Better Site Characterization through Incremental Sampling Methodology Mark Bruce Ph. D. 2014, TestAmerica Laboratories, Inc. All rights reserved. TestAmerica & Design are trademarks of TestAmerica Laboratories, Inc. October 29, 2014
Chasing Uncertainty Sources Instrumental analysis Sample preparation Laboratory sub-sampling Field sample collection 2
Does the decision unit fit in the sample jar? Representative subsampling 3
Why is this important? _ X with known and less uncertainty Representative subsampling 4
Cost Implications of ISM Lower remediation cost Lower envir. liability cost Lower measurement cost 5
Incremental Sampling Systematic Random Design Random starting locations in first grid Increment collection point 6
ITRC Guidance Incremental Sampling Methodology Team ~ www.itrcweb.org/ism-1/ ~ Formed Jan. 2009 TestAmerica recruited by the US Army Corps of Engineers 7
Paint the Big Picture Plan Implement Assess 8
Focus on Planning Plan Implement Assess Develop Conceptual Site Model Identify Data Quality Objectives Consider end use of data Base decision on mean or UCL? Develop Decision Units 9 Determine increments & replicates
Implement the Plan Plan Implement Assess Collect field samples Process samples at laboratory 10
Assess the results Plan Implement Assess Are the data sufficient for decision making? Manage risk / remediate as needed 11
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 12 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
Introduction ISM for Environmental Sampling Limitations of Traditional Approaches How ISM addresses these Limitations Comparing ISM to Compositing Frequently Asked Questions Document Organization Key Terms 13
ISM Principles Introduction Soil Heterogeneity Foundation Concepts of Sampling Sampling Considerations Gy Theory and the Source of Sampling Error Sampling Approaches ~ (Discrete, Composite, Incremental) 14
Systematic Planning Introduction Intent of Decision Units Decision Units (DU) Establishing new DUs Hot spots 15
Statistical ISM Design Estimating the Mean Concentration Uncertainty in Estimates of the Mean Evaluate Sampling Performance Areas for Further Study 16
Field Implementation Introduction Sampling Tools Field Collection Field Processing Options 17
Laboratory Processing & Analysis Introduction Laboratory Processing Laboratory Analysis Quality Assurance 18
Making Decisions Introduction Decision Mechanisms Error Assessment 19
Regulatory Concerns Introduction Perception Issues Regulatory Challenges State of Knowledge, Experience and Training Implementation Issues 20
Case Studies PCB Contaminated Landfill Petroleum Contaminated Soil Stockpile Former Golf Course Field Demonstration Hawaiian Homelands Development 21
Stakeholder Input Citizen, tribal, community, environmental advocate Find all contamination, facilitate cleanup Desire to understand characterization & cleanup plans Avoid averaging away hot spots Explain ISM approach early in the process 22
ISM Limitations Limitation Smaller number of replicates Effect Limits calculation methods for Upper Confidence Limit No spatial resolution within Decision Unit Limits remediation options within Decision Unit Limits multivariate comparisons 23
ISM Advantages Advantage Better spatial coverage Higher Sample Mass Optimized processing Fewer non-detects Effect Includes high & low concentrations in proper proportions Reduces errors associated with sample processing and analysis Representative subsamples for analysis Simplifies statistical analysis More consistent data More confident decision 24
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 25 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
ISM Principles Nature of Soil & Interaction of Contaminants With Soil Results In 26
ISM Principles Heterogeneity Sampling w/o addressing leads to 27
ISM Principles Sampling Errors Manifested (observed) in 28
ISM Principles Unknown Data Variability which can lead to 29
ISM Principles Decision Errors 30
Representativeness Which is more likely to represent the true mean? Average from 3 discrete samples One 30-increment sample 50 ft 50 ft 50 ft 50 ft 31
Representative Sample A representative sample is one that contains a subset of all the constituents of a population in exactly the same proportion that they are present in the target population. Represents 32
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 33 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
The Decision Unit: A Key Concept for ISM Decision Unit (DU) smallest volume of soil (or other media) for which a decision will be made based upon ISM sampling. A DU may consist of one or more Sampling Units (SUs). 34
Source Area DUs Above Ground Storage Tank Transformer Pad 35
Side Exposure Areas DUs Residential Yard Backyard DU-1 House Side 36 DU-2 Front yard DU-3
Combined DUs Future residential lots, DUs sized as exposure areas (EUs) Pesticide mixing area, DUs sized to assist remediation 37
Subsurface DUs Borings into DUs for increment collection Stacked spill area DUs 0 m -2 m -4 m -6 m DU-1 DU-2 DU-3 DU-4-8 m 38 Separate DU increments
Confirmation Sampling DUs Walls and Floor of an Excavation Site DU-2 DU-2 DU-3 DU-1 39
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 40 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
How Many Increments? Depends on heterogeneity ~ Multiple distributions ~ Large nugget effect 25-30 increments usually enough 3 replicates recommended 41
Single ISM result ~ Under estimate > 50% of the time Calculate the Upper Confidence Limit ~ Based on Student s t distribution UCL X Uncertainty in Estimates of the Mean t ( 1 )( r 1) S X r ~ Based on Chebyshev inequality UCL X 1/ 1 S X r 42
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 43 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
44 Sampling Tools
45 Surface Samples
Collection Pattern Example: 60 increment triplicate 46
47 Subsurface Samples
48 Subsurface Sample: Whole Core
Sample Mass M s = n D s ( / 2) 2 M s mass of sample (g) - soil or sediment density (g/cm 3 ) n number of increments see D s increment length (cm) - diameter of sample core (cm) 49
50 Volatile Organic Compound Sampling
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 51 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
52 No Universal Lab Sample Processing
Laboratory Analyses Choose the Best ISM Processes ISM Field Sampling Yes VOCs Analysis Needed? Laboratory Processing Separate Sample Container for % Moisture Determination Determine the Number & Mass of Soil Increments Collected in Methanol No Yes VOCs Analysis Needed? No % Moisture Determination Determine the Number & Mass of Soil Increments Collected % Moisture Determination % Moisture Reporting Weighing - Methanol vol/weight Determination Volatiles Analysis Bulk Sample Mass Reduction (Optional ; Not Recommended) Yes Processing in Field? Other on-site Processing Air Drying - Sieving - Subsampling On-Site No % Dry Weight Reporting Specify Air Drying on-site to facillitate sieving Sieving On-Site: Specify Mesh Size of Screen Subsampling On- Site: Determine the Number & Mass of Soil Increments Sample Containerized, Packaged and Process "As Received" Sample Conditioning ( Matrix and Analyte Specific ) Disaggregation Drying "Chopping" (May require drying) Air Oven Freeze Drying Drying Drying Water Addition GO TO BOX "A" A ON NEXT PAGE Sample Mixing (Optional) Certification Preparatory Methods Analytical Method Modifications Particle Size Selection and/or Reduction Methods/ Contamiants Method Considerations Limitations Sieving Dish and Puck Mill Mortar and Pestle Ball Mill Pulverizers Other Fundamental Error and Sample Mass Consideration/Requirements Analytical Subsampling Techniques Inorganics Organics Sectorial Splitters Riffle Splitters Rotary Splitters 1 or 2-Dimensional Japanese Slabcake Other Quality Assurance/Quality Control 53 OAF Sample Ready for Analytical Preparation/Extraction Blanks Laboratory Replicates Spikes (Blank or Matrix?) Reference
Include Lab Processing in Project Planning Lab 54
Choosing ISM processing options without understanding the project objectives is like writing a love letter and addressing it to whom it may concern 55
Two Fundamental Decisions What are you looking for? ~ Complete list of individual analytes Where are you looking for it? ~ What particles in the sample container should be included in the sample? ~ Exposed surfaces and pores of the particles? ~ Interior of the particles? 56
Two Trade-Off Decisions To air dry or not? Mechanical processing Analyte retention To mill or not? Improved precision Analyte representativeness 57
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 58 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
Making Decisions Introduction Decision Mechanisms Error Assessment 59
Decision Mechanisms Structured approach to making decisions avoid confusion Six Common Types of Comparisons 60
Compare One ISM Result to Action Level Decision Unit Single Result Action Level 61
Compare Average ISM Result to Action Level Decision Unit Mean of Replicates Action Level 62
Compare 95% Upper Confidence Limit to Action Level Decision Unit 95% UCL Action Level 63
Technical Summary Study allowed comparison of a number of factors that affected reproducibility of results. For analytes presented: Factor A Factor B Comments 64 Field collection techniques Laboratory subsampling and milling 8330B milling technique 8330B analytical equipment Discrete, box, and wheel Avg RSD = 115% 8330A Avg RSD = 22% ISM Avg RSD = 22% 8330B Avg RSD = 6% ~5x better precision with ISM ~3x better precision with ISM Puck mill Ball mill Difference inconclusive HPLC/UV HPLC/MS/MS Slightly better precision with HPLC/UV
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 65 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
66 Previous Anti-compositing regulations
67 States with Some Regulator Acceptance of ISM
EPA Regional Support 68
Federal Guidance Method 8330B Appendix A October 2006 Method 3050 Sw-846 Update VI? Uniform Federal Policy Quality Assurance Project Plan Template For Soils Assessment of Dioxin Sites September 2011 69 Superfund Lead- Contaminated Residential Sites Handbook Late 2012
ISM Guidance Introduction Laboratory Process Case Studies ISM Principles Making Decisions Delineation Decisions 70 Systematic Planning Statistical ISM Design Field Implementation Regulatory Concerns Case Studies Stakeholder Input
Michigan ISM contact : ~ John Bradley ~ Michigan Department of Environmental Quality ~ BRADLEYJ1@michigan.gov ~ (517) 284-5069 Analytes ~ VOCs, SVOCs, PAHs, Metals, Hg Sites ~ Parks, Superfund, manufacturing facility 71
Ohio ISM contact : ~ Martin Smith ~ Ohio Environmental Protection Agency ~ Martin.Smith@epa.state.oh.us ~ (614) 644-4829 Analytes ~ SVOCs, PAHs, Metals, Hg Sites ~ Ordinance manufacturing, urban garden, chromium process stockpiles 72
The Cost Savings of ISM Reduced overall analytical costs significantly Maintain analytical spend and structure sampling design to reduce uncertainty Large potential ripple effects for: ~ Liability insurance coverage ~ Remediation costs when site concentration near decision threshold 73
Better Precision Changes Decisions Florida Golf Course - Arsenic Data (mg/kg) Discrete n = 10 (mg/kg) Incr-30 n = 3 (mg/kg) Incr-100 n = 3 (mg/kg) Mean 2 1.8 1.7 Std Dev 1.4 0.08 0.03 95%UCL 3.0 2.0 1.8 FDEP SCTL: 2.1 mg/kg ITRC ISM team Florida golf course case study
ITRC Guidance & Training ITRC ISM guidance publically available ~ Feb 15, 2012 ~ www.itrcweb.org/ism-1 Internet based training ~ Live Nov. 4 & 6 2014 ~ Archived versions from 2012, 2013, 2014 75 ~ www.clu-in.org/conf/itrc/ism/
Purpose of ISM Representative samples Better data Better decisions 76
Contact Information Patricia McIsaac Industry Marketing Manager Patricia.McIsaac@TestAmericaInc.com 703.758.8381 Mark Bruce Ph. D. Corporate Technical Director Mark.Bruce@TestAmericaInc.com 330.966.7267 77