Abstract:In the early time of oilfield development, insufficient production data and unclear understanding of oil production presented a challenge to reservoir engineers in devising effective development plans. To address this challenge, this study proposes a method using data mining technology to search for similar oil fields and predict well productivity. A query system of 135 analogy parameters is established based on geological and reservoir engineering research, and the weight values of these parameters are calculated using a data algorithm to establish an analogy system. The fuzzy matter-element algorithm is then used to calculate the similarity between oil fields, with fields having similarity greater than 70% identified as similar oil fields. Using similar oil fields as sample data, 8 important factors affecting well productivity are identified using the Pearson coefficient and mean decrease impurity (MDI) method. To establish productivity prediction models, linear regression (LR), random forest regression (RF), support vector regression (SVR), backpropagation (BP), extreme gradient boosting (XGBoost), and light gradient boosting machine (LightGBM) algorithms are used. Their performance is evaluated using the coefficient of determination (R2), explained variance score (EV), mean squared error (MSE), and mean absolute error (MAE) metrics. The LightGBM model is selected to predict the productivity of 30 wells in the PL field with an average error of only 6.31%, which significantly improves the accuracy of the productivity prediction and meets the application requirements in the field. Finally, a software platform integrating data query, oil field analogy, productivity prediction, and knowledge base is established to identify patterns in massive reservoir development data and provide valuable technical references for new reservoir development.

Multi-dimensional Interactions in the Oilfield Market: A Jackknife Model Averaging Approach of Spatial Productivity Analysis

Total-Factor Spillovers, Similarities, and Competitions in the Petroleum Industry

Effects of Ownership and Business Portfolio on Production in the Oil and Gas Industry

Different Behaviors in Natural Gas Production Between National and Private Oil Companies: Economics-Driven or Environment-Driven?

Resource Allocation in Multi-Divisional Multi-Product Firms

Productivity Drain or Productivity Gain? the Effect of New Technology Adoption in the Oilfield Market

Non-structural and Structural Models in Productivity Analysis: Study of the British Isles During the 2007–2009 Financial Crisis

Productivity spillovers among linked sectors

Estimation of industry-level productivity with cross-sectional dependence by using spatial analysis

The Impact of the Digital Economy on the High-Quality Development of Green Marketing in China's Manufacturing Industry: Applications of the Spatial Durbin Model

Industrial Productivity Performance in Chinese Regions (1987–2002): a Decomposition Approach

Oilfield analogy and productivity prediction based on machine learning: Field cases in PL oilfield, China

A Fractional Step Method to Solve Productivity Model of Horizontal Wells Based on Heterogeneous Structure of Fracture Network

Dynamics of Material Productivity and Socioeconomic Factors Based on Auto-Regressive Distributed Lag Model in China

Grain industrial agglomeration and grain green total factor productivity in China: A dynamic spatial durbin econometric analysis

Shock absorber and shock diffuser: the multiple roles of industrial diversity in shaping regional economic resilience after the Great Recession

Density,Distance,Division,and Urban Labor Productivity:Empircal Study Based on Panel Data of 277 Chinese Cities During 2004-2009

Catch-Up Growth and Inter-industry Productivity Spillovers: Evidence from Trade Data

Techniques for Assessment of Productivity Sensitive Factors in Wells with fracturing Operations in Fractured Reservoirs

Shale gas well productivity potential evaluation based on data-driven methods: case study in the WY block

Dynamic Metafrontier Malmquist–Luenberger Productivity Index in Network DEA: An Application to Banking Data