Title: Managing Research to Evaluate if a Risk is Acceptable: Applied to Nuclear Waste Siting
Authors: Rex Brown, George Mason University
Date: February 2000
Status: working paper

Evaluating whether a potentially hazardous facility is safe enough can cost much unnecessary time, resources and even public embarrassment, unless there is firm and defensible guidance on how to manage the supporting research.  This requires a test of acceptable risk and some way of assessing whether the test has been passed—or will be passed after research.  Current regulatory practice rarely does this soundly.

A topical example is how to control the massive research effort being undertaken to determine whether a proposed nuclear waste site should be approved.  According to regulations, radioactive release from an acceptable site should be below some limit with 90% probability.  The difficulty with this test is that, even if a site passes it now based on current evidence, there is no assurance that the site will still pass after additional research.  This test gives no motivation to “look for bad news” and the researcher’s parochial interests can drive the research program.

A more complex test is proposed that takes the “firmness” of risk assessment into account, along with a corresponding research management principle.  It extends the familiar  "performance allocation" procedure.  A set of second-order risk assessments of individual risk variables are specified as targets.  These are such that, if all targets are met, overall risk is acceptable.  Among possible target sets, the one most economical of research effort is chosen.  As research proceeds and evidence accumulates, the assessment targets are reconfigured and research effort is reallocated.  Research is halted when the test of acceptable second order risk is met or is found to be unachievable, and the site is either approved or disapproved.
 

Click here to access a pdf version of the paper

Back to the Decision Analysis Working Paper Index