We wanted to reduce risk to very low levels in our manufacturing operations. We also needed to predict the effect on risk of potential future process changes. A survey was done on all processes to determine where flammable and combustible materials may be used. Process vessels were screened to determine which may be operating above the flash point without inert blanketing and with sufficient vapor space to generate a significant fire or explosion. This work was needed to ascertain whether or not to inert the process vessel, as well as its potential effect on the vent manifold system. An accurate assessment requires data on chemical flash points, but available data was limited primarily to pure components, not the mixtures—organic acids with water or anhydrides—with which we were dealing. Assessments using pure component data would have been very conservative, but also would have resulted in process modifications requiring large capital expenditures. To generate more accurate data we ran a four-component experiment designed for organic acid: water mixtures. Two different models were required to describe the results due to boundary issues. Excellent correlations were achieved. Results from this study helped us achieve our safety objectives with minimal capital expenditures.