Production Forecast Models Research Articles

Prediction of Wells Performance and Producing Life Applying Production Decline Analysis

Analyzing the production data became a common method in modern reservoir engineering practice, which is used to dynamically indicate the reservoir's characteristics, predict long-term performance, and determine the reservoir's criteria through daily production analysis. Input data required to analyze production data is production rate and sometimes flow pressure (if any). The decline curve analysis method analyzes well production data to history matching and forecasts the future well performance. The decline curve analysis (DCA) method is a non-linear regression method for production history matching and prediction. In this study, we used the DCA method to analyze eight oil wells' production data produced from the Mishrif reservoir in an Iraqi oil field. This work dealt with the prediction of the reservoir’s performance and future production rate using the DCA method tool integrated with IPM- MBAL software. Production decline rate trend analysis was performed by generating standard curves utilizing exponential, hyperbolic, and harmonic decline model equations based on historical production data. Afterward, a production forecast model was built for production to 2025. The producing life for each well at the economic flow rate (300 bbl/Day) is estimated. Results from the production forecast showed that the exponential and hyperbolic decline models yielded 54.057 MMSTB and 80.73 MMSTB, respectively. The decline rates for hyperbolic and exponential models till 2025 were estimated as 4614.25 STB/Day and 1153.1STB/Day.

Day‐ahead wind power ramp forecasting using an image‐based similarity search strategy

With the increase in penetration of wind generation on interconnected power systems, the importance of wind power ramp forecasting has continuously grown. Large power ramps caused by sudden weather changes raise more concerns due to their significant impact on the power system economics and stability. Correct wind power ramp forecasts can help the system operators and utility companies to tradeoff the risks when scheduling wind energy in the electricity market. In this paper, a day-ahead wind power ramp forecasting algorithm is developed to provide probabilistic ramp forecasts for look-ahead times up to 48 h using hourly wind speed forecasts from Environment Canada High Resolution Deterministic Prediction System (HRDPS). An image-based similarity search strategy has been designed to build a direct link between the wind speed forecasts and the wind power ramp prediction, thus reducing the impact of the uncertainty from both the power production forecast model and the ramp identification process on the forecasting accuracy. A performance assessment and validation of the proposed ramp event forecasting method is conducted by using the forecast and operation data from six investigated wind farms across Canada.

Proxy Capacitance-Resistance Modeling for Well Production Forecasts in Case of Well Treatments

Summary We disclose a new-age field-scale production forecast model that handles complex treatment of wellbores during their life cycle. Predictive production models have been an object of increased interest and research for a long time due to the need for a fast tool for forecasting production rates or choosing an optimal field development scheme. The existing approaches based on the material balance equation have several limitations and are not very applicable for real objects. Full-scale reservoir modeling is relatively slow and requires large computing resources. In this paper, we propose a proxy model based on advanced capacitance-resistance approach. The model predicts multiphase flow rates based on the available historical data of field production and information about well treatments. In addition, it provides preferable transmissibility trends, the presence of sealed or leaking faults, and the degree of dissipation between injector-producer well pairs. The advanced feature of the model is time-dependent weight coefficients, which have not been studied previously. They help in accounting the shut-in and workover periods and can be found during the optimization procedure simultaneously. Another feature is fast calculations due to a vectorized form of the model and application of modern optimization techniques. All these options allow modeling real oil fields with a large number of wells and a complex system of production control.

Data-driven model for hydraulic fracturing design optimization. Part II: Inverse problem

We describe a stacked model for predicting the cumulative fluid production for an oil well with a multistage-fracture completion based on a combination of Ridge Regression and CatBoost algorithms. The model is developed based on an extended digital field data base of reservoir, well and fracturing design parameters. The database now includes more than 5000 wells from 23 oilfields of Western Siberia (Russia), with 6687 fracturing operations in total. Starting with 387 parameters characterizing each well, including construction, reservoir properties, fracturing design features and production, we end up with 38 key parameters used as input features for each well in the model training process. The model demonstrates physically explainable dependencies plots of the target on the design parameters (number of stages, proppant mass, average and final proppant concentrations and fluid rate). We developed a set of methods including those based on the use of Euclidean distance and clustering techniques for offset wells selection (search for similar wells in terms of certain metrics), which is useful for a field engineer to analyze earlier fracturing treatments on similar wells. These approaches are also adapted for obtaining the margings for optimization parameters for a particular pilot well, as part of the field testing campaign of the methodology. An inverse problem (selecting an optimum set of fracturing design parameters to maximize production) is formulated as optimizing a high dimensional black box approximation function constrained by boundaries and solved with four different optimization methods: surrogate-based optimization (pSeven), sequential least squares programming, particle swarm optimization and differential evolution. A recommendation system containing all of the above methods is designed to advise a production stimulation engineer on an optimized fracturing design. • Workflow for data-drive fracturing design optimization is developed based on digital database and production forecast model. • Inverse problem is solved with several gradient-free optimization methods (with probability of improvement) and surrogate-based optimization. • Methodology for offset wells selection is presented as an algorithm of optimization based on best practices on similar wells. • Recommendation system containing these methods is designed to assist production stimulation engineer.

CO2-prepad injection EOR simulation and sensitivity analysis considering miscibility and geomechanics in tight oil reservoirs

Injecting CO2 pre-pad into tight reservoirs before hydraulic fracturing is a promising way to improve the productivity of single well. In this method, CO2 should be injected into the formation at a high rate. Also the miscibility of CO2 and oil with the changes of intricate geomechanics during this process are complex. Meanwhile, the geological and operating parameters have a dramatic influence to oil production and is lack of systematic research. In this study, a fluid model of crude oil was established to evaluate the impact of CO2 on oil volume and viscosity. Then, a compositional numerical model coupled miscibility and geomechanics were set up to analyze enhanced oil recovery (EOR) potential under different miscible degree with in-situ stress field. Ultimately, in accordance with multiple nonlinear regression production forecast models, Sobol, FAST (Fourier Amplitude Sensitivity Test), and DGSM (derivation-based Global Sensitivity Measure) methods were utilized to study the main factors affecting cumulative production. The results shown that CO2 almost has the same effect on the volume of these tight oil and has the most significant effect on the viscosity of crude oil we studied by comparing. The production of miscible flooding is 38.2% and 16.4% higher than that of immiscible and near-miscible flooding, respectively. The formation were compacted during production especially in the early stage, hence the production of the model with geomechanics was 23.3% less than the model of non-geomechanics. The first-order sensitivity indexes of bottom-hole flow pressure, reservoir thickness and fracture conductivity are about 0.808, 0.176 and 0.004 respectively obtained by Sobol method, consistent with two other methods. This result shows that bottom-hole flow pressure, reservoir thickness and fracture conductivity are the most important factors affecting production, and bottom-hole flow pressure has the most significant impact on production. This work investigates how miscibility and geomechanics effect on production as well as clarified the factors affecting production for CO2 pre-pad injection, providing vital insights into operating in tight reservoirs for oil field.

Oil production forecast models based on sliding window regression

The article presents a method of statistical modeling and forecasting of oil production rate at a real field that has been in operation for a long time. Nowadays such fields form the basis of the oil production industry. The “aging” of oil production on such fields requires continuous change of the management of the oil production system. Therefore, production modeling and forecast must be dynamic, varying over time. Moreover the system for measuring production parameters, like production and injection series that are used to fit the proposed model, does not work regularly and has a large number of random gaps. So the development of our model starts with combination of interpolation and averaging procedures to produce a regular grid of data points. On such a regular grid, we construct the desired dynamic regression model using a machine learning method that we refer to as the sliding window regression. Our model predicts the total oil production rate for the next month taking into account some manually designed features of the oil field from the bounded history interval. For convenience, the features of the model are divided into four groups: integral, local, pressure and autoregressive. After prediction for the next month the history window moves one step forward and the model is retrained to fit the new data. Further we generalize the idea of sliding window to obtain a model with stable coefficient dynamics and ability for long-term forecast. In particular, we consider two approaches: introducing some regularization terms to the objective loss functional or weighing sample objects with weights decreasing in the past. The model is verified on data from a real oil field and shows good results in terms of relative error.

Irradiance forecast model for PV generation based on cloudiness web service

Among renewable energies that play a critically important role to achieve a sustainable development in the future, solar photovoltaic (PV) power have been growing dramatically recently. Despite the enormous potential of solar PV, its drawback, which is intermittent in generation, imposes a significant issue on electrical system, such as stability or operation planning of the system. One of the solutions for the problem is PV generation forecast that have been researched in various studies. This paper presents a methodology to obtain an indirect PV production forecast model. Processes of model formation, validation and test are performed in the study using real data from a Campbell Scientific weather station Grenoble, France.

Effect of introducing weather parameters on the accuracy of milk production forecast models

The objective of this study was to analyze the effect of adding meteorological data to the training process of two milk production forecast models. The two models chosen were the nonlinear auto-regressive model with exogenous input (NARX) and the multiple linear regression (MLR) model. The accuracy of these models were assessed using seven different combinations of precipitation, sunshine hours and soil temperature as additional model training inputs. Lactation data (daily milk yield and days in milk) from 39 pasture-based Holstein-Friesian Irish dairy cows were selected to compare to the model outputs from a central database. The models were trained using historical milk production data from three lactation cycles and were employed to predict the total daily milk yield of a fourth lactation cycle for each individual cow over short (10-day), medium (30-day) and long-term (305-day) forecast horizons. The NARX model was found to provide a greater prediction accuracy when compared to the MLR model when predicting annual individual cow milk yield (kg), with R2 values greater than 0.7 for 95.5% and 14.7% of total predictions, respectively. The results showed that the introduction of sunshine hours, precipitation and soil temperature data improved the prediction accuracy of individual cow milk prediction for the NARX model in the short, medium and long-term forecast horizons. Sunshine hours was shown to have the largest impact on milk production with an improvement of forecast accuracy observed in 60% and 70% of all predictions (for all 39 test cows from both groups). However, the overall improvement in accuracy was small with a maximum forecast error reduction of 4.3%. Thus, the utilization of meteorological parameters in milk production forecasting did not have a substantial impact on forecast accuracy.

Research on a Model of Oil Production Forecast Based on Technical Development and Dynamic Change of Reserves

The Hubbert model is widely used in forecasting the oil production. However, many scholars compared the history data to the data calculated by Hubbert model and got the conclusion that the Hubbert model was not suitable for forecasting oil production. In this paper, the Hubbert model is modified and the technical progress is considered in the model. An oil production forecast system is presented, which can be used to analyze the technical progress and dynamic change of reserves. This forecast system is decomposed into three models, namely the technical model, dynamic reserves model and production forecast model. The purpose of this paper is to show the powerful effect of technical progress and dynamic reserves on oil production. It is an innovative view that less research focus on this side. Furthermore, seismic technology, exploration theory, drilling technology and oil production technology are considered in technical model. These models are used to forecast world oil production from 2018 to 2030.

Effect of parity weighting on milk production forecast models

The objectives of this study were to compare the prediction accuracy of two milk prediction models at the individual cow level and to develop, compare and evaluate six input data preprocessing treatments designed to factor parity information into the milk prediction model configuration process. The two models were a nonlinear auto-regressive model with exogenous input and a polynomial curve fitting model. These were tested using six different parity data input treatments. Different combinations of static parity weight, dynamic parity weight and removal of the first lactation data were selected as input treatments. Lactation data from 39 individual cows were extracted from a sample herd of pasture-based Holstein-Friesian cattle located in the south of Ireland and situated in close proximity. The models were trained using three years of historical milk production data and were employed for the prediction of the total daily milk yield of the fourth lactation for each individual cow using a 305-day forecast horizon. The nonlinear auto-regressive model with exogenous input was found to provide higher prediction accuracy than the polynomial curve fitting model for individual cows using each input treatment. An improvement in forecast accuracy was observed in 62% of test cows (24 of 39). However, on average across the entire population, only part of the treatments delivered an increase in accuracy and the success rate varied between test groups. Prediction performance was strongly influenced by the cows' historical milk production relative to parity and also the prediction year. These results highlighted the importance of examining the accuracy of milk prediction models and model training strategies across multiple time horizons. Removal of the first lactation and applying static parity weight were shown to be the two most successful input treatments. The results showed that historical parity weighting trends had a substantial effect on the success rate of the treatments for both milk production forecast models.

An automatic model configuration and optimization system for milk production forecasting

The objective of this study was to develop and implement the Milk Production Forecast Optimization System (MPFOS) for the purpose of comparing the effectiveness of multiple herd milk yield prediction models for an Irish pasture-based dairy herd. The MPFOS was populated by nine milk production models that were categorized into three types: curve fitting, regression and auto-regressive models. The Adaptive Stratified Sampling Approach (ASSA) was introduced for data filtering, processing and for randomly selecting each member of the 100 cow sample herd. The MPFOS calculated optimal model parameters, statistical analysis and milk production forecasts for each chosen model using input data combinations based on animal, herd and milk production records. The model evaluations were based on historical milk production data between the years 2004 and 2009 from dairy farms in the south of Ireland situated in close proximity. Milk yield records from 2004 to 2008 were used for model training, whereas the milk production records for 2009 were set for model evaluation and validation. The ASSA randomly selected the representative herd population based on the required criteria. The MPFOS automatically generated the optimal configuration for each of the nine milk production forecast models and benchmarked their performance over a short, medium and long term prediction horizon. The Root Mean Square Error (RMSE) value of the nine prediction models varied substantially (from 68.5kg to 210.4kg per day). The surface fitting model performed better (10% in RPE and R2) than the dynamic NARX model for the same prediction horizon (365-day and 30-day). However, the NARX model provided more accurate results for shorter (10-day) prediction horizons. The MPFOS found the most accurate model based on prediction horizon length and on number of input parameters. The results of this study demonstrate the effectiveness of the MPFOS as a model configuration and comparison tool. The MPFOS may also be employed for selecting the optimal milk production forecast model for a specific application.

A semi-analytical mathematical model for predicting well performance of a multistage hydraulically fractured horizontal well in naturally fractured tight sandstone gas reservoir

Multiple fractured horizontal well (MFHW) is one of the key technologies to exploit low permeability reservoirs. The accuracy of productivity evaluation after fracturing not only affects the effectiveness of fracturing treatment, but also affects the optimum design of parameters in horizontal wells. This paper is the first investigation into the semi-analytical mathematical modeling for gas productivity calculation of multi-staged horizontal wells in fractured tight sandstone gas reservoirs that emphatically considers the impact of flow in fracture network (natural fracture system). Herein, a production forecast model of complex hydraulic fractures was created where the impact of nature fractures was simulated, and a semi-analytical solution was obtained by using Gaussian elimination method. As extension of historical work, this novel model could be applicable to more actual conditions such as fractures at arbitrary angles and unequal, asymmetric fractures.In the subsequent work, a numerical simulation of a simple conceptual model has been conducted, which validated the accuracy of this novel model. Then a real case application was successfully implemented. The modeling calculation result and actual production data of well X in Surig gas field is in good accordance. Besides, the impact of reservoir parameters upon production and morphology optimization were specifically analyzed. Using this new model, the production performance decline curves were drawn up based on computing results and were thoroughly analyzed for the sensitivity of reservoir parameters, especially the natural fractures. As presented in the performance curves, the post-frac production is greatly influenced by the change of permeability in fracture network, also confirming the applicability of this new model in fractured tight sandstone gas reservoirs.

Production Forecast of SAGD Considering Steam Loss Along Horizontal Well

Steam Assisted Gravity Drainage (short for SAGD, raised by Butler) is a successful thermal recovery method for extra heavy oil reservoir. Many scholars have finished much research on its productivity prediction, all of these work were based on the hypothesis that the distribution of steam along the injector is uniform and hence the height of the steam chamber is constant which is not reasonable. On the basis of Butler's oblique pane drainage theory and production prediction approach of SAGD, and the non-uniform distribution of steam in its horizontal steam chamber is taken into account, this paper will present a new SAGD production forecast model for horizontal well, which will consider the steam loss along its processing and will be further compared with the result of numerical simulation software that considering the wellbore divergence that to prove its correction. The simulation results indicate the height of the steam chamber along the horizontal wellbore is not constant, which is different from the uniform steam chamber hypothesis and coincident with the reality. The production prediction results of the simulation were compared with the calculation results of the new model, It is concluded that the SAGD forecast result that considering the steam loss along the gas injector is closer to the real situation.

Improved Production Forecasts and History Matching Using Approximate Fluid-Flow Simulators

Summary Forecasts of production with associated uncertainties must be based on a stochastic model of the reservoir variables and a fluid-flow simulator. The latter is usually very computationally demanding to activate. Approximate fluid-flow simulators based on upscaling of the reservoir variables are frequently used to assess the forecasts with uncertainties. The upscaling introduces biases and changes the error structures in the production forecasts, however. A production forecast model that accounts for these biases and changed error structures is defined, and estimators for the model parameters are specified. Forecasts both at the appraisal stage and the production stage including history matching are discussed. The so-called "ranking problem" is formalized and solved as a part of the study. The results are demonstrated and verified on a large case study inspired by the Troll field in the North Sea.