Multi-task learning for virtual flow metering

TL;DR

The findings show that MTL improves robustness over single-task methods, without sacrificing performance, and yields a 25-50% error reduction on average for the assets where single- task architectures are struggling.

Abstract

Virtual flow metering (VFM) is a cost-effective and non-intrusive technology for inferring multiphase flow rates in petroleum assets. Inferences about flow rates are fundamental to decision support systems that operators extensively rely on. Data-driven VFM, where mechanistic models are replaced with machine learning models, has recently gained attention due to its promise of lower maintenance costs. While excellent performances in small sample studies have been reported in the literature, there is still considerable doubt about the robustness of data-driven VFM. In this paper, we propose a new multi-task learning (MTL) architecture for data-driven VFM. Our method differs from previous methods in that it enables learning across oil and gas wells. We study the method by modeling 55 wells from four petroleum assets and compare the results with two single-task baseline models. Our findings show that MTL improves robustness over single-task methods, without sacrificing performance. MTL yields a 25-50% error reduction on average for the assets where single-task architectures are struggling.

Figures (11)

• Figure 1: Asset with $J$ wells sharing a single separator.
• Figure 2: Choke valve with instrumentation.
• Figure 3: Scatter plot of choke opening and upstream pressure for all wells. Observations from a single well is highlighted and coloured by days since first observation. Choke is continuously adjusted to counteract the declining reservoir pressure.
• Figure 4: Box plot of pressure upstream observations for each well. The dotted vertical lines and coloring indicate which wells are from the same asset.
• Figure 5: Block diagram of the model architecture. The model is composed of two functions. A task specific domain adaptation $g$, and a flow computation $h$, which takes both task parameters and shared parameters.