No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Abstract
Since production curtailment for other than engineering reasons is graduallydisappearing, and more and more wells are now producing at capacity and showingdeclining production rates, it was considered timely to present a brief reviewof the development of decline-curve analysis during the past three or fourdecades.
Several of the commoner types of decline curves were discussed in detail andthe mathematical relationships between production rate, time, cumulativeproduction and decline percentage for each case were studied.
The well-known loss-ratio method was found to be an extremely valuable toolfor statistical analysis and extrapolation of various types of curves. Atentative classification of decline curves, based on their loss ratios, wassuggested. Some new graphical methods were introduced to facilitate estimationof the future life and the future production of producing properties wherecurves are plotted on semilogarithmic paper.
To facilitate graphical extrapolation of hyperbolic-type decline curves, aseries of decline charts was proposed, which will make straight-lineextrapolation of both rate-time and rate-cumulative curves possible.
Introduction
During the period of severe production curtailment, which is now behind us,production-decline curves lost most of their usefulness and popularity inprorated areas because the production rates of all wells, except those in thestripper class, were constant or almost constant.
While production-decline curves were thus losing in importance forestimating reserves, an increasing reservoir consciousness and a betterunderstanding of reservoir performance developed among petroleum engineers.This fact, together with intelligent interpretation and use of electric logs,core-analysis data, bottom-hole pressure behavior and physical characteristicsof reservoir fluids, eliminated a considerable part of the guesswork inprevious volumetric methods and put reserve estimates, based on this method, ona sound scientific basis. At the same time, a number of ingenious substituteswere developed for the regular production-decline curve, which made it possibleto obtain an independent check on volumetric estimates in appraisal work, eventhough the production rates were constant.
With the now steadily increasing demand for oil to supply the hugerequirements of this global war, proration for reasons other than prevention ofunderground waste is gradually disappearing. More and more wells are, or willbe, producing at capacity or at their optimum rates, as determined by soundengineering practice.
T.P. 1758
... There are lots of empirical methods that can be used with unconventional reservoirs, these methods were discussed and compared with their characteristics and limitations which can be found in recent review papers [3,22,25,31,33.34] One of the most used decline methods is Arps' method [4]. However, researchers suggested that Arps' decline curves are not suitable for tight/shalegas wells [36]. ...
... To address this urgent issue, this study focuses on comparing and evaluating three empirical decline methods commonly used in the industry from different empirical methods [2,4,5,6,7,8,12,13,14,23,29,30,32,37,38]. These three methods include Arps' decline method [4], Stretched Exponential Production Decline (SEPD) method [29,30], and Duong's method [6,7]. ...
... To address this urgent issue, this study focuses on comparing and evaluating three empirical decline methods commonly used in the industry from different empirical methods [2,4,5,6,7,8,12,13,14,23,29,30,32,37,38]. These three methods include Arps' decline method [4], Stretched Exponential Production Decline (SEPD) method [29,30], and Duong's method [6,7]. By examining the principles and characteristics of these methods, we aim to identify the most reliable approach for accurate production and EUR forecasting. ...
... Over the years, production decline models for shale gas wells have emerged one after another and been used widely. Arps (1945) proposed the approach for analyzing production decline with empirical formulas, and developed exponential decline, hyperbolic decline, and harmonic decline models according to the production data of actual gas wells, which enable the qualification of production decline of gas wells. Later, other scholars put forward more production decline curve analysis methods. ...
... Arps (Arps, 1945) q qi [1+bDi(t−ti)] 1/b q i , D i , and t i are the initial production, decline rate, and time, respectively; b is the decline index Boundary-dominated flow period, or the whole production period depending on the production data The production period is limited when predicting the cumulative shale gas production. Otherwise, the production is overestimated PLE (Ilk et al., 2008) ...
To effectively develop the shale gas in the southern Sichuan Basin, it is essential to accurately predict and evaluate the single-well production and estimated ultimate recovery (EUR). Empirical production decline analysis is most widely used in predicting EUR, since it is simple and can quickly predict the gas well production. However, this method has some disadvantages, such as many parameters of the model, difficulty in fitting and large deviation. This paper presents an efficient process of EUR prediction for gas wells based on production decline models. Application of nine empirical production decline models in more than 200 shale gas wells in the Changning block of the Sichuan Basin was systemically analyzed. According to the diagnosis of flow regime, it was determined that all models are applicable in the prediction of production and EUR in this area, with the fitting degree higher than 80% for gas wells producing for more than 12 months. Based on the fitting and prediction results, the parameter distribution charts of the nine production decline models with initial parameters constrained were plotted for shale gas wells, which greatly improved the prediction accuracy and efficiency. Coupled with the probability method, the EUR was evaluated and predicted effectively, and the average EUR of more than 200 shale gas wells in the Changning block is 1.21 × 10 ⁸ m ³ . The EUR of Well CNH1 predicted by the proposed process and charts is believed reliable. The study results provide meaningful guidance for the efficient prediction of gas well production and EUR in the Changning block.
... Predicting the production rates of oil, gas, and water from a hydrocarbon reservoir poses a time series forecasting challenge. One common approach to predicting production rates from conventional reservoirs with high porosity and high permeability, such as limestone or sandstone reservoirs, is to use a decline curve model developed by Arps (1945) The decline curve analysis (DCA) method involves developing a mathematical equation through regression of historical data, mainly the flow production rate versus time (Panja & Wood, 2022). This equation is then used to predict future production rates and cumulative production. ...
... Arp's empirical model, Arps Hyperbolic Decline (AHD) in this study, is the basis for many DCA applications in the oil and gas industry today (Arps, 1945) based on solutions to the equation ...
The traditional Arp's decline model failed to predict production from many oil and gas reservoirs due to some inherent assumptions like boundary-dominated flow contrary to long transient flow. Fundamentally, this is a time series curve fitting and forecasting problem. Advanced machine learning (ML) algorithms can be used to capture the unusual trend in hydrocarbon production decline. The objective of this study is to develop various ML algorithms such as Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) in forecasting future production performance and estimating ultimate recovery (EUR).
Decline curve analysis (DCA) is a straightforward and rapid way to estimate future production simply by suitable curve fitting. However, the traditional Arp's method overestimates the production from many reservoirs, resulting in new empirical methods such as Power Law Exponential Analysis (PLE by Ilk, 2008), Logistic Growth Analysis (LGA by Clark 2011), and Duong Method (DM by Duong 2011). The outcomes of these recent models also depend on the quality of the data and the reservoir flow regimes. The machine learning algorithm is applied to overcome the drawbacks and limitations of the empirical decline curve models. Machine learning algorithms such as RNN, LSTM, and GRU are compared. The first 80% of time-series data is used for training the models and the last 20% is used for testing. The trained models are employed to forecast future rates and to calculate EUR. The value of NSE close to unity suggests good model performance. A normalized Nash-Sutcliffe model efficiency coefficient (NNSE) and Normalized Root Mean Squared Error (NRMSE) are selected for assessing the efficacy of different models. The LSTM models have several unique advantages over typical supervised machine learning algorithms. The models are flexible in handling multiple inputs in time series. The ML models developed in this work can be coupled with an economic model considering the future oil price and operational costs.
Machine learning is a research area quickly growing across several industries providing valuable insights. Machine algorithm for time series forecasting in the oil and gas industry has not been comprehensively explored. Results from this work will provide the literature with another application perspective with strong opportunities in production data analysis.
... Data is from well TM-20 in the "X" field. Data Table 3 Arps' model for decline curve analysis (Arps, 1945) is obtained in the form of flow data produced for approximately ten years. Flow forecasting is performed until a feed flow of 20 bbl/day is reached. ...
... The linear relationship between cumulative oil produced and pressure does not exist in most actual conditions. Pressures are usually not proportional to the amount of remaining oil, but they appear to decrease at a gradual and slow rate as the amount (Amega Yasutra, et al.) DOI of remaining oil decreases Arps., (1945). As shown in the tail of the production profile on Figure 6, this condition results in a slower decline in the production rate at the end of the well's life, which is still successfully modeled by Arps' decline model. ...
There are many methods for predicting the production performance of oil wells, using the simplest method by looking at the declining trend of production, such as Decline Curve Analysis (DCA), Material Balanced, or using reservoir simulations. Each of these methods has its advantages and disadvantages. The DCA method, the Arps method, is often used in production forecast analysis to predict production performance and estimate remaining reserves. However, the limitation of this method is that if the production system changes, the trend of decline will also change. At the same time, the application in the field of taking the trend of decreasing production does not pay attention to changes in the production system. This study aims to see that changes in the well production system will affect the downward trend of well production, estimated ultimate recovery (EUR) value, and well lifetime. To see the effect of these changes, the initial data tested used the results of reservoir simulations and field data. From the evaluation results, it is found that if the production system changes during the production time, for example, from changing natural flow using artificial lifting assistance, the trend taken from the production profile will follow the behaviour of the reservoir if the trend is taken in the last system from the production profile, not from the start of production. If the downward trend is taken without regard to the changing system, then the prediction results will not be appropriate
... They came to the conclusion that the composite model developed by them can give some idea of the mechanism of fluid movement in gas reservoirs and help analyze the decrease in productivity of such wells and reservoirs. A significant part of the models used to predict the gas well performance of the dip curve analysis models is based on the Arps equation models [12]. ...
... The Arps prediction model is a common empirical model used for forecasting oil and gas well production [24]. It is suitable for oil and gas wells with exponential decline characteristics. ...
With the development of artificial intelligence technology, machine learning-based production forecasting models can achieve the rapid prediction and analysis of production. However, these models need to be built on a large dataset, and having only a small amount of data may result in a decrease in prediction accuracy. Therefore, this paper proposes a transfer learning prediction method based on the hierarchical interpolation model. It uses data from over 2000 shale gas wells in 22 blocks of the Marcellus Shale formation in Pennsylvania to train the transfer learning model. The knowledge obtained from blocks with sufficient sample data is transferred and applied to adjacent blocks with limited sample data. Compared to classical production decline models and mainstream time-series prediction models, the proposed method can achieve an accurate production decline trend prediction in blocks with limited sample data, providing new ideas and methods for studying the declining production trends in shale gas.
... The first part corresponds to the initial stages of hydration when the cement matrix is in a saturated state, while the second part exhibits an almost linear decrease in IRH with the degree of hydration, indicating the consummation of combined water with the progression of hydration reactions. To predict the IRH variation with f, we employ the classical analysis of decline curves (Arps, 1945). ...
Introduction: This study focuses on autogenous shrinkage in cement pastes and presents a novel calculation method considering variations in internal relative humidity (IRH). IRH significantly influences autogenous shrinkage, and its evolution is modeled based on decline curves. The proposed method accurately evaluates autogenous shrinkage and aligns well with experimental data. Additionally, we calculate capillary depression and meniscus radius using the Laplace–Kelvin equation. Methods: To address early autogenous shrinkage in construction materials, we developed our calculation method, emphasizing IRH variation. We analyzed decline curves to model IRH and validated our model using literature-based experimental data. Results: Our validated model for predicting IRH and autogenous shrinkage in Portland cement pastes, based on cement paste hydration degree, water-to-cement ratio (w/c), and the critical degree of hydration (αcr), closely aligns with experimental data and existing models.
... Comparison of the Arps Model. The Arps prediction model 30 is a common empirical model used for forecasting oil and gas well production. In the application process of this article, the parameter estimation of the Arps decline model is conducted using the method of linear fitting between cumulative production and production. ...
Forecasting oil production is crucially important in oilfield management. Currently, multifeature-based modeling methods are widely used, but such modeling methods are not universally applicable due to the different actual conditions of oilfields in different places. In this paper, a time series forecasting method based on an integrated learning model is proposed, which combines the advantages of linearity and nonlinearity and is only concerned with the internal characteristics of the production curve itself, without considering other factors. The method includes processing the production history data using singular spectrum analysis, training the autoregressive integrated moving average model and Prophet, training the wavelet neural network, and forecasting oil production. The method is validated using historical production data from the J oilfield in China from 2011 to 2021, and compared with single models, Arps model, and mainstream time series forecasting models. The results show that in the early prediction, the difference in prediction error between the integrated learning model and other models is not obvious, but in the late prediction, the integrated model still predicts stably and the other models compared with it will show more obvious fluctuations. Therefore, the model in this article can make stable and accurate predictions.
... The graph of production rate versus time was plotted on the semi-log graph(see fig.2.). A straight line relationship on the semi-log graph shows that the data undergoes the empirical model of Arps, J.J. 6 ...
The lifestyle of the 20 th century man has been influenced more by oil and gas than any other natural resource and indications are that oil and gas reserve will increase in importance the remainder of this century. Based on the foregoing, many petroleum engineers spend a major part of their professional capabilities, along with new methods and techniques for improving these estimates.
... The graph of production rate versus time was plotted on the semi-log graph(see fig.2.). A straight line relationship on the semi-log graph shows that the data undergoes the empirical model of Arps, J.J. 6 ...
The lifestyle of the 20 th century man has been influenced more by oil and gas than any other natural resource and indications are that oil and gas reserve will increase in importance the remainder of this century. Based on the foregoing, many petroleum engineers spend a major part of their professional capabilities, along with new methods and techniques for improving these estimates.
... First the dimensionless time, t D for E1 was given by the formula (7) the reason being that the aquifer oil leg area A was given (9) Where U = water influx constant, rb/psi; W D (t D ) = the dimensionless water influx read from the Van Everdingen and Hurst water influx chart. ...
... Concept of DCA equation(Arps, 1945). ...
In Indonesia, for a half decade, the decrease of oil and gas production from 2016 is 4.23% and 3.53% respectively (ESDM, 2021). This production decrease has a domino effect on the investment loss. According to the International Trade Administration, investment in Indonesia’s oil and gas industry in 2019 reached around US$ 12 billion, which was decreasing from around US$ 16 billion in 2016. Such loss is a serious disaster, thus applying digital transformation such as machine learning to the most-used method, well stimulation, is immediately needed. Unfortunately, the implemented well stimulations nowadays are prone to short-lived effects due to the unreliable selection methods, as they do not have any integrated database. This research, as the pilot project, focuses on field data collected in West Indonesia from sandstone and carbonate lithologies, and the type of stimulation used is acidizing. This tool, OliFANT, defines the success of stimulation based on the productivity index before and after stimulation. The method uses geostatistical approaches and optimizing decline curve analysis for analysing and modelling spatially correlated data. The accuracy of the model is validated at a minimum of 75%, which shows its high reliability. It can also forecast the duration effect of the stimulation, additionally it provides the estimation of profit scenarios. The proposed machine learning model adopts an empirical working principle by utilizing reservoir parameters and test data of stimulation, which are inputted into a user-friendly interface after filling in a comprehensive database. In conclusion, the main benefits of using this tool are cutting evaluation time and achieving higher cost-efficiency. This software can be continuously improved by adding more data to widen the variety of the methods. Considering that each field has different types of properties, this tool is built to be adaptable to every reservoir condition. Over and above that, this tool can be implemented for other stimulated wells and be modified for other methods and operations, such as drilling and workover. In the future, it can be a one-stop solution for stimulation plan validation, where data-driven solutions pave the way for success.
... DCA, a prevalent empirical approach, relies on historical production data to anticipate future production outcomes [10,11]. This method assumes that the production decline of a well or a reservoir follows a certain mathematical function. ...
Shale gas has revolutionized the global energy supply, underscoring the importance of robust production forecasting for the effective management of well operations and gas field development. Nonetheless, the intricate and nonlinear relationship between gas production dynamics and physical constraints like shale formation properties and engineering parameters poses significant challenges. This investigation introduces a hybrid neural network model, GRU-MLP, to proficiently predict shale gas production. The GRU-MLP architecture can capture sequential dependencies within production data as well as the intricate nonlinear correlations between production and the governing constraints. The proposed model was evaluated employing production data extracted from two adjacent horizontal wells situated within the Marcellus Shale. The comparative analysis highlights the superior performance of the GRU-MLP model over the LSTM and GRU models in both short-term and long-term forecasting. Specifically, the GRU model’s mean absolute percentage error of 4.7% and root mean squared error of 120.03 are notably 66% and 80% larger than the GRU-MLP model’s performance in short-term forecasting. The accuracy and reliability of the GRU-MLP model make it a promising tool for shale gas production forecasting. By providing dependable production forecasts, the GRU-MLP model serves to enhance decision-making and optimize well operations.
... Production history fitting serves as the basis for forecasting production dynamics [15]. ARPS [16] proposed a systematic approach to the decline laws of oil and gas fields in the 1950s. Yu et al. [17] defined three new capacity reduction rates based on the instantaneous decline rate, derived the expression for capacity reduction under the three types of reductions in Arps' decline theory, and provided a clearer representation of the relationship among capacity, pressure, and production degree. ...
The development of shale gas reservoirs often involves the utilization of horizontal well segmental multi-stage fracturing techniques. However, these reservoirs face challenges, such as rapid initial wellhead pressure and production decline, leading to extended periods of low-pressure production. To address these issues and enhance the production during the low-pressure stage, pressurized mining is considered as an effective measure. Determining the appropriate pressurization target and method for the shale gas wells is of great practical significance for ensuring stable production in shale gas fields. This study takes into account the current development status of shale gas fields and proposes a three-stage pressurization process. The process involves primary supercharging at the center station of the block, secondary supercharging at the gas collecting station, and the introduction of a small booster device located behind the platform separator and in front of the outbound valve group. By incorporating a compressor, the wellhead pressure can be reduced to 0.4 MPa, resulting in a daily output of 12,000 to 14,000 cubic meters from the platform. Using a critical liquid-carrying model for shale gas horizontal wells, this study demonstrates that reducing the wellhead pressure decreases the critical flow of liquid, thereby facilitating the discharge of the accumulated fluid from the gas well. Additionally, the formation pressure of shale gas wells is estimated using the mass balance method. This study calculates the cumulative production of different IPR curves based on the formation pressure. It develops a dynamic production decline model for gas outlet wells and establishes a relationship between the pressure depletion of gas reservoirs and the cumulative gas production before and after pressurization of H10−2 and H10−3 wells. The final estimated ultimate recovery of two wells is calculated. In conclusion, the implementation of multi-stage pressurization, as proposed in this study, effectively enhances the production of, and holds practical significance for, stable development of shale gas fields.
... Besides such approaches, the traditional Arps method and decline curve analysis (DCA) have also been predominantly used in predicting gas production. 161 To forecast productivity, the declining trend with a DCA is analyzed after employing production data obtained from gas wells. However, the recovery data obtained from shale wells for the first few years displays only a transient flow pattern. ...
Decline-curve models inherently assume that the bottomhole flowing pressure (BHP) is constant. This is a poor assumption for many unconventional wells. For this reason, the application of decline-curve models might lead to incorrect flow regime identification and estimated ultimate recovery (EUR). This work presents a novel technique that combines variable BHP conditions with decline-curve models and compares its results with traditional decline-curve analysis (DCA) for both synthetic and tight-oil wells.
Using superposition, we generate a synthetic rate example using the constant-pressure solution of the diffusivity equation for a slightly compressible fluid (decline-curve model) along with a BHP history. However, we validate the technique using bottomhole and initial reservoir pressures that contain errors. The algorithm consists of three sequential optimizations. In each optimization, the algorithm estimates (1) the decline-curve model parameters, (2) the BHP, and (3) the initial reservoir pressure. The result of the synthetic example leads to an accurate production history match and corrected estimates of the initial reservoir pressure and the BHP history. Finally, we compare the results of the technique with traditional DCA in terms of (a) the model parameters, (b) flow regime identification, (c) production history matches, and (d) EUR for tight-oil wells using three decline-curve models: 1D single-phase constant-pressure solution of the diffusivity equation for a slightly compressible fluid, logistic growth model, and Arps hyperbolic relation.
For the synthetic case, the algorithm estimates the model parameters and the true initial reservoir pressure within a 2% error. In addition, the method regenerates the true BHP history and provides an excellent production history match. The analysis of the tight-oil wells shows that the new approach clearly identifies the flow regimes present in the well, which can be difficult to detect using traditional DCA when the BHP varies. In contrast, the application of traditional DCA shows considerable errors in the estimation of the model’s parameters and a poor history match of the production data. Finally, this work shows that incorporating variable BHP into the decline-curve models leads to more accurate production history matches and EUR values compared to using only rate-time data.
This paper illustrates a workflow that incorporates variable BHP conditions for any decline-curve model. Moreover, the approach can handle errors in both the BHP and the initial reservoir pressure and provides corrected estimates of these variables. The technique is computationally fast and history matches and forecasts the production of unconventional wells more accurately than traditional DCA. The major contribution of this work is the remarkable simplicity yet robustness of our solution to variable-pressure DCA. Finally, we developed a web-based application to provide the readers with a hands-on experience of this new technique.
Within the Queensland energy industry, unconventional gas, such as Coal Seam Gas (CSG) production holds significant importance. This study employs the Continuous Wavelet Transform (CWT) to analyze variations in CSG production, aiming to comprehend its complex production dynamics. The primary objectives of this research are to identify and categorize gas rate production variations, quantify their attributes—such as frequencies, durations, and magnitudes—and explore their integration into forecasting models.
The study's results demonstrate successful identification and categorization of significant variations, specifically categorizing two types of variations impulse and step events. Through the CWT, the attributes of these events are quantified. Moreover, the obtained attributes are used to refine production forecasting via applying recreated production variations to conventionally smooth forecasting models such as type curves or history matching.
The broader impact of the study lies in its ability to provide deeper insights into CSG reservoir behavior. By uncovering patterns often obscured by traditional analysis methods, the CWT approach allows for better understanding of the operational dynamics. Quantifying the production variation into meaningful metrics results in more accurate and reliable production forecasts and more informed decision-making in optimizing production strategies.
Decline curve analysis (DCA) is a curve fitting procedure used for analyzing declining production rates and forecasting future performance of oil and gas wells. Certain types of standard curves are fit, based on the past production performance which are then extrapolated to predict the future well performance. It is the basic tool for estimating recoverable reserves and can be used when the production history is long enough that a decline trend can be identified.
One of the key technologies used in the process is an automated DCA methodology which can be automatically run on a machine without any human interaction and with minimum errors. The traditional DCA is derived by Arp's equation, where three types of decline curves have been identified: exponential, hyperbolic, and harmonic.
In this research work, authors have developed a python code that apply DCA methods to wells in an unbiased, systematic, and automated manner. This method contrasts with manual DCA, which was the widespread practice of the industry. As industry has entered digital transformation era, it is imperative that digital capabilities be coupled with conventional knowledge of DCA. This can be used as a handy tool to generate type curves for fitting decline curves and to further analyze using business intelligence (BI) tools, for a batch of oil wells under study.
Accurately forecasting oil and gas well production, especially in complex unconventional reservoirs, is vital. Leveraging advanced techniques like machine learning and deep learning is becoming more common due to ample historical data availability. While traditional methods work for conventional reservoirs, they struggle in unconventional scenarios. Modern machine and deep learning models excel in such challenges, offering insights while bypassing temporary disruptions or pressure issues. This study compares ten empirical production forecast models with state-of-the-art deep learning and time series models (ARIMA, LSTM, GRU) in the Bakken shale play of the Williston Basin. After thorough calibration using extensive data, model efficacy is assessed using R2-score and MSE. Results highlight well-specific performance, with no single model consistently outperforming across all wells. Notably, optimally adjusted ARIMA produced commendable results for many wells. This research aids reservoir engineers by simplifying production decline trend identification, reducing reliance on intricate decline curve analyses. It ushers in a streamlined and dependable paradigm for production forecasting.
Technology Assisted Review (TAR), which aims to reduce the effort required to screen collections of documents for relevance, is used to develop systematic reviews of medical evidence and identify documents that must be disclosed in response to legal proceedings. Stopping methods are algorithms which determine when to stop screening documents during the TAR process, helping to ensure that workload is minimised while still achieving a high level of recall. This paper proposes a novel stopping method based on point processes, which are statistical models that can be used to represent the occurrence of random events. The approach uses rate functions to model the occurrence of relevant documents in the ranking and compares four candidates, including one that has not previously been used for this purpose (hyperbolic). Evaluation is carried out using standard datasets (CLEF e-Health, TREC Total Recall, TREC Legal), and this work is the first to explore stopping method robustness by reporting performance on a range of rankings of varying effectiveness. Results show that the proposed method achieves the desired level of recall without requiring an excessive number of documents to be examined in the majority of cases and also compares well against multiple alternative approaches.
The influence of geological and engineering factors results in the complex production characteristics of shale gas wells. The productivity evaluation method is effective to analyze the production decline law and estimate the ultimate recovery in the shale gas reservoir. This paper reviews the production decline method, analytical method, numerical simulation method, and machine learning method. which analyzes the applicable conditions, basic principles, characteristics, and limitations of different methods. The research found that the production decline method can mainly account for the gas well production and pressure data by fitting type curve analysis. The analytical method is able to couple multiple transport mechanisms and quantify the impact of different mechanisms on shale gas well productivity. Numerical simulation builds multiple pore media in shale gas reservoirs and performs production dynamics as well as capacity prediction visually. Machine learning methods are a nascent approach that can efficiently use available production data from shale gas wells to predict productivity. Finally, the research discusses the future directions and challenges of shale gas well productivity evaluation methods.
Consistent Estimated Ultimate Recovery (EUR) forecasting has been a much-studied topic for the development and reserves estimation of unconventional resources. It is widely known that unconventional well EUR forecasts using Arp's Decline Curve Analysis (DCA) are not very accurate in early time. This is because Arp's empirical equations were originally defined for constant operating pressure and unconventional wells have significant early-time variations in operating pressure. Consequently, EURs based on early DCA are often overestimated. This paper develops the Pressure Normalized Rate Decline Curve Analysis (PNR DCA) method and demonstrates its ability to provide consistent EUR forecasts for Permian multiphase unconventional wells through multiple case studies and dynamic reservoir simulation.
The authors developed PNR DCA method by normalizing actual oil rate to a rate at constant flowing pressure (PNR) and then applying decline curve analysis using the PNR to forecast EUR of Permian unconventional wells with limited production history. This method was first developed for a single-phase gas reservoir, the Haynesville shale (Xie et al., 2012). Lacayo and Lee (2014) introduced this method to forecast wells in four different shale plays (Eagle Ford, Woodford, Marcellus, and Bakken) for more accurate early forecasts, mostly for single-phase oil or gas reservoirs. This paper describes how the method was modified for more complicated Permian reservoirs, in which the resource contains mobile water and many wells feature early gas/oil ratio (GOR) increases, resulting in multiphase flow in the reservoirs.
The method was validated with the data from over 100 wells with significant history (1-5 years) in the Bone Spring, Spraberry, and Wolfcamp formations in both the Midland and Delaware basins. The EUR estimated using the PNR DCA method was also found consistent with forecasts using history-matched dynamic reservoir simulation. The validation results show the PNR DCA method can provide relatively consistent EUR forecasts with limited production data, whereas Arp's DCA method can only forecast consistent EUR after flowing pressures stabilize, which can take a year to achieve. Results indicate that the PNR DCA method works for several sub-regional areas of the Permian Basin, including areas where water cut is as high as 80-90%.
The majority of production forecasting methods currently used are point forecasting methods developed in the setting of individual well forecasting. For an actual oilfield, instead of needing to predict individual production time series, one is faced with forecasting thousands of related time series and the uncertainty can be assessed. The objective of this work is to enable global modeling and probabilistic forecasting of a large number of related production time series using Deep Autoregressive Recurrent Neural Networks (DeepAR). The DeepAR model consists of three parts. First, the auxiliary data such as static classification covariates and dynamic covariates are encoded. Second, establish a forward model based on an autoregressive recurrent neural network. Third, the normal distribution is defined as the output distribution function. And the variance and mean are obtained by solving the maximum log-likelihood function using the gradient descent algorithm. We demonstrate how the application of DeepAR to forecasting can overcome many of the challenges(e.g. frequent well shut-in and opening, probabilistic prediction, classification prediction) that are faced by widely-used classical approaches to the problem. In this work, history fitting and prediction were performed on a dataset from more than 2000 tight gas reservoir wells in the Ordos Basin, China. The DeepAR and conventional methods were tested and compared based on the datasets. We show through extensive empirical evaluation on several real-world forecasting data sets accuracy improvements of around 30% compared to RNN-based networks. In the case of frequent well shut-ins and openings, the RNN-based network structure cannot capture the fast pressure response and extreme fluctuations, which eventually leads to high errors. In contrast, DeepAR is more stable to frequent or significant well variations, can learn different dynamic and static category features, generates calibrated probabilistic forecasts with high accuracy, and can learn complex patterns such as seasonality and uncertainty growth over time from the data. This study provides more general production forecasting and analysis of production dynamics methods from a big data perspective. Instead of performing costly well tests or shut-ins, reservoir engineers can extract valuable long-term reservoir performance information from predictions estimated by DeepAR trained on an extensive collection of related production time series data.
The scope of this work is to assess the flow beyond boundary-dominated flow, defining the existence of a third flow regime (after transient linear flow and boundary-dominated flow) that we call "exterior flow" for further use in decline curve analysis in unconventional reservoirs. Thus, exterior flow is defined as the linear flow of gas from the non-stimulated matrix feeding into the edges of the depleted stimulated reservoir volume (SRV), at late times, far away (predicted by Lee 2021; and anticipated by Marder et al. 2021). Sometimes referred as "flow beyond the tips" or post-SRV flow (Blasingame 2019).
Because of the existence of this third flow regime, the production curves of over-pressured shale gas reservoirs, such as the Haynesville formation, seem to not be fitted by a single hyperbolic model or any of the modern rate-time relations (Power law exponential, Stretched exponential, Duong, etc.). We believe that the over-pressured condition of the formation—close to the lithostatic gradient in the Haynesville Shale for example—yields such high pressure drawdowns that all flow regimes (transient linear flow, boundary-dominated flow, and exterior flow) occur sooner compared to other basins. Additionally, this paper shows that a new member of the logistic growth family of curves, called the "modified Logistic Growth Model" in its 2023 version (m-LGM 2023) solves the problem of curve-fitting production data from horizontal wells in the Haynesville Shale.
Furthermore, two novel diagnostic plots are presented for rate-time analysis to obtain the characteristic time of switching from boundary-dominated to exterior flow, which enables the prediction of additional volume to be produced under this flow regime. Finally, given that the current literature of the SPE-PRMS 2022 does not provide specific guidelines for the categorization of probable reserves (P2) in shale gas, we believe this work could signify a contribution for future reserves using this category for unconventional formations. The volume expected from exterior flow could be justified as probable reserves (P2) in shale gas wells.
Carbon capture, utilization and sequestration (CCUS) is playing an important role in mitigating greenhouse gas emissions and reversing global warming. Finding suitable sequestration sites is a key stage for CCUS, however, numerous potential sites do not hold sufficient data for evaluation based on conventional criteria. This study presents a feasible and efficient composite-scoring system using data analytics method to provide feeding data for machine learning studies to assess the potential depleted oil/gas reservoirs for carbon storage with limited data available. BOEM database in Gulf of Mexico (GoM) is used to apply the scoring system as a case study. The paper first presents a description of the reservoir properties and geological properties in GoM; then, the paper introduces the analytics method that maximize values of the available data and generate the final composite score for machine learning models; finally, the paper applies unsupervised machine learning method to cluster the reservoirs based on the location and composite scores. The findings of this study present that the scoring system quantifies and generalizes different types of geological and petroleum data to be feasibly applied in machine learning studies. Besides, with the BOEM database, the study revealed the best scored reservoirs located along the Louisiana coastline with large capacity, good injectivity potential and great performance in safety and economy. The result is consistent with the previous multi-criteria screening studies.
Several studies have used machine learning-based techniques to improve the production behavior prediction in existing shale gas wells. However, few studies have investigated production prediction in new wells wherein no prior information is available. This is challenging because these predictions are generally based on the analysis of data available on existing wells. Therefore, in this study, data-driven analytics is utilized to analyze the production characteristics of existing wells and improve the predictive performance of the production for new wells.
Field data on the Marcellus Shale wells with production histories exceeding 48 months were collected from Enverus’ Drillinginfo. We derived production-related attributes of these wells and identified the key factors using principal component analysis to establish the production dependency on them. Subsequently, the prediction reliability was improved by classifying different production characteristics into groups, and using their trend lines to estimate the cumulative production of the existing wells. For new wells, we developed a model to classify groups based on key factors, and utilized probabilistic values from the classified groups to predict stochastic ranges of cumulative production using an artificial neural network.
The field data were normalized with respect to lateral length or the number of stages to enable comparison between multiple wells. Outliers of each input factor were excluded during pre-processing. An analysis of production characteristics was performed by classifying the existing wells into three groups. Results indicated that Group 2 was highly productive, with evident influence of normalized fluid volume during the middle and late phases of production. Further, initial variations in production tendency were observed in Groups 1 and 3 by spacing. Trend lines of classified groups were used to forecast the cumulative production per unit length (NP) of the existing wells. The observed error was less than 10 % in the prediction of NP 48 based on the analysis of NP 6 and NP 12. Additionally, high production variability in shale play is known to induce a rapid reduction in the production trend after the initial production. Therefore, a prediction model with NP of 6, 12, and 48 months was developed. To validate the model, probabilistic values of spacing and decline factors were used in the predictions of NP 6, 12, and 48, yielding an accuracy exceeding 80% and an error of approximately 10%.
The proposed multi-well productivity analysis is a trial-and-error process based on data-driven analytics, which can be used to predict shale production in any shale play. In addition, the range of the predicted probabilistic production includes the actual values; therefore, the prediction errors are small compared to those of previous methods for new wells. Consequently, time and resources expended for data acquisition are reduced, and the reliability of productivity forecasts in shale development is improved.
The evaluation of estimated ultimate recovery (EUR) in tight gas reservoirs holds paramount significance within the domain of unconventional oil and gas development. However, the accuracy of EUR prediction using traditional decline models is hampered by the complexity of the percolation environment after the compaction of tight sandstone and the limitations of commonly employed models.
This paper proposes a new and rapid EUR evaluation method for tight gas reservoirs based on production data analysis (PDA). First, an improved model is utilized to fit the production dynamic history, enabling the determination of reservoir and fracture parameters. The introduction of the rate-normalized pressure (RNP) technique reduces the occurrence of multiple solutions during parameter inversion, simplifies the calculation of the linear flow parameter [(LFP= total fracture area × square root of permeability)], and facilitates the estimation of EUR through productivity simulation. Secondly, to validate the reliability of the proposed method, an application analysis is conducted using the Sulige gas field as a case study.
The results demonstrate that the LFP and EUR of the JA well are 1196.09 m2·md0.5 and 3.17×107 m3, respectively. Furthermore, the EUR range of four representative wells is examined, revealing an actual range of 1.98×107 m3 to 4.77×107 m3, while the EUR range obtained through linear analysis is 2.00×107 m3 to 5.07×107 m3, with relative errors of 1.10%, 3.50%, 6.30%, and 1.05%, respectively. The average error remains within 5%. Additionally, correlation analysis conducted with over ten typical wells confirms a positive correlation LFP with EUR. In conclusion, this paper presents a novel and efficient methodology approach for predicting production and estimating ultimate recoveries in shale gas wells. By overcoming the limitations of traditional decline models, the proposed method offers improved accuracy and reliability in EUR evaluation.
These findings enhance our understanding of EUR estimation in tight gas reservoirs and contribute to informed decision-making in the development of these valuable energy resources.
This study focuses on the optimization of child well spacing to mitigate fracture hits caused by hydraulic fracturing near depleted fractured rocks. Fracture hits lead to damage in the parent wells and significantly reduce the productivity of child wells due to the substantial loss of fracture fluids and proppant to the parent well. The reduction in child well productivity can exceed 50% and is influenced by well spacing and parent well(s) depletion time.
To achieve this optimization, the study combines decline curve analysis (DCA) and reservoir volumetric calculations to estimate the depleted pore volume impacted by the parent well(s). This allows for the estimation of depletion radii for the parent wells.
By analyzing production history in an actual field case, depletion radii for shale wells were estimated. The preliminary results indicated that child wells with spacing greater than the depletion radius of parent wells did not experience frac hits, in contrast to those within the depletion radius, and these findings were further confirmed through fracture-driven interactions (FDIs).
By estimating the depletion radii of parent wells, this study presents a more efficient and cost-effective method for optimizing well spacing in unconventional reservoirs. This approach helps to avoid detrimental frac hits, minimize parent well damage, and improve the productivity of child wells, thus maximizing the overall performance of hydraulic fracturing operations near depleted fractured rocks.
Shale gas, as an important unconventional natural gas resource, is the main force to increase natural gas reserves and production in the future. For shale gas with huge resources, it is particularly important to accurately evaluate its development potential and realize scale benefit development. Given the importance of shale gas production evaluation in the context of digital transformation of the oil and gas industry, this paper presents a systematic review and examination of the application of data-driven technology in shale gas production evaluation to provide an overview of their current status. With the deepening of shale gas development theory and the maturity of data-driven technology, the existing data-driven technology has made great progress in shale gas production evaluation. It is worth noting that the application of these technologies in actual production needs to be further refined to assess shale gas production more accurately in various production demand scenarios, thereby effectively guiding production optimization and development plans. Therefore, in view of the limitations of the application of data-driven technology in shale gas production evaluation, this paper puts forward some possible priorities and trends in the future, aiming to optimize the application of data-driven technology in shale gas production evaluation in the future as much as possible.
Traditional machine learning methods are difficult to accurately forecast oil production when development measures change. A method of oil reservoir production prediction based on normalized mutual information and a long short-term memory-based sequence-to-sequence model (Seq2Seq-LSTM) was proposed to forecast reservoir production considering the influence of liquid production and well spacing density. First, the marine sandstone reservoirs in the Y basin were taken as the research object to establish the sample database. Then, the feature selection was carried out according to the normalized mutual information, and liquid production, production time, equivalent well spacing density, fluidity and original formation pressure were determined as input features. Finally, a Seq2Seq-LSTM model was established to forecast reservoir production by learning the interaction from multiple samples and multiple sequences, and mining the relationship between oil production and features. The research showed that the model has a high accuracy of production prediction and can forecast the change of production when the liquid production and well spacing density change, which can provide scientific recommendations to help the oilfield develop and adjust efficiently.
Current methods for predicting output, such as material balancing and numerical simulation, need years of production history, and the model parameters employed determine how accurate they are. The use of artificial neural network (ANN) technology in the production forecasting of a deep offshore field under water injection/water flooding in Nigeria’s Niger-Delta region is investigated in this study. Oil, water, and gas production rates were predicted using well models and engineering features. Real-world field data from producer and water injection wells in deep offshore is used to test the models’ performance. Ninety percent (90%) of the historical data were utilised for training and validating the model framework before being put to the test with the remaining information. The predictive model takes little data and computation and is capable of estimating fluid production rate with a coefficient of prediction of more than 90%, with simulated results that match real-world data. The discoveries of this work could assist oil and gas businesses in forecasting production rates, determining a well’s estimated ultimate recovery (EUR), and making informed financial and operational decisions.
Breagh, a moderately tight gas field in the Southern North Sea started production in 2013. Until compression is available, Breagh has been constrained by minimum turn-down constraints and the need to keep wells flowing above their loading rates leading to complicated well scheduling optimisation.
This paper describes a tool that makes hourly forecasts for cycling and non-cycling wells used to rapidly assess a proposed well scheduling for loading and minimum turn-down compliance.
The Breagh short term forecasting tool uses empirical equations to match historical pressures and production rates. Parameter values have been found for each of the existing nine flowing wells which match build-up pressures, predict the initial gas rate via a pseudo-PI approach and decline the gas rate over its flowing period. By matching the wells’ hourly rates over multiple months, these same parameters are then used to predict future well flows and display the results in a number of graphics for easy history matching and well scheduling assessments.
Due to the steep decline of the cycling wells, an hourly forecast was required. Excellent matches were obtained over multiple month periods. A single framework was found to be sufficient for both cycling and non-cycling wells, but with different sets of parameters: the Breagh short term forecasting tool was born and has proved to be very useful for: short term forecasting over a few months; rapid assessment of well schedules, including cases where a well issue has required a well to be closed in at short notice and alternative well schedules devised to enable uninterrupted production; and to assess the need for additional production tests to further lower minimum turn-down.
A new framework is provided for the performance forecasting of cycling gas wells, some with significant declines on an hourly basis, thus enabling rapid response times to well schedule optimisation issues. The following are the key findings:Hourly data rather than daily were required to make quality forecasts for cycling wells with extending a few months ahead.New potential well cycling schedules could be rapidly assessed, and more optimal schedules selected.The Breagh Short Term Forecasting Tool was one of three key contributors to returning Breagh from weekly cyclic to continuous production.
Micro to Nano pore scale and strong stress-sensitive effects are the main characteristics of shale reservoirs. Restricting rate and frequently changing working system are the commonly production pattern in shale oil developments. It makes the flowing characteristics of shale reservoir extremely complex and brings new challenges to the dynamic analysis and accurately estimated ultimate recovery (EUR) prediction of shale oil. In this study, a new method is proposed for EUR evaluation of shale oil wells by combining empirical method, modern production decline method and analytical method. The production data of shale oil wells is firstly normalized to couple the effects of variable rate and pressure, and analyze the flowing characteristic based on the regularized production data to reduce the impact of data fluctuations on EUR evaluation under changes of working system. Finally, empirical decline method, modern decline method and analytical model are adopted to comprehensively evaluate the EUR of shale oil wells and reduce the uncertainty of EUR evaluation. The practical application results show that the prediction of the proposed method is similar to numerical simulation, which can effectively improve the accuracy and reliability of the productivity evaluation of shale oil. The results also demonstrated that the proposed method is suitable for the rapid evaluation of EUR in shale oil wells.KeywordsShale oil and gasProductivity evaluationEmpirical decline methodModern production decline methodAnalytical methodComprehensive evaluation method
As the world transitions towards a more sustainable energy landscape, the need for a less environmentally harmful transition fuel becomes increasingly crucial. This has led to renewed interest in developing previously abandoned natural gas fields. In this paper, we present Field Surface Planner Algorithm (FSPA), a DCA-based model implemented in Python, which aims to generate an effective investment timing strategy for bringing new wells online in large gas fields. The model considers various surface facilities and commercial constraints, while prioritizing the fulfillment of contractual gas demand.
FSPA's algorithm revolves around a sales error concept, which quantifies the disparity between the current sales gas rate and the nominated gas demand. By comparing this sales error against a predetermined error tolerance, the model determines the optimal timing for introducing new wells into production. This approach ensures that the gas field's total sales gas rate remains in line with the nominated gas demand, while minimizing any potential environmental impact.
To ensure efficient operation, FSPA operates within the bounds of nominated gas demand constraints and facility constraints. Nominated gas demand constraints serve as a limit against which the on-stream dates and DCA forecasts for future wells are optimized, minimizing the need for flaring gas. Facility constraints restrict the volume of gas supply that the algorithm optimizes, aligning it with the capabilities of the surface facilities.
During a well's initial contribution, FSPA generates accurate DCA forecasts based on several key factors, including the exact amount of sales gas required to close existing gaps, the well's reserves, and economic limitations. This information aids in making informed decisions regarding short-term production shortfalls and facilitates effective negotiation with customers, ultimately delivering substantial value for the organization involved in gas field development.
By providing an efficient solution for well sequencing, the proposed model significantly reduces gas flaring costs, minimizes opportunity costs associated with production shortfalls, and assists in short-term decision-making processes. Overall, it offers a valuable contribution to sustainable energy transition efforts and promotes responsible development in the gas industry.
ResearchGate has not been able to resolve any references for this publication.