Integration of Production Data Analysis in Ensemble History Matching

One key limitation of conventional production data analysis (PDA) is non-uniqueness of solution. This can cause huge uncertainty in reservoir parameter estimates and hydrocarbon forecasts. Hence, there is need for an alternative approach to PDA. This paper tackles this issue using flowback data analysis (FDA) as a constraint to guide PDA. We recently extended the existing linear dual-porosity model (DPM) to develop a flowback analysis model (FAM). Here, we intend to i) test the accuracy of FAM to predict gas production in the field, and ii) investigate how the results from its application on flowback data could reduce the uncertainty associated with conventional PDA. This study applies FAM to history match 2-phase flowback data from three multifractured shale-gas wells completed in the Horn River Basin. The history match estimatesFAM estimates effective half-length and initial gas-volume in hydraulic-fractures before flowback. It is also used to forecasts post-flowback gas production. This forecast is verified by comparing it actual post-flowback production data. An independent single-phase PDA on post-flowback production data from the same shale-gas wells overestimates hydraulic fracture(HF) half-lengths compared to the FAM history-match results. This could be a result of the presence of water plugged secondary fractures with negligible communication interface with the hydraulic fractures during flowback. The interface significantly increases during post-flowback production periods as water gives way to gas from the matrix. One key result of this study is that HF half-lengths from PDA could significantly differ from those from post-flowback PDA. It is more appropriate to interpret both flowback and production data using a single multiphase model. FAM is an example of such models that captures the "necessary" physics for both flowback and production operations. PDA without considering flowback data could lead to inaccurate results. This study shows how to perform a robust flowback-data-analysis that complements PDA for comprehensive fracture/reservoir characterization. These results encourage the industry to start careful and accurate measurement of rate + pressure immediately after opening wells for flowback.

An integrated model-building exercise was conducted for Field P in the Malay Basin. A model-based estimate of a robust and realistic no-further-activity (NFA) forecast was essential for planning the effective implementation and monitoring of Enhanced Oil Recovery (EOR) processes. Field P has been in production for 40 years. Over 100 wells have been drilled through three target zones targeted for enhanced oil recovery (EOR), providing a clear structural understanding. Being a shore-facies deposit, strong connectivity was observed despite several north-south trending faults. Until now, production has been carried out through a total of 90+ strings, with some allowing commingled production across different zones. Analysis of probe-based pressure data during various drilling campaigns helped provide a qualitative mapping of the nature of shale barriers and their role in controlling water movement. Thorough analysis of production data in a spatial sense aided in understanding the overall recovery process, particularly the water encroachment pattern. Challenges such as the absence of relative permeability data, complex or unique pressure trends during production, poor cement bonds in wells, and commingled production resulting in inadequate allocation to zones were overcome through systematic analysis of reservoir and well performance. A combination of classical reservoir engineering techniques (such as data visualization and material balance approaches) and advanced modeling approaches, like assisted history matching, was utilized to generate multiple realizations of history matches. A total of eighteen uncertainty parameters were identified. The objective functions contain too many entities and parameters. Therefore, using a single objective function to eliminate uncertain parameters was considered misleading. The first phase of AHM focused on regional pressure matches to reduce uncertainty related to the aquifer, ensuring a reasonable match of the drive mechanism. The second phase of AHM concentrated on other parameters (such as relative permeability parameters, vertical permeability, and transmissibility across zones) to match oil rate and pressure data. These parameters were grouped under different objective functions. Having multiple independent objective functions provided a unique advantage in achieving simultaneous matching of these functions. This study contributed to achieving a robust and reasonable history match addressing the typical challenges of a mature field. This was done through the application of classical reservoir engineering techniques, well-designed visualization, preconditioning of static models with qualitative observations, and a customized approach in AHM. Most significantly, a systematic approach to the problem in phases and manageable pieces was adopted. This detailed and appropriately scaled approach greatly assisted the team in achieving a successful history match for the field. This approach could be replicated in many other mature fields.

The most common data that engineers can count on, specially in the case of mature fields is production data. Practical methods for production data analysis have come a long way since their introduction to the industry several decades ago and they all fall into two categories: Decline Curve Analysis (DCA) and Type Curve Matching (TCM). While Decline Curve Analysis is independent of any reservoir characteristics Type Curve Matching is a very subjective procedure. State of the art in production data analysis can provide reasonable reservoir characteristics but it has two major shortcomings: For reservoir characterization, the process requires bottom - hole or well-head pressure data in addition to rate data. Bottom-hole or well-head pressure data are not usually available in most of the mature fields. A technique that would allow the integration of results from hundreds of individual wells into a cohesive field-wide or reservoir-wide analysis for business decision making is not part of today's production data analysis toolkit. In order to overcome these shortcomings a new methodology is introduced in this paper that has three unique specifications: It does not require pressure data (bottom-hole or well - head); It integrates decline curve analysis, type curve matching, and numerical reservoir simulation (history matching) in order to iteratively converge to a near unique set of reservoir characteristics for each well; It uses fuzzy pattern recognition technology in order to achieve field-wide decisions from the findings of the analysis.

As the role of reservoir flow simulation increasingly impacts existing operations and field development decisions, it follows that rigor, fitness and consistency should be imposed on the calibration of reservoir flow models to dynamic data through history matching. Although a wealth of history matching techniques exist in the petroleum literature that propose novel algorithms or share case studies, seldom does the content guide the modeler in fit-for-purpose reservoir model calibration for an operating asset. To evaluate the applicability of these diverse techniques against standards required for reservoir management, an internal study was performed to benchmark four assisted history matching (AHM) techniques commonly promoted in the oil and gas industry. The techniques were vetted against a comprehensive suite of modeling requirements for multiple asset classes, integrating a variety of historical dynamic data types through the calibration of reservoir properties that control flow behavior from the field to inter-well scale. The methods benchmarked were: (1) Design of Experiments (DoE)-based, (2) Ensemble Kalman Filter and Ensemble Smoother, (3) Genetic Algorithm and (4) Generalized Travel Time Inversion. This manuscript focuses solely on the DoE-based technique.In order to consistently benchmark the techniques, a set of standards was defined against which each was evaluated to determine its suitability for widespread history matching applications. The standards involve: the capacity to parameterize (and therefore calibrate) a diversity of reservoir flow model attributes, the capacity to integrate different types of dynamic data, the level of independence from the flow simulator and the capability to provide probabilistic outcomes for predictive uncertainty assessment. Of the four techniques, the DoE-based approach uniquely satisfied all requirements. Its history matching workflow has the flexibility to incorporate any form of reservoir model parameter and to assimilate a history matching error metric for any individual or group of historical data types; therefore, benchmarking established DoE-based techniques as unambiguously the most compliant with generic asset modeling requirements. The approach was also identified as the most straightforward, both theoretically and in practical computation, and therefore applicable to the broadest range of practitioners. Perhaps most importantly, the approach demonstrated the capacity for accurate quantification of uncertainty (or non-uniqueness) in reservoir quality resulting from an exhaustive, although approximate, exploration of model parameter space and the associated history matching error metric(s).This manuscript compiles the results and insights gained from benchmarking of the DoE-based techniques through the proposal of a comprehensive assisted history matching workflow. The workflow is designed for generality while providing best practices that guide the modeler in fit-for-purpose application. Limitations of the workflow are also recognized. Key components include: selection and screening of calibration parameters based on statistical significance, development of surrogate models to characterize the relationship between parameters and the simulated historical data being integrated into the reservoir model, use of the surrogate models for exhaustive yet efficient exploration of parameter space to identify (non-unique) history matched models, and (deterministic or probabilistic) discrete reservoir model selection for use in forecast-based decision making. Each step of the AHM workflow is presented from a conceptual and applied perspective, and field applications are provided to demonstrate key concepts. Although the applications presented include two deepwater Gulf of Mexico assets, the workflow and insights provided are developed from benchmarking across a diverse suite of asset types.

First and foremost, production analysis techniques require accurate rate and bottomhole pressure histories. In most cases the pressure history of the well is not measured directly at the bottomhole condition, but is calculated from surface measurements by the use of single or multiphase flow correlations. In some cases significant error is introduced through the use of these correlations. This paper evaluates the magnitude of such errors for oil and gas producers with regard to the estimation of flow capacity, completion efficiency, and effective drainage area. Synthetic cases are used as control sets in order to evaluate the sensitivity of the results to the various multiphase flow correlations and flowing conditions. In addition to synthetic (simulated) performance behavior, field cases are presented and the variance in estimated reservoir and completion properties is evaluated. The technical contributions of this paper are: Systematic evaluation of the effect of errors in flow rates and bottomhole flowing pressures on production data analysis — using both synthetic and field derived well performance data. Qualitative guidelines as the effect of errors in rate and pressure on estimated reservoir properties. Establishment of the most relevant pressure drop correlations for use in practice.

Relative permeability is one of the most important properties in Coal bed methane/gas (CBM/CBG) fields for reliable reservoir modelling due to two-phase fluid flow. In CBM reservoirs, one of the biggest challenges remains with the high unreliability of laboratory experimental relative permeability data due to the inherent nature of coal. The scope of this study is to define the shape of the relative permeability curve using available production data analysis. The interplay between water and gas within the coal cleats affects the mobility of both water and gas. In Sohagpur CBM blocks of India, there is enough production data associated with wells. In this study, a comprehensive production data analysis (PDA) was done on wells to quantify the relative permeability curve. Though gas and water rates are available from the production history, it becomes essential to determine the well static pressure at different times so that the pressure-dependent gas and water properties can be evaluated. Conventional Z function is modified by King (1990) for CBM to include sorption behaviour and known as Z* function which, when used with pressure and plotted with cumulative gas production (Gp), gives a straight line like conventional gas reservoir behaviour. The plot gives static well pressure for any known Gp in the production history of the well and gas water properties can be determined at every pressure using standard correlations. Thus, the relative permeability ratio of gas and water krg/krw can be obtained at every pressure using the relationship described above. It is found that the application of material balance study which, when combined with Production Data Analysis (PDA), gives useful information on relative permeability. This helped quantifying relative permeability which is used for reservoir modelling and simulation. History matching exercises are carried out for the wells for part of the available production data and forecasts are generated for rest of the production history. The forecasts of both gas and water are matching with the history which acts as validation points and provides good confidence on reliability of the generated relative permeability curve. Further, forecast period of the existing wells are extended to more years and sensitivities are carried out with additional infill and/or new to optimize further development planning. The present study brings uniqueness in the form of two innovative workflows integrating production data, compositional data, and pressure. This has been used along with existing conventional knowledge of material balance with some modification for coal reservoir adsorption phenomena which adds considerable value to the existing knowledge. These workflows can be applied in other reservoirs or basins for more reliable reservoir modelling and forecasting which may help both CBM evaluation and exploitation time and cost effective and optimize the field development.

Effective history matching of a simulation model with real data from the field requires the resolution of three outstanding issues. First, a conflict may exist between the production data and the existing geological model built from static information. Second, during model updates, geological consistency must be maintained by honoring the prior geologic information. Third, uncertainties related to mechanical and operational issues must be taken into account.;The objective of this study is to build a realistic simulation model able to predict future behavior of a complex CBM reservoir for a CO2 sequestration pilot project in Marshall County, WV by using a systematic approach for history matching. Our study is limited to building a history matched model for the Methane production.;As a general overview, the field project covered in our work includes two known coal seams, Pittsburgh coal (upper layer) and Upper Freeport coal (lower layer). Production is from six horizontal wells, two of them completed in Pittsburgh coal and the other four completed in Upper Freeport coal. This project is part of the national plan of CO2 sequestration with the goals of improving the quality of the environment and increasing methane production from coal.;In order to accomplish the objective of this study, a systematic history matching approach was devised. In this approach, data from the field are gathered and analyzed in order to be used as input in a commercial reservoir simulator software. Then, effective strategies and key reservoir parameters are defined to realistically history match the field production. The resulted approach propose a new systematic methodology for performing history matching complex reservoirs. This approach accounts for production data inconsistency related to reservoir characteristics heterogeneity, mechanical, and operational issues within different scenarios.;The systematic approach included a detail analysis of the field production data in order to learn as much about the reservoir behavior as possible. The analysis included a normalized production comparison done based on the amount of gas produced in a per-foot of the well, in a per-foot of thickness around the well, and based on the average gas content. This information will be used extensively in the history matching process, which included sensitivity analysis of conventional parameters (permeability, porosity, relative permeability, and water-gas contact), unconventional parameters (Gas content, desorption time, and Langmuir parameters), and finally wellbore parameters (skin factor, bottom-hole pressure, and well length). Using the results of the sensitivity analysis above, an analysis table is formed to be

Production data analysis of unconventional gas wells: Review of theory and best practices

Rate normalized pressure (RNP) and derivative (RNP') method uses the log-log plot of RNP and RNP' versus material balance time (te) to interpret formation and hydraulic parameters like permeability and fracture half-length. It's a very useful methodology and sometimes the only way to conduct transient rate and pressure analysis for shale gas wells because pressure buildup is rarely recorded. However, the method may be not applicable because material balance time is not a monotonic increasing function of time especially when the one well has a short production history and has big vibration of production rate. A new method is proposed to solve the above problem for shale gas wells. In the new method, the average material balance time (cumulative production (Np) divided by average gas rate) is used to replace te. Therefore, the increasing monotonicity with production time is guaranteed. Curves of te vs. production time and average te vs. production time are contrasted. Different field cases with smooth decreasing gas rate and the cases with sharp change rates are used to check the applicability for the proposed new method. The plots of RNP & RNP' vs. average te, RNP & RNP' vs. te are drawn for different shale gas wells data for accuracy comparison. Results show that the new method has similar type-curve with the RNP & RNP' vs. te method and can be used to interpret the formation and hydraulic parameters by diagnosing typical flow regimes for shale gas wells. Compared with the RNP & RNP' vs. te method, the new method theoretically improved the accuracy production date analysis because the x axis is a monotonic increasing function of production time while not showing inflection points in RNP & RNP' vs. te plots. Field case with smooth decreasing gas rate shows that the new method and RNP & RNP' vs. te method can both get satisfactory log-log plot match and rate & pressure history match. Although the RNP & RNP' vs. average te may give smaller SRV volume, the interpretation results for both methods are comparable. The shale gas case with sharp vibrated gas rate shows that the new method can get both better log-log plot match and rate & pressure history match, in return, the interpretation results may be more confident. The RNP & RNP' vs. average te method may improve the production data analysis (PDA) accuracy when the shale gas wells have huge rate vibrations which may be caused by fluid flow-back, choke size change, temporary well shut-in and etc. However, more shale gas wells data should be used to test the method.

Reservoir simulation models should always be built for specific business goals. It is an accepted rule that models used for production forecasts should reproduce the production on history. Although, most history matching processes are often the result of a complex team effort, objectives for using simulation models and the required level of detail are quite diverse. Applications range from prospect evaluation with limited available calibration data to designing detailed production planning scenarios for mature fields with highly constraining well production data. In either case, applied techniques, workflow requirements and the level of complexity will naturally differ. In recent years assisted history matching techniques and optimization workflows have been established and included in best-practice guidelines in an increasing number of companies in the oil and gas industry. The application of assisted history matching techniques is often motivated by the need to handle increasingly complex problem statements as well as the desire to improve workflow efficiency and transparency. Initially, the focus was given to finding single best models. Modelling paradigms, however, are changing. More recently, the industry has given a stronger interest to understanding a distribution of alternative scenarios which more realistically captures the uncertainty-envelope. This step is non-trivial, since there is no natural extension from the paradigm of single best history-matched models with deterministic forecasting capabilities to the paradigm of establishing a distribution of alternative production forecasts. This defines a major challenge to the reservoir engineering workflow and the question of handling multiple models with alternative outcomes. This talk reviews selected techniques used in history matching workflows. It discusses practical considerations for finding a compromise between “accurate” history-matched models with deterministic forecasting capabilities and the newer paradigm of a sufficient coverage of the uncertainty space for establishing uncertainty distributions.<br&gt

This paper presents a field case history of the integrated analysis and interpretation developed using continuously measured rate from production ticket and surface pressure and static reservoir pressure data from pre-tertiary basement gas reservoir in South Sumatra, Indonesia. The primary objective of this work is to estimate areal distributions of flow properties (effective permeability and skin factor) as well as volumetric properties (original gas in place, gas reserves and reservoir drainage area – per well basis). The outcomes are then used to evaluate the effectiveness of well spacing and to identify additional potential area for infill well. Throughout this study, a type curve approach is used to analyze production data, which enable us to perform real-time evaluation and monitoring of the well performance. This approach is cross-validated with conventional material balance analysis. Combination of the two analyses should provide more consistent results for characterizing the pre-tertiary basement gas reservoir. In addition, the evaluation of production data can identify the occurrence of possible well interference.

Activities in exploitation and development of unconventional resources have grown tremendously in recent years. Production is further improved by applying multi-stage hydraulic fracture stimulation in horizontal wells from tight oil and gas reservoirs. Due to complex mechanical behavior including rock deformation and fracture closure during the development of multi-stage hydraulically fractured horizontal wells (MFHW), geometrical fracture parameters in MFHW may partially change. In this paper, a method of determination of fracture properties based on production data analysis combined with numerical simulation of multi-stage hydraulic fracturing in horizontal wells is presented. Accordingly, an analysis procedure is provided to determine fracture properties in the stimulated reservoir volume with two field cases from a tight gas reservoir and a tight oil reservoir. The results prove that the proposed method works well in determining fracture properties by analyzing production data.

Production data analysis is a viable tool for reservoir characterization and estimation of initial gas in place (IGIP) and reserves. Several methods are available to analyse production data starting with Arps classical decline curve analysis (DCA) in 1945 all the way to more sophisticated analytical and advanced DCA techniques. Most of these methods are applicable only for single phase flow in porous media. In this paper, we present a simple analytical decline curve analysis (ADCA) model that takes into account the effect of water influx on gas reservoir performance. We introduced the water influx effect into the pseudo-steady state flow equation which enables us to estimate the reservoir pressure and the IGIP for water drive gas reservoirs. The model is based on coupling the material balance equation for gas reservoirs, aquifer models, and the gas flow equation to calculate the well’s production rate versus time. The model can also estimate reservoir pressure, gas saturation, water production rate, and gas production rate with time. When the model is run in history-match mode to match gas and water production, we can estimate the IGIP, well’s productivity index, and aquifer parameters. The model can also be run in prediction mode to predict gas and water production at any conditions of bottom-hole flowing pressure (BHFP) (or surface tubing pressure) and reserves can be calculated. The model was validated with several simulated cases at variable conditions of rate and pressure. The model was then used to perform decline curve analysis in several field cases. This technique is fast and requires minimum input data. The paper will also present the application of this technique to analyse production data and predict reserves for gas wells producing both gas and water.

An iteration-free approach for determining the average reservoir pressure and original gas in place by production data analysis: Methodology and field cases

An efficient deep learning-based workflow for real-time CO2 plume visualization in saline aquifer using distributed pressure and temperature measurements