01-08 Practical oil & gas production forecasting using machine learning

Objectives

The forecasting of oil and gas production is critically important for asset managers and subsurface staff in order to anticipate the profitability of their businesses. Oil and gas fields need ongoing production forecasts for planning purposes and to understand their performance. It is necessary to develop a range of production forecasts (Base-Low-High cases) based on actual production data. We are pleased to offer a practical production forecasting course using Python based machine learning (ML) techniques based on in Pycaret TS Alpha and KATS – two Python time series machine learning libraries. KATS is a lightweight, easy-to-use, and generalizable framework to perform time series analysis developed by Facebook while Pycaret TS Alpha is an automated machine learning (AutoML) library – a low code library that minimizes the amount of coding one has to do. Pycaret TS Alpha could a strong contender once it matures from Alpha status and is integrated into Pycaret. This is envisaged to occur in mid 2022. In this course, actual production data from the Volve Field in Norway is used. The analysis is focused on monthly oil production data. KATS provides a full set of tools for forecasting that includes 10+ individual forecasting models, ensembling, a self-supervised learning (meta-learning) model, back testing, hyperparameter tuning, and empirical prediction intervals. KATS currently supports the following 10 forecasting models; Linear, Quadratic, ARIMA, SARIMA, Holt-Winters, Prophet, AR-Net, LSTM, Theta and VAR.

Who should attend

Prerequisites

Duration

Languages

Geoid Training available