Site characterization is a process of reducing uncertainty, with the eventual aim of developing an accurate conceptual site model (CSM) that is appropriate for the remedial objectives of the site. The economic cost of a CSM is highly variable and dependent on many factors, and little in the way of formal guidance exists on balancing the production of data from intensive and costly site characterization activities, the associated cost of the activities, and the ultimate benefit to achieving remedial goals more cost effectively. A similar lack of guidance exists for understanding which elements of site characterization translate to the highest value data for remediation planning, as well as in understanding which performance metrics are the most critical in assessing operational performance and in predicting long-term effectiveness of the applied remedy. 

The overall research objective of the DIVER project was to develop a framework and associated methodologies for the optimization of the site characterization process, such that the life cycle cost of any remedy is minimized while remaining protective of human health and the environment.

Technical Approach

The project developed detailed, large-scale data sets through simulation, which acted as “sites” under investigation and remediation. These “virtual” sites provided the basis for assessment of “Experienced Based” and “Decision Theoretic (DT)” approaches to remediation design. The “sites” were investigated (virtually) by several decision maker (DM) teams, comprised of some of the most experienced and senior practitioners in the industry. The DM teams developed site investigation plans, followed by investigation of the large-scale data sets to produce site investigation data. The DM teams then developed conceptual site models for each site with a focus on selected quantitative parameters and developed remediation plans based on their CSM to achieve pre-defined performance objectives. The CSM and remediation performance were assessed against “true” values (known from the simulations) and optimal remediation designs based on this “perfect data”. Effective practice for selected CSM parameters was then developed with stochastic modeling used to determine the value of additional site characterization information in reducing uncertainty in the accuracy of the parameters. In the DT approach, the CSM parameters developed by the DM teams were used in a stochastic framework, where the most likely remedial design was identified through the minimization of cost functions given probabilistic distributions of the input data. The final stage of the project tested the validity of the effective practice on another virtual site based on an existing Department of Defense (DoD) site where remediation performance has been found to be sub-optimal.


The DIVER Project (Data Information Value to Evaluate Remediation) has developed technical guidance on the value of data in both the site characterization and remediation contexts based on detailed site data, empirical evidence gathered from some of the most respected and successful practitioners in the field, highly detailed virtual site investigations, and stochastic approaches to quantifying both the value of additional information and optimal remediation designs.


This project will have the immediate benefit of improving remediation outcomes across the DoD, by identifying the key parameters which, through reduction of uncertainty, can lead to more cost-effective and efficient remediation of dense non-aqueous phase liquid-impacted sites. The project has also developed highly beneficial effective practice based on scientific principles that determine the value of data and the value of current site investigation practices, which in turn can be used by the DoD to assess the cost effectiveness of proposals and work plans submitted by contractors for applicable impacted groundwater sites. (Project Completion - 2023)


Mumford, K., S. Bryck, B.H. Kueper, S. Mancini, M. Kavanaugh, and D. Reynolds. 2022. Virtual Site Investigation to Evaluate Conceptual Site Model Development at DNAPL-Impacted Sites. National Groundwater Association, 4(3):44-58. doi.org/10.1111/gwmr.12537.