Quality Soft Sensor Design for Crude Oil Desalting/Dehydration Unit Using Local Instrumental Variable (LIV) Approach
Volume Title: 1
1Center for Process Integration and Control (CPIC), Department of Chemical Engineering
2Center for Process Integration and Control (CPIC), Department of Chemical Engineering, University of Sistan and Baluchestan, Zahedan, Iran
3Center for Process Integration and Control (CPIC), Department of Chemical Engineering, University of Sistan and Baluchestan, Zahedan
Oil produced in most oil fields is accompanied by water which normally contains soluble salts such as chlorides of sodium, calcium, and magnesium that must be treated. If crude oil is left untreated, when it is processed in a refinery the salt can cause various operating problems. Desalting/dehydration plants (DDP) are often installed in crude oil production units in order to remove water-soluble salts from an oil stream. The product quality is only obtained through laboratory analysis with significant delays that makes it difficult to maintain continuous monitoring and accurate control of the unit. This paper describes the development of data-driven soft sensor for product quality of the oil desalting/dehydration process. The product quality is determined using salt removal efficiency, which is expected to be affected by five secondary variables as heating, settling time, chemical dosage, mixing time and wash water. The soft sensors are constructed based on novel state-dependent parameter (SDP) modelling method by local instrumental variable (LIV). The results showed that proposed soft sensor can be applied for DDP quality estimation as an alternative to laboratory testing. So, it becomes possible to continuously estimate product quality and to reach robust plant control system.
Data-driven soft sensor; Crude oil desalting/dehydration unit; quality prediction; State dependent parameter (SDP); Local instrumental variable (LIV) approach