METHOD AND SYSTEM FOR EMISSIONS-BASED ASSET INTEGRITY MONITORING AND MAINTENANCE
A method involves obtaining current asset data for an asset, the current asset data including process data. The method further involves predicting, using a machine learning model, a methane emissions event associated with the asset, based on the current asset data, and reporting the predicted methane emissions event in a user visualization.
1 . A method, comprising:
obtaining current asset data for an asset, the current asset data comprising process data;
predicting, using a machine learning model, a methane emissions event associated with the asset, based on the current asset data; and
reporting the predicted methane emissions event in a user visualization.
2 . The method of claim 1 , wherein the asset comprises at least one petrochemical asset.
3 . The method of claim 1 , wherein the current asset data further comprises at least one selected from a group consisting of:
environmental data,
historical data associated with the asset, and
methane sensor data.
4 . The method of claim 1 , wherein the prediction of the methane emissions event comprises a classification performed between multiple categories of methane emissions events of different magnitude.
5 . The method of claim 1 , wherein the prediction of the methane emissions event comprises a prediction of at least one selected from a group consisting of a timing, a location, and a quantification of the methane emissions event.
6 . The method of claim 1 , further comprising:
predicting, using the machine learning model, a mitigation action for the methane emissions event.
7 . The method of claim 6 , wherein the mitigation action comprises adjusting a setting of a valve associated with the asset.
8 . The method of claim 6 , further comprising:
performing the mitigation action such that an actual occurrence of the predicted methane emissions event is avoided.
9 . The method of claim 1 , further comprising, prior to performing the prediction:
obtaining, for the asset, archived asset data comprising process data and methane sensor data; and
training the machine learning model to predict methane emissions events based on the archived asset data used as training data.
10 . The method of claim 9 , further comprising, prior to training the machine learning model:
preprocessing the archived asset data, comprising at least one selected from a group consisting of removing outliers and removing false positives.
11 . The method of claim 9 , further comprising, prior to training the machine learning model:
standardizing the archived asset data for sensor-agnostic operation of the machine learning model.
12 . A system, comprising:
a computing environment that:
obtains current asset data for an asset, the current asset data comprising process data,
predicts, using a machine learning model, a methane emissions event associated with the asset, based on the current asset data; and
a dashboard comprising a user visualization that reports the predicted methane emissions event.
13 . The system of claim 12 , wherein the machine learning model is a digital twin that establishes a virtual model that reflects characteristics of a physical environment related to the methane emissions event.
14 . The system of claim 13 ,
wherein the asset is in the physical environment reflected by the virtual model, and
wherein the asset is one selected from a group consisting of a vapor recovery unit, a compressor, storage tank, a power unit, a valve, a flange, and a seal.
15 . The system of claim 14 , wherein the physical environment comprises sensors that obtain the current asset data for the asset in the physical environment.
16 . The system of claim 15 , wherein the sensors comprise at least one selected from a group consisting of a fenceline sensor, a thermal camera, a non-thermal camera, an optical gas imaging camera, a drone-based sensor, a robot-based sensor, a helicopter-based sensor, an airplane-based sensor, and a satellite-based sensor.
17 . The system of claim 15 ,
wherein the computing environment comprises an edge computing platform that receives the current asset data from the sensors, and forwards the current asset data to the digital twin.
18 . The system of claim 13 ,
wherein the computing environment comprises a cloud computing platform, and
wherein the digital twin is executed on the cloud computing platform.
19 . The system of claim 13 ,
wherein the computing environment comprises a supervisor control and data acquisition (SCADA) system that obtains the process data associated with the asset and forwards the process data to the digital twin.
20 . The system of claim 12 ,
wherein the user visualization in the dashboard is configurable to enable monitoring of the methane emissions on a global, regional, side-wide, and asset-specific level.