Hill, 1973. -
A- 13. Code of Federal Regulations 40. Prni-PrHnn of the Environ-
oe o eera eguations 40. Prni-PrHnn of the Envi
ment, Pts . by-99. Revised ed. Washington, DC, General
Services Admin . , Office of the Federal Register, July 1976.
A-14. American Society for Testing and Materials, 1973 Annual
Book of AS TM Standards. Ft. 23. Water; Atmospheric
Analysis; Philadelphia, PA, 1973. - -
A- 15. American Society for Testing and Materials, "Standard
Recommended Practice for Sampling Industrial Chemicals,"
E300-73, in ASTM Book of Standards 1973. Philadelphia,
PA, 1973, p.~6Trr — ' -
A- 16. Environmental Protection Agency, Office of Technology
Transfer, Handbook for Monitoring Industrial Waste Water.
August 1973.
A- 17. Hamersma, J. W. , S. L. Reynolds, and R. F. Maddalone ,
IERL-RTP Procedures Manual : Level 1 Environmental Assess-
ment. EPA 600/2-76-106a. Redondo Beach, CA, TRW Systems
Group, June 1976.
A- 18. Hamersma, J. W., and S. L. Reynolds, Tentative Procedures
for Sampling and Analysis of Coal Gasification Processes.
Redondo Beach, CA, TRW Systems Group, March 1975.
A-19. Oldham, Ronald G. , and Robert G. Wetherold, Assessment.
Selection, and Development of Procedures for Determining
the Environmental Acceptability of Synthetic Fuel Plants
Based on Coal, Revised Report, 4 vols. FE-1795-3. Austin,
TX, Radian Corp. , May 1977.
A-20. Radian Corporation, Characterization of the Effluents
From the C02 Acceptor Process Sampling Plan. Tech. Note
200-154-04. Austin, TX, December 19767
A-21. Harris, W. E., and Byron Kratochvii , "Sampling Variance In
Analysis for Trace Components in Solids . Preparation of
Reference Samples," Anal. Chem. 46(2), 313 (1974).
A-22 American Society for Testing and Materials, 1974 Annual
Book of ATSM Standards. Pt. 26, Gaseous Fuel?; Coal~
and Coke; Atmospheric Analysis^. Philadelphia, PA, 1974.
-177-
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing!
1 PEPCRT NO.
EPA-600/7-78-134
2.
4. TIT|_E ANDSUBTITLE
Guidelines for Preparing Environmental Test Plans
for Coal Gasification Plants
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
July 1978
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
G.C. Page, W.E. Corbett. andW.C. Thomas
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
8500 Shoal Creek Boulevard
Austin. Texas 78766
10. PROGRAM ELEMENT NO.
EHE623A
11. CONTRACT/GRANT NO.
68-02-2147, Exhibit A
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: 5/76-12/77
14. SPONSORING AGENCY CODE
EPA/600/13
15 SUPPLEMENTARY NOTESIERL_RTP project officer is William J. Rhodes, Mail Drop 61.
919/541-2851.
is. ABSTRACT
rep0rt outlines a philosophy and strategy for preparing environmental
assessment sampling and analysis (test) plans. Five major points of test plan devel-
opment are addressed: (1) defining the test objectives , (2) performing an engineering
analysis of the test site, (3) developing a sampling strategy, (4) selecting analytical
methods, and (5) defining data management procedures. The important considera-
tions involved in each area are discussed in relation to three types of environmental
tests: (1) waste stream (Levels 1, 2, and 3), (2) control equipment, and (3) process
stream characterization. Specific sampling and analytical methods are presented,
with numerous references cited for more detailed information.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATl 1 icld/Group
Pollution
Coal Gasification
Tests
Assessments
Samp ling
Analyzing
Properties
Pollution Control
Stationary Sources
Environmental Assess-
ment
Characterization
13B
13I-I
14B
1. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
187
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-178-
------- |