|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Feder, Judy (JPT Technology Editor)
This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE-191779-18ERM-MS, “A Fast-Paced Work Flow for Well Spacing and Completions Design Optimization in Unconventional Reservoirs,” by Hoss Belyadi, SPE, and Malcolm Smith, EQT Corporation, prepared for the 2018 SPE Eastern Regional Meeting, Pittsburgh, Pennsylvania, USA, 7–11 October. The paper has not been peer reviewed.
Well spacing optimization is one of the more important considerations in unconventional field development. The essence of field development and optimization is to use completions design and well spacing to optimize the net present value (NPV) of the field on the basis of current commodity pricing, capital expenditure (CAPEX), operating cost, cycle time, and net revenue interest. A substantial variation in any of these essential factors must be studied to make sure the appropriate changes are accounted for in field development and optimization. A fast-paced and dynamic work flow has been developed that can be applied in different shale reservoirs to maximize the NPV of these assets. This paper describes the work flow, starting with a fracture model, then coupled with a production model using numerical simulation to obtain a calibrated model, and, finally, a detailed economic and sensitivity analysis to obtain the well spacing and completions design that will yield the highest NPV of the field.
When well spacing systems (interlateral spacing) for various unconventional basins were developed, commodity pricing was much higher and completions job sizes were smaller than they are today. The majority of wells were completed with less than 1,300 lbm/ft of proppant. As operators increased job sizes and seized the benefit of higher production performance, discussions regarding increasing well spacing also took place.
After 2014, operators sought ways to stay economical at lower commodity pricing and began to consider feasible ways to reduce operational costs, improve well productivity, raise NPV/acre, automate processes and work flows, use machine learning (ML) to improve predictability, and optimize workforce efficiency. Optimal well spacing for any unconventional well depends on many factors, including gas price, capital and operating expense, acreage position and inventory, completions design, production performance, and lateral length. There is no one-size-fits-all well spacing for various completions designs. Performing a full analysis, therefore, is crucial to finding the optimal well spacing for each area, either analytically or numerically.
Factors such as geology, engineering, and economic analysis must be considered. For instance, optimal well spacing and completions design for a geologically noisy and complex reservoir will be invalid in a discreet and quiet area. Similarly, if well spacing and completions design were developed for a high-commodity-pricing environment, performing the same work flow and evaluation at lower commodity pricing would yield an increase in well spacing. The work flow described in the complete paper addresses all these factors and uses modeling, numerical simulation, ML, and linear programming to optimize NPV.
Feder, Judy (JPT Technology Editor)
This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 190277, “Mechanistic Study for the Applicability of CO2-EOR in Unconventional Liquids-Rich Reservoirs,” by Dheiaa Alfarge, SPE, Iraqi Ministry of Oil and Missouri University of Science and Technology, and Mingzhen Wei and Baojun Bai, Missouri University of Science and Technology, prepared for the 2018 SPE Improved Oil Recovery Conference, Tulsa, 14–18 April. The paper has not been peer reviewed.
Improved oil recovery (IOR) methods for shale-oil reservoirs are considered relatively new concepts compared with IOR for conventional oil reservoirs. Different IOR methods—including CO2, surfactant, natural gas, and water injection—have been investigated for unconventional reservoirs using laboratory experiments, numerical simulation studies, and limited pilot tests. For a variety of reasons, CO2 injection is the most-investigated option. In this paper, numerical simulation methods of compositional models were incorporated with logarithmically spaced, locally refined, and dual-permeability reservoir models and local grid refinement (LGR) of hydraulic-fracture conditions to investigate the feasibility of CO2 injection in shale oil reservoirs.
Advancements in horizontal drilling and hydraulic fracturing enabled unconventional liquids-rich reservoirs (ULRs), such as shale and source-rock formations and very tight reservoirs, to change the oil industry. ULRs are characterized by pore throats of micro- to nanomillimeters and an ultralow permeability. Although different studies re-ported that these ULRs contain billions of recoverable oil barrels in place, it is estimated that less than 7% of the original oil in place can be recovered during the primary depletion stage. Production sustainability is the main problem behind the low oil recovery in these unconventional reservoirs. Oil wells in ULRs typically start with a high production rate, but show a steep decline rate in the first 3–5 years of production life because of the rapid depletion in the natural fractures combined with a slow recharge from the rock matrix.
The logical steps of academic research such as experimental investigation, simulation studies, and pilot tests for examining the applicability of different unconventional IOR methods have just begun in the past decade. Applying one of the feasible IOR methods in most oil and gas reservoirs should be mandatory to increase the oil-recovery factor. However, the mechanisms of IOR methods in unconventional reservoirs are not necessarily the same as those in conventional reservoirs. The primary characteristics of unconventional reservoirs that might impair performing IOR operations are low porosity and ultralow permeability. As a result, finding IOR methods that are insensitive to these very small pore throats is a priority.
Carpenter, Chris (JPT Technology Editor)
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 190323, “Gas Injection for EOR in Organic-Rich Shale: Part I—Operational Philosophy,” and paper URTeC 2903026, “Gas Injection for EOR in Organic-Rich Shale: Part II—Mechanisms of Recovery,” by Francisco D. Tovar, SPE, Maria A. Barrufet, SPE, and David S. Schechter, SPE, Texas A&M University. Paper SPE 190323 was prepared for the 2018 SPE Improved Oil Recovery Conference, Tulsa, 14–18 April; paper URTeC 2903026 was prepared for the 2018 Unconventional Resources Technology Conference, Houston, 23–25 July. The papers have not been peer reviewed.
This synopsis contains elements of two papers. In the first, the authors describe their comprehensive experimental evaluation of gas injection for enhanced oil recovery (EOR) in organic-rich shale. Experiments in preserved core demonstrated the potential of CO2 to extract the naturally occurring oil in organic-rich shale reservoirs, whereas tests in resaturated core plugs were used to compute accurate recovery factors, and evaluate the effect of soak time, operating pressure, and the relevance of slimtube minimum miscibility pressure (MMP) on recovery. In the second paper, the authors focus on the effect of fluid transport in organic-rich shale on recovery mechanisms under gas injection, and provide the rationale behind the proposed operational philosophy.
Part I—Operational Philosophy
Background. The notion that industry experience in the implementation of gas-injection methods in conventional reservoirs can be applied to unconventional reservoirs is a problematic one. A lack of understanding exists regarding the effect of contrast in mechanisms at the pore scale on the implementation of a gas-injection process. Experimental research so far, though encouraging, suffers from serious limitations. Also, there still is a significant lack of understanding of the mechanisms of recovery under gas injection for enhanced recovery in organic-rich shales.
In this paper, the authors base their investigation on experimental observations made in core plugs extracted from the reservoir interval, and show the development of a coreholder configuration that enables the physical simulation of the injection of gas through a hydraulic fracture in the laboratory. Then, this configuration is used to perform coreflooding experiments at the pressure and temperature conditions seen in the reservoir. Detailed descriptions and results of the experimental work are provided in the complete paper.
Summary. The authors begin by demonstrating that direct gas injection through an organic-rich shale matrix is not possible in a reasonable time frame. That discovery triggered the construction of specialized equipment and the development of a novel injection technique that resembles that of injection through hydraulic fractures. Using that technique,
nine experiments injecting CO2 in preserved organic-rich shale cores were performed. Only three of those experiments recovered a significant volume of oil, and the recovery factor was estimated to be between 18 and 62% of the initial crude-oil volume in the cores.
This demonstrated CO2 can be used to extract the naturally occurring oil in core plugs with extremely low permeability, where gas cannot be injected directly. Also, by coupling the coreflooding equipment developed in-house to a computed-tomography (CT)-scanner, this technology proved able to track the changes in density resulting from the mass exchange between CO2 and crude oil.
Carpenter, Chris (JPT Technology Editor)
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper URTeC 2890074, “Laboratory Investigation of EOR Techniques for Organic-Rich Shales in the Permian Basin,” by Shunhua Liu, SPE, Vinay Sahni, and Jiasen Tan, SPE, Occidental Oil and Gas, and Derek Beckett and Tuan Vo, CoreLab, prepared for presentation at the 2018 Unconventional Resources Technology Conference, Houston, 23–25 July. The paper has not been peer reviewed.
Commercial production from light oil, organic-rich shales in the Permian Basin has largely come from a solution-gas-drive recovery mechanism as a result of horizontal drilling and multistage hydraulic fracturing. These onshore, capital-intensive developments feature steep production declines and low expected ultimate recoveries. This paper involved laboratory experiments introducing miscible gases into core samples to investigate enhanced oil recovery (EOR) mechanisms for Permian Basin shales to provide information to design field tests for a huff ’n’ puff (HNP) recovery process.
The average recovery factor in the un-conventional resources is typically less than 10% with very steep decline rates, indicating enormous potential for EOR. In recent years, research efforts and field pilots of unconventional EOR have targeted the Bakken and Eagle Ford shales. Most focused on miscible-gas (either CO2 or produced gas) injection, while others investigated water-based chemical injection. This paper provides EOR fluid and core analyses in Permian Basin organic-rich shale, an unconventional hydrocarbon growth play with different geological, rock, and fluid properties from those of the Bakken and Eagle Ford plays. The experimental results from this paper were used to calibrate the operator’s unconventional EOR reservoir simulation and field pilot design.
Fluid properties such as equation-of-state (EOS) and minimum miscibility pressure (MMP) are extremely important because they are the fundamental designing parameters for any gas EOR project. In this study, oil and gas samples were collected in the well from perforations inside the Wolfcamp formation of the Permian organic-rich shale. A gas/oil ratio (GOR) of 1230 scf/bbl was chosen to recombine the separator oil and gas on the basis of observed solution GOR values before any increase caused by the flowing bottomhole pressure falling below the bubblepoint pressure.
The pressure/volume/ temperature (PVT) laboratory-testing program consisted of a constant-composition-expansion (CCE) test and a series of swelling tests with CO2. Using the recombined reservoir fluid (with a GOR of 1230 scf/bbl), a CCE test was performed at the reservoir temperature of 162°F to measure the bubblepoint pressure, single-phase oil density, and compressibility. The swelling test results were performed to tune an EOS to be used to calculate oil properties with increasing CO2 concentration during a CO2 flood.
An EOS model was generated to match the CCE data, viscosity data, and CO2 swelling-test data. To use this EOS for CO2 reservoir simulation, the reported system components were grouped, but the CO2 component was left ungrouped. Otherwise, it would be grouped with component C2. The minor component N2 was grouped with C1. All C4s and C5 were grouped together, as were the C6s. The C7+ components were divided into three pseudocomponents.
Feder, Judy (JPT Technology Editor)
This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 191272, “Optimization Under Uncertainty for Reliable Unconventional Play Evaluation: A Case Study in Vaca Muerta Shale Gas Blocks, Argentina,” by Reza Mehranfar, Leonardo Marquez, SPE, Raphael Altman, SPE, Hassan Kolivand, Rodrigo Orantes, and Oswaldo Espinola, SPE, Schlumberger, prepared for the 2018 SPE Trinidad and Tobago Section Energy Resources Conference, Port of Spain, Trinidad and Tobago, 25–26 June. The paper has not been peer reviewed.
Asset evaluation embraces the integrated analysis of a hydrocarbon-bearing field, and the identification of suitable strategies for its future development, to add incremental value for the investor(s). Optimizing the evaluation process under uncertainty is important particularly in unconventional reservoirs, which hold large quantities of oil and gas resources but also exhibit large degrees of uncertainty. This paper describes a comprehensive optimization-under-uncertainty work flow that combines a simulation-based approach with semiautomatic work flows and high-speed computers to facilitate the process of decision-making for investors, using data from the Vaca Muerta Formation in Argentina as an example.
An asset evaluation depends on many input parameters, some of which are partially known, partly analyzed, or un-available. Yet a go/no-go decision must be made, frequently within short time frames, because of competition, changing conditions, or the chance to take advantage of the business opportunity. The decision usually is based on preliminary assumptions and a conscientious analysis of several possible outcomes. Identifying the suitable future development strategies and the estimation of un-certainties in the input variables is crucial. Knowing the possible variability of the input and how the field mechanisms function will allow probabilistic forecasts of parameters such as production, costs, prices, and revenues. In the case of unconventional reservoirs with very limited history and high development costs, optimization under uncertainty plays a significant role in maximizing profit, reducing investment risk, and facilitating the decision-making process.
The authors summarize the optimization-under-uncertainty work flow that was implemented for this study. The starting point is the development of a base-case, single-well, matched simulation model, and, where available, an extended model with history-matched offset wells. This is followed by sensitivity analysis to identify the most-influential parameters; uncertainty analysis and proxy modeling for developing probabilistic forecasting profiles (type wells); and optimization of key parameters under existing uncertainty, which is the final objective of this paper. The model and the uncertainty and optimization work flows have been built in the Petrel platform and all the simulations have been executed in the Eclipse compositional reservoir simulator, using published Vaca Muerta data.
Base-Case Single-Gas-Well Model
A critical element of any single-well simulation study is developing a base-case simulation model that correctly captures all the fluid-flow mechanisms that take place during the life of the well, while running as fast as possible. The characteristics of the base-case single-well simulation model and the mechanisms that were considered are discussed briefly, because the main objective of this work was to focus on probabilistic forecasting and optimization.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 191457, “Coupling Geomechanical Effects and Reservoir Dynamics for Modeling Rejuvenation in Unconventional Plays,” by R. Dutta, SPE, Drilling Info; R. Pinto, Sciences Po University; and J.C. Flores, S.M. Stolyarov, SPE, and J. Yang, Baker Hughes, a GE Company, prepared for the 2018 SPE International Hydraulic Fracturing Technology Conference and Exhibition, Muscat, Oman, 16–18 October. The paper has not been peer reviewed.
An integrated understanding of geomechanical effects, fracture propagation, and reservoir dynamics is critical in the efficient and cost-effective application of rejuvenation technologies for unconventional plays. While various reservoir models depicting the hydraulic-fracturing process are available in the industry, many tend to be simplified or do not capture the numerous parameters that affect both the initial and restimulation processes. This work takes a further step toward building a more-realistic picture of fracturing in unconventional plays.
A common assumption in reservoir simulation is that the proppant-fluid mixture is present in the hydraulic fracture before flowback and production. The quantity of water assumed to be present in the hydraulic fractures is a conjecture and is calibrated generally with production-logging tools. These assumptions may skew the results of hydrocarbon recovery.
A method of incorporating geomechanical aspects of fracturing into the model involves the concept of pressure-dependent permeability variation in natural fractures that results in formation of pressure-dependent stimulated reservoir volume (SRV). Hysteretic permeability models employed in numerical modeling can offer a description of the SRV and also can be used in addressing longer-term geomechanical effects in a practical manner. While this concept has matured in the context of modeling hydraulic fracturing in reservoir simulation, it is being newly applied in modeling refracturing treatments.
Because the importance of capillary effect in low-permeability formations is recognized, the authors also incorporate capillary pressure in their model. In addition to pressure-dependent permeability variation, results explain how capillarity is significant in understanding fluid migration, the trapping of fluid in the matrix, and, consequently, restimulation.
The main challenge in selecting good candidate wells for this study was in finding wells that targeted the same formation, used varying refracturing technologies, and had sufficient data to build a reservoir-simulation model with input for the reservoir properties.
After studying a large number of wells, the authors focused on two horizontal gas wells producing from the Barnett Shale. One well was identified to be refractured with a selective zone-treatment method, while the other used a method of fluid diversion. The wells are located approximately 3 miles from each other and approximately 1,600 ft from neighboring wells. These wells have differing production signatures, but this is not indicative of a difference in the performance of two technologies. Understanding the difference in performance may be key to planning a successful refracturing operation.
Carpenter, Chris (JPT Technology Editor)
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 190860, “Evaluating the Impact of Lateral Landing, Wellbore Trajectory, and Hydraulic Fractures To Determine Unconventional Reservoir Productivity,” by Piyush Pankaj, Priyavrat Shukla, SPE, Ge Yuan, and Xu Zhang, Schlumberger, prepared for the 2018 SPE Europec featured at the 80th EAGE Annual Conference and Exhibition, Copenhagen, Denmark, 11–14 June. The paper has not been peer reviewed.
Inconsistent production performance from wells completed in similar pay zones has been observed when shale formations are exploited through horizontal wells. This paper demonstrates the need to couple the wellbore model to the reservoir-simulation and hydraulic-fracturing model in shale formations to optimize well landing, trajectory profile, and long-term productivity. The authors aim to demonstrate and deconvolute the well-trajectory plan with an integrated parametric study that helps to improve well productivity.
To plan a well profile, two critical pieces of information are required: Lateral landing depth, and well trajectory originating from the landing depth. To reach the targeted landing depth, the well trajectory undergoes a certain buildup of curvature deviating from the vertical section and, eventually, when the landing depth is reached, the designed trajectory profile is maintained and continued for the horizontal wellbore. The authors evaluated well trajectory and well productivity on the basis of the effect of the following parameters to guide well-trajectory planning:
Geological Review of the Model
A 3D earth model in the Permian Basin for the Wolfcamp shale was used to develop a work flow for determining well landing and well trajectory. The Wolfcamp shale covers most of the Midland Basin and ranges in thickness from 200 ft in the north of the basin to 2,600 ft in the south. The entire play is dominated by a fine-grained, naturally fractured source rock. The depths range from 5,500 to 11,000 ft. The Wolfcamp is slightly overpressured, with the pressure gradient varying between 0.55 and 0.70 psi/ft. In the past few years, the Wolfcamp has become one of the most profitable and exploited unconventional plays in the US. Almost all of the operators are collecting a significant share of their well inventory, which yields over 1,000 BOPD routinely in initial-production rate. The production declines within a short period (6 to 9 months). The recovery factors remain in the single digits for most operators. The Wolfcamp, Spraberry, and Bone Spring formations are the most prolific in the basin.
Defining Landing Depth
The proposed solution considers applying an end-to-end cycle of a streamlined work flow that starts with sampling engineered landing location points in the geomodel defined by the user on the basis of reservoir-quality (RQ) cutoffs. The first step is building the geological model around the sweet spot. This geological model contains petrophysical and mechanical properties of the rock along the depth of the targeted interval.
Feder, Judy (JPT Technology Editor)
This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 194180, “An Analytical Model for Double-Shouldered Connection Strengths,” by Grant Pettit, SPE, Bureau Veritas, prepared for the 2019 SPE/IADC International Drilling Conference and Exhibition, The Hague, 5–7 March. The paper has not been peer reviewed.
The extra torsional capacity and clearance available when proprietary, double-shouldered connections are used instead of public-domain, single-shouldered connections enables well teams to drill farther, faster, and with less damage to the drillstring. No analytical model as yet can accurately calculate the tensile and torsional capacities of double-shouldered connections. This paper presents a set of equations that extends the approach of the original single-shouldered equation to account for a second shoulder, and helps to understand connection strengths better.
Single-shouldered connections such as API drillstem connections use a straightforward analytical equation to determine the capacities of any connection. This equation may not be perfect—it relies on linear assumptions that are probably not descriptive of the connection loading—but several decades of use have made the industry confident in the equation’s strengths and aware of its shortcomings.
Because no analytical model is available to calculate these capacities in proprietary double-shouldered connections, the connection designer or manufacturer typically creates an empirical formula that is calibrated through laboratory and field testing. Although this approach is acceptable, it hinders tool designers who need something different from what is offered on the public market. It is rarely cost-efficient to perform laboratory and field testing for a one-off connection design.
The complete paper states that what is needed is to extend the approach of the original single-shouldered equation to account for a second shoulder. According to the author, although the mathematical complexity increases, the assumptions are the same, lending confidence to the extended equation.
The complete paper presents the derivation of a set of equations that use the same basic assumptions as the original single-shouldered connection torque equations, in the same way as do the API equations, to advance understanding of the torsion and tension capacities for double-shouldered equations. The implicit assumptions present in the original equations are discussed; then the same ideas are applied to a double-shouldered connection. The full set of new equations is developed and described, including tips on their practical use gained from tool-design experience.
Current Single-Shouldered Connection Model
Connection strength formulas contained in API RP 7G Appendix A are the basis for the connection torsion and tensile capacities in that standard and in many other oilfield references. These equations and resulting numbers have been in place for decades, creating industry confidence in their usefulness and awareness of their weaknesses.
For an industry that is just over 50 years old, liquefied natural gas (LNG) has matured rapidly and is playing an ever-growing role in the global energy system. From the start of international trade in the 1960s, demand reached 50 MTPA in 1990, then 100 MTPA in 2000, and 240 MTPA in 2012, according to the Center for Liquefied Natural Gas (CLNG).
LNG has become the world’s fastest-growing gas supply source and is now part of an upheaval in the global energy market. Trade has quadrupled over the past two decades and is set to double over the next two. Consumption and share growth have set new records for three consecutive years. New markets for both demand and supply are developing rapidly. Liquefaction facilities and vessel fleets are expanding. Innovative and collaborative trade models are emerging. And, new technologies are being pursued to enhance LNG’s flexibility and competitiveness with other primary energy forms (Fig. 1).
LNG is not without risks. These range from project economics, fuel competition, and politics on a local, national, and global scale to partner priorities, marketing, and contractual arrangements. Given the growing world population, the rising demand for more energy, and the need to mitigate climate change, the consensus is that the long-term outlook is bullish. But, says industry technical advisor DNV GL, the sector stands at a crossroads, and the industry must adopt new thinking to address current and future needs of buyers, sellers, and consumers both globally and regionally.
What’s Behind the Growth?
Three primary factors are behind the current wave of growth in LNG, according to the Shell 2019 LNG Outlook: the energy challenge, new countries choosing LNG for various reasons, and supply and demand.
The world’s current population of 7.6 billion people is expected to reach 8.6 billion in 2030, 9.7 billion in 2050, and 11.2 billion in 2100, according to United Nations estimates. Of today’s population, nearly one person in three either has no electricity or struggles with unreliable supplies. As the population grows and more people seek to improve their quality of life, Shell projects that, by 2070, the world is likely to use at least 50% more energy than it does today.
Natural gas has an advantage over other energy sources to meet current and future energy challenges because it is abundant, relatively clean, and cost-effective to produce. For these reasons and the role it can play in lowering carbon dioxide emissions, it is being touted as a bridge fuel to renewables as well as a foundational source of energy.
Carpenter, Chris (JPT Technology Editor)
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper IPTC 19311, “Data-Mining Approaches for Casing-Failure Prediction and Prevention,” by Christine Noshi, SPE, Samuel Noynaert, SPE, and Jerome Schubert, SPE, Texas A&M University, prepared for the 2019 International Petroleum Technology Conference, Beijing, 26–28 March. The paper has not been peer reviewed. Copyright 2019 International Petroleum Technology Conference. Reproduced by permission.
Recent casing failures in the Granite Wash play in the western Anadarko Basin have sparked deep concerns for operators in North Texas and Oklahoma. Hydrostatic tests made in the field show that current API standards do not assure adequate joint and bursting strength to meet deep-well requirements. This paper is part of an ongoing effort to minimize the likelihood of failure using data-mining and machine-learning algorithms.
Casing failure has long presented a challenge to the industry. The combined effects of design, dynamic borehole conditions, metallurgy, and handling have been challenging to quantify and predict accurately. Additionally, most casing-string challenges have been handled reactively instead of proactively; the total number of failures have been underreported and overlooked.
The authors focus on the effects of poor cement as a primary factor; this translates into the absence of cement in a case study presented in the complete paper. Additional factors are the pumping of corrosive acids and poor standardized casing design that does not account for varied formations along with cyclical temperatures.
Casing with partial cementation and sheaths with voids can contribute to excessive buckling-related collapse and tensile failures. Large pressure loadings, along with significant change in temperature, contribute to significant stresses in the intercasing annuli. In fragmentally cemented casing, tensile loading can show a great discrepancy between compression and high tension, with instances of failures in both the outer and inner strings. Additionally, cement thickening by downward flow could lack uniformity and could be prone to channeling. Air entrapment might occur, establishing bridges that hinder the process. Some authors in the literature related cementing failures with hole enlargements and washouts in long cement depths. The lack of cement support in those significant intervals exposed the casing to movement during drillpipe rotation, which triggered wear and ultimate buckling.
The data were descriptively visualized using methods such as box plots, mosaic plots, and trellis charts, while predictive techniques included artificial neural networks (ANNs) and boosted-ensemble trees. A statistical software package was used along with Python coding to implement the models and choose the most-significant factors contributing to failure. Data-preprocessing techniques were implemented. The process began with data cleaning to account for missing data, re-move the bias incurred by noise, and remove outliers. For missing values, multivariate normal imputation on the basis of all samples belonging to the same class was used. Then, several parameters from different databases were integrated. Data transformation involved standardizing the data by the subtraction of the mean value and the subsequent division by standard deviation from each feature. Categorical variables were converted to numerical values because models such as neural nets, regression, and nearest-neighbor involve only numeric inputs. The compiled data set comprised 78 wells. Caution should be taken when assessing its statistical significance.