Adjoint Approach for Assessment of Chemical Risk

摘要

An efficient adjoint approach is presented here to predict the potential risk caused by chemical accidents or chemical terrorist attacks in open spaces. Traditional method to assess chemical risk is via the direct approach, which applies the atmospheric dispersion model to compute the concentration of poisonous gas and then, the risk value. However, the location where the accident takes place is usually unknown beforehand, so the direct approach means that one must run the model repeatedly for every possible location of sources. An alternative approach—the adjoint method is developed. It is shown that with the adjoint method, the formulation of the risk value can be transformed equivalently. With the new formulation, it needs to solve the adjoint problem only once. It is obviously that time and computer expense can be considerably saved. A numerical simulation is presented to illustrate the adjoint method. The method developed here is a good tool to support decision-makers to locate chemical factories or storages, to design optimal paths for transporting dangerous materials, and to assign guard forces for chemical accidents.