Large Scale Ammonia Process Optimization in Real-Time

At the Agricultural Chemical Complex (ACC) of Esso Chemical Alberta Limited (ECAL) in Redwater, Alberta, a real time optimization (RTO) project was initiated in 1985 with the objective of developing an application to optimize the production of a world scale ammonia unit. Due to the low energy design concept, the process is highly integrated with strong interaction between the production, energy consumers and the control set points. This forms the need to use a large scale "one piece" approach for the application development. A rigorous model-based RTO application was initially developed by writing the heat, material, pressure, stoichiometry, and other relationships of the process in dense format supported by Exxon Chemical\´s Equation Manager & Solver (EMS) package. For efficient computing in the real time environment, the dense equations were further transformed into sparse, residual equation format and a very significant reduction of execution time was obtained through the use of "intelligent" differencing and optimization method. Since rigorous principles were used, the RTO model can be used to perform process optimization, engineering simulation studies, and process performance monitoring and analysis. In order to update the RTO model to match the process at steady-state operating conditions, actual plant operating data was used to back calculate the model parameters through the same rigorous model relationships. Statistical analysis of the parameter history can reveal useful information about the process performance and model quality. These applications are targeted for implementation on an RTO computer control system designed to interface with the existing process control system.