Coalbed Methane Reservoir Simulation: An Evolving Science

The demand for indigenous supplies of natural gas in the UK and other countries in Europe has driven operations in the unconventional gas arena since the 1990s. In Coal Bed Methane (CBM) there have been attempts in the UK, Belgium, Germany, France and Poland. Dart Energy and its heritage company, Composite Energy, have been active in CBM in the UK since 2004, drilling 25 CBM wells in the UK, including 10 appraisal and development wells on the Airth field in central Scotland. A further 4 appraisal wells were drilled by the previous operator of the Airth field in the 1990s. The local Carboniferous coal geology, as with most European coals of Carboniferous age, is characterised by thin, numerous, low permeability, undulating coal seams. A number of different well designs were tried over the 14 Airth wells to meet the challenges of the local geology: initially vertical fracture stimulated wells; moving to geosteered multi-lateral horizontal wells, either intersecting a vertical well at the end of the horizontal section or as ‘updip’ branches off a motherbore without an intersection and finally multilateral geosteered horizontal wells intersecting a vertical well at the start of the horizontal section. The evolution in well design incorporated learnings from drilling operations, reservoir geology and production operations for each type of well architecture and advances in drilling technology in other CBM provinces around the world, adapting them to answer the particular subsurface problems encountered in the Airth field. Eventually, through this evolutionary learning process, Dart Energy was able to announce a commercial flowrate of 0.7MMscf/d from Airth 12 in January 2013, a multi-lateral horizontal well with a vertical well intersection at the start of the horizontal section. This paper describes the journey to that success, charting the evolution of CBM production well design on the Airth field, recognising the geological factors that drive well design, an evolution that can be applied to Carboniferous coal systems with thin, numerous and structurally complex coal seams found in other parts of the UK and Northern Europe.

C oalbed methane (CBM) is well recognised as an important natural gas resource. However, producing it commercially has proven to be a challenging process, primarily due to the complex nature of its gas storage and flow mechanisms. Although CBM reservoirs are naturally fractured due to the presence of face and butt cleats, gas stor­age and saturation depend on the adsorption capacity of the coal matrix. The ability to predict permeability variations and its evolu­tion in a CBM reservoir, which is affected by several dynamic CBM properties and operating conditions, also define the design, optimization, and analysis of the gas recovery process. As the permeability is naturally low, interventions are necessary in order to achieve commercially viable gas production rates. CBM are also dual-porosity media where the vast majority of the gas is stored in the low permeability coal matrix by sorp­tion. The flow to production wells, however, occurs through the coal’s natural fracture system, that stores relatively small amounts of gas, because coal matrix practically has no perme­ability. The result is that properties of the coal matrix have the greatest effect on the estimates of gas in place and recovery. It is well understood that the key parameter controlling gas flow in CBM reservoirs are cleat networks, gas storage, desorp­tion and permeability anisotropy (coal shrinkage and swelling). While the role of reservoir simulation is well under­stood for conventional oil and gas reservoirs and its related techniques for prediction and production optimisation, for unconventional reservoirs (CBM, Shale, etc.) current reservoir simulation software needs to be further developed and adapted to represent the performance of these types of fields and gener­ate more acceptable results. For example, the basic governing fluid flow regime in conventional reservoir simulations is limited to Darcy flow. For unconventional reservoirs, however, another flow regime term comes into the equation that reflects the flow of gas in the coal matrix and stands for diffusion. This new flow regime is not very well defined in the current simulators and needs further studies to appropriately address the production performance of a reservoir. Furthermore, the concept of multi-phase flow in reservoirs is well defined by the introduction of relative permeability equations in the flow equations. However, measuring relative permeability in coal is relatively complex (Ham and Kantas, 2008) due to its friability, heterogeneity, stress dependence and porous morphology. Young et al. (1992), for example, noticed that the derived relative permeability from a history match does not resemble the laboratory measurement. This phenomenon could be due to the change of matrix structure on phase flow in addition to wettability and capil­larity. In other words, both phases may flow through different passages over time due to heterogeneity. Thus dynamic rela­tive permeability (phase saturation) is required to accurately predict recovery from a CBM reservoir. As the focus on CBM has increased in the past decades, so has the quest for more effective technologies to help bring the reserves in such complicated conditions into production and to identify sweet spots. Reservoir modelling and reservoir simulation are technologies that are likely to play a key role in achieving this. Firstly, however, it is necessary to look at the challenges of CBM property variations.

Thermal performance of an aquifer thermal energy storage system: Insights from novel multilateral wells

Multi-lateral wells have been increasingly used in recent years by different industries including oil- and gas industry along with coal bed methane- and water production. The common purpose of these wells is to achieve a higher production rate per well. More and more sophisticated well patterns and geometries can be implemented in practice which calls for improved modelling techniques. Complicated well geometries and small lateral diameters require high resolution models in the vicinity of the wells. With structured finite difference grids this can only be achieved by unnecessary refinements even far away from the wellbores. However the model may still suffer from orientation problems if laterals do not coincide with the rows or columns of the rectangular mesh.In the present work, we applied unstructured grids to model multi-lateral wells and compared the results to structured models. We used the MODFLOW-USG code, which simulates groundwater flow using a generalized control volume finite-difference approach, allowing grids other than orthogonal structured grids to be applied. This offers a solution for orientation and resolution problems. The second part of the paper aims to optimize multi-lateral well geometry by evaluating the effect of length, angle and number of laterals.

The primary objectives of this study are to design a gas injection pilot in the Eagle Ford and to estimate the benefits of gas injection under different operational scenarios. This pilot design study entails the construction of multiple reservoir simulation models to understand the hydraulic fracturing and flow dynamics of multiple wells and gas injection operations in the Eagle Ford. Two DSUs with multiple hydraulically fractured wells were studied to achieve the proposed objectives. One of the DSUs was identified as the main study area to design a huff-and-puff gas injection pilot. Having an existing gas injection operation, the other DSU was selected to improve our understanding of the physics associated with gas injection. A dual porosity numerical reservoir simulation model coupled with geo-mechanics was built to replicate the historical well performances of the pilot area using a sophisticated numerical reservoir simulator. Another dual porosity simulation model was constructed to assimilate the existing huff-and-puff performance of the second DSU in which data was only publically available. The methodology used in this study integrates the hydraulic fracturing process, multi-phase flow, geo-mechanics, and proppant transport within the reservoir simulation. The simulation model was calibrated to match the historical hydraulic fracture treatment, fluid flow back and post-stimulation production. The proppant entrapment and migration from child well to the parent well was captured. The calibrated simulation model was then utilized to design a huff-and-puff gas injection pilot. Learnings and observations obtained from modeling of the existing gas injection operation in the second DSU were integrated into the pilot model. Additional sensitivity runs were performed to examine the potential benefits of gas injection under different operational scenarios. The calibration results indicated that the stimulated rock volume geometries of pilot study wells vary based on their completion practices. The historically observed well interference and frac hits between parent and child wells were captured by establishing a proper connectivity between wells during calibration. Proppant entrapment and movement of the proppant impacted the well performance. The results showed that significant amount of depletion leads to considerable matrix permeability reduction around wells. The most important knowledge gained from the calibration of the second DSU with huff-and-puff data is the identification of reservoir model characteristics that have the largest impact on the huff-and-puff performance. This study allows us to identify opportunities to design and improve huff-and-puff operation as well as estimating benefits of gas injection under different operational scenarios. The utilized technology in this study is unique and novel as it solves the geomechanics and flow in a single process. Proppant flow and entrapment was captured successfully. The multi-well calibration of the simulation model provides physics-based explanations for the historical well performances in the Eagle Ford.

Modeling fluid flow around horizontal wellbores with conventional reservoir simulators is somewhat inefficient due to the fact that the radial flow patterns found in the vicinity of the wellbore must be simulated with rectangular coordinate grid-blocks. In this research, we present an efficient hybrid grid model for horizontal well simulation by incorporating radial gridblocks to simulate the radial fluid flow around the wellbore and introduce the concept of spherical gridblocks to model fluid flow around the end of the well. This configuration can dramatically reduce the number of grid-blocks (and therefore computer time) required to accurately simulate the physical model. The spherical gridblocks consist of hemispheres, corresponding to each radii in the cylindrical coordinate system region, which are intersected by planes that allow coupling to the reservoir region gridblocks. Transmissibilities for the spherical gridblocks are calculated by use of the integrated finite difference method. Two problems from the literature were chosen to demonstrate the effectiveness of the model. The first problem is a one-phase, constant rate pressure drawdown test for which an analytical solution is available. The second problem is a water cresting problem where the results for comparison were obtained from a commercial reservoir simulator using a grid that reportedly minimized numerical dispersion.

Multilateral wells require a sophisticated type of well model to be applied in reservoir simulators to represent them. The model must be able to determine the flow rate of each fluid and the pressure throughout the well. The production rate calculations are very important because they give an indication about some main issues associated with multi-lateral wells such as one branch may produce water or gas before others, no production rate from one branch, and selecting the best location of a new branch for development process easily. This paper states the way to calculate production rate of each branch of a multilateral well-using multi-segment well model. The pressure behaviour of each branch is simulated dependent on knowing its production rate. This model has divided a multi-lateral well into an arbitrary number of segments depending on the required degree of accuracy and run time of the simulator. The model implemented on a field example (multi-lateral well HF-65ML) in Halfaya Oil Field/Mishrif formation. The production rate and pressure behaviour of each branch are simulated during the producing interval of the multilateral well. The conclusion is that production rate of the main branch is slightly larger than a lateral branch.

The Provost Dina 'X' Pool is a turbidite sand with a pay zone of up to 20 m at a depth of 200 mSS. The field contains approximately 1.2 million m3 of oil and there is a bottom water aquifer with partial pressure support in parts of the field. Most of the vertical wells in the field shut-in due to high water cuts after a relatively short period of production. Therefore, a numerical simulation study of the whole field was undertaken in order to improve recovery by infill drilling as well as by investigating different operating strategies. This paper describes the integrated team approach taken with input from the geological, production and reservoir engineering specialists to construct a reservoir model. The model has been successfully history matched and used, in turn, for predicting the performance of the field under existing conditions as well as a variety of infill drilling and operating scenarios. Results indicate that significant water coning, which would be observed with new vertical wells in certain areas of the pool, can be reduced through horizontal infill wells. Despite the potential technical success of horizontal infill wells, they were not economically feasible in this particular reservoir and vertical infill wells were the preferred choice. It is shown that a combination of carefully planned vertical infill wells and a pressure maintenance program is able to boost the ultimate recovery by 34% while keeping the development costs to a minimum. Results of various operating strategies are presented and the optimum field development scenario implemented in the field is outlined. The initial actual reservoir behaviour is discussed in the light of the results from the simulation study. Introduction The Provost Dina 'X' pool was discovered in 1988 at an initial pressure of 6450 kPa. The reservoir is located in east central Alberta, south of the town of Consort. The formation is a turbidite sand with a limited bottom water aquifer support. The production from the pool began with the discovery well 05-15-35-06 W4M and, to date, a total of 11 wells have delineated a reservoir area of slightly more than 32 ha. (see Figure 1). Historically, most of the vertical wells in the field experience high water cuts after a relatively short period of production. Thus, a significant portion of the reserves cannot be produced because of water coning. Infill drilling can compete favourably with BOR processes for the recovery of the bypassed oil for much less investment and operating cost. As the existing vertical wells have to be operated above the critical water coning rates to be economic, horizontal wells become another option for infill drilling. The field experience indicates that horizontal wells can reduce the water coning tendency while increasing the oil production. However, further improvement to oil production rate of horizontal wells is still limited by the encroachment of the water cone when the bottom water exists. Therefore a simulation study was undertaken to investigate the feasibility of increasing reserves by infill drilling with horizontal and vertical wells. Model Construction The black oil simulation module (SimBest II) of The Petroleum WorkBench reservoir management package of Scientific Software-Intercomp was used to conduct this simulation study. SimBest II is a three-dimensional, fully implicit reservoir simulator which is capable of handling as many as three fluid phases and five components in a dual and/or single porosity mode. P. 555

Integration of geologic, geophysical and petroleum engineering disciplines has facilitated a reservoir characterization of Cedar Hill field, a naturally-fractured coalbed gas reservoir. A 3-D multicomponent (3-D, 3-C) seismic survey was collected over a portion of Cedar Hill field. The seismic survey was successful in defining detailed structural features in addition to identifying zones of high fracture permeability within the reservoir. By incorporating the results from the seismic interpretation with the detailed vertical mapping of the reservoir into a simulation model, improved reservoir exploration and exploitation can be achieved. Reservoir simulation techniques verify faults and areas of increased reservoir pressure in the study area. These areas of higher pressure indicate blocks of the reservoir which are not currently adequately drained by the present well spacing. Three-dimensional reservoir flow simulation used in combination with 3-D, 3-C seismic gives a more complete description of this coalbed reservoir and may provide a cost effective alternative in the early development of coalbed gas fields. Introduction Cedar Hill field is located 25 miles northeast of the city of Farmington in San Juan County, New Mexico, U.S.A. (Figure 1). It is the oldest producing coalbed gas field in the San Juan Basin and has an aerial extent of approximately 15 square miles. Production is extremely variable ranging from 0.1 to 5 MMSCFGPD per well. Wells of greatest productivity produce from zones of high fracture permeability in the Fruitland coal. Due to the high compressibility of the coal, natural fractures or cleats are generally closed at reservoir depths. Local stress relief by faulting, folding, and over-pressuring can cause fracture enhanced permeability in the coal. A portion of the western edge of the field, centered on Section 30, T32N, R10W, was selected for a detailed study. This area was chosen to include a portion of the basin which lies between two known pressure regimes, a high pressure area and a transitional pressure zone. A three-dimensional (3-D), three-component (3-C), seismic survey was acquired from October 25 - November 6, 1991 over Section 30. The area used for engineering analysis and reservoir simulation was enlarged beyond the seismic survey area to eliminate boundary effects in the primary area of interest. Data Acquisition A large quantity of geological, geophysical, and engineering data has been acquired in Cedar Hill field. The Hamilton No. 3 well (SW1/4 Section 30) has one of the most complete suites of well logs and cores available within the field. Geophysical data collected include a 3-C Vertical Seismic Profile (VSP) in the Hamilton No. 3 Well and a 3-D, 3-C seismic survey encompassing 1- 1/4 square miles, with a bin size of 55x55 feet (Figure 2). Three-component receivers as well as compressional (P) wave and two orthogonal shear (S) wave vibrator sources, were used in the acquisition. Amoco began production at Cedar Hill field in May, 1977. Since then, 23 production wells have been completed in the field. The eastern portion of the field was drilled on irregular spacing until 1984 when regulations required that wells be drilled on 320-acre spacing with the first well drilled in the NE1/4 or SW1/4 of a section. Thirty-one wells are completed in the Fruitland coals in the study area. While there is a long documented production history, there is limited pressure data available from the Fruitland coals. A total of 46.6 BSCFG has been produced from the study area through December 1993. P. 483

This paper describes application of Project INTERSECT, a next generation highly scalable reservoir simulator on real large scale field models. High resolution reservoir simulation is required to better define and describe fluid flow and enable improved field development and tactical operational planning. Massively parallel computing techniques overcome limitations of problem size and space resolution. This paper demonstrates that large-scale simulation models can be performed on commodity hardware taking advantage of evolution in multicpu hardware architecture and software engineering. This allows both geologists and reservoir engineers to include more realistic geologic and engineering detail for better and more reliable production optimization. Intense computer simulation is essential for effective reservoir management. The advances in reservoir characterization techniques and the industry drive towards the ‘smart oilfield’ with rapid model updates will require more efficient model processing to achieve timely field operational decisions. Parallel reservoir simulators have the potential to solve larger, more realistic problems than previously possible. The size and application of reservoir simulation problems have been limited by the availability of computing hardware, reservoir simulation architecture and of solution methods for solving large-scale heterogeneous problems. The next generation reservoir simulator demonstrates that key modeling challenges has been overcome by a software architecture and capability to model more realistic subsurface and surface models. Applications of the new reservoir simulator illustrates how typical reservoir engineering options such as local grid refinement, local grid coarsening, multilateral wells and aquifer modeling affect the overall parallel performance and scalability using highly heterogeneous large-scale models. Application of new modeling techniques highlight increased accuracy of modeling results and more reliable field development planning and reservoir management decisions.

Production characteristics and reservoir analysis of coalbed methane reservoirs

Recently Imperial Oil Limited has been pursuing gas plays in North-Eastern British Columbia where single wells have the potential to recover gas in the 30-300 Bcf range. The tie-in costs for these wells are very high due to acid gas content, mountainous terrain and long distances to tie-in points. A reliable prediction of the recoverable reserves helps determine if a tie-in is economic and manage the business over the long-term. The intent is to demonstrate how the long-term recovery from naturally fractured gas wells can be accurately evaluated with buildup tests and a conventional finite difference reservoir simulator. The decision to tie in a well is strongly based upon its initial deliverability. However, this is not generally appropriate since the initial deliverability and the longterm recovery are controlled by different factors. The initial deliverability is a function of the fracture system around the wellbore while the long-term recovery is related to the fracture system's extent and the ability of the reservoir matrix to recharge the fracture system. Other evaluation techniques such as open hole logs and core can only evaluate the fracture system at the wellbore which again only relates to the initial deliverability. Properly planned and executed production tests, in conjunction with reservoir simulation, can provide accurate evaluations of long-term recovery potential. However, the current suite of dual permeability/porosity simulators are either too simplistic to be of any use or very complex and difficult to use. Considerable lime and money could be saved if a conventional finite difference reservoir simulator could be used. Two example wells will first be presented to demonstrate how a conventional reservoir simulator can be used to model the long-term production from a naturally fractured reservoir. It will then be demonstrated how a properly conducted flow and buildup test used in conjunction with a conventional simulator can be used to predict the longterm deliverability.

A significant percentage of hydrocarbon reservoirs around the world is fractured. Moreover, the major part of gas reservoirs in Iran is also fractured type, so the existence of an in-house software is necessary. In this study, an efficient, user-friendly, and indigenous simulation of a three-dimensional black oil fractured dry gas reservoir has been developed through IMPES method with the two-phase flow of gas and water. The presented simulator, which was written by C++ language and was known as fracture dry gas reservoir simulator, uses the implicit pressure and explicit saturation method for solving the equations. Also, effect of gravity pressure is neglected and effect of the capillary is considered in equations. By this simulator, we can investigate the dry gas reservoirs behavior with fractures. Darcy or non-Darcy fracture and matrix flow, Cartesian, cylindrical, and combination of Cartesian–cylindrical reservoir gridding, single porosity, dual porosity–single permeability, and dual porosity–dual permeability modeling are abilities of this simulator too. Additionally, this simulator is able to make outputs (such as pressure) at any given specific radius and time interval as numerical and/or graphical output in so little run time. Also, this simulator has PVT box and gridding box for doing the calculation of PVT and gridding. PVT box contains new correlations and EOS in comparison with another reservoir simulator. Gridding box makes us be able to simulate fractured dry gas reservoirs and hydraulically fractured well reservoirs too. Finally, the validity of this simulator was verified by comparing the simulation results with the other reservoir simulator (Eclipse) and showed a good compatibility between the developed software and Eclipse results in each time with different conditions such as various gridding conditions, various fluid data conditions and also various well configuration conditions.

Development of a reservoir simulation model involves collection and analysis of an enormous amount of reservoir and well data for initial input. Computer graphics is the most effective tool for geologists and engineers to visualize, verify, and analyze the input data. After each simulation run, simulation engineers then have to depend heavily on computer graphics to analyze an even more intimidating amount of simulation results data. This is especially true at Saudi Aramco because of the large simulation models used to simulate the world's largest hydrocarbon reservoirs. Saudi Aramco's EXPEC Computer Center supports a wide spectrum of computer graphics applications to aid the exploration and engineering staffs through the entire cycle of reservoir characterization. The graphics applications help during the pre-simulation model data analysis as well as in the post-simulation results analysis. Larger and more accurate simulation models are currently being developed In Saudi Aramco. Computer technology (both software and hardware) needs to advance continually such that computer graphics can be utilized as an effective and efficient engineering tool by reservoir simulation professionals. This paper addresses Saudi Aramco's experience in dealing with very large simulation models and using computer graphics as an engineering tool. The program efficiency, device dependence, and data management problems will be discussed. The computer speed, data storage and access methods, data transmission speed, and graphics devices will also be detailed.

This paper presents results of a reservoir and near wellbore modelling study to assist with the evaluation of inorganic scale risk management for a cold sandstone reservoir in the Norwegian sector of the North Sea. Reservoir temperature is in the range 32-48°C and poses a challenge to manage the scale risk. The field is heavily supported by aquifer water, and the development includes five production wells and seven seawater injection wells. The small number of high-cost, highly productive subsea multilateral wells with long horizontal sections, some with sliding sleeves to enhanced flow management, coupled with a high barium sulphate scaling tendency (up to 350ppm barium concentration in the formation water) requires effective scale management. This paper addresses questions of timing and nature of seawater breakthrough, when scale inhibitor squeeze treatments will be required, what is the impact of any reservoir reactions between the formation water and injected seawater on the scale risk at the production wells, what squeeze placement profiles look like under a range of squeeze treatment scenarios (various injection rates and sleeve position options). The treatment of multilateral wells by bullhead scale squeeze treatments is not a subject extensively covered in the literature, so the focus of this paper is to understand the impact on squeeze treatment life of bullhead deployment as a function of pump rate with both laterals open and the impact on treatment lifetime and treatment cost of selective placement into each lateral by treating one lateral at a time via activation of the sliding sleeve mechanism. The paper draws on a wide range of technical inputs to make scale management decisions including: thermodynamic modelling techniques, reservoir simulation of fluid mixing and reaction, laboratory-generated coreflood data to assess chemical selection for scale inhibitor squeeze, and squeeze treatment design drawing on all the information outlined above. Sharing the scale management technical challenges associated with such low temperatures and treating multilateral wells provides an example of the factors that should be taken into account during project design in similar fields under development in Northern Norway, offshore West Africa and offshore Brazil.