An orthogonal design of experiment/exploratory data analysis for plutonium contamination

This paper describes a statistical-experimental design approach for identifying sources of Pu contamination in the radioactivity material analysis of dried shellfish at the National Institute of Standards and Technology (NIST). Identification and elimination of sources of contamination were required for certifications of the radioactivity concentration of actinides in the proposed reference material. Sources of contamination included the fume hood, glassware, and reagents. The experimental design employed reduced number of required experiments by a factor of 100. The result is an orthogonal design of experiment (ODEX) and subsequent data was analyzed using Exploratory Date Analysis (EDA), which narrowed the 200+ experiments to a manageable 16. This design made it possible to identify the sources and interactions of sources of contamination in a timely and cost effective manner. (author)