Forecasting Oil Production Research Articles

Debiasing probabilistic oil production forecasts

Exploration and production companies in the hydrocarbon industry have every interest in producing unbiased production forecasts at the time of the investment decision, since it is an intrinsic part of making oil field development profitable. However, recent results show presence of significant biases in the uncertainty models which support these decisions. Some important questions which are addressed in this study are i) whether there are simpler and more robust approaches to forecasting than what is the practice in this industry today, ii) whether forecasts can be calibrated for bias, and iii) what the consequences are for valuation of investments in new oil fields. In this study, 71 oil fields on the Norwegian continental shelf with production start between 1995 and 2020 are analyzed. Three robust bias reduction methods are proposed: a pure reference class forecast and two calibration models for the field operators' own forecasts. These show that expected production early in the field lifetime must be shifted down and that the uncertainty range must be expanded. The results are also consistent across field sizes and over time. The findings in this study demonstrate the need to draw on results in behavioral economics to improve uncertainty quantification - reference class forecasting is an inexpensive and powerful way to avoid cognitive biases. An important conclusion is also that the discounted revenue stream from new oil fields is far more uncertain and has a lower expected value than companies lay to ground.

Forecasting Oil Production Flowrate Based on an Improved Backpropagation High-Order Neural Network with Empirical Mode Decomposition

Developing a forecasting model for oilfield well production plays a significant role in managing mature oilfields as it can help to identify production loss earlier. It is very common that mature fields need more frequent production measurements to detect declining production. This study proposes a machine learning system based on a hybrid empirical mode decomposition backpropagation higher-order neural network (EMD-BP-HONN) for oilfields with less frequent measurement. With the individual well characteristic of stationary and non-stationary data, it creates a unique challenge. By utilizing historical well production measurement as a time series feature and then decomposing it using empirical mode decomposition, it generates a simpler pattern to be learned by the model. In this paper, various algorithms were deployed as a benchmark, and the proposed method was eventually completed to forecast well production. With proper feature engineering, it shows that the proposed method can be a potentially effective method to improve forecasting obtained by the traditional method.

Prediction of well performance in SACROC field using stacked Long Short-Term Memory (LSTM) network

Forecasting oil production can be very challenging, especially for reservoirs with sparse data or other complexities. If traditional decline curve analysis or time series models fail to capture production rate variabilities, a machine learning model for time series data may be effective. A temporal machine learning model, a long short-term memory network model (LSTM) in specific, may be trained to predict oil, gas, and water production rates. We develop an LSTM for such an application and evaluate its efficacy with a case study of the Scurry Area Canyon Reef Operators Committee (SACROC) unit, an active CO2 enhanced oil recovery (EOR) field in western Texas, USA. The monthly averaged production rates of oil, gas and water from 23 producers are obtained from simulations of the SACROC reservoir model that also includes 22 injectors for 5-spot injection of water and CO2 (alternating annually). The bottom hole pressure (BHP) of the producers, BHPs of surrounding injectors, and historical production rates are used as input data for the LSTM. Blind predictions of LSTM test sets show promising outcomes for data that otherwise traditional time series models are not effective. Stacked LSTM models are efficient in multi-step-ahead forecasting. Such an LSTM approach may also be effective for quantitative analysis of unconventional oil and gas reservoirs like shales or other tight formations. Critical aspects of the LSTM workflow include optimization of machine learning parameters and quantification of the relative impacts of different variables on forecasted outcomes.

Modified aquila optimizer for forecasting oil production

ABSTRACT Oil production estimation plays a critical role in economic plans for local governments and organizations. Therefore, many studies applied different Artificial Intelligence (AI) based methods to estimate oil production in different countries. The Adaptive Neuro-Fuzzy Inference System (ANFIS) is a well-known model that has been successfully employed in various applications, including time-series forecasting. However, the ANFIS model faces critical shortcomings in its parameters during the configuration process. From this point, this paper works to solve the drawbacks of the ANFIS by optimizing ANFIS parameters using a modified Aquila Optimizer (AO) with the Opposition-Based Learning (OBL) technique. The main idea of the developed model, AOOBL-ANFIS, is to enhance the search process of the AO and use the AOOBL to boost the performance of the ANFIS. The proposed model is evaluated using real-world oil production datasets collected from different oilfields using several performance metrics, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE), coefficient of determination (R2), Standard Deviation (Std), and computational time. Moreover, the AOOBL-ANFIS model is compared to several modified ANFIS models include Particle Swarm Optimization (PSO)-ANFIS, Grey Wolf Optimizer (GWO)-ANFIS, Sine Cosine Algorithm (SCA)-ANFIS, Slime Mold Algorithm (SMA)-ANFIS, and Genetic Algorithm (GA)-ANFIS, respectively. Additionally, it is compared to well-known time series forecasting methods, namely, Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Seasonal Autoregressive Integrated Moving Average (SARIMA), and Neural Network (NN). The outcomes verified the high performance of the AOOBL-ANFIS, which outperformed the classic ANFIS model and the compared models.

Possibility of using machine learning algorithms to assess the effectiveness of acid hydraulic fracturing

At the fields of the Perm Territory, there is a significant application of hydraulic fracturing technology to stimulate the production of carbonate reservoirs. The analysis of the efficiency of the acid hydraulic fracturing technology by the example of the fields of the Perm Territory is presented in the article. The effectiveness of the technology is largely determined by the geological and physical conditions of the fields. Using the methods of mathematical statistics, the main geological and physical parameters which most significantly affect the effectiveness of measures for acid hydraulic fracturing have been identified. A multiple regression model was built to find out the complex influence of the parameters on the increase in oil production after the measures. Various machine learning algorithms have been used to predict the increase in oil and fluid input rates after acid fracturing. Different algorithms showed different results. For the problems of forecasting oil production after the application of acid fracturing, the most successful one was the support vector method.

Forecasting oil production for Oligocene C sequence, X field, Cuu Long basin using logistic growth model

Hydrocarbon production forecasting for the field lifetime in the short and long term is an important phase, the accuracy of this process plays a tremendous role in giving the decision of reasonable field management and development. In this article, the logistic growth models using the function MATLAB’s ‘nlinfit’ have been built to forecast oil production yield for the Oligocene C sequence, X field, Cuu Long basin. Thanks to the combination with the history matching process, the logistic growth model expressed high accuracy, the results of the model are very close to the actual production data with a relative error of 1,85%. The article analyzed and evaluated the production parameters of wells obtained when building logistic growth models such as the time at which half of the carrying capacity has been produced, the steepness of the decline of the rate, and the production rate of the wells at the forecast time. Without applying any improved oil recovery method, the decline of the rate of all wells approaches 100 bbl/d before reaching the validity period of the oil and gas contract. This is the basis for operators to establish and improve field development plans.

An artificial intelligence method for improving upscaling in complex reservoirs

The upscaling of geological properties is a fundamental requirement to construct a suitable simulation model since the geological model typically contains millions of cells, which makes it computationally difficult to simulate it on such a scale. The process consists of a scale transfer that adapts the petrophysical properties of a high-resolution grid to a coarser grid. Nevertheless, this is not a trivial operation, especially for absolute permeability, which is a non-additive property. One of the most reliable methods is to perform a flow simulation on fine-scale cells that correspond to the coarse block and derive the single permeability value that reflects the same flow value. Still, this is a very time-consuming method and depends on the imposed boundary conditions. This work aims to take advantage of recent advances in artificial intelligence to produce results with similar or better quality of flow-based methods, and more adaptive. It handles models with multiple geological realizations by employing machine learning to capture patterns from a subset of scenarios and use it to generalize for all others. The methodology is divided into two stages — local and global. In the first one, a deep neural network is trained with a fraction of geological realizations using the flow-based results as a reference. In the second stage, clustering analysis is employed along with a neural network optimization to learn an adjustment procedure for the coarse simulation model concerning the cumulative field production predicted by the fine model simulation. Afterward, the trained network performs the upscaling of remaining realizations, but more efficiently in terms of computational time and providing better output in terms of production and injection. The method was applied to a benchmark model and the experimental results demonstrate that the local technique was capable of reproducing very similar values to the flow-based upscaling using only one scenario in training and at the global stage the coarse model was improved even further matching the field oil production forecast of the fine model. Different scenarios were used for training and testing and the results were consistent showing no bias towards a specific configuration and capability of generalization to different scenarios. Ultimately, the proposed artificial intelligence approach performed an accurate upscaling, surpassing the reference approach with forecast production similar to the fine model, and also fast to compute when considering multiple geological realizations, since it reduces the required numerical simulations to a fraction of the total.

A Novel Exponential Time Delayed Fractional Grey Model and Its Application in Forecasting Oil Production and Consumption of China

Based on particle swarm optimization (PSO), a new exponential time delay fraction order grey prediction model is proposed in this paper. Firstly, the original data is preprocessed by fractional-order accumulation; on the basis of fractional-order accumulation, it is proved that the initial value of the original sequence satisfies the fixed point theorem. On the basis of GM(1,1) model, a new model is established by adding exponential time delay term. The model is discretized by integral, the least square estimation of the linear parameters and the approximate time response equation are obtained. Finally, PSO is used to search the optimal parameters of the model and the experimental results are verified by Wilcoxon rank sum test. In order to test the good adaptability and strong prediction ability of the new model, two groups of data are verified for the production and consumption of oil and electricity in China. The results show that the new model has better prediction accuracy and adaptability than the other existing six grey models.

Bayesian Deep Decline Curve Analysis: A New Approach for Well Oil Production Modeling and Forecasting

Summary Following the rapid growth of unconventional resources, many models and methods have been proposed for forecasting the performances of unconventional wells. Several studies have attempted to use machine learning (ML) for improving the forecasting. However, owing to limitations of ML in regard to long-term forecasts (e.g., the occurrence of unphysical results), most of these ML forecasts are not satisfactory. In this work, we propose, demonstrate, and discuss a new ML approach able to rapidly provide probabilistic, long-term forecasts of oil production rates from individual wells in a decline curve analysis (DCA) manner. The novelties of the proposed approach are as follows: (1) it combines an automated ML (AutoML) method for supervised learning and a Bayesian neural ordinary differential equation (BN-ODE) framework for time-series modeling; (2) it uses the DCA model to inform the BN-ODE framework of “physics” and regulate the BN-ODE forecasts; and (3) several completion parameters (such as locations, lengths, and slickwater volume) of individual wells are analyzed and included as the inputs of model building, in addition to measured oil production rate data. Specifically, AutoML method is first used to model the relationship between the well location, completion parameters, and the DCAs parameters, and the BN-ODE framework is then used to model the relationship between the DCAs parameters and the time-series oil production rates. A publicly accessible data set, consisting of completion parameters and oil production rates, of 396 horizontal wells in the Bakken Shale Formation is used to train and test the model of the proposed approach. The results lead to the conclusion that the proposed approach is practical for providing probabilistic, long-term forecasts of oil production from individual wells, given data of existing wells in the reservoir.

Data-driven deep-learning forecasting for oil production and pressure

Production forecasting plays an important role in oil and gas production, aiding engineers to perform field management. However, this can be challenging for complex reservoirs such as the highly heterogeneous carbonate reservoirs from Brazilian Pre-salt fields. We propose a new setup for forecasting multiple outputs using machine-learning algorithms and evaluate a set of deep-learning architectures suitable for time-series forecasting. The setup proposed is called N-th Day and it provides a coherent solution for the problem of forecasting multiple data points in which a sliding window mechanism guarantees there is no data leakage during training. We also devise four deep-learning architectures for forecasting, stacking the layers to focus on different timescales, and compare them with different existing off-the-shelf methods. The obtained results confirm that specific architectures, as those we propose, are crucial for oil and gas production forecasting. Although LSTM and GRU layers are designed to capture temporal sequences, the experiments also indicate that the investigated scenario of production forecasting requires additional and specific structures.

Application of Rough Neural Network to forecast oil production rate of an oil field in a comparative study

As real production data of a well have an irregular pattern, accurate prediction of oil rate demands a powerful model to capture the non-linear behavior of data. In addition to the traditional methods such as Decline Curve Analysis (DCA), Machine Learning methods can be considered as one of the effective predictive tools of engineers. This paper employs a comparative approach to find an appropriate network to forecast the oil production rate of an Iranian Oil field. The performance of several networks such as Rough Neural Network (Rough NN), Long-Short-Term Memory (LSTM), Artificial Neural Network (ANN) with only dense layer, and 1D Convolutional Neural Network (CONV-1D) were monitored by investigating various statistical parameters such as error value, the cross plot of real data and predicted data and error distribution. A combination of five inputs static and dynamic parameters was considered as input to the model. All networks were trained for 80% of all data (10025 points) and the rest were divided equally for testing and validation. The highest performance was observed in the results of Rough-NN with a coefficient of determination of 0.82 for predicting test data. The results showed an accuracy slightly less than Rough-NN for the case of CONV-1D (R 2 = 0.79). However, the worst performance referred to ANN and LSTM where their R-squared was about 0.54. Finally, the precision of the superior network (Rough-NN) was compared with DCA for several wells. It was revealed that DCA cannot predict the oil production rate accurately specifically when the well comes into production after a long shut-in. In contrast, Rough-NN is capable of predicting the oil production rate accurately regardless of the production history of the well. • Handling the noisy and non-linear behavior of real data by Rough neural network. • Accurate prediction of oil rate under various dynamic parameters. • Comparing the performance of various networks for oil rate prediction. • Demonstrating the preference of machine learning method respect to traditional decline curve analysis. • Applying surface constraints on well oil rate prediction.

Generalized Extreme Value Statistics, Physical Scaling and Forecasts of Oil Production from All Vertical Wells in the Permian Basin

We analyze nearly half a million vertical wells completed since the 1930s in the most prolific petroleum province in the U.S., the Permian Basin. We apply a physics-guided, data-driven forecasting approach to estimate the remaining hydrocarbons in these historical wells and the probabilities of well survival. First, we cluster the production data set into 192 spatiotemporal well cohorts based on 4 reservoir ages, 6 sub-plays, and 8 completion date intervals. Second, for each cohort, we apply the Generalized Extreme Value (GEV) statistics to each year of oil production from every well in this cohort, obtaining historical well prototypes. Third, we derive a novel physical scaling that extends these well prototypes for several more decades. Fourth, we calculate the probabilities of well survival and observe that a vertical well in the Permian can operate for 10–100 years, depending on the sub-play and reservoir to which this well belongs. Fifth, we estimate the total field production of all existing vertical wells in the Permian by replacing historical production from each well with its prototype. We then time-shift and sum up these prototypes together, obtaining 34 billion barrels of oil as estimated ultimate recovery (EUR). Our most notable finding is that the rate of finding big reservoirs in the Permian has been declining drastically and irreversibly since the 1970s. Today, operators need to drill wells that are twice as deep as the 1930s’ wells, yet they produce 4–12 times less.

Statistical modeling of oil reservoir life cycle

Based on statistical analysis, a method of stage-by-stage structuring and allocation of the boundaries of the stages of oil field development is proposed. The procedures for calculating the remaining recoverable reserves and the time to reach the maximum level of current oil production are given. Based on the values of accumulated oil production at the final stage of development, using the adaptive Kalman filter, forecasts of oil production are constructed in discrete time and estimates of the possibility of achieving design indicators with the existing development system are given. Keywords: Stage-to-stage structuring; Development stages; Forecast of recoverable reserves; Predicted oil production; Kalman filter.

Statistical modeling of oil reservoir life cycle

Based on statistical analysis, a method of stage-by-stage structuring and allocation of the boundaries of the stages of oil field development is proposed. The procedures for calculating the remaining recoverable reserves and the time to reach the maximum level of current oil production are given. Based on the values of accumulated oil production at the final stage of development, using the adaptive Kalman filter, forecasts of oil production are constructed in discrete time and estimates of the possibility of achieving design indicators with the existing development system are given. Keywords: Stage-to-stage structuring; Development stages; Forecast of recoverable reserves; Predicted oil production; Kalman filter.

Approaches to forecast oil field performance at the well pads level and beta testing on oilfields of RN-Nyaganneftegaz (ROSNEFT)

Background. The study performed by Rosneft employees shown in this paper demonstrates approach and analytical methods that allows to forecast oil production at the level of minimal infrastructure units. These approaches are used to forecast long-term oil production and predict infrastructure blockage. The approach was partially automated by the authors. This made it possible to testing at giant Krasnoleninskoye oilfield. 
 
 Aim. The study was performed in order to develop and test an approaches to forecast oil production of large oil fields with high detail levels. 
 
 Materials and methods. Common methods of decline curve analysis and water-into-oil curve analysis were used in this work to analyze the precondition. The main feature of the approach is the analysis of precondition at the level of large well clusters and transfer it to the level of wells. Some of the actions were automated by new proprietary software and were tested at the giant brown field. The software was integrated with the corporate database. 
 
 Results. An author’s approach has been developed. The approach allows to forecast oil production at the level of infrastructure units using analytical methods. Oil production of the giant brown field with high detail levels were forecasted using the proposed approaches and developed software. 
 
 Conclusions. The results show that the developed approaches and software can be used to forecast mediumand long-term performance of producing oil fields in the conditions of existing external and infrastructural constraints.

Forecast of Economic Tight Oil and Gas Production in Permian Basin

We adopt a physics-guided, data-driven method to predict the most likely future production from the largest tight oil and gas deposits in North America, the Permian Basin. We first divide the existing 53,708 horizontal hydrofractured wells into 36 spatiotemporal well cohorts based on different reservoir qualities and completion date intervals. For each cohort, we fit the Generalized Extreme Value (GEV) statistics to the annual production and calculate the means to construct historical well prototypes. Using the physical scaling method, we extrapolate these well prototypes for several more decades. Our hybrid, physico-statistical prototypes are robust enough to history-match the entire production of the Permian mudstone formations. Next, we calculate the infill potential of each sub-region of the Permian and schedule the likely future drilling programs. To evaluate the profitability of each infill scenario, we conduct a robust economic analysis. We estimate that the Permian tight reservoirs contain 54–62 billion bbl of oil and 246–285 trillion scf of natural gas. With time, Permian is poised to be not only the most important tight oil producer in the U.S., but also the most important tight gas producer, surpassing the giant Marcellus shale play.

Production Forecasting with the Interwell Interference by Integrating Graph Convolutional and Long Short-Term Memory Neural Network

Summary Accurate production forecasting is an essential task and accompanies the entire process of reservoir development. With the limitation of prediction principles and processes, the traditional approaches are difficult to make rapid predictions. With the development of artificial intelligence, the data-driven model provides an alternative approach for production forecasting. To fully take the impact of interwell interference on production into account, this paper proposes a deep learning-based hybrid model (GCN-LSTM), where graph convolutional network (GCN) is used to capture complicated spatial patterns between each well, and long short-term memory (LSTM) neural network is adopted to extract intricate temporal correlations from historical production data. To implement the proposed model more efficiently, two data preprocessing procedures are performed: Outliers in the data set are removed by using a box plot visualization, and measurement noise is reduced by a wavelet transform. The robustness and applicability of the proposed model are evaluated in two scenarios of different data types with the root mean square error (RMSE), the mean absolute error (MAE), and the mean absolute percentage error (MAPE). The results show that the proposed model can effectively capture spatial and temporal correlations to make a rapid and accurate oil production forecast.

Influence of reservoir lithology on porous flow resistance of gas-bearing tight oil reservoirs and production forecast

The release of dissolved gas during the development of gas-bearing tight oil reservoirs has a great influence on the effect of development. In this article, the high-pressure mercury intrusion experiment was carried out in cores from different regions and lithologies of the Ordos Basin and the Sichuan Basin. The objectives are to study the microscopic characteristics of the porous throat structure of these reservoirs and to analyze the porous flow resistance laws of different lithology by conducting a resistance gradient test experiment. A mathematical model is established and the oil production index is corrected according to the experiment results to predict the oil production. The experimental results show that for tight reservoirs in the same area and lithology, the lower the permeability under the same back pressure, the greater the resistance gradient. And for sandstone reservoirs in different areas, the resistance gradients have little difference and the changes in the resistance coefficients are similar. However, limestone under the same conditions supports a much higher resistance gradient than sandstone reservoirs. Furthermore, the experimental results are consistent with the theoretical analysis indicating that the PVT (pressure–volume-temperature) characteristics in the nanoscale pores are different from those measured in the high-temperature, high-pressure sampler. Only when the pressure is less than a certain value of the bubble point pressure, the dissolved gas will begin to separate and generate resistance. This pressure is lower than the bubble point pressure measured in the high-temperature and pressure sampler. The calculation results show that the heterogeneity of limestone reservoirs and the mismatch of fluid storage and flow space will make the resistance, generated by the separation of dissolved gas, have a greater impact on oil production.

Solving problems of the oil and gas sector using machine learning algorithms

The article describes the tasks of the oil and gas sector that can be solved by machine learning algorithms. These tasks include the study of the interference of wells, the classification of wells according to their technological and geophysical characteristics, the assessment of the effectiveness of ongoing and planned geological and technical measures, the forecast of oil production for individual wells and the total oil production for a group of wells, the forecast of the base level of oil production, the forecast of reservoir pressures and mapping. For each task, the features of building machine learning models and examples of input data are described. All of the above tasks are related to regression or classification problems. Of particular interest is the issue of well placement optimisation. Such a task cannot be directly solved using a single neural network. It can be attributed to the problems of optimal control theory, which are usually solved using dynamic programming methods. A paper is considered where field management and well placement are based on a reinforcement learning algorithm with Markov chains and Bellman's optimality equation. The disadvantages of the proposed approach are revealed. To eliminate them, a new approach of reinforcement learning based on the Alpha Zero algorithm is proposed. This algorithm is best known in the field of gaming artificial intelligence, beating the world champions in chess and Go. It combines the properties of dynamic and stochastic programming. The article discusses in detail the principle of operation of the algorithm and identifies common features that make it possible to consider this algorithm as a possible promising solution for the problem of optimising the placement of a grid of wells.

Box–Jenkins Methodology Application to Improve Crude Oil Production Forecasting: Case Study in a Colombian Field

Box–Jenkins Methodology Application to Improve Crude Oil Production Forecasting: Case Study in a Colombian Field