The overall goal of the planned research projects is the integration of process design and process control. Specifically, the development of economically viable approaches to be used at the design stage of a process to analyze for controllability/operability problems. The PI plans to look at three areas: (1) the effect of process structure on process controllability/operability, (2) correlation of transient performance with process parameters (available from steady state data) and (3) the use of composition measurements to improve distillation column control. Chemical plants can be viewed as being built up from a small number of basic structures. Irrespective of the type of unit or units involved, this structure affects overall process performance. A defining set of basic structures might include series-parallel, recombining, and recycle structures. The latter are widely used in the chemical industry for economic reasons and can have a drastic effect on both the open loop and the closed loop performance of a system. The goal of this research is to determine the effect that recycle processing configurations have on process operability/controllability and to correlate this effect with design parameters. Two examples will be used to illustrate the control problem caused by recycle structures: evaporator configurations and recycle systems involving distillation towers, heat exchangers, and industrial reactors. Control system performance will be related to key design variables such as throughput, holdup, heat integration, and recycle ratio of mass and/or energy. The PI has developed a method using a dimensionless number called the relative disturbance gain (RDG) which when used in conjunction with the calculated relative gain (RGA) can be employed to assess a process' transient behavior based on steady state information. In essence what the RDG measures is whether the closure of other loops in a multiloop system produces favorable or unfavorable interaction in terms of transient performance of the loop under consideration. Ultimately the PI plans to relate correlating parameters to key design variables and to integrate RGA, RDG and measured peak properties into an expert system.
