In this article we consider strategies for exploiting supercomputer
technology in solving the sparse matrix problems arising in process simulation
and optimization. A new multifrontal solver for use in simulating equilibrium-stage
processes and a new parallel frontal solver for large-grained parallel
solution of process simulation and optimization problems are described.
Results for several problems, including large-scale industrial problems,
are given.
in Computational Chemistry and Chemical Engineering, eds., G. Cisneros,
J. A. Cogordon, M. Castro, and C. Wang, World Scientific,
pp. 377-393 (1997).