ADMINISTRATIVE  AND  TECHNICAL  ASPECTS
OF  SOURCE  SAMPLING  FOR  PARTICULATES
                  PREPARED BY
    PEDCo-ENVIRONMENTAL  SPECIALISTS,  INC.
         SUITE  8,  ATKINSON SQUARE
           CINCINNATI,  OHIO 45246
             CONTRACT NO. CPA 70-124
     ENVIRONMENTAL  PROTECTION AGENCY
              TECHNICAL CENTER
    RESEARCH TRIANGLE  PARK, NORTH CAROLINA
                  MAY 1971

-------
  ADMINISTRATIVE AND TECHNICAL ASPECTS

  OF SOURCE SAMPLING FOR PARTICULATES
              Prepared by
 PEDCo-Environmental Specialists,  Inc.
       Suite 8, Atkinson Square
       Cincinnati,  Ohio  45246
       Contract No.  CPA 70-124
 ENVIRONMENTAL PROTECTION AGENCY
         Technical Center
Research Triangle Park, North Carolina
              May 1971

-------
This report was furnished to the Environmental Protection
Agency by PEDCo-Environmental Specialists, Inc. in ful-
fillment of contract number CPA 70-124.   The contents of
this report are reproduced herein as received from the
contractor.  The opinions, findings, and conclusions are
those of the authors and not necessarily those of the
Environmental Protection Agency.  Mention of company or
product names does not constitute endorsement by the
Environmental Protection Agency.
                         11

-------
                 Acknowledgments

     The principal authors of this report were Messrs
Richard W. Gerstle and Donald J. Henz of PEDCo-
Environmental Specialists, Inc.

     Mr. Gene W. Smith of the Office of Air Programs,
Environmental Protection Agency  was the project officer,
                        111

-------
                TABLE OF CONTENTS

Section                                           Page
List of Figures                                   vii
List of Tables                                     ix
INTRODUCTION                                        1
1.0  SOURCE SAMPLING PROGRAM                        2
2.0  FUNCTIONS OF THE SOURCE SAMPLING UNIT          3
     2.1  Specific Duties Assigned to the           4
          Source Sampling Unit
3.0  REGULATIONS REQUIRED TO CONDUCT SOURCE         6
     SAMPLING
     3.1  Statutory Authorization to                6
          Establish Program
          3.1.1  State Programs                     6
          3.1.2  Local Programs                     7
          3.1.3  Litigation of Source               8
                 Sampling Regulations
     3.2  Regulations Requiring Source              8
          Sampling and Monitoring
          3.2.1  Tests by the Agency                8
          3.2.2  Tests by the Owner-Operator       10
     3.3  Search Warrants                          10
     3.4  Typical Statute Codes and Regulations    11
          3.4.1  State Statutes                    11
          3.4.2  Regulations of State and Local    13
                Agencies
4.0  LEGAL USE OF SOURCE SAMPLING INFORMATION      18
     4.1  Taking the Sample                        18
          4.1.1  Test Equipment                    19
          4.1.2  Test Personnel                    21
     4.2  Transportation of the Sample             22
     4.3  Identification of the Sample             23
          4.3.1  Identification of Filters and     23
                 Containers
                         IV

-------
                TABLE OF CONTENTS
Section                                           Page
     4.4  Handling and Chain of Custody            23
     4.5  Laboratory Analysis and Calculations     24
     4.6  Custody of Final Report and Data         25
5.0  ORGANIZATION AND ADMINISTRATION OF A          27
     SOURCE SAMPLING UNIT
     5.1  Organizational Plans                     27
     5.2  Personnel Requirements                   34
          5.2.1  Manpower Needs                    34
          5.2.2  Test Team Personnel               36
          5.2.3  Personnel Costs                   37
     5.3  Equipment and Space Requirements,        38
          and Associated Costs
          5.3.1  Equipment and Costs               38
          5.3.2  Space Requirements and Costs      39
     5.4  Administrative Procedures                39
          5.4.1  Request for Source Test           39
6.0  PRELIMINARY PROCEDURES REQUIRED IN            44
     CONDUCTING A STACK TEST
     6.1  Pre-Survey Process Information           44
     6.2  Selection of Test Site                   51
     6.3  Preliminary Determination of Test        51
          Parameters
7.0  PARTICULATE SAMPLING PROCEDURES               53
     7.1  Measurement of Stack Gas Velocity and    53
          Related Parameters
          7.1.1  Location of Traverse Points       53
          7.1.2  Velocity Head Measurements        58
          7.1.3  Temperature and Static            61
                 Pressure Measurements
          7.1.4  Gas Density and Moisture          64
                 Determination
                         v

-------
                TABLE OF CONTENTS
Section                                           Page
          7.1.5  Calculation of Velocity and       66
                 Total Gas Flow
     7.2  Determination of Isokinetic Sampling     67
          Rates
          7.2.1  Calculation Aides                 69
     7.3  Non-Ideal Sampling Conditions            73
          7.3.1  Poor Flow Distribution            74
          7.3.2  Non-Isokinetic Sampling           74
                 Conditions
          7.3.3  Cyclic Flow Conditions            75
     7.4  Particulate Sampling Equipment           76
          7.4.1  Description of Sampling Train     76
          7.4.2  Assembling and Testing the        80
                 Train
     7.5  Sampling Procedure                       84
          7.5.1  Location of Sampling Points       84
          7.5.2  Length of Sampling Period         84
          7.5.3  Operation of Sampling Train       84
          7.5.4  Recording Data During Test        85
                 Period
          7.5.5  Sampling Problems                 92
     7.6  Disassembly and Particulate Clean-out    92
          Procedure
     7.7  Particulate Analysis                     94
     7.8  The Test Report                          97
          7.8.1  Format of the Test Report         97
          7.8.2  Presenting the Results            98
          7.8.3  Example Calculations              99
8.0  SIGNIFICANCE OF ERRORS IN SOURCE SAMPLING    107
APPENDIX  A  - Nomographs  for Sampling              115
APPENDIX  B  - Cleaning  of  Sampling Train           118
             Components
APPENDIX  C  - Orifice Calibration                  120
LIST OF SYMBOLS                                   123
REFERENCES                                        124
                         VI

-------
                  LIST OF FIGURES

Figure                                            Page

5.1     Organization Chart - State of New
        Jersey Bureau of Air Pollution
        Control                                    28

5.2     Current Organization Chart - State
        of New Jersey Air Pollution Control
        Bureau, Technical Services and
        Special Investigation Section              29

5.3     Organization Chart - Bay Area Air
        Pollution Control District                 30

5.4     Current Organization Chart - Bay
        Area APCD, Engineering Section             31

5.5     Organization Chart - City of Chicago
        Department of Environmental Control        32

5.6     Current Organization Chart - City of
        Chicago Dept. of Environmental
        Control Technical Services Division        33

5.7     Request for Source Test or Sample
        Analysis - Form Used in Los Angeles        41
5.8     Automated Source Test Request Form -
        State of New Jersey                        43

6.1     Pre-Survey Form for Combustion
        Sources                                    46

6.2     Pre-Survey Form for Incinerators           48
6.3     Pre-Survey Form for Industrial Process     49

7.1     Number of Test Points                      55
7.2     Cross-Section of a Circular Flue
        Divided into Three Concentric Equal
        Areas, Showing Location of
        Sampling Points                            56

7.3     Cross-Section of Rectangular Flue
        Divided into Twelve Equal Areas With
        Sampling Points Located at the
        Center of Each Area *                       56

7.4     Pitot Tubes Usually Used to Measure
        Velocity Head                              58
                        VII

-------
                 LIST OF FIGURES
Figure                                            Page
7.5     Gas Velocity and Volume Data               62
7.6     Correction Factor Nomograph                71
7.7     Operating Nomograph                        72
7.8     Expected Errors Incurred by Non-
        Isokinetic Sampling                        75
7.9     Particulate Sampling Train Used By
        Office of Air Programs                     78
7.10    Particulate Field Sampling Meter Data      86
7.11    Boiler Operating Data                      88
7.12    Incinerator Operating Data                 89
7.13    Process Operating Data                     91
7.14    Particulate Analysis Data                  96
7.15    Format for Presenting Emissions from
        Fuel Combustion Units                     100
7.16    Summary of Emission Test Data             101
7.17    Particulate Sampling Calculations         104
Al      Correction Nomograph for Use With
        Figure A2                                 116
A2      Operating Nomograph                       117
Cl      Orifice Calibration Form                  122
                        Vlll

-------
                 LIST OF TABLES
Table                                             Page
5.1     Examples of Source Sampling Staffs
        of Various Agencies                        35
5.2     Relative Pay Scales of Technical
        Personnel By Region                        37
5.3     Space Requirements for Source
        Sampling Programs                          39
7.1     Percent of Circular Stack Diameter
        From Inside Wall to Traverse Point         57
7.2     Example Determination of Type S
        Pitot Tube Correction Factor               59
7.3     Typical Format for Test Report             97
                        IX

-------
INTRODUCTION
     This manual is provided as a service by the Division
of Applied Technology, .Office of Air Programs, Environ-
mental Protection Agency to assist state and local air
pollution control agencies to better understand the
purposes and procedures of source sampling.  Presented
in this booklet are, general guidelines to show how
source sampling can be part of an agency's program, the
organization and approximate cost of such a program;
regulations required to permit source sampling, a
detailed description of the Office of Air Program's
procedure for particulate sampling; and other related
material.
     Organizational structures and functional duties of
the source sampling group cannot be exactly defined
since this will vary with the overall structure of the
control agency.  Example organization charts and
functions are however presented.  Sampling and analysis
procedures can likewise not be exactly defined for all
cases since these will vary with the purpose of the test
and the process sampled.  Procedures currently used by
the Office of Air Programs are however presented, and
should be followed in order to obtain comparable results.

-------
1.0  SOURCE SAMPLING PURPOSES
     Source sampling or emission testing, as applied to
air pollution studies, is the procedure whereby a
representative sample is removed from some larger,
contaminant bearing gas stream confined in a duct or
stack.  This sample is then subjected to further
analysis and the contaminant concentrations are re-
lated to the parent gas stream to determine total
quantities.  Since the sample extracted from the main
gas stream usually represents a very small fraction of
the total volume, extreme care is required in obtaining
a representative sample.  Due to the many and variable
factors encountered in sampling gas streams, complex
methods must frequently be used to obtain these repre-
sentative samples.
     Source sampling, frequently answers a variety of
questions, the main one being - what are the quantities
and concentrations of emissions?  Subsequent questions
which can be answered from this basic determination in-
clude :
     1.  Is the process in compliance with present
         or expected emission regulations?
     2.  What is efficiency of existing pollution
         control equipment?
     3.  What effect do various process variables
         have on emissions?
     4.  Is a valuable product or by-product being
         emitted?
     5.  What are the potential  (uncontrolled)
         emissions of various processes?

-------
2.0  FUNCTIONS OF THE SOURCE SAMPLING UNIT3
     The primary function of the source sampling unit is
to obtain reliable emission data.  The exact duties
assigned to the source sampling unit to perform this
function will vary widely from agency to agency depending
on the potential work load for this unit, the emission
regulations, and the availability of other agency person-
nel when required.  In small agencies, where source
sampling may not be a full-time activity, source sampling
personnel may actually be part of some other unit such as
engineering or technical services.  In this case, when
sampling is required, personnel will have to reschedule
their other work, perform the test work and analysis,
and then return to their routine duties.
     In contrast to this part-time activity, a large
agency with many requirements for source testing will
have a full-time staff performing tests.  This staff will
include chemists as well.  Engineering technicians can
maintain the sampling equipment, perform calibrations,
assist in stack testing, and make routine calculations.
The engineering staff will perform sampling site surveys,
plan the test procedures, set the schedule, supervise the
actual tests, review calculations, and prepare the final
report.  In large source sampling units, a chemist or
senior chemical technician may be assigned to the sampling
group.  This person would be responsible for all routine
lab analysis and serve as coordinator between the
laboratory and sampling units.
a)  For simplicity, the group of people comprising
    the source sampling function are referred to as
    a unit.  They could be referred to as a section,
    group, etc., depending on the agency's adminis-
    trative breakdown.

-------
     The stack sampling unit can also perform duties
closely related to source sampling, such as determin-
ing or checking emission factors for various processes,
developing and/or improving test methods and equipment,
developing particle size distribution data, and prepar-
ing summary reports of emission data and related factors
for presentation at technical meetings.
     Section 5.0 presents organizational plans and
personnel assignments for various types of control
agencies.
2.1  Specific Duties Assigned to the Source Sampling
     Unit
     Specific duties to be performed by the source
sampling unit in a larger agency include:
     Technical Duties
     1)  Develop and up-date reliable source test-
         ing procedures for particulate and gaseous
         emissions*
     2)  Calibrate and maintain all equipment
     3)  Plan and conduct source tests as required
     4)  Perform and check all test calculations
     5)  Prepare test reports and summaries of
         emission data
     6)  Review source tests conducted by private
         firms
     Administrative Duties
     1)  Train personnel
     2)  Procure equipment to conduct source tests
     3)  Maintain a file of all source test data
     4)  Prepare annual reports and budget require-
         ments
*)  All methods used for compliance tests are subject
    to approval by the Office of Air Programs

-------
     5)   Make contacts with plant personnel
     6)   Schedule tests
     7)   Coordinate source test data with other
         agency activities
     The functions assigned to the source sampling
group in a small agency are more varied since other
duties will be performed in the interim between con-
ducting source tests.
     In addition to the technical duties connected
with source sampling, the following additional duties
for example can be performed:
     1)   Conduct a limited ambient air monitoring
         program
     2)   Conduct emission inventories
     3)   Assist in plan review and site inspections
     4)   Perform routine laboratory analyses
     Some of the engineering and administrative functions
may be assumed by higher levels of supervision in smaller
agencies.  Alternatively the entire sampling furction can
be contracted out to a reliable consultant with the
administrative duties handled by the agency.

-------
3.0  REGULATIONS REQUIRED TO CONDUCT SOURCE SAMPLING
3.1  Statutory Authorization to Establish Program
     Air Pollution control agencies possess only those
powers specified by the legislative body through some
type of enabling legislation.  Generally two steps are
required before the agency can embark on a source
sampling program, namely:  1) enabling legislation is
adopted and 2) regulations are promulgated.  The
enabling legislation should establish that the air
pollution control agency is empowered to maintain a
source testing program.  The regulations detail the
program and refer to the test procedures, testing
requirements, test frequencies, emission limits, and
the like.
     3.1.1  State Programs
     Most state air pollution control agencies have
authority to inspect processes and equipment to deter-
mine compliance with equipment specifications and
emission regulations.  However, a deficiency may exist
where inspection powers are granted without specific
mention of the administration of a testing program.
In the absence of specific language authorizing source
sampling, it is possible that the statute is suffi-
ciently broad to reasonably infer that a testing program
is to be implemented.  Perhaps such an inference may be
drawn from the stated purpose of the legislative
power-grant.  However, to guard against possible mis-
interpretations, enabling legislation should specifically
mention inspection powers and source sampling adminis-
tration.  The Federal Clean Air Act of 1970 requires
that a state have authority to make inspections and

-------
test emissions.  A source sampling program is essential
to the enforcement aspects of an implementation plan.
     After legal advice has been obtained as to the
adequacy of the enabling legislation, the state agency
should develop administrative regulations consistent
with the legislation.  While there are many existing
regulations upon which administrative regulations can
be based there is no substitute for the assistance of
legal counsel at the outset.  He can, however, benefit
from a study of existing regulations, and they should
not be ignored.
     3.1.2  Local Programs
     Many states have delegated to their various
political subdivisions the authority to establish and
maintain air pollution control programs.  The Federal
Clean Air Act however specifies that the primary
responsibility for controlling air pollution lies with
the states.  When a state does delegate this authority,
it must be ready to step in if the local entity fails
to meet its obligations.  For local programs, the specific
entity - usually a county or health district - has to
adopt emission source sampling regulations.  Local
regulations must be no less stringent than the state's
regulations.  As with the state agency, the local
political subdivision must determine that it has
adequate authority to establish a source sampling
program and then adopt compatible regulations or ordi-
nances.
     In some states it may not be necessary for the
state legislature to sanction local programs.  That is,

-------
in the absence of statutory authority it may be possible
to establish and maintain a source sampling program
through the powers given to or retained by various state
political subdivisions under the state constitution.
Thus, various cities may maintain programs on the basis
of their constitutionally granted home-rule powers.  As
previously stated, local program regulations must be no
less stringent than the state regulations.
     3.1.3  Litigation of Source Sampling Regulations
     Through 1969, no cases have been reported concern-
ing the litigation of source sampling regulations.
However, the related area of search and seizure has been
very active since the Supreme Court decision in See v.
City of Seattle, 87 S. Ct. 1737 (1967).  Search warrant
requirements are discussed in Section 3.3
3.2  Regulations Requiring Source Sampling and Monitoring
     State regulations requiring periodic reports on the
nature and amount of emissions, and the installation of
emission monitoring equipment are mandated by the Clean
Air Act as amended in 1970 [Sec. 110(a) (2)  (F)] .  The
Act, as amended, also provides the Administrator of EPA
with authority to promulgate regulations regarding the
periodic testing and monitoring of emissions by the
owner or operator of any stationary source.  [Sec. 114(a)]
Authority also exists for the Administrator to conduct
source tests under certain conditions.  [Sec. 114(a)  (2)
(B)].  Basically, both the regulators and the regulated
will conduct source tests.
     3.2.1  Tests by the Agency
     While the primary responsibility for source testing

-------
ing rests with the process owner, the agency must have

authority to conduct its own tests as a back-up measure,

The Agency's  regulations should consider the following

     0  Test Methods - Standardized testing methods
        are required.  Regulations should specify
        that tests will be conducted in a manner
        determined by the director of the Agency.
        These methods in turn should be approved by
        the Office of Air Programs.
     0  Equipment and Processes to be Sampled -
        Regulations should specify that all
        stationary sources are subject to being
        tested by the Agency.
     0  Frequency of Tests - The director of the
        Agency should have the discretion to
        require source tests.  Provisions should
        be made for testing when the agency has
        good cause to suspect emissions in excess
        of the regulatory limitations as deter-
        mined by field inspections.

     0  Employment of Independent Testers - It may
        be desirable to provide for the employment
        of qualified independent testers.  This is
        especially pertinent to the smaller agencies.

     0  Access to Facilities - Sampling ports,
        electrical power, platforms, ladders, aud
        the like are all necessary for source
        sampling.  These facilities should be
        provided at the Owner's expense and should
        be specified for all operations subject to
        the source sampling requirements.  Reason-
        able access to the test facilities should
        also be specified.  Installation of these
        facilities can be incorporated in a permit
        system.
     0  Test Costs - Regulations should specify an
        equitable allocation of costs.  A general
        guideline might be to require full payment
        by the Owner-Operator in all cases where
        the test indicates emissions are in excess
        of the regulatory limitations and where
        the test is being conducted pursuant to
        the issuance of the first operation permit.
        Where emissions are below the regulatory

-------
        limit, the Owner-Operator would not be
        charged.

     3.2.2  Tests by the Owner-Operator

     The Owner-Operator will be required to conduct

tests pursuant to state and Federal regulations.   The

following items should be considered in preparing

regulations:

     0  Frequency of Tests - Tests should be made
        to provide the Agency with information
        regarding the nature, extent, and quantity
        of emissions.  After the initial test, the
        Agency should be given the discretion to
        require additional tests.

     0  Test Certification - All tests should be
        certified by a professional engineer or
        witnessed by an agency representative.

     0  Test Costs - The Owner-Operator should
        bear all costs incurred in making his
        own tests.

     0  Test Methods - Standardized testing methods
        should be required.  Regulations should
        specify that tests must be conducted in a
        manner determined by the director of the
        agency.

3.3  Search Warrants

     The necessity for the procurement of a search warrant
as a condition precedent to source sampling must be con-
sidered in preparing a regulation.  This is a very fluid

area at present.   Leading cases in this area are See v.

City of Seattle,  87 S.Ct. 1737  (1967); People v.  White,
65 Cal. Rptr.923  (1968); United States v. Kramer Grocery

Co., 418 F. 2d 987  (1969); and Colonnade Catering Corp.
v. United States, 25 L Ed 2d 60  (1970).

     The decision to design a regulation which will
alleviate the need for search warrants is up to the
                         10

-------
  agency.  Such  a  regulation will  require the  advice of

  the agency's legal counsel and should consider the

  following factors:

       0  The entrance, inspection and testing should
          be connected to a bone fide licensing or
          permit system.

       0  Penalty  provisions should not be designed
          so as  to indicate that they are the  sole
          sanction,  without a warrant, to enter.
       0  Consent  to test should be obtained in
          advance  with the issuance of the license
          or permit.

  3.4  Typical Statutes, Codes  and Regulations

       Typical statutes and regulations promulgated in

  jurisdictions  which have established air pollution
  control agencies are presented in this Section to show

  how the various  factors discussed in the first three
  Sections of this chapter may  be  integrated into statutes,

  codes, and regulations.  These statutes and  regulations
  cover state statutes, regulations of state agencies,  and

  regulations of local agencies, all of which  pertain to

  source sampling.

       3.4.1  State Statutes

       The enabling legislation of the State of Ohio

  (ORC §3704.03) reads as follows:
 Sec.  3704.03 Powers of board.
 The air pollution control board may:
  (K)  Through -any" indiyiduir member•or any  	 Entry  by the board,  an
representative authorized by the board, enter ^J          .*  .   ,    .
upon private or  public property,  including         authorized employee,  or
Improvements thereon, at any reasonable time for         consultant
the purpose of determining if there are any .
emissions from such premises, and if so, to
determine the sources and extent of such emissions; <]	 Source sampling
                             11

-------
The  New Jersey  law
                                                     §26:2c^9L  provide^
    The department shall control air pollu-
 tion in accordance with the provisions of any
 applicable code,  rule or regulation  promul-
 gated by the department and for this purpose
 thall have power  to—
   (d) Enter and inspect any building or place,
   except private residences, for the. purpose of
   investigating an  actual or suspected source
   of air pollution  and ascertaining compliance
   or noncompliance with any code, rules and
   regulations of the department. Any informa-
   tion relating to  secret processes or  methods
   of manufacture  or production obtained in
   the course of such  inspection, investigation
   or determination, shall be kept confidential
   and shall not be admissible in evidence in
   any court or in  any other proceeding except
   before the department  as herein defined. If
   samples are taken for analysis, a duplicate
   of the analytical report shall  be furnished
   promptly to the person suspected of causing
   air  pollution;
                                    Promulgation of  regulations
                                    Source  sampling
          The Kentucky  law  (KRS  §224^3701  readsi
   224.370  Inspection of premises; inter-
 ference  unlawful.  Any duly authorized
 officer, employe, or representative of the
 commission may enter and inspect  any
 property, premise, or place at any reason-
 able time for the purpose of investigating
 either an actual or suspected source of air
 pollution or of ascertaining the state of
 compliance with KRS 224.310 to 224.460
 and 224.991 and regulations enforced pur-
 suant thereto. No person shall refuse entry
 or access to any authorized representative
 of the commission who requests  entry for
 purposes of inspection, and who presents
 appropriate credentials; nor shall any per-
 son obstruct, hamper, or  interfere  with
 any such inspection.  (1966, e. 22, § 9)
                              O—Entry  by  the  commission,
                                   an authorized employee,.
                                   on consultant
                              <]— Source  sampling
          Illinois  has just passed  a comprehensive  Environmental
   Protection  Act.   Section  10  of that  Act reads  in part:
  Section  10. The Board,  pursuant to procedures pre-
scribed in Title VII of this Act, may adopt regulations to
promote the  purposes  of this Title. Without limiting the
generality of this authority, such regulations may among
other tilings prescribe;
  (0 Requirements and procedures  for the inspection of
any equipment, facility, vehicle, vessel, or aircraft that may
cause or contribute to air pollution;
                                                    Adoption  of  regulations
                                                    including broad  power
                                                    for  inspection
                                       12

-------
   3.4.2  Regulations of State and Local Agencies

      The Commonwealth of Kentucky Air Pollution Control

  Commission has  adopted  the following  testing  require-

  ments  for indirect  heat exchangers  in the  commission's

  Regulation 7:


(1)  Whenever  the Kentucky Air Pollution i
Control Commission has reason to believe
that  the   emission   limits   of  this
Regulation  are  being  violated, it may 
-------
        Kentucky's Regulation  8, Section 6,  also provides  for

   source  testing  as a condition of the  issuance  of a use
   permit.
Permits  issued  hereunder  shall be  sub-
ject   to  such  terms and conditions set
forth   and  embodied  in the  permit  as the
Commission    shall    deem   necessary  to
insure  compliance   with  its standards.
Such  terms   and  conditions  may  include
maintenance  and availability of  records
relating  to  operations which may cause
or  contribute to air pollution including  <]—  Periodic  sampling
periodic source or  stack  sampling  of the       ky  licensee
air contaminant sources.
Acceptance   of  a   permit  conditioned  as
described   herein  shall  denote agreement  
-------
 twecn  the commissioner and the' person.
 All tests and calculations shall be made
 under   the  direction  of  a  competent
 engineer. Any  test or tests to be con-
 ducted  on  the  premises  where  such
 equipment or device is  located  shall  be
 made  during   reasonable   hours,  after
 written  notice  to, and with the  coopera-           Owner Cooperation/
 tion of, the owner or operator. The cost  <]	  Test COSt
 of any test or tests and calculations shall
 be a debt due the city from any person
 responsible as  owner, operator  or other-
 wise of  such fuel-burning, combustion or
 process  equipment  or device in all cases '
 when such test or tests shall have proven
 any  emission  of  particulate  matter  in
 violation of any provision of this chapter,
 and such unpaid  debt shall be  recover-
 able in any court of competent jurisdic-
 tion. If  any  such emission is  shown by
such test or tests to be within the limits
 of emission  prescribed in  this  chapter,
the cost  of such  test or tests shall be
charged  to the  annual appropriation  of
the department.
          The  City of  Cleveland's  authority  to  test  is given
   in Chapter  5 of  the  Air  Pollution  Code.    Section  84.0502
   reads  in part:
     §4.0502.  Duties of Commissioner.
     The  Commissioner  of Air  Pollution  Control
under the supervision and direction of the Director of
Public Health and Welfare shall:
     F. Make inspections and tests of existing and
newly installed equipment subject to this ordinance
to determine whether such equipment complies with
this code;

-------
          Complete  details  of  source  sampling  requirements  are

   then  given  in  Chapter  17  of  the  same Code,:
     §4.1702.  Sampling and Testing.
     (A) The Commissioner of Air Pollution Con-
trol is hereby authorized to conduct, or cause to be
conducted, any test or tests of any new or existing
process,   fuel-burning,  refuse-burning,  or  control
equipment the operation of which in his judgment
may result in emissions  in excess of the limitations
contained in this ordinance or when the emissions
from any such equipment  may exceed .the limits of
emissions provided for  herein. All tests shall be
conducted in a manner determined by the Commis-
sioner and a complete detailed test report of such test
or tests  shall  be  submitted to him. When tests are
taken by the owner or independent testers employed
by the owner, the Commissioner shall require that the
•aid  tests be  conducted  by  reputable,  qualified
personnel and shall stipulate that  a qualified repre-
sentative or  representatives of the Division of  Air
Pollution Control be present during the conduct of
such tests. The Commissioner may stipulate a reason-
able time limit for the completion of such test and
the submission of test reports.
     (B) Nothing in  this section concerning tests
conducted by  and  paid for by any  person or his,
authorized agent shall be deemed to abridge the rights
of the Commissioner or his representatives to conduct
separate or additional tests of any process, fuel--
burning,  refuse-burning, or control equipment on
behalf of the City of Cleveland, whether or not such
tests relate to emissions controlled by specific limita-
tions under this code.

      14.1703. Test Facilities and Access.
     (A) It shall be the responsibility of the owner
or operator of the equipment tested to provide, at his
expense, utilities, facilities and reasonable and neces-
sary openings in  the system or stack, and safe and
easy access thereto, to permit samples and measure-
ments to be taken. All new sources of  air contami-
nants  created after the  effective date of this ordi-
nance may be required by the Commissioner of Air
Pollution Control to provide  utilities, facilities and
adequate openings in  the system or stack, and safe
and easy access thereto,  to permit measurements and
samples  to be taken.
     (B) When any process equipment, fuel-burning
equipment or refuse-burning equipment has caused an
air pollution nuisance, as determined by the Commis-
sioner, or has violated a provision of Chapter 11, 13
or 16 of this code, the Commissioner may, at his
discretion require that said equipment be equipped
with an  air  contaminant recording device  with an
audible  alarm set so as to become  activated upon
reaching prohibited levels of emission, which device
shall be  maintained in proper operating conditions at
•II times. Records from such recording device shall be
made available to the Commissioner for periods up to
one year.
           Tests  by  the Commi-ssioner
           or authorized  represen-
           tatives
O-
           Test procedures
-Tester  qualifications
           Testing  facilities
           and  access

-------
      §4.1704.  Test Costs.
      If emission tests conducted as a result of the
action of the Commissioner of Air Pollution Control
substantiate that a  violation  exists,  the person or
persons responsible for the violation shall be respons-
ible for paying  all attendant costs for conducting said
tests.  If said tests do not show that a violation exists,
then the City shall be responsible for paying all costs
for conducting  the said test. In no event shall the city
assume  costs  of providing facilities, utilities  and
access for such testing. When 'the person responsible
O— Test  costs
<]— Cost  of  providing
      facilities
test
elects to conduct his-own stack emission tests, then
the person so electing shall pay for the test or tests
notwithstanding other provisions of this section, and
irrespective  of the result. The costs of emission tests
required by the Commissioner on newly installed
equipment for the issuance of the initial  permit  to
install and  the issuance of the initial  certificate  of
operation shall not be at the  expense of the City  of
Cleveland regardless of results. The tests for existing  '
sources  relating  to contaminants  not specifically
controlled by this code shall be at the expense of the  <]— Test  COSts
City of Cleveland except  for facilities, utilities and
access required to be provided by this Chapter.

      §4.1705. Circumvention and Right  of Entry.
      (A) No person shall  build, erect,  install, or use
any article,  machine, equipment, or  other contri-
vance,  the  sole  purpose  of  which is to dilute  or
conceal an emission without resulting in a reduction
in  the  total release  of  air  contaminants to the
atmosphere  nor  shall a person do any  thing nor
commit any act  with the  intent to distort stack test
emission results.
      (B) Any person who in any manner  hinders,
obstructs, delays, resists, prevents, or in any manner
interferes or attempts to interfere with the Commis-
sioner or his representatives  in the performance  of
any duty  enjoined, or  shall refuse to permit the
Commissioner or such  representatives to perform
their duty   by  refusing them, or either of  them,
entrance at reasonable hours to any premises in which
the provisions of this ordinance are being violated, or
are suspected of being violated, or refuse to permit
testing, or permit the inspection  or examination  of
such premises for the purpose of the enforcement of
this ordinance shall  be subject to cancellation of the
certificate of operation, or such other action as may
be provided at law or by provisions of this code.
                                             17

-------
4.0  LEGAL USE OF SOURCE SAMPLING INFORMATION
     Every test should be conducted as if it will
ultimately be used as evidence in court.  The collection
and analysis of source samples should become a routine
matter to the agency personnel involved.  However, it
must be remembered that this routine procedure is
esoteric to the layman and therefore subject to greater
scrutiny whenever the agency has to rely on these
results.  It is imperative that source sampling and
analysis be done under standard procedures and that
each step be well documented.  In short, the report may
ultimately be subjected to the requirements of the Rules
of Evidence.
     This chapter will discuss the standardization of
source sampling procedures relative to taking the sample,
chain of custody, laboratory analysis, report custody,
and disposition of the original work sheets.
4.1  Taking the Sample
     In attacking the validity of source sampling
results, the adverse party will concentrate on four main
items relative to taking the sample:  1) The sampling
procedure, 2) The recorded data and calculations, 3)
The test equipment, and 4) Ttie qualifications of the test
personnel.
     The agency must keep in mind the possibility of
adverse inferences that may arise from the use of un-
orthodox or new procedures.  Therefore deviations from
the standard procedure must be kept to a minimum and
applied only where absolutely necessary to obtain an
accurate sample.  Changes in methodology must be based
                         18

-------
on sound engineering judgment and must be carefully
documented.  Standard procedures which should receive
particular attention are:
     1)  Location of sampling station
     2)  Number and size of sampling zones in the duct
     3)  Use of recommended sampling equipment
     4)  Careful determination of gas velocities
     5)  Maintenance of isokinetic sampling conditions
     6)  Proper handling of the collected sample and
         recording of container and filter numbers.
     Close scrutiny is also focused upon the recorded
field data since it is these data which form part of the
physical evidence.  Standardized forms should be utilized
to insure that there is no lack of necessary information.
Example forms designed for this purpose are included in
Chapter 7.  These forms consist of field forms, labo-
ratory forms, and calculation forms.  Only the field
forms are utilized when taking the sample.  This form
is designed to clearly identify the process tested,
the date and time, location of test station, sampling
personnel, and the person who recorded the data.  During
the actual test period, the meter readings, temperature
readings, and other pertinent data should be recorded in
the provided spaces immediately upon observation.  These
data determine the accuracy of the test, and should not
be erased or altered.  Any errors should be crossed out
with a single line and the correct value recorded above
the crossed-out number.
     4.1.1  Test Equipment
     Faulty test equipment can also invalidate a test.
In general there are two types of field test equipment,
                         19,

-------
the gas sampling equipment and process measuring equip-
ment.
     The process measuring equipment consists of any of
the metering devices from which test data is obtained.
These are scales for weighing fuel or raw materials,
orifices and gages for measuring product flow, and the
like.  It cannot be assumed that these devices are
accurate since proper maintenance and calibration
procedures are often lacking.  In any case check and
record the date on which the device was serviced.
     Ideally, the use of process measuring equipment
should be kept to a minimum.  However, process weight
regulations may frequently require the use of such
equipment - especially scales.  Such scales can only
be properly serviced and calibrated by specially trained
personnel.  The scale manufacturer most generally provides
this service.  A stamp affixed to the scale by the
service crew as a standard procedure will note the date
of calibration or inspection.  If the scale has not been
recently calibrated an engineering judgment must be made
concerning its accuracy.   A material balance will some-
times provide a check on scale readings.
     Other equipment such as flow meters and gages should
be properly maintained and used.  If there is reason to
believe that the equipment is defective, note the reason
on the Field Data Form and make an engineering judgment
on the validity of the data.
     Among gas sampling equipment which requires mainte-
nance and calibration is the Pitot tube, manometers,
thermometers, flow meters, and dry gas meter.  The main-
tenance of these instruments is subject to even greater
                         20

-------
scrutiny in court.  Therefore written maintenance

records must be kept.   Suggested maintenance procedures

are as follows:

     Pitot tube - The  Pitot tube should be calibrated
     when acquired.  Subsequent calibration is not
     required but a visual check should be made and
     noted prior to each test series.  (See Section
     7.1.2)

     Manometers - The  inside of the tubes are
     subjected to the  flue gas and the specific
     gravity of the oil may change due to evapora-
     tion.  Readings also become difficult as dirt
     coats the glass tube.  It is suggested that
     the manometers be washed with soapy water and
     the oil replaced  after approximately every
     sixth test series.  Note that the specified
     oil must be used.
     Thermometers - Dial type thermometers are
     frequently used in the field.  These are
     easily damaged and therefore should be
     checked prior to  each test series.  The
     check should be made against a mercury
     thermometer at approximately 1/4 and 3/4
     of full scale.  Thermocouples and associated
     recording equipment must also be periodically
     calibrated.  Six month intervals are re-
     commended as a minimum.

     Dry Gas Meter - The meter should be cali-
     brated prior to each test series.  This high
     frequency of calibration is recommended due
     to the relatively severe conditions under
     which the meter is used.  It is subject to
     being bumped, dropped, vibrated, or even
     being carried upside down.  The best method
     of testing is with a positive displacement
     calibrator such as a Bell-type Prover or a
     calibrated orifice.

     4.1.2  Test Personnel

     The sample must be taken by experienced personnel.
While it is not necessary that the chief of the field
team be a professional engineer, he must have special
                         21

-------
training which qualifies him for source sampling.   If
the report is used in court, the chief of this field
team may be called as a witness.  Poor data may be in-
admissible as evidence.  Therefore the chief of the
field team should have previous experience as an aide
on field tests, and he should preferably have received
special training in source sampling.   (Section 5.2
describes personnel duties in greater detail.)
     One cannot usually perform a source test alone.
Two men are normally required for one test station and
a minimum of three are required for two stations.   It
is often difficult to accurately record the large amount
of required data if the team is inadequately manned
4.2  Transportation of the Sample
     Of primary importance in transporting the sample
to the laboratory is that precautions be made to
eliminate the possibility of tampering, accidental
destruction, and/or physical and chemical action on the
sample.
     To reduce the possibility of invalidating the
results, all components of the sample must be carefully
removed from the sampling train and placed in sealed,
non-reactive, numbered containers.  The sample should
then be delivered to the laboratory for analysis.  It is
recommended that this be done on the same day as the
sample was taken.  If this is impractical, all the samples
should be placed in a carrying case (preferably locked)
in which they are protected from breakage and contamination
as well as avoiding the possibility of loss.
                         22

-------
4.3  Identification of the Sample
     Care must be taken in properly marking the sample
for positive identification throughout the test and
analysis procedures.  The Rules of Evidence require
impeccable procedures for identification of samples,
the analysis of which is the basis for future evidence.
An admission by the lab analyst that he could not be
positive whether he analyzed sample No. 6 or sample
No. 9, for example, could destroy the validity of the
entire report.  Positive identification must be provided
for the filters and the containers used in any specific
test.
     4.3.1  Identification of Filters and Containers
     Filters should be marked for positive identification.
Three digits should insure the unique identification of
filters for many years.  The ink on the filter must be
indelible and unaffected by the gases and temperatures
to which it will be subjected.  Filters must be marked
before taring.  If another method of identification is
desired by the agency, it should be kept in mind that
the means of identification must be positive and must
not impair the ability of the filter to function.
     Each container should also have a unique identifi-
cation to preclude the possibility of interchange.  The
number of the container should be recorded on the analysis
data sheet  (See Figure 7.14) and thereby associated with
the sample throughout the test and analysis.
4.4  Handling and Chain of Custody
     In no case should the sample be handled by persons
not associated in some way with the task.  A good
                         23

-------
general rule to follow is, "the fewer hands the better",
even though a properly sealed sample may pass through a
number of hands without affecting its integrity.
     It is generally impractical for the analyst to
perform the field test.  However, the Rules of Evidence
require that the prosecution be able to prove the chain
of custody of the sample.  For this reason each person
must be able to remember from whom he received the sample
and to whom he delivered it.  This requirement is best
satisfied by having each recipient sign a receipt or
the data sheet for the sample or set of samples.  The
process owner should also be given a receipt for the
collected sample.
4.5  Laboratory Analysis and Calculations
     Potential sources of error in the analysis of the
sample lie in the analyzing equipment, procedures,
documentation of results, and the qualifications of the
analyst.
     Laboratory equipment, especially the analytical
balance, should be subjected to a routine maintenance
program just as the field equipment is.
     Analytical Balance - Balances are extremely
     sensitive and therefore require periodic
     calibration.  It is recommended that cali-
     bration be done at least biannually, with
     Class M weights.  A record should be kept.
     Reagents - Only reagent grade chemicals should
     be used.  Reagents used in an Orsat or similar
     gas analyzer should be replaced periodically
     depending on their use.
     As with the field procedures, the laboratory data
and calculations must be well-documented.  The use of
                         24

-------
standardized forms is recommended.   In all cases the
person who performs the analysis and/or calculations
should sign the data sheet
4.6  Custody of Final Report and Data
     The team chief is responsible for the compilation
of the test report under the supervision of a senior
engineer who then reviews it for content and technical
correctness.  The ultimate use of the report as evidence
of a violation is the responsibility of the agency's
supervisory management.  It is that echelon which
finally approves the report as a correct representation
of the field conditions.
     Written documents are, generally speaking, considered
to be hearsay and therefore not admissible as evidence
without a proper foundation.  A proper foundation consists
of introducing the report by the principal author(s).
Thus the chief of the field team and the laboratory
analyst would both be required to lay the foundation for
the introduction of the test report as evidence.  However,
the foundation laying is greatly simplified, though still
required, under statutory exceptions to the Hearsay Rule
found in the Official Reports as Evidence Acts and
Business Records as Evidence Acts which various states
have adopted.
     The rationale of the Official Reports exception is
that it is assumed that a public officer performing a
particular duty, performs that duty-properly and is under
no motive to distort the truth.  Basically, the Official
Reports exception exists to avoid the necessity and
expense of calling as witnesses various persons who may
                         25

-------
have collaborated in making the records.
     To insure the benefit of these statutory exceptions
to the Hearsay Rule, the source test reports should be
filed in a safe place by a file custodian who has
responsibility for the files.  Once the report is
approved, a summary copy is sent to the requestor for
further disposition.  Generally, the field notes and
calculations need not be included in the summary report.
However all this material may be required at a future
date to bolster the acceptability and credibility of the
report as evidence in an enforcement proceeding.  There-
fore the full report including all original notes and
calculation sheets should be kept in the file.  Signed
receipts for all samples should also be filed with the
test data.
     Public records are subject to the Best Evidence Rule
which basically states that the original of a document is
the best evidence and therefore a mere copy is not
admissible as evidence.  Microfilm, snap-out carbon
copies and similar contemporary business methods of
producing copies are acceptable in many jurisdictions if
the original is not reasonably available, its unavaila-
bility is adequately explained, and the copy was made in
the ordinary course of business.
                         26

-------
5.0  ORGANIZATION AND ADMINISTRATION OF A SOURCE
     SAMPLING UNIT

5.1  Organizational Plans
     The source sampling unit must fit into the agency's
organization such that it meets both the needs and
available resources of the control agency.  Since these
parameters vary so widely from area to area, it is
impossible to define an ideal overall organizational
structure.  However the main variables in the organ-
ization of source sampling operations are the number
and complexity of the process which must be tested
and the functions to be performed by the unit.
     Structuring the agency's source sampling operations
requires considerations of so many variables that no one
type of organization can be recommended.  Figures 5.1
thru 5.6 show the structure of three agencies which have
established, comprehensive source sampling programs.
These figures are presented as examples to show how the
test program relates to the overall program, and to show
the actual organization of the source sampling unit.
     Where many diverse, well defined processes are
located within an agency's jurisdiction, it is frequently
advisable to utilize personnel who have expertise in
these specific processes as supervisors of the source
test teams.  The number and designation of the super-
visors will, of course, depend on the processes to be
sampled.  Unless an extreme amount of specialized test-
ing is required, all purpose teams are more efficient.
a)  See Chapter 2.0
                         27

-------
                                      Bureau  of Air
                                      Pollution Control
             Planning  and
              Evaluation
Chief Enforcement
    Officer
Local Program
 Development
Research and
Development
ISJ.
06
                                     Field Control
                                      Operations
           Permit  and
           Certifica-
           tion Sectioh
Technical Services
and Special Inves-
tigation Section
 (See Figure 5.2)
        Enforcement
          Section
                                                        Southern
                                                          Field
                                                         Office
                                   Central
                                    Field
                                   Office
                         Metro
                         Field
                         Office
                                 Figure  5.1  Organization  chart  -

                        State of New Jersey Bureau of Air Pollution Control.

-------
to
vo
Supervisor of
Technical Services
CSee Figure 5.1)


Secretary " —





Supervisor Supervisor Supervisor Supervisor
of Crews of Crews of Crews of Crews


n~l
Crew E Crews G


^^^ i— .Lew 15 1 v-tew u 1 t-itJW r i


Equipment
Maintenance

Building
Maintenance


Laboratory
u, 	 1


& H
staffed



                           Figure  5.2  Current  organization  chart  -


             State of  New Jersey Air Pollution Control  Bureau,  Technical  Services
                               and Special  Investigation  Section

-------
U)
O


Controller


Supervis
Air Polli
Enginee
(See Fiq
I
Public In-
formation
& Education
Off jce


I
,ory Supervisory
jtion AirPolkrtion
;r Inspector
O • TI / 1

Chief
Administra-
tive
Officer




1
Director
of
Enforcement

Direct
Counsel of
Technic
Servic

I i
Meteorology
Statistician Monitoring and Data
Analyses

.or
al
es

1
Laboratory
                                 Figure 5.3  Organization chart

                             Bay Area Air Pollution Control District.

-------
Senior Air
Pollution
Engineer
                  Supervisory
                Air Pollution
                  Engineer
                 (See Fig. 5.3)
Senior Air
Pollution
Engineer
                    Air
                    Pollution
                    Engineer
                    Assistant
                    Air Poll.
                    Engineer
                    L
                    Air
                    Pollution
                    Technician
Senior Air
Pollution
Engineer
   Figure  5.4  Current organization chart

     Bay Area APCD, Engineering Section.
                     31

-------
                   Department of
                   Environmental
                      Control
  Director
     of
Engineering
   Director
      of
  Technical
   Services
(See  Fig.  5.6)
 Director
    of
Enforcement
          Figure 5.5  Organization chart

City of Chicago Department of Environmental Control.
                        32

-------
                  Director of
                  Technical
                   Services
                 (See Fig. 5.5)
Meteorology
Monitoring
and Testing
Laboratory
     Stack Test
     Supervisor
               Monitoring
     Chief
     Technician
     Technician
        Figure  5.6  Current organization chart
    City of Chicago Dept. of Environmental Control
              Technical Services Division.
                      33

-------
Since an engineer functionally supervises the test team,
his expertise coupled with the testing team's basic back-
ground should result in reliable and efficient tests.

     The title of the source sampling supervisor will  depend
on the size and organization of the agency.   For the smaller
agencies, the supervisor could have several units under him -
perhaps the laboratory and ambient air monitoring in addition
to source sampling.  He may sometimes be the Chief Air
Pollution Control Officer.

5.2  Personnel Requirements

     5.2.1  Manpower Needs
     Estimating the manpower needs of a source sampling
operation is difficult since so many factors are involved
and the final determination of the number of people will
depend on the specific work load.  Table 5.1 shows the manpower
needs of three existing programs - state-wide, multi-county,
and city-wide.  Since these are successful programs, the two
factors used in a recent manpower model are also displayed
in the Table to provide some perspective.  It should be
noted that the current manpower needs shown in Table 5.1
are compared with 1963 data.  The manpower model predicts
that an agency must service at least 4,000 manufacturing
establishments having annual capital expenditures of over
$200 million before more than three source test personnel are
required.  State-wide capital expenditures of manufacturing
establishments during 1967 were less than $200 million in
31 states and less than $400 million in 26 states.
                         34

-------
                     Table  5.1  EXAMPLES OF SOURCE  SAMPLING  STAFFS  OF  VARIOUS  AGENCIES
CO
en
          Agency
          State  of
          New Jersey
San Francisco
Bay Area Air
Pollution
Control
District
          City  of
          Chicago
                 Actual no.  of
                 source testing
                 personnel3
                 4 Team chiefs
                12 Technicians0
                   (6 teams)
                 1 Equipment
                   maintenance
3 Sr.  engineers
1 Team chief
4 Technicians
  (2 teams)
                 1 Sr.  engineer
                 3 Technicians
                   (1 team)
                          Number of
                          manufacturing ,
                          establishments
                             15,200
                                                         6,000
                              9,200
Annual capital
expenditures of
manufacturing
establishments
    $106	
    525
    (785)'
    250
    230
          a)   Excludes  supervisor, secretary,  laboratory,  and  other  personnel  not directly
              related to testing.   (1970 data)
          b)   Source:   County and City Data Book, 1967, U.S. Department  of  Commerce.
              Reported  data is rounded off for purposes of this  table; 1963 data.
          c)   Current plans are for the addition of two more technicians.
          d)   1967 data shown for comparison only.  Source:  U.S.  Bureau of the Census,
              Statistical Abstract of the United States:   1969,  Table No. 1110.

-------
     The 1970 Amendments to the Clean Air Act however,
require greater source test efforts by state and local
agencies.  Since the manpower model was based on the pre-1970
Act, it may tend to underpredict source testing personnel.
This factor should be kept in mind when using the model.

     5.2.2  Test Team Personnel
     Location and access to the sampling ports determines the
size of the basic testing team.  In general, two men are
required for each sampling station.  Agencies currently find
that the team chief can adequately supervise two two-man
teams testing at different sites.  The added work of taking
concurrent samples before and after the control equipment
may however require a full-time fifth man.

     The technicians serving on the test teams should have a
basic understanding of source sampling principles.  Technicians
usually take the sample, record field data, and they may
weigh the sample and the filters.  Technicians are generally
responsible for maintenance and calibration of the test
equipment.

     The team chief is directly responsible for all the field
work and should have a background in engineering or industrial
hygiene.  In smaller programs, he should be an engineer since
he will then report directly to the program supervisor.  In
general he plans the test, supervises the actual extraction
of the sample and may transport the samples to the laboratory.
The team chief should check all calculations.

     In larger programs, the team chiefs will usually report
to a senior engineer.  The senior engineer should preferably
be a professional engineer having a broad knowledge of the
                         36

-------
various industrial processes within the agency's jurisdiction.

He is responsible for all tests and should be experienced in

source sampling and have rapport with process operators.  The

senior engineer is usually the agency's expert witness in

matters involving emission testing.


     5.2.3  Personnel Costs

     Estimated salary requirements for the various functional

positions are presented to indicate the approximate personnel
costs of a source sampling program.  In all cases ranges are

given.  The greatest range exists at the technician level to

represent the spread between the novice and the experienced

technician.  Other positions are affected mainly by experience,

agency size, and geographical location.  Approximate base
                             2
salary ranges are as follows:
     Supervisor                      $15,000 - $23,500
     Senior Engineer                 $13,000 - $19,500
     Team Chief                      $11,000 - $17,000
     Technician                      $ 7,000 - $13,000
     Secretary                       $ 5,000 - $ 7,000

     Overhead rates associated with the base salary must be

included for budgeting.  In general, the New England and

Middle Atlantic regions are areas of highest pay.  Table 5.2

lists the various geographic areas with relative salaries

based on the New England scale being 100%.

     Table 5.2  RELATIVE PAY SCALES OF TECHNICAL
                PERSONNEL BY REGION3
Region
Relative salaries (percent)
New England
Middle Atlantic
South
Midwest
Plains
Southwest
West
           100
           100
            92
            89
            84
            87
            92
a)  Based on the 1969 survey conducted by the National Society
    of Professional Engineers.
                         37

-------
5.3  Equipment and Space Requirements,  and Associated Costs

     5.3.1  Equipment and Costs
     This section describes the major items of equipment
required for a source sampling program.  Incidental items
such as clamps, heating wire, safety equipment, miscellaneous
hardware, and the various pieces of workshop equipment will
not be discussed.  Section 7.4 presents a detailed discussion
of specific equipment needs.

     At least two complete sampling trains as described in
Section 7.4 are required.  These trains include the nozzles
and probe, cyclone/filter collector, impingers, pump and meter
assembly, and associated equipment.  In addition, the agency
will minimally require a desiccator and analytical balance
for drying and weighing the samples.  Provisions must be made
for calibrating the dry gas meters.  A spirometer or bell-type
prover is the best equipment for this purpose.  However these
devices are very expensive and, whenever possible, arrangements
with the local gas utility company for periodic calibration
should be made.  A carefully calibrated orifice may also be
used for calibration.  If the regulations require correction
to 12% C02 or similar basis, an Orsat apparatus will be required.

     Each team requires a vehicle for transportation of equip-
ment (a panel truck or station wagon will suffice).  In addition,
the senior engineer or team chief may require a vehicle for
field use.

     A complete single set of particulate sampling equipment
costs approximately $3,500.  Associated laboratory equipment
and miscellaneous hardware which can be shared by more than
one team costs about $2,000.  Associated equipment costs should
                         38

-------
consider depreciation and maintenance on all equipment,
including motor vehicles, used by the program.   However such
items as office supplies and furnishings are not included
herein.  Travel costs, personnel overhead, and other adminis-
trative costs must of course be figured into the total budget.

     5.3.2  Space Requirements and Costs
     The source sampling unit requires office and workshop
space.  Table 5.3 shows the space requirements of three
existing groups and shows approximately 70 square feet of
shop area per man.  Office space is actually determined by
administrative policy, and must be considered on a basis
of the number of desks.  Minimally 50 sq. ft. is required
for each desk.  Private offices require at least 80 sq. ft.
Space costs in leased buildings are on the order of $5 to
$6 per square foot per year.

Table 5.3  SPACE REQUIREMENTS FOR SOURCE SAMPLING PROGRAMSa
Agency
Chicago
Bay Area
New Jersey
Personnel
5
9
18
Space allocation, square feet
Workshop
-
625
1200
Office
400
400
360
a)  1969 Data.
b)  Does not include clerical or laboratory personnel.

5.4  Administrative Procedures

     5.4.1  Request for Source Test
     Generally, the source testing program exists as a
service to the enforcement, engineering, and permit programs.
As such, requests for source tests are initiated outside the
                         39

-------
unit.  As a rule the enforcement section will request a test
based on information received from its inspectors.   The
request may also be motivated by the agency's counsel or by
citizen complaints especially where visual inspection, both
inside and outside the plant, reveals no apparent violation.
Tests may also be requested to develop emission factors or
for emission inventory purposes.  Oftentimes a source test
will be requested prior to the issuance of an initial permit
to operate.

     As programs progress, the source test unit becomes
more and more knowledgeable of the individual processes.
Therefore, as a practical matter the supervisor of the source
sampling unit will influence the decision to test and the
priority of the test.  After these decisions are made, the
senior engineer takes steps to effect the test and determines
the type of testing desired.

     The form used for requesting a source test should contain
such information as is required to determine test methods,
priority, purpose, and status of the action requested.
Figure 5.7 shows the type of request form used in Los Angeles.
In addition to the basic required information, provision is
made for special instructions to the tester and for determining
the status of the test.

     Figure 5.8 illustrates an automated form used by the
State of New Jersey.  This form also identifies the inspector
and provides space for his comments.  Status of the file is
determined at a glance and a tickler device is incorporated.

     Upon completion of the test report, it should be approved
by the agency's chief and submitted to the requester.
                         40

-------
 Figure 5.7   Request for source test or sample analysis -
               form used in Los Angeles.

                 SOURCE LOCATION DATA
 1.   Firm Name 	 Phone No.
 2.   Address 	 City 	
 3.   Representative to Contact 	 Title 	
                REQUEST INITIATION DATA
 4.   Request Initiated by 	 Division 	
 5.   Request Approved by 	 Date 	
 6.   Reason for Request:
      |  1  Court or Hearing Board        | ]  	>	
          Action Case No. 	
      Q  Permit Pending Appli-         '—'  	
          cation No. 	        I—i
      ||  Suspected Violation

                SOURCE AND SAMPLE DATA
 7.   Type of Request:  [] Source Test || Sample Submitted for Analysis
 8.   Basic Equipment: (incl.  Index Code No.)  	
 9.   Control Equipment:  (incl. Index Code No.) 	
10.  Points to be Tested or Description and Source of Each Example
     Submitted:
11.  Test for Following Constituents:
               (Continued on Next Page)
                          41

-------
      Request for source test or sample analysis
                      (continued)

12.   Special Instructions:
             ACTION BY SOURCE TESTING UNIT
13.  Date Received 	 Priority
14.  Date Sent to Analytical Laboratory     •	
15.  Date Report Issued 	
16.  Distribution of Copies
                        REMARKS
                          42

-------
1
•' • -\
•;y  icsj o


2 1


7421


7421


7 I'' 2*
PRIORITY
1
1 1 T" * 2 1
INSPECTOR

U 3 -L, :•
7421

LOCATION ADDRESS
                 PRIVATE CITIZEN 1  F]
                                                             I INOV. I	| III COUP. |	I
                                                             II PART. FJ  IV GOV.
                                                      NEW EQUIP. INST. 4   | |
                                                      xx.  OPERATIONAL CHANGE [~] 1
MOVE OR PROJ. COMP. [J '
    EOUIP. CHANGE [~\ i
   FOLLOW-UP (11 OR iv)
   SERVICE OF PROCESS
 INSPECTOR'S COMMENTS:
SUPERVISOR'S COMMENTS/INSTRUCTIONS:
                                                                     \rtts
     Figure  5.. 8   Automated source  test  request  form  -
                           State  of  New Jersey.
                                           43

-------
6.0  PRELIMINARY PROCEDURES REQUIRED IN CONDUCTING A
     STACK TEST
     In order to properly plan the stack testing program,
a preliminary survey of the process and test site should
be made.  Information obtained during this pre-survey
will help insure the selection of the proper testing and
analytical procedures, and will provide for a more
organized test plan.
     Except in the most routine cases, an bn-site inspec-
tion or pre-survey will be required to determine certain
physical elements which must be known for stack sampling.
These elements can be sub-divided into process information,
test-site location, and emission parameters.
     Much of this information will be readily obtained from
an on-site inspection.  Gas flow rates and compositions can
frequently be estimated from process throughputs, emission
factors, material balances, and fan and motor type and
size.  Source registration forms and permit applications
will also provide information on the expected emission
characteristics.
     The use of construction permits for new sources can
also insure the proper location of test ports and necessary
scaffolding for future tests.
6.1  Pre-Survey Process Information
     Process information is required in order to determine
approximate emission constituents, volumes, and concen-
trations; and to determine the regulation which applies to
the particular process being investigated.  This information
in turn will also have a bearing on the type of sampling
equipment to be used and on the sampling schedule.
                         44

-------
     A successful stack testing program requires an intimate
knowledge of the process to be tested.   This can only be
obtained through a careful examination of the process and
thorough discussions with plant personnel.  A single
personal contact must be available at the plant.  This
person must have an understanding of the process and must
also have authority to obtain required information and the
cooperation of other plant personnel.  A member of the
plant manager's or plant engineer's staff is a desirable
contact.
     Pre-surveys are greatly facilitated by the use of
questionnaires which list the necessary process parameters.
Figures 6.1, 6.2, and 6.3 are suggested forms for combustion
sources, incinerators, and industrial processes respectively
which can be used for pre-surveys.  These questionnaires
are general guides, and in many specific cases additional
information will be available.  In general, the more
preliminary information obtained, the better.
     The cyclic operation of a process must also be
determined during the pre-survey.  If a process varies
with time over a defined cycle, the variation in emission
parameters during this cycle should be investigated.  In-
formation must be obtained to decide whether to sample over
part of a cycle, over a whole cycle, or over several cycles.
When the process is steady the desired level of operation
must be determined.  Any seasonal variation in process
conditions should also be obtained.
     The exact wording of the applicable regulations may
also have a bearing on the desired process operating
condition during the proposed tests.
                         45

-------
Figure 6.1  Pre-Survey Form for Combustion Sources.
Name of Company

Address
Phone 	 Person to Contact

Date of Survey 	 By 	

Entry Requirements 	
Location and Designation of Boiler to be Tested
Type of Boiler 	 Capacity 	 1000 Ibs
                                                          steam/hr

Type of Fuel 	' _   Steam Pressure 	 psig

   Btu Value 	 Steam Temp. 	 °F

Sulfur Content, % by Weight 	

Fuel Composition-Proximate Analysis	

Fuel Composition-Ultimate Analysis 	
Type and Efficiency of Air Pollution Control Equipment
Is Fly Ash Reinjected?
Collection Efficiency, %
Approximate Opacity of Stack Gasesf% 	

Normal Range of Steam Flucations 	 to

Can Constant Load be Maintained? 	

   If So, How Long?	
                         46

-------
Figure 6.1 .Continued
Conditions Under Which Boiler can be Tested:
     Maximum Steam Load 	
     Expected Fuel Rate 	    	  	
        Can This be Measured?
     Excess Air Rate
     Ash Withdrawal Schedule
     Soot Blowing Schedule
     Provide complete sketchs of entire boiler and flue gas ducting.
Indicate proposed location of test points, sampling port size,
location of fans, location of pollution control equipment,
obstructions at sampling site, necessary scaffolding, final exit
stack dimensions, location of electrical power and typesvpf..v?ts
sockets.
                         47

-------
Figure 6..J2  Pre-Survey Form far Incinerators
Name of Company
Address
Phone	 Person to Contact
r     ~————————^^——
!)
Date of Survey 	 By 	
Entry Requirements	
Location and Name of Unit to be Tested
Type of Incinerator
Capacity 	Tons/hr
Type of Air Pollution Control Equipment 	
Collection Efficiency, % 	
Normal Charging Rate 	
How is Charging Rate Measured?
Operating Schedule	
Type and Quantity of Auxiliary Fuel
Excess Air Rate
Overfire and Underfire Air Rates
T^emperature of Flue Gases at Proposed Test Points
Provide complete sketchs of entire incinerator and flue gas ducting.
Indicate proposed location of test points, sampling port  size,
location of fans, location of pollution control equipment,
obstructions at sampling site, necessary scaffolding, final exit
stack dimensions, location of required electrical power  and
 type of  socket.
                          48

-------
Figure 6;3  Pre-Survey Form for Industrial  Process
  •*      -          ,..,*_•<_... .  . _..,,,.	   ;}.,.	  .     .	:		
Name of Company
Address 	
Phone
Date of Survey
Entry Requirements
Type of Process
Location of Process
Operating Schedule
Process Description
  Person to Contact
 	 By 	
Process Feed  Rates
Expected Emissions
       Type
Concentration
Quantity
                          49

-------
Figure 6.3 ... Continued
Type and Efficiency of Air Pollution Control Equipment
Opacity of Exit Gases
Expected Stack Gas Parameters at Test Location
     Temp. °F
     Pressure, psig



     Volume, acfm
     Composition, %
                  % o2
Ambient Conditions at Test Site(s)



     Temperature 	
     Noxious Gases
     Weather Protection
     Required Safety Gear
Provide complete sketchs of entire process and exit gas ducting.



Indicate proposed location of test points, sampling port size,



location of fans, location of pollution control equipment,



obstructions at sampling site, necessary scaffolding, final exit



stack dimensions, location of electrical power and water, etc.
                         50

-------
6.2  Selection of Test Site
     The primary criterion in selecting the test site is
that the sample extracted from this site be representative
of the main gas stream.  Relatively little is known about
the disposition of particulate within any specific moving
gas stream.  Therefore, every effort is made to obtain a
site in which the particulate/gas mixture is as homogeneous
as possible.  Homogenity is best achieved in straight
vertical ducts.  Ideally, the gas flow should not be
disturbed by any obstruction or change in direction for
approximately 7 to 8 hydraulic diameters  upstream and
2 to 4 diameters downstream from a proposed test location.
     In addition to flow considerations, accessibility to
the site is an important consideration.  Safety, as well
as clearance for the probe and sampling apparatus,
availability of electricity, weather exposure, presence of
toxic or explosive gases, etc., must all be considered in
selecting a site.
     Because of these many considerations, compromises
must be made in test site selection.  However, one should
at least strive for ideal flow conditions.  In some cases,
a suitable test site may not be available without major
changes in the duct work.  If these changes cannot be made,
a meaningful sample may not be practical, and only approxi-
mate emission results will be obtained.
6.3  Preliminary Determination of Emission Parameters
     In addition to general process related information,
more detailed information of the gas stream parameters
at the test site is desirable.  This is especially true
+ Hydraulic diameter = Area of duct cross-section
                                  -   -    ••         X
                              Duct perimeter
                         51

-------
for non-typical processes.  In many cases, the exit gas
composition, volume, and temperature can be approximated
by material balance calculations, readings from process
instrumentation, or by comparison to similar processes
for which data are available.  When no data can be
obtained, exit gas parameters may be determined by in-
serting a probe into the duct at or near the test
site.  In this manner, an approximate velocity and
temperature can be determined.  Color-change detection
tubes using a squeeze bulb sampler can be used to
determine approximate concentrations of a wide variety
of gases.
     A list of the more important items required in
conducting pre-surveys includes:
     1.  50-1200°F dial thermometer (12" stem)
     2.  Velometer
     3.  50-foot tape measure
     4.  Set of basic tools
     5.  Polaroid Camera
     6.  Detection tube samplers
     7.  Pre-Survey forms
     8.  Safety equipment
                         52

-------
7.0  PARTICULATE SAMPLING PROCEDURES
     The particulate sampling procedure used by the Office
of Air Programs utilizes specialized sampling equipment
and analytical procedures to obtain both a filterable and
a non-filterable or condensible fraction of particulate.
Special procedures are also used to insure maintenance of
isokinetic sampling rates.
                                                             4 5
7.1  Measurement of Stack Gas Velocity and Related Parameters '
     Prior to performing any particulate measurements, the
velocity of the gas flowing through the duct at the test
location must be determined.  This velocity determination
is a very important measurement and is composed of a
number of mathematically related parameters; as shown in
Equation 7.1.  Not only is the total gas flow determined
from this velocity measurement, but the sampling rates
also depend upon this value.
     Vs =
                P M
                 s s
Where:  V  = Stack gas          P  = Stack gas pressure
             velocity           M  _ stack gas moiecular weight
        Ts = Temperature        cs = pitQt tube constant
        Ap = Velocity head      / = Constant depending on
                                     units used
                   Equation 7.1

     7.1.1  Location of Traverse Points
     Since the velocity through any cross-sectional plane
area perpendicular to the flow direction is not uniform,
the area must be divided into a number of equal-sized sub-
areas.  The various parameters which affect velocity should
then be determined at the centroid of each of these areas.
                         53

-------
The average velocity is determined by taking the arithmetic
average of the individual velocities; namely:
               V_ =       .
                8        S1
                    Where:  V   = Average velocity
                             s
                            V .  = Average velocity in any subarea
                             SI
                            N   = Number of test points
     The number of subareas required to obtain a reliable
average velocity is not well defined.  When the test site
is at least 8 hydraulic diameters downstream and 2 diameters
upstream of any flow disturbances, twelve areas should be used.
When these conditions cannot be met, Figure 7.1 should be used
to determine the number of required points.  Sampling sites
less than 2 diameters downstream or one-half diameter upstream
from a flow disturbance should be avoided if possible.

     In circular ducts, the cross-sectional area is sub-
divided into concentric areas and measurements are made at
four locations in each area as shown in Figure 7.2.  The
distances to these points which are located along the centroid
of the areas are calculated by Equation 7.2.
     Pj = 50 [I'A/ (2j =
                 v   2a
     Where:  Pj = Percent of diameter from inside of
                  duct wall to measurement point
             a  = Total number of areas selected = N/4
             j   = Number of area  being calculated such
                  as 1,2,3,4 ..... numbered from the center
                  outward
                      Equation 7 . 2
     Equation 7.2 provides only half of the distances needed.
The remaining distances are obtained by taking the difference
                         54

-------
ui
tn
           EH
           Z
           H
           O
           W
           EH
               10
                                       NUMBER OF HYDRAULIC DIAMETERS  UPSTREAM
                                                FROM FLOW DISTRUBANCE
                       NUMBER OF HYDRAULIC DIAMETERS DOWNSTREAM FROM
                               FLOW DISTURBANCE  TO TEST SITE
                          Figure 7.1  Number  of test points.6

-------
between each calculated percentage  and 100.   Table 7.1
presents the percentages determined from Equation 7.2 for
up to 12 areas

     For rectangular ducts, the area should be divided into
approximately square subareas as shown in Figure 7.3.
Measurements are made at the center of each subarea.
                                Indicates sampling point
                                   = P, x Ds
                                        X
                                Where:?, _ is determined
                                       1,2
                                from Equation 7.2 or Table 7.1
    Figure 7.2  Cross-section of a circular flue
     divided into three concentric equal areas,
        showing location of sampling points.
    JJ'lJlf
            "I
             A-
                         Where:  d, =
                     -L  and  0.5 < / D.
Number of areas across
flue width
Number of areas across
flue perpendicular to
width
      \
         < 2
    Figure 7.3  Cross-section of rectangular flue
   divided into twelve equal areas with sampling
     points located at the center of each area.
                         56

-------
 Table 7.1  PERCENT OF CIRCULAR STACK DIAMETER FROM

            INSIDE WALL TO TRAVERSE POINT
Traverse
point
number
along
diameter
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Number of traverse points on a single diameter
6
4.4
14.7
29.5
70.5
85.3
95.6


















8
3.3
10.5
19.4
32.3
67.7
80.6
89.5
96.7
















10
2.5
8.2
14.6
22.6
34.2
65.8
77.4
85.4
91.8
97.5














12
2.1
6.7
11.8
17.7
25.0
35.5
64.5
75.0
82.3
88.2
93.3
97.9












14
1.8
5.7
9.9
14.6
20.1
26.9
36.6
63.4
73.1
79.9
85.4
90.1
94.3
98.2










16
1.6
4.9
8.5
12.5
16.9
22.0
28.3
37.5
62.5
71.7
78.0
83.1
87.5
91.5
95.1
98.4








18
1.4
4.4
7.5
10.9
14.6
18.8
23.6
29.6
38.2
61.8
70.4
76.4
81.2
85.4
89.1
92.5
95.6
98.6






20
1.3
3.9
6.7
9.7
12.9
16.5
20.4
25.0
30.6
38.8
61.2
69.4
75.0
79.6
83.5
87.1
90.3
93.3
96.1
98.7




22
1.1
3.5
6.0
8.7
11.6
14.6
18.0
21.8
26.1
31.5
39.3
60.7
68.5
73.9
78.2
82.0
85.4
88.4
91.3
94.0
96.5
98.9


24
1.1
3.2
5.5
7.9
10.5
13.2
16.1
19.4
23.0
27.2
32.3
39.8
60.2
67.7
72.8
77.0
80.6
83.9
86.8
89.5
92.1
94.5
96.8
98.9
a)   Points numbered from outside wall toward opposite wall.

b)   The total number of points along two diameters would be
    twice the number of points along a single diameter.
                         57

-------
     7.1.2  Velocity  Head  Measurements
     A Pitot tube  and inclined manometer are commonly used
to measure velocities equivalent to at least 400 feet per
minute at 60°F.  A Stauscheibe or type S Pitot tube  is
recommended for velocity head measurements.  This instrument
is shown in Figure 7.4.
Off
\
3! \STAT 1C PRESSURE
ii HOLES OUTER PIPE
! ONLY
|
STAINLESS STEEL TCBING
8d
9d
Stand
             \HPACT PRESSURE OPENING
                                STATIC PRESSURE CONNECTION
                          Standard type Pitot tube
                                      TUBING ADAPTER
                  STAINLESS STEEL TUBING
                   Type S Pitot tube
  Figure  7.4   Pitot tubes usually used to measure
                     velocity head.

     When using  the type S Pitot tube, a correction  factor
must be applied  to the velocity head readings.  This factor
is usually  about 0.85, but can vary between 0.8 and  0.9
depending on  the exact configuration of the openings.  This
correction  factor should be checked by comparing velocity
head readings taken at a point of constant air flow  with a
standard  type Pitot tube which is also shown  in Figure 7.4.
                          58

-------
The correction factor is the ratio of the square root of
the velocity head readings obtained with the Standard Pitot
which has a correction factor of 1.0* divided by the square
roots of the readings obtained by the Type S Pitot tube.  A
sample calibration calculation is shown in Table 7.2.

Table 7.2  EXAMPLE DETERMINATION OF TYPE S PITOT TUBE
                  CORRECTION FACTOR
Standard
Pitot reading
HO
0.3
0.5
1.0
~T^7~
0.5477
0.7071
1.000
Type S Pitot reading
Hl
0.415
0.700
1.44
~"y H,
0.642
0.837
1.200
"VH? = C
Ratiol/iq* P
0.853
0.844
0.833
C = 0.843
     The velocity head is the arithmetic difference between
the total pressure and the static pressure in the duct.  This
difference in pressures is read on an inclined manometer by
connecting the two leads on the Pitot tube to the two ends
of the inclined manometer with flexible tubing, (1/4" O.D.
rubber or Tygon have proven adequate).   The impact or upstream
end of the Pitot tube measures the total pressure and is
connected to the zero end of the inclined manometer.  The
other leg of the Pitot tube measures static pressure and is
connected to the manometer's high side.
*  This should be calibrated by the manufacturer and could
   vary from 0.98 to 1.02.
                         59

-------
     Any suitable manometer may be used to read velocity
head.  However, the accuracy of the velocity determination
depends on the accuracy of the readings obtained.   As shown
in Section 8, the velocity readings are the single most
important factor leading to errors in source sampling work.
Therefore a sensitive, easily read, instrument must be used.
A manometer which can be read to within 1% of the highest
expected reading is desirable.

     Actual velocity head readings should not be taken until
the process has been at the desired operating conditions for
at least 30 minutes.  During this period the distances to
the required measurement points can be calculated and the
Pitot tube marked.  The Pitot tube can be marked with a high
temperature crayon or masking tape.  If the duct has a thick
wall, or a pipe fitting protrudes from the wall, this dimension
must be added to the distances calculated from the duct's
inside wall to the test points.

     Before and during the velocity traverse, the following
precautions should be taken:
     a)  The manometer connections and tubing should be
         checked for leaks, kinks, or foreign matter.
     b)  The manometer should be carefully leveled and the
         liquid column set exactly on zero.  This should be
         done after the Pitot tube has been connected to avoid
         disturbing the manometer.  To prevent any air movement
         from affecting the zero setting, a cloth should be
         held over the end of the Pitot tube.  The zero setting
         and level of the manometer should be checked during
         the test work.
     c)  The Pitot tube must be held at right angles to the
         gas flow and the impact opening of the tube must point
         directly into the gas stream.
                         60

-------
     d)  The test ports should be kept sealed to prevent
         any effect on the readings by air flow.
     e)  In ducts with erratic velocity head readings (a
         common occurrence),  an average value must be taken
         by visual observation.  In taking a visual average
         reading, try to ignore the rapid extreme fluctuations
         in pressure.  Glass  capillary tubes inserted in the
         Pitot tube connecting lines will dampen out some
         fluctuations.
     f)  Only take readings at the designated subarea centers,
         and not at the duct  edges or center.

     Always use a standardized form to record velocity head
readings and other pertinent  test data.  A suggested form
is shown in Figure 7.5.  Any  readings which appear to be
unusually high or low should  be rechecked immediately.

     For very low velocity measurements, a sensitive micro-
manometer must be used to ob,tain accurate velocity head
readings.  With this instrument readings of 0.001 inches of
water can be read.  Micromanometers are very sensitive to
leveling and zeroing errors.   If accurate velocity head
readings are still not obtained, a hot-wire thermoanemometer
or vane-type anemometer may be tried.  These devices must be
calibrated at the temperature at which they are to be used.
These devices do not give accurate readings if particulate
deposits on the device.  When this occurs, the flow must be
estimated based on material balance and/or fan data.

     7.1.3  Temperature and Static Pressure Measurements
     A long (36") dial stem thermometer with a range of 50
to 750°F will provide the best overall temperature measurements
in ducts up to about 40 square feet in area.  Though temper-
atures are usually fairly uniform across any cross-sectional
area, a traverse with the thermometer should be made to
check uniformity.
                         61

-------
a\
                                 Figure  7-5  Gas Velocity  and  Volume  Data-


              VELOCITY TRAVERSE DATA
                               Test No.

                               Location
Point

i
!

1



















Position.
Inches rt
























Total
Average
Reading, A p
"H2°
























—
—
w


























Ts°F


























                                                 Stack Inside Dimensions 	

                                                 Stack Area = 	 sq. ft.

                                                 Barometric Pressure, P,  = 	
                                                 Pressure =          "  „
                                                            	  H20

                                                 Stack Abs. Pressure, P  = 	
                                                                       s
                                 "Hg, Stack  Gage
                                  "H20 + PK =
                                  T3T6"    b
             "Hg
                                                 Stack Gas Temp.,  T  =
                              'F. + 460 =
          'R
                                                 Molecular Weight of Stack Gas, M
1.  V  = 174 C 1/Spfx 29.92 x 28.96 ft/min
     s        P " * s   ^5	   	
                                                                                 M
                                                     Vs =

                                                 2.   Volume =
                      ft/min x
sq. ft. =
cfm
    Standard Volume at 70 °F and 29.92 "Hg:

3.   cfm x 530 x Ps    = _ x 530 x
          Ts   29792
                                                                                           29'92
                                                                                                    scfm
      a) From outside of port to sampling point.
        Pitot tube
        Manometer
        Thermometer
                       Data Recorder

                       Date

-------
     For larger ducts and for high temperatures,  a
thermocouple and potentiometer will be required to
measure temperatures.  In this case, the temperature
readings should be taken at the same points and pref-
erably at the same time that the velocity head readings
are made.  For temperatures in excess of about 750°F a
shielded thermocouple should be used.  When temperature
variations occur, a continuous recording of the thermo-
couple readings will be useful to define the cyclic
nature of the process.
     All temperature data including identification of
the instrument used, should be recorded on the velocity
traverse data sheet  (Figure 7.5).
     Approximate static pressure measurements in the duct
may be made by connecting one leg of a type S Pitot tube
to a vertical U-tube manometer containing either water or
mercury, depending on the expected range of pressure, and
turning the Pitot tube sideways in the duct.  The other
leg of the manometer is open to the atmosphere.  This
static gage pressure may be either a positive  (pressure)
or a negative (vacuum) reading.  It should be determined
to the nearest one-tenth inch of water.*  The absolute
pressure in the duct is obtained by adding this value to
the atmospheric  (barometric) pressure at the test location
(add the value for gases under pressure and subtract for
vacuum readings).  Equation 7.3 illustrates this cal-
culation.
     The atmospheric pressure is determined with an aneroid
barometer, or if available, a Fortin mercury-in-glass
barometer.  The aneroid barometer should be checked and
calibrated with a Fortin barometer.  Temperature corrections
* This measurement is not critical to the velocity deter-
  mination, and may be ignored if it is less than about
  14" of water.  However, this measurement is frequently
  useful from a process or equipment standpoint.
                         63

-------
must be applied to the Fortin barometer when the ambient
temperature is not 32°F.*
           P  = PK
            s    b
                         13.6
Where:     P  = Absolute pressure in stack in inches of
            s   mercury
           P.  = Atmospheric pressure at test site, inches
                of water, measured with a barometer
           p  = Stack gas gage pressure measured in inches
            s   of water
                    Equation 7.3
     7.1.4  Gas Density and Moisture Determinations
     In addition to the temperature and pressure of the stack
gases, their density depends on composition.  Many exit gas
compositions are similar to air.  However, for various chemical
processes, this assumption may not be valid and a chemical
analysis will be required.  For example, gas streams from a
chlorine plant may contain high concentrations of chlorine
which has  a molecular weight of 71.  Since the molecular
weight of air upon which the basic velocity equations are
based is 28.96, a large difference in gas densities could lead
to an error in the velocity determination.

     For most combustion gases, the density is fairly close
to that of air, and if no correction is applied, only a small
error will result.  However, the density may be checked with
an Orsat analysis and the following calculation procedure:
      (Percent CO^ by volume dry basis) x 0.44 =
      (Percent CO by volume dry basis)  x 0.28 =
   Temperature corrections and other useful data are contained
   in ASME PTC 19.2 - 1964 - Pressure Measurement.
                         64

-------
     (Percent 02 by volume dry basis)   x 0.32 =
     (Percent N_ by volume dry basis)   x 0.28 =
     ». • "a l
Where percent C02/ CO, 02 are measured by Orsat apparatus,
and percent N2 = 100 - (% C02 + % GO + % 02),
     M, = Average molecular weight - dry basis
     W  = Volume percent moisture in flue gas  (see Equation 7.4)
     M  = Average molecular weight of actual flue gas
      s
Moisture Content
     Moisture content is best determined after a particulate
sample has been taken since the train used to  collect particu-
late will also collect moisture.  However, a preliminary
estimate of moisture content can be obtained through a knowledge
of the process, a material balance, wet and dry bulb readings,
by passing a measured quantity of gas through an accurately
weighed desiccant, or by condensation techniques.

     Passing a measured volume of stack gas through a container
with an accurately weighed quantity of silica gel has been
used to determine moisture content.  In this case the quantity
of moisture collected, divided by the sample volume will yield
the percent moisture as shown in Equation 7.4.  Care must be
taken not to saturate the silica gel and to provide sufficient
contact time for water vapor absorption.
                         65

-------
     W =       V     x 100
           m    wl
     V
       ,  = (Weight gain of silica gel, grains)  x 0.0474*

Where:  V ,  = ft3 of moisture collected at 70°F and 29.92" Hg
         wl
        V   = Metered volume of dry gas at 70°F and 29.92" Hg
         m                          3
        W   = % moisture in stack gas by volume
                    Equation 7.4

     When the full particulate sampling train is used as
described in Section 7.4,  moisture will be condensed in the
impingers and also absorbed by the silica gel, and this
quantity must be included in the total moisture calculation.
Thus, if V 2 = (moisture condensed out in impingers, ml)
x 0.0474, then:

     w = v—^Tv	T%—  x 10°         Equation 7.5
          m    wl    w2

     7.1.5  Calculation of Velocity and Total Gas Flow
     Calculation of stack gas velocity is not required prior
to sampling if the recommended sampling method is used.  The
average velocity head and other stack gas parameters are
however required.  If needed, the velocity may be calculated
according to Equations 7.6 and 7.7.
          ,  N      \
V =
s ^ j
Vsi = I
7" V . J / N
L = l S1/

-------
Where:  V .   = Stack gas velocity at point i, feet per minute
         S J.
        C    = Pitot tube correction factor, (dimensionless)
        Ap   = Velocity head, inches of water
        T .   = Stack gas temperature at point i, °R
         S X
        M    = Molecular weight of stack gas
         S
        P    = Stack gas absolute pressure, inches of mercury
         S
        N    = Number of sampling points
        K    =174 when units listed above are used

     If the molecular weight of the gas is similar to air,
and the stack pressure is approximately 29.92, this equation
simplifies to:
        V    = 174 C ~\/Ap T  .
         s.          p V r  si

     The average velocity is then the arithmetic average of
all the V  .  If the temperature is similar throughout the
         si
duct cross-section, the average velocity in the duct is
obtained by:  V  = 174 C ~V Ap  x(avg7^/ T  ).  Figure 7.5
               S        P               ®
(the velocity data sheet) provides a convenient form for
computing velocity and total gas volume in a duct.

7.2  Determination of Isokinetic Sampling Rates
     During isokinetic sampling, the velocity of the gas
entering the sampling nozzle is identical to the velocity
in the duct at the sampling point.  In most gas streams,
isokinetic sampling is required to prevent segregation of
the particulate matter and a biased sample.

     Isokinetic sampling rates may be calculated if the
gas velocity, temperature, pressure, nozzle area and gas
metering conditions are known.  These variables are related
as shown in Equation 7.8.
                         67

-------
                 T
                  m     P      100-W,
     Qm.  = Vs.An T-*  X f-  X (±T5CT>.       Equation 7.8
       i     is-      m
     Where :
     Q   = Sampling rate at meter conditions at point i,
       i   ft.-* per minute.
         = Stack gas velocity at point i,  feet per minute
                           2
si   (eq.  7.7)
     A   = Nozzle area,  ft.
     T   = Average temperature of gas passing through dry gas meter, °R
     T   = Average temperature of stack gas at point i, °R
      o *
     P   = Average absolute pressure of stack gas,  in. of Eg
      s
     W   = % moisture in stack gas (eq. 7.4 or 7.5)

     The basic orifice flow rate equation is:
     Qm  = Knm                     Equation 7.9
                PmMs
     Where:
     Q   = Meter flow rate, ft.   per minute
     Km  = Orifice calibration constant, includes orifice
           coefficient and unit conversions
     AH  = Pressure drop across orifice, inches of water

     This relationship is obtained by calibrating the orifice
and plotting the values of Q  vsAH on log-log graph paper.

     When using the procedures described herein, an orifice
with a pressure drop of about 1.84 inches of water at a flow
of 0.75 cfm is recommended.

     The basic isokinetic flow rate equation was given in
Equation 7.8 as
                         68

-------
     Q  =VA  T   xP  x (100-W)            Equation 7.8
      m    s n^    gS.    Too"
                s     m
     V  however was given by Equation 7.7
      S
     V  = 174 C  "\/AP T  x 28.96  x 29.92 *   Equation 7.7
      s        p V    s   -g       -5
                            s        s
     Substituting Equation ,7-7 into 7.8 gives:
     Q  = 174 C I/APT C,N   x  A,, Tm  x PS  x (100-W)
      m        p V   s 1       n              ___
                                  S     III
     Where:  C, = 28.96  x  29.92
                   Ms        Ps      .
     To determine the nozzle size, a sampling rate of 0.75
cfm* is substituted for Q  and Equation 7.8 is rearranged
and solved for A , the nozzle area (Eq. 7.10).  An available
nozzle size near the value calculated is then used to cal-
culate the actual sampling rates at the individual traverse
points.

                                            Equation 7.10
     7.2.1  Calculation Aides
     The determination of isokinetic sampling rates requires
a separate calculation for every traverse point and this can
be quite laborious; especially if stack gas flow conditions
vary with time.  Various aids have been developed to assist
in this calculation.  These aids, if properly used, will reduce
computational errors and time, and provide a more reliable
procedure for obtaining isokinetic rates.
   Any desired sampling rate may be used.  With the equipment
   described here a rate of 0.75 cfm is recommended.
                         69

-------
     A straight line relationship exists between the velocity
head measurements Ap, and the pressure drop AH, across the
orifice flow rate meter used in the sampling train.  This
relationship along with the related variables has been
plotted on nomographs as shown in Figures 7.6 and 7.7.

     Variations in the assumptions used in preparing the
operating nomograph  (Figure 7.7)  are compensated for by the
auxiliary correction nomograph shown in Figure 7.6.  The
only variable not taken into account is the dry molecular
weight of the stack gas which is assumed to be approximately
equal to that of air.

     The following example illustrates the procedure for
using the correction nomograph and the operating nomograph
with these assumed conditions:

     Orifice pressure drop
     AH @ 0.75 cfm = 2.7" HJD from Orifice calibration,
                     see Appendix C.  •
     Pg = Pm = 29.9 in. Hg
     T  = 600°F
      s
     T  = 100°F
      m
     W  = 20% H20
     Avg. Ap = 0.36 in. H20

     Figure 7.6 contains instructions for obtaining the
correction factor C.  With the correction factor determined,
the sliding portion of the operating nomograph is placed so
that C is set opposite the reference mark as shown in Figure 7.7
                         70

-------
           1. DRAW LINE FROM. AH@ TO Tm TO OBTAIN POINT A ON REF. 1,
           2. DRAW LINE FROM POINT A TO % HgO AND READ B ON REF. 2.
           3. DRAW LINE FROM POINT B TO Ps/Pm, AND OBTAIN ANSWER
                  FOR C.
Figure 7.6  Correction Factor  Nomograph.

-------
                                                      0.001—i
ORIFICE READING
   AH
10—-3
0.2
0.1
         Figure 7.7   Operating Nomograph.
                          72

-------
The K factor must now be determined on the operating nomo-
graph.  This point will then be the pivot point which must
be on the straight line connecting Ap and the desired AH.

     Figure 7.7 illustrates this procedure.
     1.  Connect stack temperature to average Pitot reading
          P-
     2.  Select a probe tip diameter as close as possible
         to that indicated in Step 1.  (In this example use
         1/4").
     3.  Aline the actual probe tip diameter with the stack
         gas temperature to determine an artificial Pitot
         reading.
     4.  Aline the artificial Pitot reading with the reference
         mark on the AH axis.
     5.  This line crosses the K axis at the desired K pivot
         point.

     During sampling, AH is determined by connecting the
observed Pitot reading through K, to the AH scale.  As long
as the meter and stack gas parameters do not change very much,
this K factor remains constant.  If large changes are noted,
a new C and K must be determined.

     Unmarked copies of these two nomographs are provided in
Appendix A for the reader's use.

7.3  Non-Ideal Sampling Conditions
     In practice, non-ideal sampling conditions are frequently
encountered due to non-uniform flow distributions and/or flow
variations with time.  Non-uniform flow patterns are caused by
obstructions to the flow caused by fans, bends, dust collectors,
duct transitions, etc.  Cyclic conditions are due to the
operation of the process.  The degree of non-uniformity of flow,
                         73

-------
though usually evident from the configuration of the duct,
can only be determined by making a traverse of the duct with
the Pitot tube as discussed in Section 7.1.1.

     7.3.1  Poor Flow Distribution
     When sampling less than 8 hydraulic diameters downstream
or two diameters upstream from a flow disturbance, the number
of sampling points should be increased in accordance with the
procedures in Section 7.1.1.  Sampling errors will be reduced
if a greater number of subareas are used to determine average
emissions.  When the flow pattern is tangential or spiral in
nature, only approximate results will be obtained, and
modification of the existing duct work should be considered.

     7.3.2  Non-Isokinetic Sampling Conditions
     If isokinetic sampling conditions cannot be maintained,
due to stack gas flow variations or sampling train problems,
a certain error in particulate measurement may occur.  The
degree of error will depend on the departure from isokinetic
conditions, and on the particle size.  Figure 7.8 presents
experimental data on the expected range of error.  Further
work however, is still required to quantify the magnitude
of these errors.  In the interim, one can only try to
achieve isokinetic sampling rates as closely as possible.
     Particulate concentration and emission  are usually
determined by computing the concentration of particulate
and multiplying by the volume of gas emitted  (See Section
7.8).  Emissions may also be computed by determining the
emission rate per unit time (pounds per minute) and
multiplying this ratio by the ratio of stack area divided by
sampling nozzle area.  By calculating the mass emission
                         74

-------
EH
as
u
H
W

§
Q
W
EH
U
W
O
U

EH
ffi
O
M
D
EH
U
O
H
     2.5J
2.0-
1.5-
1.0
                    80-100 micron
5-25 -""''
 micron
                  0.5       1.0       1.5       2.0

        RATIO OF NOZZLE VELOCITY TO ACTUAL STACK VELOCITY IN DUCT
        Figure 7.8  Expected Errors Incurred by Non-Isokinetic
                    Sampling8.
                    (These data should not be used to correct
                    concentrations obtained under non-isokinetic
                    conditions since a wide variety of particle
                    sizes is usually present
        rate by these two methods and comparing the results, the
        ratio of isokinetic rate actually achieved may be determined.

        The two methods yield identical results under exactly
        isokinetic sampling conditions.  By selecting one calculation

        method or the other, or by averaging the two, more accurate
                                      9
        emission data can be obtained.

             7.3.3  Cyclic Flow Conditions

             When gas flows and emissions vary with time, each
                                 75

-------
point should be sampled for a complete cycle.   For long
cyclic periods, each point may be sampled for  a 3 to
5 minute period and the entire duct traversed  2 to 3
times.  At times, exit gas particulate concentrations
and flows will be non-uniform and unsteady.  This presents
a difficult sampling condition since the duct  should be
traversed and each sample should extend over a whole cycle
or specified number of cycles.  For large ducts, or when
long cycles are encountered, the total sampling time can
become very long.  The use of a number of sampling trains,
operating simultaneously will reduce total sampling time.
7.4  Particulate Sampling Equipment
     A wide variety of sampling trains are available for
determining particulate emissions.  These trains have
been described in the literature, and each has its
particular advantages and disadvantages depending on the
                                               4511
sampling conditions and the object of the test. ' '
In all cases however, the trains consist of a  carefully
sized sampling nozzle or probe tip, a probe to convey
the gases, a filter or solid/gas separating device, a
pump, and a gas meter.  When hot gases (greater than
about 150°F) are sampled, a condenser or similar cooling
device is also used to protect the pump and meter.
     7.4.1  Description of the Office of Air Programs'
            Sampling Train 7»12
     The particulate sampling train recommended and used
by the Office of Air Programs is designed to measure both
solid and non-filterable or condensible matter.  The
sampling apparatus consists of a removable probe tip, a
heated probe, cyclone, heated filter, four impingers
connected in series, air-tight vacuum pump, dry gas meter,
                         76

-------
and an orifice flow meter as shown in Figure 7.9.   The use
of the cyclone is optional and is only used for high
expected grain loadings of particles greater than about
5 microns in size.  This train is designed for high
particulate collection efficiency and for ease in main-
taining isokinetic sampling rates.
     This train collects particulate matter on a filter
maintained at about 250°F, and additional matter in the
cooled impingers which operate in the range of 50-70°F.
Thus, both a filterable and a non-filterable fraction
of particulate matter are obtained.  The use of a filter
outside of the stack requires heating of the probe and
filter to prevent condensation on the filter and sub-
sequent high pressure drop.  The use of an air-tight
vacuum pump before the meter simplifies the calculations
needed to determine and maintain isokinetic flow rates.
     As seen from Figure 7.9, the train consists of a
button-hook type nozzle or probe tip which is connected
with a coupling to the probe sheath.  A glass probe is
inside the metal sheath.
     The probe connects to a cyclone and flask when used
in the train.  The cyclone connects to a very coarse
fritted glass filter holder, which holds a 2-1/2 inch
diameter, tared glass fiber filter?  When the cyclone is
not used, the probe connects directly to the filter holder
through an adapter tube.  The cyclone, flask, and filter
are contained in an electrically heated, enclosed and
insulated box, which is thermostatically maintained at a
minimum temperature of 250°F to prevent water condensation.
Attached to the heated box is the ice-water bath contain-
* MSA type 1106 or equivalent.
                         77

-------
                                       HEATED AREA
00
THERMOMETER J  CHECK
          /   VALVE
               ~-PROBE

               ^
              ^*\
              •
-------
ing four impingers connected in series with glass ball
and socket joints.  The first impinger receives the gas
stream from the filter.  This impinger is of the
Greenburg-Smith design, but is modified by replacing
the tip with a 1/2 inch ID glass tube extending to 0.5
inch from the bottom of the flask.   This impinger is
initially filled with distilled-deionized water*  The
second impinger is  a  standard  Greenburg-Smith
impinger with tip, and is also filled with distilled-
dionized water.  The third impinger, is a dry Greenburg-
Smith impinger modified like the first.  The fourth
impinger is modified like the first, and contains
approximately 175 grams of accurately weighed, dry
silica gel.
     From the fourth impinger the sampled gas flows
through a check valve; flexible rubber vacuum tubing;
vacuum gauge; a needle valve; and air tight vacuum pump,
rated at 4 cubic feet per minute at 0 inches of mercury
gauge pressure, and connected in parallel with a by-pass
valve; and a dry gas meter rated at 1 cubic foot per
revolution.  A calibrated orifice completes the train and
is used to measure instantaneous flow rates.  The three
thermometers used in this train are dial type, with a
range of 25°F to 125°F.  A fourth thermometer in the
heated portion of the box has a range up to 500°F.  The
manometer is an inclined type graduated in hundreths of
an inch of water.  A similar manometer, depending on the
expected range, is used to read the velocity head sensed
by the Pitot tube.
     Depending on the source being sampled, various
*  Usually 100 ml are used.  Other liquids may also
   be used if a particular gas is to be absorbed.
                         79

-------
modifications of this train may be made.  For example
for a very tarry particulate the filter may be moved to
a point after the third impinger to avoid plugging.
     7.4.2  Assembly and Testing the Train
     Before assembling the various sampling components,
the following procedures should be performed in the
laboratory.  These procedures should be completed before
each test series.
     It is especially important that all components which
contact the sampling stream be carefully cleaned.  Proper
cleaning and lubrication, as described in Appendix B
will also insure a leak-tight assembly.
     Any other suspected malfunctions in the sampling
train are also best diagnosed and fixed in the laboratory
or shop.  Frequent sources of mechanical problems include
defective pumps  (usually broken or stuck vanes),  dry gas
meter (erratic dial readings), timer or clock malfunctions,
loose or broken electrical wires, damaged nozzle or Pitot
tube openings, and cracked glass parts.
     7.4.2.1  Calibration of Train Components - In
addition to the Pitot tube calibration described in
Section 7.1.2, the following calibrations should be
made periodically.
     Meter and Orifice - The dry gas meter should be
     checked against a primary standard such as a
     large wet test meter, an accurate orifice, or a
     bell-prover.  With the meter accuracy determined,
     the orifice mounted on the meter outlet can be
     calibrated as described in Appendix C.
                          80

-------
     Sample Box Thermostat - The thermostat in the
     sample box heated compartment is calibrated by
     comparing its set temperature with a mercury
     thermometer.   To accomplish this, the heater
     and blower are turned on and the thermostat
     is set at 250°F.  When the temperature
     stabilized, the reading is noted and the
     thermostat adjusted if required, to yield a
     value of 250°F.  The thermostat scale is then
     also adjusted to indicate 250°F.
     Thermometers  - All thermometers used in the
     sampling train, and the stack gas thermometer
     should be periodically calibrated at a point
     which is near their expected operating range.
     For lower temperature ranges, this may be
     accomplished by placing the stem into hot
     water and comparing the readings to a mercury
     bulb thermometer in the same water.  For
     higher temperatures, an oven or hot gas stream
     may be used for calibration checks.
     7.4.2.2  Assembling Train Components - The basic
assembly of the sampling train for field use is
facilitated by the use of two basic units or modules
and connecting hardware.

     The first module is the sample box.  It consists of
the probe; the cyclone (if used) and filter which are
placed in the heated portion of the box; and the four
impingers in the cooled portion of the box.  Before
loading, a numbered  and  tared fiber glass  filter  is placed
                         81

-------
into the filter holder (rough face should face upstream)
and the filter number is recorded on the meter data
sheet.  The cyclone and filter holder are clamped to-
gether at the ground glass ball and socket connectors
with positive lock pinch clamps.  A light coating of
silicone grease is applied to the outer portion of the
male ground glass joints.  The inlet to the cyclone is
then temporarily sealed until the train has been
completely assembled and checked (a glass ball held
in place with a pinch clamp has been found useful for
sealing the cyclone).
     The impingers are then placed into the cold section
of the sampling train and the first two impingers are
each filled with 100 ml of distilled, high resistance,
deionized water.  The third impinger is dry and the
last impinger is filled with about 175 grams of weighed
( + 1 gm) dry silica gel,  (indicating type, 6-16 mesh).
The impingers are connected with U shaped connectors and
positive lock pinch clamps.  The first impinger is
connected to the filter outlet with a glass "L1 shaped
adapter.
     The second module consists of the meter box and
contains the vacuum pump, dry gas meter, manometers,
flow control valves, the orifice flow meter, timer, and
associated connecting tubing and wiring.  This module
is preassembled and requires no internal field connections.
     The sample box and meter box are connected by an
umbilical line consisting of the main vacuum line, the
two Pitot tube connection lines, four electrical wires
(2 for the probe heater, and two for the sample box),
and a ground wire.  The vacuum line is attached to a
                         82

-------
check valve-thermometer assembly on the last impinger
and connected to the meter box pump inlet with quick
disconnects.
     With the meter box and sample box connected, the
heaters can be started and ice and water placed into
the ice-water portion of the sample box.  The train
can then be checked as described in Section 7.4.2.4.
A probe of the desired length is then selected and
the appropriate nozzle is attached to the front of the
probe and the probe is marked with crayon or tape to
indicate the described sampling point locations.  If
not permanently mounted, the Pitot tube and stack gas
thermometer are now also attached to the probe.
     7.4.2.3  Testing the Sampling Train - To test the
sampling tra.ln for proper functioning prior to a field
test, the train should be completely assembled, the
heaters turned on and the manometers set to zero.  The
vacuum pump is then turned on (with the vacuum line
not connected) and a quick check is made of the orifice
calibration at  0.75  cfm to determine any malfunction
in the meter or orifice connections.  All thermometers
should be checked at this time to make sure they are
reading approximately the correct values.  The train is
then checked for leaks by plugging the cyclone inlet,
adjusting the vacuum to 10 and 15 inches of Hg respectively
by closing the pump by-pass valve, and checking for flow
through the dry gas meter.  If a leakage rate greater than
0.02 c.fm is obtained, the train should be checked for
leaks.  The final leakage rate should be carefully noted
since the leakage volume should be subtracted from the
actual sample volume during actual test work.  After the
                         83

-------
leak test, release the vacuum by slowly unplugging the
cyclone inlet before opening the by-pass valve and
shutting off the pump.  Failure to follow this procedure
may cause air to flow backwards through the train and
will rupture the filter.  The probe assembly should now
be connected to the cyclone.
7.5  Sampling Procedure
     7.5.1  Location of Sampling Points - Under most
conditions the location of sampling points will be the
same as those used for the velocity traverse as deter-
mined in Section 7.1.1.  However, points with no positive
velocity readings should not be sampled.  When this occurs
read the velocity and proceed to the next point.  Only
points which lie at the centroid of the sub-areas should
be sampled.
     7.5.2  Length of Sampling Period - Each traverse
point should be sampled for an equal time increment.  A
five minute sampling period per point is desirable,
however 3 minutes is an acceptable minimum.  A one-hour
total sampling period is usually the minimum total sampl-
ing time for one test.  However, this may vary considerably
depending on the process.  At least two tests should be
made.  Any test, which upon completion, is found to have
contained an error in sampling or analysis, or which is
not within + 20% of the calculated isokinetic rate should
be repeated.  During cyclic operation, at least one
complete cycle should be sampled to obtain an average
particulate emission value.
     7.5.3  Operation of Sampling Train - After all the
equipment has been checked and found to be functioning
                          84

-------
properly, the data at the top of the Particulate Field
Data Meter sheet (Figure 7.10) should be filled in.  The
initial dry gas meter reading should now also be care-
fully taken.
     The cover is then removed from the nozzle tip, and
the probe, along with the temperature-indicating device
and Pitot tube is placed in the duct until the nozzle
reaches the first sampling point.  The Pitot reading and
the desired AH found on the nomograph are recorded along
with the stack temperature.  Stack pressure can be ignored
if it was found to be insignificant during the preliminary
traverse.
     To begin testing, the on-off valve is placed in the
off position, the bypass valve is completely opened, and
the timer is set at zero*  The clock time is recorded,
and the vacuum pump is turned on.  The actual AH is
adjusted to match the desired AH by first turning the
on-off valve to on and adjusting the pump by-pass valve.
     7.5.4  Recording Data during Test Period - During
the test period pertinent data relating to the operation
of the sampling train must be recorded in order to insure
the proper operation of the train, validity of the
sample, and to provide necessary data for subsequent
computations.
     Figure 7.10 is a sample field data form which may
be used while the particulate sample is obtained.  All
data should be carefully entered immediately by the
operator.  In addition, any unusual observations in meter
readings or process conditions should be noted since
these might explain any results which appear to be
   When the stack gas is under more than about 1" of Hg
   gage pressure, the on-off valve should be left in the
   on or open position to avoid pressure build-up in the
   train.  Sampling must then start as soon as the probe
   is inserted.
                         85

-------
                           Figure 7.10  Particulate Field Sampling Meter Data.
        Plant
                                    Filter No.
        Run Number
        Location	
        Date
                                    Barometric Pressure, in. Hg
                                    Assumed Moisture, %
Time
        Operator
Assumed Meter Temp., °F_
Stack Gage Pressure	
                                                                            -"H2°
        Sample  Box Number
        Meter Box Number	
        AH@
                                    Probe Tip Diameter,  in.
                                    'C1  Correction Factor
Point












Time
Min.












Dry Gas Meter
Volume
ft3












Inlet
Temp.
oF












Outlet
Temp.
op












Velocity
Head Ap
"H2°












Orifice AH
"H2°












Pump
Vacuum
"Hg












Box
Temp.
OF












Impinger
Temp.
oF












Stack
Temp.
oF












00
a\
        COMMENTS:
                                                                 Leakage Rate @   "Hg =

-------
anomolous.  These readings should be taken at the beginning
of each sample point, or if sampling at only one point, at
five minute intervals.
     The initial and final dry gas meter readings are most
important.  The Pitot readings and the stack temperature
readings are also important since they will be used to
compute stack gas flow upon completion of the test.
     When testing has been completed, the vacuum pump
should be turned off and the final set of readings taken.
The heater, blower, and probe heat switches are turned
off and the probe is removed from the sampling port.  The
nozzle tip is covered as soon as possible in order to
avoid contamination or loss of sample.  The probe clamp
on the front of the sample box is loosened, and the probe
is disconnected from the cyclone inlet.  Both the end of
the probe and the inlet to the cyclone are covered.  After
the umbilical cord has been disconnected from the sample
box, the last impinger is covered, and the probe and sample
box are moved to the sample cleanup area.
     Various process parameters must also be recorded
during the test period.  The exact type of process data
to be obtained will of course vary with the process.  As
a general guideline, all factors which have a bearing on
the emissions should be recorded on approximately a 15
minute interval.  These factors will include process or
fuel weight rate, production rate, temperature and pressure
in the reactor and/or a boiler, control equipment, fen and/
or damper settings, pressure drop or other indicator of
particulate collection efficiency and opacity of exit
plume.  Figures 7.11 through 7.13 provide sample forms for
                         87

-------
Figure 7.11  Boiler Operating Data,
Test No.
Date
Name of Company
Location and Description of Boiler
Type of Boiler

Type of Fuel 	

Date Recorder


Time
 Capacity
1000 Ibs
steam/hr
Fuel Rate

Steam Rate
 1000 Ibs/hr

Combustion
 Air Rate,
 1000 Ibs/hr

Steam Pressure

Steam
 Temperature

I.D. Fan - RPM

I.D. Fan - Amps

Pressures H_0

  Furnace Outlet
  Collector Inlet
  I.D. Fan Inlet

Plume Opacity

Fuel Composition  (As Weighed)
   Btu/lb
  % Moisture
  % Ash

  % S
  % Volatile Matter
  % Fixed Carbon
         Ultimate Fuel Analysis
                         88

-------
Figure 7.12  Incinerator Operating Data.
Test No. 	 Date



Name of Company	
Location and Designation of Unit
Type of Incinerator
Type of Control Equipment



Type of Grate 	



Grate Speed 	
Type of Refuse Burned
Approx. Moisture Content
                         89

-------
        Figure  7 .12 (Continued)


        Data Recorder  	
rime











rot.
Material
Charged
(Ibs)











Tot.
Primary Chamber Draft
Overf ire
("H20)











Avg.
Underf ire
("H20)











Avg.
Secondary Chamber
Draft
("H20)











Avg.
Temp.
(°F)











Avg.
Plume
Opacity
(%)











Avg.
I.D. Fan
RPM











-
Amps












vo
o
        % of time Afterburners are in operation

          Fuel rate to afterburner 	

-------
Figure  7.13  Process Operating Data.


 Test No.  	 Date

 Name of Company 	
 Location and Description of Process
 Capacity and Characteristics of Process and/or Product
 Raw Materials

 Fuel Used
 Time
 Raw Material Feed Rate
 Fuel Rate
 Reactor Temp.
 Reactor Pressure
 Product Rate
 Sidestream Rates
 Recycle Stream
  Rates

 Exit Plume
  Opacity
                         91

-------
combustion, incineration, and process sources respectively.
Pertinent data obtained in the preliminary plant survey
(Section 6.1)  should also be checked at this time.
     7.5.5  Sampling Problems - Problems encountered
during actual sampling are equipment malfunctions and in-
ability to maintain isokinetic flow due to a high pressure
drop through the train.  Malfunctions can best be pre-
vented through a comprehensive, routine maintenance program
and a careful check of the equipment before starting to
sample.
     Increased pressure drop through the sampling train
is usually caused by a build-up of particulate on the
filter.  To try to prevent this, the temperature in the
filter box can be increased to about 300-350°F.  Spare
filters, mounted in their holder should also be prepared
prior to testing in order to facilitate replacement with
a new filter.   If the filter is kept in a preheated box,
sampling can be restarted almost immediately.  The
number of the new filter must be recorded immediately on
the field data sheet, and the time of test interruption
noted on the sheet.
7.6  Disassembly and Particulate Clean-out Procedure
     Upon completion of the sampling run, the sample box
is disconnected from the meter box and allowed to cool.
The probe may be disconnected for ease in handling and
its open ends carefully sealed.  The inlet and outlet of
the sampling train should also be sealed and the train
transported to a clean area for disassembly.  The various
sampling train components are disconnected, one at a time
and the collected sample is removed and placed in a
                         92

-------
numbered container.  A record of the containers and
the samples should be made, and the record should
accompany the samples to the lab (see Figure 7.14).

     First Container - Filter Holder - Remove the
fiber glass filter paper from the holder and place
it into a glass or inert plastic container.  Use forceps
in handling the filter.  Any segments of the filter
which adhere to the holder should be scraped off and
included with the filter.  Seal the container with
masking tape and mark it appropriately.

     Second Container - Probe, cyclone, cyclone flask,
front half of filter holder and connecting tubing -
The inside of these components is wiped with a rubber
policeman and any loose particulate placed into the
containers holding the probe contents.  To remove all
particulate and organic matter adhering to the inside
walls, these parts should be rinsed with acetone and
washed into the same container.

     Third Container - Impinger liquids - Carefully pour
the water from the first three impingers into a graduated
flask and record the volume to within + 2 milliliters.
When determination of condensibles is desired, this water
should be quantitatively poured into a container.  The
first three impingers and all connecting tubing are then
rinsed with distilled-deionized water into the  same
container.  If any visible particulate appears on the
fritted glass filter support or the back half of the
filter holder, these should also be added to this
                         93

-------
container.  The container is then sealed with masking
tape and labelled.  If the irapinger contents are not
to be measured, the impinger solution may be discarded
after measuring its volume.

     Fourth Container - Silica Gel - The silica gel
from the fourth impinger should be quantitatively trans-
ferred to a glass or inert plactic container designated
as No. 4 and sealed.  Use only dry brushing to remove
the silica gel - no washing.

     Fifth Container - Organic Matter - To insure re-
moval of any condensed organic matter which tends to
adhere to the inside walls of the glassware, the fritted
glass filter support, the back of the filter holder, the
first 3 impingers, and all connectors are rinsed with
acetone into a container.  This container is also sealed
and labelled.  This step may be omitted if this sample
fraction is not desired.

7.7  Particulate Analysis
     After the particulate fractions have been placed in
their respective sealed containers, the containers should
be carefully packed in a locked box, and promptly trans-
ferred to the laboratory.  In the laboratory, the follow-
ing analytical procedures should be performed on each of
the sample containers.

     First Container - The filter and any loose particulate
or pieces of filter in this container should be quanti-
tatively transferred to a tared weighing dish.  This
                         94

-------
material is then dried in a desiccator until a constant
weight is obtained*  For highly organic particulate
matter a drying period of 2 to 3 days should be used.
After drying, weigh the sample and weighing container
on an analytical balance to the nearest 0.5 milligram.
Record the weights on a standard laboratory form as
shown in Figure 7.14.
     Second Container - The acetone washings from the
container are quantitatively rinsed with acetone into
a clean, small tared beaker and evaporated to dryness
at 70°F + 10°F and at atmospheric pressure.  The beaker
and residue are then placed in a desiccator for 24
hours, and then weighed to the nearest 0.5 milligram.
Record the data on the form illustrated in Figure 7.14.
     Third Container - The water solution from the
impingers is placed in a separatory funnel, and extracted
with three - 25 cc portions of chloroform followed by
three 25 cc portions of ethyl ether.  The ether and
chloroform extracts are combined and transferred to  a
clean tared beaker and evaporated at 70°F + 10°F and
one atmosphere pressure to dryness under a hood.  The
sample is then dried in a desiccator for 24 hours and
weighed to the nearest 0.5 milligram.  The water remain-
ing after extraction is placed into a tared beaker and
evaporated at 212°F.  The residue is dried and weighed.
     Fourth Container - The silica gel and its container
is weighed to the nearest gram.
     Fifth Container - The acetone washings in this
container are quantitatively rinsed into a clean tared
beaker and evaporated to dryness at 70°F +_ 10°F and  one
 *  Desiccate  at  70°F +  10°F under  an atmosphere with  a
   moisture content of  less than 0.75% by volume.
                          95

-------
            (This  form to  accompany  samples  from field to lab)
 WEIGHT OF PARTICULATE COLLECTED
Plant
FIELD CONTAINER
NUMBER
Filter
NO.
Probe and
NO. cyclone
Impinger
Organics
No.
Water
No> Solubles
Acetone
Washings
No. y
TOTAL FINAL
WEIGHT
(mg. )





TARE WEIGHT
(mg. )
Filter
Container





WEIGHT
GAIN
(mg. )





Run No.
Location
Rec'd froirv Rec'd by
Analyzer
Date
Analyzed

TOTAL WEIGHT OF PARTICULATE COLLECTED, W = x 2.2 X 10~6 = Pounds
x 15.4 x 10 = Grains
VOLUME OF MOISTURE COLLECTED

Impingers
Silica Gel
Container
No.
FINAL
VOLUME
(ml.)


INITIAL
VOLUME OR
WEIGHT
-200 ml.
gm.
VOLUMETRIC
GAIN
(ml.)



TOTAL VOLUME OF WATER COLLECTED, V = ml. x 0.047 = scf
                            Figure 7.14  Particulate Analysis Data.

-------
atmosphere pressure under a hood.   The beaker and residue
are then desiccated for 24 hours and weighed to the
nearest 0.5 milligram.
7.8  The Test Report
     The emission test  report should contain all of the
pertinent data leading  up to the test, a description
of the process and the  operating conditions under which
tests were made, the results of the tests, and test
procedures.  The test report should enable a technically
trained person to understand what was done, and what the
results were.  The test report may at times represent a
piece of legal evidence and must therefore be carefully
prepared.  Summaries of field test data should be included
in order to allow a knowledgeable engineer to check the
results and to obtain an idea of the accuracy.
     7.8.1  Format of Test Report
     The exact format of the test report and the extent
to which each section of the report is developed, will
vary widely from agency to agency and depend mainly on
the intended use of the finished report.  A suggested
format of the test report is presented in Table 7.3.
     Table 7.3  TYPICAL FORMAT FOR TEST REPORT
     1.  Test Objective
     This introductory section presents the reasons for
performing the test series, the location of the tests,
the process(es) which were tested, the location of test
sites, the emissions measured, the test team and owner's
personnel, the dates of the test work, and any other
special comments or background information which are
pertinent to the test purpose and the results.
     2.  Summary of Results
     A summary will serve to provide a reader with a
                         97

-------
short synopsis of the tests and a tabular summary of
pertinent operating and emission data.

     3.  Process Description

     A description of the process and a schematic
diagram of the flow of materials through the process
are desirable in order to provide the reader with an
understanding of the process.  The test locations
should be clearly indicated on this schematic diagram.
Tables of process weight rates, temperature, gas
flows, production rate, etc. which occurred during
the test period should be included in this section.
Capacity of the process equipment should also be
included.

     4.  Test Results and Discussion

     A detailed summary of all test results must be
presented for each test run.  A discussion of these
results pertaining to their reliability and their
relation to the process may also be presented.
Variations in emissions should be explained.

     5.  Sampling and Analytical Procedures

     The sampling techniques and analytical procedures
used to obtain all emission results should be listed
and referenced to a standard method.  Modifications
to the sampling techniques should be carefully
explained when used.

     6.  Appendices
     The appendices should contain summaries of the
detailed field test data and they may contain a
summary of applicable regulations.

     7.8.2  Presenting the Results

     Emission data should be presented in readily under-
stood tabular form.  The results should be related to the

particular process or test condition in the summary

tabulations.  The units used to express the results will

vary with the object of the stack test.  In most cases,

emissions on a pounds per hour or pounds per ton of process

weight should be presented, in addition to a concentration

value.  In all cases, units identical to those used in the
                          98

-------
local regulation should be used.  Specify the temperature
and pressure used to convert gas volumes from stack
conditions to standard conditions*  Clearly indicate if
a concentration value has been converted to the dry basis,
and/or to a certain excess air and/or % CO. value.
     Figure 7.15 presents a suggested data summary for
particulate emissions from fuel combustion processes.
Similar tabulations should be used in presenting emission
data from other processes.  A summary table for present-
ing test results which may be used in the Appendix to the
report is shown in Figure 7.16.
     7.8.3  Example Calculations
     Reporting of emission results in a useable form
always requires some calculations.  These calculations
are best illustrated by the following example which uses
equations previously presented in this section.
     7.8.3.1  Determination of Stack Gas Volume - Assume
the following parameters were measured as explained in
Section 7.1.
     Stack dimension  60" x 72"
     Stack Area, A  =30 ft2
                  5
     Barometric Pressure, P,  = 30.0" Hg
     Stack Gage Pressure, P  =1.4 "H_0; Ps = 30.0 - 1.4 = 29.9
                           3         2              1376
     Stack Temperature, T  = 600°F = 1060°R
                         O
     Average Square Root of Velocity Head~yAp' = 0.55" Hg.
     As determined from Pitot tube traverse (Figure '/.5)
     M.W., Molecular Weight of Stack Gas is similar to
     air (28.96)
     Used Type S Pitot Tube Cp = 0.85
   70°F  (equivalent to 530°R) and 29.92 "Hg are usually
   used as standard temperature and pressure.
                          99

-------
             Figure 7.15  Format for Presenting Emissions from Fuel Combustion Units.
o
o
Test
No.

Steam Rate
1000 Ibs/hr

Fuel
Rate

Flue Gas
Volume Temp
SCFMa °F

%
co2

%
H20

Grs/
SCF

Lbs/
1000
Ibs

Ibs/ fi
10°BTU

             (a)   Standard  cubic  feet  per minute  at  70°F  and  29.92"  Hg.

-------
                                                          4
                 Figure 7.16 Summary of  Emission Test Data
 Test No.
 Date 	
 Name
 Address 	
                                          Process  Tested
 1.   Sampling Station 	
 2.   Material Collected 	
 3.   Operating Condition 	
 4.   Avg.  Flue Gas Velocity,
     Ft./min.  	
 5.   Avg.  Flue Gas Temp.,  °]
 6.   Area  of Duct, Sq.Ft.  _
 7.   Gas Flow Rate,  SCFM 	
 8.   Sampling Nozzle Diam.,
     Inches
     Avg.  Meter Sampling Rate,
     CFM
10.   Testing Time,  Min.  _
11.   Avg.  Meter Temp.,  °F
12.   Sample Gas Vol.  -
     Meter Cond.,  CF  	
13.   Sample Gas Vol.-
     Standard Cond.,  SCF
14.   Water Vapor
     Condensate, ml.  	
     Volume, SCF
15.  Total Sample Gas
     Volume,  SCF
16.  Weight Collected,
     Grams
17.  Total Weight, Grams
18.  Concentration, Grains/SCF
19.  Concentration, Grains/SCF
     @ 12% C02
20.  Concentration, percent
     by volume 	
21.  Concentration, PPM
     by volume
22.  Emission Rate, Lbs/Hr.

     COLLECTOR EFFICIENCY

23.  Material to Collector,
     Lbs/Hr.	
24.  Loss to Atmosphere,
     Lbs/Hr.
25.  Efficiency, %

 Test Conducted By
 Analysis By
 Calculations By
                              101

-------
     Gas Velocity, V  = 174 Cp~V~ATx  29.92 x  28.96'=
                    S             P  S    ~P       M.W.
                                           s
     174 • 0.85  • 0.55""\/1060 x 29.91P
                        V        29.9
     V  = 80.5~\/1060'= 2620 feet/min
                                  o
     Q  = Volume = A  x V  =30 ft  x 2620 ft/min =  78,500  cfm
      S             S    S
     The volume at standard conditions of 70°F and 29.92  "Hg
is :
     Q  =Q  x 530 x P  = 78,500 x 530 x 29.9  = 39,200 scfm
          5   Ts   29^92         106°   29'92
     This volume has the composition of the actual gas
stream, but has been converted to standard conditions.
Frequently it is desired to express this  volume on a dry
basis.  This may be done by factoring out the fraction of
this volume due to moisture.  Thus if the gas has a
moisture content, W, of 10% (as determined in Section
7.1.4) the dry volume would be Q   (100 - W) or
                                ss    __
39,200 (100 - 10) = 35,300 scfm (dry).
          100
     The gas volume may be converted to a weight basis by
multiplying by its density at a given temperature and
pressure.  Densities are usually determined by comparing
the molecular weight of the gas with that of air, i.e.,
density of gas = M.W. of gas x density of air.
                    28.96
     In this example the molecular weight of the gas is
very similar to air and therefore its density is similar;
namely 0.075 pounds per ft3 at 70°F and 29.92 "Hg*   The
quantity of dry gas emitted on a weight basis is therefore:
     35,200 ft3/min x 0.075 pounds/ft3 =  2640 pounds of
                                          dry gas per minute
     7.8.3.2  Determination of Sample Gas Volume - The
sample gas volume is equal to the gas which passed through
*  Density of air at other conditions is  obtained by the
   equation: 0.075 x 530°R x Pressure
                     Temp.   29.92 "Hg

                        102

-------
the dry gas meter and the equivalent volume of water vapor
trapped in the sampling train.  If significant, the air
in leakage should be subtracted from the meter volume.
This quantity is then converted to a standard temperature
and pressure condition as shown in Figure 7.17.
     The particulate concentration is equal to the
particulate weight divided by the sample gas volume, and
the total particulate emission is equal to the product
of the concentration and total stack gas flow.  One must
be very careful not to multiply concentrations or volume
factors which are not at the same temperature, pressure,
and moisture conditions.  Any basis may be used, but it
must be consistent throughout the calculation procedure.
     7.8.3.3  Check on Isokinetic Flow Rate - A check
on the rate of isokinetic flow actually maintained during
the test period can be estimated on an average basis.
This calculation does not insure that isokinetic flow was
maintained at every instant, but it does give the average
percent of isokinetic flow maintained at each sampling
point.  This equation which appears on the bottom of
Figure 7.17. determines the ratio of the average stack
gas velocity to the average velocity in the nozzle and
should be between 80 and 120 at each point.  Before
computing individual points, a check on the test as a
whole should be made to see if it falls within the specified
limits.
     7.8.3.4  Converting to Other Emission Standards -
Frequently emissions must be expressed on a basis other
than pounds per hour or grains per SCF.  Other emission
standards are especially popular in combustion processes
where emissions are related to fuel or heat input, and to
excess air rates.
                         103

-------
Figure 7.17  Particulate Sampling Calculations

Plant No. 	   Calculated by	
Run No. 	   Checked by	
Location                     Date
Meter Volume	(Figure 7.10)
Leakage Volume  -	(Leakage rate x
                                       sampling time)
Net Sample Volume, Qm	ft3
Average Meter Temp. T	°F + 460=	°R.
                              m
Qms = 17'7 x Qm x Pb = 17'7
Standard Sample Volume,
                  P_l
                  Tm
Equivalent Moisture Volume, Q = _ SCF (Figure 7.14)
Total Sample Volume Q  = Q   + Q  =       = _ SCF
                     t    ins    V           '  '""
Particulate Sample Weight, W  = _ Grains  (Figure 7.14)
Particulate Concentration, C = _ =      grains/SCF
                                 Qt
Particulate Concentration,
dry basis C,=C x 100     =       grains/SCF-dry
                 100 - W
Emission Rate,
Ibs/hr, E = C
% Isokinetic, I =           s             x 100
Ibs/hr, E = C x Q   x 0.00857 =        x 0.00857 = Ibs/hr
                 s s
                  _  _
                  Q.      x  T    x  29.92
                  t x An    530"      Pb

     t  = sample time
                                    2
     A  = area of sample nozzle, ft.
     V  = Stack gas velocity-Figure 7.5
      5
                        104

-------
     Grains/SCF at a Specified Excess Air - Conversion to
this basis requires measurement of the excess air rate.
This can be determined by measuring the CO ,  0_, and CO
content of the exit gases.  Excess air is then computed
from the equation:
                    -  1/2% CO
        % X =
           s  0.264% N2 - (% 02 - 1/2% CO)
        Where % N2 = 100-(% C02 + % 02 + % CO)

     Correction to 50% excess air for example at standard
conditions is obtained by multiplying the grain loading
computed at STP by the ratio:  100 + measured X
                                     150
                                        180
     For 80% excess air  C5Q%    = C x  ,5Q
                               S
     Grains/SCF at Specified Oxygen Content - Converting
a grain loading to a specified 0~ content is accomplished
by:
            C   20.9-specified % 0?
              x 20.9-measured % 02
     Thus, if the basis is 6% O-, and 10% 0~ was actually
measured the corrected grain loading is: C,-a 0 _ „    14.9
                                          b% 02- c x  1Q>9
     Grains/SCF at Specific Carbon Dioxide Content -
Converting a grain loading to a certain C02 content is
accomplished by:
                      Specified % C02
                  C x
                      Measured % CO-
     If the specified CO  content is 12%, and the measured
C0? content in the exit gas stream was 4%, then the
corrected concentration would be: C,„„ _~  = C x 12
                        105

-------
     Pounds of Emission per 1000 pounds of Dry Flue Gas -
For gas streams similar to air in composition, i.e., with
a molecular weight between about 28 to 30, concentration
can be approximately converted to this basis at standard
conditions as follows:
     C,x 1.90 at standard conditions of 70°F and 29.92 "Hg,
If correction to an excess air value, or % CO- is also
required, these corrections are applied in the same manner
as previously explained.  For other gas compositions or
non-standard temperature or pressure conditions, the gas
volume should be converted to a weight basis by multiply-
ing by the appropriate density.  The emission on a  Ibs/hr
basis is then divided by this value.
     Pounds per Million Btu - This emission expression
is commonly used for.combustion processes and is obtained
by dividing the emission in Ibs/hr by the heat input,
expressed in millions of Btu entering a unit in the same
hour.  For bituminous coal fired units, emissions ex-
pressed on this basis.can be approximated by:
                          (C12% C02> X 1'9
                         106

-------
8.0  SIGNIFICANCE OF ERRORS IN SOURCE SAMPLING
     The procedure for determining pollutant emission
rates by stack sampling involves the measurement of a
number of parameters.  Errors of measurement associated
with each parameter combine to produce an error in the
calculated emission rate.  Measurement errors are of two
types; bias and random.  Bias usually occurs as a result
of poor technique in which measured value tends to
differ from the true value in one direction.  Typically
this operator error can be minimized by proper calibration
and adequate training in instrument operation.  Random
errors result from a variety of factors which causes the
measured value to be either higher or lower than the
true value.  They are caused by the inability to read
scales very precisely, as well as the quality and sensi-
tivity of the measurement device.  The usual assumption
is that random errors are normally distributed with a
known  (or unknown) mean and standard deviation.
     The emission rate of particulates from a stack can
be expressed as follows:
     E = K, C Q                          Equation 8.1
          JL    S S
where;
     E = emission rate - Ib/hr
     C - particulate concentration - grains/SCF
     Q   = volume of gas in stack - SCFM
      o 5
     K- = constant to yield proper units
but;
          W
     C = -r2—                           Equation 8.2
                         107

-------
where;


     W  = weight of particulate sample - grains



     Q.  = total sample volume - SCF


and


          Ko A  Cp P     Ap .T

     Q  = _±—|	?_    —_§_          Equation 8.3
     uss      T       I   p  M -
               s      •—  s  s —'


     K~ = Constant to yield



     A  = Area of Stack - ft.2
      S


     Cp = Pitot tube coefficient


     Ap = Velocity head of stack gas - inches H20


     T  = Absolute temperature of stack gas - °R



     P  = Absolute pressure of stack gas - inches Hg


     M  =  Molecular weight of stack gas
      S


Substituting 8.2 and 8.3 into 8.1 yields:


        K W  A  CD P  f~ AD T   ~l
           p  s  F  s    ^  s            Equation 8.4




This is equivalent to:



           !E

           Q
            t
 Ps  1

 Ms - J
        K WP As CP          s            Equation 8.5
     The maximum relative error can be determined by use


of the logarithmic differential of these equations.
                          +       .
                      Cp  +   Qt
             d Ps _ d Ts    " "a I        Equation 8.6
              P      T
               s      s
  d M ~|

	§
   M.
    s -I
     The weight of particulate (W )  is determined by the


use of an analytical balance with sensitivity about +


0.1 mg.  For an industrial process the total sample weight
                        108

-------
is typically about 100 milligrams, while for some combustion
processes the typical sample may be around 200 milligrams*
Thus the relative error is:

     d W
     n—2. = ±  0*1 mg = 0.001 or 0.1% (Industrial process)
     W         100 mg
      P

     d W         ,
     n—£ = +  "if mg* = 0.0005 or 0.05% (Power plant)
     w      —  200 mg
      P
     The area of the stack A  is determined by actual
                            s
measurement of length and width for a rectangular cross-
section and the diameter for a circular cross-section.
The areas of each type of duct are:
     A  = L W
      s
or   A  = TT  (D)2
      s   4
then
     A"~  = V1 + V*       '             (rectangular)
      S
and
     d As _  2 d D                        (circular)
      As  "    D
     A typical procedure for determining  the inside
measurement of a stack is to insert a rod into the stack,
mark the rod, and measure with a steel rule.  Such a
procedure should yield a measurement correct to about 0.25
inches.  Thus for a circular stack with diameter 36 inches
the relative error is:

          L^.  = 2 d D  , 2 (25) =
          A         D      ^6^      U.uu or x.j*
           s
* These values could of course vary widely and these
  values are used only as examples.
                         109

-------
     Naturally the relative error would decrease with
stacks having larger inside dimensions.
     The coefficient of a type S pitot tube, Cp, varies
from 0.83 to 0.87.  If the average is assumed to be 0.85,
the maximum error is +_ 0.02.  The relative error is;

     ^T^2- = +  7r4!  = + x 0.024 or 2.4%
      v»p    ~  (J • o_>    ~~
The total sample volume Q  is determined by;
           17.7 Q  P
     Qt =  	T        + Qv              Equation 8.7
               m
where
     Q  = net sample metered volume - ft.
     T  = average absolute meter temperature -  °R
     P,  = barometric pressure - inches Hg
     Q  = equivalent moisture volume - SCF

       17-7 Pb d °m   ".7 p., Pb d Tm    17.7 Qm d Pb    d Qv
                           ^                 Tm
a Q.                        m
Qt
                           Qt
                                         Equation  8.8
     The volume of gas metered  (Q ) is typically between
40 and 50 cubic feet and the meter can be read to the
nearest 0.01 ft.   (d Q ).  Likewise the barometric pressure
                      m
(P.) is generally near 29.9 inches Hg and can be read  to
the nearest 0.01 inch Hg  (d P ).  The equivalent moisture
                             b
volume Q  is determined by;
     Q  = 0.0473 Qn
                         110

-------
where
then
     Q, = moisture collected - ml
     d Qv = 0.0473 d
     The amount of moisture collected is quite often near
100 ml and the precision of measurement is about + 2 ml.
Thus ,
     d Q  = 0.0473 x 2
        v
          = 0.0946 or 9.46%
     The absolute temperature of the meter (T )  is
determined by :
     T  = T + 460
      m
where
     T = meter temperature

     dTm     d
      Tm       Tm
     This measurement of temperature is usually made with
a bi-metallic thermometer with a precision of + 2°F.  The
range of temperature readings is from 80-120°F.  Assume
an average temperature of about 100°F or 560°R:
     d T      ,
     -^  = -^  = 0.0036 or 0.36%
       m
     Substituting these quantities in 8.8 and using the
algebraic signs of each error term to produce the maximum
error yields :
     d Q
           = 0.0006 or 0.06%
                         111

-------
     Differential pressure (Ap) is usually measured with
an inclined manometer.  The sensitivity is generally
assumed to be about + .01.  For Ap readings near 0.05
the maximum error is:
     d Ap     0.01    ,. on    __.
     -Ap   =  0705  = °'20 or 20%
     The absolute pressure of the stack gas (P )
                                              s
determined by Equation 7.3; namely:
          +
     P  = ~ Ps
     ^c   _ 2.  + p
      S   13.6    Fb
where
     p  = stack gage pressure - inches H.O

               d ps
     dPs  _  T3T6£+  dPb
      P            P
       s            s
     Stack gas pressure (p )  is measured with a manometer
                          S
which can be read to the nearest 0.1 inches of water  (d P) .
Typically the stack gas pressure is around +_ 2 inches of
water then;
     d P
     p — -  = 0.0004 or 0.04%
      s
     Stack gas temperature (T ) measurements are usually
                             S
made by mercury-glass thermometers , thermocouples , liquid-
filled bulb- thermometers , or bi-metallic thermometers.
Typical properly calibrated thermometers are accurate to
within + 5°F from 32°F to 500°F, + 10°F from 500 to 1000°F,
and + 20°F from 1000°F to 2000°F.  The maximum relative
error would occur at about 1000 °F.
d T
      = 4.      on
                °
 Ts     -  1000   460  = °-014 or
                         112

-------
     The equation for dry molecular  weight  in  terms  of
Orsat readings for a typical combustion process  is:


Md = Iffo  fMco?  (Rco?  - V  + %  (Ro,  - Rco > +
          122           222
     M^  (R-rt - Rn )  + M,,   (100-1^

                                         Equation  8.9
-y]
where:

     M     = 44-molecular weight C02
        c*

     M     = 32-molecular weight 0«
      U2
     M     = 28-molecular weight CO

     M,^    = 28-molecular weight N,,


     RO    = Initial reading of ORSAT

     RCO '  RO '  RCO  are ORSAT readings  for each gas


     Substituting the molecular weights  into equation  8 .9
and differentiating yields:
        -0.44 d R  + 0.12 d R     +  0.04 d R_
  d«.              U           wL/n            U A
  M,                           22
	d  _  	
 M                     M,
  d                     d
                                         Equation 8.10

     The error in reading the gas burette is generally
+_ 0.2 percent by volume and a typical dry molecular weight

is about 29.  Thus the maximum error is:

           d M,
           --^  = 0.0042 or + 0.42%
                        112

-------
     The maximum relative error in the emission rate
(equation 8.6) can be found by summation of all of the
above errors.*

^p = (0.1)  + (1.3) + (2.4) + (0.06) + i £(20) +  (0.04) +

     (1.4) +  (0.42)"]  = 14.8 percent

     Again,  it should be emphasized that 14.2% is the
maximum relative error if all of the individual errors
are additive and not random.
     A more realistic way of expressing error is to
consider the error in terms of standard deviations.  In
this case, the error is expressed as 3 deviation   (3a)
                     14
units about the mean.    The probable error can be
calculated from;
                     1/2
3o =fzh   A2 (30 )2~|                    Equation  8.11
thus
3a
    =  f(0.1)2 + (1.3)2 +  (2.4)2 +  (0.06)2 + 1/4  |~(20)
                2        2         2l ) 1/2
        + (0.04r + (1.4P +  (0.42PJ f
     =  0.104 or 10.4 percent
     On the basis of this error analysis, the determination
of emission rates by isokinetic stack sampling can be
expected to be within'10.4 percent of the true mean 99.6
percent of the time.  It is apparent that most of the
sources of error contribute only in a very small way to
the total error in the calculated emission rate.  The most
significant error results from the measurement of differen-
tial pressure  (Ap) with the pitot tube.
   The error associated with the dry molecular weight, M,
   is used as the error for the actual stack gas, M  .
                                                   s
                       114

-------
              APPENDIX A






NOMOGRAPHS FOR USE WITH SAMPLING  TRAIN
                   115

-------
AH,


3.0


2.5

2.0



1.5




1.0


REF 1 *H2°
0 	 Z3


T
m
150 	
A —
100 — ^
—
50 — —
—
_. —>
n ^~~^~
__
—
-50 	
- REF 2 Z
C —



B


^ — 2.0 10 — —
E — 1-5 -
E_1.0 20—-
| 	 0.8 ~
| 	 0.6

E 	 O.b 30 _



40 	 	


50 	



VPm
	 1.2
	 1.1

— — i.o


^ 	 0.9
	
	
1 " '"
	 0.8


DRAW LINE FROM AH« TO T TO OBTAIN POINT A ON REF.
(? m
DRAW 1 TNP FOnM DOTMT A TO « U.fl ANH QPAH R nw Off 9
                                   DRAW LINE FROM POINT B TO Ps/Pm§ AND OBTAIN ANSWER
                                        FOR C.
             Figure Al. Correction Nomograph For Use  With Figure  A2.

-------


ORIFICE READING
AH
10— s

97|
7 ~-
* S
—
6- •-

^-_
5 — =
-Hi
=
E


^
3~
~

2-—
	
—
—



1— g
0.9— S
0.8-3
0.7--J

O.G— I

0.5—|

0.4—=

0.3— E


0.2 —




0.1 —



Ref.
















— Ref.

























1 A
	 c. . v
r
\,
. j
CORRECTION
FACTOR
1 0

— 0.0

• 	 0.13

— 0.7

— O.fi
STACK
— 0.5 TEMPERATURE

2500

2000

1500


1000

800
600
500
400
300
200

100

0
—^
~ T$
—
—
^—~
, —
—
—
—
—
—
r
u 	
1—
1-
^~
s —
~
= —




Sliding
Scale




*- cut along lines •*•


0.001 —
K FACTOR p(TOT R£AO/NG





































AH = in. H20
C = dimensionles
Ts = °F
K = dimensionles

D = in.
aP = in. H20



AP

0.002-

0.003-
0.004 —

0.005-

O.OOP-
PROBE -.
TIPDIA/METER °-0082
D 0.01-


































— 1.0

— — 0.9

§— 0.8 °-02-

^-0.7 0.03-
I 0.04-
	 0.6
= 0.05-
E~ o'.oe-
=—0.5
= 0.08-
=- 0 ^
E— 0.4

~ 0.2-

~°-3 0.3-
- 0.4-

~ 0.5-
I 0.6-
— 0.2 0.8-
I T-0"
-
2-
3 .

4-
— 0.1 5-
6-
8
10"


Figure A2.  Operating Nomograph.
         117

-------
                   APPENDIX B
           Cleaning of Train Components
     Small Metal Parts - Small stainless steel parts
including quick connects, nozzles, check valves, unions,
and socket joints should be cleaned with water and a
detergent by hand, or with a sonic cleaning device and
the recommended cleaner.  The parts are then rinsed
with distilled deionized water and then with acetone
to remove organics and promote drying.  Quick'connects
and check valves should be lubricated very lightly with
silicone grease and the openings covered.

     Probe Sheath and Pitot Tube - The probe is first
stripped of the stainless steel union and quick connects.
These parts are cleaned together with the small metal
parts.  The rubber o-ring is cleaned using first water
and then acetone.  The Pitot tube and probe sheath are
scrubbed with acetone and water and the Pitot tube is
blown out with compressed air.  After cleaning, the
unions and quick connects are reassembled.  The glass
probe is inserted in the metal sheath and the openings
covered until ready for use.

     Glass Probe - Grease is wiped from the ground glass
ball joint and the probe brushed and rinsed first with
distilled, deionized water  and then with acetone.  A
visual inspection is made to determine if the probe is
thoroughly clean inside.  The dried glass probe is placed
in the cleaned stainless steel probe sheath and the ends
covered to avoid contamination.

     Glass Parts - All ground glass joints are wiped to
                        118

-------
remove any remaining grease.  All pieces are then soaked
in a cleaning solution of dichromate and acid for twenty-
four hours.  The parts are then washed in soap and water,
rinsed with distilled, deionized water then acetone.  A
very thin coat of acetone insoluble silicone stopcock
grease is then applied to all of the inside (female)
ground glass joints.  The impingers are then reassembled.
The glass, field sample containers and related glass
clean-up equipment should be cleaned using this procedure.
All openings on the glass parts should be covered to avoid
contamination.

     Filter Frit - The extra course glass frit from the
filter holder is cleaned by placing it in boiling hydro-
chloric acid (under a hood) for two hours and rinsing
in distilled, deionized water followed by an acetone
rinse.  If the frit does not appear clean, it should be
boiled for two hours in H?SO. containing a few drops of
sodium or potassium nitrite and rinsed in distilled,
deionized water and acetone, and left to dry.

     Miscellaneous - Manometers should be cleaned using
either soap, naphtha, or gasoline.  No other solution
should be used to clean the manometer unless recommended
by the manufacturer.  The manometers are then refilled
with the appropriate liquids.
                        119

-------
                     APPENDIX C
                                         15
            ORIFICE CALIBRATION PROCEDURE
     The meter box containing the vacuum pump and dry gas
meter is connected to a large capacity wet test meter
(1 cubic foot per revolution) by connecting the meter box
inlet to the outlet of the wet test meter.  The orifice
manometer is carefully zeroed.  The vacuum pump is then
turned on, the orifice AH is set at 0.5" of water and the
system is run for 15 minutes to equilibrate the temperatures.
The following readings are taken during the meter/orifice
calibration.  (1) Cubic feet of air registered by the dry
gas meter (CF,), (2) temperature of the wet test meter in
°F (T ),  (3) inlet temperature of the dry gas meter in °F
     W
(IT,), (4) outlet temperature of the dry gas meter in °F
(OT,), (5) time in minutes (t) required for 5 cubic feet
of air to flow through the train, and (6) barometric pressure
in inches of mercury (P.).  The same procedure is used with
the manometer orifice setting at a AH of 1 inch of water,
and the same data are recorded.  With the manometer orifice
set at AH readings of 2, 4, 6, and 8 inches of water,
respectively, 10 cubic feet of air are allowed to flow
through the wet test meter at each of these settings, and
the same data are recorded.  From those data, Y and AHQ are
                                                      @
determined for each calibration point.  Y is the ratio of
accuracy of the wet test to the dry gas meter.  AH0  (inches
of H20)  is the orifice differential that gives 0.75 cfm of
air at standard conditions of 70°F and 29.92 inches of
mercury.  Figure C-l illustrates a convenient form for
recording these data and also gives the formulas used to
calculate Y and AHQ.
                  e
                       120

-------
     If the calculated value for Y is not between 0.99 and
1.01, the dry gas meter will require adjustment as per the
manufacturer's instructions.  If the flow through the
orifice at a setting of 1.84 + 0.25 inches of H-O is not
0.75 cfm, the orifice diameter should be increased or de-
creased as the case may be.

     Once determined, AH~ is constant for a given meter
and orifice assembly, and should be recorded on the meter
box.
                       121

-------
Figure C-l  Orifice Calibration Form.
Date      Box No.               Meter No.
  AH,     CF      CF,      T     IT,   OT,   T,     Time,
Tn  H 0     W       d       W,     d,    d,   d,         '
in. n«u                    op    01?    „„    Op       T:
                                                    (Min.)
0.5
1.0
2.0
4.0
6.0
8.0
5
5
10
10
10
10
Calculation Y and  AH» at manometer orifice setting  of  2.0

Y =  CF   .  P,   .  (T, + 460)
       w _ b _ a _  = _ . _
     CF,  .  (Pb +  AH ) (T  + 460)
       d         1176   W

AHfl = 0.0317 AH        (T  -I- 460) t
  @   - =  w _ = _
      P,    (OT, + 460)     CF
       b     d              w

Y - Ratio  of accuracy of wet test meter  to dry  gas meter.

AH0 = Orifice pressure differential that gives  0.75  cfm of  air
  L   at  70 °F and  29.92 inches of mercury, in.  H20.

P,  = Barometric pressure, in. Hg.

AH = Manometer orifice setting, in. H-O.

CF  = Cubic feet of air measured by the  wet test meter,  cubic
  w   feet.

CF, = Cubic feet of air measured by the  dry gas meter,  cubic
  a   feet.

T  = Temperature at the wet test meter,  °F.

IT, = Inlet temperature at the dry gas meter, °F.
OT, = Outlet temperature at the dry  gas meter,  °F.

Td = Average of the inlet  (ITdJ and  outlet  (OTdJ  temperatures at
     the dry gas meter,  °F.

 t = Time of test,  minutes.

 Tolerances

 Y = 0.99 - 1.00 - 1.01

 AH@ = 1.6 - 1.84 - 2.1

                      122

-------
                  LIST OF SYMBOLS





                                 2
A    Area of Sampling Nozzle, ft.


                              2
A    Inside Area of Stack, ft.
 o


C    Particulate Concentrations, Grains/SCF



Cp   Pitot Tube Correction Factor, No Units



E    Emission Rate, Ibs/hr.



M    Stack Gas Molecular Weight



M,   Stack Gas Molecular Weight Dry Basis



N    Number of Sampling Points



P,   Barometric Pressure, Inches of Hg



p    Stack Gage Pressure, Inches of Water
 o


P    Stack Absolute Pressure, Inches of Hg



P    Average Pressure at Dry Gas Meter, Inches of Mercury

 m   (as used in this text P  = P, )
                            m    b


Q    Stack Gas Volume, cfm
 S


Q    Stack Gas Volume SCFM
 S S


Q    Meter Volume, CF or rate, CFM



Q    Volume of Condensed Moisture, SCF



Q    Meter Volume, SCF
wms


Q    Total Sample Volume, SCF



T    Meter Temperature, °F.
 m


T    Stack Gas Temperature,  °F.
 O


V    Stack Gas Velocity, ft/min.
 O


W    Moisture Content of Stack Gas, %



Ap   Velocity Head, Inches of Water



AH   Pressure Drop Across Orifice, Inches of Water



W.    Particulate. Weight, Grains or Grams




                      123

-------
                        REFERENCES
 1.    Walsh,  G.  W,   and D.  Von  Lehmden.   Resources  for  Air
      Quality Control  Regions.  (Presented at  EPA Workshop
      on Regional  Implementation Plans.   Raleigh, North
      Carolina,  November 1969).

 2.    Ermenc, E. D.  Air Pollution  Control Governmental
      Careers.   Chemical Engineering.  ^(1):  122-125,
      January 11,  1971.

 3.    Standard Test  Code for Centrifugal  and  Axial  Fans.
      Published by National Association of Fan Manufacturers.
      Bulletin No. 10.   1938.

 4.    Devorkin,  H.,  etal.   Source Testing Manual, Los Angeles,
      Los Angeles  County Air Pollution Control District,
      November,  1963.   179  p.

 5.    Sampling Stacks  for Particulate Matter.   ASTM Method
      D 2928-71.  Philadelphia,  Pa. January  1971.

 6.    Sample  and Velocity Traverses for Stationary  Sources
      — Method 6.   (Tentative). Office of Air Programs, EPA
      Raleigh,  North Carolina.   May 1971.  4  p.

 7.    Smith,  W.  S.,  etal.   Stack Gas Sampling Improved  and
      Simplified with  New Equipment. Paper No. 67-119.
      (Presented at  the Air Pollution Control Association
      Annual  Meeting,  Cleveland, Ohio, June 11-16,  1967).

 8.    Hemeon, W.C.   Magnitude of Errors in Stack Sampling.
      Air Repair.  £:  159-164,  November 1954.

 9.    Smith,  W.  S.,  R.  T. Shigehara, W. F. Todd. EPA
      A Method to  Interpret Sampling Data.  Paper No.  70-34.
      (Presented at  the Air Pollution Control Association
      Annual  Meeting,  St. Louis, June 14-18,  1970).

10.    Achinger,  W. C.  and R. T.  Shigehara.  A Guide for
      Selecting Sampling Methods for Different Source  Con-
      ditions.   J. Air Pollution Control  Association.
      .18:605-609,  September 1968.

11.    Bulletin WP-50.   Western  Precipitation  Company,  Los
      Angeles,  California.   6th Edition.    27p.
                          124

-------
12.    Martin,  R.  M.   Construction Details  of Isokinetic
      Sampling Equipment As  Developed and  Used by  Office
      of Air Programs -  In press.  Office  of Air Programs,
      EPA.   Raleigh,  North Carolina.

13.    Shigehara,  R.  T.,  W. F.  Todd, and W.  S. Smith.   Significance
      of Errors in Stack Sampling Measurements. Paper No.
      70-35. (Presented  at the Air Pollution Control  Association
      Annual Meeting, St. Louis,  June 14-18, 1970).

14.    Wine,  R.  L.  Statistics  for Scientists and Engineers.
      Englewood Cliffs,  New  Jersey, Prentice-Hall, Inc.
      1964.

15.    Rom,  J.   Source Sampling Using APCO  Designed Equipment.
      In Press.  Office  of Air Programs, EPA.  Raleigh, North
      Carolina.  1971.
                          125

-------
DIllLIOGRAPHIC DATA   1- Report No.
SHEET;     	    •	
I. Title an
-------