Abstract: A computer-implemented method for real-time adaptive control of demulsifier injection using self-learning AI models is described. The method includes obtaining sensor data in a gas-oil separator plant and calculating demulsifier input parameters. The method also includes predicting process variables comprising at least water content and determining an injection rate in real time based on the predicted process variables.

Abstract: Systems and methods for mitigating flow variability and slugging in pipelines (e.g., trunk-lines leading to gas-oil separation plants (GOSP)) use supervised machine learning algorithms to develop operational strategies for controlling inflows to facilities such as GOSPs.

Abstract: A method for fluid catalytic cracking (FCC) of petroleum oil feedstock includes reacting the petroleum oil feedstock with a catalyst mixture in a reaction zone of an FCC unit to obtain a product stream including desulfurized hydrocarbon product, unreacted petroleum oil feedstock, and spent catalyst. During the reacting a process control system develops a process model based on data collected during the reacting, the process model characterizing a relationship among the feed rate of the base cracking catalyst, the feed rate of the FCC additive, the operating conditions, the composition of the product stream, and emissions from the reaction; and one or more of (i) a target feed rate of the base cracking catalyst, (ii) a target feed rate of the FCC additive, and (iii) one or more target operating conditions of the reaction in the reaction zone to reduce the emissions from the FCC unit and to increase a yield of the desulfurized hydrocarbon product in the product stream are determined.

Abstract: A method for fluid catalytic cracking (FCC) of petroleum oil feedstock includes reacting the petroleum oil feedstock with a catalyst mixture in a reaction zone of an FCC unit to obtain a product stream including desulfurized hydrocarbon product, unreacted petroleum oil feedstock, and spent catalyst. During the reacting a process control system develops a process model based on data collected during the reacting, the process model characterizing a relationship among the feed rate of the base cracking catalyst, the feed rate of the FCC additive, the operating conditions, the composition of the product stream, and emissions from the reaction; and one or more of (i) a target feed rate of the base cracking catalyst, (ii) a target feed rate of the FCC additive, and (iii) one or more target operating conditions of the reaction in the reaction zone to reduce the emissions from the FCC unit and to increase a yield of the desulfurized hydrocarbon product in the product stream are determined.

Abstract: A method for fluid catalytic cracking (FCC) of petroleum oil feedstock includes reacting the petroleum oil feedstock with a catalyst mixture in a reaction zone of an FCC unit to obtain a product stream including desulfurized hydrocarbon product, unreacted petroleum oil feedstock, and spent catalyst. During the reacting a process control system develops a process model based on data collected during the reacting, the process model characterizing a relationship among the feed rate of the base cracking catalyst, the feed rate of the FCC additive, the operating conditions, the composition of the product stream, and emissions from the reaction; and one or more of (i) a target feed rate of the base cracking catalyst, (ii) a target feed rate of the FCC additive, and (iii) one or more target operating conditions of the reaction in the reaction zone to reduce the emissions from the FCC unit and to increase a yield of the desulfurized hydrocarbon product in the product stream are determined.

Abstract: The current invention is an empirical optimization method based on statistical modeling relating NGL plant process variables with the refrigeration system's electricity usage. The method identifies the key process control variables in an NGL plant to be optimized. This method is applicable to an NGL plant that uses dual refrigeration systems.