George Mason University
George Mason University Mason
George Mason University

Chemical and Biological Test and Evaluation-Detector Agent Simulant Relationship

by Andrew Loerch

Publication Details MORE LESS

  • Published Date: December 1, 2010
  • Volume/Issue: 31/4
  • Publisher: International Test and Evaluation Association

Abstract

Realistic testing of chemical and biological defense systems requires an actual warfare agent. But use of such an agent is restricted to laboratory containment chambers, which are not realistic. This state of affairs has driven the chemical and biological defense community to integrate developmental testing and operational testing. Systems are challenged with both agent and simulant in laboratory containment chambers during developmental testing. A simulant is a substance that resembles the agent from the perspective of the system under test. A three-step procedure is described in this article to relate performance when challenged with simulant during operational testing to performance when challenged with agent. The procedure is based on classical logistic regression and judgment. If there is no statistical difference in performance between the agent and the simulant, then the results of the field test with the simulant can be used to predict agent performance. If there is statistical difference in performance between the agent and the simulant, but that difference is small and the system under test performs better when challenged with the agent than with the simulant, then the simulant performance is a lower bound to agent performance. What is defined as small difference is a matter of judgment. A graphical method is provided to provide insight as to the magnitude of the difference. In all other cases, the logistic regression can be used to predict performance based on operational test challenge concentrations and other parameters from the operational test.

Other Contributors

Charlie Homan
Expertise