In this paper we address the stochastic capacitated hazardous materials facility location problem in which demands and variable production and transportation costs are uncertain. We introduce a robust optimization model for the siting of noxious facilities under uncertainty. The problem is formulated as a nonlinear program. Plans developed in the model will be both solution robust and model robust. In other words, the optimal solution from the model is `close' to being optimal for any realization of the model scenarios, and the optimal solution has `close' to no error with respect to uncertain parameters. In order to show that the model works in a real world application, we used our model to plan facility siting decisions for the Albany, New York region.