Combining Machine Learning and Classic Drilling Theories to Improve Rate of Penetration Prediction

Abstract Theoretically, rate of penetration (ROP) model is the basic to drilling parameters design, ROP improvement tools selection and drill time & cost estimation. Currently, ROP modelling is mainly conducted by two approaches: equation-based approach and machine learning approach, and machine learning performs better because of the capacity in high-dimensional and non-linear process modelling. However, in deep or deviated wells, the ROP prediction accuracy of machine learning is always unsatisfied mainly because the energy loss along the wellbore and drill string is non-negligible and it's difficult to consider the effect of wellbore geometry in machine learning models by pure data-driven methods. Therefore, it's necessary to develop robust ROP modelling method for different scenarios. In the paper, the performance of several equation-based methods and machine learning methods are evaluated by data from 82 wells, the technical features and applicable scopes of different methods are analysed. A new machine learning based ROP modelling method suitable for different well path types was proposed. Integrated data processing pipeline was designed to dealing with data noises, data missing, and discrete variables. ROP effecting factors were analysed, including mechanical parameters, hydraulic parameters, bit characteristics, rock properties, wellbore geometry, etc. Several new features were created by classic drilling theories, such as downhole weight on bit (DWOB), hydraulic impact force, formation heterogeneity index, etc. to improve the efficiency of learning from data. A random forest model was trained by cross validation and hyperparameters optimization methods. Field test results shows that the model could predict the ROP in different hole sections (vertical, deviated and horizontal) and different drilling modes (sliding and rotating drilling) and the average accuracy meets the requirement of well planning. A novel data processing and feature engineering workflow was designed according the characteristics of ROP modelling in different well path types. An integrated data-driven ROP modelling and optimization software was developed, including functions of mechanical specific energy analysis, bit wear analysis and predict, 2D & 3D ROP sensitivity analysis, offset wells benchmark, ROP prediction, drilling parameters constraints analysis, cost per meter prediction, etc. and providing quantitative evidences for drilling parameters optimization, drilling tools selection and well time estimation.

Download Full-text

Application of Machine Learning Techniques for Real Time Rate of Penetration Optimization

10.2118/202184-ms ◽

2021 ◽

Author(s):

Asad Mustafa Elmgerbi ◽

Clemens Peter Ettinger ◽

Peter Mbah Tekum ◽

Gerhard Thonhauser ◽

Andreas Nascimento

Keyword(s):

Machine Learning ◽

Genetic Algorithm ◽

Real Time ◽

Predictive Models ◽

Data Driven ◽

Machine Learning Techniques ◽

Time Rate ◽

Rate Of Penetration ◽

Drilling Parameters ◽

Learning Techniques

Abstract Over the past decade, several models have been generated to predict Rate of Penetration (ROP) in real-time. In general, these models can be classified into two categories, model-driven (analytical models) and data-driven models (based on machine learning techniques), which is considered as cutting-edge technology in terms of predictive accuracy and minimal human interfering. Nevertheless, most existing machine learning models are mainly used for prediction, not optimization. The ROP ahead of the bit for a certain formation layer can be predicted with such methods, but the limitation of the applications of these techniques is to find an optimum set of operating parameters for the optimization of ROP. In this regard, two data-driven models for ROP prediction have been developed and thereafter have been merged into an optimizer model. The purpose of the optimization process is to seek the ideal combinations of drilling parameters that would lead to an improvement in the ROP in real-time for a given formation. This paper is mainly focused on describing the process of development to create smart data-driven models (built on MATLAB software environment) for real-time rate of penetration prediction and optimization within a sufficient time span and without disturbing the drilling process, as it is typically required by a drill-off test. The used models here can be classified into two groups: two predictive models, Artificial Neural Network (ANN) and Random Forest (RF), in addition to one optimizer, namely genetic algorithm. The process started by developing, optimizing, and validation of the predictive models, which subsequently were linked to the genetic algorithm (GA) for real-time optimization. Automated optimization algorithms were integrated into the process of developing the productive models to improve the model efficiency and to reduce the errors. In order to validate the functionalities of the developed ROP optimization model, two different cases were studied. For the first case, historical drilling data from different wells were used, and the results confirmed that for the three known controllable surface drilling parameters, weight on bit (WOB) has the highest impact on ROP, followed by flow rate (FR) and finally rotation per minute (RPM), which has the least impact. In the second case, a laboratory scaled drilling rig "CDC miniRig" was utilized to validate the developed model, during the validation only the previous named parameters were used. Several meters were drilled through sandstone cubes at different weights on bit, rotations per minute, and flow rates to develop the productive models; then the optimizer was activated to propose the optimal set of the used parameters, which likely maximize the ROP. The proposed parameters were implemented, and the results showed that ROP improved as expected.

Download Full-text

Data-Driven Intelligent Application for Youtube Video Popularity Analysis using Machine Learning and Statistics

10.5121/csit.2021.110807 ◽

2021 ◽

Author(s):

Wenxi Gao ◽

Ishmael Rico ◽

Yu Sun

Keyword(s):

Machine Learning ◽

Data Processing ◽

Web Application ◽

Data Driven ◽

Accuracy Evaluation ◽

Video Content ◽

The Public ◽

Left Behind ◽

The World ◽

Web App

People now prefer to follow trends. Since the time is moving, people can only keep themselves from being left behind if they keep up with the pace of time. There are a lot of websites for people to explore the world, but websites for those who show the public something new are uncommon. This paper proposes an web application to help YouTuber with recommending trending video content because they sometimes have trouble in thinking of the video topic. Our method to solve the problem is basically in four steps: YouTube scraping, data processing, prediction by SVM and the webpage. Users input their thoughts on our web app and computer will scrap the trending page of YouTube and process the data to do prediction. We did some experiments by using different data, and got the accuracy evaluation of our method. The results show that our method is feasible so people can use it to get their own recommendation.

Download Full-text

Micro-Testing While Drilling for Rate of Penetration Optimization

Volume 11: Petroleum Technology ◽

10.1115/omae2020-18838 ◽

2020 ◽

Author(s):

Magnus Nystad ◽

Alexey Pavlov

Keyword(s):

Noisy Data ◽

Optimization Method ◽

Extremum Seeking ◽

Data Driven ◽

High Fidelity ◽

Drilling Process ◽

Optimization Scheme ◽

Rate Of Penetration ◽

Drilling Parameters ◽

Weight On Bit

Abstract The Rate of Penetration (ROP) is one of the key parameters related to the efficiency of the drilling process. Within the confines of operational limits, the drilling parameters affecting the ROP should be optimized to drill more efficiently and safely, to reduce the overall cost of constructing the well. In this study, a data-driven optimization method called Extremum Seeking is employed to automatically find and maintain the optimal Weight on Bit (WOB) which maximizes the ROP. To avoid violation of constraints, the algorithm is adjusted with a combination of a predictive and a reactive approach. This method of constraint handling is demonstrated for a maximal limit imposed on the surface torque, but the method is generic and can be applied on various drilling parameters. The proposed optimization scheme has been tested on a high-fidelity drilling simulator. The simulated scenarios show the method’s ability to steer the system to the optimum and to handle constraints and noisy data.

Download Full-text

Drill-bit displacement-source model: Source performance and drilling parameters

Geophysics ◽

10.1190/1.2227615 ◽

2006 ◽

Vol 71 (5) ◽

pp. F121-F129 ◽

Cited By ~ 6

Author(s):

Flavio Poletto ◽

Cinzia Bellezza

Keyword(s):

Near Field ◽

Source Model ◽

Rock Properties ◽

Signal Frequency ◽

Drill Bit ◽

Rate Of Penetration ◽

Axial Forces ◽

Drilling Parameters ◽

Geological Conditions ◽

Impact Area

The emission properties of a drill-bit source and the signature of the drill-bit seismograms depend on the dynamic action of the drill bit, which in turn is a function of the geological conditions, rock properties, and drilling parameters. In fact, drilling is a variable dynamic process, in which the vibrations produced by the drill bit are transmitted into the drillstring and formation with a partition of energy determined by the impedances at the bit. In this analysis, we study drill-bit performance and calculate the forces produced by an ideal drill bit to determine the variations in the signal signature and, ultimately, the conditions of source repeatability as a function of average drilling parameters and signal frequency. The model includes near-field effects and assumes the drill bit acts as a displacement source, producing axial bit forces with relative bit/formation displacement and vertical penetration in the formation. The analysis shows that the expected bit signature, which is equivalent to the ground force in Vibroseis, depends on average tooth impact area, percussion frequency, rate of penetration, formation strength, bit wear, and downhole pressure. In particular, the results show that larger amplitudes of the axial forces are expected at a lower rate of penetration, higher formation strength, and lower formation pressure.

Download Full-text

Enhancing Drilling Parameters in Majnoon Oilfield

Iraqi Journal of Chemical and Petroleum Engineering ◽

10.31699/ijcpe.2019.2.9 ◽

2019 ◽

Vol 20 (2) ◽

pp. 71-75

Author(s):

Majid M. Majeed ◽

Ayad A. Alhaleem

Keyword(s):

Focal Point ◽

Parameters Optimization ◽

Rate Of Penetration ◽

The Past ◽

Drilling Performance ◽

Optimum Parameters ◽

Drilling Parameters ◽

Drilling Data ◽

Weight On Bit ◽

Drilling Time

The objective of drilling parameters optimization in Majnoon oilfield is to arrive for a methodology that considers the past drilling data for five directional wells at 35 degree of inclination as a baseline for new wells to be drilled. Also, to predicts drilling performance by selecting the applied drilling parameters generated the highest rate of penetration (ROP) at each section. The focal point of the optimization process is to reduce drilling time and associated cost per each well. The results of this study show that the maximum ROP could not be achieved without sufficient flow rate to cool and clean the bit in clay intervals (36" and 24") hole sections. Although the influence of combination of Weight on Bit (WOB), Round per minute (RPM), and hydraulic horsepower on the bit in (16", 12 1/4" and 8 1/2") hole sections is a key to reduce drilling time, therefore, the drilling parameters produced the fastest ROP per each section was considered as optimum parameters likely to apply for the future wells.

Download Full-text

How to Land Modern Data Science in Petroleum Engineering

10.2118/205689-ms ◽

2021 ◽

Author(s):

Hongbao Zhang ◽

Yijin Zeng ◽

Lulu Liao ◽

Ruiyao Wang ◽

Xutian Hou ◽

...

Keyword(s):

Machine Learning ◽

Prior Knowledge ◽

Language Processing ◽

Data Science ◽

Parameters Optimization ◽

Data Driven ◽

Petroleum Engineering ◽

Great Success ◽

Well Production ◽

Data Product

Abstract Digitalization and intelligence are attracting increasing attention in petroleum engineering. Amounts of published research indicates modern data science has been applied in almost every corner of petroleum engineering where data generates, however, mature products are few or the performance are not up to peoples’ expectations. Despite the great success in other industries (internet, transportation, and finance, etc.), the "amazing" data science algorithms seem to be challenged when "landing" in petroleum engineering. It is time to calmly analyze current situations and discuss the methodology to apply modern data science in petroleum engineering, for safety ensuring, efficiency improvement and cost saving. Based on the experiences of several data products in petroleum engineering and wide investigation of literatures, the methodology is summarized by answering some important questions: what is the difference between petroleum engineering and other industries and what are the greatest challenges for algorithms "landing"? how could we build a data product development team? why the machine learning models didn't work well in real world, which are derived by typical procedures in textbooks? are current artificial intelligent algorithms perfect and is there any limit? how could we deal with the relationship between prior knowledge and data-driven methods? what is the key point to keep data product competitive? Several specific scenarios are introduced as examples, such as ROP modelling, drilling parameters optimization, text mining of drilling reports and well production prediction, etc. where deep learning, traditional machine learning, incremental learning and natural language processing methods, etc. are used. Besides detailed discussions in the paper, conclusions are summarized as: 1) the strengths and weakness of current artificial intelligence should be viewed objectively, practical suggestions to make up the weakness are provided; 2) the combination of prior knowledge (from lab tests or expert experiences) and data-driven methods are always necessary and methods for the combination are summarized; 3) data volume and solution portability are the key points to improve data product competitiveness; 4) suggestions on how to build a multi-disciplinary R&D team and how to plan a product are provided. This paper conducts an objective analysis on challenges for modern data science applying in petroleum engineering and provides a clear methodology and specific suggestions on how to improve the success rate of R&D projects which apply data science to solve problems in petroleum engineering.

Download Full-text

Enhancing Well, Reservoir and Facilities Management WRFM Opportunity Identification with Data Driven Techniques

10.2118/205596-ms ◽

2021 ◽

Author(s):

Manu Ujjwal ◽

Gaurav Modi ◽

Srungeer Simha

Keyword(s):

Machine Learning ◽

Data Processing ◽

Data Gathering ◽

Data Driven ◽

Petroleum Engineering ◽

Integration Time ◽

Facilities Management ◽

Learning Approaches ◽

Opportunity Identification ◽

Ranked List

Abstract A key to successful Well, Reservoir and Facilities Management (WRFM) is to have an up-to-date opportunity funnel. In large mature fields, WRFM opportunity identification is heavily dependent on effective exploitation of measured & interpreted data. This paper presents a suite of data driven workflows, collectively called WRFM Opportunity Finder (WOF), that generates ranked list of opportunities across the WRFM opportunity spectrum. The WOF was developed for a mature waterflooded asset with over 500 active wells and over 30 years of production history. The first step included data collection and cleanup using python routines and its integration into an interactive visualization dashboard. The WOF used this data to generate ranked list of following opportunity types: (a) Bean-up/bean-down candidates (b) Watershut-off candidates (c) Add-perf candidates (d) PLT/ILT data gathering candidates, and (e) well stimulation candidates. The WOF algorithms, implemented using python, largely comprised of rule-based workflows with occasional use of machine learning in intermediate steps. In a large mature asset, field/reservoir/well reviews are typically conducted area by area or reservoir by reservoir and is therefore a slow process. It is challenging to have an updated holistic overview of opportunities across the field which can allow prioritization of optimal opportunities. Though the opportunity screening logic may be linked to clear physics-based rules, its maturation is often difficult as it requires processing and integration of large volumes of multi-disciplinary data through laborious manual review processes. The WOF addressed these issues by leveraging data processing algorithms that gathered data directly from databases and applied customized data processing routines. This led to reduction in data preparation and integration time by 90%. The WOF used workflows linked to petroleum engineering principles to arrive at ranked lists of opportunities with a potential to add 1-2% increment in oil production. The integrated visualization dashboard allowed quick and transparent validation of the identified opportunities and their ranking basis using a variety of independent checks. The results from WOF will inform a range of business delivery elements such as workover & data gathering plan, exception-based-surveillance and facilities debottlenecking plan. WOF exploits the best of both worlds - physics-based solutions and data driven techniques. It offers transparent logic which are scalable and replicable to a variety of settings and hence has an edge over pure machine learning approaches. The WOF accelerates identification of low capex/no-capex opportunities using existing data. It promotes maximization of returns on already made investments and hence lends resilience to business in the low oil price environment.

Download Full-text