engleski

Soft Sensor Applications in Refinery Production

One of the common problems in industrial plants is inability of the real-time and continuous measurement of key process variables. As an alternative, the use of soft sensors as a substitute for process analyzers and laboratory testing is suggested. With the soft sensors the objective is to develop an inferential model to estimate infrequently measured variables and laboratory assays using the frequently measured variables. In this poster review, three soft sensors for the refinery application are presented. The models are developed using data from refinery DCS system and from laboratory database. First soft sensor is developed for estimation of cold filter plugging point of diesel fuel as the crude distillation column side product. Second soft sensor estimates naphtha initial boiling point and end boiling point in the crude distillation unit. Both soft sensor models have been developed using multivariate regression technique and artificial neural networks. Statistical data analysis has been carried out and the results were critically judged. In third example, soft sensor is developed for dynamic model identification and process control of Sulphur Recovery Unit (SRU). The results are soft sensor models for optimal SRU control with aim to minimize SO2 and H2S emissions. The soft sensors were developed using multiple linear regression technique and using neural network -based and fuzzy logic models. Within MLP neural networks different learning algorithms are used (back propagation with variations of learning rate and momentum, conjugate gradient descent, Levenberg-Marquardt) as well as pruning and Weigend regularization techniques. Statistics and sensitivity analysis is given.

soft sensors; crude distillation column; SRU; neural network

nije evidentirano