Measurement of Gaseous Emission Rates from
Land Surfaces Using an Emission
Isolation Flux Chamber. User's Guide
Radian Corp., Austin, TX
Prepared for

Environmental Monitoring Systems Lab.
Las Vegas, Jiv
Peb 86

-------
                                      EPA/600/8-86/008
                                      February 1986
      MEASUREMENT OF GASEOUS EMISSION
          RATES FROM LAND SURFACES
             USING AN EMISSION
           ISOLATION FLUX CHAMBER

                USER'S GUIDE
                     by
              M. R. Klenbusch
             Radian Corporation
               P.O. Box 9948
            Austin, Texas 78766
        EPA Contract No. 68-02-3889
             Work Assignment 18
   Project Officer:  Shelly J.  Williamson
        Exposure Assessment Division
Environmental  Monitoring Systems Laboratory
          Las Vegas, Nevada  89114
 ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
             LAS VEGAS, NV 89114

-------
                            -      TECHNICAL RETORT DATA    ^
                            (Htnt nti InttnttHam on Mr rrrcne trfort compltlittl
 REPORT no.
  EPA/600/8-86/008
                                                            5. RECIPIENT? ACCESSION NO.
                                                              P9     n  o o  -1 /?
 TITLE AND SUBTITLE
 MEASUREMENT OF GASEOUS  EMISSION RATES  FROM LAND
 SURFACES USING AS EMISSION ISOLATION FLUX CHAMBER
 USER'S GUIDE
                                                            S. REPORT DATE
                                                              February 1986
                                                            t. PtHf OftUINO ORGANIZATION COOC
 AUTHOR(S)

 M. R.  Klenbusch
                                                            a. PERFORMING ORGANIZATION REPORT NO,
                                                            1O. PROGRAM ELEMENT NO.
                                                              ABSD1A
. PERFORMING ORGANIZATION NAME AND ADDRESS
 Radian  Corporation
 P.O.  Box 9948
 Austin,  TX  78766
                                                            n. CONTRACT/GRANT NO.
                                                              Contract, Radian Corp.
                                                              Ho.  69-02-3889
t2. SPONSORING AOENCY NAME ANO ADDRESS
 Environmental Monitoring  Systems Laboratory - LV, NV
 Office  of Research and Development
 U.S.  Environmental Protection Agency
 Las Vegas,  NV  89114
                                                            1*. TYPE OF REPORT AND PERIOD COVERED
                                                              Uaer'g Gu-Mo  .    	
                                                            14. SPONSORING AGCMCY COOE
                                                              EPA/600/07
15. SUPPLEMENTARY NOTES
 6. ABSTRACT

       A promising method  for monitoring ground emissions  involves the use of an
 emission isolation flux  chamber.  The method is simple,  easily available, and
 inexpensive.  Applications  would include  RCRA and CERCLA facilities.  To date,
 a uniform method operations does not exist.   For this reason,  an operations
 guide has been developed.   This guide presents literature surveys,  operation
 protocols, a case study, and references for  further reading.   The use of this
 protocol will aid in unifying flux chamber measurements  and increase data
 comparability.
17.
                                KEY WORDS ANO DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b,IOENTIFIERS/OPEN ENDED TERMS  C. COSATI Field/Croap
18. DISTRIBUTION STATEMENT

   RELEASE TO PUBLIC
                                               1>. SECURITY CLASS (T
                                                  UNCLASSIFIED
                                                                            » NO* Vr PAQ6S
60
                                               JO. SECUBITYCLASS (THt M*J
                                                  UNCLASSIFIED	
                                                                          12. PRIC
EPA F*ml220-t (••». 4-77)   PKBVIOU* KOITIOM i» OMOLKTC

-------
                                                     mKTRUCTIOItt

   I.  DEPORT NUMBER
       Imrrt the EPA report number at it appears on the cam of At pubtkttk*.

   2.  LEAVE BLANK

   J.  RECIPIENTS ACCESSION NUMBER
       Reserved fot usf by «cfi report recipient.

   4,  TITLE AND SUBTITLE
       Till* should indicate  dearly and briefly tht subject coverage of the report, and be disnlarrd prmuanrtr. Srt wblftle. If uard. i* mutter
       type ot otherwise subordinate it to mam title. When a report a prepared in more ihM OIK vohntt. repeat UK primary litk. add volume
       number and mctude subtitle for the specific title                                                   r-    »            -«

   S.  REPORT DATE
       Each report shall carry a date indicating at kail month and year.  Indicate the hob on which it wa* ithcttd (r*., dttr ofiatr dot of
       tpprottl. date o{ fttfantton. tie.I.                                                             ^      '     '      '

   (.  PERFORMING ORGANIZATION CODE
       Lean blank.

   r.  AUTHORS)
       Give lumen) m ronvraiioaal otde» (John R. Dot. J. Robtn Dot. rle.J. Ul auUioc'i ifTUiiikM (f H dMTen from (be perfo«miB» otfUt-


   t.  PERFORMING ORGANIZATION REPORT NUMBER
       Inxn if perfortninf otpnizatioa viitici to aitifn thu numbct.

   «.  PERFORMING ORGANIZATION NAME AND AOOREGS
       oive name, (tnet, city, itate. and ZIP code,  list  no more thin (wo levrb of an organizational (Ureaitky.

   tO.  PROGRAM ELEMENT NUMBER
       UK UK ptojrwn element number under which the report was prepared. Subordinate number* any be MkHudcU in parenlbcict.

   11.  CONTRACT/GRANT NUMBER
       InieTt contnci or grant number under which report was prepared.

   12.  SPONSORING AGENCY NAME AND ADDRESS
       Include IIP code,

   13.  TYPE OF REPORT AND PERIOD COVERED
       Indicate interim final, etc., and if applicable, dates covered.

   14.  SPONSORING AGENCY CODE
       Insert appropriate code.

   IS.  SUPPLEMENTARY NOTES
       Enwr information not included elsewhere but usetwl, such as: Prepared in cooperation with. Translation of, Ptcvnlcd at COOIVU-IHX «>r.
       To be published in. Supersedes. Supplement!, etc.

   1».  ABSTRACT
       Include a bntf (200 »crdi or lea/ factual wmmary of the mmt sipiiTKant infonaalion contained in the report.  If the report cunmim a
       Ufni/icjm bibjkijnphy o> literature survey, mention it be»e.

   17.  KEY «VOROS AND DOCUMENT ANALYSIS
       (a) DESCRIPTORS -  Select from the Theuunii of Enguieermg and Sdenlific Term (he proper a»lhoened tcrm» that identify the major
       concept of the research and are sufficiently specific and precise to be tued as index entries for cataloging.

       (b) IDENTIFIERS AND OPEN-ENDED TERMS • Use identifiers for project IUIMS, cede names, equipment dctifjulots, etc.  UK open-
       ended tttros written in descriptor form.foi ihoie subjects for which BO descriptor exist*.

       (c) COSAT1 HLLD GROUT - Field and group assignment! are to be taken from the 196$ COSATI Subject Category Usl. Since the ma-
       jority of documents are moltidflaplinary in natunt, the Primary FieM/Group asugnmetilfi) iriB 6e wdTic disoplmc. are* of hum™
       endearor, or type of pbyskal object. The appUc*Uon
-------
                                   NOTICE
     The information in this  document has been funded wholly or in part
by the United States Environmental Protection Agency under Contract
No. 68-02-3889 to Radian Corporation.  It has been subject to the Agency's
peer and administrative review, and it has been approved for publication
as an EPA document.
                                    11

-------
                                   OF CONTENTS
faction.                                                               Page

   1      INTRODUCTION ........................  1-1

   2     BACKGROUND .........................  2-1
         2.1  Emission Processes ...................  2-1
         2.2  Measurement Techniques ......... ........  2-2
         2.3  Flux Chamber Operation. ... ........ .....  2-3

   3     MEASUREMENT OF GASEOUS EMISSION RATES FROM LAND SURFACES
         USING AN EMISSION ISOLATION FLUX CHAMBER - PROPOSED METHOD .  3-1
         3.1  Applicability and Principle ..............  3-1
              3.1.1  Applicability ..................  3-1
              3.1.2  Principle ....................  3-1
         3.2  Precision, Accuracy, Sensitivity, and Range ......  3-1
              3,2.1  Precision ....................  3-1
              3.2.2  Accuracy ....................  3-2
              3.2.3  Sensitivity ...................  3-2
              3.2.4  Range. ....... ..... .........  3-2
         3.3  Interferences .....................  3-2
              3.3.1  Flux Chamber Method ...............  3-2
              3.3.2  Emission Process ................  3-4
         3.4  Apparatus and Materials ................  3-4
              3.4.1  Flux Chamber and Supporting Equipment ......  3-4
              3.4.2  Discrete Sample Collection ...........  3-6
              3.4.3  Analysis , ...................  3-6
         3.5  Procedure ....................... 3-11
              3.5.1  Flux Chamber Operation .... ..... .... 3-11
              3.5.2  Sample Collection. ................ 3-11
              3.5.3  Sample Analysis ..... ....... ..... 3-13
              3.5.4  Sampl Ing Strategy ................ 3-15
         3.6  Calibration  ...................... 3-18
              3.6.1  Equipment .................... 3-18
              3.6.2  Analyzers. . .................. 3-19
         3.7  Qua I Ity Control .................... 3-20
              3.7.1  Sampling Equipment ............... 3-20
              3.7.2  Sampl ing .................... 3-22
              3.7.3  Analytical ................... 3-22
         3.8  Calculations ...................... 3-23
              3.8.1  Definitions ................... 3-23
              3.8.2  Percent Recovery ........ . ....... 3-25
              3.8.3  Calculation of the Dilution Factor Involved
                     In Gas Canister Analysis ............ 3-27
              3.8.4  Area  Source Emission Rate Equations ....... 3-27

   4     CASE STUDY .........................  4-1

   5     ADDITIONAL INFORMATION ...................  5-1

         REFERENCES .........................  R-1

         APPENDIX A - SELECTION OF A RANDOM SAMPLE. . . .......  A-t

-------
                                 SECTION 1

                                INTRODUCTION
     Volatilization of organic compounds from contaminated soil or ground-
water Into the air represents a major potential source of exposure which has
not been assessed.  In order to assess this exposure potential, a method Is
needed to directly measure gas emission rates frcm contaminated soils and/or
groundwater.  Additionally, It Is recognized that an understanding of the
volatilization, transport, and emission processes could lead to a predictive
tool for exposure assessment.  The Information provided by direct mea-
surement *nd/or predictive modeling will allow state and local regulatory
agencies to develop programs to assess and define the need to control gas
emissions from area sources contaminated by organic compounds.

     The purpose of this User's Guide Is to present an approach and proto-
col , namely the emission Isolation flux chamber (or flux chamber) technique,
for measuring emission rates of volatile organic compounds from contaminated
soils and/or groundwater.  Presented Is the theory of operation, specifica-
tions, sensitivities,  method of operation, and data reduction procedures for
this technique.  It Is assumed that the Individuals who will use the proto-
col are. In general, familiar with sample collection and analysis of vola-
tile organic compounds.  Also Included  In this document Is a case study that
demonstrates the measurement and data reduction processes around a spill
site.

     The flux chamber  technique Is applicable to the measurement of emission
rates frcm Resource Conservation and Recovery Act (RCRA) facilities (hazard-
ous waste land-treatment, and landfill facilities), and from Comprehensive
Environmental Response, Compensation, and Liability Act (CERCLA) area
sources contaminated by losses of volatile organic compounds from spills,
from leaking underground storage tanks, frcm pipelines, and/or from surface
Impoundments.

    This protocol  does itot present the vast amount of work that was required
to develop this document.  Rather, the protocol Is a result of IIterature
reviews selecting a measurement technique and field applications demon-
strating the technique and developing a data base and validation studies
Identifying the method of flux chamber operation.  References to the other
area sources where this technique was applied, the work performed to vali-
date the technique, and the Investigations of variables which control the
emission process are also given for those Individuals desiring further
Information.
                                   1-1

-------
                                 SECTION 2

                                 BACKGROUND
     The following subsections discuss the process by which volatile organic
compounds are emitted froa contaminated land surfaces, the basis upon which
the flux chamber technique was selected as an approach for measuring such
emission rates, and the principle of the technique.

2.1  Emission Processes

     The rate of volatile organic compound (VOC) emissions from contemlnated
soils Is generally believed to be controlled by the diffusion rate of the
chemical compound through the air-filled pore spaces of the soil.(1,2,3)
The exception occurs when the contaminated material Iles on or very near the
sotl surface.  Such Is the case when spills occur or  Immediately after waste
Is surface-applied to a landtreatment site.  In these cases, the emission
process will be controlled by the rate of evaporation.

     Evaporation Is a surface phenomenon, and trte parameters that affect the
evaporation process are the properties of the waste Itselt as well as those
that have an effect on the air-surface Interface (I.e. wind, surface rough-
ness).  The Important parameters Include the volatility or vapor pressure of
the waste, ambient meteorological conditions (solar Insolation, air and
waste temperature, surface wind speed, relative humidity), surface coarse-'
ness, and the bulk concentration of the volatile components In the air
(although this Is usually very low and generally assumed to be negligible).

     There are two major types of soil emission processes.  Each are treat-
ment dependent.  One type occurs In Iandtreatment facilities and the other
at underground facilities such as landfills.  In land-treatment applications,
the emission rate Is generally highly time-dependent,  When a fixed amount
of waste Is applied to the soil surface.  It penetrates the soil to a certain
depth.  The vaporization rate  Is maximum  Immediately  after waste applica-
tion, as the material nearest the surface  Is vaporized and diffuses through
a very thin layer of soil.  As the waste near the surface  Is depleted of  Its
VOC content, the volatile material deeper  In the soil aust diffuse through
an  Increasingly thick soil layer.  The soil presents  a resistance to VOC
diffusion In direct proportion to the VOC  depth.  Thus, the rate of emis-
sions from the surface decreases with time.

     It Is common practice In  Iandtreatnent to periodically till the sofl to
provide oxygen for bacterial activity.  The tilling effectively mixes the
remaining waste In a homogeneous layer near the soil  surface.  The emission
rate Is at a maximum Immediately following each titling episode since
                                   2-1

-------
volatile waste Is again present very near the surface, and resistance to
diffusion Is at a minimum.

     Although also diffusion controlled, the emission process from under-
ground sources such as landf11 led waste or  material present as a "lens" on
the water table has significantly different characteristics than that from
surface or near-surface sources.  The depth of the emission source Is
usually quite substantial.  Therefore, the emission rate Is Initially lover
due to the resistance to diffusion produced by the coloumn of soil.  The
Initial emission rate Is zero,  since It takes some time for the volatile
material to diffuse through the soil layer.  The adsorptlve sites on the
soil particles must also be Initially saturated.  Once the emission rate has
equilibrated, the rate Is relatively constant with time until  the under-
ground source Is exhausted.

     The diffusion process Itself through the soil  Is the same for both
types of sources, landtreatment (surface) and landfill (underground).  Con-
sequently, many of the parameters Important to the emission processes are
the same, Including dtffuslvlty of the VOC In air, soil properties (particle
size distribution, soil  type, moisture content, particle density, porosity),
soil/waste temperature,  and volatility of the VX In the waste.  Additional
parametars Important to the near surface emission processes are the amount
of material  present In the contaminated soil  layer, the Initial depth of the
contamination, the elapsed ttme from application  (or tilling) and, possibly,
ambient conditions such as surface wind speed and relative humidity.  The
depth of the soil layer above the waste is a very Important parameter In the
em sslon process from subsurface sources.  Additionally, the adsorptlve
properties of the soil may also have a significant effect on the emission
rate from this latter source type.

     An understanding of the emission processes and the Important parameters
Is necessary In the measurement of emission rates from soil surfaces and In
the proper Interpretation of the test results.  As an example, the emission
rate from a source Is affected by rain since the  porosity and, hence, the
diffusion rate are reduced with Increasing moisture content of the soil.
Thus, emission rates Immediately after a rainfall will be lower than those
from drier soils and may take substantial periods of tine to return to the
emission rate prior to the rain,(4)  Emission rates may vary with the time
of day and season, as a result of changes  In ambient and sol I/waste tempera-
tures. (4)  Emission rates from soil areas containing fissures can be higher
and much less homogeneous than those from unfractured areas.  Thus, consi-
derable care must be taken In planning and implementing a measurement pro-
gram to determine representative emission rates from such soli surfaces.

2.2  Measurement Techniques

     Based on a literature review (5>, the techniques for determining gas
emissions rates from land surfaces contaminated with organic compounds can
be divided Into three approaches: Indirect measurements, direct measure-
ments, and laboratory simulations,  indirect  techniques typically require
measurements of ambient air concentrations at or  near the site.  These
                                   2-2

-------
measurements are related to the surface area of the area source and local
meteorological conditions using a dispersion model to determine an emission
rate.  The second approach is to directly measure emission rates using for
example the flux chamber.  The third approach Is to create an emission
source In the laboratory and model the emissions by various techniques for
application to field sites.  These three approaches were compared for preci-
sion, accuracy, and sensitivity.  Other considerations Included applicabil-
ity, complexity, manpower requirements, and costs.

     The most promising technique for measuring gas emission rates from land
surfaces was determined to be the emission isolation flux chamber technique.
The advantages are:

     o    lowest (most sensitive) detection limit of the methods
          examined;

     o    easily obtained accuracy and precision data;

     o    simple and economical equipment relative to other
          techniques;

     o    minimal manpower and time requirements;

     o    rapid and simple data reduction; and

     o    applicable to a wide variety of surfaces.

2.3  Flux Chamber Operation

     The flux chamber technique has been used by researchers to measure
emission fluxes of sulfur, nitrogen, and volatile organic species
(6,7,8,9,10).  The approach uses a flux chamber  (enclosure device) to sample
gaseous emissions from a defined surface area.  Clean dry sweep air Is added
to the chamber at a fixed, control led rate.  The volumetric flow rate of
sweep air through the chamber Is recorded and the concentration of the
species of interest Is measured at the exit of the chamber.  The emission
rate is calculated as:

                                 E, - YtQ/A                          (2-1)

where:  Ej • emission rate of component I (mass/area-time),
        Y| » concentration of component I In the air flowing fron the chamber
              (mass/volume),
         0. « flow rate of air Into the chamber (volume/time),
         A » surface area enclosed fay the chamber  (area).

All parameters  In Equation 2-1 are measured directly.

     Most of the emission rate assessments are of area sources much larger
than the enclosed surface area of the flux chamber  (0.130 m2).   In these
                                   2-3

-------
cases, an overall emission rate for the  area source Is calculated  from
•ultlple measurements based on random  sampling and statistical  analysts.
                                   2-4

-------
                                 SECTION 3

          MEASUREMENT OF GASEOUS EMISSION RATES FROM LAND SURFACES

         USING AN EMISSION ISOLATION aUX CHAMBER - PROPOSED METHOD



3.1  Applicability and Principle

     3.1.1  Applicability

     The flux chamber technlquo Is applicable to the measurement of emission
rates from Resource Conservation and Recovery Act (RCRA) facilities such as
hazardous waste Iandtreatment and landfill facilities.  This technique Is
also applicable for emission rate measurements from Comprehensive Environ-
mental Response, Compensation, and Liability Act CCERCLA) waste sites such
as areas contaminated by losses of volatile organic compounds from spills,
from leaking underground storage tanks, frctn pipelines, and/or from surface
Impoundments.

     3.1.2  Principle

     Gaseous emissions are collected from an Isolated surface area with an
enclosure device called an emission Isolation flux chanter  (or flux cham-
ber).  The gaseous emissions are swept through an exit port where the con-
centration Is monitored and/or sampled.  The concentration  Is monitored
and/or sampled either continuously (I.e., "real-time") or discretely.  Real-
time measurements are typically made with portable total hydrocarbon ana-
lyzers and are useful for relative measurements (I.e.? the determination of
flux chamber steady-state operation, zoning).  Discrete samples are taken
when absolute measurements are necessary  (I.e., steady-state concentrations,
emission rate levels).  Tha emission rate Is calculated based upon the sur-
face area Isolated, the sweep air flow rate, and the gaseous concentration
measured.  An estimated average emission rate for the area source  Is calcu-
lated based upon statistical sampling of a defined total area.

3.2  Precision, Accuracy, Sensitivity, and Range

     3.2.1  Precision

     Single Camber precision (I.e., repeatabllIty) of the method  Is approx-
imately 5 percent at measured emission rates of 3,200 ug/mln-m2.  Variabil-
ity between different flux chambers (I.e., reproduclbtllty)  Is, approximately
9.5 percent within a measured emission rate range of 39,000 to
65,000 ug/mln«m2.(4>
                                   3-1

-------
     The roproduclb!IIty results were determined from a bench-scale study.
The tests were designed to eliminate temporal variations from the flux
chamber reproduclbllIty.  However, using the same bench-scale facility, a
test design was not possible for measuring flux chamber repeatability with-
out bias from temporal  variations.  As a result, the repeatability tests
were performed In the laboratory.  The differences therefore between the
stated emission rates for repeatability and reproduclbllfty reflect the
differences In laboratory simulated emission rates and those meausred from
the bench-scale facility.

     3.2.2  Accuracy

     Flux chamber recovery (Section 3.6.1.4.2) results show a recovery range
of 77 percent to 124 percent.  Table 3-1 lists measured recoveries for a
number of compounds tested.  The average recovery for the 40 compounds
tested Is 103 percent.

     Flux chamber emission rate measurements made on the soil cells range
from 50 percent to 100 percent of the predicted emission rates.  That  Is,
the measured emission rates can be expected to be within a factor of one-
half of the "true" emission rates.(4)  The flux chamber accuracy based upon
both the recovery tests and predictive modeling ranges from 50 percent to
124 percent.

     3.2.3  Sensitivity

     The sensitivity of this method depends on the detection limit of the
analytical  technique used.  When discrete samples are collected using gas
canisters and analyzed by gas chromatographIc methods, the estimated emis-
sion rate sensitivity Is 1.2 ug/mln-m^ for an analytical detection limit of
10 ppbv benzene.  When emission rates are measured In a continuous {real-
time) method, the estimated sensitivity Is 124 ug/mln-m^ for an analytical
detection Unit of 1 ppmv benzene.

     3.2.4  Range

     The range of this method depends upon the analytical technique used.
High level  emission rates are analyzed by Introducing proportional amounts
of gas sample to the analyzer.  Using this technique, high level emission
rates of 120,000 ug/mln-m* have been measured.(4)  Low levels are  limited by
the sensitivity of the analytical technique.  Gas chroraatographIc techniques
have been used to measure  low level emission rates of 1.2 ug/mln-m2 for mea-
sured concentrations of  10 ppbv benzene.

3.3  Interferences

     3.3.1  Flux Chamber Method

     Impurities  In the sweep atr and/or organic compounds outgasslng from
the transfer tines and acrylic chamber top may causa background contamina-
tion.  The emission Isolation flux chamber must be demonstrated to be  free
                                   3-2

-------
                                 TABLE 3-1
          COMPOUNDS TESTED IN  THE EMISSION ISOLATION FLUX CHAMBER
                     AND THE MEASURED PERCENT RECOVERY
       Compound
 Percent
Recovery*
Compound
 Percent
Recovery*
Total C2
Total C3
Isobutane
1-toutene
n-butane
t-2-butene
c-2-butene
Isopentane
1-pentene
2-«ethyI-1-butene
n-pentane
n-pentene
c-2-pentene
Cyclopentene
n-hexane
Isohexane
3-methyIpentane
MethyIcycIopentane
Benzene
1,2-DImethyIpentane
   100        3-«ethyIhexane               106
   108        2,2,4-trlmethyl pentane       106
   109        n-heptane                    103
   108        Methyfcyciohexane            103
   106        Toluene                      103
   107        Ethyl benzene                 94.7
   109        ntp-xylene                    88.?
   112        o-xylene                      97.3
   105        n-nonane                      99.4
   124        n-propyI benzene               95.5
   107        p-ethy I toluene                92.5
   103        1,3,5-trlroethylbenzene        93.5
   105        1,2,4-trJmethylbenzene        88.7
   105        2-«iethyl-2-butene            103
    95.1      Methyl mercaptan             107
   107        Ethyl nercaptan              107
   106        Butyl nercaptan              101
   105        Tetrahydrothlophene          115
   106        Trlchloroethylene             77.1
   105        Ethylene 
-------
from significant (
-------
                                                           FIGURE 3-1
                                A CUT/WAY  DIAGRAM OF THE EMISSION ISOLATION FLUX CHAMBER AND
                                                       SUPPORT  EQUIPMENT
                                                 TEMPERATURE
                                                  READOUT
                                                                    THERMOCOUPLE
\J»
                                                                                              SYRINGE/CANISTER
                                                                                             V  SAMPLING PORT
                                                                                                    REAL TIME
                                                                                                    ANALYZER
                     CARRIER
                       GAS
                                                                                          OUTLET LINE
STAINLESS STEEL
 OR PLEXIGLAS
                                                            CUT AWAY TO SHOW
                                                           SWEEP AIR INLET LINE
                                                           AND THE OUTLET LINE

-------
Table 5-2.  A construct ton diagram of the flux chamber Is shown In
Figure 5-2.

     The sweep air carrier gas should be dry, organic free air equal to or
better than commercial ultra high purity grade «0.1 ppmv THC).  A gas flow
meter with no Internal rubber parts and adjustable wfthln the range of 1-10
L/mln should be used to control gas flow.  Temperature measurements should
be made with an accuracy of ±1.0°C.  A fine-wire thermocouple with elec-
tronic readout Is recommended.  Caution should be taken to avoid any contact
of a thermocouple with metal.  This would give Inaccurate air temperature
readings.  A pressure release port Is required to avoid pressure bulid-up
Inside the flux chamber during operation.  This port should never be
blocked.  For system blanks, a clean Teflon" sheet should bo used to provide
a clean surface for the flux chamber.

     5.4.2  Discrete Sample Collection

     Discrete grab samples should be collected with air-tight. Inert con-
tainers.  For on-slte analysis, 100 ml precision lock, glass syringes are
recommended.  Glass plungers are recommended over Teflon" tip plungers.   If
Teflon" tip plungers are used, then special controls must be followed to
avoid cross-contamination  (Section 3.7.1.1).  For samples to be transported
or to be stored for periods longer than 1 hour, 2L stainless steel gas
canisters are recommended.

     5.4.5  Analysis

          5.4.5.1  Real Time

               Analyzer

     For real-time, continuous monitoring of the exit gas concentration,
analyzers with precision of ±10 percent of the measured value and a detec-
tion I Imlt of 1 ppmv are recommended.

               Calibration Gases

     The portable, real-time analyzers will require the following levels of
calibration gases:

     o    High-level Gas: Concentration within 50 percent to 90 per-
          cent of the span value  (maximum expected concentration or
          upper limit of  Instrument  linear range).

     o    Low-Level Gas: Concentration  less than or equal to O.Ot
          percent of the span value.

     o    Zero Grade Gas: Ultra high  purity  (UHP) air  «0.1 ppmv THC).

     The calibration gas for these jnalyers-can be the same as that used for
the on-slte discrete analyzer  (Section 5.4.5.2.2).


                                   3-6

-------
                                 TABLE 3-2
                        CHAMBER MATERIALS SPECIFICATIONS
    Item
 Description
       Specification
Carrier Gas Lines:
Inlet/Outlet       Teflon" (clear)
Sweep Afr Wrap
Perforation8
FIttlngsb
Thermocouples
 Air (!)
Flux Chamber:
 Base

 Support ring
 flange
 Dome
 Seal
 Dome to Base
Statnle-js Steel
four equidistant holes
jetting direction


Stainless steel

Stainless steel
Fine wire
K type
Stainless steel
column
Stainless steel

Acrylic
four holes

Inlet/outlet

Air temperature

Pressure release


Top gasket

Dome IIp
1/4" 00, 5« to 8' long,  thin
walled, 1/4" stainless steel
fittings

1/4" OD, 54" long, perforated
hole No. 1 (nearest Input),  5/64"
ID, holes No. 2-4, 3/32" ID,
axially, horizontally

1/4" bulkheads with teflon
washers for chamber penetration
1/4" cap to seal  wrap I Ine end
36" long, bead tfp,  teflon coated
(extensions optional),  penetrate
flux chamber 3", support with
1/4" bulkhead with septa
16" ID x 7" tall, welded to a
support ring flange
16" ID x 20" 00 x 1/4" thick

Spherical, 4" displacement at
center, 16" 10 at seal, 2" lip
for seal, 1/4* thick,  molded
Equidistant, 4* from aluminum
gasket
1/2"  ID with 1/4" stainless steel
bulkhead
1/2"  ID with 1/4" stainless steel
bulkhead
13/16" ID with 3/4" stainless
steel bulkhead

Aluminum 16" ID, 20" OD, 1/4"
thick
Below alumlnu* gasket Is the
acrylic lip of dome
                                                                (Continued)
                                     3-7

-------
                                 TABLE 3-2
                                (Continued)
    Item
 Description
       Specification
                   Seal Ing washer

                   Botton gasket
                   Fasteners
 Volume

 Surface Area

 Exit Line Probe   Teflon

 Perforation
With 1" soil
penetration
Enclosed by chamber
2 rows of holes
Teflon, 16" 10,  20" 00, 1/32"
thick
Stainless steel  support ring
20, 1/4" bolts equidistant around
Up   ,
0.03 m3 (30L)

0.130 m2

1/4" OD, 6" long, stainless
steel fitting, perforated
3/32" ID, 5 holes per row, 1"
separation, rows are positioned
orthogonally
"Avoid placement of exit line probe In Jetting path of  sweep air Inlet holes

bAII  fittings are manufactured by Swagelok* or equivalent manufacturer
 (bulkheads use Teflon" washers for sealing)
                                     5-e

-------
                                   FIGURE 3-2
                     EXPLODED  VIEW OF THE FlUX CHAMBER
                                  THERMOCOUPLE
INLET CARRIER
  GAS LINE
            SEPTA

    V." BULKHEAD
  & TEFLON WASHER
     FLUX CHAMBER
         DOME

         DOME UP

          SEALING
          WASHER
                                                             1 OF 20
                                                               BOLTS
                                                            TOP GASKET
        OUTLET CARRIER
           GAS UNE
    BULKHEAD
4 TEFLON WASHER
     BULKHEAD
 & TEFLON WASHER
(PRESSURE RELEASE)

       20 HOLES.
     EVENLY SPACED
    (SEE TOP GASKET)


 PERFORATED
SWEEP AIR WRAP
                                                          INLET HOLES

                                                          EXIT UNE PflCJBE
                                                            PERFORATED
     20 HOLES,
  EVENLY SPACED
  (SEE «DP GASKET)
                                                               SUPPORT
                                                             RING FLANGE
                                                           FLUX CHAMBER
                                                               BASE
                                     3-9

-------
               Quality Control (QC) Gas

     The portable, real-time analyzer will require a quality control «3C)
gas concentrated to fall within the span range.  The QC gas for this analy-
zer can be the same as that used for the on-slte discrete analyzer,

          3.4.3.2  Discrete

               Analyzer

     The analyzer should be sensitive with low detection limits.  For on-
slte analysis of grab samples, Instrumentation having precision of ±5 per-
cent of the measured value with a detection limit of 1 ppm  Is recommended.
Analyzers with Injection loops are recommended to reproduce the sample
volumes Injected.  For off-site analysis, Instrumentation with precision of
±30 percent at detection IImlts of J ppbv are recommended.

               CalIbratIon Gases

     The concentrations and composition of the calibration gases to be used
will vary depending on the species of Interest.  Preferably, the following
gas concentrations should be used for each species of Interest:

     o    High-level Gas: 90 percent of the span vaJue.

     o    Mid-Level Gas: Average expected concentration.

     o    Low-Level Gas: 0.01 percent of the span value.

     o    Zero Grade Gas: Ultra high purity (UHP) air, (
-------
3.5  Procedure

     3.5.1  Flux Chamber Operation

     The flux chamber Is operated Identically  for real-tine and discrete
sampI Ing.

          3.5.1.1  Preparation

     All exposed chamber surfaces should be cleaned with water and wiped dry
prior to use.  Assemble the samp I Ing apparatus and check for malfunctions
and leaks.

          3.5.1.2  Operation

     Place the flux chamber over the surface area to be sampled and work It
Into the surface to a depth of 2-3 on.  Initiate the sweep air and set the
flow rate at 5 L/mln.  Record data at time  Intervals defined by residence
times or T (tau), where U - flux chamber volume (30L)/sweep air flow rate
(5L/mln),  One T then has the value of 6 minutes under normal  operating
conditions.  At T » 0 (flux chamber placement), record the following: time,
sweep air rate, chamber Inside air temperature, ambient air temperature, and
exit gas concentration (real-time analyzer).  The data should be recorded on
the data sheet shown In Figure 3-3.  At each residence time (T, 6 minutes),
the sweep air rate shall be checked (and corrected to 5 L/mln If necossary),
and the gas concentration shall be recorded (real-time analyzer).  After 4
residence times (24 minutes). Initiate sample collection.  At this time,
record the following data: time, sweep air rate, air temperatures Inside and
outside, exit gas concentration, and sample number(s).  If sulfonated
organic compounds are of specific Interest, then meesurenents should be
taken after 10 residence times (1  hour).

     3.5.2  Sample Collection

          3.5.2.1  Real Time

     When real-time monitoring Is required, the sample Is collected by the
real-time analyzer directly from the exit gas  line.

          3.5.2.2  Discrete Sample Collection

     Sample collection should not exceed a flow rate of 2 L/mln.

               Gas Syringes

     Sample collection with syringes should be performed after purging the
syringe three times with the sample gas.  This should be performed without
removing the syringe from the sampling line manifold.  To ensure fresh
sample at each purge, a sampling manifold should be positioned prior to a
real-time analyzer (Figure 3-1).  The analyzer will then draw the sample
past the manifold for sampling.


                                   3-11

-------
                              FIGURE 3-3
   FLUX CHAMBER GAS EMISSION MEASUREMENTS FIELD DATA SHEET
               FLUX CHAMBER EMISSIONS MEASUREMENT DATA
Date.
.Samplers).
Location.
.Zone/Grid Point
Surface Description

Concurrent Activity..
Time






Sweep Air
Rate,Q
(UMIn)






Residence
No.
(Q/V)
0
1
2
3
4
5
Gas
Cone.
(ppmv)






Air Temperature
Chamber Ambient
(C) (C)












Sample
Type/No.






Comments:












Comments:.
                                                                74*24843
                                  3-12

-------
               Gas Canister

     Sample collection with evacuated gas canisters should be performed with
the real-time analyzer replaced by the gas canister (Figure 3-4).

     To collect canister samples, remove the real-time analyzer froa the
exit line sampI Ing manifold.  Securely fasten the canister sampling manifold
to the exit line manifold.  Open the flow control valve 
-------
                              FIGURE 3-4
           STAINLESS STEa  GAS CANISTER AND SAMPLING MANIFOLD
                             (NOT TO SCALE)
                             CAP, SAMPLE ELUTION/
                            PRESSURE MEASUREMENT

                              1/4' S.S. SWAGELOCK*-
 EXIT LINE
FROM FLUX i
 CHAMBER
                                                                VALVE
EXIT LINE
SAMPLING
MANIFOLD
                                  VALVE
                                   (V)
                                                    2L CANISTER

-------
gas canister,  the sample Is then Introduced Into the gas chronatograph
through cryogenic traps.  The dilution factor fs calculated by Equation 3-2
(Section 3.8.3).

     3.5.4  Sampling Strategy

     The following sampling strategy provides an accurate and precise esti-
mate of the emission rate for a total area source through random sampling In
which any location within the area source has a theoretically equal chance
of being sampled.  The sampling strategy described below provides an esti-
mated average  emission rate within 20 percent of the true mean with 95
percent confidence.

          3.5.4.1  Zones

     Based on  area source records and/or preliminary survey data, subdivide
the total area source Into zones If nonrandom chemical distribution Is
exhibited or anticipated.  The zones should be arranged to maximize the
between-zone variability and minimize the wI thin-zone variability.

          3,5.4.2  Grids

     Divide each  zone by an Imaginary grid with unit areas that depend on
zone area size (Z) as follows:

     If Z 1 500 m2, then divide the rone area Into units with areas
     equal to  5 percent of the total zone area  (I.e., 20 units total).

     If 500 m2 <  Z £ 4,000 m2, then divide the zone area Into units of
     aroa 25 m2.

     If 4000 m2 < Z <. 32000 m2, then divide the zone area Into 160
     units.

     If Z > 32000 w2, than divide the zone area  Into units with area
     equal to  200 in2.

     Assign a  series of consecutive numbers to the units In each zone.

          3,5.4.3  Sample Number

     Using Equation 3-3  (Section 3.8.4), calculate the number of units  (grid
points) to be  sampled for the Kth zone  (n«).

          3,5.4.4  Sample Locations

     Using the random numbers table  (Appendix A),  Identify n^ grid points
(units) that will be sampled  In zone K.  A grid  point shall be selected for
measurement only once.   (This  Is not to be confused with duplicate sampling.
Section 3.7.2.2.)
                                    3-15

-------
          3.5.4.5  Emission Rate Calculations

     After sample col lection, use Equations 3-4, 3-5, 5-6, 5-7, and 5-8 to
calculate the measured emission rate (Ec«|) for each grid point (I) In each
zone (K).  Research has shown an emission rate dependency upon the air
temperature Inside the flux chamber.(4)  Through a statistical analysis of
both laboratory and field data, a correction factor for temperature varia-
tions has been developed.  The correction factor compensates a measured
emission rate for chamber air temperature variations from the nominal  cham-
ber air temperature.

     The nominal chamber air temperature can be defined In two ways de-
pending on the purpose of emission rate measurements.   If emission rate
measurements are for an estimate of an area source, then the nominal chamber
air temperature should be the mean chamber air temperature of all the mea-
surements made at that area source.  If emission rate measurements are
compared between area sources, then the nominal chamber air temperature
should be 25°C (298K).

          3.5.4.6  PrelImlnary Estimates

     W_Ith Equations 5-9, 3-10, and 3-11, calculate the zone mean emission
rate  n^, then N^-IK additional
samples must be collected from zone K.  Locate these additional samples
using a random numbers table.  Do not duplicate previously sampled  loca-
tions.

     If N£ » n«. It may be most effective to rezone using the preliminary
measured emission rates as a guide.  If new zones are established, then
these new zones will need to be grldded accordingly  (Section 3.5.4.2).

          3.5.4.8  Final Estimates

     Collect any additional samples and recalculate the emission estimates
for the sample mean  (EK) and variance  (Sj£> lor each zone  (Section 3.5.4.6).
Then compute the overall area source mean  (£> and variance  (S*) for the
total site area using Equations 3-13 and 3-14, respectively.  Determine the
95 percent confidence  interval for each zone  
-------
                                 TABLE 3-3
            TOTAL  SAMPLE SIZE REQUIRED BASED ON  THE PRELIMINARY
                SAMPLE COEFFICIENT OF VARIATION ESTIMATE*
                 Coefficient of                   Number of Samples
              Variation - CV (?)**            Required  (Ng) per Zone K
0 - 19.1
19.2 - 21.6
21.7 - 24.0
24.1 - 26.0
26.1 - 28.0
28.1 - 29.7
29.8 - 31.5
31.6 - 33.1
33.2 - 34.6^
34.7 - 36.2
36.3 - 37.6
37.7 - 38.9
39.0 - 40.2
40.3 - 41.5
41.6 - 42.8
42.9 - 43.9
44.0 - 45.1
45.2 - 46.2
46.3 - 47.3
47.4 - 48.4
48.5 - 49.5
49.6 - 50.7
50.8 - 51.6
51.7 - 52.3
52.4 - 53.4
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
        given Is the sample size required to estimate the average  emission
  rate with 95 percent confidence that the estimate will  be within 20 per-
  cent of the true mean.

**Foc CVs greater than 53.4, the sample size required Is  greater or equal to
  CVV100.
                                   3-17

-------
3.6  Calibration

     3.6.1  Equipment

          3.6.1.1  Flow Meters

     The flow meter should be calibrated against an NBS-traceable bubble
meter before sampling.  The flow meter should have a working range of 2-10
L/mln.

          3.6.1.2  Thermocouple

     Fine wire K-type Insulated thermocouples are recommended for tempera-
ture measurements.  Prior to field use, the thermocouple and readout should
be calibrated against a mercury-In-glass thermometer meeting ASTM E-1 No.
63C or 63F specifications.  The thermocouple should have an accuracy within
±1°C.

          3.6.1.3  Ca11 bratIon Gases

     For checking the concentrations of the calibration gases, use calibra-
tion gases that are documented traceable to National Bureau of Standard
Reference Materials.  Use Traceablltty Protocol for Establishing True Con-
centrations of Gases Used for Calibrations and Audits of Continuous Source
Emission Monitors (Protocol  Number 1) that  Is available from the Environ-
mental Monitoring and Support Laboratory, Quality Assurance Branch, Mall
Drop 77, Environmental Protect Ior Agency, Research Triangle Park, North
Carolina 27711.  Obtain a certification frcm +he gas manufacturer that the
protocol was followed.

          3.6.1.*   Flux Chamber System

     Several tests should be performed to characterize a new flux chamber
prior to use.  These tests should be repeated  If a chamber  Is exposed to
severe conditions such as corrosive gases, extremely high  levels of organic
vapors, or organic I(quids.

               Blanks

     Check the flux chamber for background by placing the  chamber over a
clean Teflon" surface and running a test using ultra high  purity sweep air
and routine operating conditions.  Sample col lection and analysis should be
as previously described  (Sections 3.5.2 and 3.5.3).

               Recovery Efficiency

     Check the flux chamber sample recovery efficiency by  placing the cham-
ber over a flat Teflon" surface containing an  Inlet port at the canter for
Introduction of a cat(oration gas(es).  The calibration gas should be that
used for the on-slte analyzer at a concentration of at  least 1,000 ppnv
(high-level gas).  The calibration gas should be  Introduced Into the chamber


                                   3-18

-------
at a flow rate of no greater than 0.5 L/mfn.  Add ultra high pur/ty weep
air concurrently through the enclosure sweep air Inlet (5 l/mltt) and (Jeter-
nine the concentration exiting the enclosure under routine operating condi-
tions.  Compare the measured concentration to the true concentration (cor-
rected for dilution), and calculate a percent recovery using Equation 3-1.
Results for a variety of volatile organic compounds are presented In Table
3-1.  Results should be within 10 percent of the true concentration.  The
United data characterizing the recovery efficiency for halogenated com-
pounds Indicate an acceptance level tnet may be larger than 10 percent.

               Corrective Action

      If the background levels of tho flux chamber are greater than 10 per-
cent of the measured concentrations or 10 ppmv, whichever fs smaller, then
rerun the blank sample.  If high levels persist, then disassemble the flux
chamber, clean all Internal parts with water and replace those suspected to
be contaminated, and rccrsemble for another blank run.  Repeat above until
satisfactory levels are reached.

      If the recovery efficiency Is below 90 percent for non-halogenated
compounds, then rervn the recovery test.  It low recoveries persist, check
for poor sealing and/or Inlet gas shortcuttlng directly from the Input line
to the exit line and/or mlsadjusted flow rate settings.

     3.6.2  Anelyzers

     The following procedures should be performed at the recommended fre-
quency during the analysis of flux chamber samples.

          3.6.2.1  Real Time

     Real-time analyzers are used more for relative, continuous measurements
than for absolute measurements.  If these analyzers are Intended for abso-
lute measurements, then they should be calibrated according to Section
3.6.2.2.  Real-time analyzers may be used when data qua!Ity requirements are
less stringent (Section 3.1.2).  As such, these analyzers require  less
stringent quality control  practices.

     Each day prior to sampling, a three-point calibration  should  be per-
formed on each analyzer (Section 3.4.3.1.2).  Consider the  calibration
acceptable  If responses are within ±20 percent of the expected response.  If
the responses are not acceptable, then recalibrate the  Instrument.

          3.6.2.2  Discrete Analyzer

     Discrete analyzers are those that are the most relied  upon  for abso-
lute, quantitative data of the analyzers used on site.  As  su~.h, these
analyzers require more stringent quality control practices  (Section 3.1.2).
The calibration  procedure  suggested here is for linear detectors (I.e., FID,
PID).  Compensations for non-linear detectors used for analysis of sulfo-
nated compounds  (flame photometric detectors) must bo made.


                                   3-19

-------
     Prior to each field Investigation, a multipoint calibration Including
zero and at least three upscale concentrations CSubsectlon 3.4.3.2.2) should
be performed to establish the linearity of the analyzer.  The results may be
used to prepare a calibration curve for each compound.  Alternatively, If
the ratio of GC response to amount Injected (response factor) Is a constant
over tha multipoint range «10 percent coefficient of variation, standard
deviation/mean), linearity through the origin can be assumed, and the aver-
age response factor can be used In place of a calibration curve.

     Each day prior to sampling and after every fifth sample, the working
calibration curve (or response factor) must be verified by the measurement
of one or more calibration standards.  If the response for any standard
varies from the predicted response by 20 percent, the test must be repeated
using a fresh calibration standard.  If the analyzer response Is still
unacceptable, a new calibration curve (or response factor) must be prepared
for tha" compound.  A new calibration curve (or response factor) should be
calculated after each verification of calibration using the acceptable
result* of the one or more calibration standards  Injected.

3.7  V allty Control

     3.7.1  Sampling Equipment

          3.7.1.1  Syringes

     Prior to use for sample collection, all syringes should be challenged
with one or more of the calibration standards.  An acceptable response Is
within ±10 percent of the predicted response.  If the response  Is unaccept-
able, then repeat the test.  Alternatively, check for leakage around the
plunger or lock valve by pressurizing the syringe and submerging  It under
water.  Syringes should be checked after every 25 to 30 uses or whenever
leakage Is suspected.   If Teflon" tip plungers are used, then suspect memory
effects after exposure to high levels of organIcs.  In  Instances when memory
effects are apparent, the Teflon" tips should be replaced.

          3.7.1.2  Gas Canisters

     Gas canisters should be cleaned and evacuated before each  use.  The
pressure should be recorded after each evacuation.  Prior to sample collec-
tion, check the pressure and compare It to that recorded after  cleaning.
Acceptable differences are <10 percent of the post evacuation pressure.
Canisters having unacceptable pressure differences should not be  used  for
sample collection.

     To Identify gas canisters and record pressure values, each gas canister
should have a chaIn-of-custody form  (Figure 3-5).  Copies of this font
should be retained for the sampler,  laboratory, and sample control.

-------
                               FIGURE 3-5
         CHAIN-OF-CUSTODY  FORM  FOR GAS CANISTER SAMPLES

                          STAINLESS STEB. CANISTER
                              CHAIN OF CUSTODY
                                               TtNEt
	-	-T08E COMPLETED BY  HBJ) SAMPLE*
SAMPLE CONTROL NUMBER)     	
CANISTER KUNBERi           	
DATE SANPLEDi
Wat/STATION NUMBER]
OVA READING (PEAK)i
AOORCSS/RCF i tern LOOT roNt
(CJCKT/OEPTH/ROOMi
SAKPUR'S INlTIALSt
TASK!
TYPE (C(ROEOIC)i
COMCNTSi
                          AI6IENT  or  POINT SOURCE (specify)*
                     TO BE COMPIETB) .BY LAB  (PART ONE)
       OPEPATION
                             DATE
                                          INITIALS
                                                           COMMENTS
1.  Canister cleaned
2.  Filter cleaned
3.  Canister evacuated
4.  Canister shipped
5.  Canister received
6.  Analysis-coBpletetf
7.  Staple discarded
                       TO BE  COMPUTED BY LAB (PART TM>
      PARAMTTER
                        DILUTION 1    DILUTION 2    DILUTION 3    DILUTION 4
Initial  PTMSUT*
Final PTMSW*
Add UK»  Air
Dilution Factor
FINAL OlUtlon Factor
                                 3-21

-------
     3.7.2  SampI Ing

     These tests should be performed at the specified frequency during use
of the flux chamber.

          3.7.2.1   Sample Blanks

     Sample blanks should be performed once dally or after extremely high-
level samples.  The flux chamber should be cleaned and blanks rerun until
exit concentrations are <10 ppmv or <10 percent of expected concentrations,
whichever Is smaller.

          3.7.2.2   DupI leate Samples

     A minimum of  10 percent of the sampling points should be sampled In
duplicate.  Take the two samples over as brief a time span as feasible to
minimize any temporal variations In the emitting source.

          3.7.2.3  Control Point Samples

     One sampling  location (grid point or unit) In each zone should be
resampled after every ten Individual measurements (or a minimum of once per
day) when an area  source  Is being Investigated.  Preferably, this control
point should be measured at different times during the diurnal cycle (maxi-
mum difference In  ambient temperatures).  These values provide a measure of
temporal  variability of the emission rate from the area source.

     3.7.3  Analytical

          3.7.3.1   Real-Time Analyzers

     Real-time measurements are typically made with portable total hydrocar-
bon analyzers.  Real-time analyses are useful for relative measurements
(I.e., to determine  If steady-state operation of the flux chamber has been
attained or to determine the zoning boundaries).  Each day following cali-
bration, the analyzer should be challenged with the QC gas (Section
3.4.3.1.3).  Analyzer performance should be considered acceptable If the
measured concentration Is within 20 percent of the certified concentration.
If this criterion  Is not met, the QC analysis should be repeated.  If the
criterion Is still not met, then dally calibration should be repeated.

     At the conclusion of each day, the QC gas should be reIntroduced to the
analyzer.  The difference between pretesting and posttestlng responses pro-
vides a measure of upscale drift.  Drifts >30 percent should be flagged and
not relied upon.  If these data are necessary, then resample the grid points
sampled on that day.

          3.7.3.2  Discrete Analyzers

     Each day after calibration, the analyzer should be challenged with the
QC gas (Section 3.4.3.2.3).  Analyzer performance should be considered


                                   3-22

-------
acceptable If the measured concentration Is within 1.0 percent of the certi-
fied concentration.  Jf this criterion !s not met, repeat the QC gas analy-
sis.  If the criterion still cannot be met, then repeat the dally calibra-
tion (Section 3.6.2.2).

     At the conclusion of each day's testing, the QC gas and zero grade gas
should be relntroduced to the analyzer.  The differences between pretesting
and posttestlng values provide a measure of upscale and zero drifts.  Dally
drift results that show >20 percent should be flagged and tests repeated If
determined necessary.

          3.7.3.3  Analysis of Integrated Samples

     Quality control for the analysis of Integrated samples should  Include a
minimum of 10 percent analytical blanks and 10 percent duplIcate analysts.
It Is recommended that duplicate samples each be analyzed  In duplicate to
provide Information on analytical as well as sampling variation.  A con-
venient technique  Is the use of a nested sampling scheme as shown In Figure
3-6.

3.8 CALCULATIONS

     3.8.1  Definitions

      A m surface area enclosed by the flux chamber (0.130 m2)

      a * number of carbon atoms per compound molecule

     Ct * confidence Interval for the area source emission rate mean
          (±ug/mln-m2)

          confidence Interval for the zone K emission rate mean  (±ug/mln*m^)

          measured concentration of species  I (ppmv) corrected for  dilution

    CJT « theoretical concentration of species  I  (ppmv)

        * measured concentration for point I  In zone K, total NWC  (pprav-C)

        « coefficient of variance for zone K  1%)

      E * mean emission  rate for tha area  source  (ug/mln-m2)

      K * zone K emission  rate mean  (ug/nln^m2)

        * measured emission  rate for point  I  In zone K  (ug/mln-m2)

        « measured emission  rate for point  I  In zone K  (ug/mln-m2)  corrected
          for temperature  variations

     Mf « molecular weight of compound  (g/mole)


                                   3-23

-------
                                          FIGURE 3-6
                                    NESTED SAMPLING SCHEME
                                  Event
           Sample 1
                                 Duplicate
                                 saapleB
                                 (sanpling
                                 variability)
Analysis 1-1      Analyst* 1-2
Analysis 2-1
Analysis 2-2
Duplicate
analyaia
(analytical
variability)

-------
      N « total  number of grid points sampled In the area source (alt  zones)
      K " final  number of grid points (units) sampled In zone K
     nj( « Initial  number of grid points (units)  sampled fn zone K
      P » atmospheric pressure (atm)
      0 » sweep air f lor rate (L/»!n)
      R » gas constant (0.08205 L-atm/mol «K)
      S « standard error of the overall area source emission rate mean
     Sg " zone K emission rate variance
      T * temperature of laboratory where analyzer Is located (K)
   TEMP » temperature of the flux chamber sir <°C)
 f-Q.025 * tne 97.5th percentage pofnt of a student's t-dfstrfbutfon (Table
          5-4)
      V * volume enclosed by the flux chamber (301)
     WK » the fraction of the site represented by the zone K (zone area
          (m2)/slte area (m?))
    YKJ m measured concentration for point I  In zone K, total tWC (ug/L)
      o« parameter defining the level of confidence. 100X1-20) percent
      Y » total number of zones In the total  area source
      P- confidence Interval (f)
      T » measure of residence time Y/Q (mln)
     3.8.2  Percent Recovery
     The percent recovery measurements used to characterize the flux chamber
performance are calculated accordingly:
                     Percent Recovery « (CIM/CIT) x  '00               (5-1)
where:  CJM * the measured concentration of species I (ppmv) corrected for
              dilution as follows;
                              C,M « (1/DF) x C                       (5-ta)
                                   3-25

-------
                                 TABLE 3-4
                     TABULATED VALUES OF STUDENT'S »t"
Degrees of          Tabulated                Degrees of           Tabulated
 Freedom*          "t" Value**                Freedom*          »t" Value**
     1               12.706                      21                2.080
     2                4.303                      22                2.074
     3                3.182                      23                2.069
     4                2.776                      24                2.064
     5                2.571                      25                2.060

     6                2.477                      26                2.056
     7                2.365                      27                2.052
     8                2.306                      28                2.048
     9                2.262                      29                2.045
    10                2.228                      30                2.042

    11                2.201                      40                2.021
    12                2.179                      60                2.000
    13                2.160                     120                1.980
    14                2.145                       »                1.960
    15                2.131

    16                2.120
    17                2.110
    18                2.101
    19                2.093
    20                2.086
 •Degrees of freedom (df) are equal to the number of samples collected less
  one.

••Tabulated "t" values are for a two-tailed confidence Interval  and a
  probability of 0.05 (the same values are applicable to a one-tailed
  confidence Interval and a probability of 0.025).
                                  3-26

-------
              where C Is the sample concentration (ppav) and
                    OF  Is the dilution factor calculated as follows:

                                 S1/(S2+S1)

              where 8j  Is the flow rate of the trace gas and
                    $2  fs the sweep air flow rate

              the true concentration of species  I, gas cylinder value (pp«v)

     3.8.3  Calculation of the Dilution Factor Involved In Gas Canister
            Analysis

     Analyzing the gas canisters requires pressurizing the canister with
nitrogen.  This  Intrloduces a dilution which must be accounted for as
fol lows:
                         DF « (P2 - Pi)/(14.7 + P3)                  C3-2)

where:  P\ » the measured pressure after cleaning and canister evacuation
             prior to sampling (pslg)
        ?2 * the measured pressure after sample collection (pslg)
        ?3 = the measured pressure after pressurizing with nitrogen (pstg)

The temperature Is not required If all pressure Measurements used In this
equation are performed In the same laboratory (I.e., same temperature) after
the canisters have thermally equilibrated.

     3.8.4  Area Source Emission Rate Equations

     The number of units or grids (n«) to be sampled per zone (K) is depen-
dent upon the zone area as follows:
                    nK « 6 + 0.15  Varea of zone K  (m2)             (3-3)

     Flux chamber measurements taken at each of the  n« sampling units are
measured In terms of ppmv-C.  To calculate an emission rate representing the
sampled unit, the measured concentration (Cgf) oust  first be converted from
ppmv-C to ug/L as fol lows:
                          YK| - (P/CR'TWMf/aXfci                   (3-4)

where P Is pressure (atm), R Is Ryd berg's gas constant  (L-atm/mole-K), T It
the flux chamber air temperature (K) (Section 3.5.4.5), Ml  Is the species*
molecular weight Cg/mole), a Is the number of notes of  carbon per note, CKJ
Is the measured concentration of sampled unit I  In zone K  (ppmv-C), and Y«|
Is the measured concentration of sampled unit I  In zone K  (ug/L).

     The emission rate for point I In zone K (Ej<|)  Is then  calculated using
the converted gas concentration (ug/L) as follows:
                                   3-27

-------
                                    (0-YKj)/A                        (3-5)

where Q Is the flux chamber sweep air flow rate (L/mln), A Is the enclosed
surface area measured (m2), and EK| Is the emission rate measured for point
I In zone K (ug/m^-mln).

     Prior to calculating a mean emission rate for the zone neasured, the
emission rates measured for 1he Individual sampling points need to be cor-
rected for fluctuations In chamber air the temperature (I.e., atnosphere
temperature).

     The approach used to develop the correction procedure Involved devel-
oping an empirical  equation to predict emission rates as a function of
chamber air temperature. (4)  The resulting emission rate equation was then
used to define the correction factor (C), as follows:

                                C - EFs/EFa                          (3-6)

where:  EFS =• emission factor calculated at the nominal chamber air tempera-
              ture (Section 3.5.4.5)
        EFa * emission factor calculated at the measured chamber air tem-
              perature

     Both EFS and EFa are predicted using the proper chamber air tempera-
tures and the following equation:

                   EF(s or a) * exp CO.Ot3(TEMP(s or a)):            (3-7)

where TEMP Is measured In °C.

     The measured emission rate (EF«|) Is then corrected to the nominal
emission rate (EFCK|) accordingly:

                                EcKi = C-EFKi                        (3-8)

     The above procedure has a significance level (I.e., probability that
the correlation between chamber air temperature and emission rate measured
ts due to chance) of 0.4 percent.  The standard error of the coefficient  In
Equation 3-7 Is ±0.003.

     The mean emission rate for each zone  Is then calculated accordingly:


                              E  . -L  ? E                          (3-9)
where E^i Is the temperature corrected emission rates  (Equation 3-8) and
Is the number of points sampled  In zone K  (Section 3.9.4,7).

     For each zone (K) sampled, the zone variance ($£) and coefficient of
variance (CVg) must be determined as follows:


                                   3-28

-------
                                      (ECKI' - "K
                                  1-1
(3-10)
                             cvK - too • SK/EK                       (5-n>

where OK, E^m, and EK are defined In Equations 3-3, 3-8, and 3-9, respec-
tively.  The standard deviation 
require* from a given zone.  If NK > n«, then Ng-ng additional samples oust
foe collected from zone K.

     Collect any additional samples and recalculate the emission estimates
for the jone mean (EK) and variance  (3£) using Equations 3-9 and 3-10,
respectively.  If ty-ry^ additional samples  were collected., then use ^
samples Jnstead_of n^ In the recalculations.  The overall area source mean
emission rate (E) is then calculated as follows:

                                    Y
                               E«   E  vL                          (3-13)
where EK  Is defined by Equation  3-9, WK  's tne  fraction  of  site covered by
zone K  (zone area/site area) and-Y  Is the total  number of zones sampled.
                                    3-29

-------
     Ffnaffy, calculate the variance of the overat(  area source mean
ami the confidence Intervals for each zone K (C!K)  and area source (CD
emission rate mean as follows:
                          CIK
                             Cl - E ± t0.025'S
                                   3-30

-------
                                 SECTION 4

                                 CASE STUDY
     To supplement the protocol presented  In  Section 3,  a case  study will be
reviewed.  This study will Illustrate an actual  application of  the  protocol.
Calculations and pertinent decisions will  be  presented.

     The site, referred to as the Bon I fay  Spill  Site, was the scene of an
accidental spill of 5500 gallons of JP-4 aviation  fuel.  The spill  site
occurred near the Intersection of two roads.   The  majority of the contami-
nated soil was excavated.  The residual product  extended over two areas, 30
feet of unvegetated right-of-way along the highway and Into a pine  forest
containing dense underbrush.

     The free surface of the water table was  three feet  below the land sur-
face.  The thickness of the uneonso11 dated sediments that comprised the
water table aquifer at the site ranged from 20 to  50 feet.  The state
aquifer underlaid this sediment layer.  Contamination of the free water
table surface was expected since It was only  3 feet below I and surf ace.
However, the state aquifer was not considered threatened due to the contami-
nants net upward hydraulic gradient.

     A preliminary survey was performed to define  the contaminated  area.  A
series of ten borings Indicated that the contaminants had percolated down-
ward to the capillary fringe and moved laterally down gradient.  A  lens of
product several  Inches thick was detected  at  a depth of  seven feet  below
land surface.  The estimated extent of contamination at  the time of the
survey study was 7,000 square feet (Figure 4-1).

     Results from a preliminary emissions  survey performed with a portable
real-time analyzer (organic vapor analyzer) held a few Inches above ground
were used to divide the area source Into emission  zones  for grlddlng pur-
poses.  The survey Indicated only one zone was present,  and the site was
gridded accordingly.  The field data for the  survey Is shown In Table 4-1.
The grid system used Is shown In Figure 4-2.

     Surface emission measurements were made  Initially at eight sampling
grid points.  The protocol, at that tine,  called for the Minimum number of
samp I Ing points per zone, n«, to be selected  according to the following
equation (note,  this equation has since been  changed to  Equation 3-3).
                            6 + 0.1-yj zone area (m2)
                                   4-1

-------
                        FIGURE 4-1
SCHEMATIC DIAGRAM OF BONIFAY SPILL SITE, MONITOR WELLS, AND
    EXPLORATORY BORINGS {BROWN AND KIRXNER, INC., 1983)
     o
     o
     t-
                      APPROXIMATE EXTENT OF
                                AREA
                               PROJECT AREA OF
                               SUBSURFACE CONTAMINATION
                                       .3
KEY
 0 AREA OF MAXIMUM PRODUCT ACCUMULATION
    IMMEDIATELY AFTER SPILL

2m WATER-LEVEL MONITOR WELL AND
    IDENTIFICATION NUMBER

 P62 EXPLORATORY BORING AND
    IDENTIFICATION NUMBER
                            SCALE]
                                         M
                            75   WO FEET
                          4-2

-------
                                              TABLE 4-1
                                 DATA SHEET FOR UNDISTURBED SURFACE SURVEY
Operators
Heathen

Grid
Point
01
02
04
08
14
20
21
22
23
24
25
16
18
19
Well P-3
Well P-4
Well P-7
Well P-7

BME
Temperature - 45°F. Light t^reez
-------
                               FIGURE 4-2
      SCHEMATIC DIAGRAM OF SAMPLING GRID AT BOMIFAY SPILL  SITE
X
     ~l
                                 KEY
                                     POINT RANDOMLY SELECTED FOR SAMPLING
                                     USING EMISSION ISOLATION FLUX CHAMBER







R
|
0
i
Ul
s
W




/^/\
f f \
' 01 ,/
///
///.
/&
/or/i o
f%
p-i

O MONITOR WELL AND IDENTIFICATION NUMBER
P3 AREA OF MAXIMUM PRODUCT ACCUMULATION
^ IMMEDIATELY AFTEH SPILL
\ 7000 FT* SITE AREA TOTAL
\ 280 FT3 ORID AREA (4H OF TOTAL)
\ SCALE: 3/4' «
' \
/V^ 05 06* 0\
^22z..--.—_i____j
O|

\l
08* 1 09 10 11
1
-X 	 1__
_ 	 _j 	
\ :
14* ! IS* 16 17
\*
1
.P-2, 	 ^-5.._
?v
20 JNV21 22 23*
! X


w
^
\ 1
N 1
12 | 13*
\ |
...A ..
18 \| 19*
-poe-4--pi7
24 J\ 25*
S
^r \






5
- I
                               o
                               P-3
P-4
 O
           TO
         BONiFAY
                               4-4

-------
For the single zone  at Bonffay, this reduced to»
                        n« >. 6 * 0.1 -^650 m2  - 8.5
     The 8 locations were selected through the use of a random number table.
Appendix A.   Grid point 08 was selected to be the control  point (I.e., a
sampling point to be repeated each day) since It was believed that emissions
would be of  the largest magnitude at that location.  At each sampling loca-
tion a gas syringe sample was taken for on-slte analysts.   At several  sam-
p| Ing locations,  a gas canister was collected tn addition  to the syringe
samples for  off-site detailed analysis.  A sample field data sheet Is shown
In Figure 4-3. The results of the emission rate measurement are given In
Table 4-2, and a  sample calculation Is given In Table 4-3.

     Total non-methane emission rates were calculated for  each grid point
based on the on-slte analytical date.  These emission rates are also pre-
sented In Table 4-2.  The variation (spatially and temporally) In measured
emission rates over the extent of the contaminated area was large (93.8
percent coefficient of variation).  Replicate sampling at  tfie control point
allowed an estimate of the emission rate temporal variability.  The temporal
variability was also large (96.0 percent).  The major contributor to the
variation In measured emission rates from point-to-point can, therefore, be
attributed to day-to-day (temporal) variability.  The spatial variability
was then estimated to to negligible.  Using Table 3-3 to determine the total
number (%) of samples to be collected based upon the spatial variability
shows that at least 17 samples should have been collected.  Although addi-
tional samples were required to be collected, sampling was terminated due to
rain.  It was real Ized that the lack of a complete data sot would then
result tn a larger emission rate confidence  Interval.

     Using the following equation, the 95 percent confidence Interval (Cl)
for the zone emission rate was estimated.

                          Cl « ER t1,

where ER  Is the mean emission rcte of the zone, s2  Is the zone variance, NK
Is the total number erf sites sampled, and t0tj)25 is obtained from Table 3-4.
The 95 percent confidence Interval for the zone emission rate  Is from 11.3
ug/mln**2 to 55.2 uy/mln-w2.
                                   4-5

-------
                                 FIGURE 4-3
   FIELD DATA SHEET FOR ISOLATION FLUX CHAMBER SAMPLING AT GRID POINT 08
Date   1-13-84
                                  .    Sampler s_

Location   Bonlfay Spill Site,  Grid  Point 08
KtE
Concurrent Activity     None

Surface Description     Sand
Time
0858
0902
3906
0910
0914
0913
0933

Purge Air
or
FlovraCe
4.86 L/mln
4.86 L/min
4.86 L/min
4.86 L/min
4.86 L/min
4.86 L/mln
4.66 L/min

Residence
Time Num-
ber (T)
0
1
2
3
4
5
9

Temp. "F
Surface Air
46 48







Gas Data
OVA J>pnv KNU ppmv
0.15
0.16
0.16
4.0 0.16
4.0
-
4.0 0.16

Air
Sample
Number





Canister B003
Gas Svrtnge
B002
Comments      OVA background - 4 ppm.  Some trouble with syringe needle
              plugging
                                      4-6

-------
                                              TABLE 4-2
              RESULTS OF GC ANALYSIS OF GAS SYRINGE TAKEN DURING aUX CHAMBER SAMPLING
Grid
Point
4
6
8
8
8
14
15
19
23
25
Sample
No.
8004
801 7-A
8001
B002
8016
B006
8013
B009
8011
8008
Total NMHC Syringe
r* — «..»
(ppmv-C)
1/13/84 1.0
1/14/84 6.8
1/12/84 2.0
1/13/84
1/14/84
1/13/84
1/13/84
1/13/84
1/13/84
.0
.0
.0
.0
.0
.0
1/13/84 8.8
(ug/L)
0.62
4.2
1.2
0.62
0.62
0.62
0.62
0.62
0.62
5.4
Sweep Air
Rate
(L/mln)
2.60
2.60
5.00
4.86
2.60
2.60
2.60
2.60
2.60
2.60
Atmospheric
Temperature
•p oC
47
51*
42
48
52
45
51
51
50
53
8.3
10.6
5.5
8.9
11.1
7.2
10.6
10.6
10.0
11.7
Average Emission
Rate
(ug/mz*mln)
14.4
72.6
79.6
24.9
10.0
16.6
10.7
10.7
11.5
81.4
Variability
        Spatial and Temporali
        Mean
        Standard Deviation
        CV(Jf)

        Temporals (Control Points)
        Mean
        Standard Deviation
33.24
31.17
93.8
38.2
36.6
96.0
•Surface temperature used rather than the chamber air temperature due to a large temperature
 differential not present In the other measurements.  This Is suggestive of an error In chamber air
 temperature measurement.

-------
                                 TABLE 4-3
   SAMPLE CALCULATIONS OF THE EMISSION RATE FOR GRID POINT 08 ON 1/13/84
Concentration Conversion:

     Y, - (P/(R-T))(MW/a)(C|)                                 (Equation 3-4)
where:  P
        R
        T
       MM
        a
1 atm
0.08205 L-a-hn/mole-K
282.6K (average area site air temperature)
86.18 g/fflole (referenced to hexane)
6 moles of carbon/mole of hexane
1.0 ppmv-C
                       1 atm                 86.18 g/mole
       YI *                       •  .       x ———————- x  1.0 DDfliv-C
         '   (0.08205 L-atm/mole-KM282.6K)   6 mole C/mote

       YI « 0.6194 ug/L

Emission Rate (uncorrected)

     EI - (Q'Y|)/A                                            (Equation 3-5)

where: 0 " 4.86 L/mln
      YI » 0.6194 ug/L
       A * 0.130 m2

            4.86 l/mln-0.6194 ug/L

      El *        0.130 m2

      E| » 23.15 ug/mln-mZ

Emission Rate Correction Factor

     EFS - expC0.13(TEMPs)]                                   (Equation 3-7)

where:  TEMPS » 9.45*C (nominal chamber air temperature *C)
          EFS « emission factor at  nominal chamber atr temperature

     EFS - exp (0.13*9.45)
     EFS - 3.416


                                                              (Continued)
                                   4-8

-------
                                 TABLE 4-3
                                (Continued)
     EFa " exp[0.13(TEMPa)D

where:  TEMPa « 8.9*C (measured chamber air temperature *C)
          EFa » emission factor at the measured chamber air  temperature.

     EFa - exp(0.13-8.9)
     EFa « 3.160

     C « EFs/EFa                                              (Equation 3-6)
     C » 3.416/3.180
     C » 1.074

Emission Rate (corrected for temperature variation)

     EC| » C'Ej                                               (Equation 3-8)
     Eg] « 1.074-23.15 ug/mtn-m2
         - 24.86
         • 24.9 ug/mln-n»2
                                   4-9

-------
                                 SECTION 5

                            ADDITIONAL  INFORMATION
     For further Information on vapor/liquid equilibria (VIE) for organic
systems, the  following reference  Is suggested.  The Intent of this bibliog-
raphy was to  provide  a ready listing of the references for data on VLE.

     Nelson,  T.P.,  N.P.  Meserole, Annotated Bibliography of Published
     Material  on Vapor/Liquid Equilibria. EPA, July 1983.


     For further Information on the selection of the flux chamber enclosure
method for direct measurement of  gas emission rates from contaminated soils
and/or groundwater, the following reference Is suggested.

     Radian Corporation.   Soil Gas Sampling Techniques of Chemicals for
     Exposure Assessment,  Interim Report.  EPA Contract No. 68-02-3513, Work
     Assignment  32, August 1983.


     For further Information on the actual field applications of this tech-
nique, the following  references are suggested:

     Radian Corporation, Soil Gas Sampling Techniques of Chemicals for
     Exposure Assessment:  Tustln  Spill Site Data Volume. EPA Contract No.
     68-02-3513, Work Assignment  32.  July 27, 1984.

     Radian Corporation, Soil Gas Sampling Techniques of Chemicals for
     Exposure Assessment,  BonI fay Spill Site Data Volume. EPA Contract No.
     68-02-3513, Work Assignment  32, 1984.


     For further Information on the validation of the flux chamber technique
for emission  rate measurements on soil surfaces, the following reference Is
suggested!

     Ktenbusch,  M.R., D. Ranum, Validation of Flux Chamber Emission
     Measurements on  Soil  Surfaces. EPA Contract No. 68-02-3889, Work
     Assignment  18, December 1985.


     For Information  concerning the emission process  Including diffusion and
adsorption, the  following  reference  Is suggested:
                                    5-1

-------
Manos, C.G., Jr., Effects of Clay Mineral Organic Hatter Complexes on
YOG Adsorption, Draft Report.  EPA Contract No. 68-02-3889,  Work
Assignment 18, October 3, 1985.

Radian Corporation.  Soil Gas  Sampling Techniques of Chemicals for
Exposure Assessment; Laboratory  Study of Emission Rates from Soil
Columns, Draft Final Report,   EPA Contract No. 68-02-3513,  Work
Assignment 32, October 1984.
                               5-2

-------
                                  REFERENCES

1.   Farmer,  W.J.,  M.S.  Yang,  and J.  Letey.   Land  Disposal  of Hexachloro-
     benzene  Wastes—Controlling Vapor Movement  In Soil.   EPA-600/2-80-119,
     Municipal  Environmental  Research Laboratory,  Cincinnati, Ohio, August
     1981.

2.   Shen, T.T.   Estimating Hazardous Air  Emissions from  Disposal Sites,
     Pollution  Engineering, August 1981.

3.   U.S.  EPA,  Office of Solid Waste.  Guidance  Document  for Subpart F, Air
     Emission Monitoring,  Land Disposal Toxic Air  Emissions Evaluation
     Guideline.   December 1980.

4.   Klenbusch,  M.R,  D.  Ranum.  Validation of Flux Chamber Emission Measure-
     ments on Soil  Surfaces.   EPA, EMSL, Contract  No.  68-02-3889, Work
     Assignment 18, December  1985.

5.   Radian Corporation, Soil  Gas Sampling Techniques  of  Chenlea Is for
     Exposure Assessment,  Interim Report,  EPA Contract No.  68-02-3513, Work
     Assignment 32, U.S. EPA  EMSL, EAO, August 1983.

6.   Adams, O.F., M.R. Pack,  W.L. Bamesberger, and A.E. Sherrard,
     "Measurement of  Blogenlc Sulfur-Containing  Gas Emissions from Soils and
     Vegetation." In: Proceedings of  71st  Annual APCA  Meeting, Houston, TX,
     1978, p. 78-76.

7.   Adams., D.F., Sulfur Gas  Emissions from Flue Gas Desul fur Izatlon Sludge
     Ponds.   J.  Air Pol. Contr. Assoc. Vol. 29,  No. 9, p. 963-968, 1979.

8.   Denmead, O.T.   Chamber Systems for Measuring  Nitrous Oxtde  Emission
     from Soils In the Field.  Soil Sciences Soc.  of Am.  J.,  43,  p. 89-95,
     1979.

9.   Bat four, W.D.  and C.E. Schmidt,  Sampling Approaches  fcr  Measuring
     Emission Rates from Hazardous Haste Disposal  Facilities.   In:
     Proceedings of 77th Annual Meeting of the Air Pollution  Control
     Association, San Francisco, California, June  1984.

10.  Zimmerman, P.   Procedures for Conducting Hydrocarbon Emission
      Inventories of Blogenlc Sources and Sane Results  of  Recent
      Investigations.  In: Proceedings of 1977 Environmental Protection
     Agency Emission  Inventory/Factor Workshop,  Raleigh,  NC,  1977.
                                    R-1

-------
                                 APPENDIX A

                        SaECTION OF A RANDOM SAMPLE
     An illustration of the method of use of tables of randan numbers
follows.  Suppose the population consists of 87 I tans, and we wish to select
a random sample of 10,  Assign to each Individual a separate two-digit
number between 00 and 86.  In a table of random numbers, pick an arbitrary
starting place and decide upon the direction of reading the numbers.  Any
direction may be used, provided the rule Is fixed In advance and Is Indepen-
dent of the numbers occurring.  Read two-digit numbers from the table, and
select for the sample those Individuals whose numbers occur until 10 Indi-
viduals have been selected.  For example, In Table A-1, start with the
second page of the table, column 20, line 6, and read down.  The 10 Items
picked for the sample would thus be numbers 38, 44, 13, 73, 39, 41, 35, 07,
14, and 47.

     The method described  Is applicable for obtaining simple random samples
from any sampled population consisting of a finite set of  Individuals.  In
the case of an Infinite sampled population for the target  population of
weighings as comprising all weighings which might conceptually have been
made du-Ing the time while weighing was done.  Wo cannot,  by mechanical
randomization, draw a random sample from this population,  and so must recog-
nize that we have a random sample only by assumption.  This assumption will
be warranted  If previous data  Indicate that the weighing procedure  Is  In a
state of statistical control;  unwarranted If the contrary  Is Indicated; and
a  leap  In the dark  If no previous data are available.

-------
                                            TABLE  A-1
                              SHORT TABLE  OF RANDOM NUMBERS
46  96   85  77  27  92   8«  M  45  21   89  91   71   42  M   64   58  22  TS  SI   74   »1  48  4f   II
44  19   15  32  63  55   87  T7  3J  29   43  00   31   34  84   05   72  90  44  27   TJ   22  07  62   IT
3439806224338167281I3479263S342309940080SS31632791
74  97   80  30  (i  VI   71  30  01  M   47  45   89   70  74   13   M  90  SI  27   «1   34  63  87   44
22  U   81  60  M  3S   33  71  13  13   72  M   16   13  SO   M   U  SI  29  U   30   93  45  M   29

40  M   M  40  03  47   24  60  Of  21   M  U   00   0$  II   52   M  40  73  73   57   68  3<  33   91
52  33   7«  44  54  IS   47  TS  T»  T3   78  19   87   M  U   47   43  02  62  03   42   0$  32  Si   02
37  59   20  40  M  )V   82  24  19  90   80  87   32   74  59   84   24  49  79   17   23   7$  83  42   00
11  02   55  57  43  84   74  U  22  47   19  20   IS   92  53   37   13  7J  54  (9   M   73  23  39   07
10  33   79  2«  34  M   71  33  89  74   68  48   23   17  49   18   81  OS  $2  85   70   05  73  11   17

17  59   28  25  47  89   II  65  «5  20   42  23   96  41  64   20  30  89  87   64   37  93  M  M  3S
93S07S2009I8S4S46B
-------
TABLE A-1
(CONTINUED)
OS
37
a
42
OS
57
78
rt
67
W
a
16
u
98
03
M
OC
M
41
84
24
87
82
«7
32
23
12
64
44
«2
M
46
«T
28
83
M
22
29
71
27
1C
90
01
45
48
U
97
»
OS
u
7S 28 81 it
78 (7 31 06
46 T2 OS SO
I» 47 76 30
M I> 84 M
U
63
I>
2C
20
62
60
27
72
20
82
SI
47
33
50
45
02
1&
69
87
63
07
76
92
74
to
16
SI
SI
83
36
7S
S8
>S
SI
02
12
67
23
C
76
90
06
26
10
«
41
M
41
a
63
16
W
76
30
60  44   18  41   23  74   T3  51  72   90  40   52  95  41  »   89  48  98   27  J3  81   33  83  *2  94
3280647S91M09406489299946356991S0737S9290S6829324
79MS37778Mt237438271007821656S884582447893227309
4$  (3   23  32   01  09   46  36  43   M  37   IS  3$  04  88   79  83  S3   1*  13  91   S>  61  31  87
2060»7432141S422727799818330441S9024S17S663499t040

67914483432S$4332>a0995327S4I9807432S39607UM6I98
845O76938435t84S3781474492576659M144339249454644«
M  73   38  38   23  M   10  95   U   01   10   01  59  71   55  99  24  U  31  41  00  T3  1>  80   62
55  11   SO  29   17  73   97  04   20   39   20   22  71  11   4*  00  16  10  12  35  09  U  00  8f   OS
23543387929204497394S7S3S7M9309M87U«T4629SOrr85

41  43   67  7»   44  57   40  29   10   34   58   63  SI  18   07  41  02  39  79  14  46  68  10  01   61
03  97   71  72   43  27   3*  24   S»   88   «   87  26  31   11   44  28  S8  99  47  83  21  35  22   88
90  24   83  48   07  41   56  68   11   14   77   75  48  68   08  90  89  63  87  00  06  18  63  21   91
98  98   97  42   2T  II   80  51   13   13   03   42  91  14   51   22  IS  48  67  S2  09  40  34  60   M
74  20   94  21   49  9«   SI  69   99   U   43   76  SS  81   M  11  88  68  32  43  08  14  78  OS   34

94  67   48  87   U  84   00  8S   93   S«   U   99  21  74   84  13  S4  41  90  M   30  04  19  68   73
58  18   84  82   71  23   $6  33   19   2S   65   17  90  M   24  91  75  M  14  83   86  22  70  M   W
31  47   28  24   88  49   28  69   78   62   23   «S  S3  38   78  &S  87  44  91  W   91  62  78  09   20
45  62   31  06   70  92   73  27   83   57   15   (4  40  57   M  54  42  3S  40  93   55  82  08  78   87
31  49   87  12   27  41   07  91   72   64   63   «2  OC  «   82  71  28  34  4S  31   99  01  03  3S   76

69  37   22  »   46  10   75  83   62   94   44   6S  46  23   65  71  69   20  89  12   16  56  61  70   41
93  67   21  54   98  4*   S2  53   14   M   24   70  2S   18   23  a  56   24  03  M   11  06  46   10   23
77  54   18  37   01  32   20  18   70   79   20   U  77  89   28  17  77   IS  S2  47   IS  30  35   12   7S
37  07   47  79   60  75   24  IS   31   63   25   93  27  66   19  S3  52  49  98  45   12  12  06  00   32
7208710173463960375322252084M0203B6S383S04M6994

55  12   48  46   72  SO   14  «   47   47   84   37  32  84   82  64  »7   13  69  M   20  09  80  41   75
692498907029342S33231269MS0389384322896036S70901t
01  86   77  18   21  91   «  U   84   C$   4»  75  21  94   SI  40  SI   S3  36  39   77  69  06   25  «
51  40   94  0«   80  61   M  23   46   28   11   U  U  M   60  65  0«   63  71  06   1J  SS  OS   32  56
SI7I02«SM296727440T6723202StZe97S962134172T07107

3375685I003354I584»42*S0166S12615443S4U63377497S9
S860J745«20»»5M18»3422»17804t7988020M3l93139t30
721)129S3287993283«S401792S722t8MT91«23S3S4M   J*   09  85  24   59   46  OS  91   U  38  62   SI  71  47  37  38
81  94   71  90   47  41   M  3«   33  93   04  90  *»   72  85  23  0   *  TO   SI   56  03  23  84  80
44  «2   20  81   21  57   57  M   00  47   36  10  87   22  4S  72  03   31  75  23  38  38  54  77  97
63  91   i:  IS   08  02   1*  74   it   ?»   21   S3  «   41   77  IS  07   »  87   11   1»  25   62  19  30

29337760290»yM42»07IS4067S629S87J84041IS43116S3
54  13   39  19   2»  64   tr  7J   71   41   T8  03  24   02  M  M  69   74  74  28  08  98  84  08  23
75  18   85  «4   84  93   &5  «   08  84   IS  <1  S7   84  4S  U  70   13  17  «  47  80   10  13  00
34  4?   17  08   79  03   92  8*   IS  42   93  48  27   37  99  98   81   M  44   72  04  95   4Z  31   IT
29  
-------
                                            TABLE A-1
                                           CCOWTINUHJ)
«S  44  38  M  M   II  20  89   a  S2  45  41   01  Tl  SS   14  II  OS  II   W   74  94  SO   M  07
14U14572ll2U8l«704tS«»052723Mt423S20721l7I3S2ll
OS2S735123929S05S412»44<«13392133091731130442171tt
9t211024793»llOIM20l2M479«0704*29S01S  CT  9*  27   70  71  04   43  IS  44   II  75  11   TO   53  21  <0   71  30  n  54   21   02  42

U  14  «  4<  15   SI  91  23   U  24  71   19  «7  11  79   90  S3  47   M   32  U  «   97   10  17
«322MU10020S0347934<702S27M329l4I45M3*M91Tt79
42  U  20  M  19   U  M  91   21  M  U   91   74  92   31   97  tt  24   20  55  19  54   «  «  M
»  90  7«  51  51   49  25  51   21  Cl  55   SS  73  10   22  M  79  23   *0  03  SI  11   00  11  17
20  12  t7  40  25   45  94  3S   11  
-------
                                             TABLE  A-1
                                            (CONTINUED)
24  *1   04  U  98   24  93   58   43  M  56  26   24   45  *S  »I   42  W  67  42   II  74  TT   II  41
75  55   54  29  67   02  81   01   47  54  08  81   34   00  79  42   38  52  14  U   38  44  S9   41  97
41  71   80  54  37   73  34   U   74  U  91  64   82   41  02  74   12  34  71  38   43  72  84   34  27
04  19   43  35  54   M  00   41   47  44  63  13   27   SO  18  75   14  72  40  90   02  45  67   82  IS
44  IS   52  42  22   91  22   96   38  41  03  27   IS   67  26  34   81  75  U  82   94  U  62   08  94

10  80   17  17  83   05  31   23   08  07  40  00   80   44  U  70   II  31  73  OS   41  41  47   14  a
40  42   27  55  71   82  88   42   71  51  58  4t   58   75  38  23   57  04  14  19   44  M  Ot   U  tt
ff5  57   21  21  25   12  05   41   70  28  03  M   91   37  (4  4*   *>  48  St  «0   8*  71  35   83  05
57  27   M  M  M   88  M   70   8*  59  41  84   08   32  31  75   II  1»  41   11   28  41  71   79  28
80  54   W  4$  13   83  78   78   71  38  8)  51   U   47  35  84   It  M  M  88   >1  22  47   24  84

44  51   T5  51  08   17  43   53   31  09  60  34   34   81  9S  M   01  M  37  13   24  M  75   29  21
55  42   48  76  SO   13  89   69   00  05  99  45   82   01  53  86   68  81  34  50   75  20  17   94  47
80  SO   67  83  01   97  78   21   44  34  62  43   02   64  38  13   60  26  32  34   81  43  17   56  41
03  64   65  44  02   75  41   13   91  28  82  97   57   38  «9  27   26  97  34  44   26  12  00   M  24
14  53   75  37  91   43  95   IS   13  2(  33  27   45   48  33  80   80  21  19  71   04  V  83   58  32

01  M   43  34  30   71  24   75   92  73  07  81   13   35  48  88   42  80  84   69   M  25  73   92  98
39  38   79  42  17   77  99   J5   32  85  13  35   48   49  80  83   59  04  34   94   04  03  41   85  02
74  9«   24  94  89   54  64   29   35  M  50  4$   (5   SO  28  62   45  80  81   95  07  9»  57   10  54
21  18   54  55  77   48  38   33   88  Si  21  56   18   93  «  94   24  80  97   03   7»  39  73   87  70
S8SI99S39S736077«10476S978W949907S391959940S441T»

44  98   27  95  19   22  29   41   54  78  83  48   49   82  79  79  20  00  24   40  22  SO  14   30  73
58  44   34  76  19   IS  00   60   50  28  32  44   IS   35  »9  28  91  SO  S3   62  21  61  24   44  81
43  OS   50  00  20   39  2i   44   84  39  17  39   92   42  59  04  54  15  09   35  07  II  25   51  17
84  07   33  83  87   14  03   75   07  «  60  43   66   57  57   57  59  01  78   80  13  77  43   58  10
93  54   23  72  70   09  36   IS   24  04  74  05   85   29  M   67  37  28   13   98  01  48  29   75  89

54  44   72  02  34   52  81   38   $2  94  14  54   27   32  41  74   84  83  90  01   97  59  87   44  41
43  80   84  28  32   93  91   78   70  31  50  22   09   40  89  44   85  82  78  91   14  71  99   98  70
44  80   SO  18  92   44  42   48   47  22  87  14   20   45  82  01   45  21  49  80   17  39  70   74  03
78  70   39  30  04   59  65   14   64  04  62  28   44   64  05  89   81  80  09  89   54  11  27   81  44
14884703593215830401208292253488848076492510048402

49  28   04  18  54   78  97   49   14  85  01  58   31   18  20  53  74  03  27   05  80  39  15   47  49
99480994345410779SS89044S214S2S887SI31716SS3H8550
01  84   22  15  54   43  63   44   15  30  21  84   48   17  11  48  92  14  17   49  36  OS  17   80  24
47  85   26  91  23   14  28   01   76  47  45  12   58   24  27  41   59  43  20   15  93  47  30   54  27
I39!14749197834l99S04094M449782M04M»7452t4498M

95  82   20  95  52   45  9S   03   48  75  44  25   04   13  85   80  13  37  08   18  09  28   43  07  49
44  04   82  49  28   27  34   53   42  35  44  12   40  84  35   04  28  14  37   23  97  38   07  40  80
99  22   24  44  15   71  04   94   22  93  H  44   73   57  SI   22  54  82  37   99  94  27   25  87  77
08  44   26  12  87   72  42   13   57  77  61  07   94   24  62   17  76  19  45   18  98  11   47  40  31
14  94   74  04  37   32  09   72   81  22  87  70   81   93  78   93  37  22  32  25  38  43   38  OS  81

27  66   ft  S3  58   16  49   99   19  03  62  98   79  81  98   IS  03  63  32   93  68  24   14  44  SO
99  47   81  61  25   52  97   87   98  15  85  99   01   84  59   00  M  39  32   S3  49  18   42  51  45
89  14   37  94  03   22  W   45   42  61  97  83   04   26  30   48  49  40  9»   99  6»  H   13  94  21
34  13   53  15  32   42  02   58   32  14  83  73   02  82  49   25  42  91   U   M  70  n   44  SO  51
72  11   79  75  79   31  07   12   92  61  89  93   77  82  08   23  74  75   67  54  37  45   35  13  44

19  72   57  II  99   08  42   02   24  82  52  90   72  51  94   84  59  79   84  19  95  71   21  49  91
9«  99   71  43  90   27  M   94   15  70  17   74   92  31  85   24  47  55   44  SI  91  47  13  31  49
44  15   86  76   18   IS  57   29   51  62  95   84   20  W  01   11  90  66   80  81  40   43  65  87  35
34  83   94  07   SO   11  89   84   14  SO  09  97   04  74  51   41  20  54   M  20  33   53  70  10  22
S3  07   OC  16  M   M  43   40   $7  32  11  09   47  U  69   41  03  3*   24  02   16  4t   SI  3*  58

                                                   A-5

-------