NUREG-1575,Supp.l
EPA 402-R-09-001
DOE/HS-0004
MULTI-AGENCY
RADIATION SURVEY
AND ASSESSMENT OF
MATERIALS AND
EQUIPMENT
MANUAL
(MARSAME)
I
Final
January 2009
-------
NUREG-1575, Supp. 1
EPA402-R-09-001
DOE/HS-0004
Multi-Agency Radiation Survey
and Assessment of Materials and
Equipment Manual
(MARSAME)
Department of Defense
Department of Energy
Environmental Protection Agency
Nuclear Regulatory Commission
January 2009
-------
MARSAME
DISCLAIMER
This supplement was developed by four agencies of the United States Government. Neither the
United States Government nor any agency or branch thereof, or any of their employees, makes
any warranty, expressed or implied, or assumes any legal liability of responsibility for any third
party's use, or the results of such use, of any information, apparatus, product or process disclosed
in this supplement, or represents that its use by such third party would not infringe on privately
owned rights.
References within this supplement to any specific commercial product, process, or service by
trade name, trademark, or manufacturer does not constitute an endorsement or recommendation
by the United States Government.
NUREG-1575, Supp.l ii January 2009
-------
ABSTRACT
The Multi-Agency Radiation Survey and Assessment of Materials and Equipment manual
(MARSAME) is a supplement to the Multi-Agency Radiation Survey and Site Investigation
Manual (MARSSIM) providing information on planning, conducting, evaluating, and
documenting radiological disposition surveys for the assessment of materials and equipment.
MARSAME is a multi-agency consensus document that was developed collaboratively by four
Federal agencies having authority and control over radioactive materials: Department of Defense
(DOD), Department of Energy (DOE), Environmental Protection Agency (EPA), and Nuclear
Regulatory Commission (NRC). The objective of MARSAME is to provide a multi-agency
approach for planning, performing, and assessing disposition surveys of materials and
equipment, while at the same time encouraging an effective use of resources.
January 2009 iii NUREG-1575, Supp. 1
-------
-------
MARSAME
CONTENTS
Disclaimer ii
Abstract iii
Acknowledgements xxi
Acronyms and Abbreviations xxiii
Symbols, Nomenclature, and Notations xxvii
Conversion Factors xxxii
Roadmap RM-1
Introduction to MARSAME RM-1
The Goal of the Roadmap RM-1
Initial Assessment RM-2
Categorization RM-2
Standardized Survey Designs RM-2
Develop a Decision Rule RM-2
Survey Design RM-3
Measurement Quality Objectives RM-4
Implement the Survey Design RM-5
Evaluate the Results RM-5
Summary RM-5
1 Introduction and Overview 1-1
1.1 Purpose and Scope of MARSAME 1-1
1.2 Understanding Key MARSAME Terminology 1-3
1.3 Use of MARSAME 1-5
1.4 Overview of MARSAME 1-6
1.4.1 Planning Phase 1-9
1.4.2 Implementation Phase 1-10
1.4.3 Assessment Phase 1-11
1.4.4 Decision-Making Phase 1-12
1.5 Organization of MARSAME 1-12
1.6 Similarities and Differences Between MARSSIM and MARSAME 1-14
2 Initial Assessment of Materials and Equipment 2-1
2.1 Introduction 2-1
2.2 Categorize the M&E as Impacted or Non-Impacted 2-1
2.2.1 Perform a Visual Inspection 2-3
2.2.2 Collect and Review Additional Historical Records 2-4
2.2.3 Assess Process Knowledge 2-5
2.2.4 Perform Sentinel Measurements 2-7
2.2.5 Decide Whether M&E are Impacted 2-8
2.3 Design and Implement Preliminary Surveys 2-9
2.4 Describe the M&E 2-11
2.4.1 Describe the Physical Attributes of the M&E 2-11
2.4.1.1 Describe the Physical Dimensions of the M&E 2-12
January 2009 v NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
2.4.1.2 Describe the Complexity of the M&E 2-13
2.4.1.3 Describe the Accessibility of the M&E 2-14
2.4.1.4 Describe the Inherent Value of the M&E 2-14
2.4.2 Describe the Radiological Attributes of the M&E 2-15
2.4.2.1 Identify the Radionuclides of Potential Concern 2-16
2.4.2.2 Describe the Radionuclide Concentrations or Radioactivity
Associated with the M&E 2-17
2.4.2.3 Describe the Distribution of Radioactivity 2-17
2.4.2.4 Describe the Location of Radioactivity 2-17
2.4.3 Finalize the Description of the M&E 2-18
2.5 Select a Disposition Option 2-19
2.6 Document the Results of the Initial Assessment 2-19
2.6.1 Document a Standardized Initial Assessment 2-20
2.6.2 Document a Conceptual Model 2-21
3 Identify Inputs to the Decision 3-1
3.1 Introduction 3-1
3.2 Select Radionuclides or Radiations of Concern 3-3
3.3 Identify Action Levels 3-3
3.3.1 Identify Sources of Action Levels 3-6
3.3.2 Finalize Selection of Action Levels 3-7
3.3.3 Modify Action Levels When Multiple Radionuclides are Present 3-7
3.3.3.1 Modify Action Levels for Non-Radionuclide-Specific Measurement
Methods 3-8
3.3.3.2 Modify Action Levels for Non-Radionuclide-Specific
Measurements of Decay-Series Radionuclides 3-9
3.3.3.3 Modify Action Levels for Radionuclide-Specific Measurement
Methods 3-11
3.3.4 Evaluate Interface With Exposure Pathway Models 3-13
3.4 Describe the Parameter of Interest 3-14
3.5 Identify Alternative Actions 3-14
3.6 Identify Survey Units 3-15
3.6.1 Define Initial Survey Unit Boundaries 3-17
3.6.2 Modify Initial Survey Unit Boundaries 3-18
3.7 Develop a Decision Rule 3-18
3.8 Develop Inputs for Selection of Provisional Measurement Methods 3-19
3.8.1 Measurement Method Uncertainty 3-21
3.8.2 Detection Capability 3-21
3.8.3 Quantification Capability 3-22
3.8.4 Range 3-22
3.8.5 Specificity 3-22
3.8.6 Ruggedness 3-23
3.9 Identify Reference Materials 3-24
3.10 Evaluate an Existing Survey Design 3-25
4 Develop a Survey Design 4-1
4.1 Introduction 4-1
4.2 Making Decisions Using Statistics 4-1
NUREG-1575, Supp.l vi January 2009
-------
MARSAME Contents (Continued)
4.2.1 Null Hypothesis 4-2
4.2.2 Discrimination Limit 4-2
4.2.3 Scenario A 4-4
4.2.4 Scenarios 4-4
4.2.5 Specify Limits on Decision Errors 4-5
4.2.6 Develop an Operational Decision Rule 4-5
4.3 Classify the Materials and Equipment 4-6
4.3.1 Class 1 4-6
4.3.2 Class 2 4-7
4.3.3 Class 3 4-7
4.3.4 Other Classification Considerations 4-7
4.4 Design the Disposition Survey 4-8
4.4.1 Scan-Only Survey Designs 4-9
4.4.1.1 Class 1 Scan-Only Surveys 4-13
4.4.1.2 Class 2 Scan-Only Surveys 4-13
4.4.1.3 Class 3 Scan-Only Surveys 4-14
4.4.2 In Situ Survey Designs 4-15
4.4.2.1 Class 1 In situ Surveys 4-15
4.4.2.2 Class 2 In situ Surveys 4-16
4.4.2.3 Class 3 In situ Surveys 4-16
4.4.3 MARSSIM-Type Survey Designs 4-16
4.4.3.1 Class 1 MARSSIM-Type Surveys 4-17
4.4.3.2 Class 2 MARSSIM-Type Surveys 4-18
4.4.3.3 Class 3 MARSSIM-Type Surveys 4-18
4.4.4 Method-Based Survey Designs 4-18
4.4.5 Optimize the Disposition Survey Design 4-19
4.5 Document the Disposition Survey Design 4-20
4.5.1 Routine Surveys and Standard Operating Procedures 4-21
4.5.1.1 SOP Process 4-22
4.5.1.2 General Format for Disposition Survey SOPs 4-22
4.5.2 Case-Specific Applications 4-24
5 Implement the Survey Design 5-1
5.1 Introduction 5-1
5.2 Ensure Protect! on of Health and Safety 5-1
5.3 Consider Issues for Handling M&E 5-3
5.3.1 Prepare M&E for Survey 5-4
5.3.2 Provide Access 5-5
5.3.3 Transport the M&E 5-6
5.4 Segregate the M&E 5-6
5.5 Set Measurement Quality Objectives 5-7
5.6 Determine Measurement Uncertainty 5-9
5.7 Determine Measurement Detectability 5-10
5.8 Determine Measurement Quantifiability 5-12
5.9 Select a Measurement Technique and Instrumentation Combination 5-13
5.9.1 Select a Measurement Technique 5-14
5.9.1.1 Scanning Techniques 5-14
January 2009 vii NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
5.9.1.2 In Situ Measurements 5-14
5.9.1.3 Sampling 5-15
5.9.1.4 Smears 5-15
5.9.2 Select Instrumentation 5-16
5.9.2.1 Hand-Held Instruments 5-16
5.9.2.2 Volumetric Counters (Drum, Box, Barrel, 4-7i Counters) 5-16
5.9.2.3 Conveyorized Survey Monitoring Systems 5-16
5.9.2.4 In Situ Gamma Spectroscopy 5-17
5.9.2.5 Portal Monitors 5-17
5.9.2.6 Laboratory Analysis 5-17
5.9.3 Select a Measurement Method 5-17
5.9.4 Measurement Performance Indicators 5-23
5.9.4.1 Blanks 5-23
5.9.4.2 Replicate Measurements 5-23
5.9.4.3 Spikes and Standards 5-24
5.9.5 Instrument Performance Indicators 5-24
5.9.5.1 Performance Tests 5-24
5.9.5.2 Functional Tests 5-24
5.9.5.3 Instrument Background 5-25
5.9.5.4 Efficiency Calibrations 5-25
5.9.5.5 Energy Calibrations (Spectrometry Systems) 5-25
5.9.5.6 Peak Resolution and Tailing (Spectrometry Systems) 5-25
5.9.5.7 Voltage Plateaus (Gas Proportional Systems) 5-25
5.9.5.8 Self Absorption, Backscatter, and Crosstalk 5-26
5.10 Report the Results 5-26
6 Evaluate the Survey Results 6-1
6.1 Introduction 6-1
6.2 Conduct Data Quality Assessment 6-1
6.2.1 Review the Data Quality Objectives and Survey Design 6-1
6.2.2 Conduct aPreliminary Data Review 6-3
6.2.2.1 Review Quality Assurance and Quality Control Reports 6-3
6.2.2.2 Perform a Graphical Data Review 6-4
6.2.2.3 Calculate Basic Statistical Quantities 6-5
6.2.3 Selectthe Statistical Tests 6-7
6.2.3.1 Scan-Only Surveys 6-7
6.2.3.2 In Situ Surveys 6-8
6.2.3.3 MARSSIM-Type Survey Designs 6-8
6.2.4 Verify the Assumptions of the Tests 6-9
6.2.5 Draw Conclusions from the Data 6-11
6.3 Compare Results to the UBGR 6-12
6.4 Compare Results Using an Upper Confidence Limit 6-12
6.4.1 Calculate the Upper Confidence Limit 6-13
6.4.2 Upper Confidence Limit Example: Class 1 Concrete Rubble 6-14
6.5 Conduct the Sign Test 6-16
6.5.1 Apply the Sign Test to Scenario A 6-17
6.5.2 Apply the Sign Test to ScenarioB 6-17
NUREG-1575, Supp.l viii January 2009
-------
MARSAME Contents (Continued)
6.5.3 Sign Test Example: Class 1 Copper Pipes 6-17
6.6 Conduct the Wilcoxon Rank Sum Test 6-19
6.6.1 Apply the WRS Test to Scenario A 6-19
6.6.2 Apply the WRS Test to Scenario B 6-19
6.6.3 WRS Test Scenario A Example: Class 2 Metal Ductwork 6-20
6.6.4 WRS Test Scenario B Example: Class 2 Metal Ductwork 6-21
6.7 Conduct the Quantile Test 6-23
6.8 Evaluate the Results: The Decision 6-23
6.8.1 Compare Results to the UBGR 6-26
6.8.2 Compare Results Using an Upper Confidence Limit 6-26
6.8.3 Compare Results for MARSSIM-Type Surveys 6-27
6.9 Investigate Causes for Survey UnitFailures 6-27
6.10 Document the Disposition Survey Results 6-28
7 Statistical Basis for MARSAME Surveys 7-1
7.1 Overview of Statistical Survey Design and Hypothesis Testing 7-6
7.2 Statistical Decision-Making 7-9
7.2.1 Null Hypothesis 7-9
7.2.2 Discrimination Limit 7-10
7.2.3 Scenario A 7-12
7.2.4 Scenarios 7-13
7.2.5 Specify Limits on Decision Errors 7-13
7.2.6 Develop an Operational Decision Rule 7-15
7.3 Set Measurement Quality Objectives 7-16
7.3.1 Determine the Required Measurement Method Uncertainty at the UBGR 7-18
7.3.1.1 Scan-Only Survey Designs 7-18
7.3.1.2 In Situ Survey Designs 7-18
7.3.1.3 MARS SIM-Type Survey Designs 7-19
7.3.2 Determine the Required Measurement Method Uncertainty at Concentrations
Other Than the UBGR 7-20
7.4 Determine Measurement Uncertainty 7-22
7.4.1 Use Standard Terminology 7-23
7.4.2 Consider Sources of Uncertainty 7-24
7.4.3 Recommendations for Uncertainty Calculation and Reporting 7-26
7.5 Determine Measurement Detectability 7-27
7.5.1 Calculate the Critical Value 7-28
7.5.2 Calculate the Minimum Detectable Value of the Net Instrument Signal or
Count 7-29
7.5.3 Calculate the Minimum Detectable Concentration 7-31
7.5.4 Summary of Measurement Detectability 7-32
7.5.5 Measurement Detectability Recommendations 7-34
7.6 Determine Measurement Quantifiability 7-34
7.6.1 Calculate the MQC 7-35
7.6.2 Summary of Measurement Quantifiability 7-36
7.7 Establish a Required Measurement Method Uncertainty 7-37
7.7.1 Developing a Requirement for Measurement Method Uncertainty for
MARS SIM-Type Surveys 7-38
January 2009 ix NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
7.7.2 Developing a Requirement for Measurement Method Uncertainty When
Decisions are to be Made About Individual Items 7-40
7.8 Calculate the Combined Standard Uncertainty of a Measurement 7-42
7.8.1 Procedures for Evaluating Uncertainty 7-42
7.8.1.1 Identify the Measurand, 7, and all the Input Quantities, Xt, for the
Mathematical Model 7-42
7.8.1.2 Determine an Estimate, xf, of the Value of Each Input Quantity,
Xt 7-42
7.8.1.3 Evaluate the Standard Uncertainty, u(x^ for Each Input Estimate,
Xj, Using a Type A Method, a Type B Method, or a Combination of
Both 7-42
7.8.1.4 Evaluate the Covariances, u(xt,Xj), for all Pairs of Input Estimates
with Potentially Significant Correlations 7-45
7.8.1.5 Calculate the Estimate, y, of the Measurand from the Relationship
y=flxi,x2,...,xN) 7-45
7.8.1.6 Determine the Combined Standard Uncertainty, uc(y), of the
Estimate,^ 7-45
7.8.1.7 Optionally Multiply uc(y) by a Coverage Factor k to Obtain the
Expanded Uncertainty, U 7-46
7.8.1.8 Report the Result as_y ± C/with the Unit of Measurement 7-47
7.8.2 Examples of Some Parameters that Contribute to Uncertainty 7-47
7.8.2.1 Instrument Background 7-47
7.8.2.2 Counting Efficiency 7-48
7.8.2.3 Digital Displays and Rounding 7-51
7.8.3 Example Uncertainty Calculation 7-52
7.8.3.1 Model Equation and Sensitivity Coefficients 7-52
7.8.3.2 Uncertainty Components 7-53
7.8.3.3 Uncertainty Budget 7-56
7.8.3.4 Reported Result 7-56
7.9 Calculate the Minimum Detectable Concentration 7-57
7.9.1 Critical Value 7-57
7.9.2 Minimum Detectable Concentration 7-58
7.9.3 Calculation of the Critical Value 7-59
7.9.4 Calculation of the Minimum Detectable Value of the Net Instrument
Signal 7-60
7.9.5 Calculation of the Minimum Detectable Concentration 7-61
7.10 Calculate the Minimum Quantifiable Concentration 7-63
7.11 Calculate Scan MDCs 7-66
7.11.1 Calculate the Relative Fluence Rate to Exposure Rate (FRER) 7-67
7.11.2 Calculate the Probability of Interaction 7-67
7.11.3 Calculate the Relative Detector Response 7-69
7.11.4 Relationship Between Detector Response and Exposure Rate 7-69
7.11.5 Relationship Between Detector Response and Radionuclide
Concentration 7-70
7.11.6 Calculation of Scan Minimum Detectable Count Rates 7-72
7.11.7 Calculate the Scan Minimum Detectable Concentration 7-73
NUREG-1575, Supp.l x January 2009
-------
MARSAME Contents (Continued)
8 Illustrative Examples 8-1
8.1 Introduction 8-1
8.2 Mineral Processing Facility Concrete Rubble 8-1
8.2.1 Description 8-2
8.2.2 Objectives 8-2
8.2.3 Initial Assessment of the M&E 8-2
8.2.3.1 Categorize the M&E as Impacted or Non-Impacted 8-2
8.2.3.2 Describe the M&E 8-3
8.2.3.3 Design and Implement Preliminary Surveys 8-4
8.2.3.4 Select a Disposition Option 8-7
8.2.3.5 Document the Results of the Initial Assessment 8-8
8.2.4 Develop a Decision Rule 8-8
8.2.4.1 Select Radionuclides or Radiations of Concern 8-8
8.2.4.2 Identify Action Levels 8-8
8.2.4.3 Modify the Action Levels to Account for Multiple
Radionuclides 8-10
8.2.4.4 Describe the Parameter of Interest 8-11
8.2.4.5 Identify Alternative Actions 8-12
8.2.4.6 Identify Survey Units 8-12
8.2.4.7 Define the Decision Rules 8-12
8.2.4.8 Develop Inputs for Selection of Provisional Measurement
Methods 8-13
8.2.4.9 Identify Reference Materials 8-14
8.2.5 Develop a Survey Design 8-14
8.2.5.1 Classify the M&E 8-14
8.2.5.2 Design the Scanning Survey 8-15
8.2.5.3 Design the Sample Collection Survey 8-15
8.2.5.4 Develop an Operational Decision Rule 8-16
8.2.5.5 Document the Survey Design 8-16
8.2.6 Implement the Survey Design 8-16
8.2.6.1 Ensure Protection of Health and Safety 8-16
8.2.6.2 Consider Issues for Handling the M&E 8-18
8.2.6.3 Segregate the M&E 8-18
8.2.6.4 Set Measurement Quality Objectives 8-18
8.2.6.5 Determine Measurement Uncertainty for the ScanMDC 8-18
8.2.6.6 Determine Measurement Uncertainty for Concrete Samples 8-22
8.2.6.7 Collect Survey Data 8-23
8.2.7 Evaluate the Survey Results 8-24
8.2.7.1 Conduct a Data Quality Assessment 8-24
8.2.7.2 Conduct the Wilcoxon Rank Sum Test 8-24
8.2.8 Evaluate the Results: The Decision 8-25
8.3 Mineral Processing Facility Rented Equipment Baseline Survey 8-26
8.3.1 Description 8-26
8.3.2 Objectives 8-26
8.3.3 Initial Assessment of the M&E 8-26
8.3.3.1 Categorize the M&E as Impacted or Non-Impacted 8-26
January 2009 xi NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
8.3.3.2 Describe the M&E 8-28
8.3.3.3 Design and Implement Preliminary Surveys 8-30
8.3.3.4 Select a Disposition Option 8-30
8.3.3.5 Document the Results of the Initial Assessment 8-30
8.3.4 Develop a Decision Rule 8-30
8.3.4.1 Select Radionuclides or Radiations of Concern 8-30
8.3.4.2 Identify Action Levels 8-31
8.3.4.3 Describe the Parameter of Interest 8-31
8.3.4.4 Identify Alternative Actions 8-31
8.3.4.5 Develop a Decision Rule 8-31
8.3.4.6 Identify Survey Units 8-31
8.3.4.7 Develop Inputs for Selection of Provisional Measurement
Methods 8-31
8.3.4.8 Reference Materials 8-32
8.3.5 Develop a Survey Design 8-33
8.3.5.1 Select a Null Hypothesis 8-33
8.3.5.2 Set the Discrimination Limit 8-33
8.3.5.3 Specify the Limits on Decision Errors 8-33
8.3.5.4 Select a Measurement Technique 8-34
8.3.5.5 Finalize Selection of Radiations to be Measured 8-35
8.3.5.6 Develop an Operational Decision Rule 8-35
8.3.5.7 Classify the M&E 8-35
8.3.5.8 Select a Measurement Method 8-35
8.3.5.9 Optimize the Disposition Survey Design 8-37
8.3.5.10 Document the Disposition Survey Design 8-37
8.3.6 Implement the Survey Design 8-37
8.3.6.1 Ensure Protection of Health and Safety 8-37
8.3.6.2 Consider Issues for Handling M&E 8-37
8.3.6.3 Segregate the M&E 8-38
8.3.6.4 Determine the Measurement Detectability for the Scan Survey. 8-38
8.3.6.5 Determine the Measurement Detectability for the Investigation
Survey 8-38
8.3.6.6 Determine Measurement Uncertainty for the Investigation Survey
MDC 8-39
8.3.6.7 Perform Quality Control Measurements 8-41
8.3.6.8 Collect Survey Data 8-41
8.3.7 Evaluate the Survey Results 8-41
8.3.7.1 Conduct a Data Quality Assessment 8-41
8.3.7.2 Conduct aPreliminary Data Review 8-41
8.3.7.3 Conduct the Statistical Tests 8-41
8.3.8 Evaluate the Results: The Decision 8-42
8.4 Mineral Processing Facility Rented Equipment Disposition Survey 8-42
8.4.1 Description 8-42
8.4.2 Objectives 8-42
8.4.3 Initial Assessment of the M&E 8-42
8.4.3.1 Categorize the M&E as Impacted or Non-Impacted 8-42
NUREG-1575, Supp.l xii January 2009
-------
MARSAME Contents (Continued)
8.4.3.2 Describe the M&E 8-43
8.4.3.3 Select a Disposition Option 8-44
8.4.3.4 Document the Results of the Initial Assessment 8-44
8.4.4 Develop a Decision Rule 8-44
8.4.4.1 Identify Action Levels 8-44
8.4.4.2 Evaluate an Existing Survey Design 8-44
8.4.5 Develop a Survey Design 8-44
8.4.5.1 Select the Null Hypothesis 8-44
8.4.5.2 Set the Discrimination Limit 8-45
8.4.5.3 Specify Limits on Decision Errors 8-45
8.4.5.4 Classify the M&E 8-45
8.4.5.5 Optimize the Existing Survey Design 8-45
8.4.5.6 Document the Disposition Survey Design 8-46
8.4.6 Implement the Survey Design 8-46
8.4.7 Evaluate the Survey Results 8-46
8.4.7.1 Conduct a Data Quality Assessment 8-46
8.4.7.2 Conductthe Statistical Tests 8-47
8.4.8 Evaluate the Results: The Decision 8-47
Appendix A Statistical Tables and Procedures A-l
A.I Normal Distribution A-l
A.2 Sample Sizes for Statistical Tests A-2
A.3 Critical Values for the Sign Test A-4
A.4 Critical Values for the WRS Test A-6
A.5 Critical Values for the Quantile Test A-10
Appendix B Sources of Background Radioactivity B-l
B.I Introduction B-l
B.2 Environmental Radioactivity B-l
B.2.1 Terrestrial Radioactivity B-2
B.2.2 Anthropogenic Radioactive Materials B-3
B.2.3 Cosmic Radiation and Cosmogenic Radionuclides B-4
B.3 Inherent Radioactivity B-4
B.4 Instrument Background B-5
B.5 Technologically Enhanced Naturally Occurring Radioactive Material B-5
B.6 Orphan Sources B-6
Appendix C Examples of Common Radionuclides C-l
Appendix D Instrumentation and Measurement Techniques D-l
D.I Introduction D-l
D.2 General Detection Instrumentation D-l
D.2.1 Gas-Filled Detectors D-l
D.2.1.1 lonization Chamber Detectors D-2
D.2.1.2 Gas-Flow Proportional Detectors D-2
D.2.1.3 Geiger-Mueller Detectors D-3
D.2.2 Scintillation Detectors D-4
D.2.2.1 Zinc Sulfide Scintillation Detectors D-4
D.2.2.2 Sodium Iodide Scintillation Detectors D-4
January 2009 xiii NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
D.2.2.3 Cesium Iodide Scintillation Detectors D-4
D.2.2.4 Plastic Scintillation Detectors D-5
D.2.3 Solid State Detectors D-5
D.3 Counting Electronics D-5
D.4 Hand-Held Instruments D-6
D.4.1 Instruments D-6
D.4.2 Temporal Issues D-6
0.4.3 Spatial Issues D-7
D.4.4 Radiation Types D-7
0.4.5 Range D-8
D.4.6 Scale D-8
D.4.7 Ruggedness D-8
D.5 Volumetric Counters (Drum, Box, Barrel, Four-Pi Counters) D-9
D.5.1 Instruments D-9
D.5.2 Temporal Issues D-10
D.5.3 Spatial Issues D-10
D.5.4 Radiation Types D-10
D.5.5 Range D-10
D.5.6 Scale D-ll
D.5.7 Ruggedness D-ll
D.6 Conveyorized Survey Monitoring Systems D-ll
D.6.1 Instruments D-ll
D.6.2 Temporal Issues D-12
D.6.3 Spatial Issues D-12
D.6.4 Radiation Types D-13
D.6.5 Range D-13
D.6.6 Scale D-13
D.6.7 Ruggedness D-13
D.7 In Situ Gamma Spectroscopy D-14
D.7.1 Instruments D-14
D.7.2 Temporal Issues D-14
D.7.3 Spatial Issues D-15
D.7.4 Radiation Types D-15
D.7.5 Range D-15
D.7.6 Scale D-15
D.7.7 Ruggedness D-16
D.8 Hand-Held Radionuclide Identifiers D-16
D.8.1 Instruments D-16
D.8.2 Temporal Issues D-16
D.8.3 Spatial Issues D-16
D.8.4 Radiation Types D-17
D.8.5 Range D-17
D.8.6 Scale D-17
D.8.7 Ruggedness D-17
D.9 Portal Monitors D-17
D.9.1 Instruments D-17
NUREG-1575, Supp.l xiv January 2009
-------
MARSAME Contents (Continued)
D.9.2 Temporal Issues D-18
D.9.3 Spatial Issues D-18
D.9.4 Radiation Types D-19
D.9.5 Range D-19
D.9.6 Scale D-19
D.9.7 Ruggedness D-19
D.10 Sample with Laboratory Analysis D-19
D.10.1 Instruments D-20
D. 10.1.1 Instruments for the Detection of Alpha Radiation D-20
D.I 0.1.2 Instruments for the Detection of Beta Radiation D-21
D. 10.1.3 Instruments for the Detection of Gamma or X-Radiation D-21
D.10.2 Temporal Issues D-21
D.10.3 Spatial Issues D-22
D.10.3.1 Alpha Spectroscopy with Multi-Channel Analyzer D-22
D.10.3.2 Gas-Flow Proportional Counter D-22
D.10.3.3 Liquid Scintillation Spectrometer D-23
D.10.3.4 Low-Resolution Alpha Spectroscopy D-23
D.10.3.5 High-Purity Germanium Detector with Multi-Channel
Analyzer D-23
D.I0.3.6 Sodium Iodide Detector with Multi-Channel Analyzer D-23
D.10.3.7 Alpha Scintillation Detectors D-23
D.10.4 Radiation Types D-24
D.10.5 Range D-24
D.10.6 Scale D-25
D.10.7 Ruggedness D-26
Appendix E Disposition Criteria E-l
E.I Department of Energy E-l
E.I.I 10CFR835 (Non-Exhaustive Excerpts) E-l
E.I.1.1 § 835.405 Receipt of Packages Containing Radioactive MaterialE-1
E.I.1.2 § 835.605 Labeling Items and Containers E-2
E.I.1.3 § 835.606 Exceptions to Labeling Requirements E-2
E.I.1.4 §835.1101 Control of Material and Equipment E-2
E.I.1.5 § 835.1102 Control of Areas E-3
E. 1.2 Appendix D to 10 CFR 835 - Surface Contamination Values E-3
E.I.3 DOE Guidance and Similar Documents E-5
E.2 International Organizations E-7
E.2.1 International Atomic Energy Agency (IAEA) E-7
E.2.2 European Commission E-7
E.3 Nuclear Regulatory Commission E-8
E.3.1 § 20.2003 Disposal by Release into Sanitary Sewerage E-8
E.3.2 § 20.2005 Disposal of Specific Wastes E-8
E.3.3 §35.92 Decay-in-Storage E-9
E.3.4 §35.315 Safety Precautions E-9
E.3.5 §36.57 Radiation Surveys E-9
E.3.6 Appendix A to Part 40-Criteria Relating to the Operation of Uranium Mills
and the Disposition of Tailings or Wastes Produced by the Extraction or
January 2009 xv NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
Concentration of Source Material from Ores Processed Primarily for Their
Source Material Content E-9
E.3.7 §71.4 Definitions E-9
E.3.8 § 71.14 Exemption for Low-Level Materials E-ll
E.3.9 § 110.22 General License for the Export of Source Material E-12
E.3.10 § 110.23 General License for the Export of Byproduct Material E-12
E.3.11 Policies and Practices E-13
E.3.12 Issues Related to International Trade E-14
References Ref-1
Glossary GL-1
NUREG-1575, Supp.l xvi January 2009
-------
MARSAME Contents (Continued)
LIST OF FIGURES
Roadmap Figure 1 Overview of MARSAME Process RM-6
Roadmap Figure 2 The Data Life Cycle Applied to Disposition Surveys RM-7
Roadmap Figure 3 The Categorization Process as Part of Initial Assessment RM-8
Roadmap Figure 4 Assessing Adequacy of Information for Designing RM-9
Roadmap Figure 5 Assessing the Applicability of Existing SOPs RM-10
Roadmap Figure 6 Identify Inputs to the Decision RM-11
Roadmap Figure 7 Identify ActionLevels RM-12
Roadmap Figure 8 Developing Survey Unit Boundaries (Apply to All Impacted M&E for each
set of Action Levels Identified in Section 3.3) RM-13
Roadmap Figure 9 Flow Diagram for Developing a Disposition Survey Design RM-14
Roadmap Figure 10 Flow Diagram for Identifying the Number of Data Points for a MARSSIM-
Type Disposition Survey RM-15
Roadmap Figure 11 Flow Diagram for Identifying Data Needs for Assessment of Potential
Areas of Elevated Activity in Class 1 Survey Units for MARSSEVI-Type
Disposition Surveys RM-16
Roadmap Figure 12 Implementation of Disposition Surveys RM-17
Roadmap Figure 13 Assess the Results of the Disposition Survey RM-18
Roadmap Figure 14 Interpretation of Survey Results for Scan-Only and In Situ Surveys.. RM-19
Roadmap Figure 15 Interpretation of Survey Results for MARSSIM-Type Surveys RM-20
Figure 1.1 The Data Life Cycle Applied to Disposition Surveys 1-8
Figure 2.1 The Categorization Process as Part of Initial Assessment 2-2
Figure 2.2 Assessing Adequacy of Information for Designing Disposition Surveys 2-10
Figure 2.3 Documentation of the Initial Assessment 2-20
Figure 3.1 Identifying Inputs to the Decision 3-2
Figure 3.2 Identifying ActionLevels 3-4
Figure 3.3 Developing Survey Unit Boundaries 3-16
Figure 4.1 Relative Shift, A/a, Comparison for Scenario A: o is Large, but the Large A
. Results in a Large A/a and Fewer Samples 4-3
Figure 4.2 Relative Shift, A/a, Comparison for Scenario A: o is Small, but the Small A
Results in a Small A/a and More Samples 4-3
Figure 4.3 Illustration of Scenario A 4-4
Figure 4.4 Illustration of Scenario B 4-4
Figure 4.5 Flow Diagram for a Disposition Survey Design 4-10
Figure 4.6 Flow Diagram for Identifying the Number of Data Points for a MARSSEVI-Type
Disposition Survey 4-11
Figure 4.7 Flow Diagram for Identifying Data Needs for Assessment of Potential Areas of
Elevated Activity in Class 1 Survey Units for MARSSIM-Type Disposition
Surveys 4-12
Figure 4.8 Relationship Between the Relative Shift and the Amount of M&E to be
Scanned 4-13
Figure 5.1 Implementation of Disposition Surveys 5-2
January 2009 xvii NUREG-1575, Supp. 1
-------
Contents (Continued) MARSAME
Figure 5.2 The Critical Value (&) and the Minimum Detectable Value (SD) of the Net
Instrument Signal (or Count) 5-11
Figure 6.1 The Assessment Phase of the Data Life Cycle 6-2
Figure 6.2 Frequency Plot of Concrete Rubble Data 6-15
Figure 6.3 Screen Capture of Output from ProUCL Software for the Sample Data Set 6-16
Figure 6.4 Interpretation of Survey Results for Scan-Only and In Situ Surveys 6-24
Figure 6.5 Statistical Interpretation of Results for MARSSIM-Type Surveys 6-25
Figure 7.1 Relative Shift, A/a, Comparison for Scenario A: o is Large, but the Large A
Results in a Large A/a and Fewer Samples 7-11
Figure 7.2 Relative Shift, A/a, Comparison for Scenario A: o is Small, but the Small A
. Results in a Small A/a and More Samples 7-11
Figure 7.3 Illustration of Scenario A 7-12
Figure 7.4 Illustration of Scenario B 7-13
Figure 7.5 Example of the Required Measurement Uncertainty at Concentrations other than
the UBGR. In this Example the UBGR Equals the Action Level 7-21
Figure 7.6 The Critical Value of the Net Instrument Signal (Sc) and the Minimum Detectable
Net Signal (&) 7-31
Figure 7.7 Relationship Between the Critical Value of the Net Count, the Minimum
Detectable Net Counts and the MDC 7-33
Figure 7.8 Probability of Detection as a Function of Net Count (Lower X-Axis) and
Concentration (Upper X-Axis) 7-33
Figure 7.9 Relationships Among the Critical Value, the MDC, the MQC, and the Probability
of Exceeding the Critical Value 7-37
Figure 8.1 Frequency Plot of Illustrative Example Data 8-25
Figure 8.2 Cumulative Frequency Plot of Illustrative Example Data 8-25
Figure 8.3 Front Loader 8-27
Figure D.I Example Volumetric Counter (Thermo 2005) D-9
Figure D.2 Example Conveyorized Survey Monitoring System (Laurus 2001) D-12
Figure D.3 Example Portal Monitor (Canberra 2005b) D-18
NUREG-1575, Supp.l xviii January 2009
-------
MARSAME
Contents (Continued)
LIST OF TABLES
Table 1.1 The Data Life Cycle Used to Support Disposition Survey Design 1-7
Table 1.2 Similarities Between MARS SIM and MARSAME 1-15
Table 1.3 Differences Between MARS SIM and MARSAME 1-16
Table 2.1 Physical Attributes Used to Describe M&E 2-12
Table 2.2 Radiological Attributes Used to Describe M&E 2-15
Table 3.1 Example Detector Efficiency Calculation (232Th in Complete Equilibrium with its
Decay Products) Using a Gas Proportional Detector 3-11
Table 3.2 Example Alternative Actions 3-15
Table 3.3 Examples of Consensus Standards for Evaluating Ruggedness 3-24
Table 5.1 Potential Applications for Instrumentation and Measurement Technique
Combinations 5-18
Table 5.2 Survey Unit Size and Quantity Restrictions for Instrumentation and Measurement
Technique Combinations 5-18
Table 5.3 Advantages and Disadvantages of Instrumentation and Measurement Technique
Combinations 5-20
Table 6.1 Issues and Assumptions Underlying the Evaluation Method 6-10
Table 6.2 Summary of Evaluation Methods and Statistical Tests 6-11
Table 6.3 Sign Test Example Data 6-18
Table 6.4 Scenario A WRS Test Example Data 6-21
Table 6.5 Scenario B WRS Test Example Data 6-22
Table 7.1 Notation for DQOs and MQOs 7-2
Table 7.2 Notation for Uncertainty Calculations 7-3
Table 7.3 Notation for MDC Calculations 7-4
Table 7.4 Notation for MQC Calculations 7-5
Table 7.5 Recommended Approaches for Calculating the Critical Value of the Net
Instrument Signal (Count), Sc 7-28
Table 7.6 Recommended Approaches for Calculating the Minimum Detectable Net
Instrument Signal or Count 7-30
Table 7.7 Uncertainty Budget for the Efficiency Example 7-56
Table 7.8 Calculation of Detector Response to Natural Uranium 7-68
Table 7.9 Calculation of Detector Response for Natural Thorium 7-69
Table 7.10 Detector Response to Natural Uranium 7-71
Table 7.11 Detector Response to Natural Thorium 7-71
Table 7.12 Scan MDCs for FIDLER 7-74
Table 8.1 Physical Attributes of the Concrete Rubble 8-4
Table 8.2 Radiological Attributes of the Concrete Rubble 8-5
Table 8.3 Preliminary Alpha Spectrometry Results for Uranium Series Radionuclides 8-6
Table 8.4 Preliminary Alpha Spectrometry Results for Thorium Series Radionuclides 8-6
Table 8.5 Preliminary Gamma Spectroscopy Results for Uranium Series Radionuclides.. 8-7
Table 8.6 Preliminary Gamma Spectroscopy Results for Thorium Series Radionuclides .. 8-7
Table 8.7 Radionuclide-Specific Action Levels 8-9
Table 8.8 Calculation of the Gross Gamma Action Level 8-11
January 2009
xix
NUREG-1575, Supp. 1
-------
Contents (Continued)
MARSAME
Table 8.9 Job Safety Analysis for Surveying Concrete Rubble 8-17
Table 8.10 Radionuclide-Specific Required Relative Measurement Method Uncertainties8-23
Table 8.11 Sentinel Measurement Results 8-27
Table 8.12 Physical Attributes Used to Describe the Front Loader 8-28
Table 8.13 Radiological Attributes Used to Describe the Front Loader 8-29
Table 8.14 Potential Discrimination Limits 8-33
Table 8.15 Detector Efficiency for the Mineral Processing Facility (232Th in Complete
Equilibrium with its Progeny) using a Gas Proportional Detector 8-36
Table 8.16 Sentinel Measurement Results 8-43
Table A.I Cumulative Normal Distribution Function
-------
MARSAME
ACKNOWLEDGEMENTS
The Multi-Agency Radiation Survey and Site Investigation Manual (MARS SIM) and the Multi-
Agency Radiation Survey and Assessment of Materials and Equipment manual (MARSAME)
supplement came about as a result of individuals—at the management level—within the
Environmental Protection Agency (EPA), Nuclear Regulatory Commission (NRC), Department
of Energy (DOE), and Department of Defense (DOD) who recognized the necessity for a
standardized guidance document for investigating radioactively contaminated sites. The creation
of MARSSEVI and MARSAME was facilitated by the cooperation of subject matter specialists
from these agencies with management's support and a willingness to work smoothly together
toward reaching the common goal of creating a workable and user-friendly guidance manual.
Special appreciation is extended to Robert A. Meek of the NRC and Anthony Wolbarst of EPA
for developing the concept of a multi-agency workgroup and bringing together representatives
from the participating agencies.
MARSAME could not have been possible without the technical workgroup members who
contributed their time, talent, and efforts to develop this consensus guidance document:
CAPT Colleen F. Petullo, U.S. Public Health Service, EPA, Chair
DOD David P. Alberth (Army) EPA Kathryn Snead
Dennis Chambers, CHP (Army, Retired) Nidal Azzam
Gerald Falo, Ph.D., CHP (Army) Lindsey Bender
Steven Doremus, Ph.D. (Navy) Vicki Lloyd
CAPT Vincent Delnnocentiis (Navy) Eugene Jablonowski
Ramachandra Bhat, Ph.D., CHP (Air Force)
Lt Col Craig Bias, Ph.D., CHP (Air Force)
Lt Col Daniel Caputo, Ph.D. (Air Force Reserve)
DOE W. Alexander Williams, Ph.D. NRC Robert A. Meek, Ph.D.
Emile Boulos George E. Powers, Ph.D.
Harold T. Peterson, Jr., CHP (Retired) Joseph DeCicco, CHP
Amanda Anderson Anthony Huffert, CHP
Wayne Glines, CHP
DHS Carl V. Gogolak, Ph.D. (Retired)
Special mention is extended to the Federal Agency contractors for their assistance in developing
the MARSAME supplement:
Scott Hay (Cabrera Services, Inc.)
Carl V. Gogolak (Environmental Management Support, Inc.)
Nicholas Berliner (Cabrera Services, Inc.)
Robert Coleman (Oak Ridge National Laboratory)
Deborah Schneider (S. Cohen & Associates, Inc.)
Kerri Wachter (S. Cohen & Associates, Inc.)
A special thank you is extended to Mary Clark (EPA), Schatzi Fitz-James (EPA), Paul Giardina
(EPA), Bonnie Gitlin (EPA), Sally Hamlin (EPA), David Kappelman (EPA), Sophie Kastner
January 2009 xxi NUREG-1575, Supp. 1
-------
Acknowledgements (Continued) MARSAME
(EPA), Joseph LaFornara (EPA), Juan Reyes (EPA), Colby Stanton (EPA), Dennisses Valdes
(EPA), Jean-Claude Dehmel (NRC), MAJ David Pugh, CHP (Air Force), Brian Renaghan (Air
Force), Andrew Wallo III (DOE), Ethel Jacob (DHS), Kevin Miller (DHS), Peter Shebell (DHS),
Jenny Goodman (NJ Bureau of Environmental Radiation), Nancy Stanley (NJ Bureau of
Environmental Radiation), and Eric Abelquist (Oak Ridge Institute for Science and Education).
The Workgroup would also like to thank EPA's Science Advisory Board Radiation Advisory
Committee for their consultations and peer review supporting development of the MARSAME
supplement:
Chair
Bernd Kahn, Ph.D., Georgia Institute of Technology
Jill Lipoti, Ph.D., New Jersey Department of Environmental Protection (Past Chair)
Members
Thomas B. Borak, Ph.D., Colorado State University
Antone L. Brooks, Ph.D., Washington State University Tri-Cities
Faith G. Davis, Ph.D., University of Illinois at Chicago
Brian Dodd, Ph.D., Consultant
Shirley A. Fry, Ph.D., Consultant
William C. Griffith, Ph.D., University of Washington
Jonathan M. Links, Ph.D., Johns Hopkins University
Bruce A. Napier, Pacific Northwest National Laboratory
Daniel O. Stram, Ph.D., University of Southern California
Richard J. Vetter, Ph.D., Mayo Clinic
SAB Consultants
Bruce W. Church, BWC Enterprises, Inc.
Kenneth Duvall, Environmental Scientist/Consultant
Janet A. Johnson, Ph.D., Consultant
Paul J. Merges, Ph.D., Environment & Radiation Specialists, Inc.
Science Advisory Board Staff
K. JackKooyoomjian, Ph.D., Designated Federal Officer, EPA
The Workgroup acknowledges the interest of the NRC's Advisory Committee on Nuclear Waste
and Materials in the development of MARSAME.
NUREG-1575, Supp.l xxii January 2009
-------
MARSAME
ACRONYMS AND ABBREVIATIONS
AL action level
ALARA as low as reasonably achievable
ANSI American National Standards Institute
ASTM American Society for Testing and Materials
BKGD background
CERCLA Comprehensive Environmental Response Compensation and Liability Act
CFR Code of Federal Regulations
cpm counts per minute
cps counts per second
CSM conceptual site model
CSU combined standard uncertainty
CZT cadmium zinc telluride
DAC derived air concentration
DCGL derived concentration guideline level
DL discrimination limit
DOD Department of Defense
DOE Department of Energy
DOT Department of Transportation
dpm disintegrations per minute
DQA data quality assessment
DQO data quality objective
EMC elevated measurement comparison
EPA Environmental Protection Agency
EPRI Electric Power Research Institute
EU European Union
EZ exclusion zone
FIDLER field instrument for the detection of low-energy radiation
FRER fluence rate to exposure rate
GM Geiger Mueller
HASP health and safety plan
HEU high-enriched uranium
HPGe high-purity germanium
HPS Health Physics Society
HSA Historical Site Assessment
HPSR Health Physics Society Report
HWP hazard work permit
IA initial assessment
IAEA International Atomic Energy Agency
IEEE Institute of Electrical & Electronics Engineers
ISGS in situ gamma spectroscopy
ISO International Organization for Standardization
January 2009
XXlll
NUREG-1575, Supp. 1
-------
Acronyms and Abbreviations (Continued)
MARSAME
ISA job safety analysis
LBGR lower bound of the gray region
LEU low-enriched uranium
LSA low specific activity
LSC liquid scintillation cocktail
M&E materials and equipment
MARLAP Multi-Agency Radiological Laboratory Analytical Protocols manual
MARSAME Multi-Agency Radiation Survey and Assessment of Materials and Equipment
manual
MARS SIM Multi-Agency Radiation Survey and Site Investigation Manual
MCA multi-channel analyzer
MDC minimum detectable concentration
MDCR minimum detectable count rate
MDCRsurveyor MDCR by a less than ideal surveyor
MDER minimum detectable exposure rate
MQC minimum quantifiable concentration
MQO measurement quality objective
NARM naturally occurring and accelerator-produced radioactive material
NCRP National Council on Radiation Protection and Measurements
NIST National Institute of Science and Technology
NJBER New Jersey Bureau of Environmental Radiation
NORM naturally occurring radioactive material
NRC Nuclear Regulatory Commission
NUREG Nuclear Regulatory Commission technical report prepared by NRC staff
NUREG/CR Nuclear Regulatory Commission technical report prepared by NRC contractor
ORISE Oak Ridge Institute for Science and Education
OSHA Occupational Safety and Health Administration
OSWER EPA Office of Solid Waste and Emergency Response
PCB polychlorinated biphenyl
pH hydrogen ion concentration (acidity or basicity)
PIC pressurized ion chamber
PPE personal protective equipment
PVC polyvinylchloride
QA quality assurance
QAPP quality assurance project plan
QC quality control
RCA radiological control area
RCRA Resource Conservation and Recovery Act
RCSU relative combined standard uncertainty
RDR relative detector response
RESRAD RESidual RADioactivity computer code (exposure pathway model)
ROC radionuclide of concern
RTG Radioisotopic Thermoelectric Generator
NUREG-1575, Supp.l
XXIV
January 2009
-------
MARSAME
Acronyms and Abbreviations (Continued)
RWP radiation work permit
SCO surface-contaminated object
SI International System of Units (Systeme International d'Unites)
SOP standard operating procedure
TEDE total effective dose equivalent
TENORM technologically enhanced naturally occurring radioactive material
TRU transuranic
UBGR upper bound of the gray region
UCL upper confidence limit
UMTRCA Uranium Mill Tailings Radiation Control Act
UNSCEAR United Nations Scientific Committee on the Effects of Atomic Radiation
USEPA United States Environmental Protection Agency
U.S. United States
WRS Wilcoxon Rank Sum
January 2009
XXV
NUREG-1575, Supp. 1
-------
-------
MARSAME
Be
Bi
Bq
C
C
Ci
d
SYMBOLS, NOMENCLATURE, AND NOTATIONS
< less than
> greater than
< less than or equal to
> greater than or equal to
0 degrees (angle or temperature)
% percent
I-/? statistical power of a hypothesis test
a Type I decision-error rate
aQ quantile test (O.Q = a/2)
a half-width of a rectangular or triangular probability distribution
A area
A overall sensitivity of a measurement
Ac actinium (isotope listed: 228Ac)
ALt action level value an individual radionuclide (/' = 1,2, . . . , n)
ALmeas,mod modified action level for the radionuclide being measured when it is used as a
surrogate for other radionuclide(s)
ALmeas action level for the radionuclide being measured
ALinfer action level for the inferred radionuclide (in surrogate measurements)
Am americium (isotope listed: 241Am)
ft Type II decision-error rate
b background count rate
bj the average number of counts in the background interval (scanning)
beryllium (isotope listed: 7Be)
bismuth (isotopes listed: 210Bi, 212Bi, 214Bi)
becquerel
carbon (isotope listed: 14C)
radionuclide concentration or activity
curie
concentration value an individual radionuclide (/' = 1, 2, . . . , n)
sensitivity coefficient
component of the uncertainty in y due to xf
ratio of amount of the inferred radionuclide to that of the measured surrogate
radionuclide
°C degrees Celsius
cm centimeter
cm2 square centimeter
cm3 cubic centimeter
Cd cadmium (isotope listed: 109Cd)
Co cobalt (isotopes listed: 57Co, 60Co)
Cs cesium (isotope listed: 137Cs)
CsI(Tl) cesium iodide (thallium activated)
A shift (width of the gray region, UBGR-LBGR)
January 2009
XXVll
NUREG-1575, Supp. 1
-------
Symbols, Nomenclature, and Notations (Continued) MARSAME
A/er relative shift
d parameter in the Stapleton Equation for the critical net signal
d' detectability index (scanning)
Si instrument efficiency
ss surface efficiency for surveyed media
eV electron-volt
Er energy of a gamma photon of concern in kiloelectron-volts (keV)
Et energy of a photon of interest
°F degrees Fahrenheit
ft relative fraction of activity contributed by radionuclide /' to the total
ft foot (feet)
ft3 cubic foot (feet)
Fe iron (isotope listed: 55Fe)
g gram
GBq gigabecquerel (1 * 109 becquerels)
gross gamma action level
h hour
H hydrogen (isotope listed: 3H [tritium])
HO null hypothesis
HI alternative hypothesis
/' observation time interval length (scanning)
I iodine (isotopes listed: 123I, 125I, 131I)
in inch
Ir iridium (isotope listed: 192Ir)
k coverage factor for the expanded uncertainty, U
K potassium (isotope listed: 40K)
kBq kilobecquerel (1 x 103 becquerels)
keV kiloelectron-volt (1 x 103 electron-volts)
kg kilogram
kq multiple of the standard deviation defining _yg, usually chosen to be 10
L grid size spacing
L liter
Ib pound
\a micro (1CT6)
fj theoretical mean of a population distribution
(jUen/p)air mass energy absorption coefficient in air centimeters squared per gram (cm2/g)
uR microroentgen ( 1 x 1 CT6 roentgen)
m number of reference measurements (WRS test or Quantile test)
m meter
m2 square meter
MeV megaelectron-volt (1 x 106 electron-volt)
mrem millirem ( 1 x 1 CT3 rem)
NUREG-1575, Supp.l xxviii January 2009
-------
MARSAME
Symbols, Nomenclature, and Notations (Continued
mSv
n
N
Na
Nal(Tl)
Ni
Np
P
P
Pa
PA
Pb
PC
pCi
Pm
Po
Pu
q
P(X,,X])
R
R
Ra
RB
RI
Rn
r(x,,xj)
as
CT(*7)
a(y\¥=ye}
o(X,, Xj)
milliseivert(lxlO 3 Sv)
number of survey unit measurements (WRS test or Quantile test)
sample size, i.e. number of data points (or samples) for the Sign test
survey unit area divided by the maximum area corresponding to the area factor,
which yields the number of measurements needed so the scan MDC is adequate
sodium (isotope listed: 22Na)
sodium iodide (thallium activated)
nickel (isotope listed: 63Ni)
neptunium (isotope listed: 237Np)
non-Poisson variance component of the background count rate correction
coverage probability for expanded uncertainty, also used for efficiency of a less
than ideal surveyor (scanning)
probability of interaction between radiation and a detector
protactinium (isotopes listed: 234Pa, 234mPa)
probe area
lead (isotopes listed: 212Pb, 214Pb)
personal computer
picocurie (IxlCT12 curies)
promethium (isotope listed: 147Pm)
polonium (isotopes listed: 210Po, 212Po, 214Po, 216Po)
Plutonium (isotopes listed: 238P, 239Pu, 240Pu, 241Pu)
critical value for statistical tests (Table A.3, Table A.4)
density
correlation coefficient for two input quantities, Xt and Xj
ratio
roentgen (exposure rate)
radium (isotopes listed: 224Ra, 226Ra, 228Ra)
mean background count rate
mean interference count rate
radon (isotopes listed: 220Rn, 222Rn)
correlation coefficient for two input estimates, X; and Xj
theoretical total standard deviation of the population distribution being sampled
theoretical measurement standard deviation of the population distribution being
sampled, estimated by the combined standard uncertainty of the measurement
theoretical measurement variance of the population distribution being sampled
required measurement method standard deviation (upper limit)
theoretical sampling standard deviation of the population distribution being
sampled
theoretical sampling variance of the population distribution being sampled
standard deviation of the measured interference count rate
variance of the estimator y given the true concentration Y equals yq
covariance for two input quantities, Xt and Xj
January 2009
XXIX
NUREG-1575, Supp. 1
-------
Symbols, Nomenclature, and Notations (Continued)
MARSAME
s+
Sc
SD
Si
$i,surveyor
Sr
Sv
Tc
Th
Tl
fe
ts
U
U
u(x,)
u(Xi,Xj)
uc(y)
Uc(y)ly
u,{y)
UU
UMR
X,
Wr
Ws
ws
x
x,
xc
XQ
y
Sign test statistic
sample standard deviation of the input estimate, xt
critical value of the net instrument signal
mean value of the net signal that gives a specified probability, I-/?, of yielding an
observed signal greater than its critical value Sc
minimum detectable number of net source counts in the observation interval
(scanning)
minimum detectable number of net source counts in the observation interval by a
less than ideal surveyor (scanning)
strontium (isotope listed: 90Sr)
seivert
techicium (isotopes listed: 99Tc, 99mTc)
thorium (isotopes listed: 228Th: 230Th, 232Th, 234Th)
thalium (isotopes listed: 201T1,208T1)
count time for the background
count time for the source
expanded uncertainty
uranium (isotopes listed: 234U, 235U, 238U)
standard uncertainty of the input estimate, xt
relative standard uncertainty of x,
covariance of two input estimates, xt and x/
combined standard uncertainty of y
relative combined standard uncertainty of the output quantity for a particular
measurement
combined variance ofy
component of the combined standard uncertainty, uc(y\ generated by the standard
uncertainty of the input estimate xt, u(x,), multiplied by the sensitivity coefficient,
Ci
measurement method uncertainty
required measurement method uncertainty
required relative measurement method uncertainty
relative variance of the measured sensitivity
relative standard uncertainty of a nonzero input estimate, xt, for a particular
measurement. (p(x,) = u(Xj)/Xj
cumulative normal distribution function
sum of the ranks of the (adjusted) reference measurements (WRS test)
sum of the ranks of the (adjusted) sample measurements (WRS test)
weighted instrument sensitivity
estimate of the input quantity, X
an input quantity
the critical value of the response variable, x
minimum quantifiable value of the response variable, x
year
NUREG-1575, Supp.l
XXX
January 2009
-------
MARSAME Symbols, Nomenclature, and Notations (Continued
y estimate of the output quantity for a particular measurement, Y
Y output quantity, measurand
yc critical value of the concentration
yD minimum detectable concentration (MDC)
yq minimum quantifiable concentration (MQC)
yd yard
yd3 cubic yard
Z atomic number
z\.a (1 - a)-quantile of the standard normal distribution
z\.p (1 - /?)-quantile of the standard normal distribution
ZnS(Ag) zinc sulfide (silver activated)
January 2009 xxxi NUREG-1575, Supp. 1
-------
CONVERSION FACTORS
MARSAME
To Convert
From
acre
becquerel (Bq)
Bq/kg
Bq/m2
Bq/m3
centimeter (cm)
Ci
dps
dpm
dpm/100 cm2
gray (Gy)
hectare
liter (L)
To
hectare
sq. meter (m2)
sq. feet (ft2)
curie (Ci)
dps
pCi
pCi/g
dpm/100 cm2
Bq/L
pCi/L
inch
Bq
pCi
dpm
pCi
dps
pCi
Bq/m2
rad
acre
cm3
m3
ounce (fluid)
Multiply
"From"
Quantity By
0.405
4,050
43,600
2.7xlO~u
1
27
0.027
0.60
0.001
0.027
0.394
3.70xl010
IxlO12
60
27
0.0167
0.451
1.67
100
2.47
1000
0.001
33.8
To Convert
From
meter (m)
sq. meter (m2)
m3
mrem
mrem/y
mSv
mSv/y
ounce (oz)
pCi
pCi/g
pCi/L
rad
rem
seivert (Sv)
To
inch
mile
acre
hectare
sq. feet (ft2)
sq. mile
liter
mSv
mSv/y
mrem
mrem/y
liter (L)
Bq
dpm
Bq/kg
Bq/m3
Gy
mrem
mSv
Sv
mrem
mSv
rem
Multiply By
39.4
0.000621
0.000247
0.0001
10.8
3.86xlO~7
1,000
0.01
0.01
100
100
0.0296
0.037
2.22
37
37
0.01
1,000
10
0.01
100,000
1,000
100
NUREG-1575, Supp.l
XXXll
January 2009
-------
MARSAME Roadmap
ROADMAP
Introduction to MARSAME
The Multi-Agency Radiation Survey and Assessment of Materials and Equipment manual
(MARSAME) is a supplement to the Multi-Agency Radiation Survey and Site Investigation
Manual (MARSSIM 2002). MARSAME provides technical information on approaches for
planning, implementing, assessing, and documenting surveys to determine proper disposition of
materials and equipment (M&E).
The technical information in MARSAME is based on the data life cycle, similar to MARSSIM.
Survey planning is based on the data quality objectives (DQO) process and is discussed in
MARSAME Chapters 2, 3, and 4. Implementation of the survey design is described in
MARSAME Chapter 5, with discussions on selection of instruments and measurement
techniques as well as handling and segregating the M&E. MARSAME also includes the concept
of measurement quality objectives (MQOs) for selecting and evaluating instruments and
measurement techniques from the Multi-Agency Radiological Laboratory Analytical Protocols
manual (MARLAP 2004). Assessment of the survey results uses data quality assessment (DQA)
and the application of statistical tests as described in MARSAME Chapter 6. In addition to the
first six chapters, which present the MARSAME process, the MARSAME manual contains the
statistical basis for the DQOs, MQOs, and survey designs (Chapter 7) and illustrative examples
of the information and process presented in MARSAME (Chapter 8).
The scope of MARSSIM was limited to surfaces soils and building surfaces. The scope of
MARSAME is M&E potentially affected by radioactivity, including metals, concrete, tools,
equipment, piping, conduit, furniture and dispersible bulk materials such as trash, rubble, roofing
materials, and sludge. The wide variety of M&E requires additional flexibility in the survey
process, and this flexibility is incorporated into MARSAME.
The Goal of the Roadmap
The increased flexibility of MARSAME comes with increased complexity. The goal of the
roadmap is to assist the MARSAME user in negotiating the information in MARSAME and
determining where important decisions need to be made on a project-specific basis, as
summarized in Roadmap Figure 1. Roadmap Figure 2 provides additional detail and illustrates
how the data life cycle is applied to disposition surveys. (Shaded blocks within the figures depict
significant decisions or milestones.)
This roadmap is not designed to be a stand-alone document, but to be used as a quick reference
to MARSAME for users already familiar with the process of planning, implementing, and
assessing surveys. Roadmap users will find flowcharts summarizing major decision points in the
survey process combined with references to sections in MARSAME with more detailed
information. The roadmap assumes a familiarity with MARSAME terminology. Section 1.2 of
MARSAME discusses key terminology, and a complete set of definitions is provided in the
glossary.
January 2009 RM-1 NUREG-1575, Supp. 1
-------
Roadmap MARSAME
Initial Assessment
The initial assessment (IA) is the first step in the investigation of M&E, similar to the historical
site assessment (HSA) in MARSSIM. The purpose of the IA is to collect and evaluate
information about the M&E to support a categorization decision and support potential disposition
of the M&E (e.g., release or interdiction). Project managers are encouraged to use the IA to
evaluate M&E for other hazards (e.g., lead, PCBs, asbestos) that could increase the complexity
of the disposition survey design or pose potential risks to workers during subsequent survey
activities (Section 5.2), or to human health and the environment following subsequent disposition
of the M&E.
Categorization
MARSAME uses the term categorization to describe the decision of whether M&E are impacted
or non-impacted. Non-impacted is a term that applies to M&E where there is no reasonable
potential to contain radionuclide concentration(s) or radioactivity above background. Impacted is
a term that applies to M&E that are not classified as non-impacted. Roadmap Figure 3 shows the
categorization process as part of the IA.
Standardized Survey Designs
Most operating radiological facilities maintain standard operating procedures (SOPs) as part of a
quality system. In many cases these SOPs include instructions for conducting disposition
surveys. The first step in evaluating an existing SOP is to determine whether there is adequate
information available to design a disposition survey. If the existing information is inadequate to
design a disposition survey, it is inadequate for determining if an existing survey design is
adequate either. Roadmap Figure 4 addresses assessing the adequacy of existing information for
designing disposition surveys. Roadmap Figure 5 shows how implementing an existing SOP that
is applicable to the M&E being investigated takes the user from MARSAME Chapter 2 to
MARSAME Chapter 6. If a project-specific survey design needs to be developed, Roadmap
Figure 5 directs the user to the information in MARSAME Chapter 3.
In some cases, it may be possible to modify the M&E to match the assumptions used to develop
the existing SOP, or modify the existing SOP to address the M&E being investigated. M&E may
be modified by changing the physical attributes described in Table 2.1 or the radiological
attributes described in Table 2.2. Modifications to the SOP are most often associated with MQOs
such as the measurement detectability (Section 5.7) or measurement quantifiability (Section 5.8).
Modifying the MQOs may result in small changes such as an increased count time (e.g., to
account for an increase in measurement uncertainty or a decrease in counting efficiency) or
larger changes such as selecting a different instrument (e.g., a gas-proportional detector instead
of a Geiger-Mueller detector) or a different measurement technique (e.g., in situ measurements
instead of scan measurements). Information on evaluating an existing survey design to determine
if it will meet the DQOs for the M&E being investigated is provided in Section 3.10.
Develop a Decision Rule
MARSAME Chapter 3 focuses on developing a decision rule by identifying inputs to the
decision (see Roadmap Figure 6, which depicts the various inputs to the decision). A decision
NUREG-1575, Supp. 1 RM-2 January 2009
-------
MARSAME Roadmap
rule is a theoretical "if.. .then..." statement that defines how the decision maker would choose
among alternative actions. There are three parts to a decision rule:
• An action level that causes a decision maker to choose between the alternative actions (see
Roadmap Figure 7 and Section 3.3),
• A parameter of interest that is important for making decisions about the target population
(see Section 3.4), and
• Alternative actions that could result from the decision (Section 3.5).
Other inputs to the decision that are discussed in MARSAME Chapter 3 include selecting
radionuclides or radiations of concern (Section 3.2), developing survey unit boundaries (see
Roadmap Figure 8 and Section 3.6), inputs for selecting provisional measurement methods
(Section 3.8), and identifying reference materials if necessary (Section 3.9).
Survey Design
Once a decision rule has been developed, a disposition survey can be designed for the impacted
M&E being investigated. The disposition survey incorporates all of the available information to
determine the quality and quantity of data required to support a disposition decision. Roadmap
Figure 9 shows how a disposition survey design is developed.
MARSAME, like MARSSIM, provides information on using a null hypothesis that radionuclide
concentrations or activity levels associated with the M&E exceed the action level (i.e., Scenario
A). MARSAME also incorporates additional technical information from NUREG-1505 (NRC
1998a) and MARLAP for designing surveys using Scenario B where the null hypothesis is that
the radionuclide concentrations or activity levels are less than the action level. The assignment of
values to the lower bound of the gray region (LBGR) and upper bound of the gray region
(UBGR), specification of decision error rates, and classification are all similar to information
provided in MARSSIM.
MARSAME provides information on four types of survey designs:
• Scan-only survey designs (Section 4.4.1),
• In situ survey designs (Section 4.4.2),
• Survey designs that combine scans and static measurements (MARSSIM-type surveys,
Section 4.4.3), and
• Method-based survey designs (Section 4.4.4).
A method-based survey design is a special type of scan-only, in situ, or MARSSIM-type survey
design that incorporates a required measurement method or combination of measurement
technique and instrumentation, so Roadmap Figure 9 only depicts the first three. It will still need
to address all of the required components, such as number, type, location, and sensitivity of
measurements.
Scan-only survey designs use scanning techniques to measure the M&E. In general, scan-only
survey designs may be applied to all types of M&E, from small individual items to large
January 2009 RM-3 NUREG-1575, Supp. 1
-------
Roadmap MARSAME
quantities of materials to large, complex machines. Scan-only surveys range from hand-held
instruments moving over the M&E to conveyorized systems that move the M&E past the
detectors. Scan-only survey designs often require the least amount of resources to design and
implement, and are easy to incorporate into SOPs or project-specific survey designs. In many
cases it is not necessary to document the results of individual scanning measurements because it
is easy to identify results that exceed some threshold corresponding to the action level. With the
real-time feedback available during Class 1 scan-only surveys, the user can implement a "clean
as you go" practice by segregating M&E that exceed the threshold for additional investigation.
Drawbacks to scan-only surveys include increased measurement uncertainty because of
variations in scan speed and source to detector distance making it difficult to detect or quantify
radionuclides with action levels close to zero or background.
In situ surveys are characterized by limited numbers of static measurements with long count
times (relative to scan-only surveys) to measure the M&E. In situ surveys generally are
applicable to situations where scan-only surveys are determined to be unacceptable. For
example, variations in source-to-detector distance, scan speed, and surface efficiency that are
commonly associated with scanning measurements can often be effectively controlled using an
in situ survey design. There are a wide variety of in situ measurement techniques available
including box counters, portal monitors, in situ gamma spectroscopy systems, and direct
measurements with hand-held instruments. The primary difference between an in situ survey and
a MARSSIM-type survey is that an in situ survey measures 10-100% of an item (using one or
several measurements) to determine the average radionuclide concentration for that item. A
MARSSIM-type survey uses a statistically based number of measurements (that generally do not
measure 10% of the item or group of items being surveyed) to calculate an average radionuclide
concentration for that item or group of items.
MARSSIM-type survey designs combine a statistically based number of static measurements or
samples (Roadmap Figure 10) to determine average radionuclide concentrations with scanning to
identify localized areas of elevated activity (Roadmap Figure 11). MARSSIM-type surveys are
designed using the information in MARSSIM. The process of identifying survey unit sizes,
laying out systematic or random measurement grids, and calculating project- and item-specific
area factors requires significantly greater effort during planning and implementation than either
scan-only or in situ survey designs. In general, MARSSIM-type surveys of M&E are only
performed on large, complicated M&E with a high inherent value after scan-only and in situ
survey designs have been considered and rejected as inappropriate or unacceptable.
Measurement Quality Objectives
Measurement quality objectives (MQOs) are characteristics of a measurement method required
to meet the objectives of the survey. Additional information on MQOs can be found in
MARSAME Section 3.8, Section 5.5, and Section 7.3 as well as MARLAP Chapter 3. MQOs are
an important concept that was not presented in MARSSIM, and should be an important factor
when evaluating existing survey designs and SOPs for applicability to surveying M&E.
MQOs for a project include, but are not limited to—
NUREG-1575, Supp. 1 RM-4 January 2009
-------
MARSAME Roadmap
• The measurement method uncertainty at a specified concentration expressed as a standard
deviation (Sections 3.8.1, 5.5, and 7.4);
• The measurement method's detection capability expressed as the minimum detectable
concentration (MDC; see Sections 3.8.2, 5.7, and 7.5);
• The measurement method's quantification capability expressed as the minimum quantifiable
concentration (MQC; see Sections 3.8.3, 5.8, 7.6, and 7.7);
• The measurement method's range, which defines the measurement method's ability to
measure the radionuclide or radiation of concern over some specified range of concentration
or activity (see Section 3.8.4 and Appendix D);
• The measurement method's specificity, which refers to the ability of the measurement
method to measure the radionuclide or radiation of concern in the presence of interferences
(Section 3.8.5); and
• The measurement method's ruggedness, which refers to the relative stability of measurement
method performance for small variations in measurement method parameter values (see
Section 3.8.6 and Appendix D).
Implement the Survey Design
The implementation phase of the data life cycle is when the activities described in the survey
design are performed. Roadmap Figure 12 illustrates the process for implementing disposition
surveys.
MARSAME, like MARS SIM, does not provide prescriptive guidance for implementing survey
designs. Chapter 5 presents topics to be considered while implementing disposition surveys. This
approach allows MARSAME users flexibility to use either existing or new and innovative
techniques that meet the survey objectives.
Evaluate the Results
The assessment phase of the data life cycle involves evaluating the results of the survey as shown
in Roadmap Figure 13. DQA is used to evaluate the survey results. DQA is a scientific and
statistical evaluation that determines whether data are the type, quality, and quantity to support
their intended use. When individual measurement results are not recorded, as allowed in some
scan-only survey designs, the preliminary data review will be brief and based primarily on the
results of quality control (QC) measurements. To increase the flexibility and general
applicability of MARSAME, several evaluation methods have been incorporated in addition to
the Sign test and Wilcoxon Rank Sum (WRS) test used in MARSSIM. Roadmap Figure 14
presents information on interpreting survey results for scan-only and in situ surveys. Roadmap
Figure 15 presents information on interpreting survey results for MARSSIM-type surveys.
Summary
The roadmap presents a summary of the data life cycle as it applies to disposition surveys in
MARSAME and identifies where information on important topics are located in MARSAME.
Flow charts are provided to summarize major steps in the survey design process, again citing
appropriate references in MARSAME.
January 2009 RM-5 NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
Categorization
Initial Assessment
Preliminary Surveys
<
Q.
LU
^.
LU
_l
Q.
CO
CO
LJJ
CO
CO
LJJ
g
o
LU
Q
Decision Rule
Design Disposition Survey
Disposition Survey
Verification
&
Validation
Evaluate Results
Decision
Roadmap Figure 1. Overview of MARSAME Process
NUREG-1575, Supp. 1
RM-6
January 2009
-------
MARSAME
Roadmap
<
CL
Yes
Are
Preliminary Surveys
Needed to Describe
The M&E?
Are the M&E Impacted?
(Section 2.2)
Design and Implement
Preliminary Surveys
(Section 2.3)
Document Non-Impacted
Decision, If Necessary
(Section 2.2.5)
Describe the M&E
(Section 2.4)
Select Appropriate
Disposition Options
(Section 2.5)
Is the
Existing Survey Design
Applicable to the M&E?
(Section 3.10)
Is There
An Existing Survey
Design?
(Section 2.6)
Develop Decision Rule(s)
(Chapter 3)
Does the
urvey Design Meet the
DQOs?
Section 6.2.1
Develop a Survey Design
(Chapter 4)
Implement the Survey Design
(Chapter 5)
Evaluate the Survey Results
(Chapter 6)
Make a Disposition Decision
(Section 6.8)
Finalize Radionuclides of Concern
Select Action Levels
Define Parameter of Interest
Define Survey Unit Boundaries
Develop Measurement Quality Objectives
Identify Alternative Actions
Define the Null Hypothesis
Specify Limits on Decision Errors
NOTE: Shaded boxes
represent important decisions
(diamonds) or milestones
(rectangles).
Roadmap Figure 2. The Data Life Cycle Applied to Disposition Surveys
January 2009
RM-7
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
Review Existing
Relevant Information
-Yes
Is the
Existing Information
dequate to Categorize
the M&E?
1
1
Perform a
Visual Inspection
(Section 2.2.1)
Review
Historical Records
(Section 2.2.2)
i
r
Assess Process
Knowledge
(Section 2.2.3)
r
Are Sentinel
Measurements
Applicable?
Yes-
Plan and Perform
Sentinel
Measurements
(Section 2.2.4)
Documentation
of the Non-Impacted
Decision
Necessary?
Yes-
Proceed to
Preliminary Survey
(Roadmap Figure 4)
NOTE: Shaded diamonds
represent important decision
points.
Yes-
Prepare Documentation of
the Non-Impacted Decision
(Section 2.2.5)
No Further Action
Roadmap Figure 3. The Categorization Process as Part of Initial Assessment
NUREG-1575, Supp. 1
RM-8
January 2009
-------
MARSAME
Roadmap
From Roadmap Figure 3
Is the Existing
Information Adequate to Select a
Disposition Option?
Select a Disposition Option
(Section 2.5)
Is the Existing
Information Adequate
to Design a Disposition
Survey?
Identify Data Gaps
(Section 2.3)
Yes-
Proceed to Roadmap
Figure 5
Design and Implement
Preliminary Surveys
(Section 2.3)
Describe the M&E
(Section 2.4)
NOTE: Shaded diamonds
represent important decision
points.
Roadmap Figure 4. Assessing Adequacy of Information for Designing
January 2009
RM-9
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
From Roadmap Figure 4
Yes
Is an
Applicable SOP Available
forth is M&E?
Implement and Document the
Results of the Survey as
Described in the SOP
(Section 2.6.1)
Develop a Conceptual Model
and Document the IA
(Section 2.6.2)
Proceed to Roadmap Figure 13
Proceed to Roadmap Figure 6
NOTE: Shaded box represents important milestone.
Roadmap Figure 5. Assessing the Applicability of Existing SOPs
NUREG-1575, Supp. 1
RM-10
January 2009
-------
MARSAME
Roadmap
From Roadmap Figure 5
i
i
Select Radionuclides or
Radiations of Concern
(Section 3.2)
i
i
Identify Reference Materials
Identify Action Levels D
(Section 3.3) * ^
i
Describe the Parameter of Interest
(Section 3.4)
i
Identify Alternative Actions
(Section 3.5)
i
'
Identify Survey Units
(Section 3.6)
Retur
Figu
to
map
re 7
r
ifrom
map
re 7
Go
Figu
i
Develop a Decision Rule
(Section 3.7)
i
Retur
Figu
'
Develop Inputs for Selection of
Provisional Measurement Methods
(Section 3.8)
to
map
re 8
r
i from
map
re 8
(Section 3.9)
i
/"Is there a
\. Survey
Y
/Do the M&
\. Survey Rec
Y
*
n Existinjhv
Design? S
3S
E Meet trie's.
uirements? V/
3S
Implement and Document Results
as Described in the Survey
(Section 3. 10)
NOTE: Shaded boxes
represent important
milestones.
No
Roadmap Figure 6. Identify Inputs to the Decision
January 2009
RM-11
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
From Roadmap Figure 6
Identify Applicable
Regulatory Limits
(Dose-, Risk-, Activity-,
or Method Based)
Identify Applicable
Requirements
(e.g., ANSI N13.12)
Identify Applicable
Administrative Limits
(e.g., Waste Acceptance
Criteria)
Identify Applicable DOT
Requirements for
Shipping M&E
Convert Potential Action Levels
into Measurement Units
NOTE: Shaded boxes
represent important
milestones.
Finalize Selection of Action Level(s)
Modify AL Using
Equation 3-1
(Gross Activity AL)
Are There Multiple
Radionuclides?
Radionuchde-
Specific Measurements?
Will Surrogates
Infer Multiple
Radionuclides?
Surrogate
Measurements
Available?
Modify AL Using
Equation 3-4
Modify AL Using
Equation 3-3
Evaluate Survey Results Using
Equation 3-2 (Unity Rule)
Apply ALARA, as Appropriate
Document Selection of
Action Level(s)
Return to Roadmap Figure 6
NOTE: Information on ALARA
can be found in 10CFR20, 10CFR
835, DOE 1993, ICRP 1989,
NCRP 1993, NRC 1977, NRC
1982, NRC 1993, NRC 2002, and
PNL1988.
Roadmap Figure 7. Identify Action Levels
NUREG-1575, Supp. 1
RM-12
January 2009
-------
MARSAME
Roadmap
Are Survey
Unit Dimensions Specified
in a Regulation?
(Section 3.3.1)
Determine Assumptions Used to
Develop Action Levels
(Sections 3.3.1 and 3.2.4)
Yes
NOTE: Shaded boxes
represent important
milestones.
Develop Survey Unit Boundaries
Based on Assumptions
(Section 3.6.1)
Identify Parameter of Interest
Identify Target Population
Reduce Survey Unit Size Based on
Physical Characteristics of the M&E
(Sections 3.6.1, 2.3.1, and 5.3)
Handling (Size, Shape, Mass)
Acessibility
Reduce Survey Unit Size
Based on Measurement
Method Requirements
(Sections 3.6.1, 3.8, and 5.9)
Examples:
Dimensions of Box Counter or Portal Monitor
Field of View for In Situ Gamma Spectrometer
Penetrating Power of Radioactivity
Identify Final
Survey Unit Boundaries
Document Development of Survey
Unit Boundaries as Part of the
Survey Design
(Section 4.5)
Return to Roadmap Figure 6
Roadmap Figure 8. Developing Survey Unit Boundaries (Apply to all Impacted M&E for
each set of Action Levels Identified in Section 3.3)
January 2009
RM-13
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
From Roadmap Figure 6
Select a Null Hypothesis
Scenario A (Section 7.2.3)
Scenario B (Section 7.2.4)
Assign Values to the
LBGR and UBGR
NOTE: Shaded boxes
represent important decisions
(diamonds) or milestones
(rectangles).
NOTE: A method-based survey
(Section 4.4.4) is a special case of
either scan-only, in situ, or
MARSSIM-type, and so will follow
one of the three paths show on this
figure.
Action Level (Section 3.3)
Discrimination Limit (Section 4.2.2)
Specify Limits on
Decision Errors
Type I Error (Section 4.2.5)
Type II Error (Section 4.2.5)
Consequences of Decision Errors
a: and /?
Develop an Operation
Decision Rule
Statement of the Statistical
Hypothesis Test (Section 4.2.6)
Scan-Only
(Section 4.4.1)
Select a Type of
Disposition Survey
Class 1 (Section 4.3.1)
Class 2 (Section 4.3.2)
Class 3 (Section 4.3.3)
MARSSIM-Type
(Section 4.4.3)
Determine % of M&E
to be Scanned
In Situ (Section 4.4.2)
t
Determine % of M&E and
Locations to be Measured
Select M&E and Locations
to be Scanned
Select M&E and Locations
to be Measured
Select Measurement and
Scan Locations
Optimize the Survey Design
Document the Disposition
Survey Design
(Section 4.5)
Proceed to Roadmap Figure 12
Roadmap Figure 9. Flow Diagram for Developing a Disposition Survey Design
NUREG-1575, Supp. 1
RM-14
January 2009
-------
MARSAME
Roadmap
From Roadmap Figure 9
Estimate a's, the Variabilities in
the Reference Area and
Survey Unit Activities
Estimate cr, the Variability in
the Level of Radioactivity
Is the Radioactivity
Present in Bkgd?
Calculate the Relative Shift A/a
Calculate the Relative Shift A/a
Is A/a
Between
1 and 3?
Is A/a
Between
1 and 3?
Obtain Number of Data Points
for the Sign Test, N, from Table
for each Survey Unit
Obtain Number of Data Points
for WRS Test, N/2, from Table
for each Survey Unit and
Reference Area
NOTE: Shaded boxes
represent important
milestones.
Do the Number of Data
Points Need to be Adjusted
for Class 1 M&E?
-Yes-
Prepare Summary of Data
Points for M&E being
Investigated
Return from
Roadmap Figure 11
Return to
Roadmap Figure 9
Roadmap Figure 10. Flow Diagram for Identifying the Number of Data Points for a
MARSSIM-Type Disposition Survey
January 2009
RM-15
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
From Roadmap Figure 10
Establish DQOs for Areas with the
Potential for Exceeding
Als and Acceptable Risk for
Missing Such Areas
Identify Number of Data Points
needed Based on Statistical Tests, n
Determine the Acceptable
Concentration Corresponding to the
Calculated Area, A (i.e., Area
Factor x Action Level)
Determine Acceptable
Concentrations in Various
Individual Smaller Areas within a
Survey Unit (i.e., Use Area Factors)
Calculate the Area, A, Bounded by
Sample Location, n
Determine the Required Scan MDC
to Identify the Acceptable
Concentration in an Area, A
NOTE: Shaded boxes
represent important
milestones.
Evaluate MDCs for Available
Instrumentation
s the Scan MDC for Available
Instrumentation Less than the
Required Scan MDC?
Yes-
No Additional Sampling Points
are Necessary for Potential
Elevated Locations
Calculate Area Factor that
Corresponds to the Actual Scan
MDC (scan MDC/AL)
NOTE: "VOLUME" or
"MASS" replaces "AREA" in
this flow diagram as
appropriate for a specific
survey design, and scan MDC
is discussed in MARSSIM
Section 5.5.2.4
Determine the Maximum
Area, A', that Corresponds
to the Area Factor
Recalculate Number of
Data Points Needed
(nEA = Survey Unit Area//!')
Determine Grid Size Spacing, L
Return to Roadmap
Figure 10
Roadmap Figure 11. Flow Diagram for Identifying Data Needs for Assessment of Potential
Areas of Elevated Activity in Class 1 Survey Units for MARSSIM-Type Disposition
Surveys
NUREG-1575, Supp. 1
RM-16
January 2009
-------
MARSAME
Roadmap
From Roadmap Figure 9
Ensure Protection of
Health and Safety
(Section 5.2)
Handle M&E
(Section 5.3)
Prepare M&E for Survey (Section 5.3.1)
Provide Access to M&E (Section 5.3.2)
Transport the M&E (Section 5.3.3)
Segregate the M&E
(Section 5.4)
Do M&E Need
Segregation?
No
T
Set Measurement
Quality Objectives
(Section 5.5-5.8)
NOTE: Shaded boxes
represent important decisions
(diamonds) or milestones
(rectangles).
Select a Measurement
Technique and
Instrumentation Combination
(Section 5.9)
Set Quality
Control Requirements
(Section 5.10)
Perform the Survey &
Report the Results
(Section 5.11)
Proceed to Roadmap Figure 13
Roadmap Figure 12. Implementation of Disposition Surveys
January 2009
RM-17
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
From Roadmap Figure 12
Conduct Data Quality
Assessment
(Section 6.2)
Proceed to Roadmap
Figure 14
-Yes
Is the Survey Design
Scan-Only or in Situ?
Proceed to Roadmap
Figure 15
Return from Roadmap
Figure 14
Evaluate the Results and
Make a Decision
(Section 6.8)
Return from Roadmap
Figure 15
Document the Disposition
Survey Results
(Section 6.10)
NOTE: Shaded box
represents important
milestone.
No Further Action
Roadmap Figure 13. Assess the Results of the Disposition Survey
NUREG-1575, Supp. 1
RM-18
January 2009
-------
MARSAME
Roadmap
From Roadmap Figure 13
Disposition
Decision Based
on Mean of a Sampled
Population?
Is the
AL Equal to Zero
or Background?
Disposition
Decision Based on
Individual
Items?
Requires Scenario B
LBGR = AL
Scan MDC < UBGR
Recording Individual
Scan Results
Not Required
Individual Results Must
be Recorded
All
Results < S
from the
MDC?
All
Results < S
from the
UBGR?
M&E Do Not Meet the
Disposition Criterion
(Section 6.9)
M&E Do Not Meet the
Disposition Criterion
(Section 6.9)
1
F
Return to Roadmap
Figure 13
Yes
Yes
1
r
Return to Roadmap
Figure 13
Yes
/ M&E Meet the
\, Disposition Criterion
1
r
Return to Roadmap
Figure 13
Roadmap Figure 14. Interpretation of Survey Results for Scan-Only and
In Situ Surveys
January 2009
RM-19
NUREG-1575, Supp. 1
-------
Roadmap
MARSAME
NOTE: An elevated
measurement comparison
also needs to be performed
for MARSSIM-type surveys.
From Roadmap Figure 13
Radionuchde
of Concern Present in
Background?
Radionuchde-
Specific
Measurements?
S+ > q?
- LBGR)
S+ > q?
(UBGR-Xi)
/M&E Do Not Meet th^\
Disposition Criterion
q or
more of the r
Largest Values
from the Survey
Unit?
Disposition Criterion
/ M&E Do Not Meet the
Disposition Criterion
V (Section 6.9)
Roadmap Figure 15. Interpretation of Survey Results for MARSSIM-Type Surveys
NUREG-1575, Supp. 1
RM-20
January 2009
-------
MARSAME Introduction and Overview
1 INTRODUCTION AND OVERVIEW
1.1 Purpose and Scope of MARSAME
Large quantities of materials and equipment (M&E) potentially affected by radioactivity are
present throughout the United States. The potential for residual radioactivity can come from use
of source, byproduct, and special nuclear materials as well as naturally occurring radioactive
material (NORM), naturally occurring and accelerator-produced radioactive materials (NARM)
and technologically enhanced naturally occurring radioactive material (TENORM). This M&E
may be commercial, research, education, or defense related. The M&E might be—
• Used or stored at sites and facilities licensed to handle radioactivity,
• Commercial products purposely containing radionuclides (e.g., smoke detectors),
• Commercial products incidentally containing radionuclides (e.g., phosphate fertilizers), or
• Associated with NARM and TENORM.
The owners of M&E potentially affected by radioactivity need to determine acceptable
disposition options for M&E currently under their control. Industries or facilities sensitive to the
presence of radioactivity need to evaluate the acceptability of M&E coming under their control.
Regulatory agencies need to distinguish items in general commerce that are inherently
radioactive from illicit trafficking of radioactive M&E.
This Multi-Agency Radiation Survey and Assessment of Materials and Equipment manual
(MARSAME) is a supplement to the Multi-Agency Radiation Survey and Site Investigation
Manual (MARS SIM). Like MARS SIM, MARSAME is a joint effort by the Department of
Defense (DOD), Department of Energy (DOE), Environmental Protection Agency (EPA), and
Nuclear Regulatory Commission (NRC). Information on MARSSIM can be found on the World
Wide Web (MARSSIM 2002). MARSAME also incorporates information for measuring
radioactivity from the Multi-Agency Radiological Laboratory Analytical Protocols manual
(MARLAP 2004). MARSAME provides information on surveys where radiological control of
M&E could be initiated, maintained, removed, or transferred (i.e., an M&E disposition) to
another responsible party. In addition, MARSAME discusses the need for a graded approach to
surveying M&E.
MARSAME provides technical information on approaches for planning, implementing,
assessing, and documenting surveys to determine proper disposition of M&E. Release (including
clearance) and interdiction are types of disposition options in MARSAME. Detailed descriptions
of these disposition options are provided in Chapter 2.
Examples of M&E include metals, concrete, tools, equipment, piping, conduit, furniture, and
dispersible bulk materials such as trash, rubble, roofing materials, and sludge. Liquids, gases,
and solids stored in containers (e.g., drums of liquid, pressurized gas cylinders, containerized
soil) are also included in the scope of this document.
January 2009 1-1 NUREG-1575, Supp. 1
-------
Introduction and Overview MARSAME
Radionuclides or radioactivity on workers or members of the public are outside the scope of the
document, as are liquid and gaseous effluent releases and real property (e.g., fixed buildings and
structures, surface and subsurface soil remaining in place).
The purpose of this supplement is to provide information for the design and implementation of
technically defensible surveys for disposition of M&E. MARSAME provides information on
selecting and properly applying disposition survey strategies and selecting measurement
methods. The data quality objectives (DQO) process is used for selecting the best disposition
survey design based on the selected disposition option, action level, description of the M&E
(e.g., size, accessibility, component materials), and description of the radioactivity (e.g.,
radionuclides, types of radiation, surficial versus volumetric activity). Detailed information on
the DQO process can be found in EPA QA/G-4 (EPA 2006a), MARS SIM Appendix D, and
MARLAP Appendix B.
This supplement describes a number of different approaches for performing technically
defensible disposition surveys and provides information for optimizing survey designs. However,
MARSAME does not represent the only acceptable approach to radiologically evaluate M&E.
MARSAME describes a graded approach that the signatory agencies find acceptable and useful
for most situations. The signatory agencies recognize that alternative approaches or modification
of the MARSAME procedures may be appropriate or necessary for some situations. Nothing in
MARSAME should be construed to prohibit the use of other appropriate procedures.
Disposition surveys may be performed as a single event or as part of a routine process. Single
event disposition surveys are usually performed once in association with a specific project.
Surveying a backhoe at the completion of a decommissioning project is one example of a single
event disposition survey. Routine process disposition surveys are usually associated with
ongoing tasks where similar surveys are performed repeatedly. One example of a routine process
disposition survey would be a radiological survey of tools prior to removal from a controlled
area at a nuclear facility. Both single event and routine process types of surveys are included in
the scope of MARSAME.
The guidance in MARSAME is designed to incorporate existing survey methods whenever
possible, while at the same time allowing the use of new and innovative survey techniques when
appropriate. The use of previously established and accepted standard operating procedures
(SOPs) as part of a standardized initial assessment (IA) is described in Section 2.6.1. The use of
SOPs that document approved methods for performing disposition surveys along with assessing
the results of these surveys can reduce the effort required to develop new survey designs, since
the survey design effort was applied when the SOP was developed. MARSAME also allows
consideration of innovative survey techniques through the modification of existing SOPs or
development of new survey designs as described in Chapters 3, 4, and 5. Prior to
implementation, existing SOPs should be evaluated to ensure they meet the survey design
objectives.
MARSAME assumes the user has some historical knowledge of the M&E being investigated.
The historical information is gathered during the initial assessment (IA) to determine acceptable
disposition options (Chapter 2). The characteristics, history of prior use, and inherent
radioactivity of the M&E are important when determining the appropriate disposition options.
NUREG-1575, Supp. 1 1-2 January 2009
-------
MARSAME Introduction and Overview
The historical information is termed "process knowledge." The role of process knowledge
(discussed in Chapter 2) is important in providing information on the nature and amount of
radioactivity that might be expected on, or incorporated in, the M&E being investigated. If no
historical information is available, information on the current status of the M&E can be
determined using preliminary surveys (i.e., scoping, characterization, remedial action support)
prior to designing a disposition survey.
The recommendations in this supplement may be applied to a broad range of regulations,
including dose-, risk-, or radionuclide concentration-based regulations. The translation of a
regulatory dose- or risk-based limit to a corresponding concentration level is not addressed in
MARSAME. The terms dose, risk, and dose- or risk-based regulation are used throughout the
supplement, but these terms are not intended to limit the applicability of this supplement.
MARSAME can be applied to activity concentrations (e.g., Bq/m2) without associated dose or
risk values. MARSAME does not address the regulatory status of the M&E (e.g., NRC exempted
or excluded materials).
MARSAME uses the word "should" as a recommendation. This is not to be interpreted as a
requirement. The user need not assume that every recommendation in this supplement will be
taken literally and applied to every project. Rather, it is expected the survey documentation will
address how the recommendations will be applied on a project-specific basis.
1.2 Understanding Key MARSAME Terminology
In order to understand the information in MARSAME, the user should first become familiar with
the scope of this supplement, the terminology, and the concepts in this document. As a
supplement to MARSSIM, MARSAME uses terms generally consistent with MARSSIM. Some
additional terms were developed for MARSAME, while other commonly used terms were
adopted from other sources. This section explains some of the terms used in this supplement.
The terms impacted., non-impacted, and graded approach are defined in MARSSIM. These
terms are used consistently in MARSSIM and MARSAME. Unlike MARSSIM, which applies to
land, structures, or buildings, MARSAME applies to M&E. The action taken may initiate,
maintain, remove, or transfer radiological controls associated with the M&E. The decision to
take action may be largely based on the results of a radiological survey designed to evaluate the
disposition of the M&E, either through release or interdiction. Therefore, the terms release
criterion, derived concentration guideline level (DCGL), and final status survey used in
MARSSIM are replaced by the more generic terms disposition criterion, action level, and
disposition survey, respectively, in MARSAME.
Disposition is the future use, fate, or final location for something (e.g., recycle, reuse, disposal).
Disposition options range from release to interdiction:
• Release is a reduction in the level of radiological control, or a transfer of control to another
party. Release includes clearance. Examples of release (other than clearance) include recycle,
reuse, disposal as waste, or transfer of control of radioactive M&E from one authorized user
to another.
January 2009 1-3 NUREG-1575, Supp. 1
-------
Introduction and Overview MARSAME
• Interdiction is an increase in the level of radiological control or a decision not to accept
control from another party. Examples of interdiction include identification of radioactive
material that results in the initiation of radiological controls or identification of unauthorized
movement of radioactive material.
Categorization is the act of determining whether M&E are impacted or non-impacted. This is a
departure from MARSSEVI where this decision was referred to as classification. This change was
made to emphasize the difference between the decision of whether a survey is needed (i.e.,
impacted or non-impacted) and the determination of the appropriate level of survey effort (i.e.,
classification).
Classification is the act or result of separating impacted M&E or survey units into one of three
designated classes: Class 1, Class 2, or Class 3. Classification is the process of determining the
appropriate level of survey effort based on estimates of activity levels and comparison to action
levels, where the activity estimates are provided by historical information, process knowledge,
and preliminary surveys.
Measurable radioactivity is radioactivity that can be quantified using known or predicted
relationships developed from historical information, process knowledge or preliminary
measurements as long as the relationships are developed, verified, and validated as specified in
the DQOs and measurement quality objectives (MQOs). Measurability is of primary importance
in MARSAME.
Surficial radioactive material is radioactive material distributed on any of the surfaces of a solid
object. Surficial radioactive material may be removable (by non-destructive means such as
casual contact, wiping, brushing, or washing) or fixed. Surfaces may either be accessible or
difficult-to-measure. Changes to the surface (e.g., paint, dirt, oxidation) may affect the
measurability and the physical condition of Surficial radioactive material.
Survey unit for M&E is the specific lot, amount, or piece of M&E on which measurements are
made to support a disposition decision concerning the same specific lot, amount, or piece of
M&E. The survey unit defines the spatial boundaries for the disposition decision and a separate
decision is made for each survey unit, similar to MARSSIM. The survey unit boundaries also
define the population for the parameter of interest.
Volumetric radioactive material is radioactive material that is distributed throughout or within
the material or equipment being measured, as opposed to a surficial distribution. Volumetric
radioactive material may be homogeneously (e.g., uniformly activated metal) or heterogeneously
(e.g., activated reinforced concrete) distributed throughout the M&E. Volumetric radioactive
material may be distributed throughout the M&E being measured or distributed in layers. Layers
of volumetric radioactive material may start at the surface (e.g., porous surfaces penetrated by
radioactive material) or under a layer of other material (e.g., activated rebar inside a concrete
wall). By definition all radioactive liquids and gases in containers and all bulk quantities of
radioactive material when measured as a whole are volumetric radioactive material.
NUREG-1575, Supp. 1 1-4 January 2009
-------
MARSAME Introduction and Overview
The concept of whether radioactivity is measurable is the major factor in demonstrating
compliance with an action level. MARSAME does not provide an exact definition for the
transition between surficial and volumetric radioactive material. Rather, the assumptions used to
quantify the radioactivity need to be clearly defined and identified so they can be compared to
the DQOs and MQOs. Individual action levels may specify applicability to surficial or
volumetric radioactivity. In these cases, the definition of surficial and volumetric radioactivity
should be specified as part of the definition of the action level.1
Accessible area is an area that can be easily reached or obtained. In many cases an area must be
physically accessible to perform a measurement. However, radioactivity may be measurable
even if M&E are not physically accessible (e.g., energetic gamma rays may be quantified even
after passing through a layer of shielding).
Difficult-to-measure radioactivity is radioactivity that is not measurable until the M&E to be
surveyed is prepared. Preparation of M&E may be relatively simple (e.g., cleaning) or more
complicated (e.g., disassembly or complete destruction). Given sufficient resources, all
radioactivity can be made measurable; however, it is recognized that increased survey costs can
outweigh the benefit of some dispositions.
Initial assessment (IA) is an investigation to collect existing information describing M&E and is
similar to the Historical Site Assessment (HSA) described in MARSSIM. The IA provides initial
categorization of M&E as impacted or non-impacted. In addition to the HSA activities described
in MARSSIM, the IA may lead to grouping or segregating M&E with similar characteristics as
well as designing and implementing preliminary surveys. The IA also identifies the expected
disposition of the M&E (e.g., clearance, radiological control, recycle, reuse, disposal). The
results of the IA provide most, if not all, information needed to design a disposition survey for
impacted M&E. A graded approach is used to determine the level of effort applied during the IA.
Sentinel measurement is a biased measurement performed at a key location to provide
information specific to the objectives of the IA (see Section 2.2.4). Sentinel measurements
cannot be used as the only source of information to support a decision that M&E are non-
impacted. The objective of performing sentinel measurements as part of the IA is to gather
additional information to support a decision regarding further action, verify assumptions based
on process knowledge, provide additional support to a finding of impacted or non-impacted for
M&E, and to distinguish illicit or inadvertent transport of radioactive materials from items in
general commerce that are inherently radioactive (e.g., fertilizers, phosphates, sand-blasting grit).
1.3 Use of MARSAME
MARSAME provides technical information describing a framework for planning, implementing,
and assessing radiological surveys of M&E. MARSAME does not establish or supersede any
regulatory or license requirements. Federal and State regulatory agencies may have requirements
or guidance that differs from what is presented in MARSAME and may be implemented as
1 This idea is consistent with the definition of a surface soil sample provided in the MARSSIM Glossary. A surface
soil sample is a sample that reflects the modeling assumptions used to develop the DCGL for surface soil activity.
The example in MARSSIM references 40 CFR 192, which defines surface soil as the first 15 cm of soil.
January 2009 1-5 NUREG-1575, Supp. 1
-------
Introduction and Overview MARSAME
appropriate. Consequently, persons planning, implementing, and assessing disposition surveys
should also obtain appropriate regulatory approval for the procedures that are in use to maintain
regulatory compliance.
Potential users of this supplement are Federal, State, and local government officials having
authority for control of radioactive M&E, their contractors, and other parties such as
organizations with licensed authority to possess and use radioactive materials. This supplement
to MARSSIM is intended for a technical audience having knowledge of radiation health physics
and an understanding of statistics as well as experience with the practical applications of
radiation protection. Understanding and applying the recommendations in this supplement
requires knowledge of instrumentation and measurement methodologies as well as expertise in
planning, approving, and implementing radiological surveys. Certain situations and projects may
require consultation with more experienced or specialized personnel (e.g., a statistician).
MARSAME users with less professional experience than described above should still be able to
apply the majority of guidance found in this supplement although obtaining technical support is
recommended. The wide range of topics and subjects covered by MARSAME emphasizes the
need for a well rounded planning team as described in the first step of the DQO process. While it
may be difficult to identify a single person with all the required technical experience to design an
appropriate survey, it is easier to assemble a small group of experts with the required range of
knowledge. Consultation with the responsible regulatory agency is critical for the success of all
disposition surveys, and even more so for MARSAME users with less professional experience.
In addition, MARSAME provides information in Appendix B, Appendix C, and Appendix D that
may be useful to users with less professional experience.
MARSAME recommends that a graded approach be applied to the disposition of M&E. Non-
impacted M&E are removed from further consideration early in the process through
categorization. Impacted M&E are classified based on the level of residual radioactivity so that a
higher level of scrutiny can be applied to M&E with the highest potential for residual
radioactivity. Finally, MARSAME includes practical considerations such as inherent value of the
M&E and handling the M&E when evaluating options for disposition. The combination of these
considerations results in a graded approach where an appropriate level of survey effort is applied
to M&E to minimize the impacts of any decision errors.
1.4 Overview of MARSAME
The data life cycle is the foundation for the design, implementation, and assessment of surveys
for disposition of M&E in this supplement. However, before commencing survey planning the
user must select an appropriate disposition option. Multiple disposition options may exist.
Consider all of the various disposition options and develop the most appropriate option for a
given situation. Survey designs may then be planned using the DQO process, which is often
iterative. The DQO process iterations may take place at different times during the disposition
process, for example during the IA as well as during the disposition survey. The different survey
designs are compared and the most resource-effective design that meets the survey objectives is
selected for implementation. Following implementation of the selected survey design, the results
NUREG-1575, Supp. 1 1-6 January 2009
-------
MARSAME
Introduction and Overview
are evaluated using data quality assessment (DQA). A technically defensible decision regarding
disposition of the M&E can then be made.
Whenever practical, MARSAME recommends designing disposition surveys where one hundred
percent of the M&E are measurable. This means that all radioactivity associated with the M&E
has been measured and quantified (e.g., 100% scan with conventional instruments, measurement
with a box counter, or measurement using in situ gamma spectroscopy), a known or accepted
relationship was used to estimate concentrations for difficult to measure radionuclides using
surrogate measurements,2 or that a known or accepted relationship allows quantification of
radioactivity in areas that were not measured. MARSAME employs the use of a graded approach
to determine if a 100% measurable survey is practical and to ensure that a sensible,
commensurate balance is achieved between resource expenditures and risk reduction.
MARSAME uses the data life cycle to design disposition surveys. The data life cycle is
described in MARSSIM Section 2.3, and consists of four phases:
• Planning phase (MARSAME Chapters 2, 3, and 4; MARSSIM Chapters 3, 4, and 5),
• Implementation phase (MARSAME Chapter 5; MARSSIM Chapters 6 and 7),
• Assessment phase (MARSAME Chapter 6; MARSSIM Chapter 8), and
• Decision-making phase (MARSAME Chapter 6; MARSSIM Chapter 8).
A brief description of each of the phases and how they apply to the disposition survey design
process is provided in the following sections. Table 1.1 provides a simplified overview of the
principal steps in designing a disposition survey and illustrates how the data life cycle can be
used in an iterative fashion within the survey process. Figure 1.1 illustrates how the data life
cycle is applied to disposition surveys.
Table 1.1 The Data Life Cycle Used to Support Disposition Survey Design
Disposition Survey
Design Process
Categorization
Preliminary Surveys
Disposition Survey
Data Life Cycle
Categorization
Data Life Cycle
Preliminary
Survey Data Life
Cycle
Disposition Survey
Data Life Cycle
Plan
Implement
Assess
Decide
Plan
Implement
Assess
Decide
Plan
Implement
Assess
Decide
MARSAME Processes
Provides information on collecting and
assessing existing data (Section 2.2)
Discusses the purpose (i.e., filling data
gaps) and general approach to performing
preliminary surveys (Section 2.3)
Provides detailed information for planning
(Chapters 3 and 4), implementing (Chapter
5), and assessing (Chapter 6) disposition
surveys
2 The MARSSIM term "surrogate measurement" as used here is consistent with the MARLAP term "alternate
radionuclide."
January 2009
1-7
NUREG-1575, Supp. 1
-------
Introduction and Overview
MARSAME
<
CL
Yes
Are
Preliminary Surveys
Needed to Describe
The M&E?
Are the M&E Impacted?
(Section 2.2)
Design and Implement
Preliminary Surveys
(Section 2.3)
Document Non-Impacted
Decision, If Necessary
(Section 2.2.5)
Describe the M&E
(Section 2.4)
Select Appropriate
Disposition Options
(Section 2.5)
Is the
Existing Survey Design
Applicable to the M&E?
(Section 3.10)
Is There
An Existing Survey
Design?
(Section 2.6)
Develop Decision Rule(s)
(Chapter 3)
Does the
urvey Design Meet the
DQOs?
Section 6.2.1
Develop a Survey Design
(Chapter 4)
Implement the Survey Design
(Chapter 5)
Evaluate the Survey Results
(Chapter 6)
Make a Disposition Decision
(Section 6.8)
Finalize Radionuclides of Concern
Select Action Levels
Define Parameter of Interest
Define Survey Unit Boundaries
Develop Measurement Quality Objectives
Identify Alternative Actions
Define the Null Hypothesis
Specify Limits on Decision Errors
NOTE: Shaded boxes
represent important decisions
(diamonds) or milestones
(rectangles).
Figure 1.1 The Data Life Cycle Applied to Disposition Surveys
NUREG-1575, Supp. 1
1-8
January 2009
-------
MARSAME Introduction and Overview
1.4.1 Planning Phase
The planning phase is where the survey design is developed and documented using the DQO
process. The survey design documents the decision rule as well as the number, type, and location
of measurements required to support the disposition decision. Soliciting input from regulatory
agencies early in the planning phase helps ensure the disposition survey results will meet
regulatory needs.
MARSAME processes begin with the historical evaluation of the M&E being investigated. This
IA usually combines a review of process knowledge and historical records with a visual
inspection of the M&E. The results of the IA are used to develop a conceptual model describing
the physical characteristics of the M&E and providing information on the radioactivity
potentially associated with the M&E. The physical description of the M&E should include
information on the size, shape, complexity (e.g., can it be broken down or combined with other
M&E), accessibility (e.g., can the surveyor physically access areas of concern to perform
measurements), and inherent value (i.e., resources associated with reuse, recycle, repair,
remediation, replacement, and disposal). Information on radioactivity should include the
radionuclides of potential concern, the expected levels of radioactivity, the distribution of
radioactivity (e.g., uniform or not), and the location of the radioactivity (i.e., surface or volume).
The IA may also include data collection in the form of sentinel measurements. The results of
sentinel measurements can be used as the basis to reject assumptions based on process
knowledge. However, sentinel measurements alone cannot be used to justify the categorization
of M&E as non-impacted (see Section 2.2.4 for information on sentinel measurements).
There are two decisions associated with the IA. The first decision, called categorization, is
whether or not the M&E are impacted. Non-impacted M&E do not require additional
investigation, but may require documentation of the justification for the non-impacted decision.
The second decision is to select an appropriate disposition option for impacted M&E at the end
of the IA to provide direction for designing a disposition survey. Additional information may be
required before a disposition survey can be designed. Preliminary surveys (e.g., scoping,
characterization, and remedial action support surveys) may be performed as part of the IA to
collect this additional information.
For single event surveys, the IA should focus on collecting the information necessary to develop
a technically defensible disposition survey design. Information necessary to design a disposition
survey includes a description of the M&E and the radioactivity potentially associated with the
M&E. The results of the IA are carried forward and used to develop the survey design, which is
usually documented in a project-specific work plan.
For routine process surveys, the IA should lead to an existing survey design from a standard
operating procedure (SOP), if applicable, or develop a new survey design for documentation in
an SOP. The SOP should clearly state the assumptions used to develop the survey design, along
with a description of the M&E and radioactivity covered by the SOP. The selection process is
based on evaluating the M&E to determine if the survey design in a specific SOP is applicable.
Documentation of individual survey results may not be required as long as there are records
January 2009 1-9 NUREG-1575, Supp. 1
-------
Introduction and Overview MARSAME
showing that the SOP was approved, the instruments were working properly, and the personnel
performing the survey were properly trained. Development of SOPs is usually accomplished
using the same processes as those used to develop single event surveys. There may be regulatory
or site-specific guidance that specifies documentation requirements for SOPs. Information on
developing SOPs can be found in EPA QA/G-6 (EPA 2001).
Following the IA, it is necessary to develop a decision rule for the disposition of M&E being
investigated. The decision rule is an "if...then..." statement consisting of three parts:
• Action level(s),
• Parameter of interest, and
• Alternative actions.
An example of a decision rule might be "If the average surficial activity concentration is less
than a level specified by the regulator, then the M&E can be cleared, otherwise the M&E are not
cleared." The parameter of interest is closely related to the description of the M&E, the
description of the radioactivity, and the survey unit boundaries. The action level is influenced by
the selection of a disposition option. The selected disposition option defines two alternative
actions. A decision rule should be developed for each decision to be made concerning the M&E.
For example, if the action level is stated in terms of total activity, generally only one decision
rule is required. If, on the other hand, the action level provides limits for fixed, removable, and
maximum levels of radioactivity, e.g., DOE Order 5400.5, Figure IV-1 (DOE 1993), then a
decision rule is required to evaluate each action level. The measurement performance
requirements, or MQOs, are also evaluated when developing a decision rule to ensure that an
acceptable measurement technique is available to support the proposed survey design.
Once the decision rule(s) have been established, a survey design is developed. The survey design
specifies the number and quality of measurements required to support a disposition decision
recorded in the decision rule. MARSAME recommends applying a graded approach to designing
disposition surveys (Section 4.4). The survey design, definitions of decision errors, and burden
of proof are determined by the selection of a null hypothesis (Section 4.2).
The survey design should be documented in a quality document (e.g., QA Survey Plan, SOP)
that has been reviewed and accepted by the appropriate authority (e.g., technical expert,
management, or regulator). Survey designs that are often repeated may be documented in SOPs
along with supporting records on instrument performance and personnel training. Other types of
disposition surveys are usually documented in a project-specific work plan and survey results are
presented in a disposition survey report (Sections 2.5 and 4.5). If the selected survey design is
not technically or economically practical, the planning team can investigate additional
disposition options if necessary (Sections 2.4 and 4.4).
1.4.2 Implementation Phase
To ensure flexibility and encourage the use of optimal measurement techniques for a specific
project, MARSAME does not provide detailed information on specific implementation
techniques. However, detailed descriptions of several measurement techniques are provided
NUREG-1575, Supp. 1 1-10 January 2009
-------
MARSAME Introduction and Overview
(Chapter 5 and Appendix D). These descriptions serve as a template for information required to
evaluate different measurement techniques. It is important to remember that the survey design is
usually linked to a specific option for disposition of the M&E (Chapters 3 and 4).
During implementation, the descriptions of measurement techniques are compared to the MQOs
defined during survey planning. A measurement method (i.e., combination of a measurement
technique with an instrument; see Section 5.9) is selected based on its ability to meet the MQOs.
The number and type of measurements specified in the documented survey design are performed
at the locations specified in the survey design. If a measurement method is specified in the
survey design, that method should generally be used during implementation. If the specified
measurement method cannot be performed (e.g., the instrument is unavailable or the
measurement method does not meet the MQOs), another measurement method should be
selected based on the MQOs. The selection of the replacement measurement method should be
documented in the survey design and survey report.
An action level may be established for implementing disposition surveys to support disposition
decisions about individual objects or measurement locations. If this action level is in the same
units as measurements performed in the field, then the surveyor can make final disposition
decisions by directly comparing the measurement results to the action level as the measurements
are performed. This may allow the surveyor to perform remediation as required and implement a
"clean as you go" component to the survey design (Section 6.9). Clean as you go surveys may
reduce the amount of M&E requiring additional consideration following completion of the
disposition survey. This clean as you go approach to surveys is only applicable for Class 1
surveys (i.e., radionuclide concentrations or radiation levels exceed the action level and 100% of
M&E are measured) where there is high confidence in the quality and accuracy of detection
decisions.
Quality control (QC) data are collected and analyzed during implementation to provide an
estimate of the uncertainty associated with the survey results. QC measurements are technical
activities performed to measure the attributes and performance of a survey. A well-designed QC
program increases efficiency and provides for early detection of problems. This can save time
and money by averting rework and enables the user to make decisions more expeditiously
(EPA 2002c).
1.4.3 Assessment Phase
The assessment phase begins with verification and validation of the survey results. Data
verification is used to ensure the requirements documented in the survey design were
implemented as prescribed. Data validation ensures the results of the data collection activities
support the objectives of the survey (i.e., DQOs), or permit a determination that these objectives
should be modified (MARSSIM Section 9.3 and MARSSIM Appendix N).
DQA determines if the collected data are of the right type, quality, and quantity to support their
intended use. DQA helps complete the data life cycle by providing the assessment needed to
determine that the planning objectives are achieved. DQA is described in detail in EPA QA/
G-9R (EPA 2006b), MARSSIM Section 8.2, and MARSSIM Appendix E.
January 2009 1-11 NUREG-1575, Supp. 1
-------
Introduction and Overview MARSAME
The preliminary data review is performed to learn about the structure of the data (e.g.,
identifying patterns, relationships, or potential anomalies). Graphical techniques are used to help
visualize the data. Calculation of basic statistical quantities is used to help describe the
distribution of data.
The survey data are evaluated using a statistical test. A test statistic is calculated and compared
to a critical value. The critical value divides the potential values of the test statistic into two
regions. The critical region includes values for the test statistic where the null hypothesis is
rejected. The null hypothesis is not rejected for values of the test statistic outside the critical
region. Keep in mind that a statistical test could be as simple as comparing survey results directly
to the critical value to ensure no radiation is detected, or may involve using a more complex
statistical evaluation such as the Wilcoxon Rank Sum (WRS) test.
In some cases the assessment phase may be performed during survey implementation. For
example, the "clean as you go" approach described in Section 6.9 requires that field data be
assessed and a decision made concerning the M&E being measured. The M&E are "cleaned," or
remediated, as necessary and another disposition survey performed.
1.4.4 Decision-Making Phase
Following the assessment phase, a decision is made regarding the disposition of the M&E. The
decision rule defines the final decision. The statistical test or data comparison determines
whether the parameter of interest exceeds the action level. Based on the outcome, a decision can
be made regarding the alternative actions. If multiple decision rules are defined for a single
disposition survey (e.g., a MARSSIM-type survey where the average activity is evaluated using a
statistical test and small areas of elevated activity are evaluated using the elevated measurement
comparison) any one decision that the action level has been exceeded should result in additional
investigation.
In some cases the decision making phase may be performed during survey implementation. For
example, the "clean as you go" approach described in Section 6.9 requires that field data be
assessed and a decision made concerning the M&E being measured. The M&E are "cleaned," or
remediated, as necessary and another disposition survey is performed.
1.5 Organization of MARSAME
The planning, implementation, and assessment of disposition surveys in MARSAME are based
on the data life cycle. Each chapter in MARSAME provides information for specific steps in the
process. The planning phase is discussed in Chapters 2, 3, and 4. The implementation phase is
discussed in Chapter 5, and Chapter 6 discusses the assessment phase and decision-making
phase.
Chapter 2 focuses on the IA. Information is provided on categorizing whether the M&E are
impacted or non-impacted using existing data and sentinel measurements in Section 2.2.
Information on designing and implementing preliminary surveys to provide the information
needed to design a disposition survey is provided in Section 2.3. Discussions on describing the
NUREG-1575, Supp. 1 1-12 January 2009
-------
MARSAME Introduction and Overview
M&E being surveyed are provided in Section 2.4. The selection of a disposition option and
development of a conceptual model are discussed in Section 2.5. Information pertaining to
documenting the results of the IA is provided in Section 2.6.
Chapter 3 provides information on developing a decision rule and discusses other inputs needed
to design a disposition survey. Section 3.2 addresses selecting the radionuclides or radiations of
concern which must be established before forming a decision rule. There are three parts to a
decision rule-
• Action level(s), discussed in Section 3.3,
• Parameter of interest, discussed in Section 3.4, and
• Alternative actions, discussed in Section 3.5.
Section 3.7 brings these three components together to develop decision rule(s) that are used to
design the disposition survey in Chapter 4. Survey units are discussed in Section 3.6, and inputs
for selecting measurement methods are presented in Section 3.8. Section 3.9 identifies reference
materials that can be used to estimate background radionuclide concentrations or radiation levels.
The process for evaluating an existing survey design is described in Section 3.10.
Chapter 4 completes the planning phase with the development of a survey design. This chapter
discusses the selection of a null hypothesis and setting tolerable limits on decision errors (Section
4.2), determines the level of survey effort for the disposition survey (Section 4.3), and
determines the type, number, and location of measurements to support a disposition decision
(Section 4.4). Information pertaining to disposition survey design documentation is provided in
Section 4.5. The processes in Chapter 4 result in a documented survey design.
The implementation processes in Chapter 5 focus on selection of an appropriate measurement
technique. Recommendations are provided on issues related to health and safety that may impact
the implementation of disposition surveys (Section 5.2). Chapter 5 also provides information on
process control and handling of potentially radioactive M&E (Section 5.3). The use of
segregation to help improve the efficiency of measurements and detectability of radioactivity,
and as a tool to limit the uncertainty is described in Section 5.4. Sections 5.5 through 5.8 discuss
the establishment of measurement uncertainty, measurement detectability, and measurement
quantifiability as MQOs to validate the measurement method's ability to meet the established
performance objectives. Information is provided on several measurement techniques (Section
5.9) that can be used for comparison to the MQOs developed in Chapter 3. These descriptions
can also be used during the planning phase to specify a measurement technique in the survey
design. Recommendations related to QC are also provided to ensure that survey instruments are
functioning properly, and the data meet defined performance limits specified during planning
(Section 5.10). Information related to collecting and documenting survey data is discussed in
Section 5.11.
Chapter 6 provides methods for the assessment and decision-making phases. Recommendations
are provided for performing the preliminary data review, calculating statistical quantities, and
preparing graphic representations that will assist the user in exploring the data (Section 6.2).
Disposition decisions about individual items may be based on individual measurement results by
January 2009 1-13 NUREG-1575, Supp. 1
-------
Introduction and Overview MARSAME
comparing data to the upper bound of the gray region (Section 6.3). Information is also provided
for calculating the upper confidence limit (Section 6.4). Details on performing recommended
statistical tests are also included (Sections 6.5 through 6.7). This chapter also describes how to
make a disposition decision based on the survey results (Section 6.8), what to do if the selected
disposition option is not accepted (Section 6.9), and the documentation to support the decision
(Section 6.10).
Chapter 7 discusses the general concepts of statistical survey design and hypothesis testing.
Detailed discussions and calculations of MQOs, measurement uncertainty, minimum detectable
concentrations (MDCs), and minimum quantifiable concentrations (MQCs) are provided in this
chapter. Details and examples of each topic are provided. A detailed example of a scan MDC
calculation is provided that is used to support the illustrative examples in Chapter 8.
Chapter 8 provides detailed illustrative examples implementing specific concepts found
throughout MARSAME. The illustrative examples cover a range of material, equipment,
radionuclides, and disposition options. Sections of these illustrative examples are used to
illustrate specific concepts throughout the supplement.
MARSAME contains several appendices to provide additional information on specific topics.
Appendix A provides copies of statistical tables needed to implement the information in
MARSAME. Appendix B lists sources of environmental radiation such as natural background
and fallout. A list of potential radionuclides grouped by industry or type of facility is provided in
Appendix C. Appendix D provides detailed information on specific measurement systems unique
to disposition surveys. Appendix E lists and describes some of the potential sources of action
levels applicable to decisions regarding disposition of M&E.
1.6 Similarities and Differences Between MARSSIM and MARSAME
During the 1990s, there was a concerted effort to improve the planning, implementation,
evaluation, and documentation of building surface and surface soil final radiological surveys for
demonstrating compliance with standards. This effort included the preparation of NUREG-1505
(NRC 1998a) and NUREG-1507 (NRC 1998b) by the NRC and culminated in 1997 with the
issuance of MARSSIM (MARSSIM 2002). MARSSIM was a joint effort by DOD, DOE, EPA,
and NRC to develop a multi-agency approach for planning, performing, and assessing the ability
of surveys to meet dose- or risk-based standards while at the same time encouraging effective
use of resources. MARSSIM provided recommendations for developing appropriate final status
survey designs using the DQO process to ensure survey results were of sufficient quality and
quantity to support a final decision. MARSSIM (MARSSIM 2002), NUREG-1505 (NRC 1998a),
and NUREG-1507 (NRC 1998b) replaced the previous approach for such surveys contained in
NUREG/CR-5849 (NRC 1992).
This MARSAME supplement expands the scope of MARSSIM methods and processes to
provide technical information supporting the disposition decision for M&E, specifically the
design and implementation of disposition surveys, to ensure the disposition decision is
technically defensible and optimized for efficiency. MARSSIM addressed the disposition of real
property (e.g., buildings and land) where the only disposition options were unrestricted release,
restricted release, or maintaining radiological controls. MARSAME addresses the disposition of
NUREG-1575, Supp. 1 1-14 January 2009
-------
MARSAME
Introduction and Overview
non-real property (e.g., M&E) and includes additional options for future use including recycle or
disposal as radioactive waste (see Section 2.5). Increasing radiological controls and interdiction
are also included as potential disposition options. While several, or all, disposition alternatives
may be acceptable for a specific project, optimizing the disposition survey design based on the
selected disposition alternative is described in MARSAME.
MARSAME as a supplement to MARS SIM expands the scope of technically sound
measurement processes and methods to include M&E. Table 1.2 summarizes the major
similarities between MARSSIM and MARSAME, which result from application of a graded
approach to support a technically defensible decision regarding disposition. Table 1.3
summarizes the major differences between MARSSIM and MARSAME, which result from the
change from real to non-real property.
Table 1.2 Similarities Between MARSSIM and MARSAME
Parameter
Graded Approach
Data Quality
Objectives (DQO)
Process
Data Quality
Assessment (DQA)
Process Knowledge
Classification
Flexibility
Statistics
Scale of Decision
Making
Inherent Radioactivity
MARSSIM
Used to place greater survey effort on
areas that have, or had, the highest
potential for residual radioactivity.
Used to design technically defensible
surveys to support decisions on
disposition of real property.
Used to evaluate survey results and
support a decision of whether to release
real property.
Used during the Historical Site
Assessment to support the
determination of whether an area is
impacted and provide information for
designing subsequent surveys.
Determines the level of survey effort
based on the potential amount of
residual radioactivity present.
MARSSIM allows and encourages
flexibility in the design and
implementation of final status surveys
for application to diverse site
conditions.
Used to develop a technically defensible
survey design.
A separate release decision is made for
every survey unit.
Inherent radioactivity is site-specific
and generally cannot be separated from
ambient radiation.
MARSAME
Used to place greater survey effort on
M&E that have, or had, the highest
potential for residual radioactivity.
Used to design technically defensible
surveys to support decisions on
disposition of non-real property (e.g.,
M&E).
Used to evaluate survey results and
support a disposition decision for non-
real property.
Used during the Initial Assessment to
support the determination of whether
M&E are impacted and provide
information for designing subsequent
surveys.
Determines the level of survey effort
based on the potential amount of
residual radioactivity present.
MARSAME allows and encourages
flexibility in the design and
implementation of disposition surveys
for application to diverse M&E.
Used to develop a technically
defensible survey design.
A separate release decision is made for
every survey unit.
Inherent radioactivity is specific to the
M&E being investigated. Segregation
of M&E based on inherent
radioactivity can be used to reduce
measurement variability.
January 2009
1-15
NUREG-1575, Supp. 1
-------
Introduction and Overview
MARSAME
Table 1.3 Differences Between MARSSIM and MARSAME
Parameter
MARSSIM
MARSAME
Scope
Surface soil and building surface
surveys (i.e., real property).
Materials and equipment (i.e., non-real
property).
Disposition Options
Restricted or unrestricted release, or
fail to release.
Release survey (maintain, remove, or
transfer radiological control; clearance
for reuse, recycling, or disposal),
or
Interdiction survey (increase in the level
of radiological control or a decision not
to accept control from another party).
Categorization
Included as part of classification in
MARSSIM.
Separates the decision to survey from
determining level of survey effort.
Application of the
Graded Approach
Classification and survey unit size
result in varying levels of survey
effort.
Multiple disposition options result in
varying levels of survey effort.
Sentinel Measurements
Not described in MARSSIM.
Allows use of sentinel measurements
during IA to check validity of certain
process knowledge assumptions.
Documentation of
Survey Designs
Assumes project-specific survey
designs will be developed for
individual sites.
In addition to project-specific survey
design, allows SOPs for categories of
M&E to provide standard approach to
disposition surveys.
Preliminary Surveys
Scoping and characterization surveys
regularly used to obtain information
needed to design a final status survey.
Scoping and characterization surveys
rarely used to obtain information needed
to design a disposition survey. Historical
information obtained during the IA is
generally sufficient to design a
disposition survey. If not, preliminary
surveys may be used to provide the
necessary information.
Ambient Radiation
Ambient radiation is site-specific and
generally cannot be separated from
inherent radioactivity.
Ambient radiation is selected based on
location where disposition surveys are
performed, and can be separated from
inherent radioactivity.
Interdiction
Not addressed in MARSSIM.
Surveys may be performed to identify
radioactive material resulting in an
increase in the level of radiological
control or deciding not to accept control
from another party. For example,
identifying radioactive materials and
initiating radiological controls, or
identifying unauthorized movement of
radioactive material.
NUREG-1575, Supp. 1
1-16
January 2009
-------
MARSAME
Introduction and Overview
Table 1.3 Differences Between MARSSIM and MARSAME (Continued)
Parameter
MARSSIM
MARSAME
Null Hypothesis
MARSSIM recommends using the
null hypothesis: "The activity in the
survey unit exceeds the action level
(Scenario A)."
MARSSIM allows using the null
hypothesis: "The activity in the
survey unit is indistinguishable from
background (Scenario B) with
information from NUREG-1505
(NRC 1998a)."
User selects the appropriate null
hypothesis:
"The activity in the survey unit exceeds
the action level (Scenario A)."
or
"The activity in the survey unit is
indistinguishable from background
(Scenario B)."
Scan-Only Surveys
Not addressed in MARSSIM
M&E may be dispositioned based on the
results of scan-only surveys provided
the scan measurements meet the MQOs
for the survey.
In Situ Surveys
Not addressed in MARSSIM
M&E may be dispositioned based on the
results of in situ surveys provided the in
situ measurements meet the MQOs for
the survey.
January 2009
1-17
NUREG-1575, Supp. 1
-------
-------
MARS AME Initial Assessment of Materials and Equipment
2 INITIAL ASSESSMENT OF MATERIALS AND EQUIPMENT
2.1 Introduction
The initial assessment (IA) is the first step in the investigation of materials and equipment
(M&E), similar to the historical site assessment (HSA) described in the Multi-Agency Radiation
Survey and Site Investigation Manual (MARSSIM 2002). The purpose of the IA is to collect and
evaluate information about the M&E in order to determine if it is impacted or non-impacted (i.e.,
categorization). During the IA process, additional information is collected to identify and support
potential disposition of impacted M&E (e.g., clearance, increased radiological controls,
remediation, or disposal). Project managers are encouraged to use the IA to evaluate M&E for
other hazards (e.g., lead, PCBs, asbestos) that could increase the complexity of the disposition
survey design or pose potential risks to workers during subsequent survey activities (Section 5.2)
or to human health or the environment following subsequent disposition of the M&E.
There are five major activities associated with the performance of the IA:
• Categorize the M&E as impacted or non-impacted based on visual inspection, historical
records, process knowledge, and results of sentinel measurements (Section 2.2).
• Design and implement preliminary surveys to adequately describe the M&E and address data
gaps based on a preliminary description of the M&E (Section 2.3).
• Describe the physical and radiological attributes of the M&E (Section 2.4).
• Select appropriate disposition option(s) and define alternative actions applicable to impacted
M&E (Section 2.5).
• Document the results of the IA through the use of a standard operating procedure (SOP) or
development of a conceptual model (Section 2.6).
For M&E that have been categorized as impacted, an existing survey design in the form of an
SOP may be available for investigating the radiological status of the M&E. If an applicable SOP
is available, the instructions in the SOP should be followed for implementing and assessing the
results of the survey. The information on performing preliminary surveys (Section 2.3) can be
used to determine whether an SOP is applicable to the M&E being investigated. The information
on describing the M&E (Section 2.4) can be used to determine if preliminary surveys are
necessary. The information on selecting a disposition option (Section 2.5) and documenting the
results of the IA (Section 2.6) can be used for project-specific applications, or for developing a
new SOP.
2.2 Categorize the M&E as Impacted or Non-Impacted
The first decision made when investigating M&E is whether they are impacted or non-impacted.
M&E with no reasonable potential for containing radioactivity in excess of natural background,
fallout levels, or inherent levels of radioactivity are non-impacted. Impacted M&E have a
reasonable potential to contain radionuclide concentration(s) or radioactivity above background.
The decision of whether M&E are impacted or non-impacted is primarily based on existing
information. Figure 2.1 depicts the categorization process.
January 2009 2-1 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment
MARSAME
Review Existing
Relevant Information
-Yes
Is the
Existing Information
dequate to Categorize
the M&E?
1
1
Perform a
Visual Inspection
(Section 2.2.1)
Review
Historical Records
(Section 2.2.2)
i
r
Assess Process
Knowledge
(Section 2.2.3)
r
Are Sentinel
Measurements
Applicable?
Yes-
Plan and Perform
Sentinel
Measurements
(Section 2.2.4)
Documentation
of the Non-Impacted
Decision
Necessary?
Yes-
Proceed to
Preliminary Survey
(Figure 2.2)
NOTE: Shaded diamonds
represent important decision
points.
Yes-
Prepare Documentation of
the Non-Impacted Decision
(Section 2.2.5)
No Further Action
Figure 2.1 The Categorization Process as Part of Initial Assessment
NUREG-1575, Supp. 1
2-2
January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
If adequate information is readily available to support a categorization decision, the decision
maker should decide if the M&E are impacted or non-impacted. A complex single unit or group
of M&E may be divided into portions that are impacted and portions that are non-impacted. This
is illustrated in the front loader example described in Section 8.3, where the bucket and tires may
be impacted while the engine and cab interior are non-impacted. If additional information is
required to support the categorization decision, visual inspection (Section 2.2.1), collection and
review of historical records (Section 2.2.2), and assessment of process knowledge (Section 2.2.3)
are the most common sources of additional existing information. Assumptions may be made
regarding the use and interpretation of existing information. Data collection activities may be
performed during the IA to specifically address questions about these assumptions. These data
collection activities are called sentinel measurements and are discussed in Section 2.2.4.
Additional investigation is required to make technically defensible disposition decisions
regarding impacted M&E. All impacted M&E must receive some level of additional
investigation, even if the expected disposition is disposal as radioactive waste. For example,
M&E shipped for disposal as radioactive waste must meet waste acceptance criteria at the
disposal facility as well as Department of Transportation (DOT) requirements for transporting
radioactive material. The results of any additional investigation must clearly demonstrate
compliance with any applicable requirements, and be appropriately documented. Non-impacted
M&E do not receive any additional radiological investigation.
2.2.1 Perform a Visual Inspection
The purpose of the visual inspection is to identify and document the physical characteristics of
the M&E (e.g., size, kind of material, shape, condition) when this description is not readily
available to support a categorization decision. The visual inspection may be performed during a
site visit, or by reviewing photographs or videos of the M&E. Photographs and video also
provide a means for documenting the results of the visual inspection. The visual inspection
corresponds to the Site Reconnaissance presented in Section 3.5 of MARS SIM. Information will
be used to support the following activities:
• Developing survey unit boundaries (Section 3.6).
• Defining the parameter of interest during the development of a decision rule for impacted
M&E (Section 3.4).
• Verifying the requirements of an SOP are met before performing a routine survey (Section
4.5.1).
• Evaluating any health and safety concerns (Section 5.2).
• Developing handling protocols for implementation of the disposition survey (Section 5.3 and
5.4).
Prior to performing a visual inspection, the surveyor should review what is known about the
M&E. If little or no information is available describing potential hazards associated with the
M&E, care should be exercised in performing a visual inspection. Screening measurements for
radiation, chemical, and other hazards, along with the use of personal protective equipment (e.g.,
gloves, coveralls, respirators), may be necessary depending on available information. Situations
with known or expected risks (i.e., M&E that are radiologically or chemically impacted) may
January 2009 2-3 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
require preparation of a study plan or SOP anticipating activities to be performed and identifying
specific information to be collected. Casual visual inspections of M&E with an unknown history
are not recommended. Detailed visual inspections (e.g., disassembly of potentially impacted
equipment to examine interior surfaces) should not be performed without proper precautions and
are more appropriately performed by preliminary surveys (Section 2.3).
While the primary objective for performing a visual inspection is to collect information used to
design a disposition survey, the information can be used for other purposes. Evaluation of health
and safety concerns (Section 5.2) and development of handling protocols for implementation of
the disposition survey (Section 5.3) are two examples where visual inspection information would
be used.
2.2.2 Collect and Review Additional Historical Records
When information on the identity, concentration, and distribution of radioactivity are not readily
available to support a categorization decision, historical records may provide this specific
information. Information on the physical characteristics of the M&E (e.g., size, shape, condition)
and the characteristics of the radioactivity (e.g., radionuclides of concern, expected
concentrations) will be used to select a disposition option in Section 2.5 and describe initial
survey unit boundaries in Section 3.6.1. The historical information is then used to define the
action level, parameter of interest, and alternative actions during the development of a decision
rule for impacted M&E (Section 3.7, EPA 2006a).
Types of historical records that provide useful information are described in MARSSIM Section
3.4.1, and may include—
• A facility or site radioactive materials license;
• Permits or other documents that authorize use of radioactive materials;
• Other permits and environmental program files;
• Operating records (e.g., previous surveys, waste disposal records, effluent releases);
• Corporate contract files (e.g., purchasing records, shipping records);
• A site or facility description (e.g., locations of M&E, site photographs); and
• Inspection reports, incident analyses, and compliance histories maintained by currently and
formerly involved regulatory agencies.
Another source of historical information is interviews with current or previous employees.
Interviews may be conducted early in the data collecting process or close to the end of the IA.
Interviews conducted early in the IA cover general topics, and information gathered is used to
guide subsequent data collection activities. Interviews conducted late in the IA allow the
investigator to direct the investigation to specific areas that require additional information or
clarification.
Once the historical records have been collected, they should be reviewed to identify information
that supports the categorization decision. Historical information used to support the
categorization decision should be evaluated using the data quality assessment (DQA) process
(EPA 2006b). In particular, historical information should be examined carefully because—
NUREG-1575, Supp. 1 2-4 January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
• Previous data collection efforts may not be compatible with IA objectives,
• Previous data collection efforts may not be extensive enough to fully describe the M&E
being investigated,
• Measurement techniques or protocols may not be known or compatible with IA objectives, or
• Conditions may have changed since the data were collected
Additional information on evaluating data can be found in the following documents-
• The Environmental Survey Manual Appendix A - Criteria for Data Evaluation (DOE 1987)
• Upgrading Environmental Radiation Data, Health Physics Committee Report HPSR-1 (EPA
1980)
• Guidance for Data Usability in Risk Assessment, Part A (EPA 1992a)
• Guidance for Data Usability in Risk Assessment, Part B (EPA 1992b)
Historical records describing impacted M&E may include additional information that can be
used to support additional activities during the disposition process. For example, historical
records may provide descriptions of the M&E that are sufficient to design a disposition survey
(Chapter 4). On the other hand, the historical records can be used to identify data gaps that are
addressed by performing preliminary surveys (Section 2.3).
2.2.3 Assess Process Knowledge
The characteristics, history of prior use, and inherent radioactivity are critical for evaluating the
impacted status of M&E. This information is termed process knowledge. Process knowledge is
obtained through a review of the operations conducted in facilities or areas where M&E may
have been located and the processes where M&E were involved when this information is not
readily available to support a categorization decision. This information is used to evaluate
whether M&E—such as structural steel, ventilation ductwork, or process piping—had been in
direct contact with radioactive materials or had been activated, which would lead to a decision
the M&E are impacted. Descriptions of the physical attributes of the M&E (Section 2.4.1) and
radiological attributes of the M&E (Section 2.4.2) can be obtained from process knowledge. In
addition, process knowledge supports the selection of a disposition option (Section 2.5). The
disposition option is then used to identify sources of action levels, a parameter of interest, and
alternative actions during the development of a decision rule for impacted M&E (Section 3.7 of
this supplement and EPA 2006a).
Process knowledge is obtained by researching the M&E and understanding the origin, use, and
potential disposition. The level of detail required from process knowledge is project specific. The
description of M&E could be simple, such as a set of hand tools being removed from a controlled
area where the radiological conditions are well known. At the other extreme is a complex
situation that requires knowledge of the manufacturing process, investigations of multiple
processes that could impact the radiological conditions associated with the M&E, and
understanding of recycle and reuse options that include movement of radionuclides through the
environment. Sections 2.4.1 and 2.4.2 describe types of information that may be obtained from
process knowledge and are necessary to support the development of a disposition survey.
January 2009 2-5 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
In some cases, process knowledge of the equipment being investigated can be used to support
categorization decisions. Consider a pump used to circulate demineralized make-up water.
Maintenance records do not show the presence of radioactivity and operating records indicate no
events where the pump could have been used with radioactivity. Radiological samples of the
demineralized make-up water do not show the presence of radioactivity. Based on this process
knowledge, the interior of the pump is categorized as non-impacted.
Historical records (Section 2.2.2) are one source of process knowledge. Historical records,
including interviews, provide site- and project-specific information on historical use and
radiological processes that may affect the M&E. Engineering and chemistry books and journals
provide information on the origins (e.g., manufacturing) and potential disposition of the M&E.
Industry documents and company records are also potential sources of process knowledge. Other
sources of information on M&E should be considered during the IA, indicating how, where, and
when the M&E were used in areas where they potentially could have been affected by
radionuclides or activation. These sources of information include—
• Purchasing records showing when M&E were obtained,
• Maintenance records showing where and how they were used,
• Operating logs for systems that utilized or could have affected the M&E,
• Disposal records showing survey results for similar types of M&E indicating types, and
Locations of radionuclides or radioactivity.
In some instances, process knowledge may not be available for the M&E being considered for
release. For example, consider an outdoor material staging area for a nuclear facility where
various pieces of surplus equipment and metal have accumulated over the years. The origin of
these M&E is unknown. In this case, it is particularly important that preliminary surveys be
performed on the M&E to determine if excess radioactivity is present and to finalize the list of
radionuclides of concern.
Techniques used to protect equipment or prevent radioactivity from entering difficult-to-measure
areas or penetrating porous surfaces can be used to support categorization decisions. Consider
the following examples of protection and prevention techniques:
• Plan and coordinate all work to minimize exposure of equipment, tools, and vehicles to
radioactivity.
• Evaluate materials, tools, and equipment for ease of decontamination and disassembly (that
may be required for decontamination or release) prior to use.
• Use prefilters or have a separate source of outside air on the intake for internal combustion
equipment subject to airborne radionuclides or radioactivity.
• Use a filtered inlet for high volume air handling equipment such as blowers, compressors,
etc., to minimize the potential for internal contamination due to build up of low-level
radioactivity.
• Do not bring electrically driven mobile equipment into controlled areas.
• Use protective sheathing/covers, strippable coatings, or protective caps to minimize the
potential for surficial radionuclides or radioactivity.
NUREG-1575, Supp. 1 2-6 January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
• Cover and protect all openings on equipment, tools, or vehicles that may permit radioactivity
to enter difficult-to-access or difficult-to-clean areas.
• Select technologies that minimize radiological airborne emissions, secondary wastes, and
tool or equipment damage.
2.2.4 Perform Sentinel Measurements
Sentinel measurements are biased measurements performed at key locations to provide
information specific to the objectives of the IA. The objective of performing sentinel
measurements as part of the IA is to gather sufficient information to support a decision regarding
further action (e.g., categorization). Sentinel measurements may also be used to verify
assumptions based on existing information or obtain information on the current status of the
M&E. Sentinel measurements are not a risk assessment, scoping survey, or study of the full
extent of radionuclides or radioactivity associated with the M&E.
Sentinel measurements alone cannot be used to show that M&E are non-impacted. Positive
results are definitive for determining that M&E are impacted. However, negative results provide
only part of the evidence required for determining that the M&E are non-impacted. Since
radioactivity in difficult-to-measure areas cannot be measured directly without accessing the area
(e.g., disassembling equipment), sentinel measurements performed at access points to difficult-
to-measure areas could be used to indicate that it is unlikely that radioactivity entered that area.
For example, smears with elevated radioactivity, collected inside ductwork, can provide
information to support categorization of the ventilation system as impacted. Because sentinel
measurements are usually associated with difficult-to-measure areas, they are not generally
applicable to dispersible bulk materials.
If protection and prevention techniques (described in Section 2.2.3) were applied to M&E used
around radioactive material, sentinel measurements can be used in connection with process
knowledge to support a decision of whether diffi cult-to-measure areas were impacted. For
example, if prefilters are used to capture paniculate airborne radioactivity of a specific size
before the particulates enter difficult-to-measure areas, sentinel measurements can be made on
the prefilters.
Sentinel measurement methods may involve any of the measurement techniques discussed in
Section 5.9.1 combined with the instruments discussed in Section 5.9.2. Advantages and
disadvantages of different combinations of measurement techniques and instrumentation are
listed in Table 5.5 and discussed in Section 5.9.3. The selection of a measurement method for
sentinel measurements should be made based on project-specific considerations using the DQO
process.
It should be noted that access points are often modified to limit personnel radiation exposure to
difficult-to-measure areas after use (e.g., capped, sealed, cleaned). Care should be taken to avoid
performing sentinel measurements at modified access points to reduce the probability of making
an incorrect decision about the status of the M&E. QA and QC should be considered during
planning for collection of sentinel measurements. The measurement and subsequent evaluation
of the results should be consistent with the assumptions used to define sentinel measurements.
January 2009 2-7 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
2.2.5 Decide Whether M&E are Impacted
Once there is adequate information to support a categorization decision, the decision maker
needs to decide whether the M&E are impacted or non-impacted. The categorization decision is
based on four sources of information: visual inspection, historical records review, process
knowledge, and the results of sentinel measurements.l If the results for any part of the
categorization process indicate a reasonable potential for radionuclide concentrations or
radioactivity above background, the decision is the M&E are impacted. For example, if the
visual inspection, historical records, and process knowledge all indicate the M&E are non-
impacted but the sentinel measurements indicate impacted, the M&E are impacted. Similarly, if
the visual inspection and sentinel measurements indicate the M&E are non-impacted but the
historical records and process knowledge indicate the M&E are impacted, the M&E are
impacted. An important point is that sentinel measurements alone cannot be used to support a
decision in declaring M&E as non-impacted.
In most cases, the categorization decision is obvious based on the available information. In cases
where the decision is not obvious, the consequences of making a decision error usually result in a
determination that the M&E are impacted. For example, the consequence of incorrectly
categorizing M&E as impacted when they are not impacted includes performing a radiological
survey. However, the consequence of incorrectly categorizing M&E as non-impacted when they
are impacted could result in inadvertent exposure for members of the public, lack of confidence
in other radiological decisions, and potential violation of regulatory requirements. The
consequences of incorrectly categorizing M&E are also discussed in Section 4.3.4.
Collectively, this information should be used to develop survey strategies targeting different
types of materials in recognition that a single survey method or procedure may not necessarily fit
the technical requirements of all materials, given their diverse properties. For example, one
procedure may be used to address only the routine releases of tools and equipment. On the other
hand, a separate procedure may be developed to address infrequent releases of large amounts of
bulk materials, such as concrete rubble. The approach suggested here is one of compartmentali-
zing the release activities into manageable and common functional elements with each one being
optimized in the context of facility operations as to its effectiveness, while demonstrating
compliance with applicable regulations. The development of standardized survey procedures for
infrequent releases necessitates that the MARSAME user utilize processes in the remainder of
this chapter and then move to Section 3.10 for evaluating and implementing standard operating
procedures (SOPs).
If there is insufficient information available to design a disposition survey following
categorization, preliminary surveys may be performed to obtain additional information
describing the physical and radiological characteristics of the M&E (Section 2.4). These
preliminary surveys facilitate the development of an effective and efficient disposition survey
design.
If there are questions concerning the level of documentation for the categorization decision,
consult the cognizant regulatory authority. The decision maker should consider the degree to
1 Sentinel measurements are not required to support a categorization decision.
NUREG-1575, Supp. 1 2-8 January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
which documentation of the M&E categorization decision is necessary for M&E that are
categorized as non-impacted, since no additional investigation is required. In most cases it is not
necessary to document decisions that M&E are impacted since this decision will be documented
later in the disposition process (e.g., documentation of the IA results in Section 2.6,
documentation of the survey design in Section 4.5, and documentation of the disposition survey
results in Section 6.6).
2.3 Design and Implement Preliminary Surveys
If there is insufficient information available to design a disposition survey following
categorization, it may be necessary to perform preliminary surveys to obtain the required
information. Preliminary surveys of M&E correspond to scoping and characterization surveys
described in MARS SIM Sections 5.2 and 5.3.
Following a decision that the M&E being investigated are impacted, the decision maker should
determine if an applicable standardized survey design is available, usually in the form of an SOP.
If an SOP is available and applicable to the M&E being investigated, the instructions in the SOP
should be implemented and the results of the survey evaluated as specified in the SOP (see
Figure 2.2 and Section 2.6.1).
It may be necessary to evaluate the quantity and quality of data describing the M&E to determine
if the existing data are adequate for implementing an existing SOP or developing a disposition
survey design. If the data are adequate, no additional data collection is required. On the other
hand, if there are data gaps that need to be addressed prior to completing a disposition survey
design, preliminary surveys can be used to obtain the necessary data.
The purpose of performing preliminary surveys is to obtain information describing the physical
and radiological characteristics of the M&E. The ultimate goal is to minimize heterogeneity in
the subset of M&E being surveyed. Minimizing heterogeneity helps to control the measurement
uncertainties (Section 5.6), and may be helpful in selecting a disposition option (Section 2.5). For
example, if a subset of the M&E is identified as diffi cult-to-measure while the majority of the
M&E is relatively easy to measure and is considered for release, minimizing heterogeneity of all
the M&E by segregating the difficult-to-measure subset for potential disposal may simplify
measurements and be cost-effective. See Section 5.4 for information on segregation of M&E to
minimize heterogeneity during implementation of the disposition survey design.
In general, preliminary surveys are designed using professional judgment to address specific
questions concerning the existing data. Once a data gap has been identified, a survey is designed
and implemented to obtain the information required to fill that data gap. The results of the survey
are evaluated to ensure the data gap has been adequately addressed and the results are
documented. In some cases these surveys will be large and complicated, with written survey
designs reviewed by stakeholders prior to implementation. In other cases, these will be simple
surveys that quickly provide some small piece of information required to proceed with the
disposition survey design. By necessity, there is no single approach that will address all types of
preliminary surveys. However, the DQO process can be applied to successfully design a
preliminary survey (EPA 2006a).
January 2009 2-9 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment
MARSAME
From Figure 2.1
Is the Existing
Information Adequate to Select a
Disposition Option?
Select a Disposition Option
(Section 2.5)
Is the Existing
Information Adequate
to Design a Disposition
Survey?
Identify Data Gaps
(Section 2.3)
Yes-
Proceed to
Figure 2.3
Design and Implement
Preliminary Surveys
(Section 2.3)
Describe the M&E
(Section 2.4)
NOTE: Shaded diamonds
represent important decision
points.
Figure 2.2 Assessing Adequacy of Information for Designing Disposition Surveys
NUREG-1575, Supp. 1
2-10
January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
The first step in designing a preliminary survey is to identify the data gaps to be addressed.
Section 2.4.1 and Section 2.4.2 discuss the minimum information required to describe the M&E
and design a disposition survey. Any of the required information that is not available or is not of
sufficient quality represents a data gap. In addition, there may be project-specific information
needed to complete the disposition survey design that could also represent potential data gaps. In
order to complete the list of potential data gaps, it is recommended that the planning team work
through the entire disposition survey planning process (Chapters 3 and 4). Whenever a data gap
is identified, the planning team should make reasonably conservative assumptions or proceed
with multiple survey designs based on a reasonable range of values to fill the data gap.
Identifying a complete list of data gaps will help ensure the necessary additional information can
be collected effectively and efficiently, with minimal waste of limited resources. If a separate
preliminary survey is designed and implemented for every data gap as it is identified, there is an
increased possibility of duplication of effort and increased demands on limited resources. As
with all data collection activities, QA and QC should be considered during planning and
evaluated during assessment of the results.
MARSAME uses an iterative planning process for designing surveys. Changes in the available
information may result in multiple iterations of individual steps. Iteration may be necessary at
any time that an assumption used to design a survey is shown to be false. For example, if a
historical record is found that changes the description of the M&E from beta-gamma emitting
radionuclides to include alpha emitting radionuclides, it is necessary to consider additional or
different measurement techniques to account for the alpha radiation.
2.4 Describe the M&E
The M&E being investigated must be described with regards to its physical and radiological
attributes in order to establish the information necessary to design a survey approach that can
adequately measure the M&E. This description is intended to ensure that residual radioactivity
associated with the M&E will not be missed by the disposition survey, the M&E is left in a
usable condition, and that any data collected meet the objectives of the disposition survey.
2.4.1 Describe the Physical Attributes of the M&E
A description of the physical characteristics defining the investigated M&E is required to help
the user develop a disposition survey design. The preliminary physical description is usually
developed using some combination of the techniques presented in Section 2.2 (i.e., visual
inspection, historical records, and process knowledge). The physical description of the M&E is
used to help define survey unit boundaries (Section 3.6.1) and develop a decision rule (Section
3.7), which has a direct impact on the disposition survey design.
Table 2.1 lists the four attributes that should be addressed when describing the physical
characteristics of the M&E being investigated (dimensions, complexity, accessibility, and
inherent value). Questions related to the evaluation of the attributes are provided, along with a
list of minimum information expected to be provided by the IA. The planning team should
consider designing and implementing preliminary surveys (Section 2.3) to verify existing
information and investigate data gaps identified during the initial steps of the IA.2
2 The development of a planning team is discussed in MARSSIM Section 3.2.
January 2009 2-11 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment
MARSAME
Table 2.1 Physical Attributes Used to Describe M&E
Attribute
Minimum Information
Questions for Consideration
Dimensions
Size (Total Mass)
Shape (Total Surface Area)
Are there issues with size and shape that affect
how the M&E should be handled?
Complexity
M&E may require segregation to
design a technically defensible
disposition survey.
M&E may be combined into
similar groups and still allow a
technically defensible disposition
survey.
Are there situations where segregation (e.g.,
disassembly) could affect the usefulness of the
M&E?
Are there situations where segregation (e.g.,
disassembly) could result in the release of
radioactivity or hazardous chemicals to non-
impacted areas?
Are there situations where engineering controls are
required to prevent the release of radioactivity or
hazardous chemicals to non-impacted areas?
Are there component materials that are inherently
radioactive or regulated for their chemical
properties?3
Are there multiple component materials in the
M&E?
Accessibility
Identification of impacted,
difficult-to-measure areas for
performing conventional handheld
measurements.
Known or potential relationships
among radionuclide
concentrations or radioactivity in
accessible and difficult-to-
measure areas.
Are there issues with size or shape that limit
accessibility (e.g., bottom of a large, bulky
object)?
Are there porous surfaces that could allow
permeation of radioactivity?
Are there seams, ruptures, or corroded areas where
radioactivity could penetrate to difficult-to-
measure areas?
Inherent
Value
The inherent value of the M&E
being investigated.
Can the M&E be reused or recycled?
Can the M&E be repaired or remediated?
What are the replacement and disposal costs?
2.4.1.1 Describe the Physical Dimensions of the M&E
It is important to understand the dimensions of the M&E being investigated in order to define the
scale of decision making (Section 3.6 on identifying survey unit boundaries), support evaluation
of measurement techniques (Sections 3.8 and 5.9), and identify any handling issues that may
need to be addressed (Section 5.3). The dimensions generally are defined as the size and shape of
the M&E being investigated. The size is primarily related to the scale of decision-making and
may be defined as the length, width, and depth of an item, or as the quantity of M&E. Quantity
may be expressed in terms of a number (e.g., 25 pumps) or a volume (e.g., 200 cubic yards of
3 For example, materials regulated under the Resource Conservation and Recovery Act (40 CFR 261) or the Toxic
Substances Control Act (40 CFR 700-766).
NUREG-1575, Supp. 1
2-12
January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
concrete rubble), and may be related to the mass of the M&E. An estimate of the total mass of
the M&E should be provided. The shape of the M&E is primarily related to the evaluation of
measurement techniques. The description of shape should consider surface conditions (e.g., clean
or dirty, rough or smooth, curved or flat) that affect the surface efficiency for radiation
instruments. An estimate of the total surface area of the M&E should be provided when the
radionuclides of concern are, or could be, surficial.
2.4.1.2 Describe the Complexity of the M&E
The complexity of the M&E also affects the disposition survey design. Complexity refers to the
number and types of components that make up the M&E, as well as the ability to segregate or
combine the M&E into similar groups. M&E consisting of a single component is a simple case.
Consider the situation where several hundred feet of pipe are being investigated and the entire
pipe is made from steel.
A complex situation occurs when the M&E consist of a variety of component materials.
Consider the same amount of pipe, but some pipe is steel, some is copper, and some is lined with
rubber, lead, or PVC. Some types of process equipment (e.g., pipe originating from mineral
processing industries) are internally lined with rubber, lead, or PVC. The presence of such liners
can complicate the initial categorization, as well as subsequent characterization and survey of
such equipment. The presence of lead can complicate the final disposition of process equipment
(e.g., recycling as ferrous steel or disposal in landfills).
Equipment once used in process plants or systems should be checked for the presence of
internally deposited sediment, sludge, oil, grease, water, and presence of process chemicals and
reagents. The presence of such residues may require the implementation of special worker health
and safety measures, procedures to collect and properly dispose of such hazardous material, and
may restrict possible disposition options.
Complexity also comes from the ability to break down or combine the M&E into similar groups.
A steel I-beam represents a simple case, where there is one material that can be cut into the
desired lengths. Dispersible bulk materials represent a situation that is slightly more complex,
especially when different types of materials have been combined. One example is a pile of scrap
metal, where the metal can be segregated by material (e.g., aluminum versus steel) or type (e.g.,
sheet metal versus pipe versus I-beams).
Equipment tends to be more complex, because it often contains a variety of components that can
generally be broken down by disassembling the equipment. Consider the case of a power tool
consisting of a casing, an electric motor, and controls. There are different types of metal, plastic,
and possibly glass or ceramics that make up the item, but disassembly into the individual
components may render the tool unusable and may expose component materials that are
inherently radioactive or hazardous. Disassembly of certain items could also result in the release
of radioactivity or hazardous chemicals to non-impacted areas, and may require engineering
controls to prevent such releases. The disposition survey design often increases in complexity as
the equipment increases in size and complexity. However, complex M&E may also allow the
user to segregate impacted from non-impacted items or components. This segregation may
January 2009 2-13 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
reduce the amount of M&E requiring additional investigation. One example is a front loader
used to move piles of potentially radioactive material at a decommissioning or cleanup site. The
bucket and tires of the front loader may be identified as impacted while the engine and cab are
identified as non-impacted, depending on the controls in place while the equipment was being
used. However, there may be cases where an adequate survey design cannot be developed based
on decisions made earlier in the planning process. In these cases, it may be necessary to revisit
some of the decisions made earlier, for example, re-evaluating the cost to benefit analysis.
2.4.1.3 Describe the Accessibility of the M&E
Accessibility is the next attribute to consider when describing the M&E being investigated.
Accessibility has a direct impact on measurability, so it is a critical issue for making technically
defensible disposition decisions. Areas (including surfaces and individual items) are either
accessible or difficult-to-measure. Accessible areas are areas where radioactivity can be
measured, and the results of the measurement meet the DQOs and measurement quality
objectives (MQOs) defined for the survey. During the IA it is necessary to distinguish areas that
are accessible from areas that may be difficult to measure.
The determination of whether an area is physically accessible, for purposes of the IA, should be
based on whether a measurement could be performed using a conventional hand-held radiation
instrument such as a sodium iodide (NaI[Tl]) detector, or Geiger-Mueller (GM) pancake probe.
If difficult-to-measure areas are identified and these areas are categorized as impacted, the IA
should attempt to identify if there are any known or potential relationships among radionuclide
concentrations or radioactivity in accessible areas and radionuclide concentrations or
radioactivity in difficult-to-measure areas. This information will be evaluated in Section 3.3.3 for
the potential to use surrogate measurements as a method of estimating radionuclide
concentrations or radioactivity in difficult-to-measure areas.
The potential for permeation and penetration of radioactivity should also be discussed as part of
accessibility. Permeation describes the spread of radioactivity throughout a material and is
usually associated with porous materials or surfaces (e.g., wood, concrete, unglazed ceramic).
Certain chemical and physical forms can increase the permeation rate (e.g., liquids permeate
faster than solids; small particles permeate faster than large particles). Penetration describes
infiltrating into difficult-to-measure areas, and is generally associated with radioactivity entering
through access points, seams, or ruptures. Corrosion of surfaces may also result in penetration of
radioactivity into difficult-to-measure areas.
2.4.1.4 Describe the Inherent Value of the M&E
A part of describing M&E that is often overlooked during the IA is determining the inherent
value of the materials or equipment being considered for release. Estimates of the value of
materials and equipment should include the replacement cost, condition (i.e., can the materials or
equipment be reused or recycled), and disposal cost. Replacement costs may consider increased
productivity due to upgrades to existing facilities and equipment, decontamination costs for
existing and new items, and the ultimate disposal of the replacements. Condition of the materials
and equipment may include maintenance and repair costs to start or keep the items operational,
NUREG-1575, Supp. 1 2-14 January 2009
-------
MARSAME
Initial Assessment of Materials and Equipment
as well as costs to decontaminate and release the items from radiological controls. Disposal costs
may include shipping and handling of potentially hazardous material. The limited capacity of
existing radiological waste disposal facilities may need to be considered along with the monetary
cost of disposal.
2.4.2 Describe the Radiological Attributes of the M&E
A description of the radioactivity potentially associated with M&E being investigated is required
to design a disposition survey. The review of historical documents (Section 2.2.2) and process
knowledge (Section 2.2.3) are the primary sources of information on radioactivity associated
with M&E. Sentinel measurements (Section 2.2.4) and preliminary surveys (Section 2.3) may
also provide information, such as types of radiations and identity of radionuclides. The
information describing the radioactivity is used to support a decision of whether the M&E are
impacted and supports the development of a disposition survey for impacted M&E. The
description of the radioactivity is divided into four attributes: radionuclides, activity, distribution,
and location.
Table 2.2 lists the four attributes to be addressed when describing radioactivity potentially
associated with the M&E being investigated. Questions related to the evaluation of the attributes
are provided, along with a list of minimum information expected to be provided by the IA. The
planning team should consider designing and implementing preliminary surveys (Section 2.3) to
obtain information that is not provided by the IA.
Table 2.2 Radiological Attributes Used to Describe M&E
Attribute
Minimum Information
Questions for Consideration
Radionuclides
List of radionuclides of potential concern,
including major radiations and energies.
What were the potential sources and
mechanisms for the radioactivity to
come into contact with the M&E?
Activity
List of expected radionuclide concentrations or
radioactivity (e.g., average, range, variance)
associated with the M&E
List of known and potential relationships among
radionuclide activities (e.g., activation and
corrosion products, fission products, natural
decay series).
What is the basis for the expected
radionuclide concentrations or
radioactivity?
What is the basis for the known and
potential relationships (e.g., process
knowledge of similar sources,
measurements of equilibrium
conditions)?
Distribution
List of areas where the radioactivity is uniformly
distributed.
List of areas where the distribution of
radioactivity is clustered.
List of areas where the distribution is unknown.
Can the M&E be divided into sections
where the distribution of radioactivity
is uniform?
Are there areas where small areas of
elevated activity are a concern?
Location
State whether the radioactivity is surficial,
volumetric, or a combination of both.
State whether surficial radioactivity is fixed or
removable.
Is the volumetric activity uniformly
distributed, is there a gradient, or is the
activity random or clustered?
January 2009
2-15
NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
2.4.2.1 Identify the Radionuclides of Potential Concern
Identification of the radionuclides of potential concern is a critical step in making disposition
decisions. At a minimum, the planning team should review the information available from
Section 2.2 to identify the radionuclides of potential concern. The quality and completeness of
the existing information should be evaluated. Information on known or expected relationships
among radionuclides of potential concern should be identified and evaluated for applicability to
current conditions. If necessary, a study to identify a complete list of radionuclides of potential
concern and determine relationships among radionuclides may be initiated before designing the
disposition survey.
A list of radionuclides of potential concern should be developed based on existing data. The list
should consider all potential sources of radioactivity, but only include radionuclides that are
actually of concern for the M&E being investigated.
The list is designed to help focus the disposition decision. The list of radionuclides of potential
concern should include the major types of radiation (e.g., alpha, beta, photon) and their
corresponding energies. A discussion of the sources of radionuclides of potential concern, and
their chemical and physical form should also be included, if possible.
Include a description of how the M&E became impacted if it is known. For example, it is
important to document whether the potential radioactivity resulted from deposition of airborne
particulate material, or from placing the M&E in an area of neutron flux that resulted in
activation. All potential mechanisms for radioactivity that is associated with the M&E should be
described.
The description of potential radioactivity from the IA may also identify known or suspected
relationships among radionuclides (e.g., equilibrium conditions for natural decay series, relative
activities of fission products or activation products based on process knowledge). Additional
investigations (e.g., preliminary surveys) may be performed to verify the presence of
radionuclides of potential concern and provide estimates of the activity relationships among
radionuclides. These investigations may include field measurements and sample collection with
laboratory analysis.
The identification of radionuclides of potential concern may impact other decisions made during
development of a disposition survey design. Since the sources of action levels are radionuclide or
radiation-specific, the identification of radionuclides of potential concern directly affects the
selection of an appropriate action level. The planning team should consider the impact of the list
of radionuclides of potential concern on other decisions (e.g., selection of measurement
techniques or instruments) as well as the impact of other decisions on the action levels when
considering potential sources of action levels. For example, the identification of available
measurement techniques (Section 3.8) is also directly related to the radionuclides of potential
concern. The determination of surficial or volumetric radioactivity (Section 2.4.2.4) may be
based on the energy and penetrating power of the radiation emissions, which would be indirectly
related to the radionuclides of potential concern. Caution must be used in evaluating radionuclide
concentrations or radioactivity for M&E with high levels of inherent background radioactivity.
NUREG-1575, Supp. 1 2-16 January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
2.4.2.2 Describe the Radionuclide Concentrations or Radioactivity Associated with the M&E
A description of expected radionuclide concentrations or radioactivity is also important for
supporting disposition decisions for M&E. Radionuclide concentrations or radioactivity in
excess of background (see Section 3.9 and Appendix B) support a finding that the M&E are
impacted. Historical records (Section 2.2.2) and process knowledge (Section 2.2.3) are sources
of information on radionuclide activities associated with M&E. In addition, sentinel
measurements (Section 2.2.4) can provide information on radionuclide concentrations or
radioactivity. A description of the expected radionuclide concentrations or radioactivity should
be developed for each of the radionuclides of potential concern. At a minimum, the average
expected activity should be provided. Some assumption regarding the expected activity will be
required in order to design a disposition survey using the guidance in Chapter 4. If no
assumption can be made, a preliminary survey should be performed. If possible, information on
the expected range and uncertainty (o, as described in Sections 3.8.1 and 5.6) of the activity
should be provided. The description of the expected activity should include the units, an estimate
of uncertainty in the values, and a summary of how the data were obtained (e.g., purpose of data
collection efforts, actual measurements, instrument used, count time, or process knowledge).
Any known or suspected relationships among concentrations for individual radionuclides should
be included in the description. For example, there is an expected relationship among fission
products from a nuclear reactor because of the common source of the radionuclides (i.e., nuclear
fission). Similarly, there is an expected relationship for activation and corrosion products.
Members of the natural decay series (i.e., thorium series, uranium series, actinium series; see
Appendix B) are also expected to have a relationship for activities based on equilibrium
conditions.
2.4.2.3 Describe the Distribution of Radioactivity
The distribution of radioactivity is primarily concerned with whether the activity is clustered or
more uniformly distributed throughout the item. A uniform distribution of activity has little
spatial variability, so the radionuclide concentrations or levels of radioactivity are fairly constant.
A clustered distribution of activity has high spatial variability, and small areas of elevated
activity are present as well as areas with little or no activity above background. The expected
distribution of radioactivity could include areas with uniform radionuclide concentrations or
levels of radioactivity and areas where the radionuclide concentrations or radioactivity is non-
uniform. For example, airborne deposition could have produced a uniform distribution of
radioactivity on horizontal exterior surfaces, while penetration through seams and access points
could result in clustered radioactivity on interior surfaces. In addition, the interior surfaces could
have a uniform distribution of radioactivity over localized areas (e.g., areas around a vent or
cooling fan). Concentrations of radionuclides on M&E can change over time due to in-growth,
decay, or diffusion.
2.4.2.4 Describe the Location of Radioactivity
The location of radioactivity is primarily concerned with whether the activity is located on the
surface or distributed throughout the volume of the M&E. Surficial radioactivity is restricted to
the surface of the M&E and is further described as removable, fixed, or some combination of
January 2009 2-17 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
these two. Removable (or non-fixed) radioactive material is radioactive material that can be
readily removed from a surface by wiping with an absorbent material. Fixed radioactive material
is not readily removed from a surface by wiping. Surficial radioactivity is generally associated
with non-permeable solid M&E. Volumetric radioactivity is not restricted to the surface of the
M&E and is usually associated with permeable materials, surfaces, or activation by neutrons or
other particles.
The question of surficial versus volumetric radioactivity is a complicated issue that may or may
not have a significant impact on the disposition survey design. The description of the location of
radioactivity used to design the survey may be independent of where the radioactivity is
physically located. For example, consider two different methods for surveying 60Co activity
concentrations distributed on the surface of several thousand small bolts. First, the bolts may be
surveyed in a container using in situ gamma spectroscopy assuming the radioactivity is
volumetrically distributed.4 If the same bolts are surveyed individually using a convey orized
survey monitor the conceptual model may describe the 60Co as surficial radioactivity.
In some cases, the location of the residual radioactivity may be well known. For example,
surface deposition of radioactivity on a non-porous material (e.g., smooth stainless steel) will not
penetrate into the material to a significant extent under most conditions, so the residual
radioactivity could be identified as surficial. Activated materials and bulk quantities of materials
usually have volumetric residual radioactivity, although surficial radioactivity may also be
present. On the other hand, the actual location of the residual radioactivity may be less well
known or unknown.
Process knowledge is the primary source of information on the location of residual radioactivity.
The planning team should review the information from Section 2.2.3 to determine the expected
location of residual radioactivity and the level of knowledge (i.e., well known, less well known,
unknown) associated with the information.
When the location of the residual radioactivity is well known, the planning team should proceed
with a survey design based on the appropriate assumption, surficial or volumetric. When the
location is less well known or unknown, the planning team may choose to proceed with multiple
survey designs to determine the possible effect the location of the residual radioactivity may
have on the design of the disposition survey.
2.4.3 Finalize the Description of the M&E
A final description of the M&E should be prepared following implementation of any preliminary
surveys. The description of the M&E should consider the information in Table 2.1 and Table 2.2
and provide sufficient information to design the disposition survey.
4 This example does not imply that any measurement technique should be applied to every situation. The
information in Section 3.8 should be used to develop the measurement quality objectives (MQOs) for a project. The
MQOs can be used to evaluate measurement techniques against the action levels and select the techniques best
suited for a specific application.
NUREG-1575, Supp. 1 2-18 January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
2.5 Select a Disposition Option
The disposition of the materials and equipment will be a key factor in designing the disposition
survey. MARSAME broadly considers two types of disposition decisions: release and
interdiction. Release surveys are used to determine whether radiological controls can be reduced,
removed, maintained at the current level, or transferred to another qualified user. Interdiction
surveys are used to initiate radiological control, or to decide current radiological controls are
adequate.
Examples of potential disposition options for release of impacted M&E include—
1. Reuse in a controlled environment.
2. Reuse without radiological controls (i.e., clearance).
3. Recycle for use in a controlled environment (i.e., authorized disposition).
4. Recycle without radiological controls.
5. Disposal as industrial or municipal waste.
6. Disposal as low-level radioactive waste.
7. Disposal as high-level radioactive waste.
8. Disposal as transuranic (TRU) waste.
9. Maintain current radiological controls.
Examples of potential disposition options for interdiction of impacted M&E include—
1. Remove M&E from general commerce and initiate radiological controls.
2. Decide to accept M&E for a specific application.
3. Decide not to accept M&E for a specific application.
4. Continue unrestricted use of M&E (no action).
The selection of a disposition option should be based on the information available at the end of
the IA. The disposition option (e.g., reuse, recycle, disposal, initiation of control, or refusal)
defines the action level (Section 3.3). The expected radionuclide concentrations or levels of
radioactivity associated with the M&E (Section 2.4.2) are compared to the action level to
determine whether the M&E will be controlled or uncontrolled following the disposition survey.
The disposition option also defines the alternative actions for the decision rule to be developed in
Section 3.6. Different disposition options may be applied to separate parts of equipment. If so,
implementation of the different dispositions implies the necessity for total or partial disassembly.
For example, it may be possible to remove a bucket from a backhoe for disposal and allow reuse
of the rest of the equipment.
2.6 Document the Results of the Initial Assessment
The results of the IA should be documented to the extent necessary to support the decisions
made. The level of documentation required will depend on the amount of information collected,
the quantity of M&E covered by the IA, the type of assessment (e.g., standardized or project-
specific), and, as applicable, administrative and regulatory requirements. Two options for
documenting the assessment results are the Standardized IA and the conceptual model as
described in the following sections. Figure 2.3 illustrates the documentation of the IA.
January 2009 2-19 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment
MARSAME
Is an
Applicable SOP Available
forth is M&E?
Implement and Document the
Results of the Survey as
Described in the SOP
(Section 2.6.1)
Develop a Conceptual Model
and Document the IA
(Section 2.6.2)
Proceed to Figure 6.1
Proceed to Figure 3.1
NOTE: Shaded box represents important milestone.
Figure 2.3 Documentation of the Initial Assessment
2.6.1 Document a Standardized Initial Assessment
A standardized IA is a set of instructions or questions that are used to perform the IA. These
instructions are usually documented in an SOP. The SOP should be developed, reviewed, and
documented in accordance with an approved quality system. Information on developing and
documenting a functional quality system can be found in EPA QA/G-1 (EPA 2002c). Guidance
on developing SOPs as part of a quality system can be found in EPA QA/G-6 (EPA 2001).
A standardized IA is generally associated with facilities or processes that regularly evaluate
similar types of M&E. The release of small tools and personal items from an operating nuclear
plant is one example of such a process. Another example, this time describing an interdiction
process, would be evaluating truckloads of scrap metal entering a recycle facility. SOPs may be
developed to describe repeated routine disposition surveys of similar M&E for both situations.
NUREG-1575, Supp. 1
2-20
January 2009
-------
MARS AME Initial Assessment of Materials and Equipment
The documentation of the IA results is described in the SOP. The documentation should be
sufficient to demonstrate that trained personnel using an approved SOP evaluated all potentially
impacted M&E. For a standardized IA, all these records are maintained but may not be directly
associated with the IA. Individual records for each item evaluated by an IA are not required.
The SOP should clearly describe its scope and the applicable types of M&E. This information
may be useful for determining whether the M&E are impacted as well as whether the SOP can be
used to evaluate the M&E. For example, if the SOP is applicable to all M&E used for a certain
process or within a certain part of a facility, this defines what M&E can be considered impacted
by that process.
The SOP should also describe the M&E that were used to develop the instructions. The
description of the M&E being investigated (Sections 2.2 and 2.3) should be compared to the
assumptions used to develop the instructions to determine if the SOP is appropriate. For
example, it may be appropriate to apply an SOP developed for scrap metal to evaluate hand
tools, since both are made from metal and may have similar surface radioactivity. Alternatively,
it may not be appropriate to use an SOP developed for scrap metal to evaluate dry active waste or
concrete rubble, since they may have volumetric activity and different surface efficiencies. At a
minimum, the rationale for applying the SOP to M&E other than specified in the SOP should be
documented.
The SOP should include the training requirements for personnel implementing the SOP.
Personnel performing the IA should be familiar with the SOP being implemented, as well as the
potential disposition options implied or explicitly stated in the SOP.
Additional documentation may be needed when the SOP is applied to situations other than those
considered during development of the SOP. The purpose of the additional documentation is to
determine whether the SOP may be applicable to a wider range of M&E. This documentation
will help provide technical support for modifying the SOP. If incorrect decisions are made
concerning the determination of whether M&E are impacted, or inappropriate recommendations
are made for disposition options, it may be necessary to modify the SOP to reduce the number of
decision errors. The additional documentation will help identify the source of the decision errors
and help provide technical support for modifying or revising the SOP.
2.6.2 Document a Conceptual Model
If a standardized IA approach is not available for the M&E being investigated, the results of the
IA should be documented in a conceptual model. If the information in MARSAME is being used
to develop a standardized survey design (e.g., a new SOP), the information on developing a
conceptual model applies.
The conceptual model is applied in case-by-case situations and decisions. The conceptual model
describes the M&E and radioactivity expected to be present for the project. The definition of
impacted and non-impacted as it applies specifically to the project should be included in the
conceptual model. The conceptual model describes the processes involving radioactive materials,
as well as how the radioactivity could become associated with the M&E.
January 2009 2-21 NUREG-1575, Supp. 1
-------
Initial Assessment of Materials and Equipment MARS AME
The description of the M&E documents the results of the IA investigation. At a minimum the
conceptual model should include a description of the physical attributes of the M&E (see Section
2.4.1 and Table 2.1), the radiological attributes of the M&E (see Section 2.4.2 and Table 2.2),
and a list of the applicable disposition options (Section 2.5). In addition, the conceptual model
helps identify data gaps and develop potential collection strategies for filling data gaps.
The conceptual model will serve as the basis for the information and assumptions used to
develop the disposition survey design in Chapter 4. In many cases the information in the
conceptual model will be included in either the survey design documentation or in the
documentation of the results of the disposition survey. The structure and content of the
conceptual model should be based primarily on the future uses of the data.
The planning team should review the information on radionuclides of potential concern provided
by the IA for consistency with the conceptual model. If the data appear incomplete or the quality
of the data is not adequate for the disposition survey being designed, the planning team may
decide that additional information needs to be collected using preliminary surveys before
proceeding with the survey design.
NUREG-1575, Supp. 1 2-22 January 2009
-------
MARSAME Identify Inputs to the Decision
3 IDENTIFY INPUTS TO THE DECISION
3.1 Introduction
This chapter identifies sources of information needed to evaluate the disposition option, or
options, selected during the initial assessment (IA). During implementation of an existing
standard operating procedure (SOP), this information would have been considered during
development of the SOP. This chapter discusses factors affecting the selection of survey units,
provides guidance on defining spatial and temporal boundaries, and examines practical
constraints on collecting data. Figure 3.1 depicts the process of identifying the inputs to the
decision. The expected output from this chapter is a decision rule, or multiple decision rules. A
decision rule is a theoretical "if...then..." statement that defines how the decision maker would
choose among alternative actions if the true state of nature could be known with certainty (EPA
2006a). There are three parts to a decision rule (Section 3.7):
• An action level that causes a decision-maker to choose between the alternative actions
(Section 3.3),
• A parameter of interest that is important for making decisions about the target population
(Section 3.4), and
• Alternative actions that could result from the decision (Section 3.5).
Other inputs to the decision discussed in this chapter include selecting radionuclides or radiations
of concern (Section 3.2), developing survey unit boundaries (Section 3.6), inputs for selecting
provisional measurement methods (Section 3.8), and identifying reference material (Section 3.9).
Also discussed in this chapter is the evaluation of an existing survey design to determine if it will
meet the data quality objectives (DQOs; Section 3.10).
This chapter provides guidance on performing Step 3, Step 4, and Step 5 of the DQO process
(EPA 2006a) for designing a disposition survey. These steps build on the IA where members of
the planning team were identified and M&E under investigation were identified as impacted
(non-impacted M&E do not require additional investigation). A disposition option was selected
(Section 2.5) and documented (Section 2.6).
It is important to remember the DQO process is an iterative process. This means new information
can be incorporated into the planning process and outputs from previous steps can be modified to
incorporate the new information. For example, if no measurement methods are identified in
Section 3.8 that meet the data requirements for a specific disposition option, the planning team
may return to Section 2.5 to select a different disposition option. Alternatively, the selection of
an action level or survey unit boundary may be affected by the available measurement
techniques. The issues associated with surficial vs. volumetric radioactivity (see Section 2.4.2)
affect the kinds of information (i.e., action level, survey unit identification, and measurement
techniques) as well as the definition of study boundaries (i.e., target population, spatial
boundaries, practical constraints on collecting data, subpopulation for which separate decisions
will be made).
January 2009 3-1 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision
MARSAME
From Figure 2.3
Select Radionuclides or
Radiations of Concern
(Section 3.2)
Identify Action Levels
(Section 3.3)
Describe the Parameter of Interest
(Section 3.4)
Identify Alternative Actions
(Section 3.5)
Identify Survey Units
(Section 3.6)
Develop a Decision Rule
(Section 3.7)
Develop Inputs for Selection of
Provisional Measurement Methods
(Section 3.8)
NOTE: Shaded boxes
represent important
milestones.
Identify Reference Materials
(Section 3.9)
Is there an Existing
Survey Design?
Do the M&E Meet the
Survey Requirements?
Yes
Implement and Document Results
as Described in the Survey
(Section 3.10)
Figure 3.1 Identifying Inputs to the Decision
NUREG-1575, Supp. 1
3-2
January 2009
-------
MARSAME Identify Inputs to the Decision
At the end of this chapter, the planning team should have the information required to design the
disposition survey and know whether appropriate measurement techniques are available. Spatial
and temporal boundaries will be identified, along with any practical constraints on data
collection activities. Examples of practical constraints on data collection include time, budget,
personnel, or equipment. For example, a box counter is selected to perform measurements for
clearance of items from a radiologically controlled area. Assume a five-minute count time is
required to achieve the survey objectives, and another minute is required to swap items in the
detector. This means that ten measurements can be performed each hour. If more than 240 items
require clearance each day, this measurement method would be impractical since a single box
counter cannot clear all of the M&E. The decision rule(s) developed at the end of this chapter
will be used to develop survey designs in Chapter 4.
3.2 Select Radionuclides or Radiations of Concern
A list of radionuclides of potential concern was developed in Section 2.4.2.1 as part of the
description of radiological attributes associated with the M&E. Before a decision rule can be
developed or a disposition survey designed, a final list of radionuclides or radiations to be
measured must be prepared.
The selection of radionuclides or radiations of concern is linked to several inputs to the decision.
For example, the identification of an action level (Section 3.3) may determine if the survey
results need to be radionuclide-specific, forcing the planning team to identify individual
radionuclides of concern. On the other hand, the selection of a non-radionuclide specific
measurement method may allow the selection of a radiation of concern (i.e., alpha [a], beta [0],
gamma [y], x-ray, or neutron radiation) without ever finalizing a list of radionuclides of concern.
Finalizing the list of radionuclides or radiations of concern is an example of the iterative nature
of the survey design process. The planning team is expected to evaluate different survey
techniques and measurement methods. Evaluating these different survey techniques and
measurement methods will require the planning team to return to the list of radionuclides of
potential concern and go through the selection of radionuclides or radiations of concern.
Actually, the final selection of radionuclides or radiations of concern may not occur until the
disposition survey is optimized; the last step (Step 7) of the DQO process (Section 4.4.5).
3.3 Identify Action Levels
The action level is the numerical value or values that cause a decision maker to choose one of the
alternative actions. The radionuclides of concern and disposition options selected at the
completion of the IA define the alternative actions for the disposition survey.
Figure 3.2 shows the process for selecting action levels. As shown , the iterative nature of the
DQO process may result in changes to the action levels or disposition options based on other
factors (e.g., availability of appropriate measurement techniques, measurability, surficial vs.
volumetric activity). The planning team should consider the effect of action levels on other steps
in the survey design process, as well as any effects these other steps might have on the action
levels.
January 2009 3-3 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision
MARSAME
From Figure 3.1
Identify Applicable
Regulatory Limits
(Dose-, Risk-, Activity-,
or Method-Based)
Identify Applicable
Requirements
(e.g., ANSI N13.12)
Identify Applicable
Administrative Limits
(e.g., Waste Acceptance
Criteria)
i.
Identify Applicable DOT
Requirements for
Shipping M&E
Convert Potential Action Levels
into Measurement Units
NOTE: Shaded boxes
represent important
milestones.
Finalize Selection of Action Level(s)
Modify AL Using
Equation 3-1
(Gross Activity AL)
Are There Multiple
Radionuclides?
Radionuclide-
Specific Measurements?
Will Surrogates
Infer Multiple
Radionuclides?
Surrogate
Measurements
Available?
Modify AL Using
Equation 3-4
Modify AL Using
Equation 3-3
Evaluate Survey Results Using
Equation 3-2 (Unity Rule)
Apply ALARA, as Appropriate
Document Selection of
Action Level(s)
Return to Figure 3.1
NOTE: Information on ALARA
can be found in 10CFR20, 10CFR
835, DOE 1993, ICRP 1989,
NCRP 1993, NRC 1977, NRC
1982, NRC 1993, NRC 2002, and
PNL1988.
Figure 3.2 Identifying Action Levels
(Apply to Each Disposition Option Selected in Section 2.5)
NUREG-1575, Supp. 1
3-4
January 2009
-------
MARSAME Identify Inputs to the Decision
Action levels are radionuclide- or radiation-specific and in units of concentration or activity (e.g.,
Bq/kg of 137Cs, Bq/m2 of alpha radiation, Bq of 60Co). Action levels may be provided, derived
from dose- or risk-based standards, or converted into more convenient units for a specific
measurement technique.l
More than one action level may be required to demonstrate compliance with a specific standard.
For example, DOE Order 5400.5 Figure IV-1 (DOE 1993) provides limits for average total
surface activity, maximum total surface activity, and maximum removable surface activity (see
Appendix E). All three limits must be achieved to demonstrate compliance for disposition of the
M&E. Sometimes multiple regulatory requirements may apply, for example transportation
regulations combined with waste acceptance criteria and health protection standards.
Action levels may be established based on total activity or incremental activity levels relative to
background. Examples of incremental action levels include activity levels based on dose or risk
above background, or interdiction at some quantity above background. For these types of action
levels it is important to establish a representative reference material (Section 3.9) for
comparison.
Action levels may explicitly or implicitly require the use of a specific measurement technique
(Section 5.9.1) or instrument (Section 5.9.2). These "method-based" requirements should be
considered not only during survey design and implementation, but also during selection of
disposition options and action levels. For example, U.S. Department of Transportation (DOT)
regulations include package dose rate limits (49 CFR 173.441) as well as removable external
radioactive contamination limits (49 CFR 173.443 and 177.843). Dose rate limits on external
surfaces at one meter from package surfaces imply the use of in situ or direct measurements,
although the selection of specific instrumentation is not specified. Measurements of removable
contamination explicitly require the use of smears, but the procedure for collecting and analyzing
the smear is not specified. The NRC regulations for transportation of radioactive packages (10
CFR 71) replicate DOT regulations and define limits for "surface contaminated objects,"
including fixed and non-fixed surface radionuclides (Appendix E.3.7).
At this point, it is important to identify action levels appropriate for the disposition survey. If
multiple action levels are identified, the planning team may decide to continue with the
development of multiple survey designs that will be evaluated in Section 4.4. The decision maker
and the planning team will need to evaluate the action levels and select the action level that best
meets the DQOs developed for the survey. The selected action levels are used to develop
decision rules in Section 3.7. Alternatively, the planning team may decide to revisit the selection
of disposition options from the IA to further limit the scope of the disposition survey and
eliminate some of the action levels. In either case, the selection of action levels will be finalized
in Section 4.4 with the development of a disposition survey design. Information supporting the
selection of an action level(s) is discussed in Sections 3.3.1 through 3.3.4.
1 Correctly converting action levels to counts or counts per minute (cpm) using the appropriate calibration may
provide a useful comparison for real-time evaluation of field measurement results as long as field results (e.g., cpm)
are converted to and recorded in the same radiological units as the action levels.
January 2009 3-5 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
3.3.1 Identify Sources of Action Levels
There are many potential sources of action levels available for use in developing disposition
surveys. An action level may be based on—
• Dose- or risk-based regulatory standard (i.e., disposition criterion),
• Waste acceptance criteria at a disposal site,
• Regulatory threshold standard (e.g., indistinguishable from background or no detectable
radioactivity),
• DOT regulations for shipping radioactive M&E,
• Activity-based standard,
• As low as reasonably achievable (ALARA) considerations,
• Administrative limits, or
• Limitations on technology (performance criteria for an analytical method).
Appendix E provides information on some of the federal sources of action levels that can be
applied to M&E. The list of sources for action levels is not exhaustive, but is intended to provide
examples of different types of action levels that are referred to throughout this supplement.
National and International organizations have published recommendations for action levels (e.g.,
NCRP 2002, ANSI 1999). These recommendations may be a useful source of action levels if
approved by the appropriate authorities.
As previously stated, in many cases the action levels will be dictated by the disposition option
selected during the IA. For example, the action levels for M&E being considered for clearance
may be a regulatory standard, whereas the action levels for M&E being considered for disposal
as radioactive waste will often use the waste acceptance criteria for a disposal site.
Multiple sources of action levels may be identified for a single disposition option. Waste
acceptance criteria can be evaluated for several potential burial sites.
In addition, a single source of action levels could be acceptable for more than one disposition
option. Dose- and risk-based regulatory standards can be applied to both release and recycle
scenarios, as well as for surficial or volumetric radioactivity. On the other hand, activity-based
standards may have limited applicability, such as DOE Order 5400.5 (DOE 1993) Figure IV-1
that only applies to release of M&E with surficial radioactivity.
The identification of sources for action levels may affect other decisions made during
development of a disposition survey design. Identification of survey units and spatial boundaries
for a survey are often directly linked to the action levels. In addition, the expected levels of
residual radioactivity identified during the IA (Section 2.6) will often suggest which disposition
options are feasible.
At a minimum the planning team should identify at least one source of action levels applicable to
the disposition option(s) selected during the IA. Any information related to the action levels that
may affect other decisions should also be listed. A partial list of information that may be
available from sources of action levels includes—
NUREG-1575, Supp. 1 3-6 January 2009
-------
MARSAME Identify Inputs to the Decision
• Radionuclides of concern or types of radiation,
• Assumptions regarding surficial (fixed and removable) or volumetric residual radioactivity,
• Area or volume over which the residual radioactivity can be averaged,
• Assumptions about potential disposition of the M&E (e.g., exposure scenarios, reuse vs.
recycle), and
• Conversions from dose or risk to activity or concentration (e.g., modeling and modeling
assumptions).
3.3.2 Finalize Selection of Action Levels
In cases where more than one source of action levels is identified, it is necessary to select an
action level as the basis for the disposition survey design. Generally, the source that provides the
most restrictive action levels (i.e., the most protective of human health and the environment) will
be appropriate for designing the disposition survey. If the planning team cannot determine which
action levels are most restrictive, multiple survey designs should be developed and the selection
of action levels will be determined by the selection of the most effective survey design (Section
4.4).
The expected location of residual radioactivity is an important factor in the selection of
appropriate action levels. Some sources of action levels are only applicable for surficial
radioactivity (e.g., Figure IV-1 of DOE 1993, DOT regulation 49 CFR 173.433). Other sources
of action levels (e.g., ANSI 1999) or dose assessments for deriving action levels (e.g., NRC
2003a) make assumptions about whether the residual radioactivity is surficial or volumetric, or a
combination of both. Section 2.4.2.4 discusses the location of radioactivity associated with the
M&E.
While the location of residual radioactivity is important in determining the most restrictive action
levels, other physical and radiological characteristics should also be considered. The final
selection of action levels should be supported by the description of the M&E provided by the IA
(Section 2.6).
3.3.3 Modify Action Levels When Multiple Radionuclides are Present
The implementation of action levels should be considered when evaluating whether they will be
applied to a specific survey unit or project. Section 3.3.1 discusses potential sources for action
levels, and Section 3.2 discusses the approach for selecting the radionuclides of concern.
Calculating the relative ratios among multiple radionuclides and determining the state of
equilibrium for decay series radionuclides is discussed in MARSSEVI Section 4.3. This section
describes how individual action levels can be combined and applied when more than one
radionuclide is present.
Action levels are often provided for types of radioactivity or groups of radionuclides. For
example, DOE Order 5400.5 Figure IV-1 (DOE 1993) provides surface activity action levels for
four groups of radionuclides (Appendix E). For the simple case in which the activity is entirely
attributable to one radionuclide, the action levels for that radionuclide are used for comparison to
January 2009 3-7 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
survey data. In these examples, the disposition survey data may be obtained from direct
measurements of activity, scanning with data logging, conveyorized survey monitor surveys, or
other appropriate methods.
Dose- or risk-based action levels may be radionuclide-specific. Each radionuclide-specific action
level corresponds to the chosen disposition criterion (e.g., regulatory limit in terms of dose or
risk). For example, ANSI 1999 provides surface and volumetric activity action levels for
individual radionuclides. When multiple radionuclides are present at concentrations equal to the
action levels, the total dose or risk for all radionuclides would exceed the disposition criterion. In
these cases it is possible to modify the action levels based on relationships between the
radionuclides of concern and still demonstrate compliance with the disposition criterion.
The method used to modify the action levels depends on the radionuclides of concern and the
selected measurement method. If the measurement method reports total activity for a type of
radiation (e.g., gross a, P, or y assays) the method is non-radionuclide specific and the guidance
in Section 3.3.3.1 should be applied. If the measurement reports activity for individual
radionuclides (e.g., gamma spectroscopy, alpha spectrometry) the method is radionuclide
specific and the guidance in Section 3.3.3.3 should be applied.
3.3.3.1 Modify Action Levels for Non-Radionuclide-Specific Measurement Methods
For situations in which there are radionuclide-specific action levels and multiple radionuclides
are present, a gross activity action level can be developed. Gross activity action levels are also
discussed in Section 4.3.4 of MARSSIM. This approach enables field measurement of gross
activity (using static direct measurements or scans), rather than determination of individual
radionuclide activity, for comparison to the action levels. The gross activity action level for
M&E with multiple radionuclides is calculated as follows:
1. Determine the relative fraction (/) of the total activity contributed by the radionuclide.2
2. Obtain the action level for each radionuclide present.
3. Substitute the values off and action levels in the following equation.
Gross Activity AL = -7 T (3 -1)
Iff f }
\J^+j2_+_Jn_\
(AL, AL, ALJ
Where:
ft = relative fraction of total activity contributed by radionuclide /' (/' = 1, 2,..., n)
AL, = action level for radionuclide /
2 The determination of relative fractions may be based on process knowledge, empirical data, or a combination of
both. It may be difficult or impractical to determine the relative fractions contributed by all radionuclides of concern.
The alternatives are to analyze each radionuclide independently, or use conservative assumptions to determine the
relative fractions. Additional guidance is provided in MARSSIM Section 4.3.
NUREG-1575, Supp. 1 3-8 January 2009
-------
MARSAME Identify Inputs to the Decision
Example 1: Assume that 40% of the total radioactivity was contributed by a radionuclide
with an action level of 1.4 Bq/cm2 (8,400 dpm/100 cm2). An additional 40% of the total
radioactivity was contributed by a radionuclide with an action level of 0.28 Bq/cm2 (1,700
dpm/100 cm2), and the final 20% of the radioactivity was contributed by a radionuclide with
an action level of 0.14 Bq/cm2 (840 dpm/100 cm2). Using Equation 3-1:
Gross Activity AL = -, T = 0.32 Bq/cm2 (1,900 dpm/100 cm2)
y (0.40 0.40 0.20^ H v , F ;
+ +
I 1.4 0.28 0.14
Equation 3-1 may not be appropriate for survey units with radioactivity from multiple
radionuclides having unknown or highly variable concentrations of radionuclides. In these
situations, the best approach may be to select the most restrictive surface activity action level
from the mixture of radionuclides present.3 If the mixture contains radionuclides that cannot be
measured using field survey equipment, such as 3H or 55Fe, laboratory analyses of M&E samples
may be necessary.
3.3.3.2 Modify Action Levels for Non-Radionuclide-Specific Measurements of Decay-Series
Radionuclides
Demonstrating compliance with surface activity action levels for radionuclides of a decay series
(e.g., radium, thorium, uranium) that emit both alpha and beta radiation may be demonstrated by
assessing alpha, beta, or both radiations. However, relying on the use of alpha surface activity
measurements often proves problematic because of the highly variable level of alpha attenuation
by rough, porous, uneven, and dusty surfaces. Beta measurements typically provide a more
accurate assessment of thorium and uranium (and their decay products) on most building
surfaces because surface conditions cause significantly less attenuation of beta particles than
alpha particles. Beta measurements, therefore, may provide a more accurate determination of
surface activity than alpha measurements.
The relationship of beta and alpha emissions from decay chains or various enrichments of
uranium should be considered when determining the surface activity for comparison with the
action level value(s). When the initial member of a decay series has a long half-life, the
radioactivity associated with the subsequent members of the series will increase at a rate
determined by the individual half-lives until all members of the decay series are present at
activity levels equal to the activity of the initial member. This condition is known as secular
equilibrium. Pages 4-6 and 4-7 in MARSSIM also provide a discussion on secular equilibrium.
The difficulty with radionuclides that are part of a natural decay series is that time must pass for
a sufficient number of half-lives of the longest-lived decay product that intervenes between a
radionuclide and the initial member of a decay series in order to establish secular equilibrium. In
Of")
the case of Th, the time to establish secular equilibrium is almost 40 years. This is because
232Th decays into 228Ra, which has a half-life of 5.75 years. In the case of 238U, the time to
establish secular equilibrium is approximately 2 million years. This is because 234U has a half-
3 In Example 1, the most conservative action level is 0.14 Bq/cm2.
January 2009 3-9 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
life of approximately 250,000 years. 226Ra, another member of the 238U decay series, presents
special problems. 226Ra decays into 222Rn, which is a noble gas that can escape the matrix and
disrupt equilibrium. It is important to remember the reason for determining relationships between
radionuclides. If the relationships are known or can be estimated,4 the costs and amount of time
required for performing measurements can be significantly reduced. The alternative to
determining the relationships between radionuclides is performing radionuclide-specific
measurements for each radionuclide of concern.
Example 2: the radionuclide of concern is 232Th, and all of the decay products are in secular
equilibrium. Assume that a gas proportional detector will be used for surface activity
measurements. The detector's efficiency is dependent upon the radionuclide mixture
measured and the calibration source area. Guidance from the International Organization for
Standardization (ISO 1988) states:
"The dimensions of the calibration source should be sufficient to cover the
window of the instrument detector. Where, in extreme cases, sources of such
dimensions are not available, sequential measurements with smaller distributed
sources of at least 100 cm2 active area shall be carried out. These measurements
shall cover the whole window area or at least representative fractions of it and
shall result in an average value for the instrument efficiency."
The concentration of 232Th is inferred from a measurement that includes the initial member of
the decay series and all of its decay products. The efficiency of such measurements, relative to
Of) ooo
each decay of Th, can be greater than 100%. The efficiency, relative to each decay of Th,
is calculated by weighting the individual efficiencies from each of the radionuclides present
(Table 3.1).
It is important to recognize that if the action level for 232Th includes the entire 232Th decay series,
the total efficiency for 232Th must account for all of the radiations in the decay series. The total
weighted efficiency calculated in Table 3.1 may be used to modify action levels for non-
radionuclide specific measurements using a gas proportional counter to measure thorium series
radionuclides. The total weighted efficiency can be substituted into an equation (e.g., MARSSEVI
Equations 6-1, 6-2, 6-3, or 6-4) to convert the action level (e.g., activity units) into measurement
units (e.g., counts or cpm). The modified action level can then be compared directly to the
measurement results for a real time assessment of the data.
4 There are risks and tradeoffs associated with using estimated values. The planning team should compare the
consequences of potential decision errors with the resources required to improve the quality of existing data to
determine the appropriate approach for a specific project.
NUREG-1575, Supp. 1 3-10 January 2009
-------
MARSAME
Identify Inputs to the Decision
Table 3.1 Example Detector Efficiency Calculation ( Th in Complete Equilibrium with its
Decay Products) Using a Gas Proportional Detector
Radionuclide
232Th
228Ra
228Ac
228Th
224Ra
220Rn
216Po
212pb
212Bl
212Bl
212p0
208r-p.
Energy*
(keV)
4.00 MeV alpha
7.2 keV beta
377 keV beta
5.40 MeV alpha
5. 67 MeV alpha
6.29 MeV alpha
6.78 MeV alpha
102keVbeta
769 keV beta
6.05 MeV alpha
8.78 MeV alpha
557 keV beta
Fraction
1
1
1
1
1
1
1
1
0.64
0.36
0.64
0.36
Instrument
Efficiency
0.40
0
0.54
0.40
0.40
0.40
0.40
0.40
0.66
0.40
0.40
0.58
Surface
Efficiency
0.25
0
0.50
0.25
0.25
0.25
0.25
0.25
0.50
0.25
0.25
0.50
Weighted
Efficiency
0.1
0
0.27
0.1
0.1
0.1
0.1
0.1
0.211
0.036
0.064
0.104
Total efficiency = 1 .29
Alpha energies are weighted averages based on relative abundance of major particle emissions totaling at least 90%
of the total emissions. Beta energies are average energies. Source: Japanese Atomic Energy Research Institute data
from NRC Radiological Toolbox Version 1.0.0 (NRC 2003b). Table adapted from NUREG-1761 Table 4.3 (NRC
2002a).
3.3.3.3 Modify Action Levels for Radionuclide-Specific Measurement Methods
In many cases action levels correspond to a disposition criterion (e.g., a regulatory limit) in terms
of dose or risk. When multiple radionuclides are present at concentrations equal to the action
levels, the total dose or risk for all radionuclides would exceed a dose- or risk-based disposition
criterion. In this case, the individual action levels would need to be adjusted to account for the
presence of multiple radionuclides contributing to the total dose or risk. The surrogate
measurements discussed in this section describe adjusting action levels to account for multiple
radionuclides when radionuclide-specific analyses of media samples or radionuclide-specific in
situ measurements (e.g., in situ gamma spectroscopy) are performed. The use of surrogate
measurements is also described in Section 4.3.2 of MARSSEVI. Other methods used to account
for the presence of multiple radionuclides include the use of the unity rule (MARS SIM Section
4.3.3) and development of a gross activity action level to adjust the individual radionuclide
action levels (see Section 3.3.3.1 and MARSSEVI Section 4.3.4).
The unity rule is satisfied when radionuclide mixtures yield a combined fractional concentration
limit that is less than or equal to one. The unity rule can be described by Equation 3-2:
C
L-i
AL AL,
AL.
-------
Identify Inputs to the Decision MARSAME
d = concentration or activity value for each individual radionuclide (/' = 1,2, ..., ri)5
= action level value for each individual radionuclide (/' = 1, 2, ..., n)
For the disposition of M&E that contain multiple radionuclides, it may be possible to measure
just one of the radionuclides and still demonstrate compliance for all of the radionuclides present
in the M&E through the use of surrogate measurements. In the use of surrogates, it is often
difficult to establish a "consistent" ratio between two or more radionuclides. Rather than follow
prescriptive guidance on acceptable levels of variability for the surrogate ratio, the planning team
should review the data collected to establish the ratio (e.g., from preliminary surveys or process
knowledge) and account for the variability as a measurement quality objective (MQO) during
selection of a measurement method (see Sections 3.8, 5.5, and 7.3). The action levels must then
be modified to account for the fact that one radionuclide is being used to account for the
presence of one or more other radionuclides.
Action levels for the measured radionuclide are modified (ALmeaS:moci) to account for a single
inferred radionuclide (e.g., inferring 55Fe based on the presence of 60Co) using Equation 3-3
(modified from Equation 6.2 in Abelquist 2001):
AL
infer
C }
infer \ AT i AT
^ V^^meas + ^Mnf er
y V meas j
Where:
ALmeaStmod = modified action level for the radionuclide being measured
ALmeas = action level for the radionuclide being measured
ALinfer = action level for the inferred radionuclide (i.e., not measured)
surrogate ratio of the inferred to the measured radionuclide
(1 O\
3-3)
When the measured radionuclide will be used as a surrogate for more than one radionuclide,
ALmeas,modC%n- be calculated using Equation 3-4 (MARSSIM Equation 1-14):
ALmeas.^=-f— - - T-^T (3-4)
1 KI K^ K 1
ALn
\_ i ^ ~j n /
Where:
ALi = action level for the measured radionuclide by itself
AL2 = action level for the second radionuclide (or first radionuclide being inferred)
that is being inferred by the measured radionuclide
R2 = ratio of concentration of the second radionuclide to that of the measured
radionuclide
5 C (radionuclide concentration) must be in the same units as the action level. If the action level is provided in
activity units, C will also be in units of activity.
NUREG-1575, Supp. 1 3-12 January 2009
-------
MARSAME Identify Inputs to the Decision
AL3 = action level for the third radionuclide (or second radionuclide being inferred)
that is being inferred by the measured radionuclide
RS = ratio of concentration of the third radionuclide to that of the measured
radionuclide
ALn = action level for subsequent radionuclides being inferred by the measured
radionuclide
Rn = ratio of concentration of subsequent radionuclides to that of the measured
radionuclide
Recall that the benefit of using surrogates is the avoidance of costly laboratory-based analytical
methods to provide estimates of activity for individual radionuclides of concern. Surrogates often
emit y-rays, which enable the use of noninvasive and nondestructive methods. However, a- and
p-emitting radionuclides can also be used as surrogates, depending on the objectives of the
survey and project-specific information. The surrogates come in two forms: (1) surrogates by
virtue of a decay series, and (2) surrogates by virtue of association. Surrogates that are part of a
decay series are discussed in Section 3.3.3.2. Radionuclides that are not part of a decay series
have the potential to be surrogates when they are produced by the same nuclear process (usually
fission or activation) and have similar chemical properties and release mechanisms. However,
this type of surrogate needs special attention because there must be a consistent ratio between the
measured radionuclide and surrogate, which is not always easy to demonstrate. For example, in
the case of nuclear power reactors, 60Co can be used as a surrogate of 55Fe and 63Ni because both
117
are activation-corrosion products with similar chemical properties. Similarly, Cs can be used
as a surrogate for the P-emitting 90Sr because both are fission products and generally are found in
soluble cationic forms. While 137Cs has been suggested as a possible surrogate for 99Tc, it must
be noted that 99Tc has different chemical properties and, in nuclear power reactors, it has
different release mechanisms. Additional information is available on surrogates and establishing
ratios (MARSSIM 2002, NRC 2000, and EPRI 2003).
3.3.4 Evaluate Interface With Exposure Pathway Models
Disposition criteria may be provided in units that cannot be measured directly, for example total
effective dose equivalent (TEDE) or lifetime risk of cancer incidence. These criteria are usually
converted into action levels with concentration or activity units. This conversion is typically
accomplished using exposure pathway models, such as RESRAD-Recycle for metals (DOE
2005). While the selection and application of these models is outside the scope of MARSAME,
the assumptions used to develop action levels should be considered during development of a
disposition survey design.
Alternatively, disposition criteria may be provided in units more easily measured. In general,
there are assumptions used in the development of these types of action levels. It is the
responsibility of the authority issuing the action levels to ensure regulatory involvement in their
development and to document and make assumptions available to users.
The assumptions used to design the disposition survey (Section 4.4) need to match the
assumptions used to develop the action levels. Examples of parameters that could affect
disposition survey designs and could be inputs to exposure pathway models include—
January 2009 3-13 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
• Volume, mass, or surface area of M&E;
• Accessibility;
• Physical and chemical characteristics of radionuclides or radiations of concern (types of
emissions, energies, half-lives, known or expected relationships);
• Distribution of radioactivity (uniform or variable);
• Location of radioactivity (surficial or volumetric); and
• Fixed or removable radioactivity, or some combination of both.
3.4 Describe the Parameter of Interest
The parameter of interest is the population parameter (e.g., mean, median, percentile, or total
amount) that the planning team considers to be important for making decisions about the target
population (EPA 2006a). The target population is the collection of all possible measurement
results that could be used to support a disposition decision concerning the M&E being
investigated. The target population is defined by the selection of survey unit boundaries (see
Section 3.6), since a separate disposition decision will be made for each survey unit.
The parameter of interest may be specified as part of the action level. For example, DOE Order
5400.5 Figure IV-1 (DOE 1993) lists action levels (i.e., surface concentration limits in dpm per
100 cm2), parameters of interest (i.e., mean and maximum values), and target populations (i.e.,
1 m2 for average concentration and 100 cm2 for maximum and removable limits).
Alternatively, the planning team may need to select the parameter of interest based on project-
specific needs and considerations. The most common parameter used in decision-making is the
mean because the mean is frequently used to model random exposure to environmental
contamination (EPA 2006a). The more complex the parameter of interest, the more complex will
be the decision rule (Section 3.7) and accompanying survey design.
3.5 Identify Alternative Actions
Before decision rules can be developed, the planning team needs to identify the alternative
actions based on the disposition options identified in Section 2.5. Alternative actions are the
possible actions that may be taken for disposition of M&E, including an alternative that requires
no action. Table 3.2 lists examples of alternative actions for disposition options provided in
Section 2.5.
NUREG-1575, Supp. 1 3-14 January 2009
-------
MARSAME
Identify Inputs to the Decision
Table 3.2 Example Alternative Actions
Disposition Option
Alternative Actions
Release for reuse
Reuse without radiological controls
Reuse with radiological controls
Maintain current level of radiological control and do not reuse (no action)
Release for recycle
Recycle without radiological controls
Recycle with radiological controls
Maintain current level of radiological control and do not recycle (no action)
Release for disposal
Dispose of M&E as municipal or industrial waste
Dispose of M&E as low-level radioactive waste
Dispose of M&E as high-level radioactive waste
Dispose of M&E as transuranic (TRU) waste
Maintain current level of radiological control without disposal (no action)
Interdiction
Remove M&E from general commerce and initiate radiological controls
Decide to use or accept M&E for a specific application
Decide not to use or accept M&E for a specific application
Continue unrestricted use of M&E (no action)
3.6 Identify Survey Units
To make a decision concerning the disposition of M&E it is necessary to describe the total
collection of M&E being investigated and define what segment of the total will be considered for
individual decisions. In other words, the planning team must specify the amount of M&E for
which a separate disposition decision will be made. When the M&E consist of discrete items
surveyed individually (e.g., hand tools) this task is simple. However, disposition decisions are
often required for more complex situations (e.g., bulk dispersible materials, excavation
equipment).
Survey unit boundaries should be clearly defined in order to know exactly what amount of M&E
is covered by a single decision. This clear and unambiguous definition will make data
interpretation more straightforward.
An M&E survey unit is the specific lot, amount, or piece of M&E on which measurements are
made to support a disposition decision concerning that specific lot, amount, or piece of M&E.
The purpose of this section is to identify the information that will be used to define the survey
unit boundaries. The expected output from this section is the identification of survey unit
boundaries that will be used to develop the decision rule in Section 3.7. Figure 3.3 shows the
process used to develop survey unit boundaries.
January 2009
3-15
NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision
MARSAME
Are Survey
Unit Dimensions Specified
in a Regulation?
(Section 3.3.1)
Determine Assumptions Used to
Develop Action Levels
(Sections 3.3.1 and 3.2.4)
Yes
NOTE: Shaded boxes
represent important
milestones.
Develop Survey Unit Boundaries
Based on Assumptions
(Section 3.6.1)
Identify Parameter of Interest
Identify Target Population
Reduce Survey Unit Size Based on
Physical Characteristics of the M&E
(Sections 3.6.1, 2.3.1, and 5.3)
Handling (Size, Shape, Mass)
Acessibility
Reduce Survey Unit Size
Based on Measurement
Method Requirements
(Sections 3.6.1, 3.8, and 5.9)
Examples:
Dimensions of Box Counter or Portal Monitor
Field of View for In Situ Gamma Spectrometer
Penetrating Power of Radioactivity
Identify Final
Survey Unit Boundaries
Document Development of Survey
Unit Boundaries as Part of the
Survey Design
(Section 4.5)
Return to Figure 3.1
Figure 3.3 Developing Survey Unit Boundaries
(Apply to All Impacted M&E for Each Set of Action Levels Identified in Section 3.3)
NUREG-1575, Supp. 1
3-16
January 2009
-------
MARSAME Identify Inputs to the Decision
Survey unit boundaries are affected by many variables associated with the action level, physical
properties of the M&E, characteristics of the radionuclides of concern, and available
measurement techniques. Variables affecting the definition of survey units include—
• Acti on Level ( S ecti on 3.3 )
o Assumptions used to develop the action level (e.g., surficial [fixed or removable] or
volumetric, Section 3.3.1)
o Modeling assumptions used to convert from dose or risk to concentration or activity
(Section 3.3.4)
• Physical Properties of the M&E (Section 2.4.1)
o Dimensions (i.e., size, shape, surface area)
o Complexity (i.e., number and type of components)
o Accessibility (i.e., measurability)
o Inherent value
• Radiological Attributes of the M&E (Section 2.4.2)
o Radionuclides of concern (e.g., major radiations and energies, half-life)
o Expected activity levels (e.g., average, range, variance, known or potential relationships)
o Distribution (i.e., uniform or non-uniform)
o Location (i.e., surficial [fixed or removable] or volumetric)
• Available Measurement Methods (Section 3.8, Section 5.9)
o Measurement quality objectives (Section 3.8, Section 5.5, Section 7.3)
o Measurement performance characteristics (Section 5.5)
3.6.1 Define Initial Survey Unit Boundaries
Initial survey unit boundaries should be developed based on one primary factor and modified, as
needed, using additional variables. MARSAME recommends using the assumptions used to
develop the action levels as the primary factor used to develop survey unit boundaries. The
modifying variables will usually be specific to a measurement technique, or determined by the
M&E being investigated.6
In many cases the action levels will define the survey unit boundaries. For example, DOE Order
5400.5 Figure IV-1 (DOE 1993) provides action levels for surface activity. The survey unit
boundaries are restricted to the surface of the M&E being investigated. Alternatively, NUREG-
1640 (NRC 2003a) provides modeling assumptions used to develop the action levels for different
materials. Radionuclide-specific action levels are provided for separate materials (e.g., ferrous
metals, concrete) for both surficial and volumetric radioactivity. In addition, each action level
lists the limiting exposure scenario. For example, exposure scenarios for concrete (NRC 2003a)
include—
6 This approach differs from guidance found in MARSSIM Section 4.6. While MARSSIM also uses the assumptions
used to develop the action levels (i.e., derived concentration guideline levels [DCGLs] in MARSSIM) as the primary
factor in developing survey unit boundaries, the modifications are different. MARSSIM guidance allows increasing
and decreasing survey unit size based on classification. In MARSSIM, Class 1 survey units generally are smaller
than the area assumed in the exposure pathway model, while MARSSIM allows Class 3 survey units to be larger in
area. Additional modifications to survey unit boundaries in MARSSIM can be made based on site-specific variables
(e.g., room size, topography).
January 2009 3-17 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
• Worker processing concrete rubble at a satellite facility,
• Truck driver hauling concrete rubble,
• Worker building a road using recycled concrete,
• Driver on a road built using recycled concrete,
• Worker handling concrete rubble at an industrial landfill,
• Worker handling concrete rubble at a municipal landfill,
• Individual drinking groundwater contaminated with leachate from an industrial landfill, and
• Individual drinking groundwater contaminated with leachate from a municipal landfill.
Each exposure scenario assumes different conditions that help define survey unit boundaries. For
example, a truck driver hauling concrete rubble would be exposed to one truckload of concrete
rubble, so the survey unit boundaries would be defined by a truckload of concrete rubble (i.e.,
2xl04 kg [22 tons] or 8.3 m3; NRC 2003a).
3.6.2 Modify Initial Survey Unit Boundaries
Modifications to survey unit boundaries are expected based on practical constraints for data
collection activities. In most cases smaller survey units will be acceptable, since a reduction in
size would not result in an increased dose or risk. Increasing the size of the survey unit may
result in increased dose or risk, and therefore requires approval of the planning team and
stakeholders.
Constraints on collecting data are often associated with specific measurement techniques, which
could affect the survey unit boundaries. For example, using in situ gamma spectroscopy may
restrict survey unit sizes based on the field of view of the detector, the penetrating power of the
gamma energies being measured, or the assumptions used to develop the instrument efficiency.
Alternatively, using a box counter or portal monitor may restrict survey unit sizes based on what
will fit inside or through the detector. Information on measurement parameters affecting
disposition survey design is provided in Section 3.8. Section 5.9 and Appendix D provide
detailed information on specific measurement methods.
The M&E being investigated may also cause modifications to survey unit boundaries. These
modifications are often associated with physical characteristics (e.g., size, shape). Identification
of actual survey units as part of the final disposition survey design is discussed in Chapter 4.
3.7 Develop a Decision Rule
In order to design a disposition survey, the user should define a decision rule describing the
conditions for selecting between alternative actions. The planning team should assume that ideal
data are available and there is no uncertainty in the decision making process. The available data
are integrated into an "if...then..." statement, which is the theoretical decision rule.7
7 This is called a theoretical decision rule because it is stated in terms of the true value for the parameter of interest,
even though in reality this value cannot be known. An operational decision rule that is based on an estimate of the
target population parameter of interest will be incorporated as part of the final disposition survey design selected and
documented in Chapter 4.
NUREG-1575, Supp. 1 3-18 January 2009
-------
MARSAME Identify Inputs to the Decision
The theoretical decision rule is constructed by combining the action level (Section 3.3) and the
parameter of interest (Section 3.4) with the alternative actions (Section 3.5) in an "if...then..."
statement.
For example:
Hypothetically, if the mean concentration of 226Ra in 20,000 kg (8.3 m3, one
truckload) of concrete rubble is less than the clearance action level of 0.34 Bq/g
for volumetric radioactivity, then the concrete rubble can be cleared, otherwise
radiological control of the concrete will continue.
It may be necessary to develop more than one decision rule. For example, if more than one
action level is selected in Section 3.3, a separate decision rule needs to be developed for each
action level. In addition, selection of multiple disposition options in Section 2.5 (e.g., release and
disposal as low-level radioactive waste) may result in multiple alternative actions requiring
multiple decisions and multiple decision rules. For example,
Hypothetically, if the mean concentration of 226Ra in 20,000 kg (8.3 m3, one
truckload) of concrete rubble is less than the clearance action level of 0.34 Bq/g
for volumetric radioactivity, then the concrete rubble can be cleared, otherwise
the concrete will be considered for disposal as low-level radioactive waste. If the
concrete rubble meets the waste acceptance criteria for the low-level radioactive
waste disposal facility (e.g., mean and total activity levels, chemical and physical
form, toxicity) the concrete will be packaged and transported for disposal,
otherwise radiological control of the concrete will continue.
3.8 Develop Inputs for Selection of Provisional Measurement Methods
The identification and evaluation of provisional measurement methods is an important step in
developing a disposition survey design. A measurement method is the combination of
instrumentation (e.g., GM detector, NaI[Tl] scintillation detector, gamma spectrometer) with a
measurement technique (i.e., scan, in situ, sample collection). The selection of a measurement
method is discussed in more detail in Section 5.9. The availability of measurement methods and
the amount of resources required to implement specific measurement methods is an important
factor in selecting between different survey designs, or in reducing the number of options to be
considered when developing potential disposition survey designs.
There are two potential results of this evaluation of provisional measurement methods. First, the
evaluation may identify specific measurement methods that will be included in the final
documentation of the selected disposition survey design (see Section 4.5). For example, scanning
100% of a piece of equipment using a 2-inch by 2-inch Nal(Tl) detector at a specified height
above the surface using a specified scan speed may be identified as the measurement method.
Second, the evaluation may identify characteristics of a measurement method required to meet
the objectives of a survey. These characteristics are called measurement quality objectives
(MQOs). Section 5.5 and Section 7.3 provide additional information on MQOs applied to
disposition surveys.
January 2009 3-19 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
Examples of MQOs are described in the following sections. A list of minimum MQOs required
for a survey can be developed and documented in the final disposition survey design (see Section
4.5). The selection of a measurement technique that meets the MQOs is accomplished during
implementation of the survey design.
This section focuses on measurability. Most of the variables that need to be considered for the
identification of measurement techniques have been discussed earlier in this chapter. The
identification of measurement methods is directly or indirectly related to—
• Identification of radionuclides of concern,
• Location of residual radioactivity,
• Application of action levels,
• Physical properties of the M&E,
• Distribution of residual radioactivity,
• Expected levels of residual radioactivity,
• Relationships between radionuclide activities,
• Equilibrium status of natural decay series, and
• Background radioactivity.
Measurable radioactivity is radioactivity that can be quantified and meets the DQOs and MQOs
established for the survey. Radioactivity that is quantified using known or predicted relationships
developed from process knowledge or preliminary measurements is considered measurable as
long as the relationships are developed and verified as specified in the DQOs and MQOs. The
Multi-Agency Radiological Laboratory Analytical Protocols manual (MARLAP 2004)8 lists
method performance characteristics that should be considered when establishing MQOs for a
project. This list is not intended to be exhaustive:
• The method uncertainty at a specified concentration (expressed as a standard deviation);
• The method's detection capability (expressed as the minimum detectable concentration, or
MDC);
• The method's quantification capability (expressed as the minimum quantifiable
concentration, or MQC);
• The method's range, which defines the method's ability to measure the radionuclide of
concern over some specified range of concentration;
• The method's specificity, which refers to the ability of the method to measure the
radionuclide of concern in the presence of interferences; and
• The method's ruggedness, which refers to the relative stability of method performance for
small variations in method parameter values.
Project-specific method performance characteristics should be developed as necessary and may
or may not include the characteristics listed here. Once lists of performance characteristics that
affect measurability have been identified, the planning team should develop MQOs describing
8 MARLAP was developed for selecting laboratory protocols. Applying the framework and performance-based
approach for planning and conducting radiological work from MARLAP to the selection of field measurement
techniques is an expansion of the original scope and purpose of MARLAP.
NUREG-1575, Supp. 1 3-20 January 2009
-------
MARSAME Identify Inputs to the Decision
the project-specific objectives for potential measurement techniques. Potential measurement
techniques should be evaluated against the MQOs to determine if they are capable of meeting the
objectives for measurability.
3.8.1 Measurement Method Uncertainty
The required measurement method uncertainty is perhaps the most important MQO to be
established during the planning process. Section 4.2 discusses the rationale involved in setting
the required measurement method uncertainty and developing statistical hypothesis tests for the
implementation of disposition decision rules using measurement data. Section 5.5 discusses the
application of MQOs, including the measurement method uncertainty, to disposition surveys for
M&E. Section 7.3 discusses procedures for determining the required measurement method
uncertainty and whether or not it has been achieved.
MARLAP uses the term method uncertainty to refer to the predicted uncertainty of a measured
value that would likely result from the performance of a measurement at a specified
concentration, typically the action level. Reasonable values for method uncertainty can be
predicted for a particular measurement technique based on typical values for specific parameters
(e.g., count time, efficiency) and process knowledge for the M&E being investigated (see
Sections 5.5 and 7.3). The MQO for measurement method uncertainty is related to the width of
the gray region (Section 4.2.2). The required measurement method uncertainty is directly related
to the MDC and the MQC discussed below.
The distinction between imprecision and bias as a data quality indicator depends on context.
Additional information on data quality indicators can be found in MARS SIM Appendix N and
EPA QA/G-5 (EPA 2002a). A reliable estimate of bias requires a data set that includes many
measurements, so MARSAME and MARLAP focus on developing an MQO for measurement
method uncertainty. Measurement method uncertainty effectively combines imprecision and bias
into a single parameter whose interpretation does not depend on context. This approach assumes
that all potential sources of bias present in the measurement process have been considered in the
estimation of the measurement uncertainty and, if not, that any appreciable bias would only be
detected after a number of measurements of quality control (QC) and performance evaluation
samples have been performed (see the QC discussion in Section 5.10). MARLAP Appendix C
provides examples on developing MQOs for measurement method uncertainty of laboratory
measurement techniques.
3.8.2 Detection Capability
The MDC (see Sections 5.7 and 7.5) is recommended as the MQO for defining the detection
capability, and is an appropriate MQO when decisions are to be made based on a single
measurement as to whether excess radioactivity is present or not. The MDC must not exceed the
action level if the MDC is to be used as a decision parameter. Chapter 5 provides guidance on
implementation of the selected measurement technique, including calculation of the MDC.
Additional information on calculating the MDC can be found in MARS SIM (Section 6.7,
examples in Appendix H) and MARLAP (Chapter 19, Appendix C).
January 2009 3-21 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
3.8.3 Quantification Capability
When the average of several measurements will be compared to a disposition criterion, an MQO
more stringent than the MDC is required. The MQC (see Sections 5.8 and 7.6) is recommended
as the parameter for defining the measurement capability for making quantitative comparisons of
averages to a limit. An MQO for the required measurement method uncertainty (Section 5.6) is
related to an MQO for the quantification capability because an MQC is defined as the
concentration at which a specified relative standard uncertainty is achieved. MARLAP presents
three reasons why it is important to consider this measurement method performance
characteristic:
1. To emphasize the importance of the quantification capability of a measurement technique for
instances when the issue is not whether a radionuclide is present or not (e.g., measuring 238U
in soil where the activity is inherent) but rather how precisely the radionuclide can be
measured,
2. To promote the MQC as an important measurement method performance characteristic for
comparison of measurement techniques, and
3. To provide an alternative to the overemphasis on establishing required MDCs in instances
where detection (i.e., reliably distinguishing a radionuclide concentration from zero) is not
the key analytical issue.
The MQC must not exceed the action level if the MQC is to be used as a decision parameter.
Chapter 5 provides guidance on implementation of the selected measurement technique,
including calculation of the MQC. Section 5.8 discusses issues related to measurement
quantifiability. Section 7.6 provides information on the statistical basis of the MQC calculation
including example calculations. Additional information on calculating the MQC can be found in
MARLAP Chapter 19, with examples in MARLAP Appendix C.
3.8.4 Range
The expected concentration range for a radionuclide of concern (see Section 2.4.2) may be an
important measurement method performance characteristic. Most radiation measurement
techniques are capable of measuring over a wide range of radionuclide concentrations. However,
if the expected concentration range is large, the range should be identified as an important
measurement method performance characteristic and an MQO should be developed. The MQO
for the acceptable range should be a conservative estimate. This will help prevent the selection of
measurement techniques that cannot accommodate the actual concentration range.
3.8.5 Specificity
Specificity is the ability of the measurement method to measure the radionuclide of concern in
the presence of interferences. To determine if specificity is an important measurement method
performance characteristic, the planning team will need information on expected concentration
ranges for the radionuclides of concern and other chemical and radionuclide constituents, along
with chemical and physical attributes of the M&E being investigated (see Section 2.4). The
importance of specificity depends on—
NUREG-1575, Supp. 1 3-22 January 2009
-------
MARSAME Identify Inputs to the Decision
• The chemical and physical characteristics of the M&E being investigated,
• The chemical and physical characteristics of the residual radioactivity, and
• The expected concentration range for the radionuclides of concern.
If potential interferences are identified (e.g., inherent radioactivity, similar radiations), an MQO
should be established for specificity.
If inherent radioactivity is associated with the M&E being investigated, a method that measures
total activity may not be acceptable. Consider concrete, which contains measurable levels of
naturally occurring radioactivity and emits radiation in the form of alpha particles, beta particles,
and photons. If the action level for the radionuclide of concern is close to background (e.g.,
within a factor of 3) gross measurement methods may not meet the survey objectives.
Performing gross alpha measurements using a gas proportional detector may not provide an
acceptable MDC or MQC for plutonium isotopes, where a more specific measurement method
such as alpha spectrometry following radiochemical separation would be acceptable.
Radionuclides have similar radiations if they emit radiations of the same type (i.e., alpha, beta,
photon) with similar energies. For example, both 226Ra and 235U emit a gamma ray with energy
of approximately 186 keV. Gamma spectroscopy may not be able to resolve mixtures of these
two radionuclides, which are both associated with naturally occurring radioactivity. More
specific methods involving ingrowth of 226Ra decay products or chemical separation prior to
measurement can be used to accurately quantify the radionuclides.
Documented measurement methods should include information on specificity. MARSSEVI Table
7.2 lists examples of references providing laboratory measurement methods. NUREG-1506
(NRC 1995) provides generic information on field measurement techniques, but most field
measurement methods are documented in proprietary SOPs. If specificity is identified as an
important issue for a project, consultation with an expert in radiometrics or radiochemistry is
recommended.
3.8.6 Ruggedness
For a project that involves field measurements that are performed in hostile, hazardous, or
variable environments, or laboratory measurements that are complex in terms of chemical and
physical characteristics, the measurement method's ruggedness may be an important method
performance characteristic. Ruggedness refers to the relative stability of the measurement
technique's performance when small variations in method parameter values are made. For field
measurements the changes may include temperature, humidity, or atmospheric pressure. For
laboratory measurements, a change in pH or the quantity of a reagent may be important. In order
to determine if ruggedness is an important measurement method performance characteristic, the
planning team needs detailed information on the chemical and physical characteristics of the
M&E being investigated and operating parameters for the radiation instruments used by the
measurement technique. Information on the chemical and physical characteristics of the M&E is
available as outputs from the IA. Information on the operating parameters for specific
instruments should be available from the instrument manufacturer. Generic information for
radiation detector operating parameters may be found in consensus standards. A limited list of
examples of consensus standards is provided in Table 3.3.
January 2009 3-23 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
Table 3.3 Examples of Consensus Standards for Evaluating Ruggedness
Standard Number
ANSI N42. 12- 1994
ANSIN42.17A-2003
ANSIN42.17C-1989
ANSI N42.34-2003
IEEE 309- 1999/
ANSI N42.3- 1999
ASTM El 169-2002
Title
American National Standard Calibration and Usage of Thallium-Activated
Sodium Iodide Detector Systems for Assay of Radionuclides
American National Standard Performance Specifications for Health Physics
Instrumentation - Portable Instrumentation for Use in Normal Environmental
Conditions
American National Standard Performance Specifications for Health Physics
Instrumentation - Portable Instrumentation for Use in Extreme
Environmental Conditions
American National Standard Performance Criteria for Hand-held Instruments
for the Detection and Identification of Radionuclides
Institute of Electrical and Electronics Engineers, Inc. Standard Test
Procedures and Bases for Geiger Mueller Counters
Standard Guide for Conducting Ruggedness Tests
If it is determined that measurement method ruggedness is an important performance
characteristic, an MQO should be developed. The MQO may require performance data that
demonstrate the measurement technique's ruggedness for specified changes in select
measurement method parameters. Alternatively, the MQO could list the acceptable ranges for
select measurement method parameters and monitor the parameters as part of the QC program
for the project (Section 5.10). For example, sodium iodide detectors are required to perform
within 15% of the calibrated response between 0 and 40 °C (32 and 104 °F, respectively) (ANSI
1994). The disposition survey design may call for a work stoppage at temperatures outside this
range, or an increase in the frequency of QC measurements at temperatures outside this range.
3.9 Identify Reference Materials
Action levels may be developed that are related to background radioactivity, either based on an
incremental dose or risk above background, as an administrative limit based on background, or
as a limit on technology (e.g., minimum detectable concentration). For situations where the
action levels are incremental above background, reference materials should be identified to
provide an estimate of background. MARSSEVI Section 4.5 provides guidance on determining
when a reference material is required.
Reference materials are used to develop an estimate of the distribution of background
radioactivity that can be compared to the measurements performed in a comparable survey unit.
The reference material is selected to provide information on the level of radioactivity that would
be present if the M&E being investigated had not been radiologically impacted.
Whenever possible, reference data should be obtained by performing a survey of the M&E
before it comes in contact with radiological materials. The M&E can then be surveyed prior to
leaving the area to determine the level of residual radioactivity. This works especially well for
decommissioning or cleanup applications where M&E are brought into a radiologically
NUREG-1575, Supp. 1 3-24 January 2009
-------
MARSAME Identify Inputs to the Decision
controlled area for a limited time and a specific application. Unfortunately, there are numerous
situations where pre-contact surveys are not possible.
If the M&E cannot be used as its own reference material, it is necessary to identify reference
material that is representative of the M&E being investigated. Non-impacted M&E that closely
resembles the impacted M&E being investigated (i.e., similar chemical, physical, and
radiological characteristics) will generally be acceptable as reference material. For example, if
the conceptual model shows that only surficial activity is expected, the impacted surface may be
removed and the non-impacted volume used as the reference material. When similar materials
are not available, the best match available should be used as reference material. It may be
necessary to evaluate more than one source of reference material before an acceptable match is
identified. It may be important to perform reference material surveys in areas of low ambient
background. Consider M&E consisting of individual objects that are small relative to the size of
the detector used to perform the measurements. When each object receives a separate
measurement, the ambient background may have a larger impact on the measurement than the
background contributed by the M&E itself.
As shown in Table B. 1 in Appendix B, background radionuclide concentrations for materials can
vary significantly. For example, concentrations for thorium series radionuclides in concrete can
range from 15 to 120 Bq/kg (Eicholz 1980), so it is important to identify an appropriate reference
material.
The planning team should understand that background is variable. Ambient background can
change with location and over time. It may be possible to simply move the M&E being
investigated to an area with a lower ambient background to improve the detection capability of a
measurement method. Local conditions (e.g., temperature, barometric pressure, precipitation)
can cause variations in ambient background as discussed in NUREG-1501 (NRC 1994).
NUREG-1505 (NRC 1998a) Chapter 13 provides information on accounting for variability in
background.
The planning team should evaluate the process knowledge from the IA and use professional
judgment to identify M&E that require reference materials, and identify potential reference
materials to support the disposition survey.
3.10 Evaluate an Existing Survey Design
It is not necessary to develop a new survey design for all M&E being investigated. Existing
survey designs are often available for routine or repetitive applications. If an existing survey
design is identified, the planning team or decision maker should evaluate the applicability of the
existing design to the current investigation.
Standardized survey designs for operating facilities are often documented in the form of standard
operating procedures (SOPs, see Section 4.5.1). In other cases, existing survey designs may have
been developed for similar projects. A description of the M&E that can be measured should be
included in each existing SOP or survey design. If the description matches the M&E being
investigated, the existing SOP or survey design can be used to perform the disposition survey. If
January 2009 3-25 NUREG-1575, Supp. 1
-------
Identify Inputs to the Decision MARSAME
the description of the M&E is incomplete or vague, or the M&E do not match the description, a
more detailed evaluation may be performed to determine the acceptability of the existing survey
design.
Personnel familiar with the existing survey design and the proposed application should perform
the detailed evaluation of an existing survey design. All supporting documentation used to
develop the existing survey design should be available for the evaluator(s), not just the SOP or
survey design being reviewed.
The detailed evaluation should determine whether the M&E are measurable using the existing
survey design. If the M&E are measurable, the existing survey design can be used. Detailed
evaluations should include a review of each step in the survey development process, including-
• Selection of a disposition option (Section 2.5),
• Identification of action levels (Section 3.3),
• Specification of the population parameter of interest (Section 3.4),
• Development of survey unit boundaries (Section 3.6),
• Selection of measurement methods (Section 3.8 and Section 5.9),
• Identification of alternative actions (Section 3.5), and
• Development of a decision rule (Section 3.7 and Section 4.2.6).
The results of the evaluation should be documented. The documentation may require a
modification to the existing survey design. For example, the description of M&E that can (or
cannot) be measured using a specific SOP may be expanded for M&E that are routinely or
repeatedly surveyed. Alternatively, the documentation may consist of a notation in a survey log
(including a name, title, and date) for unique items.
NUREG-1575, Supp. 1 3-26 January 2009
-------
MARSAME Develop A Survey Design
4 DEVELOP A SURVEY DESIGN
4.1 Introduction
Once a decision rule has been developed, a disposition survey can be designed for the impacted
materials and equipment (M&E) being investigated. The disposition survey incorporates all of
the available information to determine the quantity and quality of data required to support a
disposition decision. This chapter provides information on selecting the type, number, and
location of measurements required to support a decision regarding the disposition of the M&E.
Facilities or installations can use the process in this chapter and following chapters to develop a
standard operating procedure (SOP) so multiple surveys can be performed for similar M&E to
avoid costly and time-consuming development of redundant survey designs. The evaluation of
existing SOPs for usability is discussed in Section 3.10. The output from this chapter is a
documented disposition survey design that integrates measurement, data collection, and data
analysis techniques.
The information in this chapter builds on the information collected and decisions made in
Chapters 2 and 3. The disposition option selected in Section 2.5 and the action levels (ALs)
identified in Section 3.3 are incorporated into the decision rules developed in Section 3.7. A
decision rule is the basis for the disposition survey design. If multiple survey designs address the
same decision rule and meet the data quality objectives (DQOs), the decision-maker needs to
determine the most effective design for that decision rule. If none of the survey designs meet the
DQOs for a specific decision rule, it may be necessary to reconsider decisions made earlier in the
survey design process and adjust the DQOs.1 If there are multiple decision rules (e.g., one for
total radioactivity and one for removable radioactivity) more than one survey design may need to
be developed to meet all of the DQOs for the project or a single survey design may be developed
to incorporate all of the decision rules.
The complexity of a survey design generally reflects the complexity of the statistics used to
interpret the results (Chapter 6). Survey designs range from simple (e.g., scan 100% of the M&E
for surface radioactivity at a specified AL) to complex (e.g., develop a MARSSIM-type survey
design). Simple survey designs typically require few resources for planning, but may require
significant resources to implement. Complex survey designs typically require more resources
during planning, with fewer resources required during implementation. If the planning and
implementation portions of the data life cycle are performed correctly, the assessment and
decision making stages should require few resources. This chapter provides information on
statistical decision-making and how it is used during development of survey designs.
4.2 Making Decisions Using Statistics
In Section 3.6, the planning team assumed the levels and distribution of radioactivity associated
with the M&E were known with no uncertainty. A theoretical decision rule was developed using
this assumption to help focus the attention of the planning team on how they would make
1 Refer to Section 2.3 for information on performing preliminary surveys to help ensure at least one survey design
will meet the DQOs.
January 2009 4-1 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
decisions. In this chapter the planning team accounts for uncertainty in decisions when ideal data
are not available by establishing a statistical test to implement the decision rule. Decisions
regarding the disposition of M&E are based on data with uncertainties. Through the use of
statistics, the disposition survey design attempts to control the probability of making a decision
error because of these uncertainties. MARSSIM Section 2.3 provides additional discussions on
the use of statistics for making decisions based on environmental data. These steps are discussed
briefly below and in further detail in Section 7.1. MARSAME recommends the planning team
complete the following steps:
• Select a null hypothesis (Section 4.2.1),
• Choose a discrimination limit (Section 4.2.2),
• Define Type I and Type II decision errors (Section 4.2.5),
• Set a tolerable Type I decision error rate at the action level (Section 4.2.5), and
• Set a tolerable Type II decision error rate at the discrimination limit (Section 4.2.5).
4.2.1 Null Hypothesis
In hypothesis testing, two assertions about the actual level of radioactivity associated with the
M&E are formulated. The two assertions are called the null hypothesis (Ho) and the alternative
hypothesis (Hi). H0 and HI together describe all possible radionuclide concentrations or levels of
radioactivity under consideration. The survey data are evaluated to choose which hypothesis to
reject or not reject, and by implication which to accept.2 In any given situation, one and only one
of the hypotheses must be true. The null hypothesis is assumed to be true within the established
tolerance for making decision errors (Section 4.2.5). Thus, the choice of the null hypothesis also
determines the burden of proof for the test.
4.2.2 Discrimination Limit
Action levels were defined in Section 3.3 based on the selected disposition option and applicable
regulatory requirements. The planning team also chooses another radionuclide concentration or
level of radioactivity that can be reliably distinguished from the action level by performing
measurements (i.e., direct measurements, scans, in situ measurements, samples and laboratory
analyses). This radionuclide concentration or level of radioactivity is called the discrimination
limit (DL). An example where the discrimination limit is defined is provided in Section 8.4.5.
The gray region is defined as the interval between the action level and the discrimination limit
(Figures 4.1, 4.2, and 4.3 provide visual descriptions of the gray region). The width of the gray
region is called the shift and denoted as A. The objective of the disposition survey is to decide
whether the concentration of radioactivity is more characteristic of the DL or of the AL, i.e.,
whether action should be taken, or if action is not necessary. The width of the gray region
expressed as a multiple of the measurement standard deviation, cr, is called the relative shift, ACT.
Survey effort will increase as the relative shift decreases.
2 In hypothesis testing, to "accept" the null hypothesis only means not to reject it. For this reason many statisticians
avoid the word "accept." A decision not to reject the null hypothesis does not imply the null hypothesis has been
shown to be true.
NUREG-1575, Supp. 1 4-2 January 2009
-------
MARSAME
Develop A Survey Design
In Figure 4.1, it may be seen that a large o can be tolerated because A is large enough that the
resulting relative shift A/a is large. In Figure 4.2, even though o is small, either more accuracy or
more samples are needed because A/a is also small.
o
I
o
o
O
O
c
a>
3
Mean
Action
Level
Concentration
Figure 4.1 Relative Shift, A/a, Comparison for Scenario A:
a is Large, but the Large A Results in a Large A/a and Fewer Samples
o
I
o
o
O
o
c
a>
Mean
Action
Level
Concentration
Figure 4.2 Relative Shift, A/a, Comparison for Scenario A:
a is Small, but the Small A Results in a Small A/a and More Samples
January 2009
4-3
NUREG-1575, Supp.l
-------
Develop A Survey Design
MARSAME
4.2.3 Scenario A
The null hypothesis for Scenario A specifies that the radionuclide concentration or level of
radioactivity associated with the M&E is equal to or exceeds the action level. For Scenario A
(H0: X > AL), the upper bound of the gray region (UBGR) is equal to the AL and the lower
bound of the gray region (LBGR) is equal to the DL. As a general rule for applying Scenario A,
the DL should be set no higher than the expected radionuclide concentration associated with the
M&E. The DL and the AL should be reported in the same units. Figure 4.3 illustrates Scenario
A.
LBGR
UBGR
Gray Region
Discrimination Limit
Activity Level
Action Level
Figure 4.3 Illustration of Scenario A
4.2.4 Scenario B
The null hypothesis for Scenario B specifies the radionuclide concentration or level of
radioactivity associated with the M&E is less than or equal to the action level. For Scenario B
(H0: X < AL), the UBGR is equal to the DL and the LBGR is equal to the AL. The DL defines
how hard the surveyor needs to look, and is determined through negotiations with the regulator.3
In some cases the DL will be set equal to a regulatory limit (e.g., 10 CFR 36.57 and DOE 1993).
The DL and the AL should be reported in the same units. Figure 4.4 illustrates Scenario B. This
description of Scenario B is based on information in MARLAP and is fundamentally different
from the description of Scenario B in NUREG-1505 (NRC 1998a).
LBGR
UBGR
Gray Region
Action Level
Discrimination Limit
Activity Level
Figure 4.4 Illustration of Scenario B
1 In some cases, setting the discrimination limit may include negotiations with stakeholders.
NUREG-1575, Supp. 1
4-4
January 2009
-------
MARSAME Develop A Survey Design
In NUREG-1505 (NRC 1998a) the gray region is defined to be below the AL in both Scenario A
and Scenario B. In MARSAME and MARLAP the gray region is defined to be above the AL in
Scenario B. The difference lies in how the action level is defined.
4.2.5 Specify Limits on Decision Errors
There are two possible types of decision errors:
• Type I error: rejecting the null hypothesis when it is true, and
• Type II error: failing to reject the null hypothesis when it is false.
Because there is always uncertainty associated with the survey results, the possibility of decision
errors cannot be eliminated. So instead, the planning team specifies the maximum Type I
decision error rate (a) that is allowable when the radionuclide concentration or level of
radioactivity is at or above the action level. This maximum usually occurs when the true
radionuclide concentration or level of radioactivity is exactly equal to the action level. The
planning team also specifies the maximum Type II decision error rate (/?) that is allowable when
the radionuclide concentration or level of radioactivity equals the discrimination limit.
Equivalently, the planning team can set the "power" (I-/?) when the radionuclide concentration
or level of radioactivity equals the discrimination limit. See MARSSIM Appendix D, Section
D.6 for a more detailed description of error rates and statistical power.
It is important to clearly define the scenario (i.e., A or B) and the decision errors for the survey
being designed. Once the decision errors have been defined, the planning team should determine
the consequences of making each type of decision error. For example, incorrectly deciding the
activity is less than the action level may result in increased health and ecological risks.
Incorrectly deciding the activity is above the action level when it is actually below may result in
increased economic and social risks. The consequences of making decision errors are project
specific.
4.2.6 Develop an Operational Decision Rule
The theoretical decision rule developed in Section 3.6 was based on the assumption that the true
radionuclide concentrations in the M&E were known. Because the disposition decision will be
made based on measurement results and not the true but unknown concentration, an operational
decision rule needs to be developed to replace this theoretical decision rule. The operational
decision rule is a statement of the statistical hypothesis test, which is based on comparing some
function of the measurement results to some critical value. The theoretical decision rule is
developed during Step 5 of the DQO process (Chapter 3), while the operational decision rule is
developed as part of Step 6 and Step 7 of the DQO process. For example, a theoretical decision
rule might be "if the results of any measurement identify surface radioactivity in excess of
background, the front loader will be refused access to the site; if no surface radioactivity in
excess of background is detected, the front loader will be granted access to the site." The related
operational decision rule might be "any result that exceeds the critical value associated with the
MDC set at the discrimination limit will result in rejection of the null hypothesis, and the front
loader will not be allowed on the site" (see more examples in Chapter 7).
January 2009 4-5 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
4.3 Classify the Materials and Equipment
Classification is used to determine the level of survey effort for the disposition survey. The level
of survey effort is linked to the potential to exceed the action level(s) (i.e., classification), and is
a graded approach to survey design. Impacted M&E with the highest potential to exceed the
action level(s) (i.e., Class 1) receive the greatest effort for the disposition survey, while M&E
with a lower potential to exceed the action level(s) (i.e., Class 2 or Class 3) require less survey
effort. Classification in MARSAME is analogous to classification in MARSSIM. The planning
team needs to remember that classification is based on estimated radionuclide concentrations or
radioactivity relative to the AL.
There are tradeoffs (costs and benefits) associated with classification based on estimated4 or
known radionuclide concentrations or levels of radioactivity relative to the action levels. This
means that some knowledge of radionuclide concentrations is required before M&E can be
classified. Known radionuclide concentrations or levels of radioactivity may be available from
historical data identified during the initial assessment (IA; see Section 2.2), or performance of
preliminary surveys (Section 2.3). Estimates of radionuclide concentrations can be developed
based on historical data or process knowledge (Section 2.2). In the absence of information on the
radionuclide concentrations, the default assumption is that all impacted M&E are Class 1.
Because classification of impacted M&E is based in part on an action level, classification cannot
be performed until potential action levels have been identified (Section 3.3). For projects where
multiple potential action levels have been identified, classification and selection of an
appropriate action level may be an iterative process used to reduce the number of survey options.
Alternatively, multiple survey designs can be developed to address all potential action levels. In
the final step of the DQO process the most resource efficient survey design that meets the survey
objectives is selected (Section 4.4.4).
4.3.1 Class 1
Class 1 M&E are impacted M&E that have, or had, the following: (1) highest potential for, or
known, radionuclide concentration(s) or radioactivity about the action level(s); (2) highest
potential for small areas of elevated radionuclide concentration(s) or radioactivity; and (3)
insufficient evidence to support reclassification as Class 2 M&E or Class 3 M&E. Such potential
may be based on historical information and process knowledge, while known radionuclide
concentration(s) or radioactivity may be based on preliminary surveys. This class of M&E might
consist of processing equipment, components, or bulk materials that may have been affected by a
liquid or airborne release, including, for example, inadvertent effects from spills.
Class 1 M&E are those that may have been in direct contact with radioactive materials during
operations or may have become activated and are likely to exceed the action level. Additionally,
M&E that have been cleaned to remove residual radioactivity above the action level generally
are considered to be Class 1. An exception to Class 1 classification may be considered if there
4 There are risks and tradeoffs associated with using estimated values. The planning team should compare the
consequences of potential decision errors with the resources required to improve the quality of existing data to
determine the appropriate approach for a specific project.
NUREG-1575, Supp. 1 4-6 January 2009
-------
MARSAME Develop A Survey Design
are no difficult-to-measure areas and any residual radioactivity is readily removable using
cleaning techniques. Examples of such methods may include vacuuming, wipe downs, or
chemical etching that quantitatively remove sufficient amounts of radionuclides such that
surficial activity levels would be less than the release criteria. Documented process knowledge of
cleaning methods directly applicable to the particular M&E should be provided to justify this
exception.
4.3.2 Class 2
Class 2 M&E are impacted M&E that have, or had, (1) low potential for radionuclide
concentration(s) or radioactivity above the action level(s); and (2) little or no potential for small
areas of elevated radionuclide concentration(s) or radioactivity. Such potential may be based on
historical information, process knowledge, or preliminary surveys. This class of materials might
consist of electrical panels, water pipe, conduit, ventilation ductwork, structural steel, and other
materials that might have come in contact with radioactive materials. Radionuclide
concentration(s) and radioactivity above the action level, including small areas of elevated
radionuclide concentration(s) or radioactivity, are not expected in Class 2 M&E.
4.3.3 Class 3
Class 3 M&E are impacted M&E that have, or had, (1) little, or no, potential for radionuclide
concentration(s) or radioactivity above background; and (2) insufficient evidence to support
categorization as non-impacted. Radionuclide concentration(s) and radioactivity above a
specified small fraction of the UBGR are not expected in Class 3 M&E. The specified fraction
should be developed by the planning team using a graded approach and approved by the
regulatory authority.
4.3.4 Other Classification Considerations
The planning team should review any historical data used to provide information on radionuclide
concentrations or radioactivity and evaluate whether or not the data meet the objectives of the
disposition survey, as illustrated in the following examples. Representativeness (see MARSSEVI
Appendix N) is a key data quality indicator when evaluating historical data. Ideally, the IA
should provide information on the radionuclides of potential concern, expected radionuclide
concentrations or radioactivity, distribution of radioactivity, and locations where radioactivity is
expected (e.g., surficial or volumetric, see Section 2.4.3). In addition, the data should meet the
criteria for measurability (e.g., MQC) or detectability (e.g., MDC) established for the project (see
Sections 3.8 and 5.5). Historical data that do not meet the objectives of the disposition survey
may still be used to provide estimates for radionuclide concentrations or levels of radioactivity.
The results of the IA may provide estimated radionuclide concentrations or levels of
radioactivity based on process knowledge, historical data, sentinel measurements, or preliminary
surveys. In some cases, a survey is performed to develop adequate estimates for levels and
variability of radionuclide concentrations or radioactivity. Again, the planning team should
evaluate the data used to develop the estimated radionuclide concentrations or levels of
radioactivity. In general, estimated data will have a higher associated uncertainty than known
data that meet the objectives of the project. The planning team should keep this in mind when
January 2009 4-7 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
developing estimates for radionuclide concentrations or radioactivity to be used in classifying
M&E.
If the action level is defined in terms of average activity, the average radionuclide concentration
or radioactivity should be compared to the action level to determine the appropriate
classification. Similar comparisons should be developed for action levels provided in terms of
maximum activity or total activity. For example, DOE Order 5400.5 (DOE 1993) provides three
surface activity action levels for each group of radionuclides: average total surface activity,
maximum total surface activity, and maximum removable surface activity. These action levels
must be evaluated prior to disposition of the M&E. Classification would be determined by
comparing the average total surface activity, maximum total surface activity, and maximum
removable surface activity (or appropriate conservative estimates) to the corresponding action
level. The overall classification would be determined by the most restrictive case. If the
maximum total surface activity indicates the M&E is Class 1, while the maximum removable
surface activity indicates the M&E is Class 3, the M&E should be classified as Class 1.
The improper classification of M&E has serious implications, particularly when it leads to the
release of material with residual radioactivity in excess of the AL. For example, if material were
mistakenly thought to have a very low potential for having residual radioactivity, the material
will be subjected to a survey with lesser scrutiny. This misclassification might result in releasing
material that should not be released. The opposing possibility (i.e., when M&E is misclassified
as impacted when it is non-impacted) involves the stakeholders expending potentially substantial
resources involved in unnecessarily surveying non-impacted M&E.
4.4 Design the Disposition Survey
MARSAME recommends design of disposition surveys that measure 100% of the M&E being
investigated whenever practical. This includes survey designs where all of the M&E are
physically measured. Survey designs where physical measurements are performed for less than
100% of the M&E may be acceptable if the radioactivity is measurable. Measurable radioactivity
is radioactivity that can be quantified and meets the DQOs and MQOs established for the survey.
Radioactivity that is quantified using known or predicted relationships developed from process
knowledge, historical data, sentinel measurements, or preliminary measurements is considered
measurable as long as the relationships are developed and verified as specified in the DQOs and
MQOs. An example of such a relationship could be the immobile progeny of the measured
radionuclides.
Survey designs that measure 100% of the M&E being investigated reduce the uncertainty in the
final decision. Because 100% of the M&E are measured, for practical purposes spatial variability
can be ignored. Attention should be given to ensure that all impacted surfaces are measured in
100% scan surveys. Surveys that use known or predicted relationships to estimate radionuclide
concentrations or levels of radioactivity need to account for the contribution of spatial variability
to total uncertainty.
To make the best use of limited resources, MARSAME places the greatest level of survey effort
on M&E that have, or had, the greatest potential for residual radioactivity (i.e., Class 1). This is
NUREG-1575, Supp. 1 4-8 January 2009
-------
MARSAME Develop A Survey Design
referred to as a graded approach. As noted in Section 1.3, survey designs that measure 100% of
the M&E are often neither practical nor cost-effective, and could drive the user to dispose of any
material that is potentially impacted without considering the benefits of reuse or recycle. The use
of a graded approach to ensure that a sensible, commensurate balance is achieved between cost
and risk reduction should always be incorporated into MARSAME survey designs. The
following sections describe the four basic disposition-survey designs:
• Scan-only survey designs (Section 4.4.1),
• In situ survey designs (Section 4.4.2),
• Survey designs that combine scans and static measurements (MARSSIM-type surveys,
Section 4.4.3), and
• Method-based survey designs (Section 4.4.4).
Figures 4.5, 4.6, and 4.7 illustrate the process of designing a disposition survey. Classification
can be used to provide a graded survey approach to individual survey designs. Information on
adjusting the level of survey effort based on classification is provided for each type of survey
design. Each survey design can include a variety of survey techniques (Section 5.9).
4.4.1 Scan-Only Survey Designs
Scan-only survey designs use scanning techniques to measure the M&E. The detector is moving
at a constant speed relative to the M&E being surveyed while maintaining a constant distance
relative to the M&E. In general, scan-only survey designs may be applied to all types of M&E,
from small individual items to large quantities of materials to large, complex machines. Scan
techniques include hand-held instruments that are moved over the M&E, as well as systems that
move the M&E past stationary detectors (e.g., conveyor systems). For example, a scan-only
survey may involve the use of a Geiger-Mueller (GM) pancake detector to measure potential
surface radioactivity on hand tools. Alternatively, a scan-only survey could involve the use of a
conveyorized system that measures large quantities of M&E (e.g., bulk material or laundry).
Scan-only surveys generally are applicable to all types of disposition surveys.
Scan-only survey designs often require the least amount of resources to design and implement,
and are easy to incorporate into SOPs or project-specific survey designs. In many cases it is not
necessary to document the results of individual scanning measurements because it is easy to
identify results that exceed some threshold corresponding to the action level. With the real-time
feedback available during Class 1 scan-only surveys, the user can implement a "clean as you go"
practice by segregating M&E that exceed the threshold for additional investigation. Drawbacks
to scan-only surveys include increased measurement uncertainty because of variations in scan
speed and source to detector distance making it difficult to detect or quantify radionuclides with
action levels close to zero or background.
Scan-only surveys are characterized by large numbers of measurements with relatively short
count times. Measurement uncertainty should account for variations in source-to-detector
distance, scan speed, and surface efficiency that are commonly associated with scanning
measurements.
January 2009 4-9 NUREG-1575, Supp.l
-------
Develop A Survey Design
MARSAME
From Figure 3.1
Select a Null Hypothesis
Scenario A (Section 4.2.3)
Scenario B (Section 4.2.4)
Assign Values to the
LBGRandUBGR
NOTE: Shaded boxes
represent important decisions
(diamonds) or milestones
(rectangles).
NOTE: A method-based survey
(Section 4.4.4) is a special case of
either scan-only, in situ, or
MARSSIM-type, and so will follow
one of the three paths show on this
figure.
Action Level (Section 3.3)
Discrimination Limit (Section 4.2.2)
Specify Limits on
Decision Errors
Type I Error (Section 4.2.5)
Type II Error (Section 4.2.5)
Consequences of Decision Errors
a and /3
Develop an Operation
Decision Rule
Statement of the Statistical
Hypothesis Test (Section 4.2.6)
Scan-Only
(Section 4.4.1)
Select a Type of
Disposition Survey
Class 1 (Section 4.3.1)
Class 2 (Section 4.3.2)
Class 3 (Section 4.3.3)
MARSSIM-Type
(Section 4.4.3)
Determine % of M&E
to be Scanned
In Situ (Section 4.4.2)
1
Determine % of M&E and
Locations to be Measured
Select M&E and Locations
to be Scanned
Select M&E and Locations
to be Measured
Select Measurement and
Scan Locations
Optimize the Survey Design
Document the Disposition
Survey Design
(Section 4.5)
Proceed to Figure 5.1
Figure 4.5 Flow Diagram for a Disposition Survey Design
NUREG-1575, Supp. 1
4-10
January 2009
-------
MARSAME
Develop A Survey Design
From Figure 4.5
Estimate as, the Variabilities in
the Reference Area and
Survey Unit Activities
Estimate a, the Variability in
the Level of Radioactivity
Is the Radioactivity
Present in Bkgd?
Calculate the Relative Shift A/a
Calculate the Relative Shift A/a
Is A/a
Between
1 and 3?
Obtain Number of Data Points
for the Sign Test, N, from Table
for each Survey Unit
Obtain Number of Data Points
for WRS Test, N/2, from Table
for each Survey Unit and
Reference Area
NOTE: Shaded boxes
represent important
milestones.
Do the Number of Data
Points Need to be Adjusted
for Class 1 M&E?
Prepare Summary of Data
Points for M&E being
Investigated
Return to
Figure 4.5
Figure 4.6 Flow Diagram for Identifying the Number of Data Points for a MARSSIM-Type
Disposition Survey
January 2009
4-11
NUREG-1575, Supp.l
-------
Develop A Survey Design
MARSAME
From Figure 4.6
Establish DQOs for Areas with the
Potential for Exceeding
Als and Acceptable Risk for
Missing Such Areas
Identify Number of Data Points
needed Based on Statistical Tests, n
Evaluate MDCs for Available
Instrumentation
sthe Scan MDC for Available
Instrumentation Less than the
Required Scan MDC?
Calculate the Area, A, Bounded by
Sample Location, n
Determine Acceptable
Concentrations in Various
Individual Smaller Areas within a
Survey Unit (i.e., Use Area Factors)
Determine the Acceptable
Concentration Corresponding to the
Calculated Area, A (i.e., Area
Factor x Action Level)
Determine the Required Scan MDC
to Identify the Acceptable
Concentration in an Area, A
Yes
No Additional Sampling Points
are Necessary for Potential
Elevated Locations
Calculate Area Factor that
Corresponds to the Actual Scan
MDC (scan MDC/AL)
Determine the Maximum
Area, A', that Corresponds
to the Area Factor
Recalculate Number of
Data Points Needed
n£A = Survey Unit Area//!')
Determine Grid Size Spacing, L
Return to
Figure 4.6
NOTE: Shaded boxes
represent important
milestones.
NOTE: "VOLUME" or
"MASS" replaces "AREA" in
this flow diagram as
appropriate for a specific
survey design, and scan MDC
is discussed in MARSSIM
Section 5.5.2.4
Figure 4.7 Flow Diagram for Identifying Data Needs for Assessment of Potential Areas of
Elevated Activity in Class 1 Survey Units for MARSSIM-Type Disposition Surveys
NUREG-1575, Supp. 1
4-12
January 2009
-------
MARSAME
Develop A Survey Design
Evaluation of scan-only survey data depends on whether or not individual measurement results
are recorded (Section 6.2.5). The decision of whether to record individual measurement results
will impact the selection of instrumentation (Section 5.9) and survey documentation require-
ments (see Sections 4.5, 5.11, and 6.9), and may impact handling of the M&E (Section 5.3).
4.4.1.1 Class 1 Scan-Only Surveys
Class 1 scan-only surveys require that physical measurements be performed for 100% of the
M&E being investigated. For individual items this may require scanning both sides of flat items
(e.g., sheet metal, boards) and changing the surveyor's grip on the item to ensure all areas are
surveyed (e.g., handles). For conveyor systems this may require flipping or rotating the M&E
and performing additional measurements. Conveyor systems can also be designed with detectors
surrounding the M&E (e.g., above and below a conveyor belt) to provide 100% measurability.
4.4.1.2 Class 2 Scan-Only Surveys
Class 2 scan-only surveys use information about the M&E to reduce the total area surveyed
using a graded approach. The amount of the M&E surveyed is calculated based on the relative
shift (i.e., A/a). The percent of the M&E to be surveyed is 10%, or the result using Equation 4-1,
whichever is larger:
% Scan =
10
x 100%
(4-1)
The amount of M&E to be scanned should be rounded up to the next 10 percent, and at least 10%
of the M&E must be surveyed. For example, if the percent scan is 51%, then 60% of the M&E
will be surveyed. This means that between 10 to 100% of Class 2 M&E would be measured
during the disposition survey. Figure 4.8 shows the relationship between the relative shift and the
amount of M&E to be scanned.
The scanned percentages
need to represent spatially
uniform coverage of the
survey unit and coincide
with the conceptual model
for the M&E. Consider
spatially uniform coverage
when scanning 30% of a
desk and 30% of a bucket
of bolts. For the desk
example, 30% coverage
during scanning may be
derived from performing
scans on the top surface,
the legs, inside the drawers,
etc., so that essentially 30%
1!
4 5 6 7 i
Relative Shift (A/a)
10 11 12
Figure 4.8 Relationship Between the Relative Shift and the
Amount of M&E to be Scanned
January 2009
4-13
NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
of each surface is scanned, yielding 30% total coverage of the entire desk. For the bucket of bolts
example, 30% scanning coverage means laying out all the bolts and scanning 30% of them as
well as 30% of the bucket itself. Alternatively, if the conceptual model for the desk showed a
higher potential for contamination on the top, bottoms of legs, and drawer handles, 100% of
these areas could be scanned with smaller amounts of the areas with a lower potential for
radioactivity scanned to provide a total of 30% coverage for the entire desk. The graded
approach should be applied to all aspects of the survey design.
The selection of M&E to survey as part of a Class 2 survey is project specific and is determined
based on what is known about the M&E. For example, if all of the M&E is accessible and is
expected to have uniform radionuclide concentrations or levels of radioactivity, the M&E to be
surveyed should be selected randomly. However, there may be areas that are difficult-to-access
with the instrumentation selected to perform the survey. If there is a known and accepted
relationship between radionuclides in difficult-to-access areas and radionuclides in accessible
areas, the Class 2 measurements may be biased to only accessible areas (i.e., representative of
measurements in difficult-to-access areas).
If elevated radionuclide concentrations or levels of radioactivity are restricted to areas that can be
readily identified (e.g., discolored areas, corners, cracks, access points) the Class 2
measurements may be designed to concentrate on these biased areas. The Class 2 survey design
should include a combination of biased and random areas to check assumptions used to support
the survey design.
The selection of M&E to survey may also depend on the physical characteristics of the M&E.
For example, surveying 40% of the inside of a railroad car would be different from surveying
40% of a pile of rubblized concrete. Section 5.3 provides information on handling M&E and
determining what will be measured during implementation of the survey design.
4.4.1.3 Class 3 Scan-Only Surveys
Class 3 scan-only survey designs are identical to Class 2 scan-only survey designs. The planning
team may decide that some Class 3 scan-only disposition surveys require that less than 10% of
the M&E will be measured. The decision to design a survey requiring less than 10% of the M&E
to be measured should be based on the total uncertainly associated with the disposition decision.
The determination of total uncertainty should be based on process knowledge, historical data,
and the results of preliminary and disposition surveys.
In addition, some Class 3 scan-only survey designs may be based solely on biased
measurements. In other words, random measurement locations are not required for Class 3 scan-
only survey designs. However, if biased measurements are reasonable, they should be performed,
keeping in mind that Class 3 M&E have very little or no potential for exceeding the AL.
NUREG-1575, Supp. 1 4-14 January 2009
-------
MARSAME Develop A Survey Design
4.4.2 In Situ Survey Designs
In situ survey designs use static measurements to measure 100% of an item. The detector and the
item being measured are held in a fixed geometry5 for a specified count time to meet the MQOs.
There are a wide variety of in situ measurement techniques available. Examples include box
counters, portal monitors, and in situ gamma spectrometry systems, as well as direct
measurements with hand-held instruments (e.g., Nal(Tl), ZnS, GM pancake, and portable gas
proportional detectors). In situ surveys generally are applied to situations where scan-only
surveys are determined to be unacceptable. For example, variations in source-to-detector
distance, scan speed, and surface efficiency that are commonly associated with scanning
measurements can often be effectively controlled using an in situ survey design.
In situ surveys are characterized by limited numbers of measurements with long count times
(relative to scan-only surveys). Measurement uncertainty will incorporate spatial uncertainty
because of the source geometry assumed in the calibration. Thus, special attention needs to be
made to the assumptions made in the calibration of in situ systems. Potential deviations from
these assumptions need to be propagated through the calibration equation to assess the total
measurement uncertainty (see Sections 5.6 and 7.4). Count times are determined by the MQOs
rather than the time constant of the measurement system. In situ measurements provide a 100%
measurement for some portion of the M&E being investigated. The M&E may be an individual
item or piece of equipment, or some fraction of a large quantity of material determined by the
solid angle coverage of the detector.
In situ surveys may consist of a single measurement, or a series of measurements. Single
measurement surveys typically are performed on individual items or relatively small batches of
M&E. A series of in situ measurements may be used to evaluate larger quantities of M&E. In
some cases, a series of in situ measurements may be performed of a single item or batch of M&E
to provide several estimates of the radionuclide concentrations from different angles. The
primary difference between an in situ survey and a MARSSEVI-type survey is that an in situ
survey measures 100% of an item (using one or several measurements) to determine the average
radionuclide concentration for that item. A MARSSIM-type survey uses a statistically based
number of measurements (that generally do not measure 100% of the item or group of items
being surveyed) to calculate an average radionuclide concentration for that item or group of
items.
4.4.2.1 Class 1 In situ Surveys
Class 1 in situ surveys require that physical measurements be performed for 100% of the M&E
being investigated. Placing an item inside a 4-n measurement system, performing a series of
measurements with overlapping fields of view that incorporate all of the M&E, or rotating the
M&E within the field of view of the detector so 100% of the M&E are measured are examples
where 100% of the M&E are measured.
5 There are situations where the levels of radioactivity for M&E being measured are expected to be inhomogeneous.
Certain measurement systems can rotate the M&E during a measurement to provide an estimate of the average
activity. For the purposes of this section, these are considered fixed geometries. Additional discussion on the
limitations of these systems is provided in Chapter 5.
January 2009 4-15 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
4.4.2.2 Class 2 In situ Surveys
Class 2 in situ surveys use information about the M&E to reduce the total area surveyed using a
graded approach. The amount of the M&E surveyed is calculated based on the relative shift (i.e.,
A/a). The percent of the M&E to be surveyed is 10% or the result using Equation 4-2, whichever
is larger:
% Measured or % Solid Angle Coverage = - — x 100% (4-2)
The fraction of the M&E or the solid angle coverage of the M&E to be surveyed should be
rounded up to the next 10 percent. If the % coverage is 51%, then 60% of the M&E will be
surveyed. This means that 10 to 100% of Class 2 M&E would be measured during the
disposition survey. Figure 4.8, on page 4-13, shows the relationship between the relative shift
and the amount of M&E to be surveyed.
The selection of M&E to survey as part of a Class 2 survey is project specific and is determined
based on what is known about the M&E. For example, if all of the M&E is accessible and is
expected to have uniform radionuclide concentrations or levels of radioactivity, the M&E to be
surveyed should be selected randomly. However, there may be areas that are difficult-to-access
with the instrumentation selected to perform the survey. If there is a known and accepted
relationship between radionuclides in difficult-to-access areas and radionuclides in accessible
areas, the Class 2 measurements may be biased to only accessible areas (i.e., representative of
measurements in difficult-to-access areas). If elevated radionuclide concentrations or levels of
radioactivity are restricted to areas that can be readily identified (e.g., discolored areas, corners,
cracks, access points) the Class 2 measurements may be designed to concentrate on these biased
areas. The Class 2 survey design should include a combination of biased and random areas to
check assumptions used to support the survey design.
4.4.2.3 Class 3 In situ Surveys
Class 3 in situ survey designs are identical to Class 2 in situ survey designs. The planning team
may decide that some Class 3 in situ disposition surveys require that less than 10% of the M&E
will be measured. The decision to design a survey requiring less than 10% of the M&E to be
measured should be based on the total uncertainty associated with the decision based on process
knowledge, historical data, and the results of preliminary and disposition surveys.
4.4.3 MARSSIM-Type Survey Designs
MARSSIM-type survey designs combine a statistically based number of static measurements to
determine average radionuclide concentrations or radioactivity levels with scanning to identify
areas of elevated radionuclide concentrations or radioactivity for specified quantities of M&E
(i.e., survey units). Identifying survey unit sizes, laying out systematic measurement grids, and
calculating project- and item-specific area factors requires a significant effort. Section 5.3
discusses considerations for handling M&E, including locating measurements. The planning
team should consider that MARSSIM-type survey designs might be more complex and require
NUREG-1575, Supp. 1 4-16 January 2009
-------
MARSAME Develop A Survey Design
more resources than scan-only or in situ survey designs that meet the DQOs. Information on
designing MARSSIM-type surveys is found in MARSSIM Section 5.5. In general, MARSSIM-
type surveys of M&E are only performed on large, complicated M&E with a high inherent value
after scan-only and in-situ surveys have been considered and rejected.
4.4.3.1 Class 1 MARSSIM-Type Surveys
Class 1 MARSSIM-type surveys calculate the required number of measurements in each survey
unit based on the shift (i.e., A), the variability in the radionuclide concentrations or levels of
radioactivity (i.e., o), and the Type I and Type II decision error rates (i.e., a and /?). The number
of measurements per survey unit is adjusted to account for small areas of elevated activity using
the information in MARSSIM Section 5.5.2.4. In addition, scan measurements are required for
100% of the M&E being investigated.
The development of survey unit boundaries is discussed in Section 3.6. The quantity of M&E in
each survey unit should be determined based on the modeling assumptions used to develop the
action levels.
The variability in the radionuclide concentrations in each survey unit can be estimated using the
standard deviation of preliminary measurements or the uncertainties from individual
measurements, whichever is larger. Whenever practical, preliminary data should be used to
provide estimates of variability. As a last resort when preliminary data are not available,
MARSSIM states that assuming a coefficient of variation on the order of 30% may be reasonable
(MARSSIM Section 5.5.2.2, Page 5-26). This 30% is used as a starting point for the DQO
process, and should be adjusted iteratively during the development of a final survey design. For
M&E, MARSAME recommends using a higher percentage value.
Area factors are specified in a regulation or other guidance, or developed based on the changes in
dose or risk associated with changing the area (or volume) of activity to be less than the entire
survey unit. For example, DOE Order 5400.5, Figure IV-1 (DOE 1993) allows use of an area
factor of up to 3.0 for total surficial radioactivity for all radionuclides, NUREG-1640 (NRC
2003a) is only concerned with average activity and total inventory of radioactivity, which
implies that within the survey unit relatively high localized concentrations of radioactivity could
exist. This implication does not mean that a large part of the survey unit may be used to
intentionally "dilute" high concentrations of radioactivity. Rather, in the course of normal
processing there is a non-prescriptive flexibility allowed for inhomogeneity of radionuclide
concentrations. Nevertheless, mixing different classes of M&E (Class 1, 2, and 3) is not allowed.
The physical characteristics of the M&E combined with potential future exposures based on the
selected disposition option mean that area factors (and possibly exposure pathway dose or risk
models) need to be developed for each project. In the absence of regulation-specific area factors,
assuming an area factor of 1.0 for all radionuclides would be the most conservative approach.
Depending on the basis of the action level, an area factor may or may not be applicable.
MARSSIM uses completely different scenarios for real property to develop area factors in
contrast to those scenarios used for M&E in NUREG-1640 (NRC 2003a). Area factors may be
derived on a project-specific basis using project-specific scenarios.
January 2009 4-17 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
If the radioactivity being measured is present in background, Table 5.3 in MARSSIM provides
the number of measurements required in each survey unit as well as in each reference area.
MARSSIM Section 5.5.2.2 andNUREG-1505 (NRC 1998a) Sections 9.4 and 9.5 provide
information on calculating the number of required measurements when the radioactivity being
measured is present in background.
If the radioactivity being measured is not present in background, Table 5.5 in MARSSIM
provides the number of measurements required in each survey unit. MARSSIM Section 5.5.2.3
and NUREG-1505 (NRC 1998a) Sections 9.2 and 9.3 provide information on calculating the
number of required measurements when the radioactivity being measured is not present in
background. For convenience, statistical sample size and critical value tables for the Sign and
Wilcoxon Rank Sum (WRS) tests taken from MARSSIM Appendix I are given in MARSAME
Appendix A. In addition, Appendix A contains a table of critical values for the Quantile test,
taken from NUREG-1505.
Whenever area factors other than 1.0 are used to design the disposition survey, a systematic grid
should be used to determine measurement locations. The systematic grid determines the largest
area that could be missed by the measurements which is used to determine the required scan
MDC. Section 5.3 provides information on handling M&E, including setting up systematic grids.
4.4.3.2 Class 2 MARSSIM-Type Surveys
Class 2 MARSSIM-type surveys are similar to Class 1 MARSSEVI-type surveys. The numbers of
measurements in each survey unit are determined in the same manner, although the expected
radionuclide concentrations or levels of radioactivity and the decision error rates may change.
Unlike MARSSIM, the survey unit size remains the same and does not change based on
classification. The portion of the survey unit where scan surveys are required is reduced to
between 10 and 100%. The information in Section 4.4.1.2 for Class 2 scan-only surveys should
be used to determine the areas to be scanned. This recommendation is provided for M&E only,
and is not intended to update the guidance in MARSSIM for surface soils and building surfaces.
4.4.3.3 Class 3 MARSSIM-Type Surveys
Class 3 MARSSIM-type surveys are similar to Class 1 MARSSEVI-type surveys. The numbers of
measurements in each survey unit are determined the same way, although the expected
radionuclide concentrations or levels of radioactivity and the decision error rates may change.
Unlike MARSSIM, the survey unit size does not change based on classification. The portion of
the survey unit where scan surveys are required is reduced to less than 10% and is based on
professional judgment. The information in Section 4.4.1 for scan-only surveys should be used to
determine the areas to be scanned. This recommendation is provided for M&E only, and is not
intended to update the guidance in MARSSIM for surface soils and building surfaces.
4.4.4 Method-Based Survey Designs
The action level selected in Section 3.3 may implicitly or explicitly require using a specific
measurement technique (Section 5.9.1) or instrument (Section 5.9.2). A survey design that is
NUREG-1575, Supp. 1 4-18 January 2009
-------
MARSAME Develop A Survey Design
based on a required measurement method, or combination of measurement technique and
instrumentation, is called a "method-based" survey design.
A method-based survey design is a scan-only, in situ, or MARSSIM-type survey design that
incorporates the required measurement method. The survey design will still need to address all of
the required components, such as number, type, location, and sensitivity of measurements.
Survey components that are not specified as part of the required measurement method should be
identified and addressed using the DQO process.
4.4.5 Optimize the Disposition Survey Design
The disposition survey design process described in this supplement could result in the
development of multiple potential disposition survey designs. For example, consider the case
when simultaneous compliance with more than one action level is required (e.g., DOE 1993). In
other cases the decision resulting from one survey may lead to the requirement of another survey,
such as failure to demonstrate compliance with the disposition criterion for release resulting in a
survey to comply with radioactive waste acceptance criteria. Multiple survey designs could result
from selection of multiple potential disposition options, action levels, survey techniques,
measurement systems, decision rules, or some combination of these factors. Before the planning
team can proceed, all of the potential disposition survey designs need to be reviewed to select a
final disposition survey design.
The final step in the DQO process ("Develop the Detailed Plan for Obtaining Data," Step 7) is
designed to produce the most resource-efficient survey design that is expected to meet the
DQOs. It may be necessary to revisit previous steps in the DQO process and work through this
step more than once.
There are five activities included in this step:
1. Review existing data (e.g., historical data, sentinel measurement results, preliminary survey
results). Use existing data to support the data collection design. If no existing data are
available, consider performing preliminary surveys to acquire estimates of variability to
determine numbers of measurements. Evaluate potential problems regarding detection limits
or interferences. If new data will be combined with existing data, determine if there are data
gaps that need to be filled or deficiencies that can be mitigated prior to implementing the
disposition survey design.
2. Evaluate operational decision rules. The theoretical decision rules developed in Section 3.6
were based on the assumption that the true radionuclide concentrations or radioactivity
present in the M&E were known. Operational decision rules based on the statistical tests
(Chapter 6) should replace the theoretical decision rule (see Sections 3.5 and 4.2.6). Review
the parameter of interest (e.g., maximum measured value, mean or median radionuclide
concentration) and the possible statistical tests that could be applied to the data to evaluate
the operational decision rules.
3. Develop general data collection design alternatives. Sections 4.4.1, 4.4.2, and 4.4.3 provide
information on general data collection design alternatives applicable to disposition surveys.
January 2009 4-19 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
Consider individual instruments and measurements techniques (Section 5.9) combined with
general data collection designs to develop alternative survey approaches.
4. Calculate the number of measurements or amount of M&E to be surveyed. Sections 4.4.1,
4.4.2, and 4.4.3 provide general information on determining the level of survey effort for the
general data collection design alternatives based on classification. Determine the estimated
resources required for each of the alternative survey approached.
5. Select the most resource-effective survey design. Evaluate each of the survey approaches
based on the required resources and the ability to meet the DQO and MQO constraints within
the tolerable decision error limits. The survey design that provides the best balance between
cost and meeting survey objectives while considering the non-technical economic and health
factors imposed on the project is usually the most resource-effective. The statistical concept
of a power curve (MARSSIM Appendix 1.9) is extremely useful in investigating the
performance of alternative survey designs.
If none of the alternative survey designs meet the survey objectives within the tolerable decision
error limits while considering the budget or other constraints, then the planning team will need to
relax one or more of the constraints. Examples include—
• Increasing the budget for implementing the survey;
• Using exposure pathway modeling to develop site-specific action levels;
• Increasing the decision error rates, not forgetting to consider the consequences associated
with making an incorrect decision;
• Increasing the width of the gray region for Scenario A surveys by decreasing the average
activity associated with the M&E which may require remediation, or negotiating a higher
UBGR for Scenario B which may require additional reference area investigations;
• Relaxing other project constraints (e.g., schedule);
• Changing the boundaries—it may be possible to reduce measurement costs by changing or
eliminating survey units that will require different decisions;
• Segregating the M&E based on physical or radiological attributes (Section 5.4);
• Evaluating alternative measurement techniques with lower detection limits or lower survey
costs;
• Adjusting the list of radionuclides or radiations of concern (Section 3.2); and
• Considering other disposition options that will result in higher action levels.
4.5 Document the Disposition Survey Design
Documentation of the disposition survey design should provide a complete record of the selected
survey design. The documentation should include all assumptions used to develop the survey
design, a detailed description of the M&E being investigated, along with the DQOs and MQOs
for the survey (e.g., MQC, MDC, count time). The regulatory basis for the disposition criterion
and calculations showing the derivation of action levels should also be provided. Sufficient data
and information should be provided to enable an independent re-creation and evaluation of the
disposition survey design. The documentation should provide information on the following
topics:
NUREG-1575, Supp. 1 4-20 January 2009
-------
MARSAME Develop A Survey Design
• Information on who developed, reviewed, and approved the survey design, as well as training
and qualification requirements for such individuals, should be included, along with any
requirements for who can implement the survey design.
• Information on what M&E were considered when developing the survey design along with a
description of M&E to which the survey design applies.
• Information on when the survey design was developed along with when the survey design
will be implemented including restrictions on time of day, time of year, and count times
when applicable.
• Information on where the survey design can be applied (including restrictions on local
background levels) along with measurement locations including fraction of M&E to be
surveyed and locations of direct measurements or samples or methods for selecting locations
during implementation,
• Information on why a survey should be performed including justification for impacted and
non-impacted decisions and assignment of classifications,
• Information on how the survey will be performed including measurement techniques and
instruments along with instructions for segregating and handling the M&E during the survey.
There are two methods for documenting surveys described in the following sections, based on
the type of project—
• Routine or repetitive surveys, and
• Case-specific applications.
4.5.1 Routine Surveys and Standard Operating Procedures
Routine (or repetitive) surveys are disposition surveys that are routinely performed on M&E
entering or leaving an operating facility. Examples of routine surveys include-
• Clearance of tools from radiological control areas at a radiation facility,
• Preparation of low-level radioactive waste for disposal, and
• Interdiction of scrap metal entering a recycling facility.
Documenting routine survey designs, for example as SOPs, can be consistent with MARSAME
recommendations. SOPs detail the work processes that are conducted or followed within an
organization and document the way activities are performed. SOPs that also meet the DQOs for
the disposition survey can be used to document routine survey designs. The development and use
of SOPs facilitates consistent conformance to technical and quality system requirements. They
promote quality through consistent implementation of a process within an organization, even if
there are temporary or permanent personnel changes. The benefits of a valid SOP are reduced
work effort combined with improved data comparability, credibility, and legal defensibility
(EPA 2001). Additional guidance on developing SOPs, including example SOPs, is provided in
EPA QA/G-6 (EPA 2001).
January 2009 4-21 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
4.5.1.1 SOP Process
The organization developing the SOP should have a procedure in place for determining what
procedures or processes need to be documented. SOPs documenting these procedures or
processes should be written by individuals knowledgeable with the activity and the
organization's internal structure. For disposition survey designs, a team approach to writing
SOPs is often used. This allows input from subject-matter experts with information critical to the
survey process, and promotes acceptance of the SOP once it is completed.
SOPs should be concise and provide step-by-step instructions in an easy-to-read format. They
should provide sufficient detail so that a technician with limited experience, but with a basic
understanding of the process, can successfully implement the survey design when unsupervised.
Disposition survey SOPs should be reviewed and validated by one or more individuals with
appropriate training and experience in performing surveys of M&E before they are implemented.
It may be helpful to have the draft SOP field tested by someone not directly involved in the
development of the SOP. The review process for disposition surveys should include a regulatory
review and appropriate stakeholder involvement.
SOPs need to remain current. SOPs should be updated and re-approved whenever survey
procedures are changed. SOPs should be systematically reviewed on a periodic basis to ensure
that the policies and procedures remain current and appropriate.
Many disposition survey activities use checklists or forms to document completed tasks (e.g.,
daily instrument checks). Any checklists or forms included as part of the disposition survey
should be referenced at the points in the procedure where they are used and attached to the SOP.
Remember that the checklist or form is not the SOP, but a part of the SOP.
The organization should have a system for developing, reviewing, approving, controlling, and
tracking documents. This process is usually documented in the Quality Management Plan.
4.5.1.2 General Format for Disposition Survey SOPs
In general, disposition survey SOPs consist of five elements:
• Title Page,
• Table of Contents,
• Procedures,
• Quality Assurance and Quality Control, and
• References.
The title page should include a title that clearly identifies the activity, an identification number,
date of issue or revision, and the name of the organization to which the SOP applies. The
signatures and signature dates of individuals who prepared and approved the SOP also should be
included.
NUREG-1575, Supp. 1 4-22 January 2009
-------
MARSAME Develop A Survey Design
The table of contents lists the major section headings and the pages where the information is
located. This provides a quick reference for locating the desired information and identifies
changes or revisions made to individual sections.
The procedures are specific to the disposition survey design and may include some or all of the
following topics:
• Scope and applicability. This section should provide a detailed description of the M&E to
which the SOP can be applied. In addition, it is often important to clearly identify M&E to
which the SOP does not apply.
• Summary of method. This section briefly describes the overall survey design, identifies the
disposition option, lists the action levels, and provides their regulatory basis. The details on
the development of the action levels based on the disposition criterion in the regulations is
generally referenced or included as an attachment.
• Definitions. This section identifies and defines any acronyms, abbreviations, or specialized
terms used in the SOP.
• Health and safety warnings. This section indicates operations that could result in personal
injury, loss of life, or uncontrolled release to the environment. Explanations of what could
happen if the procedure is not followed or if it is followed incorrectly should appear here as
well at the critical steps in the procedure.
• Cautions. This section identifies activities that could result in equipment damage,
degradation of data, or possible invalidation of results. Explanations of what could happen if
the procedure is not followed or if it is followed incorrectly should appear here as well as the
critical steps in the procedure.
• Interferences. This section describes any component of the process that may interfere with
the final decision regarding disposition of the M&E.
• Personnel qualifications. This section lists the minimum experience required for individuals
implementing the SOP. Any required certifications or training courses should be listed. For
many routine surveys the training records of the personnel implementing the survey design
are used to document compliance with the SOP.
• Equipment and supplies. This section lists and specifies the equipment, materials, reagents,
and standards required to implement the SOP. At a minimum, this section must identify the
model number and manufacturer of instruments that will be used to perform the survey.
• Roles and responsibilities of project personnel. This section identifies the decision-maker for
the project as well as identifying who is responsible for performing specific tasks. An
organizational chart documenting the chain of command and reporting authority (including
quality control, health and safety, and any subcontractors) is a useful tool for showing
potential interactions between project team members.
• Procedure. This section provides all pertinent steps, in order, and materials needed to
implement the survey design. This section should include—
o Instrument or method calibration and standardization (generally requires a check of the
instrument calibration date and lists the appropriate MQOs such as MQC or MDC and
references the details for these processes),
o Type, number, and location of measurements,
o Data acquisition, calculations, and data reduction requirements,
o Troubleshooting, and
January 2009 4-23 NUREG-1575, Supp.l
-------
Develop A Survey Design MARSAME
o Computer hardware and software.
• Data and records management. This section describes the forms to fill out, reports to be
written, and data and record storage information. At a minimum routine survey records
should identify the personnel performing measurements and the instruments used to perform
the measurements (i.e., model and serial number for all components of the measurement
system). These records should show that the personnel performing the survey were properly
trained and the instruments used to collect the data were calibrated and operating properly.
This section should clearly state whether individual measurement results will be recorded,
because this information is not always required.
The QA/QC section describes the activities required to demonstrate the successful performance
of the disposition survey. For many organizations the QC activities for individual instruments are
provided in separate SOPs describing the proper use of that instrument, so the daily checks of the
instruments are included by reference. The QA/QC section should identify QC requirements for
the disposition survey such as blanks, replicates, splits, spikes, and performance evaluation
checks. The frequency for each QC measurement should be listed along with a discussion of the
rationale for decisions. Specific criteria should be provided for evaluating each type of QC
measurement, as well as actions required when the results exceed the QC limits. The procedures
for reporting and documenting the results of QC measurements should be listed in the QA/QC
section. Section 5.10 provides additional information on QC for disposition surveys.
The reference section should list all documents or SOPs that interface with the routine survey
SOP. Full references (including SOP versions and dates) should be provided. Published literature
and instrument manuals that are not readily available should be attached.
4.5.2 Case-Specific Applications
There are M&E that may require a disposition survey that are not covered by routine surveys.
These are collectively referred to as case-specific applications. Case-specific applications include
project-specific applications such as decommissioning or cleanup surveys, as well as unique
applications involving one-time disposition of special equipment from a facility.
Ideally, documentation of case-specific survey designs involves a comparable level of effort
associated with routine surveys. This is obviously the case for large decommissioning or cleanup
projects where survey designs are documented as SOPs using a process analogous to routine
surveys. The major differences are seen in the requirements for approval and maintenance of
SOPs, which generally are less for decommissioning or cleanup projects compared to operating
facilities. Disposition survey designs that will be applied during decommissioning or cleanup
activities typically are documented as part of the survey design. However, a survey design needs
to provide all of the information supporting the development of the disposition survey design,
where SOPs typically focus on one aspect of the survey design or implementation. Historical
information, process knowledge, description of the M&E, and assumptions used in the
disposition survey design need to be included and not referenced.
The assumptions used to develop survey designs for routine surveys cannot be applied to all
M&E, so situations will arise where a disposition survey design needs to be developed for
NUREG-1575, Supp. 1 4-24 January 2009
-------
MARSAME Develop A Survey Design
special items or unique applications. These types of surveys are often associated with M&E that
have a high inherent value (e.g., large quantities of valuable materials, unique or very expensive
equipment) to offset the resources required to develop a unique disposition survey design. These
special survey designs need to be inclusive, providing all of the information supporting the
development of the disposition survey design. Detailed discussions should be provided for all
parts of the survey design, including selection of a disposition option, selection and development
of action levels, development of MQOs and selection of instruments, and QA/QC requirements
for individual measurement systems as well as for the entire disposition survey.
For most applications the disposition survey design is expected to be documented as a stand-
alone survey plan or as a series of SOPs. However, the planning team may determine that the
survey design documentation can be combined with the results of the survey into a single
document. At a minimum, instructions on the type, number, and location of measurements
should be documented to provide instructions to the technicians performing the survey.
Documenting the entire disposition decision process in a single document is most appropriate for
unique applications where there is sufficient historical information or survey precedent such that
there is little uncertainty associated with the development of a survey design. The benefit of
documenting all of the survey decisions (e.g., design, implementation, and assessment) in one
document is the savings in resources to develop multiple documents. The risk associated with not
documenting the survey design process until after implementation is that the assessment will
identify some problems with the survey design requiring additional data collection which could
impact project costs and schedule.
January 2009 4-25 NUREG-1575, Supp.l
-------
-------
MARSAME Implement The Survey Design
5 IMPLEMENT THE SURVEY DESIGN
5.1 Introduction
This chapter discusses the implementation phase of the data life cycle and focuses on controlling
measurement uncertainty and associated MQOs. The information in this chapter describes
approaches for safely implementing the final disposition survey design developed in Chapter 4,
methods for controlling uncertainty, and techniques to determine whether the measurement
results achieve the survey objectives. Figure 5.1 illustrates the implementation phase of the data
life cycle.
Similar to MARSSIM, MARSAME excludes specific recommendations for implementing
disposition surveys. Instead, MARSAME provides recommendations and information to assist
the user in selecting measurement techniques for implementing the survey design. This approach
encourages consideration of innovative measurement techniques and emphasizes the flexibility
of the information in MARSAME.
Implementation begins with health and safety considerations for the disposition survey (Section
5.2). Section 5.3 provides information on handling M&E, while Section 5.4 discusses
segregating M&E based on physical and radiological attributes. Section 5.5 continues the
discussion of measurement quality objectives (MQOs) from Chapters 3 and 4. Measurement
uncertainty (Section 5.6), detectability (Section 5.7), and quantifiability (Section 5.8), are three
MQOs that are described in greater detail. Combining an instrument with a measurement
technique to ensure the MQOs are achieved is discussed in Section 5.9. Section 5.10 provides
information on quality control (QC), and information on data reporting is provided in
Section 5.11.
5.2 Ensure Protection of Health and Safety
Health and safety is emphasized as an issue potentially affecting the implementation of
MARSAME disposition surveys. The focus of minimizing hazards is shifted away from
environmental hazards (e.g., confined spaces, unstable surfaces, heat and cold stress) and
towards scenarios where health and safety issues may affect how a disposition survey is designed
and performed. Work areas and procedures that present potential safety hazards must be
identified and evaluated to warn personnel of potential hazards. Personnel must be trained with
regard to potential physical and chemical safety hazards (e.g., inhalation, adsorption, ingestion,
injection/puncturing) and the potential for injury (e.g., slips, trips, falls, burns).
A job safety analysis (ISA) should be performed prior to implementing a disposition survey. The
ISA offers an organized approach to the task of locating problem areas for material handling
safety (OSHA 2002). The ISA should be used to identify hazards and provide inputs for drafting
a health and safety plan (HASP). The HASP will address the potential hazards associated with
M&E handling and movement and should be prepared concurrently with the survey design. The
HASP identifies methods to minimize the threats posed by the potential hazards. The information
in the HASP may influence the selection of a measurement technique and disposition survey
procedures. Radiation work permits (RWPs) may be established to control access to
January 2009 5-1 NUREG-1575, Supp. 1
-------
Implement The Survey Design
MARSAME
From Figure 4.1
Ensure Protection of
Health and Safety
(Section 5.2)
Handle M&E
(Section 5.3)
Segregate the M&E
(Section 5.4)
-Yes
Do M&E Need
Segregation?
Prepare M&E for Survey (Section 5.3.1)
Provide Access to M&E (Section 5.3.2)
Transport the M&E (Section 5.3.3)
NOTE: Shaded boxes
represent important decisions
(diamonds) or milestones
(rectangles).
Set Measurement
Quality Objectives
(Section 5.5-5.8)
Select a Measurement
Technique and
Instrumentation Combination
(Section 5.9)
Set Quality
Control Requirements
(Section 5.10)
Perform the Survey &
Report the Results
(Section 5.11)
Proceed to Figure 6.1
Figure 5.1 Implementation of Disposition Surveys
NUREG-1575, Supp. 1
5-2
January 2009
-------
MARSAME Implement The Survey Design
radiologically controlled areas. RWPs contain requirements from the ISA, such as dosimetry and
personal protective equipment (PPE), as well as survey maps illustrating predicted dose rates and
related radiological concerns (e.g., removable or airborne radioactivity). Hazard work permits
(HWPs) may be used in place of RWPs at sites with primarily physical or chemical hazards. The
mineral processing facility concrete rubble example presented in Chapter 8 (see Table 8.9)
provides an example of a ISA.
The ISA systematically carries out the basic strategy of accident prevention through the
recognition, evaluation, and control of hazards associated with a given job as well as the
determination of the safest, most efficient method of performing that job. This process creates a
framework for deciding among engineering controls, administrative controls, and PPE for the
purpose of controlling or correcting unsafe conditions (Hatch 1978). Examples of these controls
include—
• Engineering controls, which are physical changes in processes or machinery (e.g., installing
guards to restrict access to moving parts during operation), storage configuration (e.g., using
shelves in place of piles or stacks);
• Administrative controls, which are changes in work practices and organization (e.g.,
restricted areas where it is not safe to eat, drink, smoke, etc.) including the placement of
signs to warn personnel of hazards; and
• Personal protective equipment, which are clothing or devices worn by employees to protect
against hazards (e.g., gloves, respirator, full-body suits).
Correction measures may incorporate principles of all of the controls listed above. The preferred
method of control is through engineering controls, followed by administrative controls, and then
personal protective equipment.
Proper handling procedures for hazardous M&E are documented in site-specific health and
safety plans. Compliance with all control requirements is mandatory to maintain a safe working
environment. Personnel must regard control requirements as a framework to facilitate health and
safety, while still taking responsibility for their own well being. Being wary of safety hazards
remains an individual responsibility and personnel must be aware of their surroundings at all
times in work areas.
5.3 Consider Issues for Handling M&E
Materials and equipment handling is addressed in this document as a process control issue. M&E
handling requirements are determined by the final integrated survey design (Section 4.4) and the
combination of instrumentation and measurement technique used to perform the survey (Section
5.9). M&E may also require handling to more closely match the assumptions used to develop
instrument calibrations used to determine measurement uncertainty (Section 5.6), measurement
detectability (see Section 5.7), and measurement quantifiability (Section 5.8). Typically, M&E
will be handled to—
• Prepare a measurement grid or arrange M&E to perform a survey,
• Provide access for performing measurements, and
January 2009 5-3 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
• Transport the M&E to a different location.
5.3.1 Prepare M&E for Survey
Depending on the survey design, or assumptions used to develop the survey design, it may be
necessary to prepare the M&E for survey. The amount of preparation required is determined by
the DQOs and MQOs, and ranges from identifying measurement locations to adjusting the
physical characteristics of the M&E (e.g., disassembly, segregation, physical arrangement).
The performance of a MARSSIM-type survey requires determining the location where the
measurements are to be performed. The DQOs will determine the level of effort required to
identify, mark, and record measurement locations.
Identifying measurement locations can be problematic because MARSSIM-type surveys
recommend samples to be located either randomly (Class 3) or on a systematic grid (Class 1 and
Class 2). Class 2 and Class 3 scan-only and in situ surveys do not require 100% of the M&E to
be measured, so a method of identifying which portions will be measured is required.
Bulk materials or M&E consisting of many small, regularly shaped objects can be spread out in a
uniform layer, and a two-dimensional grid can be superimposed on the surface to identify
measurement locations. However, it is virtually impossible to identify random or systematic
locations on M&E that consist of relatively few, large, irregularly shaped objects. The reason is
that it is virtually impossible to establish a reference grid for these M&E. It is important to note
that the objective for random locations is to allow every portion of the survey unit the same
opportunity to be measured. Alternatively, the objective of systematic locations is to distribute
the measurement locations equally. It is only necessary to establish a reference grid to
sufficiently identify the measurement locations to meet the survey objectives.
One way to approximate a reference grid for locating measurements is to establish a grid in the
area where the survey will be performed. The M&E to be surveyed are laid out in a single layer
within the grid. The grid can then be used to identify measurement locations. Another option for
locating measurements involves superimposing a grid on top of the M&E. A net could be laid
over the M&E to be surveyed, ropes could be laid over the M&E to form a grid, or lights on a
grid could be directed onto the M&E to approximate a grid and identify measurement locations.
If measurement locations cannot be identified with a grid, there may be no alternative but to
perform biased measurements. Measurements would be performed preferentially in locations
more likely to contain radionuclides or radioactivity, based on the results of the initial
assessment (IA) (Section 2.5). This process involves professional judgment and may result in
overestimating the average radionuclide concentration or level of radioactivity. In all cases, it is
important to document the criteria used for identifying measurement locations and to document
that these criteria were followed.
Marking measurement locations, once they have been identified, should be done in a way that
will not interfere with the measurement. For example, using paint to mark the location of an
alpha measurement could end up masking the presence of alpha activity. Using arrows, marking
NUREG-1575, Supp. 1 5-4 January 2009
-------
MARSAME Implement The Survey Design
borders, or using an alternate method for marking locations (e.g., encircling with chalk) should
be considered for these types of situations.
Recording measurement locations may be required as part of the survey objectives if the
measurements may need to be repeated. For example, a large piece of equipment is surveyed
prior to use on a decommissioning or cleanup project. If the exact same locations will be
surveyed at the completion of the project, it will be necessary to record the measurement
locations. Permanent or semi-permanent markings can be used to identify the measurement
locations. Video or photographic records of measurement locations can also be used to return to
a specific measurement location.
5.3.2 Provide Access
Large pieces of equipment may require special handling considerations. Large, mobile
equipment (e.g., front loader, bulldozer, or crane) typically requires a specially trained operator.
The operator may need to be available during the disposition survey to provide access to all areas
requiring survey (e.g., move the equipment to provide access to the bottom of tires or treads).
Other large items may require special equipment (e.g., a crane or lift) to provide access to all
areas requiring survey. Special health and safety issues (Section 5.2) may be required to ensure
protection of survey personnel from physical hazards (e.g., personnel or items falling from
heights, or large items dropping on personnel or equipment). It may be necessary to partially or
totally disassemble large pieces of equipment to provide access and ensure measurability.
Piles of M&E may involve special handling precautions. Piles of dispersible M&E (e.g.,
excavated soil or concrete rubble) may need to be rearranged to match the assumptions used to
develop the instrument efficiency. For example, a conical pile of excavated soil may need to be
flattened to a uniform thickness to ensure measurability. If the M&E consists of or contains a
significant amount of dust, precautions against generating an airborne radiation hazard may be
necessary. Because many dust control systems use liquids to prevent the dust from becoming
airborne, it may be necessary to account for dust control impacts on measurability of the M&E.
For example, adding water to control dust will make it more difficult to measure alpha
radioactivity. Piles of scrap may also present other health and safety concerns along with issues
related to measurability. Sharp edges, pinch points, and unstable piles are examples of handling
problems that may need to be addressed.
Small pieces of M&E may be surveyed individually or combined into groups for survey. Care
should be taken when combining items to prevent mixing impacted and non-impacted items, or
mixing items with different physical or radiological attributes (see Section 2.2 and Section 5.4).
The moving of materials at a given site may require labeling as a quality control measure to
ensure M&E movement is tracked and documented. Labeling will help avoid the commingling of
impacted and non-impacted materials, and facilitate the staging and storage of impacted and non-
impacted M&E in appropriate areas.
January 2009 5-5 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
5.3.3 Transport the M&E
Identification of impacted and non-impacted areas within a facility will assist in selecting areas
for storing, staging, and surveying impacted M&E. In general, impacted M&E should be stored,
staged, and surveyed in impacted areas. Care should be taken when moving or handling impacted
M&E to prevent the spread of radionuclides to non-impacted areas. M&E in areas with airborne
radioactivity issues should be moved to protect the personnel conducting surveys and reduce the
possibility of contaminating survey instruments.
Disposition surveys can be performed with the M&E in place, or the M&E can be moved to
another location. For example, work areas with high levels of radioactivity may make it difficult
or resource intensive to meet the MQOs for measurement detectability (Sections 5.7 and 7.5) or
quantifiability (Sections 5.8 and 7.6). Moving the M&E to areas with lower levels of
radioactivity will help reduce radiation exposure for personnel conducting surveys and facilitate
meeting the survey objectives.
5.4 Segregate the M&E
The purpose of segregation is to separate M&E based on the estimated total measurement
uncertainty, ease of handling, and disposition options. Segregation is based on the physical and
radiological attributes determined during the IA (Chapter 2), not only on radionuclide
concentrations or radiation levels (i.e., classification).
In general, segregation based on measurement uncertainty should consider the physical and
radiological attributes that affect efficiency (i.e., geometry and fluence rate). M&E with simple
geometries, such as drums (cylinder) and flat surfaces (plane), should be separated from M&E
with complex geometries. Fluence rate is affected by location of the radioactivity (i.e., surficial
or volumetric) as well as surface effects (e.g., rough or smooth), density of the M&E, and type
and energy of radiation. High fluence rates are associated with surface radioactivity with high
energy on flat smooth surfaces made from materials with high atomic number (due to increased
backscatter). Volumetric activity, shielded surfaces, alpha or low energy or beta radiations,
irregular shapes, or rough surfaces can cause lower fluence rates. All of these factors should be
considered when segregating M&E.
Segregation of M&E should be performed conservatively. This means that the user should
separate M&E when they are not obviously similar. It is always possible to combine M&E but it
is not always practical or possible, to separate M&E once they have been combined. For
example, consider a facility where all the waste materials (e.g., paper, wood, metal, broken
equipment) are combined into a single "trash pile." When the planning team considers different
measurement methods and disposition options, they identify an innovative measurement method
that only applies to non-ferrous scrap metal. This would allow for recycling of these materials
with significant cost recovery as opposed to disposal. If the cost of re-segregating the M&E is
not offset by the value of recycling these materials, it may not be practical to segregate the non-
ferrous metals.
NUREG-1575, Supp. 1 5-6 January 2009
-------
MARSAME Implement The Survey Design
It is important to note that segregation does not require physical separation. Consider a generic
large box geometry, such as an empty shipping container or railroad car. The large, flat sides
could be considered separate survey units from the corners. Therefore, separate surveys would be
designed for the corners and the sides even though the entire railroad car would remain intact
throughout implementation of the disposition survey. Alternatively (or additionally), obvious
flaws, corrosion areas, or damaged areas could be segregated from the areas in good condition.
Even if the entire object is eventually surveyed using a single in situ measurement (e.g., in situ
gamma spectroscopy) it is important to segregate the M&E (at least conceptually) so an adequate
evaluation of alternate measurement methods can be performed (Section 5.9).
Handling of M&E during disposition surveys should also be considered during segregation
(Section 5.3). Physical characteristics of the M&E should be considered when segregating based
on handling requirements. Small, light items are easier to move and gain access to all surfaces
than large, massive items. M&E that will require preparation (e.g., disassembly, crushing,
chopping) prior to survey should be segregated from M&E that can be surveyed in their present
form. Disposition options should also be considered when segregating M&E. M&E that can be
reused or recycled should be segregated from M&E that is being considered for disposal.
Selection of disposition options is discussed in Section 2.4.
5.5 Set Measurement Quality Objectives
A number of terms with specific statistical meanings are used in this and subsequent sections.
These terms are defined in Chapter 7. The concept of Measurement Quality Objectives (MQOs)
and in particular the required measurement method uncertainty is introduced in Section 3.8.1.
These ideas are discussed in greater detail in the Multi-Agency Radiological Laboratory
Analytical Protocols manual (MARLAP 2004) Chapter 3 and Appendix C. While MARLAP is
focused on radioanalytical procedures, these concepts are applicable on a much broader scale and
will be used in MARSAME to guide the selection of measurement methods for disposition
surveys for materials and equipment.
Section 4.2 discusses the DQO process for developing statistical hypothesis tests for the
implementation of disposition decision rules using measurement data. These concepts are further
developed in Chapter 7. This includes formulating the null and alternative hypotheses, defining
the gray region using the action level and discrimination limit, and setting the desired limits on
potential Type I and Type II decision error probabilities that a decision-maker is willing to
accept for project results. Decision errors are possible, at least in part, because measurement
results have uncertainties. Because DQOs apply to both sampling and measurement activities,
method performance characteristics specifically for the measurement process of a particular
project are needed from a measurement perspective. These method performance characteristics
(Section 3.8) are the measurement quality objectives (MQOs).
DQOs define the performance criteria that limit the probabilities of making decision errors by—
• Considering the purpose of collecting the data,
• Defining the appropriate type of data needed, and
• Specifying tolerable probabilities of making decision errors.
January 2009 5-7 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
DQOs apply to both sampling and measurement activities. MQOs can be viewed as the
measurement portion of the overall project DQOs (Section 3.8). MQOs are —
• The part of the project DQOs that apply to the measured result and its associated uncertainty,
• Statements of measurement performance objectives or requirements for a particular
measurement method performance characteristic (e.g., measurement method uncertainty and
detection capability),
• Used initially for the selection and evaluation of measurement methods, and
• Used subsequently for the ongoing and final evaluation of the measurement data.
Measurement method uncertainty refers to the predicted uncertainty of a measured value that
would be calculated if the method were applied to a hypothetical sample with a specified
concentration. Measurement method uncertainty is a characteristic of the measurement method
and the measurement process. Measurement uncertainty, as opposed to sampling uncertainty, is a
characteristic of an individual measurement.
The true measurement method standard deviation, crM, is a theoretical quantity and is never
known exactly, but it may be estimated using the methods described in Section 7.4. The
estimated value of ou will be denoted here by ou and called the "measurement method
uncertainty." The measurement method uncertainty, when estimated by uncertainty propagation,
is the predicted value of the combined standard uncertainty ("one-sigma" uncertainty) of the
measurement for material with concentration equal to the upper bound of the gray region
(UBGR). Note that the term "measurement method uncertainty" and the symbol uu actually
apply not just to the measurement method but also to the entire measurement process: it should
include uncertainties in how the measurement method is actually implemented. This definition of
measurement method uncertainty is independent of the null hypothesis and applies to both
Scenario A and Scenario B.
The true standard deviation of the measurement method, CTM, is unknown, but the required
measurement method uncertainty, OMR, is intended to be an upper bound for OM. In practice, OM is
actually used as an upper bound for the method uncertainty, CTM, which is an estimate of ou.
Therefore, the estimated value of OUR will be called the "required measurement method
uncertainty" and denoted by UMR. Note that when referring to a theoretical population standard
deviation, the symbol a is used. Estimates of the value of a in specific cases are denoted by the
symbol u, for uncertainty. An uncertainty is not a standard deviation because its evaluation
involves concepts from metrology as well as statistics, however, in many cases it is treated
mathematically as if it were a standard deviation.
The principal MQOs in any project will be defined by the required measurement method
uncertainty, UMR, at and below the UBGR and the relative required measurement method
uncertainty, <^, at and above the UBGR:
UBGR
NUREG-1575, Supp. 1 5-8 January 2009
-------
MARSAME Implement The Survey Design
Section 7.7 provides the rationale and guidance for establishing project-specific MQOs for
controlling UM.
Note: When making decisions about individual measurement results, UMR usually should be about
0.3A, and when making decisions about the mean of several measurement results, UMR usually
should be about 0.1 A, where A is the width of the gray region, A = UBGR - LBGR. These rules
of thumb require certain assumptions as discussed in Chapter 7.
This check of measurement quality against the required measurement method uncertainty relies
on having realistic estimates of the measurement uncertainty. Often reported measurement
uncertainties are underestimated, particularly if they are confined to the estimated Poisson
counting uncertainty (Section 7.8). Tables of results are sometimes presented with a column
listing simply "+" without indicating how these numbers were obtained. Often it is found that
they simply represent the square root of the number of counts obtained during the measurement.
The method for calculating measurement uncertainty, approved by both the International
Organization for Standardization (ISO) and the National Institute of Standards and Technology
(NIST) is discussed in the next section.
5.6 Determine Measurement Uncertainty
This section discusses the evaluation and reporting of measurement uncertainty. Measurements
always involve uncertainty, which must be considered when measurement results are used as part
of a basis for making decisions. Every measured and reported result should be accompanied by
an explicit uncertainty estimate. One purpose of this section is to give users of data an
understanding of the causes of measurement uncertainty and of the meaning of uncertainty
statements; another is to describe procedures that can be used to estimate uncertainties. Much of
this material is derived from MARLAP Chapter 19.
In 1980, the Environmental Protection Agency published a report entitled "Upgrading
Environmental Radiation Data," which was produced by an ad hoc committee of the Health
Physics Society (EPA 1980). Two of the recommendations of this report were that-
1. Every reported measurement result (x) should include an estimate of its overall uncertainty
(MX) that is based on as nearly a complete assessment as possible, and
2. The uncertainty assessment should include every significant source of inaccuracy in the
result.
The concept of traceability is also defined in terms of uncertainty. Traceability is defined as the
"property of the result of a measurement or the value of a standard whereby it can be related to
stated references, usually national or international standards, through an unbroken chain of
comparisons all having stated uncertainties" (ISO 1996). Thus, to realistically make the claim
that a measurement result is "traceable" to a standard, there must be a chain of comparisons
(each measurement having its own associated uncertainty) connecting the result of the
measurement to that standard.
January 2009 5-9 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
This section considers only the measurement standard deviation, OM. Reducing sampling
standard deviation,
-------
MARSAME
Implement The Survey Design
such. The rate at which this decision error occurs is denoted by a. The critical value depends
directly on the value of a. The planning team has to make a choice about the establishment of the
acceptable rate for mischaracterizing a background count for a real detection count, i.e., establish
a Type I error rate, a, for mistakenly deciding a background measurement is really a detection of
additional radiation or radioactivity.
Radioactivity measurements are often recorded as counts or count rates. Radiation exposure
measurements are often expressed in different terms, e.g., ionization current. The term
"instrument signal" is used in the following so that all types of measurement are included.1
The relationship between the critical value of the net instrument signal (or count), Sc, and the
minimum detectable net instrument signal, SD, is shown in Figure 5.2. More detail on the
calculation of the minimum detectable value of the net instrument signal (or count), SD, is given
in Section 7.9. The net instrument signal obtained for a blank sample will usually be distributed
around zero as shown. Occasionally, a net instrument signal above Sc may be obtained by
chance. The probability that this happens is controlled by the value of a, the Type I decision
error rate, shown as the lightly shaded area in Figure 5.2. Smaller values of a result in larger
values of Sc and vice versa. The minimum detectable value of the net instrument signal SD is that
value of the mean net instrument signal that results in a detection decision with probability I - ft.
That is, there is only a probability ft, the Type II decision error rate shown as the more darkly
shaded area in Figure 5.2, of yielding an observed instrument signal less than Sc. Smaller values
of ft result in larger values of SD and vice versa. The planning team has to decide what an
acceptable value of ft should be, i.e. when additional radiation or radioactivity is present, at what
rate is it acceptable to mistakenly attribute the measurement result to only background. Note that
SD depends on the values of both a and/?.
Figure 5.2 The Critical Value (Sc) and the Minimum Detectable Value (SD) of the Net Instrument
Signal (or Count)
The MDC is usually obtained from the minimum detectable value of the net instrument signal (or
count), SD. The MDC is by definition an estimate of the true concentration of the radiation or
radioactivity required to give a specified high probability that the measured response will be
1 "Net instrument signal," is used here as a general term, because many radiation-detection instruments may have
output other than "counts" (e.g., current for ionization chambers). In cases where the instrument output is in counts,
the term "net counts" can be substituted for the term "net instrument signal."
January 2009
5-11
NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
greater than Sc. The common practice of comparing a measured concentration to the MDC,
instead of Sc, to make a detection decision is incorrect.
To calculate the MDC, the minimum detectable value of the net instrument signal, SD, must first
be converted to the detectable value of the net instrument signal rate (often a count rate), Sn/ts
(s *), where ^s is the duration of the measurement in seconds. This in turn must be divided by the
instrument efficiency, e (s lfBq) to get the minimum detectable activity, yo- Finally, the
minimum detectable activity can be divided by the sample volume or mass to obtain the MDC.
At each stage in this process, additional uncertainty may be introduced by the uncertainties in
time, efficiency, volume, mass, etc. Prudently conservative values of these factors should be used
so that the desired detection power, 1 - ft, at the MDC is maintained. Another approach would be
to recognize that>t> itself has an uncertainty which can be calculated using the methods of
Section 7.8. Thus, any input quantity that is used to convert from SD ioyn that has significant
uncertainty, can be incorporated to assess the overall uncertainty in the MDC.
MARSAME recommends that when a detection decision is required, it generally should be made
by comparing the net instrument signal to its corresponding critical value. Expressions for Sc and
SD should be chosen that are appropriate for the structure and statistics of the measurement
process. An appropriate background should be used to predict the instrument signal produced
when there is no radioactivity present in the sample. The MDC should be used only as a MQO
for the measurement method. To make a detection decision, a measurement result should be
compared to Sc and never to the MDC. Finally, additional discussion of the calculation of the
MDCs is given in Section 7.9.
5.8 Determine Measurement Quantifiability
This section discusses issues related to measurement quantifiability. Much of this material is
derived from the MARLAP Chapter 20.
Action levels are frequently stated in terms of a quantity or concentration of radioactivity, rather
than in simply in terms of whether radioactivity is detected. In these cases, project planners may
need to know the quantification capability of a measurement method, or its capability for precise
measurement. The quantification capability is expressed as the smallest concentration of
radiation or radioactivity that can be measured with a specified relative measurement standard
deviation. This section explains an MQO called the minimum quantifiable concentration (MQC),
which may be used to describe quantification capabilities.
The MQC, _yg, is defined as the concentration at which the measurement process gives results
with a specified relative standard deviation, llkq, where kq is usually chosen to be 10 for
comparability. Thus, the MQC is generally the concentration at which the relative measurement
uncertainty is 10%.
Historically much attention has been given to the detection capabilities of radiation and
radioactivity measurement processes, but less attention has been given to quantification
capabilities. For some projects, quantification capability may be a more relevant issue. For
example, suppose the purpose of a project is to determine whether the 226Ra concentration on
NUREG-1575, Supp. 1 5-12 January 2009
-------
MARSAME Implement The Survey Design
material at a site is below an action level. Because 226Ra can be found in almost any type of
naturally occurring material, it may be assumed to be present in every sample, making detection
decisions unnecessary. The MDC of the measurement process obviously should be less than the
action level, but a more important question is whether the MQC is less than the action level.
A common practice in the past has been to select a measurement method based on the MDC,
which is defined in Section 5.7 and Section 7.5. For example, MARSSIM says:
During survey design, it is generally considered good practice to select a
measurement system with an MDC between 10-50% of the DCGL [action level].
Such guidance implicitly recognizes that for cases when the decision to be made concerns the
mean of a population that is represented by multiple measurements, criteria based on the MDC
may not be sufficient and a somewhat more stringent requirement is needed. The requirement
that the MDC (approximately 3-5 times CTM) be 10% to 50% of the action level is tantamount to
requiring that ou be 0.02 to 0.17 times the action level. In other words, the relative measurement
standard deviation should be approximately 10% at the action level. However, the concentration
at which the relative measurement standard deviation is 10% of the MQC when kq assumes its
conventional value of 10. Thus, a requirement that is often stated in terms of the MDC may be
more naturally expressed in terms of the MQC (e.g., by saying that the MQC should not exceed
the action level). Further details on calculating the MQC can be found in Section 7.10.
5.9 Select a Measurement Technique and Instrumentation Combination
The combination of a measurement technique with instrumentation, or measurement method, is
selected to implement a disposition survey design based on the ability to meet the MQOs (see
Sections 3.3.2 and 5.5). Note that measurement techniques are separate from survey designs. The
relationship between the two is explained in Sections 5.9.1.1, 5.9.1.2, and 5.9.1.3. A realistic
determination of the measurement method uncertainty (Section 5.6) is critical to demonstrating a
method meets the MQOs. Other considerations when selecting a measurement method include—
• Health and safety concerns (Section 5.2),
• M&E handling issues (Section 5.3),
• Segregation (Section 5.4),
• Measurement detectability (Section 5.7), and
• Measurement quantifiability (Section 5.8).
The measurement techniques discussed in Section 5.9.1 all can be classified as scanning
measurements (constant motion involved in the surveying procedure) or fixed measurements
(surveying discrete locations without motion). Fixed measurements consist of in situ
measurements (the detection instrument moves to the M&E or measures the M&E in its
entirety), and sampling (removing part of the M&E for separate analysis).
Instrumentation for performing radiological measurements is varied and constantly being
improved. Section 5.9.2 provides an overview of some commonly used types of instruments and
how they might be applied to disposition surveys. The purpose of the discussions on
January 2009 5-13 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
instrumentation is not to provide an exhaustive list of acceptable instruments, but to provide
examples of how instrumentation and measurement techniques can be combined to meet the
survey objectives. Additional information on instrumentation is found in Appendix D.
Section 5.9.3 provides information on selecting a combination of measurement technique and
instrumentation to provide a measurement method. It is necessary that the selected measurement
method meet the MQOs established during survey design (Section 3.8). Selection of
instrumentation can be an iterative process. The appropriate MQO (e.g., MDC, MQC) may not
be attainable with some measurement methods. In some cases selection of a different instrument
may be all that is necessary, while in other cases a different measurement technique or an
entirely different measurement method will need to be considered.
5.9.1 Select a Measurement Technique
A measurement technique describes how a measurement is performed. The detector can be
moved relative to the M&E (i.e., scanning), used to perform static measurements of the M&E in
place (i.e., in situ or direct measurements), or some representative portion of the M&E can be
taken to a different location for analysis (i.e., sampling). These three measurement techniques are
described in Sections 5.9.1.1, 5.9.1.2, and 5.9.1.3, respectively. Smears are a type of sampling,
where a portion of the removable radioactivity is collected (Section 5.9.1.4).
5.9.1.1 Scanning Techniques
Scanning techniques generally consist of moving portable radiation detectors at a specified
distance above the physical surface of a survey unit at some specified speed to meet the MQOs.
Alternatively, the M&E can be moved past a stationary instrument at a specified distance and
speed (e.g., conveyorized systems or certain portal monitors). Scanning techniques can be used
alone to demonstrate compliance with a disposition criterion (i.e., scan-only surveys, Section
4.4.1), or combined with sampling in a MARSSIM-type survey design (Section 4.4.3). Scanning
is used in MARSSIM-type surveys to locate radiation anomalies by searching for variations in
readings, indicating gross radioactivity levels that may require further investigation or action.
Scanning techniques can more readily provide thorough coverage of a given survey unit and are
often relatively quick and inexpensive to perform. Scanning often represents the simplest and
most practical approach for performing MARSAME disposition surveys.
Maintaining the specified distance and speed during scanning can be difficult, especially with
hand-held instruments and irregularly shaped M&E. Variations in source-to-detector distance
and scan speed can result in increased total measurement method uncertainty. Determining a
calibration function for situations other than surficial radionuclides uniformly distributed on a
plane can be complicated, and may also contribute to the total measurement method uncertainty.
5.9.1.2 In Situ Measurements
In situ measurements are taken by placing the instrument in a fixed position at a specified
distance2 from the surface of a given survey unit of M&E and taking a discrete measurement for
Measurements at several distances may be needed. Near-surface or surface measurements provide the best
indication of the size of the area of elevated radionuclide concentrations or radioactivity, and are useful for model
NUREG-1575, Supp. 1 5-14 January 2009
-------
MARSAME Implement The Survey Design
a pre-determined time interval. Single in situ measurements can be performed on individual
objects or groups of M&E. Multiple in situ measurements can be combined to provide several
different views of the same object, or used to provide measurements for a specified fraction of
the M&E. In situ measurements can also be performed at random or systematic locations,
combined with scanning measurements, in a MARSSIM-type survey design. In situ
measurements are used generally to provide an estimate of the average radionuclide
concentration or level of radioactivity over a certain area or volume defined by the calibration
function.
Determining a calibration function for situations other than radionuclides uniformly distributed
on a plane or through a regularly shaped volume (e.g., a disk or cylinder) can be complicated and
may contribute to the total measurement method uncertainty. In situ techniques are not typically
used to identify small areas or volumes of elevated radionuclide concentration or activity.
5.9.1.3 Sampling
Sampling consists of removing a portion of the M&E for separate laboratory analysis. This
measurement technique, when combined with laboratory analysis, surpasses the detection
capabilities of measurement techniques that may be implemented with the M&E left in place.
This facilitates the analysis of complicated radioisotope mixtures, difficult-to-measure
radionuclides, and extremely low concentrations of residual radioactivity. Sampling is used to
provide an estimate of the average radionuclide concentration or level of radioactivity for a
specified area or volume. The sample locations may be located using a random or systematic
grid, depending on the objectives of the survey. Sampling is typically combined with scanning in
a MARSSIM-type survey design, where sampling is used to evaluate the average concentration
or activity and scanning is used to identify small areas or volumes with elevated radionuclide
concentrations or radioactivity. Sampling may also be used to validate data collected using other
measurement techniques.
Sampling (combined with laboratory analysis) typically requires the most time for data
generation of all the surveying techniques discussed in this chapter and is often the most
expensive. Sampling is not an effective technique for identifying small areas or volumes of
elevated radionuclide concentrations or levels of radioactivity.
5.9.1.4 Smears
Smears are used to provide an estimate of removable surface radioactivity. Smears are also
referred to as smear tests, swipes, or wipes. Smears are a type of sample where a filter paper or
other substance is used to wipe a specified area of a surface. The filter paper or other substance is
then tested for the presence of radioactivity.
Individual smear results collected by hand usually have a high uncertainty because the fraction of
surface radioactivity transferred to the smear is unknown and variable and the surface area
implementation. Gamma measurements at one meter provide a good estimate of potential direct external exposure
(MARSSIM 2002).
January 2009 5-15 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
covered by the smear is variable. In addition, the results may vary with time due to
environmental factors or interactions of surface activity with the surface itself. Action levels for
removable activity based on smear measurements may include assumptions about the fraction of
surface radioactivity transferred to a single smear or specify a surface area to be smeared. For
example, DOT surface contamination guidelines assume that 10% of the surface radioactivity is
transferred to a single smear. Also, DOE Order 5400.5 Figure IV-1 (DOE 1993) provides
instructions for using smears to measure removable radioactivity. These instructions specify
wiping an area of 100 cm2 with a dry filter or soft absorbent paper while applying moderate
pressure. The instructions also discuss how to account for minor variations from the procedure.
Determining a collection or removal fraction for smears can be complicated. The uncertainty and
variability in the removal fraction estimate and surface area smeared can result in increased total
measurement method uncertainty. Using a template or cutout with a known area can help control
the variability in the area covered by a smear. Using a tool that applies consistent pressure while
collecting smears can reduce the variability in the fraction of radioactivity removed.
Implementing a protocol for preparing surfaces and sorting materials prior to survey can reduce
variability in surface textures and conditions resulting in lower variability in smear collection
conditions.
5.9.2 Select Instrumentation
This section briefly describes the typical types of instrumentation that may be used to conduct
MARSAME disposition surveys. More detailed information relevant to each type of instrument
and measurement method is provided in Appendix D.
5.9.2.1 Hand-Held Instruments
Hand-held instruments typically are composed of a detection probe (utilizing a single detector)
and an electronic instrument to provide power to the detector and to interpret data from the
detector to provide a measurement display. They may be used to perform scanning surveys or in
situ measurements. Hand-held measurements also allow the user the flexibility to constantly vary
the source-to-detector geometry for obtaining data from difficult-to-measure areas.
5.9.2.2 Volumetric Counters (Drum, Box, Barrel, 4-7i Counters)
Box counting systems typically consist of a counting chamber, an array of detectors configured
to provide 4-n counting geometry, and microprocessor-controlled electronics that allow
programming of system parameters and data-logging. Volumetric counters are used to perform in
situ measurements on entire pieces of small M&E.
5.9.2.3 Conveyorized Survey Monitoring Systems
Conveyorized survey monitoring systems automate the routine scanning of M&E. Conveyorized
survey monitoring systems typically perform scanning surveys by moving M&E through a
detector array on a conveyor belt. Conveyorized survey monitoring systems may be utilized to
NUREG-1575, Supp. 1 5-16 January 2009
-------
MARSAME Implement The Survey Design
take in situ measurements by halting the conveyor and continuing the measurement to improve
the detection efficiency.
5.9.2.4 In Situ Gamma Spectroscopy
Some in situ gamma spectroscopy (ISGS) systems consist of a small hand-held unit that
incorporates the detector and counting electronics into a single package. Other ISGS systems
consist of a semiconductor detector, a cryostat, a multi-channel analyzer (MCA) electronics
package that provides amplification and analysis of the energy pulse heights, and a computer
system for data collection and analysis. ISGS systems typically are applied to perform in situ
measurements, but they may be incorporated into innovative detection equipment set-ups to
perform scanning surveys.
5.9.2.5 Portal Monitors
Portal monitors utilize a fixed detector array through which M&E are passed to typically perform
scanning surveys (objects may also remain stationary within the detector array to perform in situ
measurements). Portal monitors typically are used to perform scanning surveys of vehicles.3 In
situ measurements may be utilized with portal monitors by taking motionless measurements to
improve the detection efficiency.
5.9.2.6 Laboratory Analysis
Laboratory analysis consists of analyzing a portion or sample of the M&E. The laboratory will
generally have recommendations or requirements concerning the amount and types of samples
that can be analyzed for radionuclides or radiations. Communications should be established
between the field team collecting the samples and the laboratory analyzing the samples. More
information on sampling is provided in Section 5.9.1.3. Laboratory analyses can be developed
for any radionuclide with any material, given sufficient resources. Laboratory analyses typically
require more time to complete than field analyses. The laboratory may be located onsite or
offsite. The quality of laboratory data typically is greater than data collected in the field because
the laboratory is better able to control sources of measurement method uncertainty. The planning
team should consider the resources available for laboratory analysis (e.g., time, money), the
sample collection requirements or recommendations, and the requirements for data quality (e.g.,
MDC, MQC) during discussions with the laboratory.
5.9.3 Select a Measurement Method
Table 5.1 and Table 5.2 illustrate the potential applications and associated size restrictions for
combinations of the instrument and measurement techniques discussed in Sections 5.9.1 and
5.9.2, respectively. Sampling followed by laboratory analysis is not included in these tables, but
is considered "GOOD" for all applications. Please note the following qualifiers:
3 Specialized vehicle monitors are available that monitor rates of change in ambient background to account for
differences in vehicles being scanned to improve measurement detectability.
January 2009 5-17 NUREG-1575, Supp. 1
-------
Implement The Survey Design
MARSAME
GOOD The measurement technique is well-suited for performing this application
FAIR The measurement technique can adequately perform this application
POOR The measurement technique is poorly suited for performing this application
NA The measurement technique cannot perform this application
Few A relatively small number, usually three or less
Many A relatively large number, usually more than three
Table 5.1 illustrates that most measurement techniques can be applied to almost any M&E and
type of radioactivity. The quantity of M&E to be surveyed becomes a major factor for the
selection of measurement instruments and techniques described in this chapter. Hand-held
measurements and techniques generally are the most efficient technique for surveying small
quantities of M&E.
Table 5.1 Potential Applications for Instrumentation and Measurement Technique Combinations
Radiation
Type
Hand-Held
Instruments
Volumetric
Counters
Portal
Monitors
In Situ Gamma
Spectroscopy
Conveyorized Survey
Monitoring Systems
In Situ Measurements
Alpha
Beta
Photon
Neutron
FAIR
GOOD
GOOD
GOOD
FAIR
FAIR
GOOD
FAIR
POOR
FAIR
GOOD
GOOD
NA
NA
GOOD
NA
FAIR
GOOD
GOOD
GOOD
Scanning Surveys
Alpha
Beta
Photon
Neutron
POOR
GOOD
GOOD
FAIR
NA
NA
NA
NA
POOR
FAIR
GOOD
FAIR
NA
NA
GOOD
NA
POOR
FAIR
GOOD
FAIR
Table 5.2 Survey Unit Size and Quantity Restrictions for Instrumentation and Measurement
Technique Combinations
Size of
Items
Number
of Survey
Units or
Items
Hand-Held
Instruments
Volumetric
Counters
Portal
Monitors
In Situ
Gamma
Spectroscopy
Conveyorized
Survey
Monitoring
Systems
In Situ Measurements
> 10m3
1 to 10 m3
-------
MARSAME
Implement The Survey Design
Table 5.2 Survey Unit Size and Quantity Restrictions for Instrumentation and Measurement
Technique Combinations (Continued)
Size of
Items
Number
of Survey
Units or
Items
Hand-Held
Instruments
Volumetric
Counters
Portal
Monitors
In Situ
Gamma
Spectroscopy
Conveyorized
Survey
Monitoring
Systems
Scanning Surveys
>10m3
1 to 10 m3
-------
Implement The Survey Design
MARSAME
Table 5.3 Advantages and Disadvantages of Instrumentation and Measurement Technique
Combinations
Instrument
Measurement
Technique
Advantages
Disadvantages
Hand-Held
Instruments
In Situ
Generally allows flexibility in media
to be measured
Detection equipment is usually
portable
Detectors are available to efficiently
measure alpha, beta, gamma, x-ray,
and neutron radiation
Generally acceptable for performing
measurements in difficult-to-
measure areas
Measurement equipment is
relatively low cost
May provide a good option for small
quantities of M&E
Requires a relatively large
amount of manual labor as a
surveying technique; may
make surveying large
quantities of M&E labor-
intensive
Detector windows may be
fragile
Most do not provide nuclide
identification
Hand-Held
Instruments
Scanning
Generally allows flexibility in media
to be measured
Detection equipment is usually
portable
Detectors are available to efficiently
measure beta, gamma, x-ray, and
neutron radiation
Generally good for performing
measurements in difficult-to-
measure areas
Measurement equipment is
relatively low cost
May provide a good option for small
quantities of M&E
Requires a relatively large
amount of manual labor as a
surveying technique; may
make surveying large
quantities of M&E labor-
intensive
Detector windows may be
fragile
Most do not provide nuclide
identification
Incorporates more potential
sources of uncertainty than
most instrument and
measurement technique
combinations
Potential ergonomic injuries
and attendant costs
associated with repetitive
surveys.
Hand-Held
Instruments
Smear
Only measurement technique for
assessing removable radioactivity
Removable radioactivity can be
transferred and assessed in a low
background counting area.
Instrument background may
not be sufficiently low.
Detectors with counting
sensitive region larger than
the smear surface area may
require counting adjustments
to account for inherent
backgrounds associated with
other media located under the
detector sensitive region.
NUREG-1575, Supp. 1
5-20
January 2009
-------
MARSAME
Implement The Survey Design
Table 5.3 Advantages and Disadvantages of Instrumentation and Measurement Technique
Combinations (Continued)
Instrument
Measurement
Technique
Advantages
Disadvantages
Volumetric
Counters
In situ
Able to measure small items
Designs are available to efficiently
measure gamma, x-ray, and alpha
radiation
Requires relatively small amount of
labor
May be cost-effective for measuring
large quantities of M&E
May not be suited for
measuring radioactivity in
difficult-to-measure areas
Size of instrumentation
may discourage portability
Portal
Monitors
In situ
Able to measure large objects
Designs are available to efficiently
measure gamma, x-ray, and neutron
radiation
Requires relatively small amount of
labor
May be cost-effective for measuring
large quantities of M&E
Not ideal for measuring
alpha or beta radioactivity
May not be ideal for
measuring radioactivity in
difficult-to-measure areas
Size of detection
equipment may discourage
portability
Portal
Monitors
Scanning
Able to measure large objects
Efficient designs available for
gamma, x-ray, and neutron radiation
Residence times generally are short
May not require objects to remain
stationary during counting
Requires relatively small amount of
labor
May be cost-effective for measuring
large quantities of M&E
Not ideal for measuring
alpha or beta radioactivity
Source geometry is an
important consideration
May not be ideal for
measuring radioactivity in
difficult-to-measure areas
Size of detection
equipment may discourage
portability
In Situ
Gamma
Spectroscopy
(ISGS)
In situ
Provides quantitative measurements
with flexible calibration
Generally requires a moderate
amount of labor
May be cost-effective for measuring
large quantities of M&E
Instrumentation may be
expensive and difficult to
set up and maintain
May require liquid
nitrogen supply (with
ISGS semiconductor
systems)
Size of detection
equipment may discourage
portability
In Situ
Gamma
Spectroscopy
(ISGS)
Scanning
Provides quantitative measurements
with flexible calibration
Generally requires a moderate
amount of labor
May be cost-effective for measuring
large quantities of M&E
Instrumentation may be
expensive and difficult to
set up and maintain
May require liquid
nitrogen supply (with
ISGS semiconductor
systems)
Size of detection
equipment may discourage
portability
January 2009
5-21
NUREG-1575, Supp. 1
-------
Implement The Survey Design
MARSAME
Table 5.3 Advantages and Disadvantages of Instrumentation and Measurement Technique
Combinations (Continued)
Instrument
Measurement
Technique
Advantages
Disadvantages
Conveyorized
Survey
Monitoring
Systems
In situ
Requires relatively small amount of
labor after initial set up
May be cost-effective for measuring
large quantities of M&E
Instrumentation may be
expensive and difficult to
set up and maintain
May not be ideal for
assessing radioactivity in
difficult-to-measure areas
Size of detection
equipment may discourage
portability
Typically does not provide
nuclide identification
Conveyorized
Survey
Monitoring
Systems
Scanning
• Requires relatively small amount of
labor after initial set up
• May be cost-effective for measuring
large quantities of M&E
Instrumentation may be
expensive and difficult to
set up and maintain
May not be ideal for
assessing radioactivity in
difficult-to-measure areas
Size of detection
equipment may discourage
portability
Typically does not provide
nuclide identification
Laboratory
Analysis
Sampling
Generally provides the lowest
MDCs and MQCs, even for
difficult-to-measure radionuclides
Allows positive identification of
radionuclides without gammas
Most costly and time-
consuming measurement
technique
May incur increased
overhead costs while
personnel are waiting for
analytical results
Great care must be taken
to ensure samples are
representative
Detector windows may be
fragile
Laboratory
Analysis
Smear
Only measurement technique for
assessing removable radioactivity
Removable radioactivity can be
transferred and assessed in a low
background counting area.
Instrument background
may not be sufficiently
low.
Detectors with counting
sensitive region larger than
the smear surface area may
require counting
adjustments to account for
inherent backgrounds
associated with other
media located under the
detector sensitive region.
NUREG-1575, Supp. 1
5-22
January 2009
-------
MARSAME Implement The Survey Design
5.9.4 Measurement Performance Indicators
Measurement performance indicators are used to evaluate the performance of the measurement
method. These indicators describe how the measurement method is performing to ensure the
survey results are of sufficient quality to meet the survey objectives.
5.9.4.1 Blanks
Blanks are measurements of materials with little or no radioactivity and none of the
radionuclide(s) of concern present, and performed to determine whether the measurement
process introduces any increase in instrument signal rate that could impact the measurement
method detection capability. Blanks should be representative of all measurements performed
using a specific method (i.e., combination of instrumentation and measurement technique).
When practical, the blank should consist of the same or equivalent material(s) as the M&E being
surveyed.
Blanks typically are performed before and after a series of measurements to demonstrate the
measurement method was performing adequately throughout the survey. At a minimum, blanks
should be performed at the beginning and end of each shift. When large quantities of data are
collected (e.g., scanning measurements) or there is an increased potential for radionuclide
contamination of the instrument (e.g., removable or airborne radionuclides), blanks may be
performed more frequently. In general, a blank should be collected whenever enough
measurements have been performed such that it is not practical to repeat those measurements if a
problem is identified.
A sudden change in a blank result indicates a condition requiring immediate attention. Sudden
changes are caused by the introduction of a radionuclide, a change in ambient background, or
instrument instability. Gradual changes in blank values indicate a need to inspect all survey areas
for sources of radionuclides or radioactivity. Gradual build up of removable radionuclides over
time or instrument drift and deterioration can result in slowly increasing blank values. High
variability in blank values can result from instrument instability or improper classification (i.e.,
high activity and low activity M&E combined into a single survey unit. It is important to correct
any problems with blanks to ensure measurement detectability (see Sections 5.7 and 7.5) is not
compromised.
5.9.4.2 Replicate Measurements
Replicate measurements are two or more measurements performed on the same M&E, and
performed primarily to provide an estimate of precision for the measurement method. The
reproducibility of measurement results should be evaluated by replicates to establish this
component of measurement uncertainty (see Sections 5.6 and 7.4).
Replicates typically are performed at specified intervals during a survey (e.g., 5% of all
measurements or once per day), and should be used to evaluate each batch of data used to
support a disposition decision (e.g., one replicate per survey unit). For single measurement
January 2009 5-23 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
surveys or scan-only surveys where decisions are made based on every measurement, typically
5% of all measurements are replicated.
Precision exhibits a range of values and depends in part on the material being measured and the
activity level. Small changes in precision are expected, and the acceptable range of variability
should be established prior to initiating data collection activities. The main causes for lack of
precision include problems with repeating measurements on irregularly shaped M&E, the
material being measured, counting statistics when the activity levels are low, and instrument
contamination.
5.9.4.3 Spikes and Standards
Spikes and standards are materials with known composition and radioactivity, used to evaluate
bias in the measurement method, and typically performed periodically during a survey (e.g., 5%
of all measurements or once per day). When spikes and standards are available, they should be
used to evaluate each batch of data used to support a disposition decision (i.e., at least one spike
or standard per survey unit).
M&E cover a broad range of physical forms and materials that can change a measurement
method's expected bias. Tracking results of measurements with known activity can provide an
indication of the magnitude of bias. In general, activity levels near the action levels (or
discrimination limits in Scenario B) will provide adequate information on the performance of the
measurement system.
5.9.5 Instrument Performance Indicators
Instrument performance indicators provide information on how an instrument is performing.
Evaluation of these indicators provides information on the operation of the instruments.
5.9.5.1 Performance Tests
Performance tests should be performed periodically and after maintenance to ensure that the
instruments continue to meet performance requirements for measurements. An example of a
performance test is a test for response time. Performance requirements should be met as
specified in the applicable sections of ANSI N323A (ANSI 1997), ANSI N42.17A (ANSI
2003b), and ANSI N42.17C (ANSI 1989). These tests may be conducted as part of the
calibration procedure.
5.9.5.2 Functional Tests
Functional tests should be performed prior to initial use of an instrument. These functional tests
should include—
• General condition,
• Battery condition,
NUREG-1575, Supp. 1 5-24 January 2009
-------
MARSAME Implement The Survey Design
• Verification of current calibration (i.e., check to see that the date due for calibration has not
passed),
• Source and background response checks (and other tests as applicable to the instrument), and
• Constancy check.
The effects of environmental conditions (temperature, humidity, etc.) and interfering radiation on
an instrument should be established prior to use. The performance of functional tests should be
appropriately documented. This may be as simple as a checklist on a survey sheet, or may
include more detailed statistical evaluation such as a chi-square test.
5.9.5.3 Instrument Background
All radiation detection instruments have a background response, even in the absence of a sample
or radiation source (Section 3.4.2). Inappropriate background correction will result in
measurement error and increase the uncertainty of data interpretation.
5.9.5.4 Efficiency Calibrations
Detector efficiency is critical for converting the instrument response to activity (MARSAME
Section 7.8.2.2, MARSSIM Section 6.5.4, MARLAP Chapter 16). Routine performance checks
may be used to demonstrate the system's operational parameters are within acceptable limits, and
these measurements typically are included in the assessment of bias. The system's operational
parameters may be tracked using control charts.
5.9.5.5 Energy Calibrations (Spectrometry Systems)
Spectrometry systems identify radionuclides based on the energy of the detected radiations. A
correct energy calibration is critical to accurately identify radionuclides. An incorrect energy
calibration may result in misidentification of peaks, or failure to identify radionuclides present in
the M&E being investigated.
5.9.5.6 Peak Resolution and Tailing (Spectrometry Systems)
The shape of the full energy peak is important for identifying radionuclides and quantifying their
activity with Spectrometry systems. Poor peak resolution and peak tailing may result in larger
measurement uncertainty, or in failure to identify the presence of peaks based on shape.
Consistent problems with peak resolution indicate the presence of an analytical bias.
5.9.5.7 Voltage Plateaus (Gas Proportional Systems)
The accuracy of results using a gas proportional system can be affected if the system is not
operated with its detector high voltage adjusted such that it is on a stable portion of the operating
plateau.
January 2009 5-25 NUREG-1575, Supp. 1
-------
Implement The Survey Design MARSAME
5.9.5.8 Self Absorption, Backscatter, and Crosstalk
Alpha and beta measurement results can be affected by the M&E through self-absorption and
backscatter. Measurement systems simultaneously detecting alpha and beta particles using an
electronic discriminator (e.g., gas flow proportional detectors) can be affected by crosstalk (i.e.,
identification of alpha particles as beta particles and vice versa). Accurate differentiation
between alpha and beta activity depends on the assessment and maintenance of information on
self-absorption and crosstalk.
5.10 Report the Results
Once the instruments have been checked to ensure proper operation, the data should be collected
in a manner consistent with the survey design. Any field changes and deviations from survey
design should be documented and described in sufficient detail to enable an independent
recreation and evaluation at some future time.
The reported measurements should comprise raw data that includes background radioactivity
(i.e., gross measurement data). Electronic instruments with data logging capabilities should be
used when applicable. Electronic data should be exported and backed up periodically to
minimize the chance of losing data and the need for re-surveying.
Use of a measurement identification system should be considered. If required by the objectives
of the survey, the identification system should be developed and used such that each
measurement is assigned and labeled with a unique (preferably sequential) identifying number,
the collection date and time, the measurement location, and any applicable comments.
While MARSAME does not make specific recommendations with regard to approved media
formats for storing documentation, some users of MARSAME (e.g., private industry nuclear
power plants) may be required to retain documentation in media formats prescribed by State and
Federal rules of evidence. Similarly, State and Federal rules of evidence may specify retention
periods for documentation that exceed internal facility requirements. Compliance with State and
Federal rules of evidence is intrinsic to maintaining legally defensible records for insurance and
litigation-related purposes.
Projects at large, complex facilities often occur over relatively long time frames (e.g., years or
decades). In many cases the project is divided into smaller sub-projects that are performed as
resources and information become available. Retention of records, data compatibility, data
accessibility, and transfer of data between sub-projects should be considered during the
performance of individual surveys.
Documentation of the survey measurements should provide a complete and unambiguous record
of the data collected. Documentation should also include descriptions of variability and other
conditions pertaining to the M&E that may have affected the measurement capabilities of the
survey procedure, and photographs where applicable. The documentation itself should be clear,
legible, retained, retrievable, and to the level of detail required.
NUREG-1575, Supp. 1 5-26 January 2009
-------
MARSAME Implement The Survey Design
Negative results (net activity below zero) can be obtained when an instrument background is
subtracted from the measurement of a low activity sample. In the case where the activity is close
to zero, the measurement uncertainty will result in a distribution of results where approximately
one-half are less than zero and one-half are greater than zero. As long as the magnitude of
negative values is comparable to the estimated measurement uncertainties and there is no
discernible negative bias, negative results should be accepted as legitimate estimates of
radionuclide concentrations or levels of radioactivity associated with the M&E. A preponderance
of negative results, even if they are close to zero may indicate a bias or systematic error.
The inclusion of the information described above is important in creating comprehensive
documentation to make disposition surveys technically and legally defensible. The collection of
all necessary data prepares the MARSAME user to assess the results of the disposition survey,
which is discussed in Chapter 6.
January 2009 5-27 NUREG-1575, Supp. 1
-------
-------
MARSAME Evaluate the Survey Results
6 EVALUATE THE SURVEY RESULTS
6.1 Introduction
The assessment phase of the data life cycle involves the interpretation of survey results.
Interpretation of survey results is very straightforward when all of the data are below or all of the
data are above the action level, and the correct decision regarding disposition of the M&E is
obvious. In these cases very little data interpretation is required. However, formal statistical tests
provide a valuable tool when the survey results are neither clearly above nor entirely below the
action level. In either case, statistical tests always can be used to support the survey design in
helping to ensure the quantity and quality of data meet the data quality objectives (DQOs) and
measurement quality objectives (MQOs). Figure 6.1 illustrates the assessment phase of the data
life cycle.
6.2 Conduct Data Quality Assessment
Data quality assessment (DQA) is a scientific and statistical evaluation that determines whether
data are the right type, quality, and quantity to support their intended use (EPA 2006b). There are
five steps in the DQA process:
1. Review the DQOs and survey design.
2. Conduct a preliminary data review.
3. Select the statistical test.
4. Verify the assumptions of the statistical test.
5. Draw conclusions from the data.
The effort applied to DQA should be consistent with the graded approach used to develop the
survey design. More information on DQA can be found in Data Quality Assessment: A User's
Guide (EPA QA/G-9R, EPA 2006b) and Data Quality Assessment: Statistical Tools for
Practitioners (EPA QA/G-9S, EPA 2006c). Data should be verified and validated as described in
the quality assurance project plan (QAPP). Guidance on data verification and validation can be
found in MARSSIM Section 9.3 and MARLAP Chapter 8. Guidance on developing a QAPP is
available in EPA QA/G-5 (EPA 2002a) and MARLAP Chapter 4.
6.2.1 Review the Data Quality Objectives and Survey Design
The first step in the DQA process is a review of the DQO outputs used to develop the survey
design to ensure they are still applicable. The review of the DQOs and survey design should also
include the MQOs (e.g., measurement uncertainty, detectability, quantifiability). For example, if
the data show the measurement uncertainty exceeds the estimate used to design the survey, the
DQOs and MQOs should be revisited.
The survey design should be reviewed for consistency with the DQOs. For example, the review
should verify that the appropriate number or amount of measurements were performed in the
correct locations and were analyzed using measurement methods with adequate sensitivity.
January 2009 6-1 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results
MARSAME
From Figure 5.1
Conduct Data Quality
Assessment
(Section 6.2)
Proceed to Figure 6.4
-Yes
Is the Survey Design
Scan-Only or in Situ?
Proceed to Figure 6.5
Return from Figure 6.4
Evaluate the Results and
Make a Decision
(Section 6.8)
Return from Figure 6.5
Document the Disposition
Survey Results
(Section 6.10)
NOTE: Shaded box
represents important
milestone.
No Further Action
Figure 6.1 The Assessment Phase of the Data Life Cycle
In cases where the survey did not involve taking discrete measurements or samples (i.e., scan-
only, conveyor systems, or in situ surveys), it is imperative that the minimum detectable
concentrations (MDCs) be calculated realistically and they truly reflect at least 95% probability
that concentrations at or about the MDC were detected. Clearly, MDCs must be capable of
detecting radionuclide concentrations or levels of radioactivity at or below the upper bound of
the gray region (UBGR). When detection decisions are made for individual items (i.e., Scenario
B) the MDC should be less than or equal to the UBGR.
The minimum quantifiable concentration (MQC) is defined as the radionuclide concentration or
level of radioactivity at which the measurement method gives results with a specified relative
standard deviation l/£g, where kq is usually chosen to be 10 (see Section 5.8, MARLAP Section
NUREG-1575, Supp. 1
6-2
January 2009
-------
MARSAME Evaluate the Survey Results
19.4.5, MARLAP Section 19.7.3). MARSAME recommends that the MQC should be no larger
than the upper bound of the gray region (UBGR) when making quantitative comparisons of the
mean survey data to the action level (i.e., Scenario A). This is an expression of the fact that the
MQC, unlike the MDC used for a simple detection decision, addresses the relative uncertainty of
the data value obtained. If the objective of the disposition survey is to quantify radionuclide
concentrations near the UBGR, the MQC should be no larger than the UBGR.1
For MARSSIM-type surveys (Section 4.4.3) it is important to collect sufficient data to support a
disposition decision. This is particularly important in cases where the radionuclide
concentrations are near the action level. This can be done prospectively during survey design to
test the efficacy of a proposed survey design (see Chapter 4), or retrospectively during
interpretation of survey results to demonstrate the objectives of the survey design have been
achieved. The procedure for generating power curves for the Sign test and the Wilcoxon Rank
Sum test are provided in Appendix I of MARSSIM. Note that the accuracy of a prospective
power curve depends on estimates of data variability and the planned number of measurements.
After the data are analyzed, the sample standard deviation provides an estimate of data
variability and the actual number of valid measurements are known, and these two parameters
are used to generate a retrospective power curve (see MARSSIM Appendix I). The consequence
of inadequate power is an increased Type II decision error rate. For Scenario A, this means M&E
that actually meet the release criteria have a higher probability of being incorrectly determined
not to meet the release criterion. For Scenario B, this means M&E that actually do not meet the
release criterion have a higher probability of being incorrectly determined to meet the release
criterion.
6.2.2 Conduct a Preliminary Data Review
A preliminary data review is performed to learn more about the structure of the data by
identifying patterns, relationships, or potential anomalies. The preliminary data review includes
reviewing quality assurance (QA) and quality control (QC) reports, performing a graphical data
review, and calculating basic statistical quantities.
6.2.2.1 Review Quality Assurance and Quality Control Reports
Quality assurance reports describing data collection and reporting processes provide valuable
information about potential problems with or anomalies in the data. EPA QA/G-9R (EPA 2006b)
recommends a review of (1) data validation reports that document the data collection, handling,
analysis, reduction, and reporting procedures; (2) QC reports from laboratories or field stations
that document measurement system performance including data from blanks, replicates, spikes,
standards, and certified reference materials, or other internal QC measures; and (3) technical
systems reviews, performance evaluation audits, and audits of data quality including data from
performance evaluation measurements. EPA QA/G-9R (EPA 2006b) also suggests paying
particular attention to information that can be used to check assumptions made during survey
design using the DQO process, especially any anomalies in recorded data, missing values,
deviations from SOPs, or the use of nonstandard data collection methods (e.g., new, emerging, or
"cutting edge" technology). Verification of instrument calibrations and review of MQOs are
1 The UBGR is either the action level for Scenario A or the discrimination limit for Scenario B (see Section 4.2).
January 2009 6-3 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
particularly important to disposition surveys. Periodic measurements must be made to ensure the
measurement systems remain within acceptable calibration and control limits.
Quality control measurements are performed during implementation of the survey design to
monitor performance of the measurement methods, identify problems, and initiate corrective
actions when necessary. The evaluation of QC measurements used to control measurement
methods is distinct from the evaluation from survey results. MARLAP Section 18.3 ("Evaluation
of Performance Indicators"), Attachment ISA ("Control Charts"), and Attachment 18B
("Statistical Tests for QC Results") provide information on the evaluation of quality control
measurements.
Reviewing QA and QC reports is the only preliminary data review performed for surveys where
individual measurements are not recorded (e.g., scan-only surveys with hand-held instruments).
This increases the importance of the QA and QC reports and should be considered during survey
planning to ensure data quality is adequate to meet the survey objectives.
6.2.2.2 Perform a Graphical Data Review
Preparing and evaluating graphs and other visual depictions of the data may identify trends in the
data that go unnoticed using purely numerical methods. The graphical data review may include
posting plots, frequency plots, quantile plots, or other methods for visually interpreting data.
General guidance on performing a graphical data review and exploratory data analysis is
provided in EPA QA/G-9R (EPA 2006b) and by the National Institute of Science and
Technology (NIST 2006). A graphical data review cannot be performed unless the measurement
results are recorded. Surveys where recording individual measurement results is not required
(e.g., scan-only surveys with hand-held instruments) do not receive a graphical data review.
A posting plot is simply a map of the survey unit with the data values entered at the measurement
locations. This type of plot potentially reveals heterogeneities in the data, especially possible
clusters of elevated radionuclide concentrations. For a reference material survey a posting plot
can reveal spatial trends in background data that might affect the results of the statistical tests.
If the posting plot reveals systematic spatial trends in the M&E, the cause of the trends should be
investigated. In some cases the trends could be attributable to residual radioactivity, but they may
also be caused by inhomogeneities in the ambient background in the area the survey is
performed. EPA QA/G-9S (EPA 2006c) provides additional diagnostic tools for examining
spatial trends. The role of a posting plot for a conveyorized system would be a time series
display of the data showing any trends between adjacent batches of M&E conveyed past the
detector.
The geometric configuration of most M&E survey units composed of a few large irregularly
shaped pieces of M&E is transitory. The arrangement of tools and piles of scrap metal, for
example, changed as volumes of material were moved, or even as individual pieces were handled
during the survey (Section 5.3). In these cases some identifying marks, numbers, or bar-code
labels should have been used to identify and track where measurements were made, at least until
it is determined that the M&E meet the disposition criteria. Such marking and labeling need not
be permanent, but may be made with materials such as chalk or removable labels.
NUREG-1575, Supp. 1 6-4 January 2009
-------
MARSAME Evaluate the Survey Results
A frequency plot, or histogram, is a useful tool for examining the general shape of a distribution.
This plot is a bar chart of the number of data points within a certain range of values. A frequency
plot reveals any obvious departures from symmetry, such as skewness or bimodality (two peaks),
in the data distributions for the M&E or reference material.
The presence of two peaks in the M&E data set frequency plot may indicate the presence of
small areas of elevated activity. In some cases it may be possible to identify an appropriate
background distribution within the M&E data set. This type of data interpretation generally
depends on site-specific considerations and should only be pursued after consultation with the
responsible regulatory agency.
The presence of two peaks in the M&E or reference material frequency plots may also indicate a
mixture of materials with different intrinsic radiation backgrounds. The greater variability in the
data caused by the presence of such a mixture reduces the power of the statistical tests. These
situations should be avoided whenever possible through segregation of M&E (see Section 5.4)
and carefully matching the reference materials to the M&E being surveyed.
When data are obtained from scan-only surveys incorporating data loggers, large quantities of
data are usually recorded. In essence, 100% of Class 1 M&E are measured. While the survey
coverage may be less than 100% for Class 2 and Class 3 M&E, the number of data points is still
likely to be large. As long as there was no bias in the selection of areas that were scanned, the
frequency plot will be close to the population distribution of radioactivity levels in the M&E.
The mean and standard deviation calculated from these logged values should be very close to the
corresponding population values.
For conveyorized survey monitors, the data may be interpreted batch-by-batch as it is scanned. In
this case, the data treatment would be most similar to a single in situ measurement used to
evaluate all of the M&E. If, on the other hand, the data were logged continuously the data
treatment would be similar to a scan-only survey using data loggers.
6.2.2.3 Calculate Basic Statistical Quantities
Radiological survey data are usually obtained in units (e.g., counts per unit time) that have no
intrinsic meaning relative to the action levels. For comparison of survey data to action levels,
survey data from laboratory and field analyses are converted into action level units. MARSSIM
Section 6.6 provides guidance on data conversion. Any uncertainty associated with data
conversion should be included in the estimate of measurement uncertainty (Section 5.6). For
surveys where individual results are not recorded (e.g., scan-only surveys with hand-held
instruments) the uncertainty is associated with converting the action level into the units provided
by the instrument in the field. Because individual results are not recorded, no statistical quantities
can be calculated.
Basic statistical quantities that should be calculated for the sample data set include the mean,
standard deviation, and the median. Other statistical quantities may be calculated based on the
survey objectives.
January 2009 6-5 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
Example 1: Suppose the following 10 measurement results are obtained from a disposition
survey:
9.1, 10.7, 13.6, 3.4, 13.3, 7.9, 4.5, 7.7, 8.3, 10.4
The mean of the data (u) is 8.89 and the standard deviation (o) is 3.3231.
The next 10 measurement results are from an appropriate matching reference material:
6.2, 13.8, 15.2, 9.3, 6.7, 4.9, 7.1, 3.6, 8.8, 8.9
The mean of the reference data (u) is 8.45 and the standard deviation (o) is 3.6713.
The means of the two data sets can be compared to provide a preliminary indication of the
survey unit status. 2 The difference is 0.44, with the M&E being investigated having a higher
mean concentration. If the mean for the M&E exceeds the mean for the reference material by
more than the action level, the M&E clearly do not meet the disposition criterion. On the other
hand, if the difference between the largest M&E measurement (13.6 for this example) and the
smallest reference material measurement (3.6 for this example) is below the action level, the
M&E will pass the Wilcoxon Rank Sum (WRS) test (Section 6.6), but will have to meet other
criteria as well.
The value of the sample standard deviation is especially important. If the standard deviation is
too large compared to what was assumed for variability during development of the survey
design, this may indicate an insufficient number of samples were collected to achieve the desired
power for the statistical test. As previously mentioned, inadequate power can lead to an increase
in the Type II decision error rate.
The median is the middle value of the data set when the number of data points is odd or the mean
of the two middle values when the number of data points is even. A large difference between the
mean and the median indicate a potential skew in the data. This would also be evident in a
histogram of the data.
Examining other statistical quantities such as the maximum, minimum, and range may provide
additional useful information. When there are 30 or fewer data points, range values greater than
4 or 5 standard deviations would be unusual.
Example 2: For the example M&E data set the minimum is 3.4 and the maximum is 13.6.
The range is 13.6 - 3.4 = 10.2. The range is equal to 3.1 standard deviations (i.e., 10.2/3.3).
Thus, the range for this example data set is not unusually large. The range may be greater for
larger data sets.
Note the use of significant digits in this example. Because all of the numbers in the text are interim values in
calculating the difference between two means, they are not rounded. If the mean and standard deviation values were
to be reported as results they would be rounded to two significant digits because the original data is a mixture of
numbers with two and three significant digits. If the data were rounded after each calculation, the difference in the
rounded means appears to be 0.4 (i.e., 8.9 minus 8.5), but the actual difference is 0.44 based on the un-rounded
means (i.e., 8.89 minus 8.45). This is an example of how rounding numbers too early in the process can result in
additional uncertainty.
NUREG-1575, Supp. 1 6-6 January 2009
-------
MARSAME Evaluate the Survey Results
6.2.3 Select the Statistical Tests
In most cases the selection of a statistical test is determined by the survey design used to collect
the data. The most appropriate procedure for summarizing and analyzing the data is chosen
based on the preliminary data review. If the preliminary data review indicates that the
assumptions used to develop the survey design are valid, the statistical tests and evaluation
methods determined should then be applied. If the assumptions used to develop the survey
design are determined to be invalid, it may be necessary to consult a statistician to determine the
most appropriate statistical test for evaluating the survey results.
6.2.3.1 Scan-Only Surveys
Scan-only surveys generate large amounts of data. Class 1 surveys measure all of the M&E.
When less than 100 percent of the M&E are measured (i.e., Class 2 or Class 3 surveys) the areas
that are measured are assumed representative of the areas that are not measured. This assumption
should be checked during the preliminary data review (Section 6.2.2). The radionuclide
concentrations or radioactivity in the areas that are not measured can be inferred based on the
measurement results in the areas that are measured. Data indicating this inference may not be
reasonable should result in re-evaluation of the survey design. For example, suppose the survey
design specifies that 137Cs is the radionuclide of concern and scanning 50% of the M&E is
appropriate based on the expected distribution of radionuclide concentrations, expected levels of
radioactivity, and the beta-gamma emissions from the radionuclide of concern. If additional
historical data is found showing 239Pu is also a radionuclide of concern, the survey design should
be re-evaluated based on the presence of an alpha emitting radionuclide as well.
If disposition decisions will be made for individual items or based on individual measurement
results, all of the results should be compared to the action level. Comparison to the action level
based on a detection decision or measurement (Section 5.7) is discussed in Section 6.3.
Individual measurement results can be recorded for scan-only surveys. The benefit of logging
individual measurement results is the ability to statistically evaluate the data (e.g., calculate a
mean and an upper confidence limit). If disposition decisions will be made based on the mean of
logged data, an upper confidence limit for the mean is calculated and compared to the UBGR.
This means that compliance with the disposition criterion can be demonstrated for the entire
survey unit, even if some of the results exceed the UBGR. Evaluations using the upper
confidence limit are discussed in Section 6.4. When less than 100% of the M&E are measured
(i.e., Class 2 and Class 3 surveys), the total uncertainty includes both spatial and measurement
uncertainty. Measuring 100% of the M&E (i.e., Class 1 survey) accounts for spatial variability,
but there is still an uncertainty component resulting from variability in the measurement process.
Conveyorized systems that continually log the survey results also generate large amounts of data.
An upper confidence limit for the mean can be used for the evaluation of data from these types of
systems (see Section 6.4) in the same manner as logged scan data. Conveyorized systems that
operate in a batch mode are essentially treated as single in situ measurements of small batches of
M&E. The results generated by these types of systems are evaluated as a series of comparisons
to the UBGR; using detection decisions based on the MDC (Section 6.3).
January 2009 6-7 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
6.2.3.2 In Situ Surveys
In situ surveys may consist of a series of isolated measurements covering all or part of the M&E,
a series of measurements with overlapping fields of view incorporating all (Class 1) or a portion
(Class 2 or Class 3) of the M&E, or a single measurement incorporating all of the M&E (Section
4.4.2).
Similar to scan-only surveys, if disposition decisions will be made for individual items or based
on individual measurement results, all of the results should be compared to the action level.
Comparison to the action level based on a detection decision (Section 5.7) is discussed in Section
6.3. Unlike scan-only surveys, in situ surveys are likely based on a limited number of data points.
To perform in situ measurements, assumptions were made about the distribution of radioactivity
within the volume of M&E being measured. These assumptions are inherent in the calibration of
in situ measurement systems and the validity of these assumptions determines the
appropriateness of the measurement. It is important to account for uncertainty in these
assumptions when calculating the MDC and to evaluate these assumptions using QC
measurements performed during the survey. If there is uncertainty about the true MDC or critical
value, use conservative values for the efficiency as described in Section 7.5.2.
6.2.3.3 MARSSIM-Type Survey Designs
MARSSIM-type survey designs generally are used when instrumentation for scan-only or in situ
measurement surveys do not provide sufficient sensitivity (e.g., the MDC is greater than the
UBGR). A statistically based number of measurements is used to provide an estimate of the
mean activity in each survey unit, and scanning is used to identify small areas of elevated
activity between sample locations.
The number of measurements is determined by the statistical test. In most cases the statistical
tests used in MARS SIM are appropriate for Scenario A. The criteria for choosing between the
Sign test and the WRS test are described in MARSSIM Section 8.2.3. In general, when the
radionuclide is not present in background (or its background concentration is negligible
compared to the action level) and radionuclide-specific measurements are made, the Sign test
(Section 6.5) is used. Otherwise, the WRS (Section 6.6) test should be used. The Sign test is
designed to detect whether there is radioactivity in the M&E above the action level. The WRS
test is used to compare measurements of the M&E to measurements performed on the reference
material.
When Scenario B is used, the statistical tests described in NUREG-1505 (NRC 1998a) generally
are used. The Sign test and the WRS test are still used, but the application of the test is adjusted
to account for the difference in the null hypothesis. When using Scenario B, there is a potential
for the WRS test to miss non-uniform radioactivity (i.e., slightly elevated radionuclide
concentrations or levels of radioactivity over a portion of the survey unit). Randomization of the
M&E through mixing or homogenization can eliminate this possibility. If randomization is not
practical, the Quantile test (Section 6.7) should be used to evaluate survey units when the WRS
test fails to reject the null hypothesis.
NUREG-1575, Supp. 1 6-8 January 2009
-------
MARSAME Evaluate the Survey Results
The results of scanning measurements performed as part of a MARSSIM-type survey are
evaluated using the elevated measurement comparison (EMC). The EMC is simply a comparison
to an action level (see Section 6.3). The action level used for the EMC is the action level for
small areas of elevated activity. If there is no action level for elevated activity, the scanning
results are compared to the action level for the mean activity in the survey unit. Additional
information on the EMC is available in MARSSIM Section 8.5.1 and NUREG-1505 Chapter 8
(NRC 1998a).
6.2.4 Verify the Assumptions of the Tests
An evaluation to determine the data are consistent with the underlying assumptions of the
statistical tests helps to validate the use of a particular test. One may also determine that certain
departures from these assumptions are acceptable when given the actual data and other
information about the project. The nonparametric tests described in this chapter assume that the
data from the M&E or the reference material consist of independent measurements from each
distribution. The primary issue associated with the evaluation of scan-only and single in situ
measurement survey data is the MDC or MQC as discussed in Section 6.2.1.
Asymmetry in the data can be identified using a histogram or a Quantile plot. Information on
histograms and Quantile plots is provided in MARSSIM Appendix I and NUREG-1505 Section
4.2.2 (NRC 1998a). As discussed in Section 6.2.2.3, data transformations can sometimes be used
to minimize the effects of asymmetry.
One of the primary advantages to using the nonparametric tests is that they involve fewer
assumptions about the data than their parametric counterparts. If parametric tests are used (e.g.,
Student's t test) any additional assumptions made in using these tests should be verified (e.g.,
testing for normality). These issues are discussed in detail in EPA QA/G-9S (EPA 2006c).
One of the more important assumptions made in the survey design is that the number of
measurements is sufficient to achieve the DQOs set for the Type I (a) and Type II (/?) decision
error rates. Verification of the power of the statistical tests (I-/?) may be of particular interest.
Methods for assessing power are discussed in Appendix 1.9 of MARSSIM. If there is not
reasonable assurance the DQOs have been achieved, additional investigations including
repeating the survey may be needed. The planning team can develop survey designs cautiously to
avoid unnecessary and potentially costly decision errors by—
• Estimating the potential data variability conservatively,
• Taking more measurements than suggested by the DQO process, and
• Estimating the MDCs conservatively.
In the absence of other data, each of these estimates could be multiplied by a safety factor of 1.2
(i.e., increase the estimate by 20%). Examples of assumptions and possible methods for
evaluating and verifying these assumptions are summarized in Table 6.1.
January 2009 6-9 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results
MARSAME
Table 6.1 Issues and Assumptions Underlying the Evaluation Method
Evaluation Method
Compare single
measurements to a limit
(Section 6.3)
Compare an upper
confidence limit for the
mean to a limit
(Section 6.4)
Statistical Tests
(Sections 6.5, 6.6, 6.7)
Issue
Verify the MDC and
Measurement Uncertainty
Verify the MQC and
Measurement Uncertainty
Verify the Assumptions of the
Statistical Test (e.g., spatial
independence, symmetry, data
variance, power)
Verification Method
Review the MDC
Review QA/QC Reports
Review IA and DQOs
Review the Measurement
Uncertainty
Review QA/QC Reports
Review IA and DQOs
Preliminary Data Review
(e.g., posting plot,
histogram, summary
statistics, power curve)
Survey Type
Scan-Only
In situ
Scan Only
In situ
MARSSIM-
Type Survey
Verification of scan-only and in situ survey results focuses on the estimates of the MDC and
MQC values used to design the survey. If the assumptions used to estimate these values are
incorrect, the survey design may be invalid.
The first step in evaluating the MDC and MQC is to review the assumptions used to develop
these values. In general, the key assumptions are made in determining the source and detector
efficiencies. QA and QC reports should be reviewed to evaluate measurement performance (e.g.,
scan speed, source geometry, distance from M&E to the detector, non-uniform response of large
area detectors). The description of the M&E from the IA should be compared to the assumptions
used to develop the efficiency.
In some cases it may be possible to compare the survey results of multiple measurement
techniques. For example, if there are multiple radiations associated with the M&E it may be
possible to compare gamma measurement results to alpha or beta measurement results to verify
the survey results. Direct measurements may provide more quantitative results for areas of
elevated activity identified during scan-only surveys.
It may be possible to use an entirely different survey method to provide information to support
verification of assumptions used to design a survey. For example, smears or surface scrapings
can be used to verify the presence of radionuclides or radioactivity on the surface.3
In situ measurements or sample collection and analysis may be used to verify the results of scan-
only survey designs. Care must be taken to ensure comparability of survey methods before
evaluating the results to avoid generating conflicting results. For example, consider an in situ
survey used to demonstrate the mean activity is less than the action level. A scan-only survey
method is used to verify the results and identifies an area of elevated activity. This discrepancy
in results warrants additional investigation of the small area of elevated activity. The additional
investigation should determine if the activity in this area actually causes the mean activity to
exceed the disposition criterion.
! This smear procedure does not rule out additional volumetric activity.
NUREG-1575, Supp. 1
6-10
January 2009
-------
MARSAME
Evaluate the Survey Results
6.2.5 Draw Conclusions from the Data
The types of measurements performed on M&E are—
• Scans,
• In situ or direct measurements at discrete locations, and
• Samples collected at discrete locations.
Specific details for conducting the Sign test and the WRS test are provided in Sections 6.5 and
6.6, respectively. When the data clearly show that the M&E meets or exceeds the disposition
criterion, the result is often obvious without performing the formal statistical analysis. This is the
expected outcome for Class 2 and Class 3 surveys. Table 6.2 summarizes examples of
circumstances leading to specific conclusions based on a simple examination of the data.
Table 6.2 Summary of Evaluation Methods and Statistical Tests
Evaluation Method or
Statistical Test
Comparison to a Limit (AL=0)
- Scenario B only
- Results may or may not be
recorded
- Scan-only or In situ surveys
Comparison to a Limit (AL^O)
- Scenario A or B
- Results not recorded
- Scan-only or In situ surveys
Comparison to Upper
Confidence Limit
- Scenario A or B
- Results must be recorded
- Scan-only or In situ surveys
Sign Test
- Radionuclide not in
background
- Nuclide-specific
measurements
- Scenario A or B
- MARSSIM-type surveys
Wilcoxon Rank Sum Test
- Radionuclide in background
- Nuclide non-specific
measurements
- Scenario A or B
- MARSSIM-type surveys
Survey Result
All measurements less than the critical value
corresponding to the MDC (e.g., does not
exceed alarm set point)
Any measurement exceeds the critical value
corresponding to the MDC
All measurements less than the critical value
corresponding to the UBGR
Any measurement exceeds the critical value
corresponding to the UBGR
Upper confidence limit less than UBGR
Upper confidence limit greater than UBGR
All measurements less than the action level
Mean greater then the action level
Any measurement greater than the action level
and the mean less than the action level
Difference between maximum survey unit
measurement and minimum reference area
measurement is less than the UBGR
Difference of survey unit mean and reference
area mean is greater than the action level
Difference between any survey unit
measurement and any reference area
measurement greater than the action level or
the difference of survey unit mean and
reference area mean is less than the action
level
Conclusion
M&E meet the disposition
criterion
M&E do not meet the
disposition criterion
M&E meet the disposition
criterion
M&E do not meet the
disposition criterion
M&E meet the disposition
criterion
M&E do not meet the
disposition criterion
M&E meet the disposition
criterion
M&E do not meet the
disposition criterion
Conduct Sign test (and
elevated measurement
comparison, if necessary)
M&E meet the disposition
criterion
M&E do not meet the
disposition criterion
Conduct WRS test (and
elevated measurement
comparison, if necessary)
January 2009
6-11
NUREG-1575, Supp. 1
-------
Evaluate the Survey Results
MARSAME
Table 6.3 Summary of Evaluation Methods and Statistical Tests (Continued)
Evaluation Method or
Statistical Test
Quantile Test
- Test for non-uniform
radioactivity
- Combine with WRS test
- Scenario B only
- MARSSIM-type surveys
Survey Result
Difference between maximum survey unit
measurement and minimum reference area
measurement is less than the UBGR
Difference of survey unit mean and reference
area mean is greater than the action level
Difference between any survey unit
measurement and any reference area
measurement greater than the action level or
the difference of survey unit mean and
reference area mean is less than the action
level
Conclusion
M&E meet the disposition
criterion
M&E do not meet the
disposition criterion
Conduct Quantile test (and
elevated measurement
comparison, if necessary)
6.3 Compare Results to the UBGR
When disposition decisions will be made about individual items, or decisions will be based on
individual measurement results, each result (plus or minus a multiple of its combined standard
uncertainty) will be compared to the action level (see MARLAP Appendix C.4). In practice, this
means that any result that exceeds the critical value (Sc, see Section 5.7 and Section 7.5.1) when
the minimum detectable level (SD, see Section 5.7 and Section 7.5.2) equals the UBGR provides
evidence that the result exceeds the UBGR.
For Scenario A, if all the results are less than the action level, then the mean and the maximum
activity must also be below the action level. Thus, the radionuclide concentrations or levels of
radioactivity associated with the M&E demonstrate compliance with the disposition criterion.
For Scenario B when the action level is not zero or background, all of the results must be below
the critical value corresponding to the MDC set equal to the UBGR. If the action level is zero or
background, Scenario B must be used and any indication of the presence of radionuclide
concentrations or radioactivity above background (i.e., above the discrimination level) would
result in rejecting the null hypothesis. For this situation, any measurement result exceeding the
critical value corresponding to the required MDC indicates the potential presence of
radionuclides or radioactivity above background. This applies to single in situ measurements as
well as series of in situ measurements.
If there is an action level based on small areas of elevated activity or the maximum allowable
value, the individual results can be compared directly to the action level. This applies primarily
to the evaluation of scanning results for MARSSIM-type surveys (i.e., the EMC), but may be
applied to scan-only survey data as well.
6.4 Compare Results Using an Upper Confidence Limit
The use of the upper confidence limit (UCL) can apply to both Scenario A and B for scan-only
or in situ surveys where individual results are recorded. When disposition decisions are made
about the estimated mean of a sampled population, the assessment of the survey results is
accomplished by comparing a UCL for the mean to the UBGR. For scan-only surveys where
NUREG-1575, Supp. 1
6-12
January 2009
-------
MARSAME Evaluate the Survey Results
there are a large number of data points, a simple comparison of the mean activity to the UBGR
may be sufficient.
If individual scan-only survey results are recorded, a non-parametric confidence interval can be
used to evaluate the results of the disposition survey. Similarly, a confidence interval can be used
to evaluate a series of in situ measurements with overlapping fields of view. A one-tailed version
of Chebyshev's inequality or software (e.g., EPA's ProUCL software) can be used to evaluate
the probability of exceeding the UBGR (i.e., using a UCL). The use of a UCL applies to both
Scenario A (where the UBGR equals the action level) and Scenario B (where the UBGR equals
the discrimination limit).4
6.4.1 Calculate the Upper Confidence Limit
Chebyshev's inequality calculates the probability that the absolute value of the difference of the
true but unknown mean of the population and a random number from the data set is at least a
specified value. That is, given a specified positive number («), a mean (w), and a random number
from the data set (r), then the probability that \ju-r is greater than or equal to n is equal to a. In
addition, a one-tailed version of the inequality can be used to calculate a UCL for a data set that
is independent of the data distribution (i.e., there is no requirement to verify the data are from a
normal, lognormal, or any other specified kind of distribution) by letting the inequality equal the
UCL, as described in the following steps:
1. Calculate the mean (u) and standard deviation (o) of the number of results (n) in the data set.
2. For Scenario A, retrieve the Type I error rate (a ) used to design the survey.
3 . Using Chebyshev' s inequality, calculate the maximum UCL using equation 6-1 :
na n
4. For Scenario B, substitute the Type II error rate (ft) used to design the survey for a in
Equation 6-1.
5. If the maximum UCL is less than the UBGR, the survey demonstrates compliance with the
disposition criterion (i.e., reject the null hypothesis for Scenario A or fail to reject the null
hypothesis for Scenario B).
Chebyshev's inequality must be used with caution when there are very few points in the data set.
This is because the population mean and standard deviation in the Chebyshev formula are being
estimated by the sample mean and sample standard deviation. In a small data set from a highly
skewed distribution, the sample mean and sample standard deviation may be underestimated if
the high concentration but low probability portion of the distribution is not captured in the
sample data set. EPA has issued guidance on calculating UCLs for exposure point concentrations
(EPA 2002b).5 Software for implementing EPA's guidance is available (EPA 2006d).
4 In the case of Scenario B, if the action level is zero and the radionuclide of concern does not appear in background,
any positive radionuclide-specific detection would result in a rejection of the null hypothesis that there is zero
activity.
5 In MARSAME, "exposure point concentration" is used to mean a conservative estimate of the mean radionuclide
concentration(s) in or on M&E.
January 2009 6-13 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
6.4.2 Upper Confidence Limit Example: Class 1 Concrete Rubble
This example illustrates the survey design for concrete rubble using 3 inch x 3 inch Nal(Tl)
detectors mounted on a conveyorized survey system to measure 137Cs. A pile of concrete rubble
was loaded on the conveyor and passed beneath the detectors at a pre-determined speed. Each
one-second count recorded by a detector corresponds to approximately 9,800 cm3 of concrete
rubble (i.e., a 5-cm thick disk with a 50-cm diameter). The following information was used to
design the survey:
• The selected disposition option was clearance, using Scenario A with the null hypothesis that
the residual radioactivity exceeds the action level.
• The IA indicated the concrete was potentially volumetrically contaminated prior to being
converted to rubble.
• The concrete rubble had a maximum particle dimension of less than 0.5 cm.
• The average background count rate was estimated to be 38,000 cpm based on preliminary
surveys of non-impacted concrete, and was used for the LBGR.
• The action level was set at 20,000 cpm above the average background count rate, so the
UBGR was set at 58,000 cpm.
• The estimated standard deviation of background count rate is 2,500 cpm based on
preliminary survey data.
• The Type I decision error rate was set at 0.10, or 10%.
The survey consisted of 9,616 independent, one-second measurements that were recorded using a
data logger. The mean count rate for the survey was 39,252 cpm, with a standard deviation (a) of
5,465 cpm. The standard deviation of the mean, ov.was calculated using the following equation:
a 5 ,465 cpm rr ^ .^ _.
= 55'7 cpm (6"2)
As noted earlier, with such a large data set, one can expect that the sample mean and standard
deviation should be fairly close to their population values. The minimum count rate was 30,080
cpm, and the maximum count rate was 72,805 cpm. Note that although the mean concentration is
well below the action level, there are data points that exceed the action level. Thus, a test against
an UCL for the mean is warranted. Figure 6.2 shows a frequency plot of the survey results.
NUREG-1575, Supp. 1 6-14 January 2009
-------
MARSAME
Evaluate the Survey Results
1200 -
1000 -
1" 800 -
-------
Evaluate the Survey Results
MARSAME
P UCL Statistics for CPU
A B
1
2
3
4
5
6
Data File
C
D
E F G H
Variable: CPM
Raw Statistics
Number of Valid Samples
Number of Unique Samples
Minimum
7 Maximum
8
9
10
11
12
13
14
15
16
17
Mean
Median
Standard Deviation
Variance
Coefficient of Variation
Skewness
9616
Normal Distribution Test
Liliiefors Test Statisitic
6441 Liliiefors 5% Critical Value
30080 Data not normal at 5% significance level
72805
39251 .847
38267
5465.0563
29866840
0.1392306
12.4504964
1
0.2044466
0.0090352
95% UCL (Assuming Normal Distribution)
Student's-t UCL 39343.497
Gamma Distribution Test
A-D Test Statistic
A-D 5% Critical Value
K-S Test Statistic
Gamma Statistics K-S 5% Critical Value
k hat
k star (bias corrected)
18 Theta hat
19 ,Theta star
20
21
22
23
I
nu hat
nu star
Approx.Chi Square Value (.05)
Adjusted Level of Significance
Adjusted Chi Square Value
61 .605609
61 .586458
637.14729
637.34542
1184799.1
1184430.8
1181899.7
0.049975
1181899.4
26 Log-transformed Statistics
27 Minimum of log data
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Maximum of log data
Mean of log data
Standard Deviation of log data
Variance of log data
10.311616
11.19554
10.569616
0.1224578
0.0149959
RECOMMENDATION
Data are Non-p
arametric (C
.05)
Use Student's-t UCL
or Modified-t UCL
Data do not follow gamma distribution
at 5% significance level
585.96505
0.7522512
0.1788147
0.01814
95% UCLs (Assuming Garnrna Distribution)
Approximate Garnrna UCL
Adjusted Gamma UCL
Lognormal Distribution Test
39335.904
39335.917
Liliiefors Test Statisitic 0. 1 656286
Liliiefors 5% Critical Value
0.0090352
Data not lognorrnal at 5% significance level
95% UCLs (Assuming Lognormal Distribution)
95% H-UCL N/A
95% Chebyshev (MVUE) UCL
97.5% Chebyshev (MVUE) UCL
99% Chebyshev (MVUE) UCL
95% Non-parametric UCLs
CLT UCL
Adj-CLT UCL (Adjusted for skewness)
Mod-t UCL (Adjusted for skewness)
Jackknife UCL
Standard Bootstrap UCL
Bootstrap-t UCL
Hall's Bootstrap UCL
Percentile Bootstrap UCL
BCA Bootstrap UCL
95% Chebyshev (Mean, Sd) UCL
97.5% Chebyshev (Mean, Sd) UCL
99% Chebyshev (Mean, Sd) UCL
39441.02
39533.758
39715.923
39343.516
39345.004
39343.729
39343.497
39341 .993
39342.895
39343.322
39337.588
39344.539
39494.773
39599.887
39806.364
Figure 6.3 Screen Capture of Output from ProTJCL Software for the Sample Data Set
6.5 Conduct the Sign Test
The Sign test is used to compare the measurement results from each survey unit with the
applicable disposition criterion. The Sign test can be applied to either Scenario A or Scenario B.
The Sign test should only be used if the radionuclide being measured is not present in
NUREG-1575, Supp. 1
6-16
January 2009
-------
MARSAME Evaluate the Survey Results
background or if the radionuclide being measured is present at such a small fraction of the action
level as to be considered insignificant. Otherwise, the WRS test described in Section 6.6 should
be applied. Additional information on the Sign test can be found in Section 8.3 of MARSSIM
and Chapter 5 of NUREG-1505 (NRC 1998a).
6.5.1 Apply the Sign Test to Scenario A
The Sign test is applied to Scenario A by counting the number of measurements from each
survey unit that are less than the action level (i.e., UBGR). Each result is subtracted from the
action level (AL -X[), and the number of positive values is summed. The result is the test
statistic S+. Discard any measurement that is exactly equal to the action level and reduce the
sample size, N, by the number of such measurements. The value of S+ is compared to the critical
values in A.3. If S+ is greater than the critical value (q) in the table, the null hypothesis is
rejected.
6.5.2 Apply the Sign Test to Scenario B
The Sign test is applied to Scenario B in a manner similar to that used for Scenario A. However,
for Scenario B the action level (i.e., LBGR) is subtracted from each result (Xt - AL), and the
number of positive values is summed. The result is the test statistic S+. Discard any
measurement that is exactly equal to the action level and reduce the sample size, N, by the
number of such measurements. The value of S+ is compared to the critical values in Table A.3.
If S+ is greater than the critical value (q) in the table, the null hypothesis is rejected.
6.5.3 Sign Test Example: Class 1 Copper Pipes
This example illustrates the disposition survey design for copper pipe sections using a gas-flow
proportional counter to measure 239Pu. Because the alpha background on the copper material is
essentially zero, it was decided the Sign test would be used to determine whether the material
meets the disposition criterion. The sample size was determined using the DQO Process and
inputs such as the disposition option, action level, expected standard deviation of the
measurement results, and the acceptable probability of making Type I and Type II decision
errors.
The following inputs were used to develop the survey design-
• The selected disposition option was clearance.
• The survey was designed using Scenario A, with the null hypothesis that the residual
radioactivity exceeds the action level.
• The IA indicated that the inside surfaces of the pipes potentially came in contact with liquids
containing 239Pu, but the outside surfaces were non-impacted.
• The gross activity action level was 100 dpm/100 cm2. When converted to cpm the gross
activity action level was 10 cpm (i.e., total efficiency = 0.10 counts per disintegration).
• The LBGR (i.e., the DL) was set at the expected activity level on the copper pipe sections
(i.e., 5 net cpm, the same as the gross mean for an alpha background of 0).
• The standard deviation for the measurements was estimated at 2 cpm.
January 2009 6-17 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results
MARSAME
• The relative shift was calculated as (10-5)72 = 2.5.
• The Type I and Type II decision error rates were both set at 0.05.
Table A.2a shows the number of measurements estimated to be needed for the Sign test, TV, is 15
(a=0.05, /5=0.05, and A/er=2.5). Therefore, 15 surface activity measurements were randomly
collected from the inside surfaces of the copper pipe sections. Survey results are shown in
Table 6.3.
Table 6.3 Sign Test Example Data
Surface Concentration
(cpm/100 cm2)
4
3
11
1
1
4
6
3
9
6
14
1
4
10
2
Surface Concentration
(dpm/100 cm2)
40
30
110
10
10
40
60
30
90
60
140
10
40
100
20
< Action Level?
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
Number of measurements less than the action level (S+) =12
The surface activity values in Table 6.3 are determined by dividing the measured cpm by the
total efficiency (0.10). No probe area correction is necessary. The mean count rate is 5 cpm,
compared to the estimate of 5 cpm used for the LBGR, and the median is 4 cpm. The standard
deviation is 4 cpm, which is higher than the value of 2 used to develop the survey design.6 Thus,
the power of the test is lower than planned. With the actual value of the relative shift
(10-5)74=1.2, 23 measurements should be collected.
With the 15 measurements collected, the actual Type II decision error rate is between 0.10 and
0.25 (the closest entries in Appendix A, Table A.2a are for o=0.05, /?=0.10, and A/o=1.2 with
7V=18, and «=0.05, /?=0.25, and A/er=1.2 with 7V=12). Three measurements exceed the action
level. The portion of the material associated with these measurements merits further
investigation using the elevated measurement comparison described in MARSSIM Section 8.5.1.
6 Values are reported to one significant figure based on the data in Table 6.3. Interim calculations generally carry
extra figures, so rounding to the appropriate number of significant figures only occurs for the final calculation.
Rounding results too soon in the calculation may result in unnecessarily deleting individual results (i.e., when the
result is exactly equal to the UBGR) resulting in lower statistical power.
NUREG-1575, Supp. 1
6-18
January 2009
-------
MARSAME Evaluate the Survey Results
The value of S+, 12, was compared to the appropriate critical value, q, in Appendix A, Table
A.3. In this case, for 7V=15 and o=0.05, the critical value is 11. Because S+ exceeds q, reject the
null hypothesis that the survey unit exceeds the action level. In this case, the slight loss of power
attributable to underestimating the standard deviation did not affect the result. Pending the
outcome of the investigation of the three elevated measurements, this survey unit has satisfied
the disposition criteria established for clearance.
6.6 Conduct the Wilcoxon Rank Sum Test
The WRS test is used to compare each material survey unit with an appropriately chosen
reference material. Each reference material should be selected on the basis of its similarity to the
survey unit material, as discussed in Section 3.9. The WRS test can be applied to either Scenario
A or Scenario B. Further information on the WRS test can be found in Section 8.4 of MARSSIM
and Chapter 6 of NUREG- 1505 (NRC1998a).
6.6.1 Apply the WRS Test to Scenario A
The WRS test is applied to Scenario A as outlined in the following steps and further illustrated
by the example in Section 6.6.2.
1. Obtain the adjusted reference material measurements, Z;, by adding the action level to each
reference material measurement, Xt. Zt = Xt+ AL.
2. The m adjusted reference sample measurements, Z;, from the reference material and the n
sample measurements, Yt, from the survey unit are pooled and ranked in order of increasing
size from 1 to N, where N = m + n.
3. If several measurements are tied (i.e., have the same value), they are all assigned the mean
rank of that group of tied measurements.
4. If there are t "less than" values, they are all given the mean of the ranks from 1 to t.
Therefore, they are all assigned the rank t(t +l)/(2 f) = (t +l)/2, which is the mean of the first
t integers. If there is more than one MDC,7 all observations below the largest MDC should be
treated as "less than" values. If more than 40% of the data from either the reference material
or the survey unit are reported as less than detectable, the WRS test cannot be used.
5. The sum of all the ranks, which is the sum of the first TV positive integers, is N(N+l)/2, which
equals Wr added to Ws. Thus, one needs only to sum the ranks of the either the adjusted
reference measurements (Wr) or the sum of the ranks of the sample measurements (Ws).
6. Compare Wr with the critical value (q) given in Table A.4 for the appropriate values of n, m,
and a. If Wr is greater than the tabulated value for q, reject the hypothesis that the survey unit
exceeds the disposition criterion.
6.6.2 Apply the WRS Test to Scenario B
The WRS test is applied to Scenario B as outlined in the following steps:
7 Examples of situations where there could be more than one MDC include using multiple laboratories to perform
sample analyses and using different instruments with different backgrounds and different efficiencies to perform
measurements.
January 2009 6-19 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
1. Obtain the adjusted survey unit measurements, Z;, by subtracting the LBGR from each survey
unit measurement, Yt. Zt = Yt - LBGR.
2. The n adjusted survey unit measurements, Z;, and the m reference material measurements, Xf,
are pooled and ranked in order of increasing size from 1 to N, where N = m + n.
3. If several measurements are tied (i.e., have the same value), they are all assigned the mean
rank of that group of tied measurements.
4. If there are t "less than" values, they are all given the mean of the ranks from 1 to t.
Therefore, they are all assigned the rank t(t +l)/(2 f) = (t +l)/2, which is the mean of the first
t integers. If there is more than one MDC, all observations below the largest MDC should be
treated as "less than" values. If more than 40% of the data from either the reference material
or the survey unit are reported as less than detectable, the WRS test cannot be used.
5. Sum the ranks of the adjusted measurements from the survey unit, Ws. The sum of all the
ranks, which is the sum of the first TV positive integers, is N(N+l)/2, which equals Wr added
to Ws. Thus, one needs only to sum the ranks of the either the adjusted reference
measurements (Wr) or the sum of the ranks of the sample measurements (Ws).
6. Compare Ws with the critical value (q) given in Table A.4 for the appropriate values of n, m,
and a. (Note that when using this table for Scenario B, the roles of m and n are reversed. If
the Quantile test is being used in addition to the WRS test, then a/2 should be used rather
than a.) If Ws is greater than the tabulated value for q, reject the hypothesis that the difference
in the median concentration between the survey unit and the reference area is less than the
LBGR.
6.6.3 WRS Test Scenario A Example: Class 2 Metal Ductwork
This example illustrates the use of the WRS test for releasing Class 2 metal ductwork. Assume
that a gas-flow proportional detector was used to make gross (non-radionuclide-specific) surface
activity measurements.
The DQOs from this survey unit include a = 0.05 and/? = 0.05, and the action level converted to
units of gross cpm is 2,300 cpm, which is the UBGR. In this case, the WRS test is used because
the estimated background level (2,100 cpm) was large compared to the action level. The
estimated standard deviation of the measurements, a, is 375 cpm. The estimated added activity
level is 800 cpm; the LBGR is set at this value, and represents the DL. The relative shift is
calculated as A/er, which is (action level - LBGR)/er, which equals 4.
The sample size needed for the WRS test can be found in Table A.2b for these DQOs. The result
is nine measurements in each survey unit and nine in each reference material a = 0.05, and /? =
0.05, and A/er = 4). The ductwork was laid flat onto a prepared grid, and the 9 measurements
needed in the survey unit were made using a random-start triangular grid pattern. For the
reference materials, the measurement locations were chosen randomly on a suitable batch of
material. Table 6.4 lists the gross count rate data obtained.
NUREG-1575, Supp. 1 6-20 January 2009
-------
MARSAME
Evaluate the Survey Results
Table 6.4 Scenario A WRS Test Example Data
Data
(cpm)
2180
2398
2779
1427
2738
2024
1561
1991
2073
2039
3061
3243
2456
2115
1874
1703
2388
2159
Area
R
R
R
R
R
R
R
R
R
S
S
S
S
S
S
S
S
S
Adjusted
Data
4480
4698
5079
3727
5038
4324
3861
4291
4373
2039
3061
3243
2456
2115
1874
1703
2388
2159
Sum =
Ranks
15
16
18
10
17
13
11
12
14
3
8
9
7
4
2
1
6
5
171
Reference Material
Ranks
15
16
18
10
17
13
11
12
14
0
0
0
0
0
0
0
0
0
126
In the "Area" column, the code "R" denotes a reference material measurement and "S" denotes a
survey unit measurement. The adjusted data were obtained by adding the action level to the
reference material measurements (see Section 6.6.1, Step 1). The ranks of the data range from 1
to 18, because there are a total of 9+9 measurements (see Section 6.6.1, Step 2). Note that the
sum of all of the ranks is still 18(18+1)72 = 171. Checking this value with the formula in Step 5
of Section 6.6.1 is recommended to guard against errors in the rankings.
The total of the ranks belonging to the reference material measurements is 126. This is compared
with the entry for the critical value of 104 in Table A.4 for a = 0.05, with n = 9 and m = 9.
Because the sum of the reference material ranks is greater than the critical value, the null
hypothesis (i.e., that the mean survey unit concentration exceeds the action level) is rejected, and
the ductwork is released.
This conclusion can be reached quickly by noting the difference between the largest survey unit
measurement (3,243 cpm) and the smallest reference area measurement (1,427 cpm). This
difference (3,243 - 1,427 = 1,816 cpm) is less than the action level of 2,300 cpm. Because the
largest possible difference is less than the action level, the mean difference must also be less than
the action level.
6.6.4 WRS Test Scenario B Example: Class 2 Metal Ductwork
This example illustrates the use of the Scenario B WRS test for releasing Class 2 metal
ductwork, using the same data as in Section 6.6.3. The null hypothesis for Scenario B is that
there is no detectable radioactivity above background.
January 2009
6-21
NUREG-1575, Supp. 1
-------
Evaluate the Survey Results
MARSAME
In this case, the action level is set at no radioactivity detectable above the estimated background
level (2,100 cpm). The LBGR is equal to the action level, and is set to zero. The regulator
specified that the survey be able to detect an average excess of even 1,500 cpm being released.
This value is the DL. The UBGR is set equal to the DL (i.e., 1,500 cpm), with /? = 0.025. The
owner of the ductwork felt that there was very little if any radioactivity above background
present, and was willing to set a = 0.20. The estimated standard deviation of the measurements,
o, was 375 cpm. The relative shift is A/a = (UBGR - LBGR~)A7 = (1,500 - 0)/375 = 4.
The sample size needed for the WRS test can be found in Table A.2b. The result is 9
measurements in each survey unit and 9 in each reference material a/2 = 0.10, and/? = 0.025, and
A/er = 4. The data were obtained as in Section 6.6.3. Table 6.4 (on the previous page) lists the
gross count rate data obtained. These data were reanalyzed using Scenario B and the results are
shown in Table 6.5.
Table 6.5 Scenario B WRS Test Example Data
Data
(cpm)
2180
2398
2779
1427
2738
2024
1561
1991
2073
2039
3061
3243
2456
2115
1874
1703
2388
2159
Area
R
R
R
R
R
R
R
R
R
S
S
S
S
S
S
S
S
S
Adjusted Data
2180
2398
2779
1427
2738
2024
1561
1991
2073
2039
3061
3243
2456
2115
1874
1703
2388
2159
Sum =
Ranks
11
13
16
1
15
6
2
5
8
7
17
18
14
9
4
3
12
10
171
Survey Unit Ranks
0
0
0
0
0
0
0
0
0
7
17
18
14
9
4
3
12
10
94
In the "Area" column, the code "R" denotes a reference material measurement and "S" denotes a
survey unit measurement. The adjusted data would be obtained by subtracting the LBGR from
the survey unit measurements (see Section 6.6.2, Step 1), but because the LBGR is zero, no
adjustment is needed. The ranks of the adjusted data range from 1 to 18, because there are a total
of 9+9 measurements (see Section 6.6.2, Step 2). Note that the sum of all of the ranks is still
18(18+1)72 = 171. Checking this value with the formula in Step 5 of Section 6.6.2 is
recommended to guard against errors in the rankings. The total of the ranks belonging to the
survey unit measurements is 94. This is compared with the entry for the critical value of 100 in
Table A.4 for a = 0.10, with n = 9 and m = 9. Because the sum of the reference material ranks is
less than the critical value, the null hypothesis (i.e., that there is no detectable radioactivity above
background) is not rejected, and the ductwork may be released if the Quantile test is passed.
NUREG-1575, Supp. 1
6-22
January 2009
-------
MARSAME Evaluate the Survey Results
6.7 Conduct the Quantile Test
The Quantile test was developed to detect differences between the surveyed M&E and the
reference material that consist of a shift to higher values in only a fraction of the surveyed M&E.
The Quantile test is only performed when Scenario B is used, and only if the null hypothesis is
not rejected for the WRS test. Using the Quantile test, in tandem with the WRS test, results in
higher power to identify M&E that do not meet the disposition criterion than either test by itself.
Apply the Quantile test as follows:
1. Calculate aQ (O.Q = a/2).
2. Obtain the adjusted survey unit measurements, Z;, by subtracting the LBGR from each survey
unit measurement, Yt. Zt = Yt - LBGR.
3. The n adjusted survey unit measurements, Z;, and the m reference material measurements, Xf,
are pooled and ranked in order of increasing size from 1 to N, where N = m + n.
4. If several measurements are tied (i.e., have the same value), they are all assigned the mean
rank of that group of tied measurements.
5. Look up the values for r and q in Table A. 5 based on the number of measurements in the
survey unit («), the number of measurements in the reference area (m), and OQ. The
operational decision described in the next step is made using the values for r and q.
6. Ifq or more of the r largest measurements in the combined ranked data set are from the
survey unit, the null hypothesis is rejected.
This form of the Quantile test gives only approximate results, Because Table A. 5 provides a
limited number of combinations of n, m, and O.Q. It is recommended that several combinations of
w, m, and O.Q be considered when interpreting the results of the Quantile test. Sections 7.2 and 7.3
of NUREG-1505 (NRC 1998a) provide additional guidance on interpreting the results of the
Quantile test.
As an example, the Quantile test can be applied to the Class 2 Metal Ductwork example of
section 6.6.4. Using n = 9, m = 9, and O.Q= 0.10, the nearest entry in Table A.5d has for r = 3
q = 3 with OQ = 0.105 when n = 10 and m = 10. This means that all three of the highest
measurement would have to be from the survey unit in order to reject the null hypothesis. From
Table 6.5, one can see that the two largest measurements are from the survey unit, but the third
largest is from the reference area. Because the ductwork has passed both the WRS and the
Quantile test in the Scenario B example, one would conclude that it could be released from
radiological controls.
6.8 Evaluate the Results: The Decision
Once the data and results of the tests are obtained, the specific steps required to make a
disposition decision depends on the procedures approved by the regulator. The following
considerations are suggested for the interpretation of the test results with respect to the
disposition criteria. Note that the tests need not be performed in any particular order.
The interpretation of results from the data evaluation or statistical test is the decision to reject or
not to reject the null hypothesis. For some of the survey designs the decision is straightforward,
while for other designs the interpretation is more complex. Figures 6.4 and 6.4 summarize the
interpretation of results.
January 2009 6-23 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results
MARSAME
From Figure 6.1
Disposition
Decision Based
on Mean of a Sampled
Population?
Is the
AL Equal to Zero
or Background?
Disposition
Decision Based on
Individual
Items?
Requires Scenario B
LBGR = AL
Scan MDC < UBGR
Recording Individual
Scan Results
Not Required
Individual Results Must
be Recorded
All
Results < Sc
from the
MDC?
All
Results < S
from the
UBGR?
M&E Do Not Meet the \
Disposition Criterion
(Section 6.9) /
M&E Do Not Meet the
Disposition Criterion
(Section 6.9)
Return to Figure 6.1
Return to Figure 6.1
/ M&E Meet the
Y Disposition Criterion
Return to Figure 6.1
Figure 6.4 Interpretation of Survey Results for Scan-Only and In Situ Surveys
NUREG-1575, Supp. 1
6-24
January 2009
-------
MARSAME
Evaluate the Survey Results
NOTE: An elevated
measurement comparison
also needs to be performed
for MARSSIM-type surveys.
From Figure 6.1
Radionuchde
of Concern Present in
Background?
Radionuchde-
Specific
Measurements?
S+ > q?
- LBGR)
S+ > q?
(UBGR-Xi)
/M&E Do Not Meet th^\
Disposition Criterion
q or
more of the r
Largest Values
from the Survey
Unit?
Disposition Criterion
/ M&E Do Not Meet the
Disposition Criterion
V (Section 6.9)
Figure 6.5 Statistical Interpretation of Results for MARSSIM-Type Surveys
January 2009
6-25
NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
6.8.1 Compare Results to the UBGR
The process for interpreting results compared to the UBGR depends on the action level used to
develop the survey design. Refer to Table 6.1 for issues and assumptions underlying this
evaluation method.
If the action level is zero or background, Scenario B must be used:
• Compare every measurement result to the critical value corresponding to the required scan
MDC.
• If all results are below the critical value, the M&E demonstrate compliance with the
disposition criterion.
• Any results that exceed the critical value provide evidence of radionuclide concentrations or
radioactivity levels exceeding the disposition criteria, so the M&E do not demonstrate
compliance with the release criterion.
If the action level is not zero or background—
• Compare every measurement result to the critical value corresponding to the UBGR.
• If all results are below the critical value, the M&E demonstrate compliance with the
disposition criterion.
• Any results that exceed the critical value provide evidence of radionuclide concentrations or
radioactivity levels exceeding the disposition criteria, so the M&E do not demonstrate
compliance with the release criterion.
Scan-only results are usually available as the data are collected. This real-time availability of
results allows the surveyor to make decisions as the data are collected. M&E that exceed the
action level can be identified and segregated during implementation of the survey. This "clean as
you go" approach to surveys is only applicable for Class 1 surveys where there is high
confidence in the quality and accuracy of detection decisions around the UBGR. Extensive
documentation of the measurement process, previous applications of the process to the same or
similar M&E, and verification of MDCs and MQCs is generally necessary to implement a "clean
as you go" survey design.
6.8.2 Compare Results Using an Upper Confidence Limit
When decisions are made based on the mean of a sampled population, the survey results should
be evaluated by comparison to a UCL (refer to Table 6.1 for issues and assumptions underlying
this evaluation method):
• Compare every measurement result to the critical value corresponding to the UBGR.
• If all results are below the critical value, the M&E demonstrate compliance with the
disposition criterion.
• If any results are above the critical value, calculate the UCL (Section 6.4.1).
• If the UCL is less than the UBGR, the M&E demonstrate compliance with the disposition
criterion.
NUREG-1575, Supp. 1 6-26 January 2009
-------
MARSAME Evaluate the Survey Results
• If the UCL exceeds the UBGR, the M&E do not demonstrate compliance with the disposition
criterion.
• Investigate measurements exceeding the UBGR.
• Results above the UBGR trigger a reevaluation of classification as Class 2.
• Results above the MDC trigger a reevaluation of classification as Class 3.
6.8.3 Compare Results for MARSSIM-Type Surveys
The process for evaluating MARSSIM-type survey results is more complicated. This process is
explained in more detail in MARSSIM Section 8.5 (refer to Table 6.1 for issues and assumptions
underlying this evaluation method):
• Calculate the test statistics (see Section 6.5.1, 6.6.1, 6.6.2, and 6.7).
• Look up the critical value in the appropriate statistical table in Appendix A.
• Evaluate the results of the statistical test as described in Figures 6.3 and 6.4.
• Evaluate individual results using the elevated measurement comparison (EMC).
• M&E must pass the statistical test and the EMC (if applicable) to demonstrate compliance.
If the null hypothesis is rejected under Scenario A, there is sufficient evidence to show the
median radionuclide concentrations or radiation levels are below the disposition criterion. Under
Scenario B, failing to reject the null hypothesis means there is insufficient evidence to overturn
the initial assumption the M&E demonstrate compliance with the disposition criterion.
If the null hypothesis is rejected under Scenario B, additional investigations are required to
determine the final disposition of the M&E (see Section 6.8.2). Failure to reject the null
hypothesis under Scenario A also requires additional investigations.
6.9 Investigate Causes for Survey Unit Failures
When M&E fail to demonstrate compliance with the disposition criterion, the first step is to
review and confirm the data that led to the decision. Once this is done, the DQO process can be
used to evaluate potential problem areas leading to failure.
If the level of radioactivity on or in some Class 1 M&E exceeds the UBGR, the simplest solution
might be to segregate those items for a different disposition decision. The concept of "clean as
you go" for Class 1 M&E was discussed in Section 6.8.1 where individual objects or sample
locations were identified during implementation of the survey design. A simple modification to
this approach is to physically segregate the objects exceeding the action level as they are
identified, or after reanalysis shows the cleaning was not effective. The segregated M&E can
then be evaluated for a different disposition option (e.g., reuse, disposal).
Sometimes activity in excess of background can be removed from the M&E, or remediated,
followed by re-evaluation or re-survey of the M&E. This approach may include evaluation of
alternatives for remediation and a remedial action support survey prior to performing another
final disposition survey.
January 2009 6-27 NUREG-1575, Supp. 1
-------
Evaluate the Survey Results MARSAME
If the radionuclides of concern have short half-lives, storage of the M&E until the radionuclides
have decayed to acceptable levels, or "decay in place," may be an option. The planning team
should consider the intrinsic value of the M&E along with storage and disposal costs when
considering this option. When multiple radionuclides are present with significantly different half-
lives (e.g., order of magnitude) radionuclide-specific measurements may be required to fully
evaluate the acceptability of this option.
In other cases, a different disposition option (e.g., reuse, disposal) may be selected. If such a
situation were encountered in evaluating Class 2 or Class 3 M&E, the classification would be
questioned and the M&E would be reclassified and surveyed as Class 1 M&E. This may also
bring other classification decisions into question.
As a general rule, it may be useful to anticipate possible modes of failure. These can be
formulated as the problem to be solved using the DQO Process. Once the problem has been
stated, the decision concerning the failing survey unit can be developed into a decision rule. For
example, decide whether to attempt to remove the radioactivity or simply segregate certain types
of M&E for low-level waste disposal. Next, determine the additional data, if any, needed to
document that a survey unit where pieces with elevated measurements have been removed or
areas of added activity removed demonstrates compliance with the disposition criterion.
Alternatives to resolving the decision rule should be developed for each type of M&E that may
fail the surveys. These alternatives can be evaluated against the DQOs, and a disposition survey
design that meets the objectives of the project can be selected.
6.10 Document the Disposition Survey Results
Documentation of survey results is an important part of the disposition survey process. The form
of this documentation can vary greatly depending on the survey objectives and regulatory or
administrative requirements. Documentation of disposition survey results should be considered
during survey design to ensure adequate records are provided during implementation. Generally,
survey documentation requirements are provided as part of the documented survey design.
Documented items may include—
• A description of the final disposition, such as disposal in a landfill, return to manufacture for
refurbishment, sold as salvage, recycled as ferrous metal, etc.;
• A release statement to the transport carrier and recipient of the material indicating that the
M&E described in the bill of laden meet(s) applicable state and federal regulations; and
• Results of QC measurements made during the conduct of release surveys and confirmation of
compliance with facility SOPs and action levels.
In both routine and non-routine surveys, the documentation should comply with all applicable
regulatory requirements. Development of survey documentation should allow for any necessary
or required reviews.
If the disposition survey is a routine survey, then the survey will be documented as specified in
the SOP. For example, routine surveys performed to clear M&E from a facility may require
documentation that the instruments were calibrated and functioning properly and that trained
NUREG-1575, Supp. 1 6-28 January 2009
-------
MARSAME Evaluate the Survey Results
personnel were on duty to perform the surveys. Quality assurance reviews and audits would be
performed periodically (typically under a separate SOP) to document that the clearance surveys
were being performed properly and that no M&E were cleared without first being surveyed.
These records would document that properly trained personnel had adequately surveyed all M&E
leaving the facility using properly functioning instruments. Documentation of individual
measurement results may not be required or necessary.
If the survey is not routine, significantly more documentation may be required. This
documentation should provide a complete and unambiguous record of the radiological status of
the M&E relative to the selected action levels. In addition, sufficient data and information should
be provided to enable an independent evaluation of the survey results, including repeating
measurements at some future time Additional information on documentation is provided in
Section 2.5, Section 3.6, Section 4.5, MARSSIM Sections 3.8 and 8.6, and MARSSIM
Chapter 5.
January 2009 6-29 NUREG-1575, Supp. 1
-------
-------
MARSAME Statistical Basis For MARSAME Surveys
7 STATISTICAL BASIS FOR MARSAME SURVEYS
The statistically rigorous quantitative application of measurement quality objectives (MQOs)
plays a central role in the MARSAME process. MQOs did not appear explicitly in Multi-Agency
Radiation Survey and Site Investigation Manual (MARSSIM 2002), but were subsequently
developed for radioanalytical chemistry measurements as part of the Multi-Agency Radiological
Laboratory Analytical Protocols (MARLAP) manual. However, these concepts apply equally
well to field measurements of radiation and radioactivity. The MARSAME process incorporates
these ideas and extends them to these measurements.
A major development since the publication of MARS SIM was the publication of the Guide to the
Expression of Uncertainty in Measurement, or "GUM" (ISO 1995). The procedures described in
this document have become a de facto standard for estimating the uncertainty associated with
measurements of any type. The GUM methodology is essential for the assessment of
measurement uncertainty, but was not previously treated in MARS SIM.
Data quality objectives (DQO) form the backbone of the MARSAME process, and are discussed
in detail in Chapters 2 and 3. A number of terms with specific statistical meanings are used in
this and subsequent sections. The concept of measurement quality objectives (MQOs) and in
particular the required measurement method uncertainty was introduced in Section 3.8. These
ideas are discussed in greater detail in MARLAP Chapter 3 and Appendix C. While MARLAP is
focused on radioanalytical procedures, these concepts are applicable on a much broader scale and
are used in MARSAME in Sections 5.5 through 5.8 to guide the selection of measurement
methods for disposition surveys for materials and equipment.
In Section 7.1 the general concepts of statistical survey design and hypothesis testing are
discussed, with more detail in Section 7.2. In Sections 7.3, 7.4, 7.5, and 7.6, calculation of
measurement quality objectives (particularly the required method uncertainty), measurement
uncertainty, minimum detectable concentrations (MDCs) and minimum quantifiable
concentrations (MQCs), respectively, are introduced. Further details and examples of these
topics for the interested reader are then given in Sections 7.7, 7.8, 7.9, and 7.10. This advanced
material is optional on initial reading, and may be referred to later as needed. Section 7.11 shows
a detailed calculation of a scan MDC, which is used in Chapter 8. This process was described
and used in MARS SIM, but a systematic example was constructed for M&E. These calculations
are detailed, and are also optional on first reading.
In developing the results in this chapter, a number of new and sometimes only subtly different
definitions and symbols are used. For the convenience of the reader, many of these are
summarized in the tables below. Table 7.1 provides a summary of notation used for DQOs and
MQOs, used primarily in Sections 7.1 and 7.2. Table 7.2 contains notation used for setting
MQOs for required method uncertainties (Sections 7.3 and 7.7) and in uncertainty calculations
(Sections 7.4 and 7.8). MDC calculations (Sections 7.5 and 7.9) and MQC calculations (Sections
7.6 and 7.10) use the notation added in Table 7.3 and Table 7.4, respectively. Symbols may not
have an entry for both formula or reference and type.
January 2009 7-1 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
Table 7.1 Notation for DQOs and MQOs
Symbol
a
P
A
VMR
Sc
s
a
aN
as
0M
<*MR
MMR
«C2W
«cG>)
ZI-OL
(zi-p)
Definition
Probability of a Type I
decision error
The probability of a
Type II decision error
Width of the gray
region
Required relative
method uncertainty
above the UBGR
The critical value of
the net instrument
signal (e.g., net count)
net signal
The total standard
deviation of the data
The standard deviation
of the mean of N
independent
measurements
Standard deviation due
to sampling
Standard deviation of
the measurement
method
Required method
standard deviation at
and below the UBGR
Required method
uncertainty at and
below the UBGR
Combined variance of
y
Combined standard
uncertainty of y
1-«(or 1-/7) quantile
of a standard normal
distribution function
Formula or reference
(UBGR-LBGR)
HMK/UBGR
Calculation ofSc requires the choice of a
significance level for the test. The
significance level is a specified upper bound
for the probability, a, of a Type I error. The
significance level is usually chosen to be
0.05.
(0S + 0M2)'/2
aN = o/VjV
Upper bound to the value of ou
Upper bound to the value of UM
Uncertainty propagation
Uncertainty propagation
Table of standard normal distribution
Type
Chosen during DQO
process
Chosen during DQO
process
Chosen during DQO
process
Chosen during DQO
process
If a measured value
exceeds the critical
value, a decision is made
that radiation or
radioactivity has been
detected
Experimental
Theoretical population
parameter
Theoretical population
parameter
Theoretical population
parameter
Theoretical population
parameter
Chosen during DQO
process
Calculated
Calculated
Theoretical
NUREG-1575, Supp. 1
7-2
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
Table 7.2 Notation for Uncertainty Calculations
Symbol
a
C^
f(xh x2,...,xN)
Ml,
X2,...,XN)
k
P
r(Xi,x})
s(x,)
Ufri)
ut(y)
ttc(»
«C2W
U
u(x1^j)
Definition
Half-width of a bounded
probability distribution
Sensitivity coefficient
The calculated value of the
output quantity from
measurable input quantities
for a particular measurement
Model equation expressing
the mathematical
relationship, between the
measurand, Y and the input
quantities^
Coverage factor for
expanded uncertainty
Coverage probability for
expanded uncertainty
Correlation coefficient for
two input estimates, xt and x}-
Sample standard deviation
of the input estimate xt
Type B standard uncertainty
of the input estimate xt
Component of the combined
standard uncertainty uc(y)
generated by the standard
uncertainty of the input
estimate xit u(x,)
Combined standard
uncertainty of y
Combined variance of y
Expanded uncertainty
Covariance of two input
estimates, xt and x}-
Formula or reference
Type B evaluation of
uncertainty
dfl dxf, the partial derivative of
/with respect to xt
y =f(xlt x2,...,xN)
Y=f(X1,X2,...,XN)
Numerical factor used as a
multiplier of the combined
standard uncertainty in order to
obtain an expanded uncertainty
Probability that the interval
surrounding the result of a
measurement determined by
the expanded uncertainty will
contain the value of the
measurand
u(x^j) 1 (u(xt) u(Xj))
V(»-l)w '
tt;0) = Ci U(X,)
Uncertainty propagation
Uncertainty propagation
"Defining an interval about the
result of a measurement that
may be expected to encompass
a large fraction of values that
could reasonably be attributed
to the measurand" (GUM)
Type
Estimated
Evaluated at the measured
values Xi,x2,...,xN
Experimental
Theoretical
Chosen during DQO
process
Chosen during DQO
process
Experimental
Experimental
Estimated
Estimated
Calculated
Calculated
Calculated
Experimental
January 2009
7-3
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
Table 7.2 Notation for Uncertainty Calculations (Continued)
Symbol
ttcO)/y
U (Xj)/Xj
W i,W2, ...,Wjv
Xl, X2, ••• >X]\[
Y Y Y
•^1 ; ^2 ' ' ' ' ' TV
7
j
Z;,Z2, ...,Zjv
Definition
Relative combined standard
uncertainty of the output
quantity for a particular
measurement
Relative standard
uncertainty of a nonzero
input estimate xt for a
particular measurement
Input quantities appearing in
the numerator of y =f(xlt
X2,...,XN)
Measurable input quantities
Estimates of the measurable
input quantities for a
particular measurement
The output quantity or
measurand
Estimate of the output
quantity for a particular
measurement
Input quantities appearing in
the denominator of y =f(xlt
X2,...,XN)
Formula or reference
See "z;,z2, ... ,ZAT" below
N = n+m
Type
Experimental
Experimental
Theoretical
Experimental
Theoretical
Experimental
Experimental
Table 7.3 Notation for MDC Calculations
Symbol
NB
%
^5
tB
RB
d
£
F
Kl
Definition
Background count
Gross sample count
Count time for the test source or
sample
Count time for the background
Mean count rate of the blank
Parameter in the Stapleton
equation for the critical value of
the net instrument signal
Efficiency
Calibration function
Evaluation function
Formula or reference
R -N"
'- t,
Usually has the value 0.4
Calibration
X=F(Y)
Y =F\X), closely related to the
mathematical model
Y=f(X1,X2,...,XN)
Type
Experimental
Experimental
Experimental
Experimental
Experimental or
Theoretical
NUREG-1575, Supp. 1
7-4
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
Table 7.3 Notation for MDC Calculations (Continued)
Symbol
Sc
SD
X
xc
Y
yc
yo
^B
*A
Definition
Critical value of the net
instrument signal
Minimum detectable value of
the net instrument signal
Observable response variable,
measurable signal
The critical value of the
response variable
State variable, measurand
Critical value of the
concentration
Minimum detectable
concentration (MDC)
Relative systematic error in the
background determination
Relative systematic error in the
sensitivity
Formula or reference
Net instrument signal is
calculated from the gross signal
by subtracting the estimated
background and any
interferences
Net instrument signal that gives
a specified probability, l-(3, of
yielding an observed signal
greater than its critical value Sc
Calculation of yc requires the
choice of a significance level for
the test. The significance level is
a specified upper bound for the
probability, a, of a Type I error.
The significance level is usually
chosen to be 0.05.
Uncertainty propagation
yc = rl(xc)
yD-^
£
Type
Experimental
If a measured value
exceeds the critical value,
a decision is made that
radiation or radioactivity
has been detected
Experimental
Experimental
Table 7.4 Notation for MQC Calculations
Symbol
ke
°\y Y = ye)
yQ
RI
Definition
Multiple of the standard
deviation defining JQ
usually chosen to be 10
The variance of y given the
true concentration Y equals
yQ
Minimum quantifiable
concentration (MQC)
Mean interference count
rate
Formula or reference
, ^\y\Y = ye)
Q V
yQ
The concentration at which the
measurement process gives results with a
specified relative standard deviation \lkQ,
where kQ is usually chosen to be 10
Type
Chosen during
DQO process
Theoretical
Theoretical
Experimental
January 2009
7-5
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
Table 7.4 Notation for MQC Calculations (Continued)
Symbol
a(^)
fi
Definition
Standard deviation of the measured interference count
rate
Relative variance of the measured efficiency, s
Formula or
reference
Type
Experimental
Experimental
7.1 Overview of Statistical Survey Design and Hypothesis Testing
Designing a MARSAME survey involves the following key statistical parameters:
(1) The uncertainty in the measurement method. The measurement method uncertainty can be
affected by changes to the measurement method, such as changing counting times, or
performing repeated measurements. Generally, the measurement method uncertainty is
characterized by its standard deviation, OM. This value may be a constant, meaning that all
measurements will have the same standard deviation. Alternatively, this value may vary with
the level of radionuclide concentration or radioactivity, such that the standard deviation
increases with increasing radionuclide concentration or radioactivity.
(2) The uncertainty in the distribution of radionuclide concentrations or radioactivity in the
population of materials and equipment (M&E) to be measured. This variation of radionuclide
concentrations or radioactivity in space and time can be characterized by the sampling
standard deviation, as.
(3) The number of samples, N, from the population of radionuclide concentrations or
radioactivity that comprises the survey unit.
(4) The null (Ho) and alternative (Hi) hypotheses to be examined. The symbol A represents the
detectable difference between the null hypothesis concentration value (the action level, or
AL), and the alternative hypothesis concentration value (the discrimination limit, or DL). The
range of concentrations between the AL and the DL is referred to as the gray region.
(5) The values of a and /?that quantify acceptable limits for Type I and Type II decision errors,
respectively. A Type I decision error occurs when the null hypothesis is rejected when it is
actually true. A Type II decision error occurs when the null hypothesis is not rejected but
should have been rejected. The value of I-/? is termed the power, or the ability of the
statistical test to reject the null hypothesis, when appropriate. For a specific survey design,
the power (1-/7) of the survey can be compared at different values of a, since the power is
the probability of rejecting the null hypothesis at a given value of a.
Note: Designing a survey involves collecting a number of measurements, N, that will yield the
desired a and power (I-/?), given a detectable difference A, the o^for the measurement method
selected and the as for the distribution of radionuclide concentrations or radioactivity in the
population of materials and equipment (M&E) to be measured. The relationships between these
parameters are complex and interrelated. The choice or determination of one parameter affects
the choice or determination of the other parameters.
NUREG-1575, Supp. 1
7-6
January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
When a single measurement is taken, the variance of that measurement will equal:
-------
Statistical Basis For MARSAME Surveys MARS AME
measurements, N, is found to fulfill the desired limits on decision error rates a and /?. If any of
these are changed, it will affect the others.
In MARSAME, the null and alternative hypotheses concern the true difference in the M&E
between containing radionuclide concentrations or radioactivity in excess of the AL above the
appropriate background reference M&E. l Scenario A uses a null hypothesis that assumes the
radionuclide concentration or radioactivity associated with the M&E exceeds the AL. Scenario A
is sometimes referred to as "presumed not to comply" or "presumed not clean." Scenario B uses
a null hypothesis that assumes the radionuclide concentration or radioactivity associated with the
M&E is less than or equal to the AL. Scenario B is sometimes referred to as "indistinguishable
from background" (when the AL is zero) or "presumed clean."
Note: Under Scenario A, the M&E are only deemed suitable for release if the null hypothesis is
rejected, whereas under Scenario B, the M&E are suitable for release only if the null hypothesis
is not rejected.
For example, under Scenario A, if the true, but unknown, value of the radionuclide concentration
or radioactivity in excess of background is less than or equal to the DL, then the hypothesis test
upon which the survey is designed will have power I-/? to reject the null hypothesis that the true,
but unknown, value is greater than or equal to the AL at Type I error rate a. Under Scenario B, if
the true, but unknown, value of radionuclide concentration or radioactivity in excess of
background is greater than the DL (AL + A), then the hypothesis test upon which the survey is
designed will once again have power I-/? to reject this null hypothesis at Type I error rate a.
For a given a and I-/?, A depends on a, so it is important that the measurement method (and
sampling fraction, where appropriate) selected is sensitive enough to provide a small enough cr,
in order that A meets survey design requirements for the DL. This ensures that the DL is not set
too low in Scenario A or too high in Scenario B. For normally distributed measurements.
A/o- =(zi-/j + zi-«) (7-5)
Segregation according to likely radionuclide concentrations or radioactivity or a measurement
method with a longer counting time may improve crand therefore A. Hypothesis testing (i.e.,
accepting or rejecting the null hypothesis) consists of comparing an estimate of the radionuclide
concentration or radioactivity to a "critical value," Sc. The result indicates whether the observed
estimate is consistent with the null value for a given Type I error rate a, after taking account of
the uncertainty a of the measurement. For Scenario A, the critical value is
Sc = AL-zi-a
-------
MARSAME Statistical Basis For MARSAME Surveys
Where zi_ais the 1-orquantileof the standard normal distribution. In situations where the
distribution of the estimate may not be normally distributed, more specialized statistical analysis
may be needed. By definition, the power l-/?is the probability as computed under the alternate
hypothesis of rejecting the null hypothesis, or that the probability that the observed estimate is
less than the critical value Sc for Scenario A, and greater than Sc for Scenario B.
7.2 Statistical Decision-Making
In Section 4.2, MARSAME recommends the planning team complete the following steps:
• Select a null hypothesis,
• Choose a discrimination limit,
• Define Type I and Type II decision errors,
• Set a tolerable Type I decision error rate at the action level, and
• Set a tolerable Type II decision error rate at the discrimination limit.
7.2.1 Null Hypothesis
In hypothesis testing, two assertions about the actual level of radioactivity associated with the
M&E are formulated. The two assertions are called the null hypothesis (Ho) and the alternative
hypothesis (Hi). HO and HI together describe all possible radionuclide concentrations or levels of
radioactivity under consideration. The survey data are evaluated to choose which hypothesis to
reject or not reject, and by implication which to accept.2 In any given situation, one and only one
of the hypotheses must be true. The null hypothesis is assumed to be true within the established
tolerance for making decision errors (Section 7.2.5). Thus, the choice of the null hypothesis also
determines the burden of proof for the test.
If the action level (AL) is not zero, the planning team generally assumes the radionuclide
concentration or level of radioactivity (X) exceeds the action level unless the survey results
provide evidence to the contrary. In other words, surveys are designed to provide sufficient
evidence to disprove H0. In this case, the null hypothesis is that the radionuclide concentration or
level of radioactivity is greater than or equal to the action level (i.e., HQ: X> AL). The alternative
hypothesis is the radionuclide concentration or level of radioactivity is less than the action level
(i.e., Hi: X< AL). MARSSIM and NUREG-1505 (NRC 1998a) describe this as Scenario A, and
the burden of proof falls on the owner of the M&E. Scenario A is sometimes referred to as
"presumed not to comply" or "presumed not clean."
On the other hand, the planning team may choose to assume the action level has not been
exceeded unless the survey results provide evidence to the contrary. The null hypothesis
becomes H0: X< AL, and the alternative hypothesis is HI: X> AL. MARSSIM and NUREG-
1505 (NRC 1998a) describe this as Scenario B, and the burden of proof falls on the regulator.
Scenario B is sometimes referred to as "indistinguishable from background" or "presumed
2 In hypothesis testing, to "accept" the null hypothesis only means not to reject it. For this reason many statisticians
avoid the word "accept." A decision not to reject the null hypothesis does not imply the null hypothesis has been
shown to be true.
January 2009 7-9 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
clean." This is the only practical approach when the action level is equal to zero (above
background); because it is technically impossible to obtain statistical evidence that the
radionuclide concentration or level of radioactivity is exactly zero. However, Scenario B can be
applied to situations other than "indistinguishable from background." The example in Section 8.4
uses Scenario B to support an interdiction decision.
7.2.2 Discrimination Limit
Action levels were defined in Section 3.3 based on the selected disposition option and applicable
regulatory requirements. The planning team also chooses another radionuclide concentration or
level of radioactivity that can be reliably distinguished from the action level by performing
measurements (i.e., direct measurements, scans, in situ measurements, samples and laboratory
analyses). This radionuclide concentration or level of radioactivity is called the discrimination
limit (DL). An example where the discrimination limit is defined is provided in Section
8.4.5.The gray region is defined as the interval between the action level and the discrimination
limit (Figures 7.1, 7.2, 7.3, and 7.4 provide visual descriptions of the gray region). The width of
the gray region is called the shift and denoted as A. The objective of the disposition survey is to
decide whether the concentration of radioactivity is more characteristic of the DL or of the AL,
i.e., whether action should be taken, or if action is not necessary. Figures 7.1 and 7.2 show
examples that would fall under Scenario A (discussed in Section 7.2.3). In Figure 7.1 (top) the
difference in concentration between the AL and the DL (i.e., A) is large; but the variability in the
measured concentration (i.e., a) is also large. In Figure 7.2 (bottom) the difference in
concentration between the AL and the DL (i.e., A) is relatively small. However, the variability in
the measured concentration (i.e., a) is also smaller. Figures 7.1 and 7.2 illustrate that
determining the level of survey effort depends not just on the width of the gray region, but also
in the ratio of that width to the expected variability of the data. This ratio, A/cr, is called the
relative shift in MARSSEVI. In situations where A/cris small, i.e., less than 1, it may be
impracticable to achieve the required accuracy of measurements or the number of samples to
meet the Type I error rate in the DQOs. Section 4.4.4 presents options for relaxing project
constraints to optimize the survey design in such cases. In Figure 7.1, A/cris greater than 4; while
in Figure 7.2, A/cris approximately 1.
As discussed in MARSSIM, generally, the larger A/cr, the easier the survey effort. When A/cris
greater than three, the survey effort will be minimal, and any effort to increase it by either
widening the gray region or reducing the measurement variability usually would not be
worthwhile.
NUREG-1575, Supp. 1 7-10 January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
o>
u
u
u
O
•5
>
u
c
0)
Action
Level
Concentration
Figure 7.1 Relative Shift, A/o, Comparison for Scenario A:
o is Large, but the Large A Results in a Large A/o and Fewer Samples
o>
u
c
£
3
U
U
O
ii-
o
>
u
c
0)
Mean
Action
Level
Concentration
Figure 7.2 Relative Shift, A/o, Comparison for Scenario A:
o is Small, but the Small A Results in a Small A/o and More Samples
On the other hand, when A/a is less than one, the survey effort will become substantial, and any
effort to increase it by either widening the gray region or reducing the measurement variability
will be worthwhile. The measurement variability is thus just as important as the width of the gray
region when designing disposition surveys. In MARSSEVI surveys, the total variability had two
components: sampling and analytical. For some MARSAME surveys this will also be the case.
However, in many MARSAME surveys the sampling variability will be of less importance,
either because 100% of the survey unit is being measured, or because disposition decisions are
being made on the basis of single measurements on single items or single locations. In such
January 2009
7-11
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
cases, the required measurement method uncertainty discussed in Section 3.8.1 will be of
paramount importance in the survey planning. The details for determining the required
measurement method uncertainty and how to determine if it is being met are discussed in detail
in Section 7.7.
Depending on the survey design, the combination of action levels, expected radionuclide
concentrations or levels of radioactivity, instrument sensitivity, and local radiation background
contribute to defining the width of the gray region. Reducing the radionuclide concentrations or
levels of radioactivity known or assumed to be associated with the M&E can affect the selection
of a discrimination limit, so remediation costs may need to be considered. Increasing the
sensitivity of a measurement method to reduce the measurement method uncertainty generally
involves increased instrument costs or increased counting times.
The lower bound of the gray region is denoted by LBGR and the upper bound of the gray region
is denoted by UBGR. The association of either the UBGR or the LBGR with the DL or AL will
depend on the scenario selected (see Sections 7.2.3 and 7.2.4). The width of the gray region
(UBGR - LBGR) is denoted by "A" and is called the "shift" or the "required minimum
detectable difference" in activity or concentration (MARSSIM Section 5.5.2 and Section D.6,
MARLAP Section C.2, NRC 1998a, and EPA 2006a,).
7.2.3 Scenario A
The null hypothesis for Scenario A specifies that the radionuclide concentration or level of
radioactivity associated with the M&E is equal to or exceeds the action level. For Scenario A
(H0: X> AL), the UBGR is equal to the AL and the LBGR is equal to the DL. As a general rule
for applying Scenario A, the DL should be set no higher than the expected radionuclide
concentration associated with the M&E. The DL and the AL should be reported in the same
units. Figure 7.3 illustrates Scenario A. Note that the Type I (a) and Type II (ft) error rates need
not be equal. This is discussed further in Section 7.2.5, and an example can be seen in Section
7.5.2.
Discrimination Critical
Limit Value
Scenario A
(H0: Net Activity > Action Level)
Figure 7.3 Illustration of Scenario A
NUREG-1575, Supp. 1
7-12
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
7.2.4 Scenario B
The null hypothesis for Scenario B specifies the radionuclide concentration or level of
radioactivity associated with the M&E is less than or equal to the action level. For Scenario B
(H0: X< AL), the UBGR is equal to the DL and the LBGR is equal to the AL. For example, if
the AL=0 (sometimes called indistinguishable from background), then the LBGR will be zero.
The DL defines how hard the surveyor needs to look, and is determined through negotiations
with the regulator.3 In some cases, the DL will be set equal to a regulatory limit (e.g., 10 CFR
36.57 and DOE 1993). The DL and the AL should be reported in the same units. Figure 7.4
illustrates Scenario B. As above, note that the Type I (a) and Type II (fl) error rates need not be
equal. This is discussed further in Section 7.2.5, and an example can be seen in Section 7.5.2.
Critical Discrimination
Value Limit
Scenario B
(H0: Net Activity < Action Level)
Figure 7.4 Illustration of Scenario B
This description of Scenario B is based on information in MARLAP and is fundamentally
different from the description of Scenario B in NUREG-1505 (NRC 1998a).
In NUREG-1505 (NRC 1998a) the gray region is defined as being below the AL in both
Scenario A and Scenario B. In MARSAME and MARLAP the gray region is defined as being
above the AL in Scenario B. The difference lies in how the action level is defined.
7.2.5 Specify Limits on Decision Errors
There are two possible types of decision errors:
• Type I error: rejecting the null hypothesis when it is true.
• Type II error: failing to reject the null hypothesis when it is false.
1 In some cases setting the discrimination limit may include negotiations with stakeholders.
January 2009
7-13
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Because there is always uncertainty associated with the survey results, the possibility of decision
errors cannot be eliminated. So instead, the planning team specifies the maximum Type I
decision error rate (a) that is allowable when the radionuclide concentration or level of
radioactivity is at or above the action level. This maximum usually occurs when the true
radionuclide concentration or level of radioactivity is exactly equal to the action level. The
planning team also specifies the maximum Type II decision error rate (/?) that is allowable when
the radionuclide concentration or level of radioactivity equals the discrimination limit.
Equivalently, the planning team can set the "power" (I-/?) when the radionuclide concentration
or level of radioactivity equals the discrimination limit. See MARSSIM Appendix D, Section
D.6, for a more detailed description of error rates and statistical power.
The definition of decision errors depends on the selection of the null hypothesis. For Scenario A
the null hypothesis is the radionuclide concentration or level of radioactivity exceeds the action
level. A Type I error for Scenario A occurs when the decision maker decides the radionuclide
concentration or level of radioactivity is below the action level when it is actually above the
action level (i.e., mistakenly decides the M&E are clean when they are actually not clean). A
Type II error for Scenario A occurs when the decision maker decides the radionuclide
concentration or level of radioactivity is above the action level when it is actually below the
action level (i.e., mistakenly decides the M&E are not clean when they are actually clean).
For Scenario B, the null hypothesis is that the radionuclide concentration or level of radioactivity
is less than or equal to the action level. A Type I error for Scenario B occurs when the decision
maker decides the radionuclide concentration or level of radioactivity is above the action level
when it is actually below the action level (i.e., mistakenly decides the M&E are not clean when
they are actually clean). A Type II error for Scenario B occurs when the decision maker decides
the radionuclide concentration or level of radioactivity is below the action level when it is
actually above the action level (i.e., mistakenly decides the M&E are clean when they are
actually not clean). It is important to clearly define the scenario (i.e., A or B) and the decision
errors for the survey being designed.
Once the decision errors have been defined, the planning team should determine the
consequences of making each type of decision error. This process should be revisited as more
information is obtained. For example, incorrectly deciding the activity is less than the action
level may result in increased health and ecological risks. Incorrectly deciding the activity is
above the action level when it is actually below may result in increased economic and social
risks. The consequences of making decision errors are specific to the actual situation at a
particular site and could vary significantly from one site to another, reflecting the major concerns
of the various stakeholders.
Once the consequences of making both types of decision errors have been identified, acceptable
decision error rates can be assigned for both Type I and Type II decision errors. Historically, a
decision error rate of 0.05, or 5%, often has been acceptable for decision errors. However,
assigning the same tolerable decision error rate to all projects does not account for the
differences in consequences of making decision errors. This becomes evident with M&E where
there are wide ranges of disposition options generating a wide range of consequences. For
example, a Type I decision error for Scenario A could have different consequences for a
NUREG-1575, Supp. 1 7-14 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
clearance decision compared to a low-level radioactive waste disposal decision. Not all
consequences of decision errors are the same, and it is unlikely that applying a fixed value to all
decision error rates will result in reasonable survey designs resulting in comparable decisions.
Project-specific decision error rates should be selected based on the project-specific
consequences of making decision errors.
7.2.6 Develop an Operational Decision Rule
The theoretical decision rule developed in Section 3.7 was based on the assumption that the true
radioactivity concentrations or radiation levels associated with the M&E were known. Since the
disposition decision will be made based on measurement results and not the true but unknown
concentration level, an operational decision rule needs to be developed to replace the theoretical
decision rule. The operational decision rule is a statement of the statistical hypothesis test, which
is based on comparing some function of the measurement results to some critical value. The
theoretical decision rule is developed during Step 5 of the DQO Process (Chapter 3), while the
operational decision rule is developed as part of Step 6 and Step 7 of the DQO Process. For
example, a theoretical decision rule might be "if the results of any measurement identify surface
radioactivity in excess of background, the front loader will be refused access to the site; if no
surface radioactivity in excess of background is detected, the front loader will be granted access
to the site." The related operational decision rule might be "any result that exceeds the critical
value associated with the MDC, set at the discrimination limit, will result in rejection of the null
hypothesis and the front loader will not be allowed on the site" (see more examples in Chapter
8).
Chapter 6 provides guidance on using statistical tests to evaluate data collected during the
disposition survey to support a disposition decision. The planning team should evaluate the
statistical tests and possible operational decision rules and select one that best matches the intent
of the theoretical decision rule with the statistical assumptions. Each operational decision rule
will have a different formula for determining the number of measurements or fraction of M&E to
be measured to meet the DQOs.
Developing an operational decision rule incorporates all relevant information available
concerning the M&E (Section 2.4.3), selected instrumentation and measurement technique
(Section 5.9), selected statistical tests (Section 6.2.3), and any constraints on collecting data
identified by the planning team. The operational decision rule will need to specify a
measurement technique (e.g., scan-only, in situ, sample collection and analysis) and a statistical
test. Examples of statistical tests include comparison to the UBGR (Section 6.3), comparison to
an upper confidence interval (Section 6.4), the Sign test (Section 6.5), the Wilcoxon Rank Sum
test (Section 6.6), and the Quantile test (Section 6.7). At this point in the survey design process it
is not necessary to select a specific instrument to perform the measurements. However, selection
of a measurement technique will assist the planning team in identifying the appropriate statistical
test. For example, if a scan-only measurement method is selected it is not appropriate to select
the Wilcoxon Rank Sum test to determine the number of measurements. However, if no scan-
only or in situ measurement methods are available that meet the measurement quality objectives
(MQOs), a MARSSEVI-type survey (which combines scan and static measurements, see Section
4.4.3) should be developed.
January 2009 7-15 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
The planning team uses the combination of the selected instrumentation and measurement
technique (see Section 5.9) with a data evaluation method (see Section 6.2.5) to establish an
operational decision rule. Then, from the operational decision rule, the planning team can
determine the number of measurements or the fraction of the M&E that needs to be measured
during the disposition survey. There is no formal structure for stating an operational decision
rule. The structure of the operational decision rule is generally defined in terms that meet the
needs of a particular project. An operational decision rule can be simple or complex. A simple
example could be "If 100% of the surfaces of hand tools are surveyed using a scan-only
technique that meets the DQOs, and none of the results exceed the action level for release, then
the tools can be released." The statistical test for this simple example is a comparison of the
mean to the action level; however, since all of the values are below the action level, the mean
value must also be below the action level. Therefore, it is not necessary to perform the actual
statistical test. This represents a conservative approach to data interpretation that may not always
be appropriate. More complex operational decision rules can-
• Account for different types of measurements and multiple radionuclides of concern,
• Specify critical values and test statistics for the statistical tests, and
• Incorporate multiple decisions (e.g., average and maximum values, fixed and removable
radioactivity) depending on the project.
7.3 Set Measurement Quality Objectives
Section 4.2 briefly discussed the DQO process for developing statistical hypothesis tests for the
implementation of disposition decision rules using measurement data. This included formulating
the null and alternative hypotheses, defining the gray region using the action level and
discrimination limit, and setting the desired limits on potential Type I and Type II decision error
probabilities that a decision maker is willing to accept for project results. Decision errors are
possible, at least in part, because measurement results have uncertainties. The effect of these
uncertainties is expressed in the size of the relative shift, A/cr, introduced in Section 7.2.2. The
overall uncertainty, a, has components that may be due to sampling variability in radioactivity
concentration, as, but also because of uncertainty in the measurement method, ou. Because
DQOs apply to both sampling and measurement activities, what are needed from a measurement
perspective are method performance characteristics specifically for the measurement process of a
particular project. These method performance characteristics (see Section 3.8) are the
measurement quality objectives (MQOs).
DQOs define the performance criteria that limit the probabilities of making decision errors by-
• Considering the purpose of collecting the data,
• Defining the appropriate type of data needed, and
• Specifying tolerable probabilities of making decision errors.
DQOs apply to both sampling and measurement activities.
MQOs can be viewed as the measurement portion of the overall project DQOs (see Section 3.8).
MQOs are:
NUREG-1575, Supp. 1 7-16 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
• The part of the project DQOs that apply to the measured result and its associated uncertainty.
• Statements of measurement performance objectives or requirements for a particular
measurement method performance characteristic, for example, measurement method
uncertainty and detection capability.
• Used initially for the selection and evaluation of measurement methods.
• Subsequently used for the ongoing and final evaluation of the measurement data.
A number of MQOs were introduced in Section 3.8, but for survey planning the single most
important MQO is the required measurement method uncertainty, UMR. Other MQOs, such as
range, ruggedness, and specificity, if not controlled, will lead to increased measurement
uncertainty. In this sense, the required measurement method uncertainty encompasses many of
the effects of other MQO parameters that could impact decision making. MDCs and MQCs are
closely related to the measurement uncertainty, have a long history of use for comparing the
appropriateness of competing measurement techniques, and can contribute much to survey
planning. These concepts are developed further in the later sections of this chapter (Sections 7.5
and 7.6). However, essentially the same information can be conveyed by specifying the required
measurement method uncertainty, which is a more general concept. Thus, in this section and the
next, it is this MQO that will be emphasized.
Measurement method uncertainty refers to the predicted uncertainty of a measured value that
would be calculated if the method were applied to a hypothetical sample with a specified
radioactivity concentration or radiation level. Measurement method uncertainty is a characteristic
of the measurement method and the measurement process. Measurement uncertainty, as opposed
to sampling uncertainty, is a characteristic of an individual measurement.
The true measurement method standard deviation, crM, is a theoretical quantity and is never
known exactly, but it may be estimated using the methods described in Section 7.4. The estimate
of
-------
Statistical Basis For MARSAME Surveys MARSAME
When making decisions about individual measurement results UMR should ideally be 0.3A, and
when making decisions about the mean of several measurement results UMR should ideally be
0. 1 A, where A is the width of the gray region, A = UBGR - LBGR.
7.3.1 Determine the Required Measurement Method Uncertainty at the UBGR
This section provides the rationale and guidance for establishing project-specific MQOs for
controlling OM. This control is achieved by establishing a desired maximum measurement method
uncertainty, u^., at the upper boundary of the gray region. This control also will assist in both the
measurement method selection process and in the evaluation of measurement data. Approaches
applicable to several situations are detailed below.
Four basic survey designs were described in Chapter 4: scan-only, in situ, MARSSEVI-type, and
method-based. The relative shift, A/cr, is important in determining the level of survey effort
required in the first three survey designs. For a given width of the gray region, A, the relative
shift, A/cr, can only be controlled by controlling a. The overall standard deviation of the
measurement results, cr, may have both a measurement component, CTM, and a sampling
component, as. Segregation and classification may help in controlling as (Sections 4.3 and 5.4).
7.3.1.1 Scan-Only Survey Designs
For 100% scan-only surveys, the decision uncertainty associated with crsis essentially eliminated
because the entire survey unit is measured. In class 2 survey units, the scan coverage can vary
from 10% to nearly 100% depending on the value of A/cr. This is a reflection of the fact that for a
fixed measurement variability, CTM, smaller values of A/cr imply larger sampling variability.
Larger sampling variability demands higher scan coverage to reduce the decision uncertainty.
That is, more of the survey unit must be measured to lower the standard deviation of the mean. In
such cases, it will be desirable to reduce aM until it is negligible in comparison to as. OM can be
considered negligible if it is no greater than as /3. Therefore, MARSAME recommends the
requirement UMR
-------
MARSAME Statistical Basis For MARSAME Surveys
If a = ,9 = 0.05, then
uMr< A = - * - = ^~0.3A (7-9)
z095+z095 1.645 + 1.645 3.29
Therefore, MARSAME recommends the requirement UMR< 0.3 A. The details are discussed in
Section 7.7.2.
For the special case where the LBGR = 0, then A = UBGR and OMR = A / (zi_a + z\-p) implies
(7-10)
z095 + z095 1.645 + 1.645 3.29
This is equivalent to requiring that the MDC (see Section 7.9.2) be less than the action level. The
MDC is defined as the concentration at which the probability of detection is 1 -ft and the
probability of false detection in a sample with zero concentration is at most a.
Example 1: Suppose the action level is 10,000 Bq/m2 and the lower bound of the gray region is
5,000 Bq/m2, a = 0.05, and P = 0.10. If decisions are to be made about individual items, then the
required measurement method uncertainty at 10,000 Bq/m2 is
A 10,000 Bq/m2 -5,000 Bq/m2 5,000 Bq/m2 , ^nn „ . 2
= 1, 700 Bq/m
z095+z090 1.645 + 1.282
7.3.1.3 MARSSIM-Type Survey Designs
When a decision is to be made about the mean of a sampled population, generally the average of
a set of measurements on a survey unit is compared to the disposition criterion. For MARSSIM-
type designs, the ratio A/a, called the "relative shift," determines the number of measurements
required to achieve the desired decision error rates a and/?. The target range for this ratio should
be between 1 and 3, as explained in MARSSIM (MARSSIM 2002) and NUREG-1505 (NRC
1998a). Ideally, to keep the required number of measurements low, the DQOs are aimed at
establishing A/a ~ 3. The cost in number of measurements rises rapidly as the ratio A/afalls
below 1, but there is little benefit from increasing the ratio much above 3. One of the main
objectives in optimizing survey design is to achieve a relative shift, A/cr, of at least one and
ideally three. Values of A/a greater than three, while desirable, should not be pursued at
additional cost. If A/a is 3 and aM is negligible in comparison to as, then aM will be A/10. The
details are discussed in Section 7.7.1.
Therefore, MARSAME recommends the requirement UMR < A / 10 by default when decisions are
being made about the mean of a sampled population. If the LBGR is zero, this is equivalent to
requiring that the MQC be less than the UBGR (Section 7.7.1).
January 2009 7-19 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Example 2: Suppose the action level is 10,000 Bq/m2 and the lower bound of the gray region is
2,000 Bq/m2. If decisions are to be made about survey units based on measurements at several
locations, then the required measurement method uncertainty (MM?) at 10,000 Bq/m2 is
A 10,000-2,000 2
— = = 800 Bq/m
Example 3: Suppose the action level is 10,000 Bq/m2, but this time assume the lower bound of
the gray region is 0 Bq/m2. In this case the required method measurement uncertainty, UMR, at
10,000 Bq/m2 is
A 10,000-0 innnn , 2
— = —' = I,000 Balm
f
The recommended values of UUR are based on the assumption that any known bias in the
measurement process has been corrected and that any remaining bias is well less than 10% of the
shift, A, when a concentration near the gray region is measured.
Achieving a required measurement method uncertainty UUR less than the recommended limits
may be difficult in some situations. When the recommended requirement for UMR is too difficult
to meet, project planners may allow UMR to be larger. In this case, project planners may choose
UMR to be as large as A/3 or any calculated value that allows the data quality objectives to be met
at an acceptable effort. Two situations that may make this possible are if os is believed to be less
than A/10 or if it is not difficult to make the additional measurements required by the larger
overall data variance (a2M +
-------
MARSAME
Statistical Basis For MARSAME Surveys
(7.11)
This is illustrated in Example 5 and Figure 7.5.
Example 5: Suppose the action level is 10,000 Bq/m and the discrimination limit is 3,000.
Scenario A is used, so the UBGR = AL = 10,000 Bq/m2 and the LBGR = DL = 3,000 Bq/m2.
Thus the width of the gray region, A= 10,000 - 3,000 = 7,000. If decisions are to be made about
individual items, a = 0.05, and/? = 0.05, then the required measurement uncertainty at 10,000
Bq/m2 is
A 10,000 Bq/m2-3,000 Bq/m2 7,000 Bq/m2
zi-a+zi-P
95 Z0
1.645 + 1.645
2,- UUU
The required measurement method uncertainty, MMR-, is 2,000 Bq/m at 10,000 Bq/m . Thus, for
any measured result less than 10,000 Bq/m2, the reported CSU, uc, should be less than or equal to
2,000 Bq/m2. For example, a reported result of 4,500 Bq/m2 with
a CSU of 1,900 Bq/m2 would meet the requirement. A reported result of 7,700 Bq/m2 with a
CSU 2,500 Bq/m2 would not meet the requirement.
The required relative measurement method uncertainty ((PMR) is 2,000 Bq/m2 / 10,000 Bq/m2 =
20% at 10,000 Bq/m2. Thus, for any measured result greater than 10,000 Bq/m2, the reported
RCSU should be less than or equal to 20%. For example, a reported result of 14,500 Bq/m2 with
a CSU of 2,900 Bq/m2 would meet the requirement because 2,900/14,500 = 20%. A reported
result of 18,000 Bq/m2 with a CSU 4,500 Bq/cm2 would not meet the requirement because
4,500/18,000 = 25%.
Below the action level, the bound
on the absolute uncertainty is
constant and equal to uMR = 2,000
Bq/m2.
Above the action level, the bound on
the relative uncertainty is constant
and equal to cpMR = uMR/AL =
2,000/10,000 = 20%.
The required method uncertainty, UMR =
is specified at the Action
5000
10000
15000
20000
25000
30000
True Concentration (Bq/m2)
Figure 7.5 Example of the Required Measurement Uncertainty at Concentrations other than the
UBGR. In this Example the UBGR Equals the Action Level
January 2009
7-21
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
7.4 Determine Measurement Uncertainty
Checking the measurement quality against the required measurement method uncertainty relies
on having realistic estimates of the measurement uncertainty. Often reported measurement
uncertainties are underestimated, particularly if they are confined to the estimated Poisson
counting uncertainty (Section 7.8). Tables of results are sometimes presented with a column
listing "+" without indicating how these numbers were obtained. Often, the "+" represents the
square root of the number of counts obtained during the measurement. The method for
evaluation calculation and reporting of measurement uncertainty, approved by both the
International Organization for Standardization (ISO) and the National Institute of Standards and
Technology (NIST) is discussed in this section. Further details of the method are given in
Section 7.8.
Measurements always involve uncertainty, which must be considered when measurement results
are used as part of a basis for making decisions. Every measured and reported result should be
accompanied by an explicit uncertainty estimate. One purpose of this section is to give users of
data an understanding of the causes of measurement uncertainty and of the meaning of
uncertainty statements; another is to describe procedures that can be used to estimate
uncertainties. Much of this material is derived from MARLAP Chapter 19.
In 1980, the Environmental Protection Agency published a report entitled Upgrading
Environmental Radiation Data, which was produced by an ad hoc committee of the Health
Physics Society (EPA 1980). Two of the recommendations of this report were that:
1. Every reported measurement result (x) should include an estimate of its overall uncertainty
(MX) that is based on as nearly a complete an assessment as possible.
2. The uncertainty assessment should include every significant source of inaccuracy in the
result.
The concept of traceability is also defined in terms of uncertainty. Traceability is defined as the
"property of the result of a measurement or the value of a standard whereby it can be related to
stated references, usually national or international standards, through an unbroken chain of
comparisons all having stated uncertainties" (ISO 1996). Thus, to realistically make the claim
that a measurement result is "traceable" to a standard, there must be a chain of comparisons
(each measurement having its own associated uncertainty) connecting the result of the
measurement to that standard.
This section considers only measurement variability,
-------
MARSAME Statistical Basis For MARSAME Surveys
resulting from the random nature of radioactive decay is only one component of measurement
method uncertainty. If only this component of uncertainty is accounted for, rather than
performing a full uncertainty analysis, the result will be misleading because it is at best only a
lower bound of the uncertainty and may lead to incorrect decisions based on overconfidence in
the measurement. Software is available to perform the mathematical operations for uncertainty
evaluation and propagation, eliminating much of the difficulty in implementing the mathematics
of uncertainty calculations. There are several examples of such software (McCroan 2006, GUM
Workbench 2006, Kragten 1994, and Vetter 2006).
7.4.1 Use Standard Terminology
The methods, terms, and symbols recommended by MARSAME for evaluating and expressing
measurement uncertainty are described in the GUM (ISO 1995). The ISO methodology is
summarized in the NIST Technical Note TN-1297 (NIST 1994).
The result of a measurement is generally used to estimate some particular quantity called the
measurand. The difference between the measured result and the actual value of the measurand is
the error of the measurement. Both the measured result and the error may vary with each
repetition of the measurement, while the value of the measurand (the true value) remains fixed.
The error of a measurement is unknowable, because one cannot know the error without knowing
the true value of the quantity being measured (the measurand). For this reason, the error is
primarily a theoretical concept. However, the uncertainty of a measurement is a concept with
practical uses. According to the GUM and NIST Technical Note 1297, the term "uncertainty of
measurement" denotes the values that could reasonably be attributed to the measurand. In
practice, there is seldom a need to refer to the error of a measurement, but an uncertainty should
be stated for every measured result.
The first step in defining a measurement process is to define the measurand clearly. The
specification of the measurand is always ambiguous to some extent, but it should be as clear as
necessary for the intended purpose of the data. For example, when measuring the activity of a
radionuclide on a surface, it is generally necessary to specify the activity, the date and time, what
area of the surface was measured, and where.
Often the measurand is not measured directly but instead an estimate is calculated from the
measured values of other input quantities, which have a known mathematical relationship to the
measurand. For example, input quantities in a measurement of radioactivity may include the
gross count, blank or background count, counting efficiency, and area measured. The
mathematical model measurement process specifies the relationship between the output quantity,
7, and measurable input quantities, X],X2, ...XN, on which its value depends: Y=f(X],X2, ...XN).
The mathematical model for a radioactivity measurement may have the simple form:
(Gross Instrument Signal) - (Blank Signal ) ._ , _.
Measurement = - —'— - (7-12)
Efficiency
Each of the quantities shown here may actually be a more complicated expression. For example,
the efficiency may be the product of factors such as surveyor efficiency, surface roughness
January 2009 7-23 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
efficiency correction, and the instrument counting efficiency. Interferences may be due to
ambient background or other radionuclides that have interactions with the detector in a manner
that contributes spuriously to the gross instrument signal.
When a measurement is performed, a specific value Xjis estimated for each input quantity, Xt,
and an estimated value, y, of the measurand is calculated using the relationship y =J(xi, X2, • • • ,XN)-
Since there is an uncertainty in each input estimate, xf, there is also an uncertainty in the output
estimate, y. Determining the uncertainty of the output estimate^ requires that the uncertainties of
all the input estimates Xibe determined and expressed in comparable forms. The uncertainty of xt
is expressed in the form of an estimated standard deviation, called the standard uncertainty and
denoted by u(xi). The ratio u(xi) I \Xt\ is called the relative standard uncertainty ofxf, where \Xt\ is
the absolute value
The partial derivatives, dfl dxt, are called sensitivity coefficients, and are usually denoted by ct.
The ct measure how much/changes when xt changes. The standard uncertainties are combined
with sensitivity coefficients to obtain the component of the uncertainty my due to xi: ci u(xi) .
The square of the CSU, denoted by u2c (y), is called the combined variance. It is obtained using
the formula for the propagation of uncertainty:4
N ( arV N
U2(y) = y\^-\u2(x.) = yc2u2(x.) (7-13)
c \s f / i ^ I \ i s Z i i \ i f \ '
i=\
dx.
The square root of the combined variance is the CSU ofy, denoted by uc(y). Further details of
this process are given in Section 7.8.1.
7.4.2 Consider Sources of Uncertainty
The following sources of uncertainty should be considered:
• The random nature of radioactive decay (e.g., counting statistics),
• Instrument calibration (e.g., counting efficiency),
• Variable instrument backgrounds,
• Variable counting efficiency (e.g., due to the instrument or to source geometry and
placement), and
• Interferences, such as crosstalk and spillover.
Other sources of uncertainty could include:
• Temperature and pressure.
• Volume and mass measurements,
• Determination of counting time and correction for dead time,
If the input estimates are potentially correlated, covariance estimates ufaxj) must also be determined. The
covariance u(xtjCj) is often recorded and presented in the form of an estimated correlation coefficient, r(x,,x,), which
is defined as the quotient u(xt,Xj) I u(x,)u(Xj). See Section 7.8.
NUREG-1575, Supp. 1 7-24 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
• Time measurements used in decay and ingrowth calculations,
• Approximation errors in simplified mathematical models, and
• Published values for half-lives and radiation emission probabilities.
•
There are a number of sources of measurement uncertainty in gamma-ray spectroscopy,
including:
• Poisson counting uncertainty,
• Compton baseline determination,
• Background peak subtraction,
• Multiplets and interference corrections,
• Peak-fitting model errors,
• Efficiency calibration model error,
• Summing,
• Density-correction factors, and
• Dead time.
Additional discussion of some major sources of uncertainty may be found in Section 7.8.2.2.
The following example may appear complex, but all but the most casual users will use software
to perform these calculations. Some possibilities are listed after the example. A complete
example is worked out to here to illustrate the underlying principles.
Example 6: Consider a simple measurement of a sample. The activity will be calculated from
(N8/ts)-(NB/tB)
£
Where:
y = sample activity (Bq)
e = counting efficiency 0.4176 (s'VBq)
Ns = gross count observed during the measurement of the source, (1 1578)
ts = source count time (300 s)
NB = observed background count (87)
IB = background count time (6,000 s)
The CSU of e is given by uc(e) = 0.005802. This is shown in Example 2 in Section 7.8.2.2.
Assume the radionuclide is long-lived; so, no decay corrections are needed. The uncertainties of
the count times are also assumed to be negligible. The standard uncertainties in Ns and NB will be
estimated as ^Ns and ^JNB using the Poisson assumption.
Then (^/^)-(^/^)
0.4179
January 2009 7-25 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
d-
(Nslts)-(NBltB)'
dNs
d-
(Nslts)-(NBltB)
u2(s)
ds J
'ts)-(NB
0.4176
= 0.73768 + 0.00001+ 1.64745 = 2.38515.
Note that these calculations show which input quantities are contributing the most to the
combined variance. Ns contributes 0.73768/2.38515 ~ 31%. NB contributes virtually nothing. The
uncertainty in the efficiency contributes 1.64745/2.38515 ~ 69%. An analysis such as this is
called an uncertainty budget, and quickly points out where improvements in the measurement
may be made.
Taking the square root of the combined variance we find uc(y) = 1.54439. Usually the CSU is
rounded to two significant figures and the result is rounded to match the same number of decimal
places. So the result would be reported as 92.3 Bq with a CSU of 1.5 Bq.
Note that if the uncertainty in the efficiency had been neglected, the CSU would have been
underestimated as 0.86 Bq, and would have been attributed entirely to the uncertainty in the
sample counts. This illustrates the importance of including all significant sources of uncertainty
in the calculations. Many of these calculations can be done using computer software programs
mentioned earlier.
A much more detailed and involved example is given in Section 7.8.3.
Again, it should be noted that software (e.g., McCroan 2006, GUM Workbench 2006, Kragten
1994, Vetter 2006) is available to perform the partial derivatives, insert the proper mean and
standard uncertainty for each input, and perform the algebra for uncertainty evaluation and
propagation. This eliminates much of the tedium in implementing the uncertainty calculations,
and frees the analyst to carefully examine the model equation to be sure that significant sources
of uncertainty are not omitted.
7.4.3 Recommendations for Uncertainty Calculation and Reporting
• Use the terminology and methods of the GUM (ISO 1995) for evaluating and reporting
measurement uncertainty.
NUREG-1575, Supp. 1 7-26 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
• Follow QC procedures that ensure the measurement process remains in a state of statistical
control, which is a prerequisite for uncertainty evaluation.
• Account for possible blunders or other spurious errors. Spurious errors indicate a loss of
statistical control of the process and are not part of the uncertainty analysis described above.
• Report each measured value with either its CSU (or its expanded uncertainty, see Section
7.8.1.7).
• Reported measurement uncertainties should be clearly explained. (In particular, when an
expanded uncertainty is reported, the coverage factor should be stated and the basis for the
coverage probability should also be given, see Section 7.8.1.7).
• Consider all possible sources of measurement uncertainty and evaluate and propagate the
uncertainties from all sources believed to be potentially significant in the final result.
• Each uncertainty should be rounded to either one or two significant figures and the measured
value should be rounded to the same number of decimal places as its uncertainty.
• Results should be reported as obtained together with their uncertainties (whether positive,
negative, or zero).
7.5 Determine Measurement Detectability
This section summarizes issues related to measurement detection capabilities. Much of this
material is derived from MARLAP Chapter 20. More detail may be found in see Section 7.9.
Environmental radioactivity measurements may involve material with very small amounts of the
radionuclide of interest. Measurement uncertainty often makes it difficult to distinguish such
small amounts from zero. Therefore, an important MQO of a measurement process is its
detection capability, which is usually expressed as the smallest concentration of radioactivity that
can be reliably distinguished from zero. Effective project planning requires knowledge of the
detection capabilities of the measurement method that will be or could be used. This section
explains an MQO called the minimum detectable concentration (MDC) and describes
radioactivity detection capabilities, as well as methods for calculating it.
The method most often used to make a detection decision about radiation or radioactivity
involves the principles of statistical hypothesis testing. It is a specific example of a Scenario B
hypothesis testing procedure described in Section 7.2.4. To "detect" the radiation or radioactivity
requires a decision on the basis of the measurement data that the radioactivity is present. The
detection decision involves a choice between the null hypothesis (H0): There is no radiation or
radioactivity present (above background), and the alternative hypothesis (Hi): There is radiation
or radioactivity present (above background). In this context, a Type I error is to conclude that
radiation or radioactivity is present when it actually is not, and a Type II error is to conclude that
radiation or radioactivity is not present when it actually is.5 Making the choice between these
hypotheses requires the calculation of a critical value. If the measurement result exceeds this
critical value, the null hypothesis is rejected and the decision is that radiation or radioactivity is
present.
5
Note that in any given situation only one of the two types of decision error is possible. If the sample does not
contain radioactivity, a Type I error is possible. If the sample does contain radioactivity, a Type II error is possible.
January 2009 7-27 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
7.5.1 Calculate the Critical Value
The critical value defines the lowest value of the net instrument signal6 (count) that is too large
to be compatible with the premise that there is no radioactivity present. It has become standard
practice to make the detection decision by comparing the net instrument count to its critical
value, Sc. The net count is calculated from the gross count by subtracting the estimated
background and any interferences.7
The mean value of the net instrument count typically is positive when there is radioactivity
present (i.e., above background). The gross count must be corrected by subtracting an estimate of
the count produced under background conditions. See Section 7.8.2 for more information on
instrument background.
Table 7.5 lists some formulas that are commonly used to calculate the critical value, Sc, together
with the major assumptions made in deriving them. Note that the Stapleton formulas given in
rows 3 through 5 especially are appropriate when the total background is less than 100 counts.
These formulas depend on NB (background count), tB (background count time), ts ( sample count
time), and z\.a (the (1 - a)-quantile of the standard normal distribution). The value of a
determines the sensitivity of the test. It is the probability that a detection decision is made when
no radioactivity above background is actually present.
More detail on the calculation of critical values is given in Section 7.9.3. Software (Strom 1999)
is available for calculating fusing the equations recommended here, among others.
Table 7.5 Recommended Approaches for Calculating the Critical Value of the Net
Instrument Signal (Count), Sc
1
2
Critical Value Equation
^•-M'i)
5, =2.33^
Assumptions
Poisson
Poisson
a = 0.05
Background
Count
>100
>100
"Net instrument signal," is used here as a general term, because many radiation-detection instruments may have
output other than "counts" (e.g., current for ionization chambers). In cases where the instrument output is in counts,
the term "net counts" can be substituted for the term "net instrument signal."
"Interference" is the presence of other radiation or radioactivity or electronic signals that hinder the ability to
analyze for the radiation or radioactivity of interest.
Q
These particular expressions for the critical value of the net instrument signal (in this case the net count) depend
for their validity on the assumption of Poisson counting statistics. If the variance of the blank signal is affected by
interferences, or background instability, then the Equation 20.7 of MARLAP may be more appropriate.
NUREG-1575, Supp. 1
7-28
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
-2
A
5
Critical Value Equation
e /jM ?s 1 I Zi-« - 1 i ?s i - If V l /A tg 1 l ?s
U J 4 I 'J f u w
5 01"f/s lliL6452"fli 'OiieiS/fV lOl^fli /sl
Sc 0.4x^- 1J, 4 x^ + -J + 1.645^,+0.4)-^ + -J
Sc =1.35 + 2.33^/^+0.4
Assumptions
Stapleton
tB±tS
Stapleton
tB + ts
a =0.05
d = 0.4
Stapleton
ts ts
a =0.05
d = 0.4
Background
Count
< 100
< 100
<100
d = the critical value of the net instrument signal parameter in the Stapleton Equation
Example 7: A 600-second background measurement is performed on a proportional counter and
108 beta counts are observed. A sample is to be counted for 300 s. Estimate the critical value of
the net instrument signal (i.e., net count) when a = 0.05.
=14.8 net counts
Therefore, if 15 or more net counts are observed, the decision will be made that the sample
contains radioactivity above background. Values of Sc should be rounded up when necessary to
make sure that the specified Type I error probability, a, is not exceeded.
7.5.2 Calculate the Minimum Detectable Value of the Net Instrument Signal or Count
Table 7.6 lists some formulas that are commonly used to calculate the minimum detectable net
count, SD, together with the major assumptions made in deriving them. SD, is defined as the mean
value of the net instrument signal or count that gives a specified probability, 1 - ft, of yielding an
observed net instrument signal or count greater than its critical value Sc. Therefore, Sc must be
calculated before SD. Note specifically that the Stapleton formulas given in rows 4 and 5 are
especially appropriate when the total background is less than 100 counts. Generally, the
Stapleton methods may be used for both high and low total background counts as they agree well
with the more traditional methods when the background counts are over 100. The simpler, more
familiar formulas have been included for completeness.
It is important that the assumptions used to calculate SD are consistent with those that were used
to calculate Sc. The equations for SD depend on the same variables as Sc, namely NB, tB, and ts.
Notice that neither a nor z\.a appears explicitly, rather they enter the calculation through Sc.
However, ft now enters the calculation of SD through Zj -p. The value of ft, like a, is usually
chosen to be 0.05 or is assumed to be 0.05 by default if no value is specified.
January 2009
7-29
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
Table 7.6 Recommended Approaches for Calculating the Minimum Detectable Net
Instrument Signal or Count9
1
2
3
4
5
Minimum Detectable Net Signal Equation
2 1 2 f ~\
o o , Zi-/? , _ p-/7 , o | Y ls i ts\
^D ^c ' 9 ' zi-/?J , ! ^c ! yvs * 1
/ \ ^ *B V B )
b>D = Zj_^ + 2oc
SD =2.71 + 2SC = 2.71 + 2(2.33-7^7) = 2.71 + 4.66^7
^ (*l-a+^)2f1 O f , L ^L O
4 rd ^r^rd
^=5.41 + 4.65^/7^
Assumptions
Poisson
tB±ts
Poisson
tB+tS
a = fi
Poisson
a = yg = 0.05
tB = ts
Stapleton
Stapleton
a = /? = 0.05
tB = ts
Background
Count
> 100
>100
>100
<100
<100
Example 8 A 600-second background measurement on a proportional counter produces 108 beta
counts and a source is to be counted for 300 s. Assume the background measurement gives the
available estimate of the true mean background count rate and use the value 0.05 for Type I and
Type II error probabilities. From section 7.5.1, Example 7, the critical net count, Sc, equals 14.8,
so SD = z\_p + 2SC = 1.6452 + 2 (14.8) = 32.3 net counts. Values of SD should be rounded up when
necessary to make sure that the specified Type II error probability, /?, is not exceeded.
The relationship between the critical value of the net instrument signal (or count), Sc, and the
minimum detectable net instrument signal (or count), SD, is shown in Figure 7.6. Figure 7.6
illustrates a case where alpha is greater than beta. The net instrument signal (or count) obtained
for a blank sample will usually be distributed around zero as shown. Occasionally, a net count
rate above Sc may be obtained by chance. The probability that this happens is controlled by the
value of a, shown as the lightly shaded area in Figure 7.6. Smaller values of a result in larger
values of Sc and vice versa. The minimum detectable value of the net instrument signal (or
count) SD is that value of the mean net instrument signal (or count) that results in a detection
decision with probability I-/?. That is, there is only a probability equal to/?, shown as the more
These expressions for the critical value of the net count depend for their validity on the assumption of Poisson
counting statistics. If the variance of the blank signal is affected by interferences, or background instability, then
Equation 20.7 of MARLAP may be more appropriate. "Interference" is the presence of other radiation or
radioactivity or electronic signals that hinder the ability to analyze for the radiation or radioactivity of interest.
NUREG-1575, Supp. 1
7-30
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
darkly shaded area in Figure 7.6, of yielding an observed count less than Sc. Smaller values of ft
result in larger values of SD and vice versa.
Figure 7.6 The Critical Value of the Net Instrument Signal (Sc) and the Minimum
Detectable Net Signal (SD)
More information detail on the calculation of the minimum detectable value of the net instrument
signal, SD, is given in Section 7.9.
7.5.3 Calculate the Minimum Detectable Concentration
The MDC is usually obtained from the minimum detectable value of the net instrument signal (or
count), SD. The MDC is by definition an estimate of the true concentration of the radiation or
radioactivity required to give a specified high probability that the measured response will be
greater than the critical value. The common practice of comparing a measured concentration to
the MDC, instead of to the Sc, to make a detection decision is incorrect. To calculate the MDC,
the minimum detectable value of the net signal (or count), SD, must first be converted to the
detectable value of the net instrument signal per unit time (or count rate), SDI ts(s~'). This in turn
must be divided by the counting efficiency, e (s~')fBq to get the minimum detectable activity, yD.
Finally, the minimum detectable activity can be divided by the sample volume or mass to obtain
the MDC. At each stage in this process, additional uncertainty may be introduced by the
uncertainties in time, efficiency, volume, mass, etc. Thus, prudently conservative values of these
factors should be used so that the desired detection power, I-/?, at the MDC is maintained.
Another approach would be to recognize that_yD itself has an uncertainty which can be calculated
using the methods of Section 7.4. Thus any input quantity that is used to convert from SDtoyD
that has significant uncertainty can be incorporated to assess the overall uncertainty in the MDC.
Additional discussion of the calculation of the MDCs is given in Section 7.9.5.
Example 9: Continuing Example 8, SD = 32.3 net counts.
Assuming negligible uncertainty in the count time, the net count rate is
ts = 32.3/300 = 0.1077 (r1) .
The mean efficiency from Example 6 in Section 7.4.2 was 0.4176 (5 :)/(Bq) with a CSU of uc
(s) = 0.005802.
January 2009
7-31
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
In Example 8, the value 0.05 was specified for both Type I and Type II error probabilities. So the
specified power was I-/? = 1 - 0.05 = 0.95.
Assume a normal distribution for e, to obtain a 95% probability of detection for the MDC. To
account for the variability in the efficiency, the value used for s should be the 5th percentile, i.e.,
0.4176 - 1.645(0.005802) = 0.4081.
S It
Thus, the minimum detectable activity,yD = D s =0.1077/0.4081 = 0.2639Bq.
£
Using the mean value of the efficiency would potentially underestimate the minimum detectable
activity as yD =^^ = 0.1077 70.4176 = 0.2578Bq.
£
These values for yD would then be divided by the mass or volume of the sample to yield the
MDC.
7.5.4 Summary of Measurement Detectability
The concepts surrounding the MDC and the critical value are illustrated in Figure 7.7, using
familiar formulae for Sc and SD discussed above, assuming a background count ofNs =100 with
a = ft = 0.5. In this case, the equation in row 2 of Table 7.5 was used to obtain Sc = 23.3, and the
corresponding equation in row 3 of Table 7.6 to obtain SD = 49.3. The use of these equations
implies a = ft = 0.05 and tB = t$. It is important to note that traditionally the values a = ft = 0.05
are used for MDC calculations, so that the MDCs for different methods are comparable.
However, when developing a standard operating procedure for a survey, other values for a and ft
may be more appropriate. A case where this typically occurs is in the calculation of scan MDCs
(Section 7.11.6) where a may be much greater than /?, because the consequences associated with
misidentifying a background area as elevated are much lower than the consequences associated
with missing a true elevated area.
Note, the upper abscissa scale is in concentration and the lower abscissa scale is in net count.
These are related by the efficiency at the point where the MDC corresponds to the minimum
detectable net instrument signal (or count), SD. Each of the curves illustrates the distribution of
mean net counts (or concentration) that may exist for a measurement. The width of these curves
represents the variation due to counting statistics. The variability due to other factors is
associated with uncertainty in £. Changes in the relationship between the lower and the upper
scales result from changes in £. This illustrates the importance of choosing realistic, or even
conservative, values of £. Note that the probability of making a detection decision (which is
proportional to the area of each curve to the right of Sc) depends on the concentration, increasing
from 5% at background to 95% at the MDC, passing through 50% at Sc. This is perhaps more
clearly shown in Figure 7.8, which plots the probability of making a detection decision as a
function of net instrument signal, count, or concentration.
Figure 7.8 shows that for concentrations corresponding to net counts between 0 and Sc the
probability of a non-detect is greater than 50%. For concentrations corresponding to net counts
between Sc and SD the probability of detection is greater than 50%, but less than 95%.
NUREG-1575, Supp. 1 7-32 January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
Concentrations above the MDC (with net counts greater than SD) are highly likely to be detected,
but will have relative standard uncertainties that are somewhat large.
Concentration = net count rate/efficiency
o MDC
Upper x-axis in units of
concentration, lower x-
axis in units of net
counts
Figure 7.7 Relationship Between the Critical Value of the Net Count, the
Minimum Detectable Net Counts and the MDC
Concentration
Probability
Measure
>S
0 Sc
Net Count
Figure 7.8 Probability of Detection as a Function of Net Count (Lower X-Axis) and
Concentration (Upper X-Axis)
January 2009
7-33
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
7.5.5 Measurement Detectability Recommendations
• When a detection decision is required, generally it should be made by comparing the net
instrument signal (or count) to its corresponding critical value.
• Expressions from Tables 7.5 and 7.6 for Sc and SD should be chosen to match the
assumptions and background for the measurement method.
• An appropriate background should be used to predict the instrument signal produced when
there is no radioactivity present in the sample.
• The Minimum Detectable Concentration (MDC) should be used only as a MQO for the
measurement method. To make a detection decision, a measurement result should be
compared the critical value and never to the MDC.
• The validity of the Poisson approximation for the measurement process should be confirmed
using the methods described in MARLAP Chapter 20 before using an expression for the
critical value that is based on Poisson statistics. When the Poisson approximation is
inappropriate for determining the critical value, estimating o by the sample standard
deviation of replicated background measurements is preferable to using the square root of the
number of counts.
• Consider all significant sources of variance in the instrument signal (or other response
variable) when calculating the critical value, Sc, and minimum detectable value, SD.
• Report each measurement result and its uncertainty as obtained even if the result is less than
zero. Never report a result as "less than MDC" or "less than Sc"
• The MDC should not be used for projects where the issue is a quantitative comparison of
measurements to a limit rather than just a detection decision made for a single measurement.
For these projects, the minimum quantifiable concentration is a more relevant MQO for the
measurement process (see Section 7.6).
7.6 Determine Measurement Quantifiability
This section discusses issues related to measurement quantifiability. Much of this material is
derived from the MARLAP Chapter 20. Further details and an additional example are given in
Section 7.10.
Action levels are frequently stated in terms of a quantity or concentration of radioactivity, rather
than in terms of detection. In these cases, project planners may need to know the quantification
capability of a measurement method, or its capability for precise measurement. The
quantification capability is expressed as the smallest concentration of radiation or radioactivity
that can be measured with a specified relative standard deviation. This section explains an MQO
called the minimum quantifiable concentration (MQC), which may be used to describe
quantification capabilities.
The MQC of the concentration, _yg, is defined as the concentration at which the measurement
process gives results with a specified relative standard deviation llkq where kq is usually chosen
to be 10 for comparability.
NUREG-1575, Supp. 1 7-34 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
Historically much attention has been given to the detection capabilities of radiation and
radioactivity measurement processes, but less attention has been given to quantification
capabilities. For some projects, quantification capability may be a more relevant issue. For
example, suppose the purpose of a project is to determine whether the 226Ra concentration on
material at a site is below an action level. Since 226Ra can be found in almost any type of
naturally occurring material, it may be assumed to be present in every sample, making detection
decisions unnecessary. The MDC of the measurement process obviously should be less than the
action level, but a more important question is whether the MQC is less than the action level.
A common practice in the past has been to select a measurement method based on the minimum
detectable concentration (MDC), which is defined in Section 7.5. For example, MARSSIM
(2002) says:
During survey design, it is generally considered good practice to select a measurement
system with an MDC between 10-50% of the DCGL [action level].
Such guidance implicitly recognizes that for cases when the decision to be made concerns the
mean of a population that is represented by multiple measurements, criteria based on the MDC
may not be sufficient and a somewhat more stringent requirement is needed. The requirement
that the MDC (approximately 3-5 times CTM) be 10% to 50% of the action level is tantamount to
requiring that OM be 0.02 to 0.17 times the action level - in other words, the relative standard
deviation should be approximately 10% at the action level. However, the concentration at which
the relative standard deviation is 10% is the MQC when kq assumes its conventional value of
10. Thus, a requirement that is often stated in terms of the MDC may be more naturally
expressed in terms of the MQC, e.g., by saying that the MQC should not exceed the action level.
7.6.1 Calculate the MQC
The minimum quantifiable concentration, when there are no interferences, can be calculated
from:
1 / I O O /" / N \ '
kn 4(l-£'fV t ( t \]
ye = o n /2,2, 1 + J1+ /2 #B- 1 + - C7-14)
2tss(l-kQ(f).)\^ ^ kQ (_ tB{ tBJJ^
Where:
ts = count time for the source, s
tB = count time for the background, s
NB = background count
l = relative variance of the measured efficiency, s (see Section 7.8.2.2)
kQ = relative percent standard deviation at the MQC, usually assumes a conventional
value of 10 for purposes of comparison among methods
If k2Q $ > 1, this equation has no solution.
January 2009 7-35 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Example 10: Continuing Example 9, ts = 300, tB = 600, NB = 108, $ = (0.005802/0.4176)2 =
0.0001932, andkQ= 10. So,
I I Z Z X 7 N N. N
I | 4(l-100(0.0001932))rios300ri | 300^
2(300)(0.4176)(1- 100(0.0001932)) \ 100 ^ 600 ^ 600 j^
J
= 1.239 Bq. This value for yQ would then be divided by the mass or volume of the sample to
yield the MQC. _
The next example is given to verify that the equation for yq does indeed produce a value with a
relative uncertainty of 10%. It also provides an opportunity to give another illustration of the
methodology for the calculation of measurement uncertainty developed in Sections 7.4 and 7.8.
Additional information on the calculation of MQCs is given in Section 7.10.
Example 11: The calculations of Example 10 can be verified by calculating the uncertainty of a
measurement made at the MQC. The expected number of counts for a sample at the MQC
counted for 300 s:
tsltB} = (1.239 Bq)(300 s)(0.4176) + (108 s"1)(300/600) = 209,
rounded to the nearest whole number.
The model equation is the same as was used in Example 6, Section 7.4.2:
y = (__s — s) ( B — B)_ ^ so the equation for tne CSU is the same:
2,v^
u2(NB)
^0.4176 J ^ 0.4176 J ^ 0.4176"
= 1.332xlO"2+1.72xlO"3+2.95xlO"4=1.534xlO"2
uc (y) = Vl.534xlO~2 = 0.124. Thus, the relative uncertainty at the MQC is 0.124/1.239 =
0.09995. This means, apart from some small difference due to rounding, the relative
measurement uncertainty at^g is 10%, as should be the case for the MQC.
7.6.2 Summary of Measurement Quantifiability
Figure 7.9 is a modification of Figure 7.8, illustrating the relationships between the critical value,
the MDC, the MQC and the probability of exceeding the critical value. As can be seen, the issue
of detection is almost moot at the MQC. The probability of detection is near 100%. However, the
NUREG-1575, Supp. 1 7-36 January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
MQC specifies a concentration with a defined relative standard uncertainty, making comparisons
between measurements or comparisons between measurements and regulatory criteria
meaningful.
c
o
0)
a
£1
O
0.95
0.5
0.05
Gray Region
Net Count: °
N/A
Concentration: 0
N/A
MDC
MQC
Multiple of the uncertainty: Q 1.5-2.5
3-5
10
Figure 7.9 Relationships Among the Critical Value, the MDC, the MQC, and the
Probability of Exceeding the Critical Value
Three x-axis scales are shown in Figure 7.9 for net count, concentration, and multiple of
measurement uncertainty. This figure emphasizes, for example, that the minimum detectable net
count, SD, corresponds to the MDC, but has different units. It also shows that the MQC is by
convention 10 times the measurement uncertainty at that concentration. The critical value of the
net count, Sc, has no corresponding common term in concentration units. This is because
detection decisions are usually made on the basis of the net counts (instrument reading). These
are inherently qualitative "yes or no" decisions. The relationship between Sc and SD and the
multiple of the uncertainty varies according to which set of assumptions are used and which
equations in Table 7.5 and Table 7.6 are appropriate to those assumptions. Therefore, an
approximate range is shown for these quantities on the multiple of uncertainty axis.
7.7 Establish a Required Measurement Method Uncertainty
This section provides the rationale and guidance for establishing project-specific MQOs for
controlling
-------
Statistical Basis For MARSAME Surveys MARSAME
7.7.1 Developing a Requirement for Measurement Method Uncertainty for MARSSIM-
Type Surveys
When, as in MARSSEVI-type surveys, a decision is to be made about the mean of a sampled
population, generally the average of a set of measurements on a survey unit is compared to the
disposition criterion.
The total variance of the data, cr2, is the sum of two components
a2=a2M+a2s (?_15)
Where:
a2M = measurement method variance (M for measurement)
cr2 = variance of the radionuclide concentration or activity concentration in the
sampled population (S for sampling)
The spatial and temporal distribution of the concentration (i.e., the variation of the true but
unknown concentrations from place to place and from time to time), the extent of the survey unit,
the physical sizes of the measured material, and the choice of measurement locations may affect
the sampling standard deviation, as. The measurement standard deviation, aM, is affected by the
measurement methods. The value of aM is estimated in MARSAME by the CSU of a measured
value for a measurement of material whose concentration equals the hypothesized population
mean concentration. The calculation of measurement uncertainties is covered in Sections 7.4 and
7.8.
Four cases are considered below where target values for ox/can be suggested depending on what
is known about as. Cases 1 and 2 treat the desired overall objective of keeping A/cr~ 3 or higher.
When this is not possible, Cases 3 and 4 treat the less desirable alternative of attempting to
prevent A/'a from going lower than 1. If A/cr< 1 then a large number of measurements will be
required to meet the Type I and II decision error rates specified in the DQO process. If cr» A, it
may be necessary to re-evaluate the error rates specified in the DQO process.
Case 1: o$ is known relative to A / 3
Generally, it is easier to control A/3
If as > A/3, the requirement that the total crbe less than A/3 cannot be met regardless of
aM. In this case, it is sufficient to make aM negligible in comparison to as. Generally, aM
can be considered negligible in comparison to crsif aM < crs/3.
NUREG-1575, Supp. 1 7-38 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
Case 2: a8 is not known relative to A/3
Often one needs a method for choosing OM in the absence of specific information about as. Since
it is desirable to have a< A/3, this condition is adopted as a primary requirement. Assume for the
moment that as is large. Then aM should be made negligible by comparison. As mentioned
above, aM can be considered negligible if it is no greater than crs/3. When this condition is met,
further reduction of aM has little effect on a and therefore is usually not cost-effective. So, OM <
as/3 is adopted as a secondary requirement.
Starting with the definition a2 = o2M + a2s and substituting the secondary requirement OM < os/3
we get a2 > a2M + 9a2M = \Qa2M , thus
Substituting the primary requirement that A/cr> 3 (i.e., o< A / 3) we get 3) will be met if crsis small. When the measurement
standard deviation aM is less than a^, the primary requirement will be met unless the sampling
variance,
-------
Statistical Basis For MARSAME Surveys MARSAME
It may be that the primary requirement that A/crbe at least 3 is not achievable. Suppose that the
primary requirement is relaxed to achieving A/crat least 1 (i.e., cr< A). This leads to
consideration of-
Case 3: c%is known relative to A
As in Case 1, it is generally easier to control aM than as. If as is known (approximately), a target
value for A
If Us > A, the requirement that the total crbe less than A cannot be met regardless of CTM. In
this case, it is sufficient to make aM negligible in comparison to as. Generally, aM can be
considered negligible if it aM < crs/3.
Case 4: crs is not known relative to A
Suppose cr< A is adopted as the primary requirement. As in Case 2, if crsis large, then a2M + 9a2M = lOcr^ , thus-
"»*-k (7-20)
Substituting the primary requirement that A/cr> 1 (i.e., cr< A) we get
-------
MARSAME Statistical Basis For MARSAME Surveys
so, the multiplier of uc(x) equals zi_a, the (l-a)-quantile of the standard normal distribution (see
MARLAP appendix C).
Alternatively, if AL = 0, so that any detectable amount of radioactivity is of concern, the
decision may involve comparing the net instrument signal (e.g., count rate) to the critical value
of the net instrument signal, Sc, as defined in Section 7.5.1.
Two cases are considered below where target values for o^can be suggested depending on what
is known about the width of the gray region and the desired Type I and Type II decision error
rates. Case 5 is for Scenario A, and Case 6 is for Scenario B.
Case 5: Suppose the null hypothesis is X> AL (see Scenario A in Chapter 4), so that the action
level is the upper bound of the gray region. Given the measurement variance a2M, only a
measured result that is less than (UBGR - z\-a LBGR +
ZI-O.OM + ZI-P
as an MQO for method uncertainty when decisions are to be made about individual items or
locations and not about population parameters.
If a =/? = 0.05, one may use the value UMR = 0.3 A. Other combinations of a and ft may lead to a
similar result, but the relationship is nonlinear (depending on the standard normal distribution
function) so one cannot simply apply a proportionality factor. Equation 7-25 must be used,
January 2009 7-41 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
The recommended value of MM? is based on the assumption that any known bias in the
measurement process has been corrected and that any remaining bias is well less than a third of
the method uncertainty.
7.8 Calculate the Combined Standard Uncertainty of a Measurement
This section expands upon the material in Section 7.4. Calculations of combined standard
uncertainties (CSUs) can be complex, and typically would be carried out using a software
package. For the purpose of illustration and clarity, fully worked out examples are included in
this section.
7.8.1 Procedures for Evaluating Uncertainty
The usual eight steps for evaluating and reporting the uncertainty of a measurement are
summarized in the following subsections (adapted from Chapter 8 of GUM):
7.8.1.1 Identify the Measurand, 7, and all the Input Quantities, Xt, for the Mathematical Model
Include all quantities whose variability or uncertainty could have a potentially significant effect
on the result. Express the mathematical relationship, Y=f(X\, X2,...rXN), between the measurand
and the input quantities.
The procedure for assessing the uncertainty of a measurement begins with listing all significant
sources of uncertainty in the measurement process. A good place to begin is with the input
quantities' mathematical model Y=f(X\, X2,...rXN). When an effect in the measurement process
that is not explicitly represented by an input quantity has been identified and quantified, an
additional quantity should be included in the mathematical measurement model to correct for it.
The quantity, called a correction (additive with a nominal value of zero) or correction factor
(multiplicative with a nominal value of one), will have an uncertainty that should also be
evaluated and propagated. Each uncertainty that is potentially significant should be evaluated
quantitatively.
7.8.1 .2 Determine an Estimate, xf, of the Value of Each Input Quantity, Xt
This involves simply determining for the particular measurement at hand, the specific value, xt,
that should be substituted for the input quantity Xt in the mathematical relationship, Y=f(Xi,
7.8.1.3 Evaluate the Standard Uncertainty, u(Xj), for Each Input Estimate, xt, Using a Type A
Method, a Type B Method, or a Combination of Both
Methods for evaluating standard uncertainties are classified as either "Type A" or "Type B"
(NIST 1994). Both types of uncertainty need to be taken into consideration. A Type A evaluation
of an uncertainty uses a series of measurements to estimate the standard deviation empirically.
Any other method of evaluating an uncertainty is a Type B method. A Type B evaluation of
standard uncertainty is usually based on scientific judgment using all the relevant information
available, which may include:
NUREG-1575, Supp. 1 7-42 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
• Previous measurement data,
• Experience with, or general knowledge of, the behavior and property of relevant materials
and instruments,
• Manufacturer's specifications,
• Data provided in calibration and other reports, and
• Uncertainties assigned to reference data taken from handbooks.
The Type A standard uncertainty of the input estimate xt is defined to be the experimental
standard deviation of the mean:
Example 12: Type A uncertainty calculation using Equation 7-26:
Ten independent one-minute measurements of the counts from a check source Xt were made with
a digital survey meter, yielding the values:
12,148, 12,067, 12,207, 12,232, 12,284, 12,129, 11,862, 11,955, 12,044, and 12,150.
The estimated value xf is the arithmetic mean of the values Xtk.
J_£ 121078 *
1 nii '•* 10
The standard uncertainty of xt is
U(X) = \ V (X x)2 = 1 V (X -12107.8)2
l) VW(W-l)tf a '' VlO(10-l)tT a
= Vl 6628.84 =128.95
There are other Type A methods, but all are based on repeated measurements.
Any evaluation of standard uncertainty that is not a Type A evaluation is a Type B evaluation.
Sometimes a Type B evaluation of uncertainty involves making a best guess based on all
available information and professional judgment. Despite the reluctance to make this kind of
evaluation, it is almost always better to make an informed guess about an uncertainty component
than to ignore it completely.
There are many ways to perform Type B evaluations of standard uncertainty. One example of a
Type B method is the estimation of counting uncertainty using the square root of the observed
counts. If the observed count is N, when the Poisson approximation is used, the standard
uncertainty of TV may be evaluated as u(N) =
January 2009 7-43 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARS AME
Example 13: The standard uncertainty of the first value in Example 12, (12,148 counts), could
be estimated as V12148 = 1 10.218 counts. When TV may be very small or even zero, the equation
u(N) = \/N + 1 may be preferable.
Another Type B evaluation of an uncertainty u(x) consists of estimating an upper bound, a, for
the magnitude of the error of x based on professional judgment and the best available
information. If nothing else is known about the distribution of the measured result, then after a is
estimated, the standard uncertainty may be calculated using the equation
u(x) = -?= (7-27)
V3
which is the standard deviation of a random variable uniformly distributed over the interval
(x - a, x + a). The variable a is called the half-width of the interval.
Example 14: Suppose in Example 12, all that was given was the observed range of the data from
an analog survey meter dial (i.e., from 11,862 to 12,284), a difference of 422. If it was assumed
that the data came from a uniform distribution across this range, then the average is
(11,862+12,284)72= 12,073
the half- width is 21 1, and an estimate of the standard uncertainty would be
u(x) = = 121. 821
"s/3
Given the same information on the range, if values near the middle of the range were considered
more likely than those near the endpoint, a triangular distribution may be more appropriate. The
standard uncertainty for a triangular distribution is calculated using the equation
u(x) = -^ (7-28)
which represents the standard deviation of a random variable with a triangular distribution over
the interval (x - a, x + a). Given the same information on the range, if values near the middle of
the range were considered more likely than those near the endpoints, a triangular distribution
may be more appropriate. The mean would be the same as above, 12,073. However the standard
uncertainty then be calculated using the equation
86.14 (7-29)
Example 15: As in Example 14, all that is given was the observed range of the data from an
analog survey meter dial, i.e., from 1 1,862 to 12,284, a difference of 422. If it was assumed that
the data came from a triangular distribution across this range, then the average is
(1 1,862+12,284)72 = 12,073, the half-width is 211, and an estimate of the standard uncertainty
would be
NUREG-1575, Supp. 1 7-44 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
When the estimate of an input quantity is taken from an external source, such as a book or a
calibration certificate, the stated standard uncertainty can be used.
7.8.1.4 Evaluate the Covariances, u(xt,Xj), for all Pairs of Input Estimates with Potentially
Significant Correlations
A Type A evaluation of the covariance of the input estimates xt = and x, = is
u(x x ) = - . - Y(r., -x)(x,t -x) (7-30)
V 11 j! , i\Z— (v ',k !'v J,k ]' V '
n\n ~ 1) k=l
An evaluation of variances and covariances of quantities determined by the method of least
squares may also be a Type A evaluation. Evaluation of the covariance of two input estimates, xt
and Xj, whose uncertainties are evaluated by Type B methods may require expert judgment. In
such cases it may be simpler to estimate the correlation coefficient,
)-u(Xj)] (7-3 1 )
first and then multiply it by the standard uncertainties, u(Xj) and u(x}) to obtain the covariance,
u(Xi,xJ).
A covariance calculation is demonstrated in Example 16 in Section 7.8.2.2.
7.8.1.5 Calculate the Estimate,^, of the Measurand from the Relationship^ =f(x],x2, ...,XN)
This involves simply substituting, for the particular measurement at hand, the specific values of
Xj for the input quantity Xt into the mathematical relationship, Y =f(X\,X^ . . . J(N\ and calculating
the result^ =f(x^x^ . . . ,XN).
7.8.1.6 Determine the Combined Standard Uncertainty, uc(y), of the Estimate, y
The CSU of y is obtained using the following formula:
N ( flf\2 N-l N flf fit
u2(v) = y — U/2(x)+2y y — -^—u(x x} (7-32)
i=l { OX j j=i ]=i+\ OX OX •
Here, u2(Xj) denotes the estimated variance of x,, or the square of its standard uncertainty; u(Xj,Xj)
denotes the estimated covariance of x, and xy; df/dxj (or dy/dx^ denotes the partial derivative of/
with respect to xt evaluated at the measured values Xi,x2,... ,XN, and u2 (y) denotes the combined
variance ofy, whose positive square root, uc(y), is the CSU ofy. The partial derivatives, dfJdxf,
are called sensitivity coefficients, usually denoted ct. The sensitivity coefficient measures how
January 2009 7-45 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
much/changes when xt changes. Equation 7-32 is called the "law of propagation of uncertainty"
in GUM (ISO 1995).
If the input estimates x\,X2,.. .,XN are uncorrelated, the uncertainty propagation formula reduces to
(7-33)
Suppose the values xi,X2,...,x^ are composed of two groups wi,W2,...,wn andzi,Z2,...,zm with
N=n+m. If all the variables, w and z, are uncorrelated and nonzero, the CSU of_y =
1 2 " ' — may be calculated from the formula:
\Z1 •••Zm
(7.34)
The symbols Zi,z2,. . .,zm have been introduced simply to differentiate those values appearing in
the denominator of the model equation from the wi,W2,. . .,wn appearing in the numerator.
f (w w w ^)
= - - — 2''"' — — , where/is some specified function ofwi,W2,...,wn, all the z, are nonzero,
and all the input estimates are uncorrelated. Then:
Zl Z2 Zm
An alternative to uncertainty propagation is the use of computerized Monte Carlo methods to
propagate not the uncertainties of input estimates but their distributions. Given assumed
distributions for the input estimates, the method provides an approximate distribution for the
output estimate, from which the CSU or an uncertainty interval may be derived.
7.8.1.7 Optionally Multiply uc(y) by a Coverage Factor k to Obtain the Expanded Uncertainty
U
The interval \y-U,y+ U], constructed using the expanded uncertainty U= k-uc(y), can be
expected to contain the value of the measurand with a specified probability,/?. The specified
probability, p, is called the "level of confidence" or the "coverage probability" and generally is
only an approximation of the true probability of coverage. When the distribution of the measured
result is approximately normal, the coverage factor often is chosen to be k = 2 for a coverage
probability of approximately 95%. An expanded uncertainty calculated with k = 2 or 3 is
sometimes informally called a "two-sigma" or "three-sigma" uncertainty, respectively. The
GUM recommends the use of coverage factors in the range of 2 to 3 when the CSU represents a
good estimate of the true standard deviation. Attachment 19D of MARLAP describes a more
general procedure for calculating the coverage factor that gives a desired coverage probability/?
when there is substantial uncertainty in the value of wc(y).
NUREG-1575, Supp. 1 7-46 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
7.8.1.8 Report the Result as^ ± f/with the Unit of Measurement
At a minimum, state the coverage factor used to compute U and the estimated coverage
probability. Alternatively, report the result, y, and its CSU, uc(y\ with the unit of measurement.
The number of significant figures that should be reported for the result of a measurement
depends on the uncertainty of the result. A common convention, recommended by MARLAP, is
to round the uncertainty (standard uncertainty or expanded uncertainty) to two significant figures
and to report both the measured value and the uncertainty to the same number of decimal places.
Only final results should be rounded in this manner. Intermediate results in a series of calculation
steps should be carried through all steps with additional figures to prevent unnecessary round-off
errors. Additional figures are also recommended when the data are stored electronically.
Rounding should be performed only when the result is reported. Many of the values in the
examples given in MARSAME carry more significant digits so that the calculations can be
reasonably reproduced by the reader. All results, whether positive, negative, or zero, should be
reported as obtained, together with their uncertainties.
A measured value y of a quantity 7 that is known to be positive may be so far below zero that it
indicates a possible blunder, procedural failure, or other quality control problem. Usually, if
y + 3uc(y) < 0, the result may be invalid. For example, ify = -10 and uc(y) = 1, this would imply
that 7 is negative with high probability, which is known to be impossible. However, ify = -1 and
wc(y) = 1, the expanded uncertainty covers positive values with reasonable probability. The
accuracy of the uncertainty estimate uc(y) must be considered in evaluating such results,
especially in cases where only few counts are observed during the measurement and counting
uncertainty is the dominant component of uc(y). (See MARLAP Chapter 18 and Attachment
19D).
7.8.2 Examples of Some Parameters that Contribute to Uncertainty
The sources of uncertainty described in the following sections, drawn from MARLAP Section
19.5, should be considered.
7.8.2.1 Instrument Background
Single-channel background measurements are usually assumed to follow the Poisson model, in
which the uncertainty in the number of counts obtained, N, is given by \IN . There may be effects
that increase the variance beyond what the model predicts. For example, cosmic radiation and
other natural sources of instrument background may vary between measurements, the instrument
may become contaminated, or the instrument may simply be unstable. Generally, the variance of
the observed background is somewhat greater than the Poisson counting variance, although for
certain types of instruments, the Poisson model may overestimate the background variance
(Currie et al., 1998). If the background does not closely follow the Poisson model, its variance
should be estimated by repeated measurements.
The "instrument background," or "instrument blank," is usually measured under the same
conditions that will be encountered in the field. Ambient background sources should be
January 2009 7-47 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
minimized, and kept constant during the measurements of M&E. Periodic checks should be
made to ensure that the instrument has not picked up additional radioactivity from the M&E
during the measurements. If the background drifts or varies non-randomly over time (i.e., is non-
stationary), it is important to minimize the consequences of the drift by performing frequent
background measurements.
If repeated measurements demonstrate that the background level is stable, then the average, x ,
the results of n similar measurements performed over a period of time may give the best estimate
of the background. In this case, if all measurements have the same duration, the experimental
standard deviation of the mean, s (x), is also a good estimate of the measurement uncertainty.
Given the Poisson assumption, the best estimate of the uncertainty is still the Poisson estimate,
which equals the square root of the summed counts, divided by the number of measurements,
Inx/ - x,
, but the experimental standard deviation may be used when the Poisson
assumption is invalid. It is always wise to compare the value of s(x) to the value of the Poisson
uncertainty when possible to identify any discrepancies.
7.8.2.2 Counting Efficiency
The counting efficiency for a measurement of radioactivity (usually defined as the detection
probability for a particle or photon of interest emitted by the source) may depend on many
factors, including source geometry, placement, composition, density, activity, radiation type and
energy and other instrument-specific factors. The estimated efficiency is sometimes calculated
explicitly as a function of such variables (in gamma-ray spectroscopy, for example). In other
cases a single measured value is used (e.g., alpha-particle spectrometry). If an efficiency function
is used, the uncertainties of the input estimates, including those for both calibration parameters
and sample-specific quantities, must be propagated to obtain the CSU of the estimated
efficiency. Calibration parameters tend to be correlated; so, estimated covariances must also be
included. If a single value is used instead of a function, the standard uncertainty of the value is
determined when the value is measured. An example of the calculation of the uncertainty in
counting efficiency is given in Example 16.
Example 16: A radiation counter is calibrated, taking steps to ensure that the geometry of the
source position, orientation of the source, pressure, temperature, relative humidity, and other
factors that could contribute to uncertainty are controlled, as described below:
The standard source is counted 15 times on the instrument for 300 s.
The radionuclide is long-lived; so, no decay corrections are needed. The uncertainties of the
count times are assumed to be negligible.
Within the range of linearity of the instrument, the mathematical model for the calibration is:
(7.36)
Where:
s = counting efficiency
n = number times the source is counted (15)
_ NS,J = gross count observed during the /th measurement of the source
NUREG-1575, Supp. 1 7-48 January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
ts = source count time (300 s)
NB = observed background count (87)
IB = background count time (6,000 s)
as = activity of the standard source (150.0 Bq). (The standard uncertainty of the
source, 2.0 Bq, was given in the certificate for the source.)
The CSU of e can be evaluated using Equation 7-36. For the purpose of uncertainty evaluation, it
is convenient to rewrite the model as:
R
£ = •
Where:
and Rt=(N8J/ts)-(NB/tB),
The values Rt and their average, R , are estimates of the count rate produced by the standard,
while R/as is an estimate of the count rate produced by 1 Bq of activity. The standard
uncertainty of R can be evaluated experimentally from the 15 repeated measurements:
_ _ 1 " _
u2(R) = s2(R) = - TX^ ~^)2 • Since only one background measurement was made, the
n(n-\}^
input estimates Rt are correlated with each other. The uncertainty of NB, u(NB} = V87 , using a
Type B evaluation based on an assumption of a Poisson distribution for the number of
background counts.
The covariance between Rt and R,-, for /' ^y, may be estimated as
dN dN
t t
60002
However, the correlation is negligible here because the uncertainty of the background count, NB,
is much smaller than the uncertainty of each source count, NSJ. So, the correlation of the input
estimates Rt will be approximated as zero (i.e., treated as if they were uncorrelated), and the
correlation terms dropped from Equation 7-32. This means the evaluation used to calculate the
CSU of s can proceed using equation 7-33.
a,.
u\R)-
dR J ^
V()Z)Y 2|V(a,)"
a.
da,.
u\as}= - u\R)+ -^ u\as}
(al (a:
a
a
. Therefore, uc (s) =,
\u\R)
January 2009
7-49
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Assume the following data were obtained
for the 1 5 separate counts
of the calibration source.
Count Number, i Gross count, Nx; Ri (s"1)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
18,375
18,644
18,954
19,249
19,011
18,936
18,537
18,733
18,812
18,546
18,810
19,273
18,893
18,803
18,280
Average, R (s"1)
Experimental standard deviation, s(Ri) (s *)
Experimental standard deviation
Then the estimated counting efficiency is
R
as
And the CSU of e is given by
, . /(0.2449s-1)2
ii i r\ — /
c\ ) \\ 2
Which may be rounded to 0.0058.
of the mean, s(R ) (s *)
62.6202s-1 nA,^
— f\ A 1 7^.
— VJ.H- i / o
150.0 Bq
2 (2.0 Bq)2
'" (150.0 Bq)2
61.236
61.236
61.236
64.149
63.356
63.106
61.776
62.429
62.692
61.806
62.686
64.229
62.962
62.662
60.919
62.6202
0.9483
0.2449
0 OOS809
\J . \J\JJ O\JZ,
The true counting efficiency may vary because of variations in geometry, position and other
influence quantities not explicitly included in the model. These sources of uncertainty may not
be controlled as they were in the above example. If this is the case, the standard uncertainty of e
should include not only the standard uncertainty of the estimated mean, as calculated in the
example, but also another component of uncertainty due to variations of the true efficiency
during subsequent measurements. The additional component may be written as efy, where $ is
the coefficient of variation (i.e., the standard deviation divided by the mean) of the true
efficiency. Then the total uncertainty of e is obtained by squaring the original uncertainty
estimate, adding e
-------
MARSAME Statistical Basis For MARSAME Surveys
(7-37)
In the example above, the experimental variance of the count rates, Rt, may be used to
estimate^. Section 18B.2 of Attachment 18B of MARLAP describes an approach for estimating
such "excess" variance in a series of measurements.
Variations in counting efficiency due to source placement should be reduced as much as possible
through the use of positioning devices that ensure a source with a given geometry is always
placed in the same location relative to the detector. If such devices are not used, variations in
source position may significantly increase the measurement uncertainty.
Calibrating an instrument under conditions different from the conditions under which M&E
sources are counted may lead to large uncertainties in the activity measurements. Source
geometry in particular tends to be an important factor for many types of radiation counters. If
correction factors are used, their uncertainties should be evaluated and propagated, as mentioned
in Section 7.8.1.1.
7.8.2.3 Digital Displays and Rounding
If a measuring device has a digital display with readability10 d, the standard uncertainty of a
measured value is at least S/2-J3 , which is the variance of a random variable uniformly
distributed over the interval (x - 6/2, x + 6/2). Note that this is the same result as given by
equation 7-24 with a = 6/2. This uncertainty component exists even if the instrument is
completely stable.
A similar Type B method may be used to evaluate the standard uncertainty due to computer
round-off error. When a value x is rounded to the nearest multiple of 10" (where n is an integer),
the component of uncertainty generated by round-off error is 10" /(2V3) . This component of
uncertainty should be kept small in comparison to the total uncertainty of x by performing
rounding properly and printing with an adequate number of figures. In a long calculation
involving mixed operations, carry as many digits as possible through the entire set of
calculations and then round the final result appropriately as described in MARLAP Section
19.3.7 (MARLAP 2004).
Example 17: The readability of a digital survey dose rate meter is 1 nGy/h. Therefore, the
minimum standard uncertainty of a measured absorbed dose rate is l/2>/3 = 0.29 nGy/h.
10 Readability is the smallest difference that can still be read on a display. For instruments with an analog indicating
device, the readability is equal to the smallest fraction of a scale interval that can still be estimated with reasonable
reliability or which can be determined by an auxiliary device. For instruments with a numeric indicator (digital
display), the readability is equal to one digital step.
January 2009 7-51 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Example 18: Suppose the results for Rt in Example 16 had been rounded to the nearest whole
number before the analysis. Then the average would be computed as 62.6 instead of 62.6202 and
the standard deviation would be computed as 0.9103 instead of 0.9483. This demonstrates the
effect that rounding intermediate results can have on subsequent calculations. If this rounding to
the nearest positive integer had already occurred prior to receiving the data, and the original data
were no longer available, a correction for it could be made when estimating the CSU of Rt. The
component of uncertainty generated by round-off error is 1/(2V3) :
u(Ri}= 0.91032+ —U =0.9549.
7.8.3 Example Uncertainty Calculation
To illustrate how the uncertainty calculations are performed in practice, the following example is
given based on that of Lewis et al. (2005). The calculation will be that of the CSU in the
calibration of a surface contamination monitor.
7.8.3.1 Model Equation and Sensitivity Coefficients
Surface contamination monitors are calibrated in terms of their response to known rates of
radioactive emissions. In practice this is achieved by using large-area, planar sources that have a
defined area and whose emission rates have been determined in a traceable manner. The
calibration is usually determined in terms of response per emission rate per unit area. In this
example, the source is positioned with its active face parallel to and at a distance of 3 mm from
the face of the detector. The monitor detector area (50 cm2) is smaller than the area of the
calibration source, which is a 10 cm x 10 cm layer of 14C on a thick aluminum substrate. The
monitor has an analog display and has a means to set the detector voltage.
The efficiency, s, is defined by:
_(M-B)xfvxfdxfuxfbs
Where:
M = ob served monitor reading, s"1
B = background reading, s"1
E = emission rate of the calibration source, s"1
A = area of the active portion of the calibration source, cm2
/v = plateau voltage factor
fd = source-detector separation factor
fu = source uniformity factor
fbs = backscatter factor
The sensitivity coefficients of Equation 7-38 are given by:
fjC1 C*
. = (A/E)xf,xfdxfuxfbs=—— (7-39)
,dubs
dM (M - B)
NUREG-1575, Supp. 1 7-52 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
— = -(A/E)xLxfdxfxfh = - — (7-40)
dB v d u bs (M-B)
fjp _ p
— = -(M-B)(A/E?)xfrXfdxfuxfb.=— (7-41)
±- (7-42)
A
f-jp p
= (M-B)(A/E)xfdxfuxfbs= — (7-43)
OJv Jv
p
-^- (7-44)
fd
r}p p
= (M-B)(A/E)xfvxfdxfbs= — (7-45)
dfu fu
f-jp p
= (M-B)(A/E)xfyxfdxfu=— (7-46)
OJbs Jbs
Under normal conditions, the factorsfv,fd,fu and/^ are each assumed to have a value of one. If
the uncertainties are to be calculated in relative terms, the uncertainty equation becomes (see
Equation 7-34):
~~ + ~~ + ~~ + ~ + (~ + (~ + (~
If the relative uncertainties are all expressed as percentages, —— , where xf is an input quantity,
V Xi )
then the CSU will be a percentage. The relative sensitivity coefficients, ct are the terms
fO
multiplying each relative uncertainty term —^ in Equation 7-47. This approach produces
U- J
relative sensitivity coefficients of unity for the last 6 terms.
7.8.3.2 Uncertainty Components
Monitor Reading of Source, M (Type A)
Several techniques can be used to determine the mean observed monitor reading, M, and its
uncertainty. Assume a snap-shot technique is used whereby six successive, but randomly timed,
readings are recorded, giving 350, 400, 400, 325, 350, 350 s"1. The mean and standard deviation
January 2009 7-53 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
of the mean becomes 362.5 ± 12.5 s l. This equates to a percentage uncertainty inMof 3.45%
M
and the relative sensitivity coefficient from Equation 7-47, , is 362.57(362.5 - 32.5),
(M-B)
which is equal to 1.10. The distribution is assumed to be normal.
Monitor Reading of Background, B (Type B)
In this case, an eye-averaging technique was used whereby the highest and lowest count rates
were recorded over a given period of time. These count rates were 40 and 25 s"1 respectively,
giving a mean value of 32.5 s"1. This value is assumed to have a rectangular distribution with a
half-width of 7.5 s"1, and an uncertainty of 7.57 V3 = 4.330, equating to a percentage uncertainty
of 4.330/32.5 = 0.1332 or 13%. The relative sensitivity coefficient from Equation 7-47,
D
, is 32.57(362.5 - 32.5), which gives a value of 0.098.
(M-B)
Emission Rate of Calibration Source, E (Type B)
The emission rate of the source and its uncertainty were provided on the calibration certificate by
the laboratory that calibrated the source using a windowless proportional counter. The statement
on the certificate was:
"The measured value of the emission rate is E = 2,732±13 s"1"
The reported uncertainty is based on a standard uncertainty multiplied by a coverage factor of
k = 2, which provides a level of confidence of approximately 95%. The standard uncertainty onE
is therefore 13/2 = 6.5 s"1 or 0.24%. Unless the certificate provides information to the contrary, it
is assumed that the uncertainty has a normal distribution.
Source Area, A (Type B)
In the absence of an uncertainty statement by the manufacturer, the only information available is
the product drawing that shows the active area dimensions to be 10 cm x 10 cm. On the
assumption that the outer bounds of the length, L, and the width, W, are 9.9 and 10.1 cm, the
uncertainty of the linear dimensions may be taken to be a rectangular distribution with a half-
width of 0.1 cm.
L= Wandu(L)= 0.1/-/3 = 0.0577 . W= Wandu(W)= 0.1/>/3 = 0.0577 . Because^ =LW,we
get u2(A) = u2(LW) = L2u2(W) + W2u2(L) = 2(10)2(0.0577)2 = 0.665858, therefore
u(A) = 0.816cm2 or 0.816%.
Plateau Voltage Factor. fv_(Type B)
This applies only to those instruments where voltage adjustments are possible. If the setting is
not checked and/or adjusted between calibrations, then this has no effect. Changing the plateau
voltage without performing a recalibration is not recommended. If, however, the user is allowed
to do this, the setting may not be returned to exactly that used during the calibration. In this
particular example, the slope of the response curve in this region is taken to be 10% / 50 v. It is
NUREG-1575, Supp. 1 7-54 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
assumed that an operator is more likely to set the voltage nearer to the optimum than the
extremes and that ± 50 v represents the range at the 100% confidence level. Accordingly, a
triangular distribution is assumed with a half-width of 50 v, equating to an uncertainty for the
voltage of 507 V6 = 20.4124 and an uncertainty for the voltage factor of 20.4124(10%)/ 50 =
4.0825%.
Source-Detector Separation Factor, /j/Type B)
This effect arises from the uncertainty in mounting the calibration source exactly 3 mm from the
detector face. Experimental evidence has shown that, for the particular 14C source at 3 mm
source-detector separation, the change in response was 2.6% per mm. It is assumed that the
deviation from the nominal 3-mm separation is no greater than 1 mm but that all values are
equally probable between 2 and 4 mm, a rectangular distribution. The uncertainty in the
separation is thusl/v3 = 0.5774 . The uncertainty of the separation factor is thus 0.5774 mm x
2.6% / mm, equal to 1.5011%.
Non-Uniformity of Calibration Source, /»_(Type B)
Large area sources may have a non-uniform activity distribution across their surfaces. For the
14C source, the uniformity is assumed to be better than ± 10%. This is based on comparing 10
cm2 sections of the source. For a typical monitor with a detector area of 50 cm2 and a calibration
source area of 100 cm2, a worst-case condition could be that the area under the detector has an
activity per unit area that is 10% greater than the mean value for the whole source. (The outer
area correspondingly will be 10% less than mean value.) Assuming a rectangular distribution,
this represents an uncertainty of 10/V3 = 5.774% for the source non-uniformity factor.
Backscatter Factor, //^(Type B)
Variations in backscatter effects arise from factors such as the nature of the surface on which the
calibration source is resting and the proximity to scattering surfaces such as walls. This effect
can be quite marked for photon emitters, but for 14C on aluminum substrates the effect is
negligible.
7.8.3.3 Uncertainty Budget
An important part of the uncertainty analysis is to determine which factors are contributing the
most to the overall uncertainty.
To do this, each component of uncertainty ufy)=Ci ut(Xj) is squared to give its component of
variance (ufy)) . These are totaled to obtain the total variance, 69.07. Finally, the ratio of each
component of variance to the total is computed. The relative sensitivity coefficients, ct are the
f
-------
Statistical Basis For MARSAME Surveys
MARSAME
The last column of the uncertainty budget (Table 7.7) shows that the major source of uncertainty
is due to source non-uniformity (48%) followed by the voltage factor (24%) and the reading of
the source (21%). Thus, to decrease the overall uncertainty, attention should be paid to those
factors first.
Table 7.7 Uncertainty Budget for the Efficiency Example
Source of
Uncertainty
Standard deviation of
mean of M
Standard deviation of
mean of B
Standard uncertainty
of calibration source
emission rate, E
Half -width of source
length, L and width
W on the area A
Half -width of voltage
factor, fv
Half -width of source -
detector separation
factor, y}
Half-width of
calibration source
non- uniformity
factor,/^
Uncertainty of
backscatter factor,^
Combined standard
uncertainty
Expanded uncertainty
(k=2)
Type
A
B
B
B
B
B
B
B
Probability
Distribution
Normal
Rectangular
Normal
Product of 2
independent
rectangular
Triangular
Rectangular
Rectangular
n.a.
Normal
Normal
Relative
Sensitivity
Coefflent,ci
1.10
0.098
1.0
1.0
1.0
1.0
1.0
1.0
—
—
Ui(Xj)
(%)
3.45
13.32
0.24
0.816
4.08
1.50
5.77
0.0
—
—
ufy)=
Ci Ui(Xi)
(%)
3.80
1.31
0.24
0.816
4.08
1.50
5.77
0.0
8.31
V69.07
2-8.31=
16.6
(ufy))2
14.44
1.72
0.06
0.666
16.65
2.25
33.29
0.00
Total=
69.07
—
(ufy))2 /Total
0.21
0.02
0.00
0.01
0.24
0.03
0.48
0.00
0.99
—
7.8.3.4 Reported Result
Using the formula above, the calibration factor in terms of emission rate becomes:
= (M-£)x/Fx/,x/Hx/fa=(362.5-32.5)xlxlxlxl =
(2132
NUREG-1575, Supp. 1
7-56
January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
The CSU is (12.1)x(0.0831) = 1.0056. The reported expanded uncertainty will be 2.0, based on a
standard uncertainty of 1.0 multiplied by a coverage factor of k= 2, which provides a level of
confidence of approximately 95%.
7.9 Calculate the Minimum Detectable Concentration
This section is intended to expand on the material in Section 7.5. It contains more statistical
detail and more complex examples. This advanced material may be deferred on a first reading of
MARSAME.
7.9.1 Critical Value
In the terminology of ISO 11843-1 (1997), the measured concentration is the state variable,
denoted by 7, which represents the state of the material being analyzed. The state variable
usually cannot be observed directly, but it is related to an observable response variable, denoted
by X, through a calibration function F, the mathematical relationship being written as X = F(Y).
The response variable Xis most often an instrument signal, such as the number of counts
observed. The inverse, Y = F~l( X) of the calibration function is sometimes called the evaluation
function. The evaluation function, which gives the value of the net concentration in terms of the
response variable, is closely related to the mathematical model Y = f(Xl,X2,...,XN) described
in Section 7.8.1.1.
Either the null or the alternative hypothesis is chosen on the basis of the observed value of the
response variable, X. The value of Xmust exceed a certain threshold value to justify rejection of
the null hypothesis and acceptance of the alternative hypothesis. This threshold is called the
critical value of the response variable and is denoted by xc.
The calculation of xc requires the choice of a significance level for the test. The significance
level is a specified upper bound for the probability, a, of a Type I error. The significance level is
usually chosen to be 0.05. This means that when there is no radiation or radioactivity present
(above background), there should be at most a 5% probability of incorrectly deciding that it is
present.
The critical value of the concentration, yc, is defined as the value obtained by applying the
evaluation function, Kl, to the critical value of the response variable, xc. Thus,_yc= F~l (xc).
When x is the gross instrument signal, this formula typically involves subtraction of the
background signal and division by the counting efficiency, and possibly other factors.
A detection decision can be made by comparing the observed gross instrument signal to its
critical value, xc, as indicated above. However, it has become standard practice to make the
decision by comparing the net instrument signal to its critical value, Sc. The net signal is
calculated from the gross signal by subtracting the estimated blank value and any interference.
The critical value of the net instrument signal, Sc, is calculated from the critical gross signal, xc,
by subtracting the same correction terms; so, in principle, either approach should lead to the
same detection decision.
January 2009 7-57 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Because the term "critical value" alone is ambiguous, one should specify the variable to which
the term refers. For example, one may discuss the critical (value of the) radionuclide
concentration, the critical (value of the) net signal, or the critical (value of the) gross signal. In
this document, the signal is usually a count, and the critical value generally refers to the net
count.
The response variable is typically an instrument signal, whose mean value generally is positive
even when there is radioactivity present (i.e., above background). The gross signal must be
corrected by subtracting an estimate of the signal produced under those conditions. See Section
7.8.2.1 (Instrument Background).
7.9.2 Minimum Detectable Concentration
The minimum detectable concentration (MDC) is the minimum concentration of radiation or
radioactivity that must be present in a sample to give a specified power, 1 —ft. It may also be
defined as:
• The minimum radiation or radioactivity concentration that must be present to give a
specified probability, 1 - /?, of detecting the radiation or radioactivity; or
• The minimum radiation or radioactivity concentration that must be present to give a
specified probability, 1 —ft, of measuring a response greater than the critical value, leading
one to conclude correctly that there is radiation or radioactivity present.
The power of any hypothesis test is defined as the probability that the test will reject the null
hypothesis when it is false, i.e., the correct decision. Therefore, if the probability of a Type II
error is denoted by ft, the power is 1 - ft. In the context of radiation or radioactivity detection, the
power of the test is the probability of correctly detecting the radiation or radioactivity
(concluding that the radiation or radioactivity is present), which happens whenever the response
variable exceeds its critical value. The power depends on the concentration of the radiation or
radioactivity and other conditions of measurement; so, one often speaks of the "power function"
or "power curve." Note that the power of a test for radiation or radioactivity detection generally
is an increasing function of the radiation or radioactivity concentration (i.e., the greater the
radiation or radioactivity concentration, the higher the probability of detecting it).
In the context of MDC calculations, the value of ft that appears in the definition, like a, is usually
chosen to be 0.05 or is assumed to be 0.05 by default if no value is specified. The minimum
detectable concentration is denoted in mathematical expressions by y^. The MDC is usually
obtained from the minimum detectable value of the net instrument signal, SD. SD, is defined as
the mean value of the net signal that gives a specified probability, 1 -ft, of yielding an observed
signal greater than its critical value Sc. The relationship between the critical value of the net
instrument signal, Sc, and the minimum detectable net signal, SD, is shown in Figure 7.6 in
Section 7.5.2.
The term MDC must be carefully and precisely defined to prevent confusion. The MDC is by
definition an estimate of the true concentration of the radiation or radioactivity required to give a
specified high probability that the measured response will be greater than the critical value.
NUREG-1575, Supp. 1 7-58 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
The common practice of comparing a measured concentration to the MDC, instead of to the Sc,
to make a detection decision is incorrect. If this procedure were used, then there would be only a
a 50% chance of deciding that radioactivity was present when the concentration was actually at
the MDC. This is in direct contradiction to the definition of MDC. See MARLAP Appendix B,
Attachment Bl for a further discussion of this issue.
Since the MDC is calculated from measured values of input quantities such as the counting
efficiency and background level, the MDC estimate has a CSU, which in principle can be
obtained by uncertainty propagation. To avoid confusion, it may be useful to remember that a
detection decision is usually made by comparing the instrument response to the critical value,
and that the critical value generally does not even have the units of radiation or radioactivity
concentration.
7.9.3 Calculation of the Critical Value
If the net signal is a count, then in many circumstances the uncertainty in the count can be
estimated by a Type B evaluation using the fact that for a Poisson distribution with mean NB, the
variance is also NB. Thus, the uncertainty in the background count is estimated as ^]Ng and the
critical value is often an expression involving ^NB .
The most commonly used approach for calculating the critical value of the net instrument signal,
Sc, is given by the following equation. u
S -z IN --1+-- (7-49^
°C ~~ L\-a \l B ^ '
Where:
NB = background count
ts = count time for the sample
IB = count time for the background
zi.a = (1 -a)-quantile of the standard normal distribution
Example 19: A 6,000-second background measurement is performed on a proportional counter
and 108 beta counts are observed. A sample is to be counted for 3,000 s. Estimate the critical
value of the net count when a = 0.05.
This expression for the critical net count depends for its validity on the assumption of Poisson counting statistics.
If the variance of the blank signal is affected by interferences, or background instability, then Equation 20.7 of
MARLAP may be more appropriate. Interference is the presence of other radiation or radioactivity or electronic
signals that hinder the ability to analyze for the radiation or radioactivity of interest.
January 2009 7-59 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Sc=1.645jl08xM9«ii + =14.8 ne.count.
V i 6,000 s
If a = 0.05 and tB = ts, Equation 7-49 leads to the well-known expression 2.33^JNB for the
critical net count (Currie, 1968).
When the background count is high (e.g., 100 or more), Equation 7-49 works well, but at lower
background levels it can produce a high rate of Type I errors. Because this is a Scenario B
hypothesis test, this means that too often a decision will be made that there is radiation or
radioactivity present when it actually is not.
When the mean background counts are low and tB ^ t$, another approximation formula for Sc
appears to out-perform all of the other approximations reviewed in MARLAP, namely the
Stapleton approximation:
(7-50)
When a = 0.05, setting the parameter d= 0.4 yields the best results. When, in addition, IB =
the Stapleton approximation gives the equation
Sc =1.35 + 2.33^/^+0.4 (7-51)
7.9.4 Calculation of the Minimum Detectable Value of the Net Instrument Signal
The traditional method for calculating the MDC involves three steps: first calculating critical
value of the net instrument signal, then calculating the minimum detectable value of the net
instrument signal and finally converting the result to a concentration using the mathematical
measurement model.
The minimum detectable value of the net instrument signal, denoted by SD, is defined as the
mean value of the net signal that gives a specified probability, 1 -/?, of yielding an observed
signal greater than its critical value, Sc.
Note: The MDC may be estimated by calculating the minimum detectable value of the net
instrument signal, SD, and converting the result to a concentration.
Counting data rarely, if ever, follow the Poisson model exactly, but the model can be used to
calculate SD if the variance of the background signal is approximately Poisson and a conservative
value of the efficiency constant, s, is used to convert SD ioyo- The equation below shows how to
calculate fusing the Poisson model.
(7-52)
NUREG-1575, Supp. 1 7-60 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
Where:
Sc = critical value
RB = mean count rate of the blank, RB = ——
NB = background count
ts = count time for the test source
IB = count time for the background
z\.p = (1 -/?)-quantile of the standard normal distribution
When Equation 7-49 is appropriate for the critical net count, and a = /?, this expression for SD
simplifies to zf_p + 2SC. If in addition, a = ft = 0.05 and tB = ts then
SD=2.7\ + 2SC = 2.71 + 2(2.33 JN^) = 2.71 + 4.66 JW^ (7-53)
Example 20: A 6,000-s background measurement on a proportional counter produces 108 beta
counts and a source is to be counted for 3,000 s. Assume the background measurement gives the
available estimate of the true mean background count rate, RB and use the value 0.05 for Type I
and Type II error probabilities. From Section 7.9.3, Example 19, the critical net count, Sc, equals
14.8, so SD = zlfl + 2Sc =1.6452+2(14.8) = 32.3 net counts.
When the Stapleton approximation (Equation 7-51) is used for Sc, the minimum detectable net
count SD may be calculated using the Equation 7-53, but when the Poisson model is assumed, a
better estimate is given by the equation:
( * ^ I 7 ~
(7-54)
This equation is the same as that recommended by ISO 11929-1 (ISO 2000) in a slightly
different form.
When a = ft = 0.05 and fe = ts, the preceding equation becomes:
SD = 5.41 + 4.65^JRBts (7-55)
Consult MARLAP Chapter 20 for a discussion of the calculation of SD and_yD when both Poisson
counting statistics and other sources of variance are considered.
7.9.5 Calculation of the Minimum Detectable Concentration
The MDC is often used to compare different measurement procedures against specified
requirements. The calculation of the nominal MDC is complicated by the fact that some input
quantities in the mathematical model, such as interferences, counting efficiency, and instrument
background may vary significantly from measurement to measurement. Because of these
variable quantities, determining the value of the radiation or radioactivity concentration that
January 2009 7-61 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
corresponds to the minimum detectable value of the net instrument signal, SD, may be difficult in
practice. One common approach to this problem is to make conservative choices for the values of
the variable quantities, which tend to increase the value of the MDC.
The mean net signal, S, is usually directly proportional to 7, the true radiation or radioactivity
concentration present. Hence, there is a efficiency constant, s, such that S = sY. The constant s is
typically the mean value of the product of factors such as the source count time, decay-correction
factor, and counting efficiency. Therefore, the value of the minimum detectable concentration,
.yo, is
yD=^ (7-56)
s
The preceding equation is only true if all sources of variability are accounted for when
determining the distribution of the net signal, S . Note that ensuring the MDC is not
underestimated also requires that the value of snot be overestimated.
Using any of the equations in Section 7.5.2 to calculate SD is only appropriate if a conservative
value of the efficiency constant, s, is used when converting SD to the MDC.
Example 21: Consider a scenario where t& = 6,000 s, ^s = 3,000 s, and RB -0.018 s"1. Let the
measurement model be Y = •
Ns-(NBtsltB)
tss
Where:
Y = activity of the radionuclide in the sample and
s = counting efficiency (counts per second)/(Bq/cm2)
Assume the source count time, t$, has negligible variability, the counting efficiency has mean
0.42 and a 10% relative CSU, and from Example 20, SD = 32.3 net counts.
S 32 3
The mean minimum detectable concentration is yn = — 2- = - : - = o 0256 Bq/cm2
D tss (3000)(0.42)
Adjusting for the 10% variability in the counting efficiency, the uncertainty is (0.10)x(0.42) =
0.042. Assuming that the efficiency is normally distributed, the lower 5th percentile for e is
(0.42)-(1.645)(0.042) = 0.35, where -1.645 is the 5th percentile of a standard normal
distribution. Therefore, a conservative estimate of the efficiency constant is s = 0.35 and a
conservative estimate of the minimum detectable concentration is:
32.3 B 2
tss (3000)(0.35)
An alternative procedure could be to recognize that because of the uncertainties in the input
estimates entered into the measurement model to convert from SD to 7, that the MDC is actually
a random variable. Then the methods for propagation of uncertainty given in Section 7.8 can be
applied. Using the same assumptions as above, we would find that>t> = 0.0256 ± 0.0051 with
95% confidence based on a coverage factor of 2. Therefore the 95% upper confidence level for
yD would be 0.0307 Bq.
NUREG-1575, Supp. 1 7-62 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
More conservative (higher) estimates of the MDC may be obtained by following NRC
recommendations (NRC 1984), in which formulas for the MDC include estimated bounds for
relative systematic error in the background determination ( 4^ ) and the sensitivity ( ^ )• The
critical net count Sc is increased by 4^B NB (ts/ts), and the minimum detectable net count <$D is
increased by 2 $>B NB (ts/ts)- Next, the MDC is calculated by dividing SD by the efficiency and
multiplying the result by 1+4^4 • The conservative approach presented in NRC 1984 treats
random errors and systematic errors differently to ensure that the MDC for a measurement
process is unlikely to be consistently underestimated, which is an important consideration if it is
required by regulation or contract to achieve a specified MDC.
7.10 Calculate the Minimum Quantifiable Concentration
This section is intended to expand on the material in Section 7.6. It contains more statistical
detail and more complex examples. This advanced material may be deferred on a first reading of
MARSAME.
Calculation of the MQC requires that one be able to estimate the standard deviation for the result
of a hypothetical measurement performed on a sample with a specified radionuclide
concentration. The MQC is defined symbolically as the value .yethat satisfies the relation:
= yQ) (7-57)
Where the specified relative standard deviation of _yeis l/kQ (usually chosen to be 10% so that
kQ =10). a2(y | Y = yQ) is the variance of the estimator y given the true concentration Y equals
ye. If the function
-------
Statistical Basis For MARSAME Surveys
MARS AME
yo =
(7-59)
Where:
fc
IB
RB
RI
count time for the source, s
count time for the background, s
mean background count rate, s"1
mean interference count rate, s"1
2~ = relative variance of the measured efficiency, s
If the efficiency emay vary, then a conservative value, such as the 0.05-quantile SQ.OS, should be
substituted fore in the formula. Note that $ denotes only the relative variance of s due to
subsampling and measurement error; it does not include any variance of the efficiency s itself
(see discussion in Section 7.8).
Note that Equation 7-59 defines the MQC only if 1 - k*$ > 0. If 1 - k^ < 0, the MQC is
infinite, because there is no concentration at which the relative standard deviation of y fails to
exceed 1 / kQ . In particular, if the relative standard deviation of the measured efficiency s
exceeds 1 / kQ , then 1 - k^l < 0 and the MQC is infinite.
If there are no interferences, Equation 7-59 simplifies to:
(7-60)
Example 22: Consider the scenario of Example 21, where fe = 6,000 s, ^s = 3,000 s, and
"1
RB ~ 0.018 s". Suppose the measurement model is 7 =
^ B s B'
tss
Where:
Y = specific activity of the radionuclide in the sample
s = counting efficiency (cps/Bq)/(Bq/cm2)
Assume the source count time, t$, has negligible variability, the counting efficiency has a mean
of 0.42 and a 5% relative CSU, and SD = 32.3 net counts.
So/ts =32.3/3000 is the net count rate and the counting efficiency, 8, is 0.42.
S 32 3
The mean minimum detectable concentration is yn = —^- = : = 0.0256 Bq/cm2.
-' U . //-* r\r\r\\ /r\ A r\\ l
tss (3000)(0.42)
Also assume:
=10
NUREG-1575, Supp. 1
7-64
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
4 = 0.05
42 = 0.052
l-k2Q(f>l =1-100 x(0.052) = 0.75
There are no interferences so that Equation 7-60 can be used.
Note that if the counting efficiency had a mean of 0.42 and a 10% relative standard uncertainty
as in Example 11, then \-k2Q2 = l-100x (0.102) = 0 and the MQC would be infinite. Therefore it
was necessary to change the procedure for evaluating the efficiency in this example so that the
relative CSU could be reduced. In this example it is assumed to be 5%.
The MQC can be calculated as:
100
2 (3000X0.42)(0.75)
= 0.151Bq/cm2
1+. 1 +
4(0.75)1
100 I
(0.018 s^XSOOO s) 1 +
(3000 s)
(6000 s)
+ 0
As a check, yq can be calculated in a different way. lfyQ is the MQC and kQ = 10, then the
relative CSU of a measurement of concentration^ is 10%. The procedure described in Section
7.4 can be used to predict the CSU of a measurement made on a hypothetical sample whose
concentration is exactly^ = 0.151 Bq/cm2.
The measurement model is Y =
Ns-(NBts/tB)
tss
f(x x x ^)
Recall from Section 7.8.1.6 that if^ = " 2'"'' " ; where/is some specified function of
Z\Z1 •••Zm
x\,X2,.. .,*„, all the zt are nonzero, and all the input estimates are uncorrelated that the CSU may
be calculated using Equation 7-35:
Substituting
zl = s, and
ul(Ns-(NBtsltB)lts) = ul(
_ Ns+NB(t2sltl)
January 2009
7-65
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
Results in:
tss
„„ \N,+(NBt2Jt2B)
"c(y)=J g \\8 B
Inserting the values
Y= yQ=0.l5lEq/cm2
tB = 6,000 s
ts = 3,000 s
s = 0.42 (counts per second)/(Bq/cm2)
NB =RBtB = (0.018 s^)(3,000s) = 108 and
Ns =xQtss + RBtB = (0.151 Bq)(3000 s)(0.42) +(0.018 s^)(3,000 s) = 244.26
yields
24426+(108)(3.000)'/(6.000f 5
cV ' V (3000)2(0.42)2 V ' } 4
Thus, the uncertainty at_yg= 0.151 is 0.0151 and the relative uncertainty is 0.1, so_ygis verified
to be the MQC.
As above in this example, we adjust for the (now) 5% relative CSU in the counting efficiency.
The uncertainty is (0.05) x (0.42) = 0.02142. Assuming that the efficiency is normally
distributed, the lower 5th percentile is (0.42) - (1.645)(0.021) = 0.385. Therefore a conservative
estimate of the efficiency is s = 0.385 and a conservative estimate of the minimum detectable
concentration is: yn = — ——- = 0.165 Bq/cm2.
8 0.385
7.11 Calculate Scan MDCs
The methodology used to determine the scan MDC is based on NUREG-1507 (NRC 1998b).
This procedure is quite complex as it requires, among other skills, a familiarity with radiation
transport calculations for its implementation. The information developed here will be used in the
example in Section 8.2, "Mineral Processing Facility Concrete Rubble." However, the details
given in this section are not essential to understanding the example.
The radionuclides of concern are the members of the natural uranium and thorium series. The
instrument used is a "Field Instrument for the Detection of Low Energy Radiation" (FIDLER).
The approach used would be similar for other instruments and radionuclides.
The approach to determine scan MDCs includes:
NUREG-1575, Supp. 1 7-66 January 2009
-------
MARSAME Statistical Basis For MARSAME Surveys
• Calculate the fluence rate relative to the exposure rate (FRER) for the range of energies of
interest (Section 7.11.1).
• Calculate the probability of interaction (P) between the radiation of interest and the detector
(Section 7.11.2).
• Calculate the relative detector response (RDR) for each of the energies of interest (Section
7.11.3).
• Determine the relationship between the detector's net count rate to net exposure rate in
cpm/nR/h, Section 7.11.4).
• Determine the relationship between the detector response and the radionuclide concentration
(Section 7.11.5).
• Obtain the minimum detectable count rate (MDCR) for the ideal observer, for a given level
of performance, by postulating detector background and a scan rate or observation interval
(Section 7.11.6).
• Relate the MDCR for the ideal observer to a radionuclide concentration (in Bq/kg) to
calculate the scan MDC (Section 7.11.7).
7.11.1 Calculate the Relative Fluence Rate to Exposure Rate (FRER)
For particular gamma energies, the relationship of Nal scintillation detector count rate
and exposure rate may be determined analytically (in cpm/|jR/h). The approach is to
determine the gamma fluence rate necessary to yield a fixed exposure rate (|j,R/h) as a
function of gamma energy. The fluence rate, following NUREG-1507 (NRC 1998b), is
directly proportional to the exposure rate and inversely proportional to the incident
photon energy and mass energy absorption coefficient:
Fluence tfate(FRER) oc X- - (7-61)
Er (Vm/ P\ir
Where:
X = exposure rate (set equal to 1 |j,R/hr for these calculations)
E7 = energy of the gamma photon of concern (keV)
(//en/p)air = mass energy absorption coefficient in air at the gamma photon energy of
concern (cm2/g)
The mass energy absorption coefficients in air are presented in Table 7.8 (natural uranium) and
Table 7.9 (natural thorium) along with the calculated fluence rates (up to a constant of
proportionality, since only the ratios of these values are used in subsequent calculations). Note
that while the mass energy absorption coefficients in air, (jJen/p)nir, are tabulated values (NIST
1996), the selected energies are determined by the calculation of the detector response based on
radionuclide concentration (Section 7.11.5).
7.11.2 Calculate the Probability of Interaction
Assuming that the primary gamma interaction producing the detector response occurs through
the end of the detector (i.e., through the beryllium window of the detector, as opposed to the
sides), the probability of interaction (P) for a gamma may be calculated using Equation 7-52:
January 2009 7-67 NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
P =
Where:
P
(////?)NaI
X
P
probability of interaction (unitless)
mass attenuation coefficient of FIDLERNal crystal at the energy of interest
(e.g., 0.117 cm2/g at 400 keV)
thickness of the thin edge of the FIDLER Nal crystal (0.16 cm)
density of the Nal crystal (3.67 g/cm3)
The mass attenuation coefficients for the Nal crystal and the calculated probabilities for each of
the energies of interest are presented in Table 7.8 (natural uranium) and Table 7.9 (natural
thorium). The mass attenuation coefficients for Nal were calculated using the XCOM program
(NIST 1998).
Table 7.8 Calculation of Detector Response to Natural Uranium
Energy
(keV)
15
20
30
40
50
60
80
100
150
200
300
400
500
600
662
800
1,000
1,500
2,000
(Wen//>)air
(cm2/g)
1.334
0.5389
0.1537
0.06833
0.04098
0.03041
0.02407
0.02325
0.02496
0.02672
0.02872
0.02949
0.02966
0.02953
0.02931
0.02882
0.02789
0.02547
0.02345
FRER
(Section
7.11.1)
0.04998
0.09278
0.2169
0.3659
0.4880
0.5481
0.5193
0.4301
0.2671
0.1871
0.1161
0.08477
0.06743
0.05644
0.05154
0.04337
0.03586
0.02617
0.02132
(W//>)NaI
cm2/g
47.4
21.8
7.36
18.8
10.5
6.45
3.00
1.67
0.611
0.328
0.166
0.117
0.0950
0.0822
0.0766
0.0675
0.0588
0.0470
0.0415
P
(Section
7.11.2)
1.000
1.000
0.9867
1.000
0.9979
0.9773
0.8282
0.6249
0.3015
0.1752
0.09288
0.06640
0.05426
0.04712
0.04398
0.03886
0.03394
0.02722
0.02407
RDR
(Section
7.11.3)
0.04998
0.09278
0.2140
0.3659
0.4870
0.5356
0.4301
0.2688
0.08052
0.03278
0.01078
0.005629
0.003659
0.002660
0.002267
0.001685
0.001217
0.0007125
0.0005133
cpm per
jiR/h
(Section
7.11.4)
28,374
52,678
121,498
207,725
276,511
304,123
244,204
152,606
45,717
18,613
6,120
3,196
2,077
1,510
1,287
957
691
405
291
NUREG-1575, Supp. 1
7-68
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
Table 7.9 Calculation of Detector Response for Natural Thorium
Energy
(keV)
40
60
80
100
150
200
300
400
500
600
662
800
1,000
1,500
2,000
3,000
("en//>)air
(cm2/g)
0.06833
0.03041
0.02407
0.02325
0.02496
0.02672
0.02872
0.02949
0.02966
0.02953
0.02931
0.02882
0.02789
0.02547
0.02343
0.02057
FRER
(Section
7.11.1)
0.3659
0.5481
0.5193
0.4301
0.2671
0.1871
0.1161
0.08477
0.06743
0.05644
0.05154
0.04337
0.03586
0.02617
0.02134
0.01620
(w/pW
cm2/g
18.8
6.45
3.00
1.67
0.611
0.328
0.166
0.117
0.0950
0.0822
0.0766
0.0675
0.0588
0.0470
0.0415
0.0368
P
(Section
7.11.2)
1.000
0.9773
0.8282
0.6249
0.3015
0.1752
0.09288
0.06640
0.05426
0.04712
0.04398
0.03886
0.03394
0.02722
0.02407
0.02138
RDR
(Section
7.11.3)
0.3659
0.5356
0.4301
0.2688
0.08052
0.03278
0.01078
0.005629
0.003659
0.002660
0.002267
0.001685
0.001217
0.0007125
0.0005137
0.0003464
cpm per
jiR/h
(Section
7.11.4)
207,725
304,123
244,204
152,606
45,717
18,613
6,120
3,196
2,077
1,510
1,287
957
691
405
292
197
7.11.3 Calculate the Relative Detector Response
The relative detector response (RDR) for each of the energies of interest is determined by
multiplying the FRER by P. The results are presented in Table 7.8 (natural uranium) and Table
7.9 (natural thorium).
7.11.4 Relationship Between Detector Response and Exposure Rate
Using the same methodology described in Sections 7.11.1 through 7.11.3, FRER, P, and RDR
are calculated at the 137Cs energy of 662 keV and are also presented in Table 7.8 and Table 7.9.
The manufacturer of the FIDLER Nal detector provides an estimated response of the crystal in a
known radiation field, which is 1,287 cpm per |jR/h at the 137Cs energy of 662 keV. The
response at 662 keV can be used to determine the response at all other energies of interest using
Equation 7-63:
cpm _[ 1,287 cpm
uR/hj
RDR137
(7-63)
Cs
Where:
= energy of the photon of interest (keV)
January 2009
7-69
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys MARSAME
= response of the detector for energies of interest, Table 7.8 and Table 7.9
uR/hE_
RDRE = RDR at the energy of interest, Table 7.8 and Table 7.9
RDR137cs = RDR for 137Cs, Table 7.8 and Table 7.9
The responses in cpm per |jR/h for each of the decay energies of interest are presented in Tables
7.8 and 7.9.
7.11.5 Relationship Between Detector Response and Radionuclide Concentration
The minimum detectable exposure rate is used to determine the MDC by modeling a specific
impacted area. The relationship between the detector response (in cpm) and the radionuclide
concentration (in Bq/kg) uses a computer gamma dose modeling code to model the presence of a
normalized 1 Bq/kg total activity source term for natural uranium and natural thorium. The
following assumptions from NUREG-1507 (NRC 1998b) were used to generate the computer
gamma dose modeling runs:
• Impacted media is concrete,
• Density of concrete is 2.3 g/cm3,
• Activity is uniformly distributed into a layer of crushed concrete 15 cm thick,
• Measurement points are 10 cm above the concrete surface,
• Areas of elevated activity are circular with an area of 0.25 m2 and a radius of 28 cm,
• 0.051 cm beryllium shield simulates the window of the FIDLER detector, and
• Normalized 1 Bq/kg source term decayed for 50 years to allow ingrowth of decay progeny.
The weighted cpm per |j,R/h response (weighted instrument sensitivity [WSj]) for each decay
energy is calculated by multiplying the |jR/h at 1 Bq/kg (exposure rate with buildup, R,) by the
cpm per |j,R/h and dividing by the total |j,R/h (at 1 Bq/kg) for all decay energies of interest
(Equation 7-64):
^x(cpmper//R/h)
RT
Where:
WSt = weighted instrument sensitivity (cpm per |jR/h)
Ri = exposure rate with buildup (|jR/h)
RT = Total exposure rate with buildup (|jR/h)
Calculate the percent of FIDLER response for each of the decay energies of interest by dividing
WSt by the total weighted cpm per |j,R/h and multiplying by 100 percent (Equation 7-62):
Percent of FIDLER response = — • - (7-65)
WT
Where:
WT = Total WSt weighted instrument sensitivity (cpm per |jR/h)
NUREG-1575, Supp. 1 7-70 January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
The exposure rates for each of the decay energies of interest are presented in Table 7.10
(assuming natural uranium for the source term) and Table 7.11 (assuming natural thorium for the
source term).
Table 7.10 Detector Response to Natural Uranium
Energy
keV
15
20
30
40
50
60
80
100
150
200
300
400
500
600
800
1,000
1,500
2,000
Total
Ri
(tiRlh)
(Section 7.11.5)
4.473xlO~10
3.597xlO~12
2.623xlO~07
1.299xlO~10
1.052xlO~07
5.065xlO~06
1.518xlO~06
2.309xlO~05
5.138xlO~06
2.881xlO~05
2.237xlO~07
2.434xlO~07
4.208xlO~07
2.048xlO~06
1.478xlO~05
5.759xlO~05
1.695xlO~06
2.841xlO~07
1.413xlO~04
cpm per uR/h
(Section 7.11.4)
28,374
52,678
121,498
207,725
276,511
304,123
244,204
152,606
45,717
18,613
6,120
3,196
2,077
1,510
957
691
405
291
WSi
(cpm per ull/h)
(Section 7.11.5)
0
0
226
0
206
10903
2625
24938
1663
3796
10
6
6
22
100
282
5
1
44,923
Percent of
FIDLER
Response
(Section 7.11.5)
0.00%
0.00%
0.504%
0.00%
0.460%
24.3%
5.86%
55.7%
3.71%
8.48%
0.0216%
0.0123%
0.0138%
0.0489%
0.224%
0.629%
0.0108%
0.00131%
100%
Table 7.11 Detector Response to Natural Thorium
Energy
keV
40
60
80
100
150
200
300
Ri
(uR/h)
(Section 7.11.5)
1.299xlO~06
1.816xlO~06
1.989xlO~04
5.027xlO~05
5.862xlO~05
1.135xlO~03
8.922xlO~04
cpm per uR/h
(Section 7.11.4)
207,725
304,123
244,204
152,606
45,717
18,613
6,120
WSi
(cpm per uR/h)
(Section 7.11.5)
10
21
1855
293
102
807
209
Percent of
FIDLER
Response
(Section 7.11.5)
0.266%
0.544%
47.8%
7.55%
2.64%
20.8%
5.37%
January 2009
7-71
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
Table 7.11 Detector Response to Natural Thorium (Continued)
Energy
keV
400
500
600
800
1,000
1,500
2,000
3,000
Total
Ri
(tiRlh)
(Section 7.11.5)
1.105xlO-°4
8.146xlO~04
2.218xlO~03
2.892xlO~03
6.443xlO~03
2.062xlO~03
5.822xlO~05
9.249xlO~03
2.619xlO~02
cpm per itRJh
(Section 7.11.4)
3,196
2,077
1,510
957
691
405
292
197
WSi
(cpm per ull/h)
(Section 7.11.5)
13
65
128
106
170
32
1
69
3881
Percent of
FIDLER
Response
(Section 7.11.5)
0.348%
1.67%
3.30%
2.72%
4.38%
0.821%
0.0167%
1.79%
100%
7.11.6 Calculation of Scan Minimum Detectable Count Rates
In the computer gamma dose modeling, an impacted area with a radius of 28 cm or
approximately 0.25 m was assumed. Using a scan speed of 0.25 m/s provides an observation
interval of one second.
A typical background exposure rate is 10 //R/h. Using a conversion factor based upon field
measurements of 1,287 cpm per //R/h for 137Cs (see 7.11.4) results in an estimated background
count rate of 12,870 cpm. Converting this value from cpm to counts per second (cps) using
Equation 7-66 results in a background of 214.5 cps.
,, . Imin 1,287 cpm 1 min
o(cpm) x x /(sec) = — x 10 //R/h x x 1 sec = 214.5 cps (7-66)
60 sec 1 //R/h 60 sec
Where:
b = background count rate (12,870 cpm)
/' = observation interval length (1 s)
The MDCR is calculated using the methodology in NUREG-1507 (NRC 1998b) shown in
Equations 7-67 and 7-68:
s. =d'Jbt= 1.38x^214.5 =20.21 counts
(7-67)
d'Jb, 1.38x^214.5
, surveyor = ~^ = - 7= - = 28.58 COUntS
V0.5
MDCR = *,.x (60//) = 20.21 x (60/1) = 1,212 cpm
(7-68)
NUREG-1575, Supp. 1
7-72
January 2009
-------
MARSAME
Statistical Basis For MARSAME Surveys
surveyor i, surveyor
;(60//) = 28.58x(60/l) = 1,715 cpm
Where:
bi
i
P
d'
Si
Si, si,
'irveyor
MDCR
MDCR
•surveyor
average number of counts in the background interval (214.5 counts)
observation interval length (one second)
efficiency of a less than ideal surveyor, range of 0.5 to 0.75 from
NUREG-1507 (NRC 1998b); a value 0.5 was chosen as a conservative
value
detectability index from Table 6.1 of NUREG-1507 (NRC 1998b); a
value of 1.38 was selected, which represents a true positive detection
rate of 95% and a false positive detection rate of 60%12
minimum detectable number of net source counts in the observation
interval (counts)
minimum detectable number of net source counts in the observation
interval by a less than ideal surveyor
minimum detectable count rate (cpm)
MDCR by a less than ideal surveyor (cpm)
7.11.7 Calculate the Scan Minimum Detectable Concentration
The scan minimum detectable concentration (MDC) can be calculated from the minimum
detectable exposure rate (MDER). The MDER can be calculated using the previously calculated
total weighted instrument sensitivities (WS,), in cpm per |jR/h, for natural uranium and natural
thorium as shown in Equations 7-69 and 7-70:
MDER
MDCR,
(7-69)
Scan MDC = C x
MDER
(7-70)
Where:
MDER
MDCR
WT
RT
C
= MDER for the "ith" source term, by a less than ideal surveyor, (|j,R/h)
•surveyor = MDCR rate by a less than ideal surveyor (cpm), from Section 7.11.6
= Total weighted instrument sensitivity (cpm per |j,R/h, Table 7.10 and
Table 7.11)
= Total exposure rate with buildup (|jR/h, Table 7.10 and Table 7.11)
= concentration of source term (set at 1 Bq/kg in Section 7.11.5)
A Type I error, misidentifying a background area as elevated will have the consequence that a longer reading will
be needed to verify the initial decision. This will happen with probability a. A Type II error, missing a true elevated
area, may lead to incorrectly exceeding the limit for the chosen disposition option. This will happen with probability
P. Since in this instance the consequences of a Type I error are often considered much lower than the consequences
associated with a Type II error. Thus, a may be set higher than /?. Setting both very low could result in slow
scanning speeds and operator fatigue.
January 2009
7-73
NUREG-1575, Supp. 1
-------
Statistical Basis For MARSAME Surveys
MARSAME
Scan MDC = minimum detectable concentration (Bq/kg)
The Scan MDCs for the FIDLER were calculated using Equations 7-69 and 7-70 and the
instrument response information from Table 7.10 (assuming natural uranium as the source term)
and Table 7.11 (assuming natural thorium as the source term). The scan MDCs for natural
uranium and natural thorium using a FIDLER are listed in Table 7.12.
Table 7.12 Scan MDCs for FIDLER
Source
Term
Natural
Uranium
Natural
Thorium
MVCRsurveror
(cpm)
Section 7.11.6
1,715
1,715
WT
(cpm per
uR/h)
Section
7.11.5
44,786
3,881
MDER
(uR/h)
Section
7.11.7
0.03829
0.4419
RT
(uR/h)
Section
7.11.5
1.413xlO"04
2.619xlO"02
C
(Bq/kg)
Section
7.11.5
1
1
Scan MDC
(Bq/kg)
Section
7.11.7
271 « 300
16.9 « 20
The scan MDCs of approximately 300 Bq/kg for uranium and 20 Bq/kg for thorium are both less
than their respective action levels of 38,000 and 330 Bq/kg, respectively.
NUREG-1575, Supp. 1
7-74
January 2009
-------
MARSAME Illustrative Examples
8 ILLUSTRATIVE EXAMPLES
8.1 Introduction
This chapter presents illustrative examples providing examples of applications of the information
in the Multi-Agency Radiation Survey and Assessment of Materials and Equipment manual
(MARSAME) supplement to the Multi-Agency Radiation Survey and Site Investigation Manual
(MARSSIM). The purpose of these illustrative examples is to illustrate applications of the
information in conditions that are frequently encountered and cover a broad range of situations.
The general format for each illustrative example mirrors as closely as possible the information
presented in MARSAME. References to information, tables, figures, and equations from Chapter
2 through Chapter 6 are provided throughout the illustrative examples.
MARSAME contains both procedural as well as informative sections. The illustrative examples
provide a practical use of the MARSAME process and, as such, generally apply only the
procedural sections. In addition, much of the information in MARSAME is designed to be
applied iteratively. In some illustrative examples, the information is applied in a different
sequence than it is presented in MARSAME because of this iterative nature.
Section 8.2 provides an example of a disposition survey for a large quantity of bulk material at a
mineral processing facility. This example establishes gross activity action levels based on
normalized effective dose equivalents. These action levels are applied with multiple decision
rules using a MARSSIM-type survey design to collect scan survey data as well as systematic and
judgmental samples for laboratory analysis.
Section 8.3 and Section 8.4 are based on the same mineral processing facility that serves as the
basis for Section 8.2. Section 8.3 provides an example of an interdiction survey for rented heavy
equipment that is designed to establish a "baseline" estimate of the residual radioactivity
associated with a front loader before it is brought into a radiological control area (RCA) for the
impacted bulk material. This baseline survey establishes zero net activity as the lower bound of
the gray region (LBGR) and applies MARSAME processes to a Scenario B survey design.
Section 8.4 demonstrates the clearance of the same rented front loader that was brought on to the
site in Section 8.3. Section 8.4 describes a Scenario A clearance survey based on the same
surface activity action levels to clear the front loader. Sections that contain redundant
information are presented in Section 8.3 only and are omitted from Section 8.4.
8.2 Mineral Processing Facility Concrete Rubble
This illustrative example is provided for information purposes only and presents a theoretical
application of MARSAME guidance. This illustrative example discusses the process of
designing and implementing a MARSSIM-type disposition survey design for a large quantity of
bulk material at a mineral processing facility. This example includes discussions on most of the
guidance provided in MARSAME, including establishing gross activity action levels based on
normalized effective dose equivalents. Calculations of uncertainties associated with scanning
measurements are included. The MARSSIM-type survey design includes scanning, systematic
January 2009 8-1 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
samples, and judgmental samples to support a disposition decision. The text is provided to
illustrate the application of MARSAME guidance, and should not be considered an example
survey plan. The amount of discussion provided in this example is based on the complexity of
the problem and the relative difficulty expected from applying or interpreting specific portions of
MARSAME guidance. The amount of discussion for this example is not related to, and should
not be used as an estimate of, the level of effort associated with planning, implementing, or
assessing an actual disposition survey.
8.2.1 Description
An abandoned mineral processing facility is being redeveloped for commercial/industrial use.
The facility processed mineral ores for various metals for over 30 years and was abandoned more
than 10 years ago. The processing equipment and existing stockpiles of ore were transferred to
another facility when site renovations began. The receiving facility discovered radioactivity
levels in excess of background on exterior portions of processing equipment using hand-held
Geiger-Mueller (GM) "pancake" detectors.
Prior to discovery of the radioactivity on the processing equipment, the concrete floors had been
removed from the processing buildings and stockpiled on-site. Note that if the buildings were
still intact, they could be surveyed using a MARSSIM survey. An investigation is performed to
trace the source of the radioactivity to the appropriate portion(s) of the mineral processing
facility.
8.2.2 Objectives
The objective is to make an appropriate disposition decision regarding the concrete rubble from
the impacted portions of the mineral processing facility. It is anticipated that leaks of potentially
radioactive processing liquids could have occurred throughout the operating lifetime of the
facility. Airborne radioactive concrete dust may have been released during demolition activities,
which could have exposed construction personnel and contacted components of the demolition
equipment.
8.2.3 Initial Assessment of the M&E
8.2.3.1 Categorize the M&E as Impacted or Non-Impacted
As part of the initial assessment (IA), it is necessary to determine whether the concrete rubble is
impacted or not. A visual inspection of the concrete rubble was performed. Historical records
from the facility concerning sources of ore, ore processing techniques, waste disposal practices,
industrial accidents, as well as building and equipment repairs, modifications, and upgrades were
reviewed. Interviews with key facility personnel were also performed. In addition, research into
mineral processing techniques and radionuclide content of raw ores was performed to obtain
additional process knowledge.
Process knowledge indicated the facility processed ilmenite ore (iron titanium oxide,
and produced titanium dioxide. A sentinel measurement of a small amount of ilmenite ore
NUREG-1575, Supp. 1 8-2 January 2009
-------
MARSAME Illustrative Examples
remaining at the site was analyzed by alpha spectrometry and found to contain elevated levels of
natural uranium and thorium. Additional measurements performed on the radioactive processing
equipment reported concentrations of uranium and thorium greater than expected from
background.
Site history indicates that the general layout of the process was unchanged over the lifetime of
the facility, and it is likely that spills occurred repeatedly in discrete locations. Processing liquids
and slurries were considered hazardous because of their low pH; radioactivity was not
considered an issue. Limited information regarding site history and operations was obtained
through interviews with former employees and review of historical documentation. Former
employees stated that spills and leaks of process liquids and slurries occurred periodically in
several areas of the processing plant; these represent the only potential source of radioactivity in
the plant. Fluid spills were quickly corrected by neutralizing the acid to protect employees and
equipment. Spills frequently resulted from seal failure within the various pumps in use at the
processing operation.
Results from the visual inspection indicated there was a reasonable potential for radioactivity
from plant activities to be associated with the concrete rubble. Several chunks of concrete rubble
are obviously discolored from plant operations, indicating possible locations of spills. The
facility floor consisted of reinforced concrete on a gravel base mat. Portions of the rubble contain
possible evidence of staining. The rubble still contains rebar which, for operational reasons, must
be segregated and treated as a separate waste stream.
The concrete rubble is considered to be impacted due to the discovery of residual radioactivity
on exterior portions of the processing equipment, historical records that acidic process fluids may
have spilled on the concrete floor, and process knowledge that the acidic process fluids were
mixed with raw ore containing elevated levels of naturally occurring radioactive material
(NORM) from the uranium and thorium radioactive decay series. The results of the sentinel
measurement performed on the raw ore support the categorization as impacted.
8.2.3.2 Describe the M&E
Table 8.1 lists the physical attributes of the concrete rubble. No data gaps associated with the
physical attributes were identified.
Table 8.2 lists the known radiological attributes associated with the concrete rubble, as well as
data gaps showing where additional information is required to design a disposition survey. As
presented, the existing information is not adequate to design a disposition survey. Preliminary
surveys were designed and implemented to address the data gaps identified in Table 8.2. The
results of the preliminary surveys were used to modify the conceptual site model by filling some
of the data gaps.
January 2009 8-3 NUREG-1575, Supp. 1
-------
Illustrative Examples
MARSAME
Table 8.1 Physical Attributes of the Concrete Rubble
Attribute
Dimensions
Complexity
Accessibility
Inherent Value
Description
Total Mass
400 ft x 100 ft x 1 ft « 40,000 ft3
40,000 ft3 x 0.0283 m3/ft3 « 1,132 m3
The approximate density of crushed concrete is 2.3 x 1Q6 g/m3
1,132 m3 x 2.3 x IQ6 g/m3 = 2.60 x 109 g = 2.60 x 106 kg
Shape
The concrete has been broken into chunks less than one meter in the largest dimension.
The concrete is stored in three piles. Each pile is approximately 1.5 m high, 6 m wide,
and 40 m long.
Rebar used to reinforce the floor is present in the concrete rubble. The rebar will be
segregated and removed, and treated as a separate waste stream.
The concrete rubble may require further reduction in size to ensure measurability.
The concrete represents inherent value for several potential disposition options. Crushed
concrete serves many useful purposes, including recyclable use as roadbed material.
This option presents potential cost savings over using virgin materials in place of
recycled concrete and a reuse scenario that avoids the relatively high cost for disposal.
The radionuclides of potential concern are the uranium (238U) and thorium (232Th) natural
radioactive decay series. Based on process knowledge, radionuclide concentrations in the raw ore
average between 750 and 1,100 Bq/kg for members of the uranium series, and between 200 and
400 Bq/kg for members of the thorium series. Following processing, some 238U and 232Th decay
products may not have been in equilibrium with the parents. The amount of time since the plant
ceased operations (i.e., 10 years) indicates there is a potential for the thorium series radionuclides
to have re-established secular equilibrium. Preliminary survey measurements are required to
determine the equilibrium status of the uranium and thorium series radionuclides.
8.2.3.3 Design and Implement Preliminary Surveys
Limited scanning of concrete rubble was performed using a GM detector. The purpose of the
scanning was to determine how the radioactivity associated with the concrete was distributed.
The scanning survey also included additional visual inspection of the concrete.
Intermittent staining within the concrete rubble and scanning surfaces of concrete chunks
demonstrates that the radioactivity was heterogeneously deposited on the processing building
floor. Higher levels of radioactivity were found in areas where spills occurred historically (i.e.,
discolored concrete). The staining did not appear to have penetrated more than one-quarter inch
into the concrete when the floor was intact. Prior to demolition, the presence of cracks and other
structural irregularities in the concrete floor provided preferential pathways for activity to
penetrate to greater depths. This resulted in some variance in activity with depth of the original
concrete floor.
NUREG-1575, Supp. 1
January 2009
-------
MARSAME
Illustrative Examples
Table 8.2 Radiological Attributes of the Concrete Rubble
Attribute
Radionuclides
Activity
Distribution
Location
Description
Uranium Series
Radionuclides
238U
234Th
234mPa
234Pa
234U
230Th
226Ra
222Rn
218Po
214Pb
214Bi
214Po
210Pb
210Bi
210p0
Thorium Series
Radionuclides
232Th
228Ra
228Ac
228Th
224Ra
220Rn
216p0
212pb
212Bi
212Po (64%)
208T1 (36%)
Principal
Emission
Particle
Alpha
Beta
Beta/Gamma
Beta
Alpha
Alpha
Alpha/Gamma
Alpha
Alpha
Beta/Gamma
Beta/Gamma
Alpha
Beta
Beta
Alpha
Principal
Emission
Particle
Alpha
Beta
Beta/Gamma
Alpha
Alpha
Alpha
Alpha
Beta/Gamma
Alpha/Beta
Alpha
Beta
Emission
Energy
(MeV)
4.20
0.1886
2.28/1.001
0.224
4.77
4.688
4.78/0.186
5.49
6.00
0.67/0.352
1.54/0.609
7.687
0.016
1.161
5.305
Emission
Energy
(MeV)
4.01
0.0389
1.17/0.911
5.42
5.686
6.288
6.78
0.334/0.238
6.05/2.246
8.785
1.80
Activity levels range from background
(approximately 40 Bq/kg) to 4,000 Bq/kg from
isolated portions of the concrete rubble where
spills occurred.
The radioactivity is heterogeneously
distributed throughout the mass of concrete
rubble.
The concrete rubble is considered a
volumetrically impacted mass. The residual
radioactivity that is present is a combination of
fixed and removable.
Data Gaps
The radioactivity is likely to have come
in contact with the M&E through spills
of process fluids and dumping of solid
tailings on the concrete floor.
Equilibrium status of the decay series is
unknown, although sufficient time has
elapsed since site closure for the
thorium series to have re-established
secular equilibrium.
The expected range of activity is an
estimate. Nature and extent of activity
needs to be investigated to provide
better estimates of average and
maximum activity. Better estimates of
background are needed.
No data gaps were identified. The
current distribution is not a concern
because the concrete will be crushed to
2-3 cm size prior to survey.
The distribution of radioactivity with
depth may provide useful information
for selecting measurement methods
because it can impact the total
measurement efficiency.
January 2009
8-5
NUREG-1575, Supp. 1
-------
Illustrative Examples
MARSAME
Samples were collected from the crushed concrete from the processing mill floor to determine
concentrations of residual radioactivity using alpha spectrometry and gamma spectroscopy.
Concrete samples were collected from four biased locations, including two areas of elevated
gross activity within the concrete rubble with GM readings as high as 250 cpm and visible
staining (Samples 1 and 2), and two samples with readings consistent with the average readings
observed during scanning (40 to 45 cpm) (Samples 3 and 4). Process knowledge and limited
historical site information indicates that radiological materials were never used or stored within
the on-site administrative building. Reference Samples 1 and 2 were collected from the concrete
floor of the onsite administrative building to provide information on background activities in
non-impacted concrete for the uranium and thorium decay series for the conceptual model. The
six samples were sent to a radioanalytical laboratory for analysis, and the results of the analyses
are provided in Tables 8.3 through 8.6.
Table 8.3 Preliminary Alpha Spectrometry Results for Uranium Series Radionuclides
Sample ID
Sample 1
Sample 2
Sample 3
Sample 4
Reference
Sample 1
Reference
Sample 2
234U
7,000
7,200
21
25
19
13
csu1
±2,100
± 2,300
±7.4
±8.1
±5.2
±3.7
MDC2
1,900
1,900
3.7
3.7
3.7
3.7
235U
340
320
0.74
0.74
0.37
0.37
CSU1
± 1,900
± 1,700
±1.9
±3.0
±0.74
±0.74
MDC2
1,600
1,600
0.74
0.74
0.74
0.74
238U
7,600
7,000
21
21
20
11
CSU1
± 2,400
±2,100
±7.0
±7.0
±5.6
±3.3
MDC2
1,900
1,900
3.7
3.7
3.7
3.7
All units in Bq/kg
1 CSU is the combined standard uncertainty of the measurement result reported by the analytical laboratory.
2 MDC is the minimum detectable concentration reported by the analytical laboratory.
Table 8.4 Preliminary Alpha Spectrometry Results for Thorium Series
Radionuclides
Sample ID
Sample 1
Sample 2
Sample 3
Sample 4
Reference Sample 1
Reference Sample 2
232Th
1,400
1,200
21
26
21
23
CSU1
±110
±130
± 1.5
± 1.1
± 1.1
± 1.1
MDC2
10
10
.1
.1
.1
.1
228Th
1,300
1,500
23
24
22
23
CSU1
±150
±190
± 1.5
± 1.1
± 1.1
± 1.1
MDC2
110
110
1.1
1.1
1.1
1.1
All units in Bq/kg
1 CSU is the combined standard uncertainty of the measurement result reported by the
analytical laboratory.
2 MDC is the minimum detectable concentration reported by the analytical laboratory.
NUREG-1575, Supp. 1
January 2009
-------
MARSAME
Illustrative Examples
Table 8.5 Preliminary Gamma Spectroscopy Results for Uranium Series Radionuclides
Sample
ID
Sample 1
Sample 2
Sample 3
Sample 4
Reference
Sample 1
Reference
Sample 2
214Bi
93
740
21
22
17
20
csu1
±920
± 1,000
± 1.1
± 1.1
± 1.1
± 1.1
MDC2
1,400
1,300
3.6
4.1
3.1
3.4
214pb
530
1,000
21
23
17
20
CSU1
±780
±870
± 1.1
± 1.1
± 1.1
± 1.1
MDC2
1,300
1,200
6.3
7.0
7.0
5.6
226Ra
47
192
64
68
36
52
CSU1
± 1,100
± 1,200
±9.6
±8.5
±6.3
±7.1
MDC2
1,500
1,400
16
19
18
17
All units in Bq/kg
1 CSU is the combined standard uncertainty of the measurement result reported by the analytical laboratory.
2 MDC is the minimum detectable concentration reported by the analytical laboratory.
Table 8.6 Preliminary Gamma Spectroscopy Results
for Thorium Series Radionuclides
Sample ID
Sample 1
Sample 2
Sample 3
Sample 4
Reference
Sample 1
Reference
Sample 2
228Ac
1,600
1,400
14
21
15
16
CSU1
± 180
± 130
±2.6
±3.1
±3.3
±3.4
MDC2
52
41
4.4
6.3
5.9
3.4
All units in Bq/kg
1 CSU is the combined standard uncertainty of the measurement
result reported by the analytical laboratory.
2 MDC is the minimum detectable concentration reported by the
analytical laboratory.
Note the results provided in Tables 8.3 through 8.6 are from actual samples collected from a real
site. However, the sample results included as part of this illustrative example were selected to
provide specific information supporting the application of MARSAME guidance and represent a
portion of the total amount of information available. The number and type of samples collected
as part of preliminary survey should be determined using the DQO Process as discussed in
Section 2.3.
8.2.3.4 Select a Disposition Option
The preferred disposition of the concrete rubble is clearance. It is expected that the concrete will
be reused as roadbed or disposed of in a municipal landfill. If the activity levels exceed the
project action levels, then the concrete may need to be disposed of as discrete naturally occurring
or accelerator-produced (NARM) waste. If the activity is below the alternate action levels, the
concrete may either be reused or disposed of as diffuse NARM waste.
January 2009
8-7
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.2.3.5 Document the Results of the Initial Assessment
The results of the IA were documented in a letter report. The purpose of the letter report was to
document the categorization decision and all supporting information. The letter report was
reviewed and finalized by the facility owner. Detailed results of the IA will be included in the
final documentation of the survey design.
8.2.4 Develop a Decision Rule
Following completion of the IA, additional information was needed to develop the disposition
survey design.
8.2.4.1 Select Radionuclides or Radiations of Concern
The list of radionuclides of concern was finalized based on the preliminary survey results.
Uranium-23 8, 234U, and 226Ra are the radionuclides of concern for the uranium natural decay
series. The alpha spectrometry results indicate that 238U and 234U are in equilibrium (i.e., have
equal concentrations). Because alpha spectrometry for uranium isotopes provides results for both
238U and 234U, both isotopes (and their decay products with half-lives less than six months) will
be kept as radionuclides of concern. There is no indication of enrichment or depletion of uranium
as a result of site activities based on the uranium alpha spectrometry results listed in Table 8.3.
Radium-226 decay products, including 210Pb, are assumed to be out of secular equilibrium with
the other uranium series radionuclides (e.g., 238U and 234U) because process knowledge shows
the chemical processing at the plant would separate uranium from radium. Bismuth-214 and
214Pb can be used as beta or gamma emission surrogates for 226Ra, because the decay products of
226Ra should be in secular equilibrium with one another. However, a 21-day ingrowth period may
be required to confirm this assumption. The planning team determined an ingrowth study was
not required for this project following discussions with the regulators.
Thorium-232 is the radionuclide of concern for the thorium natural decay series. Based on the
alpha spectrometry and gamma spectroscopy results shown in Table 8.3, all members of the
thorium natural decay series are in secular equilibrium. Actinium-228 emits gamma rays that are
easy to quantify using gamma spectroscopy, and can be used as a surrogate for the members of
the thorium series.
8.2.4.2 Identify Action Levels
For the purposes of this illustrative example, an action level of 0.01 mSv/y was selected based on
discussions with the planning team. Using information provided in NUREG-1640 (NRC 2003),
the action levels were converted into concentration units based on clearance as the disposition
option. Incorporating the concrete rubble into roadbed material would provide the highest
potential doses following clearance. The mean values from NUREG-1640 (NRC 2003),
Table 11.13 ("Normalized effective dose equivalents from all pathways: Driving on road [uSv/y
per Bq/g]"), are the basis for the action levels.
NUREG-1575, Supp. 1 8-8 January 2009
-------
MARSAME
Illustrative Examples
Radionuclide of concern
Mass-based EDE mean values
(uSv/y per Bq/g)
238U
0.26
234U
8.2 x 1Q~4
30
226Ra
22
The action levels from Table 11.13, NUREG-1640 (NRC 2003) are expressed in units of uSv/y
per Bq/g, but the preliminary survey measurement results are in Bq/kg. To make a direct
comparison, the action levels were converted to units of Bq/kg. Note that a hypothetical dose for
clearance was selected only for the purpose of showing example calculations for this illustrative
example. Clearance criteria should be provided by the regulator for actual applications of this
guidance. The action levels were converted to concentrations by inverting the action levels and
multiplying by the hypothetical dose limit (i.e., the inverted action levels in units of Bq/g per
uSv/y are multiplied by 0.01 mSv/y, 1,000 g/kg, and 1,000 uSv/mSv providing action levels in
Bq/kg). Table 8.7 lists the action levels in concentration units of Bq/kg.
Table 8.7 Radionuclide-Specific Action Levels
Radionuclide
238U
234U
232Th
226Ra
Mass-Based EDE Mean Values
(Bq/g per uSv/y)
"° v 0 01 rn^v/vYl vl O6 ^9
0.26uSv/y
IrJq/g mqWvYl Ylfl6 1°
8.2xlQ-4uSv/y
1Bq/S xOOlmSv/vxlxlO6
S.OxlO1 uSv/y
"° Y 0 01 mSWvvl vl O6 —
2.2xl01uSv/y
000
000,000
330
450
Action Level
(Bq/kg)
38,000
12,000,000
330
450
The unity rule (Equation 8-1) is used to account for the individual radionuclide action levels. The
unity rule is satisfied when the summed analyses of each radionuclide against its respective
action level yields a value less than one:
The Unity Rule =
C C
AL
AL
(8-1)
Where:
C = concentration of each individual radionuclide (1,2, ... ri)
AL = action level value for each individual radionuclide (1,2, ... ri)
Equation 8-1 is used to calculate the sum of fractions for each of the preliminary survey results:
(- 234TT
C238
The Unity Rule = u
zTh
°Ra
<1
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
Sample 1= ' | 7,OOOBq/kg | l,400Bq/kg | 47 Bq/kg
38,OOOBq/kg 12,000,000 Bq/kg 330 Bq/kg 450 Bq/kg
Sample 2 = 6' 9°° Bq/kg I ?' 2°° Bq/g I *' 23° Bq/kg I 192 Bq/kg - 1 2
38, 000 Bq/kg 12,000,000 Bq/kg 330 Bq/kg 450 Bq/g
Sample3= + + 21 Bq/kg + 64 Bq/kg ^
38, 000 Bq/kg 12,000,000 Bq/kg 330 Bq/kg 450 Bq/g
Sample 4= 21**** + 2S B"/k« + 26 Bq/kg + 68 Bq/kg ^ p 2,
38, 000 Bq/kg 12,000,000 Bq/kg 330 Bq/kg 450 Bq/g
The results of the calculations for Samples 1 and 2 exceed a sum of fractions of 1.0, and indicate
the presence of small volumes of concrete with elevated activity. Note that the reported MDCs
for gamma spectroscopy for 226Ra in Samples 1 and 2 would not meet the MQOs for clearance
(i.e., the MDC exceeds the action level). However, the radionuclide concentrations in these two
samples clearly exceed the action level. Therefore, the quality of these results is acceptable to
support the disposition survey design.
The results of the calculations for Samples 3 and 4 indicate that, on average, the concrete rubble
is expected to have radionuclide concentrations below the action levels. Therefore, the average
activity in the concrete rubble is expected to be below the action level. Large blocks containing
elevated levels of radioactivity may be visually identified via staining, verified with a GM
detector, and segregated prior to removal of the rebar.
8.2.4.3 Modify the Action Levels to Account for Multiple Radionuclides
Radionuclide-specific action levels need to be combined into a single gross gamma action level
for evaluating the field instrument for detection of low-energy radiation (FIDLER) scan
measurements. The information in Section 3.3.3.1 requires an estimate of the relative fraction of
the total activity contributed by each radionuclide. A consistent relationship between 238U and
232Th concentrations is not expected based on the IA, because different ore bodies could contain
different ratios of these radionuclides. Rather than develop a preliminary survey attempting to
develop this relationship, a conservative approach was adopted for this project.
Assuming the entire radioactivity detected by the FIDLER results from the presence of the most
restrictive radionuclide will provide the most conservative gross gamma action level. The ratios
of exposure rate to radionuclide concentration (uR/h per Bq/kg) and instrument response to
exposure rate (cpm per uR/h) were developed in Section 7.11 during development of the scan
MDC for both 238U and 232Th. These ratios can be used to calculate the count rate above
background associated with a radionuclide activity equal to the action level as shown in
Equation 8-2.
NUREG-1575, Supp. 1 8-10 January 2009
-------
MARSAME
Illustrative Examples
GGAL = AL x
Bq/kg.
cPm
—-—
(8-2)
Where:
GOAL
AL
gross gamma action level (cpm)
action level value for each individual radionuclide (Bq/kg)
Equation 8-2 was used to calculate a gross gamma count rate above background for the FIDLER
assuming each radionuclide of concern was present at a concentration equal to the action level.
The gross gamma count rates were divided by two to account for uncertainty associated with the
detector efficiency calculation and added to the background count rate. The result is a gross
gamma action level for the FIDLER to identify locations with unexpectedly high gamma activity
that could result in doses near the action level of 0.01 mSv/y. The results of the calculations are
shown in Table 8.8. The 232Th gross gamma action level of 30,000 cpm is more conservative
than the 238U gross gamma action level of 140,000 cpm, so 30,000 cpm was selected as the gross
gamma action level.
Table 8.8 Calculation of the Gross Gamma Action Level
Action Level
(Bq/kg)
238U
38,000
232Th
330
jiR/h per
Bq/kg
1.413xlO~4
2.619xlO~2
cpm per
jiR/h
45,593
3,923
Gross
Gamma
Count Rate
(cpm)
244,807
33,905
Adjusted Gross
Gamma Count
Rate (cpm)
122,404
16,953
Background
Count Rate
(cpm)
12,870
12,870
Gross Gamma
Action Level
(cpm)
140,000
30,000
FIDLER readings that exceed the 232Th gross gamma action level indicate locations where
radionuclide concentrations could result in doses exceeding the 0.01 mSv/y used for this
ooo^^
illustrative example if all of the activity results from Th.
232.
232n
Because Th has decay products in secular equilibrium that can be used to estimate the Th
activity, gamma spectroscopy can be used to quantify 232Th concentrations. FIDLER readings
that exceed 140,000 cpm identify locations where radionuclide concentrations could result in
doses exceeding 0.01 mSv/y if all of the activity results from 238U. Alpha spectrometry is
required to quantify 238U concentrations.
8.2.4.4 Describe the Parameter of Interest
Because the disposition option is stated in terms of dose, the parameter of interest is the mean
radionuclide concentration. The target population is all of the possible measurement results that
could be obtained within a survey unit. This means the target population will be defined by the
survey unit boundaries (Section 8.2.4.6) and the selected measurement method (Section 8.2.4.8).
January 2009
8-11
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.2.4.5 Identify Alternative Actions
The alternative actions identify the results of decisions based on the measurement results. If the
radionuclide concentrations do not result in a dose that exceeds the action level, the material is
cleared. If the dose exceeds the action level, materials exceeding the action level will be
segregated and investigated for disposal as NARM waste.
8.2.4.6 Identify Survey Units
Survey unit boundaries are based primarily on the modeling assumptions used to develop the
action levels. The volume of concrete used to model exposures for building a road is 83 m3
(NUREG-1640 [NRC 2003] Volume 2, Appendix B, Tables B-8 and B-l 1). Each survey unit
will consist of approximately 80 m3 of crushed concrete (approximately 25 m x 22 m x 0.15 m).
The volume of concrete poured to create the floor of the processing mill was approximately
1,100 m3. Crushing the concrete and removing the rebar is expected to result in approximately a
25% increase in volume due to air gaps, for a total volume of 1,400 m3 of crushed concrete.
Using these calculations, there will therefore be a total of 18 survey units plus one reference area.
The concrete rubble can be spread into a relatively uniform layer approximately 15 cm thick and
scanned. This adapts an approach used in MARSSIM to survey the top 15 cm of surface soil as a
two-dimensional object.
8.2.4.7 Define the Decision Rules
MARSSEVI-type surveys are designed to evaluate the average radionuclide concentration in a
survey unit using samples or direct measurements, as well as small areas of elevated activity
using scans. Small areas of elevated activity receive additional investigation. Because there are
multiple action levels and multiple decisions to be made, there are multiple decision rules for the
disposition survey. The first two decision rules address how small areas of elevated activity are
identified by scans and what investigations will be performed. The third decision rule evaluates
the results of the investigations of small areas of elevated activity. The fourth decision rule
evaluates the average activity in each survey unit.
1. If any FIDLER scanning measurement result exceeds the gross gamma action level of 30,000
cpm (see Section 8.2.5.4), a biased sample will be collected for laboratory analysis by
gamma spectroscopy, otherwise no biased samples will be collected.
2. If any FIDLER scanning measurement exceeds 140,000 cpm, the biased sample collected for
gamma spectroscopy analysis will also be analyzed by alpha spectrometry for uranium and
thorium isotopes, otherwise the concrete will be held awaiting the results of the gamma
spectroscopy analysis.
3. If the results from a biased sample result in a sum of fractions for 238U, 234U, 226Ra, and 232Th
exceeding 1.0, the concrete will be segregated and investigated for disposal as NARM waste.
Otherwise, the survey unit will be evaluated based on the WRS test results for the samples
taken over a systematic grid.
NUREG-1575, Supp. 1 8-12 January 2009
-------
MARSAME Illustrative Examples
4. If the mean sum of fractions in a survey unit exceeds 1.0, the concrete will be segregated and
investigated for disposal as NARM waste. Otherwise, the WRS test will be performed to
support the final disposition decision for that survey unit.
8.2.4.8 Develop Inputs for Selection of Provisional Measurement Methods
The selected measurement method will be required, at a minimum, to detect radionuclide
concentrations at or below the action levels in Table 8.8 (page 8-11). The survey planners
considered each of the possible measurement techniques (Section 5.9.1).
Scan-only techniques have the ability to detect surface activity at concentrations below the action
levels. In situ measurement techniques are also expected to have the ability to measure
radionuclide concentrations at the action levels. However, uncertainties associated with the
efficiency for both techniques will be large. In order to reduce these uncertainties to a level
where the radionuclide concentrations are measurable, the concrete would need to be pulverized
and mixed rather than just crushed to 2-3 cm size. Because the cost of processing the concrete
this way would be a major cost associated with the disposition survey, a MARSSIM-type survey
design was selected for the disposition survey. No method-based survey designs were identified
that matched the description of the M&E, so no method-based survey designs were considered.
Concrete samples will be analyzed in a laboratory using alpha spectrometry for uranium isotopes
91zl 0^8 01A
(i.e., U and U) as well as gamma spectroscopy for other radionuclides of concern (i.e., Bi,
214Pb, and 228Ac). Sample sizes must be sufficient to allow quantification of radionuclide
concentrations at the action levels. By convention, the MQC for each radionuclide of concern is
selected so the measurement method uncertainty at concentrations equal to the action levels in
Table 8.7 is 10%. Alternatively, the samples can be sealed in airtight containers for at least
twenty-one days to allow secular equilibrium to be reestablished prior to analysis by gamma
spectroscopy so decay products can be used as surrogate radionuclides.
Due to the rough, irregular shape of the concrete rubble, alpha radiation is attenuated easily and
is difficult to measure. Beta and gamma measurements typically provide a more accurate
assessment of thorium and uranium activity on most building surfaces because surface conditions
cause significantly less attenuation of beta and gamma particles than alpha particles. For this
reason, scanning will be performed using instruments that detect beta or gamma radiation.
Surface scans, using a 12.7-cm by 0.16-cm FIDLER sodium iodide (NaI[Tl]) scintillation
detector, are used to scan for gamma emissions. The approximate detection sensitivity of the
FIDLER is 300 Bq/kg for natural uranium and 20 Bq/kg for natural thorium when activity is
present at the surface. The FIDLER is a large detector and can detect gammas from a greater
height above the crushed concrete than alpha or beta detection equipment, making it a more
practical choice for surveying large volumes of material. The selection of the FIDLER over more
conventional Nal(Tl) detectors (e.g., a three-inch by three-inch gamma scintillation detector) is
primarily based on the FIDLER's ability to detect low-energy gamma radiation, which comprises
the majority of the gamma radiation from the radionuclides of concern.
January 2009 8-13 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.2.4.9 Identify Reference Materials
Concrete from the administrative building contains non-impacted materials, as established by the
process knowledge discussed in Section 8.2.3.1. The reference material measurements will be
performed on the floor in the administrative building. The geometry of the floor is similar
enough to the concrete rubble (after crushing to 2-3 cm size and arrangement into a 15-cm thick
layer) that modifications to the building are not required.
8.2.5 Develop a Survey Design
The concrete rubble from the mineral processing facility is surveyed for clearance using a
MARSSEVI-type disposition survey. The survey includes scanning to identify small areas of
elevated activity combined with collection and analysis of samples to evaluate the average
activity in the concrete rubble.
Scenario A is used to design the survey, because decisions will be made based on average
radionuclide concentrations and radioactivity levels in each survey unit. The null hypothesis is
that the radionuclide concentrations in the concrete rubble will result in a dose that exceeds
0.01 mSv/y. There are two decisions for MARSSEVI-type surveys. The first decision is based on
the average radionuclide concentrations in the survey unit, and the second decision is based on
the scanning survey results and subsequent biased sample results from flagged locations. The
same null hypothesis applies to both decisions.
A Type I decision error would occur if the decision-maker decided the activity levels in the
concrete rubble were below the action level when they actually exceeded the action level. The
consequence of making this decision error could result in increased doses to members of the
public and failing to identify small areas of elevated radionuclide concentrations. The members
of the planning team agreed to a Type I decision error rate of 5% based on the consequence of
making this decision error. This Type I error rate applies to both the scanning portion of the
survey design as well as sampling on a systematic grid.
A Type II decision error would occur if the decision-maker decided the activity levels in the
concrete rubble exceeded the action level when they were actually below the action level. The
consequence of making this decision error could result in increased disposal costs. The members
of the planning team agreed to a Type II decision error rate of 10% based on the consequence of
making this decision error for sampling. However, during scanning the consequence of making
this decision error is the need to perform additional investigation., As such, a Type II decision
error rate of 40% is selected for the scanning surveys.
8.2.5.1 Classify the M&E
All of the concrete rubble from the floor of the processing facility has the potential to exceed one
or more of the action levels. The concrete rubble is classified as Class 1 M&E.
NUREG-1575, Supp. 1 8-14 January 2009
-------
MARSAME Illustrative Examples
8.2.5.2 Design the Scanning Survey
The concrete must be crushed prior to performing the scanning survey to reduce the size of
individual particles to less than 2-3 cm in diameter. This provides a uniform matrix of material
ensuring a representative sample can be collected, and also allows the rebar to be removed. The
crushed concrete is distributed in a layer approximately 15 cm thick, and surveyed using a
FIDLER at a height of 10 cm above the surface. The scan speed is 0.25 m/s, which is consistent
with the scan MDC calculations. One hundred percent of the concrete rubble is scanned with
readings in excess of 30,000 cpm flagged for additional investigation. The additional
investigations include collection and analysis of samples using gamma spectroscopy to quantify
activity levels for the radionuclides of concern. Samples collected from locations with readings
in excess of 140,000 cpm are also analyzed for uranium and thorium isotopes by alpha
spectrometry.
8.2.5.3 Design the Sample Collection Survey
The concrete rubble is divided into survey units and a statistically based number of samples are
collected from each survey unit. Because multiple radionuclides are present, the unity rule is
used to evaluate the sample results. Because the radionuclides are present in background, the
Wilcoxon Rank Sum (WRS) test is used to evaluate the survey results.
The upper bound of the gray region (UBGR) is set equal to the action level, which is a sum of
fractions of 1.0 above background. The lower bound of the gray region (LBGR) is set equal to
the expected sum of fractions based on results from the preliminary survey. The expected
average activity in the concrete rubble is close to background, even though isolated areas have
results more than four times the action level. An LBGR value of 0.15 is selected, which is
consistent with results reported in Tables 8.3 through 8.6 for the two randomly selected samples
(i.e., samples 3 and 4). Because the values are not corrected for background, this value is
considered conservative. The shift (UBGR - LBGR) is 0.85.
The variability in the activity levels for the concrete rubble, as, is not well defined. To be
conservative, the variability in the results should be large for results near the LBGR. A value of
0.15 was selected for the variability. This value is equal to the LBGR, and represents 100%
variability in results that are at or near background. The relative shift equals 5.6 (0.85 divided by
0.15 and rounded down). Because relative shifts greater than 4.0 do not result in significantly
smaller numbers of samples, a relative shift of 4.0 was used to determine the number of samples
and also help to ensure adequate statistical power.
Table A.2b (Appendix A) lists the number of samples required for each survey unit and reference
area for use with the WRS test. Seven samples are required for each survey unit and reference
area using a relative shift of 4.0, Type I decision error rate of five percent, and Type II decision
error rate of 10 percent. The radionuclide or radioactivity concentrations derived from the dose-
based action level are based on an average radionuclide concentration or level of radioactivity
over the entire survey unit. No adjustments need to be made to the number of measurements to
01£
account for the scan MDC, because the scan MDC is less than the action level for both U and
232Th.
January 2009 8-15 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
Seven samples of approximately 1,000 g of concrete rubble are collected from each survey unit.
This mass corresponds to a cylinder with a diameter of approximately 6 cm (2.5 in) to a depth of
15 cm (6 in). This disposition survey design will be applied to all of the concrete rubble,
including the concrete segregated based on visual inspection and elevated scanning results with a
GM detector during the preliminary surveys (Section 8.2.3.3).
8.2.5.4 Develop an Operational Decision Rule
The action level is stated in terms of incremental dose above background. In a MARSSIM
survey, there are requirements for both sample measurements and scanning results. Samples will
be collected from non-impacted concrete to represent background radionuclide concentrations.
The WRS test will be used to evaluate the survey results. If the test statistic for the WRS test is
less than or equal to 65 (n = m = 7, a = 0.05), decide that the dose from that survey unit exceeds
0.01 mSv/y and the concrete will not be cleared.
For the scanning results, if any FIDLER measurement exceeds 30,000 cpm, collect a biased
concrete sample at the location of the elevated measurement for analysis by gamma
spectroscopy. If any FIDLER measurement exceeds 140,000 cpm, analyze the biased concrete
sample by alpha spectrometry as well. If the sum of fractions for any biased sample exceeds 1.0,
decide that the dose from that survey unit exceeds the 0.01 mSv/y used for this illustrative
example and the concrete will not be cleared.
8.2.5.5 Document the Survey Design
The final survey design was documented in a detailed work plan. The work plan provided the
results of the IA, as well as all of the assumptions used to develop the survey design. The DQOs
and MQOs for the survey design were also included.
The draft work plan was submitted to the planning team for review. Comments were received,
and responses to comments developed and approved. The approved responses to comments were
incorporated into a final work plan documenting the disposition survey design.
8.2.6 Implement the Survey Design
8.2.6.1 Ensure Protection of Health and Safety
A job safety analysis (ISA) was performed based on the tasks defined in the work plan
documenting the disposition survey design. Table 8.9 shows the results of the ISA. Potential
health and safety hazards identified by the ISA are addressed in a site-specific health and safety
plan. No hazards associated with the concrete rubble will notably affect how the disposition
survey is implemented.
NUREG-1575, Supp. 1 8-16 January 2009
-------
MARSAME
Illustrative Examples
Table 8.9 Job Safety Analysis for Surveying Concrete Rubble
Sequence of Basic Job Steps | Potential Hazards
1. Dividing rubble into
manageable survey units
2. Establish exclusion zone for
survey area
3. Use hand-held survey
instruments to perform survey
measurements on the crushed
concrete
4. Physical handling of larger
pieces of concrete debris to
expose underside for gamma
surveying
5. Entering Exclusion Zone (EZ)
to perform survey
6.Moving contaminated or clean
material to appropriate disposal
containers
Use of front end loader
by untrained personnel
Personnel in area could
be struck by heavy
equipment
Exposure to silica
Lower back strain
from lifting
Exposure to
radiological
contamination
None anticipated
Unstable footing may
result in slips, trips, or
falls
Rough surfaces may
cut and scrape skin on
hands
Tripping
Exposure to
radiological
contamination
Spread of radiological
contamination outside
EZ
se of front end loader
f untrained personnel
ower back strain
from lifting
Exposure to
radiological
contamination
Exposure to silica
Recommended Action or Procedure
Ensure equipment operators are adequately trained
Area workers must maintain eye contact with
equipment operators
Reflective vests will be worn to improve visibility
Use of a real-time dust monitor will document dust
levels. Respiratory protection will be used if dust
levels exceed established action levels (dependent on
silica content of concrete)
Proper lifting techniques will be used
Loads will be sized so as not to create unreasonable
weights for manual lifting
PPE including booties, Tyveks, and gloves will be
used
Spread out rubble in a way to minimize tripping
hazards by creating clear rows between rows of
concrete
Wear a set of work gloves to protect hands when
handling concrete pieces
Maintain good housekeeping in survey area
PPE including booties, Tyveks, and gloves will be
used
Establish step-off area outside of EZ
Ensure equipment operators are adequately trained
Proper lifting techniques will be used. Loads will be
sized so as not to create unreasonable weights for
manual lifting
PPE including booties, Tyveks, and gloves will be
used
Use of a real-time dust monitor will document dust
levels. Respiratory protection will be used if dust
levels exceed established action levels (dependent on
silica content of concrete)
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.2.6.2 Consider Issues for Handling the M&E
The concrete rubble must be crushed to a uniform size of less than one inch to implement the
disposition survey design and meet the MQOs. The crushing process will generate dust
potentially containing radioactive material. Controls to limit dust generation were implemented
during concrete crushing activities. Equipment involved in handling the concrete during crushing
activities (e.g., front loader, crusher, rebar separator, conveyor belts, dump trucks) is categorized
as impacted and will require a disposition survey before the equipment can be released. Surveys
of the front loader are discussed in Sections 8.3 and 8.4.
8.2.6.3 Segregate the M&E
Concrete rubble with visible stains and pitting on the floor surface is segregated as having higher
activity concentrations. Stained and unstained concrete were grouped into separate survey units.
Following segregation, the concrete was crushed to 2-3-cm diameter pieces, and the rebar was
removed.
8.2.6.4 Set Measurement Quality Objectives
The two most important MQOs for this survey design are the required measurement method
uncertainty, UMR, for the scan MDCs for the FIDLER measurements (Section 8.2.6.5) and the
concrete samples collected on the systematic grid (Section 8.2.6.6). Other MQOs were
established during the development of the survey design to support selection of measurement
methods. These included setting the MQC for each radionuclide of concern so the relative
measurement method uncertainty, <^, at concentrations equal to the action levels in Table 8.7 is
10% (Section 8.2.4.8) and calculating the scan MDCs for the FIDLER (Section 7.11).
8.2.6.5 Determine Measurement Uncertainty for the Scan MDC
This section describes the calculation of the uncertainty for the scan MDC measurements
performed as part of this survey using the FIDLER. An upper bound for an expanded uncertainty
for the scan MDC calculation is derived to reduce the probability that the scan MDC has been
underestimated. The result is used as the investigation level for evaluating the results of the scan
survey.
The uncertainty calculations presented in this section may be performed using commercially
available statistical software (Section 5.6). Detailed solutions for this illustrative example are
provided below.
The scan MDCs for the FIDLER measurements, y, are calculated in Section 7.11. The scan MDC
for natural uranium, yu, is approximately 400 Bq/kg. The scan MDC for natural thorium, yTh, is
approximately 25 Bq/kg. Both scan MDCs are less than their respective action levels of 38,000
and 330 Bq/kg. The values used to calculate the scan MDCs for the FIDLER measurements are:
NUREG-1575, Supp. 1 8-18 January 2009
-------
MARSAME Illustrative Examples
bj = average number of counts in the background interval (214.5 counts). Here bt is
assumed to have a triangular distribution with a half-width of 30% or 64 counts,
so the mean value of b, is 215 and u(bi) = 64/ v6 = 26 .
/' = observation interval length (one second). Here /' is assumed to have a triangular
distribution with a half-width of 0.5, so the mean value of/' =1.0 and
«(/) = 0.5/V6 = 0.2.
p = efficiency of a less than ideal surveyor, range of 0.5 to 0.75 from NUREG-1507
(NRC 1998b); a value 0.5 was chosen as a conservative value. Here/? is assumed
to have a rectangular distribution with a half-width of 0.125, so the mean value of
p = 0.625 and u(p) = 0.125/^3 = 0.072.
d' = detectability index from Table 6.1 of NUREG-1507 (NRC 1998b); a value of
1.90 was selected and treated as a constant.
WT = total weighted instrument sensitivity (cpm per (iR/h)
= 44,923 for natural uranium from Table 7.10 and
= 3,881 for natural thorium from Table 7.11.
RT = total exposure rate with buildup (|j,R/h)
= 1.413 x 10"4 for natural uranium from Table 7.10 and
= 2.619x 10"2 for natural thorium from Table 7.11.
C = concentration of source term (set at 1 Bq/kg and treated as a constant).
y = Scan MDC (in Bq/kg) introduced here for simplicity of notation.
Because we are assuming there are no correlations among the input variables, the combined
standard uncertainty ofy can be calculated using Equation 7.33 from Section 7.8.1.6:
N
i uy i 7, -^ •^-^ ^ ,
;. u
i=\
rv
2 f \ v"* uy 2
uc (y) = 2. T^ \u
a*, J" -' ^
The concentration of the source term, C, and the detectability index, d\ are treated as constants
with no associated uncertainty, so this expands to:
The sensitivity coefficients, cf, are calculated as follows:
January 2009 8-19 NUREG-1575, Supp. 1
-------
Illustrative Examples
MARSAME
d\
dp
I y^GOCd^j yy
\ /I) JW 7? n3'2 \ /2/ n
a
y
di
dR,
iWTR^Jp RT
y
dWT
Therefore,
The most notable sources of uncertainty associated with WT and RT are the modeling assumptions
for the source-to-detector separation distance during scanning and the depth distribution of the
radioactivity in the crushed concrete. To calculate uncertainties, the same basic modeling
assumptions as those for the MDC calculations were applied, though with variations to both the
source-to-detector separation distance during scanning and the distribution of the radioactivity in
the crushed concrete. While the MDC calculation assumes a source-to-detector distance of 10 cm
and that the activity is uniformly distributed within a cylindrical volume of crushed concrete 15
cm thick with a radius of 28 cm, several other calculations were made using source-to-detector
separation distances during scanning of 8, 10, and 12 cm, and by varying the distribution of the
radioactivity in the crushed concrete from uniform to uniformly distributed within both the top
and bottom 7.5 cm of the cylindrical volume of crushed concrete, to assess the potential
variability in the MDC. In each calculation the total activity was the same, only the distribution
with depth was changed. The extreme cases were for a source-to-detector distance of 8 cm with
NUREG-1575, Supp. 1
January 2009
-------
MARSAME
Illustrative Examples
the activity uniformly distributed within the top 7.5 cm of the concrete versus a source-to-
detector distance of 12 cm with the activity uniformly distributed within the bottom 7.5 cm of the
concrete. While more extreme conditions might be imagined, the foregoing were considered to
represent reasonable bounds on the source-to-detector distance and the activity distribution with
depth. The other assumptions used in the calculations were the same as used in Section 7.11.
Therefore, there are three values each to describe the distribution of the possible values of WT
and RT : The estimated mean value calculated for a uniform distribution of radioactivity in the 15
cm of concrete surveyed at 10 cm above; an estimated lower bound calculated for a uniform
distribution of radioactivity in the bottom 7.5 cm of concrete surveyed at 12 cm above; and an
estimated upper bound calculated for a uniform distribution of radioactivity in the top 7.5 cm of
concrete surveyed at 8 cm above.
The values for WT and RT at the extremes considered were not equally distant from the mean, i.e.,
their distribution was not symmetric. However the GUM suggests that in the absence of more
information the simplest approximation is a symmetric rectangular distribution of the same total
width. With this approximation, u(Wr) = 6673 and u(RT) = 4.638xlO"5 for natural uranium and
u(WT) = 539 and u(RT) = 7.315xlO"3 for natural thorium.
Using the information for natural uranium in Equation 7-34 we find:
0.072
f 26 | f 0.2 , ,
[2(215)J +[ 1 J \2(0.625^
= (238)2
= 10,013 (Bq/kg)2.
1 + 4£) +[
I ( RT ) \
'4.638xlO~5^
1.413X10"4
u(WT)
W^
[_6673_|
\ 44,923 J
So, taking the square root of the variance and rounding the result, uc (yv ) = 100 Bq/kg.
Therefore the FIDLER Scan MDC for natural uranium, yu, is 400 Bq/kg with an expanded
uncertainty of 200 Bq/kg, using a coverage factor of 2 and an estimated coverage probability of
95%. The upper bound of the Scan MDC using this interval is 600 Bq/kg.
Similarly substituting the information for natural thorium into Equation 7-34 we find:
( 26 ^ f0.2
( 0.072 1 f u([L) \ \u(WT)
i i +1 1 + + -L^ + -L-iZ
(2(215)} I 1 J ^ 2(0.625 J ^ R, ) ( WT ,
= (15)2
f 26 1 (0.2\ ( 0.072 1 f7.315xlO"3l (539
I I + + + +
1. 1 J ^2(0.625j ^2.619xlO"2J ^3,881
12(215) )
= 32 (Bq/kg)2.
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
So, taking the square root of the variance and rounding the result, uc (yTh) = 6 Bq/kg.
Therefore the FIDLER Scan MDC for natural thorium, yn, is 25 Bq/kg with an expanded
uncertainty of 12 Bq/kg, using a coverage factor of 2 and an estimated coverage probability of
95%. The upper bound of the Scan MDC using this interval is 37 Bq/kg.
The upper bound of the scan MDCs of approximately 600 Bq/kg for natural uranium and 37
Bq/kg for natural thorium are both less than their respective action levels of 38,000 and 330
Bq/kg. Therefore, the FIDLER is an acceptable instrument for performing the scan
measurements.
8.2.6.6 Determine Measurement Uncertainty for Concrete Samples
The primary measurement quality objective is the required measurement method uncertainty at
the action level. MARSAME recommends UMR < A/10 by default when decisions are being made
about the mean of a sampled population.
C238 c234 c232 c226
For this illustrative example, the Unity Rule, — ^- + — ^- + — 2-^L- + — ^- < 1 , is used to
compare the sum of the ratios of the radionuclide concentrations to their respective action
levels.1 Because the results of the survey are used to calculate a sum of fractions, the action level
is normalized to 1. The required measurement method uncertainty at this action level is A/10 =
(UBGR - LBGR)/10. Because the LBGR was chosen to be 0.15, then UMR < A/10 = (UBGR -
LBGRyiO = (1.0 - O.lSyiO = 0.085.
Therefore, we require that:
" C23S C234 C232 C226
V^—^U --^*u ""^Th — "<>Ray V ^SU /34U ^Th "6Ra
Clearly, if each of the four terms in the sum is constrained to a fourth of its limit, the unity rule
will be satisfied.
If the concentrations of the radionuclides of concern are independent, then:
C C C C
*— 238rr *— 234rr *— 232^, *— 2
u" I u- + u- + ^- '
025AL226Ra^
»(c,,,,) Y , r 4%,) Y , r »(c...j Y
0.25AL2
025AL226
•\2
< (0.085)2
If the required relative measurement method uncertainty is the same for each radionuclide,
therefore, the required relative method uncertainty for each individual radionuclide is:
1 MARSSIM Section 4.3.3 and MARS SIM Appendix 1.11 provide information on applying the unity rule.
NUREG-1575, Supp. 1 8-22 January 2009
-------
MARSAME
Illustrative Examples
025AL2
w
025AL2
0.25AL2
(o.oss)2
= (0.0425)2
"Ra J
The required relative measurement method uncertainties for each radionuclide are provided in
Table 8.10.
Table 8.10 Radionuclide-Specific Required Relative Measurement Method Uncertainties
Radionuclide
238U
234U
232Th
226Ra
Modified Action Level
(Bq/kg)
38,000/4 = 9500
12,000,000/4 = 3,000,000
330/4 = 82.5
450/4= 112.5
Required Relative
Measurement Method
Uncertainty, (p^
4.25%
4.25%
4.25%
4.25%
The required measurement method uncertainty for each radionuclide was provided to the
analytical laboratory. The analytical laboratory used this information to specify sample volumes
required to ensure this MQO was achieved.
8.2.6.7 Collect Survey Data
As the concrete is removed from the crusher, it is placed in a wooden frame (measuring 8 m by
10 m by 15 cm) on a concrete pad. The wooden frame's volume (12 m3) corresponds to the
volume associated with each sample from the survey design (i.e., 83 m3 divided by 7 samples).
Therefore, 7 batches of concrete equal 1 survey unit. One sample is collected from the center of
the concrete rubble residing in the wooden form for each batch of crushed concrete. One hundred
percent of the surface is scanned to identify locations with count rates greater than 30,000 cpm to
investigate for areas of elevated activity and establish biased sampling points. A sample is
collected at each location exceeding 30,000 cpm.
If no scan results exceed 30,000 cpm, the concrete is removed from the form and placed in the
non-impacted concrete staging area awaiting laboratory analysis of the samples. If the scan
survey identifies areas exceeding 30,000 cpm, the concrete is transferred to a holding container
to control access to the concrete until the laboratory analyses are completed. A total of 126
batches of concrete are scanned (7 batches for each of the 18 survey units). Seventeen batches of
concrete are segregated as potentially containing elevated levels of radioactivity based on the
scan survey results, and one additional sample is collected from each batch as part of the
investigation. No areas exceeding 100,000 cpm are identified during implementation of the
disposition survey.
Five additional samples are collected from random locations on the floor of the administrative
building to provide a total of seven reference area samples. The results of the two samples
collected from the administrative building during the preliminary surveys are reviewed and
determined to be of adequate quality for the disposition survey.
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
All of the concrete samples collected during implementation of the disposition survey are sent to
a laboratory for analysis by gamma spectroscopy and alpha spectrometry for uranium isotopes.
Thorium-232 is quantified based on the 228Ac gamma spectroscopy results. Radium-226 is
quantified based on the 214Bi gamma spectroscopy results. A total of 150 samples are analyzed,
including seven samples from the reference area. The 17 biased-sample locations identified by
the scan survey were analyzed by gamma spectroscopy.
Performance checks of the FIDLER were made at the beginning and end of collection activities
for each survey unit. These performance checks included a blank measurement in an area away
from potential sources of radioactivity and a source check. Control charts were constructed to
monitor the performance of the FIDLER throughout the survey. One FIDLER was dropped while
performing a scan survey and the window was damaged. The instrument was removed from
service and all scan measurements were repeated using a replacement FIDLER for that survey
unit. No quality related problems were identified during the performance of the scan surveys.
The offsite laboratory provided the results of the laboratory analyses. The quality control
measurements specified in the work plan were performed. All of the QC results were within the
limits specified in the work plan. No quality related issues were identified during the
performance of the sampling surveys.
8.2.7 Evaluate the Survey Results
8.2.7.1 Conduct a Data Quality Assessment
The disposition survey design for the concrete rubble is verified as having been executed very
closely to the survey design, with the appropriate number of measurements collected for each of
the survey units.
The quality control sample results from the laboratory are reviewed and the data are deemed
acceptable. An exploratory data analysis of the entire data set is performed to gain an
understanding of the structure of the data.
The sum of fractions for each sample is calculated using the results for 238U, 234U, 232Th (228Ac),
and 226Ra (214Bi) and the radionuclide specific action levels. Only two samples result in sums of
fractions greater than 1.0 without correcting for background. Both of these samples came from
batches that were segregated prior to crushing based on visual evidence of staining within the
concrete rubble; these were also the two locations with the highest scan survey results. A
frequency plot (Figure 8.1) and normal cumulative frequency plot (Figure 8.2) were constructed
to provide visual representations of the data.
8.2.7.2 Conduct the Wilcoxon Rank Sum Test
The Wilcoxon Rank Sum test was used to compare the reference area data to the survey unit
data. In each case the test statistic exceeded the critical value of 65, so the null hypothesis was
NUREG-1575, Supp. 1 8-24 January 2009
-------
MARSAME
Illustrative Examples
rejected for all 17 survey units. It was concluded that the average activity in all the crushed
concrete exceeds background by less than a sum of fractions of 1.0.
90 -
80 -
70 -
60 -
I 50-
o 40 -
LL
30 -
20 -
10 -
0 -
0.1
0.3 0.5
0.7
™
i i — I
0.9 1.1 1.3
Sum of Fractions
Figure 8.1 Frequency Plot of Illustrative Example Data
1.6
1.4 -
1.2 -
in
I 1 -
•c
03
i 0.8 -
•5
E 0.6 -
U)
0.4 -
0.2 -
0 -
-3
-2
-1 0 1
Normal Cumulative Frequency
Figure 8.2 Cumulative Frequency Plot of Illustrative Example Data
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.2.8 Evaluate the Results: The Decision
In every survey unit, including those with stained concrete, the test statistic for the WRS test
exceeded the critical value in Table A.4 in Appendix A. The null hypothesis that the mean sum
of fractions in the survey unit exceeds 1.0 is rejected. Even though the standard deviation of the
survey unit results (0.287) exceeded the variability used to design the survey (i.e., 0.15), it did
not significantly impact the ability to make a decision about the concrete rubble. Based on the
results of the disposition survey, all the crushed concrete can be cleared.
8.3 Mineral Processing Facility Rented Equipment Baseline Survey
This illustrative example is provided for information purposes only and presents a theoretical
application of MARSAME guidance. This example describes a scan-only interdiction survey
using Scenario B with an action level of no detectable radioactivity. The text is provided to
illustrate the application of MARSAME guidance, and should not be considered an example
survey plan. The amount of discussion provided in this example is based on the complexity of
the problem and the relative difficulty expected from applying or interpreting specific portions of
MARSAME guidance. The amount of discussion for this example is not related to, and should
not be used as an estimate of, the level of effort associated with planning, implementing, or
assessing an actual disposition survey.
8.3.1 Description
Heavy equipment is required to move the piles of concrete rubble at the mineral processing
facility discussed in Section 8.2. A front loader is rented to assist with the work. The radiological
history of the rented front loader is unknown.
8.3.2 Objectives
The objective is to apply interdiction controls to prevent the introduction of offsite radioactive
materials to the mineral processing facility. In addition, surveying the front loader before it
enters the site may provide reference area data for use in clearing the front loader at the end of
the project (Section 8.4). The scope of this illustrative example is limited to a rented front loader
being brought to the site for on-site transport of impacted concrete rubble.
8.3.3 Initial Assessment of the M&E
8.3.3.1 Categorize the M&E as Impacted or Non-Impacted
The material to be assessed is a rented front loader (Figure 8.3). A review of the existing
information shows it is not adequate to categorize the front loader (see Figure 2.1 in Chapter 2).
A visual inspection of the front loader as it is delivered to the site shows the equipment has been
used, but there are no notable quantities of soil. No detailed historical records pertaining to the
usage history of the front loader are available for review, other than that available from the rental
company pertaining to the types of sites where heavy equipment is rented and used. Natural
radionuclides are present in or commingled with soil, sediment, rubble, debris, and water. Heavy
NUREG-1575, Supp. 1 8-26 January 2009
-------
MARSAME
Illustrative Examples
equipment is in direct contact with natural uranium and thorium during operations. Because there
is a possibility the M&E may contain radionuclide concentrations or radioactivity exceeding the
background at the mineral processing facility, the front loader is categorized as impacted.
Figure 8.3 Front Loader
Sentinel measurements were performed to provide information on whether the difficult-to-
measure portions of the front loader, specifically the engine, were impacted by activities
conducted prior to arrival at the site. The existing air filter was removed and a sentinel
measurement of the used air filter was performed to determine if any radioactivity was associated
with the air filter. A smear sample was taken from the air intake beyond the air filter to test for
removable radioactivity. A second smear sample was taken inside the exhaust pipe to test for
removable radioactivity exiting the difficult-to-measure engine areas. Measurements were
performed using a hand-held gas proportional detector with an effective probe area of 100 cm2, a
detection limit less than 1,000 dpm per 100 cm2 (Section 8.3.5.2), and counting for 1 minute.
Smear measurements were made using a dual phosphor detector with a detection limit less than
1,000 dpm per 100 cm2 (Section 8.3.5.2), and counting for 2 minutes. The results of the sentinel
measurements are shown in Table 8.11.
Table 8.11 Sentinel Measurement Results
Sample
Description
Air Filter
Air Intake
Smear
Exhaust
Smear
Reference Material
Counts (Before Use, NB)
a
2
0
0
ft
154
66
66
Sample Counts
(After Use, Ns)
a
5
1
0
I
156
74
68
Net Count
(NS-NB)
a
3
1
0
ft
11
8
2
Critical Value of the
Net Instrument
Signal (Sc, Table 7.5)
«
4.96
2.82
2.82
ft
28.1
18.9
18.9
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples
MARSAME
The sentinel measurement results are below the critical value of the net instrument signal, so no
radioactivity was detected by the sentinel measurements. As long as the results of the interdiction
survey do not detect any radioactivity, the decision will be that the difficult-to-measure areas of
the front loader are non-impacted.
8.3.3.2 Describe the M&E
The information available after categorizing the front loader is not adequate to select a
disposition option (see Figure 2.2 in Chapter 2). The data gaps for the front loader are associated
with describing the physical and radiological attributes of the front loader. The scoping survey
design includes scanning external and easily measurable areas of the front loader that have the
highest potential to contact radioactive materials.
A description of the physical attributes of the front loader is listed in Table 8.12 (per Table 2.1).
The front loader is a large, complicated piece of machinery. It incorporates four wheels that are
50 centimeters (cm) (1 feet [ft], 8 inches [in]) wide and 150 cm (5 ft) tall, a wheelbase of 345 cm
(11 ft, 4 in), an additional section of 246 cm (8 ft, 1 in) behind the rear wheels for the engine
housing, and a height of 363 cm (11 ft, 9 in) to the top of the operator cab.
Table 8.12 Physical Attributes Used to Describe the Front Loader
Attribute
Description
Dimensions
Size: Total Mass «25,490 kg (56,196 Ibs)
hape: Total Surface Area «180 m2
Complexity The front loader is composed of multiple materials. Most external components are
painted steel. However, the tires are rubber, the cab is comprised of large sections of
jlass, hydraulic fluid hoses are composed of high-pressure silicon, and the joints are
coated with grease.
Disassembly would ideally be avoided for the considerable time and expense it adds to
)erforming disposition surveys on the equipment.
Options for surveying interior surfaces include surveying of the engine air filters and
interior surfaces of the exhaust plumbing to determine whether it is likely radioactive
materials have spread into the engine.
Accessibility The inside corners of the bucket and portions of each tire and wheel are difficult to
measure using conventional hand-held measurements, even with a relatively small
land-held GM detector. The large height of the front loader, the underside of the front
oader, and the varying orientation of surfaces associated with the equipment represent
a scenario that makes accessibility difficult.
There are only a few porous surfaces that allow permeation of radioactivity, such as the
;rease used on external hinges and joints.
Air inlets, grease used on external hinges and joints, and air vents in the external panels
represent areas where radioactivity could penetrate to difficult-to-measure areas.
Inherent Value The front loader can be decontaminated, reused, or recycled. The costs associated with
ither replacing impacted portions of the front loader, or disposing of the front loader
and replacing it, are very high. As long as only exterior surfaces of the front loader
become impacted, the cost of decontamination to allow unrestricted release and reuse
Isewhere will probably not be substantial.
NUREG-1575, Supp. 1
January 2009
-------
MARSAME Illustrative Examples
The front loader uses a 320-cm wide (10 ft, 6 in), 4.7-m3) capacity bucket (6 yd3). The overall
length with the bucket is 914 cm (30 ft, 0 in).
The surface area was estimated by dividing the front loader into components with regular
geometric shapes and rounding to the nearest square meter. For example, the tires were modeled
as cylinders and the cab was modeled as a box. The bucket has a surface area of 13.5 m2, which
is applied to the inside and outside surfaces for a total of 27 m2. The exterior surfaces of the body
have a surface area of approximately 76 m2. The tires have a surface area of 24 m2, and the
inside of the cab is estimated at 16 m2. Because the surfaces are not actually regular geometric
shapes, a contingency factor of 25% (35 m2) was used to account for irregular surfaces, hoses,
etc. This contingency factor was based on professional judgment and approved through
discussions with the regulators. The rounded total surface area is 180 m2.
The front loader is composed of multiple materials. Most external components are painted steel.
However, the tires are rubber, the cab is comprised of large sections of glass, hydraulic fluid
hoses are composed of high-pressure silicon, and the joints are coated with grease. The front
loader is deemed accessible, as the areas most likely to contain radioactivity are all accessible
(though some portions of the front loader are more accessible than others) for conducting
measurements with hand-held instruments. Internal areas of the front loader are inaccessible
without disassembly.
The radiological attributes of the front loader are listed in Table 8.13 (per Table 2.2).
Radionuclides of potential concern include any radionuclides that may be present. Members of
the uranium and thorium radioactive decay series are used as a preliminary list of radionuclides
because these are the radionuclides of concern for the site (Appendix C lists types of sites where
uranium and thorium series radionuclides may be present). These are the radionuclides that are
known to be present at the mineral processing facility. Radioactivity associated with the front
loader is anticipated to be present at near-background concentrations. Materials may have built
up in specific locations on the front loader (e.g., joints with external grease, tires, corners of the
bucket) resulting in small areas of elevated radioactivity. The distribution of radioactive material
is expected to be concentrated on the underside and lower edges of the front loader. Horizontal
surfaces also present areas for the potential deposition of airborne radioactivity (angled and
vertical surfaces also present areas for the potential deposition of airborne radioactivity but
deposition of radioactivity is less likely in these areas due to surface orientation).
Table 8.13 Radiological Attributes Used to Describe the Front Loader
Attribute
Radionuclides
Activity
Distribution
Location
Description
Radionuclides of potential concern are any radionuclides that can be identified. The
uranium and thorium series radionuclides are used as a preliminary list, because these
are the radionuclides of concern for the mineral processing facility.
Radionuclide concentrations are expected to be close to background or zero.
Radioactivity is expected to be associated with materials that have come in contact
with the front loader. These materials will likely build up in specific locations
resulting in small areas of elevated activity that can be visually identified.
Radioactivity associated with the front loader is expected to be surficial and
removable.
January 2009 8-29 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
Given the unknown use history of the front loader, professional judgment and process knowledge
are used to develop a likely scenario for the potential distribution of radioactivity. Radioactivity
associated with the front loader is expected to be surficial only. Because the radioactivity is
expected to be associated with materials from the site, the radioactivity is also expected to be
removable.
Process knowledge does not provide a likely scenario for activation or other method for
volumetrically impacting the front loader.
8.3.3.3 Design and Implement Preliminary Surveys
A Geiger-Mueller (GM) meter is used to collect initial scanning survey data to help address data
gaps on the bucket and tires (i.e., external and easily measurable areas of the front loader that
have the highest potential for residual radioactivity). The maximum reading from the bucket was
80 counts per minute (cpm), and the maximum reading from the tires was 65 cpm. A collimated
in situ gamma spectrum made of the front loader showed no gamma lines other than those
associated with natural uranium, potassium, and thorium. Although one might expect some trace
amounts of 137Cs from atmospheric fallout, there was not enough to show up in the spectrum.
A non-impacted section of steel I-beam approximately one foot long (which resembles the
majority of the surfaces of the front loader) is used as a reference material to establish the GM's
background count rate. Scanning measurements are collected from flat surfaces, edges, and
inside corners of the I-beam; count rates of 30 to 35 cpm are observed. Daily quality control
checks were performed to ensure the instruments were operating properly.
8.3.3.4 Select a Disposition Option
The disposition options for the front loader are to accept it for use at the mineral processing
facility following an interdiction survey, or to return it to the rental company.
8.3.3.5 Document the Results of the Initial Assessment
The results of the IA were documented in a letter report to the project manager. The decision to
categorize the front loader as impacted was included in the report, along with the descriptions of
the physical and radiological attributes of the front loader. The letter report described the scoping
survey and listed the results of the measurements.
8.3.4 Develop a Decision Rule
Following completion of the IA, additional information needed to develop the disposition survey
design is collected.
8.3.4.1 Select Radionuclides or Radiations of Concern
The initial assessment indicates that natural uranium and natural thorium are the radionuclides of
potential concern.
NUREG-1575, Supp. 1 8-30 January 2009
-------
MARSAME Illustrative Examples
8.3.4.2 Identify Action Levels
The action level selected for the interdiction survey is no detectable surface radioactivity above
background. Because there are multiple radionuclides to be evaluated during the interdiction
survey, additional discussion of action levels may be necessary.
8.3.4.3 Describe the Parameter of Interest
The parameter of interest for an interdiction survey with an action level of no detectable activity
is the level of radioactivity above background reported for each measurement. Any measurement
that detects the presence of radioactivity above background indicates the action level has been
exceeded.
8.3.4.4 Identify Alternative Actions
The alternative actions are determined by the disposition option. If the front loader is refused
access to the site, it will be returned to the rental company. If the front loader is granted access to
the site, it will be used to transport concrete rubble.
8.3.4.5 Develop a Decision Rule
The decision rule incorporates the action level, parameter of interest, and alternative actions into
an "if.. .then" statement.
If the results of any measurement identify surface radioactivity in excess of background, then the
front loader will be refused access to the site. If no surface radioactivity in excess of background
is detected, then the front loader will be granted access to the site.
8.3.4.6 Identify Survey Units
A survey unit is defined as the quantity of M&E for which a separate disposition decision will be
made. The front loader is the survey unit. The decision rule will be applied by comparing
individual measurement results to the critical value for detection. All measurements must be
below the critical value (i.e., no surface radioactivity in excess of background detected) in order
to accept the front loader.
8.3.4.7 Develop Inputs for Selection of Provisional Measurement Methods
The selection of a measurement method depends on the list of radionuclides or radiations of
concern and will affect the survey unit boundaries. Establishing performance characteristics for
the measurement method (i.e., measurement quality objectives [MQOs]) will help ensure the
measurement results are adequate to support the disposition decision. Three provisional
measurement methods were identified by the planning team for consideration; scan-only, in situ,
or a combination of both methods in a MARSSIM-type survey design. No method-based survey
designs were identified that matched the description of the M&E, so no method-based survey
designs were considered.
January 2009 8-31 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
Detection Capability
Because the action level is stated in terms of detection capability, the detection capability is
critical in selecting an acceptable measurement method. The detection capability is defined as the
minimum detectable concentration (MDC). The survey design will need to specify how hard to
look (i.e., select an appropriate discrimination limit) before the MQO for detection capability can
be established. The MDC for the selected measurement method must be less than or equal to the
discrimination limit.
Measurement Method Uncertainty
The measurement method uncertainty is also important in selecting a measurement method. The
MQO for detection capability will determine the acceptability of a measurement method, but it
will also include information on the measurement method uncertainty. The measurement method
uncertainty at background concentrations is used to calculate the MDC, as well as the critical
value for the detection decision.
Range
The selected measurement method must be able to detect radionuclide concentrations or
radioactivity at the discrimination limit. However, the measurement method must also be able to
operate and quantify radionuclide concentrations or radioactivity at levels equal to those
identified in the M&E at the site.
Specificity
The requirement for specificity will be tied to the list of radionuclides and radiations of concern.
If radionuclide specific measurements are required, the measurement method must be able to
identify radioactivity associated with specific radionuclides. If radionuclide specific
measurements are not required, methods that measure gross activity may be acceptable.
Ruggedness
Ruggedness is not expected to be a major concern for selecting a measurement method. Because
only surficial radioactivity is expected, in situ measurements of front loader surfaces will be used
to collect data for comparison to the action levels. The selected measurement method must be
able to perform these surface measurements in the field where the front loader is located. The
environmental conditions will depend on the site location (e.g. northeast versus southwest) and
the time of the year (e.g., winter versus summer).
8.3.4.8 Reference Materials
The majority of the surfaces on the front loader are metal (e.g., steel), although there are several
rubber surfaces as well (e.g., tires, hoses). The small steel I-beam used to estimate background
during the preliminary surveys will be used as the reference materials for the disposition survey.
There is no inherent radioactivity from the uranium or thorium decay series expected in steel or
NUREG-1575, Supp. 1 8-32 January 2009
-------
MARSAME Illustrative Examples
rubber, so the selection of the reference material is not expected to result in any bias during
interpretation of the results.
8.3.5 Develop a Survey Design
8.3.5.1 Select a Null Hypothesis
The hypotheses being tested are:
• Null Hypothesis: The front loader contains no detectable radionuclide concentrations or
radioactivity above background levels (i.e., indistinguishable from background).
• Alternative Hypothesis: The front loader contains detectable radionuclide concentrations or
radioactivity above background levels.
MARSAME processes require the use of Scenario B when the action level is zero, which is the
case for indistinguishable from background.
8.3.5.2 Set the Discrimination Limit
The discrimination limit is the radionuclide concentration or level of radioactivity that can be
reliably distinguished from the action level by performing measurements. Under Scenario B, the
discrimination limit determines how hard the surveyor needs to look to determine there is no
detectable radioactivity.
Acceptable surface activity levels derived from the relevant regulatory agency were selected as
the discrimination limits for radionuclides of potential concern. Table 8.14 lists the potential
discrimination limits based on the preliminary list of radionuclides of concern.
Table 8.14 Potential Discrimination Limits
Radionuclide of Potential Concern
Average (dpm/100 cm2)
Maximum (dpm/100 cm2)
Natural U
5,000
15,000
Natural Th
1,000
3,000
Based on the preliminary selection of radionuclides of potential concern, the discrimination
limits for natural thorium represent the limiting case.
8.3.5.3 Specify the Limits on Decision Errors
A Type I decision error occurs when the null hypothesis is rejected when it is true. For this
survey, a Type I decision error would be refusing to allow the front loader onto the site even
though there is no radioactivity present that exceeds background. The consequences of this
decision error may include unnecessarily returning the front loader and taking additional time to
locate a replacement, or possibly deciding to decontaminate the front loader prior to use on the
site. During scanning, the consequence of making a Type I decision error is the need to perform
an investigation to determine the reason for the elevated reading, A Type I decision error rate of
25% is selected for the scanning survey to balance the potential of additional rental costs for the
January 2009 8-33 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
front loader while additional investigations are performed and evaluated against the additional
time required to scan at slower speeds to achieve this DQO.
A Type II decision error occurs when the null hypothesis is not rejected when it is false. For this
survey, a Type II decision error would be allowing the front loader to be used on the site when
there is radioactivity above background. The consequences of a Type II decision error may
include introducing additional radionuclides on to the site and slightly increased exposures to
workers. It may also make it difficult to clear the front loader and return it to the rental company
when the work is complete. For this reason a Type II decision error rate of 5% is selected for the
scanning.
8.3.5.4 Select a Measurement Technique
At this point in the survey design process, the planning team decides to evaluate each of the three
provisional measurement methods from Section 8.3.4.7 to determine what might be feasible for
surveying the front loader. Final selection of a measurement technique will help determine the
final survey design and decide between the multiple options currently available for the survey.
A scan-only survey approach requires that the measurement method be capable of detecting
radioactivity at the discrimination limit. Any results exceeding the critical value would provide
evidence of radioactivity levels exceeding background. There would be no need to record
individual measurement results, because every result would be compared to the critical value.
The calculation of the total efficiency is expected to be a major source of measurement method
uncertainty. Additional measurements or assumptions are required to select a source term as the
basis for the efficiency calculations. Scanning can be performed for alpha, beta, gamma, or some
combination of the types of radiation. The amount of the front loader requiring scanning (i.e. 10
to 100%) would be determined by the classification. It is unknown if any scan-only measurement
methods are available that meet the MQOs.
In situ survey approaches also require that the measurement method be capable of detecting
radioactivity at the discrimination limit. In situ techniques allow identification of specific
radionuclides, if necessary. The major source of measurement method uncertainty will likely be
the model used to calculate the efficiency. Additional measurements or assumptions are required
to select a source term as the basis for the efficiency calculations. The amount of the front loader
requiring measurement (i.e., 10 to 100%) would be determined by the classification. The final
number of measurements will be linked to the field of view of the detector. For example, a
detector with a 1-m2 field of view would require more than 180 measurements to measure 100%
of the external surfaces of the front loader. An instrument such as the GM detector used during
the scoping survey with a field of view of less than 100 cm2 would require thousands of
measurements to measure the minimum 10% of the front loader.
A MARSSEVI-type approach would use a combination of direct measurements or samples with
scanning to support a disposition decision. Sampling could damage the front loader, so direct
measurements would be preferred. Locating measurements on the surface of the front loader will
be problematic. Similar to scan-only and in situ designs, the scanning and direct measurements
should be capable of detecting radioactivity at the discrimination limit. The MARSSIM-type
survey design would require the most resources to implement.
NUREG-1575, Supp. 1 8-34 January 2009
-------
MARSAME Illustrative Examples
Based on the evaluation of measurement techniques, a scan-only survey design is the preferred
approach. Assumptions about the radionuclides of concern will need to be established and the
availability of scan-only measurement methods needs to be verified.
8.3.5.5 Finalize Selection of Radiations to be Measured
Scan-only measurement methods are available for alpha, beta, and gamma radiations. The higher
background associated with scanning for gamma radiation makes it unlikely that the
measurement method could detect radioactivity at the discrimination limit. Alpha particles are
attenuated more than beta particles, increasing the uncertainty caused by variations in source to
detector distance. Scan-only measurement methods for beta radiation should provide the
optimum survey design. However, the lower detection limits associated with alpha
measurements may be required to meet the detection capability MQO. Any radioactivity in
excess of background is assumed to result from natural thorium, which is the limiting
radionuclide.
8.3.5.6 Develop an Operational Decision Rule
A scan-only survey will be performed for beta (and possibly alpha) radiation. Any result that
exceeds the critical value associated with the MDC set at the discrimination limit will result in
rejection of the null hypothesis, and the front loader will not be allowed on the site. Additional
constraints on data collection activities include that the front loader be clean and dry when the
measurements are performed.
8.3.5.7 Classify the M&E
The expected levels of radioactivity are background (see Table 8.13). No radioactivity in excess
of background is expected, so the front loader is classified as Class 3.
8.3.5.8 Select a Measurement Method
The planning team decided to verify the availability of an acceptable measurement method prior
to finalizing the survey design. The GM detector used to perform the preliminary survey is
evaluated first. The expected range of radioactivity based on the reference material and
preliminary survey data is approximately 35 cpm (i.e., background) to 80 cpm.
Based on the scanning survey data collected using the GM detector during the preliminary
surveys, the anticipated Scan MDC of the GM detector may not be capable of detecting
radioactivity at the discrimination limit of 1000 dpm/100 cm2 (see Table 8.14).
An alpha-beta gas proportional detector utilizing a larger effective probe area will help achieve a
lower scan MDC. The maximum reading for measurements from the bucket is 250 cpm; and the
maximum reading from the tires is 220 cpm. Measurements collected from flat surfaces, edges,
and inside corners of the reference material I-beam provide count rates between 180 and 190
cpm. The maximum background count rate is converted to scan MDC using NUREG-1761
(NRC 2002a) Equations 4-3 and 4-4.
January 2009 8-35 NUREG-1575, Supp. 1
-------
Illustrative Examples
MARSAME
5. =
= 2.32 x
= 6.7 counts
MDCR = Sjx — = 6.7x — = 201 cpm
ScanMDC =
MDCR
201
/0.5 xl.29
= 220dpm/WOcm2
Where:
/'
p
d'
MDCR
SjSs
average number of background counts in the observation interval
2(250/60) = 8.3 counts)
the interval length (2 s) based on a scan speed of 5 cm/s
efficiency of a less than ideal surveyor, range of 0.5 to 0.75 from
NUREG-1507 (NRC 1998b); a value 0.5 was chosen as a conservative
value
detectability index from Table 6.1 of NUREG-1507 (NRC 1998b); a
value of 2.32 was selected, which represents a true positive detection rate
of 95% and a false positive detection rate of 25%
minimum detectable number of net source counts in the observation
interval (counts)
minimum detectable count rate (cpm)
weighted total alpha-beta efficiency for natural thorium in equilibrium
with its progeny on the surveyed media (1.29, see Table 8.15)
The scan MDC for activity is now below 1,000 dpm/ 100 cm2 and is good enough to detect
radioactivity at the 232Th discrimination limit.
232
Table 8.15 Detector Efficiency for the Mineral Processing Facility (Th in Complete
Equilibrium with its Progeny) using a Gas Proportional Detector
Radionuclide
232Th
228Ra
228Ac
228Th
224Ra
220Rn
216Po
212Pb
212^
212Bi
212p0
208yi
Average Energy
(keV)
alpha
7.2 keV beta
377 keV beta
alpha
alpha
alpha
alpha
102keVbeta
770 keV beta
alpha
alpha
557 keV beta
Fraction
0.64
0.36
0.64
0.36
Instrument
Efficiency
0.40
0
0.54
0.40
0.40
0.40
0.40
0.40
0.66
0.40
0.40
0.58
Surface
Efficiency
0.25
0
0.50
0.25
0.25
0.25
0.25
0.25
0.50
0.25
0.25
0.50
Total efficiency =
Weighted
Efficiency
0.1
0
0.27
0.
0.
0.
0.
0.
0.211
0.036
0.064
0.104
1.29
From NUREG-1761 (NRC 2002a), Table 4.3
NUREG-1575, Supp. 1
January 2009
-------
MARSAME Illustrative Examples
8.3.5.9 Optimize the Disposition Survey Design
A scan-only interdiction survey will be performed of the exterior surfaces of the front loader.
Because the front loader is Class 3, approximately 10% of the external surface area will be
surveyed. Professional judgment will be used to select the locations for the scans in the locations
with the highest potential for radioactivity (i.e., the bucket, tires, and floor of the cab).
Approximately 50% of each of these areas will be surveyed, for a total of approximately 18m2
(7 m2 of the bucket, 10m2 of the tires, and 1 m2 of the cab floor). Experienced technicians will
be used to perform the surveys. The scan speed will be 5 cm per second, so the scan should take
approximately one man-hour to complete. The scans will be performed using a 100 cm2 active
probe area alpha-beta gas-proportional detector.
If while scanning, an area is perceived to exceed background (i.e., exceeds the scan MDC), the
surveyor will suspend the scan survey and perform an investigation survey consisting of a one-
minute measurement to verify the result of the scan measurement. The one-minute time interval
was chosen to meet the DQOs and MQOs for this measurement. If the results of the one-minute
verification measurement exceed the critical value calculated in 8.3.6.5, the radioactivity at that
location exceeds background and should be recorded on a log sheet. The location of any one-
minute verification measurement that exceeds the critical value will be clearly marked.
Quality control (QC) measurements will be performed prior to the start of the survey and at the
completion of the survey. These QC measurements will demonstrate that the instruments were
working properly while the survey was being performed. In addition, approximately 5% of the
survey will be repeated using a different surveyor to confirm the results of the initial survey.
8.3.5.10 Document the Disposition Survey Design
The interdiction survey design was documented in a letter report to the project manager. The
results of the IA were also included in this letter report.
8.3.6 Implement the Survey Design
8.3.6.1 Ensure Protection of Health and Safety
Protection of health and safety was performed as part of the survey implementation, but is not
included in this illustrative example (see Section 8.2.6.1 for an example Job Safety Analysis.)
8.3.6.2 Consider Issues for Handling M&E
Because only a portion of the front loader needs to be accessed to implement the survey design,
the front loader does not need to be moved to provide access to additional areas during the
survey (e.g., bottom of tires, underside of bucket). Areas included in the survey do not need to be
marked, outside of the small area that will be re-surveyed as part of the QC checks and locations
of measurements exceeding the critical value. The front loader will not be parked adjacent to
areas known to contain radionuclide concentrations or radioactivity in excess of background
(e.g., piles of concrete rubble) while the survey is performed.
January 2009 8-37 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.3.6.3 Segregate the M&E
No segregation of the front loader is required to implement the survey design.
8.3.6.4 Determine the Measurement Detectability for the Scan Survey
Section 8.3.4.7 established the MQO for the measurement detectability. The scan MDC must be
less than or equal to the discrimination limit.
8.3.6.5 Determine the Measurement Detectability for the Investigation Survey
As indicated in Section 8.3.5.9, an investigation survey will be performed for any result that
exceeds the scan survey investigation level (i.e. scan MDC). Both Type I and Type II errors that
might occur during the investigation survey are equally undesirable. The consequence of
incorrectly alleging that the front loader contains radioactivity in excess of background (Type I
error) may raise unnecessary regulatory concerns. On the other hand, accepting a front loader
that has radioactivity detectable above facility background (Type II error) may make it difficult
to clear when the work is finished. Thus it is desirable to initially set a = ft = 0.01. The critical
value for the one-minute measurement may be calculated from the equation in line 1 of Table
7.5:
— l + —=2.326>/2x250= 2.326V500 = 52 net counts,
t \ t I
1B \ 1B J
Where:
Sc = the critical value
NB = the mean background count (250 counts)
ts = the count time for the test source (one minute)
IB = the count time for the background (one minute)
Z]_a = the (1 - or)-quantile of the standard normal distribution (2.326 when or =0.01).
The minimum detectable net count can be calculated from the equation in line 1 of Table 7.6:
000/^2
= 52 + - + 2.326J— + 52+ 250 (2) =109 net counts,
Where:
zi-p = the (1 - /?)-quantile of the standard normal distribution (2.326 when /?=0.01)
<5b = the minimum detectable value of the net instrument signal (discrimination limit, 7
cpm)
The MDC can be calculated from Equation 4-1 in NUREG-1761 (NRC 2002a):
NUREG-1575, Supp. 1 8-38 January 2009
-------
MARSAME
Illustrative Examples
MDC =
detection limit
S
D
total efficiency x sample size
of natural thorium.
_ (109) _ 1Q9
Probe Area / ^ 100 1.29
X 1 1 . 2* V I X
= 84.5dpm/100cm2
100
100
8.3.6.6 Determine Measurement Uncertainty for the Investigation Survey MDC
MDC =
Sr
Probe Area
(S,SJX 10Q
Assuming a negligible uncertainty in the probe area, the combined standard uncertainty of the
MDC is (see Equation 7-33):
MC2(MDC) =
3MDC
dSn
dMDC] ,
Note that SjSs is treated as a single input variable because it is the weighted total alpha-beta
efficiency for natural thorium in equilibrium with its progeny on the surveyed media.
Because the MDC is of the form of a ratio of products, Equation 7-34 may be used:
S2r
+N - 1 + -
T
z, aiNB-*-
\\ t \
ll *z?\
4
2.3262
t,
-^- +
t \ 2
*z?/ -^
+ 2.326
2.326'
Where the formula for Sc and the values of the constants have been inserted. The uncertainties in
the times are assumed to be negligible, so these have also been treated as constants. Thus, the
uncertainty in SD will be due entirely to the uncertainty in the background count:
u\SD) =
u2(NB)
The sensitivity coefficient for SD at NB = 250 is
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples
MARSAME
as;
(2)
-(2326jNB(2)) + 2NB
' d(232&jNB(2)Y
57Vn
(2))-
l326jNa(2) +27V,
(2.326VI)
+ 0 + 2.326
2.3262
2.3262
4
(2.326^2)) + 2NB
d\ ^l + (2.326j7V^2))
?*(-
1.6447
+ 2.326
.2897
d (0. 5 8 1 5 + 3 .289^/7^ + 27VB
d o. 5 815 + 3.289 + 27V
1.163
1.6447 ( A/N
= 0.104 + - = 0.208
23.5
Suppose the spatial variability in NB can be described by a triangular distribution with a mean of
250 and a half-width of 50, then,
U(NB) =
and
u(SD} = \
= (0.208)(20.4) =
A complete analysis of the uncertainty in SiSs, the weighted total alpha-beta efficiency for natural
thorium in equilibrium with its progeny on the surveyed media involves propagation of
uncertainty through all of the input quantities in Table 8.15. The uncertainty in the weighted total
alpha-beta efficiency is
Putting this information together into Equation 7-34 for the combined total variance of the MDC
we have:
NUREG-1575, Supp. 1
8-40
January 2009
-------
MARSAME Illustrative Examples
1.292
= 7,140(0.000148 + .024)
= 172.4
So the estimated combined standard uncertainty in the MDC is wc(MDC) = 13.1.
8.3.6.7 Perform Quality Control Measurements
The required QC measurements are performed as described in the survey design.
8.3.6.8 Collect Survey Data
Data from the survey of the front loader is collected consistent with the survey design and
provides a complete record of the data collected. Thirty-seven locations were flagged during the
survey for investigations using one-minute measurements. None of the one-minute measurement
results exceeded the critical value.
8.3.7 Evaluate the Survey Results
8.3.7.1 Conduct a Data Quality Assessment
The surveying procedure utilized for the front loader was verified as having been executed very
closely to the survey design, with the appropriate survey coverage. The results of the QC
measurements demonstrated that the instruments were working properly and a different surveyor
could duplicate the results of the survey. Control charts used to check the performance of the
survey instruments did not identify any potential problems with the instruments.
8.3.7.2 Conduct a Preliminary Data Review
The preliminary data review for this baseline survey does not yield identifying patterns,
relationships, or potential anomalies. The locations of the additional investigations appear to be
randomly located based on visual inspection of the front loader.
8.3.7.3 Conduct the Statistical Tests
The statistical test selected for this scanning survey is direct comparison to the critical level. If
all the results are below the critical level associated with the discrimination limit, there is no
detectable radioactivity above background. All of the scanning results that exceeded the critical
value were subjected to additional investigation. All of the results of the additional investigations
were below the critical value.
January 2009 8-41 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.3.8 Evaluate the Results: The Decision
Based on the results of the baseline survey, the front loader is determined to have no detectable
radioactivity above background and is therefore allowed to enter the site.
8.4 Mineral Processing Facility Rented Equipment Disposition Survey
This illustrative example is provided for information purposes only and presents a theoretical
application of MARSAME guidance. This example describes a scan-only disposition survey
using Scenario A. Because this example uses the same M&E and the same survey design used in
Section 8.3, it points out the similarities and differences between interdiction and release
surveys. The examples in Sections 8.3 and 8.4 also point out the similarities and differences
between surveys designed using Scenario A and surveys designed using Scenario B. The text is
provided to illustrate the application of MARSAME guidance, and should not be considered an
example survey plan. The amount of discussion provided in this example is based on the
complexity of the problem and the relative difficulty expected from applying or interpreting
specific portions of MARSAME guidance. The amount of discussion for this example is not
related to, and should not be used as an estimate of, the level of effort associated with planning,
implementing, or assessing an actual disposition survey.
8.4.1 Description
The radiological surveys at the mineral processing facility described in Section 8.2 have been
completed. The front loader that was brought on site to assist with handling the concrete rubble
(Section 8.3) is no longer being used. The front loader must be cleared before it can be returned
to the rental company.
8.4.2 Objectives
The objective is to demonstrate the front loader can be cleared. The scope of this illustrative
example is limited to the rented front loader used for the on-site transport of impacted concrete
rubble.
An interdiction survey was performed to demonstrate there was no detectable radioactivity above
background associated with the front loader when it entered the site. This illustrative example
provides a comparison between interdiction and clearance surveys performed on the same piece
of equipment.
8.4.3 Initial Assessment of the M&E
8.4.3.1 Categorize the M&E as Impacted or Non-Impacted
The existing information is adequate to categorize the front loader. The front loader was used to
transport concrete rubble containing radionuclides with concentrations exceeding background.
The front loader is impacted. Following use, the front loader was steam cleaned to remove loose
dirt and grease (together with any associated radioactivity) for acceptance by the rental company.
NUREG-1575, Supp. 1 8-42 January 2009
-------
MARSAME
Illustrative Examples
Sentinel measurements were performed to provide information on whether the difficult-to-
measure portions of the front loader, specifically the engine, were impacted by site activities. In
addition to sentinel measurements, dust control measures were used to minimize the potential for
airborne radioactivity from soil parti culates throughout the project. Air monitoring of the work
zone and the breathing zone of the front loader operator was performed throughout the project to
estimate inhalation exposure for project workers.
A new air filter was installed at the beginning of the project and a single measurement of
radioactivity associated with the air filter was performed prior to use to provide an estimate for
background. Following completion of soil handling activities the air filter was removed and
stored for 72 hours to allow for decay of short-lived radon decay products. A sentinel
measurement of the used air filter was performed following storage to determine if any
radioactivity was associated with the air filter after being used at the site. A smear sample was
taken from the air intake beyond the air filter to determine if there was any removable
radioactivity. Measurements were performed using a hand-held gas proportional detector with an
effective probe area of 100 cm2, a detection limit less than 1,000 dpm per 100 cm2 (see Section
8.3.5.2), and counting for 1 minute. Smear measurements were made using dual phosphor
detector with a detection limit less than 1,000 dpm per 100 cm2 (see Section 8.3.5.2), and
counting for 2 minutes. The results of the sentinel measurements are shown in Table 8.16.
Table 8.16 Sentinel Measurement Results
Sample
Description
Air Filter
Air Intake
Smear
Reference Material
Counts (Before Use, NB)
a
2
0
I
145
66
Sample Counts
(After Use, Ns)
a
4
1
I
168
79
Net Count
(NS-NB)
a
2
1
I
23
13
Critical Net Signal
(Sc, Table 7.5)
«
4.96
2.82
I
28.1
18.9
The engineering controls minimized the potential for airborne contamination. The work zone and
breathing zone air monitoring results reported no detectable radioactivity with detection limits
below the acceptable derived air concentrations (DACs). The sentinel measurement results are
below the critical net signal, so no radioactivity was detected by the sentinel measurements. The
combination of engineering controls, air monitoring measurements, and sentinel measurements
support categorization of the difficult-to-measure portions of the front loader as non-impacted.
8.4.3.2 Describe the M&E
The description of the physical attributes associated with the front loader has not changed (see
Table 8.7). The uranium series and thorium series radionuclides listed in Table 8.2 are the
radionuclides of potential concern for the front loader. The existing information is adequate to
select a disposition option, and there are no data gaps.
January 2009
NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
8.4.3.3 Select a Disposition Option
The preferred disposition option for the front loader is clearance. The existing interdiction survey
design used to allow the front loader access to the site will be evaluated for applicability as a
clearance survey (Section 8.4.4.2).
8.4.3.4 Document the Results of the Initial Assessment
The decision to categorize the front loader as impacted will be documented with the results of the
survey. The planning team determined that no other documentation is necessary.
8.4.4 Develop a Decision Rule
8.4.4.1 Identify Action Levels
The action level selected for the interdiction survey was no detectable surface radioactivity
above background. The action levels in this case are the limits shown in Table 8.13 The limiting
value is 1000 dpm/100 cm2 for natural thorium.
8.4.4.2 Evaluate an Existing Survey Design
Because the same front loader is being surveyed, the measurement method is still adequate. The
scan MDC of 132 dpm/100 cm2 for natural thorium is well below the action level. There were no
problems identified during the interdiction survey that would prevent using the measurement
method for a clearance survey. The population parameter of interest and the survey unit
boundaries are linked to the measurement method (see Sections 8.3.4.3 and 8.3.4.6).
The alternative actions are different for the clearance survey. If the front loader is cleared, it will
be returned to the rental company. If the front loader is not cleared, it will remain on site. This
results in a change to the decision rule. If the results of any measurement identify surface
radioactivity in excess of background, the front loader will remain on site and radiological
controls will remain in place. If no surface radioactivity in excess of 1,000 dpm/100 cm2 over
background is detected, the front loader will be cleared and returned to the rental company.
8.4.5 Develop a Survey Design
8.4.5.1 Select the Null Hypothesis
Scenario A is being used, so the hypotheses being tested are:
• Null Hypothesis: The front loader contains detectable radionuclide concentrations or
radioactivity equal to or in excess of 1,000 dpm/100 cm2 above background levels
• Alternative Hypothesis: The front loader contains radionuclide concentrations or
radioactivity less than 1,000 dpm/100 cm2 above background levels.
NUREG-1575, Supp. 1 8-44 January 2009
-------
MARSAME Illustrative Examples
8.4.5.2 Set the Discrimination Limit
The discrimination limit is the radionuclide concentration or level of radioactivity that can be
reliably distinguished from the action level by performing measurements. Under Scenario A, the
discrimination limit should represent a prudently conservative estimate of any amount of natural
thorium that may be present on the front loader in excess of background.
8.4.5.3 Specify Limits on Decision Errors
A Type I decision error occurs when the null hypothesis is rejected when it is true. For this
survey, a Type I decision error would be clearing the front loader when there is radioactivity
detectable more than 1,000 dpm/100 cm2 above background. The consequence of a Type I
decision error may include releasing radionuclides from the site and increased exposures to
members of the public. The existing survey design specifies a Type I decision error rate of 5%
for scanning measurements for this decision error.
A Type II decision error occurs when the null hypothesis is not rejected when it is false. For this
survey, a Type II decision error would be refusing to clear the front loader even though the
radioactivity present exceeds background by less than 1,000 dpm/100 cm2. The consequence of
this decision error may include the need to perform an investigation to determine the reason for
the elevated reading, unnecessarily remediating the front loader, incurring additional costs for
extra rental time, or even purchasing the front loader and disposing of it as low-level radioactive
waste. The existing survey design specifies a Type II decision error rate of 25% for the scanning
measurements for this decision error. Note that the definitions of Type I and Type II decision
errors are reversed compared to the existing survey design from Section 8.3.
8.4.5.4 Classify the M&E
The potential for radioactivity exceeding background has increased because the front loader is
known to have contacted concrete rubble containing radionuclides at concentrations that exceed
background. This increased potential for radioactivity exceeding background results in a higher
classification for portions of the front loader for the clearance survey. The inside of the bucket is
now classified as Class 1. The remaining external surfaces are considered Class 3 so professional
judgment can still be used to determine where surveys will be performed.
8.4.5.5 Optimize the Existing Survey Design
The front loader will be scanned with an alpha-beta gas proportional detector. Experienced
technicians will perform the surveys. If while scanning, an area is perceived to exceed
background, a one-minute measurement will be performed at that location to verify the scan
results. If the results of the one-minute count exceed 1,000 dpm/100 cm2 above background the
front loader will require further remediation before it can be released. The results of all one-
minute verification counts will be recorded on a log sheet. The location of any one-minute count
that exceeds the critical value will be clearly marked.
January 2009 8-45 NUREG-1575, Supp. 1
-------
Illustrative Examples MARSAME
Based on the classification of the inside of the bucket as Class 1, 100% of the inside of the
bucket will be surveyed. In addition, 25% of the outside surface of the bucket will be surveyed,
concentrating on the bottom where the bucket frequently came in contact with the concrete
rubble. Similar to the interdiction survey, 50% of the tires and the floor of the cab will be
surveyed. In addition, 10% of the bottom and 5% the top (i.e., horizontal surfaces) will be
included in the clearance survey. Areas to be scanned will be biased to locations where residual
dirt or grease is visible. The increased surface area to be scanned is expected to increase the scan
time to approximately three man-hours. Based on professional judgment, four times as many
investigations are expected for the clearance survey, or approximately 150 one-minute
measurements. The additional investigations are expected to require an additional three man-
hours.
Implementation of this survey design will likely identify locations on the front loader bucket
with radioactivity levels exceeding 1,000 dpm/100 cm2 above background. To minimize these
occurrences, the front loader will be steam cleaned and dried prior to implementing the survey
design. Locations on the bucket (which is a Class 1 survey unit) where the additional
measurement exceeds the action level will be delineated using scanning techniques, scrubbed
clean to remove any surface radioactivity, and re-surveyed (i.e., clean-as-you-go). Locations with
radioactivity exceeding 1,000 dpm/100 cm2 above background are not expected anywhere else
on the front loader.
8.4.5.6 Document the Disposition Survey Design
The modified survey design was documented in a letter report to the project manager. The letter
report included the results of the categorization decision (Section 8.4.3.1).
8.4.6 Implement the Survey Design
The front loader was positioned on a concrete pad during steam cleaning operations. The water
was collected and containerized for survey prior to release. The bucket was lifted off the ground
and supported with wooden beams to provide access to the bottom of the bucket.
The survey was implemented as described in the survey design. The beta background in the area
underneath the bucket was higher than expected (i.e., 350 cpm instead of the 250 cpm used to
design the survey). The bucket was lifted higher off the ground (i.e., 1.5 meters instead of 15 cm)
and the scan survey was repeated with a lower background. The survey results were documented
in a letter report to the project manager.
8.4.7 Evaluate the Survey Results
8.4.7.1 Conduct a Data Quality Assessment
The surveying procedure utilized for the front loader was verified as having been executed very
closely to the survey design. The surveys included the appropriate scan coverage and number of
additional investigations. The preliminary data review for this baseline survey does not yield
identifying patterns, relationships, or potential anomalies. Control charts documenting the results
NUREG-1575, Supp. 1 8-46 January 2009
-------
MARSAME Illustrative Examples
of quantitative QC checks and performance checks indicate the DQOs have been achieved for
this clearance survey.
8.4.7.2 Conduct the Statistical Tests
The statistical test selected for this scanning survey is direct comparison to the action level of
1,000 dpm/100 cm2 above background. If all of the measurement results are below the action
level, the average natural thorium above background cannot exceed 1,000 dpm/100 cm2 above
background.
At 83 locations the scan MDC of 132 dpm/100 cm2 above background appeared to be exceeded.
However, none of the one-minute follow up counts at those locations exceeded 500 dpm/100 cm2
above background.
8.4.8 Evaluate the Results: The Decision
Based on the results of the disposition survey, the front loader is determined to have no
radioactivity above the action level and so can be cleared.
January 2009 8-47 NUREG-1575, Supp. 1
-------
-------
MARSAME
Appendix A
A.
STATISTICAL TABLES AND PROCEDURES
A.I Normal Distribution
Table A.I Cumulative Normal Distribution Function
z
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
2.10
2.20
2.30
2.40
2.50
2.60
2.70
2.80
2.90
3.00
3.10
3.20
3.30
3.40
0.00
0.5000
0.5398
0.5793
0.6179
0.6554
0.6915
0.7257
0.7580
0.7881
0.8159
0.8413
0.8643
0.8849
0.9032
0.9192
0.9332
0.9452
0.9554
0.9641
0.9713
0.9772
0.9821
0.9861
0.9893
0.9918
0.9938
0.9953
0.9965
0.9974
0.9981
0.9987
0.9990
0.9993
0.9995
0.9997
0.01
0.5040
0.5438
0.5832
0.6217
0.6591
0.6950
0.7291
0.7611
0.7910
0.8186
0.8438
0.8665
0.8869
0.9049
0.9207
0.9345
0.9463
0.9564
0.9649
0.9719
0.9778
0.9826
0.9864
0.9896
0.9920
0.9940
0.9955
0.9966
0.9975
0.9982
0.9987
0.9991
0.9993
0.9995
0.9997
0.02
0.5080
0.5478
0.5871
0.6255
0.6628
0.6985
0.7324
0.7642
0.7939
0.8212
0.8461
0.8686
0.8888
0.9066
0.9222
0.9357
0.9474
0.9573
0.9656
0.9726
0.9783
0.9830
0.9868
0.9898
0.9922
0.9941
0.9956
0.9967
0.9976
0.9982
0.9987
0.9991
0.9994
0.9995
0.9997
0.03
0.5120
0.5517
0.5910
0.6293
0.6664
0.7019
0.7357
0.7673
0.7967
0.8238
0.8485
0.8708
0.8907
0.9082
0.9236
0.9370
0.9484
0.9582
0.9664
0.9732
0.9788
0.9834
0.9871
0.9901
0.9925
0.9943
0.9957
0.9968
0.9977
0.9983
0.9988
0.9991
0.9994
0.9996
0.9997
0.04
0.5160
0.5557
0.5948
0.6331
0.6700
0.7054
0.7389
0.7704
0.7995
0.8264
0.8508
0.8729
0.8925
0.9099
0.9251
0.9382
0.9495
0.9591
0.9671
0.9738
0.9793
0.9838
0.9875
0.9904
0.9927
0.9945
0.9959
0.9969
0.9977
0.9984
0.9988
0.9992
0.9994
0.9996
0.9997
0.05
0.5199
0.5596
0.5987
0.6368
0.6736
0.7088
0.7422
0.7734
0.8023
0.8289
0.8531
0.8749
0.8944
0.9115
0.9265
0.9394
0.9505
0.9599
0.9678
0.9744
0.9798
0.9842
0.9878
0.9906
0.9929
0.9946
0.9960
0.9970
0.9978
0.9984
0.9989
0.9992
0.9994
0.9996
0.9997
0.06
0.5239
0.5636
0.6026
0.6406
0.6772
0.7123
0.7454
0.7764
0.8051
0.6315
0.8554
0.8770
0.8962
0.9131
0.9279
0.9406
0.9515
0.9608
0.9686
0.9750
0.9803
0.9846
0.9881
0.9909
0.9931
0.9948
0.9961
0.9971
0.9979
0.9985
0.9989
0.9992
0.9994
0.9996
0.9997
0.07
0.5279
0.5674
0.6064
0.6443
0.6808
0.7157
0.7486
0.7794
0.8078
0.8340
0.8577
0.8790
0.8980
0.9147
0.9292
0.9418
0.9525
0.9616
0.9693
0.9756
0.9808
0.9850
0.9884
0.9911
0.9932
0.9949
0.9962
0.9972
0.9979
0.9985
0.9989
0.9992
0.9995
0.9996
0.9997
0.08
0.5319
0.5714
0.6103
0.6480
0.6844
0.7190
0.7517
0.7823
0.8106
0.8365
0.8599
0.8810
0.8997
0.9162
0.9306
0.9429
0.9535
0.9625
0.9699
0.9761
0.9812
0.9854
0.9887
0.9913
0.9934
0.9951
0.9963
0.9973
0.9980
0.9986
0.9990
0.9993
0.9995
0.9996
0.9997
0.09
0.5359
0.5753
0.6141
0.6517
0.6879
0.7224
0.7549
0.7852
0.8133
0.8389
0.8621
0.8830
0.9015
0.9177
0.9319
0.9441
0.9545
0.9633
0.9706
0.9767
0.9817
0.9857
0.9890
0.9916
0.9936
0.9952
0.9964
0.9974
0.9981
0.9986
0.9990
0.9993
0.9995
0.9997
0.9998
Negative values of z can be obtained from the relationship O(- z) = 1 - O(z)
January 2009
A-l
NUREG-1575, Supp. 1
-------
Appendix A
MARSAME
A.2 Sample Sizes for Statistical Tests
Table A.2a Sample Sizes for Sign Test
(Number of measurements to be performed in each survey unit)
A/o
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.5
3.0
(«,/?) or (/?,«)
0.01 0.01 0.01 0.01 0.01
0.01 0.025 0.05 0.1 0.25
4,095 3,476 2,984 2,463 1,704
1,035 879 754 623 431
468 398 341 282 195
270 230 197 162 113
178 152 130 107 75
129 110 94 77 54
99 83 72 59 41
80 68 58 48 34
66 57 48 40 28
57 48 41 34 24
50 42 36 30 21
45 38 33 27 20
41 35 30 26 17
38 33 28 23 16
35 30 27 22 15
34 29 24 21 15
33 28 24 20 14
32 27 23 20 14
30 26 22 18 14
29 26 22 18 12
28 23 21 17 12
27 23 20 17 12
0.025 0.025 0.025 0.025
0.025 0.05 0.1 0.25
2,907 2,459 1,989 1,313
735 622 503 333
333 281 227 150
192 162 131 87
126 107 87 58
92 77 63 42
70 59 48 33
57 48 39 26
47 40 33 22
40 34 28 18
35 30 24 17
32 27 22 15
29 24 21 14
27 23 18 12
26 22 17 12
24 21 17 11
23 20 16 11
22 20 16 11
22 18 15 10
21 18 15 10
20 17 14 10
20 17 14 9
0.05 0.05 0.05
0.05 0.1 0.25
2,048 1,620 1,018
518 410 258
234 185 117
136 107 68
89 71 45
65 52 33
50 40 26
40 32 21
34 27 17
29 23 15
26 21 14
23 18 12
21 17 11
20 16 10
18 15 10
17 14 9
17 14 9
16 12 9
16 12 9
15 12 8
15 11 8
14 11 8
0.1 0.1 0.25
0.1 0.25 0.25
1,244 725 345
315 184 88
143 83 40
82 48 23
54 33 16
40 23 11
30 18 9
24 15 8
21 12 6
18 11 5
16 10 5
15 9 5
14 8 4
12 8 4
11 8 4
11 6 4
10 6 4
10 6 4
10 6 4
10 6 3
953
953
NUREG-1575, Supp. 1
A-2
January 2009
-------
MARSAME
Appendix A
Table A.2b Sample Sizes for Wilcoxon Rank Sum Test
(Number of measurements to be performed on the reference material and for each survey unit)
A/o
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.25
2.5
2.75
3.0
3.5
4.0
(«,/?) or (/?,«)
0.01 0.01 0.01 0.01 0.01
0.01 0.025 0.05 0.1 0.25
5,452 4,627 3,972 3,278 2,268
1,370 1,163 998 824 570
614 521 448 370 256
350 297 255 211 146
227 193 166 137 95
161 137 117 97 67
121 103 88 73 51
95 81 69 57 40
77 66 56 47 32
64 55 47 39 27
55 47 40 33 23
48 41 35 29 20
43 36 31 26 18
38 32 28 23 16
35 30 25 21 15
32 27 23 19 14
30 25 22 18 13
28 24 20 17 12
26 22 19 16 11
25 21 18 15 11
22 19 16 14 10
21 18 15 13 9
20 17 15 12 9
19 16 14 12 8
18 16 13 11 8
18 15 13 11 8
0.025 0.025 0.025 0.025
0.025 0.05 0.1 0.25
3,870 3,273 2,646 1,748
973 823 665 440
436 369 298 197
248 210 170 112
162 137 111 73
114 97 78 52
86 73 59 39
68 57 46 31
55 46 38 25
46 39 32 21
39 33 27 18
34 29 24 16
30 26 21 14
27 23 19 13
25 21 17 11
23 19 16 11
21 18 15 10
20 17 14 9
19 16 13 9
18 15 12 8
16 14 11 8
15 13 10 7
14 12 10 7
14 12 10 6
13 11 9 6
13 11 9 6
0.05 0.05 0.05
0.05 0.1 0.25
2,726 2,157 1,355
685 542 341
307 243 153
175 139 87
114 90 57
81 64 40
61 48 30
48 38 24
39 31 20
32 26 16
28 22 14
24 19 12
22 17 11
19 15 10
18 14 9
16 13 8
15 12 8
14 11 7
13 11 7
13 10 7
11 9 6
11 9 6
10 8 5
10 8 5
985
975
0.1 0.1 0.25
0.1 0.25 0.25
1,655 964 459
416 243 116
187 109 52
106 62 30
69 41 20
49 29 14
37 22 11
29 17 8
24 14 7
20 12 6
17 10 5
15 9 4
13 8 4
12 7 4
11 7 3
10 6 3
963
953
853
853
742
742
642
642
642
642
January 2009
A-3
NUREG-1575, Supp. 1
-------
Appendix A
MARSAME
A.3 Critical Values for the Sign Test
Table A.3 Critical Values for the Sign Test Statistic, S+
N
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
Alpha
0.005 0.01 0.025
444
555
665
766
777
887
998
10 9 9
10 10 9
11 11 10
12 11 11
12 12 11
13 13 12
14 13 12
14 14 13
15 14 14
16 15 14
16 16 15
17 16 16
18 17 16
18 18 17
19 18 17
19 19 18
20 19 19
21 20 19
21 21 20
22 21 20
0.05 0.1 0.2
433
443
554
655
665
766
876
887
987
998
10 9 9
11 10 9
11 11 10
12 11 10
12 12 11
13 12 11
14 13 12
14 13 12
15 14 13
15 15 14
16 15 14
17 16 15
17 16 15
18 17 16
18 17 16
19 18 17
19 19 17
0.3 0.4 0.5
322
332
433
443
544
554
655
665
766
776
877
987
998
10 9 8
10 10 9
11 10 9
11 11 10
12 11 10
12 12 11
13 12 11
13 13 12
14 13 12
14 14 13
15 14 13
15 15 14
16 15 14
16 16 15
NUREG-1575, Supp. 1
A-4
January 2009
-------
MARSAME
Appendix A
Table A.3 Critical Values for the Sign Test Statistic, S+ (continued)
N
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Alpha
0.005 0.01 0.025
23 22 21
23 23 22
24 23 22
24 24 23
25 24 23
26 25 24
26 26 24
27 26 25
27 27 26
28 27 26
29 28 27
29 28 27
30 29 28
30 30 28
31 30 29
32 31 30
32 31 30
33 32 31
33 33 31
34 33 32
0.05 0.1 0.2
20 19 18
21 20 18
21 20 19
22 21 19
22 21 20
23 22 21
23 22 21
24 23 22
25 23 22
25 24 23
26 25 23
26 25 24
27 26 24
27 26 25
28 27 25
29 27 26
29 28 26
30 28 27
30 29 27
31 30 28
0.3 0.4 0.5
17 16 15
17 17 16
18 17 16
19 18 17
19 18 17
20 19 18
20 19 18
21 20 19
21 20 19
22 21 20
22 21 20
23 22 21
23 22 21
24 23 22
24 23 22
25 24 23
25 24 23
26 25 24
26 25 24
27 26 25
For TV greater than 50, the table (critical) value can be calculated from:
2 2
(A-l)
where:
i-a = (1-a) percentile of a standard normal distribution (page A-9)
January 2009
A-5
NUREG-1575, Supp. 1
-------
Appendix A
MARSAME
A.4 Critical Values for the WRS Test
The parameter, m, is the number of reference area samples and the parameter, «, is the number of
survey unit samples. When using this table under Scenario A, m is the number of reference area
samples and n is the number of survey unit samples. When using this table for Scenario B, the
roles of TO and n in this table are reversed.
Table A.4 Critical Values for the WRS Test
m
2
3
4
5
6
«
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
n
2345
7 9 11 13
7 9 11 13
7 9 11 13
7 9 11 13
7 9 11 12
7 8 10 11
12 15 18 21
12 15 18 21
12 15 18 21
12 15 18 20
12 14 17 19
11 13 16 18
18 22 26 30
18 22 26 30
18 22 26 29
18 22 25 28
18 21 24 27
17 20 22 25
25 30 35 40
25 30 35 39
25 30 34 38
25 29 33 37
24 28 32 35
23 27 30 34
33 39 45 51
33 39 44 49
33 39 43 48
33 37 42 47
32 36 41 45
31 35 39 43
6 7 8 9 10
15 17 19 21 23
15 17 19 21 23
15 17 19 21 23
15 17 18 20 22
14 16 17 19 21
13 15 16 18 19
24 27 30 33 36
24 27 30 32 35
24 26 29 31 34
22 25 27 30 32
21 24 26 28 31
20 22 24 27 29
34 38 42 46 49
33 37 40 44 47
32 36 39 42 46
31 34 37 41 44
30 33 36 39 42
28 31 34 36 39
45 50 54 58 63
43 48 52 56 60
42 46 50 54 58
41 44 48 52 56
39 43 46 50 53
37 41 44 47 51
57 63 67 72 77
54 59 64 69 74
53 58 62 67 72
51 56 60 64 69
49 54 58 62 66
47 51 55 59 63
11 12 13 14 15
25 27 29 31 33
25 27 29 31 33
25 27 28 30 32
23 25 27 29 31
23 24 26 27 29
21 22 24 26 27
39 42 45 48 51
38 40 43 46 48
37 39 42 45 47
35 37 40 42 45
33 36 38 40 43
31 33 35 37 40
53 57 60 64 68
51 54 58 61 64
49 52 56 59 62
47 50 53 56 59
45 48 51 54 57
42 45 48 50 53
67 72 76 81 85
64 68 72 77 81
62 66 70 74 78
60 63 67 71 75
57 61 64 68 71
54 57 61 64 67
82 88 93 98 103
79 83 88 93 98
77 81 86 91 95
73 78 82 87 91
70 75 79 83 87
67 71 75 79 83
16 17 18 19 20
35 37 39 41 43
35 37 39 40 42
34 36 38 39 41
33 34 36 38 40
31 33 34 36 38
29 30 32 33 35
54 56 59 62 65
51 54 57 59 62
50 52 55 58 60
47 50 52 55 57
45 47 50 52 54
42 44 46 48 50
71 75 78 82 86
68 71 75 78 81
66 69 72 76 79
62 66 69 72 75
59 62 65 68 71
56 59 61 64 67
89 94 98 102 107
85 89 93 97 101
82 86 90 94 98
79 82 86 90 94
75 79 82 86 89
71 74 77 81 84
108 113 118 123 128
103 107 112 117 122
100 104 109 114 118
95 100 104 109 113
91 96 100 104 108
87 91 94 98 102
NUREG-1575, Supp. 1
A-6
January 2009
-------
MARSAME
Appendix A
Table A.4 Critical Values for the WRS Test (continued)
m
1
8
9
10
11
12
13
«
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
«
2345
42 49 56 63
42 49 55 61
42 48 54 59
42 47 52 57
41 46 51 56
40 44 49 54
52 60 68 75
52 60 66 73
52 59 65 71
51 57 63 69
50 56 62 67
49 54 60 65
63 72 81 88
63 71 79 86
63 70 77 84
62 69 76 82
61 67 74 80
60 66 71 77
75 85 94 103
75 84 92 100
75 83 91 98
74 81 89 96
73 80 87 93
71 78 84 91
88 99 109 118
88 98 107 115
88 97 105 113
87 95 103 111
86 93 101 108
84 91 98 105
102 114 125 135
102 112 122 131
102 111 120 129
100 109 118 126
99 108 116 124
97 105 113 120
117 130 141 152
117 128 139 148
116 127 137 146
115 125 134 143
114 123 132 140
112 120 129 137
6 7 8 9 10
69 75 81 87 92
66 72 77 83 88
65 70 76 81 86
63 68 73 78 83
61 65 70 75 80
58 63 67 72 76
82 89 95 102 109
79 85 92 98 104
77 84 90 96 102
75 81 86 92 98
73 78 84 89 95
70 75 80 85 91
96 104 111 118 126
93 100 107 114 121
91 98 105 111 118
88 95 101 108 114
86 92 98 104 110
83 89 94 100 106
111 119 128 136 144
108 115 123 131 138
106 113 121 128 135
103 110 117 124 131
100 107 114 120 127
97 103 110 116 122
127 136 145 154 163
124 132 140 148 157
122 130 138 146 153
118 126 134 141 149
115 123 130 137 144
112 119 126 133 139
145 154 164 173 183
140 149 158 167 176
138 147 156 164 173
135 143 151 159 168
132 140 147 155 165
128 135 143 150 158
163 173 183 193 203
158 168 177 187 196
156 165 174 184 193
152 161 170 179 187
149 157 166 174 183
145 153 161 169 177
11 12 13 14 15
98 104 110 116 122
94 99 105 110 116
92 97 102 108 113
88 93 98 103 108
85 90 94 99 104
81 85 90 94 99
115 122 128 135 141
110 116 122 129 135
108 114 120 125 131
104 109 115 121 126
100 105 111 116 122
96 101 106 111 116
133 140 147 155 162
127 134 141 148 155
125 131 138 144 151
120 126 133 139 145
116 122 128 134 140
112 117 123 129 134
152 160 167 175 183
146 153 160 168 175
142 150 157 164 171
138 145 151 158 165
133 140 147 153 160
128 135 141 147 153
171 180 188 197 206
165 173 181 189 197
161 169 177 185 193
156 164 171 179 186
152 159 166 173 180
146 153 160 167 173
192 202 210 220 230
185 194 202 211 220
181 190 198 207 215
176 184 192 200 208
171 179 186 194 202
165 172 180 187 194
213 223 233 243 253
206 215 225 234 243
202 211 220 229 238
196 205 214 222 231
191 199 208 216 224
185 193 201 209 217
16 17 18 19 20
128 133 139 145 151
121 127 132 138 143
118 123 129 134 139
113 118 123 128 133
109 113 118 123 128
103 108 112 117 121
148 154 161 167 174
141 147 153 159 165
137 143 149 155 161
132 137 143 149 154
127 132 138 143 148
121 126 131 136 141
169 176 183 190 198
161 168 175 182 188
157 164 170 177 184
151 158 164 170 176
146 152 158 164 170
140 145 151 157 162
191 199 207 215 222
183 190 197 205 212
178 186 193 200 207
172 179 186 192 199
166 173 179 186 192
160 166 172 178 184
214 223 231 240 248
205 213 221 229 237
200 208 216 224 232
194 201 208 216 223
187 195 202 209 216
180 187 194 201 207
238 247 256 266 275
228 237 246 254 263
223 232 240 249 257
216 224 232 240 248
209 217 225 233 240
202 209 216 224 231
263 273 282 292 302
253 262 271 280 290
247 256 265 274 283
239 248 257 265 274
233 241 249 257 266
224 232 240 248 256
January 2009
A-7
NUREG-1575, Supp. 1
-------
Appendix A
MARSAME
Table A.4 Critical Values for the WRS Test (continued)
m
14
15
16
17
18
19
20
«
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
0.001
0.005
0.01
0.025
0.05
0.1
«
2345
133 147 159 171
133 145 156 167
132 144 154 164
131 141 151 161
129 139 149 158
128 136 145 154
150 165 178 190
150 162 174 186
149 161 172 183
148 159 169 180
146 157 167 176
144 154 163 172
168 184 197 210
168 181 194 206
167 180 192 203
166 177 188 200
164 175 185 196
162 172 182 192
187 203 218 232
187 201 214 227
186 199 212 224
184 197 209 220
183 194 205 217
180 191 202 212
207 224 239 254
207 222 236 249
206 220 233 246
204 217 230 242
202 215 226 238
200 211 222 233
228 246 262 277
227 243 258 272
226 242 256 269
225 239 252 265
223 236 248 261
220 232 244 256
250 269 286 302
249 266 281 296
248 264 279 293
247 261 275 289
245 258 271 284
242 254 267 279
6 7 8 9 10
182 193 204 215 225
177 187 198 208 218
175 185 194 204 214
171 180 190 199 208
167 176 185 194 203
163 171 180 189 197
202 212 225 237 248
197 208 219 230 240
194 205 215 226 236
190 200 210 220 230
186 196 206 215 225
182 191 200 209 218
223 236 248 260 272
218 229 241 252 264
215 226 237 248 259
210 221 232 242 253
206 217 227 237 247
202 211 221 231 241
245 258 271 284 297
239 252 264 276 288
236 248 260 272 284
232 243 254 266 277
228 238 249 260 271
223 233 243 253 264
268 282 296 309 323
262 275 288 301 313
259 272 284 296 309
254 266 278 290 302
250 261 273 284 295
244 255 266 277 288
292 307 321 335 350
286 300 313 327 340
283 296 309 322 335
278 290 303 315 327
273 285 297 309 321
267 279 290 302 313
317 333 348 363 377
311 325 339 353 367
307 321 335 349 362
302 315 329 341 354
297 310 322 335 347
291 303 315 327 339
11 12 13 14 15
236 247 257 268 278
228 238 248 258 268
224 234 243 253 263
218 227 236 245 255
212 221 230 239 248
206 214 223 231 240
260 271 282 293 304
251 262 272 283 293
247 257 267 278 288
240 250 260 270 280
234 244 253 263 272
227 236 246 255 264
284 296 308 320 332
275 286 298 309 320
270 281 292 303 314
264 274 284 295 305
257 267 278 288 298
250 260 269 279 289
310 322 335 347 360
300 312 324 336 347
295 307 318 330 341
288 299 310 321 332
282 292 303 313 324
274 284 294 305 315
336 349 362 376 389
326 339 351 364 376
321 333 345 357 370
313 325 337 348 360
307 318 329 340 352
299 309 320 331 342
364 377 391 405 419
353 366 379 392 405
348 361 373 386 399
340 352 364 377 389
333 345 356 368 380
325 336 347 358 370
392 407 421 435 450
381 395 409 422 436
376 389 402 416 429
367 380 393 406 419
360 372 385 397 409
351 363 375 387 399
16 17 18 19 20
289 299 310 320 330
278 288 298 307 317
272 282 291 301 311
264 273 282 292 301
257 265 274 283 292
248 257 265 273 282
316 327 338 349 360
304 314 325 335 346
298 308 319 329 339
289 299 309 319 329
282 291 301 310 319
273 282 291 300 309
343 355 367 379 390
331 342 353 365 376
325 336 347 357 368
316 326 337 347 357
308 318 328 338 348
298 308 317 327 336
372 384 397 409 422
359 371 383 394 406
353 364 376 387 399
343 354 365 376 387
335 345 356 366 377
325 335 345 355 365
402 415 428 441 454
388 401 413 425 438
382 394 406 418 430
372 383 395 406 418
363 374 385 396 407
352 363 374 384 395
433 446 460 473 487
419 431 444 457 470
411 424 437 449 462
401 413 425 437 450
392 403 415 427 439
381 392 403 415 426
464 479 493 507 521
450 463 477 490 504
442 456 469 482 495
431 444 457 470 482
422 434 446 459 471
410 422 434 446 458
NUREG-1575, Supp. 1
January 2009
-------
MARSAME
Appendix A
Reject the null hypothesis if the test statistic (Wr) is greater than the table (critical) value.
For n or m greater than 20 with few or no ties, the table (critical) value can be calculated from:
„ . . 1Tr i
Critical Value =
+ z
nm(n
12
(A-2)
If there are ties, the critical value can be calculated from:
Critical Value =
m(n + m +
2
1) Inm
\ " \
\12
(n + m-
i-l^
*-)
Zj j
j=\ (n + m)(n + m + l)
(A-3)
Where:
g
tj
z
number of groups of tied measurements
number of tied measurements in the/h group
(1-a) percentile of a standard normal distribution (see list below)
a
0.001
0.005
0.01
0.025
0.05
0.1
z
3.090
2.575
2.326
1.960
1.645
1.282
Other values for z can be obtained from Table A. 1.
January 2009
A-9
NUREG-1575, Supp. 1
-------
Appendix A
MARSAME
A.5 Critical Values for the Quantile Test
Tables A.5a-d contain values of the parameters r and k needed for the Quantile test calculated by
Gilbert and Simpson (Gilbert 1992) for certain combinations of m (the number of measurements
in the reference area) and n (the number of measurements in the survey unit). The value of a
listed is that obtained from simulation studies,
Table A.5a Values of r and k for the Quantile Test When a Is Approximately 0.01
m
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
SO
85
90
95
100
Number of Survey Unit Measurements, n
5 10 15 20 25
r,k
a
3,3
0.009
6,4
0.005
4,3
0009
4,3
0006
2,2
0.013
2,2
0.01
2,2
0008
r,k
a
6,6
0005
7,6
0.007
4,4
0.008
7,5
0012
3,3
0012
3,3
0.008
3,3
0.006
6,4
0008
4,3
0.013
4,3
0.01
4,3
0008
4,3
0.007
2,2
0.014
2,2
0.013
2,2
0.011
2,2
0.01
11,11
0.008
7,7
0013
6,6
0.008
5,5
0.009
4,4
0015
4,4
0009
4,4
0.006
7,5
0.013
3,3
0013
3,3
0.01
3,3
0.008
3,3
0007
3,3
0.006
6,4
0.008
4,3
0.014
4,3
0.012
4,3
0.01
4,3
0.009
4,3
0.008
4,3
0.007
13,13
0.015
9,9
0012
7,7
0.011
6,6
0.01
5,5
0013
5,5
0007
4,4
0.014
4,4
0.01
4,4
0007
4,4
0.005
7,5
0.013
3,3
0014
3,3
0.012
3,3
0.01
3,3
0.008
3,3
0.007
3,3
0.006
3,3
0.005
6,4
0.008
4,3
0.014
16,16
0.014
11,11
0011
8,8
0.014
7,7
0.011
6,6
0011
6,6
0006
5,5
0.01
5,5
0.006
4,4
0014
4,4
0.01
4,4
0.008
4,4
0006
6,5
0.006
7,5
0.013
3,3
0.014
3,3
0.012
3,3
0.011
3,3
0.009
3,3
0.008
3,3
0.007
30 35 40 45 50
19,19
0.013
13,13
001
10,10
0.009
8,8
0.011
7,7
001
6,6
0012
6,6
0.007
5,5
0.012
5,5
0008
5,5
0.006
4,4
0.014
4,4
0011
4,4
0.009
4,4
0.007
4,4
0.006
6,5
0.006
7,5
0.013
3,3
0.014
3,3
0.013
3,3
0.011
22,22
0.013
14,14
0014
11,11
0.011
9,9
0.011
8,8
0009
7,7
001
6,6
0.012
6,6
0.008
5,5
0014
5,5
0.01
5,5
0.007
5,5
0006
4,4
0.013
4,4
0.011
4,4
0.009
4,4
0.008
4,4
0.006
4,4
0.005
6,5
0.005
7,5
0.013
25,25
0.013
16,16
0013
12,12
0.013
10,10
0.011
9,9
0009
80
,o
0008
7,7
0.009
6,6
0.013
6,6
0009
5,5
0.015
5,5
0.011
5,5
0009
5,5
0.007
5,5
0.005
4,4
0.013
4,4
0.011
4,4
0.009
4,4
0.008
4,4
0.007
4,4
0.006
28,28
0.012
18,18
0012
13,13
0.014
11,11
0.011
9,9
0014
8,8
0013
7,7
0.014
7,7
0.009
6,6
0013
6,6
0.009
6,6
0.007
5,5
0013
5,5
0.01
5,5
0.008
5,5
0.006
5,5
0.005
4,4
0.013
4,4
0.011
4,4
0.01
4,4
0.008
19,19
0015
15,15
0.011
12,12
0.011
10,10
0012
9,9
0011
8,8
0.011
7,7
0.013
7,7
0009
6,6
0.013
6,6
0.01
6,6
0007
5,5
0.014
5,5
0.011
5,5
0.009
5,5
0.007
5,5
0.006
5,5
0.005
4,4
0.013
4,4
0.011
55 60 65 70 75
21,21
0014
16,16
0.012
13,13
0.011
11,11
0011
10,10
0009
9,9
0.009
8,8
0.01
7,7
0013
7,7
0.009
6,6
0.014
6,6
001
6,6
0.008
5,5
0.015
5,5
0.012
5,5
0.01
5,5
0.008
5,5
0.007
5,5
0.006
4,4
0.015
23,23
0013
17,17
0.013
14,14
0.012
12,12
0011
10,10
0013
9,9
0.013
8,8
0.014
8,8
0009
7,7
0.012
7,7
0.009
6,6
0014
6,6
0.011
6,6
0.008
6,6
0.007
5,5
0.013
5,5
0.011
5,5
0.009
5,5
0.008
5,5
0.007
25,25
0012
18,18
0.014
15,15
0.012
12,12
0015
11,11
0011
10,10
0.01
9,9
0.011
8,8
0012
8,8
0.009
7,7
0.012
7,7
0009
6,6
0.014
6,6
0.011
6,6
0.009
6,6
0.007
5,5
0.014
5,5
0.012
5,5
0.01
5,5
0.009
26,26
0015
19,19
0.015
16,16
0.012
13,13
0014
1211
0014
10,10
0.014
9,9
0.014
9,9
0009
8,8
0.011
8,8
0.008
7,7
0011
7,7
0.009
6,6
0.014
6,6
0.011
6,6
0.009
6,6
0.008
5,5
0.015
5,5
0.013
5,5
0.011
28,28
0014
21,21
0.012
17,17
0.012
14,14
0013
12,12
0013
11,11
0.011
10,10
0.011
9,9
0012
8,8
0.014
8,8
0.01
7,7
0014
7,7
0.011
7,7
0.009
6,6
0.014
6,6
0.012
6,6
0.01
6,6
0.008
6,6
0.007
5,5
0.013
80 85 90 95 100
30,30
0013
22,22
0.013
18,18
0.012
15,15
0012
13,13
0012
11,11
0.015
10,10
0.014
10,10
0009
9,9
0.011
8,8
0.013
8,8
001
7,7
0.014
7,7
0.011
7,7
0.009
6,6
0.014
6,6
0.012
6,6
0.01
6,6
0.008
6,6
0.007
23,23
0.014
19,19
0.012
16,16
0011
14,14
0011
12,12
0.012
11,11
0.012
10,10
0012
9,9
0.013
9,9
0.009
8,8
0012
8,8
0.009
7,7
0.013
7,7
0.011
7,7
0.009
6,6
0.014
6,6
0.012
6,6
0.01
6,6
0.008
24,24
0.015
19,19
0.015
16,16
0014
14,14
0014
13,13
0.011
11,11
0.014
10,10
0015
10,10
0.01
9,9
0.012
8,8
0015
8,8
0.011
8,8
0.009
7,7
0.013
7,7
0.01
7,7
0.008
6,6
0.014
6,6
0.012
6,6
0.01
26,26
0.013
20,20
0.015
17,17
0014
15,15
0012
13,13
0.013
12,12
0.012
11,11
0012
10,10
0.012
9,9
0.014
9,9
001
8,8
0.014
8,8
0.011
8,8.
0.008
7,7
0.013
7,7
0.01
7,7
0.008
6,6
0.014
6,6
0.012
r,k
a
27,27
0.013
21,21
0.015
18,18
0013
15,15
0015
14,14
0.012
12,12
0.014
11,11
0014
10,10
0.015
10,10
0.011
9,9
0013
9,9
0.01
8,8
0.013
8,8
0.01
7,7
0.015
7,7
0.012
7,7
0.019
7,7
0.008
6,6
0.014
NUREG-1575, Supp. 1
A-10
January 2009
-------
MARSAME
Appendix A
Table A.5b Values of r and k for the Quantile Test When a Is Approximately 0.025
m
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
SO
85
90
95
100
Number of Survey Unit Measurements, «
5 10 15 20 25
r,k
a
11,5
0.03
8,4
0023
2,2
0.023
6,3
0.026
7,3
003
3,2
0029
3,2
0.023
r,k
a
7,6
0.029
6,5
0.023
3,3
003
8,5
0.027
6,4
0.026
4,3
003
4,3
0022
8,4
0.029
2,2
0.025
2,2
0022
14,5
0022
6,3
0.028
6,3
0.024
11,4
0022
7,3
0028
3,2
0.029
9,9
0.03
6,6
0.028
5,5
0.021
4,4
0026
6,5
0.021
9,6
0.026
3,3
0023
8,5
0028
6,4
0.036
6,4
0.022
4,3
0029
4,3
0024
7,4
0.021
2,2
0.029
2,2
0026
2,2
0024
2,2
0.021
5,3
0.02
10,4
0.029
6,3
0.029
12,12
0.024
8,8
0.022
6,6
0.024
5,5
0024
7,6
6.023
4,4
0.021
6,5
002
11,7
0025
3,3
0.026
3,3
0.021
8,5
0028
8,5
0021
6,4
0.025
6,4
0.021
4,3
0028
4,3
0024
4,3
0.021
11,5
0.027
2,2
0.029
2,2
0.027
15,15
0.021
9,9
0.029
7,7
0.026
6,6
0022
5,5
0.025
7,6
0.029
4,4
0026
6,5
0028
8,6
0.021
11,7
0.077
3,3
0028
3,3
0023
10,6
0.025
8,5
0.028
8,5
0022
6,4
0028
6,4
0.023
9,5
0.023
4,3
0.028
4,3
0.025
30 35 40 45 50
17,17
0.026
11,11
0.024
8,8
0.027
7,7
002
6,6
0.02
5,5
0.026
10,8
0022
4,4
003
4,4
0.023
6,5
6.026
8,6
0021
11,7
0029
3,3
0.029
3,3
0.025
3,3
0022
10,6
0024
8,5
0.028
8,5
0.023
6,4
0.029
6,4
0.025
20,20
0.024
12,12
0.029
9,9
0.028
12,11
0021
10,9
0.026
9,8
0.024
5,5
0027
10,8
0026
7,6
0.025
4,4
0.026
4,4
002
6,5
0024
8,6
0.021
13,8
0.026
9,6
0028
3,3
0027
3,3
0.023
3,3
0.021
10,6
0.023
8,5
0.028
22,22
0.028
14,14
0.025
10,10
0.029
13,12
0024
7,7
0.027
6,6
0.029
9,8
0024
5,5
0027
5,5
0.02
7,6
0.028
4,4
0029
4,4
0023
6,5
0.029
6,5
0.023
8,6
0021
13,8
0027
9,6
0.03
3,3
0.028
3,3
0.025
3,3
0.022
25,25
0.025
17,17
0.025
11,11
0.03
9,9
0028
8,8
0.023
7,7
0.023
6,6
0027
9,8
0023
5,5
0.028
5,5
0.021
10,8
0021
7,6
0023
4,4
0.026
4,4
0.022
6,5
0027
6,5
0023
8,6
0.02
13,8
0.028
11,7
0.026
3,3
0.029
18,18
0.029
13,13
0.022
10,10
0026
13,12
0.027
12,11
0.021
7,7
002
6,6
0026
9,8
0.023
5,5
0.028
5,5
0022
10,8
0024
7,6
0.026
4,4
0.028
4,4
0024
4,4
002
6,5
0.026
6,5
0.022
8,6
0.02
13,8
0.028
55 60 65 70 75
20,20
0.026
15,15
0.023
11,11
0024
9,9
0.027
8,8
0.025
7,7
0027
10,9
0028
6,6
0.024
9,8
0.022
5,5
0028
5,5
0023
10,8
0.026
7,6
0.028
7,6
0023
4,4
0026
4,4
0.022
6,5
0.029
6,5
0.025
6,5
0.022
21,21
0.029
14,14
0.023
12,12
0023
10,10
0.024
9,9
0.021
8,8
0021
7,7
0024
10,9
0.026
6,6
0.023
9,8
0022
5,5
0029
5,5
0.023
10,8
0.027
7,6
003
7,6
0024
4,4
0.028
4,4
0.024
4,4
0.021
6,5
0.028
23,23
0.026
16,16
0.024
13,13
0022
11,11
0.022
9,9
0.027
8,8
0027
12,11
002
7,7
0.022
6,6
0.029
6,6
0092
9,8
0022
5,5
0.029
5,5
0.024
10,8
0029
10,8
0023
7,6
0.026
4,4
0.029
4,4
0.026
4,4
0.023
24,24
0.029
17,17
0.025
13,13
0029
11,11
0.028
10,10
0.023
9,9
0022
80
,o
0023
7,7
0.027
7,7
0.02
6,6
0028
6,6
0022
9,8
0.022
5,5
0.029
5,5
0024
5,5
007
10,8
0.024
7,6
0.028
7,6
0.024
4,4
0.027
26,26
0.026
18,18
0.025
14,14
0027
12,12
0.025
10,10
0.029
9,9
0027
8,8
0029
8,8
0.02
7,7
0.025
10,9
0029
6,6
0027
6,6
0.021
9,8
0.022
5,5
0029
5,5
0025
5,5
0.021
10,8
0.026
7,6
0.029
7,6
0.025
80 85 90 95 100
27,27
0.029
19,19
0.026
15,15
0026
13,13
0.823
11,11
0.025
10,10
0022
9,9
0022
8,8
0.025
12,11
0.02
7,7
0023
10,9
0027
6,6
0.026
6,6
0.021
9,8
0021
5,5
0029
5,5
0.025
5,5
0.022
10,8
0.027
10,8
0.022
21,21
0.021
16 16
0025
13,13
0.628
11,11
0.03
10,10
0027
9,9
0027
8,8
0.03
8,8
0.022
7,7
0027
7,7
0021
10,9
0.026
6,6
0.025
6,6
0021
9,8
0021
5,5
0.029
5,5
0.025
5,5
0.022
10,8
0.028
21,21
0.027
17,17
0024
14,14
0.025
12,12
0.026
11,11
0022
10,10
0021
9,9
0.023
8,8
0.026
12,11
0023
7,7
0025
7,7
0.020
6,6
0.029
6,6
0024
6,6
002
9,8
0.021
5,5
0.03
5,5
0.026
5,5
0.022
22,22
0.027
17,17
0029
15,15
0.023
13,13
0.023
11,11
0027
10,10
0026
9,9
0.027
13,12
0.027
8,8
0023
7,7
003
7,7
0.024
10,9
0.03
6,6
0028
6,6
0024
6,6
0.02
9,8
0.021
5,5
0.03
5,5
0.026
r,k
a
23,23
0.027
18,18
0028
15,15
0.028
13,13
0.027
12,12
0023
11,11
0021
10,10
0.021
9,9
0.023
8,8
0027
8,8
0021
7,7
0.028
7,7
0.022
10,9
0028
6,6
0027
6,6
0.023
9,8
0.025
9,8
0.021
5,5
0.03
January 2009
A-ll
NUREG-1575, Supp. 1
-------
Appendix A
MARSAME
Table A.5c Values of r and k for the Quantile Test When a Is Approximately 0.05
in
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
SO
85
90
95
100
Number of Survey Unit Measurements, n
5 10 15 20 25
r,k
a
2,2
0.053
9,4
0.04
6,3
0.041
3,2
0.047
8,3
0046
4,2
0055
4,2
0.045
r,k
a
4,4
0.043
3,3
0.052
8,5
0.056
6,4
0.043
2,2
0.058
2,2
0045
5,3
0048
9,4
0.047
6,3
0.051
3,2
0059
3,2
0.052
.3,2
0.045
8,3
0.057
8,3
0049
4,2
0059
4,2
0.054
8,8
0.051
5,5
0.057
4,4
0.05
6,5
0.04
3,3
0.046
10,6
0.052
6,4
0058
4,3
0057
2,2
0.059
2,2
0.05
2,2
0043
5,3
0.052
5,3
0.043
9,4
0.048
6,3
0056
6,3
0048
3,2
0.058
3,2
0.053
3,2
0.048
3,2
0.044
10,10
0.057
14,12
0.045
5,5
0.048
4,4
0.053
6,5
0.052
3,3
0.058
3,3
0043
10,6
0059
8,5
0.052
6,4
0.051
4,3
0056
4,3
0.046
2,2
0.053
2,2
0.047
2,2
0043
5,3
0053
5,3
0.047
5,3
0.041
9,4
0.048
6,3
0.057
13 13
0.043
8,8
0.046
6,6
0.046
5,5
0.043
4,4
0.055
11,8
0.045
6,5
0041
3,3
0053
3,3
0.042
12,7
0.05
8,5
0058
6,4
0.059
6,4
0.048
4,3
0.055
4,3
0047
2,2
0055
2,2
0.05
2,2
0.046
2,2
0.042
5,3
0.054
30 35 40 45 50
15 15
0.048
9,9
0.052
7,7
0.045
9,8
0.052
5,5
0.041
4,4
0.056
4,4
004
6,5
0048
8,6
0.041
3,3
0.049
3,3
0041
3,3
0.035
10,6
0.05
8,5
0.05
6,4
0054
6,4
0046
4,3
0.054
6,4
0.059
2,2
0.056
2,2
0.052
17,17
0.051
10,10
0.058
8,8
0.052
6,6
0.056
5,5
0.059
8,7
0.044
4,4
0057
4,4
0043
6,5
0.054
8,6
0.049
5,4
0041
3,3
0.047
3,3
0.04
5,4
0.041
10,6
0053
8,5
0055
4,3
0.048
6,4
0.051
4,3
0.059
4,3
0.053
19,19
0.054
12,12
0.046
9,9
0.043
7,7
48
6,6
0.046
5,5
0.054
8,7
0043
4,4
0058
4,4
0.045
6,5
0.059
6,5
0046
8,6
0.043
3,3
0.052
3,3
0.046
3,3
004
5,4
0041
10,6
0.056
8,5
0.058
8,5
0.05
6,4
0.056
21,21
0.056
13,13
0.05
9,9
0.06
8,8
0043
11,10
0.042
6,6
0.04
5,5
0051
8,7
0042
4,4
0.058
4,4
0.047
9,7
0042
6,5
51
6,5
0.041
3,3
0.057
3,3
0051
3,3
0045
5,4
0.049
5,4
0.042
10,6
0.058
10,6
0.049
14,14
0.054
10,10
0.057
8,8
0057
7,7
0.05
6,6
0.053
9,8
0052
5,5
0048
8,7
0.041
4,4
0.059
4,4
0048
9,7
0046
6,5
0.055
6,5
0.045
8,6
0044
3,3
0055
3,3
0.049
3,3
0.044
5,4
0.048
5,4
0.043
55 60 65 70 75
15,15
0.057
11,11
0.055
9,9
0051
8,8
0.042
7,7
0.041
6,6
0047
9,8
0047
5,5
0.046
8,7
0.041
4,4
0059
4,4
0049
4,4
0.042
6,5
0.058
6,5
0049
6,5
0041
3,3
0.059
3,3
0.053
3,3
0.048
3,3
0.043
17,17
0.049
12,12
0.054
10,10
0046
8,8
0.053
7,7
0.052
6,6
0057
6,6
0042
5,5
0.057
5,5
0.045
8,7
004
4,4
0059
4,4
0.05
4,4
0.043
9,7
0041
6,5
0052
6,5
0.044
8,6
0.045
3,3
0.056
3,3
0.051
18,18
0.052
13,13
0.052
10,10
0057
9,9
0.045
8,8
0.042
7,7
0043
6,6
0051
9,8
0.056
5,5
0.054
5,5
0043
13,10
0052
4,4
0.06
4,4
0.051
4,4
0044
9,7
0043
6,5
0.055
6,5
0.047
6,5
0.041
3,3
0.059
19,19
0.055
14,14
0.051
11,11
0052
9,9
0.055
8,8
0.051
7,7
0053
11,10
0042
6,6
0.047
9,8
0.051
5,5
0052
5,5
0042
13,10
0.052
4,4
0.06
4,4
0052
4,4
0045
9,7
0.046
6,5
0.058
6,5
0.05
6,5
0.044
20,20
0.057
15,15
0.05
12,12
0048
10,10
0.048
9,9
0.042
8,8
0041
7,7
0045
6,6
0.055
6,6
0.043
9,8
0048
5,5
005
5,5
0.041
13,10
0.051
5,5
006
4,4
0053
4,4
0.046
4,4
0.041
9,7
0.040
6,5
0.053
80 85 90 95 100
21,21
0.059
16,16
0.049
12,12
0057
11,11
0.042
9,9
0.05
8,8
0049
7,7
0053
11,10
0.046
6,6
0.05
6,6
004
5,5
0058
5,5
0.048
5,5
0.041
13,10
0051
7,6
0058
4,4
0.053
4,4
0.047
4,4
0.042
9,7
0.042
23,23
0.053
16,16
0.058
13,13
0053
11,11
0.05
9,9
0.059
8,8
0057
80
,o
0041
7,7
0.047
6,6
0.058
6,6
0047
9,8
0054
5,5
0.055
5,5
0.047
8,7
0047
13,10
0051
7,6
0.059
4,4
0.054
4,4
0.048
4,4
0.043
17,17
0.057
14,14
0049
11,11
0.058
10,10
0.049
9,9
0046
8,8
0048
7,7
0.054
7,7
0.041
6,6
0054
6,6
0044
9,8
0.051
5,5
0.054
5,5
0046
87
0046
10,8
0.06
7,6
0.059
4,4
0.054
4,4
0.049
18,18
0.056
14,14
0057
12,12
0.052
10,10
0.057
9,9
0053
80
,0
0055
8,8
0.041
7,7
0.048
11,10
0043
6,6
005
6,6
0.041
9,8
0.048
5,5
0052
5,5
0045
8,7
0.045
10,8
0.06
7,6
0.59
4,4
0.055
r,k
a
19,19
0.055
15,15
0054
12,12
0.06
11,11
0.049
10,10
0044
9,9
0043
8,8
0.047
7,7
0.054
7,7
0043
6,6
0056
6,6
0.047
9,8
0.057
5,5
0058
5,5
0051
5,5
0.044
8,7
0.041
10,8
0.059
7,6
0.059
NUREG-1575, Supp. 1
A-12
January 2009
-------
MARSAME
Appendix A
Table A.5d Values of r and k for the Quantile Test When a Is Approximately 0.10
m
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
SO
85
90
95
100
Number of Survey Unit Measurements, «
5 10 15 20 25
r,k
a
9,4
0.098
3,2
0091
4,2
0.119
4,2.
0.089
5,2
0 109
5,2
0087
6,2
0.103
r,k
a
3,3
0.105
10,6
0.106
2,2
0 103
7,4
0.084
5,3
0.089
3,2
0 119
3,2
0098
3,2
0.082
7,3
0.083
4,2
0 109
4,2
0.095
4,2
0.084
5,2
0.115
5,2
103
5,2
0093
5,2
0.084
7,7
0.083
4,4
0.108
3,3
0.112
5,4
0093
8,5
0.112
2,2
0.106
2,2
0086
5,3
0 119
5,3
0.094
9,4
0.115
3,2
0 114
3,2
0 1
3,2
0.089
7,3
0.101
7,3
0088
4,2
0 116
4,2
0.106
4,2
0.097
4,2
0.089
4,2
0.082
8,8
0.116
5,5
0.109
4,4
0.093
3,3
0 115
3,3
0.08
14,8
0.111
6,4
0 12
2,2
0 107
2,2
0.091
7,4
0.097
5,3
0 114
5,3
0097
5,3
0.082
9,4
0.106
3,2
0 111
3,2
0 101
3,2
0.092
3,2
0.085
7,3
100
7,3
0.09
10,10
0.109
6,6
0.109
5,5
0.081
4,4
0085
3,3
0.117
3,3
0.088
5,4
0091
12,7
0 109
6,4
0.115
2,2
0.108
2,2
0095
2,2
0084
7,4
0.090
5,3
0.112
5,3
0098
5,3
0086
9,4
117
3,2
0.119
3,2
0.11
3,2
0.102
30 35 40 45 50
12,12
0.104
7,7
0.109
5,5
0.117
4,4
0 119
4,4
0.08
3,3
0.119
3,3
0093
5,4
0 102
7,5
0.086
10,6
0.112
6,4
0 112
2,2
0 109
2,2
0.097
2,2
0.088
7,4
0 101
7,4
0086
5,3
0.111
5,3
0.099
5,3
0.089
5,3
0.08
14,14
0.1
8,8
0.109
6,6
0.102
5,5
0093
4,4
0.107
9,7
0.116
3,3
0 12
3,3
0097
5,4
0.112
5,4
0.09
14,8
0 111
8,5
0 119
6,4
0.11
2,2
0.109
2,2
0099
2,2
009'
2,2
0.083
7,4
0.095
7,4
0.084
5,3
0.109
15,15
0.117
9,9
0.109
7,7
0.092
10,9
0084
8,7
0.108
4,4
0.1
9,7
0 112
6,5
0 100
3,3
0.1
3,3
0.084
5,4
0098
5,4
0082
12,7
0.113
8,5
0.114
2,2
0 119
2,2
0 109
2,2
0.101
2,2
0.093
2,2
0.086
2,2
0.08
17,17
0.112
10,10
0.109
7,7
0.118
6,6
0099
5,5
0.101
8,7
0.093
4,4
0094
9,7
0 109
6,5
0.101
3,3
0.103
3,3
0088
5,4
0 105
5,4
0.089
7,5
0.081
10,6
0 117
8,5
0 111
2,2
0.118
2,2
0.109
2,2
0.102
2,2
0.095
11,11
0.109
8,8
0.106
7,7
0083
10,9
0.088
5,5
0.088
4,4
0 114
4,4
009
9,7
0.107
6,5
0.102
3,3
0 104
3,3
0091
5,4
0.111
5,4
0.096
5,4
0083
14,8
0 11
10,6
0.112
8,5
0.108
2,2
0.117
2,2
0.11
55 60 65 70 75
12,12
0.109
9,9
0.098
7,7
0 102
6,6
0.096
5,5
0.106
8,7
0 107
4,4
0 107
4,4
0.087
9,7
0.105
6,5
0 103
3,3
0 106
3,3
0.093
3,3
0.083
5,4
0 102
5,4
0089
7,5
0.084
12,7
0.114
10,6
0.08
6,4
0.118
13,13
0.109
9,9
0.118
8,8
0088
6,6
0.114
6,6
0.08
5,5
0094
8,7
0097
4,4
0.102
4,4
0.084
9,7
0 104
6,5
0 103
3,3
0.108
3,3
0.096
3,3
0085
5,4
0 107
5,4
0.094
5,4
0.083
14,8
0.117
12,7
0.109
14,14
0.109
10,10
0.109
8,8
0 105
7,7
0.093
6,6
0.095
5,5
0 11
5,5
0086
4,4
0.117
4,4
0.098
4,4
0082
9,7
0 102
6,5
0.104
3,3
0.109
3,3
0098
3,3
0088
5,4
0.111
5,4
0.099
5,4
0.088
7,5
0.086
15,15
0.109
11,11
0.101
9,9
0092
7,7
0.108
6,6
0.11
6,6
0081
5,5
0099
8,7
0.107
4,4
0.112
4,4
0095
4,4
0081
9,7
0.101
6,5
0.104
3,3
0 11
3,3
0099
3,3
0.09
3,3
0.082
5,4
0.103
5,4
0.093
16,16
0.109
11,11
0.118
9,9
0 107
8,8
0.091
7,7
0.087
6,6
0094
5,5
0 112
5,5
0.091
8,7
0.099
4,4
0 107
4,4
0092
7,6
0.084
9,7
0.191
6,5
0 105
3,3
0 111
3,3
0.101
3,3
0.092
3,3
0.084
5,4
0.08
80 85 90 95 100
1712
0.109
12,12
0.11
10,10
0095
8,8
0.104
7,7
0.1
6,6
0 107
6,6
0082
5,5
0.103
5,5
0.084
4,4
0 12
4,4
0 103
4,4
0.09
7,6
0.082
9,7
0.1
6,5
0 105
3,3
0.112
3,3
0.102
3,3
0.094
3,3
0.086
18,18
0.109
13,13
0.104
10,11
0 108
8,8
0.117
7,7
0.113
6,6
0 12
6,6
0093
5,5
0.115
5,5
0.95
8,7
0 107
4,4
0 115
4,4
0.1
4,4
0.088
7,6
0.081
6,5
0 12
6,5
0.105
3,3
0.113
3,3
0.103
3,3
0.095
13,13
0.118
11,11
0098
9,9
0.1
8,8
0.092
7,7
0094
6,6
0 104
6,6
0.083
5,5
0.105
5,5
0088
8,7
0.1
4,4
0.11
4,4
0.097
4,4
0.086
9,7
0 116
6,5
0.119
6,5
0.105
3,3
0.113
3,3
0.104
14,14
0.111
11,11
0 11
9,9
0.112
8,8
0.103
7,7
0 105
6,6
0 116
6,6
0.093
5,5
0.116
5,5
0098
5,5
0.083
8,7
0.094
4,4
0.107
4,4
0.095
4,4
0084
9,7
0.114
6,5
0.119
6,5
0.106
3,3
0.114
r,k
a
15,15
0.106
12,12
0 1
10,10
0.098
8,8
0.115
7,7
0 116
7,7
0089
6,6
0.103
6,6
0.083
5,5
0 108
5,5
0.092
8,7
0.107
4,4
0.117
4,4
0.104
4,4
0093
4,4
0.083
9,7
0.113
6,5
0.118
6,5
0.106
January 2009
A-13
NUREG-1575, Supp. 1
-------
-------
MARSAME Appendix B
B. SOURCES OF BACKGROUND RADIOACTIVITY
B.I Introduction
Background radioactivity can complicate the disposition decision for M&E. Background
radioactivity may be the result of environmental radioactivity, inherent radioactivity, instrument
noise, or some combination of the three. Special consideration is given to issues associated with
technologically enhanced naturally occurring radioactive materials (TENORM) and orphan
sources as contributors to background. The planning team should consider these potential
sources of background activity and determine what effect, if any, they may have on the design of
the disposition survey.
Information on background radioactivity can be obtained from many sources, including:
• The Nuclear Regulatory Commission (NRC) provides information concerning background
radioactivity in Background as a Residual Radioactivity Criterion for Decommissioning
NUREG-1501 (NRC 1994).
• The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR)
has published a report on Sources and Effects of Ionizing Radiation (UNSCEAR 2000) and
provides a searchable version of the report on the World Wide Web at www.unscear.org.
• The National Council on Radiation Protection and Measurements (NCRP) has published
reports on Exposure of the Population in the United States and Canada from Natural
Background Radiation, NCRP Report No. 94 (NCRP 1988a) and Radiation Exposure of the
U.S. Population from Consumer Products and Miscellaneous Sources, NCRP Report No. 95
(NCRP 1988b).
B.2 Environmental Radioactivity
Environmental radioactivity is radioactivity from the environment where the M&E is located.
There are three sources contributing to environmental radioactivity; terrestrial (Section B.2.1),
manmade (Section B.2.2), and cosmic and cosmogenic (Section B.2.3). Although background
radiation is present everywhere, the component radionuclide concentrations and distributions are
not constant. Certain materials have higher concentrations of background radiation, and varying
environmental and physical conditions can result in accumulations of background radiation.
Information on environmental radioactivity is usually available from historic measurements
identified during the initial assessment (IA).
If high levels of environmental radioactivity interfere with the disposition decision (e.g., action
level less than environmental background, variability in environmental radioactivity determines
level of survey effort), the planning team may consider moving the M&E being investigated to a
location with less environmental radioactivity (see Sections 3.3.1.3 and 5.3). If the level of
environmental radioactivity is unknown, it may be necessary to collect data during a preliminary
survey (see Section 2.3) to provide this information.
January 2009 B-l NUREG-1575, Supp. 1
-------
Appendix B
MARSAME
B.2.1 Terrestrial Radioactivity
The naturally occurring forms of radioactive elements incorporated into the Earth during its
formation that is still present are referred to as "terrestrial radionuclides." The most significant
terrestrial radionuclides include the uranium and thorium decay series, potassium-40 and
rubidium-87. Virtually all materials found in nature contain some concentration of terrestrial
radionuclides. Table B.I lists average and typical ranges of concentrations of terrestrial
radionuclides. Although the ranges in the table are typical, larger variations exist in certain areas
(e.g., Colorado).
Bulk materials containing elevated concentrations of terrestrial radionuclides as well as
equipment used to handle or process these materials should be identified during the IA even if
these materials and equipment were not impacted by site activities.
Radon is an element that occurs as a gas in nature. Isotopes of radon are members of both the
uranium and thorium natural decay series. These radon isotopes decay to produce additional
radioactive isotopes, which are collectively called radon progeny.
Table B.I Typical Average Concentration Ranges of Terrestrial Radionuclides
Material
Soil, U.S.
Phosphate Fertilizer
Concrete
Concrete Block
Brick
Coal Tar
Fly Ash-Bottom Ash
Coal, U.S.
Tile
Porcelain, Glazed
Ceramic, Glazedb
Radium-226
(Bq/kg)a
40 (8-160)b
200C - 100,000d
(19-89)e
(41-780)e
(4-180)e
(100-300)e
200e
~
~
~
Uranium-238
(Bq/kg)a
35 (4-140)b
200-l,500b
(19-89/
(41-780/
(4-180)f
(100-300)b
200b
18 (l-540)g
(550-8 10)h
Thorium-232
(Bq/kg)a
35 (4-130)b
20b
(15-120/
(37-81/
(l-140)f
200b
21 (2-320)g
650h
(180-37,000)h'1
Potassium-40
(Bq/kg)a
370 (100-700)b
~
(260-1, 100)f
(290-1, 100)f
(7-l,200)f
~
52 (l-710)g
~
~
(79-l,200)h)1
i To convert Bq/kg to pCi/g, multiply by 0.027.
3 UNSCEAR, Sources and Effects of Ionizing Radiation (UNSCEAR 2000).
: U.S. Environmental Protection Agency (EPA), 2000. Evaluation of EPA 's Guidelines for Technologic
Enhanced Naturally Occurring Radioactive Materials (TENORM).
d National Academy of Sciences (NAS). 1999. Evaluation of Guidelines for Technologically Enhanced
Naturally Occurring Radioactive Materials (TENORM), Committee on Evaluation of EPA Guidelines for
Exposure to Naturally Occurring Radioactive Materials Board on Radiation Effects Research Commission
on Life Sciences National Research Council, National Academy Press, p. 72.
e 226Ra is assumed to be in secular equilibrium with 238U.
f Eicholz G.G., Clarke F.J., and Kahn, B., 1980. Radiation Exposure From Building Materials, in "Natural
Radiation Environment III," U.S. Department of Energy CONF-780422.
g Beck H.L., Gogolak C.V., Miller K.M., and Lowder W.M., 1980. Perturbations on the Natural Radiation
Environment Due to the Utilization of Coal as an Energy Source, in "Natural Radiation Environment III,"
U.S. Department of Energy CONF-780422.
h Hobbs T.G., 2000. Radioactivity Measurements on Glazed Ceramic Surfaces, J. Res. Natl. Inst. Stand.
Technol. 105,275-283.
i Values reported as total radioactivity without identification of specific radionuclides.
NUREG-1575, Supp. 1
B-2
January 2009
-------
MARSAME Appendix B
Radon emissions vary significantly over time based on a wide variety of factors. For example,
relatively small changes in the relative pressure between the source material and the atmosphere
(indoor or outdoor) can result in large changes in radon concentrations in the air. Soil moisture
content also has an affect on the radon emanation rate.
Radon progeny tend to become fixed to solid particles in the air. These particles can become
attached to surfaces as a result of electrostatic charge or gravitational settling. Air flow through
ventilation ducts can produce an electrostatic charge that will attract these particles. A decrease
in atmospheric pressure often precedes a rainstorm, which increases the radon emanation rate.
Immediately prior to an electrical storm, an electrostatic charge can build up on equipment
resulting in elevated radiation levels from radon progeny. Rainfall acts to scavenge these
particles from the air, potentially resulting in elevated dose rates and surface activities during and
immediately following rainfall.
Pb-210 is a decay product of 222Rn and 238U. The 22-year half-life provides opportunities for
buildup 210Pb and progeny in sediments and low-lying areas. As mentioned previously, rain acts
to scavenge radon progeny from the air. Areas where rain collects and concentrates can result in
elevated levels of 210Pb and progeny over time. In addition, lead is easily oxidized and can
become fixed to surfaces through corrosion processes. Rust or oxide films on equipment can be
indicators of locations with a potential for elevated background radioactivity.
B.2.2 Anthropogenic Radioactive Materials
Nuclear weapons testing and nuclear power reactors have produced large quantities of
radionuclides through the fissioning of uranium and other heavy elements and the activation of
various elements. Examples of anthropogenic radionuclides that could be in the environment are
137Cs, 90Sr, and various isotopes of plutonium.
Prior to the 1963 Limited Test Ban Treaty, fallout from atmospheric nuclear tests distributed
large quantities of anthropogenic radionuclides around the globe. Following the 1963 treaty most
nuclear weapons tests were conducted underground, although China and France continued
atmospheric testing of nuclear weapons into the late 1970s. In 1996 a Comprehensive Test Ban
Treaty was negotiated with the help of the United Nations. The Comprehensive Test Ban Treaty
has not been ratified by China or the United States and was broken by Pakistan and India in
1998. However, worldwide fallout concentrations have been declining since the mid 1960s.
In 1964 a Department of Defense weather satellite containing a radiation source failed to achieve
orbit. The Space Nuclear Auxiliary Power (SNAP) 9-A Radioisotopic Thermoelectric Generator
(RTG) burned up on re-entry and dispersed the nuclear inventory (primarily plutonium-238) into
the atmosphere. Incidents involving Soviet satellites with radioisotopes or nuclear reactors
occurred in 1969, 1973, 1978, and 1983. In April 1986 there was a non-nuclear steam explosion
and fire at the number four reactor at the nuclear power plant in Chernobyl in north-central
Ukraine. Large quantities of radioactive material were released into the environment as a result
of the catastrophe. The radionuclides from these incidents have been inhomogenously deposited
around the world.
January 2009 B-3 NUREG-1575, Supp. 1
-------
Appendix B MARSAME
Isolated pockets with elevated concentrations of anthropogenic radionuclides can still be found.
For example, ventilation systems that were installed prior to 1963 collected fallout radionuclides.
If these systems are still in use and the ducts have not been thoroughly cleaned, there is a
potential for elevated background radiation. Another potential source of elevated background
radiation from anthropogenic radionuclides is wood ash. Trees filter and store some airborne
pollutants, including 137Cs from fallout. When the wood is burned the 137Cs is concentrated in the
wood ash. Materials or equipment associated with the ash could have elevated levels of
background radiation.
B.2.3 Cosmic Radiation and Cosmogenic Radionuclides
Cosmic radiation consists of highly energetic particles that are believed to originate from
phenomena such as solar flares and supernova explosions. The Earth's atmosphere serves as a
shield for these particles, although on rare occasions a solar flare is strong enough to produce a
significant radiation dose in the lower reaches of the atmosphere.
Cosmic radiation is also responsible for the production of radioactive elements called
cosmogenic radionuclides. These radionuclides are produced from collisions between the highly
energetic cosmic radiation with stable elements in the atmosphere. Cosmogenic radionuclides
include 3H, 7Be, 14C, and 22Na. Background concentrations of cosmic radiation and cosmogenic
radionuclides generally do not impact disposition surveys.
B.3 Inherent Radioactivity
Inherent radioactivity, or intrinsic radioactivity, is radioactivity that is an integral part of the
M&E being investigated. For example, concrete is made from materials that contain terrestrial
radionuclides and is inherently radioactive. Some equipment is constructed from radioactive
components, such as electron tubes or night vision goggles containing thorium components.
Information on inherent radioactivity is usually obtained from process knowledge or historical
measurements identified during the IA. Manufacturers of equipment that incorporates radioactive
components can usually provide the radionuclide and the activity incorporated into the
equipment. Information on radionuclides and activity levels for other types of equipment or bulk
materials that are inherently radioactive is usually more generic. Table B.I lists ranges of
terrestrial radionuclide concentrations in some common materials (e.g., concrete, soil, brick).
The wide range of radionuclide concentrations observed in these materials prevents establishing
any general rules of thumb, so it is usually necessary to obtain project-specific information. For
release scenarios, it is strongly recommended that all M&E be surveyed before it enters a
controlled area. This provides project-specific information on inherent radioactivity and
minimizes complications when designing the disposition survey. For interdiction scenarios, it is
important to understand the types of M&E being investigated and the potential for inherent
radioactivity. It may be necessary to establish an administrative action level that defines the
upper end of acceptable inherent radioactivity for different types of M&E (see Section 3.2).
NUREG-1575, Supp. 1 B-4 January 2009
-------
MARSAME Appendix B
B.4 Instrument Background
Instrument background is a combination of radioactivity in the constituent materials of the
detector, ancillary equipment, and shielding, and electronic noise contributing to the instrument
response. Instruments designed to measure low levels of radioactivity generally are constructed
from materials with very low levels of inherent radioactivity to minimize instrument background.
The electronics in radiation instruments are also designed to minimize the signal-to-noise ratio,
also reducing instrument background. Instrument background becomes the primary contributor
to background only for radionuclide-specific measurements for radionuclides not contributing to
environmental or inherent background (e.g., 60Co in bulk soil measured by gamma
spectroscopy). Note that radiation from M&E can interact with instrument shielding to produce
secondary effects that may contribute to instrument background (e.g., Compton backscatter,
generation of secondary photons and characteristic x rays, photoelectric absorption). Additional
information on instrument background is available in Chapter 20 of Radiation Detection and
Measurement (Knoll 1999).
B.5 Technologically Enhanced Naturally Occurring Radioactive Material
Technologically Enhanced Naturally Occurring Radioactive Material (TENORM) is any
naturally occurring material not subject to regulation under the Atomic Energy Act whose
radionuclide concentrations or potential for human exposure have been increased above levels
encountered in the natural state by human activities (NAS 1999). Some industrial processes
involving natural resources concentrate naturally occurring radionuclides, producing TENORM.
Much TENORM contains only trace amounts of radioactivity and is part of our everyday
landscape. Some TENORM, however, contains very high concentrations of radionuclides. The
majority of radionuclides in TENORM are found in the uranium and thorium natural decay
series. Potassium-40 is also associated with TENORM. Radium and radon typically are
measured as indicators for TENORM in the environment. TENORM is found in many industrial
waste streams (e.g., scrap metal, sludges, slags, fluids) and is being discovered in industries
traditionally not thought of as being affected by radiation. Examples of products and processes
affected by TENORM include:
• Uranium overburden and mine spoils,
• Phosphate industrial wastes,
• Phosphate fertilizers and potash,
• Coal ash, slag, cinders,
• Oil and gas production scale and sludge,
• Sludge and other waste materials from treatment of drinking water and waste water,
• Metal mining and processing waste,
• Geothermal energy production waste,
• Paper and pulp,
• Scrap metal recycling,
• Slag from industrial processes (metal and non-metal),
• Abrasive mineral sands, and
• Cement production.
January 2009 B-5 NUREG-1575, Supp. 1
-------
Appendix B MARSAME
Radon and radon progeny are concerns when dealing with TENORM. Radon-222 is a decay
OOQ OOO
product of U. The 3.8-day half-life means that Rn is capable of migrating through several
decimeters of soil or building materials and reaching the atmosphere before it decays. The
radioactive progeny of unsupported 222Rn have short half-lives (e.g., 27 minutes for 214Pb) and
usually decay to background levels within a few hours. 220Rn, which has a 55-second half-life, is
a decay product of 232Th. The short half-life limits the mobility of 220Rn since it decays before it
can migrate to the atmosphere. However, 232Th activity that is located on or very near the surface
can produce significant quantities of 220Rn in the air. The radioactive progeny of unsupported
220Rn can result in elevated levels of surface radioactivity for materials and equipment used or
stored in these areas. The 10.6-hour half-life of 212Pb means that this surface radioactivity could
take a week or longer to decay to background levels.
B.6 Orphan Sources
Radiation sources are found in certain types of specialized industrial devices, such as those used
for measuring the moisture content of soil and for measuring density or thickness of materials.
Usually, a small quantity of the radioactive material is sealed in a metal casing and enclosed in a
housing that prevents the escape of radiation. These sources present no health risk from
radioactivity as long as the sources remain sealed, the housing remains intact, and the devices are
handled and used properly.
If equipment containing a sealed source is disposed of improperly or sent for recycling as scrap
metal, the sealed source may be "lost" and end up in a metal recycling facility or in the
possession of someone who is not licensed to handle the source. Specially licensed sources bear
identifying markings that can be used to trace these sources to their original owners. However,
some sources do not have these markings or the markings become obliterated. In these cases, the
sources are referred to as "orphan sources" because no known owner can be identified. They are
one of the most frequently encountered sources of radioactivity in shipments received by scrap
metal facilities.
Scrap yards and disposal sites attempt to detect orphan sources and other contaminated metals by
screening incoming materials with sensitive radiation detectors before they can enter the
processing stream and cause contamination. Housings that make the sources safe also make
detection difficult. Further, if the source is buried in a load of steel, the steel acts as further
shielding and thus these sources may elude detection. Consequently, there is always a potential
for sources to become mixed within and impact scrap metal.
NUREG-1575, Supp. 1 B-6 January 2009
-------
MARSAME
Appendix C
C. EXAMPLES OF COMMON RADIONUCLIDES
Table C.I Examples of Common Radionuclides at Selected Types of Facilities
Facility Type
Accelerator/Cyclotron
Aircraft Manufacturing and Maintenance Facility
Cement Production Facility
Ceramic Manufacturing Facility
Fertilizer Plant
Fuel Fabrication Facility
Gaseous Diffusion Plant
Medical Imaging and Therapy Facility
Metal Foundry
Munitions and Armament Manufacturing and
Testing Facility
Common Radionuclides
22Na
Activation products (e.g., 60Co)
3H (dials and gauges)
Magnesium-thorium alloys
Nickel-thorium alloys
147Pm (lighted dials and gauges)
226Ra and progeny (radium dials)
Depleted uranium
Thorium series radionuclides
Uranium series radionuclides
Thorium series radionuclides
Uranium series radionuclides
40K
Uranium series radionuclides
99Tc (reprocessing only)
Enriched uranium
Transuranics (e.g., 237Np, 239Pu) (reprocessing only)
"Tc
Enriched uranium
Transuranics (e.g., 237Np, 239Pu)
60Co
90Sr
99mTc
131j
137Cs
192jr
201T1
226Ra and progeny
Depleted uranium collimators
40K
60Co
137Cs
Thorium series radionuclides
Uranium series radionuclides
3H (fire control devices)
226Ra and progeny
Depleted uranium
January 2009
C-l
NUREG-1575, Supp. 1
-------
Appendix C
MARSAME
Table C.I Examples of Common Radionuclides at Selected Types of Facilities (continued)
Facility Type
Nuclear Medicine Laboratory
or Pharmaceutical Laboratory
Nuclear Power Reactor
Oil and Gas
Optical Glass Facility
Paint and Pigment Manufacturing Facility
Paper and Pulp Facility
Radium Dial Painting
Rare Earth Facility
R&D Facility with Broad Scope License
Research Laboratory
Scrap Metal Recycling Facility
Common Radionuclides
99mTc
131j
137Cs
192^
201 y,
226Ra and progeny
Activation products (e.g., 55Fe, 60Co, 63Ni)
Fission products (e.g., 90Sr, 137Cs)
Transuranics (e.g., 237Np, 239Pu)
226Ra and progeny
Thorium series radionuclides
Uranium series radionuclides
Thorium series radionuclides
Uranium series radionuclides
Thorium series radionuclides
Uranium series radionuclides
226Ra and progeny
40K
Thorium series radionuclides
Uranium series radionuclides
3H
14C
3H
14C
22Na
24Na
32p
57Co
63Ni
123j
125j
60Co
90Sr
137^
Cs
226Ra and progeny
Thorium series radionuclides
Uranium series radionuclides
NUREG-1575, Supp. 1
C-2
January 2009
-------
MARSAME
Appendix C
Table C.I Examples of Common Radionuclides at Selected Types of Facilities (continued)
Facility Type
Common Radionuclides
Sealed Source Facility
137,
241
'Co
'Sr
Cs
Am
Transuranic Facility
241Am
238, 239, 240, 241p
Uranium Mill
238
230
'U
Th
226
'Ra and progeny
Thorium series radionuclides
Uranium series radionuclides
Waste Water Treatment Facility
Thorium series radionuclides
Uranium series radionuclides
40
57
60
Widely Distributed General Commerce
H (exit signs)
K (naturally-occurring)
Co (lead paint analyzer)
Co (radiography source)
63Ni (chemical agent detectors)
109Cd (lead paint analyzer)
137Cs (soil moisture density gauge, liquid level
192T
gauge)
Ir (radiography source)
226Ra (watch dials)
241Am (AmBe soil moisture density gauge,
smoke detectors)
Orphan sources
January 2009
C-3
NUREG-1575, Supp. 1
-------
-------
MARSAME Appendix D
D. INSTRUMENTATION AND MEASUREMENT TECHNIQUES
D.I Introduction
This appendix provides information on various field and laboratory equipment used to measure
radiation levels and radioactive material concentrations. The descriptions provide information
pertaining to the general types of available radiation detectors and the ways in which those
detectors are utilized for various circumstances. Similar information may be referenced from
MARSSEVI Appendix H, "Description of Field Survey and Laboratory Analysis Equipment"
(MARSSIM 2002), andNUREG-1761 Appendix B, "Advanced/Specialized Information" (NRC
2002). The information in this appendix is specifically designed to assist the user in selecting the
appropriate radiological instrumentation and measurement technique during the implementation
phase of the data life cycle (Chapter 5).
The following topics will be discussed for each instrumentation and measurement technique
combination:
• Instruments - A description of the equipment and the typical detection instrumentation it
employs;
• Temporal Issues - A synopsis of time constraints that may be encountered through use of
the measurement technique;
• Spatial Issues - Limitations associated with the size and portability of the instrumentation as
well as general difficulties that may arise pertaining to source-to-detector geometry;
• Radiation Types - Applicability of the measurement technique for different types of
ionizing radiation;
• Range - The associated energy ranges for the applicable types of ionizing radiation;
• Scale - Typical sizes for the M&E applicable to the measurement technique; and
• Ruggedness - A summary of the durability of the instrumentation (note that this is
frequently limited by the detector employed by the instrumentation; e.g., an instrument
utilizing a plastic scintillator is inherently more durable than an instrument utilizing a sodium
iodide crystal); suitable temperature ranges for proper operation of the instrumentation and
measurement technique have been provided where applicable.
D.2 General Detection Instrumentation
This section summarizes the most common detector types used for the detection of ionizing
radiation in the field. This will include many of the detector types incorporated into the
measurement methods that are described in later sections of this chapter.
D.2.1 Gas-Filled Detectors
Gas-filled detectors are the most commonly used radiation detectors and include gas- ionization
chamber detectors, gas-flow proportional detectors, and Geiger-Muller (GM) detectors. These
detectors can be designed to detect alpha, beta, photon, and neutron radiation. They generally
consist of a wire passing through the center of a gas-filled chamber with metal walls, which can
be penetrated by photons and high-energy beta particles. Some chambers are fitted with Mylar
January 2009 D-l NUREG-1575, Supp. 1
-------
Appendix D MARSAME
windows to allow penetration by alpha and low-energy beta radiation. A voltage source is
connected to the detector with the positive terminal connected to the wire and the negative
terminal connected to the chamber casing to generate an electric field, with the wire serving as
the anode, and the chamber casing serving as the cathode. Radiation ionizes the gas as it enters
the chamber, creating free electrons and positively charged ions. The number of electrons and
positively charged ions created is related to the properties of the incident radiation type (alpha
particles produce many ion pairs in a short distance, beta particles produce fewer ion pairs due to
their smaller size, and photons produce relatively few ion pairs as they are uncharged and
interact with the gas significantly less than alpha and beta radiation). The anode attracts the free
electrons while the cathode attracts the positively charged ions. The reactions among these ions
and free electrons with either the anode or cathode produce disruptions in the electric field. The
voltage applied to the chamber can be separated into different voltage ranges that distinguish the
types of gas-filled detectors described below. The different types of gas-filled detectors are
described in ascending order of applied voltage.
D.2.1.1 lonization Chamber Detectors
lonization chamber detectors consist of a gas-filled chamber operated at the lowest voltage range
of all gas-filled detectors.l lonization detectors utilize enough voltage to provide the ions with
sufficient velocity to reach the anode or cathode. The signal pulse heights produced in ionization
chamber detectors is small and can be discerned by the external circuit to differentiate among
different types of radiation. These detectors provide true measurement data of energy deposited
proportional to the charge produced in air, unlike gas-flow proportional and GM detectors which
are detection devices. These detectors generally are designed to collect cumulative beta and
photon radiation without amplification and many have a beta shield to help distinguish among
these radiation types. These properties make ionization detectors excellent choices for measuring
exposure rates from photon emission radiation in roentgens. These detectors can be deployed for
an established period of time to collect data in a passive manner for disposition surveys.
lonization chamber detectors may assist in collecting measurements in inaccessible areas due to
their availability in small sizes.
Another form of the ionization chamber detector is the pressurized ion chamber (PIC). As with
other ionization chamber detectors, the PIC may be applied for M&E disposition surveys when a
exposure-based action level is used. The added benefit of using PICs is that they can provide
more accurate dose measurements because they compensate for the various levels of photon
energies as opposed to other exposure rate meters (e.g., micro-rem meter), which are calibrated
to a 137Cs source. PICs can be used to cross-calibrate other exposure rate detectors applicable for
surveying M&E, allowing the user to compensate for different energy levels and reduce or
eliminate the uncertainty of underestimating or overestimating the exposure rate measurements.
D.2.1.2 Gas-Flow Proportional Detectors
The voltage applied in gas-flow proportional detectors is the next range higher than ionization
chamber detectors, and is sufficient to create ions with enough kinetic energy to create new ion
1 At voltages below the ionization chamber voltage range, ions will recombine before they can reach either the
cathode or anode and do not produce a discernable disruption to the electric field.
NUREG-1575, Supp. 1 D-2 January 2009
-------
MARSAME Appendix D
pairs, called secondary ions. The quantity of secondary ions increases proportionally with the
applied voltage, in what is known as the gas amplification factor. The signal pulse heights
produced can be discerned by the external circuit to differentiate among different types of
radiation. Gas-flow proportional detectors generally are used to detect alpha and beta radiation.
Systems also detect photon radiation, but the detection efficiency for photon emissions is
considerably lower than the relative efficiencies for alpha and beta activity. Physical probe areas
for these types of detectors vary in size from approximately 100 cm2 up to 600 cm2. The detector
cavity in these instruments is filled with P-10 gas which is an argon-methane mixture (90%
argon and 10% methane). Ionizing radiation enters this gas-filled cavity through an aluminized
Mylar window. Additional Mylar shielding may be used to block alpha radiation; a lower voltage
setting may be used to detect pure alpha activity (NRC 1998b).
D.2.1.3 Geiger-Mueller Detectors
GM detectors operate in the voltage range above the proportional range and the limited
proportional range.2 This range is characterized by extensive gas amplification that results in
what is referred to as an "avalanche" of ion and electron production. This mass production of
electrons spreads throughout the entire chamber, which precludes the ability to distinguish
among different kinds of radiation because all of the signals produced are the same size. GM
detectors are most commonly used for the detection of beta activity, though they may also detect
both alpha and photon radiation. GM detectors have relatively short response and dead times and
are sensitive enough to broad detectable energy ranges for alpha, beta, photon, and neutron
emissions (though they cannot distinguish which type of radiation produces input signals) to
allow them to be used for surveying M&E with minimal process knowledge.3
GM detectors are commonly divided into three classes: "pancake", "end-window", and "side-
wall" detectors. GM pancake detectors (commonly referred to as "friskers") have wide diameter,
thin mica windows (approximately 15 cm2 window area) that are large enough to allow them to
be used to survey many types of M&E. Although GM pancake detectors are referenced beta and
gamma detectors, the user should consider that their beta detection efficiency far exceeds their
gamma detection efficiency. The end-window detector uses a smaller, thin mica window and is
designed to allow beta and most alpha particles to enter the detector unimpeded for concurrent
alpha and beta detection. The side-wall detector is designed to discriminate between beta and
gamma radiation, and features a door that can be slid or rotated closed to shield the detector from
beta emissions for the sole detection of photons. These detectors require calibration to detect for
beta and gamma radiation separately. Energy-compensated GM detectors may also be cross-
calibrated for assessment of exposure rates.
The limited proportional range produces secondary ion pairs but does not produce reactions helpful for radiation
detection, because the gas amplification factor is no longer constant.
3 GM detectors may be designed and calibrated to detect alpha, beta, photon, and neutron radiation, though they are
much better-suited for the detection of charged particles (i.e., alpha and beta particles) than neutral particles (i.e.,
photons and neutrons).
January 2009 D-3 NUREG-1575, Supp. 1
-------
Appendix D MARSAME
D.2.2 Scintillation Detectors
Scintillation detectors (sometimes referred to as "scintillators") consist of scintillation media that
emits a light "output" called a scintillation pulse when it interacts with ionizing radiation.
Scintillators emit low-energy photons (usually in the visible light range) when struck by high-
energy charged particles; interactions with external photons cause scintillators to emit charged
particles internally, which in turn interact with the crystal to emit low-energy photons. In either
case, the visible light emitted (i.e., the low-energy photons) are converted into electrical signals
by photomultiplier tubes and recorded by a digital readout device. The amount of light emitted is
generally proportional to the amount of energy deposited, allowing for energy discrimination and
quantification of source radionuclides in some applications.
D.2.2.1 Zinc Sulfide Scintillation Detectors
Zinc sulfide detector crystals are only available as a polycrystalline powder that are arranged in a
thin layer of silver-activated zinc sulfide (ZnS(Ag)) as a coating or suspended within a layer of
plastic scintillation material. The use of these thin layers makes them inherently dispositioned for
the detection of high linear energy transfer (LET) radiation (radiation associated with alpha
particles or other heavy ions). These detectors use an aluminized Mylar window to prevent
ambient light from activating the photomultiplier tube (Knoll 1999). The light pulses produced
by the scintillation crystals are amplified by a photomultiplier tube, converted to electrical
signals, and counted on a digital scaler/ratemeter. Low LET radiations (particularly beta
emissions) are detected at much lower detection efficiencies than alpha emissions and pulse
characteristics may be used to discriminate beta detections from alpha detections.
D.2.2.2 Sodium Iodide Scintillation Detectors
Sodium iodide detectors are well-suited for detection of photon radiation. Energy-compensated
sodium iodide detectors may also be cross-calibrated for assessment of exposure rates. Unlike
ZnS(Ag), sodium iodide crystals can be grown relatively large and machined into varying shapes
and sizes. Sodium iodide crystals are activated with trace amounts of thallium (hence the
abbreviation Nal(Tl)), the key ingredient to the crystal's excellent light yield (Knoll, 1999).
These instruments most often have upper- and lower-energy discriminator circuits and when
used correctly as a single-channel analyzer, can provide information on the photon energy and
identify the source radionuclides. Sodium iodide detectors can be used with handheld
instruments or large stationary radiation monitors.
D.2.2.3 Cesium Iodide Scintillation Detectors
Cesium iodide detectors generally are similar to sodium iodide detectors. Like Nal(Tl), cesium
iodide may be activated with thallium (CsI(Tl)) or sodium (CsI(Na)). Cesium iodide is more
resistant to shock and vibration damage than Nal, and when cut into thin sheets it features
malleable properties allowing it to be bent into various shapes. CsI(Tl) has variable decay times
for various exciting particles, allowing it to help differentiate among different types of ionizing
radiation. A disadvantage of Csl scintillation detectors is due to the fact that the scintillation
emission wavelengths for Csl are longer than those produced by sodium iodide crystals; because
NUREG-1575, Supp. 1 D-4 January 2009
-------
MARSAME Appendix D
almost all photomultiplier tubes are designed for Nal, there are optical incompatibilities that
result in decreased intrinsic efficiencies for Csl detectors. Additionally, Csl scintillation
detectors feature relatively long response and decay times for luminescent states in response to
ionizing radiation (Knoll 1999).
D.2.2.4 Plastic Scintillation Detectors
Plastic scintillators are composed of organic scintillation material that is dissolved in a solvent
and subsequently hardened into a solid plastic. Modifications to the material and specific
packaging allow plastic scintillators to be used for detecting alpha, beta, photon, or neutron
radiation. While plastic scintillators lack the energy resolution of sodium iodide and some other
gamma scintillation detector types, their relatively low cost and ease of manufacturing into
almost any desired shape and size enables them to offer versatile solutions to atypical radiation
detection needs (Knoll 1999).
D.2.3 Solid State Detectors
Solid state detection is based on ionization reactions within detector crystals composed of an
electron-rich (n-type or electron conductor) sector and an electron-deficient (p-type or hole
conductor) sector. Reverse-bias voltage is applied to the detector crystal; forming a central
region absent of free charge (this is termed the depleted region). When a particle enters this
region, it interacts with the crystal structure to form hole-electron pairs. These holes and
electrons are swept out of the depletion region to the positive and negative electrodes by the
electric field, and the magnitude of the resultant pulse in the external circuit is directly
proportional to the energy lost by the ionizing radiation in the depleted region.
Solid state detection systems typically employ silicon or germanium crystals4 and utilize
semiconductor technology (i.e., a substance whose electrical conductivity falls between that of a
metal and that of an insulator, and whose conductivity increases with decreasing temperature and
with the presence of impurities). Semiconductor detectors are cooled to extreme temperatures to
utilize the crystal material's insulating properties to prevent thermal generation of noise. The use
of semiconductor technology can achieve energy resolutions, spatial resolutions, and signal-to-
noise ratios superior to those of scintillation detection systems.
D.3 Counting Electronics
Instrumentation requires a device to accumulate and record the input signals from the detector
over a fixed period of time. These devices are usually electronic, and utilize sealers or rate-
meters to display results representing the number of interaction events (between the detector and
radionuclide emissions) within a period of time (e.g., counts per minute). A sealer represents the
total number of interactions within a fixed period of time, while a rate-meter provides
information that varies based on a short-term average of the rate of interactions.
4 Solid state detection systems may also utilize crystals composed of sodium iodide, cesium iodide, or cadmium zinc
telluride in non-semiconductor applications.
January 2009 D-5 NUREG-1575, Supp. 1
-------
Appendix D MARSAME
Sealers represent the simpler of these two counting approaches, because they record a single
count each time an input signal is received from the detector. Scaling circuits typically are
designed with sealers to allow the input signals to be cut by factors of 10, 100, or 1,000 to allow
the input signals to be counted directly by electromechanical registers when counting areas with
elevated radioactivity. Sealers generally are used when taking in situ measurements and are used
to determine average activities.
Contemporary rate-meters utilize analog-to-digital converters to sample the pulse amplitude of
the input signal received from the detector and convert it to a series of digital values. These
digital values may then be manipulated using digital filters (or shapers) to average or "smooth"
the data displayed. The counting-averaging technique used by rate-meters may be more helpful
than sealers in identifying elevated activity. When using sealers in performing scanning surveys
to locate areas of elevated activity, small areas of elevated activity may appear as very quick
"blips" that are difficult to discern, while rate-meters continue to display heightened count rates
once the detector has moved past the elevated activity, and display "ramped up" count rates
immediately preceding the elevated activity as well. Rate-meters have the inherent limitation in
that the use of their counting electronics varies the signals displayed by the meter because they
represent a short-term average of the event rate. It is conceivable that very small areas of
elevated activity (e.g., particle) might have their true activity concentrations "diluted" by the
averaging of rate-meter counting electronics.
D.4 Hand-Held Instruments
This section discusses hand-held instruments, which may be used for in situ measurements or
scanning surveys.
D.4.1 Instruments
In situ measurements with hand-held instruments typically are conducted using the detector types
described in Section D.2. These typically are composed of a detection probe (utilizing a single
detector) and an electronic instrument to provide power to the detector and to interpret data from
the detector to provide a measurement display.
The most common types of hand-held detector probes are GM detectors, ZnS(Ag) alpha/beta
scintillation detectors, andNal(Tl) photon scintillation detectors. There are instances of gas-flow
proportional detectors as hand-held instruments, though these are not as common because these
detectors operate using a continuous flow of P-10 gas, and the accessories associated with the
gas (e.g., compressed gas cylinders, gauges, tubing) make them less portable for use in the field.
D.4.2 Temporal Issues
Hand-held instruments generally have short, simple equipment set-ups requiring minimal time,
often less than ten minutes. In situ measurement count times typically range from 30 seconds to
two minutes. Longer count times may be utilized to increase resolution and provide lower
minimum detectable limits. Typical scanning speeds are approximately 2.5 centimeters per
second. Slower scanning speeds will aid in providing lower minimum detectable concentrations.
NUREG-1575, Supp. 1 D-6 January 2009
-------
MARSAME
Appendix D
DAS Spatial Issues
Detectors of hand-held instruments typically are small and portable, having little trouble fitting
into and measuring most M&E. Spatial limitations are usually based on the physical size of the
probe itself. The user must be wary of curved or irregular surfaces of M&E being surveyed.
Detector probes generally have flat faces and incongruities between the face of the detector and
the M&E being surveyed have an associated uncertainty. ZnS scintillation and gas-flow
proportional detectors are known to have variations in efficiency of up to 10% across the face of
the detector. Therefore, the calibration source used should have an area at least the size of the
active probe area.
D.4.4 Radiation Types
Assortments of hand-held instruments are available for the detection of alpha, beta, photon, and
neutron radiations. Table D.I illustrates the potential applications for the most common types of
hand-held instruments.
Table D.I Potential Applications for Common Hand-Held Instruments
lonization chamber detectors
Gas-flow proportional
detectors
Geiger-Muller detectors
ZnS(Ag) scintillation
detectors
Nal(Tl) scintillation detectors
Nal(Tl) scintillation detectors
(thin detector, thin window)
CsI(Tl) scintillation detectors
Plastic scintillation detectors
BF3 proportional detectors5
3He proportional detectors5
Alpha
NA
GOOD
FAIR
GOOD
NA
NA
NA
NA
NA
NA
Beta
FAIR
GOOD
GOOD
POOR
POOR
FAIR
POOR
FAIR
NA
NA
Photon
GOOD
POOR
POOR
NA
GOOD
GOOD
GOOD
GOOD
NA
POOR
Neutron
NA
POOR
POOR
NA
NA
NA
NA
NA
GOOD
GOOD
Detectable Energy Range
Low End
Boundary
40-60 keV
5-50 keV
30-60 keV
30-50 keV
40-60 keV
lOkeV
40-60 keV
40-60 keV
0.025 eV
0.025 eV
High End
Boundary
1.3-3MeV
8-9 MeV
1-2 MeV
8-9 MeV
1.3-3 MeV
60-200 keV
1.3-3 MeV
1.3-3 MeV
100 MeV
100 MeV
Notes:
GOOD The instrument is well-suited for detecting this type of radiation.
FAIR The instrument can adequately detect this type of radiation.
POOR The instrument may be poorly suited for detecting this type of radiation.
NA The instrument cannot detect this type of radiation.
5 The use of moderators enables the detection of high-energy fast neutrons. Either BF3 or 3He gas proportional
detectors may be used for the detection of fast neutrons, but 3He are much more efficient in performing this function.
BF3 detectors discriminate against gamma radiation more effectively than 3He detectors.
January 2009
D-7
NUREG-1575, Supp. 1
-------
Appendix D MARSAME
DAS Range
The ranges of detectable energy using hand-held instruments are dependent upon the type of
instrument selected and type of radiation. Some typical detectable energy ranges for common
hand-held instruments are listed above in Table D.I. More detailed information pertaining to the
ranges of detectable energy using hand-held instruments are available in the European
Commission for Nuclear Safety and the Environment Report 17624 (EC 1998).
D.4.6 Scale
There is no definitive limit to the size of an object to be surveyed using hand-held instruments.
Hand-held instruments may generally be used to survey M&E of any size; constraints are only
placed by the practical sizing of M&E related to the sensitive area of the probe. Limitations may
also be derived from the physical size of the detector probes used for surveying. The largest
hand-held detector probes feature effective detection surface areas of approximately 175 to 200
cm2. Detection probes larger than this may be of limited use with hand-held instruments.
D.4.7 Ruggedness
All varieties of hand-held instruments discussed here typically are calibrated for use in
temperatures with lower ranges from -30 ° to -20 °C and upper ranges from 50 ° to 60 °C. The
durability of a hand-held instrument depends largely upon the detection media (crystals, such as
sodium iodide and germanium crystals are fragile and vulnerable to mechanical and thermal
shock) and the presence of a Mylar (or similar material) window:
• lonization chamber detectors - lonization chamber detectors are susceptible to physical
damage and may provide inaccurate data (including false positives) if exposed to
mechanical shock.
• Gas-flow proportional detectors - Detection gas used with gas-flow proportional detectors
may leak from seals such that these detectors are usually operated in the continuous gas flow
mode; the use of flow meter gauges to continuously monitor the gas flow rate is
recommended along with frequent quality control checks to ensure the detector still meets
the required sensitivity; gas-flow proportional detectors may also use fragile Mylar windows
to contain the detection gases, which renders the detectors vulnerable to puncturing and
mechanical shock.
• Geiger-Muller detectors - GM tubes typically use fragile Mylar windows to contain the
detection gases; the presence of a Mylar window renders the detector vulnerable to
puncturing and mechanical shock.
• ZnS(Ag) scintillation detectors - Zinc sulfide is utilized as thin-layer polycrystalline
powder in detectors and are noted for being vulnerable to mechanical shock; zinc sulfide
detectors may use fragile Mylar windows, in which case the detector is vulnerable to
puncturing and mechanical shock.
• Nal(Tl) scintillation detectors - Sodium iodide crystals are relatively fragile and can be
damaged through mechanical shock; sodium iodide is also highly hydroscopic such that the
crystals must remain environmentally sealed within the detector housing.
NUREG-1575, Supp. 1 D-8 January 2009
-------
MARSAME
Appendix D
• Plastic Scintillation Detectors - Plastic scintillators typically are robust and resistant to
damage from mechanical and thermal shock.
D.5 Volumetric Counters (Drum, Box, Barrel, Four-Pi Counters)
The term "box counter" is a generic description for a radiation measurement system that
typically involves large area, four-pi (4-7i) radiation detectors and includes the following industry
nomenclature: tool counters, active waste monitors, surface activity measurement systems, and
bag/barrel/drum monitors. Box counting systems are most frequently used for conducting in situ
surveys of M&E that is utilized in radiologically controlled areas. These devices are best-suited
for performing gross activity screening measurements on Class 2 and Class 3 M&E (NRC 2002).
Typical items to be surveyed using box counters are hand tools, small pieces of debris, bags of
trash, and waste barrels. Larger variations of box counting systems can count objects up to a few
cubic meters in size. Because of potential problems with self-shielding, materials may need to be
opened or partially disassembled prior to placing into a box counting system.
D.5.1 Instruments
Box counting systems typically consist of a counting chamber, an array of detectors configured
to provide a 4-n counting geometry, and microprocessor-controlled electronics that allow
programming of system parameters and data-logging. Systems typically survey materials for
photon radiation and usually incorporate a shielded counting chamber and scintillation detectors
(plastic scintillators or sodium iodide scintillation detectors). These systems most commonly
utilize four or six detectors, which are situated on the top, bottom, and sides of the shielded
counting chamber (Figure D. 1). Some systems monitor M&E for beta activity, using a basic
design similar to photon radiation detection systems, but utilizing gas-flow proportional
counters. In rare cases, neutron detection
has been used for criticality controls and
counter-proliferation screening.
Box counting systems for alpha activity
feature a substantial departure in design
from beta/gamma detection systems.
Alpha activity systems do not require
heavy shielding to filter out ambient
sources of radiation. These devices
utilize air filters to remove dust and
particulates from air introduced into
counting chambers that incorporate
airtight seals. Filtered air introduced into
the counting chamber interacts with any
surface alpha activity associated with the
M&E.
Each alpha interaction with a surrounding
air molecule produces an ion pair. These , r, , .. , . _ .„,. ~nn-s
1 ' Figure D-l Example Volumetnc Counter (Thermo 2005)
January 2009
D-9
NUREG-1575, Supp. 1
-------
Appendix D MARSAME
ion pairs are produced in proportion to the alpha activity per unit path length. This air (i.e., the
ion pairs in the air) is then counted using an ion detector for quantification of the specific
activity. The specific activity of the air in the counting chamber provides a total surface activity
quantification for the M&E (BIL 2005).
D.5.2 Temporal Issues
Typically, box counting systems require approximately one to 100 seconds to conduct a
measurement (Thermo 2005). The count times are dependent on a number of factors to include
required measurement sensitivity and background count rates with accompanying subtraction
algorithms. The count times for box counting typically are considered relatively short for most
disposition surveys.
D.5.3 Spatial Issues
Because box counters typically average activity over the volume or mass of the M&E, the spatial
distribution of radioactivity may be a significant limitation on the use of this measurement
technique. The design of box counting systems is not suited to the identification of localized
elevated areas, and therefore may not be the ideal choice when the disposition criteria is not
based on average or total activity.
Some systems incorporate a turntable inside the counting chamber to improve measurement of
difficult-to-measure areas or for heterogeneously distributed radioactivity. When practical,
performing counts on objects in two different orientations (i.e., by rotating the M&E 90 or 180
degrees and performing a subsequent count) will yield more thorough and defensible data.
Proper use of box counters includes segregating the M&E to be surveyed and promoting accurate
measurements through uniform placement of items to be surveyed in the counting chamber. For
example, a single wrench placed on its side in a box counter has different geometric implications
from a tool of similar size standing up inside the counting chamber. Counting jigs for sources
and M&E to be surveyed are frequently employed to facilitate consistent, ideal counting
positions between the M&E and the counting chamber detector array.
D.5.4 Radiation Types
Box counting systems are intrinsically best-suited for the detection of moderate- to high- energy
photon radiation. As described in Section D.5.1, specific systems may be designed for the
detection of low-energy photon, beta, alpha, and in some cases neutron radiation. For proper
calibration and utilization of box counters, it is often necessary to establish the radiation types
and anticipated energy ranges prior to measurement.
D.5.5 Range
Photon radiation can typically be measured within a detectable energy range of 40 to 60 keV up
to 1.3 to 3 MeV. For example, typical box counters positioned at radiological control area exit
points are configured to alarm at a set point of 5,000 dpm total activity. The precise count time is
adjusted automatically by setting the predetermined count rate to limit the error. Measurement
NUREG-1575, Supp. 1 D-10 January 2009
-------
MARSAME Appendix D
times will range from 5 to 45 seconds in order to complete counts of this kind, depending on
current background conditions (Thermo 2005). Lower detection capabilities are achievable by
increasing count times or incorporating background reduction methodologies.
D.5.6 Scale
Size limitations pertaining to the M&E to be surveyed are inherently linked to the physical size
of the counting chamber. Smaller box counting systems have a counting chamber of less than
0.028 cubic meters (approximately one cubic foot) and are often used for tools and other
frequently used small items. The maximum size of box counters is typically driven by the
logistics of managing the M&E to be measured, and this volume is commonly limited to a 55-
gallon waste drum. Some box counting systems allow counts to be performed on oversized items
protruding from the counting chamber with the door open.
D.5.7 Ruggedness
Many volumetric counter models feature stainless steel construction with plastic scintillation
detectors and windowless designs, which translates to a rugged instrument that is resistant to
mechanical shock.
D.6 Conveyorized Survey Monitoring Systems
Conveyorized survey monitoring systems automate the routine scanning of M&E. Conveyorized
survey monitoring systems have been designed to measure materials such as soil, clothing
(laundry monitors), copper chop (small pieces of copper), rubble, and debris. Systems range
from small monitoring systems comprised of a single belt that passes materials through a
detector array, to elaborate multi-belt systems capable of measuring and segregating material
while removing extraneously large items. The latter type comprises systems that are known as
segmented gate systems. These automated scanning systems segregate materials by activity by
directing material that exceeds an established activity level onto a separate conveyor. Simpler
Conveyorized survey monitoring systems typically feature an alarm/shut-down feature that halts
the conveyor motor and allows for manual removal of materials that have exceeded the
established activity level.
D.6.1 Instruments
A typical Conveyorized survey monitoring system consists of a motorized conveyor belt that
passes materials through an array of detectors, supporting measurement electronics, and an
automated data-logging system (Figure D.2). Systems typically survey materials for photon
radiation and usually incorporate scintillation detectors (plastic scintillators or sodium iodide
scintillation detectors) or high-purity germanium detectors. Scintillation detector arrays are often
chosen for gross gamma activity screening. Conveyorized survey monitoring systems designed
to detect radionuclide mixtures with a high degree of process knowledge work best using plastic
scintillators, while systems categorizing material mixtures where the radionuclide concentrations
are variable are better-suited to the use of sodium iodide scintillation detectors. Conveyorized
survey monitoring systems designed for material mixtures where the radionuclide concentrations
January 2009 D-ll NUREG-1575, Supp. 1
-------
Appendix D MARSAME
are unknown may be suitable for more expensive and maintenance-intensive high-purity
germanium detector arrays, which will allow for quantitative measurement of complex photon
energy spectra. An alternative method for screening materials for different photon energy regions
of interest is to incorporate sodium iodide detector arrays with crystals of varying thickness to
target multiple photon energies. Systems may also be fitted with gas flow proportional counters
for the detection of alpha and beta emissions. Laundry conveyorized survey monitoring systems
typically are designed for the detection of alpha and beta radiation, as the nature of clothes
allows the survey media to be compressed, allowing the detector arrays to be close to or in
contact with the survey media.
Figure D-2 Example Conveyorized Survey Monitoring System (Laurus 2001)
D.6.2 Temporal Issues
Typically, conveyorized survey monitoring systems require approximately one to six seconds to
count a given field of detection (Novelec 2001a). Systems are designed to provide belt speeds
ranging from 0.75 meters up to 10 meters (2.5 to 33 feet) per minute to accommodate the
necessary response time for detection instrumentation (Thermo 2008; Eberline 2004). This yields
processing times of 15 to 45 metric tons (16 to 50 tons) of material per hour for soil or
construction demolition-type material conveyorized survey monitoring systems (NRC 2002).
D.6.3 Spatial Issues
The M&E that typically are surveyed by conveyorized survey monitoring systems may contain
difficult-to-measure areas. Most systems employ the detector arrays in a staggered, off-set
configuration, which allows the sensitive areas of the detectors to overlap with respect to the
direction of movement. This off-set configuration helps to eliminate blind spots (i.e., locations
where activity may be present but cannot be detected because the radiation cannot reach the
detectors). Some systems are designed specifically for materials of relatively small particles of
uniform size (e.g., soil), while others have been designed to accommodate heterogeneous
materials like rubble and debris.
NUREG-1575, Supp. 1 D-12 January 2009
-------
MARSAME Appendix D
The data logging system accepts the signal pulses from the detector systems and stores the pulse
data in counting sealers. The recorded values are continuously compared with pre-set alarm
values corresponding to the selected action level(s). The detectors incorporate integral amplifiers
which are routed to a PC containing multi-channel sealer hardware. The multi-channel sealer
hardware allows data to be collected in a series of short, discrete sealer channels known as "time
bins". The count time for each time bin is selected as a function of the speed of the conveyor
belt. The time bin length is frequently set up to be half the length of "dwell time," which is the
time the material aliquot to be surveyed spends within the detection field (Miller 2000).
The approach cited in the paragraph above ensures that activity present within the survey unit
will be in full view of the detector for one complete time bin. Data collection is optimized by
performing the measurement when the activity is concentrated (i.e., within an area of elevated
activity) as well as when the activity is approximately homogenously distributed within a given
material aliquot.
D.6.4 Radiation Types
Conveyorized survey monitoring systems generally are best-suited for the detection of photon
radiation. Specific systems may be tailored for the detection of beta emissions of moderate
energy and even alpha radiation by employing gas flow proportional counter detector arrays.
D.6.5 Range
Photon radiation can typically be measured with a detectable energy range from 50 keV up to 2
MeV. Conveyorized survey monitoring systems equipped to measure alpha and beta emissions
can typically measure from 100 keV up to 6 MeV.
D.6.6 Scale
Most conveyorized survey monitoring systems are designed for soils or laundry, both of which
are compressible media. Applicable sample/material heights range from 2 cm to 30 cm (Fuji
2008, Canberra 2008).
D.6.7 Ruggedness
Conveyorized survey monitoring systems have typical operating ranges from -20 °C to 50 °C.
Conveyorized survey monitoring systems are often constructed from steel and with plastic
scintillation detectors and windowless designs, which makes them generally resistant to damage
from extraneous pieces of debris during scanning. Mechanical shock is not a typical concern for
conveyorized survey monitoring systems because there is little need for moving these systems.
For this reason conveyorized survey monitoring systems are seldom transported from one
location to another.
January 2009 D-13 NUREG-1575, Supp. 1
-------
Appendix D MARSAME
D.7 In Situ Gamma Spectroscopy
In situ gamma spectroscopy (ISGS) systems combine the peak resolution capabilities of
laboratory methods with instrumentation that is portable and rugged enough to be used in field
conditions. These solid state systems can perform quantitative, multi-channel analysis of gamma-
emitting isotopes in both solid and liquid media over areas as large as 100 m2, enabling
spectrographic analysis of M&E that assists the user in identifying constituent radionuclides and
differentiating them from background radiation. ISGS system measurements can also provide
thorough coverage within broad survey areas, minimizing the risk of failing to detect isolated
areas of elevated radioactivity that could potentially be missed when collecting discrete samples.
D.7.1 Instruments
ISGS systems consist of a semiconductor detector, a cryostat, a multi-channel analyzer (MCA)
electronics package that provides amplification and analysis of the energy pulse heights, and a
computer system for data collection and analysis. Semiconductor detection systems typically
employ a cryostat and a Dewar filled with liquid nitrogen (-196 °C). The cryostat transmits the
cold temperature of the liquid nitrogen to the detector crystal, creating the extreme cold
environment necessary for correct operation of the high-resolution semiconductor diode. ISGS
systems may have electronic coolers as well.
ISGS systems use detectors referred to as N- and P-type detectors. N-type detectors contain
small amounts of elements with five electrons in their outer electron shell (e.g., phosphorus,
arsenic) within the germanium crystal (the inclusion of these elements within the germanium
crystal is called "doping"). These result in free, unbonded electrons in the crystalline structure,
providing a small negative current. P-type detectors utilize elements with less than four electrons
in their outer electron shell (e.g., lithium, boron, gallium) are also used in doping to create
electron holes, providing a small positive current. Use of these two varieties of doped germanium
crystals provide different detection properties described below in Section D.7.5.
D.7.2 Temporal Issues
Setup for ISGS semiconductor systems may require one full day. The systems often require one
hour to set up physically, six to eight hours for the semiconductor to reach the appropriate
temperature operating range after the addition of liquid nitrogen, and quality control
measurements may require another hour.6 Count times using ISGS semiconductor systems tend
to be longer than those associated with simpler detector systems for conducting static
measurements, though this may be offset by enlarging the field-of-view. A measurement time of
several minutes is common, depending on the intensity of the targeted gamma energies and the
presence of attenuating materials.
Count times can be shortened by reducing the distance between the area being surveyed and the
detector to improve the gamma incidence efficiency or by using a larger detector. Each option
will ultimately help the detection system see more gamma radiation in a shorter time. Yet either
6 It is important not to move the apparatus prematurely, as failure to allow the ISGS system to cool and equilibrate to
its proper operating temperatures as may cause damage to the semiconductor detector.
NUREG-1575, Supp. 1 D-14 January 2009
-------
MARSAME Appendix D
approach creates greater uncertainty associated with the source-to-detector geometry. A slight
placement error (e.g., a 0.5-cm placement error) will result in significantly higher quantification
error at a distance of one centimeter than at a distance of 10 centimeters. Additionally, this
technique for decreasing count times promotes an effect called cascade summing, a phenomena
affecting detection of gamma radiation from radionuclides that emit multiple gamma photons in
a single decay event (e.g., 60Co, which yields gamma particles of 1.17 and 1.33 MeV). If both
incident gammas deposit their energy in a relatively short period of time (i.e., when compared to
the detector response time and/or the resolving time for the associated electronics), limitations of
the detection system may prevent these individual photons from being distinguished (Knoll
1999).
D.7.3 Spatial Issues
ISGS semiconductor systems require calibration for their intended use. While ISGS
semiconductor systems can be calibrated using traditional prepared radioactive sources, some
ISGS systems have software that enables the user to calculate efficiencies by entering parameters
such as elemental composition, density, stand-off distance, and physical dimensions. Supplied
geometry templates assist in generating calibration curves that can be applied to multiple
collected spectra. The high resolution of these systems coupled with advanced electronic controls
for system parameters allows them to overcome issues related to source-to-detector geometry
and produce quantitative concentrations of multiple radionuclides in a variety of media (e.g.,
soil, water, air filters). Because ISGS systems integrate all radioactivity within their field-of-
view, lead shielding and collimation may be required to "focus" the field-of-view on a specified
target for some applications.
D.7.4 Radiation Types
ISGS systems can accurately identify and quantify only photon-emitting radionuclides.
D.7.5 Range
ISGS systems can identify and quantify low-energy gamma emitters (50 keV with P-type
detectors, 10 keV with N-type detectors) and high-energy gamma emitters (ISGS systems can be
configured to detect gamma emissions upwards of 2.0 MeV). Specially designed germanium
detectors that exhibit very little deterioration in resolution as a function of count rate use N-type
detectors or planar crystals with a very thin beryllium window for the measurement of photons in
the energy range 5 to 80 keV.
D.7.6 Scale
These systems therefore offer functional quantitative abilities to analyze small objects (e.g.,
samples) for radionuclides. They can also effectively detect radioactivity over areas as large as
100 m2 or more (Canberra 2005a). With the use of an appropriate Dewar, the detector may be
used in a vertical orientation to determine gamma isotope concentrations in the ground surface
and shallow subsurface.
January 2009 D-15 NUREG-1575, Supp. 1
-------
Appendix D MARSAME
D.7.7 Ruggedness
ISGS semiconductor systems are fragile, because the extremely low temperatures utilized by the
cryostat render portions of the system brittle and susceptible to damage if not handled with care.
Some ISGS systems are constructed of more rugged materials and their durability is comparable
to most hand-held instruments.
D.8 Hand-Held Radionuclide Identifiers
Hand-held radionuclide identifiers represent a relatively new addition to the radiation detection
market, merging the portability of hand-held instruments with some of the analytical capabilities
of ISGS systems. Hand-held radionuclide identifiers also feature data logging and storage
capabilities (including user-definable radionuclide libraries) and the ability to transfer data to
external devices. These devices are most commonly used for nuclear non-proliferation, where
immediate isotope identification is more critical than low-activity detection sensitivity. Design
parameters for hand-held radionuclide identifiers required by ANSI N42.34 (ANSI 2003) are
user-friendly controls and intuitive menu structuring for routine modes of operation, enabling
users without health physics backgrounds (e.g., emergency response personnel) to complete
basic exposure rate or radionuclide identification surveys. These units also feature restricted
"expert" survey modes of operation to collect activity concentration data for more advanced
applications, including disposition surveys.
D.8.1 Instruments
Hand-held radionuclide identifiers consist of two general types: integrated systems and modular
systems. The integrated systems have the detector and electronics contained in a single package;
modular systems separate the detector from the electronics. These spectrometers employ small
scintillators, typically Nal(Tl) or CsI(Tl), or room temperature solid semiconductors, such as
cadmium zinc telluride (CZT), linked to multi-channel analyzers and internal radionuclide
libraries to enable gamma-emitting radionuclide identification.
D.8.2 Temporal Issues
Hand-held radionuclide identifiers require minimal time to set up.7 Depending upon the
conditions in which data is being collected (i.e., climatic, environmental, the presence of sources
of radiological interference), it may require seconds to several minutes for the unit to stabilize
the input signals from the field of radiation and properly identify the radionuclides.
D.8.3 Spatial Issues
Detectors of hand-held radionuclide identifiers typically are small and portable. Spatial
limitations are usually based on the physical size of the probe itself, and whether the probe is
coupled internally within the casing or externally via an extension cord.
7 The use of multi-point calibrations may add an estimated one to two hours to the time required for instrument set
up.
NUREG-1575, Supp. 1 D-16 January 2009
-------
MARSAME Appendix D
D.8.4 Radiation Types
Hand-held radionuclide identifiers are most commonly used for the detection of photon
radiation, although many devices have capabilities for detecting neutron and beta emissions (the
detection of neutron radiation requires a different probe from the photon radiation probe).
D.8.5 Range
Photon radiation can typically be measured within a detectable energy range of 10 to 30 keV up
to 2.5 to 3 MeV. Neutron radiation can typically be measured within a detectable energy range of
0.02 eV up to 100 MeV.
D.8.6 Scale
There is no definitive limit to the size of an object to be surveyed using hand-held radionuclide
identifiers. Hand-held radionuclide identifiers may generally be used to survey M&E of any size;
practical constraints are only imposed by the size of M&E related to the sensitive area of the
probe.
D.8.7 Ruggedness
All varieties of hand-held radionuclide identifiers discussed here typically are calibrated for use
in temperatures from -20 °C to 50 °C and feature seals or gaskets to prevent water ingress from
rain, condensing moisture, or high humidity. Most hand-held radionuclide identifiers have a
limited resistance to shock, though the durability of an instrument depends largely upon the
detection media (e.g., Nal(Tl) crystals are fragile and vulnerable to mechanical and thermal
shock).
D.9 Portal Monitors
Portal monitors screen access points to controlled areas, and are designed for detecting
radioactivity above background. These systems are used for interdiction-type surveys, and
generally do not provide radionuclide identification. Portal monitors are primarily designed to
monitor activity on vehicles.
Historically, portal monitors have been used to detect radioactive materials at entrance points to
scrap metal facilities and solid waste landfills, and radiological control area exit points within
nuclear facilities to screen for the inadvertent disposal of radionuclides. The proximity of other
items to be surveyed containing high concentrations of activity may influence the variability of
the instrument background, because portal monitors survey activity by detecting small variations
in ambient radiation (NRC 2002).
D.9.1 Instruments
Portal monitors can easily be arranged in various geometries that maximize their efficiencies.
Most national and international standards, for example ANSI 42.35 (ANSI 2004) require both
January 2009 D-17 NUREG-1575, Supp. 1
-------
Appendix D
MARSAME
gamma- and neutron-detecting capabilities, but gamma-only versions are available. Portal
monitors typically use large-area polyvinyl toluene scintillators (a form of plastic scintillators) to
detect photon radiation and 3He proportional tubes to detect neutrons.8 Individual detectors may
be cylindrical or flat. The detectors are usually arranged to form a detection field between two
detectors, and items to be surveyed pass through the detection field (i.e., between the detectors)
as shown in Figure D.3.
The system usually consists of one
or more detector array(s), an
occupancy sensor, a control box, and
a monitoring PC. The control box
and monitoring PC store and analyze
alarm and occupancy data, store and
analyze all gamma and neutron
survey data, and may even send data
through an integrated internet
connection. The monitoring PC also
manages software that operates
multiple arrangements of detector
arrays as well as third party
instruments. For example, security
cameras can take high-resolution
images of objects that exceed a
radiation screening level (Novelec
200 Ib).
Figure D-3 Example Portal Monitor (Canberra 2005b)
D.9.2 Temporal Issues
Count or integration times are very short, typically just a few seconds (NRC 2002). Set-up time
in the field is variable, because temporary systems may require two hours to one half-day to set
up, while permanent systems may require one week to install. For vehicular portal monitor
systems, objects may typically pass through the field of detection at speeds of 8 to 9.5 kilometers
per hour (Canberra 2005b). Most systems use speed correction algorithms to minimize the
effects of variations in dwell time (i.e., the time a given area to be surveyed spends within the
detection field).
D.9.3 Spatial Issues
There are a large number of factors that affect portal monitor performance. The isotopic content
of a radioactive material can determine the ease of detection. For example, high-enriched
uranium (HEU) is easier to detect in a gamma portal than low-enriched uranium (LEU) or
OIS
natural uranium because of the larger gamma emission rate from U.
8 Neutron detectors use materials that detect thermal neutrons, which may be fast neutrons that are thermalized for
detection through the use of moderators.
NUREG-1575, Supp. 1
D-18
January 2009
-------
MARSAME Appendix D
The chemical composition of a material is also important; background levels of radioactivity
must also be considered. Neutron portals are an effective method for detecting plutonium in
areas with large gamma backgrounds. The surface area and size of the detectors and distance
between the detectors all affect the geometry and response of the system. In a large area system
set-up, the closer together the detector arrays are, the better the geometric efficiencies are going
to be. Finally, for each system there is a maximum passage speed through the portal that gives a
counting time necessary to meet the required detection sensitivity.
D.9.4 Radiation Types
Portal monitors typically detect gamma radiation and can also be equipped to detect neutron
radiation. Gamma portals often use integrated metal detectors to provide an indication of
suspicious metal containers that could be used to shield radioactive materials. If the gamma
radiation is not shielded adequately, the detector's alarm will sound. Portal monitors can detect
radioactive material even if it is shielded with a material with a high atomic number, like lead.
D.9.5 Range
Photon radiation can typically be measured within a detectable energy range of 60 keV up to 2.6
MeV. Neutron radiation can typically be measured within a detectable energy range of 0.025 eV
up to 100 MeV. Required detection sensitivities for gamma and neutron sources are described in
ANSI 42.35, Table 3 (ANSI 2004). Portal monitors provide gross counts and cannot compute
quantitative measurements (e.g., activity per unit mass).
D.9.6 Scale
Most systems are designed to monitor items ranging in size from bicycles and other small
vehicles to tractor trailers, railroad cars, and even passenger airplanes (Canberra 2005b). The
width of the detection field (i.e., space between the detector arrays) can usually be modified.
D.9.7 Ruggedness
Portal monitors have typical operating ranges from -20 ° to 55 °C, and some systems may be
functional in temperatures as low as -40 °C according to ANSI 42.35 (ANSI 2004). Portal
monitors are usually designed with weatherproofing to withstand prolonged outdoor use and
exposure to the elements.
D.10 Sample with Laboratory Analysis
Laboratory analysis allows for more controlled conditions and more complex, less rugged
instruments to provide lower detections limits and greater delineation among radionuclides than
any measurement method that may be utilized in a field setting. For this reason, laboratory
analyses are often applied as quality assurance measures to validate sample data collected using
field equipment.
January 2009 D-19 NUREG-1575, Supp. 1
-------
Appendix D MARSAME
D.10.1 Instruments
This section provides a brief overview of instruments used for radiological analyses in a
laboratory setting. For additional detail on these instruments, please refer to the accompanying
section references in MARLAP.
D. 10.1.1 Instruments for the Detection of Alpha Radiation
• Alpha Spectroscopy with Multi-Channel Analyzer - This system consists of an alpha
detector housed in an evacuated counting chamber, a bias supply, amplifier, analog-to-digital
converter, multi-channel analyzer, and computer. Samples are placed at a fixed distance from
the solid state partially implanted silica for analysis, and the multi-channel analyzer yields an
energy spectrum that can be used to both identify and quantify the radionuclides. The overall
properties of the instrumentation allow for excellent peak resolution, although this technique
often requires a complex chemical separation to obtain the best results.
• Gas-Flow Proportional Counter - The system consists of a gas-flow detector, supporting
electronics, and an optional guard detector for reducing the background count rate. A thin
window can be placed between the gas-flow detector and sample to protect the detector from
contamination, or the sample can be placed directly into the detector. This system does not
typically provide data useful for identifying radionuclides unless it is preceded by nuclide-
specific chemical separations.
• Liquid Scintillation Spectrometry - Typically, samples will be subjected to chemical
separations and the resulting materials placed in a vial with a scintillation cocktail. When the
alpha particle energy is absorbed by the cocktail, light pulses are emitted, which are detected
by photomultiplier tubes. One pulse of light is emitted for each alpha particle absorbed. The
intensity of light emitted is related to the energy of the alpha. This system can provide data
useful for identifying radionuclides if the system is coupled to a multi-channel analyzer.
• Low-Resolution Alpha Spectrometry - The system consists of a small sample chamber,
mechanical pump, two-inch diameter silicon detector, multi-channel analyzer, readout
module, and a computer. Unlike alpha spectroscopy with multi-channel analyzer, this method
allows the technician to load samples for analysis without drying because the presence of
moisture generally has negligible effects on the results. This method is therefore estimated to
substantially reduce the time for analysis. However, the low resolution may limit the ability
to identify individual radionuclides in a sample containing multiple radionuclides and thus
may limit the applicability of this method (Meyer 1995).
• Alpha Scintillation Detector - This system is used primarily for the quantification of 226Ra
by the emanation and detection of 222Rn gas. The system consists of a bubbler system with
gas transfer apparatus, a vacuum flask lined with scintillating material called a Lucas Cell,9 a
photomultiplier tube, bias supply, and a sealer to record the count data.
9 One end of a Lucas cell is covered with a transparent window for coupling to a photomultiplier tube and the
remaining inside walls are coated with zinc sulfide.
NUREG-1575, Supp. 1 D-20 January 2009
-------
MARSAME Appendix D
D. 10.1.2 Instruments for the Detection of Beta Radiation
• Gas-Flow Proportional Counter - The system consists of a gas-flow detector, supporting
electronics, and an optional guard detector for reducing the background count rate. A thin
window can be placed between the gas-flow detector and sample to protect the detector from
non-fixed activity, or the sample can be placed directly into the detector. This technique does
not provide data useful for identifying individual radionuclides unless it is preceded by
nuclide-specific chemical separations.
• Liquid Scintillation Spectrometry - Typically, samples will be subjected to chemical
separations and the resulting materials placed in a vial with a scintillation cocktail. When the
beta particle energy is absorbed by the cocktail, light pulses are emitted, which are detected
by photomultiplier tubes. One pulse of light is emitted for each beta particle absorbed. The
intensity of light emitted is related to the energy of the beta. This system can provide data
useful for identifying radionuclides if the system is coupled to a multi-channel analyzer. This
system must be allowed to darken (i.e., equilibrate to a dark environment) prior to
measurement.
D. 10.1.3 Instruments for the Detection of Gamma or X-Radiation
• High-Purity Germanium Detector with Multi-Channel Analyzer - This system consists
of a germanium detector connected to a cryostat (either mechanical or a Dewar of liquid
nitrogen), high voltage power supply, spectroscopy grade amplifier, analog to digital
converter, and a multi-channel analyzer. This system has high resolution for peak energies
and is capable of identifying and quantifying individual gamma peaks in complex spectra. It
is particularly useful when a sample may contain multiple gamma-emitting radionuclides and
it is necessary to both identify and quantify all nuclides present.
• Sodium Iodide Detector with Multi-Channel Analyzer - This system consists of a sodium
iodide detector, a high voltage power supply, an amplifier, an analog to digital converter, and
a multi-channel analyzer. This system has relatively poor energy resolution and is not
effective for identifying and quantifying individual gamma peaks in complex spectra. It is
most useful when only a small number of gamma-emitting nuclides are present or when a
gross-gamma measurement is adequate.
D.10.2 Temporal Issues
Laboratory analysis is usually controlled by the turnaround time involved in preparing and
accurately measuring the collected samples. The sample matrix impacts the preparation time,
because soils and bulk chemicals typically require more extensive preparation than liquids or
smears. Table D.2 describes the typical preparation and counting times associated with the
various analytical instruments and methods described in Section D.10.1. Additional issues that
may result in extended time for sample preparation and analysis are described in MARLAP.
January 2009 D-21 NUREG-1575, Supp. 1
-------
Appendix D
MARSAME
Table D.2 Typical Preparation and Counting Times
Alpha Spectroscopy with Multi-
Channel Analyzer
Gas-Flow Proportional Counter
Liquid Scintillation Spectrometer
Low-Resolution Alpha
Spectroscopy
High-Purity Germanium (HPGe)
Detector with Multi-Channel
Analyzer
Sodium Iodide (Nal) Detector with
Multi-Channel Analyzer
Alpha Scintillation Detector
Typical Preparation Time
1 to 7 days
Hours to days
Minutes,10
hours to 2 days11
Minutes (DOE, 1995)
Minutes to 1 day
Minutes to 1 day
1 to 4 days;
4 to 28 days12
Typical Counting Time
100 to 1,000 minutes
10 to 1,000 minutes
>60 to 300 minutes
10 to 1,000 minutes
10 to 1,000 minutes
1 to 1,000 minutes
10 to 200 minutes
D.10.3 Spatial Issues
This section addresses issues related to detector-M&E geometry and provides information on the
range of impacts resulting from dissenting geometries between the calibration source and the
measured sample. Other topics may include detector dimensions and problems positioning
instruments.
D.I0.3.1 Alpha Spectroscopy with Multi-Channel Analyzer
Sample geometry (lateral positioning on a detector shelf) in some detectors may be a small
source of additional uncertainty. Uncertainty in the preparation of the actual calibration standards
as well as the applicability of the calibration standards to the sample analysis should also be
considered.
D. 10.3.2 Gas-Flow Proportional Counter
Even deposition of sample material on the planchette is critical to the analytical process. In some
analyses, ringed planchettes may aid in the even deposition of sample material. An uneven
deposition may result in an incorrect mass-attenuation correction as well as introducing a
position-dependent bias to the analysis. The latter situation arises from the fact that gas-flow
proportional counters are not radially symmetric, so rotation of an unevenly deposited sample by
45° may drastically change the instrument response.
10 Minimal preparation times are possible if the sample does not require concentration prior to being added to the
liquid scintillation cocktail vial.
1: Longer preparation times are necessary for speciation of low-energy beta emitters.
12 Longer count times represent the necessary time for in-growth of 222Rn for 226Ra analyses.
NUREG-1575, Supp. 1
D-22
January 2009
-------
MARSAME Appendix D
D. 10.3.3Liquid Scintillation Spectrometer
For gross counting, samples (e.g., smears and filters) can be placed directly into a liquid
scintillation counter (LSC) vial with liquid scintillation cocktail, and counted with no
preparation. There are samples with more complicated matrices that require chemical separation
prior to being placed and counted in LSC vials. Calibration sources are also kept and counted in
these vials, so the geometry of the source and the sample compared to the detector generally are
similar.
D.I0.3.4Low-Resolution Alpha Spectroscopy
Sample geometry (lateral positioning on a detector shelf) in some detectors may be a small
source of additional uncertainty. Uncertainty in the preparation of the actual calibration standards
as well as the applicability of the calibration standards to the sample analysis should be
considered.
D. 10.3.5 High-Purity Germanium Detector with Multi-Channel Analyzer
Geometry considerations are most important for spectroscopic gamma analyses. Sample
positioning on the detector may significantly affect the analytical results, depending on the size
and shape of the germanium crystal. Moreover, the instrument is calibrated with a source that
should be the same physical size, shape, and weight as the samples to be analyzed.13 Discrepan-
cies between the volume or density of the sample and the source introduce additional uncertainty
to the analytical results.
Sample homogeneity is a critical factor in gamma spectroscopy analyses, particularly with
relatively large samples. For example, sediment settling during the course of analysis of a turbid
aqueous sample will result in a high bias from any activity contained in the solid fraction.
Likewise, the positioning of areas containing elevated activity in a solid sample will create a bias
in the overall sample activity (the activity will be disproportionately high if the particle is located
at the bottom of the sample, and the activity will be disproportionately low if it is located at the
top of the sample).
D. 10.3.6 Sodium Iodide Detector with Multi-Channel Analyzer
The spatial considerations for Nal detectors are the same as those listed above for high-purity
germanium detectors.
D.I0.3.7 Alpha Scintillation Detectors
Accurate sample analysis depends heavily on the complete dissolution of the 226Ra or other
radionuclides of interest in the bubbler solution. Adequate sample preparation will help ensure
that spatial issues do not influence results, as the apparatus itself minimizes any other potential
geometry-related sources of error or uncertainty.
13 Some software packages allow a single calibration geometry to be modeled to assimilate the properties of other
geometries.
January 2009 D-23 NUREG-1575, Supp. 1
-------
Appendix D
MARSAME
D.I0.4 Radiation Types
Table D.3 describes the types of radiation that each laboratory instrument and method can
measure.
Table D.3 Radiation Applications for Laboratory Instruments and Methods
Alpha Spectrometry with a
Multi-Channel Analyzer
Gas-Flow Proportional Counter
Liquid Scintillation
Spectrometer
Low-Resolution Alpha
Spectroscopy
High-Purity Germanium
Detector with Multi-Channel
Analyzer
Sodium Iodide Detector with
Multi-Channel Analyzer
Alpha Scintillation Detector
Alpha
GOOD
GOOD
POOR
GOOD
NA
NA
GOOD
Beta
NA
GOOD
GOOD14
NA
NA
NA
NA
Photon
NA
POOR
POOR
NA
GOOD
GOOD
NA
Neutron
NA
NA
NA
NA
NA
NA
NA
Differentiate
Radiation
Types
NA
FAIR
FAIR
NA
NA
NA
NA
Identify Specific
Radionuclides
GOOD
POOR
FAIR
FAIR15
GOOD
FAIR
FAIR
Notes:
GOOD
FAIR
POOR
NA
The instrumentation and measurement technique is well-suited for this application
The instrumentation and measurement technique can adequately perform this application
The instrumentation and measurement technique may be poorly suited for this application
The instrumentation and measurement technique cannot perform this application
D.10.5 Range
All of the instrumentation discussed here has physical limitations as to the amount of activity
that can be analyzed. This limitation arises primarily from the ability of the detector to recover
after an ionizing event, and the speed with which the component electronics can process the data.
Typically, a count rate on the order of 106 counts per second taxes the physical limitations of
most detectors. Other practical considerations, (such as the potential to impact the detector with
non-fixed activity) often override the physical limitations of the counting system.
There are energy range limitations as well. For example: window proportional counters are poor
choices for very low energy beta emitters; some gamma spectrometers have poor efficiencies at
low energies; and some systems are not calibrated for high-energy gammas. Table D.4 describes
the energy range that each instrument and method can be used to determine, and the maximum
activity per sample that the method can be used to count.16
14
This system is designed for the detection of low-energy beta particles.
15 The low resolution may limit the ability to identify individual radionuclides in a sample containing multiple
radionuclides.
16 David Burns, Paragon Analytics, Inc., private communication with Nick Berliner, Cabrera Services, Inc., March
2005.
NUREG-1575, Supp. 1
D-24
January 2009
-------
MARSAME
Appendix D
Table D.4 Typical Energy Ranges and Maximum Activities
Alpha Spectrometry with Multi-
Channel Analyzer
Gas-Flow Proportional Counter
Liquid Scintillation Spectrometer
Low-Resolution Alpha Spectrometry
High-Purity Germanium (HPGe)
Detector with Multi-Channel Analyzer
Sodium Iodide (Nal) Detector with
Multi-Channel Analyzer
Alpha Scintillation Detector
Energy Range
3 to 8 MeV
3 to 8 MeV (a)
100 to 2,000 keV((3)
>3MeV
15 to 2,500 keV ((3);
> 1.5 MeV ((3)17
3 to 8 MeV (a)
50 to >2,000 keV (P-type
detector);
5 to 80 keV (N-type detector)
>80 to 2,000 keV
All a emission energies
Maximum Activity
<10 Bq (<270 pCi)
35 Bq (946 pCi)
100,000 Bq (2.7 uCi)
<10 Bq (<270 pCi)
370 Bq (10,000 pCi)
370 Bq (10,000 pCi)
<10 Bq (<270 pCi)
D.10.6 Scale
There is no minimum sample size required for a given analysis. Smaller sample sizes will
necessarily result in elevated detection limits. Minimum sample sizes (e.g., 0.1 gram) may be
specified in order to ensure that the sample is reasonably representative given the degree of
homogenization achieved in the laboratory. Typical liquid and solid sample sizes are noted in
Table D.5.
Table D.5 Typical Liquid and Solid Sample Sizes
Alpha Spectrometry with Multi-
Channel Analyzer
Gas-Flow Proportional Counter
Liquid Scintillation Spectrometer
Low-Resolution Alpha Spectrometry
High-Purity Germanium (HPGe)
Detector with Multi-Channel Analyzer
Sodium Iodide (Nal) Detector with
Multi-Channel Analyzer
Alpha Scintillation Detector
Typical Liquid
Sample Size
1 liter
1 liter
<10milliliters; 1 liter19
1 liter
4 liters
4 liters
1 liter
Typical Solid
Sample Size
2 grams; 50 grams18
2 grams
<0.5 grams; 500 grams
2 grams; 50 grams17
1 kilogram
1 kilogram
2 grams
Very high-energy beta emitters may be counted using liquid scintillation equipment without liquid scintillation
cocktails by the use of the Cerenkov light pulse emitted as high energy charged particles move through water or
similar substances.
18 The use of sample digestion processes allows the processing of larger sample masses.
19 Direct depositing of sample material into the scintillation cocktail limits the sample size to the smaller sample
sizes noted; prepared analyses may use substantially larger sample quantities as noted (this applies to both liquid and
solid sample matrices).
January 2009
D-25
NUREG-1575, Supp. 1
-------
Appendix D MARSAME
D.10.7 Ruggedness
Ruggedness does not hold relevance to laboratory analyses, because they are performed in a
controlled environment that precludes the instrumentation from being exposed to conditions
requiring durability.
NUREG-1575, Supp. 1 D-26 January 2009
-------
MARSAME Appendix E
E. DISPOSITION CRITERIA
E.I Department of Energy
Disposition criteria specified by DOE regulations and orders are found in the Code of Federal
Regulations, Title 10 (especially 10 CFR 835, Occupational Radiation Protection) and in
applicable DOE Orders (especially DOE Order 5400.5, Radiation Protection of the Public and
the Environment). The DOE regulations and orders govern the conduct of DOE employees and
contractors in the operation of DOE facilities and in the disposition of real property (e.g.,
buildings and land) and non-real property ("personal property" such as materials, equipment,
materials in containers, clothing, etc.). The DOE Order requirements are applicable to DOE
activities only and are enforceable as contractual provisions in most DOE contracts. DOE rules
are enforceable under 10 CFR Part 820. The following list of DOE requirements is not
exhaustive. In addition, a listing of some non-mandatory guidance documents is also provided.
E.1.1 10 CFR 835 (Non-Exhaustive Excerpts)
E.I.1.1 §835.405 Receipt of Packages Containing Radioactive Material
(a) If packages containing quantities of radioactive material in excess of a Type A quantity (as
defined at 10 CFR 71.4) are expected to be received from radioactive material transportation,
arrangements shall be made to either:
(1) Take possession of the package when the carrier offers it for delivery; or
(2) Receive notification as soon as practicable after arrival of the package at the carrier's
terminal and to take possession of the package expeditiously after receiving such
notification.
(b) Upon receipt from radioactive material transportation, external surfaces of packages known
to contain radioactive material shall be monitored if the package:
(1) Is labeled with a Radioactive White I, Yellow II, or Yellow III label (as specified at
49 CFR 172.403 and 172.436-440); or
(2) Has been transported as low specific activity material (as defined at 10 CFR 71.4) on
an exclusive use vehicle (as defined at 10 CFR 71.4); or
(3) Has evidence of degradation, such as packages that are crushed, wet, or damaged.
(c) The monitoring required by paragraph (b) of this section shall include:
(1) Measurements of removable contamination levels, unless the package contains only
special form (as defined at 10 CFR 71.4) or gaseous radioactive material; and
(2) Measurements of the radiation levels, unless the package contains less than a Type A
quantity (as defined at 10 CFR 71.4) of radioactive material.
(d) The monitoring required by paragraph (b) of this section shall be completed as soon as
practicable following receipt of the package, but not later than 8 hours after the beginning of the
working day following receipt of the package.
January 2009 E-l NUREG-1575, Supp. 1
-------
Appendix E MARSAME
E.I. 1.2 § 835.605 Labeling Items and Containers
Except as provided at § 835.606, each item or container of radioactive material shall bear a
durable, clearly visible label bearing the standard radiation warning trefoil and the words
"Caution, Radioactive Material" or "Danger, Radioactive Material." The label shall also provide
sufficient information to permit individuals handling, using, or working in the vicinity of the
items or containers to take precautions to avoid or control exposures.
E. 1.1.3 § 835.606 Exceptions to Labeling Requirements
(a) Items and containers may be excepted from the radioactive material labeling requirements of
§ 835.605 when:
(1) Used, handled, or stored in areas posted and controlled in accordance with this
subpart and sufficient information is provided to permit individuals to take
precautions to avoid or control exposures; or
(2) The quantity of radioactive material is less than one tenth of the values
specified in appendix E of this part; or
(3) Packaged, labeled, and marked in accordance with the regulations of the
Department of Transportation or DOE Orders governing radioactive material
transportation; or
(4) Inaccessible, or accessible only to individuals authorized to handle or use
them, or to work in the vicinity; or
(5) Installed in manufacturing, process, or other equipment, such as reactor
components, piping, and tanks; or
(6) The radioactive material consists solely of nuclear weapons or their
components.
(b) Radioactive material labels applied to sealed radioactive sources may be excepted from the
color specifications of § 835.601(a).
E.I.1.4 §835.1101 Control of Material and Equipment
(a) Except as provided in paragraphs (b) and (c) of this section, material and equipment in
contamination areas, high contamination areas, and airborne radioactivity areas shall not be
released to a controlled area if:
(1) Removable surface contamination levels on accessible surfaces exceed the
removable surface contamination values specified in appendix D of this part;
or
(2) Prior use suggests that the removable surface contamination levels on
inaccessible surfaces are likely to exceed the removable surface contamination
values specified in Appendix D of this part.
(b) Material and equipment exceeding the removable surface contamination values specified in
Appendix D of this part may be conditionally released for movement on-site from one
radiological area for immediate placement in another radiological area only if appropriate
monitoring is performed and appropriate controls for the movement are established and
exercised.
NUREG-1575, Supp. 1 E-2 January 2009
-------
MARSAME Appendix E
(c) Material and equipment with fixed contamination levels that exceed the total contamination
values specified in Appendix D of this part may be released for use in controlled areas outside of
radiological areas only under the following conditions:
(1) Removable surface contamination levels are below the removable surface
contamination values specified in Appendix D of this part; and (2) The material or
equipment is routinely monitored and clearly marked or labeled to alert personnel of
the contaminated status.
E.I.1.5 §83 5.1102 Control of Areas
(a) Appropriate controls shall be maintained and verified which prevent the inadvertent transfer
of removable contamination to locations outside of radiological areas under normal operating
conditions.
(b) Any area in which contamination levels exceed the values specified in appendix D of this part
shall be controlled in a manner commensurate with the physical and chemical characteristics
of the contaminant, the radionuclides present, and the fixed and removable surface
contamination levels.
(c) Areas accessible to individuals where the measured total surface contamination levels exceed,
but the removable surface contamination levels are less than, corresponding surface
contamination values specified in Appendix D of this part, shall be controlled as follows
when located outside of radiological areas:
(1) The area shall be routinely monitored to ensure the removable surface contamination
level remains below the removable surface contamination values specified in
Appendix D of this part; and
(2) The area shall be conspicuously marked to warn individuals of the contaminated
status.
(d) Individuals exiting contamination, high contamination, or airborne radioactivity areas shall
be monitored, as appropriate, for the presence of surface contamination.
(e) Protective clothing shall be required for entry to areas in which removable contamination
exists at levels exceeding the removable surface contamination values specified in Appendix
D of this part.
E.1.2 Appendix D to 10 CFR 835 - Surface Contamination Values
The data presented in Appendix D are to be used in identifying the need for posting of
contamination and high contamination areas in accordance with § 835.603(e) and (f) and
identifying the need for surface contamination monitoring and control in accordance with §§
835.1101 and 835.1102.
January 2009 E-3 NUREG-1575, Supp. 1
-------
Appendix E
MARSAME
Table E.I Surface Contamination Values1 in dpm/100 cm2 as Reported in Appendix D to 10 CFR 835
Radionuclide
U-nat, U-235, U-238, and associated decay products
Transuranics, Ra-226, Ra-228, Th-230, Th-228, Pa-231, Ac-227, 1-
125,1-129
Th-nat, Th-232, Sr-90, Ra-223, Ra-224, U-232, 1-126, 1-131, 1-133
Beta-gamma emitters (nuclides with decay modes other than alpha
emission or spontaneous fission) except Sr-90 and others noted
above5
Tritium and tritiated compounds6
Removable2'4
1,0007
20
200
1,000
10,000
Total (Fixed+
Removable)2'3
5,0007
500
1,000
5,000
N/A
The values in this appendix, with the exception noted in footnote 5, apply to radioactive contamination deposited on, but not
incorporated into the interior or matrix of, the contaminated item. Where surface contamination by both alpha-and beta-gamma-
emitting nuclides exists, the limits established for alpha-and beta-gamma-emitting nuclides apply independently.
2 As used in this table, dpm (disintegrations per minute) means the rate of emission by radioactive material as determined by
correcting the counts per minute observed by an appropriate detector for background, efficiency, and geometric factors associated
with the instrumentation.
3 The levels may be averaged over one square meter provided the maximum surface activity in any area of 100 cm2 is less than
three times the value specified. For purposes of averaging, any square meter of surface shall be considered to be above the
surface contamination value if: (1) From measurements of a representative number of sections it is determined that the average
contamination level exceeds the applicable value; or (2) it is determined that the sum of the activity of all isolated spots or
particles in any 100 cm2 area exceeds three times the applicable value.
4 The amount of removable radioactive material per 100 cm2 of surface area should be determined by swiping the area with dry
filter or soft absorbent paper, applying moderate pressure, and then assessing the amount of radioactive material on the swipe
with an appropriate instrument of known efficiency. (Note: The use of dry material may not be appropriate for tritium.) When
removable contamination on objects of surface area less than 100 cm2 is determined, the activity per unit area shall be based on
the actual area and the entire surface shall be wiped. It is not necessary to use swiping techniques to measure removable
contamination levels if direct scan surveys indicate that the total residual surface contamination levels are within the limits for
removable contamination.
5 This category of radionuclides includes mixed fission products, including the Sr-90 which is present in them. It does not apply
to Sr-90 which has been separated from the other fission products or mixtures where the Sr-90 has been enriched.
6 Tritium contamination may diffuse into the volume or matrix of materials. Evaluation of surface contamination shall consider
the extent to which such contamination may migrate to the surface in order to ensure the surface contamination value provided in
this appendix is not exceeded. Once this contamination migrates to the surface, it may be removable, not fixed; therefore, a
"Total" value does not apply.
(alpha)
DOE Order 5400.5 (Non-exhaustive Excerpts) from Chapter II
5. Release of Property Having Residual Radioactive Material
(a) Release of Real Property. Release of real property (land and structures) shall be in accordance
with the guidelines and requirements for residual radioactive material presented in Chapter
IV. These guidelines and requirements apply to both DOE-owned facilities and to private
properties that are being prepared by DOE for release. Real properties owned by DOE that
are being sold to the public are subject to the requirements of Section 120(h) of the
Comprehensive Environmental Response Compensation and Liability Act (CERCLA), as
amended, concerning hazardous substances, and to any other applicable Federal, State, and
local requirements. The requirements of 40 CFR Part 192 are applicable to properties
remediated by DOE under Title I of the Uranium Mill Tailings Radiation Control Act
(UMTRCA).
NUREG-1575, Supp. 1
E-4
January 2009
-------
MARSAME Appendix E
(b) Release of Personal Property. Personal property, which potentially could be contaminated,
may be released for unrestricted use if the results of a survey with appropriate instruments
indicate that the property is less than the contamination limits presented in Figure IV-1.
(c) Release of Materials and Equipment.
(1) Surface Contamination Levels. Prior to being released, property shall be surveyed to
determine whether both removable and total surface contamination (Including
contamination present on and under any coating) are in compliance with the levels
given in Figure IV-1 and that the contamination has been subjected to the ALARA
process.
(2) Potential for Contamination. Property shall be considered to be potentially
contaminated if it has been used or stored in radiation areas that could contain
unconfined radioactive material or that are exposed to beams of particles capable of
causing activation (neutrons, protons, etc.).
(3) Surveys. Surfaces of potentially contaminated property shall be surveyed using
instruments and techniques appropriate for detecting the limits stated in Figure IV-1.
(4) Inaccessible Areas. Where potentially contaminated surfaces are not accessible for
measurement (as in some pipes, drains, and ductwork), such property may be released
after case-by-case evaluation and documentation based on both the history of its use
and available measurements demonstrate that the unsurveyable surfaces are likely to
be within the limits given in Figure IV-1.
(5) Records. The records of released property shall include:
(a) A description or identification of the property;
(b) The date of the last radiation survey;
(c) The identity of the organization and the individual who performed the monitoring
operation;
(d) The type and identification number of monitoring instruments;
(e) The results of the monitoring operation; and
(f) The identity of the recipient of the released material.
(6) Volume Contamination. No guidance is currently available for release of material that
has been contaminated in depth, such as activated material or smelted contaminated
metals (e.g., radioactivity per unit volume or per unit mass). Such materials may be
released if criteria and survey techniques are approved by EH-1.
E.1.3 DOE Guidance and Similar Documents
The following discussion summarizes DOE policy, practice, and guidance for the disposition of
personal property, including materials and equipment from several DOE guidance documents.
"Application of DOE 5400.5 requirements for release and control of property containing residual
radioactive material," a guidance memorandum dated November 17, 1995. This guidance
memorandum explains the procedures through which authorized limits can be approved for the
disposition of waste materials to sanitary waste landfills. It also discusses the disposition criteria
for certain radionuclides. Finally, it delegates some responsibilities for the approval of release of
volumetrically contaminated materials to DOE field office managers when specified conditions
are met.
January 2009 E-5 NUREG-1575, Supp. 1
-------
Appendix E
MARSAME
Table E.2 Figure IV-1, from DOE Order 5400.5, as Supplemented in November, 1995
Memorandum: Surface Activity Guidelines - Allowable Total Residual Surface Activity
(dpm/100 cm2)1
Radionuclides2
Group 1 - Transuranics, 1-125, 1-129, Ac-227, Ra -
226, Ra-228, Th-228, Th-230, Pa-231
Group 2 - Th-natural, Sr-90, 1-126, 1-131, 1-133,
Ra-223, Ra-224, U-232, Th-232
Group 3 - U-natural, U-235, U-238, and associated
decay products, alpha emitters
Group 4 - Beta-gamma emitters (radionuclides with
decay modes other than alpha emission or
spontaneous fission) except Sr-90 and others noted
above7
Tritium (applicable to surface and subsurface)8
Average3'4
100
1,000
5,000
5,000
N/A
Maximum4'5
300
3,000
15,000
15,000
N/A
Removable4'6
20
200
1,000
1,000
10,000
As used in this table, dpm (disintegrations per minute) means the rate of emission by radioactive material as determined by
correcting the counts per minute measured by an appropriate detector for background, efficiency, and geometric factors
associated with the instrumentation.
2 Where surface contamination by both alpha- and beta-gamma-emitting radionuclides exists, the limits established for alpha- and
beta-gamma-emitting radionuclides should apply independently.
3 Measurements of average contamination should not be averaged over an area of more than 1 m2. For objects of less surface area,
the average should be derived for each such object.
4 The average and maximum dose rates associated with surface contamination resulting from beta-gamma emitters should not
exceed 0.2 mrad/h and 1.0 mrad/h, respectively, at 1 cm.
5 The maximum contamination level applies to an area of not more than 100 cm2.
6 The amount of removable material per 100 cm2 of surface area should be determined by wiping an area of that size with dry
filter or soft absorbent paper, applying moderate pressure, and measuring the amount of radioactive material on the wiping with
an appropriate instrument of known efficiency. When removable contamination on objects of surface area less than 100 cm2 is
determined, the activity per unit area should be based on the actual area and the entire surface should be wiped. It is not necessary
to use wiping techniques to measure removable contamination levels if direct scan surveys indicate that the total residual surface
contamination levels are within the limits for removable contamination.
7 This category of radionuclides includes mixed fission products, including the Sr-90 which is present in them. It does not apply
to Sr-90 which has been separated from the other fission products or mixtures where the Sr-90 has been enriched.
8 Property recently exposed or decontaminated, [sic] should have measurements (smears) at regular time intervals to ensure that
there is not a build-up of contamination over time. Because tritium typically penetrates material it contacts, the surface guidelines
in group 4 are not applicable to tritium. The Department has reviewed the analysis conducted by the DOE Tritium Surface
Contamination Limits Committee ("Recommended Tritium Surface Contamination Release Guides," February 1991), and has
assessed potential doses associated with the release of property containing residual tritium. The Department recommends the use
of the stated guideline as an interim value for removable tritium. Measurements demonstrating compliance of the removable
fraction of tritium on surfaces with this guideline are acceptable to ensure that non-removable fractions and residual tritium in
mass will not cause exposures that exceed DOE dose limits and constraints.
"Control and Release of Property with Residual Radioactive Material for use with DOE Order
5400.5, Radiation Protection of the Public and the Environment," DOE G 441.1-XX, a draft
guidance document approved for interim use and issued on May 1, 2002. This guidance
document contains detailed discussions of the disposition approaches for real and personal
property, as well as summaries of DOE's policies regarding the disposition or release of
property.
"Cross-Cut Guidance on Environmental Requirements for DOE Real Property Transfers
(Update)," DOE/EH-413/97-12, originally issued October, 1997, revised March, 2005. This
NUREG-1575, Supp. 1
E-6
January 2009
-------
MARSAME Appendix E
guidance document contains a summary of various environmental requirements for the release or
transfer of real property.
"Managing the Release of Surplus and Scrap Materials," January 19, 2001, from DOE Secretary
Richardson to all DOE elements. This memorandum provides direction as well as guidance
regarding the release of property from DOE radiological control. It also restricts the release of
metal from radiological areas for recycle until certain steps are taken by DOE.
E.2 International Organizations
In general, each country establishes its own disposition criteria for materials and equipment.
These national disposition criteria may be consistent with guidance promulgated by multi-
national organizations, such as the International Atomic Energy Agency (IAEA) or the European
Commission (EC). One example of widely accepted regulations is the "Advisory Material for the
IAEA Regulations for the Safe Transport of Radioactive Material SAFETY GUIDE No. TS-G-
1.1 (ST-2)." The references listed below provide the detailed information on guidance from the
IAEA and the EC. URLs are provided for internet access of this information. Disposition criteria
from specific nations should be obtained from those nations.
E.2.1 International Atomic Energy Agency (IAEA)
Advisory Material for the IAEA Regulations for the Safe Transport of Radioactive Material
SAFETY GUIDE No. TS-G-1.1 (ST-2):
http://www-pub.iaea.org/MTCD/publications/PDF/Publl09 scr.pdf
Planning and Preparing for Emergency Response to Transport Accidents Involving Radioactive
Material, SAFETY GUIDE No. TS-G-1.2 (ST-3)
http://www-pub.iaea.org/MTCD/publications/PDF/Publ 119_scr.pdf
Application of the Concepts of Exclusion, Exemption and Clearance SAFETY GUIDE No. RS-
G-1.7: http://www-pub.iaea.org/MTCD/publications/PDF/Pub 1202 web.pdf.
E.2.2 European Commission
The publication list for radiation protection may be found on the EC website at:
http://europa.eu.int/comm/energy/nuclear/radioprotection/publication en.htm. Contact
information for most of the authorities in the European Union may be found in Annex 3, in the
last pages of publication 139, "A review of consumer products containing radioactive substances
in the European Union," which can be found at:
http://europa.eu.int/comm/energy/nuclear/radioprotection/publication/doc/139 en.pdf.
Radiation protection publications pertaining to disposition criteria for materials and equipment
include:
134: Evaluation of the application of the concepts of exemption and clearance for
practices according to title III of Council Directive 96/29/Euratom of 13 May 1996 in EU
Member States, Volume 1, Main Report:
http://europa.eu.int/comm/energy/nuclear/radioprotection/publication/doc/134 en.pdf.
January 2009 E-7 NUREG-1575, Supp. 1
-------
Appendix E MARSAME
122: Practical Use of the Concepts of Clearance and Exemption Part I: Guidance on
General Clearance Levels for Practices:
http://europa.eu.int/comm/energy/nuclear/radioprotection/publication/doc/122_partl_en.
Edf.
122: Practical Use of the Concepts of Clearance and Exemption Part II: Application of
the Concepts of Exemption and Clearance to Natural Radiation Sources:
http://europa.eu.int/comm/energy/nuclear/radioprotection/publication/doc/122_part2_en.
Edf.
114: Definition of Clearance Levels for the Release of Radioactively Contaminated
Buildings and Building Rubble:
http://europa.eu.int/comm/energy/nuclear/radioprotection/publication/doc/l 14_en.pdf
European legislation related to the transport of radioactive materials (database):
http://europa.eu.int/comm/energy/nuclear/transport/legislation_en.htm.
E.3 Nuclear Regulatory Commission
Disposition criteria specified by NRC regulations are found in the Code of Federal Regulations,
Title 10 (10 CFR). NRC regulations in 10 CFR are structured in Parts, which apply to respective
areas of applicability. For example, 10 CFR Part 20 addresses "Standards for Protection against
Radiation." The regulatory citations below indicate the specific Part by the number to the left of
the decimal point, for example, §20.2003 is in 10 CFR Part 20, and 2003 indicates a specific
portion. In this appendix only the radiological component of those criteria pertaining to
quantitative measurement attributes are listed; there are almost always additional regulatory
requirements. "Disposition criteria" refers to the quantitative radiological portion of the complete
criteria. In some circumstances, disposition criteria are not addressed in the regulations, and
these cases are handled by existing policy and practices. A list of NRC disposition criteria, which
is not necessarily exhaustive, follows.
E.3.1 § 20.2003 Disposal by Release into Sanitary Sewerage.
(2) The quantity of licensed or other radioactive material that the licensee releases into the sewer
in 1 month divided by the average monthly volume of water released into the sewer by the
licensee does not exceed the concentration listed in table 3 of appendix B to part 20; and
(4) The total quantity of licensed and other radioactive material that the licensee releases into the
sanitary sewerage system in a year does not exceed 5 curies (185 GBq) of hydrogen-3, 1
curie (37 GBq) of carbon-14, and 1 curie (37 GBq) of all other radioactive materials
combined.
E.3.2 § 20.2005 Disposal of Specific Wastes
(a) A licensee may dispose of the following licensed material as if it were not radioactive
(1) 0.05 microcurie (1.85 kBq), or less, of hydrogen-3 or carbon-14 per gram of medium
used for liquid scintillation counting; and
(2) 0.05 microcurie (1.85 kBq), or less, of hydrogen-3 or carbon-14 per gram of animal
tissue, averaged over the weight of the entire animal.
NUREG-1575, Supp. 1 E-8 January 2009
-------
MARSAME Appendix E
E.3.3 § 35.92 Decay-in-Storage
(a) A licensee may hold byproduct material with a physical half-life of less than 120 days for
decay-in-storage before disposal without regard to its radioactivity ifit-
(1) Monitors byproduct material at the surface before disposal and determines that its
radioactivity cannot be distinguished from the background radiation level with an
appropriate radiation detection survey meter set on its most sensitive scale and with
no interposed shielding.
E.3.4 § 35.315 Safety Precautions
(4) Either monitor material and items removed from the patient's or the human research subject's
room to determine that their radioactivity cannot be distinguished from the natural
background radiation level with a radiation detection survey instrument set on its most
sensitive scale and with no interposed shielding, or handle the material and items as
radioactive waste.
E.3.5 § 36.57 Radiation Surveys
(e) Before releasing resins for unrestricted use, they must be monitored before release in an area
with a background level less than 0.5 microsievert (0.05 millirem) per hour. The resins may
be released only if the survey does not detect radiation levels above background radiation
levels. The survey meter used must be capable of detecting radiation levels of 0.5
microsievert (0.05 millirem) per hour.
E.3.6 Appendix A to Part 40-Criteria Relating to the Operation of Uranium Mills and the
Disposition of Tailings or Wastes Produced by the Extraction or Concentration of
Source Material from Ores Processed Primarily for Their Source Material Content
(6) The design requirements in this criterion for longevity and control of radon releases apply to
any portion of a licensed and/or disposal site unless such portion contains a concentration of
radium in land, averaged over areas of 100 square meters, which, as a result of byproduct
material, does not exceed the background level by more than: (i) 5 picocuries per gram
(pCi/g) of radium-226, or, in the case of thorium byproduct material, radium-228, averaged
over the first 15 centimeters (cm) below the surface, and (ii) 15 pCi/g of radium-226, or, in
the case of thorium byproduct material, radium-228, averaged over 15-cm thick layers more
than 15 cm below the surface.
E.3.7 § 71.4 Definitions
The following terms are as defined here for the purpose of this part. To ensure compatibility with
international transportation standards, all limits in this part are given in terms of dual units: The
International System of Units (SI) followed or preceded by U.S. standard or customary units.
The U.S. customary units are not exact equivalents but are rounded to a convenient value,
providing a functionally equivalent unit. For the purpose of this part, either unit may be used.
January 2009 E-9 NUREG-1575, Supp. 1
-------
Appendix E MARSAME
A] means the maximum activity of special form radioactive material permitted in a Type A
package. This value is either listed in Appendix A, Table A-l, of this part, or may be derived in
accordance with the procedures prescribed in Appendix A of this part.
A2 means the maximum activity of radioactive material, other than special form material, LSA,
and SCO material, permitted in a Type A package. This value is either listed in Appendix A,
Table A-l, of this part, or may be derived in accordance with the procedures prescribed in
Appendix A of this part.
Low Specific Activity (LSA) material means radioactive material with limited specific activity
which is nonfissile or is excepted under §71.15, and which satisfies the descriptions and limits
set forth below. Shielding materials surrounding the LSA material may not be considered in
determining the estimated average specific activity of the package contents. LSA material must
be in one of three groups:
(1) LSA-I
(i) Uranium and thorium ores, concentrates of uranium and thorium ores, and other ores
containing naturally occurring radioactive radionuclides which are not intended to be
processed for the use of these radionuclides;
(ii) Solid unirradiated natural uranium or depleted uranium or natural thorium or their
solid or liquid compounds or mixtures;
(iii) Radioactive material for which the A2 value is unlimited; or
(iv) Other radioactive material in which the activity is distributed throughout and the
estimated average specific activity does not exceed 30 times the value for exempt
material activity concentration determined in accordance with Appendix A.
(2) LSA-II
(i) Water with tritium concentration up to 0.8 TBq/L (20.0 Ci/L); or
(ii) Other material in which the activity is distributed throughout and the average specific
activity does not exceed 10~4A2/g for solids and gases, and 10~5A2/g for liquids.
(3) LSA-III. Solids (e.g., consolidated wastes, activated materials), excluding powders, that
satisfy the requirements of § 71.77, in which:
(i) The radioactive material is distributed throughout a solid or a collection of solid
objects, or is essentially uniformly distributed in a solid compact binding agent (such
as concrete, bitumen, ceramic, etc.);
(ii) The radioactive material is relatively insoluble, or it is intrinsically contained in a
relatively insoluble material, so that even under loss of packaging, the loss of
radioactive material per package by leaching, when placed in water for 7 days, would
not exceed 0.1 A2; and
(iii) The estimated average specific activity of the solid does not exceed 2xlO~3A2/g.
Low toxicity alpha emitters means natural uranium, depleted uranium, natural thorium; uranium-
235, uranium-238, thorium-232, thorium-228 or thorium-230 when contained in ores or physical
or chemical concentrates or tailings; or alpha emitters with a half-life of less than 10 days.
NUREG-1575, Supp. 1 E-10 January 2009
-------
MARSAME Appendix E
Surface Contaminated Object (SCO) means a solid object that is not itself classed as radioactive
material, but which has radioactive material distributed on any of its surfaces. SCO must be in
one of two groups with surface activity not exceeding the following limit:
(1) SCO-I: A solid object on which:
(i) The non-fixed contamination on the accessible surface averaged over 300 cm2 (or the
area of the surface if less than 300 cm2) does not exceed 4 Bq/cm2 (104
microcurie/cm2) for beta and gamma and low toxicity alpha emitters, or 0.4 Bq/cm2
(10~5 microcurie/cm2) for all other alpha emitters;
(ii) The fixed contamination on the accessible surface averaged over 300 cm2 (or the area
of the surface if less than 300 cm2) does not exceed 4><104 Bq/cm2 (1.0
microcurie/cm2) for beta and gamma and low toxicity alpha emitters, or 4* 103
Bq/cm2 (0.1 microcurie/cm2) for all other alpha emitters; and
(iii) The non-fixed contamination plus the fixed contamination on the inaccessible
surface averaged over 300 cm2 (or the area of the surface if less than 300 cm2) does
not exceed 4* 104 Bq/cm2 (1 microcurie/cm2) for beta and gamma and low toxicity
alpha emitters, or 4* 103 Bq/cm2 (0.1 microcurie/cm2) for all other alpha emitters.
(2) SCO-II: A solid object on which the limits for SCO-I are exceeded and on which:
(i) The nonfixed contamination on the accessible surface averaged over 300 cm2 (or the
area of the surface if less than 300 cm2) does not exceed 400 Bq/cm2 (102
0 0
microcurie/cm ) for beta and gamma and low toxicity alpha emitters or 40 Bq/cm
(103 microcurie/cm2) for all other alpha emitters;
(ii) The fixed contamination on the accessible surface averaged over 300 cm2 (or the area
of the surface if less than 300 cm2) does not exceed 8><105 Bq/cm2 (20
microcuries/cm2) for beta and gamma and low toxicity alpha emitters, or 8* 104
Bq/cm2 (2 microcuries/cm2) for all other alpha emitters; and
(iii) The non-fixed contamination plus the fixed contamination on the inaccessible
surface averaged over 300 cm2 (or the area of the surface if less than 300 cm2) does
S 0 0
not exceed 8x10 Bq/cm (20 microcuries/cm ) for beta and gamma and low toxicity
alpha emitters, or 8><104 Bq/cm2 (2 microcuries/cm2) for all other alpha emitters.
E.3.8 § 71.14 Exemption for Low-Level Materials
(a) A licensee is exempt from all the requirements of this part with respect to shipment or
carriage of the following low-level materials:
(1) Natural material and ores containing naturally occurring radionuclides that are not
intended to be processed for use of these radionuclides, provided the activity
concentration of the material does not exceed 10 times the values specified in
Appendix A, Table A-2, of this part.
(2) Materials for which the activity concentration is not greater than the activity
concentration values specified in Appendix A, Table A-2 of this part, or for which the
consignment activity is not greater than the limit for an exempt consignment found in
Appendix A, Table A-2, of this part.
(b) A licensee is exempt from all the requirements of this part, other than §§ 71.5 and 71.88, with
respect to shipment or carriage of the following packages, provided the packages do not
January 2009 E-ll NUREG-1575, Supp. 1
-------
Appendix E MARSAME
contain any fissile material, or the material is exempt from classification as fissile material
under § 71.15:
(1) A package that contains no more than a Type A quantity of radioactive material;
(2) A package transported within the United States that contains no more than 0.74 TBq
(20 Ci) of special form plutonium-244; or
(3) The package contains only LSA or SCO radioactive material, provided—
(i) That the LSA or SCO material has an external radiation dose of less than or equal
to 10 mSv/h (1 rem/h), at a distance of 3 m from the unshielded material; or
(ii) That the package contains only LSA-I or SCO-I material.
E.3.9 § 110.22 General License for the Export of Source Material
(3) Th-227, Th-228, U-230, and U-232 when contained in a device, or a source for use in a
device, in quantities of less than 100 millicuries of alpha activity (3.12 micrograms Th-227,
122 micrograms Th-228, 3.7 micrograms U-230, 4.7 milligrams U-232) per device or source.
E.3.10 § 110.23 General License for the Export of Byproduct Material
(2) Actinium-225 and -227, americium-241 and -242m, californium-248, -249, -250, -251, -252,
-253, and -254, curium-240, -241, -242, -243, -244, -245, -246 and -247, einsteinium-252, -
253, -254 and -255, fermium-257, gadolinium-148, mendelevium-258, neptunium-235 and -
237, polonium-210, and radium-223 must be contained in a device, or a source for use in a
device, in quantities of less than 100 millicurie of alpha activity (see Sec. 110.2 for specific
activity) per device or source, unless the export is to a country listed in Sec. 110.30. Exports
of americium and neptunium are subject to the reporting requirements listed in paragraph (b)
of this section.
(3) For americium-241, exports must not exceed one curie (308 milligrams) per shipment or 100
curies (30.8 grams) per year to any country listed in Sec. 110.29, and must be contained in
industrial process control equipment or petroleum exploration equipment in quantities not to
exceed 20 curies (6.16 grams) per device or 200 curies (61.6 grams) per year to any one
country.
(5) For polonium-210, the material must be contained in static eliminators and may not exceed
100 curies (22 grams) per individual shipment.
(6) For tritium in any dispersed form, except for recovery or recycle purposes (e.g., luminescent
light sources and paint, accelerator targets, calibration standards, labeled compounds),
exports must not exceed the quantity of 10 curies (1.03 milligrams) or less per item, not to
exceed 1,000 curies (103 milligrams) per shipment or 10,000 curies (1.03 grams) per year to
any one country. Exports of tritium to the countries listed in Sec. 110.30 must not exceed the
quantity of 40 curies (4.12 milligrams) or less per item, not to exceed 1,000 curies (103
milligrams) per shipment or 10,000 curies (1.03 grams) per year to any one country, and
exports of tritium in luminescent safety devices installed in aircraft must not exceed a
quantity of 40 curies (4.12 milligrams) or less per light source.
NUREG-1575, Supp. 1 E-12 January 2009
-------
MARSAME Appendix E
E.3.11 Policies and Practices
Disposition criteria for the release of materials and equipment that are not specified in NRC
regulations are determined by the current policies and practices. NRC's current approaches for
making decisions on disposition of solid materials is different for materials licensees, i.e.,
industrial, research, and medical facilities, and for reactors, which include power, test, and
research reactors. These are summarized in Table E-3, and discussed in more detail below.
For non-reactor licensees—materials licensees—licensee requests for release of solid material
will continue to be evaluated using the nuclide concentration tables in Regulatory Guide 1.86
and its equivalent, Fuel Cycle Policy and Guidance Directive FC 83-23. Many materials
licensees obtain approval, as a license condition, to routinely use these guidelines. For residual
radioactivity within the volume of solid materials (for example, within a concrete or soil matrix),
non-reactor licensee requests for release of solid material may continue to be approved under a
disposal request (10 CFR 20.2002); a license termination plan; decommissioning plan review; or
other specific license amendment. In verifying that the dose from such release is maintained
ALARA and below the limits of our regulations in 10 Part 20, approval of a release is possible.
The disposition of materials with volumetrically distributed radioactivity from materials
licensees is considered on a case-by-case basis with a reference of an annual individual dose
criterion of a "few mrem per year (a few 0.01 mSv/a)."
Non-reactor licensees, that is, materials licensees, and reactor licensees have essentially the same
detection level criteria for surface activity. But for materials licensees, radioactivity below these
detection level criteria is allowed—detectable radioactivity is not allowed at any level for reactor
licensees.
For reactor licensees, licensees may release of solid material using the "no detectable" policy of
NRC's Inspection and Enforcement Circular 81-07 and Information Notices 85-92 and 88-22.
For reactors, the policy is that released material can have no detectable licensed radioactivity.
The levels of detection are specified by each reactor licensee's procedures and are frequently
consistent with a now discontinued Regulatory Guide issued in 1974. In practice, these detection
levels for radioactivity on surfaces are: 5/6 Bq /cm2 (5,000 dpm/100 cm2) total 0-y and Ve
Bq/cm2 (1,000 dpm/100 cm2) removable 0-y. Non-detection at these levels of detectability was
considered to result in potential doses to an individual significantly less than 5 mrem/y («0.05
mSv/a) from any non-detectable radioactivity that could remain on surfaces.
Detection levels for a-emitting radioactivity are specified as 1/60 Bq/cm2 (100 dpm/100 cm2)
total and 1/300 Bq/cm2 (20 dpm/100 cm2) for removable a-emitting radioactivity. For volumetric
radioactivity from reactors, the detection levels are from guidance written in the late 1970s and
specifies 0-y concentrations in the general range of 3-4 Bq/kg (81-108 pCi/kg).
January 2009 E-13 NUREG-1575, Supp. 1
-------
Appendix E
MARSAME
Table E.3 Summary of NRC Disposition Criteria from Current Practices for the Release of
Materials and Equipment
Reactor Licenses
Materials Licenses
Surficial Radioactivity
p-y: Non-detectable [MDC 5/6
Bq/cm2; 1/6 Bq/cm2 removable]
a: Non-detectable [MDC 1/60
Bq/cm2; 1/300 Bq/cm2
removable]
P-y: 5/6 Bq/cm2; 1/6 Bq/cm2
removable1
a: 1/60 Bq/cm2; 1/300 Bq/cm2
removable2
Volumetric Radioactivity
P-y: Non-detectable [MDC in
General range of ~ 3-4 Bq/kg]
a: Non-detectable [MDC not
indicated]
P-y: Case-by-case [Reference to a
few 0.01 mSv in a year]
a: Case-by-case [Reference to a
few 0.01 mSv in a year]
1 Except Sr-90,1-126,1-131, and 1-133, where 1/6 Bq/cm2and 1/30 Bq/ cm2 removable applies; and except 1-125,
and 1-129 where 1/60 Bq/cm2 and 1/300 Bq/cm2 removable applies.
2Except natural U, U-235, U-238, and associated decay products where 5/6 Bq/cm2 and 1/6 Bq/cm2 removable
applies; and except transuranics, Ra-226, Ra-228, Th-230, Th-228, Pa-231, and Ac-227, where 1/60 Bq/cm2 and
1/300 Bq/cm2 removable applies.
E.3.12 Issues Related to International Trade
With regard to issues relating to international trade of solid materials released from facilities,
NRC's regulations contain requirements for export and import of material and could be
considered in handling materials that meet established international clearance criteria and, at the
same time, do not meet the guidelines for NRC licensees. Among other things, these regulations
require that "the proposed import does not constitute an unreasonable risk to the public health
and safety."
NUREG-1575, Supp. 1
E-14
January 2009
-------
MARSAME References
REFERENCES
10 CFR20.1003. Standards for Protection Against Radiation, Definitions. Code of Federal
Regulations (CFR) Title 10, Energy, Part 20.1003, December 19, 2002.
http://www.nrc.gov/reading-rm/doc-collections/cfr/part020/ (accessed September 1,
2008).
10 CFR 36.57. Licenses and Radiation Safety Requirements for Irradiators, Radiation Surveys.
Code of Federal Regulations (CFR) Title 10, Energy, Part 36.57, January 1, 2005.
http://www.nrc.gov/reading-rm/doc-collections/cfr/part036/part036-0057.html (accessed
September 1, 2008).
10 CFR 50.2. Domestic Licensing of Production and Utilization Facilities, Definitions. Code of
Federal Regulations (CFR) Title 10, Energy, Part 50.2, January 13, 1998.
http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/ (accessed September 1,
2008).
10 CFR 71.4. Packaging And Transportation of Radioactive Material. Code of Federal
Regulations (CFR) Title 10 (Energy), Part 71.4, September 28, 1995.
http://www.nrc.gov/reading-rm/doc-collections/cfr/part071/full-text.html (accessed
October 24, 2008).
10 CFR 820. Procedural Rules for DOE Nuclear Activity. Code of Federal Regulations (CFR)
Title 10, Energy, Part 820, August 17, 1993.
http://www.hss.energy.gov/enforce/rands/10CFR820.pdf (accessed October 24, 2008).
10 CFR 835. Occupational Radiation Protection. Code of Federal Regulations (CFR) Title 10,
Energy, Part 835, Subpart E, "Monitoring of Individuals and Areas," 401, "General
Requirements" November 4, 1998. http://www.eh.doe.gov/radiation/10cfr835/835gpo.pdf
(accessed September 1, 2008).
10 CFR 172. Hazardous Materials Table, Special Provisions, Hazardous Materials
Communications, Emergency Response Information, Training Requirements, and
Security Plans. Code of Federal Regulations (CFR) Title 49 (Transportation), Subpart E,
"Labeling," 403, "Class 7 (radioactive) material" April 15, 1976, 436, "Radioactive
White-I Label," 438, "Radioactive Yellow-II Label," 440 "Radioactive Yellow-Ill Label"
December 20, 1991.http://ecfr.gpoaccess.gov/cgi/t/text/text-
idx?c=ecfr:sid=f5b01bccf82655893c893dl4885a307f:rgn=div5:view=text:node=49%3A
2.1.1.3.8:idno=49:cc=ecfr#49:2.1.1.3.8.9.25.22 (accessed October 24, 2008).
49 CFR 173.433. Requirements for Determining Basic Radionuclide Values and for the Listing
ofRadionuclides on Shipping Papers and Labels. Code of Federal Regulations (CFR)
Title 49, Transportation, Part 173.433, September 23, 2005.
http ://ecfr. gpoaccess. gov/cgi/t/text/text-
idx?c=ecfr:sid=f5b01bccf82655893c893dl4885a307f:rgn=div5:view=text:node=49%3A
2.1.1.3.8:idno=49:cc=ecfr#49:2.1.1.3.8.9.25.22 (accessed October 24, 2008).
Abelquist, E. 2001. Decommissioning Health Physics: A Handbook for MARSSIM Users.
Institute of Physics Publishing, Philadelphia, PA.
January 2009 Ref-1 NUREG-1575, Supp. 1
-------
References MARSAME
American National Standards Institute (ANSI) 1989. Performance Specifications for Health
Physics Instrumentation-Portable Instrumentation for Use in Extreme Environmental
Conditions. ANSIN42.17C.
American National Standards Institute (ANSI) 1994. Calibration and Usage of Thallium-
Activated Sodium Iodide Detector Systems for Assay ofRadionuclides. ANSI N42.12.
American National Standards Institute (ANSI) 1997. Radiation Protection Instrumentation Test
and Calibration, Portable Survey Instruments. ANSI N323 A.
American National Standards Institute (ANSI) 1999. Surface and Volume Radioactivity
Standards for Clearance. ANSIN13.12.
American National Standards Institute (ANSI) 2003 a. Performance Criteria for Hand-held
Instruments for the Detection and Identification ofRadionuclides. ANSI N42.34.
American National Standards Institute (ANSI) 2003b. Performance Specifications for Health
Physics Instrumentation-Portable Instrumentation for Use in Normal Environmental
Conditions. ANSIN42.17A.
American National Standards Institute (ANSI) 2004. American National Standard for Evaluation
and Performance of Radiation Protection Portal Monitors for Use in Homeland Security.
ANSIN42.35.
Beck H.L., Gogolak C.V., Miller K.M., and Lowder W.M. 1980. "Perturbations on the Natural
Radiation Environment Due to the Utilization of Coal as an Energy Source," in Natural
Radiation Environment III, U.S. Department of Energy CONF-780422.
BIL 2005. "lonSens® 208 Large Item Monitor." BIL Solutions.
http://www.bilsolutions.co.uk/pdf/datasheets new/ionsens2081argeitemsmonitor.pdf
(accessed September 1, 2008).
Canberra 2005a. "Considerations for Environmental Gamma Spectroscopy Systems." Canberra,
Inc. http://www.canberra.com/literature/972.asp (accessed September 1, 2008).
Canberra 2005b. "RadSentry™ Security Portals for SNM and Other Radionuclides." Canberra,
Inc. http://www.canberra.com/products/1211 .asp (accessed September 1, 2008).
Canberra 2008. "Automatic Conveyor Monitor for Soil and Debris." Canberra, Inc.
http://www.canberra.com/products/795.asp (accessed September 1, 2008).
Currie, L.A. 1968. "Limits for Qualitative Detection and Quantitative Determination:
Application to Radiochemistry." Analytical Chemistry 40(3): 586-593.
DOE 1987. The Environmental Survey Manual, Appendix A - Criteria for Data Evaluation.
DOE/EH-0053, DOE, Office of Environmental Audit, Washington, D.C. (DE88-000254),
August.
DOE 1993. Radiation Protection of the Public and the Environment. DOE Order 5400.5, Change
2, U.S. Department of Energy, Washington, DC, January.
http://homer.ornl.gov/oepa/guidance/risk/54005.pdf (accessed September 1, 2008).
NUREG-1575, Supp. 1 Ref-2 January 2009
-------
MARSAME References
DOE 2005. RESRAD Recycle Version 3.10. U.S. Department of Energy, Argonne National
Laboratory, http://web.ead.anl.gov/resrad/home2/index.cfm (accessed September 1,
2008).
Eberline 2004. "Segmented Gate System." Eberline Services.
http://www.eberlineservices.com/documents/SGSBrochure OOP.pdf (accessed September
1, 2008).
Eckerman, K.F., Westfall, R.J., Ryman, J.C., and Cristy, M. 1993. Nuclear Decay Data Files of
the Dosimetry Research Group, ORNL-TM-12350.
Eicholz G.G., Clarke F.J., and Kahn, B. 1980. "Radiation Exposure From Building Materials," in
Natural Radiation Environment III., U.S. Department of Energy CONF-780422.
Electric Power Research Institute (EPRI) 2003. Operational Changes and Impacts on LLW
Scaling Factors. D. James, EPRI Report 1008017, Electric Power Research Institute,
Palo Alto, CA, December.
U.S. Environmental Protection Agency (EPA) 1980. Upgrading Environmental Radiation Data,
Health Physics Society Committee Report HPSR-1. EPA 520/1-80-012, EPA, Office of
Radiation Programs, Washington, D.C. (PBS 1-100364), August.
U.S. Environmental Protection Agency (EPA) 1992a. Guidance for Data Usability in Risk
Assessment, Part A. Office of Solid Waste and Emergency Response (OSWER)
Directive 9285.7-09A, Environmental Protection Agency, Office of Emergency and
Remedial Response, Washington, D.C. (PB92-963356), April.
U.S. Environmental Protection Agency (EPA) 1992b. Guidance for Data Usability in Risk
Assessment, Part B. Office of Solid Waste and Emergency Response (OSWER)
Directive 9285.7-09B, Environmental Protection Agency, Office of Emergency and
Remedial Response, Washington, D.C. (PB92-963362), May.
U.S. Environmental Protection Agency (EPA) 2000. Evaluation of EPA 's Guidelines for
Technologically Enhanced Naturally Occurring Radioactive Materials (TENORM).
Report to Congress, EPA 402-R-00-01, U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency (EPA) 2001. Guidance for Preparing Standard
Operating Procedures, EPA QA/G-6. EPA/240/B-01/004, U.S. Environmental Protection
Agency, Office of Environmental Information, Washington, DC, March.
U.S. Environmental Protection Agency (EPA) 2002a. Guidance for Quality Assurance Project
Plans, EPA QA/G-5. EPA/240/R-02/009, U.S. Environmental Protection Agency, Office
of Environmental Information, Washington, DC, December.
U.S. Environmental Protection Agency (EPA) 2002b. Calculating Upper Confidence Limits For
Exposure Point Concentrations At Hazardous Waste Sites. Office of Solid Waste and
Emergency Response (OSWER) Directive 9285.6-10, Environmental Protection Agency,
Office of Emergency and Remedial Response, Washington, DC.
January 2009 Ref-3 NUREG-1575, Supp. 1
-------
References MARSAME
U.S. Environmental Protection Agency (EPA) 2002c. Guidance for Developing Quality Systems
for Environmental Programs, EPA QA/G-1. EPA/240/R-02/008, U.S. Environmental
Protection Agency, Office of Environmental Information, Washington, DC, November.
U.S. Environmental Protection Agency (EPA) 2006a. Guidance on Systematic Planning Using
the Data Quality Objectives Process, EPA QA/G-4. EPA/240/B-06/001, U.S.
Environmental Protection Agency, Office of Environmental Information, Washington,
DC, February.
U.S. Environmental Protection Agency (EPA) 2006b. Data Quality Assessment: Statistical Tools
for Practitioners, EPA QA/G-9R. EPA/240/B-06/002, U.S. Environmental Protection
Agency, Office of Environmental Information, Washington, DC, February.
U.S. Environmental Protection Agency (EPA) 2006c. Data Quality Assessment: A Reviewer's
Guide, EPA QA/G-9S. EPA/240/ B-06/003, U.S. Environmental Protection Agency,
Office of Environmental Information, Washington, DC, February.
U.S. Environmental Protection Agency (EPA) 2006d. Technical Support Center for Monitoring
and Site Characterization: Software for Calculating Upper Confidence Limits (UCLs),
ProUCL Version 4.00.02. http://www.epa.gov/esd/tsc/software.htm (accessed September
1, 2008).
European Commission for Nuclear Safety and the Environment (EC) 1998. Handbook on
Measurement Methods and Strategies at Very Low Levels and Activities. Report 17624,
ISBN 92-828-3163-9.
Fuji 2008. Radioactive Contamination Monitors. Hasegawa, T., Hashimoto, T. Hashimoto, M.
Fuji Electronics, http://www.fujielectric.co.jp/eng/company/tech/pdf/r50-4/05.pdf
(accessed September 1, 2008).
Gilbert, R.O. and Simpson, J.C. 1992. Statistical Methods for Evaluating the Attainment of
Cleanup Standards. Pacific Northwest Laboratory Report PNL-7409, Rev. 1,
http://www.osti.gov/energycitations/product.biblio.jsp?osti_id=6637176 (accessed
November 10, 2008).
GUM Workbench 2006. GUM Workbench Version 1.2 Demo Installer and User Manual.
http://www.gum.dk/download/download.html (accessed September 1, 2008).
Hatch, L. L., Rentes, P. G., Godbey, F. W., and Schrems, E. L. 1978. Self-Evaluation of
Occupational Safety and Health Programs. DHEW (NIOSH) 78-187," U.S. Department
of Health, Education, and Welfare.
Hobbs T.G. 2000. Radioactivity Measurements on Glazed Ceramic Surfaces. J. Res. Natl. Inst.
Stand. Technol., 105,275-283.
International Commission on Radiological Protection (ICRP) 1989. Optimization and Decision-
Making in Radiological Protection: Annals of the ICRP Volume 20/1. ICRP Publication
55.
NUREG-1575, Supp. 1 Ref-4 January 2009
-------
MARSAME References
International Organization for Standardization (ISO) 1988. Evaluation of Surface Contamination
-Part 1: Beta Emitters and Alpha Emitters. ISO-7503-1 (1st Edition), Geneva,
Switzerland.
International Organization for Standardization (ISO) 1995. Guide to the Expression of
Uncertainty in Measurement. ISO Guide 98, GUM, Geneva, Switzerland.
International Organization for Standardization (ISO) 1996. International Vocabulary of Basic
and General Terms in Metrology. ISO Guide 99, VIM, Geneva, Switzerland, 1996.
International Organization for Standardization (ISO) 1997. Capability of Detection-Part 1:
Terms and Definitions. ISO 11843-1, Geneva, Switzerland.
Knoll, G.F. 1999. Radiation Detection and Measurement, 3rd Edition. John Wiley & Sons, Inc.,
New York, NY.
Kragten 1994. J. Analyst, 119, 2161-2165.
Laurus 2001. "Gamma Solid Waste Monitor WM-295." Laurus Systems, Inc..
http://www.laurussystems.com/products/gamma solid waste monitor.htm (accessed
September 1, 2008).
Lewis V., Woods M., Burgess P., Green S., Simpson J., and Wardle J. 2000. The Measurement
Good Practice Guide No. 49: The Assessment of Uncertainty in Radiological Calibration
and Testing. National Physical Laboratory, Teddington, Middlesex, UK, TW11 OLW,
February, http://eig.unige.ch/nucleaire/articles/gpg49.pdf, (accessed September 1, 2008).
MARLAP 2004. Multi-Agency Radiological Laboratory Analytical Protocols Manual. Nuclear
Regulatory Commission NUREG-1576, Environmental Protection Agency EPA 402-B-
04-001 A, National Technical Information Service NTIS PB2004-105421, July.
http://www.epa.gov/radiation/marlap/links.html (accessed November 10, 2008).
MARSSIM 2002. Multi-Agency Radiation Survey and Site Investigation Manual (Revision 1).
Nuclear Regulatory Commission NUREG-1575 Rev. 1, Environmental Protection
Agency EPA 402-R-97-016 Rev. 1, Department of Energy DOE EH-0624 Rev. 1,
August, http://www.epa.gov/radiation/marssim/obtain.html (accessed November 10,
2008).
McCroan 2006. GUMCalc freeware. http://www.mccroan.com/GumCalc.htm (accessed
September 1, 2008).
Meyer, K. and Lucas, A. 1995. "Assays of Thick Soil Samples Using Low-Resolution Alpha
Spectroscopy." 41st Bioassay Analytical and Environmental Radiochemistry Conference,
1995. http://www.lanl.gOv/BAER-Conference/BAERCon-41p038.pdf (accessed
September 1, 2008).
Miller E,. Peters J., Nichols D. 2000. "Release Surveying of Scrap Metals with the lonSens™
Conveyor." Waste Management 2000 Conference, Tucson, AZ, February.
http://www.bilsolutions.co.uk/file_download.php?file_id=4 (accessed September 1,
2008).
January 2009 Ref-5 NUREG-1575, Supp. 1
-------
References MARSAME
National Academy of Sciences (NAS) 1999. Evaluation of Guidelines for Technologically
Enhanced Naturally Occurring Radioactive Materials (TENORM). Committee on
Evaluation of EPA Guidelines for Exposure to Naturally Occurring Radioactive
Materials, National Research Council, National Academy of Sciences, National Academy
Press, p. 72.
National Bureau of Standards (NBS) 1963. Experimental Statistics. NBS Handbook 91, National
Bureau of Standards, Gaithersburg, MD.
National Council on Radiation Protection and Measurements (NCRP) 1988a. Exposure of the
Population in the United States and Canada from Natural Background Radiation. NCRP
Report No. 94, Bethesda, MD.
National Council on Radiation Protection and Measurements (NCRP) 1988b. Exposure of the
U.S. Population from Consumer Products and Miscellaneous Sources. NCRP Report No.
95, Bethesda, MD.
National Council on Radiation Protection and Measurements (NCRP) 1993. Limitation of
Exposure to Ionizing Radiation. NCRP Report No. 116, MD.
National Council on Radiation Protection and Measurements (NCRP) 2002. Managing
Potentially Radioactive Scrap Metal. NCRP Report No. 141, Bethesda, MD.
National Institute of Standards and Technology (NIST) 1994. Guidelines for Evaluating and
Expressing the Uncertainty of NIST Measurement Results. NIST Technical Note 1297,
Gaithersburg, MD. http://phvsics.nist.gov/Document/tnl297.pdf (accessed September 1,
2008).
National Institute of Standards and Technology (NIST) 1996. Tables of X-Ray Mass Attenuation
Coefficients and Mass Energy-Absorption Coefficients. Hubbell, J.H. and Seltzer, S.M.,
Gaithersburg, MD, April. http://physics.nist.gov/PhysRefData/XrayMassCoef/cover.html
(accessed September 1, 2008).
National Institute of Standards and Technology (NIST) 1998. XCOM: Photon Cross Sections
Database. http://physics.nist.gov/PhysRefData/Xcom/Text/XCOM.html (accessed
September 1, 2008).
National Institute of Standards and Technology (NIST) 2006. "NIST/SEMATECH e-Handbook
of Statistical Methods." http://www.itl.nist.gov/div898/handbook/ (accessed September 1,
2008).
Novelec 200la. "Technical Information Systems and Machines." Novelec Nuclear
Instrumentation, 2001. http://www.novelec.fr/descriptionUK-
CRL01 .htm#T2C%20and%20T2C-AB (accessed September 1, 2008).
Novelec 2001b. "Technical Information Radiation Survey." Novelec Nuclear Instrumentation,
2001. http://www.novelec.fr/descriptionUK-MSA.htmtfMSA-P and MSA-C (accessed
September 1, 2008).
NUREG-1575, Supp. 1 Ref-6 January 2009
-------
MARSAME References
U.S. Nuclear Regulatory Commission (NRC) 1977. Operating Philosophy for Maintaining
Occupational Radiation Exposures As Low As Is Reasonably Achievable. Regulatory
Guide 8.10. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1982. Information Relevant to Ensuring that
Occupational Radiation Exposures at Medical Institutions Will Be As Low As Reasonably
Achievable. Regulatory Guide 8.18. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1984. Lower Limit of Detection: Definition and
Elaboration of a Proposed Position for Radiological Effluent and Environmental
Measurements. NUREG/CR-4007. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) and Oak Ridge Associated Universities 1992.
Manual for Conducting Radiological Surveys in Support of License Termination, Draft
Report for Comment. NUREG/CR-5849. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1993. ALARA Levels for Effluents from Materials
Facilities. Regulatory Guide 8.37. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1994. Background as a Residual Radioactivity
Criterion for Decommissioning, Draft Report. NUREG-1501. Office of Nuclear
Regulatory Research, Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1995. Proposed Methodologies for Measuring
Low Levels of Residual Radioactivity for Decommissioning. NUREG-1506, Draft
Report for Comment. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1998a. A Nonparametric Statistical Methodology
for the Design and Analysis of Final Status Decommissioning Surveys (Revision 1).
NUREG-1505 Rev. 1. Office of Nuclear Regulatory Research, Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 1998b. Minimum Detectable Concentrations with
Typical Radiation Survey Instruments for Various Contaminants and Field Conditions.
NUREG-1507. Office of Nuclear Regulatory Research, Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 2000. Low-Level Waste Classification,
Characterization, and Assessment: Waste Streams and Neutron-Activated Metals.
NUREG/CR-6567, D.E. Robertson, C.W. Thomas, S.L. Pratt, E.A. Lepel, and V.W.
Thomas, Pacific Northwest Laboratory.
U.S. Nuclear Regulatory Commission (NRC) 2002a. Radiological Surveys for Controlling
Release of Solid Materials. NUREG-1761. Office of Nuclear Regulatory Research,
Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 2002b. Information Relevant to Ensuring that
Occupational Radiation Exposures at Uranium Mills Will Be As Low As Is Reasonably
Achievable. Regulatory Guide 8.31. Washington, DC.
U.S. Nuclear Regulatory Commission (NRC) 2003a. Radiological Assessments for Clearance
of Materials from Nuclear Facilities. NUREG-1640. Office of Nuclear Regulatory
Research, Washington, DC.
January 2009 Ref-7 NUREG-1575, Supp. 1
-------
References MARSAME
U.S. Nuclear Regulatory Commission (NRC) 2003b. Radiological Toolbox Version 1.0.0.
Eckerman, K.F. and Sjoreen, A.L. http://www.nrc.gov/about-
nrc/regulatory/research/radiological-toolbox.html (accessed September 1, 2008).
Occupational Safety and Health Administration (OSHA) 2002. Job Hazard Analysis. OSHA
Publication 3071. http://www.osha.gov/Publications/osha3071 .pdf (accessed September
1, 2008).
Pacific Northwest Laboratory (PNL) 1988. Health Physics Manual of Good Practices for
Reducing Radiation Exposure to Levels that are As Low As Reasonably Achievable
(ALARA). Munson, L.H., et al. PNL-6577, UC-610,, Richland, Washington.
http://www.pnl.gov/bavesian/refs/ALARAMAN88 .PDF (accessed September 1, 2008).
Srom, DJ. Counting Statistics Utility for Comparing Eight Decision Rules.
http://www.pnl.gov/bayesian/strom/strompub.htm (accessed September 1, 2008).
Thermo 2005. "SAM11 Small articles and tools monitor for low level gamma measurement"
Thermo Fisher Scientific, Inc.
http://www.thermo.com/eThermo/CMA/PDFs/Product/productPDF_23747.pdf (accessed
September 1, 2008).
Thermo 2008. "FHT 3031 CCM Contamination Monitor." Thermo Fisher Scientific, Inc..
http://www.thermo.eom/com/cda/product/detail/0,1055,15815,00.html (accessed
September 1, 2008).
United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) 2000.
Sources and Effects of Ionizing Radiation. Report to the General Assembly, ISBN 92-1-
142238-8. http://www.unscear.org/unscear/en/publications.html (accessed September 1,
2008).
Vetter, T.W. 2006. Quantifying Measurement Uncertainty in Analytical Chemistry - A Simplified
Practical Approach. http://www.cstl.nist.gov/acd/839.03/Uncertainty.pdf (accessed
September 1, 2008).
NUREG-1575, Supp. 1 Ref-8 January 2009
-------
Glossary
GLOSSARY
Accessible Area is an area that can be easily reached or obtained. In many cases an area must be
physically accessible to perform a measurement. However, radioactivity may be measurable
even if an area is not physically accessible. See in this glossary measurable radioactivity.
Action Level is the numerical value that causes a decision-maker to choose one of the alternative
actions. In the context of MARSAME, the numerical value is the radionuclide concentration or
level of radioactivity corresponding to the disposition criterion, and the alternative actions are
determined by the selection of a disposition option.
Alternative Action is the choice between two mutually exclusive possibilities. See in this
glossary decision rule.
Ambient Radiation is radiation that is currently present in the surrounding area. Ambient
radiation may include natural background, intrinsic radiation from surrounding materials,
intrinsic radiation from the item(s) being measured, contamination, or radiation from nearby
machines (e.g., x-ray machines when operating) depending on the local conditions. Ambient
radiation changes with season, time, location, weather, and other environmental conditions.
Background Radiation (as defined in Nuclear Regulatory Commission regulations) is radiation
from cosmic sources; naturally occurring radioactive material including radon (except as a decay
product of source or special nuclear material); and global fallout as it exists in the environment
from the testing of nuclear explosive devices or from past nuclear accidents such as Chernobyl
that contribute to background radiation and are not under the control of the licensee.
"Background radiation" does not include radiation from source, byproduct or special nuclear
materials regulated by the Nuclear Regulatory Commission (10 CFR 20.1003). See in this
glossary distinguishable from background.
Biased Measurements are measurements performed at locations selected using professional
judgment based on unusual appearance, location relative to known contamination areas, high
potential for residual radioactivity, and general supplemental information. Biased measurements
are not included in the statistical evaluation of survey unit data because they violate the
assumption of randomly selected, independent measurements. Instead, biased measurement
results are individually compared to the action levels. Biased measurements are also called
judgment measurements (MARSSIM 2002).
Calibration Function is the function that relates the net instrument signal to activity (e.g.,
relates counts to disintegrations or radiations).
Categorization is the act of determining whether M&E are impacted or non-impacted. This is a
departure from MARSSIM where this decision was included in the definition of classification.
Class 1 M&E are impacted M&E that have, or had, the following: (1) highest potential for, or
known, radionuclide concentration(s) or radioactivity above the action level(s); (2) highest
potential for small areas of elevated radionuclide concentration(s) or radioactivity; and (3)
insufficient evidence to support reclassification as Class 2 or Class 3. Such potential may be
based on historical information and process knowledge, while known radionuclide
January 2009 GL-1 NUREG-1575, Supp. 1
-------
Glossary MARSAME
concentration(s) or radioactivity may be based on preliminary surveys. See in this glossary Class
2, Class 3, classification, and impacted.
Class 2 M&E are impacted M&E that have, or had, the following: (1) low potential for
radionuclide concentration(s) or radioactivity above the action level(s); and (2) little or no
potential for small areas of elevated radionuclide concentration(s) or radioactivity. Such potential
may be based on historical information, process knowledge, and preliminary surveys. See in this
glossary Class 1, Class 3, classification, and impacted.
Class 3 M&E are impacted M&E that have, or had, the following: (1) little or no potential for
radionuclide concentrations(s) or radioactivity above background; and (2) insufficient evidence
to support categorization as non-impacted. See in this glossary Class 1, Class 2, classification,
impacted, and non-impacted.
Classification is the act or result of separating impacted M&E or survey units into one of three
designated classes: Class 1, Class 2, or Class 3. Classification is the process of determining the
appropriate level of survey effort based on estimates of activity levels and comparison to action
levels, where the activity estimates are provided by historical information, process knowledge,
and preliminary surveys. See in this glossary Class 7, Class 2, Class 3, and impacted.
Clearance is the removal of radiological regulatory controls from materials and equipment.
Clearance is a subset of release. See in this glossary release, restricted release, and unrestricted
release.
Combined Standard Uncertainty is the standard uncertainty of an output estimate calculated
by combining the standard uncertainties of the input estimates. The combined standard
uncertainty of^ is denoted by uc(y). See also in this glossary expanded uncertainty, input
estimate, measurement method uncertainty, output estimate, and standard uncertainty .
Combined Variance is the square of the combined standard uncertainty. The combined variance
is denoted by [wc(y)]2. See in this glossary combined standard uncertainty .
Concentration is activity per unit mass or volume (e.g., Bq/kg, pCi/g, or Bq/m3) or activity per
unit area (e.g., Bq/m2 or dpm/100 cm2).
Conceptual Model is an idealized model or map of a component or area to be surveyed and the
associated radionuclides or radioactivity expected to be present, and is intended to aid in
describing or designing the survey. The initial conceptual model is based on the results of the
initial assessment. Additional data is used to update the conceptual model iteratively throughout
the development, implementation, and assessment of the disposition survey. See in this glossary
initial assessment.
Coverage Factor (K) is the value multiplied by the combined standard uncertainty uc(y) to give
the expanded uncertainty, U. See in this glossary combined standard uncertainty and expanded
uncertainty.
Coverage Probability is the approximate probability that the reported uncertainty interval will
contain the value of the measurand. See in this glossary measurand.
NUREG-1575, Supp. 1 GL-2 January 2009
-------
Glossary
Critical Value in the context of radiation detection is the minimum measured value (e.g., of the
instrument signal or the radionuclide concentration) required to give a specified probability that a
positive (non-zero) amount of radioactivity is present in the material being measured. The critical
value is the same as the critical level or decision level in publications by Currie (Currie 1968 and
NRC 1984).
Data Life Cycle is the process of planning the survey, implementing the survey plan, and
assessing the survey results prior to making a decision (MARSSIM 2002).
Data Quality Objectives (DQOs) are qualitative and quantitative statements derived from the
DQO process that clarify [the survey] technical and quality objectives, define the appropriate
type of data, and specify tolerable levels of potential decision errors that will be used as the basis
for establishing the quality and quantity of data needed to support decisions (MARSSIM 2002).
Data Quality Objectives Process is a systematic strategic planning tool based on the scientific
method that identifies and defines the type, quality, and quantity of data needed to satisfy a
specific use (MARSSIM 2002). See also in this glossary data quality objectives.
Data Quality Assessment (DQA) is a scientific and statistical evaluation that determines
whether data are the right type, quality and quantity to support their intended use (EPA 2006b).
Decision Rule is an "if.. .then" statement consisting of three parts: action level(s), parameter of
interest, and alternative actions. A theoretical decision rule is developed early in the planning
process assuming ideal data are available to support a disposition decision (see Chapter 3). An
operational decision rule is developed based on the measurements that will be performed as part
of the final disposition survey (see Chapter 4).
Detection Capability is a generic term describing the capability of a measurement process to
distinguish small amounts of radioactivity from zero. It may be expressed in terms of the
minimum detectable concentration. See in this glossary minimum detectable concentration.
Difficult-to-Measure Radioactivity is radioactivity that is not measurable until the M&E to be
surveyed is prepared. Preparation of M&E may be relatively simple (e.g., cleaning) or more
complicated (e.g., disassembly or complete destruction). Given sufficient resources, all
radioactivity can be made measurable; however, it is recognized that increased survey costs can
outweigh the benefit of some dispositions.
Discrimination Limit is the level of radioactivity selected by the members of the planning team
that can be reliably distinguished from the action level. The lower bound of the gray region
(LBGR) for Scenario A and the upper bound of the gray region (UBGR) for Scenario B are
examples of discrimination limits. See also in this glossary lower bound of the gray region.,
upper bound of the gray region, Scenario A, and Scenario B.
Disposition is the future use, fate, or final location for something (e.g., recycle, reuse, disposal).
Disposition Decision is the selection among alternative actions to determine acceptable future
use, fate, or final location for something (e.g., recycle, reuse, disposal)..
Disposition Survey is a radiological survey designed to collect information to support a
disposition decision.
January 2009 GL-3 NUREG-1575, Supp. 1
-------
Glossary MARSAME
Distinguishable from Background is the radionuclide concentration or radioactivity that is
statistically different from the background level of that radionuclide concentration or
radioactivity in similar M&E. See in this glossary background radiation, measurable
radioactivity, minimum detectable concentration, measurement quality objectives.
Energy Resolution is the quantifiable ability of a measurement method to distinguish between
radiations with different energies.
Environmental Radioactivity is radioactivity from the environment where the M&E are
located. Environmental radioactivity includes background radiation as well as inherent
radioactivity and radioactivity from nearby sources.
Evaluation Function is a mathematical expression that allows the user to compare options and
draw a conclusion or calculate a result.
Expanded Uncertainty is the product, U, of the combined standard uncertainty of a measured
value, y, and a coverage factor, k, chosen so that the interval fromy-Utoy+Uhas a desired
high probability of containing the value of the measurand. See in this glossary combined
standard uncertainty, coverage factor, and measurand.
Fluence is the number of photons or particles passing through a cross-sectional area. The
international standard (SI) unit for fluence is nT2.
Frequency Plot is a chart plotting the number of data points against their measured values.
Graded Approach is the process of basing the level of application of managerial controls
applied to an item or work according to the intended use of the results and the degree of
confidence needed in the quality of the results. See in this glossary data quality objectives
process.
Gray Region is the range of radionuclide concentrations or quantities between the
discrimination limit and the action level, where the consequence of making a decision error is
relatively minor. See in this glossary action level, discrimination limit, lower bound of the gray
region, and upper bound of the gray region.
Impacted is a term applied to M&E that are not classified as non-impacted. M&E with a
reasonable potential to contain radionuclide concentration(s) or radioactivity above background
are considered impacted (10 CFR 50.2). See in this glossary background radiation and non-
impacted.
Inherent Radioactivity is radioactivity resulting from radionuclides that are an essential
constituent of the material being measured (e.g., 40K in fertilizer containing potassium).
Initial Assessment (IA) is an investigation to collect existing information describing materials
and equipment and is similar to the historical site assessment (HSA) described in MARSSIM.
Input Quantity is any of the quantities in a mathematical measurement model whose values are
measured and used to calculate the value of another quantity, called the output variable.
Instrument Efficiency is the ratio between the instrument net reading and the surface emission
rate of a source under given geometrical conditions (ISO 1988). For a given instrument, the
NUREG-1575, Supp. 1 GL-4 January 2009
-------
Glossary
instrument efficiency depends on the energy of the radiations emitted by the source. See in this
glossary source efficiency and total efficiency.
Interdiction is an increase in the level of radiological control or a decision not to accept control
from another party. Examples of interdiction include identification of radioactive material that
results in the initiation of radiological controls or identification of unauthorized movement of
radioactive material.
Interdiction Survey is the collection of data to support an interdiction decision regarding M&E.
In general, interdiction surveys are used to accept or refuse to accept control of M&E that are
potentially radioactive. The goal of an interdiction survey often is to detect radioactive M&E that
should be controlled. In some cases, an interdiction survey may result in the impoundment of
radioactive M&E that represent an unacceptable risk to human health or the environment.
Interference is the presence of other radiation or radioactivity, chemicals, background noise,
instrument noise, or other factors that hinders the ability to analyze for the radiation or
radioactivity of interest.
Intrinsic Radioactivity See in this glossary inherent radioactivity.
Lower Bound of the Gray Region (LBGR) is the radionuclide concentration or level of
radioactivity that corresponds with the lowest value in the range where the consequence of
decision errors is relatively minor. For Scenario A, the LBGR corresponds to the discrimination
limit. For Scenario B, the LBGR corresponds to the action level. See in this glossary action level,
discrimination limit, gray region, Scenario A, and Scenario B.
Mathematical Model is the general characterization of a process, object, or concept in terms of
mathematics, which enables the relatively simple manipulation of variables to be accomplished
in order to determine how the process, object, or concept would behave in different situations.
Materials and Equipment (M&E) are items considered for disposition that include metals,
concrete, dispersible bulk materials, tools, equipment, piping, conduit, furniture, solids, liquids,
and gases in containers, etc. M&E are considered non-real property distinguishable from
buildings and land, which are considered real property. See in this glossary disposition and non-
realproperty'.
Measurand is a particular quantity subject to measurement (ISO 1996).
Measurement Method Uncertainty See in this glossary method uncertainty.
Measurement Quality Objectives (MQOs) are a statement of a performance objective or
requirement for a particular method performance characteristic (MARLAP 2004).
Measurement Standard Deviation See in this glossary standard deviation (of measurement).
Measurement Uncertainty See in this glossary uncertainty (of measurement).
Measurable Radioactivity is radioactivity that can be quantified using known or predicted
relationships developed from historical information, process knowledge, or preliminary
measurements as long as the relationships are developed, verified, and validated as specified in
the data quality objectives (DQOs) and measurement quality objectives (MQOs).
January 2009 GL-5 NUREG-1575, Supp. 1
-------
Glossary MARSAME
Median is the middle value of the data set when the number of data points is odd, or the average
of the two middle values when the number of data points is even.
Method Uncertainty, UM, is the predicted uncertainty of the measured value that would be
calculated if the method were applied to a hypothetical sample with specified concentration.
Minimum Detectable Activity (MDA) is the minimum detectable value of activity for a
measurement. See in this glossary minimum detectable value.
Minimum Detectable Concentration (MDC) is the minimum detectable value of the
radionuclide or radioactivity concentration for a measurement. See in this glossary minimum
detectable value.
Minimum Detectable Value is an estimate of the smallest true value of the measurand that
ensures a specified high probability, 1 - /?, of detection. This definition presupposes that an
appropriate detection criterion has been specified (e.g., critical value). See in this glossary
measurand and critical value.
Minimum Quantifiable Concentration (MQC) is the smallest value of the concentration that
ensures the relative standard deviation of a measurement of M&E with that concentration does
not exceed a specified value, usually 10%.
Non-impacted is a term applied to M&E where there is no reasonable potential to contain
radionuclide concentration(s) or radioactivity above background (10 CFR 50.2). See in this
glossary background radiation and impacted.
Non-Real Property is property that is not real property. See in this glossary real property and
materials and equipment (M&E).
Null Hypothesis, or baseline condition, is a tentative assumption about the true, but unknown,
radionuclide concentration or level of radioactivity that can be retained or rejected based on the
available evidence. When hypothesis testing is applied to disposition decisions, the data are used
to select between a presumed baseline condition (the null hypothesis) and an alternate condition
(the alternative hypothesis). The null hypothesis is retained until evidence demonstrates with a
previously specified probability that the baseline condition is false.
Output Quantity is the quantity in a mathematical measurement model whose value is
calculated from the measured values of other quantities in the model. See in this glossary input
quantity.
Planning Team is the group of people who perform the DQO process. Members may include the
decision-maker (senior manager), site manager, representatives of other data users, senior
program and technical staff, someone with statistical expertise, and a quality assurance and
quality control advisor (such as a QA manager) (EPA 2000a).
Posting Plot is a map of the survey unit with the data values entered at the measurement
locations. This type of plot potentially reveals heterogeneities in the data, especially possible
patches of elevated radioactivity.
NUREG-1575, Supp. 1 GL-6 January 2009
-------
Glossary
Preliminary Survey is any survey performed prior to the disposition survey in MARSAME, and
is generally performed to provide information required to support the design of the disposition
survey. See also in this glossary disposition survey.
Process Knowledge is information concerning the characteristics, history of prior use, and
inherent radioactivity of the materials and equipment being considered for release. Process
knowledge is obtained through a review of the operations conducted in facilities or areas where
materials and equipment may have been located and the processes where the materials and
equipment were involved.
Radioactive Materials consist of any material, equipment or system component determined or
suspected to contain radionuclides in excess of inherent radioactivity. Radioactive material
includes activated material, sealed and unsealed sources, and substances that emit radiation. See
in this glossary inherent radioactivity.
Radiological Controls are any means, method or activity (including engineered or
administrative) designed to protect personnel or the environment from exposure to a radiological
risk.
Radionuclides or Radiations of Concern are radionuclides or radiations that are present at a
concentration or activity that may pose an unacceptable risk to human health or the environment.
In MARSAME, the term radionuclides or radiations of concern is used to describe the
radionuclides or radiations that are actually measured during the disposition survey. See also in
this glossary radionuclides or radiations of potential concern and disposition survey.
Radionuclides or Radiations of Potential Concern are radionuclides or radiations that are
identified during the initial assessment as potentially being associated with the M&E being
investigated. See also in this glossary initial assessment and radionuclides or radiations of
concern.
Ratemeter is an instrument that indicates the counting rate of an electronic counter. In the
context of radiological measurements, a ratemeter displays the counting rate from a radiation
detector. The averaging time for calculating the rate is determined by the time constant of the
meter. See in this glossary sealer.
Real Property, in the MARSAME context, means developed or undeveloped land, fixed
buildings and structures, or surface and subsurface soil remaining in place. Real property is
outside the scope of MARSAME. See in this glossary materials and equipment (M&E) and non-
realproperty'.
Recycle is beneficial reuse of constituent materials incorporated within the M&E. A hammer that
is melted down as scrap metal so the component metals can be reused is an example of recycle.
Reference Material is material of similar physical, radiological, and chemical characteristics as
the M&E considered for disposition. Reference material provides information on the level of
radioactivity that would be present if the M&E being investigated had not been radiologically
impacted. See in this glossary impacted.
January 2009 GL-7 NUREG-1575, Supp. 1
-------
Glossary MARSAME
Relative Standard Uncertainty is the ratio of the standard uncertainty of a measured result to
the result itself. The relative standard uncertainty of x may be denoted by ur(x). See in this
glossary standard uncertainty.
Release is a reduction in the level of radiological control, or a transfer of control to another
party. Release includes clearance. Examples of release (other than clearance) include recycle,
reuse, disposal as waste, or transfer of control of radioactive M&E from one authorized user to
another. See also in this glossary reuse, recycle, restricted release, and clearance.
Release Survey is a type of disposition survey designed to collect information to support a
release decision. See also in this glossary disposition survey and release.
Restricted Release is a reduction in the level of radiological control, or transfer of control to
another party, where restrictions are placed on how the released items will be used or transferred.
Maintaining a tool crib in a radiologically controlled area restricts reuse of those tools to that
radiologically controlled area, and tools returned to the tool crib represent a restricted release of
those tools. See also in this glossary reuse, recycle, release, and clearance.
Reuse is the continued use of M&E for their original purpose(s). An example of reuse is a
hammer that continues to be used as a hammer.
Ruggedness is the relative stability of a measurement technique's performance when small
variations in method parameter values are made.
Sampling Standard Deviation, as,, is the theoretical true value of the variability of radionuclide
concentration or radioactivity in space and time (i.e., the variation of the true but unknown
concentrations from place to place and from time to time). The extent of the survey unit, the
physical sizes of the measured material, and the choice of measurement locations affects the
sampling standard deviation.
Sealer is an electronic counter that displays the aggregate of a number of signals, which usually
occur too rapidly to be recorded individually. In the context of radiological measurements, a
sealer records the number of counts from a radiation detector over a specified time interval. See
in this glossary ratemeter.
Scenario A uses a null hypothesis that assumes the level of radioactivity associated with the
M&E exceeds the action level. Scenario A is sometimes referred to as "presumed not to comply"
or "presumed not clean."
Scenario B uses a null hypothesis that assumes the level of radioactivity associated with the
M&E is less than or equal to the action level. Scenario B is sometimes referred to as
"indistinguishable from background" or "presumed clean."
Secular Equilibrium is the condition in which the initial member of the decay series has a
longer half-life than any subsequent members of the series. Secular equilibrium is achieved when
the activities for all members of the decay series are equal to the activity of the precursor
radionuclide.
NUREG-1575, Supp. 1 GL-8 January 2009
-------
Glossary
Segregation is the process of separating or isolating from a main body or group. In the context
of disposition surveys, segregation is based on the physical and radiological attributes of the
M&E being investigated and is used to help control measurement method uncertainty.
Sensitivity Coefficient for an input estimate, X{, used to calculate an output estimate,
y =f(x\, X2, ..., Jtjv), is the value of the partial derivative, df/dx[, evaluated at /' = x\, x2,...,xN. The
sensitivity coefficient represents the ratio of the change in^ to a small change in X[.
Sentinel Measurement is a biased measurement performed at a key location to provide
information specific to the objectives of the initial assessment (IA).
Significance Level is, in the context of a hypothesis test, a specified upper limit for the
probability of a Type I decision error.
Sign Test is a non-parametric statistical test used to evaluate radionuclide-specific disposition
survey results if the radionuclide being measured is not present in background, or is present at
such a small fraction of the action level as to be considered insignificant.
Smear is a non-quantitative test for the presence of removable radioactive materials in which the
suspected surface or area is wiped with a filter paper or other substance, which is then tested for
the presence of radioactivity. The surface area tested may be related to the release criterion.
Smear is also referred to as a smear test, swipe, or wipe.
Source Efficiency is the ratio between the number of particles of a given type above a given
energy emerging from the front face of a source or its window per unit time and the number of
particles of the same type created or released within the source (for a thin source) or its
saturation layer thickness (for a thick source) per unit time (ISO 1988). See also in this glossary
instrument efficiency and total efficiency.
Specific Activity is the radioactivity per unit mass for a specified radionuclide.
Specificity is the ability of the measurement method to measure the radionuclide of concern in
the presence of interferences.
Spectrometry is a measurement across a range of energies. The measurement of alpha particles
by energy is called alpha spectrometry.
Spectroscopy is the measurement and analysis of electromagnetic spectra produced as the result
of the emission or absorption of energy by various substances. The measurement of gamma-ray
emissions from a substance is called gamma spectroscopy.
Standard Operating Procedure (SOP) is a written document that details the method for an
operation, analysis, or action with thoroughly prescribed techniques and steps, and that is
officially approved as the method for performing certain routine or repetitive tasks (MARSSIM
2002).
Standard Deviation (of Measurement),
-------
Glossary MARSAME
Standard Uncertainty is the uncertainty of a measured value expressed as an estimated standard
deviation, often called a "1-sigma" (Icr) uncertainty (MARLAP 2004). The standard uncertainty
of a value x is denoted by u(x). See also in this glossary uncertainty (of measurement).
Standardized Initial Assessment is a set of instructions or questions that are used to perform
the initial assessment, usually documented in a standard operating procedure. See also in this
glossary initial assessment and standard operating procedure.
Surficial Radioactive Material is radioactive material distributed on any of the surfaces of a
solid object. Surficial radioactive material may be either removable by non-destructive means
(such as casual contact, wiping, brushing, or washing) or fixed to the surface.
Surrogate Measurement is a measurement where one radionuclide is quantified and used to
demonstrate compliance with the release criterion for additional radionuclide(s) based on known
or accepted relationships between the measured radionuclide and unmeasured radionuclide(s).
Survey Unit for M&E is the specific lot, amount, or piece of M&E on which measurements are
made to support a disposition decision concerning the same specific lot, amount, or piece of
M&E.
Total Efficiency is the product of the instrument efficiency and the source efficiency. See in this
glossary instrument efficiency and source efficiency.
Traceability is the "property of the result of a measurement or the value of a standard whereby it
can be related to stated references, usually national or international standards, through an
unbroken chain of comparisons all having stated uncertainties" (ISO 1996).
Type I Decision Error occurs when the null hypothesis is rejected when it is actually true. The
Type I decision error rate, or significance level, is represented by a. See in this glossary null
hypothesis and significance level.
Type II Decision Error occurs when the null hypothesis is not rejected when it is actually false.
The Type II decision error rate is denoted by /?. See in this glossary null hypothesis.
Uncertainty (of Measurement), w(x), is a parameter, associated with the result of a
measurement, x, that characterizes the dispersion of the values that could reasonably be
attributed to the measurement of x. It is the estimated value of o(x) obtained from the
propagation of uncertainty. See also in this glossary See also in this glossary standard deviation
(of measurement).
Unrestricted Release is the removal of radiological regulatory controls from materials and
equipment. See in this glossary release and clearance.
Upper Bound of the Gray Region (UBGR) is the radionuclide concentration or level of
radioactivity that corresponds with the highest value in the range where the consequence of
decision errors is relatively minor. For Scenario A, the UBGR corresponds to the action level.
For Scenario B, the UBGR corresponds to the discrimination limit. See in this glossary action
level., discrimination limit., gray region., Scenario A, and Scenario B.
NUREG-1575, Supp. 1 GL-10 January 2009
-------
Glossary
Volumetric Radioactive Material is radioactive material that is distributed throughout or within
the materials or equipment being measured, as opposed to a surficial distribution. Volumetric
radioactive material may be homogeneously (e.g., uniformly activated metal) or heterogeneously
(e.g., activated reinforced concrete) distributed throughout the M&E.
Wilcoxon Rank Sum (WRS) Test is a non-parametric statistical test used to evaluate
disposition survey results when the radionuclide being measured is present in background by
comparing the results to measurements performed using an appropriately chosen reference
material.
January 2009 GL-11 NUREG-1575, Supp. 1
------- |