Application of a Technology Risk Assessment Model to Public Decision Making. Marilyn Jacobs Quadrel, Jonathan M. Cruse, Kimberly M. Fowler, Rick Cameron, Russel J. Treat, and W. David McCormack
The U.S. Department of Energy (DOE) has embarked on a new approach for developing technologies to solve major DOE environmental management problems. One of the new "Focus Areas" is the approximately 350 underground storage tanks (UST) that are distributed across the country. This paper describes a decision and risk communication tool (the technology risk assessment model) for screening new technologies for their application to tank waste remediation systems and for their funding priority. It is comprised of a risk model and a decision interface. The former assesses the potential reductions in risk offered by new technology relative to the baseline risk of an entire system. The latter puts this information in context of other values articulated by decision makers and stakeholders in the DOE system. The model was designed to generate and communicate systematic risk and cost information during funding deliberations among technology program decision makers, technologists, and public stakeholders.
The technology risk assessment model uses a structured methodology for estimating risks and costs of new technologies for remediating DOE UST relative to a base case. The base case establishes a flowsheet and material balance for all stages of a baseline UST remediation. The model incorporates new technologies (test cases) into this complete remediation system to allow a disciplined assessment of changes in the risk and cost of specific functions as well as overall system risk and cost. Potential risks to workers and to the public associated with both routine and failure conditions are currently represented. These data are combined with additional decision variables and stakeholder values, which are elicited on line from decision participants for different configurations of the model.
This activity addresses a current gap in DOE's decision making "toolbox" for selecting what technologies to fund when those technologies must function as part of a large and complex system. Current evaluations of emerging technologies typically focus on the ability of individual technologies to satisfy the performance requirements of a specific function. This does not provide decision makers who have fiscal responsibility for program budgets to determine whether technologies provide sufficient benefit to the overall system to warrant development. On the other hand, current systems engineering or Environmental Impact Statement (EIS) analyses that do address entire systems are larger and more costly than practical for the multiple, more detailed decisions required to guide a technology development program. These more complex models are also not typically amenable to the online sensitivity analyses required to address expert disagreements regarding predictions about a new technology's performance under different conditions.
Model Objectives
The model and associated software presented here addresses these major shortcomings. It summarizes system-level information that is difficult for experts to consider when assessing individual technologies. It provides a simplified systems perspective that is more practical for technology decisions than more complicated systems engineering models. It easily accommodates differences of opinion in assumptions or predicted performance. Finally, the interface is easy to display and understand, which makes this model useable "on-line" in forums involving multiple decision makers, including public stakeholders and regulators who may not have the detailed technical knowledge required to assess the model's accuracy but who want to see (or have the opportunity to assess and differentially weight) how a technology performs on a broad range of variables.
Technology Risk Assessment Model
The baseline model consists of 13 modules that encompass the complete set of tank remediation functions, including waste retrieval, treatment, disposal, and tank and facility decontamination and decommissioning. Each module is a group of similar or related process steps linked to a specific technology, and is comprised of one or more unit operations or major components. The modular format allows users to add or replace modules to define the test cases and provides a practical breakdown of overall system risk and cost.
Risk and cost analysis are performed for each module using available data, elicited expert judgment, and simplifying assumptions where data were not available. The values for each module are summed to provide total system values. The assessment methods were selected to meet requirements that they be (1) comprehensive and "cradle to grave" in scope, (2) easy for decision makers to understand, (3) easy to apply early in the technology evaluation process (when limited information is available), and (4) amenable to changes in assumptions or estimates to facilitate discussion among users.
The risk assessment spans both operational and post-operational phases of remediation and addresses all potentially significant contaminant sources and all potential contaminant receptors. Potential risks associated with both routine and failure conditions affecting both workers and the public are addressed. Routine risks to the public are currently estimated using the Multimedia Environmental Pollution Assessment System (MEPAS) computer code (Droppo et al., 1989). Risks to workers and failure risks are currently determined using a "Risk Priority Grid" that was developed under DOE funding and is commonly used at the Hanford site. Strengths and weaknesses of these methods are discussed. All costs are transformed into total net present worth based on a 13-year operational life and 1994 dollars.
While risk and cost measures are critical to making defensible decisions, they do not represent the only important decision variables. Stakeholders at the Hanford site, which has the biggest UST remediation problem, have identified a set of 14 values that they want to guide remediation decisions. A decision interface has been added to the technology risk assessment model to facilitate discussions among decision makers, technologists, and public stakeholders or regulators.
That interface displays the risk and cost estimates from the technology risk assessment model along with performance of the baseline and test cases on the remaining relevant decision variables. These are estimated and weighted by meeting participants, using input from technical and regulatory experts who can assess the system's performance on each criterion. Preferences for aggregated or disaggregated risk and performance measures, and challenges in displaying the information for practical decision applications, will be discussed.