Jnited States
Environmental Protection
Agency
fice ot Solid Waste
and Emergency Response
Washington, DC 20460
November 1986
SW-846
Third Edition
Solid Waste
Test Methods
for Evaluating Solid Waste
Volume II: Field Manual
Physical/Chemical Methods
>^^X&;^^:-^?ivl .;?;:• i _
&^::'--\-;-.;Si&™li-:^3^--~^~~sii£*
' — — --- :, . . — — .-•.-• ~- . - — - -
* •f-z^---f^ ;-=K_ , •.-".".-•-"-• •••'/• .--
i^^S&i^'-^-^'*.--^'^--.-'--- -r-j'-U --<•£: "^n
— ~--' -
-------
.Test Methods for. Evaluating Solid Waste Physical/Chemical Methods
Third Edition
Proposed Update Package
Instructions
Enclosedjfis-tthe proposed Update ' I package for ""Test Methods .
-------
METHODS INCLUDED IN PROPOSED UPDATE PACKAGE I
.* Chapter 1
* Chapter 2
*. Chapter' 4
* Chapter'7
'* Method 1310 -
* Method 1330 -
* Methbd.;3005 k
* Method 3010 -
* Method 3020 -
*
*
*-Method
* Method
* Method
* Method
* Method
T Method
* Method
T Method
Method
Method
N Method
* Method
N Method
N Method
N Method
N Method
T Method
N Method
N Method
N Method
* Method
* Method
N Method
3050
3510
3520
3540
3600
3650
5030
6010
7000
7061
7081
7196
7211
7381
7430
7461
7760
7761
7780
7951
8000
8010
8011
* Method 8015
N Method 8021
* Method
* Method
N Method
N Method
* Method
N Method
* Method
* Method
* Method
N Method
8030
8040
8070
8110
8120
8141
8150
8240
8250
8260
Definitions
Tables .
Table 4-1. .
Reactive Cyanide and Sulfide
Extraction Procedure (EP) Toxicity Test Method and Structural
Integrity Test
Extraction Procedure -for Oily Wastes
Acid Digestion of Waters for Total Recoverable or Dissolved-
Metals for Analysis by Flame Atomic Absorption Spectroscopy
or Inductively Coupled Plasma Spectroscopy
Acid Digestion of Aqueous Samples of Extracts for -TotaT Metals
for Analysis by Flame Atonic Absorption Spectroscopy or «
Inductively Coupled Plasma Spectroscopy
Acid Digestion of Aqueous Samples and Extracts for Total Metals
- for Analysis by Furnace Atomic Absorption Spectroscopy
Acid Digestion of Sediments, Sludges, and Soils
Separatory Funnel Liquid - Liquid Extraction
Continuous Liquid-Liquid Extraction
Soxhlet Extraction
Cleanup
Acid-Base Partition Cleanup
Purge-and-Trap
Inductively Coupled Plasma Atomic Emission Spectroscopy
Atomic Absorption Methods •- •
Arsenic (AA, Gaseous Hydride)
Barium (AA, Furnace Technique) "'
Chromium, Hexavalent (Colorimetric)
Copper (AA, Furnace Technique)
Iron (AA, Furnace Technique)
Lithium (AA, Direct Aspiration)
Manganese (AA, Furnace Technique)
Silver (AA, Direct Aspiration)
Silver (AA, Furnace Technique)
Strontium (AA, Direct Aspiration)
Zinc (AA, Furnace Technique)
Gas Chromatography
Halogenated Volatile Organics
1,2-Dibromoethane and 1,2-Dibromo-3-chloropropane in Water by
Microextraction and Gas Chromatography
Nonhalogenated Volatile Organics
Volatile Organic Compounds in Water by Purge-and-Trap Capillary
Column Gas Chromatography with PID and Electroconductivity
Detector in Series
Acrolein, Acrylonitrile, Acetonitrile
Phenols
Nitrosamines
Haloetherjs
Chlorinated Hydrocarbons
Organophosphorus Pesticides Capillary Column
Chlorinated Herbicides
GC/MS for Volatile Organics
GC/MS for Semi volatile Organics: Packed Column Technique
GC/MS for Volatile Organics: Capillary Column
-------
METHODS INCLUDED IN PROPOSED UPDATE PACKAGE I (cont'd)
* Method 8270 - QC/MS for Semivolatile Organics: Capillary Column-Technique
T Method 9010 and 9010A - Total and Amenable Cyanides
N Method-9021/- ;Purgeable Organic Hal ides (POX)
T Method 9030 -' Acid-Soluble and Acid-Insoluble Sulfides
N Method 9031 - Extractable Sulfides
* Method 9090 .;- Compatibility -Test for Wastes .and Membrane" Liners.
* Indicates partial revision
N Indicates a new method
T Indicates a total revision
-------
VOLUME TWO
Revision 0
Date September 1986
-------
For sale by the Superintendent of Documents, U.S. Government Printing Office. Washinifton, D.C. 20402
-------
ABSTRACT
This manual provides test procedures which may be used to evaluate those
properties of a solid waste which determine whether the waste 1s a hazardous
waste within the definition of Section 3001 of the Resource Conservation and
Recovery Act (PL 94-580). These methods are approved for obtaining data to
satisfy the requirement of 40 CFR Part 261, Identification and Listing of
Hazardous Waste. This manual encompasses methods for collecting
representative samples of solid wastes, and for determining the reactivity,
corroslvlty, 1gn1tab1l1ty, and composition of the waste and the mobility of
toxic species present 1n the waste.
ABSTRACT - 1
Revision
Date September 1986
-------
TABLE OF CONTENTS
VOLUME ONE,
SECTION A
ABSTRACT
TABLE OF CONTENTS
METHOD INDEX AND CONVERSION TABLE
PREFACE
ACKNOWLEDGEMENTS
PART I METHODS FOR ANALYTES AND PROPERTIES
CHAPTER ONE — QUALITY CONTROL
1.1 Introduction
1.2 Quality Control
1.3 Detection Limit and Quantification Limit
1.4 Data Reporting
1.5 Quality Control Documentation
1.6 References
CHAPTER TV/0 — CHOOSING THE CORRECT PROCEDURE
2.1 Purpose
2.2 Required Information
2.3 Implementing the Guidance
2.4 Characteristics
2.5 Ground Water
2.6 References
CONTENTS - 1
Revision
Date September 1986
-------
CHAPTER THREE — METALLIC ANALYTES
3.1 Sampling Considerations
3.2 Sample Preparation Methods
Method 3005:
Method 3010:
Method 3020:
Method 3040:
Method 3050:
Acid Digestion of Waters for Total Recoverable
or Dissolved Metals for Analysis by Flame
Atomic Absorption Spectroscopy or
Inductively Coupled Plasma Spectroscopy
Acid Digestion of Aqueous Samples and Extracts
for Total Metals for Analysis by Flame
Atomic Absorption Spectroscopy or
Inductively Coupled Plasma Spectroscopy
Acid Digestion of Aqueous Samples and Extracts
for Total Metals for Analysis by Furnace
Atomic Absorption Spectroscopy
Dissolution Procedure for Oils, Greases, or
Waxes
Acid Digestion of Sediments, Sludges, and Soils
3.3 Methods for Determination of Metals
Method 6010:
Method 7000:
Method 7020:
Method 7040:
Method 7041:
Method 7060:
Method 7061:
Method 7080:
Method 7090:
Method 7091:
Method 7130:
Method 7131:
Method 7140:
Method 7190:
Method 7191:
Method 7195:
Method 7196:
Method 7197:
Method 7198:
Method 7200:
Method 7201:
Method 7210:
Method 7380:
Method 7420:
Method 7421:
Method 7450:
Method 7460:
Inductively Coupled Plasma Atomic Emission
Spectroscopy
Atomic Absorption Methods
Aluminum (AA, Direct Aspiration)
Antimony (AA, Direct Aspiration)
Antimony (AA, Furnace Technique)
Arsenic (AA, Furnace Technique)
Arsenic (AA, Gaseous Hydride)
Barium (AA, Direct Aspiration)
Beryllium (AA, Direct Aspiration)
Beryllium (AA, Furnace Technique)
Cadmium (AA, Direct Aspiration)
Cadmium (AA, Furnace Technique)
Calcium (AA, Direct Aspiration)
Chromium (AA, Direct Aspiration)
Chromium (AA, Furnace Technique)
Chromium, Hexavalent (Coprecipitation)
Chromium, Hexavalent (Colorimetric)
Chromium, Hexavalent (Chelation/Extraction)
Chromium, Hexavalent (Differential Pulse
Polarography)
Cobalt (AA, Direct Aspiration)
Cobalt (AA, Furnace Technique)
Copper (AA, Direct Aspiration)
Iron (AA, Direct Aspiration
Lead (AA, Direct Aspiration
Lead (AA, Furnace Technique
Magnesium (AA, Direct Aspiration)
Manganese (AA, Direct Aspiration)
CONTENTS - 2
Revision 0
Date September 1986
-------
Method 7470:
Method 7471:
Method
Method
Method
Method
Method
Method
Method
Method
Method
Method
Method
Method
Method
Method
Method
7480:
7481:
7520:
7550:
7610:
7740:
7741:
7760:
7770:
7840:
7841:
7870:
7910:
7911:
7950:
Mercury 1n Liquid Waste (Manual Cold-Vapor
Technique)
Mercury 1n Solid or Semi solid Waste (Manual
Cold-Vapor Technique)
Molybdenum (AA, Direct Aspiration)
Molybdenum (AA, Furnace Technique)
Nickel (AA, Direct Aspiration)
Osmium (AA, Direct Aspiration)
Potassium (AA, Direct Aspiration)
Selenium (AA, Furnace Technique)
Selenium (AA, Gaseous Hydride)
Silver (AA, Direct Aspiration)
Sodium (AA, Direct Aspiration)
Thallium (AA, Direct Aspiration)
Thallium (AA, Furnace Technique)
Tin (AA, Direct Aspiration)
Vanadium (AA, Direct Aspiration)
Vanadium (AA, Furnace Technique)
Z1nc (AA, Direct Aspiration)
APPENDIX — COMPANY REFERENCES
CONTENTS - 3
Revision 0
Date September 1986
-------
VOLUME ONE.
SECTION B
ABSTRACT
TABLE OF CONTENTS
METHOD INDEX AND CONVERSION TABLE
PREFACE
CHAPTER ONE. REPRINTED — QUALITY CONTROL
1.1 Introduction
1.2 Quality Control
1.3 Detection Limit and Quantification Limit
1.4 Data Reporting
1.5 Quality Control Documentation
1.6 References
CHAPTER FOUR — ORGANIC ANALYTES
4.1 Sampling Considerations
4.2 Sample Preparation Methods
4.2.1 Extractions and Preparations
Method 3500:
Method 3510:
Method 3520:
Method 3540:
Method 3550:
Method 3580:
Method 5030:
Method 5040:
4.2.2 Cleanup
Method 3600:
Method 3610:
Method 3611:
Method 3620:
Method 3630:
Method 3640:
Method 3650:
Method 3660:
Organic Extraction And Sample Preparation
Separatory Funnel Liquid-Liquid Extraction
Continuous Liquid-Liquid Extraction
Soxhlet Extraction
Sonication Extraction
Waste Dilution
Purge-and-Trap
Protocol for Analysis of Sorbent Cartridges from
Volatile Organic Sampling Train
Cleanup
Alumina Column Cleanup
Alumina Column Cleanup And Separation of
Petroleum Wastes
Florisil Column Cleanup
Silica Gel Cleanup
Gel-Permeation Cleanup
Acid-Base Partition Cleanup
Sulfur Cleanup
CONTENTS - 4
Revision 0
Date September 1986
-------
4.3 Determination of Organic Analytes
4.3.1 Gas Chromatographic Methods
Method 8000: Gas Chromatography
Method 8010: Halogenated Volatile Organlcs
Method 8015: Nonhalogenated Volatile Organlcs
Method 8020: Aromatic Volatile Organlcs
Method 8030: Acroleln, Acrylonltrile, Acetom'trile
Method 8040: Phenols
Method 8060: Phthalate Esters
Method 8080: Organochlorine Pesticides and PCBs
Method 8090: NHroaromatlcs and Cyclic Ketones
Method 8100: Polynuclear Aromatic Hydrocarbons
Method 8120: Chlorinated Hydrocarbons
Method 8140: Organophosphorus Pesticides
Method 8150: Chlorinated Herbicides
4.3.2 Gas Chromatograph1c/Mass Spectroscoplc Methods
Method 8240: Gas Chromatography/Mass Spectrometry for
Volatile Organlcs
Method 8250: Gas Chromatography/Mass Spectrometry for
Semi volatile Organics: Packed Column
Technique
Method 8270: Gas Chromatography/Mass Spectrometry for
Semi volatile Organics: Capillary Column
Technique
Method 8280: The Analysis of Polychlorinated Dibenzo-P-
Dioxins and Polychlorinated Dibenzofurans
Appendix A: Signal-to-Noise Determination Methods
Appendix B: Recommended Safety and Handling Procedures
for PCDD's/PCDF's
4.3.3 High Performance Liquid Chromatographic Methods
Method 8310: Polynuclear Aromatic Hydrocarbons
4.4 Miscellaneous Screening Methods
Method 3810: Headspace
Method 3820: Hexadecane Extraction and Screening of Purgeable
Organlcs
APPENDIX — COMPANY REFERENCES
CONTENTS - 5
Revision
Date September 1986
-------
VOLUME ONE,
SECTION C
ABSTRACT
TABLE OF CONTENTS
METHOD INDEX AND CONVERSION TABLE
PREFACE
CHAPTER ONE. REPRINTED — QUALITY CONTROL
1.1 Introduction
1.2 Quality Control
1.3 Detection Limit and Quantification Limit
1.4 Data Reporting
1.5 Quality Control Documentation
1.6 References
CHAPTER FIVE — MISCELLANEOUS TEST METHODS
Method 9010:
Method 9012:
Method 9020:
Method 9022:
Method 9030:
Method 9035:
Method 9036:
Method 9038:
Method 9060:
Method 9065:
Method 9066:
Method 9067:
Method 9070:
Method 9071:
Total and Amenable Cyanide (Color1metric,
Manual)
Total and Amenable Cyanide (Colorlmetrlc,
Automated UV)
Total Organic Halldes (TOX)
Total Organic Halldes (TOX) by Neutron
Activation Analysis
Sulfldes
Sulfate (Colorlmetrlc, Automated, Chloranilate)
Sulfate (Colorlmetrlc, Automated, Methyl thymol
Blue, AA II)
Sulfate (Turbldimetric)
Total Organic Carbon
Phenol1cs (Spectrophotometrlc, Manual 4-AAP with
Distillation)
Phenol1cs (Colorlmetric, Automated 4-AAP with
Distillation)
Phenol1cs (Spectrophotometrlc, MBTH with
Distillation)
Total Recoverable 011 & Grease (Gravimetric,
Separatory Funnel Extraction)
Oil & Grease Extraction Method for Sludge
Samples
CONTENTS - 6
Revision 0
Date September 1986
-------
Method 9131:
Method 9132:
Method 9200:
Method 9250:
Method 9251:
Method 9252:
Method 9320:
Total Collform: Multiple Tube Fermentation
Technique
Total Coliform: Membrane Filter Technique
Nitrate
Chloride (Colorimetric, Automated Ferricyanlde
AAI)
Chloride (Colorimetric, Automated Ferricyanide
AAI I)
Chloride (Titrimetric, Mercuric Nitrate)
Radium-228
CHAPTER SIX — PROPERTIES
Method
Method
Method
Method
Method
Method
Method
1320:
1330:
9040:
9041:
9045:
9050:
9080:
Method 9081:
Method 9090:
Method 9095:
Method 9100:
Method 9310:
Method 9315:
Multiple Extraction Procedure
Extraction Procedure for Oily Wastes
pH Electrometric Measurement
pH Paper Method
Soil pH
Specific Conductance
Cation-Exchange Capacity of Soils (Ammonium
Acetate)
Cation-Exchange Capacity of Soils (Sodium
Acetate)
Compatibility Test for Wastes and Membrane
Liners
Paint Filter Liquids Test
Saturated Hydraulic Conductivity, Saturated
Leachate Conductivity, and Intrinsic
Permeability
Gross Alpha & Gross Beta
Alpha-Emitting Radium Isotopes
PART II CHARACTERISTICS
CHAPTER SEVEN — INTRODUCTION AND REGULATORY DEFINITIONS.
7.1 Ignitabillty
7.2 Corrosltivity
7.3 Reactivity
Section 7.3.3.2: Test Method to Determine Hydrogen Cyanide
Released from Wastes
Section 7.3.4.1: Test Method to Determine Hydrogen Sulfide
Released from Wastes
7.4 Extraction Procedure Toxicity
CONTENTS - 7
Revision 0
Date September 1986
-------
CHAPTER EIGHT — METHODS FOR DETERMINING CHARACTERISTICS
8.1 Ignltabinty
Method 1010: Pensky-Martens Closed-Cup Method for Determining
• Ign1tab1l1ty
Method 1020: Setaflash Closed-Cup Method for Determining
Ign1tab1l1ty
8.2 Corros1v1ty
Method 1110: Corroslvlty Toward Steel
8.3 Reactivity
8.4 Toxlclty
Method 1310: Extraction Procedure (EP) Toxldty Test Method
and Structural Integrity Test
APPENDIX — COMPANY REFERENCES
CONTENTS - 8
Revision
Date September 1986
-------
VOLUME TWO
ABSTRACT
TABLE OF CONTENTS
METHOD INDEX AND CONVERSION TABLE
PREFACE
CHAPTER ONE. REPRINTED— QUALITY CONTROL
1.1 Introduction
1.2 Quality Control
1.3 Detection Limit and Quantification Limit
1.4 Data Reporting
1.5 Quality Control Documentation
1.6 References
PART III SAMPLING
CHAPTER NINE — SAMPLING PLAN
9.1 Design and Development
9.2 Implementation
CHAPTER TEN — SAMPLING METHODS
Method 0010: Modified Method 5 Sampling Train
Appendix A: Preparation of XAD-2 Sorbent Resin
Appendix B: Total Chromatographable Organic Material Analysis
Method 0020: Source Assessment Sampling System (SASS)
Method 0030: Volatile Organic Sampling Train
CONTENTS - 9
Revision
Date September 1986
-------
PART IVMONITORING
CHAPTER ELEVEN — GROUND WATER MONITORING
11.1 Background and Objectives
11.2 Relationship to the Regulations and to Other Documents
11.3 Revisions and Additions
11.4 Acceptable Designs and Practices
11.5 Unacceptable Designs and Practices
CHAPTER TWELVE — LAND TREATMENT MONITORING
12.1 Background
12.2 Treatment Zone
12.3 Regulatory Definition
12.4 Monitoring and Sampling Strategy
12.5 Analysis
12.6 References and Bibliography
CHAPTER THIRTEEN — INCINERATION
13.1 Introduction
13.2 Regulatory Definition
13.3 Waste Characterization Strategy
13.4 Stack-Gas Effluent Characterization Strategy
13.5 Additional Effluent Characterization Strategy
13.6 Selection of Specific Sampling and Analysis Methods
13.7 References
APPENDIX — COMPANY REFERENCES
CONTENTS - 10
Revision
Date September 1986
-------
METHOD INDEX AND CONVERSION TABLE
Method Number.
Third Edition
0010
0020
0030
1010
1020
1110
1310
1320
1330
3005
3010
3020
3040
3050
3500
3510
3520
3540
3550
3580
3600
3610
3611
3620
3630
3640
3650
3660
3810
3820
5030
5040
6010
7000
7020
Chapter Number,
Third Edition
Method Number,
Current Revision
Ten
Ten
Ten
Eight
Eight
Eight
Eight
Six
Six
Three
(8.1)
(8.1)
(8.2)
(8.4)
Three
Three
Three
Three
Four (4.2.1)
Four (4.2.1)
Four (4.2.1
Four (4.2.1
Four (4.2.1
Four (4.2.1
Four
Four
Four
Four
Four
Four
Four
Four
Four
4.2.2
4.2.2]
4.2.2;
4.2.2)
4.2.2;
4.2.2)
4.2.2]
4.2.2]
4.4)
Four (4.4)
Four (4. 2.1]
Four (4.2.i;
Three
Three
Three
Second Edition
0010
0020
0030
1010
1020
1110
1310
1320
1330
3005
3010
3020
3040
3050
None (new method)
3510
3520
3540
3550
None (new method)
None (new method)
None (new method)
3570
None (hew method)
None (new method)
None (new method)
None (new method)
None (new method)
5020
None (new method)
5030
3720
6010
7000
7020
Number
0
0
0
0
0
0
0
0
0 .
0
0
0
0
0
0
' 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
METHOD INDEX - 1
Revision 0
Date September 1986
-------
METHOD INDEX AND CONVERSION TABLE
(Continued)
Method Number,
Third Edition
Chapter Number,
Third Edition
Method Number,
Second Edition
Current Revision
Number
7040
7041
7060
7061
7080
7090
7091
7130
7131
7140
7190
7191
7195
7196
7197
7198
7200
7201
7210
7380
7420
7421
7450
7460
7470
7471
7480
7481
7520
7550
7610
7740
7741
7760
7770
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
Three
7040
7041
7060
7061
7080
7090
7091
7130
7131
7140
7190
7191
7195
7196
7197
7198
7200
7201
7210
7380
7420
7421
7450
7460
7470
7471
7480
7481
7520
7550
7610
7740
7741
7760
7770
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
METHOD INDEX - 2
Revision 0
Date September 1986
-------
METHOD INDEX AND CONVERSION TABLE
(Continued)
Method Number.
Third Edition
7840
7841
7870
7910
7911
7950
8000
8010
8015
8020
8030
8040
8060
8080
8090
8100
8120
8140
8150
8240
8250
8270
8280
8310
9010
9020
9022
9030
9035
9036
9038
9040
9041
9045
9050
Chapter Number,
Third Edition
Three
Three
Three
Three
Three
Three
Four
Four
Four
Four
4.3.1
4.3.1
4.3.1
4.3.1
Four (4.3.1)
Four
Four
Four
Four
4.3.1
4.3.1
4.3.1
4.3.1
Four (4.3.1)
Four
Four
Four
[4.3.1
4.3.1
[4.3.1
Four (4.3.2
Four
Four
Four
Four
Five
Five
Five
Five
Five
Five
Five
Six
Six
Six
Six
4.3.2)
4.3.2]
4.3.2]
(4.3.3)
Method Number.
Second Edition
Current Revision
Number
7840 0
7841 0
7870 0
7910 0
7911 0
7950 0
None (new method) 0
8010 0
8015 0
8020 0
8030 0
8040 0
8060 0
8080 0
8090 0
8100 0
8120 0
8140 0
8150 0
8240 0
8250 0
8270 0
None (new method) 0
8310 0
9010 0
9020 0
9022 0
9030 0
9035 0
9036 0
9038 0
9040 0
9041 0
9045 0
9050 0
METHOD INDEX - 3
Revision 0
Date September 1986
-------
METHOD INDEX AND CONVERSION TABLE
(Continued)
Method Number,
Third Edition
Chapter Number,
Third Edition
Method Number,
Second Edition
Current Revision
Number
9060 Five
9065 Five
9066 Five
9067 Five
9070 Five
9071 Five
9080 Six
9081 Six
9090 Six
9095 Six
9100 Six
9131 Five
9132 Five
9200 Five
9250 Five
9251 Five
9252 Five
9310 Six
9315 Six
9320 Five
HCN Test Method Seven
H2S Test Method Seven
9060
9065
9066
9067
9070
9071
9080
9081
9090
9095
9100
9131
9132
9200
9250
9251
9252
9310
9315
9320
HCN Test Method
H2S Test Method
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
METHOD INDEX - 4
Revision 0
Date September 1986
-------
PREFACE AND OVERVIEW
PURPOSE OF THE MANUAL
Test Methods for Evaluating Solid Waste (SW-846) is Intended to provide a
unified, up-to-date source of Information on sampling and analysis related to
compliance with RCRA regulations. It brings together Into one reference all
sampling and testing methodology approved by the Office of Solid Waste for use
1n Implementing the RCRA regulatory program. The manual provides methodology
for collecting and testing representative samples of waste and other materials
to be monitored. Aspects of sampling and testing covered in SW-846 include
quality control, sampling plan development and implementation, analysis of
Inorganic and organic constituents, the estimation of intrinsic physical
properties, and the appraisal of waste characteristics.
The procedures described in this manual are meant to be comprehensive and
detailed, coupled with the realization that the problems encountered 1n
sampling and analytical situations require a certain amount of flexibility.
The solutions to these problems will depend, in part, on the skill, training,
and experience of the analyst. For some situations, it is possible to use
this manual 1n rote fashion. In other situations, it will require a
combination of technical abilities, using the manual as guidance rather than
1n a step-by-step, word-by-word fashion. Although this puts an extra burden
on the user, it 1s unavoidable because of the variety of sampling and
analytical conditions found with hazardous wastes.
ORGANIZATION AND FORMAT
This manual 1s divided into two volumes. Volume I focuses on laboratory
activities and is divided for convenience into three sections. Volume IA
deals with quality control, selection of appropriate test methods, and
analytical methods for metallic species. Volume IB consists of methods for
organic analytes. Volume 1C Includes a variety of test methods for
miscellaneous analytes and properties for use 1n evaluating the waste
characteristics. Volume II deals with sample acquisition and Includes quality
control, sampling plan design and Implementation, and field sampling methods.
Included for the convenience of sampling personnel are discusssions of the
ground water, land treatment, and incineration monitoring regulations.
Volume I begins with an overview of the quality control precedures to be
Imposed upon the sampling and analytical methods. The quality control chapter
(Chapter One) and the methods chapters are Interdependent. The analytical
procedures cannot be used without a thorough understanding of the quality
control requirements and the means to implement them. This understanding can
be achieved only be reviewing Chapter One and the analytical methods together.
It 1s expected that Individual laboratories, using SW-846 as the reference
PREFACE - 1
Revision 0
Date September 1986
-------
source, will select appropriate methods and develop a standard operating
procedure (SOP) to be followed by the laboratory. The SOP should Incorporate
the pertinent Information from this manual adopted to the specific needs and
circumstances of the Individual laboratory as well as to the materials to be
evaluated.
The method selection chapter (Chapter Two) presents a comprehensive
discussion of the application of these methods to various matrices 1n the
determination of groups of analytes or specific analytes. It aids the chemist
1n constructing the correct analytical method from the array of procedures
which may cover the matr1x/analyte/concentrat1on combination of Interests.
The section discusses the objective of the testing program and Its
relationship to the choice of an analytical method. Flow charts are presented
along with tables to guide 1n the selection of the correct analytical
procedures to form the appropriate method.
The analytical methods are separated Into distinct procedures describing
specific, Independent analytical operations. These Include extraction,
digestion, cleanup, and determination. This format allows Unking of the
various steps 1n the analysis according to: the type of sample (e.g., water,
soil, sludge, still bottom); analytes(s) of Interest; needed sensitivity; and
available analytical Instrumentation. The chapters describing Miscellaneous
Test Methods and Properties, however, give complete methods which are not
amenable to such segmentation to form discrete procedures.
The Introductory material at the beginning of each section containing
analytical procedures presents Information on sample handling and
preservation, safety, and sample preparation.
Part II of Volume I (Chapters Seven and Eight) describes the
characteristics of a waste. Sections following the regulatory descriptions
contain the methods used to determine 1f the waste is hazardous because 1t
exhibits a particular characteristic.
Volume II gives background information on statistical and nonstatistlcal
aspects of sampling. It also presents practical sampling techniques
appropriate for situations presenting a variety of physical conditions.
A discussion of the regulatory requirements with respect to several
monitoring categories 1s also given in this volume. These Include ground
water monitoring, land treatment, and incineration. The purpose of this
guidance is to orient the user to the objective of the analysis, and to assist
1n developing data quality objectives, sampling plans, and laboratory SOP's.
Significant Interferences, or other problems, may be encountered with
certain samples. In these situations, the analyst 1s advised to contact the
Chief, Methods Section (WH-562B) Technical Assessment Branch, Office of Solid
Waste, US EPA, Washington, DC 20460 (202-382-4761) for assistance. The
manual 1s Intended to serve all those with a need to evaluate solid waste.
Your comments, corrections, suggestions, and questions concerning any material
contained in, or omitted from, this manual will be gratefully appreciated.
Please direct your comments to the above address.
PREFACE - 2
Revision 0
Date September 1986
-------
CHAPTER ONE,
REPRINTED
QUALITY CONTROL
1.1 INTRODUCTION
Appropriate use of data generated under the great range of analytical
conditions encountered 1n RCRA analyses requires reliance on the quality
control practices Incorporated into the methods and procedures. The
Environmental Protection Agency generally requires using approved methods for
sampling and analysis operations fulfilling regulatory requirements, but the
mere approval of these methods does not guarantee adequate results.
Inaccuracies can result from many causes, Including unanticipated matrix
effects, equipment malfunctions, and operator error. Therefore, the quality
control component of each method is indispensable.
The data acquired from quality control procedures are used to estimate
and evaluate the information content of analytical data and to determine the
necessity or the effect of corrective action procedures. The means used to
estimate information content Include precision, accuracy, detection limit, and
other quantifiable and qualitative indicators.
1.1.1 Purpose of this Chapter
This chapter defines the quality control procedures and components that
are mandatory in the performance of analyses, and Indicates the quality
control Information which must be generated with the analytical data. Certain
activities in an Integrated program to generate quality data can be classified
as management (QA) and other as functional (QC). The presentation given here
is an overview of such a program.
The following sections discuss some minimum standards for QA/QC programs.
The chapter is not a guide to constructing quality assurance project plans,
quality control programs, or a quality assurance organization. Generators who
are choosing contractors to perform sampling or analytical work, however,
should make their choice only after evaluating the contractor's QA/QC program
against the procedures presented in these sections. Likewise, laboratories
that sample and/or analyze solid wastes should similarily evaluate their QA/QC
programs.
Most of the laboratories who will use this manual also carry out testing
other than that called for in SW-846. Indeed, many user laboratories have
multiple mandates, Including analyses of drinking water, wastewater, air and
Industrial hygiene samples, and process samples. These laboratories will, In
most cases, already operate under an organizational structure that Includes
QA/QC. Regardless of the extent and history of their programs, the users of
this manual should consider the development, status, and effectiveness of
their QA/QC program 1n carrying out the testing described here.
ONE - 1
Revision 0
Date September 1986
-------
1.1.2 Program Design
The Initial step for any sampling or analytical work should be strictly
to define the program goals. Once the goals have been defined, a program must
be designed to meet them. QA and QC measures will be used to monitor the
program and to ensure that all data generated are suitable for their intended
use. The responsibility of ensuring that the QA/QC measures are properly
employed must be assigned to a knowledgeable person who 1s not directly
Involved in the sampling or analysis.
One approach that has been found to provide a useful structure for a
QA/QC program 1s the preparation of both general program plans and project-
specific QA/QC plans.
The program plan for a laboratory sets up basic laboratory policies,
Including QA/QC, and may include standard operating procedures for specific
tests. The program plan serves as an operational charter for the laboratory,
defining Its purposes, Its organization and Its operating principles. Thus,
1t 1s an orderly assemblage of management policies, objectives, principles,
and general procedures describing how an agency or laboratory Intends to
produce data of known and accepted quality. The elements of a program plan
and Its preparation are described in QAMS-004/80.
Project-specific QA/QC plans differ from program plans 1n that specific
details of a particular sampling/analysis program are addressed. For example,
a program plan might state that all analyzers will be calibrated according to
a specific protocol given in written standard operating procedures for the
laboratory (SOP), while a project plan would state that a particular protocol
will be used to calibrate the analyzer for a specific set of analyses that
have been defined 1n the plan. The project plan draws on the program plan or
Its basic structure and applies this management approach to specific
determinations. A given agency or laboratory would have only one quality
assurance program plan, but would have a quality assurance project plan for
each of Its projects. The elements of a project plan and Its preparation are
described 1n QAMS/005/80 and are listed 1n Figure 1-1.
Some organizations may find 1t Inconvenient or even unnecessary to
prepare a new project plan for each new set of analyses, especially analytical
laboratories which receive numerous batches of samples from various customers
within and outside their organizations. For these organizations, 1t 1s
especially Important that adequate QA management structures exist and that any
procedures used exist as standard operating procedures (SOP), written
documents which detail an operation, analysis or action whose mechanisms are
thoroughly prescribed and which 1s commonly accepted as the method for
performing certain routine or repetitive tasks. Having copies of SW-846 and
all Its referenced documents 1n one's laboratory 1s not a substitute for
having In-house versions of the methods written to conform to specific
Instrumentation, data needs, and data quality requirements.
ONE - 2
Revision 0
Date September 1986
-------
FIGURE 1-1
ESSENTIAL ELEMENTS OF A QA PROJECT PLAN
1. Title Page
2. Table of Contents
3. Project Description
4. Project Organization and Responsibility
5. QA Objectives
6. Sampling Procedures
7. Sample Custody
8. Calibration Procedures and Frequency
9. Analytical Procedures
10. Data Reduction, Validation, and Reporting
11. Internal Quality Control Checks
12. Performance and System Audits
13. Preventive Maintenance
14. Specific Routine Procedures Used to Assess Data
Precision, Accuracy, and Completeness
15. Corrective Action
16. Quality Assurance Reports to Management
ONE - 3
Revision
Date September 1986
-------
1.1.3 Organization and Responsibility
As part of any measurement program, activities for the data generators,
data reviewers/approvers, and data users/requestors must be clearly defined.
While the specific titles of these Individuals will vary among agencies and
laboratories, the most basic structure will include at least one
representative of each of these three types. The data generator 1s typically
the individual who carries out the analyses at the direction of the data
user/requestor or a designate within or outside the laboratory. The data
reviewer/approver is responsible for ensuring that the data produced by the
data generator meet agreed-upon specifications.
Responsibility for data review is sometimes assigned to a "Quality
Assurance Officer" or "QA Manager." This individual has broad authority to
approve or disapprove project plans, specific analyses and final reports. The
QA Officer is Independent from the data generation activities. In general,
the QA Officer 1s responsible for reviewing and advising on all aspects of
QA/QC, including:
Assisting the data requestor 1n specifying the QA/QC procedure to be used
during the program;
Making on-site evaluations and submitting audit samples to assist in
reviewing QA/QC procedures; and,
f problems are detected, making recommendations to the data requestor and
upper corporate/institutional management to ensure that appropriate
corrective actions are taken.
In programs where large and complex amounts of data are generated from
both field and laboratory activities, 1t is helpful to designate sampling
monitors, analysis monitors, and quality control/data monitors to assist in
carrying out the program or project.
The sampling monitor is responsible for field activities. These Include:
Determining (with the analysis monitor) appropriate sampling equipment
and sample containers to minimize contamination;
Ensuring that samples are collected, preserved, and transported as
specified in the workplan; and
Checking that all sample documentation (labels, field notebooks, chain-
of-custody records, packing lists) 1s correct and transmitting that
Information, along with the samples, to the analytical laboratory.
The analysis monitor 1s responsible for laboratory activities. These
Include:
Training and qualifying personnel in specified laboratory QC and
analytical procedures, prior to receiving samples;
ONE - 4
Revision
Date September 1986
-------
Receiving samples from the field and verifying that Incoming samples
correspond to the packing 11st or cha1n-of-custody sheet; and
Verifying that laboratory QC and analytical procedures are being followed
as specified 1n the workplan, reviewing sample and QC data during the
course of analyses, and, 1f questionable data exist, determining which
repeat samples or analyses are needed.
The quality control and data monitor is responsible for QC activities and
data management. These Include:
Maintaining records of all Incoming samples, tracking those samples
through subsequent processing and analysis, and, ultimately,
appropriately disposing of those samples at the conclusion of the
program;
Preparing quality control samples for analysis prior to and during the
program;
Preparing QC and sample data for review by the analysis coordinator and
the program manager; and
Preparing QC and sample data for transmission and entry Into a computer
data base, if appropriate.
1.1.4 Performance and Systems Audits
The QA Officer may carry out performance and/or systems audits to ensure
that data of known and defensible quality are produced during a program,.
Systems audits are qualitative evaluations of all components of field and
laboratory quality control measurement systems. They determine 1f the
measurement systems are being used appropriately. The audits may be carried
out before all systems are operational, during the program, or after the
completion of the program. Such audits typically Involve a comparison of the
activities given 1n the QA/QC plan with those actually scheduled or performed.
A special type of systems audit 1s the data management audit. This audit
addresses only data collection and management activities.
The performance audit 1s a quantitative evaluation of the measurement
systems of a program. It requires testing the measurement systems with
samples of known composition or behavior to evaluate precision and accuracy.
The performance audit is carried out by or under the auspices of the QA
Officer without the knowledge of the analysts. Since this 1s seldom
achievable, many variations are used that increase the awareness of the
analyst as to the nature of the audit material.
ONE - 5
Revision
Date September 1986
-------
1.1.5 Corrective Action
Corrective action procedures should be addressed 1n the program plan,
project, or SOP. These should Include the following elements:
The EPA predetermined limits for data acceptability beyond which
corrective action 1s required;
Procedures for corrective action; and,
For each measurement system, Identification of the Individual responsible
for Initiating the corrective action and the Individual responsible for
approving the corrective action, 1f necessary.
The need for corrective action may be Identified by system or performance
audits or by standard QC procedures. The essential steps 1n the corrective
action system are:
11dentification and definition of the problem;
Assignment of responsibility for Investigating the problem;
Investigation and determination of the cause of the problem;
Determination of a corrective action to eliminate the problem;
Assigning and accepting responsibility for Implementing the corrective
action;
Implementing the corrective action and evaluating Its effectiveness; and
Verifying that the corrective action has eliminated the problem.
The QA Officer should ensure that these steps are taken and that the
problem which led to the corrective action has been resolved.
1.1.6 QA/QC Reporting to Management
QA Project Program or Plans should provide a mechanism for periodic
reporting to management (or to the data user) on the performance of the
measurement system and the data quality. Minimally, these reports should
Include:
Periodic assessment of measurement quality Indicators, I.e., data
accuracy, precision and completeness;
Results of performance audits;
Results of system audits; and
Significant QA problems and recommended solutions.
ONE - 6
Revision
Date September 1986
-------
The individual responsible within the organization structure for
preparing the periodic reports should be identified in the organizational or
management plan. The final report for each project should also include a
separate QA section which summarizes data quality information contained in the
periodic reports.
Other guidance on quality assurance management and organizations is
available from the Agency and professional organizations such as ASTM, AOAC,
APHA and FDA.
1.1.7 Quality Control Program for the Analysis of RCRA Samples
An analytical quality control program develops information which can be
used to:
Evaluate the accuracy and precision of analytical data in order to
establish the quality of the data;
Provide an indication of the need for corrective actions, when comparison
with existing regulatory or program criteria or data trends shows that
activities must be changed or monitored to a different degree; and
To determine the effect of corrective actions.
1.1.8 Definitions
ACCURACY:
ANALYTICAL BATCH:
BLANK:
Accuracy means the nearness of a result or the mean (7) of
a set of results to the true value. Accuracy is assessed
by means of reference samples and percent recoveries.
The basic unit for analytical quality control 1s the
analytical batch. The analytical batch is defined as
samples which are analyzeHtogether with the same method
sequence and the same lots of reagents and with the
manipulations common to each sample within the same time
period or 1n continuous sequential time periods. Samples
in each batch should be of similar composition.
A blank 1s an artificial sample designed to monitor the
Introduction of artifacts Into the process. For aqueous
samples, reagent water 1s used as a blank matrix; however,
a universal blank matrix does not exist for solid samples,
and therefore, no matrix is used. The blank 1s taken
through the appropriate steps of the process.
A reagent blank 1s an aliquot of analyte-free water or
solvent analyzed with the analytical batch. Field blanks
are aliquots of analyte-free water or solvents brought to
the field 1n sealed containers and transported back to the
ONE - 7
Revision 0
Date September 1986
-------
CALIBRATION
CHECK:
CHECK SAMPLE:
laboratory with the sample containers. Trip blanks and
equipment blanks are two specific types of field blanks.
Trip blanks are not opened in the field. They are a check
on sample contamination originating from sample transport,
shipping and from site conditions. Equipment blanks are
opened 1n the field and the contents are poured
appropriately over or through the sample collection device,
collected 1n a sample container, and returned to the
laboratory as a sample. Equipment blanks are a check on
sampling device cleanliness.
Verification of the ratio of Instrument response to analyte
amount, a calibration check, 1s done by analyzing for
appropriate solvent. Calibration
from a stock solution which is
analyte standards in an
check solutions are made
different from the stock used to prepare standards.
A blank which has been spiked with the analyte(s) from an
Independent source in order to monitor the execution of the
analytical method is called a check sample. The level of
the spike shall be at the regulatory action level when
applicable. Otherwise, the spike shall be at 5 times the
estimate of the quantification limit. The matrix used
shall be phase matched with the samples and well
characterized: for an example, reagent grade water 1s
appropriate for an aqueous sample.
ENVIRONMENTAL
SAMPLE:
An environmental sample or field sample is a representative
sample of any material (aqueous, nonaqueous, or multimedia)
collected from any source for which determination of
composition or contamination is requested or required. For
the purposes of this manual, environmental samples shall be
classified as follows:
Surface Water and Ground Water;
Drinking Water — delivered (treated or untreated) water
designated as potable water;
Water/Wastewater — raw source waters for public drinking
water supplies, ground waters, municipal Influents/
effluents, and industrial Influents/effluents;
Sludge — municipal sludges and Industrial sludges;
Waste -- aqueous and nonaqueous liquid wastes, chemical
solids, contaminated soils, and industrial liquid and solid
wastes.
ONE - 8
Revision 0
Date September 1986
-------
MATRIX/SPIKE-
DUPLICATE
ANALYSIS:
MQL:
PRECISION:
PQL:
RCRA:
REAGENT GRADE:
In matrix/spike duplicate analysis, predetermined quanti-
tles of stock solutions of certain analytes are added to a
added to a sample matrix prior to sample extraction/
digestion and analysis. Samples are split Into duplicates,
spiked and analyzed. Percent recoveries are calculated for
each of the analytes detected. The relative percent
difference between the samples 1s calculated and used to
assess analytical precision. The concentration of the
spike should be at the regulatory standard level or the
estimated or actual method quantification limit. When the
concentration of the analyte 1n the sample 1s greater than
0.1%, no spike of the analyte 1s necessary.
The method quantification limit (MQL) 1s the minimum
concentration of a substance that can be measured and
reported.
Precision means the measurement of agreement of a set of
themselves without assumption of
the true result. Precision Is
replicate results among
any prior Information as to
assessed by means of duplicate/replicate sample analysis.
REPLICATE SAMPLE:
The practical quantltatlon limit (PQL) 1s the lowest level
that can be reliably achieved within specified limits of
precision and accuracy during routine laboratory operating
conditions.
The Resource Conservation and Recovery Act.
Analytical reagent (AR) grade, ACS reagent grade, and
reagent grade are synonomous terms for reagents which
conform to the current specifications of the Committee on
Analytical Reagents of the American Chemical Society.
A replicate sample 1s a sample prepared by dividing a
sample into two or more separate aliquots. Duplicate
samples are considered to be two replicates.
STANDARD CURVE: A standard curve 1s a
known analyte standard
the analyte.
curve which plots concentrations of
versus the Instrument response to
SURROGATE:
Surrogates are organic compounds which are similar to
analytes of interest 1n chemical composition, extraction,
and chromatography, but which are not normally found in
environmental samples. These compounds are spiked Into all
blanks, standards, samples and spiked samples prior to
analysis. Percent recoveries are calculated for each
surrogate.
ONE - 9
Revision 0
Date September 1986
-------
WATER: Reagent, analyte-free, or laboratory pure water means
distilled or deionlzed water or Type II reagent water which
1s free of contaminants that may Interfere with the
analytical test in question.
1.2 QUALITY CONTROL
The procedures Indicated below are to be performed for all analyses.
Specific Instructions relevant to particular analyses are given 1n the
pertinent analytical procedures.
1.2.1 Field Quality Control
The sampling component of the Quality Assurance Project Plan (QAPP) shall
Include:
Reference to or Incorporation of accepted sampling techniques 1n the
sampling plan;
Procedures for documenting and justifying any field actions contrary to
the QAPP;
Documentation of all pre-field activities such as equipment check-out,
calibrations, and container storage and preparation;
Documentation of field measurement quality control data (quality control
procedures for such measurements shall be equivalent to corresponding
laboratory QC procedures);
Documentation of field activities;
Documentation of post-field activities Including sample shipment and
receipt, field team de-briefing and equipment check-In;
Generation of quality control samples Including duplicate samples, field
blanks, equipment blanks, and trip blanks; and
The use of these samples in the context of data evaluation, with details
of the methods employed (including statistical methods) and of the
criteria upon which the Information generated will be judged.
1.2.2 Analytical Quality Control
A quality control operation or component 1s only useful 1f 1t can be
measured or documented. The following components of analytical quality
control are related to the analytical batch. The procedures described are
Intended to be applied to chemical analytical procedures; although the
principles are applicable to radio-chemical or biological analysis, the
procedures may not be directly applicable to such techniques.
ONE - 10
Revision 0
Date September 1986
-------
All quality control data and records required by this section shall be
retained by the laboratory and shall be made available to the data requestor
as appropriate. The frequencies of these procedures shall be as stated below
or at least once with each analytical batch.
1.2.2.1 Spikes, Blanks and Duplicates
General Requirements
These procedures shall be performed at least once with each analytical
batch with a minimum of once per twenty samples.
1.2.2.1.1 Duplicate Spike
A split/spiked field sample shall be analyzed with every analytical batch
or once intwentysamples, whichever is the greater frequency. Analytes
stipulated by the analytical method, by applicable regulations, or by other
specific requirements must be spiked into the sample. Selection of the sample
to be spiked and/or split depends on the information required and the variety
of conditions within a typical matrix. In some situations, requirements of
the site being sampled may dictate that the sampling team select a sample to
be spiked and split based on a pre-visit evaluation or the on-site inspection.
This does not preclude the laboratory's spiking a sample of Its own selection
as well. In other situations the laboratory may select the appropriate
sample. The laboratory's selection should be. guided by the objective of
spiking, which 1s to determine the extent of matrix bias or Interference on
analyte recovery and sample-to-sample precision. For soil/sediment samples,
spiking 1s performed at approximately 3 ppm and, therefore, compounds 1n
excess of this concentration 1n the sample may cause Interferences for the
determination of the spiked analytes.
1.2.2.1.2 Blanks
Each batch shall be accompanied by a reagent blank. The reagent blank
shall be carried through the entire analytical procedure.
1.2.2.1.3 Field Samples/Surrogate Compounds
Every blank, standard, and environmental sample (Including matrix
spike/matrix duplicate samples) shall be spiked with surrogate compounds prior
to purging or extraction. Surrogates shall be spiked Into samples according
to the appropriate analytical methods. Surrogate spike recoveries shall fall
within the control limits set by the laboratory (1n accordance with procedures
specified 1n the method or within +20%) for samples falling within the
quantification limits without dilution? Dilution of samples to bring the
analyte concentration Into the linear range of calibration may dilute the
surrogates below the quantification limit; evaluation of analytical quality
then will rely on the quality control embodied 1n the check, spiked and
duplicate spiked samples.
ONE - 11
Revision
Date September 1986
-------
1.2.2.1.4 Check Sample
Each analytical batch shall contain a check sample. The analytes
employed shall be a representative subset of the analytes to be determined.
The concentrations of these analytes shall approach the estimated
quantification limit 1n the matrix of the check sample. In particular, check
samples for metallic analytes shall be matched to field samples in phase and
1n generaT~matr1x composition.
1.2.2.2 Clean-Ups
Quality control procedures described here are Intended for adsorbent
chromatography and back extractions applied to organic extracts. All batches
of adsorbents (FlorlsU, alumina, silica gel, etc.) prepared for use shall be
checked for analyte recovery by running the elutlon pattern with standards as
a column check. The elutlon pattern shall be optimized for maximum recovery
of analytes and maximum rejection of contaminants.
1.2.2.2.1 Column Check Sample
f
The elution pattern shall be reconfirmed with a column check of standard
compounds after activating or deactivating a batch of adsorbent. These
compounds shall be representative of each elutlon fraction. Recovery as
specified in the methods 1s considered an acceptable column check. A result
lower than specified Indicates that the procedure is not acceptable or has
been misapplied.
1.2.2.2.2 Column Check Sample Blank
The check blank shall be run after activating or deactivating a batch of
adsorbent.
1.2.2.3 Determinations
1.2.2.3.1 Instrument Adjustment: Tuning. Alignment, etc.
Requirements and procedures are Instrument- and method-specific.
Analytical Instrumentation shall be tuned and aligned in accordance with
requirements which are specific to the Instrumentation procedures employed.
Individual determinative procedures shall be consulted. Criteria for initial
conditions and for continuing confirmation conditions for methods within this
manual are found in the appropriate procedures.
1.2.2.3.2 Calibration
Analytical Instrumentation shall be calibrated in accordance with
requirements which are specific to the Instrumentation and procedures
employed. Introductory Methods 7000 and 8000 and appropriate analytical
procedures shall be consulted for criteria for Initial and continuing
calibration.
ONE - 12
Revision 0
Date September 1986
-------
1.2.2.3.3 Additional QC Requirements for Inorganic Analysis
Standard curves used In the
prepared as follows:
determination of Inorganic analytes shall be
Standard curves derived from data consisting of one reagent blank and
four concentrations shall be prepared for each analyte. The response for each
prepared standard shall be based upon the average of three replicate readings
of each standard. The standard curve shall be used with each subsequent
analysis provided that the standard curve 1s verified by using at least one
reagent blank and one standard at a level normally encountered or expected in
such samples. The response for each standard shall be based upon the average
of three replicate readings of the standard. If the results of the
verification are not within +10% of the original curve, a new standard shall
be prepared and analyzed. I? the results of the second verification are not
within +10% of the original standard curve, a reference standard should be
employed to determine if the discrepancy 1s with the standard or with the
instrument. New standards should also be
minimum. All data used 1n drawing or
indicated on the curve or its description
verification.
prepared on a quarterly basis at a
describing the curve shall be so
A record shall be made of the
Standard deviations and relative standard deviations shall be calculated
for the percent recovery of analytes from the spiked sample duplicates and
from the check samples. These values shall be established for the twenty most
recent determinations in each category.
1.2.2.3.4 Additional Quality Control Requirements for
Organic Analysis
The following requirements shall be applied to the analysis of samples by
gas chromatography, liquid chromatography and gas chroma tography/mass
spectrometry.
The calibration of each instrument shall be verified at frequencies
specified 1n the methods. A new standard curve must be prepared as specified
in the methods.
The tune of each GC/MS system used for the determination of organic
analytes shall be checked with 4-bromofluorobenzene (BFB) for determinations
of volatlles and with decaf Iuorotr1phenylphosph1ne (DFTPP) for determinations
of semi -volatlles. The required 1on abundance criteria shall be met before
determination of any analytes. If the system does not meet the required
specification for one or more of the required ions, the Instrument must be
retuned and rechecked before proceeding with sample analysis. The tune
performance check criteria must be achieved daily or for each 12 hour
operating period, whichever 1s more frequent.
j
Background subtraction should be straightforward and designed only to
eliminate column bleed or instrument background Ions. Background subtraction
ONE - 13
Revision 0
Date September 1986
-------
actions resulting in spectral distortions for the sole purpose of meeting
special requirements are contrary to the objectives of Quality Assurance and
are unacceptable.
For determinations by HPLC or GC, the instrument calibration shall be
verified as specified in the methods. ,
1.2.2.3.5 Identification
Identification of all analytes must be accomplished with an authentic
standard of the analyte. When authentic standards are not available,
identification is tentative.
For gas chromatographic determinations of specific analytes, the relative
retention time of the unknown must be compared with that of an authentic
standard. For compound confirmation, a sample and standard shall be re-
analyzed on a column of different selectivity to obtain a second
characteristic relative retention time. Peaks must elute within daily
retention time windows to be declared a tentative or confirmed Identification.
For gas chromatographic/mass spectrometrlc determinations of specific
analytes, the spectrum of the analyte should conform to a literature
representation of the spectrum or to a spectrum of the authentic standard
obtained after satisfactory tuning of the mass spectrometer and within the
same twelve-hour working shift as the analytical spectrum. The appropriate
analytical methods should be consulted for specific criteria for matching the
mass spectra, relative response factors, and relative retention times to those
of authentic standards.
1.2.2.3.6 Quantification
The procedures for quantification of analytes are discussed 1n the
appropriate general procedures (7000, 8000) and the specific analytical
methods.
In some situations 1n the course of determining metal analytes, matrix-
matched calibration standards may be required. These standards shall be
composed of the pure reagent, approximation of the matrix, and reagent
addition of major interferents 1n the samples. This will be stipulated 1n the
procedures.
Estimation of the concentration of an organic compound not contained
within the calibration standard may be accomplished by comparing mass spectral
response of the compound with that of an internal standard. The procedure 1s
specified in the methods.
ONE - 14
Revision 0
Date September 1986
-------
1.3 DETECTION LIMIT AND QUANTIFICATION LIMIT
The detection limit and quantification limit of analytes shall be
evaluated by determining the noise level of response for each sample in the
batch. If analyte 1s present, the noise level adjacent in retention time to
the analyte peak may be used. For wave-length dispersive Instrumentation,
multiple determinations of digestates with no detectable analyte may be used
to establish the noise level. The method of standard additions should then be
used to determine the calibration curve using one dlgestate or extracted
sample 1n which the analyte was not detected. The slope of the calibration
curve, m, should be calculated using the following relations:
m = slope of calibration line
SB = standard deviation of the average noise level
MDL = KSB/m
For K = 3; MDL = method detection limit.
For K = 5; MQL = method quantitation limit.
1.4 DATA REPORTING
The requirement of reporting analytical results on a wet-weight or a dry-
weight basis Is dictated by factors such as: sample matrix; program or
regulatory requirement; and objectives of the analysis.
Analytical results shall be reported with the percent moisture or percent
solid content of the sample.
1.5 QUALITY CONTROL DOCUMENTATION
The following sections 11st the QC documentation which comprises the
complete analytical package. This package should be obtained from the data
generator upon request. These forms, or adaptations of these forms, shall be
used by the data generator/reporter for Inorganics (I), or for organlcs (0) or
both (I/O) types of determinations.
1.5.1 Analytical Results (I/O: Form I)
Analyte concentration.
Sample weight.
Percent water (for non-aqueous samples when specified).
Final volume of extract or diluted sample.
Holding times (I: Form X).
ONE - 15
Revision 0
Date September 1986
-------
1.5.2 Calibration (I: Form II; 0: Form V, VI, VII, IX)
Calibration curve or coefficients of the linear equation which
describes the calibration curve.
Correlation coefficient of the linear calibration.
Concentration/response data (or relative response data) of the
calibration check standards, along with dates on which they were
analytically determined.
1.5.3 Column Check (0: Form X)
Results of column chromatography check, with the chromatogram.
1.5.4 Extraction/Digestion (I/O: Form I)
Date of the extraction for each sample.
1.5.5 Surrogates (0: Form II)
Amount of surrogate spiked, and percent recovery of each surrogate.
1.5.6 Matrix/Duplicate Spikes (I: Form V, VI; 0: Form III)
Amount spiked, percent recovery, and relative percent difference for
each compound in the spiked samples for the analytical batch.
1.5.7 Check Sample (I: Form VII; 0: Form VIII)
Amount spiked, and percent recovery of each compound spiked.
1.5.8 Blank (I: Form III; 0: Form IV)
Identity and amount of each constituent.
1.5.9 Chromatograms (for organic analysis)
All chromatograms for reported results, properly labeled with:
- Sample Identification
- Method Identification
- Identification of retention time of analyte on the chromatograms.
ONE - 16
Revision
Date September 1986
-------
1.5.10 Quantitative Chromatogram Report (0: Forms VIII, IX, X)
Retention time of analyte.
Amount Injected.
Area of appropriate calculation of detection response.
Amount of analyte found.
Date and time of Injection.
1.5.11 Mass Spectrum
Spectra of standards generated from authentic standards (one for
each report for each compound detected).
Spectra of analytes from actual analyses.
Spectrometer Identifier.
1.5.12 Metal Interference Check Sample Results (I: Form IV)
1.5.13 Detection Limit (I: Form VII; 0: Form I)
Analyte detection limits with methods of estimation.
1.5.14 Results of Standard Additions (I: Form VIII)
1.5.15 Results of Serial Dilutions (I: Form IX)
1.5.16 Instrument Detection Limits (I: Form XI)
1.5.17 ICP Interelement Correction Factors and ICP Linear Ranges
(when applicable) (I: Form XII. Form XIII).
1.6 REFERENCES
1. Guidelines and Specifications for Preparing Quality Assurance Program
Plans, September 20, 1980, Office of Monitoring Systems and Quality Assurance,
ORD, U.S. EPA, QAMS-004/80, Washington, DC 20460.
2. Interim Guidelines and Specifications for Preparing Quality Assurance
Project Plans, December 29, 1980, Office of Monitoring Systems and Quality
Assurance, ORD, U.S. EPA, QAMS-005/80, Washington, DC 20460.
ONE - 17
Revision
Date September 1986
-------
Lab Kant
COVil'K FACE
INORGANIC ANALYSES DATA PACKACL
Case No.
Q.C. Report l.'o.
EPA No.
Sar.ple I.'urbt. r*
Lab ID No. EKA No.
Conner)i s:
Lab ID No.
ONE - 18
Revision 0
Date September 1986
-------
Forr I
Sample No.
LAB NAME
LAb SAMPLt 1U. NO.
Date
INOKGAMC ANALYSIS UATA SHLL1
CASE NO.
• Lab Receipt Date
QC REPORT NO.
Llerac-nts Identified and Measured
Matrix: Water
Soil
Sludge
Other
ug/L or iib/kt dry weight (Circle One)
1.
2.
3.
5.
fc.
7.
K.
9.
10.
11.
12.
Aluminur.
Antiffionv
Arsenic
bariurt
Beryllium
Cadmiuir.
Calciuir
Chroiriur.
Cobalt
Copper
Iron
Lead
1J.
14.
15.
16.
17.
Ib.
19.
2u.
21.
22.
2J.
Magnesiu-
Man^anese
Mercury
Nickel
Potassiur
Seleniur
Silver
Sodiuc
Thalliur
Vanadiun
Zinc
Precent Solids ( :'. )
Cyanide
Comnrnts:
Lab Manager
ONE - 19
Revision 0
Date September 1986
-------
LAB NAML
Fore II
Q. C. Report No.
INITIAL AND CONTINUING CALIBRATION VLKlFlCATlUN
CASh NO.
DATE
Compound
Metals:
1 . Aluminum
2. Antinony
3. Arsenic
4. bariuir.
5 . Be ry 1 1 i ura
b. CadtriuT?
7. Calciun
b. Chromium
V. Cobalt
JO. Copper
U. Iron
12. Lead
13. Magnesium
14. Manganese
15. Mercury
Ito. Nickel
Initial Calib
True Value
17. Potassium)
iis. Selenium
19. Silver
20. Sodium
/I. Thallium
22. Vanadium |
23. Zinc
Utlicr:
Cyanide j
Found
UNITS: . ug/L
.J Continuing' Calibration2
XR
True Value
i
i
i1
i
Found
2R
1
Found
'
%K
i
1
i
Method1- :
1 il
Initial Calibration Source
(,'ontinuinK Calibration Source
Indicate Analytical Method Used: V - JCP; A - Flame AA; F - Furnace AA
ONE - 20
Revision 0
Date September 1986
-------
LAB NAME
DATE
For IT. Ill
Q- C. Report No.
BLANKS
CASE NiJ.
UNIT*
Compound
Metals:
1 . Aluminur.
2. Antimony
3. Arsenic
A. Bariur
5. beryllium
fc. Cadciurt
7. Calcium
b. Chrotr.iur.
9. Cobalt
10. Copper
11. Iron
12. Lead
13. hagnesiurr,
14. Manganese
15. Mercury
Ib. Nickel
17. Potassium
16. Selenium
IV. Silver
20. Sodium
21. Thallium
22. Vanadium
2J. Zinc
Other:
Cyanide
Initial
Cali bration
Blank Value-
Continuing Calibration
1
Blank Value
2
3
i*
Preparation Elank
Matrix: Matrix:
.' 2
Reporting Units: aqueous,
; solid
ONE - 21
Revision 0
Date September 1986
-------
Form IV
Q. C. Report .No.
LAB NAX£
1CP INTERFERENCE CHECK SAMPLE
CASE NO.
Check Sacple l.p.
DATE
Check Sample Source
Units.' ug/L
Compound
Metals:
1 . Alur.inum
2. Antimony
3. Arsenic
A. Bariuc
5. beryllium
6. Cadmiuir
7. Calcium
b. Chrooiuir
9. Cobalt
10. Copper
11. Iron
12. Lead
13. Magnesium
14. Manganese
15. Mercury
16. Nickel
17. Potassium
IB. Selenium
IV. Silver
2U. Sodium
Control Limits'
'-- Mean"'
21. Thallium )
22. Vanadium
2J. Zinc
Other:
" Std. Dev.
'True^
-
Initial
Observed
*R
Final
Observed
*•/ *-i
XR
Mean value based on n
2 True value of tPA 1CP Interference Check Sample or contractor standard.
ONE - 22
Revision 0
Date September 1986
-------
Fore V
Q. C. Report No.
SPIKE SAMPLE RECOVERY
LAB NAME
DATE
CASE KG.
Sample No.
Lab Sarcple ID No.
Units
Matrix
Compound
Metals:
1. Aluminuc
2. Antimonv
3. Arsenic
4. Bariuc
5. Beryllium
6. Cadciuc
7. Calciun;
8. Chrociur.
9. Cobalt
10. Copper
11. Iron
12. Lead
13. Magnesium
K. Manganese
15. Mercurv
16. Nickel
17. Potassiuc
Ib. Selenium
19. Silver
20. Sodium
21. Thalliuc
22. Vanadium
23. Zinc
Other:
Cyanide
Control Limit
1R
Spiked Sample
Result (SSR)
Sample
Result (SR)
Spiked
Added (SA)
IP-1
1 *R - |(SSR - SK)/SAJ x 100
"N"- out of control
i* |j
NK - Not required
Comments:
ONE - 23
Revision p
Date September 1986
-------
LAB NAM!
DATE
Fpm VI
Q. C. Report No.
DUPLICATES
CASE NO.
Sample No.
Lab Sample ID Mo.
Units
Matrix
Compound
Metals:
1. Aluminum
2. Antimony
3. Arsenic
4. Barium
5. Beryllium
6. Cadmium
7. Calcium
8. Chromium
9. Cobalt
10. Copper
11. Iron
12. Lead
13. Magnesium
14. Manganese
15. Mercurv
16. Nickel
17. Potassium
Ib. Selenium
IV. Silver
20. Sodium
21. Thallium
22. Vanadium
23. Zinc
Other:
Cyanide
Control Limit *
\
Sample(S)
Duplicate(D)
RPD2
* Out of Control
1 To be added at a later date. 2 RPD = ||S - D|/((S + D)/2)] x 100
NC - Non calculable KPL) due to value(s) less than CRDL
ONE - 24
Revision 0
Date September 1986
-------
LAB NAME
Fore VII
Q.C. Report No.
INSTRUMENT DETECTION LIMITS AND
LABORATORY CONTROL SAMPLE
CASE NO.
DATE
LCS 1(0.
Compound
Metals:
1. Aluninurr,
2. Antimony
3. Arsenic
4. Bariuu:
5. Berylliutr
6. Cadraiun
7. Calciun
fa1. Chromium
9. Cobalt
10. Copper
11. Iron
12. Lead
13. Magnesium
14. ilanganese
15. Mercurv
16. Nickel
17. Potassium
18. Selenium
1*. Silver
20. Sodium
21. Thalliun
22. Vanadium
23. Zinc
Other:
Cyanide
Required Detection
Licits (CRDL)-up/!
Instrument Detection
Lioits (IDL)-ug/! '
1CP/AA Furnace
IDfr 1DI?
NR
I;K
Lab Control Sample
ug/L mR/kp
(circle one)
True Found XR
required
ONE - 25
Revision 0
Date September 1986
-------
Fore VI11
Q.C. Report No.
STANDARD ADDITION RESULTS
LAB
UATt
CASE NO.
UMTS .' ug/L
tPA
Sample V
tleroent
Matrix
0 ADD
AUS.
1 ADD
CON.
ABS-
'
2 ADD
CON .
AbS.^
3 ADi,
CUN.
Afci . <•'
FINAL
Cos . 3
r*
i
*• CON is the concentration added, AbS. is the instrument readout in absorbanco or
concentration.
•* Concentration as_ determined by NSA
*"r" is the correlation coefficient.
+ - correlation coefficient is outside ol control window of U.
ONE - 26
Revision 0
Date September 1986
-------
LAB NAME
DATL
Forr I).
Q- C. Report No.
1CP SERIAL DILUTIONS
CASL NO,
Sample No.
Lab Sample 1L <
Units! ug/L
Matrix
Compound
Metals:
1 . Aluminur
2. Antirconv
3. Arsenic
4. bariur
i>. Bervlliuc
t. Cadmium
/. Calciur.
C. Chrorr.iur.
V. Cobalt
lu. Copper
1 1 . Iron
12. Lead
13. Magnesiur.
K. Manganese
li. Nickel
Ib. Potassiur
17. Seleniur
Ib. Silver
IV. Sodiur
20. Thallium
^1 . Vanadiun
ii. Zinc
Ot he r :
Initial Sacple
Concentration( 1 )
Serial Dilution
Result(S)
tt
* Difference''
i
^ Diluted sample concentration corrected for 1:4 dilution (see Exhibit D)
^ Percent Difference = I1 " sl • x 1UU
1
M; - Not keqtiired, initial sample concentration less than lu times ILJL
NA - Not Applicable, analyte not determined by 1CP
ONE - 27
Revision 0
Date September 1986
-------
Form X
QC Report No.
HOLDING JIKL*
LAB NAME
DATE
CASL Ku.
LPA
Sample No.
Matrix
'
Dace
Received
,
Mercury
Prep Date
.
Mercury
Holding Time J
(Davs)
>
CN Prep
Date
CN
Holding Tine1
(Davs)
'holding time is defined as number of days between the date received and the
sample preparation date.
ONE - 28 •
Revision 0
Date September 1986
-------
Fore XI
INSTRUMENT DETECTION LIMITS
LAb NAME
DATE
ICP/Flame AA (Circle One) Model Number
Furnace AA Number
Element
1 . Aluminum
2. Antimonv
3. Arsenic
4 • barium
i. Bervlliucj
0. Cadmium
7. Calcium
b. Chromium
y. Cobalt
1U. Copper
1 1 . Iron
12. Lead
Wavelength
(nm)
| IDL
(ug/L)
Element
13. Magnesium
K. Manganese
J5. Mercury
16. Nickel
17. Potassium
Ib. Selenium
19. Silver
20. Sod i us
21. Thalliur
22. Vanadiur
23. Zinc
Wavelength
(nc)
IUL
(ur/L)
Footnotes:
• Indicate the instrument for which the IUL applies with a "f" (for ICP
an "A" (for Flame AA), or an "F" (for Furnace AA) behind the IbL valu
• Indicate elements commonly run with background correction (AA) with
a "B" behind the analytical wavelength.
• If more than one ICP/Flame or Furnace AA is used, suDmit separate
frorms XI-X11I for each instrument.
COMMtKTb:
Lab Manager
ONE - 29
Revision 0
Date September 1986
-------
Fort XII
ICP Interelement Correction Factors
LABORATORY
DATE
ICP Model Nutber
Analyte
1. Antinonv
2. Arsenic
3. Bariuc
A. Bervlliur
5. Cadciur
6. Chroiriur
7. Cobalt
£. Copper
9. Lead
'J. Manganese
1. Mercurv
2. Nickel
j. Potassiur
<. Seleniuc
). Silver
3. Sodium
J. Thalliur.
i. Vanadium
». /inc
Analyte
Wavelength
(n=)
I
Interelereen: Correction Jactcr<:
for
Al
Ca
•
•
Fe
ME
i
1
i i i •
i
•
-
•
COMMENTS:
Lab Manager
ONE - 30
Revision 0
Date September 1986
-------
Fore XII
1CP Interelement Correction Factors
LABORATOkY_
DATE
ICP Model Nucber
Analyte
1. Antimonv
2. Arsenic
3. Bariutr
4. Berylliuc
5. Cadoiur,
6. Chromlurr
7. Cobalt
fc. Copper
9. Lead
10. Manganese
11. Mercury
12. Nickel
13. Potassium
K. Selenium
15. Silver
16. Sodium
17. Thallium
18. Vanadium
IV. Zinc
Analyte
Wavelength
(nr.)
Interelement Correction Factors
for
I
1
COMMENTS:
Lab Manager
ONE - 31
Revision 0
Date September 1986
-------
LAB NAME
DATE
Form XIII
ICP Linear Ranges
ICP Model Number
Analyte
1. Aluminum
2. Antimony
3. Arsenic
4. Bariur:
5. Beryllium
6. Cadnium
7. Calciun
6. Chrociutr
9. Cobalt
10. Copper
11. Iron
12. Lead
Integration
Tine
(Seconds )
Concen-
tration
(ug/L)
Analyte
13. Magnesiuc
14. Manganese
15. Mercurv
16. Nickel
17. Potassiuc
18. Selenium
19. Silver
20. Sodiur.
21. Thalliur:
22. Vanadiur
23. Zinc
Integration
Time
(Seconds)
Concen-
tration
(ug/L)
Footnotes:
• Indicate elements not analyzed by ICP with the notation "NA"
COMMENTS .-
Lab Manager
ONE - 32
Revision 0
Date September 1986
-------
Organics Analysis Data Sheet
(Pagel)
Sample Number
Laboratory Name:
Lab Sample ID Nc
Sample Matrix
Case No:
QC Report No:
Data Release Authorized By
Date Sample Received:
Volatile Compounds
Date Extracted/Prepared:
Date Analyzed:
Conc'Dil Factor:
-PH.
Percent Moisture: (Not Decanted).
CAS ug/lorug/Kg
Number (Circle One)
74-87-3
74-83-9
75-01-4
75-00-3
75-09-2
67-64-1
75-15-0
75-35-4
75-34-3
156-60-5
67-66-3
107-06-2
78-93-3
71-55-6
56-23-5
108-05-4
75-27-4
Cnloromeihane
Bromomethaie
Vtnv! CniO'ide
Chloroethane
Metnylene Cnloride
Acetone '
Carbon Disulfide
1. 1-Dichloroethene
1. 1-Dichloroethane
Trans- 1. 2-Dichloroethene
Chloroform
1. 2-Dichloroethane
2-Butsnone
1.1, 1-Trichloroethane
Carbon Tetrachloride
Vinyl Acetate
Bromodichloromethane
CAS ug/lorug/Kg
Number (Circle One;
76-87-5
10061-02-6
79-01-6
124-48-1
79-00-5
71-43-2
10061-01-5
110-75-8
75-25-2
108-10-1
591-78-6
127-18-4
79-34-5
108-88-3
108-90-7
100-41-4
100-42-5
1. 2-Dichloropropane
Trans- 1. 3-Dichloropropene
Trichloroethene
Dibromochloromethane
1.1, 2-Trichioroethane
Benzene
cis-1. 3-Dichloropropene
2-Chloroethylvinyleihe'
Bromoform
4-Methyl-2-Penianone
2-Hexanone
Tetrachloroethene
1. 1.2. 2-Teirachioroethane
Toluene
Chlorobenzene
Ethylbenzene
Styrene
Total Xylenes
D*u Reporting QuiMien
for reporting results lo EPA. Iht following mulii qualifiers art used
Additional flags or footnotes e tentatively identified compounds
where a 1 1 response is assumed or when the mass spectral date
indicated the presence of a compound mat meets the identification
criteria but the result is less than the rpecilied detection limn BUI
greater than nro (• g . 10JI II limn of detection it 10 pg 'I and a
concentration of 3 *ig 'I is calculated report at 3J
Other
This llag applies to pesticide parameters wnere Hie idemilicaiion nas
been confirmed by GC'MS Single component pesucidei210
ng 'ul in the linal attract should be confirmed by GC MS
This llag is used when the anaiyte is louno in ine blank as wen at a
sample It indicates possible probable blank contamination and
warns the data user to lake appropriate action
Other specific Hags and lootnotes maybe required 10 properly define
the results H used, they must be fuliydescnoed and sucndescnpnon
attached to the data summary repon
Form I
ONE - 33
Revision 0
Date September 1986
-------
Laboratory Name
Cose No
Sample Number
Date Extracted/Prepared:
Date Analyzed.
Organic* Analysis Data Sheet
(Page 2j
Semivolatile Compounds
GPC Cleanup DYes DNo
Separatory Funnel Extraction DYes
Continuous Liquid • Liquid Extraction DYes
Conc/Dil Factor.
Percent Moisture (Decanted).
CAS
Number
ug/lorug/Kg
(Circle One)
108-95-2
111.44-4
95-57-8
541-73-1
106-46-7
100-51-6
95-50-1
95-46-7
39638-32-9
106-44-5
621-647
67-72-1
98-95-3
78-59-1
88-75-5
105-67-9
65-85-0
111-91-1
120-83-2
120-82-1
91-203
106-47-8
87-68-3
59-50-7
91-57-6
77-47-4
88-06-2
95-95-4
91-58-7
88-74-4
131-11-3
208-96-8
99-09-2
Pnenol
bis(-2-Chloroethyl)Ether
2-Cnlorophenol
1. 3-Dichlorobenzene
1, 4-Dichlorobenzene
Benzyl Alcohol
1. 2-Dichlorobenzene
2-Methylpheno!
bis(2-chloroisopropyl)Ether
4-Methylphenol
N-Nitroso-Di-n-Propylamme
Hexachloroeihane
Nitrobenzene
Isophorone
2-Nitrophenol
2. 4-Dimethylphenol
Benzole Acid
bis(-2-Chloroethoxy)Methane
2, 4-Dichlorophenol
1, 2, 4-Trichlorobenzcne
Naphthalene
4-Chloroaniline
Hexachlorobutadiene
4-Chloro-3-Methylphenol
2-Methylnaphthalene
Hexachlorocyclopentadiene
2. 4. 6-Trichlorophenol
2. 4. 5-Trichloropheno!
2-Chloronaphthalene
2-Nitroaniline
Dimethyl Phthalaie
Acenaphthylene
3-Nitroaniline
CAS ug/lorug/Kg
Number (Circle One!
83-32-9
51-28-5
100-02-7
132-64-9
121-14-2
606-20-2
84-66-2
7005-72-3
86-73-7
100-01-6
534-52-1
86-30-6
101-55-3
118-74-1
87-86-5
85-01-8
120-12-7
84-74-2
206-44-0
129-00-0
85-68-7
91-94-1
56-55-3
117-81-7
218-01-9
117-84-0
205-99-2
207-08-9
50-32-8
193-39-5
53-70-3
191-24-2
Acenaphthene
2. 4-Dinitrophenol
4-Nitropheno!
Dibenzofuran
2, 4-Dinitrotoluene
2, 6-Dinitrotoluene
Diethylphthalate
4-Chlorophenyl-phenyle:her
Fluorene
4-Nitroaniline
4. 6-Dmitro-2-Methylphenol
N-Nitrosodiphenylamine (1)
4-Bromophenyl-phenylether
Hexachlorobenzene
Pentachlorophenol
Pnenanthrene
Anthracene
Di-n-Butylphthalaie
Fluoranthene
Pyrene
Butylbenzylphthalaie
3, 3 -Dichlorobenzidine
Benzo(8)Anthracene
bis(2-Ethylhexyl)Pnthalate
Chrysene
Oi-n-Octyl Phthalate
Benzo(b)Fluoranthene
Benzo(k)Fluoranthene
Benzo(a)Pyrene
Indenod. 2, 3-cd)Pyrene
Dibenz(a. h)Anthracene
Benzofg h, i)Perylene
(1 ^-Cannot be separated from diphenylamine
Form I
ONE - 34
Revision 0
Date September 1986
-------
Laboratory Name
Case No
Sample Number
Date Extracted/Prepared
Date Analyzed:
Conc/Dil Factor:
Organics Analysis Data Sheet
(Page 3}
Pesticide/PCBs
GPC Cleanup DYes DNo
Separatory Funnel Extraction DYes
Continuous Liquid - Liquid Extraction DYes
Percent Moisture (decanted).
CAS og/lorug/Kg
Number (Circle One)
319-84-6
319-85-7
319-86-8
58-89-9
76-41- 6
309-00-2
1024-57-3
959-96-8
60-57-1
72-55-9
72-20E
33213-65-9
72-54-8
1031-07-8
50-29-3 '
72-43-5
53494-70-5
57-74-9
8001-35-2
12674-11-2
11104-28-2
11141-16-5
53469-21-9
12672-29-6
11097-69-1
11096-82-5
Alpha-BHC
Beia-BHC
Delia-BHC
Gamma-BHC (Lindane)
Heptachio'
Aldnn
Hepiachlor Epomde
Endosulfan 1
Oielcirin
4.4'-DDE
Endrir,
Endosulfan r.
4.4--DDD
Endosulfan Sulfaie
4. 4'- DDT
Metho»ychlor
Endrin Ketone
Chlordane
Toxaphene
Aroclor-1016
Aroclor-1221
Aroelor-1232
Aroclrr-1242
Aroclor-1248
Aroclor-1254
Aroclor-1260
OrW.
V- = Volume of extract injected (ul)
V£ - Volume of water extracted (ml)
W£ = Weight of cample extracted (g)
V, = Volume of total extract (ul)
Form 1
ONE - 35
Revision 0
Date September 1986
-------
Laboratory Name:
Case No:
Organics Analysis Data Sheet
Sample Number
CAS
Number
1-
3-
a
A
s
«
7
H
«»
10
11
15
13
14
IB
1fi
17
IP
19
30
31
33
33
34
3K
3fi
3-7
3R
38
•»»
Compound Name
Fraction
RT or Scan
Number
Estimated
Concentration
(ug/lorug/kg)
Form 1, Pan 8
ONE - 36
Revision 0
Date September 1986
-------
WATER SURROGATE PERCENT RECOVERY SUMMARY
o
r+
ft)
O
n>
-j
I— •
VO
00
o
m
OJ
30
«
<.
v> .
o
0
NO-
TOtUCMC-M
IM-III)
14 OICW.OKO-
CTMAMC-04
•
KITHO-
(31-114)
t-riumo-
BirHCHTL
(43-11*)
OI4
(31-141)
VALUES ARE OUTSIDE OF REQUIRED OC LIMITS Volat
Pesti
Comments!
EMI-VOLATIL
E— — — — ____ __ 1
(10-»«>
|p«»? out of ;
-Vi)l^tilf4-l>4>
outside of OC limits
outside of OC limits
outside of OC limits
FORM H
-------
SOIL SURROGATE PERCENT RECOVERY SUMMARY
O
&>
i-*
n>
(S>
n
0
«-••
n
§•
ESTICIOC--
B«utTi-
c~in«rNo«Tt
(rn-iiwn
outside of OC limits
outside of OC limits
outside of OC limits
-------
WATER MATRIX SPIKE/MATRIX SPIKE DUPLICATE RECOVERY
Case No.
Laboratory Name.
to
VO
FRACTION
VOA
SAMPLE NO.
B/N
SAMPLE NO.
Ann
SAMPLE NO
PEST
SAMPLE NO.
COMPOUND
1 , 1 -Dichloroethene
Trichloroethene
CMorobervcn*
Toluene
Benzene
1 ,2.4-Tr ichlnrobenzrne
Acenaohth.?no
2.4 Diniirotoluene
Di-n-Butylphthalate
Pyrene
N-Nitroso-Di-n-Propylamim
1 ,4-Oichlorobenzene
Pentachlorophenol
Phenol
2-Chlorophenol
4-CMoro-3-Methylphenol
4-Nitrophenol
Lindane
Heptachlor
Aldrin
Dieldrin
Endrin
4 4'DDT
CONC. SPIKE
ADDED (uq/L)
SAMPLE
RESULT
CONC.
MS
%
RCC
CONC.
MSO
%
REC
Don
o
npo
14
14
13
13
11
28
31
38
40
31
38
28
50
42
40 -
42
50
15
20
22
18
21
27
P LIMITS
RbCOVERY
6JLJ4S
75-130
76-125 ,
76-127
39-98 1
46-118
24-96
11-117
26-127
41-116
36 97
9 103
12-89
27-123
23 97
10 80
56 123
40 131
40 120
52 126
56 121
T8-177
O 73
O> O>
r+ <
(0 ->•
(/> O
fO 3
O
r+
n
cr
vo
oo
en
ADVISORY LIMITS
RPD: VOAt nut of nut Of! limits R/IM
ADD nitf of oMtV'l'' Of! limits ADO
PFST r,iir nf Out^ir1p OC limits PF«;T .,
out of
out of
; outside OC
; outside OC
outside OC
outside OC
limits
limits
limits
limits
FORM III
-------
SOIL MATRIX SPIKE/MATRIX SPIKE DUPLICATE RECOVERY
Case No.
Laboratory Name.
I
o
O 73
o» n
r+ <
n —•
(/> O
n 3
o
«-»•
(D
o
VO
oo
Ot
FRACTION
VGA
SAMPLE NO.
B/N
SAMPLE NO.
ACID
SAMPLE NO.
PEST
SAMPLE NO.
COMPOUND
1.1-Dicholorethene
Trichloroethene
Chlorobenrene
Toluene
Benzene
1 ,2,4-Trichlorobenzpne
Acenaphthene
2,4 Dinitrololuene
Di-n-Butylphthalatc
Pyrene
N-Nitrosodi-n-Propytamtne
1 .4-Dichlorobenj>enp
Pentachlorophcnol
Phenol
2-Chlorophenol
4-Chloro-3-Methylphpnol
4-Nitrophenol
Lindane
Heptachlor
Aldrin
Oieldrin
Endrin
4.4'-DOT
CONC. SPIKE
ADDF.D (i*|/Kq)
SAMPLE
RFSULT
CONC.
MS
%
RF.C
CONC>
MSD
%
REC
RPD
Of
RPD
22
24
21
21
21
23
19
47
47
36
38
27
47
35
50
33
50
50
31
i_ 43
38
45
50
r LIM'I?
RECOVERY
59-172
82-137
60-133
59139
66-142
38-107
31-137
28-89
29-135
35-142
41-126
28-104
17-109
2690
25-102
26103
11-114
46-127
35130
34-132
31-134
42-139
23134
ADVISORY LIMITS
RPD:
VOA* n.* nl
R/N nn« of
APin mrt nl
PEST out of
outside OC
outside OC
outside OC
outside OC
limits
limits
limits
limits
RFCOVFRY- VOA* out of
B/N out of
AC'n. ..out of
PEST out of
outside
outside
outside
outside
OC
OC
OC
OC
limit*
limit*
limit*
limits
••ii • '
FORM III
-------
METHOD BLANK SUMMARY
Case No.
Laboratory Name.
O 30
o> n
c+ <
n —
(/> o
(V 13
o
rue w
OATC or
• N«LTSIS
rn/kcriON
MATRIX
COWC.
Lf.VfA
INST. 10
CAS NUMBER
COMPOUND (MSL.TIC OA UNKHOWM)
CONC.
UNITS
CTOU
CO
en
FORM IV
-------
GC/MS TUNING AND MASS CALIBRATION
Bromofluorobenzene (BFB)
Case No..
Instrument ID
Laboratory Name
Date
Time.
Data Release Authorized By:
m/e
ION ABUNDANCE CRITERIA
^RELATIVE ABUNDANCE
50
75
95
96
173
174
175
176
177
15.0 • 40.0% of the base peak
30.0 • 60.0% of the base peak
Base peak, 100% relative abundance
5.0 • 9.0% of the base peak
Less than 1.0% of the base peak
Greater than 50.0% of the base peak
5.0 -9.0% of mass 174
Greater than 95.0%, but less than 101.6% of mass 174
5.0 -9.0% of mass 176
( )1
C )'
( >'
THIS PERFORMANCE TUNE APPLIES TO THE FOLLOWING
SAMPLES. BLANKS AND STANDARDS.
Value in parenthesis is % mass 174.
'Value in parenthesis is % mass 176.
SAMPLE ID
LAB ID
DATE OF ANALYSIS TIME OF ANALYSIS
FORM V
ONE - 42
Revision 0
Date September 1986
-------
Case No.
GC/MS TUNING AND MASS CALIBRATION
Decafluorotrlphenylphosphine (DFTPP)
Laboratory Name
Instrument ID
Date
Time .
Data Release Authorized By:
m/e
ION ABUNDANCE CRITERIA
VRELATIVE ABUNDANCE
51
68
69
70
127
197
198
199
275
365
441
442
443
30.0 • 60.0% of mass 198
less than 2.0% of mass 69
mass 69 relative abundance
lets than 2.0% of mass 69
40.0 • 60.0% of mass 19E
less than 1.0% of mass 198
base peak, 100% relative abundance
5.0 • 9.0% of mass 19E
10.0 • 30.0% of mass 19E
greater than 1.00% of mass 198
present, but less than mass 4<3
greater than 40.0% of mass 19E
17.0 • 23.0% of mass 442
c V
( )'
c )5
THIS PERFORMANCE TUNE APPLIES TO THE FOLLOWING
SAMPLES, BLANKS AND STANDARDS.
Value in parenthesis is % mass 6S.
Value in parenthesis is % mass 442.
SAMPLE ID
LAB ID
DATE OF ANALYSIS
TIME OF ANALYSIS
FORM V
ONE - 43
Revision 0
Date September 1986
-------
Case No:
Laboratory Name
Initial Calibration Data
Volatile HSL Compounds
Instrument I D: .
Calibration Date:
Minimum RFfor SPCC is 0.300
(0.25 for Bromoform)
Maximum % RSD for CCC is 30%
Laboratory ID
Compound
Chloromethane
Bromomethane
Vinyl Chloride
Chloroethane
Methylene Chloride
Acetone
Carbon Disulfide
1. 1-Bichloroethene
1. 1-DiChloroethane
Trans-1, 2-Dichloroethene
Chloroform
1. 2-Dichloroethane
2-Butenone
1.1. 1-Trichloroeihane
Carbon Tetrachloride
Vinyl Acetate
Bromodichloromethane
1, 2-Dichloropropane
Trans- 1, 3-Dichloropropene
Trichloroethene
Dibromochloromethane
1.1, 2-Trichloroethane
Benzene
cis-1. 3-Dichloropropene
2-Chloroethylvinylether
Bromoform
4-Methyl-2-Penianone
2-Hexanone
Tetrachloroethene
1.1.2. 2-Tetrachloroethanc
Toluene
Chlorobenzene
Ethylbenzene
Styrene
Total Xylenes
RF20
"FED
RF100
(
*
,
RF150
RF200
RT
% RSD
,.
CCC-
SPCC««
* •
•
»
• •
*
•
• •
• *
*
» »
•
RF -Response Factor (subscript is the amount of ug/L)
RT-Average Response Factor
%RSO -Percent Relative Standard Deviation
CCC -Calibration Check Compounds (•)
SPCC -System Performance Check Compounds (•«)
Form VI
ONE - 44
Revision Q
Date September 1986
-------
Initial Calibration Data
Volatile HSL Compounds
Case No:
Laboratory Name.
Instrument I D: .
Calibration Date:
Minimum RF for SPCC is 0.300 Maximum % RSD for CCC is 30%
(0.25 for Bromoform)
Laboratory ID
Compound
RF20
RF
100
RF150
RF200
VoRSD
CCC-
SPCC*
RF -RasportM Factor (subscript it the amount of ug/L)
HT -Average Response Factor
%RSD -Percent Relative Standard Deviation
CCC -Calibration Check Compounds (•)
SPCC -System Performance Check Compounds (••)
Form VI
ONE - 45
Revision p
Date September 1986
-------
Case No:
Laboratory Name.
Initial Calibration Data
Semivolatile HSL Compounds
(Pagel)
Instrument ID: _
Calibration Date:
Minimum RF for SPCC is 0.050 Maximum % RSD for CCC is 30%
Laboratory ID
Compound
Phenol
bis(-2-Chloroethyl)Ether
2-Chlorophenol
1. 3-Dichlorobenzene
1,4-Dichlorobenzene
Benzyl Alcohol
1. 2-Dichlorobenzene
2-Methylpheno! '
bis(2-chloroisopropyl)Ether
4-Methylphenol
N-Nnroso-Di-n-Propylamine
Hexachioroethane
Nitrobenzene
Isophorone
2-Nitrophenol
2. 4-Dimethylpheno!
Benzoic Acid
bis(-2-Chloroethoxv)Methane
2. 4-Dichlorophenol
1. 2. 4-Tnchlorobenzene
Naphthalene
4-Chloroanilme
Hexachlorobutadiene
4-Chloro-3-Methylpheho!
2-Methylnaphthalene
Hexachlorocyclopentadiene
2. 4. 6-Tnchloropheno!
2. 4. 5-Trichlorophenol
2-Chloronaphthalene
2-Nitroaniline
Dimethyl Phthalate
Acenaphthylene
3-Nitroaniline
Acenaphthene
2. 4-Dmitrophenol
4-Nitrophenol
Dibenzoturan
RF20
- '
t
T
t
t
t
t
"F50
RFso
RF120
RF160
RF
VRSD
CCC«
SPCC««
•
•
• •
•
•
•
•
• •
•
•
• •
• *
Response Factor (subscript is the amount of na nog rams)
RF -Average Response Factor
%RSD -Percent Relative Standard Deviation
CCC -Calibration Check Compounds (•)
SPCC -System Performance Check Compounds <••)
t -Not detectable at 20 ng
Form VI
ONE - 46
Revision 0
Date September 1986
-------
Case No:
Laboratory Name
Initial Calibration Data
Semivolatile HSL Compounds
(Page 2)
Instrument ID: _
Calibration Date:
Minimum RF for SPCC is 0.050 Maximum % RSD for CCC is 30%
Laboratory ID
Compound
2. 4-Dinitrotoluene
2. 6-Dinitrotoluene
Diethylphthalate
4-Chlorophenyl-phenylether
Fluorene
4-Nitroanilme
4, 6-Dmiuo-2-Methylpheno!
N-Nitrosodiphenylamine (1)
4-Bromophenyl-phenylether
Hexachlorobenzene
Pentachloropheno!
Phenanthrene
Anthracene
Di-N-Butylphthalate
Fluoranthene
Pyrene
Butylbenzylphthalate
3, 3°-Dichlorobenzidme
Benzo(a)Anthracene
bis(2-Ethylhexyl)Phthalaie
Chrysene
Di-n-Octyl Phthalate
Benzo{b)Fluoranthene
Benzo(k)Fluoranthene
Benzo(8)Pyrene
Indenod. 2. 3-cd)Pyrene
Diberu(a, h)Anthracene
Benzo(g. h. i)Perylene
RF20
t
t
t
RF50
R^BO
RF120
RF160
WT
KRSD
CCC«
SPCC««
•
»
•
•
•
Response Factor (subscript is the amount of nanoerams)
H? -Average Response Factor
%RSD -Percent Relative Standard Deviation
CCC -Calibration Check Compounds (•)
SPCC -System Performance Check Compounds (••)
t -Not detectable at 20 ng
(1) -Cannot be separated from diphenylamine
Form VI
3
ONE - 47
Revision o
Date September 1986
-------
Case No
Laboratory Name.
Initial Calibration Data
Semivolatile HSL Compounds
(Page 1)
Instrument ID _
Calibration Date.
Minimum RF for SPCC is 0.050 Maximum °/o RSD for CCC is 30c/i
Laboratory ID
Compound
RF
20
RFc
RF
120
'160
RF
%RSD
CCC«
SPCC«
Response Factor (subscript is the amount of nanograms)
R? -Average Response Factor
*oRSD -Percent Relative Standard Deviano"
CCC -Calibration Check Compounds (•}
SPCC -System Performance Check Compounds («i|
1 -Not detectable at 20 ng
Form VI
ONE - 48
Revision 0
Date September 1986
-------
Case No:
Laboratory Name.
Contract No:
Continuing Calibration Check
Volatile HSL Compounds
Calibration Date:
—, Time:
Instrument ID:
Minimum RF for SPCC is 0.300
(0.25 for Bromoform)
Laboratory ID:
Initial Calibration Date:
Maximum %D for CCC is 25%
Compound
Chloromethane
Bromomethane
Vinyl Chloride
Chloroethane
Methylene Chloride
Acetone
Carbon Disulfide
1. 1-Dichloroethene
1, 1-Dichloroethane
Trans-1. 2-Dichloroethene
Chloroform
1. 2-Dichloroe!haie
2-Butanone
1,1, 1-Tnchloroethane
Carbon Teuachloride
Vinyl Acetate
Bromodichlorometha ne
1. 2-Dichloropropane
Trans-1. 3-Dichloropropene
Trichloroethene
Dibromochloromethane
1.1. 2-Trichloroethane
Benzene
cis-1. 3-Dichloropropene
2-Chloroethylvinylether
Bromoforrr.
4-Methyl-2-Pentanone
2-Hexanone
Tetrachloroethene
1, 1.2, 2-Tetrechloroethane
Toluene
Chlorobenzene
Ethylberuene
Styrene
Total Xylenes
R~F
«FBO
VD
CCC
*
•
»
•
*
*
SPCC
• •
• •
• •
• •
• •
"F
60 -
Factor from daily standard file at 50 us/l
RF -Average Response Factor from initial calibration Form VI
%D -Percent Difference
CCC -Calibration Check Compounds (•)
SPCC -System Performance Check Compounds (••)
Form VII
ONE - 49
Revision 0
Date September 1986
-------
Continuing Calibration Check
Volatile HSL Compounds
Case No
Laboratory Name
Contract No
Instrument ID:
Calibration Date
Time '
Laboratory ID:
Initial Calibration Date
Minimum RF for SPCC is 0.300
(0.25 for Bromoform)
Maximum %D for CCC is 25%
Compound
RF
RF
60
CCC
SPCC
RFgQ -Response Factor from daily standard file at 50 ug I
RF -Average Response Factor from initial calibration Form VI
°t>D -Percent Difference
CCC -Calibration Check Compounds (•)
SPCC -System Performance Cnec*. Compounds (..I
. VII
ONE - 50
Revision 0
Date September 1986
-------
Continuing Calibration Check
Semivolatile HSL Compounds
(Pagel)
Case No:
Laboratory Name.
Instrument ID:
Calibration Date:
Time:
Laboratory JD:
Initial Calibration Date:
Minimum RF for SPCC is 0.050 Maximum %D for CCC is 25%
Compound
Phenol
bis(-2-Chloroethyl)Ether
2-Chlorophenol
1. 3-Dichlorobentene
1.4-Dichlorobenzene
Ben/yl Alcohol
1. 2-Dichlorobeniene
2-Meiiiylphenol
bis(2-Chloroisopropyl)Ether
4-Methylpheno!
N-Nitroso-Di-n-Propylamine
Hexachloroethane
Nitrobeniene
Isophorone
2-Nuropheno!
2. 4-Dimethylpheno!
Benzoic Acid ^
bis(-2-Chloroeihony|Methane
2. 4-Dichlorophenol
1. 2. 4-Tnchlorobenzene
Naphthalene
4-Chloroamhne
Hexachlornbutadiene
4-Chloro-3-Melhylpheno!
2-Meihylnaphihalene
Hexachlorocyclopeniadiene
2. 4. 6-Tnchlorophenol
2. 4. 5-Trichlotophenol ^
2-Chloronaphthalene
2-Nilroanilme |
Dimethyl Pnthalate
Acenaphthylene
3-Nitroanilme |
Acenaphihene
2. 4-Dmitrophenol
4-Nnrophenol
Dibenzoluran
RT
RF50
% D
CCC
*
*
*
*
•
*
*
•
SPCC
* •
* *
• *
* •
RFgg -Response Factor from daily standard die at concentration
indicated (SO toul n*nogr§ms)
RF -Average Response Factor from initial calibration Form VI
+>Due to low rctponse, analyze
It BO tout ninograms
1oD -Percent Difference
CCC -Calibration Check Compounds (>)
SPCC -System Performance Check Compounds (•
Form VII
ONE - 51
Revision 0
Date September 1986
-------
Continuing Calibration Check
Semivolatile HSL Compounds
(Page 2)
Case No:
Laboratory Name.
Instrument ID:
Calibration Dale:
Time:
Laboratory ID.
Initial Calibration Date:
Minimum RF for SPCC is 0.050 Maximum %D for CCC is 25%
70
Compound
2. 4-Dinitrotoluene
2, 6-Dinitrotoluene
Diethylphthalate
4-Chlorophenyl-phenylether
Fluorene
4-Nitroaniline t
4. 6-Dmitro-2-Meihylphenoi f
N-Nitrosodiphenylamine (1 )
4-Bromophenyl-phenylet>ier
Hexachlorobenzene
Pentachloropheno' t
Phenanthrene
Anthracene
Di-N-Butylphihalaie
Fluoranthene
Pyrene
Butylbenrylphthalate
3, 3'-Dichloroberuidme
Benzo(a)Anthracene
bis(2-Ethylhexyl)Pnthalate
Chrysene
Di-n-Octyl Phthalaie
Beruo(b)Fluoranthene
Benzo(k)Fluoranihene
Benzo(8)Pyrene
lndeno(1. 2. 3-cd)Pyrene
Dibenzfa. h)Anthracene
Benzo(8. h. i)Perylene
RF
RF60
%D
CCC
*
•
•
•
•
SPCC
*
RFtQ -Ri'spunsc F.u'lor (ruin d.iily st.inil.iiil lilt- .11 cuncviilf.it
indicated (50 total nanograms)
RT -Aveiiiye Respuiibf F.iclor Irom initi.il t.ililx.ihun Form VI
"«D Petcenl Dilleienff
^••Ou'e to low response, analyze
•t 80 total nanograms
CCC -CiiiUr.ilioii Check CoiiipuumJs (•)
SPCC -System Perluriiiiince Check Compoundi |.. I
(1) C«nnol !)L- kep.ifJli.-U (ruin Uiplienyl.iiiiinc
Form VII
ONE - 52
Revision 0
Date September
1986
-------
Continuing Calibration Check
Semivolatile HSL Compounds
(Pagel)
Case No:
Laboratory Name.
Instrument ID:
Calibration Date: .
Time: _
Laboratory ID:
Initial Calibration Date:
Minimum RF for SPCC is 0.050 Maximum %D for CCC is 25%
Compound
RF
50
CCC
SPCC
RFcQ -Response Faciur Irom daily st.inil.inl tile .11 co
indicoied (50 total nanojrams!
ET -Avornge Response F.iclur Irom initial c.ilibi
-------
Pesticide Evaluation Standards Summary
(Pagel)
Case No
Date of Analysis,
Laboratory Name:.
GC Column'.
Instrument ID..
Evaluation Check for Linearity
Laboratory
ID
Pesticide
Aldrm
Endrm
6.6 - DDT*'1
Dibutyl
Cniorendate
Calibration
Factor
Eval. Mix A
Calibration
Facto?
Eval. MixB
Calibration
Factor
Eval. Mix C
% RSD
( <10c/o)
Evaluation Check for 4,4'- DDT/Endrin Breakdown
(percent breakdown expressed as total degradation)
Eval Mix B
72 Hour
Eva: Mix B
Evai Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Eval Mix B
Laboratory
I.D.
•
Time of
Analysis
Endrin
4.4'- DDT
Combined '
(1) See Exhibit E. Section 7.5.4-
(2) See Exhibit E. Section 7.3.1.2.2.1
Form VIII
RCRA
4/86
ONE - 54
Revision Q
Date September 1986
-------
Pesticide Evaluation Standards Summary
(Page 2)
Evaluation of Retention Time Shift for Dibutyl Chlorendate
Report all standards, blanks and samples
Sample No
Laii
I.D
Time of
Analysis
Percent
Diff.
SMO
Sample No.
Lab
I.D
Time ot
Analysis
Percent
Diff.
RCRA
Form VIII (Continued) 4/66
ONE - 55
Revision 0
Date September 1986
-------
PESTICIDE/PCB STANDARDS SUMMARY
Case No.,
Laboratory Name.
QC Column
GC Instrument ID
COMPOUND
alpha -BHC
beta-BHC
delta -BHC
gamma— BHC
Heptachlor
AMrin
Heptachlor Epoxide
Endosuffan I
Dieldrin
4.4'-DDE
Endrin
Endosulfan I
4,4'-DDD
Endrin Aldehyde
Endosutfan Sulfate
4,4'-DDT
Methoxychlor
Endrin Ketone
Tech. Chlordane
alpha-Chlordane
gamma-Chlordane
Toxaphene
Aroclor - 1 0 1 6
Aroclor - 1 22 1
Aroclor - 1 232
Aroclor - 1 24 .
Aroclor- 1248
Aroclor - 1 254
Aroclor - 12 GO
DATE OF AN
TIME OF AN/
LABORATORY
RT
Al YRIft
1 YRIR
t in
RETENTION
TIME
WINDOW
-
CALIBRATION
FACTOR
CONF.
OR
OUANT.
DATE OF ANj
TIME OF AN/
LABORATOR'
RT
M YSIS
11 YRIR
Y 10
CALIBRATION
FACTOR
CONF.
OR
OUANT.
PERCENT
DIFF.**
I
01
O 73
O> CO
r* <
n -••
in
CO O*
O) 3
O
r+
CO
**
= CONFIRMATION «!O%
OMANT.— OHANTITATir'M (-:ir''v.
-------
Case No.
P««tlclde/PCB Identification
Laboratory Name.
I
tn
030
tu n
r+ <
n -^
VI
(/) O
n 3
n
n
vo
CD
SAMPLE
10
PRIMARY
COLUMN
PESTICIDE/
PCB
RT OF
TENTATIVE
10
RT WINDOW
OF APPROPRIATE
STANDARD
CONFIRMATION
COLUMN
RT ON
CONFIRMATORY
COLUMN
RT WINDOW OF
APPROPRIATE
STANDARD
GC/MS
CONFIRMED
(Y or N)
FORM X
-------
PART III SAMPLING
Revision 0
Date September 1986
-------
CHAPTER NINE
SAMPLING PLAN
9.1 DESIGN AND DEVELOPMENT
The initial — and perhaps most critical -- element in a program designed
to evaluate the physical and chemical properties of a solid waste 1s the plan
for sampling the waste. It is understandable that analytical studies, with
their sophisticated instrumentation and high cost, are often perceived as the
dominant element 1n a waste characterization program. Yet, despite that
sophistication and high cost, analytical data generated by a scientifically
defective sampling plan have limited utility, particularly in the case of
regulatory proceedings.
This section of the manual addresses the development and implementation
of a scientifically credible sampling plan for a solid waste and the
documentation of the chain of custody for such a plan. The information
presented in this section is relevant to the sampling of any solid waste,
which has been defined by the EPA in its regulations for the identification
and listing of hazardous wastes to include solid, semi sol id, liquid, and
contained gaseous materials. However, the physical and chemical diversity of
those materials, as well as the dissimilarity of storage facilities (lagoons,
open piles, tanks, drums, etc.) and sampling equipment associated with them,
preclude a detailed consideration of any specific sampling plan. Conse-
quently, because the burden of responsibility for developing a technically
sound sampling plan rests with the waste producer, it is advisable that he/she
seek competent advice before designing a plan. This 1s particularly true in
the early developmental stages of a sampling plan, at which time at least a
basic understanding of applied statistics is required. Applied statistics is
the science of employing techniques that allow the uncertainty of inductive
inferences (general conclusions based on partial knowledge) to be evaluated.
9.1.1 Development of Appropriate Sampling Plans
An appropriate sampling plan for a solid waste must be responsive to both
regulatory and scientific objectives. Once those objectives have been clearly
identified, a suitable sampling strategy, predicated upon fundamental statis-
tical concepts, can be developed. The statistical terminology associated with
those concepts is reviewed 1n Table 9-1; Student's "t" values for use 1n the
statistics of Table 9-1 appear 1n Table 9-2.
9.1.1.1 Regulatory and Scientific Objectives
The EPA, in its hazardous waste management system, has required that
certain solid wastes be analyzed for physical and chemical properties. It is
mostly chemical properties that are of concern, and, in the case of a number
of chemical contaminants, the EPA has promulgated levels (regulatory
thresholds) that cannot be equaled or exceeded. The regulations pertaining to
the management of hazardous wastes contain three references regarding the
NINE - 1
Revision 0
Date September 1986
-------
TABLE 9-1. BASIC STATISTICAL TERMINOLOGY APPLICABLE TO SAMPLING PLANS FOR SOLID WASTES
Terminology
Symbol
Mathematical equation
(Equation)
Variable (e.g., barium
or endrln)
Individual measurement
of variable
Mean of all possible
measurements of variable
(population mean)
Mean of measurements
generated by sample
(sample mean)
V- =
N
E >
1=1
, with N = number of
possible measurements
Simple random sampling and
systematic random sampling
x =
n
E x
1=1
n
1
with n = number of
sample measurements
(1)
(2a)
Stratified random sampling
x =
E W. x
k=l k
with XK = stratum (2b)
mean and W^ = frac-
tion of population
represented by Stratum
k (number of strata
[k] range from 1 to r)
• Variance of sample
Simple random sampling and
systematic random sampling
n 9 n
E xf - (E >
2 1=1 ^ 1=1
s =—inn—
,)2/n
(3a)
Stratified random sampling
s =
r
E W
k=l
ksk
with s£ = stratum (3b)
variance and WV =
fraction of population
represent by Stratum k
(number of strata [k]
ranges from 1 to r)
NINE - 2
Revision 0
Date September 1986
-------
TABLE 9-1. (Continued)
Terminology
Symbol
Mathematical equation
(Equation)
Standard deviation of
sample
Standard error
(also standard error
of mean and standard
deviation of mean)
of sample
sx
(4)
(5)
Confidence Interval
for &
CI
CI = 7 + t.20
, With t.20
obtained from
Table 2 for
appropriate
degrees of freedom
(6)
Regulatory threshold3
RT
Defined by EPA (e.g., 100 ppm for (7)
barium in elutriate of EP toxldty)
Appropriate number of
samples to collect from
a solid waste (financial
constraints not considered)
n =
, with A = RT - x
(8)
Degrees of freedom
df
df = n - 1
(9)
Square root transformation
+ 1/2
(10)
Arcsln transformation
Arcsln p; if necessary, refer to any (11)
text on basic statistics;
measurements must be con-
verted to percentages (p)
aThe upper limit of the CI for /* 1s compared with the applicable regulatory
threshold (RT) to determine 1f a solid waste contains the variable (chemical
contaminant) of concern at a hazardous level. The contaminant of concern 1s not
considered to be present in the waste at a hazardous level 1f the upper limit of the CI
is less than the applicable RT. Otherwise, the opposite conclusion is reached.
NINE - 3
Revision 0
Date September 1986
-------
TABLE 9-2. TABULATED VALUES OF STUDENT'S "t" FOR EVALUATING
SOLID WASTES
Degrees of
freedom (n-l)a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
40
60
120
Tabulated
"t" valueb
3.078
1.886
1.638
1.533
1.476
1.440
1.415
1.397
1.393
1.372
1.363
1.356
1.350
1.345
1.341
1.337
1.333
1.330
1.328
1.325
1.323
1.321
1.319
1.318
1.316
1.315
1.314
1.313
1.311
1.310
1.303
1.296
1.289
1.282
^Degrees of freedom (df) are equal to the number of samples (n)
collected from a solid waste less one.
tabulated "t" values are for a two-tailed confidence Interval
and a probability of 0.20 (the same values are applicable to a one-tailed
confidence interval and a probability of 0.10).
NINE - 4
Revision 0
Date September 1986
-------
sampling of solid wastes for analytical properties. The first reference,
which occurs throughout the regulations, requires that representative samples
of waste be collected and defines representative samples as exhibiting average
properties of the whole waste. The second reference, which pertains just to
petitions to exclude wastes from being listed as hazardous wastes, specifies
that enough samples (but in no case less than four samples) be collected over
a period of time sufficient to represent the variability of the wastes. The
third reference, which applies only to grounHwater monitoring systems,
mandates that four replicates (subsamples) be taken from each ground water
sample intended for chemical analysis and that the mean concentration and
variance for each chemical constituent be calculated from those four
subsamples and compared with background levels for ground water. Even the
statistical test to be employed in that comparison is specified (Student's t-
test).
The first of the above-described references addresses the issue of
sampling accuracy, and the second and third references focus on sampling
variability or, conversely, sampling precision (actually the third reference
relates to analytical variability,which,fn many statistical tests, is
indistinguishable from true sampling variability). Sampling accuracy (the
closeness of a sample value to its true value) and sampling precision (the
closeness of repeated sample values) are also the issues of overriding
importance in any scientific assessment of sampling practices. Thus, from
both regulatory and scientific perspectives, the primary objectives of a
sampling plan for a solid waste are twofold: namely, to collect samples that
will allow measurements of the chemical properties of the waste that are both
accurate and precise. If the chemical measurements are sufficiently accurate
and precise, they will be considered reliable estimatesof the chemical
properties of the waste.
It is now apparent that a judgment must be made as to the degree of
sampling accuracy and precision that is required to estimate reliably the
chemical characteristics of a solid waste for the purpose of comparing those
characteristics with applicable regulatory thresholds. Generally, high
accuracy and high precision are required if one or more chemical contaminants
of a solid waste are present at a concentration that is close to the
applicable regulatory threshold. Alternatively, relatively low accuracy and
low precision can be tolerated if the contaminants of concern occur at levels
far below or far above their applicable thresholds. However, a word of
caution is in order. Low sampling precision is often associated with
considerable savings in analytical, as well as sampling, costs and is clearly
recognizable even 1n the simplest of statistical tests. On the other hand,
low sampling accuracy may not entail cost savings and is always obscured 1n
statistical tests (i.e., it cannot be evaluated). Therefore, although 1t is
desirable to design sampling plans for solid wastes to achieve only the
minimally required precision (at least two samples of a material are required
for any estimate of precision), it is prudent to design the plans to attain
the greatest possible accuracy.
NINE - 5
Revision 0
Date September 1986
-------
The roles that Inaccurate and Imprecise sampling can play 1n causing a
solid waste to be Inappropriately judged hazardous are Illustrated 1n Figure
9-1. When evaluating Figure 9-1, several points are worthy of consideration.
Although a sampling plan for a solid waste generates a mean concentration
00 and standard deviation (s, a measure of the extent to which Individual
sample concentrations are dispersed around 7) for each chemical contaminant of
concern, 1t 1s not the variation of Individual sample concentrations that 1s
of ultimate concern, but rather the variation that characterizes 7 Itself.
That measure of dispersion 1s termed the standard deviation of the mean (also,
the standard error of the mean or standard error) and is designated as s^.
Those two sample values, 7 and sy, are used to estimate the Interval (range)
within which the true mean (/*) of the chemical concentration probably occurs,
under the assumption that the individual concentrations exhibit a normal
(bell-shaped) distribution. For the purposes of evaluating solid wastes, the
probability level (confidence interval) of 80% has been selected. That is,
for each chemical contaminant of concern, a confidence interval (CI) is
described within which p occurs if the sample is representative, which is
expected of about 80 out of 100 samples. The upper limit of the 80% CI is
then compared with the appropriate regulatory threshold. If the upper limit
1s less than the threshold, the chemical contaminant 1s not considered to be
present in the waste at a hazardous level; otherwise, the opposite conclusion
is drawn. One last point merits explanation. Even 1f the upper limit of an
estimated 80% CI is only slightly less than the regulatory threshold (the
worst case of chemical contamination that would be judged acceptable), there
is only a 10% (not 20%) chance that the threshold is equaled or exceeded.
That is because values of a normally distributed contaminant that are outside
the limits of an 80% CI are equally distributed between the left (lower) and
right (upper) tails of the normal curve. Consequently, the CI employed to
evaluate solid wastes is, for all practical purposes, a 90% interval.
9.1.1.2 Fundamental Statistical Concepts
The concepts of sampling accuracy and precision have already been
introduced, along with some measurements of central tendency (7) and
dispersion (standard deviation [s] and sy) for concentrations of a chemical
contaminant of a solid waste. The utility of 7 and sy in estimating a
confidence Interval that probably contains the true mean (/i) concentration of
a contaminant has also been described. However, it was noted that the
validity of that estimate is predicated upon the assumption that individual
concentrations of the contaminant exhibit a normal distribution.
Statistical techniques for obtaining accurate and precise samples are
relatively simple and easy to implement. Sampling accuracy is usually
achieved by some form of random sampling. In random sampling, every unit in
the population (e.g., every location in a lagoon used to store a solid waste)
has a theoretically equal chance of being sampled and measured. Consequently,
statistics generated by the sample (e.g., 7 and, to a lesser degree, sy) are
unbiased (accurate) estimators of true population parameters (e.g., the CI for
p). In other words, the sample is representative of the population. One of
the commonest methods of selecting a random sample is to divide the
NINE - 6
Revision 0
Date September 1986
-------
ACCURATE AND PRECISE SAMPLE
(Warn Appropriately Judged Nonhazardous)
ACCURATE AND IMPRECISE SAMPLE
(Watte Inappropriately Judged Hazardous)
o
'*•
1
Y OF VALUES
o o
kl W
1 1
FREQUE
o
—
True Mean (M) and Sample Mean (x)
Standard Error (15) -7
Confidence
Interval (CD
Upper
f
1 Regulatory
Threshold (RTJ
ix-11
70 75 80 85 90 95 100 105 110
CONCENTRATION OF BARIUM (ppm)
65
70 75 80 85 90 95 100 105 110
CONCENTRATION OF BARIUM (ppm)
INACCURATE AND PRECISE SAMPLE
(Waste Inappropriately Judged Hazardous)
INACCURATE AND IMPRECISE SAMPLE
(Waste Inappropriately Judged Hazardous)
P
*.
1
OF VALUES
IS-7
QU
P
L.
1
0.4 -
IS -11
O
65
85 90 95 100 105 110
CONCENTRATION OF BARIUM (ppm)
65
i i i
70 75 80 85 90 95 100 105 110
CONCENTRATION OF BARIUM (ppm)
NOTE: In All CUM, Confidence Interval for M • 5 ± t JQ »5-
Figure 9-1 .-Important theoretical relationships between sampling accuracy and precision and
regulatory objectives for a chemical contaminant of a solid waste that occurs at a concentration
marginally lea than its regulatory threshold. In this example, barium is the chemical contaminant.
The true mean concentration of barium in the elutriate of the EP toxicity test is 85 ppm, as compared
to a regulatory threshold of 100 ppm. The upper limit of the confidence interval for the true
mean concentration, which is estimated from the sample mean and standard error, must be less Than
the regulatory threshold if barium is judged to be present in the waste at a nonhazardous level.
NINE - 7
Revision 0
Date September 1986
-------
population by an imaginary grid, assign a series of consecutive numbers to the
units of the grid, and select the numbers (units) to be sampled through the
use of a random-numbers table (such a table can be found in any text on basic
statistics). It is important to emphasize that a haphazardly selected sample
is not a suitable substitute for a randomly selected sample. That is because
there is no assurance that a person performing undisciplined sampling will not
consciously or subconsciously favor the selection of certain units of the
population, thus causing the sample to be unrepresentative of the population.
Sampling precision is most commonly achieved by taking an appropriate
number of samples from the population. As can be observed from the equation
for calculating sy, precision increases (sy and the CI for /* decrease) as the
number of samples (n) increases, although not in a 1:1 ratio. For example, a
100% increase in the number of samples from two to four causes the CI to
decrease by approximately 62% (about 31% of that decrease is associated with
the critical upper tail of the normal curve). However, another 100% increase
in sampling effort from four to eight samples results in only an additional
39% decrease in the CI. Another technique for increasing sampling precision
is to maximize the physical size (weight or volume) of the samples that are
collected.That has theeffect of minimizing between-sample variation and,
consequently, decreasing sy. Increasing the number or size of samples taken
from a population, in addition to increasing sampling precision, has the
secondary effect of increasing sampling accuracy.
In summary, reliable information concerning the chemical properties of a
solid waste is needed for the purpose of comparing those properties with
applicable regulatory thresholds. If chemical information is to be considered
reliable, it must be accurate and sufficiently precise. Accuracy is usually
achieved by incorporating some form of randomness into the selection process
for the samples that generate the chemical information. Sufficient precision
is most often obtained by selecting an appropriate number of samples.
There are a few ramifications of the above-described concepts that merit
elaboration. If, for example, as in the case of semiconductor etching
solutions, each batch of a waste is completely homogeneous with regard to the
chemical properties of concern and that chemical homogeneity is constant
(uniform) over time (from batch to batch), a single sample collected from the
waste at an arbitrary location and time would theoretically generate an
accurate and precise estimate of the chemical properties. However, most
wastes are heterogeneous in terms of their chemical properties. If a batch of
waste is randomly heterogeneous with regard to its chemical characteristics
and that random chemical heterogeneity remains constant from batch to batch,
accuracy and appropriate precision can usually be achieved by simple random
sampling. In that type of sampling, all units in the population (essentially
all locations or points in all batches of waste from which a sample could be
collected) are identified, and a suitable number of samples is randomly
selected from the population. More complex stratified random sampling is
appropriate if a batch of waste is known to be nonrandomly heterogeneous in
terms of its chemical properties and/or nonrandom chemical heterogeneity is
known to exist from batch to batch. In such cases, the population is
stratified to isolate the known sources of nonrandom chemical heterogeneity.
NINE - 8
Revision 0
Date September 1986
-------
After stratification, which may occur over space (locations or points 1n a
batch of waste) and/or time (each batch of waste), the units In each stratum
are numerically Identified, and a simple random sample Is taken from each
stratum. As previously Intimated, both simple and stratified random sampling
generate accurate estimates of the chemical properties of a solid waste. The
advantage of stratified random sampling over simple random sampling is that,
for a given number of samples and a given sample size, the former technique
often results 1n a more precise estimate of chemical properties of a waste (a
lower value of sy) than the latter technique. However, greater precision 1s
likely to be realized only 1f a waste exhibits substantial nonrandom chemical
heterogeneity and stratification efficiently "divides" the waste into strata
that exhibit maximum between-strata variability and minimum within-strata
variability. If that does not occur, stratified random sampling can produce
results that are less precise than in the case of simple random sampling.
Therefore, it is reasonable to select stratified random sampling over simple
random sampling only if the distribution of chemical contaminants in a waste
is sufficiently known to allow an intelligent identification of strata and at
least two or three samples can be collected in each stratum. If a strategy
employing stratified random sampling is selected, a decision must be made
regarding the allocation of sampling effort among strata. When chemical
variation within each stratum can be estimated with a great degree of detail,
samples should be optimally allocated among strata, I.e., the number of
samples collected from each stratumshould be directly proportional to the
chemical variation encountered 1n the stratum. When detailed information
concerning chemical variability within strata 1s not available, samples should
be proportionally allocated among strata, i.e., sampling effort 1n each
stratum should be directly proportional to the size of the stratum.
Simple random sampling and stratified random sampling are types of
probability sampling, which, because of a reliance upon mathematical and
statistical theories, allows an evaluation of the effectiveness of sampling
procedures. Another type of probability sampling is systematic random
sampling, in which the first unit to be collected from a population 1s
randomly selected, but all subsequent units are taken at fixed space or time
intervals. An example of systematic random sampling is the sampling of a
waste lagoon along a transect in which the first sampling point on the
transect is 1 m from a randomly selected location on the shore and subsequent
sampling points are located at 2-m intervals along the transect. The
advantages of systematic random sampling over simple random sampling and
stratified random sampling are the ease with which samples are identified and
collected (the selection of the first sampling unit determines the remainder
of the units) and, sometimes, an increase in precision. In certain cases, for
example, systematic random sampling might be expected to be a little more
precise than stratified random sampling with one unit per stratum because
samples are distributed more evenly over the population. As will be
demonstrated shortly, disadvantages of systematic random sampling are the poor
accuracy and precision that can occur when unrecognized trends or cycles occur
in the population. For those reasons, systematic random sampling is recom-
mended only when a population is essentially random or contains at most a
modest stratification. In such cases, systematic random sampling would be
employed for the sake of convenience, with little expectation of an increase
in precision over other random sampling techniques.
NINE - 9
Revision 0
Date September 1986
-------
Probability sampling 1s contrasted with authoritative sampling, in which
an individual who is well acquainted with thesolidwaste to be sampled
selects a sample without regard to randomization. The validity of data
gathered in that manner is totally dependent on the knowledge of the sampler
and, although valid data can sometimes be obtained, authoritative sampling is
not recommended for the chemical characterization of most wastes.
It may now be useful to offer a generalization regarding the four
sampling strategies that have been identified for solid wastes. If little or
no information is available concerning the distribution of chemical
contaminants of a waste, simple random sampling 1s the most appropriate
sampling strategy. As more information is accumulated for the contaminants of
concern, greater consideration can be given (in order of the additional
Information required) to stratified random sampling, systematic random
sampling, and, perhaps, authoritative sampling.
The validity of a CI for the true mean (/*) concentration of a chemical
contaminant of a solid waste is, as previously noted, based on the assumption
that individual concentrations of the contaminant exhibit a normal
distribution. This is true regardless of the strategy that 1s employed to
sample the waste. Although there are computational procedures for evaluating
the correctness of the assumption of normality, those procedures are
meaningful only if a large number of samples are collected from a waste.
Because sampling plans for most solid wastes entail just a few samples, one
can do little more than superficially examine resulting data for obvious
departures from normality (this can be done by simple graphical methods),
keeping in mind that even if individual measurements of a chemical contaminant
of a waste exhibit a considerably abnormal distribution, such abnormality is
not likely to be the case for sample means, which are our primary concern.
One can also compare the mean of the sample (7) °with the variance of the
sample (s2). in a normally distributed population, 7 would be expected to be
greater than s2 (assuming that the number of samples [n] is reasonably large).
If that is not the case, the chemical contaminant of concern may be
characterized by a Poisson distribution (7 1s approximately equal to s2) or a
negative binomial distribution[7Ts less than s2). In the former
circumstance, normality can often be achieved by transforming data according
to the square root transformation. In the latter circumstance, normality may
be real 1 zed through use of the arcsine transformation. If either
transformation is required, all subsequentstatisticalevaluations must be
performed on the transformed scale.
Finally, it is necessary to address the appropriate number of samples to
be employed in the chemical characterization of a solid waste.As has already
.been emphasized, the appropriate number of samples is the least number of
samples required to generate a sufficiently precise estimate of the true mean
(/*) concentration of a chemical contaminant of a waste. From the perspective
of most waste producers, that means the minimal number of samples needed to
demonstrate that the upper limit of the CI for ft is less than the applicable
regulatory threshold (RT). The formula for estimating appropriate sampling
effort (Table 9-1, Equation 8) indicates that increased sampling effort is
generally justified as s2 or the "t.20" value (probable error rate) Increases
NINE - 10
Revision 0
Date September 1986
-------
and as A(RT - 7) decreases. In a well-designed sampling plan for a solid
waste, an effort 1s made to estimate the values of 7 and s^ before sampling Is
Initiated"! Such preliminary estimates, which may be derived from information
pertaining to similar wastes, process engineering data, or limited analytical
studies, are used to identify the approximate number of samples that must be
collected from the waste. It is always prudent to collect a somewhat greater
number of samples than indicated by preliminary estimates of 7 and s^ since
poor preliminary estimates of those statistics can result in an underest1mate
of the appropriate number of samples to collect. It is usually possible to
process and store the extra samples appropriately until analysis of the
initially identified samples is completed and it can be determined if analysis
of the additional samples is warranted.
9.1.1.3 Basic Sampling Strategies
It is now appropriate to present general procedures for implementing the
three previously introduced sampling strategies (simple random sampling,
stratified random sampling, and systematic random sampling) and a hypothetical
example of each sampling strategy. The hypothetical examples illustrate the
statistical calculations that must be performed in most situations likely to
be encountered by a waste producer and, also, provide some insight into the
efficiency of the three sampling strategies in meeting regulatory objectives.
The following hypothetical conditions are assumed to exist for all three
sampling strategies. First, barium, which has an RT of 100 ppm as measured 1n
the EP elutriate test, is the only chemical contaminant of concern. Second,
barium is discharged in particulate form to a waste lagoon and accumulates in
the lagoon in the form of a sludge, which has built up to approximately the
same thickness throughout the lagoon. Third, concentrations of barium are
relatively homogeneous along the vertical gradient (from the water-sludge
interface to the sludge-lagoon interface), suggesting a highly controlled
manufacturing process (little between-batch variation in barium concen-
trations). Fourth, the physical size of sludge samples collected from the
lagoon is as large as practical, and barium concentrations derived from those
samples are normally distributed (note that we do not refer to barium levels
in the samples of sludge because barium measurements are actually made on the
elutriate from EP toxicity tests performed with the samples). Last, a
preliminary study of barium levels in the elutriate of four EP toxiclty tests
conducted with sludge collected from the lagoon several years ago Identified
values of 86 and 90 ppm for material collected near the outfall (in the upper
third) of the lagoon and values of 98 and 104 ppm for material obtained from
the far end (the lower two-thirds) of the lagoon.
For all sampling strategies, it is important to remember that barium will
be determined to be present in the sludge at a hazardous level 1f the upper
limit of the CI for u is equal to or greater than the RT of 100 ppm (Table 9-
1, Equations 6 and 7).
NINE - 11
Revision 0
Date September 1986
-------
9.1.1.3.1 Simple Random Sampling
Simple random sampling (Box 1) is performed by general procedures in
which preliminary estimates of 7 and s2, as well as a knowledge of the RT, for
each chemical contaminant of a solid waste that is of concern are employed to
estimate the appropriate number of samples (n) to be collected from the waste.
That number of samples is subsequently analyzed for each chemical contaminant
of concern. The resulting analytical data are then used to conclude
definitively that each contaminant is or is not present in the waste at a
hazardous concentration or, alternatively, to suggest a reiterative process,
involving increased sampling effort, through which the presence or absence of
hazard can be definitively determined.
In the hypothetical example for simple random sampling (Box 1),
preliminary estimates of 7 and s2 indicated a sampling effort consisting of
six samples. That number of samples was collected and initially analyzed.
generating analytical data somewhat different from the preliminary data (s2
was substantially greater than was preliminarily estimated). Consequently,
the upper limit of the CI was unexpectedly greater than the applicable RT,
resulting in a tentative conclusion of hazard. However, a reestlmation of
appropriate sampling effort, based on statistics derived from the six samples,
suggested that such a conclusion might be reversed through the collection and
analysis of just one more sample. Fortunately, a resampling effort was not
required because of the foresight of the waste producer 1n obtaining three
extra samples during the initial sampling effort, which, because of their
influence in decreasing the final values of 7, 57, t.20» an£l» consequently,
the upper limit of the CI — values obtained from all nine samples — resulted
in a definitive conclusion of nonhazard.
9.1.1.3.2 Stratified Random Sampling
Stratified random sampling (Box 2) 1s conducted by general procedures
that are similar to the procedures described for simple random sampling. The
only difference is that, in stratified random sampling, values of 7 and s2 are
calculated for each stratum in the population and then integrated Into overall
estimates of those statistics, the standard deviation (s), 57, and the
appropriate number of samples (n) for all strata.
The hypothetical example for stratified random sampling (Box 2) is based
on the same nine sludge samples previously identified in the example of simple
random sampling (Box 1) so that the relative efficiencies of the two sampling
strategies can be fully compared. The efficiency generated through the
process of stratification is first evident in the preliminary estimate of
n (Step 2 in Boxes 1 and 2), which is six for simple random sampling and four
for stratified random sampling. (The lesser value for stratified sampling is
the consequence of a dramatic decrease in s2, which more than compensated for
a modest Increase 1n A.) The most relevant indication of sampling efficiency
is the value of sy, which is directly employed to calculate the CI. In the
case of simple random sampling, sy is calculated as 2.58 (Step 9 in Box 1),
and, for stratified random sampling, sy is determined to be 2.35 (Steps 5 and
7 in Box 2). Consequently, the gain in efficiency attributable to
stratification is approximately 9% (0,23/2.58).
NINE - 12
Revision 0
Date September 1986
-------
BOX 1. STRATEGY FOR DETERMINING IF CHEMICAL CONTAMINANTS OF SOLID WASTES
ARE PRESENT AT HAZARDOUS LEVELS - SIMPLE RANDOM SAMPLING
Step General Procedures
1. Obtain preliminary estimates of 7 and s2 for each chemical contaminant of
a solid waste that 1s of concern. The two above-Identified statistics
are calculated by, respectively, Equations 2a and 3a (Table 9-1).
2. Estimate the appropriate number of samples (ni) to be collected from
the waste through use of Equation 8 (Table 9-1) and Table 9-2. Derive
Individual values of ni for each chemical contaminant of concern.
The appropriate number of samples to be taken from the waste 1s the
greatest of the Individual nj values.
3. Randomly collect at least nj (or n£ - ni, n$ - r\2, etc., as will be
Indicated later 1n this box) samples from the waste (collection of a
few extra samples will provide protection against poor preliminary
estimates of 7 and s2). Maximize the physical size (weight or volume) of
all samples that are collected.
4. Analyze the ni (or r\2 - n\, n^ - r\z etc.) samples for each chemical
contaminant of concern. Superficially (graphically) examine each set of
analytical data for obvious departures from normality.
5. Calculate 7, s2, the standard deviation (s), and Sy for each set of
analytical data by, respectively, Equations 2a, 3a, 4, and 5 (Table 9-1).
6. If 7 for a chemical contaminant Is equal to or greater than the
applicable RT (Equation 7, Table 9-1) and 1s believed to be an accurate
estimator of /*, the contaminant 1s considered to be present 1n the
waste at a hazardous concentration, and the study 1s completed.
Otherwise, continue the study. In the case of a set of analytical data
that does not exhibit obvious abnormality and for which 7 1s greater than
s2, perform the following calculations with nontransformed data.
Otherwise, consider transforming the data by the square root
transformation (If 7 Is about equal to s2) or the arcslne transformation
(1f 7 1s less than s2) and performing all subsequent calculations with
transformed data. Square root and arcslne transformations are defined
by, respectively, Equations 10 and 11
(Table 9-1).
7. Determine the CI for each chemical contaminant of concern by Equation 6
(Table 9-1) and Table 9-2. If the upper limit of the CI 1s less than the
applicable RT (Equations 6 and 7, Table 9-1), the chemical contaminant 1s
not considered to be present 1n the waste at a hazardous concentration
and the study Is completed. Otherwise, the opposite conclusion Is
tentatively reached.
NINE - 13
Revision 0
Date September 1986
-------
8. If a tentative conclusion of hazard 1s reached, reestlmate the total
number of samples (n2) to be collected from the waste by vuse of
Equation 8 (Table 9-1) and Table 9-2. When_der1v1ng n?, employ the newly
calculated (not preliminary) values of x and s2. If additional
r\2 - n\ samples of waste cannot reasonably be collected, the study 1s
completed, and a definitive conclusion of hazard 1s reached. Otherwise,
collect extra i\2 - nj samples of waste.
9. Repeat the basic operations described 1n Steps 3 through 8 until the
waste 1s judged to be nonhazardous or, 1f the opposite conclusion
continues to be reached, until Increased sampling effort 1s Impractical.
Hypothetical Example
Step
The preliminary study of barium 'levels in the elutriate of four EP
toxldty tests, conducted with sludge collected from the lagoon several
years ago, generated values of 86 and 90 ppm for sludge obtained from
the upper third of the lagoon and values of 98 and 104 ppm for sludge
from the lower two-thirds of the lagoon. Those two sets of values are
not judged to be indicative of nonrandom chemical heterogeneity
(stratification) within the lagoon. Therefore, preliminary estimates of
7 and s2 are calculated as:
x =
n
E X
1=1
1
n
86 + 90 + 98+104
= 94.50, and
(Equation 2a)
s =
n ~ n ?
E Xf - (E XjVn
1=1 1 1=1 1
n - 1
(Equation 3a)
35.916.00 - 35.721.00 « nn
= —' 5 ' = DD.UU.
2. Based on the preliminary estimates of 7 and s2, as well as the knowledge
that the RT for barium is 100 ppm,
nl =—7L
2 2
_ (1.6382)(65.00) _ - 77
~~ f\ ~ D • / / •
Ac 5.50^
(Equation 8)
3. As Indicated above, the appropriate number of sludge samples (nj) to be
collected from the lagoon 1s six. That number of samples (plus three
extra samples for protection against poor preliminary estimates of 7 and
s2) is collected from the lagoon by a single randomization process
(Figure 9-2). All samples consist of the greatest volume of sludge that
NINE - 14
Revision 0
Date September 1986
-------
WASTE OUTFALL
' i i n i i • i U"i
'•I I I I I X N
ll I
16
35
52
69
86
103
89
_U_LJJ
«" I I I
17
34
96
JlL
255
425
UPPER THIRD
OF LAGOON
\
LOWER TWO-THIRDS
OF LAGOON
WASTE LAGOON
1 J V
OVERFLY P.PE
«MAG,NA«YSAMPUNGGR,D
LEGEND
1-425 Units in Sampling Grid
r"T| Barium Ceneamration* (ppm)
' ' Attoeiatad with Mint SamplM of Sludgt
Figure 9-2.—Hypothetical sampling conditions in waste lagoon containing sludge contaminated with barium.
Barium concentrations associated with samples of sludge refer to levels measured in the elutriate of EP toxicity
tests conducted with the samples.
NINE - 15
Revision 0
Date September 1986
-------
can be practically collected. The three extra samples are suitably
processed and stored for possible later analysis.
4. The six samples of sludge (ni) designated for immediate analysis
generate the following concentrations of barium in the EP toxicity
test: 89, 90, 87, 96, 93, and 113 ppm. Although the value of 113 ppm
appears unusual as compared with the other data, there is no obvious
indication that the data are not normally distributed.
5. New values for 7 and s2 and associated values for the standard deviation
(s) and sy are calculated as:
+ 87 % * 93 * 113 . 94.67. (Equation 2a)
A
r2
s
n
n
E
1=1
- 54,
6
? n ?
Xf - (E X.) /n
1 1=1 1
M 1
224.00 - 53,770.67 _ on f7
(Equation 3a)
53.770.67 _ on C7
s = Js2 = 9.52, and (Equation 4)
sy = s/Jn = 9.52/J6 = 3.89. (Equation 5)
6. The new value for 7 (94.67) is less than the RT (100). In addition, 7 is
greater (only slightly) than s2 (90.67), and, as previously Indicated,
the raw data are not characterized by obvious abnormality. Consequently,
the study 1s continued, with the following calculations performed with
nontransformed data.
7. CI = x + t 2Qs- = 94.67 + (1.476)(3.89) (Equation 6)
= 94.67 + 5.74.
Because the upper limit of the CI (100.41) is greater than the applicable
RT (100), it is tentatively concluded that barium is present 1n the
sludge at a hazardous concentration.
NINE - 16
Revision
Date September 1986
-------
8. n 1s now reestlmated as:
= _^20f! (1.4762) (90.67) . 6
-------
BOX 2. STRATEGY FOR DETERMINING IF CHEMICAL CONTAMINANTS OF SOLID WASTES
ARE PRESENT AT HAZARDOUS LEVELS - STRATIFIED RANDOM SAMPLING
Step General Procedures
i
1. Obtain preliminary estimates of 7 and s2 for each chemical contaminant of
a solid waste that 1s of concern. The two above-identified statistics
are calculated by, respectively, Equations 2b and 3b (Table 9-1).
2. Estimate the appropriate number of samples (ni) to be collected from
the waste through use of Equation 8 (Table 9-1) and Table 9-2. Derive
individual values of nj for each chemical contaminant of concern.
The appropriate number of samples to be taken from the waste is the
greatest of the individual nj values.
3. Randomly collect at least nj (or r\2 ~ nl» n3 ~ n2» etc., as will be
indicated later in this box) samples from the waste (collection of a
few extra samples will provide protection against poor preliminary
estimates of 7 and s2). If s|< for each stratum (see Equation 3b) is
believed to be an accurate estimate, optimally allocate samples among
strata (I.e., allocate samples among strata so that the number of samples
collected from each stratum is directly proportional to s|< for that
stratum). Otherwise, proportionally allocate samples among strata
according to size of the strata. Maximize the physical size (weight or
volume) of all samples that are collected from the strata.
4. Analyze the nj (or i\2 - nj, n3 - r\2 etc.) samples for each chemical
contaminant of concern. Superficially (graphically) examine each set of
analytical data from each stratum for obvious departures from normality.
5. Calculate 7, 52, the standard deviation (s), and sy for each set of
analytical data by, respectively, Equations 2b, 3b, 4, and 5 (Table 9-1).
6. If 7 for a chemical contaminant is equal to or greater than the
applicable RT (Equation 7, Table 9-1) and is believed to be an accurate
estimator of u, the contaminant is considered • to be present in the
waste at a hazardous concentration, and the study is completed.
Otherwise, continue the study. In the case of a set of analytical data
that does not exhibit obvious abnormality and for which 7 is greater than
s2, perform the following calculations with nontransformed data.
Otherwise, consider transforming the data by the square root transfor-
mation (if 7 is about equal to s2) or the arcsine transformation (if 7 is
less than s2) and performing all subsequent calculations with transformed
data. Square root and arcsine transformations are defined by,
respectively, Equations 10 and 11 (Table 9-1).
7. Determine the CI for each chemical contaminant of concern by Equation 6
. (Table 9-1) and Tab.le 9-2. If the upper limit of the CI is less than the
applicable RT (Equations 6 and 7, Table 9-1), the chemical contaminant 1s
not considered to be present in the waste at a hazardous concentration,
and the study is completed. Otherwise, the opposite conclusion 1s
tentatively reached.
NINE - 18
Revision 0
Date September 1986
-------
8. If a tentative conclusion of hazard 1s reached, reestlmate the total
number of samples (n2) to be collected from the waste by use of
Equation 8 (Table 9-1) and Table 9-2. When deriving n?, employ the newly
calculated (not preliminary) values of X and s2. If additional
r\2 - nj samples of waste cannot reasonably be collected, the study 1s
completed, and a definitive conclusion of hazard Is reached. Otherwise,
collect extra r\2 - nj samples of waste.
9. Repeat the basic operations described 1n steps 3 through 8 until the
waste 1s judged to be nonhazardous or, 1f the opposite conclusion
continues to be reached, until increased sampling effort is Impractical.
Hypothetical Example
Step
1. The preliminary study of barium levels 1n the elutriate of four EP
toxldty tests, conducted with sludge collected from the lagoon several
years ago, generated values of 86 and 90 ppm for sludge obtained from
the upper third of the lagoon and values of 98 and 104 ppm for sludge
from the lower two-thirds of the lagoon. Those two sets of values are
not judged to be Indicative of nonrandom chemical heterogeneity
(stratification) within the lagoon. Therefore, preliminary estimates of
7 and s2 are calculated as:
* , I w^ . (1) (88.00) + (21(101.00) . 96-67_ and (Equat1on 2b)
+ (2) (18.00) . u 6?
2. Based on the preliminary estimates of X and s2, as well as the knowledge
that the RT for barium is 100 ppm,
n, = ^Of! = (L3682) (14.67) - 3.55. (Equation 8)
1 tf 3.33^
3. As Indicated above, the appropriate number of sludge samples (nj) to be
collected from the lagoon is four. However, for purposes of comparison
with simple random sampling (Box 1), six samples (plus three extra
samples for protection against poor preliminary estimates of 7 and s2)
are collected from the lagoon by a two-stage randomization process
(Figure 2). Because S|< for the upper (2.12 ppm) and lower (5.66 ppm)
strata are not believed to be very accurate estimates, the nine samples
to be collected from the lagoon are not optimally allocated between the
two strata (optimum allocation would require two and seven samples to be
NINE - 19
Revision
Date September 1986
-------
collected from the upper and lower strata, respectively). Alternatively,
proportional allocation is employed: three samples are collected from
the upper stratum (which represents one-third of the lagoon), and six
samples are taken from the lower stratum (two-thirds of the lagoon). All
samples consist of the greatest volume of sludge that can be practically
collected.
4. The nine samples of sludge generate the following concentrations of
barium in the EP toxicity test: upper stratum — 89, 90, and 87 ppm;
lower stratum — 96, 93, 113, 93, 90, and 91 ppm. Although the value of
113 ppm appears unusual as compared with the other data for the lower
stratum, there is no obvious Indication that the data are not normally
distributed.
5. New values for 7 and s^ and associated values for the standard deviation
(s) and sy are calculated as:
x = E W. x. = frH*8-67) + (2) (96-00) = 93.56, (Equation 2b)
k=1 k k 3 3
S2 = J w 2 = mi^m. + (2) (73.60) . 4g>84| (Euat1on 3b)
k=1 K K J 6
s = Js = 7.06, and (Equation 4)
SY = s/Jn = 7.06/J9 = 2.35. (Equation 5)
6. The new value for 7 (93.56) is less than the RT (100). In addition, 7 is
greater than s2 (49.84), and, as previously Indicated, the raw data are
not characterized by obvious abnormality. Consequently, the study 1s
continued, with the following calculations performed with nontransformed
data.
7. CI = x + t 2Qs- = 93.56 + (1.397)(2.35) (Equation 6)
= 93.56 + 3.28
The upper limit of the CI (96.84) is less than the applicable RT (100).
Therefore, it is concluded that barium is not present in the sludge at a
hazardous concentration.
NINE - 20
Revision
Date September 1986
-------
9.1.1.3.3 Systematic Random Sampling
Systematic random sampling (Box 3) 1s Implemented by general procedures
that are Identical to the procedures Identified for simple random sampling.
The hypothetical example for systematic random sampling (Box 3) demonstrates
the bias and Imprecision that are associated with that type of sampling when
unrecognized trends or cycles exist 1n the population.
9.1.1.4 Special Considerations
The preceding discussion has addressed the major Issues that are critical
to the development of a reliable sampling strategy for a solid waste. The
remaining discussion focuses on several "secondary" Issues that should be
considered when designing an appropriate sampling strategy. These secondary
Issues are applicable to all three of the basic sampling strategies that have
been Identified.
9.1.1.4.1 Composite Sampling
In composite sampling, a number of random samples are Initially collected
from a waste and combined into a single sample, which is then analyzed for the
chemical contaminants of concern. The major disadvantage of composite
sampling, as compared with noncomposlte sampling, is that Information
concerning the chemical contaminants is lost, I.e., each Initial set of
samples generates only a single estimate of the concentration of each
contaminant. Consequently, because the number of analytical measurements (n)
is small, s* and t.20 are large, thus decreasing the likelihood that a
contaminant will be judged to occur 1n the waste at a nonhazardous level
(refer to appropriate equations in Table 9-1 and to Table 9-2). A remedy to
that situation is to collect and analyze a relatively large number of
composite samples, thereby offsetting the savings 1n analytical costs that are
often associated with composite sampling, but achieving better representation
of the waste than would occur with noncomposlte sampling.
The appropriate number of composite samples to be collected from a solid
waste 1s estimated by use of Equation 8 (Table 9-1), as previously described
for the three basic sampling strategies. In comparison with noncomposlte
sampling, composite sampling may have the effect of minimizing between-sample
variation (the same phenomenon that occurs when the physical size of a sample
is maximized), thereby reducing somewhat the number of samples that must be
collected from the waste.
9.1.1.4.2 Subsampling
The variance (s2) associated with a chemical contaminant of a waste
consists of two components in that:
S2
s2 = s2 + -f • (Equation 12)
NINE - 21
Revision 0
Date September 1986
-------
BOX 3. STRATEGY FOR DETERMINING IF CHEMICAL CONTAMINANTS OF SOLID WASTES
ARE PRESENT AT HAZARDOUS LEVELS - SYSTEMATIC RANDOM SAMPLING
Step General Procedures
1. Follow general procedures presented for simple random sampling of solid
wastes (Box 1).
Step Hypothetical Example
1. The example presented 1n Box 1 1s applicable to systematic random
sampling, with the understanding that the nine sludge samples obtained
from the lagoon would be collected at equal Intervals along a transect
running from a randomly selected location on one bank of the lagoon to
the opposite bank. If that randomly selected transect were established
between Units 1 and 409 of the sampling grid (Figure 9-2) and sampling
were performed at Unit 1 and thereafter at three-unit intervals along the
transect (I.e., Unit 1, Unit 52, Unit 103, ... , and Unit 409), it is
apparent that only two samples would be collected 1n the upper third of
the lagoon, whereas seven samples would be obtained from the lower
two-thirds of the lagoon. If, as suggested by the barium concentrations
illustrated in Figure 9-2, the lower part of the lagoon is characterized
by greater and more variable barium contamination than the upper part of
the lagoon, systematic random sampling along the above-Identified
transect, by placing undue (disproportionate) emphasis on the lower part
of the lagoon, might be expected to result in an inaccurate
(overestimated) and Imprecise characterization of barium levels in the
whole lagoon, as compared with either simple random sampling or
stratified random sampling. Such inaccuracy and imprecision, which are
typical of systematic random sampling when unrecognized trends or cycles
occur 1n the population, would be magnified if, for example, the randomly
selected transect were established solely in the lower part of the
lagoon, e.g., between Units 239 and 255 of the sampling grid.
NINE - 22
Revision
Date September 1986
-------
where s2, = a component attributable to sampling (sample) variation, s2, = a
component attributable to analytical (subsample) variation, and m = number of
subsamples. In general, s<| should not be allowed to exceed one-ninth of s§.
If a preliminary study indicates that s£ exceeds that threshold, a sampling
strategy involving subsampling should be considered. In such a strategy, a
number of replicate measurements are randomly made on a relatively limited
number of randomly collected samples. Consequently, analytical effort is
allocated as a function of analytical variability. The efficiency of that
general strategy in meeting regulatory objectives has already been
demonstrated in the previous discussions of sampling effort.
The appropriate number of samples (n) to be collected from a solid waste
for which subsampling will be employed is again estimated by Equation 8
(Table 9-1). In the case of simple random sampling or systematic random
sampling with an equal number of subsamples analyzed per sample:
n
= E x./n, (Equation 13)
where Xj = sample mean (calculated from values for subsamples) and n = number
of samples. Also,
n « n •?
E x? - (E x.r/n
s2 = — p _ 1i=1 (Equation 14)
The optimum number of subsamples to be taken from each sample (m0pt.) is
estimated as:
s
m(opt.) = i; (Equation 15)
i
when cost factors are not considered. The value for sa is calculated from
available data as:
n m ~ 7
E E xf. - (E x..r
s . i=l .1=1 1J J
a
n (m - 1) ' (Equation 16)
and s , which can have a negative characteristic, is defined as:
ss
$2
s2 - -^ , (Equation 17)
NINE - 23
Revision
Date September 1986
-------
with $2 calculated as Indicated 1n Equation 14.
In the case of stratified random sampling with subsampUng, critical
formulas for estimating sample size (n) by Equation 8 (Table 9-1) include:
x = I W.x. ,
k=l K K
(Equation 2b)
where Xfc = stratum mean and W|< = fraction of population represented by Stratum
K (number of strata, k, ranges from 1 to r). In Equation 2b, 7|< for each
stratum 1s calculated as the average of all sample means 1n the stratum
(sample means are calculated from values for subsamples). In addition, s^ is
calculated by:
v»
= E W. s
k=l K
(Equation 3b)
with s2|< for each stratum calculated from all sample means in the stratum.
The optimum subsampUng effort when cost factors are not considered and all
replication is symmetrical is again estimated as:
m
(opt.)
' w1th
(Equation 15)
ss =
r n m
III
k=l 1=1 j=l
U-^
Is m '
Xk1j " (E Xki
rn (m - 1)
j) /m
' and (Equation 18)
(Equation 17)
with s2 derived as shown 1n Equation 3b.
9.1.1.5 Cost and Loss Functions
The cost of chemically characterizing a waste is dependent on the
specific strategy that 1s employed to sample the waste. For example, in the
case of simple random sampling without subsampUng, a reasonable cost function
might be:
C(n) = C0
Cln'
(Equation 19)
NINE - 24
Revision 0
Date September 1986
-------
where C(n) = cost of employing a sample size of n, CQ = an overhead cost
(which is Independent of the number of samples that are collected and
analyzed), and Cj = a sample-dependent cost. A consideration of C(n) mandates
an evaluation of L(n), which 1s the sample-size-dependent expected financial
loss related to the erroneous conclusion that a waste 1s hazardous. A simple
loss function 1s:
L(n) = n ' (Equation 20)
with a = a constant related to the cost of a waste management program 1f the
waste 1s judged to be hazardous, s2 = sample variance, and n = number of
samples. A primary objective of any sampling strategy 1s to minimize
C(n) + L(n). Differentiation of Equations 19 and 20 Indicates that the number
of samples (n) that minimize C(n) + L(n) 1s:
(Equation 21)
As is evident from Equation 21, a comparatively large number of samples (n)
1s justified 1f the value of a or s2 1s large, whereas a relatively small
number of samples 1s appropriate 1f the value of Cj 1s large. These general
conclusions are valid for any sampling strategy for a solid waste.
9.2 IMPLEMENTATION
This section discusses the Implementation of a sampling plan for the
collection of a "solid waste," as defined by Section 261.2 of the Resource
Conservation and Recovery Act (RCRA) regulations. Due to the uniqueness of
each sampling effort, the following discussion 1s 1n the general form of
guidance which, when applied to each sampling effort, should Improve and
document the quality of the sampling and the representativeness of samples.
The following subsections address elements of a sampling effort 1n a
logical order, from defining objectives through compositing samples prior to
analysis.
9.2.1 Definition Of Objectives
After verifying the need for sampling, those personnel directing the
sampling effort should define the program's objectives. The need for a
sampling effort should not be confused with the objective. When management, a
regulation, or a regulatory agency requires sampling, the need for sampling 1s
established but the objectives must be defined.
NINE - 25
Revision 0
Date September 1986
-------
The primary objective of any waste sampling effort 1s to obtain
Information that can be used to evaluate a waste. It 1s essential that the
specific Information needed and Its uses are defined 1n detail at this stage.
The Information needed 1s usually more complex than just a concentration of a
specified parameter; 1t may be further qualified (e.g., by sampling location
or sampling time.) The manner in which the Information Is to be used can also
have a substantial Impact on the design of a sampling plan. (Are the data to
be used 1n a qualitative or
-------
0.4-
•5 0.3-H
u
0)
0.2-
0.1-
Sample Mean = True Mean
Confidence
Interval
Lower
Limit
Upper
Limit
Regulatory
Threshold (RT)
25 SO 75 1OO
Concentration of Barium (ppm)
Distance of true value from regulatory threshold
requires less accuracy and precision.
Figure 9-3. Distribution of barium concentration removed from a
regulatory threshold.
0.4-J
Sample Mean = True Mean
•S 0.3-I
re
"o
>
8 0.2_
cr
0)
0.1-
{ Regulatory
I Threshold (RT)
85 90 95 100
Concentration of Barium (ppm)
Proximity of true value from regulatory threshold
requires more accuracy and precision.
Figure 9-4. Distribution of barium concentration near a regulatory
threshold.
NINE - 27
Revision p
Date September 1986
-------
The form in Figure 9-5 can be used to document primary and specific
objectives prior to development of a sampling plan. Once the objectives of a
sampling effort are developed, it is important to adhere to them to ensure
that the program maintains its direction.
9.2.2 Sampling Plan Considerations
The sampling plan is usually a written document that describes the
objectives and details the individual tasks of a sampling effort and how they
will be performed. (Under unusual circumstances, time may not allow for the
sampling plan to be documented in writing, e.g., sampling during an emergency
spill. When operating under these conditions, it is essential that the person
directing the sampling effort be aware of the various elements of a sampling
plan.) The more detailed the sampling plan, the less the opportunity for
oversight or misunderstanding during sampling, analysis, and data treatment.
To ensure that the sampling plan 1s designed properly, it is wise to have
all aspects of the effort represented. Those designing the sampling plan
should include the following personnel:
1. An end-user of the data, who will be using the data to attain
program objectives and thus would be best prepared to ensure that
the data objectives are understood and incorporated into the
sampling plan.
2. An experienced member of the field team who will actually collect
samples, who can offer hands-on insight Into potential problems and
solutions, and who, having acquired a comprehensive understanding of
the entire sampling effort during the design phase, will be better
prepared to implement the sampling plan.
3. An analytical chemist, because the analytical requirements for
sampling, preservation, and holding times will be factors around
which the sampling plan will be written. A sampling effort cannot
succeed 1f an improperly collected or preserved sample or an
Inadequate volume of sample is submitted to the laboratory for
chemical, physical, or biological testing. The appropriate
analytical chemist should be consulted on these matters.
4. An engineer should be involved if a complex manufacturing process 1s
being sampled. Representation of the appropriate engineering
discipline will allow for the optimization of sampling locations and
safety during sampling and should ensure that all waste-stream
variations are accounted for.
5. A statistician, who will review the sampling approach and verify
that the resulting data will be suitable for any required
statistical calculations or decisions.
6. A quality assurance representative, who will review the
applicability of standard operating procedures and determine the
number of blanks, duplicates, spike samples, and other steps
required to document the accuracy and precision of the resulting
data base.
NINE - 28
Revision 0
Date September 1986
-------
Sampling Site:
Address:
Description of Waste to be Sampled;
Primary Objective:
Specific Sampling Objectives:
Specific Analysis Objectives;
Specific Data Objectives:
Figure 9-5. Form for Documenting Primary and Specific Objectives
NINE - 29
Revision 0
Date September 1986
-------
At least one person should be familiar with the site to be sampled. If
not, then a presampUng site visit should be arranged to acquire site-specific
Information. If no one 1s familiar with the site and a presampUng site visit
cannot be arranged, then the sampling plan must be written so that 1t can
address contingencies that may occur.
Even 1n those cases 1n which a detailed sampling plan 1s authored and a
comprehensive knowledge of the site exists, 1t Is unusual for a sampling plan
to be Implemented exactly as written. Waste-stream changes, Inappropriate
weather, sampling equipment failure, and problems 1n gaining access to the
waste are some reasons why a sampling plan must be altered. Thus 1t 1s always
necessary to have at least one experienced sampler as a member of a sampling
team.
The sampling plan should address the considerations discussed below.
9.2.2.1 Statistics
A discussion of waste sampling often leads to a discussion of statistics.
The goals of waste sampling and statistics are Identical, I.e., to make
Inferences about a parent population based upon the Information contained In a
sample.
Thus 1t 1s not surprising that waste sampling relies heavily upon the
highly developed science of statistics and that a sampling/analytical effort
usually contains the same elements as does a statistical experiment.
Analogously, the Harris pollster collects opinions from randomly chosen
people, whereas environmental scientists collect waste at randomly chosen
locations or times. The pollster analyzes the Information Into a useable data
base; laboratories analyze waste samples and generate data. Then the unbiased
data base 1s used to draw Inferences about the entire population, which for
the Harris pollster may be the voting population of a large city, whereas for
the environmental scientist the population may mean the entire contents of a
landfill.
During the Implementation of a waste sampling plan or a statistical
experiment, an effort 1s made to minimize the possibility of drawing Incorrect
Inferences by obtaining samples that are representative of a population. In
fact, the term "representative sample" Is commonly used to denote a sample
that (1) has the properties and chemical composition of the population from
which 1t was collected, and (2) has them 1n the same average proportions as
are found 1n the population.
In regard to waste sampling, the term "representative sample" can be
misleading unless one 1s dealing with a homogeneous waste from which one
sample can represent the whole population. In most cases, It would be best to
consider a "representative data base" generated by the collection and analysis
of more than one sample that defines the average properties or composition of
the waste. A "representative data base" is a more realistic term because the
evaluation of most wastes requires numerous samples to determine the average
properties or concentrations of parameters in a waste. (The additional
samples needed to generate a representative data base can also be used to
determine the variability of these properties or concentrations throughout the
waste population.)
NINE - 30
Revision 0
Date September 1986
-------
Statisticians have developed a number of strategies to obtain samples
that are unbiased and collectively representative of a population. A detailed
discussion of these strategies is presented in Section 9.1 of this chapter.
The following discussion of statistical considerations 1s a less technical
summary of these strategies. It was written to complement Section 9.1 and
will be most useful after Section 9.1 is read and studied.
Section 9.1 describes three basic sampling strategies: simple random,
stratified random, and systematic random sampling. It should be noted that
the word random has more than one meaning. When used 1n statistical
discussions, it does not mean haphazard; it means that every part of a waste
has a theoretically equal chance of being sampled. Random sampling, which
entails detailed planning and painstaking implementation, is distinctly
different from haphazard sampling, which may introduce bias into the
collection of samples and the resulting data.
Systematic random sampling and authoritative sampling strategies require
a substantial knowledge of the waste to ensure that: (1) a cycle or trend 1n
waste composition does not coincide with the sampling locations; or (2) 1n the
case of authoritative sampling, all or most of the assumptions regarding waste
composition or generation are true. Because the variabilities of waste
composition and the waste generation process are often unknown, systematic
random and authoritative sampling strategies are usually not applicable to
waste evaluation.
Therefore, for waste sampling, the usual options are simple or stratified
random sampling. Of these two strategies, simple random sampling 1s the
option of choice unless: (1) there are known distinct strata (divisions) 1n
the waste over time or 1n space; (2) one wants to prove or disprove that there
are distinct time and/or space strata 1n the waste of interest; or (3) one 1s
collecting a minimum number of samples and desires to minimize the size of a
hot spot (area of high concentration) that could go unsampled. If any of
these three conditions exists, it may be determined that stratified random
sampling would be the optimum strategy. To explain how these strategies can
be employed, a few examples follow:
Example 1: Simple Random Sampling of Tanks
A batch manufacturing process had been generating a liquid waste over a
period of years and storing it in a large open-top tank. As this tank
.approached capacity, some of the waste was allowed to overflow to a smaller
enclosed tank. This smaller tank allowed for limited access through an
inspection port on its top.
Because the on-site tank storage was approaching capacity, 1t was
determined that the waste would have to be disposed of off-site.
The operators of the facility had determined that the waste was
a nonhazardous solid waste when the RCRA regulations were first promulgated.
However, upon recent passage of more stringent state regulations and concerns
of potential liability, the operators determined that they should perform a
more comprehensive analysis of the waste.
NINE - 31
Revision 0
Date September 1986
-------
Because the waste was generated 1n a batch mode over a period of years,
the operators were concerned that the waste composition might have varied
between batches and that stratification might have occurred In the tank at
unknown and random depths. Based on their knowledge, the operators knew that
a grab sample would not suffice and that a sampling program would have to be
designed to address the heterogeneity of the waste.
Because the operators Intended to dispose of the entire contents of the
tank and lacked any specific information regarding stratification and
variability of the waste, it was decided that a simple random strategy would
be employed. (If the operators had treated portions of the waste differently
or had been aware of distinct strata, then stratified random sampling might
have been more appropriate.)
The large, unenclosed tank had a diameter of 50 ft, a height of 20 ft,
and an approximate volume of 295,000 gal allowed. It was encircled and
traversed by catwalks (refer to Figure 9-6), which allowed access to the
entire waste surface. The smaller tank had a diameter of 10 ft, a height of
10 ft, and an approximate volume of 6,000 gal; an Inspection port located on
the top allowed limited access. It was determined that the different
construction of the two tanks would require different simple random sampling
approaches.
In the case of the large tank, 1t was decided that vertical composite
samples would be collected because the operators were interested 1n the
average composition and variability of the waste and not 1n determining if
different vertical strata existed. It was decided to select points randomly
along the circumference (157 ft) and along the radius (25 ft). These numbers,
which would constitute the coordinates of the sampling locations, were chosen
from a random-number table by indiscriminately choosing a page and then a
column on that page. The circumference coordinates were then chosen by
proceeding down the column and listing the first 15 numbers that are greater
than or equal to 0, but less than or equal to 157. The radius coordinates
were chosen by continuing down the column and listing the first 15 numbers
that are greater than or equal to 0, but less than or equal to 25. These
numbers were paired to form the coordinates that determined the location of
the 15 randomly chosen sampling points. These coordinates were recorded in
the field notebook (refer to Table 9-3). Because no precision data on waste
composition existed prior to sampling, the number of samples (15) was chosen
as a conservative figure to more than allow for a sound statistical decision.
The actual samples were collected by employing a sampling device, which
was constructed on site from available materials, and a weighted bottle. This
device, which was used to access more remote areas of the tank, consisted of a
weighted bottle, a rope marked off at 1-ft increments, and a discarded spool
that originally contained electrical wire (refer to Figure 9-7).
Samples were collected by a three-person team. The person controlling
the weighted bottle walked to the first circumference coordinate (149 ft),
while the two persons holding the ropes attached to the spool walked along
opposing catwalks toward the center of the tank. The person controlling the
NINE - 32
Revision 0
Date September 1986
-------
Catwalks
CO
CO
O 73
QJ rt>
r+ <
n ->•
CO O
(T> 2
a
Inspection Port
2.000 Gallon
Overflow Tank
3,000 Gallon
Tank Tmck
295.000 Gallon
Storaqe Tank
VO
00
Figure 9-6. Bird's eye view of waste tank, overflow tank, tank truck and connecting plumhinq.
-------
TABLE 9-3. RANDOM COORDINATES FOR 295,000-GAL TANK
Sampling point Circumference Radius
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
149
86
94
99
23
58
52
104
23
51
77
12
151
83
99
4
22
13
0
10
2
22
16
25
4
14
5
15
23
18
NINE - 34
Revision
Date September 1986
-------
Trip Cord
Rope to radial catwalk
Rope to circumference catwalk
Rope to opposite radial catwalk
Figure 9-7. Device used to collect sample from the open tank.
NINE - 35
Revision 0
Date September 1986
-------
weighted bottle measured off the radius coordinate (4 ft). The spool was then
centered In the quadrant, the weighted bottle was lowered to the surface, and
a sample was collected from the first 2 ft of waste. This sample was then
transferred Into a large, labeled sample container, which was used for
compositing. This same process was repeated nine more times at the same
location at different 2-ft depth Intervals, resulting in the collection of a
total of 10 component depth samples that were compiled in the field into one
sample for that sampling point. This process was repeated at the remaining 14
sampling points, resulting in the collection of 15 vertical composite samples.
These vertical composite samples were taken to address any vertical
stratification that may have occurred.
The samples were properly preserved and stored, chain-of-custody
procedures were completed, and the samples were submitted to the laboratory.
A cost/benefit decision was made to composite aliquots of the samples into
five composite samples that were submitted for analysis. (Following analysis,
Equation 8 of Section 9.1 of this chapter was employed to determine if enough
samples were analyzed to make a statistically sound decision. If the number
of samples analyzed was not sufficient, then the samples would be recomposited
to a lesser degree or analyzed individually.)
Because there was no information to prove that the waste in the smaller
tank was the same as that in the larger tank, the operators decided that the
smaller tank must also be sampled. The different construction of the smaller,
enclosed tank mandated that a different sampling plan be designed. The only
access to the tank was through a small Inspection port on the top of the tank.
This port would allow sampling only of a small portion of the tank contents;
thus, to make a decision on the entire contents of the tank, one would have to
assume that the waste In the vicinity of the inspection port was
representative of the remainder of the tank contents. The operators were not
willing to make this assumption because they determined that the liability of
an incorrect decision overrode the convenience of facilitating the sampling
effort.
To randomly sample the entire contents of the tank, a different plan was
designed. This plan exploited the relatively small volume (approximately
6,000 gal) of the tank. A decision was made to rent two tank trucks and to
sample the waste randomly over time as It drained from the tank into the tank
trucks.
It was calculated that at a rate of 20 gal/min, it would take 300 m1n to
drain the tank. From the random-number tables, 15 numbers that were greater
than or equal to 0, but less than or equal to 300, were chosen In a manner
similar to that employed for the larger tank. These numbers were recorded in
the field notebook (refer to Table 9-4) at the time that they were encountered
1n the random-number table and ,were then assigned sampling point numbers
according to their chronological order.
The 15 samples were collected at the previously chosen random times as
the waste exited from a drainage, hose Into the tank trucks. These samples
were collected in separate labeled containers, properly preserved and stored;
cha1n-of-custody procedures were employed for transferral of the samples to
the laboratory.
NINE - 36
Revision 0
Date September 1986
-------
TABLE 9-4. RANDOM TIMES FOR 6,000-GAL TANK
Sampling point Time (m1n)
11 153
10 122
8 85
6 55
5 46
15 294
12 195
1 5
13 213
9 99
2 29
4 41
7 74
3 31
14 219
NINE - 37
Revision
Date September 1986
-------
The above example employed simple random sampling to determine the
average composition and variance of the waste contained 1n the two tanks. The
contents of the large tank were sampled randomly 1n space, whereas the
contents of the smaller tank were sampled randomly over time.
The following example will Involve the use of stratified random sampling,
which Is used when: (1) distinct strata are known to exist or (2) 1t 1s not
known whether different strata exist, but an objective of the sampling effort
1s to discover the existence or nonexistence of strata.
A variation of this second reason for employing stratified random
sampling 1s when cost considerations limit the number of samples that can be
collected (e.g., when the budget allows for the collection of only six samples
1n a 40-acre lagoon). In this situation, where little 1s known about the
composition of the waste, a concern exists that an area of the lagoon may be
highly contaminated and yet may not be sampled. The smaller the number of
samples, the greater the probability that an area of high contamination (a
distinct stratum) could be missed, and the greater the probability that the
sampling accuracy will suffer. Under such circumstances, a sampling plan may
employ stratified random sampling to minimize the size of a highly
contaminated area that could go unsampled.
For example, consider the situation where the budget allows only for the
collection of six samples 1n a 40-acre lagoon. If simple random sampling 1s
employed with such a small number of samples, there Is a certain probability
that large areas of the lagoon may go unsampled. One approach to minimizing
the size of areas that may go unsampled Is to divide the lagoon Into three
strata of equal size and randomly sample each stratum separately. This
approach decreases the size of an area that can go unsampled to something less
than one-third of the total lagoon area.
The following example details more traditional applications of stratified
random sampling.
Example 2: Stratified Random Sampling of Effluents and Lagoons
A pigment manufacturing process has been generating wastes over a number
of years. The pigment 1s generated In large batches that Involve a 24-hr
cycle. During the first 16 hr of the cycle, an aqueous sludge stream 1s
discharged. This waste contains a high percentage of large-sized black
particulate matter. The waste generated during the remaining 8 hr of the
manufacturing cycle 1s an aqueous-based white sludge that consists of much
smaller-sized particles than those found 1n the sludge generated 1n the first
16 hr of the batch process. This waste has been disposed of over the years
Into a 40-acre settling lagoon, allowing the particulate matter to settle out
of solution while the water phase, drains to an NPDES outfall at the opposite
end of the lagoon. The smaller white pigment particles released in the last 8
hr of the batch process settle more slowly than the much larger black
particles generated in the previous 16 hr. This settling pattern 1s quite
apparent from the distinct colors of the wastes. The sludge 1n the quadrant
closest to the waste influent pipe is black; the next quadrant is a light gray
color, resulting from settling of both waste streams. The last two quadrants
contain a pure white sludge, resulting from the settling of the small pigment
particles.
NINE - 38
Revision 0
Date September 1986
-------
Eventually, the facility operators decided that the settled particulate
matter had to be removed to keep the settling lagoon functioning. In the
past, this residual lagoon waste was found to be a hazardous waste due to Its
Teachable barium content. Further studies determined that the source of the
barium was a certain raw material that was released during the first 16 hr of
batch process.
To minimize present disposal costs, the operators wanted to determine 1f
the white sludge 1n the last two quadrants and the light gray waste were
nonhazardous. Also, the operators had recently changed raw materials, with
the Intention of removing the source of barium in an attempt to minimize
future disposal costs. Thus, the operators were interested in determining
whether the currently generated waste was hazardous. If the altered waste
stream was not hazardous, future lagoon sludge could be disposed of more
economically as a solid waste. If the waste generated during the first 16 hr
of the process remained hazardous but the waste generated during the following
8 hr was nonhazardous, the operators were willing to shift this latter waste
to a second lagoon reserved for nonhazardous wastes. By sequestering the
waste streams In this manner, the operators Intended to decrease the amount of
hazardous waste by precluding generation of additional amounts of hazardous
waste under the "mixture rule."
To decide how the lagoon sludge should be handled, the operators arranged
to have the lagoon sludge sampled. The objectives of sampling the lagoon
sludge were to determine the average concentration and variance of Teachable
barium for the sludge in the entire lagoon and for each of the different
sludges.
The dimensions of the 40-acre square lagoon were calculated to be
1,320 ft on a side, with the black and the gray sludge each covering a
quadrant measuring 1,320 ft by 330 ft, and the white sludge covering the
remaining area of the lagoon, which measured 1,320 ft by 660 ft (refer to
Figure 9-8). The sludge had settled to a uniform thickness throughout the
lagoon and was covered with 2 ft of water.
Because the Teachable barium was assumed to be associated with the black
sludge, which was concentrated in the first quadrant, a stratified random
sampling approach was chosen. (Because of the obvious strata in the lagoon
sludge, the stratified sampling strategy was expected to give a more precise
estimate of the Teachable barium, 1n addition to giving Information specific
to each stratum.)
When the actual sampling was being planned, it was decided that the
hazards presented by the lagoon waste were minimal, and, that 1f proper
precautions were employed, a stable and unsinkable boat could be used to
collect samples. The samples were collected with a core sampler at random
locations throughout each stratum. Because the cost of collecting samples was
reasonable and no historical data were available to help determine the optimum
number of samples, the operators decided to collect a total of 10 samples from
each of the smaller strata and a total of 20 samples from the larger strata.
They had confidence that this number of samples would allow them to detect a
small significant difference between the mean concentration of Teachable
barium and the applicable regulatory threshold.
NINE - 39
Revision 0
Date September 1986
-------
U)
U)
o
CO
O)
O)
o
Overflow Pipe
Figure 9-8. Schematic of the 40-acre settling lagoon displaying strata
generated by a waste stream.
NINE - 40
Revision Q
Date September 1986
-------
The locations of the random sampling points were determined by selecting
length and width coordinates from a random-number table. This was done by
Indiscriminately choosing a page from the random-number tables and then a
column on that page. The width coordinates of the two smaller quadrants were
then chosen by proceeding down the column and listing the first 20 numbers
that were greater than or equal to 0, but less than or equal to 330. The
width coordinate for the third and largest stratum was chosen by proceeding
down the column and selecting the first 20 numbers that were greater than or
equal to 0, but less than or equal to 660. Because the lengths of the three
quadrants were all 1,320 ft, the length coordinates were chosen by listing the
first 40 numbers that were greater than or equal to 0 but less than or equal
to 1,320. These coordinates were recorded 1n the field notebook (refer to
nable 9-5).
The samples were collected by a four-person team. Two people remained
onshore while two maneuvered the boat and collected the samples. The first
sample In the first quadrant was collected by launching the boat at a distance
of 41 ft from the corner, which was designated the origin, 0 ft. The boat
proceeded out Into the lagoon perpendicular to the long side of the quadrant.
The person onshore released 134 ft of a measured rope, which allowed the boat
to stop at the first sampling point (41, 134). The sample was then collected
with a core sampler and transferred to a sample container. This process was
repeated for all sampling points 1n the three strata. The samples were
properly preserved and stored, and the cha1n-of-custody records documented the
transfer of samples to the laboratory.
AHquots of the samples were composited Into five composite samples for
each stratum. The mean and variance of each stratum were calculated by
Equations 2(a) and 3(a), respectively. The mean and variance for the total
lagoon were calculated by using Equations 2(b) and 3(b), respectively.
Equation 6 was used to calculate a confidence Interval for the Teachable
barium concentration, and the upper limit of this Interval was compared with
the regulatory threshold. (See Table 9-1, Section 9.1 of this chapter, for
equations.)
As previously mentioned, the operators had recently changed their raw
materials and were also Interested in discovering if the currently generated
waste was nonhazardous or 1f portions of this waste stream were nonhazardous.
As described above, the waste effluent for the first 16 hr of the day was
different from that discharged during the last 8 hr. However, because the
same large plumbing system was used for both waste streams, there were two 2-
hr periods during which the discharged waste was a mixture of the two
different wastes.
With the above objectives 1n mind, the operators decided to employ
stratified random sampling with four strata occurring over time, as opposed to
the strata in space that were employed for sampling the lagoon. The four time
strata were from 6:00 to 8:00 hr, from 8:00 to 20:00 hr, from 20:00 to 22:00
hr, and from 22:00 to 6:00 hr the following day. The two 2-hr strata were
those time periods during which the. waste was a mixture of the two different
waste streams. The 12-hr stratum was the time period during which the large-
sized particulate black waste was being discharged. The smaller partlculate
white waste was being discharged during the 8-hr stratum.
NINE - 41
Revision 0
Date September 1986
-------
TABLE 9-5. RANDOM COORDINATES FOR EACH STRATUM
IN THE 40-ACRE SETTLING LAGOON
Stratum #1
(Black)
••
Stratum #2
(Gray) - .
Stratum #3
(White)
•
•. -
••
Sampling
Point
1
2
3
4
" 5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15-
16
17
18
19
20
Length
(ft)
41
271
968
129
472
1,198
700
286
940
151
1,173
277
438
780
525
50
26
1,207
1,231
840
54
909
1,163
1,251
1
1,126
717
1,155
668
66
462
213
1,220
1,038
508
1,293
30
114
1,229
392
Width
(ft)
134
51
32
228
137
56
261
8
26
121
109
2
302
5
135
37
127
149
325
32
374
434
390
449
609
140
235
148
433
642
455
305
541
644
376
270
38
52
570
613
NINE - 42
Revision 0
Date September 1986
-------
The flow rate was constant throughout the 24-hr period, and there were no
precision data available for the waste. Therefore, 1t was decided that the
number of samples collected in the 8- and 12-hr strata would be proportional
to time. Because the 2-hr periods were times during which the composition of
the waste was changing, it was decided to collect more samples to get a more
precise estimate of the average composition of the waste during these time
strata. Thus a total of 28 samples was collected.
The samples were collected at randomly chosen times within each time
stratum. The random sampling times were chosen by employing a random-number
table. After indiscriminately selecting a starting point, the first four
numbers greater than or equal to 0, but less than or equal to 120 were
selected for the 120-min strata from 6:00 to 8:00 hr. These minutes were then
added to the starting time to determine when the four samples would be
collected. In similar fashion, the remaining 24 sampling times were chosen.
The random-number data were recorded in a laboratory notebook (refer to Table
9-6).
The samples were collected from the waste Influent pipe with a wide-mouth
bottle at the randomly chosen sampling times. The samples were properly
preserved and stored and shipped to the laboratory, along with cha1n-of-
custody records. The samples were subjected to analysis, and the data were
evaluated 1n a manner similar to that employed for the samples of sludge
collected in the different strata of the lagoon.
9.2.2.2 Waste
The sampling plan must address a number of factors 1n addition to
statistical considerations. Obviously, one of the most Important factors 1s
the waste Itself and its properties. The following waste properties are
examples of what must be considered when designing a sampling plan:
1. Physical state; The physical state of the waste will affect most
aspects of a sampling effort. The sampling device will vary
according to whether the sample 1s liquid, gas, solid, or
multiphasic. It will also vary according to whether the liquid Is
viscous or free-flowing, or whether the solid Is hard or soft,
powdery, monolithic, or clay-like.
Wide-mouth sample containers will be needed for most solid samples
and for sludges or liquids with substantial amounts of suspended
matter. Narrow-mouth containers can be used for other wastes, and
bottles with air-tight closures will be needed for gas samples or
gases adsorbed on solids or dissolved in liquids.
The physical state will also affect how sampling devices are
deployed. A different plan will be developed for sampling a soll-
like waste that can easily support the weight of a sampling team and
Its equipment than for a lagoon filled with a viscous sludge or a
liquid waste.
NINE - 43
Revision 0
Date September 1986
-------
TABLE 9-6. RANDOM TIMES FOR THE WASTE EFFLUENT
Stratum #1
(6:00 to 8:00
hours)
Stratum #2
(8:00 to 20:00
hours)
Stratum #3
(20:00 to 22:00
hours)
Stratum #4
(22:00 to 6:00
hours
Sampl 1 ng
Point
1
2
3
4
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
1
2
3
4
5
6
7
8
Random
Minute
28
62
99
112
11
107
156
173
296
313
398
497
555
600
637
706
13
52
88
108
48
113
153
189
227
290
314
474
Time
6:28
7:02
7:39
7:52
8:11
9:47
10:36
10:53
12:56
13:13
14:38
16:17
17:15
18:00
18:37
19:46
20:13
20:52
21:28
21:48
22:48
23:53
24:33
1:09
1:47
2:49
3:14
5:44
NINE - 44
Revision 0
Date September 1986
-------
The sampling strategy will have to vary 1f the physical state of the
waste allows for stratification (e.g., liquid wastes that vary 1n
density or viscosity or have a suspended solid phase),
homogenlzatlon or random heterogeneity.
2. Volume; The volume of the waste, which has to be represented by the
samples collected, will have an effect upon the choice of sampling
equipment and strategies. Sampling a 40-acre lagoon requires a
different approach from sampling a 4-sq-ft container. Although a
3-ft depth can be sampled with a Collwasa or a drum thief, a
weighted bottle may be required to sample a 50-ft depth.
3. Hazardous properties; Safety and health precautions and methods of
sampling and shipping will vary dramatically with the toxlcity,
ignitabillty, corrosivlty, and reactivity of the waste.
4. Composition; The chosen sampling strategy will reflect the
homogeneity, random heterogeneity, or stratification of the waste in
time or over space.
9.2.2.3 Site
Site-specific factors must be considered when designing a sampling plan.
A thorough examination of these factors will minimize oversights that can
affect the success of sampling and prevent attainment of the program
objectives. At least one person involved in the design and implementation of
the sampling plan should be familiar with the site, or a presampUng site
visit should be arranged. If nobody 1s familiar with the site and a visit
cannot be arranged, the sampling plan must be written to account for the
possible contingencies. Examples of site-specific factors that should be
considered follow:
1. Accessibility; The accessibility of waste can vary substantially.
Some wastes are accessed by the simple turning of a valve; others
may require that an entire tank be emptied or that heavy equipment
be employed. The accessibility of a waste at the chosen
sampling location must be determined prior to design of a sampling
plan.
2. Waste generation and handling; The waste generation and handling
process must beunderstood to ensure that collected samples are
representative of the waste. Factors which must be known and
accounted for in the sampling plan include: if the waste is
generated 1n batches; if there 1s a change in the raw materials used
in a manufacturing process; if waste composition can vary
substantially as a function of process temperatures or pressures;
and if storage time after generation may vary.
3. Transitory events; Start-up, shut-down, slow-down, and maintenance
transients can result in the generation of a waste that is not
representative of the normal waste stream. If a sample was
unknowingly collected at one of these intervals, incorrect
conclusions could be drawn.
NINE - 45
Revision 0
Date September 1986
-------
4. Climate; The sampling plan should specify any clothing needed for
personnel to accommodate any extreme heat or cold that may be
encountered. Dehydration and extensive exposure to sun, Insects, or
poisonous snakes must be considered.
5. Hazards; Each site can have hazards — both expected and
unexpected. For example, a general understanding of a process may
lead a sampling team to be prepared for dealing with toxic or
reactive material, but not for dealing with an electrical hazard or
the potential for suffocation 1n a confined space. A thorough
sampling plan will Include a health and safety plan that will
counsel team members to be alert to potential hazards.
9.2.2.4 Equipment
The choice of sampling equipment and sample containers will depend upon
the previously described waste and site considerations. For the following
reasons, the analytical chemist will play an Important role 1n the selection
of sampling equipment:
1. The analytical chemist 1s aware of the potential Interactions
between sampling equipment or container material with analytes of
Interest. As a result, he/she can suggest a material that minimizes
losses by adsorption, volatilization, or contamination caused by
leaching from containers or sampling devices.
2. The analytical chemist can specify cleaning procedures for sampling
devices and containers that minimize sample contamination and cross
contamination between consecutive samples.
3. The analytical chemist's awareness of analyte-spedflc properties 1s
useful In selecting the optimum equipment (e.g., choice of sampling
devices that minimize agitation for those samples that will be
subjected to analysis for volatile compounds).
The final choice of containers and sampling devices will be made jointly
by the analytical chemist and the group designing the sampling plan. The
factors that will be considered when choosing a sampling device are:
1. Negative contamination; The potential for the measured analyte
concentration to be artificially low because of losses from
volatilization or adsorption.
2. Pos1tlve contamlnatlon; The potential for the measured analyte to
be artlficallyhighbecause of leaching or the Introduction of
foreign matter Into the sample by particle fallout or gaseous air
contaminants.
3. Cross contamination; A type of positive contamination caused by the
Introduction of part of one sample Into a second sample during
sampling, shipping, or storage.
NINE - 46
Revision 0
Date September 1986
-------
4. Required sample volume for physical and/or chemical analysis.
5. "Ease of use" of the sampling device and containers under the
conditions that will be encountered on-s1te. This Includes the ease
of shipping to and from the site, ease of deployment, and ease of
cleaning.
6. The degree of hazard associated with the deployment of one sampling
device versus another.
7. Cost of the sampling device and of the labor for its deployment.
This section describes examples of sampling equipment and suggests
potential uses for this equipment. Some of these devices are commercially
available, but others will have to be fabricated by the user. The Information
in this section is general 1n nature and therefore limited.
Because each sampling situation 1s unique, the cited equipment and
applications may have to be modified to ensure that a representative sample is
collected and its physical and chemical Integrity are maintained. It 1s the
responsibility of those persons conducting sampling programs to make the
appropriate modifications.
Table 9-7 contains examples of sampling equipment and potential
applications. It should be noted that these suggested sampling devices may
not be applicable to a user's situation due to waste- or site-specific
factors. For example, if a waste 1s highly viscous or if a solid is clay-
like, these properties may preclude the use of certain sampling devices. The
size and depth of a lagoon or tank, or difficulties associated with accessing
the waste, may also preclude use of a given device or require modification of
its deployment.
The most important factors to consider when choosing containers for
hazardous waste samples are compatibility with the waste, cost, resistance to
breakage, and volume. Containers must not distort, rupture, or leak as a
result of chemical reactions with constituents of waste samples. Thus, 1t is
important to have some Idea of the properties and composition of the waste.
The containers must have adequate wall thickness to withstand handling during
sample collection and transport to the laboratory. Containers with wide
mouths are often desirable to facilitate transfer of samples from samplers to
containers. Also, the containers must be large enough to contain the optimum
sample volume.
Containers for collecting and storing hazardous waste samples are usually
made of plastic or glass. Plastics that are commonly used to make the
containers include high-density or linear polyethylene (LPE), conventional
polyethylene, polypropylene, polycarbonate, Teflon FEP (fluorinated ethylene
propylene), polyvinyl chloride (PVC), or polymethylpentene. Teflon FEP is
almost universally usable due to its chemical Inertness and resistance to
breakage. However, Its high cost severely limits its use. LPE, on the other
hand, usually offers the best combination of chemical resistance and low cost
when samples are to be analyzed for Inorganic parameters.
NINE - 47
Revision 0
Date September 1986
-------
TABLE 9-7. EXAMPLES OF SAMPLING EQUIPMENT FOR PARTICULAR WASTE TYPES
Waste location or container
Waste type Drum
Storage Ponds,
Sacks Open-bed Closed- tanks Waste lagoons, Conveyor
and bags truck bed truck or bins piles & pits belt Pipe
Free-flowing
liquids and
slurries
Sludges
Moist
powders
or granules
Dry powders
or granules
Sand or
packed
powders
and granules
Large-
grained
solids
Coliwasa
Trier
Trier
Thief
Auger
Large
Trier
N/A
N/A
Trier
Thief
Auger
Large
Trier
N/A
Trier
Trier
Thief
Auger
Large
Trier
Coliwasa
Trier
Trier
Thief
Auger
Large
Trier
Weighted N/A Dipper
bottle
Trier a a
Trier Trier Trier
a Thief Thief
Thief Thief a
Large Large Large
Trier Trier Trier
N/A Dipper
Shovel Dipper
Shovel Dipper
Dipper Dipper
Trier Dipper
This type of sampling situation can present significant logistical sanpling problems, and sanpllng
equipment must be specifically selected or designed based on site and waste conditions. No general
statement about appropriate sanpling equipment can be made.
NINE - 48
Revision 0
Date September 1986
-------
Glass containers are relatively Inert to most chemicals and can be used
to collect and store almost all hazardous waste samples, exept those that
contain strong alkali and hydrofluoric add. Glass soda bottles are suggested
due to their low cost and ready availability. BoroslHcate glass containers,
such as Pyrex and Corex, are more Inert and more resistant to breakage than
soda glass, but are expensive and not always readily available. Glass
containers are generally more fragile and much heavier than plastic
containers. Glass or FEP containers must be used for waste samples that will
be analyzed for organic compounds.
The containers must have tight, screw-type Hds. Plastic bottles are
usually provided with screw caps made of the same material as the bottles.
Buttress threads are recommended. Cap liners are not usually required for
plastic containers. Teflon cap liners should be used with glass containers
supplied with rigid plastic screw caps. (These caps are usually provided with
waxed paper liners.) Teflon liners may be purchased from plastic specialty
supply houses (e.g., Scientific Specialties Service, Inc., P.O. Box 352,
RandalIstown, Maryland 21133). Other liners that may be suitable are
polyethylene, polypropylene, and neoprene plastics.
If the samples are to be submitted for analysis of volatile compounds,
the samples must be sealed 1n air-tight containers.
Prior to sampling, a detailed equipment 11st should be compiled. This
equipment list should be comprehensive and leave nothing to memory. The
categories of materials that should be considered are:
1. Personnel equipment, which will include boots, rain gear, disposable
coveralls, face masks and cartridges, gloves, etc.
2. Safety equipment, such as portable eyewash stations and a first-aid
kit.
3. Field test equipment, such as pH meters and Draeger tube samplers.
4. An ample supply of containers to address the fact that once 1n the
field, the sampling team may want to collect 50% more samples than
originally planned or to collect a liquid sample, although the
sampling plan had specified solids only.
5. Additional sampling equipment for use 1f a problem arises, e.g., a
tool kit.
6. Shipping and office supplies, such as tape, labels, shipping forms,
chain-of-custody forms and seals, field notebooks, random-number
tables, scissors, pens, etc.
Composite Liquid Waste Sampler (Coliwasa)
The Coliwasa is a device employed to sample free-flowing liquids and
slurries contained 1n drums, shallow tanks, pits, and similar containers. It
is especially useful for sampling wastes that consist of several immiscible
liquid phases.
NINE - 49
Revision 0
Date September 1986
-------
The Collwasa consists of a glass, plastic, or metal tube equipped with an
end closure that can be opened and closed while the tube 1s submerged 1n the
material to be sampled (refer to Figure 9-9).
Weighted Bottle
This sampler consists of a glass or plastic bottle, sinker, stopper, and
a line that 1s used to lower, raise, and open the bottle. The weighted bottle
samples liquids and free-flowing slurries. A weighted bottle with line Is
built to the specifications 1n ASTM Methods D270 and E300. Figure 9-10 shows
the configuration of a weighted-bottle sampler.
Dipper
The dipper consists of a glass or plastic beaker clamped to the end of a
two- or three-piece telescoping aluminum or fiberglass.pole that serves as the
handle. A dipper samples IJquids and free-flowing slurries. Dippers are not
available commercially and must be fabricated (Figure 9-11).
Thief
A thief consists of two slotted concentric tubes, usually made of
stainless steel or! brass. The outer tube has a conical pointed tip that
permits the sampler to penetrate the material being sampled. The Inner tube
1s rotated to open and close the sampler. A thief Is used to sample dry
granules or powdered wastes whose particle diameter 1s less than one-third the
width of the slots. A thief (Figure 9-12) 1s available at laboratory supply
stores.
Trier
A trier consists of a tube cut in half lengthwise with a sharpened tip
that allows the sampler to cut Into sticky sol Ids and to loosen soil. A trier
samples moist or sticky solids with a particle diameter less than one-half the
diameter of the trier. Triers 61 to 100 cm long and 1.27 to 2.54 cm 1n
diameter are available at laboratory supply stores. A large trier can be
fabricated (see Figure 9-13).
An auger consists of sharpened spiral blades attached to a hard metal
central shaft. An auger samples hard or packed solid wastes or soil. Augers
are: available at .hardware and laboratory supply stores.
Scoops and Shovels
Scoops and shovels are used to sample granular or powdered material 1n
bins, shallow containers, and conveyor belts. Scoops are available at
laboratory supply houses. Flat-nosed shovels are available at hardware
stores.
NINE - 50
Revision 0
Date September 1986
-------
-*])»— 2.66 cm (1 1/8")
t T-H«ndle >
cz
T«p*r*d _ _ _j
Stopper ~
f
II
II
11
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
1H
Ok
=3
•••
J
1
^
\
~ 6.35 cm (2 54") Locking . |
~ Block '
1.62m(5'-0")
••
n"T~
17.8 cm (7")
II
II
II
II
II
II
II
II
II
II
II
||
I
1 I
1 1
II
II
II
II
II
||
||
i
4 Stopper Rod, PVC
0.95 cm (3/8") 0. 0.
t Pipe. PVC. 4. 13cm (1 6/8") I.D.
"• 4.26 cm (1 7/8") 0. D.
H
*il
II
III
ijl Stopptr. Neoprene. No. 9 witn
4P« 3/8" S. S. or PVC Nut tnd W«h<
SAMPLING POSITION
CLOSE POSITION
Figure 9-0. Composite liquid waste sampler (Coliwasa).
NINE - 51
Revision Q
Date September 1986
-------
Washer
Pin
Nut
Figure 9-10. Weighted bottle sampler.
NINE - 52
Revision 0
Date September 1986
-------
Varigrip Gamp
I
en
O 73
01 0>
it <
00 O
O> =3
T3
r*
fD
§
(D
Beaker
150 to 600 ml
Telescoping Aluminum Pole
2.5 to 4.5 Meters (8 to 15 ft.)
Figure 9-11. Dipper.
vo
00
0>
-------
60-100 cm
1
-HH-
1.27-2.54 cm
Figure 9-12. Thief sampler.
NINE - 54
Revision 0
Date September 1986
-------
c
K
122-183 cm
(48-72")
1
>. 5.08-7.62
^^. T
cm
T
\
60-100 cm
\
1.27-2.54 cm
Figure 9-13. Sampling triers.
NINE - 55
Revision p
Date September 1986
-------
Bailer
The bailer Is employed for sampling well water. It consists of a
container attached to a cable that Is lowered Into the well to retrieve a
sample. Bailers can be of various designs. The simplest is a weighted bottle
or basally capped length of pipe that fills from the top as it is lowered into
the well. Some bailers have a check valve, located at the base, which allows
water to enter from the bottom as it 1s lowered into the well. When the
bailer is lifted, the check valve closes, allowing water in the bailer to be
brought to the surface. More sophisticated bailers are available that remain
open at both ends while being lowered, but can be sealed at both top and
bottom by activating a triggering mechanism from the surface. This allows
more reliable sampling at discrete depths within a well. Perhaps the best-
known bailer of this latter design is the Kemmerer sampler.
Bailers generally provide an excellent means for collecting samples from
monitoring wells. They can be constructed from a wide variety of materials
compatible with the parameter of interest. Because they are relatively
inexpensive, bailers can be easily dedicated to an individual well to minimize
cross contamination during sampling. If not dedicated to a well, they can be
easily cleaned to prevent cross contamination. Unfortunately, bailers are
frequently not suited for well evacuation because of their small volume.
Suction Pumps
As the name Implies, suction pumps" operate by creating a partial vacuum
1n a sampling tube. This vacuum allows the pressure exerted by the atmosphere
on the water in the well to force water up the tube to the surface.
Accordingly, these pumps are located at the surface and require only that a
transmission tube be lowered Into the well. Unfortunately, their use is
limited by their reliance on suction to depths of 20 to 25 ft, depending on
the pump. In addition, their use may result 1n out-gassing of dissolved gases
or volatile organics and is therefore limited in many sampling applications.
In spite of this, suction methods may provide a suitable means for well
evacuation because the water remaining in the well is left reasonably
undisturbed.
A variety of pumps that operate on this principle are available, but the
ones most commonly suggested for monitoring purposes are the centrifugal and
peristaltic pumps. In the centrifugal pump, the fluid 1s displaced by the
action of an impeller rotating inside the pump chamber. This discharges water
by centrifugal force. The resulting pressure drop 1n the chamber creates a
suction and causes water to enter the Intake pipe 1n the well. These pumps
can provide substantial yields and are readily available and Inexpensive. The
disadvantages are that they require an external power source and may be
difficult to clean between sampling events. In addition, the materials with
which these pumps are constructed may frequently be incompatible with certain
sample constituents. However, their substantial pumping rates make them
suitable for well evacuation.
NINE - 56
Revision 0
Date September 1986
-------
Peristaltic pumps operate 1n a manner similar to centrifugal pumps but
displace the fluid by mechanical peristalsis. A flexible transmission line is
mounted around the perimeter of the pump chamber, and rotating rollers
compress the tubing, forcing fluid movement ahead (the peristaltic effect) and
Inducing suction behind each roller. This design Isolates the sample from the
moving part of the pump and allows for easy cleaning by removal and
replacement of the flexible tubing. Unfortunately, peristaltic pumps are
generally capable of providing only relatively low yields. They are,
therefore, not Ideally suited to well evacuation.
Positive Displacement Pumps
A variety of positive displacement pumps are available for use 1n with-
drawing water from wells. These methods utilize some pumping mechanism,
placed in the well, that forces water from the bottom of the well to the
surface by some means of positive displacement. This minimizes the potential
for aerating or stripping volatile organics from the sample during removal
from the wel1.
The submersible centrifugal pump is one common example of a positive
displacement pump. It works 1n a manner similar to the centrifugal suction
11ft pump previously described, except that, in this case, both the pump and
electric motor are lowered into the well. As the impeller rotates and fluid
1s brought Into the pump, fluid is displaced up the transmission line and out
of the well. These pumps are capable of providing a high yield. However,
they require an external source of power and are frequently constructed with
materials and contain lubricants Incompatible with certain sample
constituents, particularly organics. They also require considerable equipment
and effort to move from well to well. Cleaning between sampling events 1s
difficult as well, and, until recently, they have not been available for well
diameters smaller than 3 in.
Piston-driven or reciprocating piston pumps are another example of common
positive displacement pumps. These pumps consist of a piston in a submerged
cylinder operated by a rod connected to the drive mechanism at the surface. A
flap valve or ball-check valve is located immediately above or below the
piston cylinder. As the piston 1s lowered 1n the cylinder, the check valve
opens, and water fills the chamber. On the upstroke, the check valve closes,
and water 1s forced out of the cylinder, up Into the transmission line, and to
the surface. The transmission line or piston contains a second check valve
that closes on the downstroke, preventing water from re-entering the cylinder.
These pumps are capable of providing high yields. However, moving these pumps
from well to well 1s difficult, and their use 1n monitoring programs may
require that a pump be dedicated to each well. Many of these pumps may not be
constructed with materials compatible with monitoring certain constituents.
A special adaptation of this pump has recently become available for use
1n ground water monitoring. These piston pumps use compressed gas, rather
than a rod connected to a driving mechanism at the surface, to drive the
pistons. This provides a much more convenient and portable means for
collecting samples from monitoring wells. Compressed-gas pumps provide good
yields and can be constructed with materials compatible with many sampling
programs.
NINE - 57
Revision 0
Date September 1986
-------
Another positive displacement pump applicable for monitoring purposes 1s
the gas-operated squeeze pump. This pump was originally developed by R. F.
Mlddleburg of the U.S.G.S. and consequently 1s referred to as the Mlddleburg
pump. It consists principally of a collapsible membrane Inside a long, rigid
housing, a compressed gas supply, and appropriate control valves. When the
pump 1s submerged, water enters the collapsible membrane through the bottom
check valve. After the membrane has filled, gas pressure 1s applied to the
annular space between the rigid housing and membrane, forcing the water upward
through a sampling tube. When the pressure 1s released, the top check valve
prevents the sample from flowing back down the discharge line, and water from
the well again enters the pump through the bottom check valve.
Gas-operated squeeze pumps offer a number of advantages for use in ground
water monitoring programs. They can be constructed in diameters as small as 1
in. and from a wide variety of materials. They are also relatively portable
and are capable of providing a fair range of pumping rates. Most important,
the driving gas does not contact the water sample, so that possible
contamination or gas stripping does not occur. However, they do require a gas
source, and withdrawal of water from substantial depths may require large gas
volumes and long pumping cycles.
Jet pumps, a common type of submersible pump used 1n small domestic water
wells, may in some cases be suggested for use in monitoring wells. These
pumps operate by injecting water through a pipe down into the well. A venturl
device is located at the intake portion of the pump. As the water injected
from the surface passes through the constricted portion of the venturi, the
velocity increases and pressures decrease according to Bernoulli's principle.
If the discharge velocity at the nozzle is great enough, the pressure at this
point will be lowered sufficiently to draw water into the venturi assembly
through the intake and to bring it to the surface with the original water
injected Into the well. This additional increment of water 1s then made
available at the surface as the pump's output. Because jet pumps require
priming with water and because the water taken from the well mixes with water
circulating in the system, they are clearly not applicable to collecting
samples for monitoring purposes. For similar reasons, their use is not
recommended for well evacuation.
Pressure-Vacuum Lysimeters <
The basic construction of pressure-vacuum lysimeters (Wood, 1973), shown
1n Figure 9-14, consists of a porous ceramic cup, with a bubbling pressure of
1 bar or greater, attached to a short piece of PVC pipe of suitable diameter.
Two tubes extend down into the device, as illustrated. Data by Silkworth and
Grigal (1981) indicate that, of the two commercially available sampler sizes
(2.2 and 4.8 cm diameter), the larger ceramic cup sampler is more reliable,
influences water quality less, and yields samples of suitable volume, for
analysis.
Detailed installation instructions for pressure-vacuum lysimeters are
given by Parizek and Lane (1970). Significant modification may be necessary
to adapt these instruments to field use when heavy equipment is used. To
prevent channelling of contaminated surface water directly to the sampling
device, the sampler may be installed in the side wall of an access trench.
Because random placement procedures may locate a sampler in the middle of an
NINE - 58
Revision 0
Date September 1986
-------
E ]c
S C
TUBING TO SURFACE
CONNECTORS
PIPE-THREAD SEALANT
PVC PIPE CAP
PVC PIPE
PVC CEMENT
POLYETHYLENE TUBING
BRANCH "T"
FEMALE ELBOW
POPPET CHECK VALVE
CONNECTORS
EPOXY CEMENT
POLYETHYLENE TUBING
POROUS CUP 3-8 m pore size
Figure 9-14. One example of a pressure-vacuum lysimeter (Wood, 1973).
Reprinted by permission of the American Geophysical Union.
NINE - 59
Revision 0
Date September 1986
-------
active area, the sample collection tube should be protected at the surface
from heavy equipment by a manhole cover, brightly painted steel cage, or other
structure. Another problem associated with such sampler placement 1s that Its
presence may alter waste management activities (I.e., waste applications,
tilling, etc., will avoid the location); therefore, the sampler may not yield
representative leachate samples. This problem may be avoided by running the
collection tube horizontally underground about 10 m before surfacing.
For sampling after the unit 1s 1n place, a vacuum 1s placed on the system
and the tubes are clamped off. Surrounding soil water 1s drawn Into the
ceramic cup and up the polyethylene tube. To collect the water sample, the
vacuum 1s released, and one tube 1s placed 1n a sample container. Air
pressure 1s applied to the other tube, forcing the liquid up the tube and Into
the sample container. Preliminary testing should ensure that waste products
can pass Into the ceramic cup. If sampling for organlcs, an Inert tubing,
such as one made of Teflon, should be substituted for the polyethylene pipe to
prevent organic contamination.
The major advantages of these sampling devices are that they are easily
available, relatively Inexpensive to purchase and Install, and quite reliable.
The major disadvantage 1s the potential for water quality alterations due to
the ceramic cup; this possible problem requires further testing. For a given
Installation, the device chosen should be specifically tested using solutions
containing the soluble hazardous constituents of the waste to be land treated.
This device 1s not recommended for volatlles unless a special trap device 1s
used (Hazardous Waste Land Treatment, SW-874).
Vacuum Extractor
Vacuum extractors were developed by Duke and Halse (1973) to extract
moisture from soils above the ground water table. The basic device consists
of a stainless steel trough that contains ceramic tubes packed 1n soil. The
unit 1s sized not to Interfere with ambient soil water potentials (Corey,
1974); 1t 1s Installed at a given depth 1n the soil with a slight slope toward
the collection bottle, which 1s 1n the bottom of an adjacent access hole. The
system 1s evacuated and moisture Is moved from the adjacent soil Into the
ceramic tubes and Into the collection bottle, from which 1t can be withdrawn
as desired. The advantage of this system is that 1t yields a quantitative
estimate of leachate flux as well as provides a water sample for analysis.
The volume of collected leachate per unit area per unit time 1s an estimate of
the downward movement of leachate water at that depth. The major
disadvantages to this system are: It is delicate; It requires a trained
operator; it estimates leachate quantity somewhat lower than actual field
drainage; and it disturbs the soil above the sampler. Further details about
the use of the vacuum extractor are given by Trout et al. (1975). Performance
of this device when Installed in clay soils is generally poor.
NINE - 60
Revision 0
Date September 1986
-------
Trench Lyslmeters
Trench lyslmeters are named for the large access trench, or caisson,
necessary for operation. Basic Installation, as described by Parizek and Lane
(1970), Involves excavating a rather large trench and shoring up the side
walls, taking care to leave open areas so that samplers can be placed 1n the
side walls. Sample trays are Imbedded 1n the side walls and connected by
tubing to sample collection containers. The entire trench area 1s then
covered to prevent flooding. One significant danger 1h using this system 1s
the potential for accumulation of hazardous fumes 1n the trench, possibly
endangering the health and safety of the person collecting the samples.
Trench lyslmeters function by Intercepting downward-moving water and
diverting It Into a collection device located at a lower elevation. The
Intercepting agent may be an open-ended pipe, sheet metal trough, pan, or
other similar device. Pans 0.9 to 1.2 m 1n diameter have been successfully
used 1n the field by Tyler and Thomas (1977). Because there 1s no vacuum
applied to the system, only free water 1n excess of saturation 1s sampled.
Consequently, samples are plentiful during rainy seasons but are nonexistent
during the dry season.
Another variation of this system 1s to use a funnel filled with clean
sand Inserted Into the sldewall of the trench. Free water will drain Into a
collection chamber, from which a sample 1s periodically removed by vacuum. A
small sample collection device such as this may be preferable to the large
trench because the necessary hole 1s smaller, so that Installation 1s easier
(Figure 9-15).
9.2.2.5 Quality Assurance and Quality Control
Quality assurance (QA) can briefly be defined as the process for ensuring
that all data and the decisions based on these data are technically sound,
statistically valid, and properly documented. Quality control (QC) procedures
are the tools employed to measure the degree to which these quality assurance
objectives are met.
A data base cannot be properly evaluated for accuracy and precision
unless It 1s accompanied by quality assurance data. In the case of waste
evaluation, these quality assurance data result from the Implementation of
quality control procedures during sampling and analysis. Quality control
requirements for specific analytical methods are given 1n detail 1n each
method 1n this manual; 1n this subsection, quality assurance and quality
control procedures for sampling will be discussed.
Quality control procedures that are employed to document the accuracy and
precision of sampling are:
1. Trip Blanks; Trip blanks should accompany sample containers to and
from the field. These samples can be used to detect any contami-
nation or cross-contamination during handling and transportation.
2. Field Blanks; Field blanks should be collected at specified
frequencies, which will vary according to the probability of
NINE - 61
Revision 0
Date September 1986
-------
VACUUM
SOURCE
. * •
Figure 9-15. Schematic diagram of a sand filled funnel used to collect
leachate from the unsaturated zone.
NINE - 62
Revision o
Date September 1986
-------
contamination or cross-contamination. Field blanks are often metal-
and/or organic-free water allquots that contact sampling equipment
under field conditions and are analyzed to detect any contamination
from sampling equipment, cross contamination from previously
collected samples, or contamination from conditions during sampling
(e.g., airborne contaminants that are not from the waste being
sampled).
3. Field Duplicates; Field duplicates are collected at specified
frequencies and are employed to document precision. The precision
resulting from field duplicates 1s a function of the variance of
waste composition, the variance of the sampling technique, and the
variance of the analytical technique.
4. Field Spikes: Field spikes are infrequently used to determine the
loss of parameters of interest during sampling and shipment to the
I laboratories. Because spiking is done 1n the field, the making of
! spiked samples or spiked blanks is susceptible to error. In
addition, compounds can be lost during spiking, and equipment can be
contaminated with spiking solutions. To eliminate these and other
problems, some analysts spike blanks or matrices similar to the
waste in the laboratory and ship them, along with sample containers,
to the field. This approach also has its limitation because the
matrix and the handling of the spike are different from those of the
actual sample. In all cases, the meaning of a Tow field-spike
recovery is difficult to interpret,' and thus, field spikes are not
commonly used.
In addition to the above quality control samples, a complete quality
assurance program will ensure that standard operating procedures (SOPs) exist
for all essential aspects of a sampling effort. SOPs should exist for the
following steps 1n a sampling effort:
1. Definition of objectives (refer to Section 9.2.1).
2. Design of sampling plans (refer to Section 9.2.2).
3. Preparation of containers and equipment (refer to the specific
analytical methods)..
4. Maintenance, calibration, and cleaning of field equipment (refer to
Instrument manuals or consult a chemist for cleaning protocols).
5. Sample preservation, packaging, and shipping (refer to the
analytical methods and to Section 9.2.2.7).
6. Health and safety protocols (refer to Section 9.2.2.6).
7. Chain-of-custody protocols (refer to Section 9.2.2.7).
In addition to the above protocols, numerous other QA/QC protocols must
be employed to document the accuracy of the analytical portion of a waste
evaluation program.
NINE - 63
Revision 0
Date September 1986
-------
9.2.2.6 Health and Safety
Safety and health must also be considered when implementing a sampling
plan. A comprehensive health and safety plan has three basic elements: (1)
monitoring the health of field personnel; (2) routine safety procedures; and
(3) emergency procedures.
Employees who perform field work, as well as those exposed to chemicals
in the laboratory, should have a medical examination at the initiation of
employment and routinely thereafter. This exam should preferably be performed
and evaluated by medical doctors who specialize in industrial medicine. Some
examples of parts of a medical examination that ought to be performed are:
documentation of medical history; a standard physical exam; pulmonary
functions screening; chest X-ray; EKG; urinalysis; and blood chemistry. These
procedures are useful to: (1) document the quality of an employee's health at
the time of matriculation; (2) ensure the maintenance of good health; and (3)
detect early signs of bodily reactions to chemical exposures so they can be
treated in a timely fashion. Unscheduled examinations should be performed 1n
the event of an accident, Illness, or exposure or suspected exposure to toxic
materials.
Regarding safety procedures, personnel should be aware of the common
routes of exposure to chemicals (I.e., Inhalation, contact, and ingestion) and
be instructed in the proper use of safety equipment, such as Draeger tube air
samplers to detect air contamination, and in the proper use of protective
clothing and respiratory equipment. Protocols should also be defined stating
when safety equipment should be employed and designating safe areas where
facilities are available for washing, drinking, and eating.
Even when the utmost care is taken, an emergency situation can occur as a
result of an unanticipated explosion, electrical hazard, fall, or exposure to
a hazardous substance. To minimize the impact of an emergency, field
personnel should be aware of basic first aid and have Immediate access to a
first-aid kit. Phone numbers for both police and the nearest hospital should
be obtained and kept by each team member before entering the site. Directions
to the nearest hospital should also be obtained so that anyone suffering an
Injury can be transported quickly for treatment.
9.2.2.7 Chain of Custody
An essential part of any sampling/analytical scheme 1s ensuring the
Integrity of the sample from collection to data reporting. The possession and
handling of samples should be traceable from the time of collection through
analysis and final disposition. This documentation of the history of the
sample 1s referred to as chain of custody.
Chain of custody is necessary 1f there is any possibility that the
analytical data or conclusions based upon analytical data will be used In
litigation. In cases where litigation 1s not Involved, many of the cha1n-of-
custody procedures are still useful for routine control of sample flow. The
components of chain of custody — sample seals, a field logbook, cha1n-of-
custody record, and sample analysis request sheet — and the procedures for
their use are described 1n this section.
NINE - 64
Revision 0
Date September 1986
-------
A sample 1s considered 1s considered to be under a person's custody 1f 1t
1s (1) 1n a person's physical possession, (2) 1n view of the person after
taking possession, and (3) secured by that person so that no one can tamper
with 1t, or secured by that person 1n an area that 1s restricted to authorized
personnel. A person who has samples 1n custody must comply with the following
procedures.
(The material presented here briefly summarizes the major aspects of
chain of custody. The reader Is referred to NEIC Policies and Procedures,
EPA-330/9/78/001-R [as revised 1/82], or other manual, as appropriate, for
more Information.)
Sample labels (Figure 9-16) are necessary to prevent m1s1dent1f1cation of
samples. Gummed paper labels or tags are adequate and should Include at least
the following Information:
Sample number.
Name of collector.
Date and time of collection.
Place of collection.
Labels should be affixed to sample containers prior to or at the time of
sampling and should be filled out at the time of collection.
Sample seals are used to detect unauthorized tampering of samples
following sample collection up to the time of analysis. Gummed paper seals
may be used for this purpose. The paper seal should Include, minimally, the
following Information:
Sample number. (This number must be Identical with the number on the
sample label.)
Name of collector.
Date and time of sampling.
Place of collection.
The seal must be attached In such a way that 1t 1s necessary to break 1t
1n order to open the sample container. (An example of an official sample seal
1s shown 1n Figure 9-17.) Seals must be affixed to containers before the
samples leave the custody of sampling personnel.
All Information pertinent to a field survey or sampling must be recorded
1n a logbook. This should be bound, preferably with consecutively numbered
pages that are 21.6 by 27.9 cm (8-1/2 by 11 1n.). At a minimum, entries 1n
the logbook must Include the following:
Location of sampling point.
Name and address of field contact.
Producer of waste and address, 1f different from location.
Type of process producing waste (1f known).
Type of waste (e.g., sludge, wastewater).
Suspected waste composition, Including concentrations.
Number and volume of sample taken.
NINE - 65
Revision 0
Date September 1986
-------
Collector : ' Sample No.
Place of Collection .
Date Sampled Time Sampled
Field Information -
Figure 9-16. Example of Sample Label
NINE - 66
Revision
Date September 1986
-------
NAME AND ADDRESS OF ORGANIZATION COLLECTING SAMPLES
Person Collecting Sample Sample No.
(signature)
Date Collected Time Collected
Place Collected
Figure,9-17. Example of Official Sample Seal
NINE - 67
Revision
Date September 1986
-------
Purpose of sampling (e.g., surveillance, contract number).
Description of sampling point and sampling methodology.
Date and time of collection.
Collector's sample Identification number(s).
Sample distribution and how transported (e.g., name of laboratory, UPS,
Federal Express).
References, such as maps or photographs of the sampling site.
Field observations.
Any field measurements made (e.g., pH, flammablllty, exploslvlty).
Signatures of personnel responsible for observations.
Sampling situations vary widely. No general rule can be given as to the
extent of information that must be entered in the logbook. A good rule,
however, is to record sufficient Information so that anyone can reconstruct
the sampling without reliance on the collector's memory. The logbook must be
stored safely.
To establish the documentation necessary to trace sample possession from
the time of collection,, a chain-of-custody record should be filled out and
should accompany every sample. This record becomes especially Important if
the sample 1s to be Introduced as evidence 1n a court litigation. (A chain-
of-custody record is illustrated in Figure 9-18.)
The record should contain, minimally, the following information:
Sample number.
Signature of collector.
Date and time of collection.
Place and address of collection.
Waste type.
Signature of persons Involved 1n the chain of possession.
Inclusive dates of possession.
The sample analysis request sheet (Figure 9-19) is Intended to accompany
the sample on delivery to the laboratory. The field portion of this form 1s
completed by the person collecting the sample and should include most of the
pertinent information noted in the logbook. The laboratory portion of this
form is Intended to be completed by laboratory personnel and to include,
minimally: '
Name of person receiving the sample.
Laboratory sample number.
Date and time of sample receipt.
Sample allocation.
Analyses to be performed.
The sample should be delivered to the laboratory for analysis as soon as
practicable — usually within 1 or 2 days after sampling. The sample must be
accompanied by the cha1n-of-custody record (Figure 9-18) and by a sample
analysis request sheet (Figure 9-19). The sample must be delivered to the
person In the laboratory authorized to receive samples (often referred to as
the sample custodian).
NINE - 68
Revision 0
Date September 1986
-------
CHAIN OF CUSTODY RECORD
O 73
a> n
(Si O
(D 3
00
Ptoi No.
P?O|CCt HAIIM
Sampler! ttivrururrj
Su. No.
D«e
Time
a
J
Relinqunlwd by: Kignnml
Relinquithcd by: IStymtunl
RHinquntMd by: 1Si*»tunl
2
o
Station Location
D*ttTmw
Oatt Tim
Date Time
No. of
Containeri
Received by: (S'om'urrl
Received by: (S/yraian-l
Received for Laboratory by:
(StgiHturttl
Relinquished by: ISigntturrl
Relinquished by: tStqrMturel
Remarks
Date Time Received by: (Siimiurtl
Date Time Received by: Kigiwurtl
Remarks
Figure 9-18.
-------
SAMPLING ANALYSIS REQUEST
Part I: Field Section
Collector Date Sampled Time hours
Affiliation of Sampler
Address
numberstreetcitystatezTp
Telephone ( ) Company Contact
LABORATORY
SAMPLE COLLECTOR'S TYPE OF
NUMBER SAMPLE NO. SAMPLE* FIELD INFORMATION**
Analysis Requested
Special Handling and/or Storage
PART II: LABORATORY SECTION**
Received by Title Date
Analysis Required
* Indicate whether sample 1s soil, sludge, etc.
**Use back of page for additional Information relative to sample location,
Figure 9-19. Example of hazardous waste sample analysis sheet.
NINE - 70
Revision
Date September 1986
-------
Any material that 1s Identified in the DOT Hazardous Material Table (49
CFR 172.101) must be transported as prescribed in the table. All other
hazardous waste samples must be transported as follows:
1. Collect sample in a 16-oz or smaller glass or polyethylene container
with nonmetallic Teflon-lined screw cap. For liquids, allow
sufficient air space (approximately 10% by volume) so that the
container is not full at 54'C (130°F). If collecting a solid
material, the container plus contents should not exceed 1 Ib net
weight. If sampling for volatile organic analysis, fill VOA
container to septum but place the VOA container inside a 16-oz or
smaller container so that the required air space may be provided.
Large quantities, up to 3.785 liters (1 gal), may be collected if
the sample's flash point is 23°C (75'F) or higher. In this case,
the flash point must be marked on the outside container (e.g.,
carton or cooler), and shipping papers should state that "Flash
point is 73*F or higher."
2. Seal sample and place in a 4-mil-thick polyethylene bag, one sample
per bag.
3. Place sealed bag inside a metal can with noncombustlble, absorbent
cushioning material (e.g., vermiculite or earth) to prevent
breakage, one bag per can. Pressure-close the can and use clips,
tape, or other positive means to hold the lid securely.
4. Mark the can with:
Name and address of originator.
"Flammable Liquid, N.O.S. UN 1993."
(or, "Flammable Solid, N.O.S. UN 1325".)
NOTE: UN numbers are now required in proper shipping names.
5. Place one or more metal cans in a strong outside container such as a
picnic cooler or fiberboard box. Preservatives are not used for
hazardous waste site samples.
6. Prepare for shipping: The words "Flammable Liquid, N.O.S. UN 1993"
or "Flammable Solid, N.O.S. UN 1325"; "Cargo Aircraft Only" (if more
than 1 qt net per outside package); "Limited Quantity" or "Ltd.
Qty."; "Laboratory Samples"; "Net Weight " or "Net Volume "
(of hazardous contents) should be indicated on shipping papers and
on the outside of the outside shipping container. The words "This
Side Up" or "This End Up" should also be on container. Sign the
shipper certification.
NINE - 71
Revision 0
Date September 1986
-------
7. Stand by for possible carrier requests to open outside containers
for inspection or to modify packaging. (It is wise to contact
carrier before packing to ascertain local packaging requirements.)
Remain in the departure area until the carrier vehicle (aircraft,
truck, etc.) is on its way.
At the laboratory, a sample custodian should be assigned to receive the
samples. Upon receipt of a sample, the custodian should inspect the condition
of the sample and the sample seal, reconcile the information on the sample
label and seal against that on the chain-of-custody record, assign a
laboratory number, log in the sample in the laboratory logbook, and store it
in a secured sample storage room or cabinet until it is assigned to an analyst
for analysis.
The sample custodian should inspect the sample for any leakage from the
container. A leaky container containing a multiphase sample should not be
accepted for analysis. This sample will no longer be a representative sample.
If the sample is contained in a plastic bottle and the container walls show
that the sample is under pressure or releasing gases, the sample should be
treated with caution because it may be explosive or release extremely
poisonous gases. The custodian should examine whether the sample seal is
intact or broken, because a broken seal may mean sample tampering and would
make analysis results inadmissible as evidence in court. Any discrepancies
between the information on the sample label and seal and the information that
is on the chain-of-custody record and the sample analysis request sheet should
be resolved before the sample is assigned for analysis. This effort might
require communication with the sample collector. Results of the inspection
should be noted on the sample analysis request sheet and on the laboratory
sample logbook.
Incoming samples usually carry the inspector's or collector's
identification numbers. To identify these samples further, the laboratory
should assign its own identification numbers, which normally are given
consecutively. Each sample should be marked with the assigned laboratory
number. This number is correspondingly recorded on a laboratory sample log
book along with the information describing the sample. The sample information
is copied from the sample analysis request sheet and cross-checked against
that on the sample label.
In most cases, the laboratory supervisor assigns the sample for analysis.
The supervisor should review the information on the sample analysis request
sheet, which now includes inspection notes recorded by the laboratory sample
custodian. The technician assigned to analysis should record in the
laboratory notebook the identifying information about the sample, the date of
receipt, and other pertinent information. This record should also include the
subsequent testing data and calculations. The sample may have to be split
with other laboratories in order to obtain all the necessary analytical
information. In this case, the same type of chain-of-custody procedures must
be employed while the sample is being transported and at the other laboratory.
NINE - 72
Revision 0
Date September 1986
-------
Once the sample has been received 1n the laboratory, the supervisor or
his/her assignee 1s responsible for Its care and custody. That person should
be prepared to testify that the sample was 1n his/her possession or secured 1n
the laboratory at all times, from the moment 1t was received from the
custodian until the analyses were performed.
9.2.3 Sample Plan Implementation
Prior to Implementing a sampling plan, 1t 1s often strategic to walk
through the sampling plan mentally, starting with the preparation of equipment
until the time when samples are received at the laboratory. This mental
excursion should be 1n as much detail as can be Imagined, because the small
details are the ones most frequently overlooked. By employing this technique,
Items not Included on the equipment 11st may be discovered, as well as any
major oversight that could cause the sampling effort to fall. During this
review of the sampling plan, an attempt should be made to anticipate what
could go wrong. A solution to anticipated problems should be found, and, If
necessary, materials needed for solving these problems should be added to the
equipment 11st.
The remainder of this section discusses examples of sampling strategies
for different situations that may be encountered.
Containers
Prior to discussing the sampling of containers, the term must be defined.
The term container, as used here, refers to receptacles that are designed for
transporting materials, e.g., drums and other smaller receptacles, as opposed
to stationary tanks. Weighted bottles, Collwasas, drum thlefs, or triers are
the sampling devices that are chosen for the sampling of containers. (See
Section 9.2.2.4 for a full discussion of sampling equipment.)
The sampling strategy for containers varies according to (1) the number
of containers to be sampled and (2) access to the containers. Ideally, 1f the
waste 1s contained 1n several containers, every container will be sampled. If
this 1s not possible due to the large number of containers or to cost factors,
a subset of Individual containers must be randomly selected for sampling.
This can be done by assigning each container a number and then randomly
choosing a set of numbers for sampling.
Access to a container will affect the number of samples that can be taken
from the container and the location within the container from which samples
can be taken. Ideally, several samples should be taken from locations
displaced both vertically and horizontally throughout the waste. The number
of samples required for reliable sampling will vary depending on the
distribution of the waste components 1n the container. At a minimum with an
unknown waste, a sufficient number and distribution of samples should be taken
to address any possible vertical anomalies In the waste. This 1s because
contained wastes have a much greater tendency to be nonrandomly heterogeneous
1n a vertical rather than a horizontal direction due to (1) settling of sol Ids
and the denser phases of liquids and (2) variation In the content of the waste
as It enters the container. Bags, paper drums, and open-headed steel drums
(of which the entire top can be removed) generally do not restrict access to
the waste and therefore do not limit sampling.
NINE - 73
Revision 0
Date September 1986
-------
When access to a container 1s unlimited, a useful strategy for obtaining
a representative set of samples 1s a three-dimensional simple random sampling
strategy 1n which the container 1s divided by constructing an Imaginary three-
dimensional grid (see Figure 9-20), as follows. First, the top surface of the
waste 1s divided Into a grid whose sections either approximate the size of the
sampling device or are larger than the sampling device 1f the container 1s
large. (Cylindrical containers can be divided Into Imaginary concentric
circles, which are then further divided Into grids of equal size.) Each
section 1s assigned a number. The height of the container 1s then divided
Into Imaginary levels that are at least as large as the vertical space
required by the chosen sampling device. These Imaginary levels are then
assigned numbers. Specific levels and grid locations are then selected for
sampling using a random-number table or random-number generator. (An
alternative means of choosing random sampling locations using circumference
and diameter dimensions is discussed in Section 9.2.2.1.)
Another appropriate sampling approach 1s the two-dimensional simple
random sampling strategy, which can usually yield a more.precise sampling when
fewer samples are collected. This strategy Involves (1) dividing the top
surface of the waste into an Imaginary grid as in the three-dimensional
strategy, (2) selecting grid sections for sampling using random-number tables
or random-number generators, an'd-(3)
-------
Figure 9-20. Container divided into an imaginary three-dimensional grid.
NINE - 75
Revision o
Date September 1986
-------
(Cylindrical tanks can be divided into imaginary concentric circles, which are
then further divided into grids of equal size.) Each section is assigned a
number. The height of the tank is then divided into imaginary levels that are
at least as large as the vertical space required by the chosen sampling
device. These imaginary levels are assigned numbers. Specific levels and
grid locations are then selected for sampling using a random-number table or
random-number generator.
A less comprehensive sampling approach may be appropriate if information
regarding the distribution of waste components is known or assumed (e.g., if
vertical compositing will yield a representative sample). In such cases, a
two-dimensional simple random sampling strategy may be appropriate. In this
strategy, the top surface of the waste is divided into an imaginary grid; grid
sections are selected using random-number tables or random-number generators;
and each selected grid point is then sampled in a vertical manner along the
entire length from top to bottom using a sampling device such as a weighted
bottle, a drum thief, or Coliwasa. If the waste is known to consist of two or
more discrete strata, a more precise representation of the tank contents can
be obtained by using a stratified random sampling strategy, i.e., by sampling
each stratum separately using the two- or three-dimensional simple random
sampling strategy.
Some tanks permit only limited access to their contents, which restricts
the locations within the tank from which samples can be taken. If sampling is
restricted, the sampling strategy must, at a minimum, take sufficient samples
to address the potential vertical anomalies in the waste in order to be
considered representative. This is because contained wastes tend to display
vertical, rather than horizontal, nonrandom heterogeneity due to settling of
suspended solids or denser liquid phases. If access restricts sampling to a
portion of the tank contents (e.g., in an open tank, the size of the tank may
restrict sampling to the perimeter of the tank; in a closed tank, the only
access to the waste may be through inspection ports), then the resulting
analytical data will be deemed representative only of the accessed area, not
of the entire tank .contents unless the tank contents are known to be
homogeneous.
If a limited access tank is to be sampled, and little is known about the
distribution of components within the waste, a set of samples that is
representative of the entire tank contents can be obtained by taking a series
of samples as the tank contents are being drained. This should be done in a
simple random manner by estimating how long 1t will take to drain the tank and
then randomly selecting times during drainage for sampling.
The most appropriate type of sampling device for tanks depends on the
tank parameters. In general, subsurface samplers (I.e., pond samplers) are
used for shallow tanks, and weighted bottles are usually employed for tanks
deeper than 5 ft. Dippers are useful for sampling pipe effluents.
NINE - 76
Revision 0
Date September 1986
-------
Waste Piles
In waste piles, the accessibility of waste for sampling 1s usually a
function of pile size, a key factor 1n the design of a sampling strategy for a
waste pile. Ideally, piles containing unknown wastes should be sampled using
a three-dimensional simple random sampling strategy. This strategy can be
employed only 1f all points within the pile can be accessed. In such cases,
the pile should be divided Into a three-dimensional grid system, the grid
sections assigned numbers, and the sampling points then chosen using random-
number tables or random-number generators.
If sampling 1s limited to certain portions of the pile, then the
collected sample will be representative only of those portions, unless the
waste is known to be homogeneous.
In cases where the size of a pile Impedes access to the waste, a set of
samples that are representative of the entire pile can be obtained with a
minimum of effort by scheduling sampling to coincide with pile removal. The
number of truck!oads needed to remove the pile should be estimated and the
truckloads randomly chosen for sampling.
The sampling devices most commonly used for small piles are thlefs,
triers, and shovels. Excavation equipment, such as backhoes, can be useful
for sampling medium-sized piles.
Landfills and Lagoons
Landfills contain primarily solid waste, whereas lagooned waste may range
from liquids to dried sludge residues. Lagooned waste that 1s either liquid
or semi solid 1s often best sampled using the methods recommended for large
tanks. Usually, solid wastes contained 1n a landfill or lagoon are best
sampled using the three-dimensional random sampling strategy.
The three-dimensional random sampling strategy Involves establishing an
Imaginary three-dimensional grid of sampling points 1n the waste and then
using random-number tables or random-number generators to select points for
sampling. In the case of landfills and lagoons, the grid 1s established using
a survey or map of the area. The map 1s divided Into two two-dimensional
grids with sections of equal size. (An alternative way of choosing random
sampling locations 1s presented 1n the second example described 1n Section
9.2.2.1.) These sections are then assigned numbers sequentially.
Next, the depth to which sampling will take place 1s determined and
subdivided Into equal levels, which are also sequentially numbered. (The
lowest sampling depth will vary from landfill to landfill. Usually, sampling
extends to the Interface of the fill and the natural soils. If soil
contamination Is suspected, sampling may extend Into the natural soil.) The
horizontal and vertical sampling coordinates are then selected using random-
number tables or random-number generators. If some Information Is known about
the nature of the waste, then a modified three-dimensional strategy may be
more appropriate. For example, 1f the landfill consists of several cells, a
more precise measurement may be obtained by considering each cell as a stratum
and employing a stratified three-dimensional random sampling strategy (see
Section 9.1).
NINE - 77
Revision 0
Date September 1986
-------
Hollow-stem augers combined with splIt-spoon samplers are frequently
appropriate for sampling landfills. Water-driven or water-rinsed coring
equipment should not be used for sampling because the water can rinse chemical
components from the sample. Excavation equipment, such as backhoes, may be
useful 1n obtaining samples at various depths; the resulting holes may be
useful for viewing and recording the contents of the landfill.
9.2.4 Sample Compos1t1ng
The compositing of samples, 1s usually done for cost-saving reasons,
Involves the combining of a number of samples or allquots of a number of
samples collected from the same waste. The disadvantage of sample compositing
1s the loss of concentration variance data, whereas the advantage 1s that, for
a given analytical cost, a more representative (i.e., more accurate) sample is
obtained.
It is usually most expedient and cost effective to collect component
samples in the field and to composite allquots of each sample later in the
laboratory. Then, if after reviewing the data any questions arise, the
samples can be recomposlted in a different combination, or each component
sample can be analyzed separately to determine better the variation of waste
composition over time and space, or to determine better the precision of an
average number. The fact that this recomposlting of samples can occur without
the need to resample often results in a substantial cost savings.
To ensure that recomposlting can be done at a later date, 1t is essential
to collect enough sample volume in the field so that, under normal
circumstances, enough component sample will remain following compositing to
allow for a different compositing scheme or even for an analysis of the
component samples themselves.
The actual compositing of samples requires the homogenlzation of all
component samples to ensure that a representative subsample is allquoted. The
homogenization procedure, and the containers and equipment used for
compositing, will vary according to the type of waste being composited and the
parameters to be measured. Likewise, the composite sample Itself will be
homogenized prior to the subsampUng of analytical allquots.
9.2.5 References
1. Corey, P.R., Soil Water Monitoring, Unpublished Report to Dept. of Agr.
Eng., Colorado State University, Fort Collins, Colorado, 1974.
2. Duke, H.R. and H.R. Haise, Vacuum Extractors to Assess Deep Percolation
Losses and Chemical Constituents of Soil Water, Soil Sc1. Soc. Am. Proc. 37,
963-4 (1973). ~
3. Parizek, R.R. and B.E. Lane, Soil-Water Sampling Using Pan and Deep
Pressure-Vacuum Lysimeters, J. Hydr. 11, 1-21 (1970).
NINE - 78
Revision 0
Date September 1986
-------
4. Sllkworth, D.R. and D.F. Grigal, Field Comparison of Soil Solution
Samplers, Soil Sc1. Soc. Am. J. 45, 440-442 (1981).
5. Trout, T.J., J.L. Smith, and D.B. McWhorter, Environmental Effects of
Land Application of Digested Municipal Sewage Sludge, Report submitted to City
of Boulder, Colorado, Dept. of Agr. Engr., Colorado State Univ., Fort Collins,
Colorado, 1975.
6. Tyler, D.D. and G.W. Thomas, Lysimeter Measurements of Nitrate and
Chloride Losses and No-tillage Corn, J. Environ. Qual. 6, 63-66 (1977).
7. U.S. Department of Transportation, Hazardous Materials Table, 49 CFR
172.101.
8. U.S. EPA, Office of Solid Waste amd Emergency Response, Hazardous Waste
Land Treatment, Washington, D.C., SW-874, 1983.
9. U.S. EPA, NEIC Policies and Procedures, 330/9/78/001-R, 1982.
10. Wood, W.W., A Technique Using Porous Cups for Water Sampling at Any Depth
1n the Unsaturated Zone, Water Resources Research 9, 486-488 (1973).
NINE - 79
Revision
Date September 1986
-------
CHAPTER TEN
SAMPLING METHODS
Methods appropriate for use in field sampling situations are included in
this chapter. It contains complete sampling methods for a specific purpose.
Chapter Nine contains general sampling techniques and plans.
TEN - 1
Revision
Date September 1986
-------
METHOD 0010
MODIFIED METHOD 5 SAMPLING TRAIN
1.0 SCOPE AND APPLICATION
1.1 This method is applicable to the determination of Destruction and
Removal Efficiency (ORE) of semi volatile Principal Organic Hazardous Compounds
(POHCs) from incineration systems (PHS, 1967). This method also may be used
to determine particulate emission rates from stationary sources as per EPA
Method 5 (see References at end of this method).
2.0 SUMMARY OF METHOD
2.1 Gaseous and particulate pollutants are withdrawn from an emission
source at an isokinetic sampling rate and are collected in a multicomponent
sampling train. Principal components of the train include a high-efficiency
glass- or quartz-fiber filter and a packed bed of porous polymeric adsorbent
resin. The filter is used to collect organic-laden particulate materials and
the porous polymeric resin to adsorb semi volatile organic species.
Semivolatile species are defined as compounds with boiling points >100*C.
2.2 Comprehensive chemical analyses of the collected sample are
conducted to determine the concentration and identity of the organic
materials.
3.0 INTERFERENCES
3.1 Oxides of nitrogen (NOX) are possible interferents in the
determination of certain water-soluble compounds such as dioxane, phenol, and
urethane; reaction of these compounds with NOX in the presence of moisture
will reduce their concentration. Other possibilities that could result in
positive or negative bias are (1) stability of the compounds in methylene
chloride, (2) the formation of water-soluble organic salts on the resin in the
presence of moisture, and (3) the solvent extraction efficiency of water-
soluble compounds from aqueous media. Use of two or more ions per compound
for qualitative and quantitative analysis can overcome interference at one
mass. These concerns should be addressed on a compound-by-compound basis
before using this method.
4.0 APPARATUS AND MATERIALS
4.1 Sampling train;
4.1.1 A schematic of the sampling train used in this method is
shown in Figure 1. This sampling train configuration is adapted from EPA
Method 5 procedures, and, as such, the majority of the required equipment
0010 - 1
Revision 0
Date September 1986
-------
o
o
I—>
o
I
ro
O 73
ft) (T>
n> •*
.f
co o
CO 13
O
r*
O>
(D
Temperature Sensor | | Stack Wall
Probe -*«A_C
.tf
Hedted Area
Reverse-Type Pilot Tube
X
•3]
Pitot Manometer
Recirculation Pump '
Thermometer
Filter Holder
Thermometer
Check Valve
Vacuum Line
Orifice
Thermometers
M
O
Impingers c Ice Bath
By-Pass Valve
)u
Main Valve
Dry Gas Muter Air Tight Pump
Figure 1. Modified Method 5 Sampling Train.
-------
Is identical to that used in EPA Method 5 determinations. The new
components required are a condenser coil and a sorbent module, which are
used to collect semi volatile organic materials that pass through the
glass- or quartz-fiber filter in the gas phase.
4.1.2 Construction details for the basic train components are given
in APTD-0581 (see Martin, 1971, in Section 13.0, References); commercial
models of this equipment are also available. Specifications for the
sorbent module are provided in the following subsections. Additionally,
the following subsections list changes to APTD-0581 and identify
allowable train configuration modifications.
4.1.3 Basic operating and maintenance procedures for the sampling
train are described in APTD-0576 (see Rom, 1972, in Section 13.0,
References). As correct usage is important in obtaining valid results,
all users should refer to APTD-0576 and adopt the operating and
maintenance procedures outlined therein unless otherwise specified. The
sampling train consists of the components detailed below.
4.1.3.1 Probe nozzle; Stainless steel (316) or glass with
sharp, tapered (30* angle) leading edge. The taper shall be on the
outside to preserve a constant I.D. The nozzle shall be buttonhook
or elbow design and constructed from seamless tubing (if made of
stainless steel). Other construction materials may be considered
for particular applications. A range of nozzle sizes suitable for
isokinetic sampling should be available in Increments of 0.16 cm
(1/16 in.), e.g., 0.32-1.27 cm (1/8-1/2 in.), or larger if higher
volume sampling trains are used. Each nozzle shall be calibrated
according to the procedures outlined in Paragraph 9.1.
4.1.3.2 Probe liner; Borosilicate or quartz-glass tubing with
a heating system capable of maintaining a gas temperature of 120 +
14°C (248 + 25°F) at the exit end during sampling. (The tester may
opt to operate the equipment at a temperature lower than that
specified.) Because the actual temperature at the outlet of the
probe is not usually monitored during sampling, probes constructed
according to APTD-0581 and utilizing the calibration curves of APTD-
0576 (or calibrated according to the procedure outlined 1n APTD-
0576) are considered acceptable. Either boroslHcate or quartz-
glass probe liners may be used for stack temperatures up to about
480*C (900°F). Quartz liners shall be used for temperatures between
480 and 900°C (900 and 1650*F). (The softening temperature for
borosilicate is 820°C (1508*F), and for quartz 1500°C (2732*F).)
Water-cool1ng of the stainless steel sheath will be necessary at
temperatures approaching and exceeding 500*C.
4.1.3.3 Pi tot tube; Type S, as described 1n Section 2.1 of
EPA Method 2, orother appropriate devices (Vollaro, 1976). The
pltot tube shall be attached to the probe to allow constant
monitoring of the stack-gas velocity. The impact (high-pressure)
opening plane of the pltot tube shall be even with or above the
nozzle entry plane (see EPA Method 2, Figure 2-6b) during sampling.
The Type S pltot tube assembly shall have a known coefficient,
determined as outlined 1n Section 4 of EPA Method 2.
0010 - 3
Revision 0
Date September 1986
-------
4.1.3.4 Differential pressure gauge; Inclined manometer or
equivalent device as described In Section 2.2 of EPA Method 2. One
manometer shall be used for velocity-head (AP) readings and the
other for orifice differential pressure (AH) readings.
4.1.3.5 Filter holder; Boroslllcate glass, with a glass frit
filter support and a sealing gasket. The sealing gasket should be
made of materials that will not introduce organic material Into the
gas stream at the temperature at which the filter holder will be
maintained. The gasket shall be constructed of Teflon or materials
of equal or better characteristics. The holder design shall provide
a positive seal against leakage at any point along the filter
circumference. The holder shall be attached immediately to the
outlet of the cyclone or cyclone bypass.
4.1.3.6 Filter heating system; Any heating system capable of
maintaining a temperature of 120 + 14*C (248 + 25*F) around the
filter holder during sampling. ~" Other temperatures may be
appropriate for particular applications. Alternatively, the tester
may opt to operate the equipment at temperatures other than that
specified. A temperature gauge capable of measuring temperature to
within 3*C (5.4*F) shall be installed so that the temperature around
the filter holder can be regulated and monitored during sampling.
Heating systems other than the one shown 1n APTD-0581 may be used.
4.1.3.7 Organic sampling module; This unit consists of three
sections, Including a gas-conditioning section, a sorbent trap, and
a condensate knockout trap. The gas-conditioning system shall be
capable of conditioning the gas leaving the back half of the filter
holder to a temperature not exceeding 20°C (68°F). The sorbent trap
shall be sized to contain approximately 20 g of porous polymeric
resin (Rohm and Haas XAD-2 or equivalent) and shall be jacketed to
maintain the internal gas temperature at 17 + 3*C (62.5 + 5.4*F).
The most commonly used coolant is ice water from the 1mp1nger Ice-
water bath, constantly circulated through the outer jacket, using
rubber or plastic tubing and a peristaltic pump. The sorbent trap
should be outfitted with a glass well or depression, appropriately
sized to accommodate a small thermocouple 1n the trap for monitoring
the gas entry temperature. The condensate knockout trap shall be of
sufficient size to collect the condensate following gas
conditioning. The organic module components shall be oriented to
direct the flow of condensate formed vertically downward from the
conditioning section, through the adsorbent media, and Into the
condensate knockout trap. The knockout trap is usually similar In
appearance to an empty impinger directly underneath the sorbent
module; 1t may be oversized but should have a shortened center stem
(at a minimum, one-half the length of the normal Impinger stems) to
collect a large volume of condensate without bubbling and
overflowing into the Impinger train. All surfaces of the organic
module wetted by the gas sample shall be fabricated of boroslllcate
glass, Teflon, or other Inert materials. Commercial versions of the
0010 - 4
Revision
Date September 1986
-------
complete organic module are not currently available, but may be
assembled from commercially available laboratory glassware and a
custom-fabricated sorbent trap. Details of two acceptable designs
are shown 1n Figures 2 and 3 (the thermocouple well 1s shown In
Figure 2).
4.1.3.8 Implnger train; To determine the stack-gas moisture
content, four 500-mL 1mp1ngers, connected in series with leak-free
ground-glass joints, follow the knockout trap. The first, third,
and fourth 1mp1ngers shall be of the Greenburg-Smlth design,
modified by replacing the tip with a 1.3-cm (l/2-1n.) I.D. glass
tube extending about 1.3 cm (1/2 1n.) from the bottom of the outer
cylinder. The second implnger shall be of the Greenburg-Smlth
design with the standard tip. The first and second impingers shall
contain known quantities of water or appropriate trapping solution.
The third shall be empty or charged with a caustic solution, should
the stack gas contain hydrochloric acid (HC1). The fourth shall
contain a known weight of silica gel or equivalent desiccant.
4.1.3.9 Metering system; The necessary components are a
vacuum gauge, leak-freepump, thermometers capable of measuring
temperature to within 3°C (5.4*F), dry-gas meter capable of
measuring volume to within 1%, and related equipment, as shown in
Figure 1. At a minimum, the pump should be capable of 4 cfm free
flow, and the dry-gas meter should have a recording capacity of
0-999.9 cu ft with a resolution of 0.005 cu ft. Other metering
systems capable of maintaining sampling rates within 10% of
isokineticlty and of determining sample volumes to within 2% may be
used. The metering system must be used 1n conjunction with a pltot
tube to enable checks of Isokinetic sampling rates. Sampling trains
using metering systems designed for flow rates higher than those
described in APTD-0581 and APTD-0576 may be used, provided that the
specifications of this method are met.
4.1.3.10 Barometer; Mercury, aneroid, or other barometer
capable of measuring atmospheric pressure to within 2.5 mm Hg (0.1
in. Hg). In many cases the barometric reading may be obtained from
a nearby National Weather Service station, in which case the station
value (which is the absolute barometric pressure) is requested and
an adjustment for elevation differences between the weather station
and sampling point is applied at a rate of minus 2.5 mm Hg (0.1 in.
Hg) per 30-m (100 ft) elevation increase (vice versa for elevation
decrease).
4.1.3.11 Gas density determination equipment; Temperature
sensor and pressure gauge (asdescribedin Sections 2.3 and 2.4 of
EPA Method 2), and gas analyzer, if necessary (as described in EPA
Method 3). The temperature sensor ideally should be permanently
attached to the pltot tube or sampling probe in a fixed
configuration such that the tip of the sensor extends beyond the
leading edge of the probe sheath and does not touch any metal.
0010 - 5
Revision 0
Date September 1986
-------
.S in.
or
168 mm
o
o
i—»
o
I
28/12
Ball Joint
-11/16" or 45 mm
O 73
01 m
(0 -u
00 O
i
40 RC Glass Frit
28/12 Socket Joint
Jacket
a>
Figure 2. Adsorbent Sampling System.
vo
oo
CTi
-------
Flow Direction
o
o
i—"
o
I
o> n
r+ <
n -••
w
(/) O
m 3
o
<-+•
CD I
CD I
vo
00
Retaining Spring -i
8 mm Glass Cooling Coil
28/12 Ball Joint
Glass Water Jacket
Glass Fritted Disc
Glass Wool Plug -'
Fritted Stainless Steel Disc
15 mm Solv Seal Joint
(or 28/12 Socket Joint)
Figure 3. Adsorbent Sampling System.
-------
Alternatively, the sensor may be attached just prior to use 1n the
field. Note, however, that 1f the temperature sensor 1s attached In
the field, the sensor must be placed In an interference-free
arrangement with respect to the Type S pltot tube openings (see EPA
Method 2, Figure 2-7). As a second alternative, if a difference of
no more than 1% in the average velocity measurement is to be
introduced, the temperature gauge need not be attached to the probe
or pi tot tube.
4.1.3.12 Calibration/field-preparation record; A permanently
bound laboratory notebook, in which duplicate copies of data may be
made as they are being recorded, is required for documenting and
recording calibrations and preparation procedures (i.e., filter and
silica gel tare weights, clean XAD-2, quality assurance/quality
control check results, dry-gas meter, and thermocouple calibrations,
etc.). The duplicate copies should be detachable and should be
stored separately in the test program archives.
4.2 Sample Recovery;
4.2.1 Probe Uner: Probe nozzle and organic module conditioning
section brushes; nylon bristle brushes with stainless steel wire handles
are required. The probe brush shall have extensions of stainless steel,
Teflon, or inert material at least as long as the probe. The brushes
shall be properly sized and shaped to brush out the probe liner, the
probe nozzle, and the organic module conditioning section.
4.2.2 Wash bottles: Three. Teflon or glass wash bottles are
recommended; polyethylene wash bottles should not be used because organic
contaminants may be extracted by exposure to organic solvents used for
sample recovery.
4.2.3 Glass sample storage containers: Chemically resistant,
borosilicate amber and clear glass bottles, 500-mL or 1,000-mL. Bottles
should be tinted to prevent action of light on sample. Screw-cap liners
shall be either Teflon or constructed so as to be leak-free and resistant
to chemical attack by organic recovery solvents. Narrow-mouth glass
bottles have been found to exhibit less tendency toward leakage.
4.2.4 Petrl dishes: Glass, sealed around the circumference with
wide (1-in.) Teflon tape, for storage and transport of filter samples.
4.2.5 Graduated cylinder and/or balances: To measure condensed
water to the nearest 1 mL or 1 g. Graduated cylinders shall have
subdivisions not >2 ml. Laboratory triple-beam balances capable of
weighing to +0.5 g or better are required.
4.2.6 Plastic storage containers: Screw-cap polypropylene or
polyethylene containers to store silica gel.
4.2.7 Funnel and rubber policeman: To aid 1n transfer of silica
gel to container (not necessary if silica gel is weighed 1n field).
0010 - 8
Revision 0
Date September 1986
-------
4.2.8 Funnels: Glass, to aid 1n sample recovery.
4.3 Filters; Glass- or quartz-fiber filters, without organic binder,
exhibiting at least 99.95% efficiency «0.05% penetration) on 0.3-um dloctyl
phthalate smoke particles. The filter efficiency test shall be conducted 1n
accordance with ASTM standard method D2986-71. Test data from the supplier's
quality control program are sufficient for this purpose. In sources
containing S02 or $63, the filter material must be of a type that 1s
unreactlve to SO? or $03. Reeve Angel 934 AH or Schlelcher and Schwell #3
filters work well under these conditions.
4.4 Crushed ice; Quantities ranging from 10-50 Ib may be necessary
during a sampling run, depending on ambient air temperature.
4.5 Stopcock grease; Solvent-Insoluble, heat-stable sillcone grease.
Use of sillcone grease upstream of the module is not permitted, and amounts
used on components located downstream of the organic module shall be
minimized. Silicone grease usage is not necessary if screw-on connectors and
Teflon sleeves or ground-glass joints are used.
4.6 Glass wool; Used to plug the unfritted end of the sorbent module.
The glass-wool fiber should be solvent-extracted with methylene chloride 1n a
Soxhlet extractor for 12 hr and air-dried prior to use.
5.0 REAGENTS
5.1 Adsorbent resin; Porous polymeric resin (XAD-2 or equivalent) Is
recommended"!These resins shall be cleaned prior to their use for sample
collection. Appendix A of this method should be consulted to determine
appropriate precleanlng procedure. For best results, resin used should not
exhibit a blank of higher than 4 mg/kg of total chromatographable organles
(TCO) (see Appendix B) prior to use. Once cleaned, resin should be stored in
an airtight, wide-mouth amber glass container with a Teflon-lined cap or
placed in one of the glass sorbent modules tightly sealed with Teflon film and
elastic bands. The resin should be used within 4 wk of the preparation.
5.2 Silica gel; Indicating type, 6-16 mesh. If previously used, dry at
175*C (350*F) for 2 hr before using. New silica gel may be used as received.
Alternatively, other types of desiccants (equivalent or better) may be used,
subject to the approval of the Administrator.
5.3 Implnger solutions; Distilled organic-free water (Type II) shall be
used, unless sampling Is intended to quantify a particular inorganic gaseous
species. If sampling is intended to quantify the concentration of additional
species, the implnger solution of choice shall be subject to Administrator
approval. This water should be prescreened for any compounds of Interest.
One hundred ml will be added to the specified Impinger; the third Implnger in
the train may be charged with a basic solution (1 N sodium hydroxide or sodium
acetate) to protect the sampling pump from acidic gases. Sodium acetate
should be used when large sample volumes are anticipated because sodium
hydroxide will react with carbon dioxide in aqueous media to form sodium
carbonate, which may possibly plug the Implnger.
0010 - 9
Revision 0
Date September 1986
-------
5.4 Sample recovery reagents;
5.4.1 Hethylene chloride: D1st1lled-1n-glass grade 1s required for
sample recovery and cleanup (see Note to 5.4.2 below).
5.4.2 Methyl alcohol: D1st1lled-1n-glass grade Is required for
sample recovery and cleanup.
NOTE: Organic solvents from metal containers may have a high
residue blank and should not be used. Sometimes suppliers
transfer solvents from metal to glass bottles; thus blanks shall
be run prior to field use and only solvents with low blank value
«0.001%) shall be used.
5.4.3 Water: Water (Type II) shall be used for rinsing the organic
module and condenser component.
6.0 SAMPLE COLLECTION, PRESERVATION, AND HANDLING
6.1 Because of complexity of this method, field personnel should be
trained 1n and experienced with the test procedures 1n order to obtain
reliable results.
6.2 Laboratory preparation;
6.2.1 All the components shall be maintained and calibrated
according to the procedure described 1n APTD-0576, unless otherwise
specified.
6.2.2 Weigh several 200- to 300-g portions of silica gel 1n
airtight containers to the nearest 0.5 g. Record on each container the
total weight of the silica gel plus containers. As an alternative to
prewelghlng the silica gel, 1t may Instead be weighed directly 1n the
Implnger or sampling holder just prior to train assembly.
6.2.3 Check filters visually against light for Irregularities and
flaws or plnhole leaks. Label the shipping containers (glass Petrl
dishes) and keep the filters 1n these containers at all times except
during sampling and weighing.
6.2.4 Desiccate the filters at 20 + 5.6'C (68 + 10'F) and ambient
pressure for at least 24 hr, and weigh at Intervals of at least 6 hr to a
constant weight (I.e., <0.5-mg change from previous weighing), recording
results to the nearest 0.1 mg. During each weighing the filter must not
be exposed for more than a 2-min period to the laboratory atmosphere and
relative humidity above 50%. Alternatively (unless otherwise specified
by the Administrator), the filters may be oven-dried at 105*C (220*F) for
2-3 hr, desiccated for 2 hr, and weighed.
0010 - 10
Revision 0
Date September 1986
-------
6.3 Preliminary field determinations;
6.3.1 Select the sampling site and the minimum number of sampling
points according to EPA Method 1 or as specified by the Administrator.
Determine the stack pressure, temperature, and range of velocity heads
using EPA Method 2. It 1s recommended that a leak-check of the pltot
lines (see EPA Method 2, Section 3.1) be performed. Determine the stack-
gas moisture content using EPA Approximation Method 4 or Its alternatives
to establish estimates of 1sok1netic sampling-rate settings. Determine
the stack-gas dry molecular weight, as described in EPA Method 2, Section
3.6. If Integrated EPA Method 3 sampling is used for molecular weight
determination, the integrated bag sample shall be taken simultaneously
with, and for the same total length of time as, the sample run.
6.3.2 Select a nozzle size based on the range of velocity heads so
that it 1s not necessary to change the nozzle size 1n order to maintain
1sok1net1c sampling rates. During the run, do not change the nozzle.
Ensure that the proper differential pressure gauge is chosen for the
range of velocity heads encountered (see Section 2.2 of EPA Method 2).
6.3.3 Select a suitable probe liner and probe length so that all
traverse points can be sampled. For large stacks, to reduce the length
of the probe, consider sampling from opposite sides of the stack.
6.3.4 A minimum of 3 dscm (105.9 dscf) of sample volume 1s required
for the determination of the Destruction and Removal Efficiency (ORE) of
POHCs from Incineration systems. Additional sample volume shall be
collected as necessitated by analytical detection limit constraints. To
determine the minimum sample volume required, refer to sample
calculations 1n Section 10.0.
6.3.5 Determine the total length of sampling time needed to obtain
the Identified minimum volume by comparing the anticipated average
sampling rate with the volume requirement. Allocate the same time to all
traverse points defined by EPA Method 1. To avoid timekeeping errors,
the length of time sampled at each traverse point should be an integer or
an integer plus one-half m1n.
6.3.6 In some circumstances (e.g., batch cycles) 1t may be
necessary to sample for shorter times at the traverse points and to
obtain smaller gas-sample volumes. In these cases, the Administrator's
approval must first be obtained.
6.4 Preparation of collection train;
6.4.1 During preparation and assembly of the sampling train, keep
all openings where contamination can occur covered with Teflon film or
aluminum foil until just prior to assembly or until sampling 1s about to
begin.
0010 - 11
Revision 0
Date September 1986
-------
6.4.2 Fill the sorbent trap section of the organic module with
approximately 20 g of clean adsorbent resin. While filling, ensure that
the trap packs uniformly, to eliminate the possibility of channeling.
When freshly cleaned, many adsorbent resins carry a static charge, which
will cause clinging to trap walls. This may be minimized by filling the
trap In the presence of an antistatic device. Commercial antistatic
devices include Model-204 and Model-210 manufactured by the 3M Company,
St. Paul, Minnesota.
6.4.3 If an implnger train 1s used to collect moisture, place 100
ml of water in each of the first two impingers, leave the third implnger
empty (or charge with caustic solution, as necessary), and transfer
approximately 200-300 g of preweighed silica gel from Its container to
the fourth Implnger. More silica gel may be used, but care should be
taken to ensure that it 1s not entrained and carried out from the
impinger during sampling. Place the container in a clean place for later
use in the sample recovery. Alternatively, the weight of the silica gel
plus impinger may be determined to the nearest 0.5 g and recorded.
6.4.4 Using a tweezer or clean disposable surgical gloves, place a
labeled (identified) and weighed filter in the filter holder. Be sure
that the filter is properly centered and the gasket properly placed to
prevent the sample gas stream from circumventing the filter. Check the
filter for tears after assembly is completed.
6.4.5 When glass liners are used, Install the selected nozzle using
a V1ton-A 0-ring when stack temperatures are <260*C (500*F) and a woven
glass-fiber gasket when temperatures are higher. See APTD-0576 (Rom,
1972) for details. Other connecting systems utilizing either 316
stainless steel or Teflon ferrules may be used. When metal liners are
used, Install the nozzle as above, or by a leak-free direct mechanical
connection. Mark the probe with heat-resistant tape or by some other
method to denote the proper distance into the stack or duct for each
sampling point.
6.4.6 Set up the train as 1n Figure 1. During assembly, do not use
any si 11 cone grease on ground-glass joints that are located upstream of
the organic module. A very light coating of si 11 cone grease may be used
on all ground-glass joints that are located downstream of the organic
module, but 1t should be limited -to the outer portion (see APTD-0576) of
the ground-glass joints to minimize si 11 cone-grease contamination.
Subject to the approval of ^he Administrator, a glass cyclone may be used
between the probe and the filter holder when the total particulate catch
is expected to exceed 100 mg or when water droplets are present In the
stack. The organic module condenser must be maintained at a temperature
of 17 + 3*C. Connect alii temperature sensors to an appropriate
potentiometer/display unit. Check all temperature sensors at ambient
temperature.
6.4.7 Place crushed ice around the Impingers and the organic module
condensate knockout.
0010 - 12
Revision 0
Date September 1986
-------
6.4.8 Turn on the sorbent module and condenser coll coolant
redrculatlng pump and begin monitoring the sorbent module gas entry
temperature. Ensure proper sorbent module gas entry temperature before
proceeding and again before any sampling 1s Initiated. It 1s extremely
Important that the XAD-2 resin, temperature never exceed 50*C (122*F),
because thermal decomposition will occur. During testing, the XAD-2
temperature must not exceed 20*C (68*F) for efficient capture of the
semivolatile species of Interest.
6.4.9 Turn on and set the filter and probe heating systems at the
desired operating temperatures. Allow time for the temperatures to
stabilize.
6.5 Leak-check procedures
6.5.1 Pre-test leak-check:
6.5.1.1 Because the number of additional 1ntercomponent
connections 1n the Sem1-VOST train (over the M5 Train) Increases the
possibility of leakage, a pre-test leak-check 1s required.
6.5.1.2 After the sampling train has been assembled, turn on
and set the filter and probe heating systems at the desired
operating temperatures. Allow time for the temperatures to
stabilize. If a V1ton A 0-r1ng or other leak-free connection 1s
used 1n assembling the probe nozzle to the probe Uner, leak-check
the train at the sampling site by plugging the nozzle and pulling a
381-mrn Hg (15-1n. Hg) vacuum.
(NOTE: A lower vacuum may be used, provided that 1t 1s not exceeded
during the test.)
*•
6.5.1.3 If an asbestos string 1s used, do not connect the
probe to the train during the leak-check. Instead, leak-check the
train by first attaching a carbon-filled leak-check 1mp1nger (shown
1n Figure 4) to the Inlet of the filter holder (cyclone, 1f applic-
able) and then plugging the Inlet and pulling a 381-mm Hg (15-1n.
Hg) vacuum. (Again, a lower vacuum may be used, provided that 1t Is
not exceeded during the test.) Then, connect the probe to the train
and leak-check at about 25-mm Hg (1-1n. Hg) vacuum; alternatively,
leak-check the probe with the rest of the sampling train 1n one step
at 381-mm Hg (15-1n. Hg) vacuum. Leakage rates 1n excess of 4% of
the average sampling rate or >0.00057 m^/m1n (0.02 cfm), whichever
1s less, are unacceptable.
6.5.1.4 The following leak-check Instructions for the sampling
train described 1n APTD-0576 and APTD-0581 may be helpful. Start
the pump with fine-adjust valve fully open and coarse-adjust valve
completely closed. Partially open the coarse-adjust valve and
slowly close the fine-adjust valve until the desired vacuum 1s
reached. Do not reverse direction of the fine-adjust valve; this
will cause water to back up Into the organic module. If the desired
vacuum 1s exceeded, either leak-check at this higher vacuum or end
the leak-check, as shown below, and start over.
0010 - 13
Revision 0
Date September 1986
-------
CROSS SECTIONAL
Lcik Testing Apparatus
28/12 Femalt
with Inverted Joint
m
v.v
I?:?:
xl
i
v.V.V.V,
28/12 Kale
'Activated Charcoal
Figure 4. Leak-check impinger.
0010 - 14
Revision 0
Date September 1986
-------
6.5.1.5 When the leak-check 1s completed, first slowly remove
the plug from the Inlet to the probe, filter holder, or cyclone (1f
applicable). When the vacuum drops to 127 mm (5 1n.) Hg or less,
Immediately close the coarse-adjust valve. Switch off the pumping
system and reopen the fine-adjust valve. Do not reopen the fine-
adjust valve until the coarse-adjust valve has been closed. This
prevents the water 1n the 1mp1ngers from being forced backward Into
the organic module and silica gel from being entrained backward Into
the third 1mp1nger.
6.5.2 Leak-checks during sampling run:
6.5.2.1 If, during the sampling run, a component (e.g., filter
assembly, 1mp1nger, or sorbent trap) change becomes necessary, a
leak-check shall be conducted Immediately after the Interruption of
sampling and before the change Is made. The leak-check shall be
done according to the procedure outlined 1n Paragraph 6.5.1, except
that it shall be done at a vacuum greater than or equal to the
maximum value recorded up to that point 1n the test. If the leakage
rate 1s found to be no greater than 0.00057 m3/m1n (0.02 cfm) or 4%
of the average sampling rate (whichever 1s less), the results are
acceptable, and no correction will need to be applied to the total
volume of dry gas metered. If a higher leakage rate 1s obtained,
the tester shall void the sampling run. (It should be noted that
any "correction" of the sample volume by calculation by calculation
reduces the Integrity of the pollutant concentrations data generated
and must be avoided.)
6.5.2.2 Immediately after a component change, and before
sampling 1s reinitiated, a leak-check similar to a pre-test leak-
check must also be conducted.
6.5.3 Post-test leak-check:
6.5.3.1 A leak-check is mandatory at the conclusion of each
sampling run. The leak-check shall be done with the same procedures
as those with the pre-test leak-check, except that 1t shall be
conducted at a vacuum greater than or equal to the maximum value
reached during the sampling run. If the leakage rate 1s found to be
no greater than 0.00057 m3/min (0.02 cfm) or 4% of the average
sampling rate (whichever 1s less), the results are acceptable, and
no correction need be applied to the total volume of dry gas
metered. If, however, a higher leakage rate is obtained, the tester
shall either record the leakage rate, correct the sample volume (as
shown in the calculation section of this method), and consider the
data obtained of questionable reliability, or void the sampling run.
6.6 Sampling-train operation;
6.6.1 During the sampling run, maintain an 1sok1netic sampling rate
to within 10% of true isoklnetic, unless otherwise specified by the
Administrator. Maintain a temperature around the filter of 120 + 14*C
(248 + 25*F) and a gas temperature entering the sorbent trap at a maximum
of 20*C (68*F).
0010 - 15
Revision 0
Date September 1986
-------
6.6.2 For each run, record the data required on a data sheet such
as the one shown in Figure 5. Be sure to record the initial dry-gas
meter reading. Record the dry-gas meter readings at the beginning and
end of each sampling time increment, when changes in flow rates are made
before and after each leak-check, and when sampling is halted. Take
other readings required by Figure 5 at least once at each sample point
during each time increment and additional readings when significant
changes (20% variation in velocity-head readings) necessitate additional
adjustments in flow rate. Level and zero the manometer. Because the
manometer level and zero may drift due to vibrations and temperature
changes, make periodic "checks during the traverse.
6.6.3 Clean the stack access ports prior to the test run to
eliminate the chance of sampling deposited material. To begin sampling,
remove the nozzle cap, verify that the filter and probe heating systems
are at the specified temperature, and verify that the pi tot tube and
probe are properly positioned. Position the nozzle at the first traverse
point, with the tip pointing directly Into the gas stream. Immediately
start the pump and adjust the flow to isokinetic conditions. Nomographs,
which aid 1n the rapid adjustment of the Isoklnetic sampling rate without
excessive computations, are available. These nomographs are designed for
use when the Type S pitot-tube coefficient 1s 0.84 + 0.02 and the stack-
gas equivalent density (dry molecular weight) is equal to 29 + 4. APTD-
0576 details the procedure for using the nomographs. If the stack-gas
molecular weight and the pitot-tube coefficient are outside the above
ranges, do not use the nomographs unless appropriate steps (Shigehara,
1974) are taken to compensate for the deviations.
6.6.4 When the stack 1s under significant negative pressure
(equivalent to the height of the impinger stem), take care to close the
coarse-adjust valve before Inserting the probe Into the stack, to prevent
water from backing Into the organic module. If necessary, the pump may
be turned on with the coarse-adjust valve closed.
6.6.5 When the probe is 1n position, block off the openings around
the probe and stack access port to prevent unrepresentative dilution of
the gas stream.
6.6.6 Traverse the stack cross section, as required by EPA Method 1
or as specified by the Administrator, being careful not to bump the probe
nozzle Into the stack walls when sampling near the walls or when removing
or Inserting the probe through the access port, 1n order to minimize the
chance of extracting deposited material.
6.6.7 During the test run, make periodic adjustments to keep the
temperature around the filter holder and the organic module at the proper
levels; add more ice and, if necessary, salt to maintain a temperature of
<20°C (68'F) at the condenser/silica gel outlet. Also, periodically
check the level and zero of the manometer. --
0010 - 16
Revision 0
Date September 1986
-------
Plant
Location
Operator
Date
Run Ho.
Sample Box No.
Meter Box No.
Meter H(>
C Factor
Schematic of Stack Cross Section
Pi tot Tube Coefficient Cp
Ambient Temperature
Barometric Pressure
Assumed Moisture t
Probe length, m (ft)
Nozzle Identification No.
Averaqe Calibrated Nozzle 01i
Probe Heater Setting
Leak Rate, m3/m1n, (cfm)
Probe Liner Material
Pter, cm fin)
Static Pressure, mm Hq (In. Hg)
Filter No.
o
o
i—"
o
I
O 73
tu n
rt <
c/> o
ft)
IO
00
Traverse Point
Number
Total
Average
Sampling
Time
(8) mln.
Vacuum
urn Hq
(in. Hg)
Stack
Temperature
4W,
Velocity
Head
i PJJ
inn (In) H?0
Pressure
Differential
Across
Orifice
Meter
nw (H?0)
(in H70)
Gas Sample
Volume
">3 (ft3>
Gas Sample Temp.
at Dry Gas Meter
Inlet Outlet
•C('F) 'C(F')
Avg. . Avq.
Filter Holder
Temperature
•C(F')
Temperature
of GAS
Enter inq
Sorbet
Trap Y.(F-)
Temperature
of Gas
I.eavinq
Condenser
or last
Imp ino.er
Figure 5. Partlculate field data.
-------
6.6.8 If the pressure drop across the filter or sorbent trap
becomes too high, making 1sok1net1c sampling difficult to maintain, the
filter/sorbent trap may be replaced 1n the midst of a sample run. Using
another complete filter holder/sorbent trap assembly 1s recommended,
rather than attempting to change the filter and resin themselves. After
a new fUter/sorbent trap assembly 1s Installed, conduct a leak-check.
The total particulate weight shall Include the summation of all filter
assembly catches.
6.6.9 A single train shall be used for the entire sample run,
except 1n cases where simultaneous sampling 1s required 1n two or more
separate ducts or at two or more different locations within the same
duct, or In cases where equipment failure necessitates a change of
trains. In all other situations, the use of two or more trains will be
subject to the approval of the Administrator.
6.6.10 Note that when two or more trains are used, separate
analysis of the front-half (1f applicable) organic-module and Implnger
(1f applicable) catches from each train shall be performed, unless
Identical nozzle sizes were used on all trains. In that case, the front-
half catches from the Individual trains may be combined (as may the
Implnger catches), and one analysis of front-half catch and one analysis
of Implnger catch may be performed.
6.6.11 At the end of the sample run, turn off the coarse-adjust
valve, remove the probe and nozzle from the stack, turn off the pump,
record the final dry-gas meter reading, and conduct a post-test leak-
check. Also, leak-check the pltot lines as described 1n EPA Method 2.
The lines must pass this leak-check 1n order to validate the velocity-
head data.
6.6.12 Calculate percent 1sok1net1c1ty (see Section 10.8) to
determine whether the run was valid or another test run should be made.
7.0 SAMPLE RECOVERY
7.1 Preparation;
7.1.1 Proper cleanup procedure begins as soon as the probe is
removed from the stack at the end of the sampling period. Allow the
probe to cool. When the probe can be safely handled, wipe off all
external particulate matter near the tip of the probe nozzle and place a
cap over the tip to prevent losing or gaining partlculate matter. Do not
cap the probe tip tightly while the sampling train is cooling down
because this will create a vacuum in the filter holder, drawing water
from the impingers into the sorbent module.
7.1.2 Before moving the sample train to the cleanup site, remove
the probe from the sample train and cap the open outlet, being careful
not to lose any condensate that might be present. Cap the filter inlet.
0010 - 18
Revision
Date September 1986
-------
Remove the umbilical cord from the last 1mp1nger and cap the 1mp1nger.
If a flexible line 1s used between the organic module and the filter
holder, disconnect the line at the filter holder and let any condensed
water or liquid drain Into the organic module.
7.1.3 Cap the filter-holder outlet and the Inlet to the organic
module. Separate the sorbent trap section of the organic module from the
condensate knockout trap and the gas-conditioning section. Cap all
organic module openings. Disconnect the organic-module knockout trap
from the 1mp1nger train Inlet and cap both of these openings. Ground-
glass stoppers, Teflon caps, or caps of other Inert materials may be used
to seal all openings.
7.1.4 Transfer the probe, the filter, the organic-module
components, and the Implnger/condenser assembly to the cleanup area.
This area should be clean and protected from the weather to minimize
sample contamination or loss.
7.1.5 Save a portion of all washing solutions (methanol/methylene
chloride, Type II water) used for cleanup as a blank. Transfer 200 ml of
each solution directly from the wash bottle being used and place each 1n
a separate, prelabeled glass sample container.
7.1.6 Inspect the train prior to and during disassembly and note
any abnormal conditions.
7.2 Sample containers;
7.2.1 Container no. 1: Carefully remove the filter from the filter
holder and place 1t 1n Its identified Petri dish container. Use a pair
or pairs of tweezers to handle the filter. If 1t is necessary to fold
the filter, ensure that the partlculate cake 1s inside the fold.
Carefully transfer to the Petri dish any particulate matter or filter
fibers that adhere to the filter-holder gasket, using a dry nylon bristle
brush or sharp-edged blade, or both. Label the container and seal with
l-1n.-w1de Teflon tape around the circumference of the lid.
7.2.2 Container no. 2: Taking care that dust on the outside of the
probe or other exterior surfaces does not get into the sample,
quantitatively recover particulate matter or any condensate from the
probe nozzle, probe fitting, probe liner, and front half of the filter
holder by washing these components first with methanol/methylene chloride
(1:1 v/v) Into a glass container. Distilled water may also be used.
Retain a water and solvent blank and analyze in the same manner as with
the samples. Perform rinses as follows:
7.2.2.1 Carefully remove the probe nozzle and clean the inside
surface by rinsing with the solvent mixture (1:1 v/v methanol/-
methylene chloride) from a wash bottle and brushing with a nylon
bristle brush. Brush until the rinse shows no visible particles;
then make a final rinse of the inside surface with the solvent mix.
Brush and rinse the Inside parts of the Swagelok fitting with the
solvent mix in a similar way until no visible particles remain.
0010 - 19
Revision 0
Date September 1986
-------
7.2.2.2 Have two people rinse the probe liner with the solvent
mix by tilting and rotating the probe while squirting solvent into
its upper end so that all inside surfaces will be wetted with
solvent. Let the solvent drain from the lower end into the sample
container. A glass funnel may be used to aid in transferring liquid
washes to the container.
7.2.2.3 Follow the solvent rinse with a probe brush. Hold the
probe in an inclined position and squirt solvent into the upper end
while pushing the probe brush through the probe with a twisting
action; place a sample container underneath the lower end of the
probe and catch any solvent and particulate matter that is brushed
from the probe. Run the brush through the probe three times or more
until no visible particulate matter is carried out with the solvent
or until none remains in the probe liner on visual Inspection^ With
stainless steel or other metal probes, run the brush through in the
above-prescribed manner at least six times (metal probes have small
crevices in which particulate matter can be entrapped). Rinse the
brush with solvent and quantitatively collect these washings in the
sample container. After the brushing, make a final solvent rinse of
the probe as described above.
7.2.2.4 It is recommended that two people work together to
clean the probe to minimize sample losses. Between sampling runs,
keep brushes clean and protected from contamination.
7.2.2.5 Clean the inside of the front half of the filter
holder and cyclone/cyclone flask, if used, by rubbing the surfaces
with a nylon bristle brush and rinsing with methanol/methylene
chloride (1:1 v/v) mixture. Rinse each surface three times or more
if needed to remove visible particulate. Make a final rinse of the
brush and filter holder. Carefully rinse out the glass cyclone and
cyclone flask (if applicable). Brush and rinse any particulate
material adhering to the inner surfaces of these components into the
front-half rinse sample. After all solvent washings and particulate
matter have been collected in the sample container, tighten the lid
on the sample container so that solvent will not leak out when it 1s
shipped to the laboratory. Mark the height of the fluid level to
determine whether leakage occurs during transport. Label the
container to identify its contents.
7.2.3 Container no. 3: The sorbent trap section of the organic
module may be used as a sample transport container, or the spent resin
may be transferred to a separate glass bottle for shipment. If the
sorbent trap Itself 1s used as the transport container, both ends should
be sealed with tightly fitting caps or plugs. Ground-glass stoppers or
Teflon caps may be used. The sorbent trap should then be labeled,
covered with aluminum foil, and packaged on ice for transport to the
laboratory. If a separate bottle is used, the spent resin should be
quantitatively transferred from the trap into the clean bottle. Resin
that adheres to the walls of the trap should be recovered using a rubber
policeman or spatula and added to this bottle.
0010 - 20
Revision 0
Date September 1986
-------
7.2.4 Container no. 4: Measure the volume of condensate collected
1n the condensate knockout section of the organic module to within +1 ml
by using a graduated cylinder or by weighing to within +0.5 g using a
triple-beam balance. Record the volume or weight of liquid present and
note any discoloration or film 1n the liquid catch. Transfer this liquid
to a prelabeled glass sample container. Inspect the back half of the
filter housing and the gas-conditioning section of the organic module.
If condensate is observed, transfer 1t to a graduated or weighing bottle
and measure the volume, as described above. Add this material to the
condensate knockout-trap catch.
7.2.5 Container no. 5: All sampling train components located
between the high-efficiency glass- or quartz-fiber filter and the first
wet impinger or the final condenser system (including the heated Teflon
line connecting the filter outlet to the condenser) should be thoroughly
rinsed with methanol/methylene chloride (1:1 v/v) and the rinsings
combined. This rinse shall be separated from the condensate. If the
spent resin is transferred from the sorbent trap to a separate sample
container for transport, the sorbent trap shall be thoroughly rinsed
until all sample-wetted surfaces appear clean. Visible films should be
removed by brushing. Whenever train components are brushed, the brush
should be subsequently rinsed with solvent mixture and the rinsings added
to this container.
7.2.6 Container no. 6: Note the color of the indicating silica gel
to determine 1f it has been completely spent and make a notation of Its
condition. Transfer the silica gel from the fourth Impinger to its
original container and seal. A funnel may make it easier to pour the
silica gel without spilling. A rubber policeman may be used as an aid in
removing the silica gel from the Impinger. It is not necessary to remove
the small amount of dust particles that may adhere strongly to the
impinger wall. Because the gain in weight is to be used for moisture
calculations, do not use any water or other liquids to transfer the
silica gel. If a balance is available 1n the field, weigh the container
and Its contents to 0.5 g or better.
7.3 Impinger water;
7.3.1 Make a notation of any color or film 1n the liquid catch.
Measure the liquid in the first three impingers to within +1 ml by using
a graduated cylinder or by weighing it to within +0.5 g by using a
balance (if one is available). Record the volume or weight of liquid
present. This information is required to calculate the moisture content
of the effluent gas.
7.3.2 Discard the liquid after measuring and recording the volume
or weight, unless analysis of the Impinger catch is required (see
Paragraph 4.1.3.7). Amber glass containers should be used for storage of
Impinger catch, if required.
7.3.3 If a different type of condenser is used, measure the amount
of moisture condensed either volumetrlcally or gravimetrlcally.
0010 - 21
Revision 0
Date September 1986
-------
7.4 Sample preparation for shipment; Prior to shipment, recheck all
sample containers to ensure that the capsare well secured. Seal the Hds of
all containers around the circumference with Teflon tape. Ship all liquid
samples upright on 1ce and all particulate filters with the particulate catch
facing upward. The partlculate filters should be shipped unrefrlgerated.
8.0 ANALYSIS
8.1 Sample preparation;
8.1.1 General: The preparation steps for all samples will result
1n a finite volume of concentrated solvent. The final sample volume
(usually 1n the 1- to 10-mL range) 1s then subjected to analysis by
GC/MS. All samples should be Inspected and the appearance documented.
All samples are to be spiked with surrogate standards as received from
the field prior to any sample manipulations. The spike should be at a
level equivalent to 10 times the MDL when the solvent 1s reduced 1n
volume to the desired level (I.e., 10 ml). The spiking compounds should
be the stable 1sotop1cally labeled analog of the compounds of Interest or
a compound that would exhibit properties similar to the compounds of
Interest, be easily chromatographed, and not Interfere with the analysis
of the compounds of Interest. Suggested surrogate spiking compounds are:
deuterated naphthalene, chrysene, phenol, nitrobenzene, chlorobenzene,
toluene, and carbon-13-labeled pentachlorophenol.
8.1.2 Condensate: The "condensate" 1s the moisture collected 1n
the first 1mp1nger following the XAD-2 module. Spike the condensate with
the surrogate standards. The volume 1s measured and recorded and then
transferred to a separatory funnel. The pH 1s to be adjusted to pH 2
with 6 N sulfurlc add, 1f necessary. The sample container and graduated
cylinder are sequentially rinsed with three successive 10-mL allquots of
the extraction solvent and added to the separatory funnel. The ratio of
solvent to aqueous sample should be maintained at 1:3. Extract the
sample by vigorously shaking the separatory funnel for 5 m1n. After
complete separation of the phases, remove the solvent and transfer to a
Kuderna-Danlsh concentrator (K-D), filtering through a bed of precleaned,
dry sodium sulfate. Repeat the extraction step two additional times.
Adjust the pH to 11 with 6 N sodium hydroxide and reextract combining the
add and base extracts. Rinse the sodium sulfate Into the K-D with fresh
solvent and discard the deslccant. Add Teflon boiling chips and
concentrate to 10 mL by reducing the volume to slightly less than 10 mL
and then bringing to volume with fresh solvent. In order to achieve the
necessary detection limit, the sample volume can be further reduced to 1
mL by using a micro column K-D or nitrogen blow-down. Should the sample
start to exhibit precipitation, the concentration step should be stopped
and the sample redissolved with fresh solvent taking the volume to some
finite amount. After adding a standard (for the purpose of quantltation
by GC/MS), the sample 1s ready for analysis, as discussed 1n Paragraph
8.2.
0010 - 22
Revision 0
Date September 1986
-------
8.1.3 Implnger: Spike the sample with the surrogate standards;
measure and record the volume and transfer to a separatory funnel.
Proceed as described 1n Paragraph 8.1.2.
8.1.4 XAD-2: Spike the resin directly with the surrogate
standards. Transfer the resin to the all-glass thimbles by the following
procedure (care should be taken so as not to contaminate the thimble by
touching 1t with anything other than tweezers or other solvent-rinsed
mechanical holding devices). Suspend the XAD-2 module directly over the
thimble. The glass frit of the module (see Figure 2) should be 1n the up
position. The thimble Is contained in a clean beaker, which will serve
to catch the solvent rinses. Using a Teflon squeeze bottle, flush the
XAD-2 Into the thimble. Thoroughly rinse the glass module with solvent
Into the beaker containing the thimble. Add the XAD-2 glass-wool plug to
the thimble. Cover the XAD-2 in the thimble with a precleaned glass-wool
plug sufficient to prevent the resin from floating into the solvent
reservoir of the extractor. If the resin is wet, effective extraction
can be accomplished by loosely packing the resin in the thimble. If a
question arises concerning the completeness of the extraction, a second
extraction, without a spike, is advised. The thimble is placed 1n the
extractor and the rinse solvent contained in the beaker is added to the
solvent reservoir. Additional solvent 1s added to make the reservoir
approximately two-thirds full. Add Teflon boiling chips and assemble the
apparatus. Adjust the heat source to cause the extractor to cycle 5-6
times per hr. Extract the resin for 16 hr. Transfer the solvent and
three 10-mL rinses of the reservoir to a K-D and concentrate as described
in Paragraph 8.1.2.
8.1.5 Participate filter (and cyclone catch): If particulate
loading is to be determined, weigh the filter (and cyclone catch, 1f
applicable). The particulate filter (and cyclone catch, if applicable)
1s transferred to the glass thimble and extracted simultaneously with the
XAD-2 resin.
8.1.6 Train solvent rinses: All train rinses (I.e., probe,
implnger, filter housing) using the extraction solvent and methanol are
returned to the laboratory as a single sample. If the rinses are
contained in more than one container, the intended spike is divided
equally among the containers proportioned from a single syringe volume.
Transfer the rinse to a separatory funnel and add a sufficient amount of
organic-free water so that the methylene chloride becomes immiscible and
its volume no longer increases with the addition of more water. The
extraction and concentration steps are then performed as described in
Paragraph 8.1.2.
8.2 Sample analysis;
8.2.1 The primary analytical tool for the measurement of emissions
from hazardous waste incinerators is GC/MS using fused-silica capillary
GC columns, as described in Method 8270 in Chapter Four of this manual.
Because of the nature of GC/MS instrumentation and the cost associated
0010 - 23
Revision 0
Date September 1986
-------
with sample analysis, prescreenlng of the sample extracts by gas
chromatography/flame ionization detection (GC/FID) or with electron
capture (GC/ECD) Is encouraged. Information regarding the complexity and
concentration level of a sample prior to GC/MS analysis can be of
enormous help. This information can be obtained by using either
capillary columns or less expensive packed columns. However, the FID
screen should be performed with a column similar to that used with the
GC/MS. Keep 1n mind that GC/FID has a slightly lower detection limit
than GC/MS and, therefore, that the concentration of the sample can be
adjusted either up or down prior to analysis by GC/MS.
8.2.2 The mass spectrometer will be operated in a full scan (40-
;450) mode for most of the analyses. The range for which data are
acquired 1n a GC/MS run will be sufficiently broad to encompass the major
ions, as listed in Chapter Four, Method 8270, for each of the designated
POHCs in an incinerator effluent analysis.
8.2.3 For most purposes, electron ionlzation (El) spectra will be
collected because a majority of the POHCs give reasonable El spectra.
Also, El spectra are compatible with the NBS Library of Mass Spectra and
other mass spectral references, which aid in the identification process
for other components 1n the Incinerator process streams.
8.2.4 To clarify some Identifications, chemical 1on1zatio'n (CI)
spectra using either positive Ions or negative ions will be used to
elucidate molecular-weight information and simplify the fragmentation
patterns of some compounds. In no case, however, should CI spectra alone
be used for compound identification. Refer to Chapter Four, Method 8270,
for complete descriptions of GC conditions, MS conditions, and
quantitative and quantitative identification.
9.0 CALIBRATION
9.1 Probe nozzle; Probe nozzles shall be calibrated before their
initial use in the field. Using a micrometer, measure the inside diameter of
the nozzle to the nearest 0.025 mm (0.001 1n.). Make measurements at three
separate places across the diameter and obtain the average of the
measurements. The difference between the high and low numbers shall not
exceed 0.1 mm (0.004 in.). When nozzles become nicked, dented, or corroded,
they shall be reshaped, sharpened, and recalibrated before use. Each nozzle
shall be permanently and uniquely identified.
9.2 Pi tot tube; The Type S pltot tube assembly shall be calibrated
according to the procedure outlined 1n Section 4 of EPA Method 2, or assigned
a nominal coefficient of 0.84 if it is not visibly nicked, dented, or corroded
and if 1t meets design and Intercomponent spacing specifications.
0010 - 24
Revision 0
Date September 1986
-------
9.3 Metering system;
9.3.1 Before its initial use in the field, the metering system
shall be calibrated according to the procedure outlined in APTD-0576.
Instead of physically adjusting the dry-gas meter dial readings to
correspond to the wet-test meter readings, calibration factors may be
used to correct the gas meter dial readings mathematically to the proper
values. Before calibrating the metering system, it is suggested that a
leak-check be conducted. For metering systems having diaphragm pumps,
the normal leak-check procedure will not detect leakages within the pump.
For these cases the following leak-check procedure is suggested: Make a
10-min calibration run at 0.00057 m3/min (0.02 cfm); at the end of the
run, take the difference of the measured wet-test and dry-gas meter
volumes and divide the difference by 10 to get the leak rate. The leak
rate should not exceed 0.00057 m3/min (0.02 cfm).
9.3.2 After each field use, the calibration of the metering system
shall be checked by performing three calibration runs at a single
Intermediate orifice setting (based on the previous field test). The
vacuum shall be set at the maximum value reached during the test series.
To adjust the vacuum, insert a valve between the wet-test meter and the
inlet of the metering system. Calculate the average value of the
calibration factor. If the calibration has changed by more than 5%,
recalibrate the meter over the full range of orifice settings, as
outlined in APTD-0576.
9.3.3 Leak-check of metering system: That portion of the sampling
train from the pump to the orifice meter (see Figure 1) should be leak-
checked prior to initial use and after each shipment. Leakage after the
pump will result in less volume being recorded than is actually sampled.
The following procedure is suggested (see Figure 6): Close the main
valve on the meter box. Insert a one-hole rubber stopper with rubber
tubing attached into the orifice exhaust pipe. Disconnect and vent the
low side of the orifice manometer. Close off the low side orifice tap.
Pressurize the system to 13-18 cm (5-7 in.) water column by blowing into
the rubber tubing. Pinch off the tubing and observe the manometer for 1
min. A loss of pressure on the manometer indicates a leak in the meter
box. Leaks, if present, must be corrected.
NOTE: If the dry-gas-meter coefficient values obtained before and after
a test series differ by >5%, either the test series shall be
voided or calculations for test series shall be performed using
whichever meter coefficient value (i.e., before or after) gives
the lower value of total sample volume.
9.4 Probe heater; The probe-heating system shall be calibrated before
its initial use in the field according to the procedure outlined in APTD-0576.
Probes constructed according to APTD-0581 need not be calibrated if the
calibration curves in APTD-0576 are used.
0010 - 25
Revision 0
Date September 1986
-------
o
o
I—«
o
I
PO
RUBBER
TUBING
RUBBER
STOPPER
ORIFICE
BY PASS VALVE
VACUUM
GAUGE
HOW INTO TUIINC
UNTIL MANOMfTCft
HI ADS $ TO 11NCHES
WATCH COLUMN
O 50
o» m
omnci
MANOMCTtM
MAIN VALVE CLOSED
AIRTIGHT
PUMP
v>
CO O
3
O
Figure 6. Leak-check of meter box.
CT
n
oo
o>
-------
9.5 Temperature gauges; Each thermocouple must be permanently and
uniquely marked on the casting; all mercury-1n-glass reference thermometers
must conform to ASTM E-l 63C or 63F specifications. Thermocouples should be
calibrated 1n the laboratory with and without the use of extension leads. If
extension leads are used 1n the field, the thermocouple readings at ambient
air temperatures, with and without the extension lead, must be noted and
recorded. Correction 1s necessary 1f the use of an extension lead produces a
change >1.5%.
9.5.1 Implnger, organic module, and dry-gas meter thermocouples:
For the thermocouples used to measure the temperature of the gas leaving
the Implnger train and the XAD-2 resin bed, three-point calibration at
Ice-water, room-air, and boiling-water temperatures 1s necessary. Accept
the thermocouples only 1f the readings at all three temperatures agree to
+2*C (3.6*F) with those of the absolute value of the reference
thermometer.
9.5.2 Probe and stack thermocouple: For the thermocouples used to
Indicate the probe and stack temperatures, a three-point calibration at
Ice-water, boiling-water, and hot-oil -bath temperatures must be
performed; 1t 1s recommended that room-air temperature be added, and that
the thermometer and the thermocouple agree to within 1.5% at each of the
calibration points. A calibration curve (equation) may be constructed
(calculated) and the data extrapolated to cover the entire temperature
range suggested by the manufacturer.
9.6 Barometer; Adjust the barometer Initially and before each test
series to agree to within +25 mm Hg (0.1 1n. Hg) of the mercury barometer or
the corrected barometric pressure value reported by a nearby National Weather
Service Station (same altitude above sea level).
9.7 Triple-beam balance; Calibrate the triple-beam balance before each
test series, using Class-S standard weights; the weights must be within +0.5%
of the standards, or the balance must be adjusted to meet these limits.
10.0 CALCULATIONS
10.1 Carry out calculations. Round off figures after the final
calculation to the correct number of significant figures.
10.2 Nomenclature;
An = Cross-sectional area of nozzle, m2 (ft2).
Bws = Water vapor In the gas stream, proportion by volume.
s P1tot tube coefficient (nominally 0.84 + 0.02),
dlmenslonless.
I = Percent of 1sok1net1c sampling.
0010 - 27
Revision 0
Date September 1986
-------
La = Maximum acceptable leakage rate for a leak-check, either pre-test
or following a component change; equal to 0.00057 m3/m1n (0.02
cfm) or 4% of the average sampling rate, whichever is less.
LI = Individual leakage rate observed during the leak-check conducted
prior to the "1*"" component change (i = 1, 2, 3...n) m3/min
(cfm).
LD = Leakage rate observed during the post-test leak-check, m3/min
H (cfm).
Md = Stack-gas dry molecular weight, g/g-mole (Ib/lb-mole).
Mw = Molecular weight of water, 18.0 g/g-mole (18.0 Ib/lb-mole).
Pbar = Barometric pressure at the sampling site, mm Hg (in. Hg).
Ps = Absolute stack-gas pressure, mm Hg (in. Hg).
pstd = Standard absolute pressure, 760 mm Hg (29.92 in. Hg).
R = Ideal gas constant, 0.06236 mm Hg-m3/K-g-mole (21.85 in.
Hg-ft3/*R-lb-mole).
Tm = Absolute average dry-gas meter temperature (see Figure 6), K
(*R).
Ts = Absolute average stack-gas temperature (see Figure 6), K (*R).
Tstd = Standard absolute temperature, 293K (528°R).
Vyc = Total volume of liquid collected in the organic module condensate
knockout trap, the implngers, and silica gel, mL.
Vm = Volume of gas sample as measured by dry-gas meter, dscm (dscf).
Vm(std) = Volume of gas sample measured by the dry-gas meter, corrected
to standard conditions, dscm (dscf).
vw(std) = Volume of water vapor 1n the gas sample, corrected to standard
conditions, scm (scf).
Vs = Stack-gas velocity, calculated by Method 2, Equation 2-9, using
data obtained from Method 5, m/sec (ft/sec).
Wa = Weight of residue in acetone wash, mg.
7 = Dry-gas-meter calibration factor, dimensionless.
AH = Average pressure differential across the orifice meter (see
Figure 2), mm H20 (1n. H20).
0010 - 28
Revision
Date September 1986
-------
/jw = Density of water, 0.9982 g/mL (0.002201 Ib/mL).
8 = Total sampling time, mln.
BI = Sampling time interval from the beginning of a run until the
first component change, min.
8i = Sampling time interval between two successive component
changes, beginning with the interval between the first and
second changes, mln.
Bp = Sampling time interval from the final (ntn) component change
until the end of the sampling run, mln.
13.6 = Specific gravity of mercury.
60 = sec/mi n.
100 = Conversion to percent.
10.3 Average dry-gas-meter temperature and average orifice pressure
drop; See data sheet (Figure 5, above).
10.4 Dry-gas volume; Correct the sample measured by the dry-gas meter
to standard conditions fcO'C, 760 mm Hg [68*F, 29.92 in. HgJ) by using
Equation 1:
Tstd P + AH/13'6
V ' — -- - -- Klv«7
where:
m(std) '
Tm pstd m
KI = 0.3858 K/mm Hg for metric units, or
K! = 17.64*R/in. Hg for English units.
It should be noted that Equation 1 can be used as written, unless the leakage
rate observed during any of the mandatory leak-checks (I.e., the post-test
leak-check or leak-checks conducted prior to component changes) exceeds La.
If Lp or Lf exceeds La, Equation 1 must be modified as follows;
a. Case I (no component changes made during sampling run); Replace Vm
in Equation 1 with the expression:
Vm -
0010 - 29
Revision
Date September 1986
-------
b. Case II (one or more component changes made during the sampling
run): Replace Vm 1n Equation,! by the expression:
Vn, - "l - "l ' "p
and substitute only for those leakage rates (Li or LD) that exceed
l-a-
10.5 Volume of water vapor:
Pw RTstd
Vw(std) ' Vlc ~ — - K2 vlc (2)
Mw pstd
where:
K2 = 0.001333 m3/mL for metric units, or
K2 = 0.04707 ft3/ml_ for English units.
10.6 Moisture content;
Vw(std)
Bws • (3)
Vm(std) + Vw(std)
NOTE: In saturated or water-droplet-laden gas streams, two calculations
of the moisture content of the stack gas shall be made, one from
the Impinger analysis (Equation 3) and a second from the
assumption of saturated conditions. The lower of the two values
of Bw shall be considered correct. The procedure for determining
the moisture content based upon assumption of saturated conditions
is given in the Note to Section 1.2 of Method 4. For the purposes
of this method, the average stack-gas temperature from Figure 6
may be used to make this determination, provided that the accuracy
of the 1n-stack temperature sensor 1s +1*C (2*F).
10.7 Conversion factors;
From To Multiply by
scT~ in* 0.02832
g/ft3 gr/ft3 15.43
g/ft3 Ib/ft3 2.205 x 1Q-3
g/ft3 g/m3 35.31
0010 - 30
Revision
Date September 1986
-------
10.8 Isok1net1c variation;
10.8.1 Calculation from raw data:
100 VK3Flc + (VV (Pbar+ AH/13.6)]
I = (4)
608VsPsAn
where:
KS = 0.003454 mm Hg-m3/mL-K for metric units, or
K3 = 0.002669 in. Hg-ft3/mL-*R for English units.
10.8.2 Calculation for Intermediate values:
T . TsVm(Std)Pstd100
X *•* x if o A n c.f\ i
TsVm(std)
where:
l<4 = 4.320 for metric units, or
K4 = 0.09450 for English units.
10.8.3 Acceptable results: If 90% ^ I ^ 110%, the results are
acceptable. If the results are low in comparison with the standard and
I is beyond the acceptable range, or 1f I 1s less than 90%, the
Administrator may opt to accept the results.
10.9 To determine the minimum sample volume that shall be collected, the
following sequence of calculations shall be used.
10.9.1 From prior analysis of the waste feed, the concentration of
POHCs introduced into the combustion system can be calculated. The
degree of destruction and removal efficiency that is required is used to
determine the maximum amount of POHC allowed to be present in the
effluent. This may be expressed as:
(WF) (POHC1 cone) (100-%DRE)
= Max POHC< Mass (6)
100 100 n
where:
WF = mass flow rate of waste feed per hr, g/hr (Ib/hr).
POHCj = concentration of Principal Organic Hazardous Compound (wt %)
Introduced into the combustion process.
0010 - 31
Revision 0
Date September 1986
-------
ORE = percent Destruction and Removal Efficiency required.
Max POHC = mass flow rate (g/hr [lb/hr]) of POHC emitted from the
combustion source.
10.9.2 The average discharge concentration of the POHC 1n the
effluent gas 1s determined by comparing the Max POHC with the volumetric
flow rate being exhausted from the source. Volumetric flow rate data are
available as a result of preliminary Method 1-4 determinations:
.Max POHC1 Mass
= Max POHC< cone (7)
DVeff(std)
where:
DVeff(std) = volumetric flow rate of exhaust gas, dscm (dscf).
POHC^ cone = anticipated concentration of the POHC 1n the
exhaust gas stream, g/dscm (Ib/dscf).
10.9.3 In making this calculation, 1t 1s recommended that a safety
margin of at least ten be Included:
LDLpQHC x 10
POHC1 cone
' VTBC
where:
LDLpQHC = detectable amount of POHC 1n entire sampling train.
NOTE: The whole extract from an XAD-2 cartridge 1s seldom 1f ever,
Injected at once. Therefore, 1f allquotlng factors are
Involved, the LDLpgnc ^s not tne same as the analytical (or
column) detection limit.
= minimum dry standard volume to be collected at dry-gas
meter.
10.10 Concentration of any given POHC 1n the gaseous emissions of a
combustion process;
1) Multiply the concentration of the POHC as determined 1n Method 8270
by the final concentration volume, typically 10 ml.
CPOHC (ug/mL) x sample volume (ml) = amount (ug) of POHC 1n sample (9)
0010 - 32
Revision
Date September 1986
-------
where:
CPOHC = concentration of POHC as analyzed by Method 8270.
2) Sum the amount of POHC found 1n all samples associated with a single
train.
Total (ug) = XAD-2 (ug) + condensate (ug) + rinses (ug) + 1mp1nger (ug) (10)
3) Divide the total ug found by the volume of stack gas sampled (m3).
(Total ug)/(train sample volume) = concentration of POHC (ug/m3) (11)
11.0 QUALITY CONTROL
11.1 Sampling; See EPA Manual 600/4-77-027b for Method 5 quality
control.
11.2 Analysis; The quality assurance program required for this study
Includes theanalysis of field and method blanks, procedure validations,
Incorporation of stable labeled surrogate compounds, quantltatlon versus
stable labeled Internal standards, capillary column performance checks, and
external performance tests. The surrogate spiking compounds selected for a
particular analysis are used as primary indicators of the quality of the
analytical data for a wide range of compounds and a variety of sample
matrices. The assessment of combustion data, positive identification, and
quantltatlon of the selected compounds are dependent on the Integrity of the
samples received and the precision and accuracy of the analytical methods
employed. The quality assurance procedures for this method are designed to
monitor the performance of the analytical method and to provide the required
information to take corrective action 1f problems are observed 1n laboratory
operations or 1n field sampling activities.
11.2.1 Field Blanks: Field blanks must be submitted with the
samples collected at each sampling site. The field blanks Include the
sample bottles containing aliquots of sample recovery solvents, unused
filters, and resin cartridges. At a minimum, one complete sampling train
will be assembled 1n the field staging area, taken to the sampling area,
and leak-checked at the beginning and end of the testing (or for the same
total number of times as the actual test train). The filter housing and
probe of the blank train will be heated during the sample test. The
train will be recovered as if it were an actual test sample. No gaseous
sample will be passed through the sampling train.
11.2.2 Method blanks: A method blank must be prepared for each set
of analytical operations, to evaluate contamination and artifacts that
can be derived from glassware, reagents, and sample handling in the
laboratory.
11.2.3 Refer to Method 8270 for additional quality control
considerations.
0010 - 33
Revision 0
Date September 1986
-------
12.0 METHOD PERFORMANCE
12.1 Method performance evaluation; Evaluation of analytical procedures
for a selectedserieso?compounds must Include the sample-preparation
procedures and each associated analytical determination. The analytical
procedures should be challanged by the test compounds spiked at appropriate
levels and carried through the procedures.
12.2 Method detection limit; The overall method detection limits (lower
and upper) must be determined on a compound-by-compound basis because
different compounds may exhibit different collection, retention, and
extraction efficiencies as well as Instrumental minimum detection limit (MDL).
The method detection limit must be quoted relative to a given sample volume.
The upper limits for the method must be determined relative to compound
retention volumes (breakthrough).
12.3 Method precision and bias; The overall method precision and bias
must be determined onacompound-by-compound basis at a given concentration
level. The method precision value would include a combined variability due to
sampling, sample preparation, and Instrumental analysis. The method bias
would be dependent upon the collection, retention, and extraction efficiency
of the train components. From evaluation studies to date using a dynamic
spiking system, method biases of -13% and -16% have been determined for
toluene and 1,1,2,2-tetrachloroethane, respectively. A precision of 19.9% was
calculated from a field test data set representing seven degrees of freedom
which resulted from a series of paired, unspiked Semi volatile Organic Sampling
trains (Sem1-VOST) sampling emissions from a hazardous waste Incinerator.
13.0 REFERENCES
1. Addendum to Specifications for Incinerator Testing at Federal Facilities,
PHS, NCAPC, December 6, 1967.
2. Bursey, J., Homolya, J., McAllister, R., and McGangley, J., Laboratory
and Field Evaluation of the Semi-VOST Method, Vols. 1 and 2, U.S.
Environmental Protection Agency, EPA/600/4-851/075A, 075B (1985).
3. Martin, R.M., Construction Details of Isoklnetic Source-Sampling
Equipment, Research Triangle Park, NC, U.S. Environmental Protection Agency,
April 1971, PB-203 060/BE, APTD-0581, 35 pp.
4. Rom, J.J., Maintenance, Calibration, and Operation of Isokinetlc Source-
Sampling Equipment, Research Triangle Park, NC, U.S. Environmental Protection
Agency, March 1972, PB-209 022/BE, APTD-0576, 39 pp.
i
5. Schlickenrieder, L.M., Adams, J.W., and Thrun, K.E., Modified Method 5
Train and Source Assessment Sampling System: Operator's Manual, U.S.
Environmental Protection Agency, EPA/600/8-85/003, (1985).
0010 - 34
Revision 0
Date September 1986
-------
6. Shlgehara, R.T., Adjustments In the EPA Nomography for Different P1tot
Tube Coefficients and Dry Molecular Weights, Stack Sampling News, 2:4-11
(October 1974).
7. U.S. Environmental Protection Agency, CFR 40 Part 60, Appendix A, Methods
1-5.
8. Vollaro, R.F., A Survey of Commercially Available Instrumentation for the
Measurement of Low-Range Gas Velocities, Research Triangle Park, NC, U.S.
Environmental Protection Agency, Emissions Measurement Branch, November 1976
(unpublished paper).
0010 - 35
Revision
Date September 1986
-------
METHOD 0010, APPENDIX A
PREPARATION OF XAD-2 SORBENT RESIN
1.0 SCOPE AND APPLICATION
1.1 XAD-2 resin as supplied by the manufacturer 1s Impregnated with a
bicarbonate solution to Inhibit mlcroblal growth during storage. Both the
salt solution and any residual extractable monomer and polymer species must be
removed before use. The resin Is prepared by a series of water and organic
extractions, followed by careful drying.
2.0 EXTRACTION
2.1 Method 1: The procedure may be carried out in a giant Soxhlet
extractor.An all-glass thimble containing an extra-coarse frit 1s used for
extraction of XAD-2. The frit 1s recessed 10-15 mm above a crenellated ring
at the bottom of the thimble to facilitate drainage. The resin must be
carefully retained 1n the extractor cup with a glass-wool plug and stainless
steel screen because 1t floats on methylene chloride. This process Involves
sequential extraction in the following order.
Solvent
Water
Water
Methyl alcohol
Methylene chloride
Methylene chloride (fresh)
2.2 Method 2:
Procedure
Initial rinse: Place resin in a beaker,
rinse once with Type II water, and
discard. Fill with water a second time,
let stand overnight, and discard.
Extract with ^0 for 8 hr.
Extract for 22 hr.
Extract for 22 hr.
Extract for 22 hr.
2.2.1 As an alternative to Soxhlet extraction, a continuous
extractor has been fabricated for the extraction sequence. This extractor has
been found to be acceptable. The particular canister used for the apparatus
shown in Figure A-l contains about 500 g of finished XAD-2. Any size may be
constructed; the choice is dependent on the needs of the sampling programs.
The XAD-2 is held under light spring tension between a pair of coarse and fine
screens. Spacers under the bottom screen allow for even distribution of clean
solvent. The three-necked flask should be of sufficient size (3-liter in this
case) to hold solvent
0010 - A - 1
Revision 0
Date September 1986
-------
Oriin
Optional Pump
Figure A-l. XAD-2 cleanup extraction apparatus.
0010 - A - 2
Revision 0
Date September 1986
-------
equal to twice the dead volume of the XAD-2 canister. Solvent 1s refluxed
through the Snyder column, and the distillate 1s continuously cycled up
through the XAD-2 for extraction and returned to the flask. The flow is
maintained upward through the XAD-2 to allow maximum solvent contact and
prevent channeling. A valve at the bottom of the canister allows removal of
solvent from the canister between changes.
2.2.2 Experience has shown that 1t 1s very difficult to cycle
sufficient water 1n this mode. Therefore the aqueous rinse 1s accomplished by
simply flushing the canister with about 20 liters of distilled water. A small
pump may be useful for pumping the water through the canister. The water
extraction should be carried out at the rate of about 20-40 mL/m1n.
2.2.3 After draining the water, subsequent methyl alcohol and
methylene chloride extractions are carried out using the refluxlng apparatus.
An overnight or 10- to 20-hr period 1s normally sufficient for each
extraction.
2.2.4 All materials of construction are glass, Teflon, or stainless
steel. Pumps, 1f used, should not contain extractable materials. Pumps are
not used with methanol and methylene chloride.
3.0 DRYING
3.1 After evaluation of several methods of removing residual solvent, a
fluldlzed-bed technique has proved to be the fastest and most reliable drying
method.
3.2 A simple column with suitable retainers, as shown 1n Figure A-2,
will serve as a satisfactory column. A 10.2-cm (4-1n.) Pyrex pipe 0.6 m (2
ft) long will hold all of the XAD-2 from the extractor shown 1n Figure A-l or
the Soxhlet extractor, with sufficient space for flu1d1z1ng the bed while
generating a minimum resin load at the exit of the column.
3.3 Method 1; The gas used to remove the solvent 1s the key to
preserving the cleanliness of the XAD-2. Liquid nitrogen from a standard
commercial liquid nitrogen cylinder has routinely proved to be a reliable
source of large volumes of gas free from organic contaminants. The liquid
nitrogen cylinder 1s connected to the column by a length of precleaned 0.95-cm
(3/8-1n.) copper tubing, colled to pass through a heat source. As nitrogen is
bled from the cylinder, it 1s vaporized 1n the heat source and passes through
the column. A convenient heat source 1s a water bath heated from a steam
line. The final nitrogen temperature should only be warm to the touch and not
over 40*C. Experience has shown that about 500 g of XAD-2 may be dried
overnight by consuming a full 160-Hter cylinder of liquid nitrogen.
3.4 Method 2: As a second choice, high-purity tank nitrogen may be used
to dry the XAD-2. The high-purity nitrogen must first be passed through a bed
0010 - A - 3
Revision 0
Date September 1986
-------
LOOM Wttvt Nylon
FtbrieCovtr
1O2em -
(4 Inch) Pyrtx
Fipt
Liquid Takt off
0.95 em (3/8 in) Tubing
Liquid Nitrogtn
Cytindtr
0501}
Hut Sourc*
FintScrttn
'^m&3^Cc*» Suppc
t I Vrfo
Figure A-2. XAD-2 fluidized-bed drying apparatus.
0010 - A - 4
Revision 0
Date September 1986
-------
of activated charcoal approximately 150 ml 1n volume. With either type of
drying method, the rate of flow should gently agitate the bed. Excessive
fluldlzatlon may cause the particles to break up.
4.0 QUALITY CONTROL PROCEDURES
4.1 For both Methods 1 and 2, the quality control results must be
reported for the batch. The batch must be reextracted 1f the residual
extractable organlcs are >20 ug/mL by TCO analysis or the gravimetric residue
1s >0.5 mg/20 g XAD-2 extracted. (See also section 5.1, Method 0010.)
4.2 Four control procedures are used with the final XAD-2 to check for
(1) residual methylene chloride, (2) extractable organlcs (TCO), (3) specific
compounds of Interest as determined by GC/MS, as described 1n Section 4.5
below, and (4) residue (GRAV).
4.3 Procedure for residual methylene chlorldet
4.3.1 Description: A 1+0.1-g sample of dried resin Is weighed Into
a small vial, 3 mL of toluene are added, and the vial 1s capped and well
shaken. Five uL of toluene (now containing extracted methylene chloride) are
Injected Into a gas chromatograph, and the resulting Integrated area 1s
compared with a reference standard. The reference solution consists of 2.5 uL
of methylene chloride In 100 mL of toluene, simulating 100 ug of residual
methylene chloride on the resin. The acceptable maximum content Is 1,000 ug/g
resin.
4.3.2 Experimental: The gas chromatograph conditions are as
follows:
6-ft x 1/8-1n. stainless steel column containing 10% OV-101 on
100/120 Supelcoport;
Helium carrier at 30 mL/m1n;
FID operated on 4 x 10"11 A/mV;
Injection port temperature: 250*C;
Detector temperature: 305*C;
Program: 30'C(4 m1n) 40*C/m1n 250»C (hold); and
Program terminated at 1,000 sec.
4.4 Procedure for residual extractable organlcs;
4.4.1 Description: A 20+0.1-g sample of cleaned, dried resin Is
weighed Into a precleaned alundum or cellulose thimble which 1s plugged with
cleaned glass wool. (Note that 20 g of resin will fill a thimble, and the
0010 - A - 5
Revision
Date September 1986
-------
resin will float out unless well plugged.) The thimble containing the resin
1s extracted for 24 hr with 200-mL of pesticide- grade methylene chloride
(Burdlck and Jackson pesticide-grade or equivalent purity). The 200-mL
extract 1s reduced 1n volume to 10-mL using a Kuderna-Danlsh concentrator
and/or a nitrogen evaporation stream. Five uL of that solution are analyzed
by gas chromatography using the TCO analysis procedure. The concentrated
solution should not contain >20 ug/mL of TCO extracted from the XAD-2. This
Is equivalent to 10 ug/g of TCO 1n the XAD-2 and would correspond to 1.3 mg of
TCO In the extract of the 130-g XAD-2 module. Care should be taken to correct
the TCO data for a solvent blank prepared (200 ml reduced to 10 ml) 1n a
similar manner.
4.4.2 Experimental: Use the TCO analysis conditions described 1n
the revised Level 1 manual (EPA 600/7-78-201).
4.5 GC/MS Screen: The extract, as prepared 1n paragraph 4.4.1, 1s
subjected to GC/MS analysis for each of the Individual compounds of Interest.
The GC/MS procedure 1s described 1n Chapter Four,. Method 8270. The extract 1s
screened at the MDL of each compound. The presence- of any compound at a
concentration >25 ug/mL 1n the concentrated extract will require the XAD-2 to
be recleaned by repeating the methylene chloride step.
4.6 Methodology for residual gravimetric determination; After the TCO
value and GC/MS data are obtained for the resin batch by the above procedures,
dry the remainder of the extract 1n a tared vessel. There must be <0.5 mg
residue registered or the batch of resin will have to be extracted with fresh
methylene chloride again until It meets this criterion. This level
corresponds to 25 ug/g In the XAD-2, or about 3.25 mg 1n a resin charge of
130 g.
0010 - A - 6
Revision 0
Date September 1986
-------
METHOD 0010, APPENDIX B
TOTAL CHROMATOGRAPHABLE ORGANIC MATERIAL ANALYSIS
1.0 SCOPE AND APPLICATION
1.1 In this procedure, gas chromatography is used to determine the
quantity of lower boiling hydrocarbons (boiling points between 90* and 300*C)
in the concentrates of all organic solvent rinses, XAD-2 resin and LC
fractions - when Method 1 is used (see References, Method 0010) - encountered
in Level 1 environmental sample analyses. Data obtained using this procedure
serve a twofold purpose. First, the total quantity of the lower boiling
hydrocarbons in the sample is determined. Then whenever the hydrocarbon
concentrations in the original concentrates exceed 75 ug/m3, the
chromatography results are reexamined to determine the amounts of individual
species.
The extent of compound identification is limited to representing all
materials as normal alkanes based upon comparison of boiling points. Thus the
method is not qualitative. In a similar manner, the analysis is
semlquantitative; calibrations are prepared using only one hydrocarbon. They
are replicated but samples routinely are not.
1.2 Application; This procedure applies solely to the Level 1 C7-C16
gas chromatographic analysis of concentrates of organic extracts, neat
liquids, and of LC fractions. Throughout the procedure, it is assumed the
analyst has been given a properly prepared sample.
1.3 Sensitivity; The sensitivity of this procedure, defined as the
slope of a plotof response versus concentration, is dependent on the
Instrument and must be verified regularly. TRW experience indicates the
nominal range is of the order of 77 uV'V-sec-uL/ng of n-heptane and 79
uV-sec-ul/ng of n-hexadecane. The Instrument is capable of perhaps one
hundredfold greater sensitivity. The level specified here is sufficient for
Level 1 analysis.
1.4 Detection limit; The detection limit of this procedure as written
is 1.3 ng/uL for a1uT Injection of n-decane. This limit is arbitrarily
based on defining the minimum detectable response as 100 uv-sec. This is an
easier operational definition than defining the minimum detection limit to be
that amount of material which yields a signal twice the noise level.
1.5 Range; The range of the procedure will be concentrations of 1.3
ng/uL and greater.
1.6 Limitations
1.6.1 Reporting limitations: It should be noted that a typical
environmental sample will contain compounds which: (a) will not elute in
the specified boiling ranges and thus will not be reported, and/or (b)
0010 - B - 1
Revision 0
Date September 1986
-------
will not elute from the column at all and thus will not be reported.
Consequently, the organic content of the sample as reported 1s a lower
bound and should be regarded as such.
1.6.2 Calibration limitations: Quant1tat1on 1s based on
calibration with n-decane. Data should therefore be reported as, e.g.,
mg C8/m3 as n-decane. Since response varies linearly with carbon number
(over a wide range the assumption may Involve a 20% error), 1t 1s clear
that heptane (C7) detected 1n a sample and quantltated as decane will be
overestimated. Likewise, hexadecane (C16) quantltated as decane will be
underestimated. From previous data, 1t 1s estimated the error Involved
1s on the order of 6-7%.
1.6.3 Detection limitations: The sensitivity of the flame
1on1zat1on detector varies from compound to compound. However, n-alkanes
have a greater response than other classes. Consequently, using an ri-
al kane as a callbrant and assuming equal responses of all other compounds
tends to give low reported values.
2.0 SUMMARY OF METHOD
2.1 A ml aliquot of all 10-mL concentrates 1s disbursed for GC-TCO
analysis. With boiling point-retention time and response-amount calibration
curves, the data (peak retention times and peak areas) are Interpreted by
first summing peak areas 1n the ranges obtained from the boiling point-
retention time calibration. Then, with the response-amount calibration curve,
the area sums are converted to amounts of material 1n the reported boiling
point ranges.
2.2 After the Instrument Is set up, the boiling point-retention time
calibration 1s effected by Injecting a mixture of n-C7 through n-C16
hydrocarbons and operating the standard temperature program. Response-
quantity calibrations are accomplished by Injecting n-decane in n-pentane
standards and performing the standard temperature program.
2.3 Definitions
2.3.1 GC: Gas chromatography or gas chromatograph.
2.3.2 C7-C16 n-alkanes:, Heptane through hexadecane.
2.3.3 GCA temperature program: 4 m1n Isothermal at 60*C, !0*C/m1n
from 60* to 220*C.
2.3.4 TRW temperature program: 5 m1n Isothermal at room
temperature, then program from 30*C to 250°C at !5°C/m1n.
3.0 INTERFERENCES
Not applicable.
0010 - B - 2
Revision 0
Date September 1986
-------
4.0 APPARATUS AND MATERIALS
4.1 Gas chromatpgraph; This procedure 1s Intended for use on a Varlan
1860 gas chromatograph, equipped with dual flame 1on1zat1on detectors and a
linear temperature programmer. Any equivalent Instrument can be used provided
that electrometer settings, etc., be changed appropriately.
4.2 Gases;
4.2.1 Helium: Minimum quality 1s reactor grade. A 4A or 13X
molecular sieve drying tube 1s required. A filter must be placed between
the trap and the Instrument. The trap should be recharged after every
third tank of helium.
4.2.2 A1r: Zero grade 1s satisfactory.
4.2.3 Hydrogen: Zero grade.
4.3 Syringe; Syringes are Hamilton 701N, 10 uL, or equivalent.
4.4 Septa; Septa will be of such quality as to produce very low bleed
during the temperature program. An appropriate septum 1s Supelco Mlcrosep
138, which 1s Teflon-backed. If septum bleed cannot be reduced to a
negligible level, 1t will be necessary to Install septum swingers on the
Instrument.
4.5 Recorder: The recorder of this procedure must be capable of not
less than 1 mV full-scale display, a 1-sec time constant and 0.5 1n. per m1n
chart rate.
4.6 Integrator; An Integrator 1s required. Peak area measurement by
hand 1s satlsfactory but too time-consuming. If manual Integration 1s
required, the method of "height times width at half height" 1s used.
4.7 Columns;
4.7.1 Preferred column: 6 ft x 1/8 1n. O.D. stainless steel column
of 10% OV-101 on 100/120 mesh Supelcoport.
4.7.2 Alternate column: 6 ft x 1/8 1n. O.D. stainless steel column
of 10% OV-1 (or other silicon phase) on 100/120 mesh Supelcoport.
4.8 Syringe cleaner; Hamilton syringe cleaner or equivalent connected
to a suitable vacuum source.
5.0 REAGENTS
5.1 Pentane; "D1st1lled-1n-Glass" (reg. trademark) or "Nanograde" (reg.
trademark) for standards and for syringe cleaning.
0010 - B - 3
Revision
Date September 1986
-------
5.2 Methylene chloride: "D1st1lled-1n-Glass" (reg. trademark) or
"Nanograde" (reg. trademark) for syringe cleaning.
6.0 SAMPLING HANDLING AND PRESERVATION
6.1 The extracts are concentrated in a Kuderna-Danish evaporator to a
volume less than 10 mL. The concentrate is then quantitatively transferred to
a 10-mL volumetric flask and diluted to volume. A 1-mL aliquot is taken for
both this analysis and possible subsequent GC/MS analysis and set aside in the
sample bank. For each GC-TCO analysis, obtain the sample sufficiently in
advance to allow it to warm to room temperature. For example, after one
analysis is started, return that sample to the sample bank and take the next
sample.
7.0 PROCEDURES
7.1 Setup and checkout: Each day, the operator will verify the
following:
7.1.1 That supplies of carrier gas, air and hydrogen are
sufficient, i.e., that each tank contains > 100 psig.
7.1.2 That, after replacement of any gas cylinder, all connections
leading to the chromatograph have been leak-checked.
7.1.3 That the carrier gas flow rate is 30+2 mL/min, the hydrogen
flow rate is 30+2 mL/min, and the air flow rate is 300 + 20 mL/min.
7.1.4 That the electrometer is functioning properly.
7.1.5 That the recorder and integrator are functioning properly.
7.1.6 That the septa have been leak-checked (leak-checking is
effected by placing the soap bubble flow meter inlet tube over the
injection port adaptors), and that no septum will be used for more than
20 injections.
7.1.7 That the list of samples to be run is ready.
7.2 Retention time calibration:
7.2.1 To obtain the temperature ranges for reporting the results of
the analyses, the chromatograph is given a normal boiling point-retention
time calibration. The n-alkanes, their boiling points, and data
reporting ranges are given in the table below:
0010 - B - 4
Revision
Date September 1986
-------
NBP.'C Reporting Range.*C Report As
n-heptane 98 90-110 C7
n-octane 126 110-140 C8
n-nonane 151 140-160 C9
n-decane 174 160-180 CIO
n-undecane 194 180-200 Cll
n-dodecane 214 200-220 C12
n-tr1decane 234 220-240 C13
n-tetradecane 252 240-260 C14
n-pentadecane 270 260-280 C15
n-hexadecane 288 280-300 C16
7.2.2 Preparation of standards: Preparing a mixture of the C7-C16
alkanes 1s required. There are two approaches: (1) use of a standards
kit (e.g., Polysclence Kit) containing bottles of mixtures of selected n-
alkanes which may be combined to produce a C7-C16 standard; or (2) use of
bottles of the Individual C7-C16 alkanes from which accurately known
volumes may be taken and combined to give a C7-C16 mixture.
7.2.3 Procedure for retention time calibration: This calibration
1s performed at the start of an analytical program;, the mixture 1s
chromatographed at the start of each day. To attain the required
retention time precision, both the carrier gas flow rate and the
temperature program specifications must be observed. Details of the
procedure depend on the Instrument being used. The general procedure 1s
as follows:
7.2.3.1 Set the programmer upper limit at 250*C. If this
setting does not produce a column temperature of 250*C, find the
correct setting.
7.2.3.2 Set the programmer lower limit at 30*C.
7.2.3.3 Verify that the Instrument and samples are at room
temperature.
7.2.3.4 Inject 1 uL of the n-alkane mixture.
7.2.3.5 Start the Integrator and recorder.
7.2.3.6 Allow the Instrument to run Isothermally at room
temperature for five m1n.
7.2.3.7 Shut the oven door.
7.2.3.8 Change the mode to Automatic and start the temperature
program.
7.2.3.9 Repeat Steps 1-9 a sufficient number of times so that
the relative standard deviation of the retention times for each peak
1s <5%.
0010 - B - 5
Revision 0
Date September 1986
-------
7.3 Response calibration;
7.3.1 For the purposes of a Level 1 analysis, response-quantity
calibration with n-decane 1s adequate. A 10-uL volume of n-decane 1s
Injected Into a tared 10 ml volumetric flask. The weight Injected 1s
obtained and the flask 1s diluted to the mark with n-pentane. This
standard contains about 730 ng n-decane per uL n-pentane. The exact
concentration depends on temperature, so that a weight 1s required. Two
serial tenfold dilutions are made from this standard, giving standards at
about 730, 73, and 7.3 ng n-decane per uL n-pentane, respectively.
7.3.2 Procedure for response calibration: This calibration 1s
performed at the start of an analytical program and monthly thereafter.
The most concentrated standard 1s Injected once each day. Any change 1n
calibration necessitates a full calibration with new standards.
Standards are stored 1n the refrigerator locker and are made up monthly.
7.3.2,1 Verify that the Instrument 1s set up properly.
7.3.2.2 Set electrometer at 1 x 10'10 A/mV.
7.3.2.3 Inject 1 uL of the highest concentration standard.
7.3.2.4 Run standard temperature program as specified above.
7.3.2.5 Clean syringe.
7.3.2.6 Make repeated Injections of all three standards until
the relative standard deviations of the areas of each standard are
7.4 Sample analysis procedure;
7.4.1 The following apparatus Is required;
7.4.1.1 Gas chromatograph set up and working.
7.4.1.2 Recorder, Integrator working.
7.4.1.3 Syringe and syringe cleaning apparatus.
7.4.1.4 Parameters; Electrometer setting Is 1 x 10~10 A/mV;
recorder Is set at 0.5 1n./m1n and 1 mV full-scale.
7.4.2 Steps 1n the procedure are;
7.4.2.1 Label chromatogram with the data, sample number, etc.
0010 - B - 6
Revision
Date September 1986
-------
7.4.2.2 Inject sample.
7.4.2.3 Start Integrator and recorder.
7.4.2.4 After Isothermal operation for 5 min, begin
temperature program.
7.4.2.5 Clean syringe.
7.4.2.6 Return sample; obtain new sample.
7.4.2.7 When analysis 1s finished, allow Instrument to cool.
Turn chromatogram and Integrator output and data sheet over to data
analyst.
7.5 Syringe cleaning procedure:
7.5.1 Remove plunger from syringe.
7.5.2 Insert syringe Into cleaner; turn on aspirator.
7.5.3 Fill pipet with pentane; run pentane through syringe.
7.5.4 Repeat with methylene chloride from a separate pi pet.
7.5.5 Flush plunger with pentane followed by methylene chloride.
7.5.6 Repeat with methylene chloride.
7.6 Sample analysis decision criterion; The data from the TCO analyses
of organic extract and rinse concentrates are first used to calculate the
total concentration of C7-C16 hydrocarbon-equivalents (Paragraph 7.7.3) in the
sample with respect to the volume of air actually sampled, I.e., ug/m3. On
this basis, a decision 1s made both on whether to calculate the quantity of
each n-alkane equivalent present and on which analytical procedural pathway
will be followed. If the total organic content is great enough to warrant
continuing the analysis -- >500 ug/m3 -- a TCO of less than 75 ug/m3 will
require only LC fractionation and gravimetric determinations and IR spectra to
be obtained on each fraction. If the TCO is greater than 75 ug/m3, then the
first seven LC fractions of each sample will be reanalyzed using this same gas
chromatographic technique.
7.7 Calculations;
7.7.1 . Boiling Point - Retention Time Calibration: The required
data for this calibration are on the chromatogram and on the data sheet.
The data reduction 1s performed as follows:
7.7.1.1 Average the retention times and calculate relative
standard deviations for each n-hydrocarbon.
0010 - B - 7
Revision
Date September 1986
-------
7.7.1.2 Plot average retention times as abscissae versus
normal boiling points as ordlnates.
7.7.1.3 Draw 1n calibration curve.
7.7.1.4 Locate and record retention times corresondlng to
boiling ranges 90-100, 110-140, 140-160, 160-180, 180-200, 200-220,
220-240, 240-260, 260-280, 280-300'C.
7.7.2 Response-amount calibration: The required data for this
calibration are on the chromatogram and on the data sheet. The data
reduction 1s performed as follows:
7.7.2.1 Average the area responses of each standard and
calculate relative standard deviations.
7.7.2.2 Plot response (uvsec) as ordlnate versus ng/uL as
abscissa.
7.7.2.3 Draw 1n the curve. Perform least squares regression
and obtain slope (uV-sec-uL/ng).
7.7.3 Total C7-C16 hydrocarbons analysis: The required data for
this calculation are on the chromatogram and on the data sheet. The data
reduction is performed as follows:
7.7.3.1 Sum the areas of all peaks within the retention time
range of Interest.
7.7.3.2 Convert this area (uV-sec) to ng/uL by dividing by the
weight response for n-decane (uV*sec.uL/ng).
7.7.3.3 Multiply this weight by the total concentrate volume
(10 ml) to get the weight of the C7-C16 hydrocarbons In thersample.
7.7.3.4 Using the volume of gas sampled or the total weight of
sample acquired, convert the result of Step 7.7.3.3 above to ug/m3.
7.7.3.5 If the value of total C7-C16 hydrocarbons from Step
7.7.3.4 above exceeds 75 ug/m3, calculate Individual hydrocarbon
concentrations 1n accordance with the Instructions 1n Paragraph
7.7.5.5 below.
7.7.4 Individual C7-C16 n-Alkane Equivalent Analysis: The required
data from the analyses are on the chromatogram and on the data sheet.
The data reduction 1s performed as follows:
7.7.4.1 Sum the areas of peaks In the proper retention time
ranges.
0010 - B - 8
Revision
Date September 1986
-------
7.7.4.2 Convert areas (uV-sec) to ng/uL by dividing by the
proper weight response (uV-sec-uL/ng).
7.7.4.3 Multiply each weight by total concentrate volume (10
ml) to get weight of species 1n each range of the sample.
7.7.4.4 Using the volume of gas sampled on the total weight of
sample acquired, convert the result of Step 7.7.4.3 above to ug/m3.
8.0 QUALITY CONTROL
8.1 Appropriate QC 1s found 1n the pertinent procedures throughout the
method.
9.0 METHOD PERFORMANCE
9.1 Even relatively comprehensive error propagation analysis 1s beyond
the scope of this procedure. With reasonable care, peak area reproduc1bH1ty
of a standard should be of the order of 1% RSD. The relative standard
deviation of the sum of all peaks 1n a fairly complex waste might be of the
order of 5-10%. Accuracy Is more difficult to assess. With good analytical
technique, accuracy and precision should be of the order of 10-20%.
10.0 REFERENCES
1. Emissions Assessment of Conventional Stationary Combustion Systems:
Methods and Procedure Manual for Sampling and Analysis, Interagency
Energy/Environmental R&D Program, Industrial Environmental Research
Laboratory, Research Triangle Park, NC 27711, EPA-600/7-79-029a, January 1979.
0010 - B - 9
Revision
Date September 1986
-------
METHOD 0020
SOURCE ASSESSMENT SAMPLING SYSTEM (SASS)
1.0 PRINCIPLE AND APPLICATION
1.1 Principle
1.1.1 Particulate and semi volatile organic materials are withdrawn
from a source at a constant rate near 1sok1net1c conditions and are
collected 1n a multlcomponent sampling train.
1.1.2 Three heated cyclones and a heated high-efficiency fiber
filter remove and collect the particulate material from the sample and a
packed bed of porous polymeric resin adsorbs the condenslble organic
vapors.
1.1.3 Chemical analyses of the sample are conducted to determine
the concentration and Identity of the semi volatile organic species and
gravimetric determinations are performed to approximate particulate
emissions.
1.2 Application; This method is applicable to the preparation of
semiquantitative estimates (within a factor of three) of the amounts and types
of semi volatile organic and particulate materials that are discharged from
incineration systems.
2.0 APPARATUS
2.1 Sampling Train; A schematic of the sampling train used in this
method is given 1n Figure 1. This sampling train configuration 1s that of the
Source Assessment Sampling System (SASS), as supplied by the manufacturer.
Basic operating and maintenance procedures are described in the "Operating and
Service Manual: Source Assessment Sampling System" supplied on purchase of the
sampling system (Blake, 1977). Users should refer to this document and
adopt, but not limit themselves to, Its operating and maintenance procedures.
The SASS train components and specifications are detailed below.
2.1.1 Probe nozzles: The probe nozzles are constructed of Type 316
seamless stainless steel tubing and have sharp leading edges. The
nozzles are a hybrid elbow/buttonhook design, obtainable in diameters
ranging from 0.31 to 1.91 cm (1/8 to 3/4 in.), and are interchangeable.
Each nozzle should be calibrated according to the procedure outlined in
Paragraph 7.2 of this method.
0020 - 1
Revision 0
Date September 1986
-------
Hut Controller
o
o
r\j
o
O 73
0> (0
rt- <
n -*
CO
— *•
c/i o
n 3
n
Stack T.C.
Convection Oven
Isolation Ball Valve
e Filter
Gas Cooler
Imp/Cooler Trace
Element Collector
Conrlensate Collector
Dry Gas Meter/Orifice Meter
Centralized Temperature
and Pressure Readout
Control Module
Impinger T.C.
Two 10 f|3/min Vacuum Pumps
<£>
00
Figure 1. SASS Schematic Diagram.
-------
2.1.2 Probe Liner:
2.1.2.1 The probe Uner 1s also constructed of Type 316
seamless stainless steel tubing; attached to the liner Is a
proportional temperature controller capable of maintaining the Uner
surface temperature at 204 + 20*C (400 + 36'F) during sampling. The
use of the proportional controller to control the liner surface
temperature at the control module Is preferred because the oven
often cannot be reached for adjustment during sampling.
2.1.2.2 It should be noted that the measurement of the probe
Uner surface temperature 1s not an accurate measurement of the
Internal gas stream temperature, which is the temperature of
Interest. This source of error 1s caused by the temperature
gradient that exists between the inner and outer walls of the liner.
Monitoring of the actual gas stream temperature is impractical with
the SASS trains as presently constructed. It is suggested that a
one-time calibration be conducted in which the internal gas stream
temperature is compared to the liner surface at various temperatures
and at the standard SASS flow rate of 4.0 scfm.
2.1.2.3 The probe and probe Uner can withstand points up to
370*C (700*F), at which temperature they will soften. However,
stack temperatures greater than 288*C (550*F) may result 1n gas
temperatures at the 10-um cyclone inlet greater than the recommended
204*C (400*F) and hence require the use of a special water- or
forced-air-cooled probe.
2.1.3 PItot tubes: The pi tot tubes are Type S, designed to meet
the specifications of EPA Method 2 (see Reference below); these are
attached to the probe sheath to allow constant monitoring of the stack
gas velocity. The point of attachment to the sheath is such that the
Impact (high pressure) opening plane of the pi tots is level with or above
the sampling nozzle entry plane, as required by Method 2, to eliminate
nozzle Interference 1n velocity measurements. If calibration is not
required, the pltot tubes are assigned a nominal coefficient of 0.84, as
described 1n the calibration section of this method.
2.1.4 Differential pressure gauges: Three Magnehelic-type gauges
are used. One gauge (0 to 0.5 1n. H20) monitors the pressure drop across
the orifice meter (AHj); the other two gauges (0 to 0.5 and 0 to 4.0 1n.
H20) are connected in parallel and indicate the pressure differential
across the pltot tubes used for measuring stack gas velocity.
2.1.5 Filter holder and filter support: The filter holder and
filter support screen are constructed of Type 316 stainless steel with a
0020 - 3
Revision 0
Date September 1986
-------
Teflon gasket providing an airtight seal around the circumference. The
holder 1s attached Immediately to the outlet of the 1-um cyclone or the
cyclone bypass.
2.1.6 Cyclone/Filter heating system: The cyclone/filter heating
system 1s an Insulated double-walled oven, capable of maintaining the
temperature 1n the area of the cyclones and filter holder around the
recommended 204'C (400*F). A chrbmel-alumel thermocouple for temperature
sensing allows feedback control of the temperature to within
approximately 10%.
2.1.7 Cyclone train: The cyclone train consists of three cyclone
separators 1n series, having nominal particle-size cutoff diameters of
10, 3, and 1 urn respectively. The material of construction 1s Type 316
stainless steel with Teflon gaskets sealing the hoods and collector cups.
The compact design of the 10-um cyclone 1s achieved by Incorporating
flow-Interrupting vanes 1n the collection cup.
2.1.8 Organic module: The organic module consists of a th1n-f1lm
heat exchanger/gas cooler, a sorbent cartridge, and a condensate
collection trap. The temperature of the heat exchanger fluid Is
regulated by activating an Immersion heater or routing the coolant
through another heat exchanger 1n the 1mp1nger 1ce water bath. Water
from the 1mp1nger bath 1s continually circulated through the Inner
reservoir of the gas cooler for additional cooling capacity. The sorbent
cartridge encloses the polymeric adsorbent bed 1n a cylinder covered on
both ends by 80-mesh, Type 316 stainless steel wire cloth. The cartridge
holds approximately 150 grams of XAD-2 adsorbent resin. Condensed
moisture from the gas stream 1s collected 1n a reservoir located directly
beneath the packed sorbent bed. The drain valve of the reservoir should
be coupled with a Teflon line to an appropriately sized (1- to 5-l1ter)
glass storage container, as the capacity of the reservoir will typically
be exceeded during a run.
2.1.9 Implnger train: The four 1mp1ngers have a capacity of
approximately 3 liters "each and are constructed of pyrex glass. The caps
are Teflon with stainless 'steel fittings and the 1mp1ngers are
Interconnected by flexible Teflon or stainless steel tubing. The first
two 1mp1ngers are equipped with splash guards to minimize fluid carryover
and the last Implnger with a thermocouple mounted In the cap for
monitoring the Implnger train exit gas temperature.
2.1.10 Pump/Metering system: Two leak-free vane-type vacuum pumps
connected 1n parallel are used to maintain the 4.0-scfm flow 1n the
sampling system. Vacuum and differential pressure gauges, thermocouples
capable of measuring temperature to within 3*C (5.4'F) and a dry gas
meter capable of measuring volume to within 2% are supplied as the other
necessary components for maintaining 1sok1net1c sampling rates.
0020 - 4
Revision 0
Date September 1986
-------
2.1.11 Barometer: An aneroid barometer, or other barometer capable
of measuring atmospheric pressure to within 2.5 mm Hg (0.5 1n. Hg), 1s
required, unless the barometric reading 1s obtained from a nearby
National Weather Station; the station value (I.e., the absolute
barometric pressure) must be corrected for elevation differences between
the weather station and the sampling point. The corrected value should
reflect a decrease of 2.5 mm Hg (0.1 1n. Hg) per 30-m (100-ft) elevation
Increase, and vice versa for elevation decrease. (See Paragraph 7.7).
2.1.12 Gas density determination apparatus: The length of a SASS
run 1s typically sufficient to determine the average gas stream density
during the run. EPA Method 3 should be consulted for detailed
specifications for an Integrated fixed gas sampling system. Analysis of
the collected samples should be performed with an ORSAT analyzer or a
GC/TCD system outfitted specifically for this purpose.
2.1.13 Calibration/Field preparation log: For documentation of
calibration and preparation procedures, a permanently bound laboratory
notebook 1s recommended, In which carbon copies are made of the data as
they are being recorded. The carbon copies should be detachable and used
only for separate storage 1n the test program archives.
2.2 Sample recovery;
2.2.1 Probe liner brush: The brush must have nylon bristles, a
stainless steel wire handle, and extensions of stainless steel, Teflon,
or other Inert material. The combined extensions must be equal to or
greater than the length of the probe.
2.2.2 Probe nozzle brush: The brush must have nylon bristles and a
stainless steel wire handle, and be properly sized and shaped for
cleaning the Inner surfaces of the nozzle.
2.2.3 Cyclone and filter holder brushes: The brushes must have
nylon bristles and a stainless steel wire handle, and be properly sized
for cleaning the Inner walls of these components. It 1s strongly
recommended that a separate brush be used for sample recovery from each
of these components to avoid cross contamination of one particle size
fraction by another.
2.2.4 Wash bottles: Three are needed. Teflon or glass 1s required
to avoid contamination of organic solvents; Teflon 1s preferred because
1t 1s unbreakable.
2.2.5 Glass sample storage containers: The containers must be
chemically resistant, boroslHcate glass bottles, 500-mL or 1,000-mL.
Screw-cap liners should be Teflon or constructed so as to be leak-free
and resistant to chemical attack by organic recovery solvents (narrow-
mouth glass bottles have been found to exhibit less tendency toward
leakage).
0020 - 5
Revision 0
Date September 1986
-------
2.2.6 Petrl dishes: These must be glass and sealed around the
circumference with Teflon tape for storage and transport of filter
samples.
2.2.7 Graduated cylinder and triple-beam balance: to measure
condensed water to the nearest 1 ml or 0.5 g. Graduated cylinders must
have subdivisions no greater than 2 ml. Equipment made of glass must be
used for measuring the volume of any solution that will be subject to
organic analysis. Laboratory triple-beam balances must be capable of
weighing to +0.5 g or better.
2.2.8 High-density linear polyethylene (HDLP) storage containers:
These are used for storage of the 1mp1ngers.
2.2.9 Plastic storage containers: Airtight containers are
necessary for storage of silica gel.
2.2.10 Funnels: Glass funnels must be used 1n recovering samples
for organic analysis. Glass or plastic funnels may be used in
other processes but care must be taken to segregate the two types.
3.0 REAGENTS AND MATERIALS
3.1 Filters; Glass fiber filters, 15.24 cm (6.0 In.) 1n diameter
without organic binder, exhibiting 99.95% efficiency (<0.05% penetration) on
0.3-m1cron dioctyl phthalate smoke particles, conforming to the specifications
outlined in ASTM Standard Method D2986-71. Test data from the supplier's
quality control program are sufficient for this purpose. The filter material
must also be unreactive to S02 and $03.
3.2 Adsorbent resin; Porous polymeric resin, XAD-2, 1s used. The resin
must be cleaned prior to use. The resin must not exhibit a blank higher than
4 mg/kg of total chromatographable organics (TCO) prior to use. Once cleaned,
the resin should be stored in a wide-mouth amber glass container and the
headspace purged with nitrogen to limit exposure to ambient air. Resin should
be used within 2 wk of preparation.
3.3 Silica gel; Indicating type, 6 to 16 mesh. If previously used, dry
at 175*C (350°F) for 2 hr. New silica gel may be used as received.
3.4 Impinger solutions; Since the impinger solutions are typically used
for the determination of gas-stream water-vapor content, Type II water should
be used. If specific inorganic species are to be determined (e.g.,
hydrochloric acid when burning chlorinated organic material), then other
appropriate collecting solutions (in the above example, dilute base) must be
used.
3.5 Crushed ice; Commercially available. Quantities ranging from 50 to
100 Ib may be necessary during a run, depending upon ambient air temperatures.
0020 - 6
Revision 0
Date September 1986
-------
3.6 Methanpl/Methylene chloride: D1stnied-1n-g1ass or pesticide-grade
methanol and methylene chloride are required.
4.0 SAMPLING PROCEDURE
4.1 Sampling equipment calibration;
4.1.1 The probe tips, pltot tubes, dry gas meter, thermocouples,
and any thermometers must be calibrated before and after each field
sampling trip according to the procedures outlined 1n APTD-0576 (Rom,
1972) and below 1n Section 7.0. During extended sampling trips where
the train will routinely be used more than 10 times, It 1s strongly
recommended that a calibrated orifice, a set of micrometers (Vernier
calipers), and a standard mercury-ln-glass thermometer accompany the
train to verify that the calibrations of the dry gas meter, probe
nozzles, and thermocouples, respectively, have not changed significantly
(more than +2%). The aneroid barometer should be calibrated on a dally
basis against a mercury barometer when 1n the laboratory and periodically
1n the field by consulting the local weather station and correcting for
elevation (see Paragraph 7.7).
4.2 Laboratory preparation;
4.2.1 Weigh several 700- to 800-g portions of silica gel 1n
airtight containers to +0.1 g. Record the weight of the silica gel plus
the container on the container and 1n a field sampling preparation
notebook.
4.2.2 Holding with blunt-tipped tweezers, check filters visually
against light for Irregularities, flaws, or plnhole leaks. Label the
shipping containers (glass PetH dishes) and keep the filters In these
containers at all times except during sampling and weighing. The filters
themselves need not be labeled 1f strict compliance with the above
Instruction Is ensured. Desiccate the filters 1n a desiccator over
Drlerlte or silica gel with the Petrl dishes open at 20 + 5.6*C (68 +
10'F) and ambient pressure for at least 24 hr and weigh. Thereafter
weigh at 6-hr (minimum) Intervals to a constant weight, I.e., previous
weight +0.5 mg; record the weight to the nearest 0.1 mg, along with the
date and time, 1n the field sampling preparation notebook.
Alternatively, the filters may be oven-dried at 105*C (220'F) for 2 to 3
hr, desiccated for 2 hr, weighed, and weighed thereafter at 6-hr
Intervals to a constant weight. During each weighing, the filter must
not be exposed to the laboratory atmosphere for longer than 2 m1n with a
relative humidity greater than 50%.
4.2.3 Passlvate all SASS train parts and sample storage containers,
referring to the procedure that appears In Figure 2, adapted from Level I
requirements. Passivation Is required of all new train components and
sample storage containers before their Initial use 1n the field.
Thereafter, passivation should be conducted every 6 months when the
frequency of tests 1s once per month or less, and every 3 months when the
0020 - 7
Revision 0
Date September 1986
-------
LMflAt I
CtiANMM I
ffiMflMJflflJ
UNNMMMMI
AMOAUIAMfll
•amctit
OIAM AIL MNM AMAt Mtl
AMftllYL AlC =3
•a
MVWMIAmnilft
tlMAMOfA*
INMMMIM HIMtMUt
Id CONIAMWttnQM
•AtWftNM AND
CAf
-------
frequency 1s between once per week and once per month. If testing 1s
more frequent, passivation should be conducted proportionately more
often. Whenever corrosion has occurred, the corrosion must be removed
and the passivation repeated. The passivation and rinse solutions should
be replaced every fourth use, or discarded weekly.
4.2.4 Prepare recycled sample containers by detergent washing
(using a stiff nylon brush where necessary), followed by rinsing with
Type II water, methanol and methylene chloride. As each part 1s treated
with the final solvent, dry with filtered air or dry nitrogen and Inspect
for any contaminating residue. Discard any container exhibiting visual
contamination. Cover all open surfaces with aluminum foil or Teflon
film, using elastic bands to secure.
4.2.5 Assembly and leak-checking of the entire train 1n the
laboratory 1s highly recommended to reveal the need for replacement of
gaskets or defective components. The leak-check procedure 1s described
In Paragraph 4.4.3.11. Substitution of V1ton-A gaskets for Teflon may
facilitate meeting the allowable leak rate. A length of Teflon tape
stretched around the circumference of each flanged connection underneath
the ring clamp also greatly reduces Inward air leakage.
4.3 Preliminary field determinations;
4.3.1 Select the sampling site and remove any accumulated scale and
corrosion from the sampling portholes. Determine the stack static
pressure, temperature, and velocity profile using EPA Method 2 (see
References); a leak-check of the pltot lines prior to conducting these
measurements 1s highly recommended. Approximate the moisture content
using EPA Method 4 (Approximation Method) or alternate means such as
drying tubes, wet bulb/dry bulb or condensation techniques,
stolchlometrlc calculations, or previous experience. Determine the dry
molecular weight of the stack gas by performing an ORSAT or GC/TCD
analysis for CO, C02, QZ and N2 on an average of three grab samples taken
from either the center of the duct or a point no closer to the stack
walls than 1.0 m (3.3 ft).
4.3.2 Select a nozzle size based upon the calculations below,
ensuring that 1t will not be necessary to change the nozzle during the
sampling run to maintain near-1sok1net1c sampling rates.
4.3.2.1 To calculate the required nozzle diameter, first
calculate the Average Stack Gas Velocity;
- K C fJTAPT)
Vpm jjavg PCMC
5 S
where:
(V$) = Average stack gas velocity, ft/sec;
0020 - 9
Revision
Date September 1986
-------
ft
sec
' (lb/lb-mole)(1n. Hg)
•R(1n. H20)
1/2 '
AP
(Ts)avg
PS
MS
= 85.48
P1tot tube coefficient, dlmenslonless;
Velocity head of stack gas, 1n. H20;
Average stack gas temperature, *R;
Absolute stack gas pressure, 1n. Hg; and
Molecular weight of stack gas (wet basis), Ib/lb-mole.
4.3.2.2 Then calculate the required Nozzle Diameter (Dn):
Dn = 0.831
^ s'avg
-------
Pm = Absolute meter pressure, 1n. Hg, calculated by
p - p , (AH)est.avq.
Km ' Kb 13.6
where:
(AH)est ava = Estimated average AH across orifice,
3-4 1n. H20, and
Pb = Barometric pressure (corrected), 1n. Hg; and
Tst = Standard temperature, 528*R.
None of these definitions has an English/metric equivalent.
4.3.3.2 Using this result, obtain the approximate sampling
time by dividing the required sample volume by the estimated
sampling flowrate.
4.3.4 Finally, calculate the Orifice Pressure Drop needed to
maintain near-isokinetic sampling conditions from the equation:
P
,n
°-1924 "
mo
AH1 = TT)
v nravg
where:
AHj = Required AH across the orifice, in. H20;
Pm = Absolute meter pressure, in. Hg (calculated the same way as
for Average Sampling Rate above);
(Tm)avg = Estimated average gas temperature at the dry gas meter, *R;
Jj = Orifice coefficient for orifice "1" (see Blake, 1977, and
Section 7.0 of this method for determining orifice
coefficients); and
D0j = Orifice diameter, in. (Information supplied upon purchase of
the SASS train; the largest diameter orifice is typically
best suited for the SASS sampling rate of 4.0 scfm).
4.3.5 It 1s desirable, but not required, to sample more than one
point of average velocity during a SASS run. Allocate equal Intervals of
the total sampling time estimated above to each sampling point chosen 1f
more than one point will be sampled.
0020 - 11
Revision
Date September 1986
-------
4.4 Preparation of collection train;
4.4.1 An Integral part of preparing the collection train 1s
securing sufficient electrical power to operate for an extended period of
time without Interruption. Three separate circuits — two 30-amp and one
20-amp — are required. It 1s highly recommended that one sampling pump
and one control box power cord (probe heater) be placed on one of the 30-
amp circuits, and the other sampling pump and control box power cord
(oven heater and temperature readout) be placed on the other 30-amp
circuit. The organic module coolant pump and temperature controller
should be placed on the smaller 20-amp circuit.
4.4.2 During assembly of the train, keep the Inner surfaces of each
component covered until 1t 1s Integrated Into the system and sampling 1s
about to begin. Fill the sorbent trap section of the organic module with
approximately 150 g of clean adsorbent XAD-2 resin. To avoid
contamination, the trap should be placed upon a clean surface (I.e.,
aluminum foil rinsed with methylene chloride and a1r-dr1ed) while
filling; gloves should be worn. Pack the trap uniformly to eliminate
potential channeling. Place 500 ml of Type II water or other appropriate
solution Into the first and second 1mp1ngers, leave the third 1mp1nger
empty, and place a prewelghed portion of silica gel Into the fourth.
NOTE: The choice of impinger solutions depends upon whether these will
be used to collect selected Inorganic species or simply to
condense water vapor from the gas stream to measure percent
moisture. For example, In an Incinerator combusting chlorinated
organic material, a solution of dilute base would typically be
used to collect hydrochloride add emissions.
Using blunt-tipped tweezers, place a tared filter Into the filter
holder. Ensure that the filter 1s centered and the gasket properly
placed to prevent the gas stream from circumventing the filter. On the
probe, mark the locations of the chosen sampling points with heat-
resistant tape or paint.
4.4.3 The stepwlse procedure for assembly of the train follows:
4.4.3.1 Place the oven on a table or rollers that will be used
as a support throughout the run.
4.4.3.2 Assemble the three cyclones, using the vortex breaker
supplied with the cyclone in the 10-um cyclone only. (To minimize
leaks throughout the system, a strip of Teflon tape should be
stretched around the circumference of each flanged seal and the ring
clamp placed over and secured.) Do not use the vortex breakers
supplied with the 3- and 1-urn cyclones 1n the 3- and 1-um cyclones.
Actual calibration data has shown that the use of the vortex
breakers 1n the two smaller cyclones may result in unreproduclble
particle-size cutoff diameters (the particle size at which 50%
collection efficiency is exhibited).
0020 - 12
Revision 0
Date September 1986
-------
4.4.3.3 Attach the filter holder to the outlet of the 1-um
cyclone and place the cyclones and filter holder together 1n the
oven. Preheat a spare filter holder containing a tared filter on
the oven floor; cover the holder openings with aluminum foil.
4.4.3.4 Attach the probe to the oven and to the 10-um cyclone.
4.4.3.5 Place the 1mp1ngers 1n the tray 1n the 1mp1nger case
and make the aproprlate Interconnections.
4.4.3.6 Connect the organic module Inlet to the filter housing
outlet and the organic module outlet to the first 1mp1nger Inlet.
4.4.3.7 Connect the vacuum pumps 1n parallel to the fourth
1mp1nger outlet.
4.4.3.8 Connect all temperature sensors and power lines to the
control unit. Check temperature Indicators and controllers at
ambient temperature.
4.4.3.9 Activate gas cooling system. Begin monitoring the
XAD-2 temperature. Always check coolant level before supplying
power. Ensure proper gas cooling system temperature before
proceeding.
NOTE: IT IS EXTREMELY IMPORTANT THAT THE XAD-2 RESIN TEMPERATURE
NEVER EXCEED 50*C, AS DECOMPOSITION WILL OCCUR. DURING
TESTING, THE XAD-2 TEMPERATURE MUST NOT EXCEED 20»C FOR
EFFICIENT CAPTURE OF THE SEMIVOLATILE ORGANIC SPECIES OF
INTEREST.
4.4.3.10 Heat oven and probe to 204*C (400*F).
4.4.3.11 Run gas flow leak-check. The following Instructions
will facilitate the leak-checking procedure:
a. Open the Isolation ball valve and plug the Inlet to the
probe with a rubber stopper or appropriate airtight cap.
b. Start the pumping system with the bypass valves fully open
and the coarse valves completely closed. Partially open the
coarse valves and slowly close the bypass valves until a vacuum
of 127 mm Hg (5 1n. Hg) Is reached. Do not reverse the
direction of the bypass valves as backflushing of the Implnger
solutions Into the organic module will result.If the desired
vacuum 1s exceeded, eitherleak-checkat the higher vacuum or
terminate the leak-check and begin again. Allow the system to
equilibrate and measure the leakage rate. The allowable leak
rate for the SASS train 1s 0.0014 m3/m1n (0.05 ft3/m1n) at this
vacuum. Close the Isolation ball valve and evacuate the train
to 281 mm Hg (15 1n. Hg). The leak rate through the back half
of the train should be less than 0.0014 m3/m1n (0.05 ft3/m1n)
at this vacuum.
0020 - 13
Revision 0
Date September 1986
-------
c. When the leak-check 1s complete, slowly remove the plug
& from the probe tip and then slowly open the Isolation ball
valve.
d. When the vacuum drops to 127 mm Hg (5 1n. Hg) or less,
Immediately close both coarse control valves together. Switch
off the pumping system and reopen the bypass valves. The
bypass valves should not be opened until the coarse valves have
been closed.
4.4.4 Only post-test leak-checks are mandatory; however, experience
has shown that pre-test leak-checks and leak-checks following component
changes are necessary to ensure that Invaluable sampling time 1s not
lost as a result of an oversight or defective component.
4.5 Sample Collection; Constant monitoring of train operations before,
during, and aftertHeparticulate run 1s essential 1n maintaining sample
Integrity. Listed below are sample collection guidelines:
4.5.1 With the coarse valves closed and bypass valves open, turn on
the vacuum pumps and allow them to warm up. As the probe and oven are
heating, prepare a SASS run data sheet as shown 1n 40 CFR Appendix A (see
References below). Barometric pressure data should be recorded at least
at the beginning and end of the run; once per hour 1s preferred.
4.5.2 When operating temperatures have been reached, place the
probe in the stack at the first designated sampling point, turn on the
vacuum pumps, adjust the sampling flowrate to achieve the calculated AH1,
and start the elapsed timer. If, however, the gas stream 1s under medium
or high negative pressure, it becomes extremely important to start the
vacuum pumps just before placing the probe in the gas stream, and to
continue to operate the pumps until Just after the probe has been removed
from the gas stream. This will eliminate the possibility of lifting of
the filter or backflushing of the filter and cyclone particulate catches
at any time.
4.5.3 Seal the sampling port around the probe to prevent
introduction of dilution air at this point. Record the clock time of the
start of the test.
4.5.4 Using the criteria outlined above under Paragraph 4.3,
Preliminary Field Determination, place the integrated fixed gas bag or
bulb sampling probe into the gas stream and begin sampling. Collect
three samples during the SASS run; record the initial and final clock
times of each Integrated fixed gas sample.
4.5.5 Monitor and maintain all temperatures and the calculated AH
and record the data at equal Intervals'of 10-15 min.
4.5.6 Add crushed ice to the implnger section and drain excess
water as necessary.
0020 - 14
Revision 0
Date September 1986
-------
4.5.7 Without Interrupting sampling, drain the condensate Initially
every 30-45 m1n, and afterward as necessary. Ensure that the vessel Into
which the reservoir 1s drained forms an airtight system with the
reservoir using a connecting Teflon line, and 1s placed well below the
level of the reservoir Itself. To drain the reservoir, close the
solatlon ball valve and open the drain valve. Allow the system to
evacuate for 10-20 sec. Carefully open the Isolation valve. The
condensate should siphon from the reservoir Into the storage vessel.
Close the drain valve when the siphoning action of the condensate ceases.
4.5.8 Replace the filter when 1t becomes Impossible to maintain
near-1sok1net1c sampling rates but not more frequently than every 20 to
30 mln. Always terminate and Initiate sampling by adjustment of the
coarse pump valves and then the bypass valves. A spare filter holder and
filter, 1f available, should be preheating 1n the oven at all times.
Conduct leak-checks before and after changing the filter. Recall
previous Instructions concerning removal and relntroductlon of the probe
Into the duct.
4.5.9 At the same time, check the 1-um cyclone reservoir for
remaining capacity, taking care not to contaminate the contents during
this Inspection.
4.5.10 When replacing a filter, start and stop the fixed gas
sampling concurrently with the SASS sampling; record the clock time and
dry-gas-meter reading whenever sampling 1s Interrupted.
4.5.11 Upon collection of the required 30 dscm (1,060 dscf), remove
the probe from the gas stream and shut down the pumps as previously
Instructed. Record the final dry-gas-meter reading and clock time;
turn off all heaters. Conduct the post-test leak-check when the
probe tip can be safely handled. Do not cap the probe while Initially
cooling, because this will create a vacuum Inside that will cause
disruption of the cyclone and filter particulate catches when 1t 1s
released. Instead, use aluminum foil to cover probe openings. Before
the probe 1s transported, secure the aluminum foil covers with elastic
bands. Leak-check the pi tot lines per EPA Method 2 to validate velocity
heat data.
5.0 SAMPLE RECOVERY
The sample handling and transfer procedures outlined 1n this section have
been adopted from the Level 1 procedures. The flow diagrammatic represen-
tation of the sample recovery procedures shown 1n Figures 3, 4, and 5 can be
found In the Level 1 Sampling and Analysis Procedures Manual.
5.1 Disassembly of the SASS Train; At the conclusion of the sampling
run, the train 1s disassembledan3transported to the prepared work area as
follows:
0020 - 15
Revision 0
Date September 1986
-------
• CN,DN MM,
!•>•* M •«•••>«)
|U» 1 T* ••« km* ««MMH
»•• milt *WM« MM ttmn
•llf t tot*M«1 ItMf Mf
M
T»» M< kM* MMMU ll«H
t lMM< U* MM*Mil
CN)O
•M ttratto IWMMI •» •
••»••» Itmr *•( i*tt*ttril
toi «MM
: CNlOM
Ml ItMHtci CWHtMl Mil (
•MUM Iwnr M» nutpch
M< •»•««. ••! CM}Cl} :
(HjON ^
(in I H«»»»»<1 tomt IM NMttttit
HMWM*Mfl
•«•»*« !«•«• ««»
M< IMMff CMIMV teM I
••W lt»0» IMMtMT
K«DOI tow* <•» MMttHll
ll>MM(
: CNjON
•««» 1 :
»•«. *W »
-------
INTACT SOUOENT MODULE
AFTER SAMPLING RUN
RELEASI CLAM* JOINING
SORBENT CARTRIDGE SEC-
TION TO THE UPPER CAS
CONDITIONING SECTION
REMOVt SOMf NT
TRIDCI FRO'/ HOLDER
REMOVE SCREEN FROM
TO'OF CARTRIDGE.
EMtTV RESIN INTO A
WIDE-MOUTH AMBER JAR
USING CLEAN FUNNEL
REMOVE CONOENSATE
RESERVOIR AMD DRAIN
eONDENSATE THROUGH
VALVE INTO THE
CONOENSATE DRAIN
CONTAINER USED TO
COLLECT CONDENSATE
DURING THE SASS RUN
RINSE SCREEN AND CAR
TRIDGE INTO RESIN
CONTAINER WITH CHjClj
CLOSE CONDENSATE VALVE
AND REASSEMILE TO MOD
ULi TO COLLECT WASHINGS
MEASURE VOLUME AND
RELEASE UTTER CLAMP
AND LIFT OUT INNER WELL
REASSEMBLE MODULE
TO COLLECT WASH
INGS
EXTRACT AT *M J
EXTRACT AT 9* 11
COMBINE CH,C»7
EXTRACTS AND
SMI» TO LAI IK
AMBER CLASS
CONTAINER
RINSE WITH TEFLON WASH
•OTTLE ICWjCij' ALONG
INNER WELL SURFACE
AND CONDENSER WALL
RINSE WITH
PLACE INNER WELL
ASIDE IN CLEAN AREA
MEASURE AQUEOUS
VOLUME REMAINING
MEASURE VOLUME
OF FIRST IMPINGER
COMBINE
J
RINSE ENTRANCE TUBE AND
RACK HALF OF FILTER HOUS-
ING INTO MODULE. RINSE
DOWN CONDENSER WALL
RELEASE CENTRAL CLAMP
TO SEPARATE CLEAN CON
DENSER SECTION FROM
LOWER SECTION. RINSE LOW
IR SECTION INTO CONDEN
SATE CUP. RELEASE THE
•OTTOM CLAMP AND RINSE
INTO CONOENSATE CUP.
DRAIN INTO AMBER tOTTLE
VIA DRAIN VALVE
LABEL RESIN
CONTAINER
AND RINSINGS
CLEAN ALL MODULE METAL
PARTS BV CLEANING PROCE
DURE
Figure 4. Sample Handling and Transfer XAD-2 Module.
0020 - 17
Revision p
Date September 1986
-------
ADO RINSE FROM CONNECTING
LINE LEADING MOM SOMfNT
MODULE TO FIRST IMPING Eft
WINGER HO. 1
MEASURE VOLUME
TRANSFER TO LINEAR HIGH DENSITY
POLYETHYLENE CONTAINER AND
MEASURE TOTAL VOLUME
HINSE WITH OtSTILLtD H>0
IMFINCER NO. 2
MEASURE VOLUME
RINSE WITH A KNOW* AMOUNT OF OlSTlllEC MjO
IMPINGER NO. 3
MEASURE VOLUME
RINSE WITH A KNOWN AMOUNT OF DISTILLED H20
COMBIKE AND
MEASURE TOTAL
VOLUME FOR
SIK'GIE AkUVitS
^[lMHNCERNQ4 | »TTEICH AND DISCARD OR REGENERATE
Figure 5. Sample Handling and Transfer Impingers.
0020 - 18
Revision p
Date September 1986
-------
5.1.1 Leaving the fan operating, open the cyclone oven door to
expedite cooling. When the probe can be safety handled, disconnect from
the 10-um cyclone Inlet. Wipe off external particulate matter near the
probe tip and place a cap over each end. The probe must remain level
throughout this procedure.
NOTE: CARE MUST BE TAKEN TO AVOID TIGHTLY CAPPING TRAIN COMPONENTS AS
THEY ARE COOLING FROM STACK OR OVEN TEMPERATURES. THIS WILL
ELIMINATE THE POSSIBILITY OF CREATING A VACUUM INSIDE WHICH, WHEN
RELEASED, MAY DISRUPT AND BACKFLUSH THE CYCLONE AND FILTER
PARTICULATE CATCHES INTO ONE ANOTHER.
5.1.2 Disconnect the line joining the filter outlet to the XAD-2
module and cap off:
a. The 10-um cyclone Inlet;
b. The filter holder outlet; and
c. The Inlet of the line just disconnected from the filter holder
outlet.
5.1.3 Disconnect the filter holder and cap the Inlet. Set aside
with the Inlet facing upward. Cap the outlet of the 1-um cyclone. The
cyclones must remain upright throughout this procedure. The cyclones can
now be disconnected from one another or moved to the recovery area as a
single unit.
5.1.4 Disconnect the line joining the XAD-2 module to the 1mp1nger
system at the organic module outlet. Cap the organic module outlet.
5.1.5 Disconnect the silica gel 1mp1nger outlet from the vacuum
line to the pumps; cap off the first 1mp1nger Inlet and the fourth
1mp1nger outlet.
5.2 Nozzle, Probe, Cyclones, and Filter; The step-by-step procedures
for the recovery of particulatematerialcollected 1n the nozzle, probe, and
cyclones, and on the filter are detailed below:
5.2.1 Carefully transfer the filter from the filter housing to Its
original glass Petrl dish; a pair of clean blunt-tipped tweezers and a
flat spatula should be used for handling the filter. Using a clean
nylon-bristled brush, add any particulate material from the front half of
the filter housing to the Petrl dish; seal the Petrl dish around the
circumference with l-1n.-w1de Teflon tape; store with the collected
particulate material facing upward.
5.2.2 Tap and brush any particulate material adhering to the walls
of the upper chamber of the 1-um cyclone Into the lower cup; remove the
cup and quantitatively transfer the bulk contents to a wide-mouth amber
glass jar. Rinse the brush with methanol/methylene chloride (1:1 v/v)
Into the probe rinse container.
0020 - 19
Revision 0
Date September 1986
-------
5.2.3 Recover the contents of the 3-um cyclone 1n the same manner,
using a separate wide-mouth amber glass jar.
5.2.4 Recover the contents of the 10-um cyclone 1n the same manner,
using a separate wide-mouth amber glass jar.
5.2.5 Reconnect the lower cups of each cyclone and rinse any
particulate material adhering to the walls down Into the cups with the
methanol/methylene chloride mixture until the walls appear clean. Remove
the lower cups and transfer the contents to the probe rinse container.
Rinse the Interconnecting tubing among the cyclones Into the probe rinse
In the same manner.
5.2.6 Carefully remove the probe nozzle and clean the Inside
surface by rinsing with the methanol/methylene chloride (1:1 v/v) from a
wash bottle and brushing with a nylon-bristle brush. Brush until the
rinse shows no visible particles; make a final rinse of the Inside
surface.
5.2.7 Rinse the probe liner (preferably with two people so as to
minimize the possibility of accidental sample loss) with methanol/
methylene chloride (1:1 v/v) by tilting and rotating while spraying
solvent into the upper end and allowing the lower end to drain into the
sample container. Follow rinsing with brushing and rinsing from the
upper to the lower end. Push the brush through the Uner with a twisting
action; ensure that the sample container is placed under the lower end.
Brush until the rinse appears clean; perform a final rinse. Inspect the
Inner surface of the liner for cleanliness. Rinse any particulate
material remaining on the brush Into the sample container.
5.2.8 Clearly label all containers according to the coding scheme
given in Table 1; cover each label completely with transparent tape; mark
liquid levels and store all liquid samples on ice.
5.3 XAD-2, condensate, and organic module; Sample recovery of the
entire organic module may be conducted independently from the previous steps.
The step-by-step procedure for recovery of this stage is given below:
5.3.1 Rinse a 1-ft x 1-ft square of aluminum foil (dull side) with
methylene chloride and allow to air dry.
5.3.2 Release the clamp joining the XAD-2 cartridge section to the
upper gas conditioning system (second clamp); remove the XAD-2 cartridge
from the holder and place upon the clean aluminum foil. GENTLY pry loose
or unscrew (depending upon the design) the ring securing the fine mesh
screen on the top of the cartridge. Remove the screen and quantitatively
transfer the XAD-2 to a clean glass amber jar. A large, clean glass
funnel should be used for the transfer. Rinse the Inner surfaces of the
cartridge and the funnel with methylene chloride as necessary to remove
adhering XAD-2. Any XAD-2 that escapes onto the aluminum foil should be
retrieved and added to the sample.
0020 - 20
Revision 0
Date September 1986
-------
TABLE 1. SUGGESTED FORMAT FOR SAMPLE CODING AND IDENTIFICATION
Sample
Code
Container
Size
Sample description
1C
3C
IOC
Amber glass
Amber glass
Amber glass
PF-a,b,c,... Glass Petrl dish
PR Amber glass
PRB
MRX
MRXB
CD-LE
CD-LEB
AR-I1
I1B
123
I23B
Amber glass
Amber glass
Amber glass
Amber glass
Amber glass
HDLPa
HDLP
HDLP
HDLP
100 mL, wide-mouth
100 mL, wide-mouth
100 mL, wide-mouth
>6-1n. dlam.
1 liter
500 mL
1 liter
500 mL
1 liter
500 mL
1 liter
500 mL
1 liter
500 mL
1-m cyclone catch
3-m cyclone catch
10-m cyclone catch
Particulate filter(s)
Methylene chloride/
methanol front-half
rinse
Methylene chloride/
methanol blank
Methylene chloride
back half rinse
XAD-2 resin blank
Methylene chloride
condensate extract
Methylene chloride
blank
Aqueous residue of
condensate combined
with first 1mp1nger
catch
First 1mp1nger blank
(distilled H20 or
other appropriate
solution)
Second and third
implnger catches
Second and third
implnger blank
(distilled H20 or
other appropriate
solution)
aHDLP - High Density Linear Polyethylene.
0020 - 21
Revision 0
Date September 1986
-------
5.3.3 Replace the screen on the cartridge, reinsert the cartridge
Into the module, and reassemble the module. One person can accomplish
this task by butting the lower section 1n Us proper sealing position up
against the upper section while securing the ring clamp. One or
more wooden spacers approximately 1/2 1n. thick are suggested for
this purpose.
5.3.4 Open the coridensate reservoir valve and drain the remaining
condensate Into the condensate storage container. Measure and record the
volume and pH (using narrow-range pH paper) of the entire condensate.
5.3.5 Transfer the entire condensate to an appropriately-sized
separatory funnel. Adjust the pH of the condensate (as Indicated by the
narrow-range pH paper) to 1-2 using ultrapure or reagent grade nitric
add. Extract the condensate three times with methylene chloride, each
time with fresh portions measuring 8-10% of the total condensate volume.
If the volume of the condensate Is extremely large (>1800 ml), the
condensate may be extracted 1n portions, but fresh volumes of methylene
chloride must be used for each and every extraction. After each addition
of methylene chloride to the separatory funnel, the funnel must be shaken
with periodic venting through the stopcock to relieve any vapor pressure.
For safety, the tip of the separatory funnel should always be directed
away from the face and eyes while venting. Whenno further vapor
pressure can be vented after shaking, the funnel should be mounted
upright on a ring stand, the cap removed, the layers allowed to separate,
and the methylene chloride (bottom) layer removed. If an emulsion forms
equal to more than one-third the size of the solvent layer, reagent-grade
sodium chloride should be added until the emulsion 1s broken or reduced
to meet the above criterion. The emulsion Interface should not be
Included as part of the methylene chloride extract.
5.3.6 Following the third extraction of the acidified condensate,
adjust the pH of the aqueous residue to 11-12 with a 50% w/w solution of
sodium hydroxide (as Indicated by narrow-range pH paper), extract with
methylene chloride 1n the same manner, and combine the methylene chloride
extracts of the condensate at the high and low pH readings.
5.3.7 Transfer the aqueous residue from this extraction to a clean
Nalgene container; retain for later addition of the first 1mp1nger
solution.
5.3.8 Ensure that the condensate reservoir valve 1s closed, release
the upper clamp, and 11ft the Inner well halfway out of the module.
Rinse the Inner well Into the XAO-2 module using a Teflon wash bottle
containing methylene chloride, so that the rinse travels down the module
and Into the condensate collector. Then remove the well entirely and
place to one side on a clean surface (aluminum foil prerinsed with
methylene chloride). Rinse the entrance tube Into the module Interior;
rinse the condenser wall allowing solvent to flow down through the system
and collect In the condensate cup.
0020 - 22
Revision 0
Date September 1986
-------
5.3.9 Release the centra] clamp again and separate the lower
section (XAD-2 cartridge holder and condensate cup) from the upper.
5.3.10 Lift the empty XAD-2 cartridge halfway out of the mid-section
and rinse the outer surface down Into the condensate cup. Remove the
cartridge completely to a clean surface (aluminum foil rinsed with
methylene chloride).
5.3.11 Rinse the empty XAD-2 section Into the condensate cup. Open
the condensate reservoir valve and drain Into the XAD-2 sample storage
container (wide-mouth amber glass jar).
5.3.12 Rinse the condensate reservoir and combine the rinse with the
XAD-2 resin as above.
5.3.13 Clearly label all containers according to the coding scheme
presented 1n Table 1; cover each label completely with transparent tape;
mark liquid levels and store all liquid samples on 1ce.
5.4 Implnqers; Sample recovery from the 1mp1ngers may also be accom-
plished Independently of the other two sections of the SASS train. The
procedures are described below.
5.4.1 First Implnger:
5.4.1.1 Measure the volume of liquid 1n the Implnger with a
graduated cylinder; combine with the aqueous residue from the
condensate.
5.4.1.2 Rinse the line connecting the XAD-2 module to the
first Implnger with Type II water; transfer the rinse to the same
graduated cylinder. Rinse the Implnger twice more with Type II
water, combining all rinses In the graduated cylinder. Measure the
total rinse volume and add to the sample. Rinse the graduated
cylinder with a known amount of Type II water and add to the sample.
Record all volumes on the sample recovery sheet.
5.4.2 Second and third Implngers:
5.4.2.1 Measure and record the combined volume of liquid 1n
the Implngers 1n a large (1,000-mL) graduated cylinder; transfer to
a clean sample storage container.
5.4.2.2 Rinse the line connecting the first and second
Implnger Into the second Implnger and the line connecting the second
and third Implnger Into the third Implnger. Transfer the rinses to
the same graduated cylinder. Rinse each Implnger twice again with
Type II water, combining all rinses 1n the graduated cylinder.
Measure and record the combined rinse volume and add to the sample.
Rinse the graduated cylinder with a known amount of Type II water
and add to the sample. Record this additional rinse volume and add
to the Implnger rinse volume above.
0020 - 23
Revision 0
Date September 1986
-------
5.4.2.3 Clearly label all sample containers according to the
coding scheme presented 1n Table 1; cover each label completely with
transparent tape; mark fluid levels and store all liquid samples on
1ce.
5.4.3 Fourth Implnger:
5.4.3.1 Transfer the silica gel to Its original container.
Weigh to the nearest 0.1 g on a triple-beam balance, and record the
weight.
5.4.3.2 Discard or regenerate.
6.0 SAMPLE PREPARATION FOR SHIPMENT
6.1 Prior to shipment, recheck all sample containers to ensure that the
caps are securely tightened. Seal the Hds of all Nalgene containers around
the circumference with vinyl tape and those of glass containers with Teflon
tape. Ship all liquid samples on 1ce and all partlculate filters with the
parti cul ate catch facing upward. Ship peroxide solutions (Impinged) In a
separate container.
7.0 CALIBRATION
7.1 All calibration results should be recorded on appropriate data
sheets and fastened securely Into a separate section 1n the field sampling
notebook. Samples of blank data appear 1n 40 CFR 60 (1979), Appendix A.
7.2 Probe nozzles;
7.2.1 Probe nozzles must be calibrated before each Initial use 1n
the field. Using Vernier calipers or micrometers, measure the Inside
diameter of the nozzle to the nearest 0.025 mm (0.001 In.). Perform ten
separate measurements using different diameters; obtain the average of
the ten measurements. The difference between the highest and lowest
measurement results must not exceed 0.1 mm (0.004 1n.). When nozzles
become nicked, dented, or corroded, they must be reshaped, sharpened, and
recalibrated before reuse. Recall brat Ion of the nozzle before each run
1n gas streams that are highly corrosive 1s strongly recommended, as the
nozzle diameter may be changing slightly from one run to the next. Each
nozzle must be permanently and uniquely engraved.
7.3 PI tot tube
i
7.3.1 If the Type-S pltot tube conforms to the construction
specifications (the face openings are not visibly nicked, dented, or
corroded) and the pltot tube/probe assembly meets the Intercomponent
spadngs outlined In EPA Method 2 (see References), the pltot tube need
not be calibrated to meet federal and many state testing requirements; a
correction coefficient may be assigned 1n these cases. Some states,
however, require that, once used, pltot tubes must be calibrated In a
0020 - 24
Revision 0
Date September 1986
-------
wind tunnel. Specific state requirements such as this must be
unequivocally stated prior to testing. In either case, pltot tube face
openings should be Inspected before each run to ensure that there has
been no change 1n appearance since their construction or most recent
calibration.
7.4 Metering system;
7.4.1 Before each Initial use 1n the field, the metering system
shall be calibrated using a standard bell prover of the proper size. (A
standard bell prover 1s recommended for this procedure because the
displacement volume of commercially available wet test meters 1s
typically Insufficient.) A meter stick should be used to Indicate the
distance travelled by the Inner tank during the measurement. Figure 6
Illustrates a suitable arrangement for the calibration. It 1s highly
recommended that the dry gas meter be adjusted until the ratio of the dry
gas meter volume to the standard bell prover volume equals l.QO + 0.01,
to ensure that the calculated AH^ will result 1n near-1sok1netic sampling
rates. The calibration procedure follows:
7.4.1.1 Perform both a positive (pressure) and a negative
(vacuum) leak-check of the metering system. For the negative leak-
check, Include only the orifice Magnehellcs (reg. trademark), dry
gas meter, and two vacuum pumps by removing the vacuum line
connecting the fourth 1mp1nger to the vacuum pumps at the common
side of the pump Inlet tee, and replacing the line with a plain-end
male quick connect. Tightly cap this end and leak-check In the
manner outlined above under 4.4.3.11. For the positive leak-check,
connect a short length of rubber tubing to the "gas exhaust" port on
the SASS control module. Disconnect and vent the low side of the
orifice magnehellc; close off the low-side orifice tap. Pressurize
the system to 13-18 cm H20 (5-7 1n. H20) by blowing Into the rubber
tubing; pinch off the tubing and observe the magnehellc for one
minute. The magnehellc reading must remain unchanged during that
time period. Any loss of magnehelic pressure indicates a leak that
must be corrected.
7.4.1.2 Upon obtaining satisfactory leak-checks, connect the
metering system to the standard bell prover.
7.4.1.3 Using the control box Magnehellc (reg. trademark)
indicator, set the pumping rate corresponding to a AH of 1 in. Hg.
Turn the pumps off using the switches.
7.4.1.4 Record the Initial temperature and pressure of the
bell prover and the initial temperature and reading of the dry gas
meter. Record the barometric pressure every hour.
0020 - 25
Revision
Date September 1986
-------
TANK REFILLING
PUMP
COUNTERWEIGHTS
PROVER
PRESSURE
MANOMETER
PROVER
THERMOMETER
SASS SAMPLING PUMP
SASS SAMPLING PUMP
HEIGHT
INDICATOR
TANK MAKEUP AIR
j METER STICK
I
i
7
3
3
WATER
SASS CONTROL
MODULE
Figure 6. Schematic Diagram of Standard Bell Prover Arrangement for SASS Dry-Gas-Meter
Calibration.
0020 - 26
Revision
Date September 1986
0
-------
7.4.1.5 Disconnect the metering system and pump the Inner tank
of the bell prover to a convenient height. Reconnect the metering
system and record the height.
7.4.1.6 Start the pumps and a stopwatch simultaneously;
evacuate the tank for 3 m1n.
7.4.1.7 After 3 m1n, turn off the pumps using the switches.
Record the final Inner tank height, the final dry gas meter reading
and temperature, and the bell prover final temperature and pressure.
7.4.1.8 Repeat steps 3-7 using AH settings of 2, 4, and 6 1n.
H20.
7.4.1.9 Duplicate the entire procedure as a check; repeat the
entire procedure after each adjustment of the dry gas meter.
7.4.1.10 Calculate the Dry-Gas-Meter Correction Coefficient.
the ratio of the volume of gasmeasured by the dry gas meter to the
standard bell prover, both corrected to standard conditions and on a
dry basis. The ratio reduces to:
V V P T
pvr(std) = pvr pvr dgm
7 ~ v— ~"\/—p—T
dgm(std) dgm dgm pvr
where:
7 « Dry gas meter correction coefficient, dlmenslonless;
vdgm(std) = Volume of gas measured by the dry gas meter on a dry
basis, corrected to standard conditions, dscm (dscf);
Vpvr(std) s Volume of gas measured by the standard bell prover on a
dry basis, corrected to standard conditions, dscm
(dscf);
Vdom = Volume of gas measured at dry-gas-meter conditions,
9 m3(ft3);
- Final volume reading - Initial volume reading;
Vprv = Volume of gas measured at standard bell prover
conditions, m3(ft3)
- Kpvr x (difference 1n meter stick height readings),
where:
= number of ft3 of air displaced represented by
each cm of movement along the meter stick,
m3/cm(ft3/cm);
0020 - 27
Revision 0
Date September 1986
-------
pdgm = Absolute meter pressure, mm Hg (In. Hg)
= Barometric pressure + AH/13.6;
Ppvr = Absolute prover pressure, mm Hg (1n. Hg)
= Barometric pressure - [(AP) prover manometer]/I3.6;
Tpvr = Absolute bell prover temperature, *K (*R); and
Tdgm = Absolute dry-gas-meter temperature, *K (*R).
7.4.1.11 Calculate the Orifice Constants using the following
equations:
» n - M R Vn PstdTm(avg)
a. v__ " Vi-D, ,_J U_4.j n
mo ws std pn
where:
Qmo = Sampling flowrate at orifice, ftVmln (dry);
Bws = Proportion by volume of water in ambient air,
dlmenslonless;
Qstd = Standard sampling flowrate for SASS, 4.0 scfm (wet);
Pstd = Standard absolute pressure, 29.92 1n. Hg;
Tm(avg) = Average meter temperature, *R;
Pm = Absolute meter pressure (barometric pressure + AH/13.6),
in. Hg; and
Tstd = Standard temperature, 528*R.
r 1 "1/2
h i - 5"!2 Tm(avg) DAg <>,,
' 1 ~ A P M KflH4^9r
1 Ao1 I pmMm 1 c.
where:
J| = Orifice coefficient for orifice "1";
- Sampling flowrate at orifice, ft3/m1n (dry);
0020 - 28
Revision
Date September 1986
-------
= Orifice area [ir(d1ameter)2]/4, 1n.2;
Tm(avg) = Average meter temperature, *R;
Pm = Absolute meter pressure (see equation above), 1n. Hg;
Mm = Molecular weight of air, 29.0 lbm/lb-mole;
R = Gas law constant, 1545 ft-lbf/*R Ib-mole;
AHj = Orifice "1" pressure drop, 1n. H20; and
gc = Gravitational constant, 32.17 lbm-ft/lbf sec2.
7.4.1.12 The orifice constants may be determined without the
bell prover by noting the dry-gas-meter volumes obtained by pumping
at 1, 2, 3, and 6 In. H20 for 3-m1n periods. The obtaining of
consistent values when checking orifice constants 1n the field may
be used as a rough Indication of a valid calibration during extended
field use.
7.4.2 After each series of field tests, the calibration of the
metering system must be checked by performing three calibration
measurements at a single Intermediate orifice setting at or near the
average used during the field testing. If the calibration has changed by
more than 5%, recalibrate the meter over the full range of orifice
settings. Calculations for the test series should then be performed
using whichever calibration results 1n the lower value for total sample
volume.
7.5 Probe heater: The probe heating system shall be calibrated before
each initial use 1n the field and checked after each series of field tests
according to the procedure outlined 1n APTD-0576.
7.6 Thermocouples; Each thermocouple must be permanently and uniquely
marked on the casing; all mercury-1n-glass reference thermometers must conform
to ASTM-E-1 #63C or 63F specifications. Thermocouples should be calibrated 1n
the laboratory without the use of extension leads. If extension leads are
used in the field, the thermocouple reading at ambient air temperatures, with
and without the extension lead, must be noted and recorded. Correction is
necessary 1f the use of an extension lead produces a change greater than 1.5%.
Calibration for the various kinds of thermocouples proceeds as follows:
7.6.1 Implnger and organic module thermocouples: For the
thermocouples used to measure the temperature of the gas leaving the
impinger train and the XAD-2 resin bed, a three-point calibration at ice
water, room-air, and boiling-water temperatures is necessary. Accept the
thermocouples only 1f the readings at all three temperatures agree within
2*C (3.6*F) of the absolute value of the reference thermometer.
0020 - 29
Revision 0
Date September 1986
-------
7.6.2 Dry-gas-meter thermocouples: For the thermocouples used to
Indicate the dry-gas-meter Inlet and outlet temperatures, a three-point
calibration at Ice-water, room-air, and boiling-water temperatures must
be performed. The values must be within 2*C (3.6*F) of the absolute
reference thermometer value at all three calibration points.
7.6.3 Probe and stack thermocouple: For the thermocouples used to
Indicate the probe and stack temperatures, a three-point calibration at
Ice-water, boiling-water, and boiling cook1ng-o1l temperatures must be
performed; 1t 1s highly recommended that room-air temperature be added as
a fourth calibration point. If the absolute values of the reference
thermometer and the thermocouple agree within 1.5% at each of the
calibration points, a calibration curve (equation) may be constructed
(calculated), and the data extrapolated to cover the entire temperature
range suggested by the manufacturer.
7.7 Barometer; Adjust the field barometer Initially and before each
test series to agree within 2.5 mm Hg (0.1 1n. Hg) of the mercury barometer,
or within the station barometric pressure value reported by a nearby National
Weather Service station and corrected for elevation.
7.8 Triple-Beam Balance; Calibrate the triple beam balance before each
test series using Class-S standard weights; the weights must be within 0.5 g
of the standards, or the balance adjusted to meet these limits.
7.9 Analytical Balance; Calibrate "'the analytical balance with Class-S
weights before Initially tare-weighing each set of filters. The balance must
agree or be adjusted to within 2 mg of the standards. Run at least one
standard each time one or more of the filters Is rewelghed.
8.0 CALCULATIONS
8.1 Dry gas volume;
8.1.1 From the SASS run sheet, average the dry-gas-meter
temperatures and orifice pressure drops readings throughout the run.
Calculate the Volume of Dry Gas Sampled at standard conditions (20*C,
760 mm Hg [528*R, 29.92 1n. Hg]) using the equation:
(Tstd) (Pbar + (AH/13.6))
ii = V f ——— —Par
Tm
(Pbar + (AH/13.6))
= M.,7 T
1 m Tm
0020 - 30
Revision 0
Date September 1986
-------
where:
vm(std) = Volume of dry gas sampled at standard conditions, dscm
(dscf ) ;
Vm = Volume of dry gas sampled at dry-gas-meter conditions,
don (dcf);
7 = Dry-gas-meter calibration factor, dlmenslonless;
Tm = Average dry-gas-meter temperature, *K (*R);
Tstd = Standard absolute temperature, *K (*R);
Pbar = Barometric pressure at\ the sampling site, mm Hg (1n. Hg);
Pstd = Standard absolute pressure, mm Hg (1n. Hg);
AH = Average orifice pressure drop during the sampling run,
mm H20 (1n. H20); and
KI = 0.358'K/mm for metric units
= 17.64*R/1n. Hg for English units.
8.1.2 The above equation must be modified whenever the leakage rate
observed during any of the mandatory leak-checks (I.e., the post-test
leak-checks or leak-checks made prior to component changes) exceeds the
maximum allowed. The modification follows:
8.1.2.1 Case I (No component changes have been made during
the sampling run, and the allowable leakage rate has been exceeded
during the post-test leak-check): Replace Vm with the expression:
where:
Lp = Leakage rate observed during post-test leak-check,
m3/m1n (cfm);
La = Maximum allowed leakage rate, 0.0014 m3/m1n (0.05
ft3/m1n); and
8 = Total sampling time, m1n.
8.1.2.2 Case II (One or more component changes made during the
sampling run, and the allowable leakage rate has been exceeded
0020 - 31
Revision
Date September 1986
-------
during one or more of the leak-checks prior to component changes or
during the post-test leak-check): Replace Vm with the expression:
n
E
1=1
VLP - La>
where:
Li = Leakage rate observed prior to "itn" component change
if the allowable leakage rate has been exceeded while
sampling with the "itn" component, m3/min (cfm);
La = Maximum allowed leakage rate, 0.0014 m3/min (0.05
ft3/min);
B-j-i = Sampling time interval between the successive component
changes in which the allowable leakage rate has been
exceeded, min;
Lp = Leakage rate observed during post-test leak-check, if
the allowable leakage rate has been exceeded, m3/min
(cfm); and
Bp = Sampling time interval, from the final (ntn) component
change until the end of the sampling run, 1f the
allowable leakage rate has been exceeded during the
post-test leak-check, min.
8.2 Moisture content;
8.2.1 Calculate the Volume of Water Vapor at standard conditions:
'w(std)
0*.
Mw
RTst_d
'std
where:
vw(std) = Volume of water vapor in the gas sample, corrected to
standard conditions, dscm (dscf);
V-|c = Volume of liquid collected in the condensate reservoir
added to the net increase in impinger solution volumes
and silica gel weight gain during the run, mL;
py, = Density of water, 0.9982 g/mL (0.002201 Ib/mL);
0020 - 32
Revision 0
Date September 1986
-------
MN = Molecular weight of water, 18.0 g/g-mole (Ib/lb-mole);
R = Ideal gas constant, 0.06236 mm Hg-m3/*K-g-mole (21.85
1n. Hg-ft3/*R-lb-mole);
Tstd = Standard absolute temperature, *K (*R);
pstd = Standard absolute pressure, mm Hg (1n. Hg); and
K2 = 0.001333 m3/mL for metric units
= 0.04707 ft3/mL for English units.
8.2.2 Calculate the Stack Gas Moisture Content (equal to Bws x 100
for conversion to percent) :
» __ Vw(std) _ R inn
— - - - ~ D X 100
u + ~
vm(std) + vw(std) ws
where:
Bws = Proportion of water vapor 1n the gas stream by volume,
dlmenslonless;
Vw(std) = Volume of water vapor 1n the gas sample, corrected to
standard conditions, dscm (dscf); and
vm(std) = Volume of gas measured by the dry gas meter, corrected
to standard conditions, dscm (dscf).
8.2.3 In saturated or water-droplet-laden gas streams, make two
calculations of the moisture content, one from the total volume of liquid
collected In the train and one from the assumption of saturated gas-
stream conditions. Use whichever method results In the lower value. To
determine the moisture content based upon saturated conditions, use the
average stack gas temperature In conjunction with: (1) a psychrometrlc
chart, correcting for difference between the chart and the absolute stack
pressure; or (2) saturation vapor pressure tables.
8.3 Parti cul ate concentration;
8.3.1 Calculate the Unit Methanol /Methyl ene Chloride Blank
Correction for all front-half samples:
mm
iroir mm
0020 - 33
Revision
Date September 1986
-------
where:
cmm = Methanol/methylene chloride blank correction, mg/g;
Minn, ~ Mass °f methanol/methylene chloride after evaporation, mg;
Vmm = Volume of methanol/methylene chloride used 1n wash, ml; and
Pm - Density of 50:50 mix of methanol/methylene chloride,
mg/mL (see labels on bottles).
8.3.2 Calculate the Total MethanolI/Methylene Chloride Blank Weight
Correction for each Individual front-half sample:
mm mm mnr mm
where:
wmm = Weight of residue 1n methanol/methylene chloride front-
half wash, mg;
cmm = Methanol/methylene chloride unit blank correction, mg/g;
Vmm = Volume of methanol/methylene chloride used for front-half
wash, mL; and
Pm\ = Density of 50:50 mixture of methanol and methylene
chloride, mg/mL.
8.3.3 Calculate Total Particulate Weight;
\J-f\J + W +• \ + (\J + U + W ^ + (\J W ^
p " v pf-a pf-b ' v lOc 3c lc; v pr mm;
where:
Wp = Total particulate weight, mg;
wpf-a+... = Particulate weight from filter Pf-a + Pf-b + ... ;
WlOc»w3c»wlc = Particulate weight catch from the 10-, 3-, and 1-um
cyclones, respectively, mg;
Wpr = Weight of front-half rinse residue before blank
correction, mg; and
wmm = Methanol/methylene chloride blank weight correction, mg.
0020 - 34
Revision
Date September 1986
-------
8.3.4 Calculate the Total Particulate Concentration;
Cp = (0.001 g/mg) (Wp/Vn(std))
where:
Cp = Concentration of particulate material In the stack gas,
g/dscm (gr/dscf);
Wp = Weight of partlculate material collected during run, mg;
and
vm(std) = Volume of gas sampled, dscm (dscf).
8.3.5 To convert the above concentration to units of gr/ft3 or
lb/ft3 for comparison with established or projected values, the following
conversion factors are useful:
From; To: Multiply By;
scf m3 0.02832
g/ft3 gr/ft3 15.43
g/ft3 lb/ft 2.205 x 10~3
g/ft3 g/m3 35.315
8.4 Concentration of organic material;
8.4.1 Calculate the Volumetric Flow Rate (Qsc|) during the run.
Determine the average stackgasvelocityand volumetric flow rate from
actual run data 1n the same manner that these were calculated during
preliminary determinations (see Paragraph 4.3).
8.4.2 Calculate the POHC Concentration;
r MPQhc Mcd-1e + Mmrx
P°hc " Qsd " Qsd
where:
Cp0nc = Concentration of POHCs 1n stack gas, ug/dscm;
= Total mass of POHCs collected 1n XAD-2 and organic
module rinse, and 1n the condensate extract, ug;
Mcd-le = Mass °f pOHCs extracted from the condensate (corrected
for methylene chloride blank extraction residue), ug;
0020 - 35
Revision
Date September 1986
-------
Mmrx = Mass of POHCs extracted from the XAD-2 sorbent and
organic module rinse (corrected for methylene chloride
blank extraction residue), ug; and
Qsd = Volumetric flow rate during the run, dscm.
8.5 Isok1net1c variation;
8.5.1 Having calculated Ts, Vm(std)i Vs, An, Ps, and Bws, determine
the Isok1net1c Variation using the equation:
TsVm(std)
where:
I = Isok1net1c variation, %;
Ts = Absolute average stack gas temperature, 'K (*R);
Vm(std) = Volume of gas sampled, dscm (dscf);
Ps = Absolute stack gas pressure, mm Hg (1n. Hg);
Vs = Stack gas velocity, m/sec (ft/sec); l
An = Cross-sectional area of nozzle, m3 (ft3);
8 = Net sampling time, m1n;
Bws = Proportion of water vapor 1n gas stream by volume,
dlmenslonless; and
K4 = 4.320 for metric units
= 0.09450 for English units.
8.5.2 For the accuracy of Level 1 requirements (factor of 3) for
measured particulate emissions, the 1sok1net1c variation must be within
70-150X.
8.6 Cyclone particle-size cutoff diameter;
8.6.1 The particle-size cutoff diameter represents that particle
diameter (assuming spherical particles of unit density) at which the
cyclone exhibits 50% collection efficiency; It 1s expressed as the "dso."
The range of particle size collected 1n each cyclone and on the filter 1s
dependent upon the operating temperature and flow rate through each of
0020 - 36
• Revision
Date September 1986
-------
these components. The particle-size cutoff diameters of 10, 3, and 1 urn
1n the cyclones are the result of calibration of these at 400*F and 4.0
scfm (6.5 acfm). When the determined 1sok1net1c sampling rate 1s not 4.0
scfm, or when It 1s necessary to maintain a constant sub1sok1net1c
sampling rate.(still within the limits of Level 1 accuracy) during the
SASS run, the particle-size cutoff diameters for the cyclones must be
extrapolated.
8.6.2 Existing calibration data 1s Insufficient to determine exact
mathematical relationships for variations of particle-size cutoff
diameter with temperature and with volumetric flow rate. The best
estimates (McCain, 1983) suggest that a square, and an Inverse square
root dependence, respectively, exist; the extrapolation equation Is
presented below.
8.6.2.1 Calculate the Gas Viscosity from the equation:
H = (1.68 x 10~4) + (2.292 X 1CT7) (T)
where:
p = Gas viscosity, poise; and
T = Gas temperature, *F.
8.6.2.2 Extrapolate the Particle Size Cutoff Diameter from:
2
DT 'Fa = D400,4.0
where:
^400
a J
'4.0
a .
1/2 = D400,4.0
3.37
V'a
DTa,Fa = Particle size cutoff diameter at cyclone operating a
temperature and flow rate, urn (note that the
volumetric flow rate must be corrected to standard
conditions);
^400,4.0 = Particle size cutoff diameter at an operating
temperature of 400*F a'nd flow rate of 4.0 scfm, urn;
MOO = Gas viscosity at 400°F, poise;
/*Ta = Gas viscosity at operating conditions, poise;
0020 - 37
Revision 0
Date September 1986
-------
V4.0 = Cyclone volumetric flow rate of 4.0 scfm; and
vFa = Cyclone volumetric flow rate at operating conditions,
scfm.
This equation reduces to:
1.
33.7
\
for the 10-um cyclone,
2.
10.1
for the 3-um cyclone,
3.
3.37
for the 1-um cyclone,
8.7 Cumulative partlculate weight percent less than calculated size;
8.7.1 Divide the weight collected 1n the Individual cyclones and on
the filter by the total weight of partlculate collected; express these as
a percentage, using the following equations:
%
Wpf-a + Wpf-b
- —
* 100
W
% W
lOc
lOc
W.
x 100
W3c '
* W3c - — * 10°
0020 - 38
Revision 0
Date September 1986
-------
w
Ic
% W
Ic
x 100
W_
where:
Wp = Total participate weight collected, mg;
Wpf = Wpf_a + Wpf.b + ...
= Particulate weight collected on filters PF-a + PF-b, etc.,
mg;
W10c = Participate weight collected 1n 10-um cyclone, mg;
WSG = Participate weight collected 1n 3-um cyclone, mg;
Wic = Partlculate weight collected 1n 1-um cyclone, mg; and
100 = Conversion to percent.
8.7.2 Calculate the Cumulative Weight Percent Less than the
Calculated Particle Size Cutoff Diameter byadding,toeach weight
percent, the weight percent ofaTTfractions having a smaller particle-
size cutoff diameter. Tabulate the data, using the form below as an
example:
PRESENTATION OF SASS PARTICLE SIZING DATA
Stage
Weight %
Collected
1n Stage
Cumulative Weight %
Less than Calculated
Particle Size
Cutoff Diameter
Calculated
Particle Size
Cutoff Diameter
10-um cyclone
3-um cyclone
1-um cyclone
Glass fiber
filter
0020 - 39
Revision 0
Date September 1986
-------
9.0 REFERENCES
9.1 References
1. Blake, D.W., Operating and Service Manual, Source Assessment Sampling
System, Acurex Corporatlon/Aerctherm Division, Mountain View, California,
1977.
2. Hamersma, J.W., D.G. Ackerman, M.M. Yamada, C.A. Zee, C.Y. Ung, K.T.
McGregor, J.F. Clausen, M.L. Kraft, J.S. Shapiro, and E.L. Moon, Emissions
Assessment of Conventional Stationary Source Combustion Systems: Methods and
Procedures Manual for Sampling and Analysis (final revision superseding drafts
of January and September 1977), TRW, Redondo Beach, California, 1978.
3. Rom, J.J., Maintenance Calibration and Operation of Isokinetlc Source
Sampling Equipment, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, APTD-0576, 1972.
4. U.S. Environmental Protection Agency, 40 CFR 60, Appendix A, Methods 1-5,
1979.
5. U.S. Environmental Protection Agency, IERL-RTP Procedures Manual: Level 1
Environmental Assessment, 2nd ed., Industrial Environmental Research
Laboratory, Research Triangle Park, NC, EPA-600/7-78-201, 1978.
9.2 Bibliography
1. Anderson Samplers, Inc., Operating Manual for Anderson 2000, Inc., Mark II
and Mark III Particle Sizing Stack Samplers, Anderson Samplers, Inc., Atlanta,
Georgia.
2. Blake, D.W., Source Assessment Sampling System: Design and Development,
U.S. Environmental Protection Agency, Office of Research and Development,
Research Triangle Park, NC, EPA-600/7-78-019, 1978.
3. Harris, J.C., D.J. Larsen, C.E. Rechsteiner, K.E. Thrun, Sampling and
Analysis Methods for Hazardous Waste Incineration, 1st ed., U.S. EPA Contract
No. 68-02-3111, Arthur D. Little, Inc., 1982.
4. Lentzen, D.E., et al., IERL-RTP Procedures Manual: Level I Environmental
Assessment, 2nd ed., Industrial Environmental Research Laboratory, Research
Triangle Park, NC, EPA-600/7-78-201, 1978.
5. Martin, R.M., Construction Details of Isokinetic Source-Sampling
Equipment, U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, APTD-0581, 1971.
0020 - 40
Revision
Date September 1986
-------
6. McCain, J.D., Southern Research Institute, 2000 South Ninth Avenue,
Birmingham, AL 35205, private communication, 1983.
7. von Lehmden, D.J., et al., Quality Assurance Handbook for A1r Pollution
Measurement Systems, Vol. Ill: Stationary Source-Specific Methods,
Environmental Monitoring and Support Laboratory, Research Triangle Park, NC,
EPA-600/4-77-027b, 1980.
0020 - 41
Revision 0
Date September 1986
-------
METHOD 0030
VOLATILE ORGANIC SAMPLING TRAIN
1.0 PRINCIPLE AND APPLICATION
1.1 Principle
1.1.1 This method describes the collection of volatile principal
organic hazardous constituents (POHCs) from the stack gas effluents of
hazardous waste Incinerators. For the purpose of definition, volatile
POHCs are those POHCs with boiling points less than 100*C. If the
boiling point of a POHC of Interest 1s less than 30'C, the POHC may break
through the sorbent under the conditions of the sample collection
procedure.
1.1.2 Field application for POHCs of this type should be supported
by laboratory data which demonstrate the efficiency of a volatile organic
sampling train (VOST) to collect POHCs with boiling points less than
30°C. This may require using reduced sample volumes collected at flow
rates between 250 and 500 mL/m1n. Many compounds which boll above 100*C
(e.g., chlorobenzene) may also be efficiently collected and analyzed
using this method. VOST collection efficiency for these compounds should
be demonstrated, where necessary, by laboratory data of the type
described above.
1.1.3 This method employs a 20-11ter sample of effluent gas
containing volatile POHCs which 1s withdrawn from a gaseous effluent
source at a flow rate of 1 L/m1n, using a glass-lined probe and a
volatile organic sampling train (VOST). (Operation of the VOST under
these conditions has been called FAST-VOST.) The gas stream 1s cooled to
20'C by passage through a water-cooled condenser and volatile POHCs are
collected on a pair of sorbent resin traps. Liquid condensate 1s
collected 1n an 1mp1nger placed between the two resin traps. The first
resin trap (front trap) contains approximately 1.6 g Tenax and the second
trap (back trap) contains approximately 1 g each of Tenax and petroleum-
based charcoal (SKC Lot 104 or equivalent), 3:1 by volume. A total of
six pairs of sorbent traps may be used to collect volatile POHCs from the
effluent gas stream.
1.1.4 An alternative set of conditions for sample collection has
been used. This method Involves collecting sample volume of 20 liters or
less at reduced flow rate. (Operation of the VOST under these conditions
has been referred to as SLO-VOST.) This method has been used to collect
5 liters of sample (0.25 L/mln for 20 min) or 20 liters of sample
(0.5 L/min for 40 m1n) on each pair of sorbent cartridges. Smaller
sample volumes collected at lower flow rates should be considered when
the boiling points of the POHCs of Interest are below 35*C. A total of
six pairs of sorbent traps may be used to collect volatile POHCs from the
effluent gas stream.
0030 - 1
Revision
Date September 1986
-------
1.1.5 Analysis of the traps Is carried out by thermal desorptlon
purge-and-trap by gas chromatography/mass spectrometry (see Method 5040).
The VOST 1s designed to be operated at 1 L/m1n with traps being replaced
every 20 m1n for a total sampling time of 2 hr. Traps may be analyzed
separately or combined onto one trap to improve detection limit.
However, additional flow rates and sampling times are acceptable. Recent
experience has shown that when less than maximum detection ability is
required, 1t is acceptable and probably preferable to operate the VOST at
0.5 L/min for a total of three 40-min periods. This preserves the 2-hr
sampling period, but reduces the number of cartridge changes in the field
as well as the number of analyses required.
1.2 Application
1.2.1 This method 1s applicable to the determination of volatile
POHCs 1n the stack gas effluent of hazardous waste Incinerators. This
method is designed for use in calculating destruction and removal
efficiency (ORE) for the volatile POHCs and to enable a determination
that ORE values for removal of the volatile POHCs are equal to or greater
than 99.99%.
1.2.2 The sensitivity of this method 1s dependent upon the level of
interferences in the sample and the presence of detectable levels of
volatile POHCs in blanks. The target detection limit of this method Is
0.1 ug/m3 (ng/L) of flue gas, to permit calculation of a ORE equal to or
greater than 99.99% for volatile POHCs which may be present in the waste
stream at 100 ppm. The upper end of the range of applicability of this
method is limited by breakthrough of the volatile POHCs on the sorbent
traps used to collect the sample. Laboratory development data have
demonstrated a range of 0.1 to 100 ug/m3 (ng/L) for selected volatile
POHCs collected on a pair of sorbent traps using a total sample volume of
20 liters or less (see Paragraph 1.1.4).
1.2.3 This method is recommended for use only by experienced
sampling personnel and analytical chemists or under close supervision by
such qualified persons.
1.2.4 Interferences arise primarily from background contamination
of sorbent traps prior to or after use 1n sample collection. Many
potential interferences can be due to exposure of the sorbent materials
to solvent vapors prior to assembly and exposure to significant
concentrations of volatile POHCs in the ambient air at hazardous waste
Incinerator sites.
1.2.5 To avoid or minimize the low-level contamination of train
components with volatile POHCs, care should be taken to avoid contact of
all interior surface or train components with synthetic organic materials
(e.g., organic solvents, lubricating and sealing greases), and train
components should be carefully cleaned and conditioned according to the
procedures described 1n this protocol.
0030 - 2
Revision 0
Date September 1986
-------
2.0 APPARATUS
2.1 Volatile Organic Sampling Train; A schematic diagram of the
principal components of the VOST is shownin Figure 1 and a diagram of one
acceptable version of the VOST is shown in Figure 2. The VOST consists of a
glass-lined probe followed by an isolation valve, a water-cooled glass
condenser, a sorbent cartridge containing Tenax (1.6 g), an empty impinger for
condensate removal, a second water-cooled glass condenser, a second sorbent
cartridge containing Tenax and petroleum-based charcoal (3:1 by volume;
approximately 1 g of each), a silica gel drying tube, a calibrated rotameter,
a sampling pump, and a dry gas meter. The gas pressure during sampling and
for leak-checking is monitored by pressure gauges which are in line and
downstream of the silica gel drying tube. The components of the sampling
train are described below.
2.1.1 Probe: The probe should be made of stainless steel with a
borosilicate or quartz glass Uner. The temperature of the probe is to
be maintained above 130*C but low enough to ensure a resin temperature of
20'C. A water-cooled probe may be required at elevated stack
temperatures to protect the probe and meet the above requirements.
Isokinetic sample collection is not a requirement for the use of VOST
since the compounds of Interest are in the vapor phase at the point of
sample collection.
2.1.2 Isolation valve: The isolation valve should be a greaseless
stopcock with a glass bore and sliding Teflon plug with Teflon wipers
(Ace 8193 or equivalent).
2.1.3 Condensers: The condensers (Ace 5979-14 or equivalent)
should be of sufficient capacity to cool the gas stream to 20°C or less
prior to passage through the first sorbent cartridge. The top connection
of the condenser should be able to form a leak-free, vacuum-tight seal
without using sealing greases.
2.1.4 Sorbent cartridges:
2.1.4.1 The sorbent cartridges used for the VOST may be used
1n either of two configurations: the inside-outside (I/O)
configuration in which the cartridge is held within an outer glass
tube and in a metal carrier, and the Inside-inside (I/I)
configuration in which only a single glass tube is used, with or
without a metal carrier. In either case, the sorbent packing will
be the same.
2.1.4.1.1 The first of a pair of sorbent cartridges shall
be packed with approximately 1.6 g Tenax GC resin and the
second cartridge of a pair shall be packed with Tenax GC and
petroleum-based charcoal (3:1 by volume; approximately 1 g of
each).
2.1.4.1.2 The second sorbent cartridge shall be packed so
that the sample gas stream passes through the Tenax layer first
and then through the charcoal layer.
0030 - 3
Revision 0
Date September 1986
-------
o
o
o
I
4*
Heated Probe
Glass Wool
Particulate
Filter
STACK
(or test System)
O 73
Qt CD
m
0>
V>
_j*
o
Vacuum
Indicator
—» Exhaust
Dry Gas
Mfiter
Empty Silica Gel
Implnger
Cpndensate
Trap
Impinqer
Figure 1. Schematic of Volatile Organic Sampling Train (VOST).
-------
1 Teflon Plug Valve w/Socket Joint
Condensers
Tubing
Tenax Trap
Tenax/Charcoal
Trap
Impinger
Ice Water Bath
Case
Figure 2. Volatile Organic Sampling Train (VOST).
0030 - 5
Revision p
Date September 1986
-------
2.1.4.2 The sorbent cartridges shall be glass tubes with
approximate dimensions of 10 cm by 1.6 cm I.D. The two acceptable
designs (I/O, I/I) for the sorbent cartridge are described 1n
further detail below.
2.1.4.2.1 Inside/Inside sorbent cartridge: A diagram of
an I/I sorbent cartridge 1s shown in Figure 3. This cartridge
is a single glass tube (10 cm by 1.6 cm I.D.) which has the
ends reduced in size to accommodate a 1/4- or 3/8-1n. Swagelok
or Cajon gas fitting. The resin is held in place by glass wool
at each end of the resin layer. The amounts of each type of
sorbent material used in the I/I design are the same as for the
I/O design. Threaded end caps are placed on the sorbent
cartridge after packing with sorbent to protect the sorbent
from contamination during storage and transport.
2.1.4.2.2 Inside/Outside type sorbent cartridge: A
diagram of an I/O sorbent cartridge 1s shown in Figure 4. In
this design the sorbent materials are held 1n the glass tube
with a fine mesh stainless steel screen and a C-cl1p. The
glass tube is then placed within a larger diameter glass tube
and held in place using V1ton 0-rings. The purpose of the
outer glass tube is to protect the exterior of the resin-
containing tube from contamination. The two glass tubes are
held in a stainless steel cartridge holder, where the ends of
the glass tubes are held in place by VIton 0-r1ngs placed 1n
machine grooves 1n each metal end piece. The three cylindrical
rods are secured 1n one of the metal end pieces and fastened to
the other end piece using knurled nuts, thus sealing the glass
tubes Into the cartridge holder. The end pieces are fitted
with a threaded nut onto which a threaded end cap 1s fitted
with a Viton 0-r1ng seal, to protect the resin from
contamination during transport and storage.
2.1.5 Metering system: The metering system for VOST shall consist
of vacuum gauges, a leak-free pump (Thomas Model 107 or equivalent,
Thomas Industries, Sheboygan, Wisconsin)), a calibrated rotameter (L1nde
Model 150, Linde Division of Union Carbide, Keasbey, New Jersey) for
monitoring the gas flow rate, a dry gas meter with 2% accuracy at the
required sampling rate, and related valves and equipment. Provisions
should be made for monitoring the temperature of the sample gas stream
between the first condenser and first sorbent cartridge. This can be
done by placing a thermocouple on the exterior glass surface of the
outlet from the first condenser. The temperature at that point should be
less than 20*C. If 1t 1s not, an alternative condenser providing the
required cooling capacity must be used.
2.1.6 Sample transfer lines: All sample transfer lines to connect
the probe to the VOST shall be less than 5 ft 1n length, and shall be
heat-traced Teflon with connecting fittings which are capable of forming
leak-free, vacuum-tight connections without the use of sealing grease.
0030 - 6
Revision 0
Date September 1986
-------
o
o
o
I
«-J
'10cm
-GLASS WOOL
OR
STAINLESS STEEL SCREEN
O 70
O> to
00
oo o'
m 3
TD
«-h
fD
Figure 3. Inside-inside vost cartridge
-------
o
o
u>
o
00
O 50
01 n
c-f <
co o
fD 3
a
Section cut through ql«M tubea
(showing screen, C-clip and O-ring in place)
LEGEND
A • Stainless Steel Carrier
B Glass Tube (9.84 L x 2.22 ID)
C Small Glass Tube (10cm x 1.6cm ID)
D • Fine Mesh Stainless Steel Screen
E • Stainless Steel C Clip
F O Ring (Viton)
G Nuts (+)
H • End Cap with Viton O Ring
I - Metal Rod with Threaded End (3)
J - Tenax/Charcoal Sorbcnt
K • Cajon Fitting
TOP
BOTTOM
Assanbled Trap
NTS
Figure 4. Sorbent Trap Assembly (I/O)
Volatile Organic Sampling Train (VOST)
-------
All other sample transfer lines used with the VOST shall be Teflon with
connecting fittings that are capable of forming leak-free, vacuum-tight
connections without the use of sealing grease.
3.0 REAGENTS AND MATERIALS
3.1 2,6-D1pheny1ene oxide polymer (Tenax, 35/60 mesh):
3.1.1 The new Tenax 1s Soxhlet extracted for 24 hr with methanol
(Burdlck & Jackson, pesticide grade or equivalent). The Tenax is dried
for 6 hr in a vacuum oven at 50*C before use. Users of I/O and I/I
sorbent cartridges have used slightly different thermal conditioning
procedures. I/O sorbent cartridges packed with Tenax are thermally
conditioned by flowing organic-free nitrogen (30 mL/min) through the
resin while heating to 190°C. Some users have extracted new Tenax and
charcoal with pentane to remove nonpolar impurities. However, these
users.have experienced problems with residual pentane in the sorbents
during analysis.
3.1.2 If very high concentrations of volatile POHCs have been
collected on the resin (e.g., mlcrograms of analytes), the sorbent may
require Soxhlet extraction as described above. Previously used Tenax
cartridges are thermally reconditioned by the method described above.
3.2 Charcoal (SKC petroleum-base or equivalent): New charcoal is
prepared and charcoal is reconditioned as described In Paragraph 4.4. New
charcoal does not require treatment prior to assembly Into sorbent cartridges.
Users of VOST have restricted the types of charcoal used in sorbent cartridges
to only petroleum-based types. Criteria for other types of charcoal are
acceptable 1f recovery of POHC in laboratory evaluations meet the criteria of
50 to 150%.
3.3 Viton-0-R1ng; All 0-r1ngs used in VOST shall be V1ton. Prior to
use, these 0-r1ngs should be thermally conditioned at 200*C for 48 hr.
0-rings should be stored in clean, screw-capped glass containers prior to use.
3.4 Glass tubes/Condensers; The glass resin tubes and condensers should
be cleaned with a nonionic detergent 1n an ultrasonic bath, rinsed well with
organic-free water, and dried at 110'C. Resin tubes of the I/O design should
be assembled prior to storage as described in Paragraph 4.1. Resin tubes of
the I/I design can be stored in glass culture tube containers with cotton
cushioning and Teflon-lined screw caps. Condensers can be capped with
appropriate end caps prior to use.
3.5 Metal parts; The stainless steel carriers, C-clips, end plugs, and
screens used 1n the I/O VOST design are cleaned by ultrasonication in a warm
nonlonic detergent solution, rinsed with distilled water, air-dried, and
heated in a muffle furnace for 2 hr at 400*C. Resin tubes of the I/I design
require Swagelok or equivalent end caps with Supelco M-l ferrules. These
should be heated at 190°C along with the assembled cartridges.
0030 - 9
Revision 0
Date September 1986
-------
3.6 Silica gel (Indicating type, 6-16 mesh): New silica gel may be used
as received^Silica gel which has been previously used should be dried for 2
hr at 175*C (350*F).
3.7 Cold packs; Any commercially available reusable liquids or gels
that can be repeatedly frozen are acceptable. They are typically sold 1n
plastic containers as "Blue Ice" or "Ice-Packs." Enough should be used to
keep cartridges at or near 4*C.
3.8 Water; Water used for cooling train components 1n the field may be
tap water; and water used for rinsing glassware should be organic-free.
3.9 Glass wool; Glass wool should be Soxhlet extracted for 8 to 16 hr,
using methanol, and oven dried at 110'C before use.
4.0 SAMPLE HANDLING AND PROCEDURE
4.1 Assembly;
4.1.1 The assembly and packing of the sorbent cartridges should be
carried out 1n an area free of volatile organic material, preferably a
laboratory 1n which no organic solvents are handled or stored and 1n
which the laboratory air 1s charcoal filtered. Alternatively, the
assembly procedures can be conducted 1n a glove box which can be purged
with organic-free nitrogen.
4.2 Tenax cartridges;
4.2.1 The Tenax, glass tubes, and metal cartridge parts are cleaned
and stored (see Section 3.0). Approximately 1.6 g of Tenax 1s weighed
and packed Into the sorbent tube which has a stainless steel screen and
C-cl1p (I/O design) or glass wool (I/I design) 1n the downstream end.
The Tenax 1s held 1n place by Inserting a stainless steel screen and
C-cllps 1n the upstream end (I/O design) or glass wool (I/I design).
Each cartridge should be marked, using an engraving tool, with an arrow
to Indicate the direction of sample flow, and a serial number.
4.2.2 Conditioned resin tubes of the I/O design are then assembled
Into the metal carriers according to the previously described
inside/inside or Inside/outside procedures (with end caps) and are placed
on cold packs for storage and transport. Conditioned resin tubes of the
I/I design are capped and placed on cold packs for storage and transport.
4.3 Tenax/Charcoal tubes
4.3.1 The Tenax, charcoal, and metal cartridge parts are cleaned
and stored as previously described (see Section 3.0). The tubes are
packed with approximately a 3;1 volume ratio of Tenax and charcoal
(approximately 1 g each). The Tenax and charcoal are held 1n place by
the stainless steel screens and C-clips (I/O design) or by glass wool
(I/I design). The glass tubes containing the Tenax and charcoal are then
0030 - 10
Revision 0
Date September 1986
-------
conditioned as described below (see Paragraph 4.4). Place the I/O glass
tubes 1n the metal carriers (see Paragraph 2.1.4.2.2), put end caps on
the assembled cartridges, mark direction of sample flow and serial
number, and place the assembled cartridges on cold packs for storage and
transport.
4.3.2 Glass tubes of the I/I design are conditioned, and stored in
the same manner as the I/O tubes.
4.4 Trap Conditioning - QC
4.4.1 Following assembly and leak-checking, the traps are connected
In reverse direction to sampling to a source of organic-free nitrogen,
and nitrogen is passed through each trap at a flow rate of 40 mL/min,
while the traps are heated to 190*C for 12-28 hr. The actual
conditioning period may be determined based on adequacy of the resulting
blank checks.
4.4.2 The following procedure 1s used to blank check each set of
sampling cartridges prior to sampling to ensure cleanliness. The
procedure provides semi -quantitative data for organic compounds with
boiling points below 110'C on Tenax and Tenax/Charcoal cartridges. It is
not intended as a substitute for Method 5040.
4.4.2.1 The procedure is based on thermal desorption of each
set of two cartridges, cryofocusing with liquid nitrogen onto a trap
packed with glass beads, followed by thermal desorption from the
trap and analysis by GC/FID.
4.4.2.2 The detection limit 1s based on the analysis of Tenax
cartridges spiked with benzene and toluene and 1s around 2 ng for
each compound.
4.4.2.3 The results of analyzing spiked cartridges on a daily
basis should not vary by more than 20 percent. If the results are
outside this range, the analytical system must be evaluated for the
probable cause and a second spiked cartridge analyzed.
4.4.2.4 The GC operating conditions are as follows:
GC Operating Conditions
Column: Packed column 6 ft x 1/8" stainless steel 1.0 percent
SP-1000 on Carbopack B 60/80, or equivalent.
Temperature program: 50*C for 5 min, 20*C/m1n increase to
190*C, hold 13 min.
Injector: 200'C.
Detector: F.I.D. 250°C.
Carrier Gas: Helium at 25 mL/min.
Sample valve: Valco 6-port with 40" x 1/16" stainless steel
trap packed with 60/80 mesh glass beads.
Cryogen: Liquid nitrogen.
Trap heater: Boiling water, hot oil, or electrically heated.
0030 - 11
Revision 0
Date September 1986
-------
Desorption heater: , Supelco "clam shell" (high capacity carrier
gas purifier) heater and Variac, adjusted to 180'C to
200'C.
4.4.2.5 Calibration 1s accomplished by preparing a spiked
Tenax cartridge with benzene and toluene and analyzing according to
the standard operating procedure. A standard of benzene, toluene
and bromofluorobenzene (BFB) is prepared by injecting 2.0 uL of
benzene and toluene and 1.0 uL of BFB into 10 ml of methanol. The
concentration of this stock is 175 ng/uL of benzene and toluene, and
150 ng/uL BFB. One microliter of the stock standard is injected
onto a Tenax cartridge through a heated injection port set at 150*C.
A GC oven can be used for this with the oven at room temperature.
Helium carrier gas is set at 50 mL/min. The solvent flush technique
should be used. After two min, remove the Tenax cartridge and place
1n the desorptlon heater for analysis. BFB is also used as an
internal standard spike for GC/MS analysis which provides a good
comparison between GC/FID and GC/MS. The results of this spike
analysis should not vary more than 20 percent day to day. Initially
and then periodically this spiked Tenax should be reanalyzed a
second time to verify that the 10 min desorption time and 180-200°C
temperature are adequate to remove all of the spiked components. It
should be noted that only one spiked Tenax cartridge need be
prepared and analyzed dally unless otherwise needed to ensure proper
Instrument operation.
An acceptable blank level 1s left to the discretion of the
method analyst. An acceptable level is one that allows adequate
determination of expected components emitted from the waste being
burned.
4.4.3 After conditioning, traps are sealed and placed on cold packs
until sampling is accomplished. Conditioned traps should be held for a
minimum amount of time to prevent the possibility of contamination.
4.4.4 It may be useful to spike the Tenax and Tenax/charcoal traps
with the compounds of interest to ensure that they can be thermally
desorbed under laboratory conditions. After spiked traps are analyzed
they may be reconditioned and packed for sampling.
4.5 Pretest preparation;
4.5.1 All train components shall be cleaned and assembled as
previously described. A dry gas meter shall have been calibrated within
30 days prior to use, using an EPA-suppl1ed standard orifice.
4.5.2 The VOST is assembled according to the schematic diagram in
Figure 1. The cartridges should be positioned so that sample flow 1s
0030 - 12
Revision 0
Date September 1986
-------
through the Tenax first and then the Tenax/charcoal. Cooling water
should be circulated to the condensers and the temperature of the cooling
water should be maintained near 0*C. The end caps of the sorbent
cartridges should be placed 1n a clean screw-capped glass container
during sample collection.
4.6 Leak-checking:
4.6.1 The train Is leak-checked by closing the valve at the Inlet
to the first condenser and pulling a vacuum of 250 mm (10 1n. Hg) above
the normal operating pressure. The traps and condensers are Isolated
from the pump and the leak rate noted. The leak rate should be less than
2.5 mm Hg after 1 m1n. The train 1s then returned to atmospheric
pressure by attaching a charcoal-filled tube to the train Inlet and
admitting ambient air filtered through the charcoal. This procedure will
minimize contamination of the VOST components by excessive exposure to
the fugitive emissions at hazardous waste Incinerator sites.
4.7 ' Sample Collection
4.7.1 After leak-checking, sample collection 1s accomplished by
opening the valve at the Inlet to the first condenser, turning on the
pump, and sampling at a rate of 1 I1ter/m1n for 20 m1n. The volume of
sample for any pair of traps should not exceed 20 liters.
4.7.2 Following collection of 20 liters of sample, the train 1s
leak-checked a second time at the highest pressure drop encountered
during the run to minimize the chance of vacuum desorptlon of organlcs
from the Tenax. The train 1s returned to atmospheric pressure, using the
method discussed 1n Paragraph 4.1 and the two sorbent cartridges are
removed. The end caps are replaced and the cartridges shall be placed 1n
a suitable environment for storage and transport until analysis. The
sample 1s considered Invalid 1f the leak test does not meet
specification.
4.7.3 A new pair of cartridges 1s placed 1n the VOST, the VOST
leak-checked, and the sample collection process repeated as described
above. Sample collection continues until six pairs of traps have been
used.
4.7.4 All sample cartridges should be kept on cold packs until they
are ready for analysis.
4.8 Blanks
4.8.1 Field blanks/trip blanks: Blank Tenax and Tenax/charcoal
cartridges are taken to the sampling site and the end caps removed for
the period of time required to exchange two pairs of traps on VOST.
After the two VOST traps have been exchanged, the end caps are replaced
on the blank Tenax and Tenax/charcoal tubes and these are returned to the
cold packs and analyzed with the sample traps. At least one pair of
field blanks (one Tenax, one Tenax/charcoal) shall be Included with each
0030 - 13
Revision
Date September 1986
-------
six pairs of sample cartridges collected (or for each field trial using
VOST to collect volatile POHCs).
4.8.2 Trip blanks: At' least one pair of blank cartridges (one
Tenax, one Tenax/charcoal) shall be Included with shipment of cartridges
to a hazardous waste Incinerator site. These "field blanks" will be
treated like any other cartridges except that the end caps will not be
removed during storage at the site. This pair of traps will be analyzed
to monitor potential contamination which may occur during storage and
shipment.
4.8.3 Laboratory blanks: One pair of blank cartridges (one Tenax,
one Tenax/charcoal) will remain 1n the laboratory using the method of
storage which 1s used for field samples. If the field and trip blanks
contain high concentrations of contaminants (e.g., greater than 2 ng of a
particular POHC), the laboratory blank shall be analyzed In order to
Identify the source of contamination.
5.0 CALCULATIONS (for sample volume)
5.1 The following nomenclature are used 1n the calculation of sample
volume:
PDar = Barometric pressure at the exit orifice of the dry gas meter, mm
(1n.) Hg.
pstd = Standard absolute pressure, 760 mm (29.92 1n.) Hg.
Tm = Dry gas meter average absolute temperature, K (*R).
Tstd = Standard absolute temperature, 293K (528*R).
Vm = Dry gas volume measured by dry gas meter, dcm (dcf).
vm(std) = Drv 9as volume measured by dry gas meter, corrected to standard
conditions, dscm (dscf).
7 = Dry gas meter calibration factor.
5.2 The volume of gas sampled 1s calculated as follows:
u » „ std bar _ „ m bar
"•(.«) - V -1^7 - Ki7 -1^-
where:
KI = 0.3858 K/mrn Hg for metric units, or
KI = 17.64 *R/1n. Hg for English units.
0030 - 14
Revision
Date September 1986
-------
6.0 ANALYTICAL PROCEDURE
See Method 5040.
7.0 PRECISION AND ACCURACY REQUIREMENTS
7.1 Method Performance Check
Prior to field operation of the VOST at a hazardous waste Incine-
rator, a method performance check should be conducted using either
selected volatile POHCs of Interest or two or more of the volatile POHCs
for which data are available. This check may be conducted on the entire
system (VOST/GC/MS) by analysis of a gas cylinder containing POHCs of
interest or on only the analytical system by spiking of the POHCs onto
the traps. The results of this check for replicate pairs of traps should
demonstrate that recovery of the analytes fall within 50% to 150% of the
expected values.
7.2 Performance Audit
During a trial burn a performance audit must be completed. The
audit results should agree within 50% to 150% of the expected value for
each specific target compound. This audit consists of collecting a gas
sample containing one or more POHCs 1n the VOST from an EPA ppb gas
cylinder. Collection of the audit sample 1n the VOST may be conducted
either 1n the laboratory or at the trial burn site. Anaysls of the VOST
audit sample must be by the same person, at the same time, and with the
same analytical procedure as used for the regular VOST trail burn
samples. EPA ppb gas cylinders currently available for VOST Audit are
shown 1n Table 1 below.
The audit procedure, audit equipment and audit cylinder may be
obtained by writing:
Audit Cylinder Gas Coordinator (MD-77B)
Quality Assurance Division
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
or by calling the Audit Cylinder Gas Coordinator at (919) 541-4531.
The request for the audit must be made at least 30 days prior to the
scheduled trial burn. If a POHC 1s selected for which EPA does not have
an audit cylinder, this audit 1s not required.
0030 - 15
Revision
Date September 1986
-------
8.0 REFERENCES
1. Protocol for the Collection and Analysis of Volatile POHCs Using VOST.
EPA/600/8-84/007, March 1984.
2. Sykes, A.L., Standard Operating Procedure for Blanking Tenax and
Tenax/Charcoal Sampling Cartridges for Volatile Organic Sampling Train (VOST),
Radion Corporation, P.O. Box 13000, Research Triangle Park, NC 27709.
3. Validation of the Volatile Organic Sampling Train (VOST) Protocol, Vols. I
and II, EPA/600/4-86/014a, January 1986.
0030 - 16
Revision
Date September 1986
-------
TABLE 1: Organic Gases 1n the ppb Audit Repository
Group I
5 Organlcs In N£:
Carbon tetrachlorlde
Chloroform
Perchloroethylene
Vinyl ch.lor1de
Benzene
Ranges of cylinders
currently available;
7-90 ppb
90 - 430 ppb
430 - 10,000 ppb
Group II
9 Organlcs In N£
Trlchloroethylene
l,2-D1chloroethane
l,2-D1bromoethane
F-12
F-ll
Bromomethane
Methyl ethyl ketone
1,1,1-TH chl oroethane
Acetron1tr1le
Ranges of cylinders
currently available;
7-90 ppb
90 - 430 ppb
0030 - 17
Revision 0
Date September 1986
-------
TABLE 1: Organic Gases In the ppb Audit Repository (Continued)
Group III
7 Organlcs 1n N£:
V1nyl1dene chloride
F-113
F-114
Acetone
l,4-D1oxane
Toluene
Chlorobenzene
Ranges of cylinders
currently available;
7-90 ppb
90 - 430 ppb
Group IV
6 Organlcs In N£:
Acrylon1tr1le
1,3-Butadlene
Ethylene oxide
Methylene chloride
Propylene oxide
Ortho-xylene
Ranges of cylinders
currently available;
7-90 ppb
430 - 10,000
0030 - 18
Revision 0
Date September 1986
-------
CHAPTER ELEVEN
GROUND WATER MONITORING
11.1 BACKGROUND AND OBJECTIVES
The hazardous waste management facility permit regulations were
promulgated In July, 1982 (40 CFR 265). Subpart F of these regulations,
Ground Water Protection, sets forth performance standards for ground water
monitoring systems at permitted facilities. Performance standards were
selected, rather than design and operating standards, because of the diversity
of designs and practices appropriate 1n various site-specific situations.
Performance standards provide more flexibility than design and operating
standards because site-specific conditions can be accomodated case by case
without variance procedures. However, Implementation 1s less efficient
because permit writers may need to consider a wider variety of designs and
practices; furthermore, much of the variation 1n reported values 1s
attributable to the variety of designs and practices currently 1n use.
The purpose of this Chapter 1s to Identify certain designs and practices
which meet the performance requirements 1n specified situations. One of the
Agency's reasons for doing so 1s to encourage the use of more standard
methods. The designs and practices which are Identified as acceptable 1n this
chapter are considered to be acceptable for the uses and conditions specified.
Therefore, permit applicants need not justify their selection. Use of these
designs and practices 1s not mandatory; owners and operators may submit
applications based on other approaches. The only Incentive to use the
"acceptable" designs and practices 1s that they are already recognized by the
Agency and so they need not be justified again. As this 11st matures, the
Agency 1s hopeful that sources of variance due to the variety 1n methodology
will decrease.
The provisions of this Chapter were developed recognizing that
professional judgement will always be needed 1n designing effective monitoring
systems. But, for efficiency of operation, repeated patterns of acceptance
and rejection of designs and operations are Identified so that the lengthy
documentation need not be repeated each time. Readers will note that there
are many arbitrary criteria for some "acceptable" methods and that there 1s
little or no attempt to Justify the cut-off values. This 1s Intended. This
Chapter 1s expected to be a living document, cautiously developed. As new
criteria become Identified further refinements of these values should be
expected. The purposes of listing the acceptable designs and practices are to
encourage use of standard techniques by making their use easier and to reduce
the burden on the applicants by relieving their need to justify use of proven
designs or practices. The listing establishes, 1n essence, blanket approvals
for a limited number of techniques 1n those conditions for which they are
known to be acceptable.
This Subsection establishes certain ground water sampling system designs
and practices as being acceptable under certain conditions for use 1n meeting
ELEVEN - 1
Revision 0
Date September 1986
-------
the requirements of Subpart F (264.90 et seq.). It also lists certain
practices and designs which are not acceptable. The acceptable designs and
practices are listed 1n Paragraph 11.4, below, with specified conditions for
which each may be acceptable. The proscribed practices and designs are listed
1n Paragraph 11.5. These are not acceptable for use 1n satisfaction of the
permit requirements; petitions for their use must follow normal channels.
11.2 RELATIONSHIP TO THE REGULATIONS AND TO OTHER DOCUMENTS.
The regulations 1n Subpart F will continue to be the sole location of the
performance standards for ground water monitoring systems. The provisions of
this Chapter only establish the -acceptability of a limited number of designs
or operations. The Chapter 1s not Intended to replace the regulations or the
guidance documents which explain application of the regulations 1n the
particular, or site-specific, situation. It 1s related to the guidance
documents In that 1t will promote use of the more established procedures found
1n general guidance.
The contents of this Chapter will be taken from general enforcement and
permitting guidance documents, and it Is Intended that these be consistent
with all RCRA ground water monitoring guidance. The specific conditions given
for the acceptable designs and procedures may not be found 1n the several
guidance documents from which those designs and procedures are taken. Many of
these conditions are arbitrarily selected. They are based on the experience
of permit writers and enforcement officials. Since the conditions only affect
procedural Issues (whether the selection 1s justified or not) the rigor of
their development has not been as extensive as 1f they were requirements.
There 1s one preeminent RCRA guidance document for ground water
monitoring at this time: The Technical Enforcement Guidance Document. (The
TEGD. finalized September 1986, 1s available from the Office of Waste Programs
Enforcement, (202)-475-9328). This document 1s written for enforcement
officials' use 1n Implementing the interim status provisions, 265.90 et seq.,
but most of the hydrogeologic principles apply directly to permitted
facilities as well as to those 1n interim status. The TEGD is the major
source of concepts for this chapter; 1t is and will be the major repository of
RCRA ground water monitoring principles. It is Intended that nothing In this
chapter conflicts with the TEGD.
Other ground water monitoring guidance documents are in circulation.
Several, such as "Ground Water Monitoring Guidance for Owners and Operators of
Interim Status Facilities," have been superceded by the TEGD.Others, such as
the draft "Permit Writers Guidance for Ground Water Monitoring," have never
been finalized and do not fully reflect Agency policy.
Other documents which may be of Interest are as follows:
1. Barcelona, Michael J., James P. G1bb and Robin A. Miller, A Guide to
the Selection of Materials for Monitoring Well Construction and Ground Water
Sampling, Illinois State Water Survey Contract Report (ISWS) #327, EPA
Contract No. EPA CR-809966-01, August 1983.
ELEVEN - 2
Revision 0
Date September 1986
-------
2. Benson, R.C., R.A. Glaccum, and M. R. Noel, Geophysical Techniques
for Sensing Burled Waste and Waste Migration, Technos, Inc., EPA Contract No.
68-03-3050; available from National Water Well Association, Worthlngton, OH.
3. Handbook for Analytical Quality Control 1n Water and Wastewater
Laboratories, EMSL, Cincinnati, EPA-600-4-79-019, March 1979 and subsequent
revisions; available from EMSL, Cincinnati, OH.
4. Hazardous Waste Ground Water Task Force, Protocol for Ground Water
Inspections at Hazardous Waste Treatment Storage and Disposal Facilities,
April 1986.
5. Methods for the Storage and Retrieval of RCRA Ground Water
Monitoring Data on STORET, Ref. Storet User Support (800-488-5985).
6. Methods of Chemical Analysis of Water and Wastes, EMSL, Cincinnati,
EPA-600/4-79-020, Revised March 1983; available from EMSL, Cincinnati, OH.
7. Plumb, R.H., and C.K. F1tzs1mmons, Performance Evaluation of RCRA
Indicator Parameters for Ground Water Monitoring, Proceedings of the First
Canadian-American Conference on Hydrogeology, National Water Well Association,
Worthlngton, OH, pp. 129-137, June 1984.
8. A Practical Guide for Ground Water Sampling, ERL, ADA, OK,
EPA/600/2-85/104, Sept. 1, 1985; available from Illinois State Water Survey,
Champagne, IL.
11.3 REVISIONS AND ADDITIONS
This Chapter will be revised from time to time as new technological
developments and experience dictate. Each revision will be proposed before
being finalized, and there will be ample time before the effective date for
the revisions to be Incorporated Into future designs.
Applicants desiring to add particular designs or practices to the
"acceptable" list, either for their own unique situation or as general
provisions, or to use designs or practices on the "proscribed" 11st may do so
by petitions.
11.4 ACCEPTABLE DESIGNS AND PRACTICES
The following designs and practices are acceptable, in the conditions
described and for the purposes listed, without need for justification. Permit
writers may question the existence of the condition or the definition of
purpose, but not the use of the design or practice once conditions and
purposes are established.
ELEVEN - 3
Revision 0
Date September 1986
-------
11.4.1 Site Characterisation
(a) Borehole location patterns, designed by qualified geologists, are
acceptable for site characterization. Such characterizations are for general
delineation of stratigraphy and flow paths and for establishing Initial design
of well placement, screen length, depth, etc.
Conditions: When unexpected discontinuities of major strata or pathways
do not occur.
(b) Geophysical logging and other Indirect measurement techniques may be
used 1n site characterization for the limited purpose of augmenting direct
observation of cuttings and corlngs by professional geologists.
Conditions: None.
(c) Quarterly measurements are generally satisfactory for establishing
seasonal and temporal variations 1n flow velocity and direction for purposes
of assuring that the elevations of screens are correct, of documenting the
appropriateness of background well locations, and of assuring coverage of all
possible downgradlent pathways.
Conditions: None.
11.4.2 Well Location, Design, and Construction
(a) Downgradlent well locations which result 1n placement 1n potential
pathways of contaminant migration are acceptable for routine detection
sampling programs. The density will vary based on the size of the pathway.
Conditions: When site characterization confirms simple homogenous
hydrogeology, without discontinuities or faults 1n the vicinity of the wells,
and when folds and fractures are not expected to channel flows past well
Intakes.
(b) Monitoring well screen lengths should generally not cut across
several flow zones but rather furnish depth-discrete measurements. These
conditions are acceptable for the purpose of obtaining samples which represent
ground water quality at the point of compliance.
Conditions: When the strata of concern 1s ^ 10' thick.
(c) Use of air rotary drilling methods 1s acceptable for Installing
monitoring wells.
Conditions: Except when drilling through contaminated upper horizons,
unless precautions are taken.
(d) Fluorocarbon resins (PTFE, PFA, FEP, etc.) and stainless steel (304
or 316) are acceptable materials for sample-contact surfaces 1n new or
replacement monitoring wells where potentially sorblng organlcs are of
concern.
Conditions: Stainless steel may only be used 1n non-corrosive
conditions. All new or replacement wells to be Installed at a given time
should be of the same material.
ELEVEN - 4
Revision 0
Date September 1986
-------
(e) Existing wells which do not meet the recommendations 1n guidance for
materials or Installation may be proposed for Inclusion 1n the permit.
Conditions: When documented to be free of bias by pairing new PTFE OR
stainless wells with, for Instance, at least ten percent of the old, existing
wells.
11.4.3 Sampling
(a) The field quality control procedures contained 1n Reference 4,
Section 11.2 above, and those specified in Chapters 1 and 9 of this document
are the only acceptable procedures.
(b) Well evacuation measured at three times the computed well casing
volume Is acceptable for assuring that the sample contains ground water
representative of the formation.
Conditions: Evacuation measured to +5% of the computed volume based on
water surface elevation and well bottom measured immediately prior to
evacuation.
(c) Samples containing less than 5 N.T.U. turbidity are acceptable for
analysis when the analytic method 1s sensitive to turbidity (such as the
analysis of metals). Samples containing greater than 5 N.T.U. are only
acceptable when well development is certified by a qualified hydrogeologist as
the best obtainable.
Conditions: Turbidity evaluation must accompany all potentially affected
values.
(d) The sample preservation techniques presented 1n Table 11-1 are
acceptable.
(e) The scheduled time Interval between sample collections should not be
greater than the computed time of travel either from the upgradient wells to
the point of compliance or from the point of compliance to the property
boundary.
(f) Evacuation of the well to dryness 1s an acceptable procedure to
ensure that the sample contains representative ground water.
Conditions: When the recharge is so slow that the well will yield fewer
than three well volumes before dryness but fast enough that the recharging
water will not cascade down the Inside of the casing.
11.4.4 Analysis and Reporting
The codes listed In Table 11-2 may be used for purposes and conditions
listed.
11.5 UNACCEPTABLE DESIGNS AND PRACTICES
The following designs and practices are unacceptable in the conditions or
for the purposes specified.
ELEVEN - 5
Revision
Date September 1986
-------
11.5.1 Site Characterization
Use of unsubstantiated data not meeting quality assurance criteria may
not be used other than 1n support of general trends or to establish
relationships between parameters.
Conditions: All conclusions and findings based on unconfirmed data and
unsupported by quality controlled data are 1nadm1ssable as support for permit
conditions or stipulations.
11.5.2 Well Location, Design, and Construction
Fabric filters should not be used as filter pack material.
11.5.3 Sampling
(a) The following devices are not generally acceptable for collecting
samples for analysis:
1. Gas driven piston pump.
2. Suction 11ft pumps.
3. Submersible diaphragm.
4. Gas 11ft samplers.
5. Impeller pumps.
(b) Data obtained by unsubstantiated techniques and procedures not
meeting quality assurance criteria or not conforming to quality control
procedures may not be used except when attempting to describe pre-existing
site conditions which are no longer observable.
11.5.4 Data Evaluation and Comparisons
Pooling upgradlent or background values from diverse hydrogeologlc strata
in a manner which combines data from discrete or distinct sampling locations
as though they were points along a continuous spectrum is not acceptable. All
up-down comparisons must be between samples taken from common flow paths.
ELEVEN - 6
Revision
Date September 1986
-------
TABUE 11-1
SAMPLDG AND PRESERVATION PROCEDURES FOR DETECTION MONITORING
,a
Parameter
pH
Specific conductance
TOG
TOX
Chloride
Iron
Manganese
Sodium
Phenols
Sulfate
Arsenic
, Barlun
Cadmium
Chromium
1/c.aH
Mercury
Selenium
Silver
Fluoride
Nitrate
Recommended
Container Preservative
Indicators of Ground Water Contamination0
T, P, G Field determined
T, P, G Field determined
G, Teflon-lined Cool 4°C, HC1 to
cap pH <2
G, amber, Teflon- Cool 4°C, add 1 mL of
lined cap 1.1M sodium sulfite
Ground Water Quality Characteristics
T, P, G 4°C
T, P Field acidified
to pH <2 with HNQj
G 4°C/ILSO to pH <2
T, P, G Cool, 4°C
EPA Interim Drinking Water characteristics
T, P Total Metals
Field acidified to
pH <2 with HN03
Dissolved Metals
1 . Field filtration
(0.45 micron)
Dark Bottle 2. Acidify to pH <2
with HNO
T, P Field acidified to
pH <2 with HN03
T, P, G 4°C/I^S04 to pH <2
(Continued)
ELEVEN-7
Revision 0
Holding Tine
None
None
28 days
7 days
28 days
6 months
28 days
28 days
6 months
6 months
28 days
14 days
Minimum Volume
Required for
Analysis
25 mL
100 mL
4 x 15 mL
4 x 15 mL
50 mL
200 mL
500 mL
50 mL
1,000 mL
1,000 mL
300 mL
1,000 mL
Date September 1986
-------
TABLE 11-1 (Continued)
SAMPLING AND PRESERVATION PROCEDURES FOR DETECTION MDNTIDRDC
a
Parameter
Recommended
Container
Preservative
Maximum
folding Time
Minimum Volume
Required for
Analysis
Bidrln
Lindane
Methoxychlor
Toxaphene
2,4 D
2,4,5 TP Silvex
Radiun
Gross Alpha
Gross Beta
Coliform bacteria
T, G
Cool, 4°C
P, G
Field acidified to
pH<2
PP, G (sterilized) Cool, 4°C
7 days
6 months
6 hours
2,000 mL
1 gallon
200 mL
Other Ground Water Characteristics of Interest
Cyanide
Oil and Grease
Sanivolatile,
volatile organics
P, G
G only
T, G
Cool, 4°C, NaOH to
pH>12
Cool, 4°C H,SO to
pH<2 L *
Cool, 4°C
14 days
28 days
7 days
500 mL
100 mL
1,000 mL
aReferences: Test Methods for Evaluating Solid Waste - Physical/Chemical Methods, SW-846 (3rd edition,
1986).
Methods for Chemical Analysis of Water and Wastes. EPA-600/4-79-020.
Standard Methods for the Examination of Water and Wastewater, 16th edition (1985).
b
Container Types:
P = Plastic (polyethylene)
G - Glass
T = Teflon
PP = Polypropylene
c.
Based on the requirements for detection monitoring ( 265.93), the owner/operator must collect a sufficient
volume of ground water to allow for the analysis of four separate replicates.
ELEVEN-8
Revision 0
Date September 1986
-------
M3L-H-
TABLE 11-2
A LISTBC AND DESCRIPTION OF CODES USED TO INDICATE THAT POLLUTANT
(ENCENTRATIONS WERE BELOW A CONCENIRAnON WHICH CAN BE MEASURED
ACCURATELY OR THAT THE POLLUTANTS WERE NOT PRESENT
Codes
LCO*-
LOJf
Def inition of
the Acronyms
Limit of detection
Limit of quantifi-
cation
Examples
of Use
LCD 0.421
LOQ 2.234
Used to Indicate
That the Pollutant
Was Less Than a
Limit of Detection
Yes
Yes
Used to Indicate
That the Pollutant
Was not Present
No
No
Method detection
limit
MDL 0.631
Yes
No
LT Less than
BDL Below detection
limit
<
Negative
signs
Trace*
K
ND* Not detected
Dashes*
Large
numbers*
Zeros*
Blanks*
LT, LT 0.01
LT 0.148
BDL, BDL 0.01
BDL 0.148
-------
CHAPTER TWELVE
LAND TREATMENT MONITORING
12.1 BACKGROUND
A monitoring program 1s an essential component at any land treatment unit
and should be planned to provide assurance of appropriate facility design, to
act as a feedback loop to furnish guidance on Improving unit management, and
to Indicate the rate at which the treatment capacity is being approached.
Because many assumptions must be made in the design of a land treatment unit,
monitoring can be used to verify whether the Initial data and assumptions were
correct or if design or operational changes are needed. Monitoring cannot be
substituted for careful design based on the fullest reasonable understanding
of the effects of applying hazardous waste to the soil; however, for existing
Hazardous Waste Land Treatment (HWLT) units (which must retrofit to comply
with regulations), monitoring can provide much of the data base needed for
demonstrating treatment.
Figure 12-1 shows the topics to be considered when developing a
monitoring program. The program must be developed to provide the following
assurances:
1. that the waste being applied does not deviate significantly from
the waste for which the unit was designed;
2. that waste constituents are not leaching from the land treatment
area in unacceptable concentrations;
3. that ground water is not being adversely affected by the migration
of hazardous constituents of the waste(s); and
4. that waste constituents will not create a food-chain hazard if
crops are harvested.
12.2 TREATMENT ZONE
As is depicted in Figure 12-2, the entire land treatment operation and
monitoring program revolve about a central component, the treatment zone.
Concentrating on the treatment zone is a useful approach to describing and
monitoring a land treatment system. The treatment zone is the soil to which
wastes are applied or incorporated; HWLT units are designed so that
degradation, transformation, and immobilization of hazardous constituents and
their metabolites occur within this zone.
In practice, setting a boundary to the treatment zone 1s difficult. In
choosing the boundaries of the treatment zone, soil-forming processes and the
associated decrease in biological activity with depth should be considered.
TWELVE - 1
Revision 0
Date September 1986
-------
f WASTE
POTENTIAL
SITE
1
DESIGN AND OPERATION
TREATMENT ZONE
CONCEPT
ANALYTICAL
CONSIDERATIONS
STATISTICAL
CONSIDERATIONS
TYPES OF
MONITORING
1
(FINAL SITE A
SELECTION y"~~^
\
I
V
MONITORING
I
CONTINGENCY PLANNING
AND ADDITIONAL CONSIDERATIONS
Figure 12-1. Topics to be considered in developing a monitoring program
for an HWLT unit.
TWELVE - 2
Revision 0
Date September 1986
-------
WASTE
I
U>
O 30
f* <
fb **^*
00
CO O
a
rt
DISCHARGE/
RUNOFF
(NPDES)
FOOD CHAIN CROPS
\ \ \
\ \ \ \ \ \ Y\ \
Figure 122. Various types of monitoring for land treatment units.
-------
12.3 REGULATORY DEFINITION
The current regulations (U.S. EPA, 1982a) require the following types of
monitoring:
1. Ground water detection monitoring to determine 1f a leachate plume
has reached the edge of the waste management area (40 CFR 264.98).
2. Ground water compliance monitoring to determine if the facility 1s
complying with ground water protection standards for hazardous
constituents (40 CFR 264.99).
3. Monitoring of soil pH and concentration of cadmium in the waste when
certain food-chain crops are grown on HWLT units where cadmium is
disposed of (40 CFR 264.276).
4. Unsaturated zone monitoring, including soil cores and soil-pore
liquid monitoring, to determine if hazardous constituents are
migrating out of the treatment zone (40 CFR 246.278).
5. Waste analysis of all types of waste to be disposed at the HWLT unit
(40 CFR 264.13).
12.4 MONITORING AND SAMPLING STRATEGY
As discussed earlier, the monitoring program centers around the treatment
zone.
The frequency of sampling and the parameters to be analyzed depend on the
characteristics of the waste being disposed, the physical layout of the unit,
and the surface and subsurface characteristics of the site. Table 12-1
provides guidance for developing an operational monitoring program. Each of
the types of monitoring is discussed below.
12.4.1 Waste Monitoring and Sampling Strategy
Waste streams need to be routinely sampled and tested to check for
changes in composition. A detailed description of appropriate waste sampling
techniques, tools, procedures, etc., is provided in Chapter Nine of this
manual (in Part III, Sampling). These procedures should be followed during
all waste sampling events. Waste analysis methods are provided 1n this
manual. The analyst should choose the appropriate method, based on each waste
and specific constituents to be tested for.
The frequency with which a waste needs to be sampled and the parameters
to be analyzed depend greatly on the variables that influence the quantity and
quality of the waste. When waste is generated in a batch, as would be
expected from an annual or biannual cleanout of a lagoon or tank, the waste
should be fully characterized prior to each application. When the waste 1s
TWELVE - 4
Revision 0
Date September 1986
-------
TABLE 12-1 GUI DANCE FOR AN OPERATIONAL MONITORING PROGRAM AT HULT UNITS
Media to be Monitored Purpose
Sampling Frequency
Number of Sarnies
Parameters to be Analyzed
Waste
Soil cores
(unsaturated zone)
Quality Change.
Determine slow movement
of hazardous constituents.
Quarterly composites if continuous
stream; each batch If Intermittent
generation.
Quarterly
One
One composited from
two per l.S ha (4 ac);
minimum of 3 composited
from 6 per uniform area.
At least rate and capacity
limiting constituents, plus
those within 2SS of being
limiting, principal hazardous
constituents. pH and EC.
.All hazardous constituents in
the waste or the principal
hazardous constituents,
metabolites of hazardous
constituents, and nonhazardous
constituents of concern.
m
I
01
O 70
o> n
00
o'
3
o
<-h
m
Soil-pore liquid
(unsaturated zone)
Groundwater
Vegetation (if
grown for food
chain use).
Runoff water
Soli in the
treatment zone
Air
Determine highly mobile
constituents.
Determine mobile
constituents.
Phytotoxlc and hazardous
transmitted constituents
(food chain hazards).
Soluble or suspended
constituents.
Determine degradation,
pH, nutrients, and rate
and capacity limiting
constituents.
Personnel and population
health hazards.
Quarterly, preferably following
leachate generating precipitation
snownelt.
Semiannually
Annually or at harvests.
As required for NPDES permit.
Quarterly
Quarterly
One composited from two
samplers per 1.5 ha
(4 ac); minimum of 3
composited from 6 per
uniform area.
Minimum of four
suggested--one up-
gradient, three down-
gradient.
One per l.S ha (4 ac)
or three of processed
crop before sales.
As permit requires,
or one.
7-10 composited to one
per l.S ha (4 ac).
Five
All hazardous constituents In
the waste or the principal
hazardous constituents,
mobile metabolites of hazard-
ous constituents, and important
mobile nonhazardous constituents.
Hazardous constituents and
metabolites or select
indicators.
Hazardous i
and their
etals and organlcs
etabolltes.
Discharge permit and
background parameters plus
hazardous organics.
Participates (adsorbed
hazardous constituents) and
hazardous volatiles.
-------
generated more nearly continuously, samples should be collected and composited
based on a statistical design over a period of time to ensure that the waste
1s of a uniform quality. For example, wastes that are generated continuously
could be sampled weekly or dally on a flow-proportional basis and composited
and analyzed quarterly or monthly. When no changes have been made 1n the
operation of the plant or the treatment of the waste which could significantly
alter concentration of waste constituents, the waste should, at a minimum, be
analyzed for (1) the constituents that restrict the annual application rates
(RLC) and the allowable cumulative applications (CLC), (2) the constituents
that are within 25% of the level at which they would be limiting, and (3) all
other hazardous constituents that have been shown to be present 1n the waste
in the Initial waste characterization. Because synerglsm and antagonism as
well as unlisted waste metabolites can create hazards that cannot be described
by chemical analysis alone, routine mult1gen1c1ty testing may be performed 1f
the treatment demonstration has Indicated a possible problem. In .addition,
waste should be analyzed as soon as possible after a change in operations that
could affect the waste characteristics.
12.4.2 Ground Water, Monitoring and Sampling Strategy
To ensure that Irreparable ground water damage does not occur as a result
of HWLT, It 1s necessary that the ground water quality be monitored. Ground
water monitoring supplements the unsaturated zone monitoring system but does
not replace 1t. A contamination problem first detected 1n the leachate water
may indicate the need to alter the management program, and ground water can
then be observed for the same problem. It 1s through the successful
combination of these two systems that accurate monitoring of vertically moving
constituents can be achieved. Ground water monitoring requirements are
discussed in Chapter Eleven of this manual.
12.4.3 Vegetation Monitoring and Sampling Strategy
Where food-chain crops are to be grown, analysis of the vegetation at the
HWLT unit will aid 1n ensuring that harmful quantities of metals or other,
waste constituents are not being accumulated by, or adhering to surfaces of,
the plants. Although a safety demonstration before planting is required (U.S.
EPA, 1982a), operational monitoring 1s recommended to verify that crop
contamination has not occurred. Vegetation monitoring 1s an Important
measurement during the post-closure period where the area may possibly be used
for food or forage production. Sampling should be done annually or at each
harvest. The concentrations of metals and other constituents 1n the
vegetation will change with moisture content, stage of growth, and the part of
the plant sampled, and thus results must be carefully interpreted. The number
of samples to analyze is again based on a sliding scale similar to that used
for sampling soils. Forage samples should Include all aerial plant parts, and
the edible parts of grain, fruit, or vegetation crops should be sampled
separately.
TWELVE - 6
Revision
Date September 1986
-------
12.4.4 Runoff Water Monitoring and Sampling Strategy
If runoff water analyses are needed to satisfy NPDES permit conditions
(National Pollution Discharge Elimination System, U.S. EPA, 1981), a
monitoring program should be Instituted. This program would not be covered
under RCRA hazardous waste land disposal requirements, but it would be an
Integral part of facility design. The sampling and monitoring approach will
vary, depending on whether the water 1s released as a continuous discharge or
as a batch discharge following treatment to reduce the hazardous nature of the
water. Constituents to be analyzed should be specified 1n the NPDES permit.
When a relatively continuous flow 1s anticipated, sampling must be flow
proportional. A means of flow measurement and an automated sampling device
are a reasonable combination for this type of monitoring. Flow can be
measured using a weir or flume (U.S.D.A., 1979) for overload flow-water
pretreatment systems and packaged water treatment plants, and 1n-l1ne flow
measurement may be an additional option on the packaged treatment systems.
The sampling device should be set up to obtain periodic grab samples as the
water passes through the flow-rate measuring device. A number of
programmable, automated samplers that can take discrete or composite samples
are on the market.
For batch treatment, such as mere gravity separation or mechanically
aerated systems, flow 1s not so Important as 1s the hazardous constituent
content of each batch. Sampling before discharge would, in this case, Involve
manual pond sampling, using multiple grab samples. The samples would
preferably represent the entire water column to be discharged 1n each batch
rather than a single depth Increment. Statistical procedures should again be
used for either treatment and discharge approach.
12.4.5 Unsaturated Zone Monitoring and Sampling Strategy
The unsaturated zone 1s described as the layer of soil or parent material
separating the bottom of the treatment zone and the seasonal high-water table
or ground water table and 1s usually found to have a moisture content less
than saturation. In this zone, the movement of moisture may often be
relatively slow 1n response to soil properties and prevailing climatic
conditions; however, 1n some locations, soils and waste management practices
may lead to periods of heavy hydraulic loading that could cause rapid downward
flux of moisture.
An unsaturated zone monitoring plan should be developed for two purposes:
(1) to detect any significant movement of hazardous constituents out of the
system, and (2) to furnish Information for management decisions. In light of
the variability 1n soil-water flux and the mobility of hazardous waste
constituents, the unsaturated zone monitoring plan should include sampling the
soil to evaluate relatively slow-moving waste constituents (soil core
monitoring) and sampling the soil-pore liquid to evaluate fast-moving waste
constituents. Monitoring for hazardousconstituents should be performed on a
representative background plot(s) until background levels are established and
TWELVE - 7
Revision 0
Date September 1986
-------
Immediately below the treatment zone (active portion). The number, location,
and depth of soil core and soil-pore liquid samples taken must allow an
accurate Indication of the quality of soil-pore liquid and soil below the
treatment zone and 1n the background area. The frequency and timing of soil-
pore liquid sampling must be based on the frequency, time, and rate of waste
application; proximity of the treatment zone to ground water; soil
permeability; and amount of precipitation. The data from this program must be
sufficient to determine 1f statistically significant Increases 1n hazardous
constituents (or selected Indicator constituents) have occurred below the
treatment zone. Location and depth of soil core and soil-pore liquid samples
follow the same reasoning, but the number, frequency, and timing of soil core
sampling differs somewhat from that required for soil-pore liquid sampling.
Thus, the unique aspects of these topics will be considered together with
discussions of techniques for obtaining the two types of samples.
12.4.5.1 Location of Samples
Soil characteristics, waste type, and waste application rate are all
important factors in determining the environmental impact of a particular land
treatment unit or part of a unit on the environment. Therefore, areas of the
land treatment unit for which these characteristics are similar (I.e., uniform
areas) should be sampled as a single monitoring unit. A uniform area 1s
defined as an area of the active portion of a land treatment unit which 1s
composed of soils of the same soil series (U.S.D.A., 1975) and to which
similar wastes or waste mixtures are applied at similar application rates.
If, however, the texture of the surface soil differs significantly among soils
of the same series classification, the phase classification of the soil should
be considered in defining "uniform areas." A certified professional soil
scientist should be consulted in designating uniform areas.
Based on that definition, it 1s recommended that the location of soil
core sampling or soil-pore liquid monitoring devices within a given uniform
area be randomly selected. Random selection of samples ensures a more
accurate representation of conditions within a given uniform area. It is
convenient to spot the field location for soil core and soil-pore liquid
devices by selecting random distances on a coordinate system and using the
intersection of the two random distances as the location at which a soil core
should be taken or a, soil-pore liquid monitoring device Installed. This
system works well for fields of both regular and Irregular shape because the
points outside the area of interest are merely discarded and only the points
Inside the area are used in the sample.
The location within a given uniform area of a land treatment unit (I.e.,
active portion monitoring) at which a soil core should be taken or a soil-pore
liquid monitoring device installed should be determined using the following
procedure:
1. Divide the land treatment unit into uniform areas under the
direction of a certified professional soil scientist.
TWELVE - 8
Revision 0
Date September 1986
-------
2. Set up coordinates for each uniform area by establishing two
base lines at right angles to each other which Intersect at an
arbitrarily selected origin, for example, the southwest corner.
Each baseline should extend far enough for all of the uniform area
to fall within the quadrant.
3. Establish a scale interval along each base line. The units of
this scale may be feet, yards, meters, or other units, depending on
the size of the uniform area, but both base lines should have the
same units.
4. Draw two random numbers from a random-number table (available
in most basic statistics books). Use these numbers to locate one
point along each of the base lines.
5. Locate the intersection of two lines drawn perpendicular to the
base lines through these points. This Intersection represents one
randomly selected location for collection of one soil core, or for
installation of one soil-pore liquid device. If this location at
the intersection is outside the uniform area, disregard and repeat
the above procedure.
6. For soil core monitoring, repeat the above procedure as many
times as necessary to obtain the desired number of locations within
each uniform area of the land treatment unit. This procedure for
randomly selecting locations must be repeated for each soil core
sampling event but will be needed only once in locating soil-pore
liquid monitoring devices.
Locations for monitoring on background areas should also be randomly
determined. Again, consult a certified professional soil scientist 1n
determining an acceptable background area. The background area must have
characteristics (including soil series classification) similar to those
present in the uniform area of the land treatment unit it 1s representing, but
1t should be free from possible contamination from past or present activities
that could have contributed to the concentrations of the hazardous
constituents of concern. Establish coordinates for an arbitrarily selected
portion of the background area and use the above procedure for randomly
choosing sampling locations.
12.4.5.2 Depth of Samples
Because unsaturated zone monitoring is intended to detect pollutant
migration from the treatment zone, samples should logically be obtained from
Immediately below this zone. Care should be taken to ensure that samples from
active areas of the land treatment unit and background samples are monitoring
similar horizons or layers of parent material. Because soils seldom consist
of smooth, horizontal layers, but are often undulating, sloped, and sometimes
discontinuous, it would be unwise to specify a single depth below the land
surface to be used for comparative sampling. A convenient method for choosing
TWELVE - 9
Revision 0
Date September 1986
-------
sampling depths 1s to define the bottom of the treatment zone as the bottom of
a chosen diagnostic solid horizon and not as a rigid depth. Sampling depth
would then be easily defined with respect to the bottom of the treatment zone.
At a minimum, soil core and soil-pore liquid sampling should monitor within 30
cm (12 1n.) of the bottom of the treatment zone. Additional sampling depths
may be desirable, for Instance, 1f analytical results are Inconclusive or
questionable. Core samples should Include only the 0- to 15-cm Increment
below the treatment zone, whereas soil-pore liquid samplers should be placed
so that they collect liquid from anywhere within this 30-cm zone.
12.4.5.3 Soil Core Sampling Techniques
Soil Cores
Waste constituents may move slowly through the soil profile for a number
of reasons, such as the lack of sufficient soil moisture to leach through the
system, a natural or artificially occurring layer or horizon of low hydraulic
conductivity, or waste constituents that exhibit only a low to moderate
mobility relative to water 1ri soil. Any one or a combination of these effects
can be observed by soil core monitoring. Based on the treatment zone concept,
only the portions of soil cores collected below the treatment zone need to be
analyzed. The Intent 1s to demonstrate whether there are significantly higher
concentrations of hazardous constituents 1n material below the treatment zone
than 1n background soils or parent material.
Soil core sampling should proceed according to a definite plan with
regard ,to number, frequency, and technique. Previous discussions of statis-
tical considerations should provide guidance 1n choosing the number of samples
requiredi Background values for soil core monitoring should be established by
collecting at least eight randomly selected soil cores for each soil series
present 1n the treatment zone. These samples can be composited 1n pairs (from
Immediately adjacent locations) to form four samples for analysis * For each
soil series, a background arithmetic mean and variance should be calculated
for each hazardous constituent. For monitoring the active portion of the land
treatment facility, a minimum of six randomly selected soil cores should be
obtained per uniform area and composited, as before, to yield three samples
for analysis. If, however, a uniform area 1s >5 ha (12 ac), at least two
randomly selected soil cores should be taken per 1.5 ha (4 ac) and composited
1n pairs based on location. Data from the samples 1n a given uniform area
should be averaged and statistically compared. If analyses reveal a large
variance from samples within a given uniform area, more samples may be
necessary. Soil coring should be done at least semiannually, except for
background sampling, which, after background values are established, may be
performed as needed to determine 1f background levels are changing over time.
It 1s Important to keep an accurate record of the locations from which
soil core samples have been taken. Even when areas have been judged to be
uniform, the best attempts at homogeneous waste application and management
cannot achieve perfect uniformity. It 1s probable 1n many systems that small
problem areas, or "hot spots," may occur, causing localized real or apparent
pollutant migration. Examples of "apparent" migration might Include small
TWELVE - 10
Revision 0
Date . September 1986
-------
areas where waste was applied too heavily or where the machinery on-s1te mixed
waste too deeply. The sampling procedure Itself 1s subject to error and so
may Indicate apparent pollutant migration. Therefore, anomalous data points
can and should be resampled at the suspect locatlon(s) to determine 1f a
problem exists, even 1f the uniform area as a whole shows no statistically
significant pollutant migration.
The methods used for soil sampling are variable and depend partially on
the size and depth of the sample needed and the number and frequency of
samples to be taken. Of the available equipment, oil field augers are useful
If small samples need to be taken by hand, and bucket augers give larger
samples. Powered coring or drilling equipment, 1f available, 1s the
preferable choice because 1t can rapidly sample to the desired depths and
provide a clean, minimally disturbed sample for analysis. Due to the time
Involved 1n coring to 1.5 m, and sometimes farther, powered equipment can
often be less costly than hand sampling. In any case, extreme care must be
taken to prevent cross contamination of samples. Loose soil or waste should
be scraped away from the surface to prevent 1t from contaminating samples
collected from lower layers. The material removed from the treatment zone
portion of the borehole can be analyzed, 1f desired, to evaluate conditions 1n
the treatment zone. It 1s advisable to record field observations of the
treatment zone even If no analysis Is done. Finally, boreholes absolutely
must be backfilled carefully to prevent hazardous constituents from channeling
down the hole. Native soil compacted to about field bulk density, clay
slurry, or other suitable plug material may be used.
Sample handling, preservation, and shipment should follow a cha1n-of-
custody procedure and a defined preservation method such as 1s found 1n
Chapter Nine of this manual or 1n the analytical section of EPA document SW-
874, Hazardous Waste Land Treatment (U.S. EPA, 1983). If more sample 1s
collected than Is needed for analysis, the volume should be reduced by either
the quartering or riffle technique. (A riffle 1s a sample-splitting device
designed for use with dried ground samples.)
The analysis of soil cores must include all hazardous constituents that
are reasonably expected to leach or the principal hazardous constituents
(PHCs) that generally indicate hazardous constituent movement (U.S. EPA,
1982a).
Soil-Pore Liquid
Percolating water added to the soil by precipitation, Irrigation, or
waste applications may pass through the treatment zone and may rapidly
transport some mobile waste constituents or degradation products through the
unsaturated zone to the ground water. Soil-pore liquid monitoring is Intended
to detect these rapid pulses of contaminants (often Immediately after heavy
precipitation events) that are not likely to be observed through the regularly
scheduled analysis of soil cores. Therefore, the timing of soil-pore liquid
sampling is a key to the usefulness of this technique. SeasonabiHty is the
TWELVE - 11
Revision
Date September 1986
-------
rule with soil-pore liquid sample timing (I.e., scheduled sampling cannot be
on a preset date, but must be geared to precipitation events). Given that
sampling is done soon after leachate-generating precipitation or snowmelt, the
frequency also varies depending on site conditions. As a starting point,
sampling should be done quarterly. More frequent sampling may be necessary at
units located in areas with highly permeable soils or high rainfall, or at
which wastes are applied very frequently. The timing of sampling should be
geared to the waste application schedule as much as possible.
At land treatment units where wastes are. applied infrequently (i.e., only
once or twice a year) or where leachate-generating precipitation is highly
seasonal, quarterly sampling and analysis of soil-pore liquid may be
unnecessary. Because soil-pore liquid sampling is instituted primarily to
detect fast-moving hazardous constituents, monitoring for these constituents
many months after waste application may be useless. If fast-moving hazardous
constituents are to migrate out of the treatment zone, they will usually
migrate within at least 90 days following waste application, unless little
precipitation or snowmelt has occurred. Therefore, where wastes are applied
infrequently or leachate generation is seasonal, soil-pore liquid may be
monitored less frequently (semiannually or annually). A final note about
timing is that samples should be obtained as soon as liquid is present. The
owner or operator should check the monitoring devices for liquid within 24 hr
of any significant rainfall, snowmelt, or waste application.
The background concentrations of hazardous constituents in the soil-pore
liquid should be established by installing two monitoring devices at random
locations for each soil series present in the treatment zone. Samples should
be taken on at least a quarterly basis for at least one year and can be
composited to give one sample per quarter. Analysis of these samples should
be used to calculate an arithmetic mean and variance for each hazardous
constituent. After background values are established, additional soil-pore
liquid samples should occasionally be taken to determine if the background
values are changing over time.
The number of soil-pore liquid samplers needed is a function of site
factors that influence the variability of leachate quality. Active, uniform
areas should receive, in the beginning, a minimum of six samplers per uniform
area. For uniform areas >5 ha, at least two samplers per 1.5 ha (4 ac) should
be installed. Samples may be composited in pairs based on location to give
three samples for analysis. The number of devices may have to be adjusted up
(or down) as a function of the variability of results.
To date, most leachate collection has been conducted by scientists and
researchers, and there is not an abundance of available field equipment and
techniques. The U.S. EPA (1977) and Wilson (1980) have prepared reviews of
pressure vacuum lysimeters and trench lysimeters. The pressure vacuum
lysimeters are much better adapted to field use and have been used to monitor
pollution from various sources (Manbeck, 1975; Nassau-Suffolk Research Task
Group, 1969; The Resources Agency of California, 1963; James, 1974). These
pressure vacuum samplers are readily available commercially and are the most
widely used, both for agricultural and waste monitoring uses. A third type of
TWELVE - 12
Revision 0
Date September 1986
-------
leachate sampler is the vacuum extractor as used 1n the field by Smith et al.
(1977). A comparison of in situ extractors was presented by Levin and Jackson
(1977).
These soil-pore liquid sampling devices are described 1n Chapter Nine of
this manual (in Part III, Sampling).
12.4.6 Treatment Zone Monitoring and Sampling Strategy
Treatment zone monitoring of land treatment units Is needed for two
purposes. One main purpose is to monitor the degradation rate of the organic
fraction of the waste material and parameters significantly affecting waste
treatment. Samples are needed at periodic intervals after application to be
analyzed for residual waste or waste constituents. Such measurements need to
be taken routinely, as specified by a soil scientist. These intervals may
vary from weekly to semiannual, depending on the nature of the waste, climatic
conditions, and application scheduling. The second major function of
treatment zone sampling is to measure the rate of accumulation of conserved
waste constituents to provide some Indication of the facility's life.
The sampling schedule and number of samples to be collected may depend on
management factors, but a schedule may be conveniently chosen to coincide with
unsaturated zone soil core sampling. For systems that will be loaded heavily
in a short period, more (and more frequent) samples may be needed to ensure
that the waste is being applied uniformly and that the system is not being
overloaded. About seven to ten samples from each selected 1.5-ha (4-ac) area
should be taken to represent the treatment zone, and these should be
composited to obtain a single sample for analysis. In addition, 1f there are
evidently anomalous "hot spots," these should be sampled and analyzed
, separately.
12.4.7 Air Monitoring and Sampling Strategy
The need for air monitoring at a land treatment unit is not necessarily
dictated only by the chemical characteristics of the waste. Wind dispersal of
particulates can mobilize even the most immobile, nonvolatile hazardous
constituents. Therefore, it is suggested that land treatment air emissions be
monitored at frequent intervals to ensure the health and safety of workers and
adjacent residents. This effort may be relaxed 1f the air emissions are
positively identified as innocuous compounds or too low in concentration to
have any effect. Although air monitoring 1s not currently required, 1t 1s
strongly recommended because wind dispersal is a likely pathway for pollutant
losses from a land treatment unit.
Sampling generally involves drawing air over a known surface area at a
known flow rate for a specified time interval. Low-molecular-weight volatiles
may be trapped by solid sorbents, such as Tenax-GC. The high-molecular-weight
compounds may be sampled by Florlsil, glass-fiber filters, or polyurethane
foam.
TWELVE - 13
Revision 0
Date September 1986
-------
12.5 ANALYSIS
12.5.1 Analytical Considerations
Parameters to be measured Include pH, soil fertility, residual
concentrations of degradable rate-limiting constituents (RLC), and the
concentrations of residuals that limit the life of the disposal site (CLC),
plus those that, if Increased in concentration by 25%, would become limiting.
Hazardous constituents of concern should also be monitored. Based on the data
obtained, the facility management or design can be adjusted or actions taken,
as needed, to maintain treatment efficiency. Projections regarding facility
life can also be made and compared with original design projections. Because
the treatment zone acts as an integrator of all effects, the data can be
invaluable to the unit operator.
The analyst should use specific methods in this manual for determining
hazardous.waste constituents.
12.5.2 Response to Detection of Pollutant Migration
If significant concentrations of hazardous constituents (or PHCs) are
observed below the treatment zone, the following modifications to unit
operations should be considered to maximize treatment within the treatment
zone:
1. Alter the waste characteristics.
2. Reduce waste application rate.
3. Alter the method or timing of waste applications.
4. Cease application of one or more particular wastes at the unit.
5. Revise cultivation or management practices.
6. Alter the characteristics of the treatment zone, particularly soil
pH or organic matter content.
12.6 REFERENCES AND BIBLIOGRAPHY
i
12.6.1 References
1. James, I.E., Colliery Spoil Heaps, in J.A. Coler (ed.), Ground Water
Pollution in Europe, pp. 252-255, Water Information Center, Port Washington,
New York, 1974.
2. Levin, M.J. and D.R. Jackson, A Comparison of In Situ Extractors for
Sampling Soil Water, Soil Sci. Soc.'Amer. J., 41, 535-536 (1977).
3. Manbeck, D.M., Presence of Nitrates Around Home Waste Disposal Sites,
Annual Meeting Preprint Paper No. 75-2066, Am. Soc. Agr. Engr., 1975.
TWELVE - 14
Revision 0
Date September 1986
-------
4. Nassau-Suffolk Research Task Group, Final report of the Long Island
Ground Water Pollution Study, New York State Dept. of Health, Albany, New
York, 1969.
5. Permit Guidance Manual on Hazardous Waste Land Treatment Demonstration,
EPA/530-SW-84-015.
6. Permit Guidance Manual on Unsaturated Zone Monitoring for Hazardous Waste
Land Treatment Units, EPA/530-SW-84-016.
7. Smith, J.L., D.B. McWhorter, and R.C. Ward, Continuous Subsurface
Injection of Liquid Dairy Manure, U.S. Environmental Protection Agency, EPA-
600/2-77-117, PB 272-350/OBE, 1977.
6. The Resources Agency of California, Annual Report on Dispersion and
Persistance of Synthetic Detergent 1n Ground Water, San Bernadlno and
Riverside Counties, 1n a report to the State Water Quality Control Board,
Dept. of Water Resources, Interagency Agreement No. 12-17, 1963.
7. U.S. Department of Agriculture, Soil Taxonomy, A Basic System of Soil
Classification for Making and Interpreting Soil Surveys, Soil Conservation
Service, U.S.D.A. Agriculture Handbook No. 436, U.S. Government Printing
Office, Washington, D.C., 1975.
8. U.S. Department of Agriculture, Field Manual for Research 1n Agricultural
Hydrology, U.S.D.A. Agricultural Handbook No. 224, U.S. Government Printing
Office, Washington, D.C., 1979.
9. U.S. Environmental Protection Agency, Procedures Manual for Ground Water
Monitoring at Solid Waste Disposal Facilities, Office of Solid Waste, SW-616,
1977.
10. U.S. Environmental Protection Agency, Criteria and Standards for the
National Pollutant Discharge Elimination System, Title 40 Code of Federal
Regulations Part 125, U.S. Government Printing Office, Washington, D.C., 1981.
11. U.S. Environmental Protection Agency, Hazardous Waste Management System -
- Permitting Requirements for Land Disposal Facilities, Federal Register
47(143). 32274-32388 (July 26, 1982).
12. U.S. Environmental Protection Agency, Hazardous Waste Land Treatment,
Office of Solid Waste and Emergency Response, Washington, D.C., SW-874, 1983.
13. Wilson, L.G., Monitoring 1n the Vadose Zone: A Review of Technical
Elements and Methods, U.S. Environmental Protection Agency, EPA-600/7-80-134,
1980.
TWELVE - 15
Revision
Date September 1986
-------
12.6.2 Bibliography
1. Corey, P.R., Soil Water Monitoring, Unpublished Report to Dept. of Agr.
Engr., Colorado State Univ., Ft. Collins, Colorado, 1974.
2. Duke, H.R., and H.R. Halse, Vacuum Extractors to Assess Deep Percolation
Losses and Chemical Constituents of Soil Water, Soil Scl. Soc. Am. Proc. 37,
963-4 (1973).
3. Parlzek, R.R. and B.E. Lane, Soil-water Sampling Using Pan and Deep
Pressure-Vacuum Lysimeters, J. Hydr. 1JL, 1-21 (1970).
4. Scalf, M.R., J.F. McNabb, W.J. Dunlap, R.L. Cosby, and J. Fryberger,
Manual of Ground Water Sampling Procedures, National Water Well Association,
Worthington, Ohio, 1981.
5. SUkworth, D.R. and D.F. Grigal, Field Comparison of Soil Solution
Samplers, Soil Scl. Soc. Am. J. 45, 440-442 (1981).
6. Trout, T.J., J.L. Smith, and D.B. McWhorter, Environmental Effects of
Land Application of Digested Municipal Sewage Sludge, Report submitted to City
of Boulder, Colorado, Dept. of Agr. Engr., Colorado State Univ., Ft. Collins,
Colorado, 1975.
7. Tyler, D.D. and G.W. Thomas, Lysimeter Measurements of Nitrate and
Chloride Losses and No-tillage Corn, J. Environ. Qual. 6, 63-66 (1977).
8. U.S. Department of the Interior, Ground Water Manual, Bureau of
Reclamation, U.S. Government Printing Office, Washington, D.C., 1977.
9. U.S. Environmental Protection Agency, Hazardous Waste Management Systems
-- Identification and Listing of Hazardous Waste, Federal Register 45 (98),
33084-33133 (May 19, 1980).
10. U.S. Environmental Protection Agency, Ground Water Monitoring Guidance
for Owners and Operators of Interim Status Facilities, Office of Solid Waste
and Emergency Response, Washington, D.C., SW-963, 1982b.
11. Wood, W.W., A Technique Using Porous Cups for Water Sampling at Any Depth
1n the Unsaturated Zone, Water Resources Research 9, 486-488 (1973).
TWELVE - 16
Revision
Date September 1986
-------
CHAPTER THIRTEEN
INCINERATION
13.1 INTRODUCTION
Environmental Protection Agency regulations require owners or operators
of hazardous waste incinerators to perform specific testing prior to issuance
of a final permit. These regulations are contained in 40 CFR Parts 264.340-
264.347, 270.19, and 270.62.
The regulations require that incinerated hazardous wastes be destroyed
with an efficiency of 99.99% or higher. In order to obtain a permit to
incinerate hazardous wastes, owners or operators must demonstrate that their
incinerator can operate at the required efficiency (usually referred to as
destruction and removal efficiency, or ORE). This demonstration will most
often involve a "trial" burn. Prior to the trial burn, the owner or operator
must test the hazardous waste being evaluated for incineration and determine
the presence and concentration of Appendix VIII constituents, along with other
parameters. The analytical results obtained will allow the owner or operator
to determine the principal organic hazardous constituents (POHCs) in the
waste. These POHCs will usually be those compounds in the waste that are
difficult to burn, toxic, and found at reasonably high concentrations in the
waste. During the trial burn, the POHCs are monitored to determine whether
the incinerator is meeting the required ORE.
The owner or operator will then prepare an incineration permit
application, which is submitted to the appropriate state and EPA region.
Contents of permits are listed in Sections 270.14, 270.19, and 270.62 of the
RCRA regulations. As part of the permit application, the owner or operator
will provide the waste analysis information, propose certain POHCs for the
trial burn, and specify the sampling and analysis methods that will be used to
obtain the trial burn data. This portion of the permit application is called
the "trial burn plan." The regulatory agency(ies) will review the application
and trial burn plan, make any necessary modifications, and authorize the owner
to conduct the trial burn. After the trial burn, the results are submitted to
the permit issuance authority and, assuming all requirements are met, a final
incineration permit will be issued. The permit contains all the information
pertaining to the licensed operation of the incinerator, and the owner or
operator must comply with whatever conditions are specified in the permit.
The rest of this chapter will explain the various sampling and analysis
strategies that can be used during the trial burn and how analysis data can be
used to obtain a final permit.
13.2 REGULATORY DEFINITION
As explained earlier, incinerator regulations are contained in 40 CFR
Parts 264.340-.347, 270.19, and 270.62. Because Part 264 contains general
requirements for hazardous waste incineration, it will not be discussed here.
THIRTEEN - 1
Revision
Date September 1986
-------
Parts 270.19 and 270.62 describe actual sampling and analysis requirements and
are summarized below. A summary of the major analytical requirements Is given
1n this section and 1s followed by sections detailing acceptable sampling and
analysis methods for meeting these requirements.
The trial burn plan must Include the following Items:
1. Heat value of the waste.
2. Viscosity or physical description.
3. A list of hazardous organic constituents that are listed In Appendix
VIII and that are reasonably expected to be present 1n the waste.
4. Approximate concentration of those compounds.
5. A detailed description of sampling and analysis procedures that
will be used.
During the trial burn (or as soon after as possible), the following
determinations must be made:
1. The concentration of trial POHCs in the waste feed.
2. The concentration of trial POHCs, mass emissions, oxygen, and
hydrogen chloride in the stack gases. (Determination of the oxygen
and water concentration in the stack exhaust gas concentration is
necessary for correction of measured particulate.)
3. The concentration of trial POHCs in any scrubber water, ash, or
other residues that may be present as a result of the trial burn.
4. A computation of the ORE.
For routine operation, the only explicit sampling and analysis require-
ment is the determination of carbon monoxide in the stack gas. Although the
permit writer or the state/local authorities may Impose additional monitoring
requirements 1n some Instances, 1t is not anticipated that comprehensive
sampling of the stack-gas effluent or specific analysis of POHCs will be
required, except in trial burn situations.
13.3 WASTE CHARACTERIZATION STRATEGY
13.3.1 Sampling
Acquisition of a representative sample of hazardous waste for subsequent
chemical analysis is accomplished by preparing a composite of several
subsamples of the waste. Sampling equipment and tactics for collection of the
subsamples are specified in Chapter Nine of this manual and generally Involve
grab sampling of liter- or kilogram-sized portions of waste materials. To
THIRTEEN - 2
Revision 0
Date September 1986
-------
ensure that the bulk of the waste 1s represented by the composite sample, the
sampling strategy requires collection of a minimum of four subsamples that
provide integration over both the depth and the surface area of the waste as
contained in drums, tanks, holding ponds, etc. The composite sample prepared
in the field must be mixed thoroughly and split into at least three replicate
samples prior to shipment to the analytical laboratory. This step is
primarily a precaution against breakage or loss of sample, but it also
provides the potential for a check on the homogeneity of the composite sample.
To ensure that sampling and analysis results will withstand legal scrutiny,
chain-of-custody procedures are incorporated into sampling protocols. The
sampling protocols also include explicit provisions for ensuring the safety of
the personnel collecting the samples.
13.3.2 Analysis of Hazardous Wastes
The overall strategy for waste characterization includes test procedures
(to determine the characteristics of the waste) and analysis procedures (to
determine the composition of the waste). The analysis procedures can be
divided Into three sections:
1. Characteristics (useful for storage, etc.; not required).
2. Proximate analysis (useful data but not required, except for
heat value).
3. Specific analysis (required for determination of POHCs).
Figure 13-1 provides an overview of this analytical approach. The
discussion below provides a capsule description of each major element of this
scheme and the use of the resulting information in the hazardous waste
Incineration permitting process.
13.3.2.1 Characteristics
The characteristics of the waste sample, defined in terms of
ignitability, corroslvity, reactivity (including exploslvity and toxic gas
generation), and extraction procedure toxicity, are determined according to
the procedures presented in Chapter Eight of this manual. These tests are
performed on a sample from each waste stream, unless there 1s sufficient
information from an engineering analysis to indicate the waste meets any of
these criteria. This information is relevant to the Part 264, Subpart B,
General Waste Analysis requirement in that 1t affects procedures for safely
storing, handling, and disposing of the waste at the facility. The data are
also relevant to possible exclusion from the trial burn requirements of Part
122. The data on the characteristics of each hazardous waste may be available
from the waste generator and from manifest or shipping papers received by the
facility owner/operator.
THIRTEEN - 3
Revision 0
Date September 1986
-------
Composite Waste Sample
I
m
m
O 73
D> (0
CO O
Characteristics
—Ignitability
—Gorrosivity
—Explosivity
-Reactivity
—EPToxicity
1
Composition
I
Proximate Analysis
Approximate Mass Balance:
—Moisture Content
—Solid Content
—Ash Content
—Elemental Analysis
—Heating Value of the Waste
-Viscosity
1
Specific Analysis
—Identification and
Quantification of
Hazardous Constituents
Selected from the
Appendix VIII List
-Metals
Figure 13-1 Overview of the analytical approach for waste characterization.
-------
13.3.2.2 Proximate Analysis
The proximate analysis provides data relating to the physical form of the
waste and an estimate of its total composition. This analysis includes
determination of:
1. Moisture, solids, and ash content.
2. Elemental composition (carbon, nitrogen, sulfur, phosphorus,
fluorine, chlorine, bromine, iodine to 0.1% level).
3. Heating value of the waste.
4. Viscosity.
Some or all of this information may satisfy the waste analysis
requirements of the Part 264 regulations, as well as be responsive to the
General Waste Analysis requirements of Subpart B. The elemental composition
data allow one to predict if a high concentration of potentially significant
combustion products (NOX, SOX, ?2Q5i hydrogen halides, and halogens) might be
formed during incineration. These data also facilitate an informed selection
of the Appendix VIII hazardous constituents that might be present in the waste
by indicating whether the overall waste composition and hence the types of
components present are consistent with expectations based on best professional
judgment. For example, if bromine were not present in the waste, any
organobromlne compounds from Appendix VIII at levels of 1,000 mg/kg would be
excluded from specific analysis.
13.3.2.3 Specific Analysis
The specific analysis portion of the waste characterization scheme
provides qualitative confirmation of the presence and identity of the Appendix
VIII constituents that might reasonably be expected to be present in the
waste, based on professional judgment or on the results of proximate analysis.
It is important to note that specific analysis does not involve screening
every waste sample for all Appendix VIII hazardous components. A preliminary
judgment is made as to the compounds or types of compounds that are actually
present.
For the specific organic analyses, a high-resolution separation technique
(fused-silica capillary gas chromatography) and a high-specificity detection
technique (mass spectrometry) are used wherever possible. This approach
ensures qualitative and quantitative analysis for a variety of waste types and
process chemistries.
Specific analysis methods in this manual can be used for Appendix VIII
constituents. Generally, the methods of choice for Appendix VIII components
will be:
THIRTEEN - 5
Revision 0
Date September 1986
-------
Method 6010 (Inductively Coupled Plasma Method)
Method 8270 (GC/MS Method for Semi volatile Organlcs:
Capillary Column Technique)
Method 8240 (GC/MS Method for Volatile Organlcs)
Other more specific methods contained In this manual may be used;
however, they cannot screen for a wide range of compounds. For example,
Method 8010 can detect only those volatile compounds containing halogen.
13.3.3 Selection of POHCs
The criteria for selection of POHCs (typically one to six specific
constituents per waste feed) Include:
1. The expected difficulty of thermal degradation of the various
hazardous organic constituents 1n the waste.
2. The concentration of those constituents 1n the waste.
It 1s anticipated that the designation of POHCs will be negotiated on a
case-by-case basis for each permit application. It is Important to note that
it is not necessarily, or even generally, true that all Appendix VIII
compounds present In the waste will be designated as POHCs. The intent is to
select a few specific compounds as indicators of incinerator performance. The
selected compounds should provide a sufficiently stringent test of the
incinerator's performance to ensure that incineration of the waste can be
carried out in an environmentally sound fashion. This criterion mandates
selection of the more thermally stable constituents as POHCs.
At the same time, however, 1t is necessary that the designated POHCs be
present 1n the waste 1n sufficiently high concentrations in order to be
detected 1n the stack gas. This is a particularly Important constraint for
wastes that are to be Incinerated with substantial quantities of auxiliary
fuel, which effectively dilute the POHCs in the exhaust gas. Although the
burning of auxiliary fuel might not affect the mass emission rate of POHCs, it
would lead to an increased volumetric flow of stack gas and thus to a
decreased concentration of POHCs at the stack. This lower concentration
directly affects the detection limit achievable for a given stack-gas sample
size (e.g., between 5 m3 and 30 m3).
It is recommended that, whenever possible, the permit writer select POHCs
present in the waste at 1,000 mg/kg or higher. If it is considered desirable
to designate as a POHC a thermally stable compound present at the hundreds-of-
parts-per-m1H1on level, the trial burn permit application must include
calculations and supporting data to Indicate that 0.01% of the mass feed rate
of that component in the waste ,could in fact be detected in the stack
effluent. A waste concentration of 100 mg/kg probably represents a practical
lower level below which determination of 99.99% ORE may require extraordinary
THIRTEEN - 6
Revision 0
Date September 1986
-------
sampling analysis and quality control procedures, which may significantly
Increase the sampling and analysis costs for that trial burn.
For a waste material that 1s a listed hazardous waste under RCRA
regulations (40 CFR Part 261, Subpart D), the constituents that caused the
Administrator to 11st the waste as toxic (tabulated 1n Appendix VII of 40 CFR
Part 261) would be logical candidates for designation as POHCs, 1f these
constituents are organic chemicals.
13.4 STACK-GAS EFFLUENT CHARACTERIZATION STRATEGY
The overall strategy for hazardous-waste-lncinerator stack-gas effluent
characterization to determine compliance with Part 264 performance standards
is to collect replicate 3- to 6-hr, 5- to 30-m3 samples of stack gas using a
comprehensive sampling train, such as the EPA Modified Method 5 Sampling Train
(MM5), the EPA/IERL-RTP Source Assessment Sampling System (SASS), or, for the
volatile species, the Volatile Organic Sampling Train (VOST). These three
strategies are described in detail in Chapter Ten (Methods 0010, 0020, and
0030). Any of the comprehensive sampling trains provides a sample sufficient
for determination of particulate mass loading, concentrations of particulate
and low-volatility vapor-phase organics, and concentrations of particulate and
volatile metals. The VOST is used to collect the sample to be analyzed for
volatile organic species. For burns of wastes that could also produce
significant emissions of HC1, an MM5 type of train is used to collect and
quantify HC1 in the stack gas.
Figure 13-2 shows an overview of the analysis scheme for stack-gas
samples. A separate sample (cyclone and particulate catch) will be used for
determination of particulate mass loading and extraction of nonvolatile
organic components. Heating during the particulate determination may drive
off semi volatile organics. Volatile organic components of the stack gas will
be collected using the VOST.
The directed analysis shown in Figure 13-2 is performed on triplicate
samples. Although analysis of only two samples would allow an average level
of a POHC to be determined, at least three samples should be analyzed so that
an error bound for the measured values can be computed. The incremental cost
of the replicate sampling and analysis is offset by increased confidence in
the resulting data; quantitative results from a single sampling and analysis
run should not generally be considered as an acceptable indicator of
performance.
The survey analysis, which is a qualitative screen of the collected
material to ensure that potentially hazardous but unexpected emissions do not
go overlooked, need be performed on no more than one stack-gas sample. During
a trial burn, the oxygen level in the stack gas must be measured using an
Orsat or Fyrite analyzer, as detailed 1n 40 CFR Part 60, Appendix A, Method 3,
so that the particulate loading may be corrected to a standard excess air
level.
THIRTEEN - 7
Revision 0
Date September 1986
-------
I
00
O 73
0> (D
ct <
(B -*
V)
(/» O
a
o
Probe Wash
Paniculate Catch
Conomtrau to Orynns
Dry
Wet*
Wei*
Combine Extracts
Aliquot (~10%)
Metals by ICAP
(It any metals
in waste)
Soxhlet Extraction
Concentrate*
Specific Analysis
Sorbent Trap
1
Comfcnuie
Soxhtet Extraction
LxjuHJ/Liquid
Extraction
Combine Extracts
Concentrate*
Specific Analysis
Tenax Sortenl Traps
Desorption/
Concentration
Impinges
Chloride Analysis
Metal Analysis by (CAP
(M my metals present
in waste)
Specific Analysis
"As «n alternative, the extracts Irom particulaw «nd vapor portions of lh« u«n may be combirMd prior to dialysis.
Figure 13-2 Overview of an analysis scheme for stack gas samples from a comprehensive sampling train.
-------
For both trial and operating burns, on-Hne monitors (nondisperslve
Infrared Instruments) are used to provide continuous readings of carbon
monoxide levels 1n the incinerator effluent.
13.5 ADDITIONAL EFFLUENT CHARACTERIZATION STRATEGY
The basic strategy for sampling scrubber water, ash, and other residue
(if any) is to prepare composite samples from grab subsamples, collected using
the same types of sampling devices and tactics as those used for waste
characterization. This sampling is required only during trial burns, in
accordance with 40 CFR Part 270.62. These additional effluent samples are
analyzed for POHCs to determine appropriate disposal or subsequent treatment
methods and to ensure that significant discharges of POHCs 1n other media do
not go undetected.
13.6 SELECTION OF SPECIFIC SAMPLING AND ANALYSIS METHODS
The preceding discussion has briefly described the RCRA regulations that
define sampling and analysis requirements for hazardous waste Incineration and
has presented an overview of the sampling and analysis procedures developed to
meet these requirements.
This section will illustrate, by means of a hypothetical example, the
transition from strategies, as described above, to methods, as described
below. In the interest of clarity, the example 1s oversimplified, but should
serve as a demonstration of how to develop and evaluate a hazardous waste
incineration trial burn plan. The discussion will deal with sampling and
analysis considerations only and will not address adequacy of design,
operating conditions, or other engineering considerations.
13.6.1 Scenario
The owner/operator of an incineration facility seeks an RCRA permit to
treat chlorinated organic waste material.
The facility is a liquid Injection Incinerator with a capacity of 10 x
106 Btu/hr and equipped with a wet scrubber for acid-gas removal. A waste oil
«0.1% chlorine) is burned as auxiliary fuel. The proposed operating
conditions for hazardous waste Incineration Include a combustion zone
temperature of 2000*F (1100*C) and a residence time of 2 sec with 150% excess
air.
The waste is a still bottom from the production of perchloroethylene.
Based on engineering analysis, it 1s expected to be a nonvlscous organic
liquid with a heating value >5,000 Btu/lb. The major components of the waste
are expected to be highly chlorinated species such as hexachlorobenzene,
hexachlorobutadlene, and other chlorinated aliphatic and aromatic compounds.
THIRTEEN - 9
Revision
Date September 1986
-------
13.6.2 Strategy
There are Insufficient data from other trial or operating burns to
specify operating conditions under which this type of facility, when burning
this type of waste, has been demonstrated to comply with the Part 264
performance criteria. Therefore, a trial burn will be required.
There are Insufficient data to develop the trial burn plan available from
the waste generator. Therefore, additional analyses of the waste will be
necessary to support the trial burn, permit application. The POHCs for which
destruction and removal efficiencies are to be demonstrated 1n the trial burn
must be designated, based on review of existing Information and/or additional
analysis of a representative sample of the waste.
Because the owner/operator plans to operate the facility under one set of
temperature, residence time, and excess air conditions when treating hazardous
waste, the-trial burn will consist of three replicate tests under that set of
operating conditions.
The trial burn sampling and analysis strategy must address:
1. The waste analysis requirements of 40 CFR Part 270.
2. The performance standards of 40 CFR Part 264, Subpart 0.
3. The monitoring requirements of 40 CFR Part 264, Subpart 0.
13.6.2.1 Sampling Strategy
During each of the three replicate tests, the following samples must be
obtained:
1. One composite sample of the waste actually treated.
2. One time-averaged (3-4 hr) sample of stack gas.
3. One composite sample of spent scrubber water.
No bottom ash or fly ash. streams (other than the stack particulate
emissions) are expected to be generated as effluents from this facility.
13.6.2.2 Analysis Strategy
The waste must be analyzed to determine,:
1. Quantity of designated trial burn POHCs.
2. Heating value of the waste.
THIRTEEN - 10
Revision
Date September 1986
-------
3. Viscosity or physical form.
4. Quantity of organically bound chlorine. (This analysis 1s not
mandatory; however, the data obtained may be helpful 1n determining
a potential for HC1 emissions.)
5. Identity and approximate quantity of known or suspected Appendix
VIII constituents.
The stack gas must be analyzed to determine:
1. Quantity of designated trial burn POHCs.
2. Quantity of partlculate matter emissions.
3. Quantity of hydrochloric add emissions.
4. Carbon monoxide level.
5. Excess air level (oxygen/carbon dioxide level determination).
The scrubber water must be analyzed to determine quantities of designated
trial burn POHCs.
13.6.3 Tactics and Methods
13.6.3.1 Selection of POHCs
The first step 1s to obtain a composite of the waste and to analyze it
for Appendix VIII constituents. In this case the waste was sampled from a
tank truck by taking a series of vertical cores at the available hatch
location on the truck. The cores were obtained by using a Coliwasa (see
Section 9.2.2.4 of Chapter Nine) and following the procedures. After the
waste sample was collected, it was sent to the laboratory using cha1n-of-
custody procedures (Section 9.2.2.7 of Chapter Nine) and was analyzed using
Method 8270 (Chapter Four) (1n this case the sample was directly injected with
a split ratio of 100:1). The sample was also analyzed by Method 9020, Chapter
Five. Table 13-1 summarizes the Information that was obtained for the waste
analysis. The major organic components that would appear to be candidates for
selection as POHCs are listed in Table 13-2, along with relevant
physical/chemical properties and recommended stack sampling and analysis
methods.
The permit writer has designated hexachloro-butadlene, hexachlorobenzene,
and hexachloroethane as POHCs. All three species are present in significant
concentrations 1n the waste and will remain at >1,000 mg/kg concentration even
if the waste were cut by as much as 1:10 with auxiliary fuel 1n order to limit
the total chlorine feed rate and to maintain an adequate heating value in the
total incinerator feed. Fully chlorinated species such as these are generally
considered to be highly resistant to thermal degradation and thus provide an
appropriate set of POHCs for ORE determination.
THIRTEEN - 11
Revision 0
Date September 1986
-------
TABLE 13-1. INFORMATION ON COMPOSITION OF HYPOTHETICAL WASTE
Visual Inspection; The waste was a pitch-black, nonvlscous liquid with
obvious partlculate loading. It had a pungent odor and fumed slightly when
the cap was removed.
Loss on Ignition: Ignition at 900*C resulted in a 99.8% loss of mass.
Higher Heating Value; The waste would not burn in a bomb calorimeter; Its
higher heating value is estimated at approximately 2,000 Btu/lb.
TOX: 74.4% Cl.
GC/MS; This analysis Indicates that hexachlorobutadiene 1s the major
component (65%) and hexachlorobenzene is present at about 10% of the Total
Organic Chlorine concentration. Other peaks 1n the chromatogram were
identified as hexachloroethane (approx. 4%), tetrachloroethanes (approx. 3%),
tetrachloroethylene (approx. 0.1%), plus four other chlorinated allphatlcs at
about 0.5% concentration of the CC1 concentration.
Summary; All of the available evidence suggests that this waste contains
essentially no perchloroethylene, that hexachlorobutadiene makes up about 65%
of the waste, and that there are perhaps a dozen other components at 1-5%
concentration. All of the minor components appear to be chlorinated, with
hexachlorobenzene the most abundant.
THIRTEEN - 12
Revision
Date September 1986
-------
TABLE 13-2
CANDIDATE POHCs FOR HYPOTHETICAL WASTE AND
RECOMMENDED STACK SAMPLING AND ANALYSIS METHODS
FOR HYPOTHETICAL TRIAL BURN
Stack Sampling Method
AH3 MM. Section
in waste (%) (°C) (kcal/mole) (g/mole) number Description
Analysis Method
Approx. con-
Compound centration B.P.
(POHC)
Method
number
Description
Hexachloro- 65
butadiene
Hexachloro- 6
benzene
Hexachloro- 2
ethane
Tetrachloro- 1.5
ethane0
Tetrachloro-
etnylene
O.I
215 N/A 260.76 1.2.1.8 »6 - Sorbent 8120,8250, GC/MS Extract-
or 8270 ables
323 567.7 284.8 1.2.1.8b Mtf - Particu- 8120,8250, OC/MS Extract-
late and or 8270 ables
Sorbents
186.8 173.8 236.74 1.2.1.8b »6 - Sorbent 8120,8250, GC/MS Extract-
or 8270 ables
130.5 230 167.84 1.2.1.13 VOST 8010 or GC/MS Volatiles
(146.2) (233) 8240
121 JO 197 165.85 1.2.1.13 VOST 8010 or GC/MS Volatiles
8240
The standard enthalpy of combustion.
The SASS method (Chapter Nine, Method 0020) could also be selected. A
specially fabricated glass-lined SASS train might be necessary to withstand
the hydrochloric acid expected in the stack.
cNumbers given in parentheses refer only to 1 ,1 ,2,2-tetrachloroethane.
THIRTEEN - 13
Revision 0
Date September 1986
-------
13.6.3.2 Selection of Sampling Methods
For sampling of wastes and liquid and solid effluents, the choice of
method 1s based primarily on the nature of the medium. Review of available
methods Indicates that for dipper sampling (Chapter Nine) or sampling from the
tap of the waste-feed pipe would be appropriate for collection of discrete
subsamples of waste feed and of spent scrubber water at regular time Intervals
over the duration of each trial burn. These would then be combined to form
the corresponding composite samples for each test.
For sampling of stack gas, both the nature of the medium and the nature
(volatility, stability) of the POHC or other target species affect the choice
of a sampling method. Table 13-2 summarizes these recommendations for the
candidate POHCs 1n this example. Note that designation of tetrachloroethylene
as a POHC 1n this Instance would require use of VOST, although the MM5 or SASS
approaches would collect all of the other candidate POHCs.
The MM5 train would also suffice to determine compliance with the two
other performance standards of 40 CFR Part 264. The particulate matter
emission rate can be determined from the mass of material collected 1n the
probe wash, cyclone (1f any), and filter of the MM5 train. The hydrochloric
acid emission rate can be determined by using caustic scrubbing solution In
the 1mp1nger portion of the MM5 train and determining the hydrochloric add
level as chloride.
In addition to the procedures chosen for the collection of POHCs, it
would be necessary to specify procedures for the required monitoring for
carbon monoxide and oxygen levels 1n the stack gas.
13.6.3.3 Selection of Analysis Methods
The analytical procedures used for qualitative identification and
quantitative determination of POHCs and other target species are determined
primarily by the nature (volatility, polarity) of the species sought.
This manual lists recommended analysis methods for each candidate POHC
after the appropriate sample preparation steps 1n Methods 0010, 0020, and 0030
have been performed. Table 13-2 summarizes the recommendation for analysis of
the candidate POHCs 1n this hypothetical example. Note that a single
analytical method suffices to determine all of the hexachlorospecies of
concern here although an additional method would be recommended 1f the
analysis were to Include the tetrachloroethanes and tetrachloroethylene.
13.6.4 Results and Calculations
This section Illustrates the proper methods for calculating ORE,
corrected particulate loading, and HC1 emissions for the hypothetical example
described above. Again, this example has been somewhat oversimplified for
purposes of illustration.
THIRTEEN - 14
Revision 0
Date September 1986
-------
According to 40 CFR Part 264, the ORE for each POHC 1s calculated as:
u - u
W1n "out
ORE = x 100%
W1n
where:
= mass feed rate of one POHC 1n the waste stream feeding the
Incinerator.
W0ut = mass emission rate of the same POHC present 1n stack
exhaust emissions.
13.6.4.1 Calculation of Win (Ib/hr):
Cy^ X iRyy
w. = —
1n 100
where:
Cw = Concentration of one POHC In the waste, %.
FRW = Mass feed rate of waste to the Incinerator, Ib/hr.
Assume that quantitative analysis of a representative aliquot drawn from
the composite waste sample from test No. 1 gave the following concentrations:
hexachlorobutadlene 63 %
hexachlorobenzene 9.4%
hexachloroethane 1.1%
Further, assume that the thermal capacity of the facility (10 x 106
Btu/hr) was met by blending waste 1:10 with waste oil to give a feed mixture
that was 8.2% chlorine and that had a heating value of 16,400 Btu/lb. The
total mass feed rate to the Incinerator was therefore 600 Ib/hr, of which 540
Ib/hr was auxiliary fuel (waste oil) and 60 Ib/hr was chlorinated waste.
The Wfn values for the three POHCs are therefore:
hexachlorobutadlene (.63 x 60 Ib/hr) 38 Ib/hr
hexachlorobenzene (.094 x 60 Ib/hr) 5.6 Ib/hr
hexachloroethane (.011 x 60 Ib/hr) 0.66 Ib/hr
THIRTEEN - 15
Revision
Date September 1986
-------
13.6.4.2 Calculation of Wnut (Ib/hr);
wout = cs x ERs x 1.32 x lO'4
where:
Cs = Concentration of one POHC in the stack gas effluent,
mg/dNm3.
ERS = Volumetric flow rate of stack gas, dNm3/min.
1.32 x 10~4 = Conversion factor from mg/min to Ib/hr.
Assume that quantitative analysis of the extract prepared from the time-
integrated comprehensive sampling train sample from test No. 1 gave the
following concentrations in the sampled gas:
hexachlorobutadiene 0.080 mg/m3
hexachlorobenzene 0.020 mg/m3
hexachloroethane £0.004 mg/m3
Further, assume that the average measured volumetric flow of stack gas
during test No. 1 was 3,200 scfm or 90 dNm3/min.
The Wout values for the three POHCs are therefore:
hexachlorobutadiene (.080 x 90 x 1.32 x lO'4) 9.5 x 10~4 Ib/hr
hexachlorobenzene (.020 x 90 x 1.32 x 10~4) 2.4 x 10~4 Ib/hr
hexachloroethane (£0.004 x 90 x 1.32 x lO'4) <0.48 x 1Q-4 Ib/hr
13.6.4.3 Calculation of ORE:
W1n " Wout
ORE = x 100
Win
The ORE values for the three POHCs are therefore:
hexachlorobutadiene 99.997
hexachlorobenzene 99.996
hexachloroethane >99.993
Note that compliance with a "four-9's" performance standard could not
have been demonstrated 1n this particular example for a component present at
<1% 1n the waste itself (or <1,000 mg/kg 1n the 1:10 waste:fuel blend fed to
THIRTEEN - 16
Revision
Date September 1986
-------
the Incinerator) unless the detection limit for that component In the stack
gas were <4 ug/m3.
In this example, compliance with the 99.99% ORE performance standard has
been demonstrated, In one test, for each of the three POHCs. If these results
were supported by data from the other two replicate trial burn tests, the
"four-9's" ORE could be considered to have been established.
13.6.4.4 Calculation of HC1 Emissions
An Incinerator burning highly chlorinated hazardous waste capable of
producing significant stack-gas emissions of hydrogen chloride (HC1) must
monitor and/or control HC1 emissions.
The hypothetical waste in this example contains approximately 75%
chlorine by weight (Table 13-1). At the proposed 60-lb/hr feed rate of waste
that is blended 1:10 with auxiliary fuel for a total feed of 600 Ib/hr (9.8 x
10^ Btu/hr), the maximum HC1 emission rate would be 45 Ib/hr of chlorine basis
or 46 Ib/hr as HC1. This rate exceeds the regulatory limit of 4 Ib/hr;
therefore, the scrubber efficiency must be determined.
The stack emission rate of HC1 can be calculated from measured values in
the following manner:
HC1out ' C1n x ERs x i'32 x 10'4
where:
Cjn = Concentration of HC1 in the stack-gas sample
(mg/m3).
ERS = Volumetric flow rate of the stack gas, m3/min.
1.32 x 10'4 = Conversion factor from mg/min to Ib/hr.
Assume that quantitative analysis of the impinger/condensate solution
from the time-integrated comprehensive sampling train from test No. 1 gave
34 mg/m3 HC1 in the stack effluent.
The stack emission rate of HC1 is calculated by:
HC1out = 34 m9/m3 (90 m3/min) (1.32 x 10'4)
= 0.40 Ib/hr HC1.
This emission level is <1% of the 46 Ib/hr of HC1 potentially generated
from the waste, an indication that the removal efficiency of the wet scrubber
was >99%.
THIRTEEN - 17
Revision
Date September 1986
-------
13.6.4.5 Calculation of Particulate Loading (mg/m3)
>,
An Incinerator-burning hazardous waste must not emit particulate matter
1n excess of 180 mg/dscm when corrected to an oxygen concentration of 7% 1n
the stack gas. <
Assume that prior to chemical- analysis, particul ate samples from the
stack effluent of the hypothetical waste (from probe washes and filter catches
of the time-Integrated comprehensive sample train) were dried and'weighed.
The hypothetical particulate loading from these measurements was calculated to
be 80 mg/m3 at the actual excess air level of the stack. The excess air level
was determined to be 150%, based on hypothetical measured values of oxygen
(12.8%) and carbon dioxide (6.7%). Correction to standard excess air level,
as specified 1n the Part 264 regulations, leads to a participate loading of
140 mg/m3 (0.06 gr/scf). This total particulate emission 1s in compliance
with the Part 264 performance standard that specifies £180 mg/m3
(£0.08 gr/scf).
13.6.5 Summary
Incinerator performance in this example complies with the Part 264
Subpart 0 Incinerator Standards as they relate to:
1. Destruction and Removal Efficiency. All three POHCs showed
compliance with the 99.99% ORE performance standard.
2. Limitation on HC1 Emissions. The HC1 emission rate of 0.40 Ib/hr
shows compliance with a 99% removal standard for HC1.
3. Limitation on Stack Emissions of Partlculate Material. The
corrected particulate loadingo?140mg/m3isless than the 180
mg/m3 standard for particulate loading (corrected to a standard
excess air level).
13.7 REFERENCES
1. Addendum to Specifications for Incinerator Testing at Federal Facilities,
PHS, NCAPC, Dec. 6, 1967.
2. American Society for Testing and Materials, Gaseous Fuels; Coal and Coke;
Atmospheric Analysis, Part 26 (pp. 617-622) of Annual Book of ASTM Standards,
Philadelphia, Pennsylvania, 1974.
3. Felix, L.G., G.I. CUnard, G.E. Lacey, and J.D. McCain, Inertlal Cascade
Impactor Substrate Media for Flue Gas Sampling, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, EPA-600/7-77-060, June 1977.
THIRTEEN - 18
Revision 0
Date September 1986
-------
4. Martin, Robert M., Construction Details of Isok1net1c Source-Sampling
Equipment, U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, APTD-0581, April 1971.
5. Rom, Jerome, J., Maintenance, Calibration and Operation of Isok1net1c
Source Sampling Equipment, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, APTD-0576, March 1972.
6. Shlgehara, R.T., Adjustments In the EPA Nomography for Different P1tot
Tube Coefficients and Dry Molecular Weights, Stack Sampling News 2:4-11
(October 1974).
7. Smith, W.S., R.T. Shlgehara and W.F. Todd, A Method 1n Interpreting Stack
Sampling Data, Paper presented at the 63rd Annual Meeting of the A1r Pollution
Control Association, St. Louis, Missouri, June 14-19, 1970.
8. Smith, W.S., et al_., Stack Gas Sampling Improved and Simplified with New
Equipment, APCA Paper No. 67-119, 1967.
9. Specifications for Incinerator Testing at Federal Facilities, PHS NCAPC,
1967.
10. Vollaro, R.F., A Survey of Commercially Available Instrumentation for the
Measurement of Low-Range Gas Velocities (unpublished paper), Emission
Measurement Branch, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, November 1976.
THIRTEEN - 19
Revision
Date September 1986
-------
APPENDIX
COMPANY REFERENCES
The following listing of frequently-used addresses Is provided for the
convenience of users of this manual. No endorsement 1s Intended or Implied.
Ace Glass Company
1342 N.W. Boulevard
P.O. Box 688
Vlneland, NJ 08360
(609) 692-3333
Aldrlch Chemical Company
Department T
P.O. Box 355
Milwaukee, WI 53201
Alpha Products
5570 - T W. 70th Place
Chicago, IL 60638
(312) 586-9810
Barneby and Cheney Company
E. 8th Avenue and N. Cassldy Street
P.O. Box 2526
Columbus, OH 43219
(614) 258-9501
B1o - Rad Laboratories
2200 Wright Avenue
Richmond, CA 94804
(415) 234-4130
Burdlck & Jackson Lab Inc.
1953 S. Harvey Street
Nuskegon, MO 49442
Calgon Corporation
P.O. Box 717
Pittsburgh, PA 15230
(412) 777-8000
Conostan Division
Conoco Speciality Products, Inc.
P.O. Box 1267
Ponca City, OK 74601
(405) 767-3456
COMPANIES - 1
Revision
Date September 1986
-------
Corning Glass Works
Houghton Park
Corning, NY 14830
(315) 974-9000
Dohrmann, Division of Xertex Corporation
3240 - T Scott Boulevard
Santa Clara, CA 95050-
(408) 727-6000
(800) 538-7708
E. M. Laboratories, Inc.
500 Executive Boulevard
Elmsford, NY 10523
Fisher Scientific Co.
203 Fisher Building
Pittsburgh, PA 15219
(412) 562-8300
General Electric Corporation
3135 Easton Turnpike
Fa1rf1eld, CT 06431
(203) 373-2211
Graham Manufactory Co., Inc.
20 Florence Avenue
Batavla, NY 14020
(716) 343-2216
Hamilton Industries
1316 18th Street
Two Rivers, WI 54241
(414) 793-1121
ICN Life Sciences Group
3300 Hyland Avenue
Costa Mesa, CA 92626
Johns - Manvllle Corporation
P.O. Box 5108
Denver, CO 80217
Kontes Glass Company
8000 Spruce Street
Vlneland, NJ 08360
M1ll1pore Corporation
80 Ashby Road
Bedford, MA 01730
(617) 275-9200
(800) 225-1380
COMPANIES - 2
Revision
Date September 1986
-------
National Bureau of Standards
U.S. Department of Commerce
Washington, DC 20234
(202) 921-1000
Pierce Chemical Company
Box 117
Rockford, IL 61105
(815) 968-0747
Scientific Glass and Instrument, Inc.
7246 - T Wynnwood
P.O. Box 6
Houston, TX 77001
(713) 868-1481
Scientific Products Company
1430 Waukegon Road
McGaw Park, IL 60085
(312) 689-8410
Spex Industries
3880 - T and Park Avenue
Edison, NJ 08820
Waters Associates
34 - T Maple Street
Mil ford, MA 01757
(617) 478-2000
(800) 252-4752
Whatman Laboratory Products, Inc.
Clifton, NJ 07015
(201) 773-5800
COMPANIES - 3
Revision
Date September 1986
•ft U. S. GOVERNMENT PRINTING OFFICE : 1986 0 - 169-936
------- |