vvEPA
United States      Office of Air Quality      EMB Report No. 86-MIN-02
Environmental Protection Planning and Standards     April 1987
Agency        Research Triangle Park NC 27711
Air

Municipal Waste Combustion
Multipollutant Study

Emission Test Report

Signal Environmental Systems, Inc.
North Andover RESCO
North Andover, Massachusetts

Volume I: Summary of Results

-------
DCN No. 87-222-124-06-02                               EMB Report No.  86-MIN-2
                             EMISSION TEST REPORT
                   PCDD/PCDF, METALS AND PARTICULATE TESTING
                      SIGNAL ENVIRONMENTAL SYSTEMS,  INC.
                              NORTH ANDOVER RESCO
                         NORTH ANDOVER, MASSACHUSETTS
                         VOLUME I: SUMMARY OF RESULTS
                            ESED Project No. 86/19
                          EPA Contract No. 68-02-4338
                           Work Assignments 2 and 6
                                 Prepared for:

                         Clyde E. Riley, Task Manager
                         Emissions Measurement Branch
                  Emission Standards and Engineering Division
                     U.S. Environmental Protection Agency
                 Research Triangle Park, North Carolina  27711
                                 Prepared by:

                             Carol L. Jamgochian
                              William P. Gergen
                              J. William Mayhew
                             Radian Corporation
                            Post Office Box 13000
                      Research Triangle Park, NC 27709
                                  April 1987

-------
                   DISCLAIMER

     This report has been reviewed by the
Emission Standards and Engineering Division of
the Office of Air Quality Planning and
Standards, EPA, and approved for publication.
Mention of trade names or commercial products is
not intended to constitute endorsement or
recommendation for use.  Copies of this report
are available through the Library Services
Office (MD-35), U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711, or
from the National Technical Information
Services, 5285 Port Royal Road, Springfield, VA
22161.

-------
                          RADIAN REPORT CERTIFICATION
     This report has been reviewed by the following Radian personnel and is a
true representation of the results obtained from the sampling program at
Signal Environmental Systems, Inc., North Andover RESCO,  North Andover,
Massachussetts.   The sampling and analytical methods were performed in
accordance with procedures outlined in the Revised Sampling and Analytical
Plan for Method Development and Testing for Municipal  Waste Combustion
Incinerators at the North Andover facility dated July 2,  1986. The sampling
and analytical plan was reviewed and accepted by the EPA/EMB Task Manager,
Clyde E. Riley.
     The subcontracting laboratories, N.C. State Nuclear Services and
Triangle Laboratories, Inc., have included their report certifications in
their respective laboratory reports, which are contained in the appendices.
                                   APPROVAL
Project Director:
                                                  Date:
                         Winton E. Kelly
Program Manager:
                                                  Date :
                         Robert M. Dykes
                                                              [/
QA Officer:
                                                  Date:
                         Donn
                                  Holder
                                                                  ?

-------
                               TABLE OF CONTENTS
Section                                                                   Page

  1.0  INTRODUCTION	1-1
       1.1  Background	1-2
       1.2  Objectives	1-3
       1.3  Brief Process Operation and Description	1-3
       1.4  Emission Measurement Program 	   1-5
            1.4.1  Test Matrix	1-5
            1.4.2  Laboratory Analysis 	   1-9
       1.5  Quality Assurance/Quality Control (QC/QC)	1-9
       1.6  Description of Report Sections 	   1-9

  2.0  SUMMARY OF RESULTS	2-1
       2.1  PCDD/PCDF Emissions Results	2-1
            2.1.1  2378-TCDD Toxic Equivalency 	   2-2
            2.1.2  Isomer Distributions	2-7
       2.2  Particulate Matter Results 	   2-7
       2.3  Metals Emissions Results for North Andover RESCO 	   2-14
            2.3.1  Flue Gas Metals Results	2-14
            2.3.2  ESP Ash Metals Results	2-16
       2.4  CEM Results	2-16

  3.0  PROCESS DESCRIPTION AND OPERATION 	   3-1
       3.1  Process Description	3-1
       3.2  Air Pollution Control System	3-5
       3.3  Incinerator and ESP Operating Conditions During Testing. . .   3-5

  4.0  SAMPLING LOCATIONS	4-1
       4.1  ESP Inlet Sampling Location	4-1
       4.2  ESP Outlet Sampling Location 	   4-1
       4.3  ESP Ash Sampling Location	4-5

  5.0  SAMPLING AND ANALYTICAL PROCEDURES	5-1
       5.1  PCDD/PCDF Sampling and Analysis   	   5-1
            5.1.1  Equipment and Sampling Preparation	5-1
            5.1.2  Sampling Operations 	   5-3
            5.1.3  Sample Recovery	5-6
            5.1.4  Analysis	5-6
            5.1.5  Data Reduction for PCDD/PCDF Results	5-10
       5.2  Flue Gas Trace Metals/Particulate Determination	5-11
            5.2.1  Equipment and Sampling Preparation	5-11
            5.2.2  Sampling Operations 	  5-11
            5.2.3  Sample Recovery	5-13
            5.2.4  Analysis	5-13
            5.2.5  Data Reduction for Metals/Particulate Results  ....  5-17

-------
                               TABLE OF CONTENTS
                                   (cont'd.)
Section
  5.0  SAMPLING AND ANALYTICAL PROCEDURES (continued)
       5.3  CEM Sampling and Analysis	5-17
            5.3.1  Equipment and Sampling Preparation	5-17
            5.3.2  Sampling Operation	5-18
            5.3.3  Data Reduction	5-18
       5.4  Molecular Weight Determination by EPA Method 3	5-20
            5.4.1  Sampling Operations 	  5-20
            5.4.2  Analysis	5-20
       5.5  Volumetric Flowrate Determination by EPA Method 2	5-20
            5.5.1  Sampling and Equipment Preparation	5-22
            5.5.2  Sampling Operations 	  5-22
       5.6  Moisture Determination by EPA Method 4	5-22
       5.7  Ash Sampling for Metals Analysis	5-22
            5.7.1  Sampling Operations 	  5-23
            5.7.2  Analysis	5-23

  6.0  QUALITY ASSURANCE/QUALITY CONTROL (QA/QC) 	  6-1
       6.1  Standard Quality Assurance Procedures	6-1
            6.1.1  Sampling Equipment Preparation	6-1
            6.1.2  General Sampling QC Procedures.  .  .  .  '.	6-3
            6.1.3  Sample Recovery	6-4
            6.1.4  Preparation of Samples for Analysis 	  6-5
       6.2  Method-Specific Sampling QC Procedures 	  6-6
            6.2.1  Procedures for Velocity/Volumetric Flowrate
                   Determination 	  6-6
            6.2.2  Quality Control Procedures for Molecular Weight
                   Determinations	6-6
            6.2.3  Quality Control Procedures for Moisture Determination  6-7
            6.2.4  Quality Control for PCDD/PCDF Testing 	  6-7
            6.2.5  Quality Control for Particulate Testing 	  6-18
            6.2.6  Quality Control for Metals/Particulate Testing. .  .   .  6-22
            6.2.7  Quality Control for Continuous Emissions Monitors .   .  6-27

References
VOLUME II

APPENDIX A -
         A.I
         A.2
         A.3
         A.4
         A.5
SUMMARY OF PARTICULATE, DIOXIN,  TOCL, AND TRACE METALS TEST DATA
  Dioxin Test Data Summaries - ESP Inlet
  TOCL/Particulate Test Data Summaries - ESP Outlet
  Metals/Particulate Test Data Summaries - ESP Inlet
  Metals/Particulate Test Data Summaries - ESP Outlet
  Method 5 Data Reduction Equations
  Dioxin/Furan Data Reduction Equations
  Trace Metals Data Reduction Equations
  CEM Data Reduction Equations

-------
Section

APPENDIX B
         B.
         B.
         B.
VOLUME III
         B.4
         B.5
         B.6

APPENDIX C -
         C.I
         C.2
         C.3
         C.4
         C.5
         C.6
         C.7
         C.8
APPENDIX D
         D.
         D.2
         D.3
         D.4
                               TABLE OF CONTENTS
                                   (cont'd.)
SUMMARY OF CEM TEST DATA
  GEM Test Data Summary
  Printout Data: One Minute Averages
  CEM Calibration Summary
  CEM Calibration Computer Printouts
  CEM Data Reduction Equations
  CEM Data Stripcharts
  CEM Daily Log

FIELD DATA SHEETS
  MM5 Dioxin Field Data Sheets - ESP Inlet
  M5 TOCL/Particulate Field Data Sheets - ESP Outlet
  M5 Metals/Particulate Field Data Sheets - ESP Inlet
  M5 Metals/Particulate Field Data Sheets - ESP Outlet
  K-Factor/Nomograph Field Data Sheets
  Preliminary Sampling and Velocity Traverse Sheets
  Orsat Analysis Field Data Sheets
  ESP Ash Field Data Sheets
LABORATORY ANALYTICAL RESULTS
  Dioxin Analytical Data
  Dioxin Data Summaries
  Triangle Labs Analysis Reports
  Metals Analytical Data
  Metals Data Summaries
  N.C. State Metals Analysis Reports
  Particulate Analytical Data
  Particulate Data Summaries
  Radian Analysis Data Sheets
  Laboratory Operations Logs
  Dioxin - Triangle Labs
  Particulate - Radian Corp.
APPENDIX E - CALIBRATION DATA
         E.I   Summary of Equipment Used During Test Program
         E.2   Equipment Calibration Sheets

APPENDIX F - PERTINENT CORRESPONDENCE
         F.I   Dioxin Analysis
         F.2   Metals Analysis
         F.3   Correspondence with Signal Environmental Systems, Inc.
                                       iii

-------
                               TABLE OF CONTENTS
                                   (cont'd.)
Section
APPENDIX G - INCINERATOR AND ESP OPERATING DATA
         G.I   Process Data Logsheets
         G.2   Process Feed Sheets

APPENDIX H - FIELD TEST LOGS
         H.I   Daily Sampling Log Summary
         H.2   Chain of Custody Forms
         H.3   Crew Chief Field Operations Log
               Field Laboratory Log
               Sample Identification Log
APPENDIX
H.4
H.5

I -
I.I
1.2
1.3
1.4
1.5
APPENDIX J -
         J.I
         J.2
         J.3
         J.4
         J.5
COMPLETE METALS RESULTS
  Flue Gas Metals Results
  Metals Analyte-to-Particulate Ratio
  Collection Efficiency of Impingers
  ESP Ash Metals Results
  QA/QC Results for NAA Analysis

QUALITY ASSURANCE INFORMATION
  Dioxin Quality Control Results
  Dioxin XAD and Filter Preparation Information
  Metals Quality Control Results
  Particulate Quality Control Results
  GEM Standard Gas Certification Sheets
APPENDIX K - SAMPLING AND ANALYTICAL PROTOCOL
         K.I   Summary of EPA Reference Methods Used During Test Program
         K.2   Modified Method 5 Dioxin Sampling and Analytical Procedures
         K.3   Ash Sampling and Analytical Procedure

APPENDIX L - PROJECT PARTICIPANTS
                                      iv

-------
'                                LIST OF FIGURES


Figures                                                                .   Page

  1-1   North Andover RESCO process line with sampling locations ....   1-4

  2-1   Uncontrolled PCDD/PCDF isomer distribution 	   2-9

  2-2   Oxygen concentration history - Runs 7,  8,  and 9 at
          North Andover RESCO	2-19

  2-3   Carbon dioxide concentration history -  Runs 7,  8,  and 9 at
          North Andover RESCO	2-20

  2-4   Carbon monoxide concentration history -  Runs 7, 8, and 9 at
          North Andover RESCO	2-21

  3-1   North Andover RESCO process line 	   3-2

  4-1   North Andover RESCO process line with sampling locations ....   4-2

  4-2   ESP inlet sampling location	4-3

  4-3   Traverse matrix for ESP inlet sampling location	4-4

  4-4   ESP outlet sampling location at North Andover Facility 	   4-6

  4-5   Traverse matrix for ESP outlet sampling location  	   4-7

  4-6   ESP ash handling system and sampling location	4-8

  5-1   PCDD/PCDF sampling train configuration used at
          North Andover RESCO	5-4

  5-2   PCDD/PCDF field recovery scheme	5-7

  5-3   PCDD/PCDF analytical scheme used at North Andover RESCO	5-9

  5-4   Metals/particulate sampling train configuration used at
          North Andover RESCO.  . :	5-12

  5-5   Sampling and analysis protocol for metals/particulate samples. .   5-14

  5-6   GEM analysis scheme	5-19

  5-7   Method 3 integrated bag sampling train  	   5-21

  5-8   ESP ash sampling and analysis scheme	5-24

  6-1   Validation of fixed gas analysis	6-33

-------
                                LIST OF TABLES


Table                                                                     Page

 1-1   Original Test Matrix for North Andover RESCO	1-6

 1-2   Summary of Sampling Log for EPA Testing at North
         Andover RESCO, July 8-16, 1986	1-8

 1-3   PCDD/PCDF Congeners Analyzed for North Andover Test
         Program	1-10

 1-4   Metals Detected by Neutron Activation in Flue Gas Sample Matrix .   1-11

 2-1   Summary of Uncontrolled PCDD/PCDF Emissions for North
         Andover RESCO 	   2-3

 2-2   Uncontrolled PCDD/PCDF Emissions at North Andover RESCO
         (Normalized to 12% CO.)	2-4

 2-3   Uncontrolled PCDD/PCDF Emissions for North Andover RESCO
         (Front, Back and Total Fraction Results)	2-5

 2-4   Uncontrolled PCDD/PCDF Concentrations Expressed as 2378-TCDD
         Toxic Equivalents	2-6

 2-5   Key to Isomer Coding for Figures 2-1	2-8

 2-6   Uncontrolled PCDD/PCDF Isomer Distribution at North
         Andover RESCO 	   2-10

 2-7   Summary of Uncontrolled and Controlled Particulate Emissions
         for North Andover RESCO  	   2-11

 2-8   Summary of Controlled Particulate Emissions for North
         Andover RESCO 	   2-13

 2-9   Summary of EPA Specific Metals Emissions for North Andover RESCO.   2-15

 2-10  Summary of ESP Ash Metals Results	2-17

 2-11  Summary of GEM Results	2-18

 3-1   North Andover Facility Structural Design Data  	   3-3

 3-2   North Andover Facility Airflow Design Data	3-4

 3-3   Average Process Data for North Andover Incinerator Tests  -
         July 9 through 16, 1986	3-6
                                      VI1

-------
                                LIST OF TABLES
                                   (cont'd.)

Table                                                                    Page

 3-4   Average GEM Data for North Andover Tests -
         July 9 through 16, 1986	3-8

 5-1   Sampling Methods and Analytical Procedures	5-2

 5-2   PCDD/PCDF Sampling Train Components for North Andover RESCO
         Shipped to Analytical Laboratory	5-8

 5-3   Sample Components for Particulate/Metals Train	5-15

 6-1   Summary of Equipment Used in Performing Source Sampling 	  6-2

 6-2   Dioxin Isokinetics and Leak Check Summary -  ESP Inlet,  North
         Andover RESCO 	  6-9

 6-3   Internal Standards Recovery Results for North Andover PCDD/PCDF
         Analyses	6-11

 6-4   Factors Used to Adjust Responses for Extraction Efficiency
         and Variable Instrument Performance 	  6-13

 6-5   Surrogate Recoveries for PCDD/PCDF Analyses  for North Andover .  .  6-14

 6-6   Analytical Results for North Andover Quality Control Samples.  .  .6-16

 6-7   Field Blank Dioxin/Furan Data for MM5 Samples	6-17

 6-8   TOCL/Particulate Isokinetics and Leak Check  Summary - ESP Outlet,
         North Andover RESCO	6-20

 6-9   Summary of Particulate QC Results by Sample  Fraction,
         North Andover RESCO, North Andover,  MA	  6-21

 6-10  Metals/Particulate Isokinetics and Leak Check Summary -
         ESP Inlet, North Andover RESCO	6-23

 6-11  Metals/Particulate Isokinetics and Leak Check Summary -
         ESP Outlet, North Andover RESCO 	  6-24

 6-12  Summary of Labproof Blanks and Field Blanks  for Metals Samples
         Collected at North Andover RESCO.  	  6-26

 6-13  Comparison of Minimum Run Values to Field Blank Values for
         North Andover RESCO	6-28
                                      viii

-------
                                LIST OF TABLES
                                   (cont'd.)

Table                                                                     Page

 6-14  Summary of Analytical Calibration Requirements	6-30

 6-15  Summary of Drift Check Results	6-32
                                        ix

-------
                               1.0  INTRODUCTION
     The results of an EPA emission test on a munipical solid waste
incinerator are presented in this report.   The Environmental Protection Agency
is currently developing a comprehensive source category document for the
municipal waste combustor (MWC) source  category.   The Office of Solid Waste,
(OSW),  the Office of Air Quality Planning and Standards (OAQPS) and the Office
of Research and Development (ORD) are participating in a joint effort to
assess the potential environmental impact of municipal solid waste (MSW)-fired
resource recovery facilities and to identify any emissions for which
additional regulations may be considered.

     The Emission Standards and Engineering Division (ESED) of OAQPS, through
its Industrial Studies Branch (ISB) and Emissions Measurement Branch (EMB),  is
responsible for reviewing the existing air emissions data base and gathering
additional data where necessary.  Several MSW emission tests are being
performed for this test program.  The results of one of these tests are the
subject of this report.  The data base supplemented by these test results will
then be used to estimate emission factors and to evaluate the various
emissions reduction alternatives that are available for MWC facilities.

     The emissions that are being studied in this assessment are the criteria
pollutants--particulate matter, sulfur oxides, nitrogen oxides, carbon
monoxide and hydrocarbons; other acid gases, such as HC1; chlorinated organics
including polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated
dibenzofurans (PCDF); and, specific metals including arsenic, cadmium,
chromium, mercury, nickel, lead and beryllium.

     The emissions data presented in this report were collected during  a joint
test program sponsored by EPA and Signal Environmental Systems, Inc.  This
report presents the results of the EPA-sponsored testing at the North Andover,
MA, facility.  The results of the test program sponsored by Signal
Environmental Systems, Inc., have been reported separately.   In addition to
                                        1-1

-------
this report, a summary report is being prepared by Radian Corporation that
will present the complete set of data that was collected during the joint
sampling program.  The summary report will discuss and analyze the
uncontrolled and controlled data as well as analyses of both the ESP ash and
bottom ash.  Also, a more detailed analysis of the continuous monitoring,
incinerator and ESP operating data will be presented.

     It should be noted that the trace metal data presented in this document
are the results of an in-house EPA development study.  During this study
non-established EPA reference test procedures were used to collect and analyze
the metals samples.   These trace metals data are presented for the convenience
of EPA and are not intended to represent the true trace metals emissions being
emitted from the facility.

1.1  BACKGROUND

     Signal Environmental Systems,  Inc. was required by the Massachusetts
Department of Environmental Quality Engineering (MDEQE) to conduct a program
to measure the PCDD/PCDF emissions in the flue gas and the PCDD/PCDF
concentration in the process ash streams at the North Andover RESCO municipal
solid waste resource recovery facility in North Andover, Massachusetts.
Radian Corporation was retained by Signal Environmental to conduct that
program.

     In order to provide additional data to evaluate the PCDD/PCDF and metals
removal effectiveness of emissions reduction systems, Signal Environmental and
EPA agreed to jointly sponsor an expanded program during the MDEQE-required
tests.   Signal Environmental sponsored PCDD/PCDF and total organic chlorine
(TOCL)  tests at the ESP outlet, and EPA sponsored both PCDD/PCDF tests at the
ESP inlet and metals and particulate testing at the ESP inlet and outlet
locations.  Ash sampling was sponsored by Signal during the PCDD/PCDF tests
and by EPA during the metals test runs.  Radian Corporation, under contract to
the EMB performed the EPA-sponsored tests.
                                         1-2

-------
1.2  OBJECTIVES


     The objective of the EPA-sponsored test program was to obtain PCDD/PCDF,

metals, and particulate data from a state-of-the-art MWC controlled by an

electrostatic precipitator (ESP).   The North Andover facility was selected by

EPA because the facility was a well-designed and operated mass burn,

waterwall, resource recovery system with a state-of-the-art ESP.   The

EPA-sponsored test program was designed to obtain:

          PCDD/PCDF uncontrolled flue gas emission results that could be
          compared with the Signal-sponsored PCDD/PCDF controlled results.

          Uncontrolled and controlled flue gas,  particulate and specific trace
          metals for program evaluation.

          The effect that MWC uncontrolled and controlled flue gas emission
          matrices have on the trace metals analysis as performed by the
          neutron activation analytical technique.

          The uncontrolled and controlled characteristics, and
          inter-relationship of the particulate matter, PCDD/PCDF, and trace
          metals flue gas concentrations.

          Trace metal results for the ESP flyash that was being generated
          during the trace metal air emissions test program.

          Continuous emissions monitoring (CEM) information for oxygen,
          carbon monoxide and carbon dioxin during the particulate/metals
          test program.

The results from the North Andover Facility will be incorporated into the data

base for the comprehensive study report, and will be used in support of any

future regulatory development which is undertaken for the MWC source category.


1.3  BRIEF PROCESS OPERATION AND DESCRIPTION
     Figure 1-1 presents a process diagram of the two identical incinerator
systems at the North Andover facility.  Unit No. 2 was tested during this
program.  Unit No. 2 is a reciprocating grate, mass-burning type incinerator
with a waterwall boiler that produces superheated steam.  The flue gas passes
from the incinerator into superheater, generator and economizer sections
                                     1-3

-------
Secondary
  Fan
                                                       ESP Inlet Flue Gas
                                                       Sampling Location
                                  Superheater
                                          Generator
                                                  Economizer
                                                                       ESP

^~^.


\
/



\ B_
Sampling Location
ESPOutl
Flue Ga
Sampling Lo
in ^x
Vibrating
Conveyor
                                                                                       Total Ash
                                                                                       Discharge
                                  Quench Tank
  Figure 1 -1.  North Andover RESCO Process Line with Sampling
                 Locations

-------
before the particulate emissions are controlled by an ESP.

     The refuse is typical residential and commercial solid waste.   No sorting
or shredding is performed prior to incineration.  The refuse is brought to the
enclosed tipping area by truck and unloaded into the receiving pit.   A
manually-operated overhead crane transfers the refuse from the receiving pit
to the incinerator charging chute.  A Martin inclined grate and ash discharge
system is in operation at the North Andover facility.

     Radian Corporation incorporated into this report the process description
and operation section (Section 3) that was prepared by Midwest Research
Institute (MRI).  The incinerator operating data recorded during the test
program as well as design data are summarized in Section 3.  The operation of
the ESP was also monitored during the test program.  However, Signal
Environmental Systems considers the ESP operating data to be confidential and
it is not included in this report.  The original incinerator data sheets are
included in Appendix G.

1.4  EMISSION MEASUREMENT PROGRAM

1.4.1  Test Program Matrix

     The emission measurement program at the North Andover facility was
conducted from July 8 to July 16, 1986.  Table 1-1 presents the original test
matrix that was planned for the program as well as the organizations that
sponsored each type of sample.

     Once on-site, the test matrix was modified in several ways.  Problems
with the overhead I-beam used at  the ESP inlet  to support  the vertical
sampling trains caused the first  PCDD/PCDF and TOCL/PM run at the ESP inlet to
be cancelled.  Then, due to the logistical problems presented by operating two
vertical sampling trains with 10-foot probes at an eight-port sampling
location, the TOCL/PM at the ESP  inlet was cancelled by  the EPA Task Manager.
                                       1-5

-------
           TABLE 1-1.  ORIGINAL TEST MATRIX FOR NORTH ANDOVER RESCO
Run
  PCDD/PCDF

 ESP     ESP
Inlet  Outlet
                           TOCL/PM
 ESP
Inlet
 ESP
Outlet
  METALS/PM
               ESP
 ESP    ESP    Ash
Inlet  Outlet
Bottom
 Ash
CEMs
 1    EPA    Signal
 2    EPA    Signal
 3    EPA    Signal
 4    EPA    Signal
 5    EPA    Signal
 6    EPA    Signal
                 EPA   Signal-TOCL
                       EPA-PM

                 EPA   Signal-TOCL
                       EPA-PM

                 EPA   Signal-TOCL
                       EPA-PM

                 EPA   Signal-TOCL
                       EPA-PM

                 EPA   Signal-TOCL
                       EPA-PM

                 EPA   Signal-TOCL
                       EPA-PM
                                  Signal  Signal  Signal
                                  Signal  Signal  Signal
                                  Signal  Signal  Signal
                                  Signal  Signal  Signal
                                  Signal  Signal   Signal
                                  Signal   Signal   Signal
                                            EPA

                                            EPA

                                            EPA
                                             EPA    EPA

                                             EPA    EPA

                                             EPA    EPA
                                                   EPA

                                                   EPA

                                                   EPA
 Dashes indicate that the sample was not collected.   Also,  Signal — Signal
 Environmental Systems,  Inc.

 The TOCL/Particulate samples for Runs 1-6 at the  ESP inlet was  cancelled due
 to sampling location logistical problems.

"PM = particulate matter.

 Continuous emissions monitors were used to measure  0~,  CO,,,  and CO at the
 ESP outlet.

"The ESP Inlet PCDD/PCDF sample for Run 1 was not  collected because the
 sampling location was not ready.

 The ESP Inlet PCDD/PCDF samples for Run 2 and Run 6 were not analyzed due  to
 sampling and incinerator operating problems that  occurred  during these runs.
                                      1-6

-------
     During PCDD/PCDF Runs 2, 3, 4,  5,  and 6,  sampling for at the ESP inlet
was conducted according to the December 1984 draft of the Environmental
Standards Workshop protocol for sampling PCDD's and PCDF's.   The PCDD/PCDF
sampling at the ESP inlet and ESP outlet was conducted simultaneously.
Particulate loading was measured according to EPA Method 5 at the ESP outlet
for Runs 1-6.

     As part of an EPA in-house study,  trace metals testing was conducted
simultaneously at the ESP inlet and ESP outlet during Runs 7, 8, and 9.
Sampling followed EPA Alternate Method 12, which also allows for the
determination of particulate loading concurrently in the sampling train.   The
EPA Method 12 train has been demonstrated specifically for lead and cadmium
only.  However, for the purposes of the in-house study the method was used as
a screening analysis for the other metals of interest.  The method was also
modified by using neutron activation (NAA) as the analysis method rather than
atomic adsorption.  Neutron activation produces results for the full spectrum
of detectable metals in one analytical step and is a cost-effective analysis
on a per element and per sample basis.   The results for arsenic, cadmium,
total chromium and nickel are included in this report.  The results for other
metals are included in Appendix I.

     Continuous emission monitoring (CEM) for oxygen  (0-) , carbon monoxide
(CO) and carbon dioxide (C0_) was also conducted during Runs 7, 8, and 9.  The
purpose of the continuous monitoring effort was to 1) observe fluctuations in
flue gas parameters, and 2) provide an indication of  combustion conditions.
Plant personnel collected the incinerator and ESP operating data.  The CEM
data and the process operating data were reviewed to  determine  if the
incinerator was operating at normal conditions.

     A summary of the sampling log for the test program is presented in
Table 1-2.  The summary shows the samples collected and sampling times for the
EPA-sponsored sampling as well as any problems that occurred.  A detailed log
is included in Appendix H.I.
                                       1-7

-------
          TABLE  1-2.   SUMMARY OF  SAMPLING LOG FOR EPA TESTING AT THE
                       NORTH ANDOVER FACILITY: July 8 to 16, 1986
  DATE
RUN
 SAMPLES
COLLECTED
SAMPLING
 PERIOD
                                                          NOTES
 7/8/86



 7/9/86




 7/10/86


 7/11/86



 7/12/86
 2    Uncontrolled    10:15-19:29
      PCDD/PCDF
 3    Uncontrolled    10:29-16:30
      PCDD/PCDF

 4    Uncontrolled    11:30-16:09
      PCDD/PCDF
 5    Uncontrolled    11:40-17:52
      PCDD/PCDF
7/13/86
 6    Uncontrolled    12:40-20:46
      PCDD/PCDF
7/14/86
7/15/86
7/16/86
      Uncontrolled    14:20-20:00
      and controlled
      metals,  ESP
      ash,  CEMs

      Uncontrolled    9:30-13:50
      and controlled
      metals,  ESP
      ash,  CEMs

      Uncontrolled    9:38-14:06
      and controlled
      metals,  ESP
      ash,  CEMs
                             Run 1 was cancelled because the ESP inle
                             sampling location was not ready in time
                             to test concurrently with ESP outlet.

                             For the inlet PCDD/PCDF train,  three
                             probe liners were used and recovered.
                             Two of the liners were broken during por
                             changes.

                             No sampling or incinerator operating
                             problems occurred.

                             Sampling time increased to 240 mintes
                             from 192 minutes.  No sampling or
                             incinerator operating problems occurred.

                             Incinerator developed a broken grate bar
                             during sampling.   Underfire air ports
                             were manually cleaned.  Incinerator
                             operation was determined by Signal to  be
                             normal.  The grate bar was repaired
                             overnight.

                             The incinerator was determined by Signal
                             not to be operating at normal conditions
                             The operating data basis was not provide
                             by Signal.   The PCDD/PCDF samples were
                             not collected simultaneously at the ESP
                             inlet and ESP outlet.

                             Outlet probe liner broke at the nozzle;
                             liner changed.
                             No  sampling or incinerator operating
                             problems  occurred.
                             No  sampling or incinerator operating
                             problems  occurred.
 The sampling period includes time for port changes and other breaks in
 sampling.
                                         1-8

-------
1.4.2  Laboratory Analysis

     The Laboratory analyses were performed by three organizations.   The
PCDD/PCDF analyses were performed by Triangle Laboratories,  Inc.,  Research
Triangle Park, N.C.  The trace metals analyses were performed by the Nuclear
Energy Services of North Carolina State University in Raleigh, North Carolina.
The particulate samples were analyzed in the Radian/RTP Laboratory.

     The PCDD/PCDF samples were analyzed by high resolution gas chromatography
and high resolution mass spectrometry (GC/MS).  The congeners that are
reported are listed in Table 1-3.  The total mono- through octa- chlorinated
homologues are reported, along with all the individual 2378-substituted
PCDD/PCDF isomers such as 2378-TCDD.

     The trace metals samples were analyzed by neutron activation analysis
(NAA).   With this method, the samples are exposed to neutrons causing them to
emit gamma rays which are counted and compared to standards for
quantification.  The method reports results for arsenic, cadmium,  chromium,
and nickel and well as thirty five other metals which are listed in Table 1-4.
NAA cannot be used for lead and beryllium because these metals do not emit
gamma rays.  The results for the specific metals of interest are included in
this report.  The results for the other metals are included in Appendix I.

1.5  QUALITY ASSURANCE/QUALITY CONTROL (QA/QC)

     Completeness and data quality was emphasized during the test program at
North Andover RESCO.  QA/QC,was incorporated into each sampling or analytical
task.  The QA/QC program and results were reviewed by the Radian Quality
Assurance (QA) Officer.  The QA/QC results are summarized in Section 6 and
presented in more detail in Appendix J.

1.6  DESCRIPTION OF REPORT SECTIONS

     The emissions report is presented in three volumes.  Volume I includes
the Summary of Results (Section 2.0), Process Description and Operation
                                      1-9

-------
  TABLE 1-3.  PCDD/PCDF CONGENERS ANALYZED FOR NORTH ANDOVER TEST PROGRAM
DIOXINS
   Monochloro dibenzo-p-dioxin (MCDD)
   Total dichlorinated dibenzo-p-dioxins (DCDD)
   Total Trichlorinated dibenzo-p-dioxins (TrCDD)
   2,3,7,8 Tetrachlorodibenzo-p-dioxin (2,3,7,8 TCCD)
   Total Tetrachlorinated dibenzo-p-dioxins (TCDD)
   1,2,3,7,8 Pentachlorodibenzo-p-dioxin (1,2,3,7,8 PCDD)
   Total Pentachlorinated dibenzo-p-dioxins (PCDD)
   1,2,3,4,7,8 Hexachlorodibenzo-p-dioxin (1,2,3,4,7,8 HxCDD)
   1,2,3,6,7,8  Hexachlorodibenzo-p-dioxin (1,2,3,6,7,8 HxCDD)
   1,2,3,7,8,9  Hexachlorodibenzo-p-dioxin (1,2,3,7,8,9 HxCDD)
   Total Hexachlorinated dibenzo-p-dioxins (HxCDD)
   Total Heptachlorinated dibenzo-p-dioxins (HpCDD)
   Total Octachlorinated dibenzo-p-dioxins (OCDD)

FURANS
   Monochloro dibenzofuran (MCDF)
   Total dichlorinated dibenzofurans (DCDF)
   Total Trichlorinated dibenzofurans (TrCDF)
   2,3,7,8 Tetrachlorodibenzofurans (2,3,7,8 TCDF)
   Total Tetrachlorinated dibenzofurans (TCDF)
   1,2,3,7,8 Pentachlorodibenzofuran (1,2,3,7,8 PCDF)
   2,3,4,7,8 Pentachlorodibenzofuran (2,3,4,7,8 PCDF)
   Total Pentachlorinated dibenzofurans (PCDF)
   1,2,3,4,7,8 Hexachlorodibenzofuran (1,2,3,4,7,8 HxCDF)
   1,2,3,7,8,9 Hexachlorodibenzofuran (1,2,3,7,8,9 HxCDF)
   2,3,4,6,7,8 Hexachlorodibenzofuran (2,3,4,6,7,8 HxCDF)
   Total Hexachlorinated dibenzofurans (HxCDF)
   Total Heptachlorinated dibenzofurans (HpCDF)
   Total Octachlorinated dibenzofurans (OCDF)
                                    i-io

-------
               TABLE 1-4.   METALS  DETECTED BY NEUTRON ACTIVATION
                           IN FLUE GAS  SAMPLE MATRIX
   Specific Metals of Interest'

     Arsenic (As)
     Cadmium (Cd)
     Chromium (Cr)
     Nickel (Ni)
   Other Metals of Interest

     c
Toxic

Antimony
Copper
Selenium
Silver
Zinc
Uranium
Vanadium
Conventional

Aluminum
Barium
Bromine
Chlorine
Cobalt
Iron
Magnesium
Molybdenum
Manganese
Tin
Titanium
                           Additional  Metals  Reported
                    Calcium
                    Cerium
                    Cesium
                    Europium
                    Hafnium
                    Indium
                    Lanthanum
                    Lutetium
                    Neodymium
      Potassium
      Rubidium
      Samarium
      Scandium
      Sodium
      Terbium
      Thorium
      Ytterbium
 NAA cannot be used for lead and beryllium because these metals do not emit
 gamma rays when exposed to neutrons.
 The results for these metals are reported in Appendix I.
 These metals are classified as toxic pollutants by the NPDES regulation 40CFR
 Part 122, Appendix D.
 These metals are classified as conventional pollutants by the NPDES
 regulation 40CFR Part 122, Appendix D.
g
 The results for these metals are reported in Appendix I.
                                       1-11

-------
(Section 3.0), Sampling Locations (Section 4.0), Sampling and Analytical
Procedures (Section 5.0) and Quality Assurance/Quality Control (Section 6.0).

     The supporting data and calculations for the results presented in
Volume I are included as Appendices.   The appendices are presented as two
volumes.  Volume II contains summaries of all the emissions data, sample
calculations, GEM one-minute averages, GEM stripcharts, GEM calibration data,
field data sheets and laboratory reports.  Volume III contains equipment
calibration data, correspondence, incinerator and ESP operating data, field
test logs,  quality assurance information, sampling and analytical protocols,
and a list of the project participants.
                                     1-12

-------
                            2.0  SUMMARY OF RESULTS
     The results of the PCDD/PCDF, particulates,  metals, and GEM sampling
conducted for EPA at North Andover RESCO are summarized in this section.   In
addition to the presentation of the results, variabilities and outliers in the
data are qualified.  Also, any incinerator or ESP operating abnormalities
encountered during sample collection or analysis are analyzed in relation to
the results.

     Dual units (English and Metric) are presented in each table, where
applicable.  For some results, such as PCDD/PCDF concentration, only the most
suitable units (ng/dscm) are presented and English conversion factors are
provided in the footnotes.  The results are normalized to a standard CC-
concentration basis to allow comparison of the results on an equivalent basis.
The EPA/MSW database is also normalized to this basis, since many state
regulations are based on a 12 percent CC*  basis.   The 12 percent C09 basis is
appropriate for the MSW source category because the ultimate analysis of
refuse on a combustible fraction basis has a reasonably constant carbon
        2
content.

     The equations used to calculate the results that are presented in the
following tables are presented in Appendix A.5.  The supporting data for the
results presented in this section are included in the appendices to this
report.

2.1  PCDD/PCDF EMISSION RESULTS

     The PCDD/PCDF results for the flue gas samples collected at the ESP inlet
are discussed in this section.  Although five flue gas samples were collected,
the three samples which were considered the most representative in terms of
sampling and incinerator operating parameters were analyzed.  The samples
analyzed were from Runs 3, 4, and 5.  The sample from Run 1 was not collected
because the inlet sampling location was not ready.  The sample for Run 2 was
                                        2-1

-------
 not  analyzed  because  the  probe  liner broke three times, twice during port
 changes.   Since  PCDD/PCDF samples were not collected simultaneously during
 Run  6,  the Run 6 sample was not analyzed.

     The uncontrolled PCDD/PCDF emissions results are summarized in Table 2-1.
 The  flue gas  characteristics and process operations data presented in the
 table show that  the incinerator was operating at similar conditions during
 each of the runs.  The average  total PCDD/PCDF concentration was 284 ng/dscm.
 Normalized to 12 percent  CO , the average total PCDD/PCDF concentration was
 342  ng/dscm @ 12 percent  CO-.

     Comparing the data from the three runs, the PCDD isomers from Run 3 are
 significantly higher  than Runs  4 and 5.  However, the furan isomers from Run 3
 are  in  good agreement with the  other runs and process conditions were normal.
 Therefore, the PCDD data  from Run 3 is considered to be a normal variation of
 the  uncontrolled PCDD/PCDF data.

     The homologue and  isomer-specific results are summarized in Table 2-2.
 The  results in this table are presented as total train analyses and are
 normalized to 12 percent  C0_.   The average total PCDD result was 169 ng/dscm @
 12 percent C09 and the average  total PCDF result was 172 ng/dscm @ 12 percent
 co2.

     The PCDD/PCDF results are  presented for the front half and back half
 fractions  of  the sampling train in Table 2-3.  The results show that the
majority of the  PCDDs and PCDFs  were captured in the front half of the train.
On the  average,  seventy percent of the PCDDs and fifty five percent of the
PCDFs were found in the front half of the train.  The front half of the
PCDD/PCDF  train  included  the probe, the cyclone and the filter.  The back half
included the  coil condenser, XAD trap and the impingers.

2.1.1   2378-TCDD Toxic Equivalency

     The PCDD/PCDF results are expressed in terms of 2378-TCDD toxicity
equivalents corrected to  12 percent CO- in Table 2-4.   Each isomer has
                                        2-2

-------
            TABLE  2-1.   SUMMARY OF UNCONTROLLED  PCDD/PCDF  EMISSIONS  FOR  NORTH  ANDOVER RESCO

Run No.
Date
Sampling Parameters3
Volume gas sampled (dscf)
Stack gas flow rate (dscftn)
Stack temperature (°F)
Percent moisture by volume
Percent 1sok1net1c.
CO (ppm by volume) .
C02 (percent by volume)
0, (percent by volume)
Run 3
7/10/86

75.9
84,600
580
13.1
100.1
25.4
8.9
10.5
Run 4
7/11/86

97.0
87,500
584
12.9
99.0
45.2
9.6
10.7
Run 5
7/12/86

97.3
88,600
591
14.2
98.0
25.7
9.8
10.1
Average

90.1
86,900
585
13.4
99.0
32.1
9.4
10.4
Process operations      ,
  Steam load   (Ibs/hr 10 )
        166

Front  Back
 Hal f  Hal f  Total
        166

Front  Back
 Hal f  Hal f  Total
        167

Front  Back
 Hal f  Hal f  Total
       166

Front  Back
 Half  Half  Total
D1ox1n Results
Total PCDD (ng/dscm)
Total PCDD (corrected
to 12* C02, ng/dscm)
Furan Results
Total PCDF (ng/dscm)
Total PCDF (corrected
to 12% (XL, ng/dscm)
D1ox1n-Furan Results0
Total PCDD-PCDF (ng/dscm)
Total PCDD-PCDF (corrected
to 12* C02, ng/dscm)
aConvers1on factors: dscf x 0.028317

184
221


64
77


248
297


96
115


97
116


193
232


280
336


161
193


441
529

= dscm; dscfm x 0.

39
47


69
83


108
130

028317 =

17
20


60
72


77
92

dscmm;

56
67


129
155


185
222

(°F

72
87


104
126


176
213

- 32) x 5/9

15
18


36
44


51
62

= °C

87
105


140
170


227
275



98
118


79
95


177
213



43
51


64
77


107
129



141
169


143
172


284
342


 Standard conditions are 68°F (20°C)  and  1  atm  (1.01325 x 103  Pa).
 These values are  averages of data taken  over the sampling period  from continuous emissions monitors located at  the ESP
 outlet.
CPCDD/PCDF results are adjusted for Internal standard recoveries and  sample blank results.  Results are reported as front
 half/back half  fractions for the convenience of EPA.

-------
 TABLE  2-2.  UNCONTROLLED PCOD/PCDF EMISSIONS AT NORTH
 ANDOVER  RESCO  (NORMALIZED TO 12 PERCENT C02)
                        CONCENTRATION
                  (NG/DSCM, CORRECTED TO 12% C02)
 ISOMER          RUN 03   RUN 04   RUN 05   AVERAGE

 DIOXIN
 Mono-CDD              0000
 Di-CDD                4223
 Tri-CDD              22        5        7       11
 2378 TCDD             4012
 Other TCDD           31        6        8       15
 12378 PCDD            2111
 Other PCDD           64        8       12       28
 123478 HxCOD          3111
 123678 HxCDD          7123
 123789 HxCDD          3202
 Other HxCDD          86       10       19       38
 Hepta-CDD            69       15       24       36
 Octa-CDD             42       17       30       29

 TOTAL PCDD          336       67      105      169

 FURAN
 Mono-CDF              1302
 Di-COF               27       26       13       22
 Trl-CDF              73       57       62       64
 2378 TCDF            11        9       12       11
 Other TCDF           36       27       34       32
 12378 PCDF            2222
 23478 PCDF            4444
 Other PCDF           13        9       13       12
 123478 HxCDF          5344
 123678 HxCDF          2121
 123789 HxCDF          0000
Other HXCDF           7466
Hepta-CDF            10        8       12       10
Octa-PCDF             3233

TOTAL PCDF          193      155      170      172

TOTAL PCDD/PCDF     529      222      275      341
Norm, ratio (a)    1.20     1.20     1.21
a Norm, ratio = normalization ratio of 12 percent C02
  to C02 measured which is used to normalize the
  results to a standard 12 percent C02 basis.
                          2-U

-------
     TABLE 2-3.   UNCONTROLLED PCDD/PCDF EMISSIONS  FOR  NORTH  ANDOVER RESCO
                            (FRONT,  BACK,  AND TOTAL  FRACTION RESULTS)
CONCENTRATION (NG/DSCM) a
RUN 03 RUN 03 RUN 03 RUN 04 RUN 0* RUN 04 RUN 5
FRONT BACK TOTAL FRONT BACK TOTAL FRONT
ISOMER HALF HALF HALF HALF HALF
OIOXIN
Mono-CDO 0.0
D1-COD 0.8
Trf-COD 6.5
2378 TCOD 3.1
Other TCOD 10.3
12378 PCOD 1.1
Other PCDO 29.8
123478 HxCDD 1.6
123678 HxCDO 3.9
123789 HxCOD 0.0
Other HxCDD 50.1
Hepta-CDD 46.1
Octa-COD 30.5
TOTAL PCOD 184
Fraction Detected
In Front Half:
FURAN
Mono-CDF 0.2
D1-COF 4.3
Trl-CDF 18.5
2378 TCDF 4.3
Other TCOF 12.0
12378 PCOF 1.0
23478 PCOF 2.1
Other PCOF 5.8
123478 HxCDF 2.6
123678 HxCOF 0.9
123789 HxCDF 0.0
Other HXCDF 3.7
Hepta-CDF 6.5
Octa-PCOF 2.0
TOTAL PCDF 63.8
Fraction Detected
1n Front Half:
TOTAL PCDD/PCDF 248
Dloxln Iscner Fraction
of Total PCOD/PCDF:

0.0
2.7
11.8
0.0
15.6
0.7
23.2
0.7
1.6
2.5
21.4
11.5
4.3
96.0



1.0
17.9
42.6
5.2
17.7
0.8
1.4
4.8
1.2
0.4
0.0
1.8
1.6
0.4
96.7


193



0.0
3.5
18.3
3.1
26.0
1.7
53.0
2.2
5.5
2.5
71.6
57.6
34.8
280

0.66

1.1
22.2
61.1
9.5
29.7
1.8
3.4
10.6
3.8
1.3
0.0
5.5
8.0
2.4
161

0.40
440

0.64

0.0
0.6
2.2
0.2
2.4
0.4
3.8
0.4
0.7
1.2
5.2
9.5
12.0
38.7



0.3
6.3
23.3
4.7
13.5
1.1
2.1
5.3
1.9
0.6
0.0
2.7
5.9
1.6
69.2


108



0.0
1.2
2.3
0.0
2.2
0.2
2.5
0.2
0.4
0.5
2.9
2.9
1.9
17.1



2.4
15.4
24.5
2.8
9.0
0.5
0.8
2.3
0.5
0.2
0.0
0.6
0.8
0.0
59.9


77.0



0.0
1.8
4.6
0.2
4.6
0.6
6.3
0.6
1.0
1.7
8.1
12.4
13.9
55.8

0.69

2.7
21.6
47.8
7.5
22.5
1.6
2.9
' 7.6
2.4
0.8
0.0
3.2
6.7
1.6
129

0.54
185

0.3

0.0
1.1
3.9
0.3
5.1
0.6
7.5
0.7
1.1
0.0
12.0
16.7
22.7
71.6



0.1
5.5
36.0
7.5
21.4
1.6
3.0
8.9
3.1
1.0
0.0
4.4
9.2
2.5
104


176


RUN 5
BACK
HALF

0.0
0.5
1.6
0.1
1.5
0.1
2.2
0.2
0.3
0.0
3.5
3.1
1.9
15.1



0.0
5.5
15.5
2.3
6.9
0.4
0.7
2.1
0.5
0.2
0.0
0.7
1.1
0.2
36.1


51.1


RUN 05 AVERAGE AVERAGE AVERAGE
TOTAL FRONT BACK TOTAL
HALF HALF

0.0
1.6
5.5
0.4
6.6
0.8
9.7
0.8
1.4
0.0
15.5
19.8
24.6
86.7

0.83

0.1
• 11.0
51.5
9.7
28.4
2.0
3.7
11.0
3.6
1.2
0.0
5.1
10.2
2.8
140

0.74
227

0.38

0.0
0.8
4.2
1.2
6.0
0.7
13.7
0.9
1.9
0.4
22.4
24.1
21.7
98.0



0.2
5.4
25.9
5.5
15.6
1.2
2.4
6.7
2.5
0.8
0.0
3.5
7.2
2.1
79.1


177



0.0
1.5
5.2
0.0
6.4
0.3
9.3
0.3
0.8
1.0
9.3
5.8
2.7
42.7



1.1
12.9
27.6
3.4
11.2
0.6
1.0
3.1
0.7
0.3
0.0
1.0
1.2
0.2
64.2


107



0.0
2.3
9.5
1.3
12.4
1.0
23.0
1.2
2.6
1.4
31.7
29.9
24.5
141

0.7

1.3
18.3
53.5
8.9
26.8
1.8
3.3
9.8
3.3
1.1
0.0
4.6
8.3
2.3
143

0.55
284

0.50
0.0 = Not detected. The minimum detection limit Is approximately 0.0004 ng/dson.
Specific detection limits are Included In Appendix D.I.
                                          2-5

-------
   TABLE 2-4.  UNCONTROLLED PCDD/PCDF CONCENTRATIONS
               EXPRESSED AS 2378-TCDD TOXIC EQUIVALENTS
ISOMER
DIOXIN
Mono-CDD
Di-CDD
Tri-CDD
2378 TCDD
Other TCDD
12378 PCDD
Other PCDD
123478 HxCOD
123678 HxCDD
123789 HxCDD
Other HxCDD
Hepta-CDD
Octa-CDD
TOTAL PCDD
2378 TCDD
2378 TCDD EQUIVALENT CONCENTRATION
EQUIV. (NG/DSCM @ 12% C02)
FACTORS RUN 03 RUN 04 RUN 05 AVERAGE

0.0000
0.0000
0.0000
1.0000
0.0100
0.5000
0.0050
0.0400
0.0400
0.0400
0.0004
0.0010
0.0000


0.0
0.0
0.0
3.7
0.3
1.0
0.3
0.1
0.3
0.1
0.0
0.1
0.0
6.0

0.0
0.0
0.0
0.3
0.1
0.3
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.9

0.0
0.0
0.0
0.5
0.1
0.5
0.1
0.0
0.1
0.0
0.0
0.0
0.0
1.3

0.0
0.0
0.0
1.5
0.1
0.6
0.1
0.1
0.1
0.1
0.0
0.0
0.0
2.7
FURAN
Mono-CDF         0.0000      0.0      0.0      0.0      0.0
Di-CDF           0.0000      0.0      0.0      0.0      0.0
Tri-CDF          0.0000      0.0      0.0      0.0      0.0
2378 TCDF        0.1000      1.1      0.9      1.2      1.1
Other TCDF       0.0010      0.0      0.0      0.0      0.0
12378 PCDF       0.1000      0.2      0.2      0.2      0.2
23478 PCDF       0.1000      0.4      0.4      0.4      0.4
Other PCDF       0.0010      0.0      0.0      0.0      0.0
123478 HxCDF     0.0100      0.0      0.0      0.0      0.0
123678 HxCDF     0.0100      0.0      0.0      0.0      0.0
123789 HxCDF     0.0100      0.0      0.0      0.0      0.0
Other HXCDF      0.0001      0.0      0.0      0.0      0.0
Hepta-CDF        0.0010      0.0      0.0      0.0      0.0
Octa-PCDF        0.0000      0.0      0.0      0.0      0.0

TOTAL PCDF          _        1.9      1.5      2.0      1.8

TOTAL PCDD/PCDF              7.9      2.4      3.2      4.5
                          2-6

-------
                                     3
2378-TCDD toxicity equivalency factor ,  (also presented in Table  2-8),  that
ranks the toxicity of the isomer relative to the toxicity of 2378-TCDD.   The
equivalency factors were developed by EPA.   2378-TCDD toxic equivalents  are
used in risk analysis models developed by EPA.  In terms of 2378-TCDD
equivalents,  the average concentration was 2.7 ng/dscm for the dioxins  and 1.8
ng/dscm for the furans.  The average total 2378-TCDD equivalents  concentration
was 4.5 ng/dscm.

2.1.2  ISOMER DISTRIBUTIONS

     The distributions of the PCDD and PCDF isomers on a mole fraction basis
are presented in graphical form in Figure 2-1 and in tabular form in
Table 2-6.  For the dioxin isomers, hexa-,  hepta- and octa-CDDs were the most
prevalent at about 20 mole percent, each.  For the furan isomers, tri-CDF was
the most prevalent at 40 mole percent.   Di-CDF and tetra-CDF were present at
about 17 mole percent, each.  Figure 2-1 also illustrates graphically the
repeatability of the data from the three runs except for the penta- and
hexa-CDD isomers measured during Run 1.

2.2  PARTICULATE MATTER RESULTS

     The particulate matter loading was measured at North Andover RESCO at
both the ESP inlet and ESP outlet locations.  The uncontrolled and controlled
results are summarized in Table 2-7.  Table 2-7 also includes flue gas and
process operations parameters that were measured during testing.   The average
controlled result does not include Run 7.  This result was determined to be a
statistical outlier based on the nine controlled particulate results under the
condition where extreme observations in either direction are considered
           4
rejectable.   The Run 7-controlled train developed a broken glass liner during
sampling in one of the ports.  Considering this, the concentration would be
expected to be lower than the average unless extraneous glass fragments were
recovered in the sample.  Instead, the concentration is higher.  This could be
attributed to the higher opacity during Run 7, except  that Run 6 had an even
higher average opacity while still having a controlled particulate loading
                                       2-7

-------
TABLE 2-5.  KEY TO ISOMER CODING FOR FIGURE  2-1

CODE

A =
B =
C =
D =
E =
F =
G =
H =
I =
J =
K =
L =
M =

N =
0 =
P =
Q =
R =
S =
T =
U =
V =
W = -
X =
Z =
AA =
ISOMER
Dioxins
Mono-CDD
Di-CDD
Tri-CDD
2378-TCDD
Other TCDD
12378 PCDD
Other PCDD
123478 HxCDD
123678 HxCDD
123789 HxCDD
Other HxCDD
Hepta-CDD
Octa-CDD
Furans
Mono-CDF
Di-CDF
Tri-CDF
2378-TCDF
Other TCDF
12378 PCDF
Other PCDF
123478 HxCDF
123678 HxCDF
123789 HxCDF
Other HxCDF
Hepta-CDF
Octa-CDF
                              2-8

-------
                   FLUE  GAS  - DIOXINS
                        INLET - NORTH ANDOVER
01
o
o
o
3
0.4- -
0.3 -

0.2 -







0.1 -






0 -
Mean Total














v\
v\~*

FKPS
J *
A B













/
/
/

y













/
/
/
y
y
y

/
I
C













x
X
X

X



















PT__
I
D









j

/

/
^
/

/













y

/
y
y
/
y
/
T
E











x

X
X
X

X


PCDDs

















i-f^r-i
/
/
/



/

/

/
/
y

/





f-
/

/
y
v,
y
y
y
y
/
I i
F G
= 141 ng/dscm









X

X

X
x
X

x

















r~»

pf7T~l r Kr^t fiyl
y
'
y
y
/



y

y

/
y
y

/

r7-



T
/

/

/

/
y
y


x
X

X

X

X

X
X
x

x














/
/

/

y

/

/
/
/

/
/
/
y

/

_/

/

y
y
y
y
/
/
*
X
X
X

X

X

X

X
x
><

X

















H I J K L

X









/

y
/
y

/
r*
y x
/ x
/ x

/ X
y
/X

y x

/x
/x
/ x

/x
— T 	
M
              RUN 03
                          DIOXIN ISOMERS
                         X///A RUN 04
IX>\l RUN 05
10
2
O
u
o
                    FLUE GAS  -  FURANS
                        INLET - NORTH ANDOVER
u.a -
0.4 -




0.3 -




0.2 -






0.1 -



n -
Mean Total PCDFs = 143 ng/dscm
n
•
,
,
,




f


~-







T7\
r®
/
/


y
/
/
/
/
7-
^/
y
'/
y
y

y
y
y
/
/
y
4







^ '
X
X
X '
><

' 7
' /
' /

'/
f /

/^

/
** /
' /

/ /


<',

I /

M.
i"
x
x
X
x
X
x
x
X
X

X
X
X
x

X
X
X
X
X
X



















FT
rr
rrr



















X
X
X






















7
/
/


/
/
/
/
/

*/
s


/
/
/
'/
<•
X
X
X
X

X
X
X
X
X
X







r-i
PMJ _^3

rr^n Vtffr Yw\ PfeR ^-r,^-, Pm rm r--^
                                                       M
               RUN 03
                          FTJRAN ISOMERS
                             RUN 04
                                            RUN 05
Figure 2 -1.   Uncontrolled PCDD/PCDF Isomer Distribution
                            2-9

-------
 FABLE 2-6.   UNCONTROLLED PCOD/PCDF CONGENER DISTRIBUTION
             AT NORTH  ANDOVER RESCO
                    MOLE FRACTION (%)
 ISOMER          RUN 03   RUN 04   RUN 05   AVERAGE

 DIOXIN
 Mono-CDD           0.00     0.00     0.00     0.00
 Di-CDD             0.02     0.05     0.03     0.03
 Tri-CDD            0.09     0.11     0.09     0.09
 2378 TCDD          0.01     0.00     0.01     0.01
 Other TCDD         0.11     0.10     0.09     0.10
 12378 PCDD         0.01     0.01     0.01     0.01
 Other PCDD         0.20     0.12     0.12     0.15
 123478 HxCDD       0.01     0.01     0.01     0.01
 123678 HxCDD       0.02     0.02     0.02     0.02
 123789 HxCDD       0.01     0.03     0.00     0.01
 Other HxCDD        0.25     0.14     0.18     0.19
 Hepta-CDD          0.18     0.20     0.21     0.20
 Octa-CDD           0.10     0.21     0.24     0.18
FURAN
Mono-CDF           0.01     0.03     0.00     0.01
Di-CDF             0.17     0.20     0.10     0.16
Tn'-CDF            0.41     0.39     0.41     0.40
2378 TCDF          0.06     0.05     0.07     0.06
Other TCDF         0.18     0.16     0.20     0.18
12378 PCOF         0.01     0.01     0.01     0.01
23478 PCDF         0.02     0.02     0.02     0.02
Other PCDF         0.06     0.05     0.07     0.06
123478 HxCDF       0.02     0.01     0.02     0.02
123678 HxCDF       0.01     0.00     0.01     0.01
123789 HxCDF       0.00     0.00     0.00     0.00
Other HXCDF        0.03     0.02     0.03     0.02
Hepta-CDF          0.04     0.04     0.05     0.04
Octa-PCDF          0.01     0.01     0.01     0.01
                       2-10

-------
TABLE 2-7.   SUMMARY OF UNCONTROLLED  AND CONTROLLED PARTICULATE  EMISSIONS  FOR NORTH  ANDOVER RESCO
Run No. Run 7
Date 07-14-86
Type Emissions Uncontrolled Controlled
Sampling Parameters
Volume gas sampled (dscf)
Stack gas flow rate (dscfm)
Stack temperature (°F)
Moisture (percent by volume)
IsoMnetlcs (percent)
CO (ppm by volume)0
C0_ (percent by volume)0
0_ (percent by volume)0
Average opacity (percent)

83.8 120.5
91,700 98,300
599 579
16.1 15.4
101.8 99.7
35.9
8.5
11.5
0.31
Run 8 Run 9
07-15-86 07-16-86
Uncontrolled Controlled Uncontrolled Controlled

85.3 114.9
93,500 94,800
600 575
14.3 13.6
101.7 98.6
35.7
8.6
8.3d
0.14

82. 7 119.9
92,200 97,300
609 587
14.3 13.7
102.1 100.2
27.2
9.2
10.5
0.13
Average8
Uncontrolled Controlled

—
92,500 96,050
603 581
14.9 13.7
101.9 99.4
32.9
8.8
11.0
0.19
 Process operations
   Steam load (Ibs/hr x 103)

 Partlcglate Results9
  Front Half Catch)
  (Probe, cyclone, and filter)
   mg - mass

   gr/dscf
   mg/dscm
       170
4866.4

0.8965
   gr/dscf  (corrected to 12X O>2)    1.148
                                   2,052
   mg/dscm  (corrected to 12*  C0_)    2,627

   Ibs/hr                             705

   Kg/hr                              320

 Collection Efficiency Percent             97.67
 152.2

0.0195
0.0250

  44.6
  57.1

  16.4

  7.45
  3450

0.6242
0.7366

 1,429
 1,686

   501

   277
                                168
  23.6

0.0032
0.0044

  7.25
  10.0

  2.58

  1.17
  3580

0.6682
0.9221

 1,529
 2,100

   528

   240
                                               165
  28.5

0.0037
0.0054

  8.39
  12.2

  3.06

  1.39
0.7296
0.9356

 1,670
                                                                                  168
0.0035
0.0049

  7.82
                                                                  99.49
                                                        99.42
                                                                            578        2.82

                                                                            262        1.28

                                                                                 99.46
  Values  from Run 7 - controlled are not Included  1n averages.   See  Section 2.2 for explanation.
 Conversion factors: dscf  x 0.028317 = dscm;  dscfm x 0.028317,.' dscmm; (°F - 32)  x 5/9 = °C
  Standard conditions are 68°F  (20 C) and 1 atm   (1.01325 x 10   Pa).
 °These values are averages of  data taken over the sampling period from continuous emissions monitors at  the ESP outlet location.
  This value 1s not Included In the 02 average and Is considered an  Invalid data point.
 6Part1culate  results are adjusted for blank  results.

-------
within  range of  the data which was from 0.0013 to 0.0054 grains/dscf
normalized to 12 percent CO-.  However, the metal-to-particulate ratios
dicussed  in Section 2.5 are similar for all three runs which indicates that
the ESP may have malfunctioned during Run 7.  However, the ESP operating data
for Run 7 are not available for review at this time.  Therefore, since a
broken  probe liner developed during Run 7 at the ESP outlet, and a malfunction
of the  incinerator and/or ESP may have occurred, the Run 7-controlled
particulate data are not included in the particulate averages presented.

     The average uncontrolled particulate concentration normalized to
12 percent CO- was 0.9356 grains/dscf and the average controlled concentration
was 0.0049 grains/dscf @ 12 percent CO-.  The average collection efficiency of
the ESP was 99.46 percent.

     The controlled particulate loading was measured for all nine runs
performed at North Andover RESCO.  The results are summarized in Table 2-8.
The average controlled particulate loading was 0.0036 grains/dscf normalized
to 12 percent CO .  However, the results from Runs 1, 6, and 7 are not
included in the average.

     The particulate loading results from Run 1 are considered invalid due to
port scrapings that were collected on the filter.  After Run 1, the ports were
lined with stove pipe to prevent rusty flakes from the port from being drawn
into the sampling train.

     For Run 6, the incinerator was determined by Signal Environmental
Systems, Inc.  to be operating at abnormal conditions after testing was
completed.  The incinerator had developed a broken grate bar during Run 5
which was manually cleaned until Run 5 was completed.  Then, the incinerator
was shut down overnight and repaired.   When Run 6 began the next day, the
incinerator appeared to be operating normally but Signal later decided that
that the incinerator was still in a start-up operating mode.  Run 7 was not
included in the average for the reasons discussed previously in this section.
                                      2-12

-------
                         TABLE 2-8.   SUMMARY OF  CONTROLLED  PARTICULATE EMISSIONS FOR NORTH ANDOVER RESCO
rv>

H
CO

Run No.
Date
Sampling Parameters
Volume gas sampled (dscf)
Stack gas flow rate (dscfm)
Stack temperature (°F)
Moisture (percent by volume)
Jsok1net1cs (percent)
CO (ppm by volume)0
C02 (percent by volume)0
0_ (percent by volume)0
Average opacity (percent)
Process, operations
Steam load (lbs/hr x 103)
Partlculate Results8
Front Half Catch)
(Probe and filter)
mg - mass
gr/dscf
gr/dscf (corrected to 12X CO
mg/dscm
mg/dscm (corrected to 12X CO
lbs/hr
Kg/hr
a
Values from Runs 1, 6, and 7
"Conversion factors: dscf x 0.
Standard conditions are 68 F
°These values are averages of
This value Is not Included 1n
Run 1
7-8-86

54.50
96,800
582
12.6
100.7
28.4
9.0
10.9
NR

166



31.8
0.0090
2) 0.0150
20.6
2) 26.4
7.47
3.39

Run 2 Run 3
7-9-86 7-10-86

53.4 84.3
95.000 85,800
587 559
13.3 12.8
100.6 100.0
37,4 25.4
8.9 8.9
10.9 10.5
0.10 0,12

165 166



13.5 5.8
0.0039 0.0011
0.0050 0.0013
8.92 2.43
11.4 2.94
3.17 0.781
1.44 0.354

Run 4
7-11-86

107.2
87.200
567
12.6
100.0
45.2
9.6
10.7
0.12

166



18.7
0.0027
0.0032
6.16
7.39
2.01
0.912

are not Included In the averages. See Section
028317 = dscmj dscfm x 0.028317,." dscmm; (°F -
Run 5
7-12-86

106.5
86,200
565
13.6
100.6
25.7
9.8
10.1
0.13

167



13.6
0.0020
0.0023
4.67
5.46
1.51
0.684

Run 6
7-13-86

115.9
94,300
577
13.0
100.0
31.1
9.2
10.8
0.55

163



29.3
0.0039
0.0048
8.93
10.9
3.15
1.43

Run 7
7-14-86

120.5
98,300
579
13.0
99.7
35.9
8.5
11.5
0.31

170



152.2
0.0195
0.0250
44.6
57.1
16.4
7.45

Run 8
7-15-86

114.9
94,800
575
13.6
98.6
35.7
8.6
8.3d
0.14

168



23.6
0.0032
0.0044
7.25
10.0
2.58
1.17

Run 9
7-16-86

119.9
97,300
587
13.7
100.2
27.2
9.2
10.5
0.13

165



28.5
0.0037
0.0054
8.39
12.2
3.06
1.39

Average3


__
91,050
573
13.3
100.0
32.8
9.2
10.5
0.12

166



—
0.0028
0.0036
6.30
8.23
2.19
0.99

2.2 for explanation.
32) x 5/9 - °C
(20°C) and 1 atm (1.01325 x 105 Pa).
data taken over the sampling period
the CL average
and Is considered an
from continuous emissions monitors
at the
ESP outlet
location.

1nval Id data point.
             NR = data not recorded by plant.

-------
 2.3  METALS EMISSIONS RESULTS

     In  order  to  screen the flue gas for a multiple number of metals, the
 Method 12  samples collected at the ESP inlet and ESP outlet were analyzed by
 NAA.  However,  the EPA Method 12 sampling train has only been demonstrated to
 capture  lead and  cadmium efficiently.  Thus the results for arsenic, total
 chromium,  and  nickel should be considered only as screening results.

 2.3.1  Flue Gas Metals Results

     The metals emission results for the specific metals of interest are
 summarized in  Table 2-9.  These results should be considered as screening
 results except for Cadmium for which the method has been validated.  The
 average normalized arsenic concentration was 934 ug/dscm uncontrolled, and
 10.4 ug/dscm controlled.  The average ESP collection efficiency for arsenic
 was 98.66  percent.  The average normalized cadmium concentration was 446
 ug/dscm uncontrolled, and 22.3 ug/dscm controlled.  The average ESP collection
 efficiency for cadmium was 94.69 percent.  For total chromium, the average
 normalized concentration was 4277 ug/dscm uncontrolled, and 767 ug/dscm
 controlled.  The  average ESP collection efficiency for total chromium was
 99.87 percent.  The average normalized nickel concentration was 523 ug/dscm
 uncontrolled, and 477 ug/dscm controlled.  The average ESP collection
 efficiency for nickel was 81.75 percent.

     Total chromium and arsenic demonstrated the highest collection
 efficiencies, with collection efficiencies in the greater than 96 percent
 range.   The ESP was less efficient for collecting cadmium, and was the least
 efficient  for collecting nickel.

     The specific metals results contain some outliers.  Nickel in Run 7 and
 Chromium in Run 7 have collection efficiencies that are very low or negative
 caused by  a high  controlled result.  Although the results were adjusted for
blanks,  precleaned glassware was used, and contact of the train with metal was
minimized, contamination may have occurred.  Thus, these results are not
 included in the averages reported.

-------
                           TABLE  2-9.   SUMMARY OF EPA SPECIFIC  METALS  EMISSIONS  FOR  NORTH ANDOVER  RESCOa
ro
H
Run No.
Date
Type Emissions
Sampling Parameters8
Volume gas sampled (dscf)
Stack gas flow rate (dscfm)
Stack temperature (°F)
Moisture (percent by volume)
Isoklnetlcs (percent)
CO (ppm by volume)
C0_ (percent by volume)
Qj (percent by volume)
Average opacity (percent)
Process operations
Steam load (Ibs/hr x 103)
Specific Metals Resultsc
(Qorrected to 12* C02>
Element
Arsenic
Cadml urn
Total chromium
Nickel
Conversion factors: dscf x 0.
Run 7
07-14-86
Uncon- Con-
trolled trolled

83.8
91,700
599
16.1
101.8




Uncon-
trolled
(ug/dscm)
786
402
2,494
594

120.5
98,300
579
15.4
99.7
35.9
8.5
11.5
0.31
170
Con-
trolled .
(ug/dscm) CE (XT
25.6 96.51
34.6 90.78
2,291 1.51f
1,357 e,f
028317 = dscm; dscfm x 0.028317.
Run 8
07-15-86
Uncon- Con-
trolled trolled

85.3
93,500
600
14.3
101.7




Uncon-
trolled
(ug/dscm)
981
470
3,370
831
.= dscmm;

114.9
94,800
575
13.6
98.6
35.7
8.6
8.3b
0.14
168
Con-
trolled .
(ug/dscm) CE (X)
2.35 99.79
6.93 98.72
10.4 99.73
25.6 97.32
<°F - 32) x 5/9 = °C
Run 9
07-16-86
Uncon- Con-
trolled trolled

82.7
92,200
609
14.3
102.1




Uncon-
trolled
(ug/dscm)
1,036
465
6,968
143


119.9
97,300
587
13.7
100.2
27.2
9.2
10.5
0.13
165
Con-
trolled .
(ug/dscm) CE (X)°
3.30 99.68
25.4 94.57
0.00 100.00
48.8 66.17

Average
Uncon- Con-
trolled trolled

—
92,500
603
14.9
101.9




Uncon-
trolled
(ug/dscm)
934
446
4,277
523


~
96,050
581
13.7
99.4
32.9
8.8
11.0
0.19
168
Con-
trolled
(ug/dscm) CE (X)
10.4 98.66
22.3 94.69
767 99.87
477 81.75

         Standard conditions are 68°F (20DC)  and  1  atm  (1.01325  x 10" Pa).
         This  value Is not included In the average  and Is considered an Invalid data point.
        cThese results are for the total train.   Beryllium and lead determination  not possible by  NAA analysis.   Values are corrected for blank  results.
         CE  =  collection efficiency based on  mass rates.
        eThe control efficiencies were negative for these runs.
         Negative and obviously uncharacteristic  collection efficiencies were not  Included 1n the  average values.

-------
2.3.2  ESP Ash Metals Results

     The results of the ESP ash metals analyses are presented in Table 2-10.
The most prevalent metals were aluminum, calcium, sodium, zinc, potassium,
chlorine, iron, titanium and magnesium.  These metals had concentrations from
80,000 to 17,000 ppm.  Total chromium was detected at 679 ppm, cadmium was
detected at 356 ppm, arsenic was detected at 365 ppm and nickel was detected
at 245 ppm.

2.4  GEM RESULTS

     The flue gas was continuously monitored for oxygen, carbon monoxide and
carbon dioxide at the ESP outlet sampling location.   The results are presented
in Table 2-11.  The Radian data acquisition system scanned each channel 180
times per minute and recorded one-minute averages.  The mean of the one-minute
averages is reported for each run.  Then, the mean from each run is averaged
for an overall average.  The one-minute averages for each run are listed in
Appendix B.

     The average oxygen concentration was 11.0 percent by volume.  The average
carbon dioxide concentration was 8.8 percent by volume and the average carbon
monoxide concentration was 32.9 ppm.  The standard deviations for each data
set are shown in parentheses.

     The one-minute averages are plotted against time for each parameter for
each run in Figures 2-2 to 2-4.  These plots, along with the standard
deviations, show the variability of the parameters during each test run.

     The GEM and Orsat analysis results were validated based on a combustion
stoichiometry method described in Reference 3.  This analysis shows that the
GEM data for Run 8 is an outlier.- The analysis is described in more detail in
Section 6.2.  The oxygen concentration during Run 8 is lower than Runs 7 and 9
where the same C0_ and CO concentrations vere measured.  AI  , the oxygen
concentration plot in Figure 2-2 shows the oxygen concentration decreasing
                                      2-16

-------
      TABLE 2-10.   SUMMARY OF ESP ASH METALS RESULTS
  Element
MICROGRAM OF ELEMENT PER GRAM OF ASH (PPM)a'b
Run 07      Run 08      Run 09     Average
Al urn i n urn
Calcium
Sodium
Zinc
Potassium
Chlorine
Iron
Titanium
Magnesium
Tin
Bromine
Barium
Manganese
Copper
Antimony
Chromium0
Arsenic
Cadmium
Nickel
80208
64044
42350
29631
14193
14730
22791
16665
6731
2390
1956
1354
1263
1115
1073
568
75
274
181
84888
64856
46332
34904
30123
27258
14642
17231
5600
3813
1530
1594
1233
1177
1000
441
465
392
448
78073
83180
43271
37267
16556
17392
14928
16837
6672
4099
932
1189
1169
1317
973
1029
554
401
100
81056
70693
43984
33934
20290
19794
17453
16911
6334
3434
1473
1379
1222
1203
1015
679
365
356
243
 When  using  NAA,  sodium  in  the  matrix may  interfere with  the
 metals  results.

^Results for the  remaining  metals  are presented  in Appendix  I.

'Total chromium results  are presented.
                               2-17

-------
             TABLE 2-11.  SUMMARY OF GEM RESULTS

a
Parameter
02 (%vol)
(std dev)
CO (ppm)
(std dev)
C02 (%vol)
(std dev)
07
11.5
(0.7)
35.9
(6.9)
8.5
(0.6)
08
8.3b
(l.D
35.7
(7.2)
8.6
(0.7)
09
10.5
(0.6)
27.2
(6.8)
9.2
(0.6)
Average
11.0
(0.7)
32.9
(7.0)
8.8
(0.6)
NOTE: Test run averages are the average of the one-minute
      averages. The dat.a acquisition system scans each
      channel 180 times per minute.

 All results are reported on a dry basis.

 This value is not included in the average and is considered
 invalid.  A problem with the oxygen analyzer, such as
 condensation in the instrument, may have developed during
 Run 8.
                              2-18

-------
    u
                  02 VS TIME - RUN 09
                    NORTH ANDOVER METALS FESTlNG
                        11.5      U.5

                        24 HOUR CLOCK
                  02 VS TIME - RUN 08
                    NOWTH ANDOVEB METALS TESTING
                          11.5
                        24 HOUR CLOCK
                                                  Mean = 10.5% by volume
                                                  Stddev = 0.6% by volume
                                                  OBS = 209
                                    Mean = 8.3*% by volume
                                    Stcdev = 1.1% by volume
                                    OBS = 211
                  02 VS TIME - RUN 07
                    NORTH ANDOVER METALS TESTING
                      17            19

                        24 HOUR CLOCK
                                                   Mean = 11.5% by volume
                                                   Stddev = 0.7% by volume
                                                   OBS = 242
                                   OBS = Number of observations
                                      * Run 8 is considered
                                       invalid
Figure 2-2.
Oxygen Concentration History for Runs 7,8 and 9 at
        North Andover RESCO
                      2-19

-------
                     C02 VS TIME - RUN 09
                        NORTH ANDOVER METALS TESTING
                                                     Mean = 9.2% by volume
                                                     Stddev = 0.6% by .volume
                                                     OBS = 209
                           24 HOUR CLOCK
                     C02 VS TIME - RUN 08
                        NORTH ANDOVER METALS TESTING
                                                     Mean = 8.6% by volume
                                                     Stddev = 0.7% by volume
                                                     OBS = 211
                             11.5
                           2* HOUR CLOCK
                     C02 VS TIME - RUN 07
                        NORTH ANDOVER METALS TESTING
                         17            19

                           24 HOUR CLOCK
                                                     Mean = 8.5% by volume
                                                     Stddev = 0.6% byvolurru
                                                     OBS =  242

                                                    OBS = Number of observe
Figure 2-3.
Carbon Dioxide Concentration History for Runs 7,8 and 9 at
            North Andover RESCO
                                    2-20

-------
                      CO VS TIME - RUN 09
                        NORTH ANDOVER METALS TCSTlNG
                            11.5      12-5

                            24 HOUR CLOCK
                       CO VS TIME - RUN 08
                         NORTH ANOOVER METALS TESTING
                              11.5       12.5

                             24 HOUR CLOCK
                                                      Mean = 27,2 ppm
                                                      Std dev = 6.8 ppm
                                                      OBS = 209
                                                      Mean = 35.7 ppm
                                                      Std dev = 7.2 ppm
                                                      OBS = 211
             30 -

             25 -

             20 -

             15
                       CO VS TIME - RUN 07
                         NORTH ANDOVER METALS TESTING
                           17           19

                             24 HOUR CLOCK
                                                       Mean = 35.9 ppm
                                                      Std dev = 6.9 ppm
                                                       OBS = 242
                                        OBS = Number of observations
Figure 2-4.
Carbon Monoxide Concentration History for Runs 7,8 and 9 at
             North Andover RESCO
                                     2-21

-------
with no correlating changes in CCL or CO (Figures 2-3 and 2-4, respectively).
This indicates that a problem with the oxygen analyzer, such as condensation
in the instrument, developed during Run 8 and the result should be rejected.
                                      2-22

-------
                     3.0  PROCESS DESCRIPTION AND OPERATION
     This section contains a description of the incinerator process and air
pollution control system at the North Andover facility.   The incinerator and
ESP operating conditions during testing are summarized in Section 3.3.   The
operating data have been summarized as averages calculated over each test run
interval.  The original field data sheets are included in Appendix G.

3.1  PROCESS DESCRIPTION

     The North Andover facility, which began operation in 1985, consists of
two identical mass burn waterwall incinerators.  Each unit is designed to burn
685 Mg/day (750 ton/day) of municipal waste and produce 93,000 kg/hr (198,000
Ib/hr) of steam at 4,130kPa (600 psig) and 400°C (750°F).  Steam from both
boilers drives a 40 MW turbine-generator.  Figure 3-1 presents a diagram of
the North Andover process line.  Design data for the incinerator are
summarized in Tables 3-1 and 3-2.

     The refuse is neither shredded nor sorted before it is transferred by
overhead cranes from an enclosed pit to gravity-fed hoppers.  Hydraulic rams,
located at the bottom of the feed hoppers, charge the waste onto Martin
reciprocating grates.

     Underfire and overfire air is drawn from the pit area to fuel the
combustion process, which is designed to achieve temperatures in excess of
1370°C (2500°F).  Underfire air is supplied through the grates, and overfire
air is distributed through nozzles located on the front and rear walls above
the flame zone.

                                       3           3
     Each furnace has a volume of 820 m   (29,000 ft ), and each furnace/boiler
           o           2
has 4,900 m   (53,000 ft ) of heat transfer area.  Bottom ash  is quenched
before being combined with the boiler fly ash and ESP ash.
                                       3-1

-------
Secondary
  Fan
                                                                             Total Ash
                                                                             Discharge
                              Quench Tank
          Figure 3 -1. North Andover RESCO Process Line

-------
TABLE 3-1.  NORTH ANDOVER FACILITY STRUCTURAL DESIGN DATA
Chamber configuration
Primary chamber Secondary "cKamb'er
U)
UJ
Geometr i c
config-
uration
Rectangular

Geometric Heat transfer area Grate data
Volume, config- Volume, Radla-, Convec-, No. of Pressure Capacity,
ft3 uration ftj tive ft* five, ft* Type sections drop tons/d
29,000 NA 2,700 50,700 Martin reciprocating


-------
                                    TABLE 3-2.   NORTH  ANDOVER FACILITY AIRFLOW DESIGN DATA
Under fire air
No. of
plenums
No.
of con-
trol led
flows
Flow
rate,
acfm
Flow distribution, percent
Combus-
Feed Dry tion Burnout Location

Flow
direction
Overf ire air
Nozzle data
No. Type Velocity
                             50,000
Front wall  Horizontal    30     2 3/4" diameter




Back wall   Inclined     31     2 3/4" diameter
UJ

-------
     Each unit is equipped with an in-situ GEM system for carbon monoxide
(CO),  carbon dioxide (C0») oxygen (0_),  oxides of nitrogen (NO ),  sulfur
                        »—           £                         X
dioxide (S0_), and opacity.  The GEM units are located just downstream of  the
ESP outlet sampling location.

3.2  AIR POLLUTION CONTROL SYSTEM

     The air pollution control system consists of two identical ESPs designed
to reduce the particulate matter to a level of 115 mg/NM  (0.05 grain/dscf)  at
12 percent CO,- which corresponds to about 98 percent collection efficiency.
Design data for the ESPs are considered confidential by the ESP manufacturer.

3.3  INCINERATOR AND ESP OPERATING CONDITIONS DURING TESTING

     Incinerator and ESP operating conditions were monitored by plant
personnel in the control room.  The following incinerator process parameters
were recorded every 30 minutes: steam flow; steam drum pressure; superheater
(SH) outlet temperature and pressure; economizer inlet feedwater (FW)
temperature; economizer outlet FW temperature (east and west); gas temperature
entering SH; gas temperature exiting economizer; percent oxygen exiting
economizer; primary air temperature, pressure, and flow; forced draft (FD) fan
percent damper opening; secondary air temperature, pressure (front and rear
walls), flow, and fan percent damper opening; and opacity.  Table 3-3 presents
the average values of process data recorded by the plant.  The ESP electrical
power data are considered confidential by the ESP manufacturer.

     Table 3-4 presents the average GEM data from the plant's instrumentation.
Also shown in Table 3-4, for comparison, are the results from Radian CEMs.
The instruments used by the plant are in-situ analyzers which give results on
a wet basis.  The results have been correctd to a dry basis for equivalent
comparison to the Radian GEM results.  The difference in the results can be
attributed to the different sampling and analysis systems used.  Appendix G
contains copies of the recorded process, GEM, and ESP data sheets.
                                      3-5

-------
              TABLE 3-3.   AVERAGE PROCESS DATA  FOR NORTH ANDOVER INCINERATOR TESTS
                                    July 9 through 16,  1986

Date
07/08/86
07/09/86
07/10/86
07/11/86
07/12/86
07/13/86

07/14/86
07/15/86
07/16/86
Run
1
2
3
4
5
6

7
8
9
Steam load,
Ib/h x 10
166
165
166
166
166
163

170
168
165
Steam
drum
pres-
sure,
pstg
NR
670
674
680
680
667

680
680
680
S.
Temp.,
760
759
751
755
757
745

745
748
750
H. out
Pressure,
pslg
600
600
600
600
NR
600

600
600
600
Econ.
1n FW
temp.,
°F
225
225
226
226
226
225

227
226
226
Econ.
out FW
temp.,
°F
498
498
498
498
498
498

498
498
498
Gas
In
S.H., °F
1,233
1,281
1,245
1,257
1,267
[1,027]
[1,154]
1,217
1,262
1,126
temp.
Out
econ., F
569
570
547
562
562
[607]
[568]
585
578
586
NR = Not  recorded by  plant personnel,

-------
             TABLE 3-3.   AVERAGE PROCESS DATA FOR NORTH ANDOVER INCINERATOR TESTS
                                    July 9 through 16, 1986
                                            (cont'd.)

Prlmarv air
Date
07/08/86
07/09/86
07/10/86
07/11/86
07/12/86
07/13/86
07/14/86
07/15/86
07/16/86
Run
1
2
3
4
5
6
7
8
9
Temp.,
250
179.5
187.3
184.6
148.8
82.8
94.3
94.8
100.3
Pres-
sure,
1n w.c.
16.00
16.00
16.46
16.00
16.00
16.00
16.00
16.00
16.00
Flow,
ft /mln
NR
43.9
38.3
38.8
36.8
43.9
44.5
44.7
45.9
Pressure,
In w.c.
Front
17.4
16.5
16.5
16.3
16.0
16.4
16.1
16.1
16.5
Back
15.0
15.0
15.0
15.0
15.0
15.4
15.3
15.0
15.0
Secondary
Flow. ft3/m1n
Front
NR
NR
NR
NR
NR
62.4
66.5
69.3
69.1
Back
NR
NR
NR
NR
NR
41.3
41.4
40.2
40.1
Fan damp
opening,
percent
NR
35.3
NR
NR
NR
31.9
31.9
33.6
34.7
FD fan Refuse
damp Feed rate
open, (Buckets/
percent hour)
NR
18.5
15.0
15.3
15.5
17.6
17.3
18.9
18.9
12
12
12
12
13
12
NR
15
14
NR = Not  recorded by plant personnel.

-------
             TABLE  3-4.  AVERAGE GEM DATA FOR NORTH ANDOVER TESTS
                            July 9 through 16, 1986

Run
econ. out,
Date vol %
co, co2, so2,
ppm vol % ppm
NO , Opacity,
X
ppm %
Plant CEMa (wet basis)
Run 1
2
3
4
5
6
7b
8b
9b
Radian CEMC
7
8
9
07/08/86
07/09/86
07/10/86
07/11/86
07/12/86
07/13/86
07/14/86
07/15/86
07/16/86
(dry basis)
07/14/86
07/15/86
07/16/86
8.4
8.4
7.8
8.5
8.1
8.6
8.3
(9.5)
8.4
(9.7)
8.6
(9.9)

11.5
8.3d
10.5
.-
44 9.1 18
36 10.6 32
25 10.5 40
11 10.4 37
28 9.9 42
48 8.7 46
(55) (10.1)
46 9.1 27
(53) (10.5)
..

35 8.5
36 8.6
27 9.2
NR
151 0.10
162 0.12
167 0.12
184 0.13
162 0.55
172 0.31
189 0.14
0.13

--
..
--
 Plant CEMs except for the 0_ analyzer were located about 10 ft downstream of
 the ESP outlet sampling locations.  The plant used in-situ infrared analyzers
 and results are reported on a wet basis.
 The results on a dry basis are reported in parentheses below the wet basis
 values.  The moisture content of the flue gas was determined by EPA Method 4.

CRadian CEMs were extracted at the ESP outlet sampling location.  Radian did
 not measure SO-,  NO  or opacity.  Radian used an extractive system and
 results are reported on a dry basis.

 value considered invalid due to an instrument malfunction.

NR = not recorded by plant personnel.
                                      3-8

-------
     Some operating problems did occur during testing.   A broken grate block
developed during the middle of Run 5.   The underfire air ports were cleaned
manually during the remainder of the test run.   Normal operating conditions
were maintained, and the incinerator was brought down for repair immediately
after completion of Run 5.

     Operating problems are believed to have occurred during Runs 6 and 7.
Signal Environmental Systems determined that the incinerator was still in the
start-up mode during Run 6.  The particulate and metals data from Run 7
indicate that the ESP may have malfunctioned during Run 7.
                                      3-9

-------
                            4.0  SAMPLING LOCATIONS
     The locations of the sampling points are shown on the schematic of the
process line in Figure 4-1.  Each sampling location is discussed in the
following paragraphs.

4.1  ESP INLET SAMPLING LOCATION

     The ESP inlet sampling location is shown in Figure 4-2.   PCDD/PCDF
(Runs 2 through 6) and metals/particulate (Runs 7, 8, & 9) samples were
obtained at this location.

     The breeching at this point is rectangular, measuring 136 inches wide and
76 inches deep.  Eight sampling ports with 30 inch long nipples are evenly
spaced 1 foot 5 inches apart across the top of the duct in the plane
perpendicular to the direction of flow.  The equivalent diameter of this duct
is 97.5 inches for the purpose of selecting the required number of sampling
points by EPA Method 1.  ESP inlet testing ports are located approximately
40.0 feet (4.9 duct diameters) downstream of a 30  bend in the breeching, and
approximately 6.5 feet (0.8 duct diameters) upstream of the ESP inlet.  EPA
Method 1 required a minimum of 24 points for this location.

    The sampling point locations and dimensions of the duct are shown in
Figure 4-3.  A 48 point test grid was used to traverse this duct.
The ratio of length to width of the sample areas was 1 to 1.3.  Twice as many
traverse points than were required by EPA Method 1 were used in order to
obtain a length to width ratio close to one.

4.2  ESP OUTLET SAMPLING LOCATION

     The flue gas was sampled at the ESP outlet breeching prior to the induced
draft fan.  The sampling port locations and dimensions of the duct are shown
                                       U-l

-------
Secondary
  Fan
                                                       ESP Inlet flue Gas
                                                       Sampling Location
                                  Superheater
                                          Generator
                                                  Economizer
                                                                      ESP
-^


\
J


ESP Ash
Sampling Location
ESPOull
Flat G»
Sampling Lo
zn ^
Vibrating
Conveyor
                                                                                      Total Ash
                                                                                      Discharge
                                  Quench Tank
  Figure 4-1.  North Andover RESCO Process Line with Sampling
                 Locations

-------
Platform
                     Port                      Port
                     ABCDEFGH,
                              Top View
                                                     Ports
                               Platform
                                      6' - 4
ESP-
                              -40'-0"-
                               Side View
             Figure 4 - 2.   ESP Inlet Sampling Location
                                  U-3

-------

1
3
i
6n

19" 	

4*ii

45" 	
58" —
71" 	
7ft"



3"








^—^m



L












A


• i



• >
• i
• t



J
A
*1
*,

»
*J
V,
*,
^



L'






<
(
4


••
B

•
• I



il
il
il


i—1"
1

i
3,


3
34
3,
3,


T" 	
L











«
(


«



4
«
«



\
.(

»c

»f

•c
>c
>c



p

1
'2

%
'J
•«
'!
•e
















L












D
- 1

• E

~ i
• i
• C
•c
•c



J

1
),


3
>4
>,
1.

136

L












E
_ i
• i
• E

* I

• E
• E
• E



J

-1
^j


•3
14





L












F
_ 1
• I
• I

— fl
• 1
•
• I
• 1



P

1



•3
F4

•



C












G

• <
• I


•
•(
•(
• (



P

i
3i


3
34
3»
3.



L J
H

• H,
• H2

. LJ
• Hj
• H4
• H,
• H,


Figure 4 - 3.  Traverse Matrix for ESP Inlet Sampling Location

-------
in Figure 4-4.   TOCL/particulate (Runs 1 through 6),  metals/particulate  (Runs
7, 8,  and 9) and GEM samples were obtained at this location.

     The breeching is rectangular, 124 inches wide by 76 inches deep.   Six
sampling ports with 8-inch nipples are installed at the sampling location.
The equivalent diameter of the duct is 94.2 inches for the purpose of
selecting the required number of sampling points by EPA Method 1.  The test
ports are located approximately 16 feet (2.0 duct diameters)  downstream of a
90  bend, and approximately 4.1 feet (0.5 diameters) upstream of the ID fan
inlet.  EPA Method 1 required at least 24 traverse points for this location.
     A 24 point test grid was used, which is shown in Figure 4-5.  The
ratio of length to width of the sample areas was 1 to 1.1.

4.3  ESP ASH SAMPLING LOCATION

     The ESP ash samples were collected from the screw conveyor that
transported the ash from the ESP hopper for Unit #2 to a discharge point on
the bottom ash conveyor belt.  The ash handling system for Unit #2 was
completely independent from the ash handling system for Unit #1.  The ESP ash
handling system and sampling location is shown in Figure 4-6.  A plate was
removed at the bottom of the screw conveyor and a sample was collected as the
ash fell from the conveyor.
                                        H-5

-------
                                        Gas Flow from ESP
                                           Straightening Vanes
                                                  _L Sampling
                                                    Platform
Figure 4 - 4.  ESP Outlet Sampling Location at North Andover Facility
                               U-6

-------
                                       Top View
tn
x:
••—i
Q.
0)
O
*-•
c

"o

t

5.5"

t
47.5"




1 \

1







1
28 fi*


i



T"
9.5"
1 i
^ 	
^ o-i n
* 21.0


'••


3









F2-
F,





.


F





- <:u.a


E4«
-
E2»
E,





.


E





r™ 
-------
                          ESP Hoppers
co
                                                   Sampling
                                                   Location
                 Figure 4 - 6. ESP Ash Handling System and Sampling Location

-------
                    5.0  SAMPLING AND ANALYTICAL PROCEDURES
     The sampling and analytical procedures used at North Andover RESCO  were
the most recent revisions of the published methods.  In some cases,  the
methods were modified to incorporate the most recent developments which  have
been accepted by the sampling community.  In this section, brief descriptions
of each sampling and analytical method are provided.  The detailed sampling
and analytical methods with modifications are included in Appendix K.  A
summary of the sampling and analytical methods used for each parameter is
presented in Table 5-1.

5.1  PCDD/PCDF SAMPLING AND ANALYSIS

     PCDD/PCDF sampling followed the December 1984 draft protocol for the
determination of chlorinated organic compounds in stack emissions.  The
protocol was developed by the Environmental Standards workshop sponsored by
the American Society of Mechanical Engineers (ASME) and EPA.  The method is
based on EPA Reference Method 5.

     The PCDD/PCDF analysis followed the ASME/EPA analytical procedures  to
assay stack effluent samples and residual combustion products for PCDDs  and
PCDFs also dated December 1984.  The sampling and analytical procedures  with
modifications are discussed in the following sections.

5.1.1  Equipment and Sampling Preparation

     In addition to the standard EPA Method 5 requirements, the PCDD/PCDF
sampling method includes several unique preparation steps which ensure that
the sampling train components are not contaminated with organics that may
interfere with analysis.  The glassware, glass fiber filters and XAD resins
are cleaned and checked for residuals before being packed.
                                        5-1

-------
           TABLE 5-1.  SAMPLING METHODS AND ANALYTICAL PROCEDURES
    Parameters
     Sampling Method
     Analytical Method
Dioxins/furans


Particulates

Metals

02> C02, CO, THC

Molecular weight

Moisture

Velocity

Temperature

ESP Ash
Environmental Standards
Workshop (Dec. 1984)

EPA Method 5

Alternative Method 12

Extraction

EPA Method 3

EPA Method 4

Method 2

Type K thermocouple

Composite grab sample
High resolution GC/MS


Gravimetric

Neutron Activation Analysis

Method 3A and 10 (CEM)

Orsat apparatus

Analytical balance
PCDD/PCDF: Environmental
Standards Workshop Protocol
(Dec. 1984)
High resolution GC/MS
Metals:  Neutron Activation
        Analysis
                                       5-2

-------
     The protocol was modified by replacing hexane with methylene chloride
during all aspects of sampling preparation and recovery.   Methylene chloride
was shown to recover the higher chlorinated PCDD/PCDFs better than hexane
during the EPA Tier 4 program.

     The glassware is washed in soapy water, rinsed with tap water, baked and
then rinsed with acetone and methylene chloride.   Once in the field, the train
glassware is assembled and rinsed with methylene  chloride.  The rinses are
analyzed for residual PCDD/PCDFs.

     The XAD resin and glass filber filters were  extracted in HPLC grade
water, methyl alcohol, methylene chloride and hexane, sequentially.  At the
conclusion of the soxhlet extractions, a 500 ml aliquot of the final solvent
rinse (hexane) was concentrated to 5 ml.  A 500 ml aliquot of fresh hexane  of
the same lot was also concentrated as a blank.  Both aliqouts were analyzed by
GC/FID for determination of Total Chromatographable Organics (TCO).  A
standard mixture of n-hydrocarbons that cover the TCO range of boiling points
which is from 100°C (C  - bp 98°C) to 300°C (C-  - bp 303°C) was used to
quantify the TCO results in the aliquot.  The results were reported as
milligrams of TCO per resin weight prepared (mg/g), corrected for the blank.
The results for the XAD resin used at North Andover are included in
Appendix J.2.  The XAD resin was packed in glass  traps and the filters in the
glass petri dishes for transport to the field.

     The remaining preparation includes calibration and leakchecking of all
the train equipment.  This includes meterboxes, thermocouples, nozzles, pitot
tubes, and umbilicals.

5.1.2  Sampling Operations

     The PCDD/PCDF sampling method uses the sampling train shown in
Figure 5-1.  Radian has modified the protocol to include a horizontal
condenser rather than a vertical condenser.  The horizontal condenser lowers
the profile of the train and reduces breakage.  The XAD trap following the
condenser is still maintained in a vertical position.

                                      5-3

-------
        Thermocouple

     "S" Type Pltot
T                  Filter Holder

        hermocouple—f
               xKcH~
         Probe
                      Thermocouple Thermocouple

                                       f    Check Valve
Stack Wall
         Pltot
       Manometer
I
-1=-
                    Reclrculatlon Pump
        Water Knockout
           Impinger
Thermocouples -»-Q ("^
      orifice   nOr
100ml HPLC Water

By • Pass
 valve
                                                             Main Valve
                                                                         Silica Gel
                                                                        (300 grams)
                                                      Air-Tight
                                                       Pump
                                                                       Vacuum Line
                       Figure 5 -1.  PCDD/PCDF Sampling Train Configuration Used at
                                   North Andover RESCO

-------
     The sampling locations at North Andover RESCO had six to  eight ports  and
to insure that the train remained intact during the port changes,  the  trains
were leakchecked after sampling in each port.  The trains were also
leakchecked at the start and finish of sampling as required in EPA Method  5.
In the event that the leakrate was found to be above the minimum acceptable
rate (0.02 ft /min) after a port, the sample volume was corrected for  that
port as specified in the sampling protocol.  This calculation is shown in  more
detail in Appendix A.5, but basically, the excessive leakrate was reduced  by
the minimum acceptable rate and then multiplied by the sampling time of the
port.

     At the ESP inlet, the duct was traversed following EPA Method 1.   The
traverse diagrams were shown previously in Figure 4-3.  Sampling was conducted
simultaneously with the ESP outlet.  Since the ESP outlet had a different
traverse matrix, sampling at the ESP inlet was stopped during the ESP outlet
port changes.  During these periods, the pump was turned off, and the train
rotated 180 degrees.

     Sample was collected for 4 minutes per point for a total of 192 minutes
for Runs 1-3 and for 5 minutes per point for a total of 240 minutes for
Runs 4-6.  The flowrate of the flue gases and the nozzle diameter required the
sampling rate to be an average of 0.4 dscfm to insure isokinetic sampling.
The sampling time was increased to insure that the sample volumes would be
significantly greater than 90 dscf which was the volume needed for the outlet
location based on the minimum detection limit.

     Other sampling operations that are unique to PCDD/PCDF sampling include
maintaining the gas temperature entering the XAD trap below 68 F.  The gas is
cooled by the condenser and, the XAD trap, itself, has a water jacket in which
ice water is circulated.  Otherwise, sampling followed EPA Method 5
specifications.
                                       5-5

-------
 5.1.3   Sample  Recovery

     To facilitate  transfer  from  the sampling location to the recovery train,
 the  sampling train  is disassembled  into four sections: the probe liner, the
 XAD  trap and condenser,  filter holder, and the impingers in their bucket.
 Each of these  sections are capped with methylene chloride-rinsed aluminum foil
 before  removal to the recovery trailer.  Once in the trailer, field recovery
 follows the scheme  shown in  Figure  5-2.  The liquid fractions were recovered
 into amber bottles  which helped to  protect the samples from light damage.
 Recovery results in the  sample components listed in Table 5-2.  The sample
 fractions are  shipped to the analytical laboratory via truck.

 5.1.4   Analysis

     The PCDD/PCDF  analyses  for the North Andover RESCO samples were performed
 by Triangle Laboratories, Research  Triangle Park, North Carolina.  The sample
 components were extracted, combined and concentrated according to the scheme
 in Figure 5-3.  Isotopically-labelled internal standards and surrogate
 compounds were added to  the  samples prior to the extraction process begins.
 The  internal standards included 2378-13Cl12-TCDD, 13C112-PCDD, 13Cl12-HpCDD
 and   C11_-OCDD.  Once added to the samples, the internal standards went
 through the entire  extraction process and were measured on the GC/MS.  Then
 the  recoveries of the internal standards were determined and the results of
 the  native species  were  adjusted  according to the internal standard
 recoveries.  The surrogate compounds which included   C119-TCDF,   C1.._-TCDD,
     13
 and   Cl  ,,-HxCDF were added  in a  similar manner, but the surrogate recoveries
 were not  used  to adjust  the  results of the native species.  Surrogates provide
 additional information of the extraction efficiency of the analytical
 procedure.

     The  samples were analyzed as front half and back half fractions.  This
was a modification  requested by EPA.  The validity of this separation for the
purpose of quantifying filterable and gaseous PCDD/PCDF is questioned by some
of the sampling community.  However, the purpose of this modification was to
                                       5-6

-------
VJl
I

Filter Support
and Back Half
Front Halt of Filter Slhlmplnaer
Probe Liner Cyclone Filter Housing Filter Housing Condenser XADTrap 1st Implnger 2nd Implnger 3rd Implnger 4lh Implnger (silica gel)
f
\ • 1 (I)) -r~ i i i i
Attach 250ml Brush and Brush and Carefully Rinse with Rinse with Secure XAD ? Weigh Weigh Weigh Weigh
M!. rinse with rinse with remove filter acetone acetone trap openings Weigh Implnger Implnger Implnger Implnger
lolnt acetone until acetone from support (3x) (3x) with glass Implnger
Rin»» nh paniculate Is (3 x) with tweezers balls and £mpiy Empty Empty Discard
Hinsewlth removed Rinse with Rinse with clamps E">P'y contents Into contents Into contents Into silica gel
acetone, Methylene Brush loose methylene methylene contents Into bottle bottle bottle
emptyflaak Then rinse Chloride paniculate chloride chloride Place In bottle, weigh
Into 950ml 3 limes with <3x) onto filler (3x) <3x) cooler for Rinse with Rinse with Rinse with
bollla methylene
Brush liner chlorld9
and rinse with
3allquots
of acetone
Check liner
to see If
paniculate Is
removed, If
not repeat 3
Rinse with
3allquots"
of methylene
chlo
ride













Seal In petrl
dish

































1
slo











rage Rinse with acetone acetone acetone
acetone (3X) <3x) (3x)
(3 x)
Rinse with Rinse with Rinse with
Rinse with methylene methylene methylene
methylene chloride chloride chloride
chloride (3X( (3x) (3x)









































                     PR
                                                  CR
                                                              SM
                                                                       CO
                                                                                        IR
                            Aliquot equals about 100ml. Empty all rinses into one 950ml bottle.
                            3 x = three times
                                Figure 5 - 2.   PCDD/PCDF Field Recovery Scheme

-------
          TABLE 5-2.  PCDD/PCDF SAMPLING TRAIN COMPONENTS FOR NORTH
                      ANDOVER RESCO SHIPPED TO ANALYTICAL LABORATORY
   Container/Component    Code
                     Glassware
 1. Component Number 1     F

 2. Component Number 2     PR



 3. Component Number 3     CR


 4. Component Number 4     CD
     Filter(s)

     Rinses3 of nozzle,  probe,  transfer line  (if
     used),  cylcone,  and front  half of filter
     holder

     Rinses  of back half of filter holder,  filter
     support and condenser

     First impinger contents and rinses
 5. Component Number 5
 6. Component Number 6     SM   XAD-2  resin
IR   Second,  third and fourth impinger
     contents and rinses

          R
an_-
 Rinses include acetone and methylene chloride combined in same container.
                                     5-8

-------
           PR
IR
CD
SM
CR
    •For large volumes, organic solutions are concentrated using a rotary evaporator apparatus
     For small volumes, samples are concentrated under nitrogen


Figure 5 - 3. PCDD/PCDF Analytical Scheme Used at North Andover

             RESCO
                                  5-9

-------
 study  the  internal  standards recovery of the front and back half fractions,
 since  the  different matrices of  the fractions may effect the recovery of the
 internal standards.

     The front half of the sample train consists of the probe rinse, cyclone,
 front  half filter holder rinse,  and the filter.  The back half includes the
                                                        p
 back half  filter holder rinse, coil condenser rinse, XAD  trap, and impinger
 contents and rinses.

     The samples were analyzed by high resolution gas chromatography  followed
 by high resolution  mass spectrometry (GC/MS).  The instruments are calibrated
 daily  with external standards.   The GC/MS results are recorded and stored onto
 a computer file.  The computer is used to reduce the data and to summarize the
 results such as amount detected, detection limit, retention time and internal
 and surrogate recoveries.  The analytical report from Triangle Laboratories is
 included in Appendix D.I.

 5.1.5  Data Reduction for PCDD/PCDF Results

     The data reduction for PCDD/PCDF results began with correcting for the
 reagent blank results.  Then the concentrations of PCDD/PCDFs in the flue gas
 are calculated by dividing the nanograms of analyte by the volume of flue gas
 collected.  The volume of flue gas has been corrected to EPA standard
 conditions and is reported in dscm units.

     To normalize the concentrations to 12 percent CCL, the ratio of 12 to C0?
measured in the flue gas, is calculated for each run based on EPA Method 3
results.  Then the  concentrations from each run are multiplied by the
respective normalization ratio.  The normalized concentrations are multiplied
by 2378-TCDD toxic  equivalency factors to obtain 2378-TCDD toxic equivalents.

     Mole fractions of each isomer are calculated by dividing the
concentration of each isomer by  its molecular weight.  The moles of each
isomer are summed for each run, and the fraction is obtained by dividing the
                                     5-10

-------
moles for each isomer by the total moles.   All of the PCDD/PCDF calculations
described above are done on a Lotus 1-2-3  spreadsheet and are also verified by
hand.  The isomer distribution plots (Figure 2-1) are also prepared using the
spreadsheet.  The calculations are included in Appendix A.5.

5.2  FLUE GAS TRACE METALS/PARTICULATE DETERMINATION

     Gas sampling and analysis for metals  and particulates was performed
according to EPA Alternate Method 12, "Determination of Inorganic Lead
Emissions from Stationary Sources."

5.2.1  Equipment and Sampling Preparation

     Equipment and sampling preparation for metals/particulate sampling
followed the specifications for EPA Method 5 with some additional glassware
preparation.  The glassware preparation took place on-site since the same
glassware from Runs 1, 2, and 3 were used.  The train glassware was washed in
soapy water, rinsed with tap water, rinsed with distilled water, rinsed with
0.1 N nitric acid, rinsed with distilled water, and then dried with acetone.
                                           p
The glassware was then capped with Parafilm .  Before the glassware was used
for sampling, the glassware was recovered with 0.1 N nitric acid, and the
rinses were analyzed for trace metals.

5.2.2  Sampling Operations

     The metals and particulate sampling protocol followed EPA Alternate
Method 12, where both particulate matter and metal samples were collected
using the same train.  Alternate Method 12  is a variation of EPA Method 5
where 0.1 N nitric acid replaces water in the impingers,  and filter media with
a low lead background are used.  The metals/particulate  sampling  train is
shown in Figure 5-4.  An EPA Method  12 train usually has  a Greenburg-Smith
impinger as the second impinger.  The outlet  train during Runs  7,  8 and  9 used
two modified-tip  impingers  in place  of the  Greenburg-Smith impinger.  For Run
7, the contents and rinses  from the  1st and 2nd  impingers were  recovered and
                                     5-11

-------
                                  For the ESP Inlet
                               Train, Greenburg-Smith
                                Impinger was used as
                                the second Impinger

                         /           Filter Holder

                          Thermocouple —f~
          Thermocouple

        'S"TypePltot
                                        Extra Impinger only used
                                        in the ESP Outlet Train
                                           Thermocouple

                                               /   Check Valve
H
IX)
Stack Wall    /
       Pltot
     Manometer
                          Thermocouples
                                 Orifice
                                                  Air-Tight
                                                   Pump
                                                                                     Vacuum Line
                 Figure 5 - 4. Metals/Particulate Sampling Train Configuration Used
                             at North Andover RESCO

-------
analyzed separately from the 3rd and 4th impingers.   This  modification was
made due to insufficient number of Greenburg-Smith  impingers  on-site.

     With respect to leakchecks and traversing,  the  particulate/metals trains
followed the same operations as previously discussed in Section 5.1.2.

5.2.3.  Sample Recovery

     To facilitate transfer from the sampling location to  the recovery
trailer, the sampling train was disassembled into three sections;  the probe
liner, the filter holder and the impingers in their bucket.   Each section was
                    p
capped with Parafilm  before being transferred to the recovery trailer.  Once
in the trailer, the field recovery followed the scheme shown in Figure 5-5.
First, the probe line was rinsed and brushed with acetone  until particulate
was qualitatively removed from the probe liner.  The acetone rinse was placed
in a borosilicate glass sample bottle.  The impingers were weighed and the
contents placed in polyethylene sample bottles.  The impingers were rinsed
with 0.1 N nitric acid and the rinses added to the impinger contents.  For the
Run 7-outlet train the first two impingers were recovered into a separate
bottle than the third and fourth impingers.  For Runs 8 and 9 all the
impingers were recovered into the same sample bottle.  The filter was sealed
in a petri dish.  The sample components shipped to the laboratory are listed
in Table 5-3.

     The dessication and evaporation of the particulate sample was performed
at the laboratory at Radian/RTP.  Although the nozzle was recovered for
particulate analysis, the rinse was kept separate, and was analyzed for
particulates separately.  The nozzle was not rinsed with nitric acid  and was
not included in the metals analysis because of possible contamination.

5.2.4  Analysis

     The analysis for the metals/particulate samples was executed in  two steps
which are shown in the bottom half of Figure 5-5.  First,  the acetone rinses
                                     5-13

-------
            Front Hall Sample Recovery Fractions
Back Half Sample Recovery Fractions
Nozzle
1
Brush/Rinse
with Acetone


Evaporate ;
Desslcate; Weigh
\
Participate (mg)
vn
H
Probe liner, Cyclone, or Cyclone by
1 1
No. 1 . Brush and rinse with No. 2. Rinse with three
three allquots of acetone allquots of 0.1N HNO,
±i

Evaporate ;
Desslcate; Weigh
1
Participate (mg)
I
Dissolve Residue
with 0.1N HNO,

i
Evaporate to less
than 50 ml
1
Make up to 50 ml
with water In
volumetric flask
i
Place In 50 ml bottle
for NAA Analysis

i
Metals (ug)
pass Filter
1
No. 3. Rinse with acetone
to dry, discard rinse


^ Field Recovery |T
• ' • 1
^ Laboratory ^ |
' Preparation » •
i !
Desslcate ; 1
Weigh |
1 1
Paniculate (mg)
1
Place filter In
bottle for
NAA Analysis



                                                                                           1st, 2nd, 3rd Implngers
                                                                                                Weigh
                                                                                           Rinse three times with
                                                                                               0.1 N HNO,
                                                                                             Evaporate to less
                                                                                               than 50 ml
                                                                                             Make up to 50 ml
                                                                                              with water In
                                                                                             volumetric flask
                                                                                            Place In 50 ml bottle
                                                                                             for NAA Analysis
                                                                                               Metals (ug)
                                                                                                                 4th Implnger
                            Weigh
                                                                                                               DiscardSlllcaOe,
Figure 5 - 5.   Sampling and Analysis Protocol for Metals/Particulate Samples

-------
       TABLE 5-3.  SAMPLE COMPONENTS  FOR PARTICULATE/METALS  TRAIN
Component        Code                      Glassware
              PR - Acet    Acetone rinses  of probe  liner  cyclone  or  cyclone
                           bypass, and front half of filter  housing

              PR - HNOo    0.1 N Nitric rinses  of probe liner  cyclone  or
                           cyclone by-pass,  and front half of  filter
                           housing

                  F        Filter
                 IR        0.1 N Nitric rinses  and  contents  of impingers,
                           also rinses  of back  half of filter housing,
                           filter support,  and  zigzag.
                 NZ        Acetone rinse of nozzle
                                  5-15

-------
and filter were dessicated and weighed for the particulate analysis.  Then,
the particulates were resuspended in 0.1 N Nitric acid and combined with the
nitric rinses from the front half.  The front half and back half fractions
were evaporated separately and placed in separate bottles, so that the samples
were analyzed as front half and back half fractions.   Front half/back half
analyses were performed for the purposes of the EPA study.  The sample
preparation and particulate analysis was completed by Radian/RTP.

     Following the particulate analysis, the samples were analyzed for metals
using NAA.  The analysis was performed by the Nuclear Services Laboratory at
North Carolina State University in Raleigh, N.C.  NAA can be used to analyze
for all the specific metals except lead and beryllium.  Also, the method does
not differentiate between different valence states or compounds of a metal
such as Cr (III) or Cr (VI).

     During the NAA procedure, the samples are exposed to neutrons.   The
neutrons excite the metal atoms and cause them to emit gamma rays.  The
density and wavelength of the gamma rays are measured and the information is
logged by a computer.  Calibration standards with known amounts of the metals
are also included in the sample batch, and by comparing the results from these
calibration standards to those of the samples, the type and concentration of
each metal detected is determined.

     However, since lead and beryllium do not emit gamma rays, NAA cannot
measure these metals.  Also, NAA hay have interferences with certain metals.
Depending how mercury is bound in the sample, the mercury results may be
variable.   The sample is heated during the exposure to neutrons and may cause
the mercury to vaporize and migrate out of the polyethylene sample bottles.
High levels of sodium (factors of 1000) may cause interference with the metals
nearby in the spectrum such as arsenic, bromine, potassium, antimony, samarium
and cadmium.   If the sodium level can be estimated in the sample, the exposure
time can be adjusted to minimize the interference.  The analytical report from
the Nuclear Services Laboratory is included in Appendix D.2.
                                       5-16

-------
5.2.5  Data Reduction For Metals/Particulate Results
     The particulate loading was calculated using the following equation.
(M -M, )        + (M -M. )
  p 1j/filter   v  p V
                                       .
                                      rinses
                         v
                          m(std)
          G  = particulate loading (mg/dscm)
          M  — mass of particulates (mg)
          M,  •— mass of the blank (mg)
          V    — volume of gas sampled at standard conditions
                 (dscm @ 1 atm and 68°F)
     Metals concentrations in the flue gas were determined in ug/dscm with the
following equation:
          C  .  - - (C° metal - C. )
           metal           	fcn
                        m(std)
where:
       C  — amount of metal detected in sample (ug)
       C,  — amount of metal detected in the blank (ug)
       V  .   .  — volume of gas sampled at standard
                 conditions (dscm @ 1 atm, 68 F)
     The metals concentration are normalized using the same procedure
described for PCDD/PCDFs in Section 5.1.5.
5.3  CEM SAMPLING AND ANALYSIS

5.3.1  Equipment and Sampling Preparation

     Each component of the CEM system was cleaned, leakchecked and calibrated
before going into the field.  The components of the system include the probe,
the umbilical, the gas conditioner and pump, the manifold, the analyzers,  the
computer data logger, and the strip chart recorder.

     The probe was checked for leaks, and the filter was  inspected.   The
umbilical consists of teflon tubing which has been wrapped in heat tracing
                                     5-17

-------
 within a  protective  sheath.  The umbilical was cleaned with hot water, and
 then dried and blanked with nitrogen.  The umbilical was also leakchecked.
 The  gas conditioner  pump, and manifold were leakchecked, cleaned, and the
 cooling operation was  checked prior to operation.

      A three point calibration was performed on each of the analyzers.  The
 calibrations also tested the operation of the computer data logger and
 stripcharts recorder.  A correlation coefficient was calculated from the three
 points  to  check the  linearity of the response of the analyzer.  (The criterion
 is discussed in Section 6).  Once the system is prepared, it was disassembled
 and  packed carefully for transport to the site.

 5.3.2   Sampling Operations.

     A  schematic of  the CEM system used at the North Andover facility is shown
 in Figure  5-6.  Before each run, the system was leakchecked and a system blank
was  analyzed.  Each  of the analyzers was calibrated with a commercially
prepared and certified zero and span gas and the response factor was
determined.  The calibration gases are introduced at the manifold.  The
calibration results  were transmitted to the data logger which also recorded
the  data during the  sampling run.  The data logger scanned each analyzer 180
times per minutes, and recorded one-minute averages.  The data logger stored
the  one-minute average instrument response from each analyzer on disk, and
also converted the response to a concentration using the initial calibration
data which was printed out as a hard copy.  The instrument responses were also
recorded on a stripchart as a back-up to the data logger.  At the end of the
sampling run, another zero and span calibration was performed.  The final
calibration was used to determine the daily drift of the instrument.  The
listings of the one-minute averages, stripcharts, and calibrations are
included in Appendix B.

5.3.3  Data Reduction

     The data reduction for the CEM results were performed on computer.  The
exact equations used are included in Appendix A.5 and are discussed in general
                                      5-18

-------
 Flue Gas
from Stack
vn

H
VD
           Instack Probe
             with Filter
                Heat Traced Teflon Umbilical
                    (50 ft long) (>120 °C)
                                                   Gas Conditioner
                                                  and Pump (0-5 °C)
                                                                         Overflow
                                                                                                             Exhaust
                                                                                                         .Electronic Signal
                                                                           Located In Instrument Trailer
                                             Figure 5-6.   GEM Analysis Scheme

-------
in this section.  At the end of the sampling run, the daily drift of each
instrument was evaluated by comparing the initial and final calibrations.
Then, each one-minute average was adjusted by assuming the drift was linear
throughout the day.  Also, any invalid sections of the data (such as blowbacks
of the instrument lines to prevent moisture from plugging the lines) were
deleted before the averages and standard deviations were calculated for each
parameter.

5.4  MOLECULAR WEIGHT DETERMINATION BY EPA METHOD 3

5.4.1  Sampling Operations

     The molecular weight of the flue gas was determined using EPA Method 3.
During the flue gas sampling an integrated bag sample was extracted at a
single point from the sampling location at a rate of 0.5 ml/min for a total
sample volume of approximately 1.5 cubic feet.  The sampling train used is
shown in Figure 5-7.  The moisture knockout trap allowed the analysis to be on
a dry basis.

5.4.2  Analysis

                                                           TM
     The integrated bag samples were analyzed with an Orsat   analyzer for CO-
and 0-.  Nitrogen was determined by difference.  Triplicate analyses were
performed on each bag, and an average was used as the input for the Method 5
calculations.  The absorbing solutions used in the analyzer were potassium
hydroxide (45% by volume) for carbon dioxide and pyrogallate (pyrogallol in
potassium hydroxide) for oxygen.  The Orsat data sheets are included in
Appendix C.7.

5.5  VOLUMETRIC FLOWRATE BY EPA METHOD 2

     Volumetric flowrate was measured according to EPA Method 2.  A type K
thermocouple and S-type pitot tube were used to measure flue gas temperature
and velocity, respectively.
                                     5-20

-------
             '/4" Stainless Steel
                  Probe
Tubing
vn
                                          Pump

V ' '

) 1 1
^ Ice Water Cooled
Moisture Knockout Trao


Tedlar Bag


                                Figure 5 - 7.  Method 3 Integrated Bag Sampling Train

-------
 5.5.1  Sampling and Equipment Preparation

     For EPA Method 2, the pitot tubes were calibrated before use following
 the directions in the method.  Also, the pitots were leakchecked before and
 after each run.

 5.5.2  Sampling Operations

     The volumetric flowrate data were recorded simultaneously with the other
 sampling train data.  The parameters that were measured include the pressure
 drop (  P) across the pitots, stack temperature, stack draft and ambient
 pressure.  These parameters were measured at each traverse point.  Based on
 EPA Method 2, a computer program was used to calculate the average velocity
 during the sampling period.  The calculations are included in Appendix A.5.

 5.6  MOISTURE DETERMINATION BY EPA METHOD 4

     The average flue gas moisture content was determined according to EPA
Method 4.  Before sampling, the initial weight of the impingers were recorded.
When sampling was completed, the final weight of the impingers was recorded,
and the weight gained was calculated.  The weight gained and the volume of gas
sampled were used to calculate the percent by moisture of the flue gas.  The
calculations are performed by computer, and a sample calculation is included
 in Appendix A.5.

5.7  ASH SAMPLING FOR METALS ANALYSIS

     A standard protocol was not available for sampling the ESP ash at North
Andover RESCO.  However, ASTM D2234-825, a method used in collection of a
gross sample of coal, was used for guidance in terms of sampling method,
frequency and size.   The method does not require that all grabs be of the  same
volume or weight, just that a minimum number and minimum size be taken.
                                       5-22

-------
5.7.1  Sampling Operations

     ESP ash samples were collected from the drag conveyor at an intermediate
transfer point in the screw conveyor system prior to being combined with the
other ash streams.  Since the ash was analyzed for metals, all plastic
sampling equipment was used.  Using the ASTM Method as a guideline and
considering the homogeneous nature of the ESP ash, the samples were collected
every 30 minutes starting 45 minutes after the start of flue gas sampling.
Approximately 4 pounds of ESP ash were collected during each test run which
were composited and an aliquot placed in a 950 ml amber glass bottle.  A
sampling scheme is shown in Figure 5-8.

5.7.2  Analysis

     The ESP ash was analyzed by NAA.  The analysis was the same as previously
described in Section 5.2.4 for the flue gas samples.  The NAA method does not
require any sample preparation for ash samples.  The ash which is placed
directly in a prepared sample bottle is ready for irradiation.
                                      5-23

-------
                   Grab No. 1
Grab No. 2
Grab No. 3
Grab No. 4
VJl
I
ro
      Comblne Grab Samples
          for Composite
                                                   Place 50 ml In
                                                   sample bottle
                                                   NAA Analysis
                                                    Metals (ug)
                           Figure 5 - 8.   ESP Ash Sampling and Analysis Scheme

-------
              6.0  QUALITY ASSURANCE AND QUALITY CONTROL (QA/QC)
     Quality assurance/quality control guidelines  outline  pertinent steps
during the production of analytical and emission data to ensure  and quantify
the acceptability and reliability of the data generated.   The measures
outlined in this section were followed to ensure the  production  of quality
data from the sampling and analytical efforts.

6.1  STANDARD QUALITY ASSURANCE PROCEDURES

     QA/QC procedures were followed during sampling and analysis to ensure
that the data generated were of acceptable quality.   These quality control and
quality assurance procedures were used during sampling and/or analysis.

6.1.1  Sampling Equipment Preparation

     The checkout and calibration of source sampling  equipment is vital to
maintaining data quality.  Referenced calibration  procedures were strictly
adhered to when available, and all results were documented and retained.

     Table 6-1 summarizes the parameters of interest  and the types of sampling
equipment that were used to measure each parameter.  Prior to sampling, all
equipment was cleaned and checked to ensure operability.   Equipment requiring
pretest calibration (Table 6-1) was calibrated in  accordance with "Quality
Assurance Handbook or Air Pollution Measurements Systems,  Volume III,
Stationary Source Specific Methods," (EPA 600 4-77-027b).

     Pre-test calibration data for type "S" pitot  tubes,  temperature measuring
devices, and dry gas meters can be found in Appendix E.2.   Balance calibration
data are located in Appendices D.4 (Particulate Operations Log)  and H.4 (Field
Laboratory Log).
                                        6-1

-------
                TABLE 6-1.   SUMMARY OF EQUIPMENT USED IN PERFORMING SOURCE SAMPLING

Parameter
Volumetric Flue Gas
Flow Rate
Gas Phase Composition
Moisture ,
Molecular Weight
D1oxin/Furan
TOCL/Particulate
Method
EPA 1 & 2

EPA 4
EPA 3
ASME/EPA
Protocol
ASME/EPA
Protocol
(modified)
Calibrated Equipment Used to Measure Parameters
Type "S" Temperature
Pitot Measuring
Tube Manometer Device Orsat Nozzles Balances
XXX

XXX X
X
X X X X X X
X X X X X X

Dry Gas
Meter


X

X
X
Metals/Part1cul ate
Alternate EPA
  Method 12
X
X
X
X

-------
6.1.2  General Sampling PC Procedures


     The following QC checks were conducted for each of the EPA Methods 2,  3,

4, 5, and PCDD/PCDF, TOCL/Particulate,  and Metals/Particulate sampling.


Prior To Sampling

          Sampling equipment was inspected for possible damage from
          travel,  and thoroughly checked to ensure operable components.
          Determined duct size and required dimensional measurements.
          Perform initial preliminary velocity, temperature, and moisture
          determinations to aid in conducting isokinetic sampling.
          Perform cyclonic or turbulent flow checks.
          Determined the proper sampling nozzle size.

Prior to Testing Each Day

          Assembled the train in an environment free from uncontrolled dust.
          Checked the number and location of the sampling points before taking
          measurements.
          Visually inspected the S-type pitot tube.
          Leak-checked each leg of the S-type pitot tube.
          Leak-checked the entire sampling train.
          Leveled and zeroed the oil manometer used to measure pressure across
          the S-type pitot tube.
          Checked the temperature measurement system for operability by
          measuring the ambient temperature prior to each  traverse.
          Checked the heating and cooling systems to ensure proper operation,
          and stocked cooling systems with ice.
          Visually  inspected each sampling train for proper assembly.
          Sealed sampling ports to help prevent possible air inleakage.
          Reviewed data requirements prior to each sampling run.

During Testing Each Day

          Properly maintained the roll and pitch axis  of the S-type  pitot  tube
          and sampling nozzle.
          Leak checked the  train before and after  any  move from one  sampling
          port to another during a run or  if a filter  change took place.
          Maintained the probe,  filter, sorbent trap,  and  impinger outlet  at
          the proper temperatures.
          Maintained ice  in the  impinger bath  at  all times.
          Made proper  readings of the dry  gas  meter,  pressures,  temperature,
          and pump  vacuum during sampling  at each traverse point.
          Maintained isokinetic  sampling velocity within + 10  percent of the
          duct velocity.
          Noted any unusual occurrences  involving the sampling train on the
          appropriate  data  form.
                                       6-3

-------
After Testing Each Day
          Leak checked the entire sampling train.
          Visually inspected the sampling nozzle.
          Visually inspected the S-type pitot tube.
          Leak checked each leg of the S-type pitot tube.
          Immediately recapped the sorbent trap, probe,  filter,  and impingers
          as train was disassembled.
          Collected any necessary sample train blanks.
          Checked completion of field data sheets.
          Gave field data sheets to sampling team leader for review.
6.1.3  Sample Recovery

     To ensure sample integrity, careful recovery techniques were adhered to
by experienced analysts.  This section outlines quality control  procedures
followed to ensure sample integrity.  These included:

          Recorded reagent lot numbers.
          Rinsed probe and nozzle brushes,  tweezers,  and scrapers before use
          with the proper reagent(s) to prevent sample  contamination.
          Clearly labeled reagent dispenser bottles.
          Rezeroed the toploading field balance after every ten  weighings or
          less.
          Sample trains were disassembled and the samples  recovered in clean
          areas  to prevent contamination.
          The nozzle was capped prior to and following  recovery.
          Filters were handled out of drafts and transferred with tweezers.
          Rinsed sampling glassware three  times with  each  reagent to  remove
          all of the sample.
          The samples were transferred to  appropriate storage containers and
          clearly labeled.   Liquid levels  were noted, and  completed sample
          containers were weighed and the  weights were  recorded  in the
          logbook.
          Samples were carefully labeled,  logged into the  field  logbook and
          assigned a unique identification code immediately after collection.
          Capped all sampling and recovery glassware  when not in use.
          Sets of glassware and nozzle brushes were dedicated to each sampling
          location and method (i.e.,  ESP outlet PCDD/PCDF train).
                                       6-U

-------
6.1.4  Preparation of Samples for Analysis

     Prior to analysis, each sample was  properly prepared.   This  section
outlines quality control procedures used to  ensure  proper sample  preparation.
Included are:
          Each sample identification code was crosschecked for accuracy
          against the sample logbook.
          The analytical requirements of each sample were reviewed.
          The sample containers were checked for leakage or damage and any
          anomalies were noted.

Sample Analysis

     The quality assurance/quality control  procedures followed during the
analysis task were dependent on the specific analysis being performed.  One or
more of the following steps were taken:
          Duplicate analyses were performed on 1 out of every 10 PCDD/PCDF
          samples.  Duplicate analyses were performed on audit samples for
          NAA.
          Blanks were analyzed to correct for background and/or matrix
          interferences.
          Blind QC samples were submitted to the analytical laboratories along
          with the field samples.
          For the PCDD/PCDF analyses, the samples  were spiked with known
          additions of  the species of interest and recoveries were calculated.

Data Documentation and  Verification

     Several measures were taken to verify the completeness and accuracy of
the data generated.  These included:
          All sampling  data were recorded on preformatted data sheets.
          Data tables were prepared and reviewed for completeness and
          accuracy.
          All data that appeared to be outside expected ranges were carefully
          scrutinized  for sampling, analytical or process problems that may
          have occured.
                                       6-5

-------
 6.2.  METHOD-SPECIFIC SAMPLING QC PROCEDURES

     In addition to the general QC procedures listed in Sections 6.1.1 to
 6.1.4, QC procedures specific to each sampling method were also incorporated
 into the sampling scheme.  These method specific procedures are discussed
 below.

 6.2.1  Procedures for Velocity/Volumetric Flow Rate Determination

     Data required to determine the volumetric gas flow rate were collected
 using the methodology specified in EPA Method 2.  Quality control procedures
 followed were:

          Visually inspected the S-type pitot tube before and after sampling.
          Leak-checked both legs of the pitot tube before and after sampling.
          Checked the number and location of the sampling traverse points
          before taking measurements.
          Maintained proper orientation of the S-type pitot tube while making
          measurements.
          Leveled and zeroed the oil manometer, and recorded the proper
          pressures and temperatures.

 6.2.2  Quality Control Procedures for Molecular Weight Determinations

     Samples used for determination of stack gas molecular weight were
 collected using the integrated sampling technique specified in EPA Method 3.
Quality control for the Method 3 sampling focussed on the following
procedures:

          The sampling train was leak checked before and after each run.
          A constant sampling rate (+ 10 percent) was used in withdrawing the
          integrated gas sample.
          The sampling train was purged prior to sample collection.
          The sampling port was sealed to prevent air inleakage.
                                       6-6

-------
Analytical quality control for Method 3 include the following:
          The Orsat analyzer was leak-checked prior to use.
          The Orsat analyzer was leveled and fluid levels zeroed prior to each
          use.
          The Orsat analyzer was thoroughly purged with sample prior to
          analysis.
          Analyses were performed until analysis agreed within 0.2% absolute.
          A control sample of ambient air was analyzed daily.
          Orsat solutions were changed when more than six passes were required
          to obtain a stable reading for any component.
          Data obtained with the Orsat analyzer was compared to data from
          other locations, and GEM data.
6.2.3  Quality Control Procedures for Moisture Determination

     The moisture content of the gas streams was determined using the
technique specified in EPA Method 4.  The following internal QC checks were
performed as part of the moisture determinations:
          Each impinger will be weighed to the nearest 0.1 grams before and
          after sampling.
          Rezeroed the field balance after every 10 weighings.
          Used fresh, dry silica gel in the silica gel container.
          Ice was kept in the ice bath to keep the gas exit temperature below
          68 degrees F while sampling.
          The sample train was leak checked before and after each run.
          Dry gas meter readings were correctly recorded on the proper data
          sheet.

6.2.4  Quality Control For PCDD/PCDF Testing

     This section summarizes the quality control activities for PCDD/PCDF
testing at North Andover RESCO.  The aforementioned general quality control
activities do apply, as they do to all sampling methods based on EPA Method 5.
The specific PCDD/PCDF QC activities discussed in this section include
sampling preparation, sampling operations, surrogate and internal standard
recoveries, sample blanks and audit analyses.
                                     6-7

-------
 Equipment and  Sampling Preparation

     Pre-test  calibrations or inspections were conducted on pitot tubes,
 temperature sensors, dry gas meters, and balances.  Precleaning procedures for
 sample train glassware and amber glass sample bottles were followed as
 specified in Section 5.1.1.  After cleaning, each piece of the sampling
 glassware was  sealed with aluminum foil to prevent contamination.  Sample
 bottles were sealed with teflon lined lids.

     The residual analyses results for the XAD resins and filters are
 contained in Appendix J.2.  The residual analysis is a requirement of the
 PCDD/PCDF sampling protocol.

 Sampling Operations

     PCDD/PCDF sampling operations followed standard Method 5 operating
 procedures (Appendix K.4) with the addition of the following:
     1)   Used only precleaned aluminum foil or ground glass caps to cover
          sample train components during train assembly, disassembly, and leak
          checks.
     2)   The  temperature of the gas entering the sorbent trap (XAD) remained
          below 68°F.
     3)   The  sampling rate averaged 0.4 dscf.
     4)   All  train components which were recovered were made of glass or
          teflon.
     5)   All  train components and sample bottles were marked according to the
          cleaning procedure used.

     PCDD/PCDF sampling results for isokinetics and leak checks are presented
in Table 6-2.   These results are acceptable from a QC point of view, but it
should be noted that the dry gas meter volume for Run 5 was corrected to
reflect an excessive leakrate for one port.  The sample volume correction
calculation is specified in Appendix A.5.
                                        6-8

-------
       TABLE 6-2.   DIOXIN  ISOKINETICS3 AND LEAK CHECKb SUMMARY
                  ESP  INLET, NORTH ANDOVER RESCO
                              Run 3         Run 4        Run 5

  Percent                     100.1          99.0         98.0
 Isokinetic
 Leak rate0 (cfm)
   Port A                     0.015         0.017        0.019
        B                     0.017         0.017        0.032d
        C                     0.015         0.014        0.018
        D                     0.012         0.017        0.019
        E                     0.010         0.015        0.017
        F                     0.015         0.015        0.018
        G                     0.010         0.017        0.019
        H                  -   0.010         0.017        0.019

 Volume correction (cf)        None          None         0.29
aThe QC objective for isokinetics was  100 + 10  percent.
 The QC objective for leak checks was  a leak-free train  or leakage
 rate less than or equal  to 0.02 cfm or less than 4% of  the average
 sampling rate (whichever is less).
cLeak checks were performed according  to EPA Method 5 protocol.
 The leak rate of 0.032 cfm for Port B required a volume correction of
 0.29 actual cubic feet.   This corrected volume consisted of 0.3 percent
 of the total volume of flue gas collected during Run 5.
                                 6-9

-------
 Sample Recovery


      PCDD/PCDF sample  recovery  followed  the procedure presented in Section 5.
 In addition to the  general  recovery QC procedures, these procedures were also
 implemented:

           All  instruments used  in  the recovery process were either teflon or
           stainless steel.

           All  instruments which come in  contact with the sample, or the
           sampling  surfaces were cleaned according to the PCDD/PCDF cleaning
           procedure.

           Probe liners were recovered using a flask attached to the male end
           of probe.

           No sealing grease was used on  the sampling train.  Reagent, lab
           proof,  and field  blanks  were taken.

           Each sample  container lid was  individually sealed with teflon tape
           to prevent leakage.

           Each container lid was covered with an integrity seal (strip of
           tape)  to  prevent  tampering.

           Samples were weighed  in  the field and again in the lab to indicate
           possible  sample loss.
                                                                      p
           All  sample containers were packaged for transport in ziplock
           plastic bags, wrapped in bubble wrap, and placed into a second
           ziplock bag.

 Surrogate  and  Internal Standard Recoveries of the Test Samples.
     PCDD/PCDF samples were spiked with internal standards and surrogates
prior to extraction.  The internal standards were added in the soxhlet
extraction step.  The surrogates were added to the impinger (condensate)
fraction.  The internal standard recoveries were used by Triangle Laboratories
to adjust the results of the native species reported.  The surrogate
recoveries were not used to adjust results but were used to provide additional
information on the extraction efficiency of the method.  The internal standard
recoveries are summarized in Table 6-3.  The QC objective as required by the
ASME/EPA protocol is + 50 percent.  All of the analyses met this requirement,
except for the following:
                                      6-10

-------
                   TABLE 6-3.   INTERNAL STANDARDS RECOVERY RESULTS FOR
                               NORTH ANDOVER PCDD/PCDF ANALYSES

OQ7fl
Sample Type '
MM5 Inlet Samples
Run 03
Run 04
Run 05
Quality Control Samples
Field Blank
Laboratory Proof Blank
Analytical Blank
Audit Quality Control
Blank (spikes)
(#25/#26)
13
- C- TCDE
FH/BH
90/99
101/102
87/107

107/a
91/97
99
129/116


Recovery
13 13
) C -PCDD
FH/BH
81/83
90/89
69/99

94/a
82/115
78
88/111


C12-HxCDD
FH/BH
81/75
83/91
67/100

96/a
91/106
86
107/113 '


13C12-HpCDD
FH/BH
79/78
74/82
58/82

9 I/a
81/137
84
94/140


13C12-OCDD
FH/BH
63/67
71/71
50/47

84/a
163b/120
76
72/130


FH/BH = front half/back half percentages of internal standard recoveries.
 Back half lost by Triangle Labs while processing,  due to broken sample transfer line.
 Front half duplicate analysis = 72%
                                         6-11

-------
                                        13
      1)    Internal standard recovery of   C  -OCDD in the back half of Run 5
           was 47 percent.
      2)    Internal standard recovery for one of the lab proof blank front half
                                13
           duplicate analyses of   C---OCDD was 163%.
      The  internal standards are used to adjust the responses for extraction
 efficiency and variable  instrument performance.  Since the internal standards
                                              13
 are  spiked as a known amount (2 ng except for   X-^-OCDD which is spiked a
 4 ng), the recovery is determined using the external standard response and the
 following equation.  External standards are the same compounds used for
 internal  standards except that they are in a pure organic matrix rather than a
 sample matrix.
      Internal standard _ (area counts) internal standard
           recovery       (area counts) external standard
 The  sample results are calculated:
           results reported (ng) «• RF x F„ x (area counts) sample
 where
           REG — the inverse of the internal standard recovery
           RF    = response factor determined with external standards
                  (ng per are count)
     The  adjustment is done by computer so that the results reported by
 Triangle  Laboratories are already adjusted.  Table 6-4 presents the internal
 standard  adjustment factors for the PCDD/PCDF samples.

     The  surrogate recoveries are summarized in Table 6-5.  All the analyses
met the ASME/EPA protocol QC requirement of +50 percent except for the
 13
  C-9-HxCDF recovery in  the front half of the field blank which was
42 percent.

 Sample Blanks

     Field and laboratory proof blanks were collected and analyzed to evaluate
contamination from the glassware, handling of the train and field recovery.
                                       6-12

-------
               TABLE  6-4.   FACTORS USED TO ADJUST RESPONSES FOR EXTRACTION
                           EFFICIENCY AND VARIABLE INSTRUMENT PERFORMANCE
                                                  Factors
  Sample  Type
2378-13C-12TCDD
MM5 Inlet Samples
Run 03
Run 04
Run 05
Quality Control Samples
Field Blank
Laboratory Proof
Blank
Analytical Blank
Audit Quality
Control Blank
(spikes) (#25/#26)
FH/BH
1.11/1.01
0.99/0.98
1.15/0.93

0.93/c
1.10/1.03
1.01
0.78/0.86
FH/BH
1.23/1.20
1.11/1.12
1.45/1.01

94/c
1.22/0.87
1.28
1.14/0.91
FH/BH
1.23/1.33
1.20/1.10
1.49/1.0

96/c
1.10/0.94
1.16
0.93/0.88
FH/BH
1.27/1.28
1.35/1.22
1.72/1.22

91/c
1.23/0.73
1.19
1.06/0.71
FH/BH
1.59/1.49
1.41/1.41
2.0/2.13

84/c
0.61d/0.83
1.32
1.39/0.77
12378-   C   -TCDD is used to adjust  2378-TCDD and total TCDD, 2378-TCDF,  and  total  TCDF.
 13C12-PCDD is  used to adjust 12378-PCDD,  12378-PCDF, 23478-PCDF,  total  PCDD and
 total  PCDF.
 13C12-HxCDD is used to adjust 123478-HxCDD, 123678-HxCDD,  123789-HxCDD,  123478-HxCDF,
 123678-HxCDF,  123789-HxCDF,  total  HxCDD and total HxCDF.
 13C12-HpCDD is used to adjust 1234678-HpCDD,  1234678-HpCDF, 1234789-HpCDF,  total  HpCDD
 and total HpCDF.
 13
  C  -OCDD is  used to adjust total OCDD and total OCDF.
5FH/BH  - front  half/back half factors.
'Back half lost by Triangle Labs while  processing, due to broken sample  transfer line.
 Front  half duplicate analysis - 1.39
                                          6-13

-------
TABLE 6-5.  SURROGATE RECOVERIES FOR PCDD/PCDF ANALYSES FOR NORTH ANDOVER

Sample Type
MM5 Samples - Inlet
Run 03
Run 04
Run 05
Quality Control Samples
Field Blank
Laboratory Blank
Analytical Method Blank
Audit Quality Control
(Spikes)
13
C-12TCDF
FH/BH
91/96
97/84
101/90

88/a
100/102
97
105/106
Recoveries (%)
37C1. -TCDD
4
FH/BH
103/98
106/88
104/96

102/a
103/99
98
120/125
13C12-HxCDF
FH/BH
96/79
82/64
80/62

42/a
104/89
93
93/66
FH/BH = front half/back half percentages of surrogate recoveries.

*a
 Back half lost by Triangle Labs while processing, due to broken sample

 transfer line.
                                  6-iU

-------
The field blank was an assembled and loaded train which was  capped off and
left at the sampling location for the duration of a  test run.   The glassware
used for the field blank had been used and recovered at least  once.   The  train
was then recovered at the end of the test run.   This blank was used to measure
the background levels of the analytes due to the handling of the  train and
field recovery.

     Laboratory proof blanks were also included for  each method.   The
laboratory proof blank quantifies the background levels of PCDD/PCDF from the
glassware and recovery equipment.  The laboratory proof blanks were obtained
from a complete set of sample train glassware and recovery equipment (brushes,
spatulas, etc.) that had been cleaned according to  the specified precleaning
procedure.  This glassware, which consisted of the probe liner, filter holder,
condenser coil, and impingers, was loaded and then  recovered according to the
standard recovery method.  Also, blanks of each solvent lot and filters used
at the test site (reagent blanks) were saved for potential analysis.

     Table 6-6 summarizes the analytical results reported by Triangle
Laboratories for laboratory system blanks, field blanks, laboratory proof
blanks, and reagent blanks.  The laboratory system  blank is an analysis of a
PCDD/PCDF-free silica gel sample which is carried throughout the entire
extraction and analysis procedure.  All of the blanks showed insignificant or
non-detectable levels of PCDD/PCDFs.

     Table 6-7 compares the field blank results to  the minimum test values for
each isomer attained in Runs 3, 4, and 5 as a group, in the form of a
percentage.  As the resulting percentages indicate,  the maximum bias that
background levels of one isomer may have on the final results were
insignficant.

Audit Samples

     Blind audit samples (spikes) were also submitted to Triangle Laboratories
to evaluate the recoveries reported during the analysis of North Andover
                                       6-15

-------
       TABLE 6-6.   ANALYTICAL RESULTS FOR NORTH ANDOVER QUALITY CONTROL SAMPLES'

Flue Gas Quality Control Samples
Compound
Dioxins
Mono-CDD
Di-CDD
Tri-CDD
2378-TCDD
Other TCDD
12378 PCDD
Other PCDD
123478 HxCDD
123678 HxCDD
123789 HxCDD
Other HxCDD
Hepta-CDD
Octa-CDD
Total PCCD
Furans
Mono -CDF
Di-CDF
Tri-CDF
2378 TCDF
Other TCDF
12378 PCDF
23478 PCDF
Other PCDF
123478 HxCDF
123678 HxCDF
123789 HxCDF
Other HxCDF
Hepta-CDF
Octa-CDF
Total PCDF
Laboratory Field Laboratory
System Blank , Blank
Blank MM5 Train MM5 Train

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
0.26
0.04
ND
ND
ND
ND
ND
ND
ND
0.70
1.21
2.21

ND
ND
ND
0.15
0.10
ND
ND
ND
0.12
ND
ND
0.0
0.27
ND
0.66

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Reagent Blanks
Water

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Acetone

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Methylene
Chloride

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
a                                    -9
 All values reported in nanograms (10   g) per gram of sample.



 Values reported are front half values only.  Back half lost during processing.



 ND - Not detected.  Minimum detection limits ranged form 0.001 to 0.009 nanograms

 per sample.
                                           6-16

-------
          TABLE 6-7.   FIELD BLANK DIOXIN/FURAN  DATA FOR MM5  SAMPLES'

Compound
Dloxins
Mono-CDD
Di-CDD
Tri-CDD
2378-TCDD
Other TCDD
12378 PCDD
Other PCDD
123478 HxCDD
123678 HxCDD
123789 HxCDD
Other HxCDD
Hepta-CDD
Octa-CDD
Furans
Mono -CDF
Di-CDF
Tri-CDF
2378 TCDF
Other TCDF
12378 PCDF
23478 PCDF
Other PCDF
123478 HxCDF
123678 HxXCDF
123789 HxCDF
Other HxCDF
Hepta-CDF
Octa-CDF

Field Blankb

ND
ND
0.26
0.04
ND
ND
ND
ND
ND
ND
ND
0.70
1.21

ND
ND
ND
0.15
0.10
ND
ND
ND
0.12
ND
ND
0.02
0.27
ND
Amount Detected
Minimum Test Run Value

ND
1.72
6.17
0.62
6.69
1.10
10.39
1.09
1.85
ND
14.23
26.06
22.05

0.32
9.26
39.78
9.22
25.74
2.06
4.48
12.52
5.23
1.88
ND
7.29
13.88
4.32

Percentage


0
4.2
6.4
0
0
0
0
0

0
2.7
3.7

0
0
0
1.6
0.4
0
0
0
2.3
0

0.3
1.9
0
a                                    -9
 All values reported in nanograms (10  g).


 Front half only.  Back half sample was lost during analysis by Triangle

 Laboratories.

 ND = not detected.
                                      6-17

-------
 dioxin/furan samples.   Known quantities  of  the  targeted 2378-TCDD and
 2378-TCDF isomers were  spiked into XAD-2 resin and impinger solution to
 assess the extraction and recovery procedure for  the front and back halves of
 the PCDD/PCDF train.  Excellent  recoveries  were reported:

      #25   Front Half (XAD-2R):               104% Recovery of 2378-TCDD
                                             120% Recovery of 2378-TCDF

      #26   Back Half (Impinger Solution):     100% Recovery of 2378-TCDD
                                             111% Recovery of 2378-TCDF

 These spike recoveries  were  well within  the quality control criteria of + 50
 percent required by the ASME/EPA protocol.

 6.2.5  Quality Control  for Particulate Testing

      Particulate sampling was incorporated  into the ESP outlet TOCL (Runs 1-6)
 and metals (Runs 7, 8,  and 9)  ESP inlet  and ESP outlet sampling trains.  TOCL
 results are not reported  here, but the particulate fraction of that sampling
 effort is.   Since many  of the QA procedures used  for the Metals/Particulate
 sampling  train are discussed in  the next section, (with the exception of
 results)  most of this section will be directed  towards the TOCL/particulate
 sampling.

 Equipment and Sampling  Preparation.

      Preparation for TOCL/particulate sampling  is identical to preparation for
 PCDD/PCDF with the  exceptions  of using hexane in place of methylene chloride,
 and using tared filters.   Filters were tared in accordance with EPA Method 5
 and placed  in sealed, precleaned, glass  petri dishes prior to leaving for
North Andover  RESCO.  The  analytical balance used to tare the filters, and
used  in later  particulate  analyses, was  calibrated with standard weights (NBS
Class S).   Measured values agreed within +0.1  mg.  The balance was calibrated
prior to making measurements.  This calibration data can be found in the
particulate analysis logbook  in Appendix D.4.
                                      6-18

-------
Sampling Operations

     Sampling operations for the TOCL/particulate  sampling train are identical
to PCDD/PCDF sampling operations.   TOCL/Particulate sampling results for
isokinetics and leak checks are presented in Table 6-8.   None of the sampling
trains were determined to have leakrates above the maximum allowable leakrate
(0.02 acfm).

Sample Recovery

     TOCL/particulate sample recovery followed the PCDD/PCDF QC format except
that hexane was used in the place of methylene chloride.   Hexane is actually
the recovery solvent specified in the method.

     With the particulate aspect of the TOCL/particulate  and
Metals/Particulate sampling effort in mind,  it is  important to note here that
the initial front half washes were performed with  acetone, as EPA Method 5
dictates.  The acetone and hexane washes were recovered into separate bottles.

Sample Analysis

     The QC analyses for particulate involved evaluations of the field,
laboratory proof, and reagent particulate blanks of the TOCL/Particulate,
Inlet Metals/Particulate, and Outlet Metals/Particulate sampling trains.  The
sample fractions evaluated were the nozzle and probe/filter holder acetone
washes and the filter.  Table 6-9 shows the values for these fractions, and
their respective sums.  When the samples were returned to the laboratory from
the field, the front half acetone fraction from the TOCL/particulate-Run 4
sample was determined to have leaked during shipment.  The sample result was
corrected using the calculations provided in Appendix A. 5.  Once the
particulate analyses were complete, the particulates were resuspended  in
hexane and combined with the front half hexane and back rinses for a total
train TOCL analysis.
                                      6-19

-------
       TABLE 6-8.  TOCL/PARTICULATE ISOKINETICS3 AND LEAK CHECKb SUMMARY
                       ESP OUTLET, NORTH ANDOVER RESCO


Percent
Isokinetic
Leak checks (cfm)
Port A
B
C
D
E
F
Run 1
100.7

0.011
0.007
0.006
0.004
0.012
0.003
Run 2
100.6

0.004
<0.02d
0.008
0.005
0.008
0.010
Run 3
100.0

0.011
0.010
0.008
0.009
0.008
0.008
Run 4
100.0

0.007
0.005
0.008
0.005
0.012
0.011
Run 5
100.6

0.011
0.017
0.013
0.009
0.014
0.019
Run 6
100.0

0.007
0.010
0.018
0.005
0.007
0.009
 Volume correction
       (cf)
None
None
None
None
None
None
 The QC objective for isokinetics was 100 + 10 percent.

 The QC objective for leak checks was a leak-free train or leakage rate less
 than or equal to 0.02 cfm or less than 4% of the average sampling rate
 (whichever is less).
£
 Leak checks were performed according to EPA Method 5 protocol.

 wo leak rate defined on data sheet.
                                    6-20

-------
                                       TABLE 6-9.   SUMMARY OF  PARTICULATE  QC  RESULTS  BY  SAMPLE  FRACTION
                                                       NORTH  ANDOVER  RESCO,  NORTH ANDOVER, MA
I
ro
Blank
Lab Proof Blanks
TOCL
Metal s/Partlculate
Inlet
Metal s/Partlculate
Outlet
Field Blanks
TOCL
Metal s/Partlculate
Inlet
Metal s/Part1culate
Outlet
Reagent Blanks
Filter
TOOL Acetone
(Runs 1-3)
(Runs 4-6)
Metal s
Acetone
Filter2
Weight
Gain (g)

0.0000
0.0002
0.0003

0.0002
0.0000
0.0000

0.0000
.—
—
Uncorrected
Probe
Acetone
Residue (g)

0.0003
0.0060
0.0022

0.0009
0.0045
0.0037

—
__
~
Volume
of Acetone
Probe
Wash (ml)

95
185
215

195
250
195

~
420
275
460
Blank Uncorrected Volume
Correction of Acetone of Acetone
Acetone Probe Acetone Nozzle Nozzle Acetone
Blank (g) Residue (g) Wash (g) Wash (ml) Blank (g)

0.0000 0.0003 Nozzle wash Included In probe Mash
0.0006 0.0054 Nozzle wash Included In probe wash
0.0007 0.0016 Nozzle wash Included In probe wash

0.0000 0.0009 Nozzle wash Included In probe wash
0.0009 0.0036 0.0003 150 0.0005
0.0007 0.0030 0.0007 100 0.0004

—
0.0000
0.0000
0.0017
Blank Probe and Probe
Correction of Nozzle Nozzle
Nozzle Acetone Residue & Filter
Residue (g) Total (g) Total (g)

0.0003 0.0003
0.0054 0.0056
0.0016 0.0019

0.0009 0.0011
0.0000 0.0036 0.0036
0.0003 0.0033 0.0033




         ''Blank Calculations - Corrected  acetone Weight * Uncorrected acetone weight (g) [volume acetone (ml) x acetone  blank g/ml)]
         ^Negative weight gains on reagent blanks (filter A acetone) are listed as zero as long as they are within the accuracy of the balance.  (0.0005 g).
          Actual values are located In the laboratory weigh sheets In this section.

-------
 6.2.6  Quality Control  for  Metals/Particulate Testing

      This section summarizes  the  quality control procedures for
 Metals/Particulate testing.   The  general EPA Method 5 QC procedures did apply,
 as Alternative Method 12  is based on Method 5, so only specific
 Metals/Particulate QC activities  are discussed here.

 Equipment and Sampling  Preparation

      Pre-test calibrations  and visual  inspections were performed on the
 necessary sampling equipment  as specified in EPA Method 5.  Precleaning
 procedures for the sample train glassware, amber glass, and Nalgene sample
 bottles were  specified  in Section 5.2.1.  After cleaning, each piece of
 sampling  glassware was  sealed with parafilm or rubber caps to prevent
 contamination.  Amber glass sample bottles were sealed with teflon lined lids,
 and Nalgene sample bottles were closed.  All glassware and sample bottles were
 marked  to show that they had  been prepared for metals sampling.  Whatman
 934 AH  filters were tared according to EPA Method 5 and placed in sealed,
 precleaned, glass  petri dishes.
 Sampling  Operations

     Metals/particulate sampling  operations followed standard Method 5
 operating procedures  with the addition of the following:
     1)    Use  of only round glass caps, parafilm, or rubber caps to cover
           sampling train components during assembly, leak checks and
           disassembly of the  sampling train.
     2)    The  sampling rate averaged 0.4 dscfm.
     3)   All  train components and sample bottles had been marked previously
           according to the cleaning procedure used.
     Metals/particulate sampling  results for isokinetics and leak checks are
presented in Tables 6-10 and  6-11, respectively.  The ESP outlet run 7 had two
leaks prior to breaking the probe during port E.  Both leaks occurred at the
union between  the  probe and the nozzle, and the break occurred approximately
one half  inch below this union.
                                          6-22

-------
TABLE 6-10.  METALS/PARTICULATE ISOKINETICSa AND LEAK CHECK** SUMMARY
                  ESP INLET,  NORTH ANDOVER RESCO


Percent
Isokinetic
Leakrate (cfm)
Port A
B
C
D
E
F
G
H
Run 7
101.8

0.009
0.017
0.010
0.013
0.010
0.032d
0.006
0.004
Run 8
101.7

0.009
0.002
0.004
0.004
0.007
0.006
0.006
0.002
Run 9
102.1

0.008
0.003
0.006
0.005
0.002
0.003
0.002
0.003
 Volume correction (cf)       0.29          None         None
aThe QC objective for isokinetics was 100 + 10 percent.
 The QC objective for leak checks was a leak-free train or leakage
 rate less than or equal to 0.02 cfm or less than 4% of the average
 sampling rate (whichever is less).
Q
 Leak checks were performed according to EPA Method 5 protocol.
 The leakrate of 0.032 cfm for Port F required a volume correction of
 0.29 actual cubic feet.  This corrected volume consisted of 0.3 percent
 of the total volume of flue gas collected during Run 7.
                                    6-23

-------
TABLE 6-11.  METALS/PARTICULATE ISOKINETICS3 AND LEAK CHECKb SUMMARY
                  ESP OUTLET, NORTH ANDOVER RESCO


Percent
Isokinetic
Leakrate (cfm)
Port A
B
C
D
E
F
Run 7
99.7

0.010
0.017
0.015
0 . 020+d
Broken
0.008
Run 8
98.6

0.004
0.004
0.007
0.003
0.004
0.003
Run 9
100.2

0.004
0.004
0.004
0.006
0.003
0.004
 Volume correction (cf)    •   4.0           None         None
 The QC objective for isokinetics was 100 + 10 percent.
 The QC objective for leak checks was a leak-free train or leakage
 rate less than or equal to 0.02 cfm or less than 4% of the average
 sampling rate (whichever is less).
Q
 Leak checks were performed according to EPA Method 5 protocol.

 L,eak rate increased dramatically as the probe cooled.  Based
 correction on a leak rate of 0.04 cfm.
Q
 Port E was not corrected for since a value for leak rate was not
 measurable.

 The volume correction includes the correction from Port D of 0.8 cf
 and Port A of 3.2 cf.  The volume correction of 4.0 cf consisted of
 3.2 percent of the volume of flue gas sampled.
                                  6-2H

-------
Sample Recovery


     Metals/particulate sample recovery followed the  procedure  located in

Chapter 5.   In addition to the general recovery QC procedures,  the  following

QC procedures were also implemented:

          All train components which were recovered were made of glass or
          teflon.  Nozzle washes were recovered into  a separate sample bottle
          to prevent the possibility of contaminating the front half metals
          sample.

          All instruments used in the recovery process were either  teflon or
          teflon coated.

          All instruments which came in contact with  the sample or  the
          sampling surfaces were cleaned according to the Metals/Particulate
          cleaning procedure.

          Probe liners were recovered with a flask attached to  the  male end of
          the probe liner.

          Reagent, laboratory proof, and field blanks were taken.

          Each sample container lid was individually  sealed with teflon tape
          to prevent leakage.

          Glassware sets were dedicated to each sampling location.

          Each container lid was covered with an integrity seal (strip of
          tape) to prevent tampering.

          Samples were weighed prior to and after shipping to indicate
          possible sample loss.

          All sample containers were packaged for transport in ziplock bags,
          wrapped in bubble wrap, and placed into a  second ziplock bag.

          Nitric acid rinses were placed in polypropylene bottles.

          Acetone rinses were placed in glass bottles.

Analysis
     Table 6-12 summarizes the results of the metals field and laboratory
proof blanks for sampling at the ESP inlet and outlet.  The field and
laboratory proof blanks are analyzed in order to evaluate potential
contamination from recovery and handling of the trains.
                                         6-25

-------
         TABLE  6-12.   SUMMARY OF LABPROOF BLANKS AND FIELD BLANKS FOR
                      METALS SAMPLES COLLECTED AT NORTH ANDOVER RESCO

UK/sample '
Sample
Labproof Blank
Uncontrolled:


Controlled:


Field Blank
Uncontrolled:


Controlled:




Front Half
Back Half
Total
Front Half
Back Half
Total

Front Half
Back Half
Total
Front Half
Back Half
Total
Arsenic

5.3 (3.1)
<0.14
5.3
1.7 (8.5)
2.4 (6.4)
4.1

4.3 (4.1)
<0.02
4.3
0.5 (6.1)
<0.02
0.5
Cadmium

9.0 (11.2)
1.8 (13.0)
10.8
7.6 (18.9)
2.2 (12.0)
9.8

2.3 (19.2)
<0.1
2.3
10.4 (11.0)
0.44 (18.0)
10.8
Total
Chromium

14.3 (5.3)
<0.7
14.3
18.4 (6.7)
6.2 (3.8)
24.6

10.4 (5.9)
0.64 (12.0)
11.0
7.0 (6.7)
0.83 (12.3)
7.8
Nickel

6.6 (20.0)
94.1 (4.1)
100.7
17.7 (14.1)
5.4 (17.0)
23.1

<5
1.2 (18.5)
1.2
18.1 (12.0)
<1
18.1
value in parenthesis is range of precision reported as a percent of its respective
value.
Minimum detection limits are denoted by a "<" symbol.  A range of precision is not
reported for the minimum detection limits.
                                            6-26

-------
     To determine if the field blank concentration  is  a.  significant portion of
the concentrations reported for the runs,  the  field blank is  compared to  the
minimum run value in Table 6-13.   The minimum  run and  field blanks
concentrations are correctd for the blank results.  The  comparison  indicates
that significant contamination was not caused  by recovery and handling of the
sampling trains.

Neutron Activation Analysis (NAA) of NBS  Reference  Methods

     National Bureau of Standards (NBS) Standard Reference Materials (SRM)
were analyzed by N.C. State along with EMB-North Andover metals samples.
A total of 197 analyses of different elements  were  included with the sample
batch during the phases of analysis.  Of these, ninety-one percent were within
the tolerances set by NBS.  All were within ten percent of the NBS tolerances.
The results are summarized in Appendix I.

Dulicate Analyses

     Duplicate analyses were performed on the  NBS SRM's to assess the
precision of the analytical method.  The nature of NAA is such that duplicate
analyses are performed either by splitting the sample into aliquots and
analyzing both at the same time, or by doing multiple analyses on the same
sample.  Multiple analyses require extra time since the sample must "cool
down" between analysis.  However, the NBS SRMs are easily split into aliquots
which are included with more than one batch of analyses, enabling duplicate
results to be easily obtained.

     The percent differences for all the duplicate analyses were less than ten
percent, indicating good precision for this data set.  Duplicate analyses and
metals concentrations in the NBS samples can be  found in Appendix I.

6.2.7  Quality  Control for Continuous  Emissions  Monitors

     Continuous  emissions monitors  (GEM) were  used to measure the  levels of
oxygen, carbon  dioxide, and carbon monoxide at the ESP  outlet sampling

                                       6-27

-------
              TABLE 6-13.  COMPARISON OF MINIMUM RUN VALUES TO
                           FIELD BLANK VALUES FOR NORTH ANDOVER RESCO

Element
Uncontrolled
Arsenic
Total Chromium
Nickel
Cadmium
Controlled
Arsenic
Total Chromium
Nickel
Cadmium
Minimum
Run Value (MRU)
(ug/s ample)

1459
4573
152
746

3.5
0
60.4
16.3
Fielda
Blank (FB)
(ug/sample)

0
0
0
0

0
0
0
1.1
Ratio of
FB: MRV
(%)

0
0
0
0

0
0
0
7
values have been adjusted for reagent blank results.
                                       6-28

-------
location during runs 7, 8, and 9.   The samples  were  collected following EPA
Method 10 for CO and EPA Method 3A for 0  and CO-.   The  following measures
were taken to insure the validity of the CEM data.

Leak Checks

     Prior to sampling, the CEM sampling train  was leakchecked by introducing
zero nitrogen at the probe outlet, and then measuring the oxygen concentration
with the calibrated oxygen monitor.  Any oxygen present  is considered to be
proportional to system leakage.  An oxygen concentration of 0.5 percent was
required to be acceptable.

Linearity Check

     A linearity check was also performed on each monitor (0_ and CO/CO-)
prior to sampling.  Three point (zero plus two  upscale)  measurements of
standards which span the range of expected gas  concentrations at the North
Andover site were performed to ensure linearity.  In all cases, the acceptance
criteria for the linearity checks were a correlation coefficient (r) of >
0.9950.  If this criterion was not met, the linearity checks were repeated
(following instrument maintenance if judged necessary) until r > 0.9550 was
achieved.

Daily Calibrations

     For all continuous monitors, daily calibrations were performed prior to
and at the conclusion of testing.  These calibrations consisted of an
analytical blank (zero nitrogen) and a single point  (span) calibration check.
The response factor (RF) for the span must be within 20% of the response
factor from the previously determined multipoint calibrations, in order to
meet the acceptance criterion.  The acceptance criteria for the analytical
blank was + 1 percent of the instrument span.  A summary of calibration
requirements is shown in Table 6-14.
                                        6-29

-------
                                 TABLE 6-14.  SUMMARY OF ANALYTICAL CALIBRATION  REQUIREMENTS
cr\

Type of
Parameter Method of Analysis Calibration
C0/C02 Beckman 865 (NDIR) Multipoint (Zero
plus 2 upscale)
Single point RF
Analytical Blank
Single point
drift CHECK
02 Beckman 755 Same as C02
(Paramagnetic)
Calibration
Standards
CO and C0_ 1n N_
CO and (XL 1n N2
Zero N2
CO and C02 In N2
02 In N2
Frequency
Once per site prior
to teslng
Dally (prior to
testing)
Dally (prior to
testing)
Dally (at conclusion
of analysis)
Same as CO/CO-
Acceptance Criteria
Correlation coefficient
10.9950
Agreement within 20% of
multipoint RF
Zero val ue £1% of span
RF agreement within 10%
of RF for single point
RF check
Same as C0/C02

-------
     Comparisons of the RF numbers based on the initial  and final daily
calibrations served as an instrument drift check.   The acceptance criterion
for the zero and span drift checks was agreement between initial and final RF
numbers within + 10 percent.  Table 6-15 summarizes the  zero and span drift
checks for Runs 7, 8, and 9, which all met the acceptance criterion.

     The validity of the GEM and Orsat analysis results  are confirmed based on
a combustion stoichiometry method described in Reference 6.  First,  the
ultimate CCL concentration is calculated based on  an ultimate analysis of the
fuel, if available.  Since ultimate analyses were  not performed on the refuse
from this site, data were used from a similar MSW.  The  ultimate analyses data
used was an average of twelve analyses.   Then, on an 0« versus CO.  axis, a
line is drawn connecting the 0_ intercept, 20.9 %, with  the CO- intercept, the
ultimate % CO .  The GEM and Orsat results should  fall within 10 percent of
this line.  This analysis for the GEM and Orsat data is  shown in Figure 6-1.
The figure shows that the GEM data for Run 8 does  not fall within the range.
This oxygen concentration is lower than the other  runs for the same C0_ and CO
concentrations.  This indicates that a problem with the  oxygen analyzer, such
as condensation in the instrument, may have developed during Run 8 and this
data point is not included in the averages reported.
                                      6-31

-------
                  TABLE 6-15.  SUMMARY OF DRIFT CHECK RESULTS

Test
Date
07-14-86
07-15-86
07-16-86
07-14-86
07-15-86
07-16-86
07-14-86
07-15-86
07-16-86
Test
Run
07
08
09
07
08
09
07
08
09
Parameter
CO
CO
CO
co2
co2
co2
°2
°2
°2
Zero Drift
Instrument
Drift, %a
1.08
0.45
1.71
0.20
0.31
0.25
0.25
0.28
0.34
Check
Meets
QC?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Span Drift
Instrument
Drift, %a
0.60
0.37
0.88
0.15
1.22
-1.23
0.17
3.02
Check
Meets
QC?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
c
Yes
 Instrument drift is defined as the percent difference between the instrument
 response to the input concentration at the beginning and end of the test run.


 QC criteria was daily instrument drift within + 10 percent.

Q
 Changed span gas between beginning and end of run.
                                      6-32

-------
22




20



18
.52    16 -J
CO
CO
.a

^   14 -
T3
c
a>
0
Q
a.
o
E
3
O
c
0)
O)
X
O

12



10


8


6

     4 -
      2-
                                           Key

                                        ® GEM Averages - Outlet

                                        ® ORSAT Analyses-Inlet

                                        A ORSAT Analyses-Outlet
                          Run8®
                         CEM Data
             2    4     6    8     10    12    14     16    18    20


                      C02 (volume percent, dry basis)





             Figure 6-1.  Validation of Fixed Gas Analysis
                                  6-33

-------
                                  REFERENCES
1.   Knisley, D.R.,  Jamgochian,  C.L.,  Miles,  A.J.,  Gergen,  W.P.  and D.J.
     Holder (Radian Corporation)  Draft Emissions Test Report -  Dioxin/furan
     and Total Organic Chlorides Emissions Testing  -  North  Andover Resource
     Recovery Facility. North Andover. Massachusetts.   Prepared  for RUST
     Corporation, Birmingham, Alabama.  October 10,  1986.   DCN.  86-233-015-06.

2.   National Incinerator Testing and Evaluation Program (NITEP) P.E.I.
     Testing Program: Volume II.  (Concord Scientific Corporation).  Prepared
     for Environmental Canada.  July 1985.  Table 6.2.2.

3.   Procedures for Estimated Risks Associated With Polychlorinated
     Dibenzo-p-dioxins and Dibenzofurans (CDD and CDF).  Prepared by the U.S.
     Environmental Protection Agency, Washington, D.C.  April  1986.

4.   Natrella, Mary Gibbons, Experimental Statistics.  National Bureau of
     Standards Handbook. 1963.  pp 17-2 to 17-3.

5.   Annual Book of ASTM Standards "Collecting a gross sample of coal" ASTM
     D2234 (05.05), 1984.

6.   R.T. Shigehara, R.M. Neulicht and W.S. Smith,  "Validating Orsat Analysis
     from Fossil-fuel-fired Units"  Stack Sampling Technical Information: A
     Collection of Monographs and Papers. Volume I.  United States
     Environmental Protection Agency.  EPA-450/2-78-042a.  October  1978.

7.   Reference 2.
                                       7-1

-------