Mitigating flow variability and slugging in pipelines
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.
1 . A method for controlling fluid flow for mitigation of flow variations upstream of a gas oil separation plant, the method comprising:
obtaining first sensor data associated with one or more trunk lines;
obtaining second sensor data associated with a gas and oil separation plant (GOSP) configured to receive fluid from the one or more trunk lines;
wherein the first and second sensor data are obtained for a first time period and a second time period that is different than the first time period;
extracting one or more features from the first sensor data and the second sensor data based identified features of training data that trains a machine learning model associated with the GOSP, the training data comprising labeled data representing incoming feed gas of the GOSP, the labeled data associating upstream flow volumes, input flows, and operating conditions with incoming feed gas volumes;
generating, from the extracted one or more features, a first feature vector for the first time period and a second feature vector for the second time period;
processing, by a machine learning model, the first feature vector and the second feature vector, the machine learning model being trained with the training data;
determining, based on the processing, predicted incoming feed gas volumes; and
controlling, based on the predicted incoming feed gas volumes, settings of machine operated valves upstream of the GOSP.
2 . The method of claim 1 , wherein extracting the features is based on principle component analysis of the training data comprising first sensor data and second sensor data.
3 . The method of claim 2 , wherein the first sensor data comprises values for process measurements including at least one of an upstream flow volume for each of the one or more trunk lines and an input flow volume from each of the one or more trunk lines.
4 . The method of claim 3 , wherein the second sensor data comprises one or more inlet manifold pressure values, one or more values of levels for oil, water, or gas, and a gas pressure value for at least one location in the GOSP.
5 . The method of claim 1 , wherein the first time period includes periods when crude trunk-line scraping activities are performed.
6 . The method of claim 1 , wherein controlling the settings of the machine operated valves upstream of the GOSP comprises chocking the machine operated valves of one or more trunk lines.
7 . The method of claim 1 , wherein controlling the settings of the machine operated valves upstream of the GOSP comprises directing certain trunk lines to a specific GOSP.
8 . The method of claim 7 , wherein controlling the settings of the machine operated valves upstream of the GOSP comprises splitting at least one trunk line into various GOSPs.