Drilling Data Classification via Change Point Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Interpreting drilling data during hydrocarbon well drilling operations is challenging and time-consuming, leading to inaccuracies and delays in identifying drilling characteristics and parameters, which can hinder timely adjustments and control of drilling operations.

Innovation Solution

A system that collects drilling data, applies it to a well decision tree for classification, uses majority voting to determine operation classifications, and conducts change point detection to identify changes in classifications, enabling real-time control of drilling operations based on identified parameters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual interpretation of drilling data is used to identify drilling characteristics, then the driller can understand the drilling operation, but the process is time-consuming and leads to delays in identifying drilling parameters

Engineering Contradiction:

Improveaccuracy of drilling characteristic identificationVSAvoidtime delay in identifying drilling parameters

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical interpretation of drilling data with an automated computer-based system that uses machine learning models and algorithms to automatically identify drilling characteristics and parameters from sensor data, eliminating the time delay associated with manual analysis while maintaining or improving accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service by automatically processing drilling data through trained machine learning models to generate drilling characteristic identifications without requiring continuous manual intervention, allowing the system to autonomously identify drilling parameters and characteristics in real-time

Inventive Principle:

Principle #25Self-service

2Productivity

If manual data interpretation is used to identify drilling characteristics, then the driller can make decisions, but inaccuracies occur in identifying drilling parameters

Engineering Contradiction:

Improveefficiency of drilling operation controlVSAvoidaccuracy of drilling parameter identification

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent replaces manual interpretation with automated machine learning-based systems that consistently and accurately identify drilling characteristics from sensor data, eliminating human error and variability while improving the precision of drilling parameter identification

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system implements feedback mechanisms where drilling data is continuously collected, analyzed by machine learning models, and used to generate real-time recommendations and alerts, allowing for continuous improvement and verification of drilling characteristic identifications through iterative processing

Inventive Principle:

Principle #23Feedback

3Speed

If real-time drilling data analysis is implemented, then timely control decisions can be made, but the system complexity increases

Engineering Contradiction:

Improvespeed of drilling operation responseVSAvoidcomplexity of data processing system

Core Design Contradiction:

SpeedVSDevice complexity

Solution Approach 1:

The patent segments the complex data analysis task into distinct components: data collection from sensors, preprocessing of raw data, application of specific machine learning models for different drilling characteristics, and generation of control recommendations. This modular approach manages system complexity while enabling real-time analysis

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system introduces intermediary processing layers including data preprocessing modules and feature extraction components that bridge raw sensor data and final drilling characteristic identifications, simplifying the overall system architecture by breaking down complex transformations into manageable intermediate steps

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11739626B2Systems and methods to characterize well drilling activities

Publication Date: 2023.08.29 SAUDI ARABIAN OIL CO

AI summary

Provided is a method of drilling a hydrocarbon well that includes conducting a drilling operation, collecting drilling data including characteristics of the drilling operation over a timespan, determining (based on the drilling data) drilling conditions for instants of time within the timespan, determining (based on application of the drilling conditions) preliminary classifications identifying a preliminary classification of the drilling operation for instants of time within the timespan, determining (based on the preliminary classifications) a series of classifications for the drilling operation that each indicate a determined classification for a respective instant of time, determining (based on the series of classifications) a change of classifications, conducting (in response to determining the change of classifications) a change point detection to identify a time of the change of classifications, generating drilling characteristic data indicating the time, and conducting the drilling operation in accordance with the time.