-------
v- '-^v
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
MAR 14 1989
MEMORANDUM
SUBJECT: Trial Burn Observation Guide
FROM:
TO:
Joseph Carra, Director f>
Permits and State Prog/cftns Division
RCRA Branch Chiefs
Regions I-X
MAR '• C 1389
OFFICE OF RCRA
Waste Management Division
U.S. EPA, REGION V
Attached are copies of the Trial Burn Observation Guide to
be distributed to incinerator permitting staff in your Regions
and States.
The Trial Burn Observation Guide was developed to assist
Regional and State regulatory staff in observing trial burns at
hazardous waste incinerators. It provides background
information about trial burns and identifies key personnel that
can provide additional information and assistance. The guide
also provides suggestions on how to prepare for observing the
trial burn and how to conduct on-site activities during tests.
Several checklists and data forms are included to aid in those
preparations and on-site activities. We think this guide will
be a useful tool particularly for less-experienced incineration
permit writers in the Regional and State offices. This final
report reflects valuable input and experience of the members of
the EPA Incinerator Permit Writers Workgroup.
Of the copies provided, we request that at least two copies
be sent on to each State within your Region. This guide will
soon be available to the general public through the National
Technical Information Service (NTIS). If your staff or any
State permit writers have questions about the content or
availability of this new document they may contact Lionel Vega
of my staff at FTS/202-475-8988.
Attachments
cc: Waste Management Division Directors, (Regions 1-10) w/out
attachments
Permit Section Chiefs, (Regions 1-10) w/out attachments
Incinerator Permit Writer Workgroup w/out attachments
Tom Kennedy, ASTSWMO
-------
TRIAL BURN OBSERVATION GUIDE
FINAL REPORT
Prepared by
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
Under Subcontract From
A. T. Kearney Inc.
225 Reinekers Lane
Alexandria, Virginia 22314
Submitted to
U.S. Environmental Protection Agency
Office of Solid Waste (WH-562)
401 M Street, SW
Washington, O.C. 20460
. Attn: Mr. Lionel Vega
In Response to
EPA Contract No. 68-01-7374
Work Assignment No. H20-08-01
MRI Project No. 8875-L(32)
February 16, 1989
-------
PREFACE
This guidance document was prepared under U.S. Environmental Protection
Agency (EPA) Contract No. 68-01-7374 (Work Assignment No. H20-08-01). The
work assignment manager was Mr. Paul Gorman of Midwest Research Institute
(MRI). This document was prepared under the direction of Mr. James Levin of
A. T. Kearney Inc., and a quality assurance (QA) review of the document was
performed by Mr. Burt O'Connell. Revisions have been made to reflect comments
on earlier drafts provided by members of the EPA Permit Writers Workgroup.
Approved for:
MIDWEST RESEARCH INSTITUTE
Chatten Cowherd, Director
Environmental Systems Department
February 16, 1989
111
-------
CONTENTS
Page
Preface 111
Figures v
Tables vi
Executive Summary vi 1
1. Introduction 1
2. Background on Test Observations and Available
Assistance 3
3. Preparation for Field Observations 9
3.1 Review of trial burn plan 9
3.2 Review of permitting needs 20
3.3 Arrangements/scheduling 20
4. On-Site Activities..... 29
4.1 Pretest orientation 29
4.2 Observation of process operations 31
4.3 Observation of field sampling/analysis
activities 41
4.4 Personal safety of observer 56
4.5 QA/QC audits ". 57
4.6 Major deviations or problems 57
4.7 Documentation of activities 58
5. Observation Reports 61
Appendix—Actual Problems That Have Been Encountered During
Trial Burns and Their Resolutions 63
FIGURES
Number Page
1 Volatile organic sampling train (VOST) 44
2 Method 5 particulate and HC1 sampling train 45
3 Modified Method 5 sampling train (MM5)—type 1 46
4 Modified Method 5 sampling train (MM5)~type 2 47
5A Modified Method 5 sampling train (MM5)—type 3 48
58 Modified Method 5 sampling train (MM5)—type 3 49
-------
EXECUTIVE SUMMARY
The purpose of this Trial Burn Observation Guide is to assist Regional
and State regulatory staff in observing trial burns at hazardous waste
incinerators. It provides background information about such tests and
identifies other documents and personnel that can provide additional
information or assistance. Subsequent sections of this Guide describe how to
prepare for observing the tests and how to conduct on-site activities during
the tests. Several checklists and data forms are included to aid in those
preparations and activities.
Most of the contents of this Guide represent suggestions on how to
observe trial burns. However, it does include information from EPA methods or
other accepted methods that are specific requirements. It also contains some
guidance points, taken from other guidance documents, that are usually viewed
as required or at least highly advisable.
Although this document is targeted to regulatory personnel, it should
also be useful to the the regulated community and those organizations that
provide test services by helping them to be cognizant of what is expected of
them. Distribution to members of the public might also serve in making them
more aware of the complexity of trial burns and the detailed observations and
scrutiny of the tests by the regulators.
vii
-------
SECTION 1.0
INTRODUCTION
The Trial Burn Observation Guide was prepared to assist Regional and
State hazardous waste regulatory staff in observing incinerator trial burns.
A premise of the Guide is that the observer's purposes in observing the trial
burn are:
1. To verify that the tests are conducted properly, in accordance with
the approved trial burn test protocol for the facility (i.e., the
Trial Burn Plan and Quality Assurance Plan).
2. To ensure that trial burn data are of sufficient quality to estab-
lish that the incinerator is capable of meeting the required
performance standards and to enable the permit writer to set
enforceable permit limits.
It has been assumed in writing this guide that the permit writer/observer
is familiar with hazardous waste incinerator regulations and is knowledgeable
in the permitting process. It is also assumed that a Trial Burn Plan (TBP)
and the associated Quality Assurance Project Plan (QAPP) have previously been
submitted and approved by the Agency and are in the possession of the permit
writer/observer. In this guide, it is also assumed that the reader is a
permit writer/observer who has regulatory authority/responsibility for the
test and is not merely a casual observer.
The next section presents background information on test observations.
Section 3 is a guide to preparing for the trial burn, prior to actual arrival
at the test site, while Section 4 contains a description of the on-site activ-
ities. Both Sections 3 and 4 contain checklists, 'guidance lists, and data
tables which should be helpful to the observer in preparing for and conducting
the trial burn observation. Section 5 is intended to help the permit writer
prepare a report on the observation activities and results. The Appendix
contains brief descriptions of actual problems that have been encountered by
observers at trial burns and their resolutions.
-------
SECTION 2.0
BACKGROUND ON TEST OBSERVATIONS
AND AVAILABLE ASSISTANCE
Trial burn observations require considerable preparation to be completed
before arriving at the test site. Much of this preparation involves a review
of the Trial Burn Plan (TBP) and associated Quality Assurance Project Plan
(QAPP), but also involves reviewing all the sampling methods designated in
those documents. In addition, it provides a means of becoming more familiar
with the incinerator operations.
It has not been assumed that the permit writer/observer necessarily has
detailed knowledge of sampling methods. But, he should at least be familiar
with them and their requirements, especially stack sampling methods. There-
fore, the permit writer/observer should have some experience with the methods
or have taken stack sampling training courses such as those offered by the
U.S. Environmental Protection Agency (EPA) (see Table 1A).
Several other resources and guidance documents should be reviewed if the
observer is not already familiar with them; these are listed in Table IB.
Another very valuable resource for help in answering questions that arise
before, during, or after a trial burn are the EPA personnel who are experi-
enced in trial burn observations and related issues. Some of these people are
listed in Table 1C along with their office phone numbers. They can be of
great help in answering questions or giving advice. Also, others who may be
able to lend assistance and advice are the air programs staff in regional and
state offices. They are usually well-experienced in many of the testing meth-
ods (e.g., particulate emissions) that are utilized in trial burns.
Probably the most difficult job for the permit writer/observer is using
trial burn results to develop the operating permit. Well in advance of the
trial burn, the permit writer should have already used the TBP, along with the
other guidance documents listed in Table IB, to prepare a draft operating
permit, and should have discussed it with the permit applicant to establish a
clear understanding of the anticipated operating limits for the incinerator.
Quite often, such discussions better define the conduct of the trial burn and
the data to be collected, and in some cases result in modifications to the
test scenario of the trial burn. The permit writer/observer can use this
draft to help ensure that all necessary data are collected and that the
results achieve the objectives of the trial burn. It may also aid in making
decisions about any changes requested after the observer is on site.
-------
TABLE 1A. TRAINING COURSES AVAILABLE
Air Pollution Training Institute (EPA training courses) n
Environmental Research Center, MD-17 i
Research Triangle Park, NC 27711
Andersen Samplers Inc. f
4215-C Wendell Drive 1
Atlanta, GA 30336
American Services Associates
15049 Bel-Red Road, Suite 100
Bellevue, WA 98007
-------
TABLE IB. REFERENCES AND GUIDANCE DOCUMENTS
1 "Guidance Manual for Hazardous Waste Incinerator Permits," USEPA, SW-966,
July 1983.
2 "Sampling and Analysis Methods for Hazardous Waste Combustion," A report
prepared by A. D. Little Inc. for USEPA, P884-155845, February 1984.
3 "Test Methods for Evaluating Solid Waste—Physical/Chemical Methods," USEPA,
SW-846, November 1986.
4 40 CFR 60, EPA Methods 1-10.
5 "Permit Writers Guide to Test Burn Data—Hazardous Waste Incineration,"
USEPA, EPA/625/6-86/012, September 1986.
6 "Practical Guide—Trial Burns for Hazardous Waste Incinerators," A report
prepared by Midwest Research Institute for EPA, PB86-190246, April 1986.
7 "Guidance on Carbon Monoxide Limits for Incinerator RCRA Permits," Midwest
Research Institute (tentative schedule 1989).
8 "Guidance for Continuous Monitoring of Carbon Monoxide at Hazardous Waste
Facilities," Pacific Environmental Services (Draft).
9 "Hazardous Waste Incineration Measurement Guidance Manual," Midwest Research
Institute, draft 1987 (document in progress).
10 "Guidance for Permit Writers for Limiting Metal Emissions from Hazardous
Waste Incinerators," Versar (tentative schedule 1989).
11 "Guidance on Setting Permit Conditions and Reporting Trial Burn Results"
(final is expected in early 1989).
12 "Proposed Methods for Measurement of CO, 02, THC, Hydrogen Chloride, and
Metals," Midwest Research Institute (tentative schedule 1989).
NOTE: For assistance in locating the above documents, contact Sonya Stelmack or
Lionel Vega (see Table 1C for phone numbers).
-------
TABLE 1C. EPA INCINERATION STAFF
• National staff
Larry Johnson, EMSL
Don Oberacker, HWERL
Bob Mournlnghan, HWERL
Sony a Stelmack, OSW
Bob Holloway, OSW
Lionel Vega, OSW
• Regional staff
Stephen Yee, Region I
Don Wright, Region II
John Brogard, Region II
Gary Gross, Region III
Betty Willis, Region IV
Y. J. Kim, Region V
Henry Onsgard, Region VI
Joe Galbraith, Region VII
Nat Miullo, Region VIII
Nina Churchman, Region VIII
John Hart, Region IX
Larry Bowerman, Region IX
Catherine Massimino, Region X
Area of expertise
Sampling and analysis
Engineering and emission
control systems
Engineering and emission
control systems
Policy and regulations
Policy and regulations
Policy and regulations
Office
phone no.
919-541-7943
513-569-7510
513-569-7430
202-382-4500
202-382-7936
202-475-8988
617-223-1925
201-321-6764
212-264-8682
215-597-7940
404-347-3433
312-886-6147
214-655-6785
913-236-2888
303-293-1668
303-293-1509
415-974-8142
415-974-8390
206-442-4153
FTS No.
629-7943
684-7510
684-7430
382-4500
382-7936
475-8988
573-9644
340-6764
264-8682
597-7940
257-3433
886-6147
255-6785
757-2888
564-1668
564-1509
454-8142
454-8390
399-4153
-------
Observing the process and sampling activities during the trial burn and
documenting these observations for the record is critical to developing final
permit conditions. Specific items regarding this phase are presented later in
this document, but the observer needs to be aware of some other important
aspects of observing trial burns and properly planning for the site visit.
One of these is that trial burns are usually not an 8-to-5 activity. They
often involve 12 to 16 hours per day for a number of days, including Saturday
and Sunday. This lengthy time period includes preparation for each test and
sample recovery. Also, process operating problems commonly contribute to the
time required for each test. Quite frequently, problems arise before, during,
or after each test that require decisions, or at least consultation with the
permit writer/observer. Therefore, the permit writer/observer should plan to
be on site at all times during the entire period of the test activity and
realize that this will likely involve long hours in the field. It probably
will also involve some decision making, which can be difficult, but can be
partly anticipated and prepared for before arrival. This is discussed further
in a later section. Also, there will likely need to be more than one observer
during the trial burn (i.e., one for process operations and one for sampling
activities, as has been done in many cases). It may be beneficial to have an
enforcement representative present for at least part of the trial burn to gain
familiarity with the incinerator operation in preparation for future
inspections.
-------
SECTION 3.0
PREPARATION FOR FIELD OBSERVATIONS
Preparations for field observations center on the need to become com-
pletely familiar with the approved test protocol to ensure quality data
regarding performance criteria and permit conditions, as well as coordinating
schedules and arrangements. The steps to be completed in preparing for the
field observation are described below.
3.1 REVIEW OF TRIAL BURN PLAN
A thorough review and understanding of the TBP and QAPP are necessary in
preparing for trial burn observations. One purpose of this is, of course, to
become familiar with the type of incinerator, as well as its design and com-
ponent parts (control devices, etc.). In addition, the emphasis in this
review is to formulate a clear understanding of what tests are to be
performed, what the incinerator operating conditions will be, and what samples
are to be taken and by what methods. The observer should be familiar enough
with the methods to know if the approved trial burn plan and methods specified
therein are being followed and should be able to answer questions like those
presented in Table 2 (see Note).
One of the most useful summary items normally contained in a TBP is a
table that lists each stream to be sampled, the sampling method, and the
analysis parameters for each sample (see example in Table 3). This table will
likely be referred to frequently and should be flagged or copied for easy
reference.
Note:
Table 2 and subsequent tables do not address metals sampling and analysis
which may be part of the TBP since EPA was considering regulation of CO
and metals emissions when this document was completed. See References 10
and 12 in Table IB.
-------
TABLE 2. CHECKLIST OF QUESTIONS TO BE ANSWERED
BY REVIEW OF TBP AND QAPP
A. Process
TBP/QAPP
page reference
1. How many different operating conditions are to be
tested?
2. What are the main differences in the test conditions
(and how might this be reflected in permit limits)?
Cond A
Cond B
Cond C
Cond D
3. Are 3 runs to be conducted at each test condition?
4. How long is each run?
How many runs per day?"
5. How many waste feed streams will there be during the
trial burn and what are their names?
1.
2.
3.
4. •
5.
6. Who is responsible for the recording of process data?
7. Which process data are to be recorded and how often,
and where is each monitor located in the process? (make
separate list)
8. What methods are to be used for determining waste feed
.• rates for liquids and solids?
Name Method
1.
2.
3.
4.
5.
9. Will liquid feed rates be checked based on tank level
change?
(Continued)
10
-------
TABLE 2 (Continued)
TBP/QAPP
page reference
10. What method will be used to indicate combustion gas
velocity?
11. If applicable, how are POHCs being spiked into waste
feeds?
12. What are the POHCs in each feed stream?
Name PQHCs
1.
2. '
3. II~~~~~~~~~~~
4. ~~~~~~~~~~
5.
13. What are the metals to be sampled and analyzed in each
stream or stack effluent?
Name Metals
1.
2. ~~~~~
3. HZHm
4. ~~~~~~~~~
5.
14. Is scrubber water recycled or once-through?
15. What and where is any material added to neutralize HC1
absorbed?
16. What types of instruments are used for continuous
monitoring of CO and 0,?
CO
17. What instruments are to be checked or calibrated prior
to the trial burn? How will this be documented?
(make separate list)
18. What instruments are to be checked or calibrated each
day of the trial burn (e.g., CO)?
How will this be documented?
(Continued)
11
-------
TABLE 2 (Continued)
I
I
B. Sampling >•
TBP/QAPP I
page reference
_ 1. What sampling method will be used for each feed stream? I
1. _
2. _ .
3. _
4. _ I
5. _
_ 2. What is feed sampling interval and amount sampled each |
time? (Indicate those to be composited.)
Interval Amount
2.
3.
4.
5.
3. What is scrubber water sampling interval and amount
sampled each time? (Indicate those to be composited.)
Name Interval Amount
4. What method will be used to sample ash, and what is the
sampling frequency and amount? Also, will ash be sam-
pled and composited during each run or sampled after
each run?
Method
Frequency Amount
What are the specific stack sampling methods to be
used?
MM5 for SV-POHC or PCDD/PCDF
M5 for particulate/HCl
VOST for volatile POHCs
Orsat for C02 and 02
(Continued)
12
-------
TABLE 2 (Continued)
TBP/QAPP
page reference
What individual samples are to be recovered from each
MM5 train?
Probe and front half rinse
Filter
XAD resin
Back half rinse
Condensate
Caustic solution
What is planned sampling time and sample volume for M5
and MM5 trains?
Time Volume
M5
MM5
What rinse solutions are to be used in recovery of M5
and MM5 trains?
M5
MM5
How many VOST trap pairs in each run?
Sampling time for each pair?
Sample volume for each pair?
Number of blank pairs per run?
10.
Are samples to be handled under traceability or chain
of custody?
Where is procedure specified?
11. Is all sampling planned to be done simultaneously? (If
not, list sequence.)
Comments:
13
-------
TABLE 3. EXAMPLE SUMMARY TABLE FOR SAMPLING METHODS AND ANALYSIS PARAMETERS
Sample
Sampling frequency Sampling
for each run methoda
Analysis parameter15
Liquid waste feed
Solid waste feed
Chamber ash
Scrubber water
influent and
effluent
Stack gas
Grab sample every 15 min S004
(composited)
Grab sample of each S006, S007
drum (composited)
Random sampling to col- S006
lect representative
composite after each
run
Grab sample every 30 min S004
(composite) and VOA vial
every 30 min
3-hr integrated sample
2-3 hr integrated
sample
Four pair of traps,
40 min each pair at
0.5 L/min
2-3 hr integrated
sample in Tedlar
gas bag
2-3 hr integrated
sample in mylar
gas bag
MM5
(0010)
M5
VOST
(0030)
S011
M3
V&SV-POHCs, Cl~, ash,
ult. anal., viscosity,
HHV, metals
V&SV-POHCs, Cl", ash,
HHV, metals
V&SV-POHCs, EP toxicity
V&SV-POHCs, pH, TDS
SV-POHCs
Particulate, H20, HC1,
metals
V-POHCs
V-POHCsc
CO2 and 02 by Orsat
(Continued)
14
-------
TABLE 3 (Continued)
Sampling frequency
Sample for each run
Continuous
Sampling
method4
Continuous
monitor
Analysis parameter15
CO (by plant's
monitor)
d VOST denotes volatile organic sampling train.
MM5 denotes EPA Modified Method 5.
M3 denotes EPA Method 3.
SXXX denotes sampling methods found in "Sampling and Analysis Methods for
Hazardous Waste Combustion," December 1983.
Methods 0010 and 0030 are from "Test Methods for Evaluating Solid Waste,"
SW-846, November 1986.
V-POHCs denotes volatile principal organic hazardous constituents (POHCs).
SV-POHCs denotes semivolatile POHCs.
HHV denotes higher heating value.
c Gas bag samples may be analyzed for V-POHCs, only if VOST samples are saturated
and not quantifiable.
15
-------
The TBP is intended to specify the test conditions and the sampling/
analysis methods. However, it must be realized that unforeseen circumstances
do occur and some changes from the TBP may be necessary after arriving on
site. These may be minor changes or major differences in sampling methods or
test conditions. For example, the TBP may involve two test conditions, but
the applicant may conclude that he cannot achieve a certain test condition.
He might therefore request deletion of testing at one of the test conditions,
even though the deletion will result in less desirable permit operating
limits.
The observer must make decisions, based on the on-site field situation,
about proposed changes or any alternatives by determining the following:
• Is the proposed change necessary to ensure that adequate sampling is
performed or that realistic operating conditions can be set?
Does the procedure or change compromise the test results (i.e., the
performance results)?
• Does the procedure or change impair ability to set necessary permit
limits?
For a new incinerator, are proposed changes within operating condi-
tions specified in the permit for the trial burn period?
In considering any change relative to the above criteria, the observer
can consider minor changes to be discretionary. However, major changes are
another matter because the TBP has presumably been the subject of extensive
review and negotiation prior to its approval, as well as public comment for
new facilities. Thus, even for interim status facilities, any major changes
must be made with caution and in consultation with appropriate EPA or State
staff. At new facilities, a major change could require the trial burn to be
delayed until after public notice of the changes, and no changes should be
allowed in operating conditions that are less stringent than those specified
in the trial burn permit.
Obviously, the observer cannot know all the possible procedural problems
or changes that may come up in the field, but can anticipate that such
problems probably will come up and prepare for them to some extent. That is,
in preparation for the trial burn, the observer can become familiar with guid-
ance documents and also be prepared with names and phone numbers (including
home phone numbers) for others who may be more experienced in trial burns and
sampling methodologies" or who may have higher supervisory responsibility or
authority. The observer should therefore prepare a list of names and phone
numbers (office and home) for people in his own agency who can help, as well
as his immediate supervisor. Of course, the EPA personnel listed previously
in Table 1 can also be of help. In certain situations, the guidance points
summarized in Table 4 may be helpful. Also, the Appendix to this document
briefly describes some of the actual problems that have been encountered and
their resolutions at the time. However, such problems and their resolutions
are very site-specific.
16
-------
TABLE 4. SUMMARY OF SPECIFIC GUIDANCE POINTSa
Three replicate runs are required for each set of incinerator operating
conditions. Results from each run must comply with RCRA requirements
for ORE, particulate emissions, and HC1 emissions (and any requirements
on CO and metals emissions).
Only one run per day should be scheduled, unless the sampling require-
ments are quite simple.
RCRA regulations call for continuous monitoring of combustion tempera-
tures, waste feed rate, CO, and the combustion gas velocity indicator,
and any other parameters as specified in the Trial Burn Plan.
Waste feed samples should be collected every 15 minutes and composited
over the entire period of stack sampling.
Each drum burned during the trial burn should be sampled and composited
unless the applicant can justify otherwise.
All other process samples (scrubber water, ash, etc.) should be taken
every 30 minutes over the entire period of stack sampling and
composited.
Sampling should not begin until the incinerator has reached steady state
on waste feed for at least 30 minutes.
Sampling should continue through incinerator operating abnormalities
unless the waste feed cutoff system shuts the incinerator down. If
sampling is stopped during a trial burn, the test may be completed using
the same sampling trains if the burn is completed on the same day it was
started.
Separate sampling trains should be used for semivolatile POHCs and for
particulates. This is necessary since drying the particulates and probe
rinse prior to weighing may result in loss of semivolatile POHCs.
Hydrogen chloride emissions may be determined based on analysis of
impinger aliquots from an M5 particulate train or a separate HC1 train,
but not from an MM5 train that includes the XAO resin trap.
t
Minimum stack sampling time for each run (actual sampling time not
including time for port changes, changing VOST trap pairs, etc.) should
be 1 hr for EPA Method 5 (M5), semi-VOST, and VOST. Data from less than
1 hr of sample collection would be an invalid test run. Two hours of
stack sampling time is recommended as optimal. A minimum of three VOST
trap pairs per run should be taken. A fourth pair is often taken in
case one pair is broken or lost due to analysis problems.
(Continued)
17
-------
TABLE 4 (Continued)
All sampling required for a test run in the trial burn should, whenever
possible, be conducted concurrently, with only the normal minor differ-
ences associated with different sampling methods (e.g., MM5 and VOST).
However, differences in sampling period start and finish due to sampling
problems are allowable (e.g., particulate train fails a leak check at
port change and sampling must be restarted with a new train, or a VOST
trap is broken at the end of the sampling period so another pair must be
run). However, all waste feed sampling must be continued for the entire
period, and possibly any water effluent or ash sampling. Also, the
incinerator must continue at the same process operating conditions with
collection of the operating data for the entire period.
The final leak check for VOST should be run at the highest vacuum used
during the sampling run but not less than 1 inch of mercury vacuum.
A sampling train which develops problems during a trial burn run may be
validated on a case-by-case basis if it can be shown that the results
were not significantly biased. For example, if an M5 train passed the
leak check at the end of the first port but failed the confirming leak
check at the beginning of the second port due to a probe liner being
broken during port change, the test could be allowed to continue after
replacement of the probe liner and including rinsing of the broken liner
for particulate recovery. However, if the train failed the leak check
at the end of the first port, the sample would be invalid, even if it
were believed that the probe liner was broken as the probe was removed
from the port (i.e., it is not possible to know if the probe liner was
already broken before removal from the port).
Volatile POHCs should always be sampled with the VOST if possible. Sam-
ples may be collected in bags if VOST samples cannot be performed. The
bag sample procedure is less desirable due to potential problems with
adsorption in the bag and loss of sample. Stability of the POHC to be
sampled in the bag should be checked prior to sampling, if this method
is used. Field blanks are essential with bag sampling.
VOST field blanks are required, and VOST trip blanks and laboratory
blanks are highly recommended.
--.
(Continued)
18
-------
TABLE 4 (Continued)
The front (Tenax) trap and back (Tenax/charcoal) trap must be analyzed
separately for each pair of traps from each run. Separate analysis of
each trap is required to check for indication of breakthrough. As a
guideline, breakthrough is not normally expected if the amount of ana-
lyte (POHC) collected on the back trap is less than 30% of the amount on
the front trap. For values higher than 30%, additional factors, such as
how close the calculated ORE is to the performance standard, need to be
considered. Specific criteria, including both a level where break-
through should be suspected and additional factors such as that
mentioned above are being developed for inclusion in Reference 9 on
Table IB.
Traceability procedures must be used for handling all samples. Full
chain-of-custody procedures are typically much more labor intensive but
may be used at the applicant's option.
The results of the analyses for particulate emissions, HC1 emissions and
removal efficiency, and ORE should be reported separately for each run,
and should not be averaged for the trial runs. This does not preclude
averaging multiple samples taken during each run.
VOST analytical results should be reported as an average value for each
run (as total ng/L of sample). This amounts to dividing the total
quantity (ng) on all traps by the total sample volume (L) for all
traps. Values for individual traps should also be reported in order to
evaluate the possibility of breakthrough.
The guidance points in this table will be presented in more detail in the
final version of Reference 9 listed in Table IB.
19
-------
It 1s expected that those performing the trial burn will inform the
observer about any proposed changes or problems and request his concurrence.
Also, the observer may identify some procedure or method variation of con-
cern. These should be discussed and advice should be requested from his
office or experts, if necessary. Usually they can be resolved on site without
too much difficulty. However, in preparing for the test, the observer must
recognize that it may be necessary to make some difficult decisions on the
spot, often under stress and at unusual hours. For example, if a number of
process upsets have occurred, the observer may have to consider the run
invalid. This can cause some consternation, especially when the decision is
not clear-cut. Therefore, it may be helpful if the observer recognizes this
possibility and decides in advance whom they may want to consult with and/or
who must be informed.
3.2 REVIEW OF PERMITTING NEEDS
The trial burn has two main purposes. The first is to demonstrate that
the facility can achieve the performance requirements for ORE, HC1, and par-
ticulate, and that there are no other concerns (e.g., fugitive emissions) or
problems. (It should again be noted that performance requirements for CO and
metals were being considered by EPA when this document was written.) The
second main purpose of the trial burn is to provide data necessary to
establish permit operating limits. Ordinarily, the performance requirements
are demonstrated in three replicate runs, each of which must show
compliance. It is possible, however, that problems might develop in
conducting the three replicate runs. The observer should be aware of this
possibility; and may need to "recommend" an additional run if there is doubt
about validity of one of the three replicate runs.
As mentioned in Section 2, preparing for the trial burn observation
should include using the information from the TBP and from a preliminary draft
of the permit operating conditions. One intention of this effort is to ascer-
tain that all parameters for which permit limits may be established will be
measured and reported in the trial burn results. A list of such parameters is
shown in Table 5, but others may be added to the trial burn plan at the dis-
cretion of the permit writer (see Ref. 11 in Table IB). It is partly for this
reason that a preliminary draft of the permit operating conditions should have
been prepared and discussed with the applicant before the trial burn. If a
preliminary permit . is drafted, the permit writer/observer will be better
prepared to examine how each parameter is measured and confirm that all
necessary data are being recorded during the trial burn. It is also helpful
in knowing how the data will be monitored and reported thereafter, so as to
enforce the limits imposed.
3.3 ARRANGEMENTS/SCHEDULING
Coordinating the efforts of all parties involved in the trial burn is
complex and important. Scheduling the test involves much coordination, and
the schedules often change. The permit writer/observer has many duties re-
lated to scheduling and coordination. A checklist of these duties is given in
Table 6, and some of these are explained further in the paragraphs below.
20
-------
I
I
1
I
TABLE 5. PARAMETERS FOR WHICH PERMIT
LIMITS MAY BE ESTABLISHED
Liquid waste feed rate(s)
% Cl or Cl input rate
% ash or ash input rate
Minimum HHV
Maximum heat input Btu/hr
Liquid feed atomizing fluid type
Liquid feed atomizing fluid pressure
Solid waste feed rate
Container size (or weight)
Volatile content
Frequency of feeding
% Cl or Cl input rate
% Ash or ash input rate
Indication of combustion gas velocity
Combustion chamber pressure
Operating temperature
CO concentration
02 concentration
Control device
specific parameters:
Venturi AP
Venturi water feed rate
Packed tower water feed rate
Scrubber water blowdown rate
Scrubber water pH
I
Baghouse AP
Baghouse cleaning cycle time
Spray tower reagent flow rate
Spray tower atomizing speed or pressure
Spray tower inlet gas temperature
ESP voltage
ESP amperage
ESP spark rate
ESP rapping rate
Note: Liquid waste viscosity and burner
turndown ratio may be permit limits
but are usually based on manufac-
turers' data for the feed nozzles.
21
-------
TABLE 6. CHECKLIST OF SCHEDULING/COORDINATION ACTIVITIES
Facility Name: Date completed
or comment
Call the permit applicant to find out the following:
a. Who is responsible for the trial burn activity and scheduling?
b. Who is the lead person for the field sampling (i.e., sampling and
analysis contractor)?
c. Who is your main contact at the test site?
d. Get directions to the test site.
Call all of the above to inform them of your plans and
discuss coordination of your activities.
Determine if you will be required to sign anything upon
arrival at test site. If so, ask that copies of forms
be sent to you for review, since you should not sign
secrecy agreements or injury waiver forms. You must
arrange for access to the site without signing such forms.
(Contact your Office of Regional Counsel or corresponding
State office, if needed.)
Arrange for a meeting to discuss preliminary draft permit
parameters and conditions, if that has not already been
discussed with the applicant.
Try to determine what other regulatory personnel plan to
be on site for the trial burn. Make sure qualified
personnel will be present to observe all the procedures,
if the permit writer/observer is not qualified.
Establish one of the regulatory personnel (permit writer)
as a coordinator and primary contact for plant problems
and decisions.
•%
Arrange for EPA audit cylinder, if required, or check
that arrangements have been made.
(Continued)
22
-------
TABLE 6 (Continued)
Facility Name: Date completed
or comment
Arrange for field audit or obtaining audit samples, if
required. Be sure auditors know directions to test
site, and are kept informed about changes in test
schedule.
Arrange for pretest briefing to take place as early as
possible after arrival on site. (Prepare notes on items
to be discussed in prebriefing per Table 7.)
Make all travel arrangements (but it is wise to make
return plans flexible).
Determine any special arrangements or requirements at the
test site (e.g., safety) or permission required for taking
a camera if desired.
Obtain safety shoes, safety glasses, and hard hat. Plan
on foul weather and getting dirty.
Obtain training and certification in use of respirators
or safe breathing apparatus that might be needed.
Comments:
23
-------
Two items (if they are needed) must be arranged well in advance of the
testing. One is the EPA audit cylinder. These cylinders are available from
EPA and are encouraged for use in auditing VOST sampling and analyses. If
VOST is being used, the TBP or QAPP should indicate if an audit cylinder is to
be used. Use of an audit cylinder should be arranged at least 30 days in ft
advance. Also, the permit writer/observer may wish to confirm that those ||
performing the sampling know that they will need to take samples from the
audit cylinder as specified in the TBP or QAPP, because additional VOST _
sampling traps must be prepared. I
The audit cylinders are available only to EPA and state agency staff and
their contractors. The cylinders cannot be ordered by private organizations, 1
so the permit writer must make these arrangements with the EPA contact as I
shown below. When doing so, the permit writer will need to provide
information on the facility, purpose of the audit, POHCs and their expected «
concentration, and to where and whom to send the audit cylinder. |
Audit Cylinders Available From:
Mr. Robert Lampe •
U.S. Environmental Protection Agency
Environmental Monitoring Systems Laboratory •
Quality Assurance Division y
Research Triangle Park, NC 27711
Phone: (919) 541-4531 .
The second item that must be arranged for, if needed, is any external "
field auditing of sampling activity or the need to -obtain field audit sam-
ples. Field auditing, in this context, requires special equipment (e.g., I
critical orifices) and special expertise. EPA sometimes uses experienced I
sampling contractors who have the necessary equipment. Therefore, initially
determine if field auditing is necessary or if it might have already been j
arranged. The need for such auditing may depend on the experience level of |
those conducting the trial burn testing, or sensitive issues surrounding the
test that make auditing advisable. If field auditing is necessary, the permit
writer must make all the contractual arrangements at least 30 days in J
advance. Also, he should develop an audit plan with the contractor to
determine exactly what is to be audited, so that all the necessary items can
be prepared (e.g., gas cylinders of known concentrations, critical orifices,
etc.).
Field audits may .involve taking of audit samples or check samples. This
also requires considerable planning and arrangements because all sample bot-
tles must be properly prepared, handled, stored, and shipped. Also, the trial
burn sampling contractor needs to know about the plan to take audit or check
samples since his help will likely be needed as part of that activity.
24
-------
It 1s emphasized that the Idea of taking audit samples needs to be
thought out well in advance, very early 1n the trial burn planning process.
In this regard, the permit writer/observer needs to decide:
What audit samples need to be taken and why?
How many need to be taken?
What are they to be analyzed for and by what methods?
Who will analyze them?
Again, 1f the permit writer/observer deems 1t necessary to obtain audit
samples, specific arrangements need to be made at least 30 days in advance of
the testing, along with a clearly formulated plan for this activity. This
will allow sufficient time for the auditors to prepare all the necessary
equipment and prepare sufficient sample containers that have been properly
precleaned.
Besides the preliminary drafting of the permit conditions, another key
item listed 1n Table 6 is the pretest briefing that needs to be done soon
after the observer arrives on site. A suggested outline of items that should
be discussed 1s given in Table 7.
Another key item is making final travel arrangements. The observer
should plan to arrive on site before noon on the day before the first run is
scheduled. This should provide sufficient time to meet the responsible
individuals on site, become oriented with the process and all data recording
and sampling locations, and hold the pretest briefing.
Several days prior to leaving for the site, the observer should refer to
Table 8, which provides a checklist of items that usually need to be taken to
the site or checked before leaving.
25
-------
TABLE 7. OUTLINE FOR PRETEST BRIEFING
• Give name, title, and agency.
• Explain your responsibility regarding trial burn:
- Assessing performance results
- Writing the operating permit
• Explain what activities you plan to observe:
- Taking of process data
- General plant operating conditions and procedures
• Liquid feed sampling
- Solid feed sampling
- Scrubber water sampling
- Stack testing preparation
- Stack testing procedures
- All sample recovery activities
• If your responsibility Includes observing certain test activities for which
you have very limited knowledge or experience (e.g., MM5 stack sampling),
you may want to explain this fact and ask for their help in understanding
the procedures.
• Explain that you expect all the sampling to be done in accordance with the
procedures given in the TBP and QAPP. Solicit advice about any deviations
that any attendees may already be aware of.
• Ask who 1s responsible for logging of process data.
• Inform them that you need continuous recorders clearly marked with run
number and date, and marked to show start/stop time of each run, and any
periods when sampling was stopped. Let them know you want to initial some
or all recorder sheets or data sheets, and specify them in a list.
• Ask them to seek you out to discuss any procedural variations or any
changes in the testing that may come up so that possible problems can be
resolved before they jeopardize the results.
• Explain that you and everyone else are there to obtain valid results.
• Tell them that you will Inform them immediately if you observe anything
that you believe could jeopardize acceptability of results. Ask whom to
contact. Also, find out the name of the crew chief.
• Raise any questions or procedural problems that may have come up during the
tour of facility.
• Determine plans and schedule for the next day's testing and all succeeding
tests.
• Inquire as to what safety equipment 1s required in what areas of the plant.
26
-------
TABLE 8. CHECKLIST OF ITEMS TO BE TAKEN ON SITE
• Copy of TBP and QAPP
• Copy of all sampling methods referenced 1n TBP and QAPP
• Copy of 40 CFR; Parts 264 (subpart 0) and 270
• List of names and phone numbers of authorized personnel who can provide
guidance and answer questions about trial burn procedures and sampling
methods
• Copy of table summarizing sampling locations and methods
• Copy of this observation guide and extra copies of tables to be used 1n the
field, with completion of entries that can be made before arrival on site.
• Directions to motel and from motel to test site
• Name of person to contact upon arrival at site
• Checklist of Items to discuss in pretest briefing
• Bound notebook and Indelible ink pens (black) for recording data and all
observations, etc.
• Hard hat, safety glasses, safety shoes
• Bad weather gear
• Camera and plenty of film (prior arrangements usually must be made for
taking a camera on site)
• Check to see 1f EPA audit cylinder has been delivered or arranged
• Before leaving, call to be sure the testing is "on schedule." Reaffirm
schedule with any others who need to know (e.g., auditors).
Comments:
27
-------
SECTION 4.0
ON-SITE ACTIVITIES
Once on site, the permit writer/observer has many activities to perform
and many Items to cover. Participation 1n on-site activities and attention to
all these Items are critical. These activities and Items are discussed 1n
this section in the following order:
Pretest orientation
Observation of process operations
Observation of field sampling/analysis activities
Personal safety
QA/QC audits
Major deviations/problems
Documentation of activities
4.1 PRETEST ORIENTATION
The permit writer/observer should first tour the incinerator facility
with someone who 1s knowledgeable about the process and, if possible, knowl-
edgeable about the TBP (i.e., the test conditions and the sampling loca-
tions). This also usually presents the opportunity to meet key participants
who have lead responsibility on site, including:
• Permit applicant's representative
Process operations (manager, operators)
• All sampling activities (project leader)
• Stack sampling activities (crew chief)
After the initial tour, it is advisable for the permit writer/observer to
take another review tour of the incinerator, either escorted or unescorted, in
order to concentrate on the details of -the upcoming test and the TBP. Some of
the objectives and Items to be covered in that tour are listed in Table 9.
Usually, the Incinerator tour and efforts to cover the Items listed in
Table 9 raise some questions for the permit writer/observer. The observer
must then seek out the responsible personnel to try to answer the questions
and resolve any procedural discrepancies or changes.
29
-------
TABLE 9. OBJECTIVES OF INCINERATOR FACILITY TOUR
Observe general process operations and personnel.
Identify each instrument from which process data will be recorded (scale,
units of measure, instrument number, color of ink for each parameter on
strip chart) (see Table 11). h
Examine log sheets to be used by personnel recording process data.
Determine who will be responsible for marking each strip chart to identify |
correct time, test periods, and instrument number.
^
Determine exact location of each sampling point, how samples will actually •
be taken, and by whom. *
Determine location of continuous monitors for CO, etc., and make arrange- ||
ments to be present to observe calibration procedures. •
Examine stack sampling location. Determine what test ports will be used h
for M5, MM5, VOST, and Orsat. Jj
Prepare sketch of stack sampling location.
Tour field laboratory. Observe preparation of stack sampling trains or ™
find out when these preparations will take place so you can be present to
observe. Investigate method of determining isokinetic sampling rate for MS m
and MM5 sampling trains. |
Determine location and method of storing samples. _^
Determine labeling system for samples. *
Find out how runs will be numbered so that your records can be consistent •
with sampling data. I
Record names, addresses, and phone numbers of all key participants.
I
I
30
-------
Soon after arrival on site, the observer will need to arrange the time
and place for a pretest briefing with all the responsible participants on
site. It should take place as early as possible, preferably before the end of
the observer's first day on site. In this pretest briefing, the points listed
previously 1n Table 7 need to be covered. It will also be a good time to
I bring up any questions, deviations, or problems that the observer may have
I after his tour of the facility. The final schedule for the next day's testing
and tentative schedule for succeeding tests should also be clarified at this
r time.
4.2 OBSERVATION OF PROCESS OPERATIONS
I The key to observing process operations is to circulate through the fa-
1 cillty, keeping your eyes and ears open. Try to keep questions to a minimum
once testing has begun. The observer should make every effort to have any
{questions answered, without being a nuisance or interfering with people's
jobs. Usually, someone will be available who 1s not directly Involved but
will be able to answer questions. Avoid lengthy discussions with operators or
I other observers on site during the testing so as not to be distracted from the
ongoing activities.
. When observing process operations, concentrate on the process data that
I will be used in setting permit limits and the process data critical to the
"* performance requirements. For the latter, the waste feed rates are the im-
portant process parameters used in calculating ORE. Therefore, the permit
(writer/observer should examine all the feed rate monitoring systems and
utilize the checklist shown 1n Table 10.
(The other critical part of observing process operations is ensuring that
all the data are being collected for those parameters which will be specified
as permit operating limits. A list of those parameters should be available if
a draft of permit conditions was prepared (see Table 5). They should also
H have been listed in the TBP. These sources may be used to prepare a process
' data logsheet using the logsheet form given as Table 11 and Table 12, and use
the form 1n Table 12 to record the readings for the critical process param-
Ieters and compare them against the expected values for the trial burn. Simi-
larly, logsheets for the air pollution control equipment parameters should be
used to record data each hour, or more often if possible. Example logsheets
for different types of control equipment are given in Tables 13A through 13C.
I
I
I
Some of the process instruments will be connected to automatic shutdown
devices, and the permit writer/observer may request a demonstration of some of
these automatic shutdowns as part of his observations during the trial
burns. This needs to be arranged in advance and can usually best be done
sometime shortly after a run is completed. Be aware that an unplanned
activation on one or more of the automatic shutdowns may occur inadvertently
during the testing.
31
-------
TABLE 10. CHECKLIST FOR WASTE FEED RATE MONITORING SYSTEMS
Facility Name: _ Date completed
or comment
Liquid feed
• Is the primary measurement device downstream of any _
feed recycle piping takeoff (or are such recycle
lines clearly blocked off)?
• Has the Instrument recently been zeroed and spanned? _
(Examine records to verify this.)
Has it been possible to conduct an actual calibra- _
tion of the instrument recently? If so, examine
calibration data and obtain copies. (This is
usually difficult, since it would require revisions
in the piping to direct the flow into a drum or
other vessel and weighing the amount collected in
the drum over a short time period. A good alterna-
tive is to cross-check the waste feed rate meter
with the tank level change during each test as
mentioned below.)
• What type of primary measurement device is used, _
and are measurement results a function of any waste
feed characteristics (I.e., specific gravity, vis-
cosity, temperature)?
Will data be collected to cross-check waste feed _
rate based on tank level changes? (This is usually LI
done and is quite Important.) •*
Does measurement readout agree closely with the _ M
value on strip chart recorder? Is the flow rate pi
near that specified in the TBP?
Solid feed
If batch feed? has the scale been recently call- _
brated with check-weights?
_
B
If batch feed, what is the approximate weight of _
each container and the frequency of feeding? _ m
(Continued) —
I
32
I
I
-------
TABLE 10 (Continued)
Facility Name: Date completed
or comment
• If continuous feed, has the weighing device re-
cently been zeroed and spanned? (If so, examine
records.)
If continuous feed, has the weighing device re-
cently been calibrated? (Difficult to do but
should be feasible.)
What does the device show as a reading when there
is no material being fed (e.g., empty belt feeder)?
Is there any way to obtain some gross check on
waste feed rate, such as a change in level in a
feed hopper?
Can each feed source be weighed before and after
each test (e.g., sludge fed from a tank truck)?
Comments:
NOTE: Many of 'the above items (e.g., calibration procedures) should have been
specified beforehand in the TBP.
33
-------
TABLE 11. RECORD OF PROCESS DATA INSTRUMENTS
Facility Name: Date °f •
Instrument Units Range Pen most recent
Namea No. or ID (e.g., °F) (e.g., 0-150) color calibration^
1. Organic liquid
feed rate
2.
3. Aqueous liquid
feed rate
4.
5. Solids feed rate
6.
7. Atomizing fluid
pressure
8. Combustion gas
velocity indicator
9. Combustion chamber
pressure
10. Combustion chamber
temperature
11.
12.
13. CO cone
14. 02 cone
15.
16.
17.
NOTE: Identify those instruments which are tied into automatic shutdown systems and •
record alarm and shutdown settings. g
a Numbers that are blank are provided for multiple feeds or other critical process -
instruments. I
34
I
-------
TABLE 12. PROCESS DATA LOGSHEET
Facility name Date
Run No.
By
Expected Observed values
Instrument value ^}{3T
Parameter No. per TBP
1. Organic liquid
feed rate
2.
3. Aqueous liquid
feed rate
4.
5. Solids feed rate
6.
7. Atomizing fluid
pressure
8. Combustion gas
velocity indicator
9. Combustion chamber
pressure
10. Combustion chamber
temperature
11.,
12.
13. CO cone
14. 02 cone
15.
35
-------
TABLE 13A. EXAMPLE LOGSHEET FOR DRY SCRUBBER/BAGHOUSE/ESP
Facility name
Parameter
Instrument
No.
Expected
value (
per TBP
Date
Run No.
By
Observed values
) ( ) ( )
Dry scrubber
Reagent feed rate
Atomizer speed
Nozzle pressure
Inlet temperature
Baqhouse
Pressure drop
Inlet temperature
Cleaning cycle
ESP
Secondary voltage
Secondary amperage
Spark rate
Rapping rate
-------
TABLE 13B. EXAMPLE L06SHEET FOR VENTURI/PACKED TOWER
I
I
I
I
Facility name
Parameter
Expected
Instrument value (
No. per TBP
Date
Run No.
By
Observed values
) ( ) ( )
VentuH
Pressure drop
Water flow rate
Effluent pH
Inlet temperature
Outlet temperature
Packed Tower
Pressure drop
Water flow rate
Effluent pH
Inlet temperature
37
-------
ft
TABLE 13C. EXAMPLE LOGSHEET FOR IONIZING WET SCRUBBER
======^= ft
Facility name Date
Run No. fc_
By g
Expected Observed values
Instrument value "{ j T"^ 5 I F
Parameter No. per TBP
Inlet gas temperature
Water flow rate
Pressure drop ^
Effluent pH
First stage
AC volts
AC amps
Spark rate
DC volts
DC MA
Second stage
AC volts
AC amps
Spark rate
DC volts
DC MA
I
I
1
I
I
38
I
-------
While observing the process operating conditions, the observer may need
to consider some other criteria or items that may be specified in the permit,
such as the following:
• Will data records be sufficient to determine variability in readings
during the trial burn periods (I.e., how much did the parameter vary
during the tests)?
• What permit conditions may need to be set (e.g., chart speed) to
ensure adequate data records for enforcement during continued
operation?
• Is there any reason to believe there might be a wider range in
variability of parameter during normal operations after the trial
burn?
• If a device is tied into the automatic shutdown system, what would
be a reasonable value for activating shutdown (i.e., how much above
or below the normal operating range)?
• Would a time delay period be reasonable before activating automatic
shutdown to avoid frequent shutdown caused by very brief excursions
in parameter (spikes)?
• What is a practical frequency at which the automatic waste feed
cutoffs should be checked under continued operation if the applicant
has requested to test these systems less often than on a biweekly
basis?
Other important aspects of observing process operations are somewhat sub-
jective. One of the observer's responsibilities will be to make note of any
fugitive emissions (e.g., puffing from kiln seals). Other subjective aspects
of the observation process are listed in Table 14. Any problems observed
should be noted in the observer's logbook.
One other important aspect of observing process operations is the fre-
quency that the desired operating parameters are outside the range of the de-
sired test conditions and the number of shutdowns that may occur during test
periods. These sometimes occur to the extent that the observer may feel the
test data are compromised and the test should be repeated. A few of the guid-
ance points that are contained in Table 4 may be helpful for problems of this
type. However, they are usually site-specific and depend on judgments by the
observer, considering the: criteria listed earlier on page 13.
39
-------
TABLE 14. GENERAL OBSERVATIONS OF PROCESS OPERATION
Facility Name:
• Do operators seem to be well trained and experienced?
• Are all personnel Informed about the objectives of the
trial burn and the TBP?
• Do all personnel conduct themselves in a professional ^
manner? II
• Is there a process operating logbook? Are entries
legible and understandable? H
• Is equipment checked and inspected regularly?
• Does equipment appear to be well maintained? ||
• Are all lines and equipment clearly marked? .
• Are safety facilities and equipment available (e.g., "
safety showers, fire extinguishers, etc.)?
• Are all leaks and/or spills cleaned up in a safe and V
expeditious manner?
Comments:
40
I
I
I
I
I
I
-------
One of the most common problems facing the observer is process problems
that delay startup of a test or a "temporary" shutdown 1n which sampling has
been stopped. These temporary shutdowns seem to frequently consume several
hours, and the observer may need to set a time limit after which the test is
considered an "abort" and must be repeated. In this regard, one of the
guidance points 1n Table 4 Indicates that if sampling is stopped during a
trial burn, the test may be completed using the same sampling trains, if the
burn is completed on the same day it was started. However, 1n cases where
several shorter duration shutdowns occur, the permit writer/observer might
also determine the need to invalidate the run 1f he/she believes 1t 1s neces-
sary based on the criteria on page 13. Some examples of such problems are
included in the Appendix.
4.3 OBSERVATION OF FIELD SAMPLING/ANALYSIS ACTIVITIES
Observing actual field sampling is a critical part of the observer's
responsibility, especially the stack sampling activities. The stack sampling
activities are vital to assessing the performance results.
The observer should observe all the sampling activities, besides the
stack sampling, on an intermittent or random basis. He/she should accompany
the samplers as often as possible on all of their sampling rounds to verify
that the samples are being taken in accordance with the test plan and sampling
methods specified. He/she should ascertain that the sampling locations are
appropriate and the procedures appropriate for obtaining representative sam-
ples. The observer should observe this sampling randomly during each test and
use the checklist in Table 15 to help evaluate this sampling.
Observing stack sampling activities is critical because that is where
sampling problems are most likely to occur. This activity usually involves
simultaneous sampling by Orsat, CEMs, VOST, MS, and MM5. These methods are
discussed 1n the following paragraphs.
Orsat is a bag sampling method used to collect samples for subsequent
analysis by Orsat apparatus for 02 and C02. One or more such bag samples may
be collected over approximately the same period as the other sampling pro-
cedures. The observer should inspect this apparatus and the sampling line
before the testing. During sampling, the observer should intermittently
examine the apparatus to verify that sampling is occurring at a relatively
constant rate, and the bag is not overfilling. After the test, the observer
should observe the field analysis of a bag sample at least once.
CEM instruments are usually a part of the process instrumentation but
they are sometimes provided and operated by the sampling contractor for
verification. The observer's main Interest here is to check that representa-
tive samples are reaching the Instruments and that the system is being zeroed
and spanned before and after each run (a common practice).
41
-------
TABLE 15. GENERAL OBSERVATIONS FOR WASTE FEED AND WATER SAMPLING
Facility Name:
Is the sampling being done in accordance with the methods
specified in the TBP and at the specified frequency?
Are the samples taken in a manner that ensures their
representativeness (e.g., lines purged before sampling;
VGA vials are bubble free).
Are the samples taken in a manner that minimizes chances
of contamination?
Are the samples stored properly (e.g., iced)?
Are the samples stored in a manner that minimizes cross- k
contamination (e.g., high concentration feed samples m
separate from low concentration scrubber water samples)?
Are log sheets filled out to show sampling times for I
each sample, sampler name, date, run number, etc.?
Are samples properly labeled and labels protected from 1
becoming illegible? •
Are traceability records being initiated and maintained m
for each sample? |
If samples are to be handled under chain-of-custody, are ^
the proper procedures being followed? Ji
Comments:
I
I
I
I
I
-------
In most cases, the CEM gas sample 1s withdrawn from the stack through a
sample line to a remote CO (or 02) Instrument. A second, smaller line 1s com-
monly attached so that span gas can be Injected near the stack location rather
than at the Instrument. This serves as a check against possible leaks in the
sample line. Even so, the observer should request a leak check of the sample
line and observe the calibration of the CEM instruments, as well as requesting
a brief test to check the response time for the extractable CEM system.
The observer should also intermittently examine CEM instrument data read-
outs and data recording. He should note the average reading and the extent
and frequency of spiking in the readings. In some cases, CO readings spike in
concert with batch feeding of solIds. The observer should be aware of guid-
ance on CO limits, since revised regulations were being considered when this
document was prepared.
VOST sampling is almost always the method used for quantifying emissions
of volatile POHCs for calculation of their ORE (see Figure 1). This method
involves changing trap pairs every 20 to 40 rain, whichever may have been spec-
ified in the TBP. Leak checks must be performed before and after the sampling
period for each pair. Therefore, the observer should be present to observe at
least the final leak check on some or all pairs. The most recent VOST method
(Method 0030 1n SW-846) allows a leak rate of 2.5 mm Hg after 1 rain.
The VOST method requires that the VOST traps be kept cool (with ice) and
protected from contamination. This should be verified by the observer. Also,
the fragile VOST traps may break from time to time. If a trap 1s broken when
removed from the VOST apparatus, the sampling may have to be repeated to
obtain the specified number of trap pairs in each run (three pair minimum, see
Table 4). This, of course, extends the sampling time. It also means that one
or more pairs may not be coincident with other sampling (e.g., MM5), but that
should be acceptable as long as the process continues to operate In a stable
manner. However, waste feed sampling and collection of operating data should
also be extended to cover the entire sampling period for VOST, even If MM5
sampling has been completed earlier.
M5 and MM5 sampling methods are very similar, but the trains contain
different impinger solutions, and the MM5 train includes a condenser and XAO
resin cartridge. MS is used to determine particulate and HC1 emissions while
MM5 is used to quantify semivolatile POHCs and their ORE. Figure 2 depicts an
M5 train for determination of particulate and HC1 emissions. The M5 train may
also include impingers for collection of vaporous metals (see Reference 12 in
Table IB). Figure 3 depicts an MM5 train. The configuration shown in Fig-
ure 3 may be preferred,- but optional configurations are also acceptable to
EPA, as shown in Figures 4 and 5. Other configurations may also be accept-
able, provided that there is no appreciable retention of liquid (condensate)
upstream of the resin cartridge and that all the liquid flows vertically down
through the resin during sampling. In any case, the train configuration,
impinger solutions, and recovery procedures should have been specified in the
trial burn plan, and the observer needs to verify that these procedures are
being followed.
43
-------
Purge
Valve
kJ
Sample
Valve
\
Charcc
Tube Valve
/
Charcoal
Filter
/
Stack .?
Gas In
Teflon Tubing
Latex Tubing
Tenax/Charcoal
Trap
Fittings A, B. C, and 0
are VJton O - ringed
Nickel Plated Fittings
Silica Gel
1
1
1
1
1
1
1
1
1
1
__ 1
1
I
1
Figure 1. Volatile organic sampling train (VOST).
Remote Ice Bath
with Submersible
Pump
^ Sampling
Console-
-------
T/C
Ul
Quartz/Glass Liner
Thermocouple
Nozzle
Reverse - Type
Pilot Tube
Potentiometer
Filter
T/C
Check
Valve
Sample Box
Silica Gel
Manometer Probe
T/C T/C Fine Control
n • • Valve
Main Valve
Airtight
Pump
0 Greenburg-Smith, 100mL Distilled H2O
© Greenburg-Smith, 100ml 0.1N Caustic Solution
(3) Modified Greenburg-Smith, Empty
@ Modified Greenburg-Smith, SiO2
Figure 2. Method '5 partlculate and HC1 sampling train.
-------
Quartz/Glass Liner
Note:
Condenser and XAD resin cartridge
may be connected separately.
as two components, with the frit
at the bottom of the resin cartridge
rather than at the top.
Nozzle
Reverse - Type
Pitot Tube
o»
Heated
Area T/c
Potentiometer
Recirculation Pump
T/C T/C Fine Control
Valve
Main Valve
Airtight
Pump
Modified Greenburg-Smith, 100mL of Double Distilled in glass H2O
Greenburg-Smith, 100ml of Double Distilled in glass H2O
(3) Modified Greenburg-Smith, Empty
Modified Greenburg-Smith,
Figure 3. Modified Method 5 sampling train (MMS)--type 1.
-------
Quartz/Glass Liner
\
Thermocouple
Nozzle
Reverse - Type
Pilot Tube
Cyclone (Optional)
Potentiometer \ Filter
T/C
Check
Valve
Silica Gel
Vacuum Line
T/C T/C Fine Control
M i • Valve
Modified Greenburg-Smlth, Reversed. Empty
Modified Greenburg-Smith, 100mL of Double Distilled in glass H20
(3) Greenburg-Smith, 100mL of Double Distilled in glass H2O
@ Modified Greenburg-Smith, Empty
(5) Modified Greenburg-Smilh, SiCfe
(A) Condenser
® XAD Resin Cartridge
* Ice Water Jacket
Figure 4. Modified Method 5 sampling train (MM5)—type 2.
-------
Cyclone (Optional)
Quartz/Glass Liner
Nozzle
Reverse - Type
Pilot Tube
T/C
Check
Valve
MML
Condenser with Ice Water Jacket "1 See Fig SB
XAD Resin Cartridge with Ice Water Jacket, (65 g of XAD resin) J
Modified Greenburg-Smith, 100mL of Double Distilled in Glass
Greenburg-Smith, 100mL of Double Distilled in Glass H^O
Modified Greenburg-Smith, Empty
Modified Greenburg-Smilh, SIO2
Figure 5A. Modified Method 5 sampling train (MM5)«type 3.
-------
From
Thermocouple
Well
a
To
Impinger
Submersible
Pump
Water In
Water Out
Condenser
XAD-2
Figure SB. Modified Method 5 sampling train (MM5)~type 3.
49
-------
Checklists that can be used by the observer for the major stack sampling
methods are given in Tables 16A through 160. Observing the MS and MM5 trains
is a key function of the observer because of their importance and potential
problems. First, the observer should watch the trains being prepared to
ascertain that they are in accordance with the TBP and method specified.
(Only Teflon and glass are permitted to be used in the MM5 train; no grease is
allowed for sealing ball joints, etc., except after the silica gel impinger.)
Observers usually prefer to watch the M5 and MM5 sampling throughout much
of the trial burn's duration, especially the port changes and leak checks (see
Table 16C). Leak checks are performed initially, before and after each port
change, and at the end of the sampling period. A final leak check is also
required on pitot tubes used in the trains. The allowable leak rate on MS and
MM5 trains is usually 0.02 cf in 1 min (see EPA Method 5 for details).
Slightly higher leak rates sometimes are acceptable, if the corrected sample
volumes do not cause the isokinetic sampling value to exceed 100% ± 10% (see
Section 6.3 of EPA Method 5 for details). If any of the leak checks are
unacceptably high, the sample will be invalid requiring another train to be
brought up to the stack and sampling restarted. This, however, means that
sampling times for the run may be extended significantly. More importantly,
it means that the sampling will not be coincident with other sampling (e.g.,
VOST) and that waste feed sampling, etc., must be extended, or an additional
set of all samples taken to cover the new sampling period for the MM5 train.
It is therefore helpful if a course of action or contingency plan is developed
beforehand to cover such situations (see Note).
Another important part of the MM5 sampling is the sample recovery pro-
cedures. The observer should observe the probe rinsing operation to be sure
it is done in accordance with the specified method using the specified rinse
solutions. He/she should also observe recovery of the filter/impinger sam-
ples, which is a tedious and lengthy process. It is this recovery that prob-
ably represents the highest potential for contamination of samples. Some of
the points to be observed in MM5 recovery are listed in Table 160.
Note:
It is desirable for all sampling to be coincident, but may not be a
necessity if the process operates stably, within the trial burn test
conditions, throughout the entire test period. In fact, some tests have
been done with sequential sampling due to limitations in the sampling
ports or other reasons. Variations may also occur in specific situations
(e.g., failure of leak check at port change).
50
-------
I
I
TABLE 16A. ORSAT SAMPLING CHECKLIST
. Facility Name:
• 1. Was the bag leak-checked before use?
12. Was the sample line leak-checked and purged before
start of sampling?
(3. Was sampling continuous over the duration of most of
the MM5 sampling?
4. Was sampling done at a reasonably constant rate?
* 5. Were the bag samples analyzed within 4 hr after
sampling?
I 6. Was the absorbing solution checked (e.g., 02 analysis
of ambient air)?
Comments:
51
-------
TABLE 16B. VOST SAMPLING CHECKLIST
I
I
Facility Name:
1. Were the traps kept cool before, during, and after I
sampling (< 20°C)?
2. Was each trap pair leak checked before and after I
sampling?
3. Was a glass or Teflon heated line used for sampling? j
4. Was the sample line leak-checked and purged before
the start of sampling? I
5. Were the required number of trap pairs taken?
6. Were the specified blank traps taken?
7. Was sampling data logged for each trap pair?
8. Was proper sample volume taken in each trap pair
(20 L)?
9. How is each trap pair identified (e.g., sample number)?
Comments:
52
-------
TABLE 16C. M5/MM5 SAMPLING CHECKLIST
Facility Name:
1. Was the sampling train properly prepared (Teflon/glass
and no grease used)?
2. Was the manometer system (console) leveled and zeroed
before sampling?
3. Is the probe properly marked for each sampling point?
(see EPA Method 1)
4. Was a check made for cyclonic flow in the stack?
(see EPA Method 1)
5. Is the probe and sample box maintained at proper
temperature (250° ± 25°F)?
6. Were there any problems with high vacuum in the train
(e.g., 15-20 in. Hg) which made it difficult to main-
tain the required isokinetic sampling rate?
7. Were sampling rate changes (adjustments) made in a
timely manner?
8. Was proper temperature maintained at the inlet to XAD
resin (< 20°C)?
9. Was the train leak checked before and after each port
change and a final leak check made? (Allowable leak
rate is 0.02 cfm or 4% of the average sampling rate,
whichever is less.)
10. Were the sampling ports adequately plugged during
sampling?
11. Was stack static pressure properly measured?
12. Were the pi tot tubes leak checked?
13. Was the total sampling time, and time at each point,
the same as that specified in the TBP?
(Continued)
53
-------
TABLE 16C (Continued)
14. Were the nozzle and other openings covered with foil as
appropriate to protect from contamination?
15. Were data sheets filled in completely during sampling?
Comments:
8
I
I
I
I
I
I
54
-------
TABLE 160. M5/MM5 RECOVERY CHECKLIST
Facility Name:
1. Was the probe properly rinsed and brushed, with
specified reagents, while protecting the sample from
contamination?
2. Were the train components disassembled to protect
against loss of sample and done in a clean area to
minimize contamination?
3. Was the filter carefully handled to prevent loss of
particulate?
4. Was there any evidence that partial late may have been
bypassing the filter?
5. Could the filter be recovered without tearing and
without pieces adhering to surfaces that might affect
filter weight?
6. Were all components properly rinsed and samples
recovered 1n accordance with specified method?
7. Were implnger volumes measured or weighed properly
(±1 g or mL) and recorded on data sheets?
8. Were joints sealed or covered with foil after sample .
recovery in preparation for next test?
9. Was a blank train assembled and recovered using the
same procedures as actual sample trains? (A blank
MM5 train need only be done once for each set of
3 runs)?
10. Were other blank samples taken (e.g., reagents, filter,
XAD)?
11. Were all MM5 samples properly labeled and stored on 1ce
after recovery? ^
Comments:
55
-------
The period after completion of actual sampling is one of the busiest
times of trial burn activity. Besides MM5 train recovery, all other samples
must be secured and stored properly, and all associated data sheets must be
completed along with all sample traceability records. Orsat bags must be
analyzed. Also, stack sampling data sheets must be compiled and calculations
made to check the 1sok1netic sampling rate sometime before the start of the
next run. This work is crucial to the validity of results; therefore, it is
important that the observer remain on site after the sampling period.
The end of the sampling period for each run is also a time when everyone
is usually tired and mistakes are most likely to occur. This is an excellent
time to observe the attitudes and competency of those involved.
There is one final point to be made, for the permit writer/observer. Try
to avoid touching any of the instruments and under no circumstances make any
adjustments in the instruments or valves, etc. Further, do not assist in any
of the sampling or handling of any sampling equipment. There is sometimes the
inclination to help someone who is struggling with lifting or moving a piece
of sampling equipment (e.g., an MM5 sample box). That temptation must be
resisted, in order to avoid contributing in some way to an operating problem
or damage to sampling equipment.
4.4 PERSONAL SAFETY OF THE OBSERVER
The permit writer/observer should bring proper work clothing including
coats, gloves, and rain gear and have certain safety equipment on site (hard
hat, safety glasses, safety shoes). He should also determine beforehand if
the plant has any other safety requirements (e.g., no beard, long sleeve
shirts). The observer should also allow time for the plant's safety orien-
tation, which is commonly required. However, do not sign any injury waiver
forms or secrecy agreements, as required at most plants. Some of these waiv-
ers might bind the observer not to reveal anything seen or learned on site.
Your refusal to sign must be worked out before arrival, as noted earlier in f
Table 6, and may still involve some difficulty or delay in gaining access when £
you arrive. Therefore, you should call in advance to be sure about any "spe-
cial" arrangements for your entry. Assistance can be obtained from your _
Office of Regional Counsel or corresponding State office. I
All the safety equipment is useful and should be worn at all times, but
reasonable caution and good sense are most important. Stack sampling observa- I
tion is not an inherently safe activity, and no action should be taken if the I
observer believes it to be unsafe, even if others are doing it. If an unsafe
situation arises and it prevents the observer from performing his duties, r
he/she should inform someone that the results will not be accepted as valid I
unless the observations can be performed in a safe manner. An individual
should not attempt to be an observer if he/she has a fear of heights or is not
in reasonably good physical condition. I
EPA personnel who are observers must be aware of the fact that it is EPA
policy to comply with OSHA regulations. These regulations prescribe 40 hours I
of health and safety training plus a baseline physical examination and annual [
examinations thereafter. Observer's training and the baseline physical
examination must be completed before field activities are allowed. .
56 I
-------
4.5 QA/QC AUDITS
One of the first sections of this document that pertained to preparation
for observing a trial burn briefly described arranging for field audits and/or
check samples. Arranging for this auditing activity might be done by the
permit writer/observer or by other regulatory agencies or even by the appli-
cant himself. In any case, the observer should be aware that this auditing
activity may be taking place during the trial burn. If the observer decides
such auditing is necessary, it is important that this be arranged well in
advance with a clear plan defining the auditing and the acceptable results
criteria, and what course of action will be taken if results are not within
acceptable range.
There are two types of audits—systems audits and performance audits.
The observer, himself, is actually performing a systems audit, in that he
tries to ascertain that the work is being performed in accordance with methods
specified in the TBP and QAPP. A performance audit, on the other hand,
actually checks the performance or accuracy of measurements being made. In
most trial burn audits, this Involves the following:
Gas cylinders containing known concentrations that check results of
CEM Instruments and Orsat analysis
NBS traceable thermometers and barometers to check temperatures and
barometric pressure readings, etc.
Calibrated orifices to check accuracy of dry gas meters used in MM5
apparatus
The performance audits may also include the preparation of known solutions of
analytes to be submitted for analysis along with actual samples to check the
accuracy/precision of analytical results.
4.6 MAJOR DEVIATIONS OR PROBLEMS
Several of the deviations or problems that may occur while observing
trial burns have been mentioned in other sections of this guidance document.
As noted, they can usually be worked out satisfactorily so as not to jeopar-
dize the results or seriously impair the permit writer's requirements for
setting permit operating conditions. Even so, there are some deviations or
problems that could do so.
It could be considered a "problem" 1f something occurred that required a
run to be repeated. However, the plant personnel and the samplers are usually
aware of this possibility and can take 1t 1n stride. In fact, the samplers
are usually the first to identify a problem and will notify everyone that a
repeat is needed. However, much more serious problems can occur, some of
which are listed below:
• The plant cannot complete the required number of runs due to insuf-
ficient waste or equipment failure.
57
-------
• Fugitive emissions are excessive or operating practices unaccept-
able.
On-site observations reveal that trial burn operating conditions and
subsequent operating limits cannot be maintained during certain
phases of plant operation (e.g., auxiliary fuel system used for
heat-up cannot attain minimum operating temperature required before
introduction of waste).
An instrument that must be operable during the trial burn in order
to set operating limits malfunctions before or during test and
cannot be repaired for some time (e.g., CO monitor, waste feed
flowmeter).
• CO levels exceed applicable limits.
Other situations may occur as part of a trial burn that would seriously
jeopardize the results or data needs, and some will likely not be a clear-cut
situation. Thus the guidance offered earlier still applies. That is, prepare
for this possibility by having names and phone numbers (office and home)
readily available for those you might need to consult with and those who have
supervisory responsibility and should be involved in the decision-making. As
noted earlier, some of the problems that have occurred during trial burns and
their resolution are listed in the Appendix. However, they are site-specific
and are presented here only as examples.
4.7 DOCUMENTATION OF ACTIVITIES
Throughout all of the activities, the permit writer will need to keep
comprehensive notes in a bound notebook. The observer should record the date
on each page and time of each entry, but also the run number, using the same
numbering system as the sampling crew. Note all those items observed that are
satisfactory, as well as those that are not. One of the important areas to
keep notes on is the general operating parameters of each test and the time
and cause of any upset condition.
It is expected that the permit writer/observer will immediately inform
those in charge if he/she feels some procedural problem or change compromises
the results or seriously impairs his/her ability to prepare the operating
permit. The observer has a responsibility to let the permit applicant know of
any serious problems at the time they are observed; not later. Presumab1y,
any such problem will be corrected or worked out on site. However, if the
permit writer/observer noted something and did not consider it a problem at
the time, but later-discovered that it was a major problem, it would not be
necessary to accept the test as valid. That is, the permit writer should not
accept bad data.under any circumstances.
58
-------
Quite often there are a number of minor problems or procedures that the
observer believes are not serious problems but could have been improved
upon. Therefore, after the trial burn is completed, the observer may review
his logbook and let those involved know what he felt might have been done
. better and, of course, to let them know what was done well. This is generally
a courtesy, but not a requirement.
For further documentation of activities, the observer may also want to
I obtain copies of critical sampling data sheets or process data sheets and
I strip charts before leaving the site. If it is feasible (i.e., copy machine
available), the observer may request copies be made of MM5 data sheets, VOST
I data sheets, process operating log sheets, and selected strip charts (tempera-
I ture, feed rate, CO monitor). Be sure these are clearly identifiable and
readable, especially copies of strip charts.
I The observer may also want to date and sign some of the continuously
recorded data (strip charts) and data sheets. If this is done, then some
selectivity should be established because it may be difficult to sign all the
I data sheets generated during a trial burn. It is best to make a checklist of
all the data sheets and strip charts selected to be signed. This checklist
can be given to the appropriate test site personnel so that they are aware of
• what is to be signed. The permit writer/observer should also use the check-
I list to ensure that he/she has signed all the selected data sheets after each
run and checked that they were properly labeled and completed.
59
-------
SECTION 5.0
OBSERVATION REPORTS
After the trial burn is completed and the observer has returned to
his/her office, it is advisable to prepare an Observation Report that
summarizes the events and observations. Such reports may contain the
following sections:
Demonstration test schedule and activities (i.e., log of events)
Process operation observations
Sampling activity observations
Process samples
Stack samples
The first section, dealing with the demonstration test, needs to contain
a table with a brief statement of activities, by date, for each day on site.
It should also indicate the run (run number) conducted on specific days and
the types of waste feed and test conditions for each run. It should also note
any important problems or aborted runs. The section itself should summarize
on-site activities, noting certain problems and solutions and a general over-
view of the conduct of the trial burn.
The second section deals with process operation observations and should
summarize operations, problems, or fugitive emissions. General conditions of
the plant and procedures should be noted.
The third section of the report summarizes observations regarding the
process and stack sampling. Deviations from procedures stipulated in the TBP
and sampling methods referenced therein should be listed, along with general
comments about the sampling and the personnel performing those activities.
Other sections of the report may address safety considerations and any
performance audit results or check samples taken, if applicable. Appendices
may be attached to the report comprised of the completed tables and checklists
presented earlier in this guide.
61
-------
F
u
APPENDIX
ACTUAL PROBLEMS THAT HAVE BEEN ENCOUNTERED DURING
TRIAL BURNS AND THEIR RESOLUTIONS
63
-------
Some problems in the hazardous waste incinerator process operation and
monitoring/sampling activities commonly occur during trial burns. This appen-
dix presents summaries of actual problems encountered during trial burns and
their resolution. In each case, the trial burn observer had to make a deci-
sion on-site to resolve the problem.
This Appendix is presented to give examples of some problems which may be
encountered in trial burns, but is not intended as a guide to resolving
them. Each problem must be resolved on a case-by-case basis by the trial burn
observer as the need arises and circumstances dictate.
A. Deviations from the Trial Burn Plan
It is common for the applicant to request one or more deviations from the
approved trial burn plan at the on-site pretest meeting. Deviations often re-
quested include changes in the incinerator operating conditions, such as tem-
perature, waste feed rate, concentration of the hazardous substance(s) in the
feed, and combustion gas velocity; changes in locations or methods of taking
process or stack gas samples; schedule changes; and changes in the automatic
waste feed cutoff system operating parameters. Important aspects to be con-
sidered before approving any requested changes, especially major changes for a
new facility, were discussed in Section 3.1. Changes or deviations often
require consultation with others, especially for a new facility or when the
change should have been resolved during development of the trial burn plan.
Field circumstances usually cause most of the changes or deviations
to be requested. Some actual deviations to trial burn plans requested by
applicants and the response of the trial burn observer are presented below:
Al. Deviation requested: Raise the secondary combustor temperature from
1600° to 1750°F.
Response; The higher operating temperature was allowed with the specifi-
cation that the permitted operating temperature would be based on 1750°F, not
1600°F.
A2. Deviation requested; Lower the waste feed rate from 4.3 tons/hr to
3.5 tons/hr.
Response: The lower waste feed rate was allowed with the specification
that the permitted waste feed rate would be based on a maximum of 3.5 tons/hr,
not 4.3 tons/hr.
A3. Deviation requested: Change location of liquid waste feed sampling
from tap in the feed line to a tap in the agitated storage tank.
Response; The request was denied since it was uncertain that the waste
feed mixture was homogeneous and only part of the contents of the tank would
be used in a single test run. Sampling the tank may not have been representa-
tive of the actual waste fed during each individual test run. Therefore, the
applicant was required to sample from the feed line.
64
-------
I
I
I
I
I
v
i
A4. Deviation requested: Change waste feed sampling interval from 15 to
30 minutes for the composite waste feed sample.
I
I
Response: The request was denied and waste feed sampling was conducted
at 15-minute intervals. The trial burn observer deemed that a less represen-
• " tative sample would result if the longer sampling interval were allowed.
A5i Deviation requested: Change impinger solution in MM5 train for sam-
Ipling metal emissions (lead and cadmium) in stack gas from silver-catalyzed
ammonium persulfate solution to nitric acid solution. Reason given was the
persulfate solution is difficult to analyze for metals.
Response: The trial burn observer was uncertain as to the effect of this
change on the sampling method, and consulted staff in his home office, a con-
tractor laboratory, and staff in Research Triangle Park, North Carolina.
After considering various opinions, the observer decided that both solutions
should be put in separate impingers in the MM5 train, and both solutions
should be analyzed for metals after sampling. The applicant complied with
that decision. (It must be noted that the need for this type of change should
have been recognized and resolved during development of the TBP.)
A6. Deviatiori j-equested: Change simultaneous sampling of stack gas with
MS and MM5 trains to sequential sampling (M5 followed by MM5) due to accessi-
bility problems on the stack.
Response: The change was denied since the accessibility problem could be
readily remedied, and simultaneous sampling is preferable over sequential sam-
pling.
A7. Deviation requested: Change schedule from one trial burn run per
day to two runs per day.
Response: The schedule change was allowed with the provision that the
second test would not extend the day's work schedule beyond 12 hr/day. This
limit was set to prevent excessive fatigue of the sampling team, which could
cause accidents or possibly affect performance of the sampling team during the
second run.
A8. Deviation requested: During assembly of the MM5 train, prior to the
first run, the observer noted that the glass tubing to the condenser/XAD was
connected with a short piece of rubber tubing. The observer informed the test
crew that this was a deviation from the method, which specified only glass or
Teflon, Since only a very short piece of tubing was involved, the test crew
requested that this deviation be allowed.
Response; After conferring with EPA experts by phone, the deviation was
not allowed. The test crew was able to replace the rubber tubing using Teflon
tubing and Teflon tape.
65
-------
B. Process Problems
Problems can occur in the incinerator operation during the trial burn
which must be resolved by the applicant and trial burn observer. Some actual
problems which have occurred in the past and their resolution are presented
below.
Bl. Problem: The solid waste feed system clogged during the middle of a
test run and required 1 hr to correct, while the incinerator continued opera-
tion.
Resolution; The trial burn observer stopped the stack gas sampling
shortly after the clog occurred. Sampling was resumed using the same sampling
trains 30 minutes after solid waste feed had been resumed. The samples col-
lected in the trains were deemed to be valid.
82. Problem: Water flow to the venturi scrubber decreased during a test
run and required about 1.5 hr to correct, while the incinerator continued to
operate without waste feed (the automatic waste feed system cutoff was
activated).
Resolution: The trial burn observer stopped stack gas sampling when the
waste feed was cut off, and sampling was resumed with the same sampling trains
about 20 minutes after the waste feed system had been reactivated. The sam-
ples collected in the trains were deemed valid.
B3. Problem: Temperature in the secondary combustor rose too high twice
during a 1-hr period of a trial burn run, and activated the automatic waste
feed cutoff system (high temperature cutoff). The temperature was lowered
both times and the run continued.
Resolution: Stack gas sampling was stopped after the first incident, re-
started, and stopped again after the second incident. Once the temperature
stabilized the second time, sampling was again resumed. The test run and sam-
ples were deemed valid by the trial burn observer since sampling was inter-
rupted for only about 1 hr.
84. Problem; The activator controlling the damper in the ID fan system
failed due to excessive vibration.
Resolution: The system was shut down until a new activator was in-
stalled. The test was cancelled and was run the next day.
•i
85, Problem: During a test run, the venturi damper broke and the system
was shut down.
Resolution: Stack sampling time had been under way for only 45 minutes
when the damper broke, and the trial burn observer aborted the test. The
damper was replaced overnight, and the test run was conducted the next day.
66
-------
86. Problem: The scrubber system ran out of caustic solution near the
end of a test run and the system was shut down. The problem developed because
6 hr of incinerator operation was required to maintain steady state before the
run began, and the caustic supply was to be replenished that night.
Resolution: Stack sampling had been conducted for 3 of the 4 hr sched-
uled. The trial burn observer determined that sufficient sampling time for
all samples had been completed, and the test was considered to be valid.
87.* Problem; Fugitive emissions of bottom ash, conveyed from the
incinerator and dumped into trucks, were excessive at the outset of the first
two test runs. The process used water to wet the ash for control of emis-
sions. However, the water was turned on manually, and the operator failed to
start water flow until several minutes after ash dumping had begun.
Resolution; The trial burn observer specified that the procedures be
changed to ensure that water flow begin before the ash conveyor started to
prevent further fugitive emissions of ash. The applicant agreed and also
indicated that the system would be automated following the trial burn.
88.* Problem; Bottom ash fines were dumped into a covered 55-gal drum
through a rigid pipe. Air emissions of the fines were observed during the
trial burn.
Resolution; The trial burn observer requested that the condition be cor-
rected. The applicant replaced the rigid pipe with a flexible boot which
solved the problem.
89.* Problem; Trucks receiving bottom ash were parked on undiked con-
crete pads. Ash and spray water spilled onto the.concrete on several occa-
sions and were not initially cleaned up.
Resolution; The trial burn observer noted the spills and requested that
they be cleaned up in accordance with the spill prevention and control plan
since the concentrations of hazardous constituents in the ash were unknown.
The applicant complied. In addition, plastic sheets were placed on the con-
crete and the area around the trucks was diked with sorbent material to con-
trol future spills, at the request of the observer. The applicant agreed to
follow this procedure in the future and make it a part of the process opera-
tion plan.
Note regarding examples 87, 88, and 89: If the observer is not the per-
mit writer, he or she may not have direct responsibility for resolving
these types of problems, which are not related to validity of test
results. In such cases, or where longer term corrective measures are
required, the issue should be noted and reported to both the applicant
and the appropriate regulatory office.
67
-------
BIO. Problem: The incineration system experienced numerous operational
difficulties on the first test day, and no tests were run. The second day,
operational difficulties caused the test to be aborted during the run.
Resolution: The trial burn observer met with the applicant at the end of
the second day to assess the feasibility of continuing the trial burn. By
mutual agreement, the trial burn was postponed until a later date.
Bll. Problem: Plant operating problems continued to delay the start of
the run.
Resolution; The observers conferred and determined a time when the test
would have to start in order to finish by 10:00 p.m. Plant personnel were
informed of this necessary start time. The run did not start by that time and
was postponed to the next day.
B12. Problem: Liquid waste feed flowmeter was malfunctioning, producing
erratic readings, prior to the start of the run.
Resolution; The run was carried out, using the tank level change to de-
termine waste feed rate during the run. However, it was stipulated that the
flowmeter needed to be repaired and operational for subsequent runs.
68
-------
C. Monitoring/Sampling Problems
Problems can occur with monitoring and sampling activities which must be
resolved by the applicant and trial burn observer. Some actual problems which
have occurred in the past and their resolution are presented below.
Cl. Problem; An audit of the automatic waste feed cutoff system
revealed that the system did not function properly for low temperature in the
secondary combustion chamber.
Resolution; The condition was corrected and passed a second audit prior
to the trial burn.
C2. Problem; The continuous emission monitors for CO and 02 were
audited using an audit gas cylinder. The CO monitor failed the audit twice.
Resolution; The trial burn was cancelled after it was- determined that
replacement of the monitor would take several days. A new date was set for
the trial burn, with the stipulation that the CO monitor pass an audit prior
to that date.
C3. Problem; During the middle of a test run, erratic temperature read-
ings for the rotary kiln were experienced.
Resolution; The trial burn observer requested that the waste feed and
all sampling activities be stopped until the condition was corrected. The
thermocouple was replaced and normal readings were obtained. The procedure
took about 2 hr, and the test run was allowed to continue to completion.
C4. Problem; All three sampling consoles were audited with a critical
orifice to check the dry test meter accuracy. One of the consoles failed the
audit.
Resolution; Another console was audited and passed, and was used to
replace the failed console prior to the trial burn.
C5. Problem; The MS and MM5 sampling trains for the stack gases were
set up by placing the probes in the stack and the impinger boxes on the
ground, and connecting the probes and boxes with unheated 30-ft Teflon
tubes. This procedure had not been specified in the trial burn plan.
Resolution: The trial burn observer indicated that this setup was a
major deviation from the -reference methods, and was not acceptable. The sam-
pling team placed the impinger boxes on the stack platform, but still used
unheated Teflon tubes (about 10 ft each) due to limited access to the stack
ports (the reason for the tubes in the first place). While this change
improved the situation, the observer warned the sampling team that this pro-
cedure might still be unacceptable, and he needed to consult other staff on
the matter. The first test was run that day anyway.
69
-------
After consultation the next day with the home office, other EPA experts,
and a contractor, the observer concluded that the tubes were acceptable only
if they were heated. The first run, therefore, was deemed invalid.
C6. Problem: M5 and MM5 trains were used in the trial burn runs to
simultaneously sample the stack gases. During one run, the M5 train failed
the leak check after the first traverse.
Resolution: The M5 train sample was invalidated because the leak check
failed. The observer allowed the MM5 train sampling to continue while a sec-
ond MS train was prepared for sampling. The second M5 train began sampling
after about two-thirds of the MM5 train sampling had been completed. The M5
train sampling was completed about 1 hr after completion of MM5 sampling.
Both samples were considered to be valid, even though they were not taken
simultaneously as specified in the trial burn plan, because the incinerator
was run at the same operating conditions throughout the entire sampling
period.
C7. Problem: The M5 train passed the leak check after the first tra-
verse, but failed the leak check before the second traverse of a trial burn
run. It was determined that the glass probe liner had been broken while
changing ports.
Resolution: The broken probe liner was replaced and the test contin-
ued. The samples in both probes were recovered and added together for
analysis.
C8. Problem; During recovery of the MS and MM5 trains at the end of the
first test run, the sample from the MS particulate train was lost when a flask
broke. The MM5 train was recovered successfully.
Resolution; The trial burn observer had to decide whether to (a) inval-
idate the first run or (b) allow sampling with an MS train prior to the begin-
ning of the second run and use that data for the first test run. The latter
choice was made since the incinerator was maintained at steady state and the
same operating conditions were maintained in both the first and second runs.
C9. Problem: Heavy rains began about midway through a test run and
stack sampling had to cease for about 4 hr. Sampling could not be resumed
until about 11:00 p.m.
Resolution: The test was aborted due to the long delay, remaining sam-
pling activities, and •; the late hour. Next day, the trial burn observer
requested that a cover be put over the stack platform to help prevent reoccur-
rence of the problem. The applicant complied with the request with a minimum
effort in about 2 hr.
70
-------
CIO. Problem: One of the two MM5 trains being used failed the leak
check at port change.
Resolution; Another MM5 train was brought up to the stack and started
anew. This meant that the other MM5 train and VOST finished sampling about
1 1/2 hr before the other MM5 train. However, all waste feed and scrubber
water sampling was extended to cover the entire sampling period.
Cll. Problem: One of the VOST traps was broken when it was removed from
the sampling apparatus. Another pair of traps was run, but necessarily
extended the VOST sampling time beyond completion of M5 and MM5 sampling.
Resolution; The observer required that pitot readings be taken during
the last pair of VOSJ trap sampling to provide comparative data on stack gas
velocity (i.e., confirm that there was no significant difference in stack flow
rate). <
C12. Problem; The VOST apparatus would not pass the leak check after
inserting the second pair of traps. A several hour delay occurred in trying
to correct the problem.
Resolution: The run was aborted and redone the next day.
C13. Problem: During recovery of the MM5 and M5 train, the pattern of
the particulate on the filter extended clear to the edge of the filter, indi-
cating that some particulate might have bypassed the filter.
Resolution: Based on the evidence, the run was considered invalid and
had to be repeated. The sampling team made a minor change in the sampling
train to help ensure that the problem was corrected for future runs.
71
------- |