Automated Pressure Transient Analysis Using Pattern Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic pressure transient analysis (PTA) methods for production and injection wells are inefficient due to difficulties in accurately modeling pressure data, which often deviates from theoretical models and is contaminated by non-subsurface phenomena, and require impractical training processes for pattern recognition, leading to incorrect reservoir and well property calculations.
Innovation Solution
An automated PTA process that records continuously measured well pressure and flow rate data, identifies pressure buildup or fall-off intervals, generates log-log derivative plots, and uses pattern recognition to identify flow regime sub-intervals, specifically locating mid-time radial flow sub-intervals to calculate permeability, skin factor, and extrapolated pressure, thereby avoiding manual flow scenario identification and reducing misidentification risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If automatic PTA methods use model response matching (log-log type curve matching) to analyze pressure data, then reservoir and well properties can be inverted, but the accuracy deteriorates because actual pressure data does not exactly follow theoretical models and is contaminated by non-subsurface phenomena
Solution Approach 1:
The patent extracts and removes the problematic model response matching step from the automatic PTA workflow. Instead of forcing pressure data to match theoretical models, the method directly analyzes the derivative plot to identify flow regimes and calculate properties, eliminating the source of accuracy deterioration while maintaining automation.
Solution Approach 2:
The patent segments the pressure transient analysis into distinct flow regime identification steps rather than attempting a single comprehensive model match. By dividing the analysis into identifying early-time, mid-time, and late-time flow regimes separately, the method achieves higher accuracy in property calculation for each regime without requiring full model response matching.
2Extent of automation
If automatic PTA methods use pattern recognition techniques (such as ANN) to identify flow scenarios, then automation is improved, but the complexity increases due to requiring large amounts of historical training data
Solution Approach 1:
The patent replaces complex, data-intensive pattern recognition systems (like ANN requiring large training datasets) with a simpler, rule-based derivative plot analysis method. This simpler approach achieves automation without the heavy computational infrastructure and historical data requirements, effectively using a 'lighter' method that doesn't require extensive training resources.
3Measurement precision
If manual PTA methods are used to identify flow regimes from log-log derivative plots, then accuracy is improved through expert judgment, but the productivity deteriorates due to laborious and time-consuming analysis
Solution Approach 1:
The patent enables the system to automatically perform the flow regime identification task that previously required expert manual analysis. By programming the computer to automatically analyze derivative plots, identify flow regimes, and calculate properties without human intervention, the system serves itself, eliminating the need for manual expert analysis while maintaining accuracy.
Solution Approach 2:
The patent replaces the manual mechanical process of expert analysis (visual inspection and judgment of derivative plots) with an automated computational system. The computer automatically processes the derivative data, applies identification criteria, and generates results, substituting human manual work with automated mechanical/computational processes to improve productivity.
4Productivity
If simpler approaches are used to find radial flow by locating the first flat derivative on log-log plot, then productivity is improved by avoiding complex analysis, but the reliability deteriorates due to misidentification from noise or non-subsurface phenomena
Solution Approach 1:
The patent performs preliminary segmentation of the derivative plot into distinct flow regime sub-intervals before attempting to identify radial flow. By pre-identifying early-time, mid-time, and late-time flow regimes and their characteristic patterns, the system establishes a framework that guides the subsequent radial flow identification, preventing misidentification of noise as radial flow while maintaining automation speed.
Data Source
AI summary
Herein disclosed are methods and systems related to automatic pressure transient analysis (PTA). More particularly, herein disclosed are methods and systems related to automatic pressure transient analysis (PTA) of continuously measured pressure data associated with production or injection wells which utilizes in part pattern recognition analysis methods to determine key flow regimes in the measured data for more accurate determination of automatic pressure transient analysis (PTA) and determination of reservoir and well characteristic properties.