Sandstone Drillability Prediction from Cuttings Using Crystal-Feature GBDT
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting sandstone drillability in oil and gas drilling are inadequate as they do not consider the geometric structure of mineral particles and lack a gradient boosting decision tree (GBDT) model, leading to high drilling costs and difficulty in obtaining drillability indices for the whole well section.
Innovation Solution
A GBDT prediction method is developed that utilizes crystal structure and mineralogical characteristics, involving cutting sample analysis, crystal boundary division, geometric parameter extraction, correlation analysis, and model training to accurately predict sandstone drillability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the laboratory micro-drilling test method is used to measure sandstone drillability, then the drillability index can be obtained, but the coring is difficult and the drilling cost is high
Solution Approach 1:
The patent uses image processing of sandstone cuttings to create a digital representation (copy) of the rock structure, extracting geometric parameters from images instead of performing actual micro-drilling tests. This copying approach eliminates the need for expensive and time-consuming laboratory tests while maintaining measurement capability.
Solution Approach 2:
The patent replaces the mechanical micro-drilling test system with an image-based analysis system. Instead of using physical drilling equipment to measure drillability, the system uses image processing and GBDT algorithms to predict drillability from visual characteristics of sandstone cuttings.
2Measurement precision
If the laboratory micro-drilling test method is used to measure sandstone drillability, then the drillability index can be obtained, but it is extremely difficult to obtain the drillability index of the whole well section
Solution Approach 1:
The patent develops a universal GBDT prediction model that can be applied to the entire well section using standard image processing equipment. The model once trained can predict drillability for any sandstone formation by analyzing cutting images, eliminating the need for separate micro-drilling tests at each location and enabling whole-well-section evaluation.
3Ease of operation
If the sandstone drillability is evaluated by using the content of cuttings elements, then the evaluation can be performed, but the geometric structure of mineral particles in sandstone is not considered
Solution Approach 1:
The patent transitions from evaluating drillability based solely on compositional data (mineral content) to incorporating geometric dimensional information. By extracting geometric parameters (area, perimeter, circularity, aspect ratio) from particle images, the system adds a spatial dimension to the evaluation, capturing the structural characteristics that influence drillability.
4Ease of operation
If the sandstone drillability is evaluated by texture coefficient (TC) index based on the drillability characterization method of sandstone texture, then the evaluation can be performed, but the influence of mineral differences on drillability is not considered
Solution Approach 1:
The patent applies local quality by considering both the compositional characteristics (mineral content) and geometric characteristics (particle shape and size) of different regions within the sandstone cutting images. The GBDT model integrates these locally extracted features to predict drillability, capturing the influence of mineral differences and their spatial distribution on drilling performance.
Data Source
AI summary
Disclosed is a gradient boosting decision tree (GBDT) prediction method for sandstone drillability based on crystal structure and mineralogical characteristics, including: acquiring cuttings samples of an area to be tested, dividing crystal boundaries based on the cuttings sample, and acquiring a plurality of crystal samples; numbering the plurality of the crystal samples, and extracting geometric parameters and mineral components of the plurality of the crystal samples; performing a correlation analysis on the geometric parameters, the mineral components and drillability data to obtain geometric parameters, the mineral components and the drillability; dividing the geometric parameters, the mineral components and the drillability into a training set and a testing set; training a GBDT model through the training set to obtain a trained GBDT model; and detecting accuracy of the trained GBDT model through the testing set to obtain prediction accuracy of trained GBDT model.