Machine Learning Methods for Handling Parameter Space Sampling Bias in Unconventional Well Performance Prediction

Technical Abstract
Written By GATE Energy

Oliver Rojas Conde | Henry Galvis Silva | Sebastien Matringe | Viking Engineering | Texas A&M University | Hess Corporation

Abstract

Prediction accuracy for extended-reach laterals and high-intensity completions is improved by applying a Gaussian Process Regression (GPR) model with a Matérn kernel that accounts for parameter space sampling bias. The method adjusts predictions based on local data density and provides uncertainty quantification, increasing reliability in underrepresented regions of the design space. Applied to wells from the Bakken Formation, the model outperforms traditional approaches and supports confident forecasting for non-standard development designs.

  • Document ID: URTEC-4249754-MS

  • Publisher: OnePetro

  • Source: SPE/AAPG/SEG Unconventional Resources Technology Conference

  • Publication Date: June 2025

___LZSettings_N4IghgrgLg9gwgJxgBxALigiBTANCAYxgBtjsCoBLGAOwEkATdEAVgDMAWcgZm247YBOQQEYCLbgHYWANgBGc7mxYh8RUuSq0AKgE9k2dCLUkIAWxoB1SgygALdACZJABnwNsAZ0oBzGswIwJAhPbGJVEDZKYihsBHRQKDAfZgBZGA9iShofXABBBCoogkowYgACOhpY0l9sGgI8AHksgDc48oAlGAArME9yuFoPXAA1SgBrbJ9ygFEc7Ow46YjA2J8YBF10GghSAF98H2hY+LQZNxA7bDAPBG1sAA8oZgjKM2TsPM8DCk6wLTMEQAOhU+CyPjsUDkMEe6DYZVC4MonigdA+Pi+EJoZnqLzQIDIbBeyNR6M+AGVKAAvQxobiXXFJBgAsAABRg3i0-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-B1GGypaJiFXQ7QAfW4LaREkLbHD2vbJAADhcDgOFOmQRHO27DpYQQZBAfYgA
Machine Learning Methods for Handling Parameter Space Sampling Bias in Unconventional Well Performance Prediction

Prediction accuracy for extended-reach laterals and high-intensity completions is improved by applying a Gaussian Process Regression (GPR) model with a Matérn kernel that accounts for parameter space sampling bias. The method adjusts predictions based on local data density and provides uncertainty quantification, increasing reliability in underrepresented regions of the design space. Applied to wells from the Bakken Formation, the model outperforms traditional approaches and supports confident forecasting for non-standard development designs.

VK-EXP-2024.02.0 - Figure 1.jpg
10 Steps to Safely Handling Failed Parts for Laboratory Testing

So your pipe burst, your pump shaft snapped, or your welded joint split, and you need to figure out why and how to fix it. The next steps you take can make the difference between finding answers and leaving the mystery unsolved until the next failure.

Slide1.JPG
STPA: A Systems Approach to Process Hazard Analysis

Systems Theoretic Process Analysis (STPA) is a systems approach to hazard analysis. It is based on the premise that accidents happen when we lose control. They are a control problem, not a failure problem.

Figure 1.jpg
HAZOP Risk Assessment via the RRR Matrix

The GATE approach to risk ranking makes risk assessment easier, more accurate and more repeatable by incorporating LOPA insights into the risk matrix approach.

Figure 1.jpg
Combining Stream-Based HAZOPs, LOPAs & STPAs into an All-In-One Assessment

HAZOP is the most commonly applied process hazard analysis (PHA) methodology in the processing industries. It is also the most flawed!

GATE Prho™ Simulator

Through recognizing our clients’ needs, our engineering expertise, and our proprietary in-house GATE Prho™ software engine, we provide customizable, cost-effective, easy-to use software tools to deliver better client outcomes quicker.

ARR2021-4-1.PNG
Offshore CCUS in the Gulf of Mexico: Technical Considerations

This article discusses high-level technical considerations when refitting an existing offshore infrastructure for CCUS applications.

Picture6.jpg
3 Ways to Go from Petrochemical to Renewable

Energy is changing. Because of the increasing threat of the consequences of a warming planet, new sources of energy are being developed that are expected to slow this planet-wide change. Some of these are well-known, such as nuclear, wind and solar, while others require more work, such as hydrogen, 3rd generation concentrated solar and fusion.

Picture3.png
How Oil & Gas Can Reduce Carbon Emissions to Make Way for Renewables

The overall aim of carbon capture and sequestration (CCS) is to store carbon dioxide (CO2) in geological formations in order to reduce the amount in the environment and lower the risk of catastrophic climate change. The rate of CO2 emission and the amount of it in the atmosphere has been increasing steadily since the beginning of the Industrial Revolution in the early part of the 19th century.

Download the Paper
GATE Energy
Previous

Systemic Risk in Major Capital Projects: Why Traditional Cost Estimation Fails & How to Fix It
Next

Why Better Decisions Start with Better Questions: How Structured Decision Analysis Improves Capital Project Outcomes