Abstract:
In the chemical industry, control strategies which use optimization to improve the day-to-day performance of unit operations can potentially perform badly during non-routine process upsets. Traditional controller analysis techniques that attempt to quantify the economic benefits of these controller strategies are excellent at quantifying the day to day improvements, but are ill-equipped to capture and quantify the economics of non-routine process upsets. On the other hand, risk analyses such as LOPA (Layer of Protection Analysis), are excellent at quantifying the economic risks of non-routine process upsets, but are unable to capture the economics of day to day operations. In many plants, the trade-off between optimization vs non-routine operational stability is ultimately settled by experienced plant managers and operators using their best judgment. In this work, we have proposed a combined frame work that combines LOPA with NPV (Net Present Value) analysis that will take the “guess work” out of this type of decision. We have then applied this comprehensive framework to a case study of an industrial methanol distillation column. We compared the current operating strategy used by plant management to a new steady state control strategy that can improve its day to day performance. This comprehensive approach was able to determine that, the proposed control strategy has a better combined overall financial performance than the current control strategy.
