HOUSTON, March 1, 2023 /PRNewswire/ -- Allocation of production/injection in multilayered fields is a critical task and whether it is reservoir simulation, accounting, and production reporting to authorities, reserves estimation or injection conformance, knowing how much is being injected and produced by each layer is a must-have in several key reservoir engineering workflows.
The task is not simple specially in mature fields with commingled production as the artificial lift equipment makes it impossible to run production logging tools in the borehole to measure individual layer contributions. So, operators rely on static allocation coefficients from some approximation methods like K*h determined from tracer jobs and simplified assumptions.
In 2020 Tachyus took on the challenge to design a tool that would solve the hurdles and that would provide allocations in a dynamic manner. This first approach worked well and has been proven to be accurate in several complex multilayered reservoirs.
But good enough is not enough for Tachyus and the team continued to work on improving the underlying algorithms…
As a result, we are excited to released Strateon, powered by Data Physics, the ultimate back allocation tool, combining mass balance with Darcy's Law in a machine learning workflow. Applicable to both primary and secondary production and injection allocation, Strateon is a fully automated and unbiased process that performs allocation at layer or pattern level and runs with minimum data.
In a test case in Argentina Strateon allocation in a multilayered reservoir was obtained in less than one week; conversely, it took the operator over one year to do a similar allocation using a simulation model.
As stated by Fernando Gutierrez, Tachyus CEO "Our mission is to find huge, painful problems and provide the best-in-class solutions, and Strateon is a testament to this strategy." Leading the R&D team in charge of technology development at Tachyus, Chief Technology Officer Dr. Pallav Sarma celebrated on this achievement: "Strateon is yet another success story of our recipe to combine well understood physics-based workflows with state-of-the-art machine learning algorithms in a fully automatic and unbiased manner to solve a critical problem."
