Battelle International Conference on the Remediation and Management of Contaminated Sediments

Platform Presentations:

A Reliable Dataset from a PFAS Sediment Investigation near a Former (Confidential) Manufacturing Site

Meredith Hayes, Rick Beach, Katherine McDonald, Leslie Nelson,
and Mark Westra

Use of Probabilistic Estimating Techniques to Quantify Long-Term Sediment Cap Monitoring and Maintenance Costs

Meredith Hayes, Victoria Ward, Albert Ricciardelli, and Curtis Toll (Greenberg Traurig)

Posters:

Which Way is the Finger Pointing? Lessons Learned from Allocation Assessments

Rick Beach and Meredith Hayes

Management of DNAPL-Impacted Sediments: “In the Eye of the Beholder?”

Rick Beach, Meredith Hayes, Dan Amate, Joseph Foglio, Michael Shaw, and John Oberer

PFAS Soil and Sediment Remediation

Mike Welch (Sevenson Environmental Services) and Joseph Foglio

Addressing Sediment Contamination under the LSRP Program in New Jersey: Challenges and Potential Improvements

Meredith Hayes, Sandra Huber, David Winslow, Rick Beach, and Tim Briggs

Session Chair:

NAPL and MGP Sites: Rick Beach and Gary Rose (Sevenson Environmental Services)

A Reliable Dataset from a PFAS Sediment Investigation
Near a Former (Confidential) Manufacturing Site in Michigan

Meredith Hayes (GZA GeoEnvironmental, Fairfield, NJ, USA)
Rick Beach (GZA GeoEnvironmental, Philadelphia, PA, USA) 
Katherine McDonald (GZA GeoEnvironmental, South Portland, ME, USA)
Leslie Nelson and Mark Westra (GZA GeoEnvironmental, Grand Rapids, MI, USA)

Background/Objective. Per- and polyfluoroalkyl substances (PFAS) in the environment have become the subject of substantial interest over the past decade. Concerns about the environmental and health impacts of this group of compounds have emerged due to their varying potential toxicities and persistence in the environment. There have been many challenges preventing investigations from developing a good understanding of PFAS distributions in the sediment environment. These issues include the ubiquitous PFAS presence and cross-contamination of samples during sampling, suitable analytical methods, appropriate field and laboratory QA/QC, and the normal challenges of sediment investigations in complex and heterogeneous substrates.

The majority of recent presentations, webinars, and training programs on PFAS in sediments that do not involve toxicity studies have focused on analytical methodology, hypothetical partitioning considerations, and theoretical discussions of where PFAS should reside in sediments instead of providing comprehensive site evaluations. Occasionally, smaller high-quality datasets associated with academic or limited project investigations have been presented. This presentation, however, will describe the reliable and substantial dataset of sediment PFAS that was developed as part of a broader investigation at a legacy PFAS site for a confidential client.  

Approach/Activities. Standard operating procedures (SOPs) were developed based on our experience from PFAS groundwater and soil investigations and modified based on our experience with sediment investigations.  The investigation team used the SOPs with best practices from both of the upland and aquatic investigation protocols including the integration of a “clean-hands” lead sampler.  As a result of the careful approaches, none of the PFAS results were rejected for field or laboratory QA/QC concerns. The rinsate from only one equipment blank, a saw blade used to cut long sediment cores, had detectable PFAS. However, the saw blade never contacted sediments in sample intervals collected for laboratory analyses.

Over 99 high-quality sediment samples were collected during the investigation from a river and a tributary creek. The samples were primarily generated from cores collected by vibracoring methods along 10 transects, with 3 locations per transect, and up to 4 sample intervals per core. Nine of the cores reached a sediment depth of at least 6 feet and one core was stopped at a depth of 7.5 feet to avoid over-filling the 8-foot core tube. The sediment cores were generally retrieved intact and consisted of primarily sandy silt with larger amounts of gravel and organic matter at various locations and sediment depths.  Based on visual observations some locations exhibited substantial methanogenesis and had large methane gaps (2-4 inches thick) in cores.

Results/Lessons Learned. In addition to potential contaminants of concern (COCs) evaluated as part of the broader site investigation, twenty-three PFAS compounds were analyzed in each sediment sample with detection limits as low as 0.43 ug/kg and a maximum concentration of 170 ug/kg perfluorooctane sulfonic acid (PFOS) using the U.S. DOD QSM isotope dilution method.  The spatial and vertical distributions of total PFAS and select PFAS compounds in sediments will be presented that illustrate the narrow distribution of PFAS near the shoreline. 
 

Use of Probabilistic Estimating Techniques to Quantify Long-term Sediment Cap Monitoring and Maintenance Costs

Meredith Hayes (GZA GeoEnvironmental, Inc., Fairfield, NJ, USA)

Victoria Ward and Al Ricciardelli, P.E., LSP (GZA GeoEnvironmental, Inc., Norwood, MA, USA)

Curtis B. Toll, Esq. (Greenberg Traurig, LLP, Philadelphia, PA, USA)

Background/Objectives. Many impacted sediment sites are reaching the state that many upland sites had attained a decade or more ago. Remedial construction is complete and we are now faced with long-term monitoring and maintenance of the residual impacts left behind. Understanding the uncertainties associated with those long-term obligations can present significant challenges – particularly when it comes to reaching agreements with respect to the allocation of the obligations. Risk modeling analysis can be utilized to quantify the unknown costs of conducting long-term work, which can then be translated into a fixed-price-to-closure contracting arrangement that helps parties craft a comprehensive settlement. Probabilistic cost estimating techniques have been applied to upland remedial issues for decades. However, sediment sites come with their own set of challenges – not the least of which is climate change. This presentation will review a case study of a Superfund site to illustrate how these techniques were applied to thirty years of sediment cap monitoring and maintenance costs. 

Approach/Activities. Monte Carlo techniques were used to assess the uncertainty in costs associated with the long-term monitoring and maintenance of a sediment cap. The model considered factors affecting routine operations and the potential for episodic events. The cap design was evaluated under a range of climatological models and river conditions to assess the potential for cap “failure” and the potential extent of that failure. Routine tasks such as monitoring were also included in the model utilizing current conditions. The entire technical model was then put into a financial model that considered factors such as the timing of disbursements, the effect of inflation, and rates of return (utilizing a random walk model). We will describe these factors as well as others we considered in our evaluation, how we developed values for those factors, and how we utilized the results to arrive at a probability distribution for the financial aspects of the remedial program. 

Results/Lessons Learned. The results of our analysis provided the client with an understanding of the possible financial uncertainties as well as the cash flow impacts for the obligations. This information was used to structure an agreement and funding mechanism for implementing the work that fit with their overall business strategy and risk tolerance, and at an overall cost substantially less than the conservative initial feasibility study estimate. Ultimately, when the model was translated into a fixed-price to closure contractual arrangement, the underlying cost assumptions and guaranteed maximum price were adopted by regulators in constructing the appropriate financial assurance for the site.
 

Which Way is the Finger Pointing? 
Lessons Learned from Allocation Assessments

Rick Beach (GZA GeoEnvironmental, Philadelphia, PA, USA) 
Meredith Hayes (GZA GeoEnvironmental, Fairfield, NJ, USA)

Background/Objective. Potentially responsible parties (PRPs) are frequently “invited” by regulatory agencies to help address sediment contamination at large multi-party sites. On Superfund sites, these “Notice of Liability” letters are provided by the U.S. Environmental Protection Agency (EPA). However, it is common for the liabilities described in the letters to be heavily based on information provided to EPA by the largest PRP identified for the site. Unfortunately, these liabilities are often rooted in limited (or selective) technical information and perhaps a desire to “share the pain.” The result is that the exposure to Superfund joint and several liability coupled with the risk of being labeled an “uncooperative party” for equitable allocation purposes may drive PRPs with smaller potential liabilities into expensive allocation assessments and litigation activities lasting years, or even decades, to establish minor or no contributions of contaminants to a waterway  These smaller PRPs must first discount the initial information presented about a site and also address misleading, strongly biased, and non-transparent technical reports developed by other PRPs as part of the allocation process. 
 
Approach/Activities.  Allocation assessments usually begin with the identification of contaminants of concern (COCs) associated with a site and the evaluation of the potential pathway(s) to the waterway that could create a nexus. The first task requires a thorough evaluation of past site operations. This abstract focuses solely on the second task. This typically includes an examination of historical aerial photographs. This presentation will illustrate the importance of correlating the aerial information, site conditions, elevations, and knowledge of the former operations to avoid erroneous conclusions and misleading speculation. 

The evaluation of wastewater and solid waste releases to waterways based on the available knowledge of legacy site operations is frequently the second step in an assessment.  This evaluation may yield an estimate of the largest potential contaminant mass released to the waterway. However, undefined stormwater releases are the easiest target to connect very small amounts of hypothetical contaminants from a site with a waterway.  It is not uncommon for the connections to be based on speculation and limited facts. Potential stormwater releases of contaminants can be estimated using several approaches with varying degrees of accuracy, but the results often agree within one order of magnitude when using reasonable average conditions for a site. While an order of magnitude in mass contribution may seem highly uncertain in allocations, the relative stormwater contributions are usually several orders of magnitude lower than the contaminant mass contributed by other inputs to the waterway.

Results/Lessons Learned. The use of liability letters by the EPA and other regulatory agencies based on desktop evaluations and limited site information may be a necessary step to assemble a group of PRPs.  However, a PRP with minimal or no potential liability can best respond with an allocation assessment to address the obvious technical deficiencies of liability descriptions using simple facts and multiple lines of defense when possible and minimal speculation.  Consultants should resist the temptation to select only supportive data and to use more complicated statistics than needed to skew the implications and findings.  The inclusion of simple bounding analyses and the presentation of hypothetical site releases relative to the reservoir of sediment contaminants in the assessment findings should be used when possible. 
 

Management of DNAPL Impacted Sediments – 
“In the Eye of the Beholder?”

Rick Beach (GZA GeoEnvironmental, Philadelphia, PA, USA) 
Meredith Hayes (GZA GeoEnvironmental, Fairfield, NJ, USA)
Dan Amate, Joe Foglio, Michael Shaw, and John Oberer (GZA GeoEnvironmental, Philadelphia, PA, USA)

Background/Objective. The evaluation and remediation of contaminated sediments is a mature practice that incorporates advanced methods for investigations, feasibility studies, and remediation. The management of upland dense non-aqueous phase liquid (DNAPL) impacted soils is also a mature practice. However, many challenges remain when addressing manufactured gas plant (MGP) and analogous DNAPL contamination in sediment. Applying upland approaches to the aquatic environment can be misleading at best or just plain wrong.

Remediation of DNAPL impacted sediments is currently driven at many sites by a strong “remove or stabilize” mentality that originates from the relatively straight-forward approaches used to address upland DNAPL impacted soils.  However, the primary risk at the majority of DNAPL sediment sites (with no/limited ebullition) results from dissolved compounds originating in the DNAPL that migrate via the pore water pathway and are bioavailable to the benthic community in the surficial bioactive zone (BAZ). However, the most common investigative and remedial approaches are often only focused on the delineation of the DNAPL in sediments for removal or in-situ stabilization. In addition to the potential biases introduced by upland practitioners, there are additional potential biases from risk assessors, regulators, contractors, field samplers, and even sediment experts. These different “beholders” may base their approaches solely on their personal experience that can ultimately miss-direct comprehensive site evaluations and generate poor management decisions. 

Approach/Activities. As a result of personal biases, the following mistakes have occurred on some sites:

• No evaluation of bioavailable polycyclic aromatic hydrocarbons (PAHs) when appropriate.
• Limiting bioavailable assessments to solid-phase partitioning instead of pore water evaluations, resulting in remedial footprints that have been 189% larger than needed.
• Compositing the 4-6-inch BAZ into 1 or 2-foot sampling intervals, with the resulting overestimation of PAH concentrations 4-13-fold at some sites. 
• Improper use or interpretation of sediment dating.
• Inadequate or no evaluation of DNAPL weathering, which can greatly impact DNAPL migration and the partitioning of PAHs into pore water.
• Lack of PAH and supportive data over the complete sediment profile, without which technical impracticability and remedial cost-effectiveness cannot be determined on many sites.  Not (adequately) evaluating vertical distributions of contaminants or characteristics in sediments is intended to save money but it often has the opposite effect.

Results/Lessons Learned. The presentation will include a reminder of why sediments should not be treated as wet soils in terms of investigative and remedial approaches brought to the table by upland practitioners.  Numerous project examples will be provided to demonstrate the potential dangers of single-focused “Beholders” directing DNAPL projects.

 

Addressing Sediment Contamination Under the LSRP Program in New Jersey — Challenges and Potential Improvements

Meredith Hayes, Sandra Huber, LSRP, and David Winslow (GZA GeoEnvironmental, Inc., Fairfield, NJ, USA)
Richard Beach (GZA GeoEnvironmental, Inc., Philadelphia, PA, USA)
Timothy L. Briggs (GZA GeoEnvironmental, Inc., Norwood, MA, USA)

Background/Objectives. In 2009 New Jersey enacted the Site Remediation Reform Act which established the Licensed Site Remediation Professional (LSRP) program. The LSRP program has been successful at accelerating the pace of cleanup at many sites in New Jersey. However, addressing sediment contamination within the established timeframes remains a challenge. The technical requirements for remediating contaminated sites are established under New Jersey administrative code and by a series of guidance documents. The framework of the LSRP program can lead to delays or improper decision making in addressing sediment areas of concern, counter to the objective of the program. This presentation will review work completed at a former chemical (resin) manufacturing plant in New Jersey which points to areas where the LSRP program (and associated technical guidance) may benefit from improvements, as well as possible practical solutions.

Approach/Activities. Prior to GZA’s involvement at the site, evaluations of potential migration of contaminants to sediment via former drainage lines were completed over the course of eight years (2005-2013) leading to the issuance of a remedial action outcome (RAO) by the LSRP. The RAO effectively provided for no further action without conducting sampling. However, upon later regulatory review, the RAO was rescinded, and a remedial investigation was required. The former site owner (potential responsible party) declared bankruptcy resulting in the current site owner (GZA’s client) assuming the obligation to perform the additional investigation. GZA completed sampling (surface water, sediment, porewater, and sediment toxicity) to characterize the nature and extent of contamination and reviewed historic information to develop an understanding of potential sources and migration pathways and inform the conceptual site model. The site is located in a highly urbanized area resulting in multiple potential sources and complicating the distinction between site-related contamination and urban background or other sources. An evaluation of regional concentrations of contaminants was also completed. Ultimately, the evaluation showed little to no impact from the site to the sediments, thus coming full circle to the original conclusion years later, but with a proper assessment. 

Results/Lessons Learned. This project provides several examples of the challenges of addressing sediment sites under the LSRP program. 1) Recognizing the need to initiate a sediment investigation can be delayed under the current process for identifying areas of concern. 2) Delineation and remediation to established criteria is often impractical, especially in urban industrial settings with elevated background concentrations. This effectively results in the traditional (pre-LSRP program) iterative loop of comment and response between the LSRP and regulator to reach consensus, which can delay case closure and site redevelopment. 3) Regulatory guidance focuses on ecological risk and can overlook other important processes regarding fate and transport, hydrodynamics, and sediment transport. To address these challenges, the LSRP must proactively work to build regulatory consensus during the entire remedial process and may need to evoke other expertise to develop a robust conceptual site model.