US 12,228,336 B2
AI-enabled gas plant monitoring and optimization through computer vision
Bashar Y. Melhem, Abqaiq (SA); Mishar Kumar Paul, Calgary (CA); and Eisa A. Almakrami, Dhahran (SA)
Assigned to SAUDI ARABIAN OIL COMPANY, Dhahran (SA)
Filed by SAUDI ARABIAN OIL COMPANY, Dhahran (SA)
Filed on Mar. 21, 2022, as Appl. No. 17/655,713.
Prior Publication US 2023/0296313 A1, Sep. 21, 2023
Int. Cl. F25J 3/02 (2006.01); G05B 13/02 (2006.01)
CPC F25J 3/0295 (2013.01) [G05B 13/027 (2013.01)] 21 Claims
OG exemplary drawing
 
1. A method to optimize a gas processing plant with control parameters, comprising:
acquiring a plurality of images, using a camera disposed at a first location in the gas processing plant, of an incoming contaminated fluid;
processing, by a computer processor, the plurality of images by a machine-learned model to produce a contamination estimate, the contamination estimate comprising a segmentation map and a classification of hydrocarbon liquid carryover;
adjusting the control parameters of the gas processing plant, forming adjusted control parameters based on, at least in part, the contamination estimate, wherein the adjusted control parameters comprise an amount of anti-foam agent to be injected into a sub-processes of the gas processing plant that is downstream from the first location;
receiving a plurality of data signals from a plurality of sensors configured to monitor the gas processing plant, wherein the plurality of sensors are downstream from the first location; and
validating the adjusted control parameters with the received data signals by verifying that the adjusted control parameters optimize, or at least improve the performance of, the gas processing plant.