Over the last years, there has been a continuous trend in the oil and gas industry to reduce exploration and development of new assets and putting more focus in the generation of cash flow by the optimization of existing producing assets to increase revenue without significant extra CAPEX investments. For this, Repsol is incorporating data analytics into its operations and has developed an in-house product (“pWatch”), where hybrid models, which combine data analytics and hydro-dynamic/fluid dynamic first principles, are used to optimize oil and gas production. This work describes two use cases of this methodology where Repsol assets benefit from Data Analytics tools.
The first application is called “Virtual Production Test” and it highlights the benefits of real time virtual flow metering based on hybrid models, to predict volumetric gas and liquid flowrates at any pressure and temperature conditions, considering operational parameters. Comparable accuracy to commercial multiphase flow meters (MPFM: 3-5% gas and 5-10% liquid rate deviation) with lower cost is confirmed.
The second application presents a methodology for continuous “Gas-lift optimization” through data-driven predictive models. Considering the limitations on gas availability, the routing of gas lift among all the wells requires an optimization to increase return which can be successfully implemented with presented approach. Results matched the production examples shown in this paper, showing daily liquid production increase from 1 to 3%, depending on the case.
