EPA-600/2-76-159
June 1976
Environmental Protection Technology Series
                         IERL-RTP  DATA  QUALITY  MANUAL
                                          Industrial Environmental Research Laboratory
                                                 Office of Research and Development
                                                U.S. Environmental Protection Agency
                                         Research Triangle Park, North Carolina 2771T

-------
               RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection  Agency, have been grouped into five series. These  five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:

     1.    Environmental Health Effects Research
     2.    Environmental Protection Technology
     3.    Ecological Research
     4.,   Environmental Monitoring
     5.   ,Socioeconomic Environmental Studies

This report has been  assigned  to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new  or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
                    EPA REVIEW NOTICE

This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication.   Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency,  nor does mention of trade
names or commercial products  constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                        EPA-600/2-76-159

                                        June 1976
               IERL-RTP

            DATA  QUALITY

                MANUAL
                     by

     Franklin Smith and James Buchanan

         Research Triangle Institute
               P.O. Box 12194
Research Triangle Park, North Carolina 27709
       Contract No.  68-02-1398, Task 35
        Program Element No. EHB524
     EPA Project Officer: L. D. Johnson

 Industrial Environmental Research Laboratory
   Office of Energy, Minerals, and Industry
      Research Triangle Park, NC  27711
                Prepared for

U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Research and Development
            Washington, DC 20460

-------
                           TABLE OF CONTENTS

SECTION                                                           PAGE
   I                          INTRODUCTION                        1
        1.1  DEFINITIONS                                            1
        1.2  BACKGROUND                                             2

  II     PURPOSE  AND SCOPE  OF DATA QUALITY MANUAL              3
        2.1  PURPOSE                                                3
        2.2  SCOPE                                                 3

 III            QUALITY POLICIES AND OBJECTIVES                  4
        3.1  QUALITY POLICIES                                        4
        3.2  QUALITY OBJECTIVES                                      5

  IV            ORGANIZATION FOR DATA  QUALITY                   7
        4.1  ORGANIZATION STRUCTURE                                  7
        4.2  FUNCTIONAL RESPONSIBILITIES                              7

   V         IMPLEMENTATION PLAN AND SCHEDULE                  13
        5.1  IMPLEMENTATION  SCHEDULE                                13
        5.2  ESTIMATED COST  OF IMPLEMENTATION                        15

  VI     COMPONENTS OF A  COMPREHENSIVE QUALITY CONTROL
         PROGRAM                                                  16
        6.1  GENERAL REMARKS                                       16
        6.2  FACILITIES AND  EQUIPMENT                               16
        6.3  CONFIGURATION CONTROL                                  18
                                    iii

-------
                        TABLE  OF CONTENTS (CON,)
SECTION                                                             PAGE
        6.4  PERSONNEL TRAINING                                      18
        6.5  DOCUMENTATION CONTROL                                   19
        6.6  CONTROL CHARTS                                          20
        6.7  IN-PROCESS QUALITY CONTROL                              20
        6.8  PROCUREMENT AND INVENTORY PROCEDURES                     22
        6.9  PREVENTIVE MAINTENANCE                                  22
        6.10 RELIABILITY                                            22
        6.11 DATA VALIDATION                                         23
        6.12 FEEDBACK AND CORRECTIVE ACTION                           23
        6.13 CALIBRATION PROCEDURES                                  24
  VII    GUIDELINES FOR QUALITY ASSURANCE  PROGRAMS            25
        7.1  GENERAL STATEMENT                                       25
        7.2  THE REQUEST FOR PROPOSAL - QUALITY CONTROL ASPECTS        25
        7.3  EVALUATION OF QUALITY CONTROL  IN THE PROPOSAL            26
        7-4  EVALUATION OF QUALITY CONTROL  IN THE WORK PLAN            27
        7.5  THE ON-SITE QUALITATIVE SYSTEMS AUDIT                    27
        7.6  THE PERFORMANCE AUDIT                                   28
        7.7  DATA QUALITY ASSESSMENT                                 28
 VIII   DEVELOPMENT OF QUALITY CONTROL AND QUALITY ASSUR-
        ANCE  PLANS  FOR IERL PROJECT CATEGORIES               30
        8.0  GENERAL                                                 30

-------
                      TABLE OF CONTENTS  (CON,)
SECTION                                                         PAGE
        8.1  ENVIRONMENTAL ASSESSMENTS                             30
        8.2  INDUSTRY SYSTEM STUDIES/POLLUTANT SYSTEM STUDIES         33
        8.3  FIELD STUDIES                                        36
        8.4  RESEARCH AND BENCH-SCALE PROJECTS                      38
        8.5  DEVELOPMENT OR PILOT PROGRAM                          39
        8.6  DEMONSTRATION PROJECTS                                42
  IX                          REFERENCES
                                                                46
APPENDIX A  QUALITATIVE ON-SITE SYSTEMS REVIEW
             CHECKLIST                                         47
APPENDIX B  STANDARD  TECHNIQUES USED  IN QUANTITATIVE
             PERFORMANCE AUDITS                                71
APPENDIX C  DEFINITIONS AND STATISTICAL TECHNIQUES
             USEFUL  IN QUALITY ASSURANCE PROGRAMS             77
APPENDIX D  SOME STANDARD AMBIENT AIR AND SOURCE
             SAMPLING  TECHNIQUES                               85

-------
                               LIST OF  FIGURES
FIGURE NO.                                                            PAGE
   1    IERL-RTP ORGANIZATION  CHART                                       8
   2   DATA QUALITY PROGRAM ORGANIZATION                                 9
   3   FLOWCHART OF FUNCTIONS AND RELATIONSHIPS                         12
   4   DATA QUALITY PROGRAM IMPLEMENTATION SCHEDULE                     14
   5   APPLICABILITY OF QUALITY CONTROL PROGRAM  COMPONENTS TO
       PROJECT CATEGORIES                                              17
   6   STANDARD QUALITY CONTROL CHART                                  21

-------
SECTION  I                     INTRODUCTION

1.1  DEFINITIONS
     In order to facilitate the understanding of this manual, it is necessary
to define three terms:  quality,  quality control, and quality assurance.   It
is sometimes difficult to distinguish quality control from quality assurance
unless the proper definitions are kept in mind.   The following definitions are
based on those given by the American Society For Quality Control (ref.  1)  and
those given by the Environmental Protection Agency (ref. 2).
Quality
     The totality of features and characteristics of a product or service
that bear on its ability to satisfy a given purpose.  For measurement systems,
the product is measurement data and the characteristics of major importance
are accuracy, precision, and completeness.  For  monitoring systems, complete-
ness—or the amount of valid measurements obtained relative to the amount
expected to have been obtained—is usually a very important measure of  quality.
The relative importance of accuracy, precision,  and completeness depends upon
the particular project requirements.
Qual-Lty control
     The overall system of activities, the purpose of which is to provide
a quality of product or service that meets the needs of users; also,  the use
of such a system.  Maintaining a quality control program is the responsibility
of the organization/individual implementing the  project.
     The aim of quality control is to provide quality that is adequate, depend-
able, and economical.  The overall system involves integrating the quality
aspects of several related steps, including the  proper specification of what
is wanted; production to meet the full intent of the specification; inspection
to determine whether the resulting product or service is in accordance with
the specification; and review of usage to provide for revision of specification.

-------
Quality assurance
     A system of activities, the purpose of which is to provide assurance
that the overall quality control job is in fact being done effectively.  The
system involves a continuing evaluation of the adequacy and effectiveness of
the overall quality control program with a view to having corrective measures
initiated where necessary.  For a specific product or service, this involves
verifications, audits, and the evaluation of the quality factors that affect
                      »
the specification, production, inspection, and use of the product or service.
Maintaining an IERL-RTP quality assurance program is a function of the
Process Measurements Branch.

1.2  BACKGROUND
     The IEKL-RTP has long recognized the importance of quality control as an
integral part of its research and measurement activities.  Heretofore, quality
control has been practiced on a project-by-project basis, with the preparation
and implementation of a quality control plan being the responsibility of the
EPA project officer and the contractor.  However, due to recent increased
impetus in EPA energy and industrial processes programs and subsequent substan-
tial increase in the number, scope, and importance of environmental assessment
and technology development projects within IERL-RTP, the need for a formal,
comprehensive, and integrated laboratorywide data quality program has become
more critical.  The development and implementation of a formal IERL-RTP data
quality program was initiated in December 1974 by the preparation and distri-
bution to senior staff members of the "Planning Document for an IERL-RTP
Quality Assurance Program" (ref. 3).  This planning document groups all
IERL-RTP projects into five categories.  Six categories are now existent, with
the inclusion of environmental assessments.  Further progress in developing and
implementing an IERL-RTP data quality program has been realized through the
preparation and on-site trial implementation of quality assurance procedures
at IERL-RTP projects (refs. 4,5,6).  Ultimately, quality assurance guideline
documents will be generated for each project category.

-------
SECTION  II    PURPOSE AND SCOPE OF DATA  QUALITY MANUAL

2.1  PURPOSE
     The purposes of this manual are:
     1.   To direct the establishment of and provide guidance for the imple-
          mentation of an integrated data quality program  for Industrial
          Environmental Research Laboratory - Research Triangle Park (IERL-
          RTP) projects.
     2.   To serve as a source of directive and instructive material for
          IERL-RTP project officers and contractors in the establishment and
          maintenance of project-specific quality control  programs sufficient
          to insure that project data objectives are realized in the most
          economical manner (sections VI and VIII).
     3.   To serve as guidelines for the Process Measurements Branch within
          IERL-RTP in establishing and maintaining a quality assurance program
          to monitor, assess, and document the  efficiency  of the various
          project quality control programs (sections VII and VIII).

2.2  SCOPE
     This manual describes the administrative systems pertaining to  the
establishment and maintenance of an IERL-RTP data quality  program and presents
guidelines for developing or designing quality  control and quality assurance
plans specific to given projects.
     The administrative systems include:  quality policies that provide  both
guidance for the establishment and implementation of a data quality  program
and quality objectives to guide in the designing of quality control  and  quality
assurance plans (section III); organization, naming key quality personnel and
groups (section IV); and a plan and schedule for implementing the quality
program (section V).
     Guidelines for developing project-specific quality control plans are given
in sections VI and VIII and for quality assurance plans in sections  VII  and
VIII.

-------
SECTION  III        QUALITY POLICIES  AND OBJECTIVES

     This section contains the IERL-RTP policies  to  be  followed  in  the estab-
lishment and maintenance of a data quality program.  The objectives to be
realized through a well-planned and conscientiously  applied data quality program
are also given here.   A time schedule and  details  for implementation of these
policies are given in section V.   The organizational structure for  establishing
and maintaining the data quality program is given  in section IV.

3.1  QUALITY POLICIES
     IERL-RTP policies pertaining to the development, implementation, and
maintenance of a data quality program are  described  by  category  below.
3.1.1  Coverage of the Data Quality Program
     The IERL-RTP data quality program will have  the following characteristics.
It will be complete in nature, encompassing both  in-house and contract experi-
ments, tasks, and projects that either generate or use  experimental data.  It
will be integrated, in that all experiments, tasks,  and projects must have a
quality control plan delineating the practices and procedures to be implemented
at each level (e.g.,  operator, bench chemist, project leader, etc.) and each
phase of the project.  This plan will be evaluated and  approved  by  the Process
Measurements Branch.   All experiments, tasks, and  projects will  also have a
quality assurance plan for monitoring the  effectiveness of the quality control
program to be implemented by the Process Measurements Branch.  Finally, the
data quality program will be applied on a  project-by-project basis  according
to project objectives and requirements.
3.1.2  Levels of Quality Application
     Quality practices and procedures will be implemented at two levels.
     1.    Quality control   The design and implementation of quality control
          practices and procedures required to assure that data  quality is
          sufficient  to meet project requirements  are the responsibility  of
          the individual or organization conducting  the project.  For  example,
          on projects conducted under contract,  the  quality  control plan  will
          be prepared by the contractor, and reviewed and approved by  the

-------
          EPA project officer with assistance from the Process Measurements
          Branch if desired.  Inhouse projects will have quality control plans
          prepared by the responsible EPA staff member,  with assistance from
          the Process Measurements Branch if desired.
     2.   Quality Assurance   Quality assurance procedures for independently
          monitoring and assessing the efficiency and  adequacy of individual
          quality control programs will be established and administered by  the
          IERL-RTP director through the Process Measurements Branch.   The quality
          assurance procedures shall be applied uniformly throughout  the
          duration of the project.  However, at any specific time during the
          project life, either at the request of the subject project  officer
          or if deemed necessary by the Process Measurements Branch,  the
          Process Measurements Branch may, using accepted quality assurance tech-
          niques, assess the IERL-RTP project's ongoing quality control program.

3.2  QUALITY OBJECTIVES
     The primary objective of the IERL-RTP data quality program is to assure,
assess, and document that the quality (i.e., precision,  accuracy, and complete-
ness) of measurements made by and/or experimental data used in IERL-RTP
activities and publications is commensurate with the end use of the data.
Management, administrative, statistical, investigative,  preventive, and
corrective techniques will be employed to maximize the end effectiveness of the
data.
     Specific data quality objectives are:
     1.   To establish acceptable limits on data quality as a function of
          project objectives, available resources, and measurement method
          capabilities;
     2.   To establish recommended procedures and require their use to insure
          the comparability of like data between projects;
     3.   To establish guidelines for the selection and use of additional
          measurement methods necessary to assure the collection of data of
          acceptable quality (i.e., of acceptable precision, accuracy and
          completeness) on a project-by-project basis;

-------
To develop and implement quality control programs on each specific
IERL-RTP project;
To develop and implement the quality assurance procedures necessary
to independently monitor the efficiency of the individual project
quality control programs;
To identify areas requiring new or improved measurement methods in
order to achieve the level of quality required to satisfy project
objectives.

-------
SECTION IV           ORGANIZATION  FOR DATA  QUALITY

     The organizational structure, functional responsibilities, levels of
authority, and lines of internal and external communication for management,
direction, and execution of the quality program are given here.  Individuals
and groups or organizations discussed include the quality assurance manager,
the quality assurance coordinator, the project officer, the Process Measure-
ments Branch, and contractors.

4.1  ORGANIZATION STRUCTURE
     IERL-RTP organizational structure is shown in figure 1.  Figure 2 presents
the laboratory's data quality program organization.
     The chief of the Process Measurements Branch (PMB) is designated as  the
quality assurance manager and on data quality matters reports  to the IERL-RTP
director.  The quality assurance coordinator directs the activities of the
quality assurance group, which is composed of the Process Measurements Branch
staff members and contract support.  The coordinator is directly responsible
to the quality assurance manager.
     The quality control organizational hierarchy shows the control line  moving
from the appropriate division director, to branch chief, project officer,  then
contractor.

4.2  FUNCTIONAL RESPONSIBILITIES
     The functional responsibility assignments for individuals and organizational
components are given in this section.
4.2.1  Quality Assurance Manager
     The chief of the Process Measurements Branch is the quality assurance
manager.  This person is responsible for the design development, implementation,
and maintenance of the IERL-RTP data quality program.  The chief directs  the
efforts of the quality assurance coordinator and thus the data quality
activities of the Process Measurements Branch.

-------
oo
UTILITIES AND INDUSTRIAL POWER
DIVISION
Mr. Everett Plyler
—

Emissions/Effluent Technology Branch
Mr. Mike Maxwell (Act.)

Process Technology Branch
Mr. Richard Stern

•Paniculate Technology Branch
a • .. 1- _ A t_ L -.4. A

                                                     ENVIRONMENTAL PROTECTION AGENCY
                                             INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
                                                  RESEARCH TRIANGLE PARK, NORTH CAROLINA
                                                                    IERL-HTP
                                                                     Director
                                                                  Or. John Burcbard
                                                                       -L
                                                                  Deputy Director
                                                                 Dr. Norfaert Jaworski
                                                                                             OFFICE OF PROGRAM OPERATIONS
                                                                                                       Dr. John 0. Smith
                                                                                                   Special Studies Staff
                                                                                                    Or. VV. Gens Tucker
                                                                                                 Planning, Management and
                                                                                                   Administration Staff
                                                                                                     Mr. William Rice
                                                         ENERGY ASSESSMENT AND CONTROL
                                                                    DIVISION
                                                               Mr. Robert Hangebrauck
                                                              Combustion Research Branch
                                                                  Dr. Joshua Bowen
                                                                 Fuel Process Branch
                                                                  Mr. T. Kelly Janes
                                                                Advanced Process Branch
                                                                 Mr. P.P. Turner, Jr.
INDUSTRIAL PROCESSES
DIVISION
Mr. Alfred B. Craig

-

Chemical Processes Branch
Dr. Dale Denny

Metallurgical
Processes Branch
Mr. Norman Plaks

Process Measurements
Branch
Mr. James Dorsey
                                              Figure  1.  IERL-RTP organization chart.
                                                                                                                            November 1975

-------
                                      IERL-RTP
               Quality Assurance Line
                           Quality Control  Line
                                                    DIVISION
                                                    DIRECTOR
Quality
Assurance
Manager
CHIEF, PROCESS
 MEASUREMENTS
    BRANCH
  BRANCH
  CHIEF
                         PMB
                       QUALITY
                      ASSURANCE
                     COORDINATOR
                                  PROJECT
                                  OFFICER
Quality
Assurance
Group
STAFF, PROCESS
 MEASUREMENTS
    BRANCH
CONTRACTOR
            Figure 2.  Data quality program organization.

-------
4.2.2  Quality Assurance Coordinator
     The quality assurance coordinator is  responsible  for  overseeing  all  IERL-
RTP data quality efforts.   This person coordinates  the activities  of  the
Process Measurements Branch in the data quality  program.   He/she works with  the
appropriate project officers in designing  and  implementing project-specific
quality control and quality assurance plans.
4.2.3  Process Measurements Branch
     The Process Measurements Branch is responsible for coordinating
all IERL-RTP quality activities.  It initiates measures to insure  the fulfill-
ment of the overall quality objectives of  the  laboratory,  and for  carrying out
the data quality policies in the most efficient  and economical manner commen-
surate with insuring continuing acceptable levels of completeness,  accuracy,
and precision of experimental data produced.   A  summary of the data quality
responsibilities and authority of the Process  Measurements Branch  is  as follows:
     1.   It develops quality control guidelines and quality  assurance programs,
          including statistical procedures and techniques, which will help the
          laboratory to meet desired quality standards at  minimum  cost; and
          coordinates the implementation of such programs  with the appropriate
          project officer.
     2.   It reviews all measurement programs  and insures  that appropriate
          methods have been selected for data  acquisitions.
     3.   It reviews quality control activities  of  the various projects and
          makes appropriate suggestions to the project officer regarding
          corrections and improvement.
     4.   It seeks out and evalutes new ideas  and current  developments  in the
          field of quality assurance/control and recommends  implementation
          wherever advisable.
     5.   It advises project officers in preparing  and/or  reviewing requests
          for proposals,  work plans, project implementation,  and
          reports of work with respect to  quality aspects  of  technology,
          methods,  and equipment.
     6.    It advises on packaging materials and  procedures for sample handling
          and on requirements for maintaining  sample integrity.
     7.    It advises project officers concerning schedules for system checks,
          calibrations,  and other checking procedures.
                                      10

-------
      8.    It evaluates data quality statistically and maintains related quality
           assurance records and other pertinent information.
      9.    It coordinates the IERL-RTP program with the Environmental Monitoring
           and Support Laboratory/Quality Assurance Branch (EMSL/QAB)  program.
     10.    It prepares and issues periodic quality assurance reports  on specific
           projects to the project officer and the IERL-RTP director.
     11.    It prepares and issues periodic quality assurance summaries to  the
           IERL-RTP director and to the EMSL-RTP director.
      The quality assurance function is in the Process Measurements Branch and  is
 under the direction of the quality assurance  manager.   The branch will  have in-house
 and contract technical support  as required to establish and maintain  the  data
 quality  program.
 4.2.4 Project Officer
      The IERL-RTP project officer is ultimately responsible for  the success or
 failure  of the project.   This person thus has the responsibility for  determining
 the optimum level of quality control for the  project  and  for seeing that  the
 program  is implemented and maintained.   The project officer is responsible for
 quality  control practices,  beginning with the preparation of the request  for
 proposal (RFP)  and extending through the final  report,  as discussed in  detail
 in section VIII.
 4.2.5 IERL-RTP Contractors
      For IERL-RTP projects  conducted under contract,  the  contractor is  respon-
 sible for designing,  developing,  implementing,  and maintaining a quality  control
 program  to insure that the  experimental  data  generated will be of suitable
 quality  to satisfy the project  requirements,  and also for documenting that
 quality.   The quality control plan must  be approved by  the IERL-RTP project
 officer.
 4.2.6  Functional Relationships
      Once a project has been defined or  a  contract negotiated the time  sequence
 of events  and interrelations  between the Process Measurements Branch, the project
 officer,   and contractor are  illustrated  in figure 3.  The first  column  in the
 figure shows the  data quality functions  of  the  Process Measurements Branch.  The
 second column gives the functions  of  the project officer  (in-house project) or  the
project officer/contractor  (contracts).  The  lines across columns show  the points
of interaction between different quality components of  the data  quality program.
                                      11

-------
   QUALITY ASSURANCE GROUP

(Process  Measurements Branch)
    QUALITY  CONTROL  GROUP

(Project Officer/Contractor)
         Develop  standard
         measurement
         methods
         Establish
         data quality
         specifications
         Revi ew
         measurements
         program
         Select
         measurement
         methods
         Define
         quality control
         guidelines
         Develop
         quality control
         program
         Develop
         quality
         assurance
         program
                                                       Implement
                                                       quality control
         Perform
        quality
        assurance
        audit
                                                            I
         Issue
         quality control
         reports
             I
         Issue
        quality
        assurance
        report
            Figure 3.   Flowchart of functional relationships.
                                     12

-------
SECTION  V          IMPLEMENTATION PLAN AND SCHEDULE

      The plan for developing and Implementing a comprehensive data quality
 program for IERL-RTP is predicated on the philosophy that a quality program
 should be implemented gradually and in phases such that the results from one
 phase can be used constructively in implementing the next phase.   Also,  the
 differences involved in implementing data quality programs  to ongoing projects
 as compared to new projects must be considered.
      The implementation plan requires that quality control  and quality assur-
 ance procedures be applied to new projects, both in-house and contract,  from
 the programs inception.  That is, the Process Measurements  Branch is presently
 available to assist the project officer (contract projects) in specifying
 quality control requirements for inclusion in the RFP and evaluating quality
 aspects of proposals and work plans.  For new in-house projects,  the Process
 Measurements Branch is available to help project officers prepare quality
 control plans for their projects.
      The general plan and schedule for developing and implementing a laboratory-
 wide data quality program, starting with the preparation of a planning
 document, are given below.

 5.1  IMPLEMENTATION SCHEDULE
      A schedule for developing and implementing a laboratorywide  data quality
 program is given in figure 4.  In December 1974, a planning document for an
 IERL-RTP quality assurance program was prepared under the direction of the
 Process Measurements Branch and distributed to the laboratory's senior staff
 members.  The date of the distribution of the plan-ning document  is indicated
 as the first milestone on the implementation schedule.
      Milestones 2, 3 and 4 in figure 4 show, respectively,  the preparation of
 a general quality assurance plan for demonstration-size projects  (EPA-600/2-
 76-081); making the plan specific for the EPA Wet-Limestone demonstration
 project at the Shawnee Steam-Electric Plant in Paducah, Kentucky  (EPA-600/2-76-
 082); and on-site implementation of that plan for evaluation purposes (EPA-
 600/2-76-083).  These three milestones occured in the last two quarters of
 1975.
                                      13

-------
TASKS
1. Planning document for
IERL-RTP quality
assurance
2. General QA plan for
demonstration size
project category
3. QA plan for EPA SHAWNEE
Scrubber Demonstration
Project
4. Trial implementation
of QA plan at SHAWNEE
Scrubber Demonstration
Project
5. QA group available to
implement QA to existing
demonstration projects
as requested
6. All new demonstration
projects require QC
and QA programs
7. General QC/QA plans
for the environmental
assessment project
category
8. General QC/QA plans for
development or pilot
programs
9. Implement and maintain
QA programs for
development or pilot
programs not covered
by 6 alone
10. General QC/QA plans
for field studies
11. General QC/QA plans
for research and
bench-scale projects
12. General QC/QA plans
for industry system
studies/pollutant
system studies
1974
4
A


















1975
1234


A

A

/p

L










1976
1234








5^
A
ZeX
A
A
*.. nun





1977
1234












k
A
A
A
A
/.A
A
Figure 4.  Data quality program implementation schedule.
                          14

-------
     Since completing the general quality assurance plan and testing it under
field conditions, the primary effort has been directed toward implementing
quality assurance procedures to other existing demonstration-size projects.
     Effective September 1, 1976, all new demonstration projects and those
existing demonstration projects having 1 year or longer to run must have quality
control and quality assurance plans approved by the Process Measurements Branch
(milestone 6).  These plans include both quality control procedures to be carried
out by the individual or organization conducting the project and quality
assurance procedures to be administered through the Process Measurements Branch
for monitoring the efficiency of the quality control procedures.
     After September 1, 1976, general quality assurance plans will be prepared
for the remaining five project categories, shown as milestones 7, 8, 10, 11,  and
12 in figure  4.  Because of the current interest in environmental assessments,
the second general guideline document will be for this category.  As figure 4
illustrates,  a complete data quality program covering all IEBL-RTP projects
should be in  effect by early 1978.

5.2  ESTIMATED COST OF IMPLEMENTATION
     A definitive cost estimate for implementing and maintaining a data
quality program cannot be made at this point in time.  However,  with the degree
of importance placed on data quality by IEKL-RTP, a realistic estimate of
average costs per project are:  (1) quality control costs in the range of 8
to 10 percent of the total project budget, and (2) quality assurance costs
between 2 to  5 percent of the total project budget.  From these estimates,
then, the costs of the IERL-RTP data quality program as described in this manual
when fully implemented should be in the range of 10 to 15 percent of the total
laboratory project budget.
     The above cost estimates apply to new projects.  The net additional cost,
if any,  for implementing a data quality program to an ongoing project will
depend upon the project's current level of quality control and quality assurance
activities.
                                      15

-------
 SECTION VI         COMPONENTS  OF A  COMPREHENSIVE QUALITY
                     CONTROL PROGRAM

 6.1   GENERAL  REMARKS
      It is  highly desirable  that  quality control and quality assurance programs
 be built into an IERL-RTP project from its inception; i.e., that quality control
 requirements  be  stated  in the RFP, and that the contractor's provision for
 meeting those requirements be evaluated in the proposal and in the ensuing work
 plan.   In this way,  quality  control is not treated as an extra-cost or add-on
 problem, but  rather  as  an integral part of the total project.
      Similarly,  the  project  officer must be aware of the desirability of main-
 taining an  adequate  quality  assurance program in order to insure the validity
 of the data from the project.  Individual project categories, from environmental
 assessments through  demonstration projects, have varying quality control and
 quality assurance requirements.   Section VIII will treat some of these require-
 ments  individually,  especially with regard to the RFP, proposal evaluation,
 and work plan review.
      For efficiency  of  presentation, the major components of a quality control
 program are treated  in  a general way in the following subsections.  Figure 5
 is a matrix that attempts to show the applicability of each of these components
 to each project  category.  (A description of each category is given in section
 VIII.)   The matrix is obviously arbitrary, and should serve as a general guide
 only,  since individual  projects will each have unique data requirements.  The
 coding  of the various quality control components is by the appropriate sub-
 section number;  e.g., component 6.2 in the matrix refers to subsection 6.2,
 Facilities  and Equipment.

 6.2  FACILITIES  AND  EQUIPMENT
     A  good beginning point  in the assessment of an ongoing project is a
 general  survey of  the facilities and equipment available for day-to-day opera-
 tion.  Are  they  adequate for the job at hand?  Do standards exist for evalua-
 tion of  facilities,  equipment, and materials?
     The laboratories and data processing and other operational areas should
be neat and orderly, within common-sense limits imposed by the nature of  the
                                      16

-------
                                                       QUALITY CONTROL PROGRAM COMPONENTS
                                                                       ff
                                                              ft
PROJECT CATEGORIES
ENVIRONMENTAL*
ASSESSMENTS
INDUSTRY/POLLUTANT
SYSTEM STUDIES
FIELD STUDIES
RESEARCH AND BENCH-
SCALE PROJECTS
DEVELOPMENT OR
PILOT-PLANT PROGRAMS
DEMONSTRATION
PROJECTS
1**




X
X

X
X

8





X

X
X

7_r / / W/ 7 / */"/ ,








X








X
X








X
X









X









X








X
X





1



X

x

x

X
x

X
X








X
X

x



X
x

X
x

Level I only; Levels II & III require more extensive QC programs, with Level III requiring all program
components, as with demonstration projects.
         Figure 5.  Applicability of quality control program components to project categories.

-------
 operation.   Laboratory  benches, particularly areas where critical operations
 such as  weighting  are carried  out,  should be kept clear of all but necessary
 tools, glassware,  etc.   Personal  items  (coats, hats, lunch boxes) should not
 be left  in  the work area.  Provision  should be made for storage of these items
 in personal lockers.  A neat,  well-organized laboratory area serves to inspire
 neatness and organization  among the laboratory workers.
      Good laboratory maintenance, particularly for certain types of instrumen-
 tation,  requires complete  manuals that  are kept in a convenient place so they
 are readily available to appropriate  personnel.  Responsibility for keeping up
 with any necessary manuals should be  given to an individual, with the under-
 standing that he/she must  devise  a  system (check-in/checkout) for quick
 location of each document.

 6.3  CONFIGURATION CONTROL
      For IERL projects  of  moderate  to long-term duration, the documentation of
 design changes in  the system must be  carried out unfailingly.  Procedures for
 such documentation should  be written  and be accessible to any individual
 responsible for configuration  control.  It is all too easy, as the system is
 modified repeatedly, to allow  one key person to hold, largely by memory, great
 amounts  of  vital information.  Much of  this information would be lost if this
 person were no longer available.  Engineering schematics should be maintained
 current  on  both the system and subsystem level, and all computer programs
 should be listed and flow  charted.  Changes in computer hardware and software
 must  be  documented,  even when  such  changes are apparently trivial.  Significant
 design changes must be  documented and forwarded to the EPA project officer by
 way of established procedure.

 6.4   PERSONNEL TRAINING
      For  long-term projects, it is  highly desirable that there be a programmed
 training  system for new employees.  This system should include motivation
 toward producing data of acceptable quality standards.  A part of the program
should involve "practice work" by the new employee.  The quality of the work
can be immediately verified  and discussed with the supervisor, with appropriate
corrective action  taken.  This system is to be preferred to  on-the-job
                                      18

-------
training, which may be excellent or slipshod, depending upon a number of
circumstances.
     Key personnel (laboratory supervisors, senior engineers) should be
required to document their specialized knowledge and techniques so far as
possible.  They should each be required to develop an assistant, if the
program personnel situation allows, who could take responsibility when the
senior person is unavailable.  A most undesirable situation arises when
replacement personnel must be brought in and forced to gain knowledge of the
program through the experience of trial and error (see subsection 3.3 above).
This is not an infrequent occurrence, however, when budgeting constraints
override other priorities.
     A thorough personnel training program should focus particular attention
on those people whose work directly affects data quality (calibration personnel,
bench chemists, etc.).  These people must be cognizant of  the quality standards
fixed for the project and the reasons for those standards.   They must be made
aware of the various ways of achieving and maintaining quality data.   As these
people progress to higher degrees of proficiency, their accomplishments should
be reviewed and then documented.  A motivating factor for  high performance
could be direct and obvious rewards (monetary, status, or  both), offered in a
manner visible to other comparable personnel.

6.5  DOCUMENTATION CONTROL
     If the project is of the type that generates a number of documents, pro-
cedures for making revisions to these documents must be clearly written out.
The revisions themselves should be written and distributed to all affected
parties, thus insuring that the change will be implemented and become permanent.
If a technical document change pertains to an operational activity, that change
should be analyzed for side effects.  The change should not be rendered
permanent until any harmful side effects have been controlled.
     Revisions to computer software should be written with reasons for the
changes clearly spelled out.  The revisions should be distributed to all
affected parties.
                                      19

-------
 6.6  CONTROL CHARTS
      For demonstration or pilot-plant  programs,  or  any project where data is
 taken on a long-term basis,  control  charts  are essential as a routine check on
 the consistency or "sameness"  of  the data precision.  A control chart should
 be kept for each measurement that directly  affects  the quality of the data.
 Typically, control charts are  maintained for  duplicate analyses, percent iso-
 kinetic sampling rates, calibration  constants, and  the like.  An example control
 chart is given as figure 6.  The  symbol a  (sigma) represents a difference, d,
 of one standard deviation unit in two  duplicate  measurements, one of which is
 taken as a standard or audit value.  Two a  is taken as a warning limit and 3a
 as a control limit.  If a laboratory measurement differs from the audit value
 by more than 3(7, the technique is considered  out of control.  Control charts
 are dealt with in depth in a number  of standard  texts on quality control of
 engineering processes*.

 6.7  IN-PROCESS QUALITY CONTROL
      During routine operation,  critical measurement methods should be checked
 for conformance to standard  operating  conditions (flow rates, reasonableness
 of data being produced, and  the like).  The capability of each method to pro-
 duce data within specification limits  should  be  ascertained by means of
 appropriate control charts.  When a  discrepancy  appears in a measurement method,
 it should be analyzed and corrected  as soon as possible.
      For all standard methods,  the operating  conditions must be clearly defined
 in writing,  with specific reference  to each significant variable.  Auxiliary
 measuring,  gaging,  and analytical instruments should be maintained operative,
 accurate,  and precise by regular  checks and calibrations against stable
 standards  that are  traceable to a primary standard, preferably furnished by
 the U.S.  Bureau of  Standards (if  available).
      Quality Assurance Handbook  for Air  Pollution Measurement Systems, Vol.  lj
Principles. EPA-600/9-76-005.
                                      2Q

-------
          2CT
         -2a
         -3o
 CHECK  NO.
ACTION LIMIT


WARNING LIMIT
                                                                      •UCL
                                                                      -CL
WARNING LIMIT


ACTION LIMIT
                                                                      •LCL
                          8
10
DATE/TIME
 OPERATOR
PROBLEM AND
CORRECTIVE
ACTION
                Figure 6.  Standard quality control chart.
                                  21

-------
 6.8  PROCUREMENT AND INVENTORY  PROCEDURES
      There should be well-defined and  documented purchasing  guidelines for all
 equipment and reagents having an effect  on data quality.  Performance specifica-
 tions should be documented  for  all critical items of equipment.  Chemical
 reagents considered critical  to an analytical procedure are  best procured from
 suppliers who agree to submit samples  for  testing and approval prior to initial
 shipment.  In the case of  incoming equipment, there should be an established
 and documented inspection  procedure to determine if procurements meet the
 quality control and acceptance  requirements.  The results of this  inspection
 procedure should be documented.
      Whenever discrepant materials are detected, the materials are either
 returned or disposed of, at the discretion of the quality control  supervisor.
      Once an item has been received and  accepted, it should  be documented in
 a receiving record log giving a description of the material, the date of the
 receipt, results of the acceptance test, and the signature of the  responsible
 individual.  It is then placed  in inventory, which should be maintained on a
 first-in, first-out basis.   It  should  be identified as to type, age, and
 acceptance status.  In particular, reagents and chemicals that have limited
 shelf life should be identified as to  shelf expiration date  and issued from
 stock only if they are still  within that date.

 6.9  PREVENTIVE MAINTENANCE
      For long-term projects,  it is important that preventive maintenance
 procedures be clearly defined and written  for each measurement system and its
 support  equipment.  When maintenance activity is necessary,  it should be
 documented on standard forms  maintained  in logbooks.  A history of the
 maintenance record of each  system serves to throw light on  the adequacy  of  its
 maintenance schedule and parts  inventory.

 6.10   RELIABILITY
     The  reliability of  each  component of  a measurement  system relates  directly
 to  the probability of obtaining valid  data from  that system.  It  follows that
procedures  for  reliability  data collection, processing, and reporting should
be  clearly  defined and in written form for each  system component.   Reliability
                                      22

-------
data should be recorded on standard forms and kept in a logbook.  If this
procedure is followed, the data can be utilized in revising maintenance and/or
replacement schedules.

6.11  DATA VALIDATION
     Data validation procedures must be defined for each project.  An environ-
mental assessment and a demonstration project will have entirely different
procedures, since in one case the data are taken on a "one-point" basis and in
the other a great quantity of data is accumulated over a long period of time,
usually years.  Whatever the nature of the project, it is important that the
criteria for data validation be documented.  Whenever practical, acceptance
limits should be established, these limits being subject to modification as
the program continues.  Any required data validation activities should be
recorded in standard form in a logbook.  Where possible (as in most demonstra-
tion projects), validation criteria should be programmed, so that routine data-
taking will include automatic flagging of invalid data.
     Most projects should, on a random but regular basis, be subjected to
quality audits.  These audits must be independent of normal project operations,
preferably performed by an independent organization.  Such audits should
include both systems reviews and independent measurement checks.  Section 4.0
will discuss these elements of a quality assurance program.

6.12  FEEDBACK AND CORRECTIVE ACTION
     Closely tied to the detection of invalid data is the problem of establish-
ment of a closed loop mechanism for problem detection, reporting, and correc-
tion.  Here it is important that the problems are reported to those personnel
who can take appropriate action.  For most projects, a feedback and corrective
action mechanism should be written out, with individuals assigned specific
areas of responsibility.  Again, documentation of problems encountered and
actions taken is most important.  Standard forms, kept in a logbook, are
recommended.  If appropriate, a periodic summary report on problems and correc-
tive action should be prepared and distributed to the appropriate levels of
management.  This report should include:  a listing of major problems  for  the
reporting period; names of persons responsible for corrective action;
                                      23

-------
criticality of problems; due dates; present status; trend of quality performance
(i.e., response time, etc.); and a listing of items still open from previous
reports.

6.13  CALIBRATION PROCEDURES
     All IERL-RTP project categories except paper studies involve the taking of
experimental data.  The quality of these data relates directly to the care with
which calibration procedures are carried out.  It is not an exaggeration to say
that calibration procedures are the crux of any attempt to produce quality data
from a measurement system.  For this reason it is extremely important that the
procedures be technically sound and consistent with whatever data quality re-
quirements exist for that system.  Calibration standards must be specified for
all systems and measurement devices, with written procedures for assuring, on
a continuing basis, traceability to primary standards.  Since calibration
personnel may change from time to time, the procedures must be in each instance
clearly written in step-by-step fashion.  Frequency of calibration should be
set and documented, subject to rescheduling as the data are reviewed.  Full
documentation of each calibration and a complete history of calibrations per-
formed on each system are absolutely essential.  This permits a systematic
review of each system's reliability.
     Good calibration techniques and procedures are also vital to short-term
experimental projects, since data taken from these studies may be used as
justification for expanded efforts in certain directions.
                                      24

-------
SECTION  VII      GUIDELINES FOR QUALITY  ASSURANCE  PROGRAMS

7.1  GENERAL STATEMENT
     The  objective of a quality assurance program is to independently evaluate
the quality control program of a project.  The quality assurance program must
be appropriate to the work being done and to the quality control program for
the project data.  As mentioned earlier, it is most important that each project
have provision for an adequate quality control program.  Section 7.2 will give
appropriate quality control statements for the RFP, while sections 7.3 and 7.4
will deal with the evaluation of quality control in the proposal and the work
plan.  The initial stage of a quality assurance program should be assistance
in planning and  development of the project data quality control program.
     Short-term  experimental projects must have rigorous quality control of the
ensuing data.  Quality assurance on such projects may consist of "one-shot"
audits, or there may be no formal quality assurance because of time constraints.
A minimum quality assurance approach would involve sampling a percentage of the
raw data  collected and verifying the calculations.  The techniques used and
general approach could also be reviewed for appropriateness,  and an attempt
made at comparing the work being done with that of other investigators.   This
can be done off-site, if necessary, and at minimal expense.
     For  IERL projects of moderate to long duration, the assessment of quality
control should normally consist of a series of systems and performance audits.
The frequency of such audits obviously should be dictated by the specific
project.  It is  recommended that a minimum frequency be once each calendar
year.  The initial systems and performance audit should take place within the
first quarter of the first project year.  Subsequent scheduling should be
dependent on the requirements of management and the apparent quality of the
day-to-day data  being obtained.  More frequent auditing may be necessary in
the initial stages of the project.

7.2  THE  REQUEST FOR PROPOSAL - QUALITY CONTROL ASPECTS
     The  design  of the RFP is predicated on stating as clearly as possible what
the objectives of the project are; e.g., to design, construct, and maintain a
given control system, systematically examining the interaction of appropriate
                                      25

-------
 system parameters.   The  quality  of  the  data  obtained  from the project will
 depend upon numerous factors—instrumentation, personnel,  sampling  technique,
 sample size, statistical expertise.   It is therefore  critical that  the RFP be
 as  explicit as  possible  in delineating  two things—what quality of  data is
 expected,  and how that quality is to  be insured.
      Generally  speaking,  the  RFP should require that  the  bidding organizations
 address each of the major areas  of  quality control discussed in sections 6.2
 through 6.13 of this manual.  Reference to figure 3 gives  those areas of
 quality control considered particularly appropriate for each project category.
      Since most RFP's are limited in  length,  it would in  most cases be inappro-
 priate to  include more than a brief (one- or two-paragraph) statement of
 quality control requirements. Nevertheless,  it is most important that the bid
 solicitation be as explicit as possible concerning quality control.  In those
 cases where an  RFP is quite lengthy,  the quality control  statement  may be
 several pages long.

 7.3  EVALUATION OF QUALITY CONTROL  IN THE PROPOSAL
      The proposal should contain a  statement as to the precise position the
 bidder's company takes regarding quality control programs.  This should include
 past  projects and the quality control program effectiveness in that project.
 In  particular,  there should be a clear  and explicit response to the quality
 control requirements stated in the  RFP.  This response must be compared directly,
 item-by-item, with other proposals  submitted against  the  RFP.  The  evaluation
 should result in a determination of a "figure of merit" for the bidder's
 quality control organization  and the  competence of the staff.
      There should be provision for  changes in procedures  when it is evident
 that  data  being obtained are  not sufficiently accurate or appropriate  for  the
 intent of  the project as  outlined by  the project officer.
      If a  contractor has  a good  proposal but is unclear on some phases  of  data
 quality, it  would seem worthwhile to  have him clarify his proposal  by  asking
him to answer specific questions.   If the answers to  these questions are  still
vague,  it  is a  good  indication that the quality for these phases of the project
may be  questionable  if this contractor  carries out the project.
                                      26

-------
7.4  EVALUATION OF QUALITY CONTROL IN THE WORK PLAN
     The work plan should be a detailed accounting of the actual steps to be
taken to complete the work delineated in the proposal and should be in direct
accord with the requirements of the KFP and other agreements with the project
officer.  Particular attention should be placed on the areas considered
critical with respect to quality control, in order to realize the collection
of data having acceptable precision, accuracy, representativeness, and complete-
ness.
     In cases where the submitted proposal has been accepted but lacks the
completeness required by the project officer, the problem areas should be
directly addressed in the work plan, showing the details of the work to be
done.
     The work plan must be submitted to the project officer before any work is
begun by the contractor.  The plan can be accepted in draft form, which will
allow for minor changes prior to the final plan's acceptance and approval.

7.5  THE ON-SITE QUALITATIVE SYSTEMS AUDIT
     The objective of the on-site qualitative systems audit is to assess and
document (1) facilities; (2) equipment; (3) systems; (4) recordkeeping;
(5) data validation; (6) operation, maintenance, and calibration procedures;
and (7) reporting aspects of the total quality control program for a project.
The review should accomplish the following:
     1.   Identify existing system documentation; i.e., maintenance manuals,
          organizational structure, operating procedures, etc;
     2.   Evaluate the adequacy of the procedures as documented; and
     3.   Evaluate the degree of use of and adherence to the documented
          procedures in day-to-day operations, based on observed conditions
          and a review of applicable records on file.
     To aid the auditor in performing the review, a checklist is included as
appendix A.  This checklist should be modified, as appropriate, for various
projects.
                                      27

-------
 7.6   THE  PERFORMANCE  AUDIT
      In addition  to a thorough  on-site systems review, quantitative performance
 audits should  be  periodically undertaken.  The objective of these audits is to
 evaluate  the validity of  project  data by independent measurement techniques.
 It  is convenient  to classify the  major measurement methods into three areas:
 physical  measurements,  gas  stream measurements, and liquid stream measurements
 (the latter including analysis  of any suspended solids).  Appendix B lists in
 table form a number of standard techniques for auditing in the three major areas
 just mentioned.   Table 1  (of appendix B) is a compilation of commonly measured
 physical  properties,  with a selection of possible measurement, calibration, and
 audit techniques. Table  2, concentrating on analysis of gas effluent streams,
 lists the material to be  analyzed,  and measurement, calibration, and audit
 techniques for that material.   Finally, table 3 very briefly and generally
 deals with measurement methods  appropriate to liquids and solids.  The specific
 techniques vary widely from project to project, but for the analytical phase
 the  audit technique generally involves use of reference samples of known
 composition and/or splitting a  sample among several laboratories for independent
 analyses.   It  is  desirable  to perform calibration checks on individual system
 components, and/or do side-by-side  sampling runs to compare both sampling and
 analysis  technique precision and  accuracy.

 7.7   DATA QUALITY ASSESSMENT
      Standard  methods exist for estimation of the precision and accuracy of
 measurement data.  Efficient usage  of the audit data requires that a rationale
 be followed which gives the best  possible estimates of precision and accuracy
 within the limits imposed by timing, number of samples taken, and the general
 situation  at the  project  site.
      Appendix  C lists statistical definitions and techniques often used  in
 quality assurance  work.   Other  statistical techniques exist which may apply  to
 specific projects  (or to  highly specialized areas of a given project).   For
projects of sufficient duration,  it is usually worthwhile to acquire  the
services of a  statistical consultant to most effectively treat  the available
data.
                                       28

-------
     As a general guide to expected data quality for a number of reference
methods, appendix D lists in table form both ambient air and source sampling
methods.  An estimate of the method bias and precision with comments on major
error sources is given, and the appropriate EPA quality assurance guideline
document for each method is referenced.
                                       29

-------
 SECTION VIII      DEVELOPMENT OF  QUALITY CONTROL AND QUALITY
                     ASSURANCE  PLANS FOR  IERL PROJECT CATEGORIES

 8.0  GENERAL
      One approach to  developing quality control and quality assurance plans
 within IERL is to divide  all projects  into six categories,  with projects within
 a given category having common characteristics as to size,  duration,  objectives,
 and data quality requirements.  This makes them amenable to the same  general
 set of quality control  and quality assurance practices and procedures.   Sugges-
 tions as to applicable  practices  and procedures are given for each of the
 project categories in the following subsections.
      The primary characteristics  and objectives of projects within each  project
 category are given, followed by a discussion of quality control and quality
 assurance applications  to the different phases of a project cycle. These
 phases, when the project  is performed  by a contractor, include (1) request for
 proposal, (2)  proposal  evaluation, (3) work plan review, (4) project  implementa-
 tion, and (5)  final report.  Quality control practices for in-house projects
 could begin with the  work plan review.  The categories 1 through 6 are:
      1.   Environmental assessments,
      2.   Industry system studies/pollutant system studies,
      3.   Field studies,
      4.   Research and  bench-scale projects,
      5.   Development or  pilot programs, and
      6.   Demonstration projects.

8.1   ENVIRONMENTAL ASSESSMENTS
8.1.1  Description of Project Category
      An environmental assessment  involves "(1) a systematic evaluation of the
physical, chemical, and biological characteristics of the streams associated
with  a  process;  (2) predictions of the probable effects of the streams on the
environment; (3) the prioritization of the streams; and (4) identification of
                                       3Q

-------
any necessary control technology programs."*  The objectives of an assessment
include identification and measurement of pollutants for which specific
standards have been set, and also those suspected to have harmful effects on the
environment.  The ultimate goal of environmental assessments is insurance that
the effluent streams from a process are within current environmental standards
of acceptability.
     Sampling and analysis programs for environmental assessments are extensive,
and presently are conducted in three distinct phases or levels.  Level I in-
volves a broad survey of all process streams.  Level II concentrates on those
streams identified in Level I as "high priority," i.e., having significant
amounts of harmful materials present.  These prioritized streams are subjected
to more quantitative studies, for the purpose of establishing the appropriate
control technology.  Level III is a long-term, continuous monitoring program
for selected "indicator" materials.
8.1.2  Applicability of Quality Control and Quality Assurance Procedures
Request for Proposal
     The RFP must specify the data quality requirementst for each level of the
assessment.  For purposes of process stream prioritization,  the RFP for a given
industrial or energy process should indicate, if possible, the "threshold"
concentrations of specified suspected or known pollutants.  These concentrations
would be the minimum for designating those pollutants for more quantitative
study in Level II.  It should be stated that such designated concentrations are
subject to modification in light of the results of Level I studies.
     Because of the completeness and comprehensiveness of the sampling and
analysis programs in an environmental assessment, the RFP will not generally be
able to anticipate the data quality requirements for each measurement.  Never-
theless, every attempt should be made to specify the quality control and quality
assurance requirements as completely as possible, so that prospective contractors
      Taken from a draft copy of "Guidelines for Environmental Sampling and
Analysis Programs - Historical Development and Strategy of a Phased Approach,"
by Dorsey, Lochmuller, Johnson and Statnick, IERL-EPA-RTP, 1976.
      As quantitatively as possible.
                                      31

-------
 will be realistic in their  estimations  of  the  cost of quality  control for the
 assessment.
 Proposal Evaluation
      Environmental assessments  generally require extensive sampling and analyt-
 ical capability.   Each proposal must be evaluated on the known personnel and
 equipment inventory of the  bidding  organization.  Instrumentation requirements
 may be large, and there must be experienced  technical and professional personnel
 to carry out each level of  the  assessment.   The ability to produce high-quality
 data, within whatever specifications were  made in the RFP, should enter heavily
 into the overall  evaluation of  each proposal.
 Work Plan Review
      The work plan affords  the  contractor  the  opportunity to demonstrate in
 some detail how the assessment  will be  carried out, and in particular how the
 data quality will be assured.  The  project officer, after reviewing the work
 plan with the Quality Assurance Group representative, may wish to negotiate
 more or less time or resources  to various  phases of the assessment, in line
 with EPA's conception of the program's  priorities and data quality requirements.
 The work plan should be as  definite as  possible with respect to how the data
 will be acquired  and what quality control  procedures will be employed.
 Project Implementation
      Periodic reports to the project officer should contain appropriate sample
 data,  statistical treatments, and contractor judgments as to the quality of
 data being currently obtained.  These reports  should be carefully read and
 analyzed by  the project officer and Quality  Assurance Group representative,
 since  these  reports serve as the  primary means of communication from the
 contractor.   If inadequate  data quality procedures are being used, the project
 officer should immediately  make arrangements to improve quality control at  the
 project,  and perhaps  should modify  the  nature  and/or frequency of systems
 reviews  and  performance audits.*  These quality assurance programs can, if
     *
      The assumption here  is  that  the  reviews  and audits  can be carried out by
an audit team which visits the  industrial  or energy site  and makes comparable
measurements, as well as observing the contractor team in operation.   This may
not be practical in some situations.

                                       32

-------
correctly timed, point out contractor quality control deficiencies and make  for
minimum data loss.
Final Report
     Quality control and quality assurance procedures must be included in the
final report of an environmental assessment, since the assessment is a phased
operation, and decisions as to priority will be made after each phase is
completed.  Quality control and quality assurance must be documented such that
sufficient confidence may be placed in the data to justify the decisions made.

8.2  INDUSTRY SYSTEM STUDIES/POLLUTANT SYSTEM STUDIES
8.2.1  Description of Project Category
     This category includes projects that are paper studies involving the
collection and analysis of air pollution measurement data from the literature
or personal contacts.
     An industry system study as performed in-house or by a contractor for the
IERL Laboratory is of short to medium duration (normally 3 to 6 months).
Typical objectives of an industry system study are the following:
     1.   To characterize an industry with respect to the type and magnitude
          of air pollutant emissions;
     2.   To assess the present degree of control of these emissions; and
     3.   To evaluate the technical feasibility and the economic and environ-
          mental costs and benefits of improved control methods.
     Pollutant system studies are also usually of short duration.  The purpose
of these studies is to identify the sources and to estimate the concentration
of a particular pollutant.
     The data reported in a paper study must be of sufficient quality to allow
IERL to determine the appropriate followup action.  Quality requirements will
depend upon the financial implications of the decisions to be made on the basis
of the data.  The quantitative confidence level exhibiting these characteristics
varies with the type of pollutant.  For example, an industry system study
reporting of a nitrogen oxide emission level as 50 ppm + 30 ppm would probably
provide an adequate basis from which to determine that the emissions were of
little consequence, whereas reporting the emission of vinyl chloride to 50 ppm
                                      33

-------
 + 30 ppm would most likely spur an immediate followup for additional informa-
 tion.  In general, the accuracy and precision of data reported in a paper study
 is not as critical as the data quality from, for example, a specific pilot plant
 study, since the paper study will be for general program development decisions
 and will be followed by more detailed and specific projects when necessary.
 8.2.2  Application of Quality Control and Quality Assurance Procedures
 Request for Proposal
      Quality assurance for a study begins with preparation of the RFP.  The
 project officer should consult with the Quality Assurance Group to insure that
 proposals can be evaluated on an equivalent basis.  The RFP should include the
 specific type of data to be obtained and an indication of the type of precision
 and accuracy statements to be attached to the reported data.  For example, in
 a pollutant system study (or industry system study), it may be stated that:
      1.   "All pollutant sources should be identified for which the concen-
           tration of NO  exceeds 50 ppm;"
                        -cv
 or
      2.   "Concentration of NO  is to be given for identified sources with
                               j^
           estimates of the precision of the reported levels, e.g., + 20 percent
           of the reported value."
      The RFP should specify or request the bidder to estimate the sample size
 necessary to insure that the sample is representative of the population of
 interest.   General requirements for statistical evaluation of the data obtained
 as well as guidelines for referencing and assignment of confidence levels to
 the data should be given in the RFP.
 Proposal Evaluation
     A representative from the Quality Assurance Group should evaluate the
 proposals  and  prepare a report to the project officer describing the ability of
 each contractor to accomplish the above requirements.  The contractor's proposal
 rationale  for  analyzing the data and making inferences and conclusions also
provides a preview of quality control for the project.  For example,  the means
by which a contractor may propose to estimate the data quality if 2. above is
the suggested  objective might be:  (1) to use the Federal Register guidelines
on precision and accuracy for standard reference methods when the information
                                      34

-------
is not presented in a report,  (2) to compare  to  the reference methods  if  the
results are quoted with precision statements, or  (3)  to call the author of  the
report/data tabulations if neither of  the above  is applicable.
Work Plan Review
     When the contract is awarded, preparation and review of the work  plan
allows refinement of the quality control procedures outlined in the proposal.
The project officer in conjunction with the Quality Assurance Group representa-
tive can ascertain if the contractor has allowed  sufficient time and resources
to develop a reasonable data base from which  statistical correlations  can be
developed.  The proposed methodology of analyzing and drawing conclusions from
the data can also be reviewed  at this  time for adequacy and completeness.  The
work plan is expected to be more definitive than  the proposal and should address
the data quality aspects, when required, even if  the proposal did not respond
to those.
Project Implementation
     During the course of the  project, monthly reports from the contractor to
the project officer provide the opportunity for continuing quality assurance.
Review of monthly reports provides the project officer the opportunity to
detect quality problems by comparing the interim  results against acceptance
criteria established in the work plan.  Early detection of problems can lead
to correction in a timely manner to avoid the use of questionable data or the
loss of time occasioned by correcting  deficiencies late in the program.  Correc-
tion of the problem may consist of modification  of the information collection
or correlation procedure or perhaps a  redefinition of the project objectives
in light of the accumulated experiences.
Final Report
     The results of the study  should be evaluated in view of several points of
consideration.  The degree of  depth of the evaluation clearly depends  on the
data quality objectives.  Some of the  points  to be considered are:
     1.   Data sources;
     2.   Inferences from data on:
          a.   the effect of data precision and accuracy on conclusions,
          b.   the appropriateness of  the analysis procedure  (engineering and
               statistics);

                                     35

-------
      3.    Comparison with comparable studies/research that are considered to
           be of good quality;
      4.    Limitations of data;  and
      5.    The need to make measurements.
      A summary of the applicability of the reported data along with the
 appropriate comments on their  quality should be included in the final report
 when requested.  This information should  be reviewed by the Quality Assurance
 Group and reported in abbreviated form to the laboratory director  with a  copy
 to the project officer.  These statements provide a source of  information on
 the adherence of the contractor to the data quality criteria as described in
 the request for proposal, proposal, and/or work plan.   This report by the
 Quality Assurance Group serves as a project evaluation file and a  source  of
 quality assurance information.

 8.3  FIELD STUDIES
 8.3.1  Description of Project  Category
      Projects in this category are usually of short duration,  i.e., less  than
 2 or 3 months.  There will be  a great variation in data quality requirements
 depending on the project objectives.  A field study may, for example, be  for
 evaluation of a new device for sizing submicron particles.
      The conduct of such studies will typically require a contractor familiar
 with the industry, control system, or device to be assessed or evaluated.  The
 quality  control techniques for the pertinent measurement systems will need to
 be  supplied by the project officer and Quality Assurance Group. The contractor
 should be required to use standard reference methods or an equivalent method,
 if  possible.   The burden of proof of using another equivalent  method would be
 upon the contractor.   The justification for using nonreference methods should
 be  clearly stated.
 8.3.2  Application of Quality  Control and Quality Assurance Procedures
     The degree of  precision and accuracy requested in the reported results of
field  studies  may vary considerably from  one study to another, and consequently
it  is  necessary to  specify as  carefully as possible the requirements of the
prospective contractor in the  RFP.  For example, measurement methods may be
                                      36

-------
specified, the desired precision of the results given, and  the responder asked
to indicate in the proposal how the desired data quality levels can be attained
through proper measurement methods and appropriate quality  control procedures.
The Quality Assurance Group can provide advice as to what levels of precision
are attainable, based on past experience.
Proposal Evaluation
     The proposals submitted can be reviewed and rated as to their responses
to .the data quality aspects and the quality assurance methods.  The Quality
Assurance Group would utilize its file of past experience on contractors to
aid in this evaluation.
Work Plan Review
     After contract award, the work plan becomes one of the most effective
means for assurance of data quality through the use of appropriate measurement
methods and quality control techniques, and the use of suggested standards for
checking on the precision and accuracy of the method.  The  Quality Assurance
Group will review the work plan in detail for its inclusion of applicable
quality control techniques.
Project Implementation
     During the project implementation phase, the Quality Assurance Group will
be kept informed through periodic reports from the contractor to the project
officer and from the project officer  concerning adherence to the work plan, any
problems in doing so, and suggested corrections to problems.  If the Quality
Assurance Group can advise as to the  means for assuring data quality, it should
indicate this to the project officer  for consideration/use  by the contractor.
The contractor's proficiency in making the measurements can be evaluated and
estimates of the precision and accuracy of the measurements made by the use of
on-site system reviews and audits, conducted at the beginning and periodically
throughout the project life.
Final Report
     At the termination of the contract and upon submittal  of the final report,
the Quality Assurance Group should file an evaluation  report on  the  study
briefly summarizing the pertinent data quality information, the  capability of
          V
the contractor to produce high-quality data, and general comments on the

                                      37

-------
contractor's  quality assurance procedures as appropriate.  This file of
evaluative  information provides a ready source of data for future contractor
evaluation  and data quality improvement techniques.  The good techniques are
just  as  important  to include in the evaluation as the deficient ones.

8.4   RESEARCH AND  BENCH-SCALE PROJECTS
8.4.1 Description of Project Category
      Research and  bench-scale projects are exploratory studies of a pollutant
measurement or control method, with the primary objectives of developing and
evaluating  a  successful process or device.  The duration of this type of
project  is  variable and may continue  concurrently with pilot plant or prototype
studies.
      Research projects are defined here as primarily paper studies consisting
of  the theoretical development of a new technology.  The bench-scale project is
the first experimental attempt to demonstrate the postulated results.  The
data  quality  requirements of a bench-scale study are dependent on the nature
of  the particular  project.  For example, documentation of the precision and
accuracy of measurements would be critical when bench-scale project results
support  pilot plant studies, whereas  qualitative measurements with an accuracy
of +  20  percent  may be adequate for a preliminary demonstration of the feasi-
bility of a new  pollutant control device.
8.4.2 Application of Quality Control and Quality Assurance Procedures
Request  for Proposal
      An  RFP leading to research and bench-scale studies may of necessity  be
general  since IERL cannot always anticipate the characteristics of novel  devices
and innovative techniques proposed by contractors.  However,  the RFP and
proposal evaluation should require a  discussion of measurement systems and basis
for selection.
Work Plan Review
     When the contract is awarded, more specific guidelines can be  laid  out  in
the preparation and review of the work plan, although a  stringently defined
test plan is not always applicable to a bench-scale study.  The project  may  be
                                      38

-------
conducted in a more cost-effective manner by allowing the researcher  to make
exploratory studies with inexact measurements in the early stages of  the project
and then to concentrate on additional measurement development specific to the
most promising directions.
Project Implementation
     During the course of the project, frequent communication between the
contractor, the project officer, and the Quality Assurance Group is essential
for quality assurance.  Written monthly reports are useful, but frequently more
cost-effective project management and quality assurance can be ascertained
through personal conferences to review the progress of the work and the intended
direction for continuation.  If the project is of such a nature that measurements
of high precision and accuracy are required, the project officer should require
documentation of the procedure for operations, calibrations, quality checks,
and validation of data.  Data quality may be assessed by means of performance
audits using duplicate, control, and/or blind samples.  These methods may
detect quality problems and allow for their correction early in the program.
Final Report
     Quality assurance in the final report of a research or bench-scale project
is of critical importance, since the study may lead to additional funding of
the process or equipment development at a more sophisticated level.  The methods
used for sampling and measurement of data, for analysis, and for correlation
must be rigorously presented so that the results could be duplicated in an
independent study if desired.

8.5  DEVELOPMENT OR PILOT PROGRAM
8.5.1  Description of Project Category
     This program has as one of its major objectives the development and refine-
ment of theoretical and empirical models relating the process variables and
system parameters to the response variables of interest, e.g., concentration of
particulate matter.  Hence, this is the stage of development in which an
experiment consisting of a series of runs is planned for deriving the relation-
ships of interest.  In some cases the determination will be made of "optimum"
values of control system parameters for various plant parameters.  Data quality
                                      39

-------
 aspects  are described  in the  steps  of  the  contract  procurement and  implementa-
 tion cycle.
 8.5.2 Application of  Quality Control  and  Quality Assurance  Procedures
 Request  for Proposal
      In  this stage of  control equipment  development,  the  objectives of  the
 program  should be clearly indicated as to  data  quality.   The statement  of work
 should include the request that the contractor  suggest an experimental  plan
 that will assure the attainment of  the quality  objectives (assuming these
 objectives to be compatible with the existing or developed technology).  The
 RFP should give relatively high priority to the data  quality control plan as an
 evaluation criterion of the proposal.   Types of statements which might  be
 employed in the RFP are:
      "Submit a proposed experimental plan  to relate variables X, Y, Z,
      and W to concentration of the  pollutant in the effluent stream
      and indicate how  the precision of the relationship will be checked
      by  the use of the design plan  and the associated analysis."
      "Indicate the methods of measurement  of critical variables and the
      expected precision and accuracy of  these methods."
 Proposal Evaluation
      At  this stage, the project officer  and Quality Assurance Group should
 review the data quality aspects of  the proposals for  the  purpose of rating  the
 prospective contractors on their apparent  ability to  provide data  of good
 quality.   At this point it should be noted that past  records of contractor
 performance will be very helpful.  Judgment of  new  prospective contractors  must
 be  made  on the basis of a careful evaluation of their discussion.   Absence  of
 any details on data quality aspects, even  though the  discussion may have very
 good general ideas, would be  a signal  for  concern and lower  ratings.
 Work Plan Review
      Shortly after the award  of the contract to a company, the Quality Assurance
Group should assist the project officer in the  role of  evaluating the adequacy
 of  the quality control aspects of the  work plan.  The plan should have consid-
erably more detail concerning data  quality than the proposal.  Also it is
possible  that  the selected contractor  was  strong on several other rating
                                      4Q

-------
criteria in the proposal evaluation but lower in the quality control aspects.
This would be the appropriate time to request that the contractor strengthen
the data quality program, in particular the experimental plan.  At this point,
in order to insure good quality data that will meet the needs of IERL, the
following factors should be considered:
     1.   The experimental plan to attain objectives;
     2.   The means for adherence to the plan—control of system parameters;
     3.   Measurement methods of sufficient precision and accuracy for meeting
          the project needs, and use of standards in checking the data quality;
     4.   Appropriate checks of measurements under specified conditions/
          control system parameters (i.e., independent or repeat runs should
          be included in the experimental plan, and duplicate measurements
          should be made for those results with large variation; for example,
          if the coefficient of variation [100 x standard deviation/mean] is
          greater than 5 percent).
     5.   Data reporting and quality checks included in plan;
     6.   Analysis methods with indicated checks, if necessary, to insure
          valid results.
     The ultimate benefits of a well-designed experimental program should be
the provision of the greatest information per unit cost of the experiment.
Because the cost of a single run of an experiment of the type required at this
stage of development is high, it places the greatest importance on making the
best possible use of the data, e.g., utilizing the preferred analysis methods
resulting in valid interpretations with attached measures of confidence/preci-
sion.
Project Implementation
     Throughout the conduct of the project work, the Quality Assurance Group
should be made aware of any quality control problems through appropriate
reports, e.g., monthly progress reports and quarterly reports.  The project
officer should provide for on-site checks of the quality control procedures.
This may be performed by personnel of the Quality Assurance Group or a
contractor so designated by the Group, with a written report provided to  the
Quality Assurance Group.  The Quality Assurance Group should advise the project
officer and project contractor on such matters of quality control as:
                                      41

-------
      1.    use of  specific  standards  for  checking  pollutant measurement methods,
      2.    how often checks should  be made,
      3.    calibration frequency,
      4.    use of  quality control  techniques  as  required,  and
      5.    interlaboratory  programs for pollutant  measurement methods.
 Final Report
      The final report on the project should  include  the necessary  information
 on data quality.   Raw data should  be provided as  an  appendix to  the report.
 Only in circumstances of unusually large amounts  of  data  should  they be  sepa-
 rated from the report, e.g., provided in one copy by computer printout.  The
 analysis of the data should be illustrated by example, the statement of
 precision and accuracy of  the data included, with method  of computation,
 particularly when different from well-documented  methods. If relationships
 are presented (e.g., concentration of S02 vs. selected control system parameters)
 raw data should be shown on the figures, if  two-dimensional.  Otherwise, the
 raw data and corresponding predicted mean given by the relationship are  to be
 given in tabulated form.  This allows more ready  interpretation  by IERL.  The
 raw data should be given by order  of the experimental run along  with remarks or
 conditions surrounding the run which may be  pertinent in  some later evaluation
 or analysis.  The inability to obtain data for  certain parameter values  should
 be noted.   Pertinent raw data on  the instrument checks should also be included
 in the report with supporting data maintained in  the lab  notebooks, should
 any questions arise.  A statement  of limitations  and applicability of the data
 should be  included in the  final report.
      Based on the final report, the  Quality  Assurance Group  should provide a
 summary  report to the director of  IERL with  a copy to the project  officer
 indicating the pertinent data aspects of the project. A checklist of consid-
 erations should be employed for convenience  in  such  evaluations.

 8.6   DEMONSTRATION PROJECTS
 8.6.1 Description of Project Category
      These projects have as their  goal the demonstration of  control systems,
models,  and methods in full-scale  systems,  i.e.,  under plant operational
                                      42

-------
conditions.  The control systems methods and models will have been developed
under previous research and pilot-scale operations.  The demonstration projects
usually will be of at least 1-year duration.  The values of the control system
parameters and plant operation variables will be changed very little, offering
little opportunity for experimental research.  There will be the opportunity
for the use of both quality control and quality assurance techniques.  The data
quality will be an important aspect of this type of study.  For example, quality
control chart data should be maintained on the critical parameters and response
measurements and these data provided with the final report in an appropriate
summary form.  There will be the need for using well-documented measurement
methods of desired precision/accuracy.  Standards should be used to calibrate
and maintain control of quality of the measurement methods.  The data quality
aspects should be described in reference to the steps of the procurement and
project implementation stages of the control system demonstration and evaluation.
8.6.2  Application of Quality Control and Quality Assurance Procedures
Request for Proposal
     The objectives of the demonstration program for the specified control
system and its associated models and methods of analysis should be clearly
indicated in the RFP.  The contractor should be requested to give relatively
high priority to data quality, quality control and assessment techniques,
maintenance and calibration of pollutant measurement instruments, use of
appropriate standards, and other general quality control capabilities in the
proposal.
     The Quality Assurance Group should provide inputs to the RFP based on the
previous studies of the control system and the requirements of data precision
and accuracy.  An example of such a statement would be:
     "Methods used for measuring the emissions of the pollutant
     must include the standard reference methods of the Federal
     Register—or an equivalent method, provided proof is given
     that the method employed is equally precise and accurate."
Proposal Evaluation
     The proposals should be reviewed and rated as to the discussion of the
data quality aspects of the demonstration study.  This rating should be based
on the prospective contractor's familiarity with quality control procedures,
                                      43

-------
 his  current  quality  control activities, performance on previous studies if
 available  from  the Quality Assurance Group files, and other general laboratory
 practices, such as preventive maintenance of equipment.  The Quality Assurance
 Group  should be responsible for review of this aspect of the proposals.
 Work Plan Review
     The selected contractor should prepare a work plan shortly after contract
 award,  detailing the project demonstration plan, which includes the quality
 control aspects of the  study.  The work plan should contain much more detail
 concerning quality control than the proposal.  In cases where quality control
 may  have been a weak point of the selected contractor's proposal, this aspect
 of the work  plan will need to be reviewed carefully to assure that:
     1.   The measurement program will meet the objectives of the study if
           properly implemented;
     2.   Means are  presented for controlling and monitoring the plant and
           control system parameters;
     3.   Measurement methods are sufficiently precise and accurate for
          meeting the project needs; that standards will be used to calibrate
          and maintain  instruments, and interlaboratory and intralaboratory
          measurements  will be used to assure data quality;
     4.   Quality control charts will be maintained on selected parameters
          deemed to  be  important;
     5.   Data  reporting and quality checks will be included in the plan
          and data quality will be assessed;
     6.   Problem identification, e.g., out-of-control operations, will be
          indicated, with progress reports that give the problem and its
          effect on  data quality.
Project Implementation
     During  the implementation stage, the Quality Assurance Group should be
provided a copy of all  progress reports that relate to any aspect of data
quality.  At  certain indicated milestones in the project work,  the group
should check  to  see  that the contractor is adhering to the work plan  (data
quality aspects), and should advise the project officer and contractor of  the
availability  of  certain standards, estimated precisions of specific methods
                                      44

-------
(if available and unknown to them), and uses of quality control charts.  The
Quality Assurance Group is to serve for the duration of the project both as a
resource group, because of its records of similar information from several
projects that may use the same methods, and as an advisory source of information
on general aspects of quality assurance.  The Quality Assurance Group should
design and coordinate through the project officer a program of performance
audits and on-site system inspections sufficient to assure, assess, and docu-
ment the data quality throughout the project's duration.
Final Report
     The final report serves as a means of reporting and properly summarizing
the raw data, data quality, the precision and accuracy of measurement methods,
the methods of engineering and statistical analysis, and quality control chart
information.  All raw data should be included in the final report, to the extent
possible.  If it is excessive, then one computer printout might be filed with
IERL to back up the laboratory notebook maintained by the contractor, should
any problems arise in interpretation and analysis at a later date.
     Methods of analysis of data quality should be provided unless they are
well-documented in the literature, in which case they may be referenced.
     A statement of limitation, if any, and applicability of the results should
be given in the report.  This may be part of an executive summary or in a
section entitled "Data Quality — Applicability of Methods and Results."
     The results of the project demonstration as reported in the final report
should be assessed by the Quality Assurance Group, against both the data
quality objectives set forth in the RFP and the details provided in the work
plan.  The contractor should be rated according to a checklist of items to be
used in such studies.  A copy of this evaluation should be given to the director
of IERL, to the project officer, and perhaps to QAB.  This file of project
evaluation closes the loop and provides useful information for future projects,
proposal evaluations, RFP statements, and alternate improvements of data
quality in all IERL-RTP projects in the area of activity.
                                      45

-------
SECTION  IX                     REFERENCES
1.   Glossary and Tables for Statistical Quality Control,  the American  Society
     for Quality Control, Milwaukee,  Wisconsin,  1973.

2.   Quality Assurance Handbook for Air Pollution Measurement Systems,
     Volume 1, Principles, EPA-600/9-76-005.

3.   "Planning Document for a Control Systems Laboratory Quality Assurance
     Program," Final Report for EPA Contract  No.  68-02-1398, Task  8,
     December 1974.

4.   "Guidelines for Demonstration Project Quality Assurance Programs,"
     Final Report for EPA Contract No. 68-02-1398, Task 20, January 1976.

5.   "A Quality Assurance Program for the EPA Wet Limestone Scrubber Demon-
     stration Project, Shawnee Steam-Electric Plant,  Paducah, Kentucky,"
     Final Report for EPA Contract No. 68-02-1398, Task 20, January 1976.

6.   "Development and Trial Field Application of a Quality Assurance Program
     for IERL Projects," Final Report for EPA Contract No. 68-02-1398,  Task
     20, January 1976.
                                      46

-------
                   APPENDIX A  QUALITATIVE ON-SITE SYSTEMS
                               AUDIT CHECKLIST
     This checklist gives three descriptions to each facet of a typical
quality control system.  In all cases, the "5" choice indicates the most
desirable and effective mode of operation; "3" is marginal and tolerable;
"1" is definitely unacceptable and ineffective as a mode of operation.
     It is not always possible to describe accurately all options with  only
three choices.  Therefore, a "2" or "4" rating may be selected if the
evaluator feels that an in-between score is more descriptive of the actual
situation.
     After all the applicable questions are answered, an average is computed
to give an overall indication of the quality control system effectiveness.
     Generally, a rating of 3.8 or better is considered acceptable.
     A rating between 2.5 and 3.8 indicates a need for improvement but  no
imminent threat to current project performance.
                                       47

-------
A.I  QUALITY ORGANIZATION

                                                                      SCORE
      (1.1)  Overall responsibility for quality assurance  (or
            quality control)  for the organization is:

             (a)  Assigned  to  one individual by title  (e.g.,
                 Quality Control Coordinator).                          5

             (b)  Assigned  to  a  specific  group within  the  organization.  3


             (.c)  Not  specifically assigned but left to  the discre-
                 tion of the  various operational, analytical,  inspect
                 tion, and testing personnel.                           1


      Cl.2)   The  Quality Control Coordinator is located  in the
             organization such that;

             (a)  He has direct  access  to the  top management level
                 for  the total  operation, independent of  others  in-
                 volved in operational activities.                      5

             (b)  He performs  as a peer with others involved in
                 operational  activities, with access  to top manage-
                 ment through the normal chain of command.              3

             (c)  His  primary  responsibility is in operational
                 activities,  with quality assurance as  an extra  or
                 part-time effort.                                      1
      (1.3)   Data reports  on quality are distributed by the Quality
             Control•Coordinator to;

             (a)   All levels of  management.*                            5

             (b)   One level of management only.                          3

             (c)   The quality control group only.                        1


      (1.4)   Data Quality  Reports contain;

             (a)   Information on operational trends, required
                  actions,  and danger spots.                            5

             (b)   Information on suspected data/analyses and
                  their  causes.                                          3

             (c)   Percent  of valid data per month.                       1
     Management at appropriate  levels  in all applicable organizations such
as subcontractors, prime  contractor,  EPA.

                                      48

-------
A.2  THE QUALITY SYSTEM
                                                                     SCORE

     (2.1)  The quality control system is:

            (a)  Formalized and documented by a set of procedures
                 which clearly describe the activities necessary
                 and sufficient to achieve desired quality objec-
                 tives, from procurement through to reporting data
                 to the EPA/RTP.                                       5

            (b)  Contained in methods procedures or is implicit in
                 those procedures.  Experience with the materials,
                 product, and equipment is needed for continuity
                 of control.                                           3

            Cc)  Undefined in any procedures and is left to the cur-
                 rent managers or supervisors to determine as the
                 situation dictates.                                   1
     C2.2)  Support for quality goals and results is indicated by:

            (a)  A clear statement of quality objectives by the top
                 executive, with continuing visible evidence of its
                 sincerity, to all levels of the organization.         5

            (b)  Periodic meetings among operations personnel and the
                 individual(s) responsible for quality assurance, on
                 quality objectives and progress toward their achieve-
                 ment .                                                 3

            (c)  A "one-shot" statement of the desire for product
                 quality by the top executive, after which the quality
                 assurance staff is on its own.                        1
     (2.3)  Accountability for quality  is:

            (a)  Clearly defined for all sections and operators/
                 analysts where their actions have an impact on
                 quality.                                              5

            (b)  Vested with the Quality Control Coordinator who
                 must use whatever means possible to achieve quality
                 goals.                                                3

            Cc)  Not defined.                                          1
                                       49

-------
A.2  THE QUALITY SYSTEM (continued)
                                                                     SCORE

     (2.4)  The acceptance criteria for the level of quality
            of the demonstration projects routine performance are:

            (a)  Clearly defined in writing for all characteris-
                 tics .                                                 5

            (b)  Defined in writing for some characteristics
                 and some are dependent on experience, memory
                 and/or verbal communication.                          3

            (c)  Only defined by experience and verbal communica-
                 tion .                                                 1
     (2.5)  Acceptance criteria for the level of quality of the
            project's routine performance are determined by:

            (a)  Monitoring the performance in a structured pro-
                 gram of inter- and intralaboratory evaluations.

            (b)  Scientific determination of what is technically
                 feasible.

            (c)  Laboratory determination of what can be done using
                 currently available equipment, techniques, and
                 manpower.
     (2.6)  Decisions on acceptability of questionable results are
            made by:

            (a)  A review group consisting of the chief chemist or
                 engineer, quality control, and others who can render
                 expert judgment.                                      5

            (b)  An informal assessment by quality control.            3

            (c)  The operator/chemist.                                 1
                                     50

-------
A.2  THE QUALITY SYSTEM (continued)
                                                                     SCORE

     (2.7)  The quality control coordinator has the authority to:

            (a)  Affect the quality of analytical results by in-
                 serting controls to assure that the methods meet
                 the requirements for precision, accuracy, sensi-
                 tivity, and specificity.                              5

            (b)  Reject suspected results and stop any method that
                 projects high levels of discrepancies.                3

            (c)  Submit suspected results to management for a
                 decision on disposition.                              1


A.3  IN-PROCESS QUALITY ASSURANCE
     (3.1)  Measurement methods are checked:

            (a)  During operation for conformance to operating
                 conditions and to specifications,  e.g.,  flow rates,
                 reasonableness of data, etc.                           5

            (b)  During calibration to determine acceptability
                 of the results.                                       3

            (c)  Only when malfunctions are reported.                  1

     •>
     (3.2)  The capability of the method to produce within
            specification limit is:

            (a)  Known through method capability analysis (X-R
                 Charts) to be able to produce consistently
                 acceptable results.                                   5

            (b)  Assumed to be able to produce a reasonably
                 acceptable result.                                    3

            (c)  Unknown.                                              1


     (3.3)  Method determination discrepancies are:

            (a)  Analyzed immediately to seek out the causes and
                 apply corrective action.                              5

            (b)  Checked out when time permits.                        3

            (c)  Not detectable with present controls and procedures.  1

                                      51

-------
A.3  IN-PROCESS QUALITY ASSURANCE (continued)
                                                                     SCORE

     (3.4)  The operating conditions (e.g., flow rate,  range,
            temperature, etc.) of the methods are:

            (a)  Clearly defined in writing in the method for each
                 significant variable.                                 5

            (b)  Controlled by supervision based on general guide-
                 lines .                                                3

            (c)  Left up to the operator/analyst.                      1
     (3.5)  Auxiliary measuring, gaging, and analytical
            instruments are:

            (a)  Maintained operative, accurate, and precise
                 by regular checks and calibrations against
                 stable standards which are traceable to the
                 U.S. Bureau of Standards.

            (b)  Periodically checked against a zero point or
                 other reference and examined for evidence of
                 physical damage, wear or inadequate maintenance.

            (c)  Checked only when they stop working or when ex-
                 cessive defects are experienced which can be
                 traced to inadequate instrumentation.
A.4  CONFIGURATION CONTROL
            Procedures for documenting, for the record, any design
            change in the system are:
            (a)  Written down and readily accessible to those
                 individuals responsible for configuration con-
                 trol.                                                 5

            (b)  Written down but not in detail.                       3

            (c)  Not documented.                                       1
                                      52

-------
A.4  CONFIGURATION CONTROL (continued)
                                                                     SCORE

     (4.2)  Engineering schematics are:

            (a)  Maintained current on the system and subsystem
                 levels.                                               5

            (b)  Maintained current on certain subsystems only.        3

            (c)  Not maintained current.                               1


     (4.3)  All computer programs are:

            (a)  Documented and flow charted.                          5

            (b)  Flow charted.                                         3

            (c)  Summarized.                                           1


     (4.4)  Procedures for transmitting significant design changes
            in hardware and/or software to the EPA project officer
            are:

            (a)  Documented in detail sufficient for implementation.    5

            (b)  Documented too briefly for implementation.            3

            (c)  Not documented.                                       1


A.5  DOCUMENTATION CONTROL
     (5.1)  Procedures for making revisions to technical documents
            are:

            (a)  Clearly spelled out in written form with the line
                 of authority indicated and available to all involved
                 personnel.                                            5

            (b)  Recorded but not readily available to all personnel.  3

            (c)  Left to the discretion of present supervisors/mana-
                 gers .                                                 1
                                     53

-------
A.5  DOCUMENTATION CONTROL (continued)
                                                                     SCORE

     (5.2)  In revising technical documents,  the revisions are:

            (a)  Clearly spelled out in written form and distrib-
                 uted to all parties affected,  on a controlled.basis
                 which assures that the change  will be implemented
                 and permanent.                                        5

            (b)  Communicated through memoranda to key people who
                 are responsible for effecting  the change through
                 whatever method they choose.                          3

            (c)  Communicated verbally to operating personnel who
                 then depend on experience to maintain continuity
                 of the change.                                        1
     (5.3)  Changes to technical documents pertaining to opera-
            tional activities are:

            (a)  Analyzed to make sure that any harmful side effects
                 are known and controlled prior to revision effectiv-
                 ity.                                                  5

            (b)  Installed on a trial or gradual basis,  monitoring
                 the product to see if the revision has a net bene-
                 ficial effect.                                        3

            (c)  Installed immediately with action for correcting side
                 effects taken if they show up in the final results.   1
     (5.4)  Revisions to technical documents are:

            (a)  Recorded as to date, serial number,  etc.  when the
                 revision becomes effective.                           5

            (b)  Recorded as to the date the revision was  made on
                 written specifications.                               3

            (c)  Not recorded with any degree of precision.             1
                                      54

-------
A.5  DOCUMENTATION CONTROL (continued)
                                                                     SCORE

     (5.5)   Procedures for making revisions to computer software
            programs are:

            (a)  Clearly spelled out in written form with the line
                 of authority indicated.                               5

            (b)  Not recorded but changes must be approved by the
                 present supervisor/manager.                            3

            (c)  Not recorded and left to the discretion of the
                 programmer.                                            1
     (5.6)  In revising software program documentation,  the  re-
            visions are:

            (a)  Clearly spelled out in written form,  with reasons
                 for the change and the authority for  making the
                 change distributed to all parties affected  by the
                 change.                                              5

            (b)  Incorporated by the programmer and communicated
                 through memoranda to key people.                     3

            (c)  Incorporated by the programmer at his will.           1
     (5.7)  Changes to software program documentation are:

            (a)  Analyzed to make sure that any harmful side
                 effects are known and controlled prior to
                 revision effectivity.

            (b)  Incorporated on a trial basis, monitoring  the
                 results to see if the revision has a net bene-
                 ficial effect.

            (c)  Incorporated immediately with action for detecting
                 and correcting side effects taken as necessary.
                                     55

-------
A.5  DOCUMENTATION CONTROL (continued)
                                                                     SCORE

     (5.8)  Revisions to software program documentation.are:

            (a)  Recorded as to date, program name or number,  etc.,
                 when the revision becomes effective.                  5

            (b)  Recorded as to the date the revision was made.         3

            (c)  Not recorded with any degree of precision.             1


A.6  PREVENTIVE MAINTENANCE


     (6.1)  Preventive maintenance procedures are:

            (a)  Clearly defined and written for all measurement
                 systems and support equipment.                         5

            (b)  Clearly defined and written for most of the  measure-
                 ment systems and support equipment.                   3

            (c)  Defined and written for only a small fraction of the
                 total number of systems.                              1


     (6.2)  Preventive maintenance activities are documented:

            (a)  On standard forms in station log books.               5

            (b)  Operator/analyst summary in log book.                 3

            (c)  As operator/analyst notes.                            1
     (6.3)  Preventive maintenance procedures as written appear
            adequate to insure proper equipment operation for:

            (a)  All measurement systems and support equipment.

            (b)  Most of the measurement systems and support equip-
                 ment.

            (c)  Less than half of the measurement systems and sup-
                 port equipment.
                                      56

-------
A.6  PREVENTIVE MAINTENANCE (continued)
                                                                     SCORE

     (6.4)  A review of the preventive maintenance records indicates
            that:

            (a)  Preventive maintenance procedures have been carried
                 out on schedule and completely documented.            5

            (b)  The procedures were carried out on schedule but not
                 completely documented.                                3

            (c)  The procedures were not carried out on schedule all
                 the time and not always documented.                   1
     (6.5)  Preventive maintenance records (histories) are:

            (a)  Utilized in revising maintenance schedules, de-
                 veloping an optimum parts/reagents inventory and
                 development of scheduled replacements to minimize
                 wear-out failures.                                    5

            (b)  Utilized when specific questions arise and for
                 estimating future work loads.                         3

            (c)  Utilized only when unusual problems occur.            1


A.7  DATA VALIDATION PROCEDURES


     (7.1)  Data validation procedures are:

            (a)  Clearly defined in writing for all measurement
                 systems.                                              5

            (b)  Defined in writing for some measurement systems,
                 some dependent on experience, memory, and/or
                 verbal communication.                                 3

            (c)  Only defined by experience and verbal communica-
                 tion.                                                 !
                                      57

-------
A.7  DATA VALIDATION PROCEDURES (continued)
                                                                     SCORE

     (7.2)  Data validation procedures are:

            (a)  A coordinated combination of computerized and
                 manual checks applied at different levels in the
                 measurement process.                                  5

            (b)  Applied with a degree of completeness at no more
                 than two levels of the measurement process.           3

            (c)  Applied at only one level of the measurement pro-
                 cess.                                                 1
     (7.3)  Data validation criteria are documented and include:

            (a)  Limits on:  (1)  operational parameters such as
                 flow rates; (2)  calibration data, (3)  special
                 checks unique to each measurement; e.g., succes-
                 sive values/averages; (4)  statistical tests; e.g.,
                 outliers;  (5)  manual checks such as hand calcula-
                 tions .                                                5

            (b)  Limits on the above type checks for most of the
                 measurement systems .                                  3

            (c)  Limits on some of the above type checks for only
                 the high-priority measurements.                       1
     (7.4)  Acceptable limits as set are reasonable and adequate
            to insure the detection of invalid data with a high
            probability for:

            (a)  All measurement systems.                              5

            (b)  At least 3/4 of the measurement systems.              3

            (c)  No more than 1/2 of the measurement systems.          1
                                     58

-------
A.7  DATA VALIDATION PROCEDURES (continued)
                                                                     SCORE

     (7.5)  Data validation activities are:

            (a)  Recorded on standard forms at all levels of the
                 measurement process.                                  5

            (b)  Recorded in the operator's/analyst's log book.         3

            (c)  Not recorded in any prescribed manner.                 1


     (7.6)  Examination of data validation records indicates that:

            (a)  Data validation activities have been carried out
                 as specified and completely documented.                5

            (b)  Data validation activities appear to have been
                 performed but not completely documented.               3

            (c)  Data validation activities, if performed, are not
                 formally documented.                                   1


     (7.7)  Data validation summaries are:

            (a)  Prepared at each level or critical point in the
                 measurement process and forwarded to the next level
                 with the applicable block of data.                    5

            (b)  Prepared by and retained at each level.                3

            (c)  Not prepared at each level nor communicated between
                 levels.                                               1
     (7.8)   Procedures for deleting invalidated data are:

            (a)   Clearly defined in writing for all levels of the meas-
                 urement process, and invalid data are automatically
                 deleted when one of the computerized validation cri-
                 teria is exceeded.                                   5

            (b)   Programmed for automatic deletion when computerized
                 validation criteria are exceeded but procedures not
                 defined when manual checks detect invalid data.       3

            (c)   Not defined for all levels of the measurement pro-
                 cess.                                                 1
                                     59

-------
A.7  DATA VALIDATION PROCEDURES (continued)
                                                                     SCORE

     (7.9)  Quality audits (i.e., both on-site system reviews and/or
            quantitative performance audits) independent of the normal
            operations are:

            (a)  Performed on a random but regular basis to ensure
                 and quantify data quality.                            5

            (b)  Performed whenever a suspicion arises that there
                 are areas of ineffective performance.                 3

            (c)  Never performed.                                      1


A.8  PROCUREMENT AND INVENTORY PROCEDURES


     (8.1)  Purchasing guidelines are established and documented
            for:

            (a)  All equipment and reagents having an effect on data
                 quality.                                              5

            (b)  Major items of equipment and critical reagents.       3

            (c)  A very few items of equipment and reagents.           1


     (8.2)  Performance specifications are:

            (a)  Documented for all items of equipment which have
                 an effect on data quality.                            5

            (b)  Documented for the most critical items only.          3

            (c)  Taken from the presently used items of equipment.     1


     (8.3)  Reagents and chemicals (critical items) are:

            (a)  Procured from suppliers who must submit samples
                 for test and approval prior to initial shipment.      5

            (b)  Procured from suppliers who certify they can meet
                 all applicable specifications.                        3

            (c)  Procured from suppliers on the basis of price and
                 delivery only.                                        1
                                     60

-------
A.8  PROCUREMENT AND INVENTORY PROCEDURES (continued)
                                                                     SCORE

     (8.4)  Acceptance testing for incoming equipment is:

            (a)  An established and documented inspection procedure
                 to determine if procurements meet the quality assurance
                 and acceptance requirements.  Results are document-
                 ed.                                                   5

            (b)  A series of undocumented performance tests performed
                 by the operator before using the equipment.           3

            (c)  The receiving document is signed by the responsible
                 individual indicating either acceptance or rejection.  1


     (8.5)  Reagents and chemicals are:

            (a)  Checked 100% against specification, quantity,  and
                 for certification where required and accepted
                 only if they conform to all specifications.           5

            (b)  Spot-checked for proper quantity and for shipping
                 damage.                                               3

            (c)  Released to analyst by the receiving clerk without
                 being checked as above.                               1
     (8.6)  Information on discrepant purchased materials is:

            (a)  Transmitted to the supplier with a request for
                 corrective action.                                    5

            (b)  Filed for future use.                                 3

            (c)  Not maintained.                                       1


     (8.7)  Discrepant purchased materials are:

            (a)  Submitted to a review by Quality Control and
                 Chief Chemist for disposition.                        5

            (b)  Submitted to Service Section for determination
                 on acceptability.                                     3

            (c)  Used because of scheduling requirements.              1
                                     61

-------
A.8  PROCUREMENT AND INVENTORY PROCEDURES (continued)
                                                                     SCORE

     (8.8)  Inventories are maintained on;

            (a)  First-in, first-out basis.                            5

            (b)  Random selection in stock room.                       3

            (c)  Last-in, first-out basis.                             1


     (8.9)  Receiving of materials is;

            (a)  Documented in a receiving record log, giving a
                 description of the material, the date of receipt,
                 results of acceptance test, and the signature
                 of the responsible individual.                        5

            (b)  Documented in a receiving record log with material
                 title, receipt date, and initials of the individual
                 logging the material in.                              3

            (c)  Documented by filing a signed copy of the requisi-
                 tion.                                                 1


     (8.10)  Inventories are:

            (a)  Identified as to type, age, and acceptance status.    5

            (b)  Identified as to material only.                       3

            (c)  Not identified in writing.                            1


     C8.ll)  Reagents and chemicals which have limited shelf life  are:

            (a)  Identified as to shelf life expiration  data and
                 systematically issued from stock only if they
                 are still within that date.                           5

            (b)  Issued on a first-in, first-out basis,  expecting
                 that there is enough safety factor so that the
                 expiration date is rarely exceeded.                   3

            (c)  Issued at random from stock.                          1
                                      62

-------
A.9  PERSONNEL TRAINING PROCEDURES
                                                                     SCORE

     (9.1)  Training of new employees is accomplished by:

            (a)  A programmed system of training where elements of
                  training, including quality standards,  are included
                 in a training checklist.  The employee's  work is
                 immediately rechecked by supervisors for errors or
                 defects and the information is fed back instanta-
                 neously for corrective action.                        5

            (b)  On-the-job training by the supervisor who gives
                 an overview of quality standards.  Details of
                 quality standards are learned as normal results
                 are fed back to the chemist.                          3

            (c)  On-the-job learning with training on the rudi-
                 ments of the job by senior coworkers.                 1


     (9.2)  When key personnel changes occur:

            (a)  Specialized knowledge and skills are retained in
                 the form of documented methods and descriptions.       5

            (b)  Replacement people can acquire the knowledge of
                 their predecessors from coworkers, supervisors,
                 and detailed study of the specifications  and
                 memoranda.                                            3

            (c)  Knowledge is lost and must be regained through long
                 experience or trial-and-error.                        1


     (9.3)  The people who have an impact on quality, e.g., cali-
            bration personnel, maintenance personnel, bench chemists,
            supervisors, etc., are:

            (a)  Trained in the reasons for and the benefits of
                 standards of quality and the methods by which
                 high quality can be achieved.                         5

            (b)  Told about quality only when their work falls below
                 acceptable levels.                                    3

            (c)  Are reprimanded when quality deficiencies are
                 directly traceable to their work.                     1
                                     63

-------
A.9  PERSONNEL TRAINING PROCEDURES (continued)
                                                                     SCORE

     (9.4)  The employee's history of training accomplishments
            is maintained through:

            (a)  A written record maintained and periodically
                 reviewed by the supervisor.                            5

            (b)  A written record maintained by the employee.           3

            (c)  The memory of the supervisor/employee.                 1


     (9.5)  Employee proficiency is evaluated on a continuing
            basis by:

            (a)  Periodic testing in some planned manner with the
                 results of such tests recorded.                       5

            (b)  Testing when felt necessary by the supervisor.         3

            (c)  Observation of performance by the supervisor.          1


     (9.6)  Results of employee proficiency tests are:

            (a)  Used by management to establish the need for and
                 type of special training.                             5

            (b)  Used by the employee for self-evaluation of needs.     3

            (c)  Used mostly during salary reviews.                    1


A.10  FEEDBACK AND CORRECTIVE ACTION
     (10.1) A feedback and corrective action mechanism to assure
            that problems are reported to those who can correct them
            and that a closed loop mechanism is established to assure
            that appropriate corrective actions have been taken is:

            (a)  Clearly defined in writing with individuals assigned
                 specific areas of responsibility.                     5

            (b)  Written in general terms with no assignment of
                 responsibilities.                                     3

            (c)  Not formalized but left to the present supervisors/
                 managers.                                              1
                                      64

-------
A.10  FEEDBACK AND CORRECTIVE ACTION (continued)
                                                                     SCORE

      (10.2)  Feedback and corrective action activities are:

              (a)  Documented on standard forms.                       5

              (b)  Documented in the station log book.                 3

              (c)  Documented in the operator's/analyst's notebook.    1


      (10.3)  A review of corrective action records indicates that:

              (a)  Corrective actions were systematic, timely, and
                   fully documented.                                   5

              (b)  Corrective actions were not always systematic,
                   timely, or fully documented.                        3

              (c)  A closed loop mechanism did not exist.              1


      (10.4)  Periodic summary reports on the status of corrective
              action are distributed by the responsible individual to:

              (a)  All levels of management.                           5

              (b)  One level of management only.                       3

              (c)  The group generating the report only.               1


      (10.5)  The reports include:

              (a)  A listing of major problems for the reporting
                   period; names of persons responsible for correc-
                   tive actions; criticality of problems; due dates;
                   present status; trend of quality performance (i.e.,
                   response time, etc.); listing of items still open
                   from previous reports.                              5

              (b)  Most of the above items.                            3

              (c)  Present status of problems and corrective actions.  1
                                      65

-------
A.ll  CALIBRATION PROCEDURES
                                                                     SCORES

      (11.1)  Calibration procedures are:

              (a)  Clearly defined and written out in step-by-step
                   fashion for each measurement system and support
                   device,                                             5

              (b)  Defined and summarized  for each system and device.   3

              (c)  Defined but operational procedures developed by
                   the individual.                                     1
      (11.2)  Calibration procedures as written are:

              (a)  Judged to be technically sound and consistent with
                   data quality requirements.                           5

              (b)  Technically sound but lacking in detail.             3

              (c)  Technically questionable and lacking in detail.      1


      (11.3)  Calibration standards are:

              (a)  Specified for all systems and measurement devices
                   with written procedures for assuring,  on a con-
                   tinuing basis, traceability to primary standards.    5

              (b)  Specified for all major systems with written
                   procedures for assuring traceability to pri-
                   mary standards.         -                           3

              (c)  Specified for all major systems but no procedures
                   for assuring traceability to primary standards.      1
      (11.4)   Calibration standards and traceability procedures as
              specified and written are:

              (a)   Judged to be technically sound and consistent
                   with data quality requirements.                      5

              (b)   Standards are satisfactory but traceability is
                   not verified frequently enough.                      3

              (c)   Standards are questionable.                         1
                                      66

-------
A.ll  CALIBRATION PROCEDURES (continued)
                                                                        SCORE

      (11.5)  Frequency of calibration is:

              (a)  Established and documented for each measurement
                   system and support measurement device.              5

              (b)  Established and documented for each major meas-
                   urement system.                                     3

              (c)  Established and documented for only certain
                   measurement systems.                                1


      (11.6)  A review of calibration data indicates that the
              frequency of calibration as implemented:

              (a)  Is adequate and consistent with data quality
                   requirements.                                       5

              (b)  Results in limits being exceeded a small frac-
                   tion of the time.                                   3

              (c)  Results in limits being exceeded frequently.        1


      (11.7)  A review of calibration history indicates that:

              (a)  Calibration schedules are adhered to and results
                   fully documented.                                   5

              (b)  Schedules are adhered to most of the time.          3

              (c)  Schedules are frequently not adhered to.            1

      (11.8)  A review of calibration history and data validation
              records indicates that:

              (a)  Data are always invalidated and deleted when
                   calibration criteria are exceeded.                  5

              (b)  Data are not always invalidated and/or deleted
                   when criteria are exceeded.                         3

              (c)  Data are frequently not invalidated and/or deleted
                   when criteria are exceeded.                         1
                                     67

-------
A.11  CALIBRATION PROCEDURES (continued)
                                                                     SCORE

      (11.9)   Acceptability requirements for calibration results
               are:

               (a)  Defined for each system and/or device requiring
                    calibration including elapsed time since the
                    last calibration as well as maximum allowable
                    change from the previous calibration.              5

               (b)  Defined for all major measurement systems.         3

               (c)  Defined for some major measurements systems only.  1


      (11.10)  Acceptability requirements for calibration results as
               written are:

               (a)  Adequate and consistent with data quality require-
                    ments:                                             5

               (b)  Adequate but others should be added.               3

               (c)  Inadequate to ensure data of acceptable quality.   1


      (11.11)  Calibration records (histories) are:

               (a)  Utilized in revising calibration schedules (i.e.,
                    frequency).                                        5

               (b)  Utilized when specific questions arise and re-
                    viewed periodically for trends, completeness,
                    etc.                                               3

               (c)  Utilized only when unusual problems occur.         1


A.12  FACILITIES/EQUIPMENT


      (12.1)  Facilities/Equipment are:

              (a)  Adequate to obtain acceptable results.              5

              (b)  Adequate to obtain acceptable results most of
                   the time.                                           3

              (c)  Additional facilities and space are needed.         1
                                     68

-------
A. 12  FACILITIES/EQUIPMENT (continued)
                                                                     SCORE

      (12.2)  Facilities, equipment, and materials are:

              (a)  As specified in appropriate documentation and/or
                   standards.                                          5

              (b)  Generally as specified in appropriate standards.    3

              (c)  Frequently different from specifications.           1


      (12.3)  Housekeeping reflects an orderly, neat, and
              effective attitude of attention to detail in:

              (a)  All of the facilities.                              5

              (b)  Most of the facilities.                             3

              (c)  Some of the facilities.                             1


      (12.4)  Maintenance Manuals are:

              (a)  Complete and readily accessible to maintenance
                   personnel for all systems, components, and
                   devices.                                            5

              (b)  Complete and readily accessible to maintenance
                   personnel for all major systems, components,  and
                   devices.                                            3

              (c)  Complete and accessible for only a few of the
                   sys terns.                                            1


A. 13  RELIABILITY
      (13.1)  Procedures for reliability data collection, processing,
              and reporting are:

              (a)  Clearly defined and written for all system com-
                   ponents.                                            5

              (b)  Clearly defined and written for major components
                   of the system.                                      3

              (c)  Not defined.                                        1
                                      69

-------
A.13  RELIABILITY (continued)
                                                                     SCORE

      (13.2)  Reliability data are:

              (a)  Recorded on standard forms.                         5

              (b)  Recorded as operator/analyst notes.                 3

              (c)  Not recorded.                                       1


      (13.3)  Reliability data are:

              (a)  Utilized in revising maintenance and/or replace-
                   ment schedules.                                     5

              (b)  Utilized to determine optimum parts inventory.      3

              (c)  Not utilized in any organized fashion.              1
                                      70

-------
APPENDIX B  STANDARD TECHNIQUES USED IN
            QUANTITATIVE PERFORMANCE AUDITS
                     71

-------
                                                 Table 1.   Physical measurements
           Yoperty
           Density
           Flow
ro
           Humidity
Measurement Methods
a. Vibrating U-tube
b. Mass/flow meter
c. Bubble tube
a. Orifice meter
   i.  manometer
  ii.  differential
       pressure cell
 iii.  mechanical gauges
  iv.  electrical cells

b. Pitot tube
   i.  manometer
  ii.  mechanical gauges
 iii.  electrical cells
  iv.  diffe'rential
       pressure cell
c. Venturl, meter
d. Magnetic flow meter
e. Ultrasonic flow meter
a. Wet bulb/dry bulb
   thermometers
b. Dewpoint meters
c. Electronic humidity
   cells
d. Fluidic
Calibration Methods
Take sample, get weight and
volume at process temperature
and calculate density.
Primary method is to remove
meter from process and cali-
brate on test stand.
Secondary method ia calibra-
tion of elements following
sensor.
Calculation of humidity from
wet and dry bulb measurements
and psychrometric relations.
   Audit Techniques
                                                                                                     Frequency:  Before start and at
                                                                                                       end of demonstration, monthly
                                                                                                       in between.
                                                                                                     Technique:  Use of appropriate
                                                                                                       laboratory weight and volume
                                                                                                       measures.
Frequency: Before start and at
  end of demonstration, monthly
  in between.
Remove sensor element and in-
  spect for corrosion or foul-
  ing.
Carry out manufacturer
  recommended calibration
  procedure.
For transducer and output,
  apply substitute signal and
  calibrate.
Frequency: Before start  and  at
  end of demonstration,  weekly
  for wet/dry bulb,  monthly
  for others.
Technique: Remove sensor and
  subject to air stream  having
  wet/dry apparatus  for  com-
  parison.

-------
                                      Table  1.   Physical measurements  (con.)
Property
Measurement Methods
Calibration Methods.
  Audit Techniques
Level
Pressure, differ-
ential pressure
 Temperature
a. Bubble tube
b. Float
c. Conductivity cell
d. Capacitance cell
e. Differential
   pressure cell
f. Ultrasonic
g. Sight glass
a. Mechanical gauge
b. Manometer
c. Electrical pressure
   cell
d. Differential pressure
   cell
a. Thermocouple
b. Resistance Temperature
   Detector
c. Thermistor
d. Filled bulb
e. Mercury thermometer
Measure level with sight
glass or dip stick.
Use of dead weight tester is
primary standard.  Secondary
standards are

a. Manometer with known fluid
b. 'Precision mechanical gauges
c. Standard electrical
   pressure cells
Comparison to reference point
a. Ice point H.O
b. Boiling point H.O
c. Standard thermometer
d. Electronic standard
Frequency: Before start and at
  end of demonstration, monthly
  in between.
Technique: Measure level at
  several points in range and
  compare to readout.
Frequency: Before start and at
  end of demonstration, monthly
  in between.
Technique: Pressure sensors to
  be provided with test taps
  and valves, also tap for
  secondary source for d/p
  or electrical cells.
Manometer Is preferred for
  calibration in field where
  possible

Frequency: Before start and at
  end of demonstration, monthly
  in between.
Technique: Remove sensor, insert
  in reference temperature.  For
  non-removable sensor, measure
  output of sensor," and insert
  substitute signal into instru-
  ment .input and check calibra-
  tion.

-------
                                           Table  2.  Gas effluent  streams
Material
    Measurement Method
                                                            Calibration Method
                                   Audit Techniques
Carbon Monoxide

Nitrogen Oxide
    Method 10 (Continuous)

    Method 7 (Grab)
Particulates
                            Continuous
    Method 5 (Sampling Train)
   Standard Reference
                            Optical (transtnissometer)
aterial, from the National Bure
Standard calibration gas.


Calibrate Sampling Train com-
ponents and use control samples
for analysis phase.
                                    1. Use standard calibration
                                       gases plus
                                    2. Compare results to Method 7
Calibrate components of sam-
pling train: pitot tube, dry
gas meter, orifice meter,
temperature measurement de-
vices, probe heater,, filter
holder, temperature system
                                    Filters
                                                        u of Standards.
Provide SRM* for measurement.


1. Independent duplicate
   sampling and analysis.
2. Review and observe operating
   procedures, check calibra-
   tion of train components,
   and prepare blind samples
   for field team to measure.
1. Provide NO/NO, calibration
   gas:NBS-SRM (analysis phase),
2. Compare to Method 7 (total
   measurement method).

1. Audit of total method by
   independent, simultaneous
   measurement from sampling
   through analysis.
2, Calibration check on sample
   train components per cent
   isokinetic rate, and visual
   observation of operating
   procedures.
                                   Calibration check with  inde-
                                   pendent  set of NBS  filters.

-------
                                             Table 2.   Gas effluent streams (con.)
        Material
Measurement Method
Calibration Method
Audit Techniques
        Sulfur Dioxide
Method 6 (Batch)
Calibrate sampling train com-
ponents and use standards
samples for analysis phase.
Ui
                                    Method 12 (Continuous)
                               1. Use  standard calibration
                                 gases plus
                               2. Use  calibrated absorbance
                                 filter  furnished by instru-
                                 ment manufacturer.
   Independent duplicate
   sampling and analysis.
   Review/observe operating
   procedures, check calibra-
   tion of train components,
   and analyze split samples
   and/or prepare blind sam-
   ples for field team to
   measure.
                                   1. Compare to Method 6.
                                   2. Provide SRS for measure-
                                      ments.

-------
                                 Table 3.   Liquid streams,  suspended solids
Material
                    Measurement Method
Calibration Method
                                                                   Audit Techniques
Liquid Stream Samples-
par cent solids, ionic
species
Effluent solids; e.g.,
per cent water, CaO,

-------
APPENDIX C  DEFINITIONS AND STATISTICAL TECHNIQUES
            USEFUL IN QUALITY ASSURANCE PROGRAMS
                         77

-------
I.  CENTRAL TENDENCY AND DISPERSION
    A.  The Arithmetic Mean.
        The sum of all values in a measurement set,  divided by the number
    of values summed.  Commonly called the "average."  Often denoted
    symbolically by a bar over the variable symbol,  as "X".
                               n
                           X
£ xi A
i=l    '
    B.  Range.
        The difference between the maximum and minimum values of a set of
    values.
                              R = X    - X .
                                   max    mm
    A rough indication of variability, particularly when the set of values
    is small (<10).
    C.  Standard Deviation.
        An indication of the dispersion of a  set of numbers about the
    mean value.  Normal (and other) distributions are expressed as a func-
    tion of the standard deviation.
         For a given set of values, the defining equation is:
                          n
                                         1/2
                              n-1

    For computational purposes,  it is convenient to use:
                                           1/2
                     s =  / ^ _  2       2
                                    -  X
                            n
                                    78

-------
     D.   Relative Standard Deviation, or Coefficient of Variation.
         The dispersion of a set of values, expressed as a percentage of
     the mean.

                          CV = (s/X) x 100

II.  MEASURES OF VARIABILITY
     A.   Accuracy.
         The difference (either on an absolute or percentage basis)  between
     a measured value and an assumed "true" value.  The larger the difference,
     the lower the accuracy.

                         B = X - T,  or
                        *,T> _ (X-T) x IQO
                        %B	f

                        (see "Bias")
     B.   Bias.
         A nonrandom measurement error; a consistent difference either
     between sets of results or between a measured value and a "true"
     value.  If the latter, the bias or percent bias is measured by  the
     relationships in A above.  (See III. SIGNIFICANCE TESTS, A.  t-test)
     C.   Precision.
         A measure of agreement among individual measurements of a vari-
     able, under identical or similar conditions.  Precision may be  ex-
     pressed in several ways, and care must be exercised in the defini-
     tion and use of precision measures.
     One set of such measures  follows:
          1.  Within-laboratory:  The within-laboratory standard deviation,
              s,  measures the dispersion in replicate single determina-
              tions made by one laboratory team (same field operators,
     * These definitions  are taken from EPA collaborative test result
      publications,  and  are applied to the various federal reference
      sampling and analysis techniques.   Since these techniques are
      frequently used  in evaluating emissions from IERL projects,  they
      are particularly appropriate for this guidelines document.
                                    79

-------
         laboratory analyst, and equipment) sampling the same true
         concentration.
     2.  Between-laboratory: The between-laboratory standard devia-
         tion, s, , measures the total variability in a concentration
         determination due to determinations by different labora-
         tories sampling the same true stack concentration.  The
                                       2
         between laboratory variance, s, , may be expressed as
                             2    2,2
                            S,  = S,. + S
                             b    L

         and consists of a within-laboratory variance plus a labora-
                              2
         tory bias variance, s .
                              Jj
     3.  Laboratory bias; The laboratory bias standard deviation,
         s  = /si: - sz , is that portion of the total variability
          J_j
         that can be ascribed to differences in the field operators,
         analysts and instrumentation, and due to different manners
         of performance of procedural details left unspecified in
         a technique.  This term measures that part of the total
         variability in a determination which results from the use
         of a technique by different laboratories, as well as from
         modifications in usage by a single laboratory over a period
         of time.  The laboratory bias standard deviation is esti-
         mated from the within-and between-laboratory estimates pre-
         viously obtained.
     A corresponding set of coefficients of variation would be CV,
CV, , and CV  .  These are convenient to use if the precision is pro-
  o        L
portional to the mean value of the variable.
                                80

-------
III.  SIGNIFICANCE TESTS
      A.   t-test.
           If one has an assumed "true" value, however obtained, the
      existence of a significant bias in other measurements of this value
      can be defined by a t-test:
                              d - 0
      where t = a parameter, the magnitude of which is referenced to
                tabulated values.  A t-value which exceeds the tabulated
                value for given specifications of probability and number
                of degrees of freedom indicates the existence (within the
                definition of probability specified) of a significant
                bias.  The more stringent the probability requirement;
                i.e., the smaller the probability chosen, the larger the
                tabulated t-value.

            d = the average difference between the true value and the
                measured values; the average bias.

           s, = the standard deviation of the differences, d .
            d                                               i

            n = the number of measurements made.
      B.  Chi-square test.
          If one has a reasonable estimate of the expected standard devia-
      tion of a set of measurements, the existence of a defined "excess
      variability" can be tested as follows:
                                 2         2
                                X        sd
                                       02{x>
                                     81

-------
             2
      where x /f = a random variable with tabulated values ( f = n - 1 =
                   number of degrees of freedom).
            2
           0 {x} = the expected variance of the measurements of x.

               2
           If X /f is larger than the chosen tabulated value (with specified
      probability), it is concluded that the measurements are exhibiting
      excess variability.  The chi-square test is a measure of the validity
      of a series of measurements based on an "expected" variability.  The
      test is worthwhile only whenever a measurement technique has been
      tested thoroughly, so that a realistic expectation can be estimated.

 IV.  CONFIDENCE LIMITS
     Confidence limits take two forms.  One form defines a numerical range
within which one has a (arbitrarily chosen) probability of finding the true
mean value of the measured variable.  If the measurement variability is ex-
pressed as a standard deviation, the confidence limits as defined above can
be calculated as follows:

                              CL = X + ts/ /n~
where all symbols have been previously defined.  Note that as the number of
measurements, n, increases, the magnitude of CL decreases.  Also, for higher
probabilities of containing the true mean within CL, the larger the value of
t and therefore the larger the size of CL.
     The second form of confidence limit defines an interval within which the
next individual measurement can be expected to fall with a given probability.
The calculation of this limit, sometimes called a tolerance limit, is by
the following relationship:

                                TL = X + ts
                                      82

-------
While n, the number of measurements, does not explicitly appear in the
equation for TL, it does determine  (along with the selected probability)
the value of t; i.e., as n increases, t decreases.

V.  TESTING FOR OUTLIERS
    An outlier is an extreme value, either high or low, which has question-
able validity as a member of the measurement set with which it is associ-
ated.
     Detection of outliers may be on one of the following basis:
     a)  A known experimental aberration, such as an instrument failure or
a technique inconsistency.
     b)  A statistical test for significance, such as the Dixon ratio test.
This test is described below.
         The Dixon criteria is based entirely on ratios of differences be-
tween observations where it is desirable to avoid calculation of s or
where quick judgment is called for.  For the Dixon test, the sample cri-
terion or statistic changes with sample size.  Critical values of the
statistic for various levels of significance are tabulated.
                          r
            &
     Table 1 below, presents selected significance (probability) levels
for criteria over the n range 3 to  25.  Note that the measurement values
are first, arranged in order of ascending magnitude; i.e., x  is the largest
value.
  Taken from "Processing Data  for  Outliers,"  by W. J. Dixon, Biometrics.
  Vol. 9, No. 1, 1953.
                                    83

-------
Table 1.  DIXON CRITERIA FOR TESTING OF EXTREME OBSERVATION
                      (SINGLE SAMPLE)3
n
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Criterion
HO = (x2 ~ x-|)/(xn - x-j) if smallest
value is suspected;
= (xn * xn-l)/(xn ' xl} 1f
largest value is suspected.
rll = (X2 ' xl)/(xn-l " xl) 1f
smallest value is suspected;
= (xn - xn.1)/(xn - x2) if
largest value is suspected.
r— (V — YI/IY — Y^1"F
OT \O " "1 / / \ "« 1 i /
smallest value is suspected;
= (xn- xn_2)/(xn - x2) if
largest value is suspected.
r~IY » Y 1 / I Y — Yl"1"f
™~ \"O "1 / / V « O ™ T»
c£. 3 1 n-c. \
smallest value is suspected;
= (xn - xn_2)/(xn - x3) if
largest value is suspected.







Significance Level
10%
•• 0.886
0.679
0.557
0.482
0.434
0.479
0.441
0.409
0.517
0.490
0.467
0.492
0.472
0.454
0.438
0.424
0.412
0.401
0.391
0.382
0.374
0.367
0.360
5%
0.94-1
0.765
0.642
0.560
0.507
0.554
0.512
0.477
0.576
0.546
0.521
0.546
0.525
0.507
0.490
0.475
0.462
0.450
0.440
0.430
0.421
0.413
0.406
1%
0.988
0.889
0.780
0.698
0.637
0.683
0.635
0.597
0.679
0.642
0.615
0.641
0.616
0.595
0.577
0.561
0.547
0.535
0.524
0.514
0.505
0.497
0.489
                             84

-------
APPENDIX D  SOME STANDARD AMBIENT AIR AND SOURCE
            SAMPLING TECHNIQUES
                        85

-------
                                          SOURCE SAMPLING TECHNIQUES
Pollutant EPA
Method
or
Number
S02 6
so2
and 8
S03/H2S04
NO 7
* *
CO 10
Particulates 5
Visible 9
emissions
Be 104
a TM<; table is a summary
Bias (absolute,
or percent of
mean concentra-
tion)
0
-2% (analysis
only)
-2% (analysis
only)
0
+7 ppm
No information
+1 .4% opacity
-20%, average
of information contained
Precision (absolute, or
coefficient of variation)
within
1 aboratory
3.9
0.1 g/m3
60%
7%
13 ppm
10-30%
2% opacity
44%
in the cited
between
laboratory
5.5
0.11 g/m3
65%
10%
25 ppm
20-40%
2.5%
58%
references, all
Comments
Major error source is dif-
ficulty of obtaining repro-
ducible titration end-
points. Minimum detect-
able limit is 3 ppm.
Same analysis technique
as Method 6 above.
Grab sample; largest error
source is failure to re-
calibrate spectrophoto-
meter.
Analyzer drift and COp
interference are largest
problems. Minimum detect-
able limit is 20 ppm.
Numerous small error
sources associated with
stack sampling.
Good results depend to a
great extent on the effec-
tive training of observers.

of which are Quality Assurance
Reference
EPA-650/
14-74-
005-e
EPA-650/
14-74-
005-g
EPA-650/
14-74-
005-f
EPA-650/
14-74-
005-h
EPA-650/
14-74-
005-d
EPA-650/
14-74-
005- i
EPA-650/
14-74-
005 -k
Guide-
lines Manuals.

-------
                                               AMBIENT AIR TECHNIQUES'
Pollutant EPA Bias (absolute,
Method or percent of
or mean concentra-
Number tion)
S02 6 0
NO, N0_2 Chemilumi- 0
HJQ nescent
A
Precision (absolute, or
coefficient of variation)
within
laboratory
5-13 pg/m3,
from
x = Q-1000
ug/m3
7-8% at 3
100 ug/m
(0.05 ppm)
between
laboratory
10-25 pg/m3
from x = -
0-1000 pg/nT

Comments
Lower limit of detec-
tion is 25 pg/m . Flow
rate changes, sampling
train leakage are prim-
ary error sources.
Lower limit of detec-
tion is 10 pg/m3 (0.005
ppm). Errors are asso-
Reference
EPA-R4-
73-028d
d
00
    Photochem
    ical oxi-
    dants
Chemilumi-
nescent
        CO
  NDIR
-35 to -15%
from 0.05 to
0.50 ppm
    +2.5
0.0033 +
0.0255 x
(0-0.5 ppm)
0.6 mg/nr
0.0008 +
0.0355 x
(0-0.15 ppm)
-0.0051 +
0.0690 x
(0.15-0.5 ppm)

0.8 - 1.6
mg/nr (non-
linear varia-
tion) over
0-60 mg/m3
ciated with calibration
and instrument drift
(from zero and span
settings).

Lower detection limit is
0.0065 ppm
EPA-R4-
73-028C
Lower detection limit is   EPA-R4-
0.3 mg/m3. Interference    73-028a
of water vapor is signifi-
cant.
    a  This table  is a summary of  information contained in the cited references, all  of which are Quality Assurance
       Guideline Manuals published by EPA.  Collaborative test results are cited, if available,  in the manuals.
    b  x = pollutant concentration.
    c  EPA-650/4-75/016.
    d  Guidelines  for Development  of a Quality Assurance Program for The Continuous Measurement  of Nitrogen Dioxide in
       the Ambient Air (Chemiluminescent), Smith & Nelson, Research Triangle Institute, Research Triangle Park,  N.C.  27709

-------
                                            AMBIENT AIR TECHNIQUES  (cont.)
Pollutant EPA
Method
or
Number
Particulates High
Volume



N0? Arsenite


Bias (absolute,
or percent of
mean concentra-
tion)
No information




-3% (50-300
yg/m3)

Precision (absolute, or
coefficient
within
laboratory
3%




8 yg/m3
(50-300 yg
m3)
of variation
between
laboratory
3.7%




11 ug/m3
(50-300 yg/
m3)

Comments
Minimum detectable limit
is 3 mg. Shorter samp-
ling periods give less
precise results, biased
high.
A tentative method.
Lower detectable limit
is 9 yg/m3.

Reference
EPA-R4-
73-028b



EPA-R4-
73-2800

oo
oo

-------
                                TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-600/2-76-159
                                  3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
IERL-RTP Data Quality Manual
                                  5. REPORT DATE
                                   June 1976
                                                      6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

Franklin Smith and James Buchanan
                                  8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
                                  10. PROGRAM ELEMENT NO.
                                  EHB-524
                                  11. CONTRACT/GRANT NO.
                                  68-02-1398, Task 35
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC 27711
                                  13. TYPE OF REPORT AND PERIOD COVERED
                                  Task Final;  3-5/76	
                                  14. SPONSORING AGENCY CODE
                                   EPA-ORD
 15. SUPPLEMENTARY NOTES ipro].ect officer for this manual is L.D. Johnson, Mail Drop 62,
 (919) 549-8411, Ext 2557.
 is. ABSTRACT
               manuai gives guidelines for the establishment and maintenance of an
 integrated data quality program for EPA's Industrial Environmental Research Labora-
 tory—Research Triangle Park (TERL-RTP). Administrative  systems dedicated to
 the data quality program are delineated.  These systems include quality policies and
 objectives, organizational structure and key quality personnel, and a schedule for
 implementation.   Components of both quality control programs and quality assurance
 programs  are given. IERL-RTP projects are divided into six categories.  Projects
 within a given category have common characteristics (e.g. , size, duration, objec-
 tives, and data quality requirements), making them amenable to  the same general
 set of quality control and quality assurance practices and procedures.  Quality control
 and quality assurance procedures applicable to each of the six categories are given
 for each phase of the project's life cycle.  These phases include: Request for
 Proposal preparation, proposal evaluation, work plan evaluation, project implemen-
 tation and  maintenance , and reports .
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                                               .  COSATI Field/Group
 Pollution
 Data
 Quality
 Quality Control
 Quality Assurance
 Management Systems
Evaluation
Proposals
Reporting
Pollution Control
Stationary Sources
IERL-RTP
Work Plans
13B


13H,14D

05A
18. DISTRIBUTION STATEMENT

 Unlimited
                      19. SECURITY CLASS (This Report)
                       Unclassified
                                                                   21. NO. OF PAGES
                             95
                      20. SECURITY CLASS (Thispage)
                       Unclassified
                                               22. PRICE
EPA Form 2220-1 (9-73)
                    89

-------