News Oil & Gas

Leveraging AI-ML models for smart natural gas pipelines, Energy News, ET EnergyWorld


New Delhi: Every company is talking about AI and ML technologies nowadays and every IT company, be it a start-up or a super major, has specialized offerings around AI and ML technologies. However, the extent and complexity these new age technologies address varies based on each individual companies’ outlook. Over the years, the inherent expectations from the oil and gas business applications are to build an error free application which requires good amount of feature engineering considering multiple variable factors that affect the expected output. This requires large amount of interaction between the business domain and technology domain. It is important to ensure the project team has (a) right kind of resources from each domain that are well verse with the boundaries within which the solutions need to be developed, (b) variety and veracity of data considered for the error free solutions.

We always say oil and gas industry is deeply interconnected right from the exploration, drilling and production of oil and gas to transportation, refining and distribution of oil and lot of work has been done to forecast, predict, and optimize the business variables required for analytics at each stage. The midstream segment traditionally has been a stable business connecting established demand and supply centers. To effectively serve this newly found growth and increased dynamism in the business, midstream companies are focusing on maintaining and optimizing their networks, a priority for which technology exists but that midstream companies have yet to fully integrate across their full network of pipelines and associated infrastructure.

A “smart pipeline”, much like a “connected factory”, uses sensors on assets to increase automation, safety, and throughput. This requires IT/OT integration and analytics on the IoT data streams. Midstream companies have been slower to adopt IT than most other areas of the oil and gas value chain.

However, advance analytics companies work with the instrumentation vendors to add a layer of prescriptive analytics to asset data. As part of the descriptive supply allocation plan to various downstream customers, the analytical tools help to optimize the overall company margins. It starts with forecasting natural gas demand requirement for various time periods. The sensitivity model helps on top of forecasting and optimization model in analyzing the gas supply contracts and in optimizing the gas allocations to its various consumers while maximizing the overall margins. The sensitivity model considers aforesaid gas demand forecast from each of its customer as well as various contractual obligations such as Adjusted Annual Contracted Quantity (AACQ), Agreed Makeup Gas for the next year, Make-good gas, Restoration gas, Downward flexibility, Upward flexibility, etc.

Typically, the shipper enters into long-term gas sales contract with the consumers, purchase contract with suppliers and transportation contract with the gas transporter. However, it is the gas transporter, who lifts the gas from the supplier and delivers it to the consumer on behalf of the shipper. While the contract provides guidelines for gas sale or purchase or transportation over a long period, the actual requirement for each year is finalized before start of the year, based on the anticipated demand-supply situation and the outstanding position of each company. Annual gas requirement (further broken down to quarterly, monthly, weekly, and daily) through proposed AI-ML models gets finalized among the customers and shippers, shippers and suppliers, transporter, and shippers. These forecasted requirements get utilized by the smart AI-ML based gas demand-supply management tool to generate an optimized allocation plan that helps in maximizing the overall margins of midstream gas pipeline companies.

In this way, AI-ML based smart gas pipeline demand-supply management tool helps consumers, shippers, suppliers, and transporters benefit by minimizing (a) cost of gas purchases, (b) cost of additional transport bookings, (c) Variable payments for usage of capacity bookings and maximizing (a) income from gas sales while balancing restrictions at each network nodes, pipeline capacity, min/max contract conditions and other tailored constraints in operation of the smart pipeline network.

[This piece was written exclusively for ETEnergyworld by Naveen Sikka, Principal Industry Consultant at SAS]





Source link

Leave a Comment