Encyclopedia of Environmental Science and Engineering, Volume I and II

1094

STACK SAMPLING

INTRODUCTION

It is frequently necessary to determine the amount, con-

centration, or rate of emission of various pollutants in the

exhaust streams from industrial or commercial processes.

Because this generally involves the sampling and eventual

analysis of the gas flowing through a stack into the outside

air, it is usually called “stack emission sampling”, or more

simply, “stack sampling”.

In most cases, it is not possible or practical to collect all

of the gases emitted to the outside air over any reasonable

time period. Therefore, it is necessary to collect only a frac-

tion of the overall gas stream. A representative side stream is

isolated from the main flow (usually by removing it from the

stack altogether) and processed in some way. The gas stream

may be filtered, condensed, bubbled, adsorbed, bottled, or

pumped through an automatic analyzer. The equipment used

for this purpose is called the sampling train. The end result

of this step is usually an assessment of the contents of the

stream. Meanwhile, an assessment is made about the other

characteristics of the stack gas itself, such as temperature

and flow rate. The information from these two assessments

is then combined to produce a measure of the emissions

from the stack in the desired units, such as pounds per hour,

grains/dry standard cubic foot, kg/kg of fuel burned, and so

on. Generally, the units are chosen to conform with appli-

cable regulations.

Of course, the testing must include all of the pollutants

that are of interest, under all of the process conditions that

are needed. This must also include a wide range of checks

and balances, often termed Quality Control (QC) and

Quality Assurance (QA) to ensure and measure the reliabil-

ity of the results. Most importantly, it all must be based on a

well reasoned plan, called a Stack Test Protocol or Quality

Assurance Project Plan (QAPP) that is written ahead of

time and approved by all interested parties. Finally, these

results must be reported in a way that fairly represents what

was done, to allow regulators and/or the sources to make

informed decisions.

A prime criterion for method selection is that the method

must produce data useful for the purpose intended. For

example, is the detection limit of the method low enough

to prove compliance with the emission standard, or is the

method able to measure and/or account for cyclonic flow in

the stack, or is it able to differentiate between similar chemi-

cal species, or is that necessary? These sorts of questions fall

into a general category called QA. They can be restated as

six data quality parameters:

P Precision Repeatability

A Accuracy Bias, closeness to “correct”

R Representativeness Typical of actual stack gas

C Comparability Similar to other data

C Completeness Enough information

S Sensitivity Low enough detection levels

These so-called PARCCS Parameters are simply ways

to ensure that the results of a sampling project will yield

useful results. When methods published by EPA are used,

the PARCCS parameters have already been determined and

are built into the methods. For novel methods, the PARCCS

parameters must be determined, at least qualitatively. This is

well beyond the scope of this chapter but must be borne in

mind should new methodology need to be developed to fit a

specific circumstance.

This leads to the most important warning concerning

stack testing. In all but the most dire emergencies, stack test-

ing projects should be planned and carried out by trained,

experienced stack testing teams. No one should believe that

even a close reading of this chapter would provide sufficient

background to plan or perform stack tests.

There are four recent advances in the stack testing

field that rate special mention. The first is the increasing

reliance on external quality assurance, as embodied in

audit samples and devices. Several agencies and organiza-

tions, particularly the Emissions, Monitoring and Analysis

Division in USEPA’s North Carolina facility, are now pro-

ducing reliable Performance Evaluation (PE) audit samples

that can be obtained for the purpose of checking analytical

accuracy at the time of sample analyses. They are avail-

able for many parameters, as listed in the specific Test

Methods.

The second major recent advance in stack testing is the

development and proliferation of reliable continuous emis-

sion monitoring systems (CEMS). CEMS are generally

electronic analyzers that determine and record instantaneous

concentrations of given parameters continuously.

For example, a CEMS for stack gas opacity, termed a

transmissometer, shines a light beam through the stack gas

and measures the fraction of light transmitted, taking a reading

C019_003_r03.indd 1094C019_003_r03.indd 1094 11/18/2005 11:07:13 AM11/18/2005 11:07:13 AM