Description
CHAPTER ONE
1.0 LITERATURE REVIEW
1.1 Introduction
It is common to consider a coalbed methane reservoir as a dual porosity or naturally fractured reservoir. A coalbed methane reservoir is a naturally fractured reservoir with a coal matrix that has the potency as methane gas storage. Storage mechanism in a coalbed methane reservoir could be explained by an adsorption process. An adsorption process enables gas to be attached on the internal surface area of the coal matrix. On the contrary, with a desorption process, methane gas is released and gas drainage occurs, which allows gas to be transported through permeable media or a fracture system. A fracture system in the coalbed methane reservoir is strongly related with the cleat system. The fracture system, in this case the face cleat and butt cleat, acts as a porous medium and cause reservoir anisotropy. The face cleat is more continuous and longer than the butt cleat, and it tends to exist continuously through the reservoir body. On the other side, the butt cleat is a perpendicular fracture that is shorter and discontinuous. The butt cleat is discontinuous because during natural fracture formation it is intersected by face cleats. Face cleats tend to be more continuous because they are first-formed fractures and are more systematic. The butt cleat is a secondary natural fracture system and is less systematic during its development than the face cleats, so this natural fracture system contributes to the reservoir anisotropy. The face cleats also provide a larger interface area with the matrix system than the butt cleats do. This phenomenon makes the face cleats more important in the fluid flow mechanism. It is common to assume the face cleat direction as the maximum permeability direction. However in some cases, this is not a correct assumption, such as in the case of Bowen basin, Australia11.
The storage capacity of a coal matrix can be considered as a economic resource; however the coal matrix permeability is very low. The coal fracture system, particularly the cleat system, provides media for fluid flow in the coal system. The cleat system contributes to overall formation permeability. Methane gas resulting from desorption process flows through the cleat system or natural fractures into the wellbore. The permeability anisotropy is related to the formation of face cleats and butt cleats, in this case the anisotropy creates a preferential flow. It is more common to find the maximum permeability orientation parallel with the face cleat direction. Furthermore, the drainage pattern will also be determined by permeability anisotropy.
1.2 OVERVIEW OF THE STUDY
Coalbed methane is an unconventional natural gas in coal seams. Global CBM in place is estimated to range from 36 to 230 × 1012 m3 (Ma et al., 2015). The major component of CBM is methane, accounting for 80 to 95% (Thakur et al., 2016). CBM is a kind of clean energy with high calorific value. The combustion heat of coalbed methane is equivalent to 1 kg of fuel, while the pollution caused by coalbed methane combustion is only 1/40 of oil and 1/800 of coal (Keyong, 2011). Exploration and development of CBM are of great significance for energy sustainable development and has attracted a lot of attention.
During the development, CBM well first continuously produces water at a low bottom hole pressure to depressure the reservoir. When the reservoir pressure drops below the critical desorption pressure, methane is desorbed from the coal seam and transmitted to the well; the gas rate increases, and the water rate drops. The performance of CBM well is directly affected by its production schedule. Coal seams are generally heterogeneous with widespread cleats, fractures, and matrix pores. The matrix permeability and porosity are low. The pore volume compressibility of coal rock is 1 to 2 orders of magnitude larger than that of sandstones, leads the coal rock has strong stress sensitivity (Kou et al., 2018).
Some studies and field practices have demonstrated that during the initial production period of CBM wells, the gas desorption quantity is limited and the reservoir stress sensitivity effect is dominant. An incorrect pressure drop rate can lead to a serious stress sensitivity in the coal reservoir, leading to a decline in permeability and poor gas production. Moreover, a fast pressure drop will mobilize coal dust in the reservoir, blocking the gas seepage channels and further reducing gas production. A proper production schedule is conducive to gas wells having high and stable yields (Manrique et al., 2011). Successful development experiment show that reducing the bottom hole pressure of CBM well step by step and slowly at the early stage can make the pressure wave propagates as far as possible and improve the recovery of coalbed methane reservoir (Manrique et al., 2011).
Although the above results have clearly shown the importance of determining a reasonable production schedule of CBM well, finding the optimal production schedule is still challenging. There has no one production schedule that fits all CBM wells. Many parameters, including formation and rock-liquid physical characteristics, should be fully considered in the production schedule optimization. This results in different optimal schedules for different wells. The practical experience can guide the CBM well production, and quantitative methods are still needed for fine development (Liu et al., 2013). To the best of our knowledge, research on optimizing CBM well production schedule quantitatively is still limited (Liu et al., 2013).
Production schedule planning of CBM well can be formulated as an optimization problem and finding the optimal solution using optimization algorithms. This idea has not been investigated for this problem before, but it already has many successful applications in conventional oil and gas production optimization problems, CBM well placement optimization problems, CBM EOR optimization problems (Liu et al., 2013), etc. A number of algorithms have been proposed and investigated for above optimization problems. These algorithms fall into two categories: gradient-based methods (Liu et al., 2013), e.g., steepest ascent algorithm, conjugate gradient algorithm, etc., and derivative-free methods, e.g., generalized pattern search (GPS), genetic algorithm (GA), particle swarm optimization (PSO), covariance matrix adaptation evolution strategy (CMA-ES), etc. The derivative-free methods only require the value of objective function and involve no explicit gradient calculations can be applied to CBM well production schedule optimization problem more easily. Among the derivative-free methods, CMA-ES performed very well in oil and gas production optimization and many other fields (Liu et al., 2013).
1.3 ORIGIN OF CBM RESERVOIRS
Coal originates as an accumulation of organic matter in swamps and marshes commonly associated with fluvial systems, deltas, and marine shorelines. It is critical to submerge the accumulating organic matter quickly beneath the water table to prevent oxidation. This requires a combination of basin subsidence and a rising water table sufficient to match the accumulation rate. Organic matter accumulates at an average rate of approximately a millimeter per year and compacts by a factor of seven to 10 times as it is transformed into coal.
As organic matter is buried, it is first transformed into peat, which consists of loosely compacted masses of organic material containing more than 75% moisture. This transformation takes place mainly through the compaction and expulsion of interstitial water. Biochemical reactions associated with this process transform the organic matter into humic substances, which are the precursors of coal minerals. These reactions can also generate significant amounts of biogenic methane, which often is referred to as swamp gas. Continued compaction and dehydration transform peat into a low-quality coal called lignite, which contains 30 to 40% interstitial water.
With deeper burial, temperatures increase, and geochemical processes dominate physical processes. Lignite evolves into subbituminous coal by expelling H2O, CO, CO2, H2S, and NH3, leaving behind a structure enriched in carbon and hydrogen. At temperatures greater than approximately 220°F (104.4°C), carbon-carbon bonds begin to break, generating gas and liquid hydrocarbons that become trapped in the coals. As these bituminous coals are buried more deeply, their hydrocarbons are cracked into thermogenic methane and expelled as an order of magnitude more gas is generated than the coal is capable of storing. In a typical coal, the H/C atomic ratio decreases from 0.75 to 0.25 as coals mature from high-volatile bituminous to anthracite.
The generation and expulsion of hydrocarbons is accompanied by several profound changes in coal structure and composition (Levine et al., 2013). Moisture content is reduced to just a few percent as water is expelled. Microporosity increases as the atomic structure of the coal changes, generating a huge surface area for sorbing methane. These changes also lower the bulk density from 1.5 g/cm3 in high-volatile bituminous coals to less than 1.3 g/cm3 in low-volatile bituminous coals. Coal strength decreases, making it easier for the coal to fracture as volatiles evolve and the coal shrinks. This creates closely spaced cleats, which enhance permeability.
At temperatures exceeding approximately 300°F, bituminous coals are changed to anthracite (> 92% carbon). Methane generation and expulsion decrease, and the bulk density increases from 1.3 g/cm3 to more than 1.8 g/cm3 as the coal structure becomes more compact. Methane contents in anthracites are typically quite high, but permeability is lower than bituminous coals because of cleat annealing. With further maturation, remaining volatiles are driven off and carbon structures coalesce, resulting in a dense coal with very high carbon content and a chemical composition similar to graphite.
To generate temperatures high enough to produce large quantities of hydrocarbons, coals must be buried deeply, typically to depths greater than 3000 m. Exceptions to this are coals transformed by local heat sources such as igneous intrusions. After sufficient burial and time to generate hydrocarbons, coals must be uplifted to shallower depths to be exploited commercially. At depths shallower than a few hundred meters, there is not enough pressure in the cleat system to hold economic quantities of sorbed gas in the coal. At depths greater than approximately 1200 m, permeabilities are generally too low to produce gas at economic rates.
1.4 GAS CONTENTS IN COAL
Gas contents in coal seams vary widely and are a function of coal composition, burial and uplift history, and the addition of migrated thermal or biogenic gas. Both vitrinite- and liptinite-rich coals can generate large quantities of hydrocarbons, but inertinite-rich coals, which consist of oxidized organic material, generate very little gas. The highest gas contents are found in anthracite coals, although their permeabilities are often too low to achieve commercial gas rates. High-volatile A to low-volatile bituminous coals have lower gas contents than anthracites but higher permeabilities. These bituminous coals have been the primary target of CBM exploration, primarily because coals of this rank are CBM reservoirs in the San Juan and Black Warrior basins where the modern CBM industry began.
During the 1990s, CBM reservoirs in the Uinta basin of Utah (high-volatile B) and Powder River basin of Wyoming (subbituminous B) were developed successfully despite being of lower rank than San Juan or Black Warrior coals. In the Uinta basin, gas contents have been enhanced by biogenic and migrating thermogenic gases. In the Powder River basin, the coals have low gas contents but are very thick, laterally extensive, and located close to the surface, allowing wells to be drilled and completed cheaply. These two projects have caused the industry to broaden its perspective and include lower rank coals as commercially viable targets.
Most CBM reservoirs contain both thermogenic and biogenic methane. Thermogenic methane is generated on burial, whereas biogenic methane is formed by late-stage bacteria that are introduced through groundwater flow and convert longer-chain hydrocarbons to methane. This gas augments the existing thermogenic methane and may increase gas contents significantly. Conversely, groundwater flow can reduce gas content by dissolving gas from the coal. An example of this is found in the Ferron coals located south of the Drunkard’s Wash CBM project in the Uinta basin of the western U.S. Groundwater is believed to have moved downward along the Joe’s Valley fault system, entering the coal seams at depth and pushing the gas updip where it is expelled at the outcrop (Montgomery et al., 2011).
Another mechanism for decreasing gas contents is the uplift and reburial of coal seams. For example, in the Hedong basin of China, Carboniferous coal seams are located beneath Plio-Pleistocene loess, which is up to several hundred meters deep. Before the deposition of this loess, the coal seams were closer to the surface and possibly were equilibrated to a lower pressure before reburial. As a result, the gas contents could be lower than expected, unless biogenic gas or migrated thermogenic gas augmented the existing gas fraction after reburial.
Coalbeds often contain gases other than methane, including:
- Carbon dioxide
- Ethane
- Hydrogen
- Nitrogen
Coal has a greater affinity for carbon dioxide and ethane than for methane and may contain substantial quantities of these gases. Proper coal desorption and sorption isotherm work can quantify the amount of each species and generate a composite isotherm representative of the coal’s sorption character. If carbon dioxide and ethane are present in the reservoir, it is likely that the produced gas will become enriched in these components as the reservoir is depleted.
1.5 METHANE ENHANCED RECOVERY
Methane recovery can be enhanced in CBM reservoirs by the injection of CO2 or nitrogen. Coal prefers carbon dioxide, and it will release methane to sorb injected CO2. This significantly increases the amount of methane available for production, but also causes the coal to swell, reducing permeability with time. Nitrogen reduces the partial pressure of methane, causing it to desorb from the coal (Puri et al., 2010). The injected gas reduces the partial pressure of methane more rapidly than the total pressure can be reduced by dewatering, resulting in accelerated production (Fulton et al., 2010). An additional benefit of nitrogen or CO2 injection is that the methane can be desorbed while maintaining higher reservoir pressures, resulting in added energy to drive the methane to the wellbore. Both injection processes have been tested over the past decade. The largest pilot projects are Burlington Resources’ Allison pilot and BP’s Tiffany Unit pilot, which are both located in the San Juan basin. While these pilots were primarily designed to enhance methane recovery, pilots that are more recent focus on the benefits of both CO2 sequestration and enhanced methane recovery.
1.6 REVIEW OF RELATED STUDIES
Coal Bed Methane can be optimized by developing a medel. The first dual porosity model was introduced by Barenblatt et al (2010). A further development of a dual porosity model was then presented by Warren and Root8, who proposed the application of a dual porosity model in well testing interpretation. The Warren and Root dual porosity model later became a basic concept in the development of naturally-fractured reservoir characterization techniques. Most unconventional reservoirs for gas such as tight gas, shale gas, and coalbed methane are classified as naturally-fractured reservoirs.
Warren and Root proposed a conceptual model for a naturally-fractured reservoir by modeling a homogeneous matrix block that is separated by fractures. The matrix block serves as storage for adsorbed gas and the fracture system provides media for the fluid flow within the reservoir body, from the matrix to the fracture system, which is followed by the fluid flow from the matrix system to the wellbore. The overall formation permeability is strongly related with a fracture or cleat system.
The dual porosity concept proposed by Warren and Root is also applicable in a coalbed methane reservoir. The dual porosity concept provides an idealized model of reservoir performance in two different types of media. The first medium is storage that contributes to the pore volume but with very low flow capacity. The second medium is a fracture system which contributes to fluid flow. Warren and Root classified porosity into two categories. The first one is the primary porosity controlled by deposition and lithification. The second type of porosity is the secondary porosity; a porosity that is controlled by water solution, natural fracturing, and jointing. A mathematical model for this description is presented for the application of a pressure build-up analysis. The idealized model is derived at an unsteady state condition and presented with two additional parameters to characterize the dual porosity system.
David et al (2014) presented a mathematical model to simulate methane and water flows through the coal seam and the effect of coalbed methane reservoir properties on gas drainage. This work uses single and multiple well systems. Olufemi, et al (2004) conducted numerical reservoir simulations to study the effects of coal seam properties variability in an enhanced coalbed methane project. They used a numerical simulation model to show the most influential parameters that affect recovery in an enhanced coalbed methane reservoir project. However, most of these works did not cover the development of fundamentals of the fluid flow and adsorption-desorption phenomenon in numerical modeling.
Cervik (2017), presented a basic concept of transport phenomenon for gas at a free gas and desorption state. This work showed gas dependency of gas desorption phenomenon to the coal particle size, equilibrated pressure and diffusivity coefficient. It showed that smaller particle tends to provide more gas. He proposed three classifications of gas transport phenomena. The first one was principally Fick’s law while the second one was a combination of Fick’s and Darcy’s law, and the third one was predominantly Darcy’s law. Base on the results, it was not recommended to use the same basic concept for conventional gas reservoir engineering in a coalbed methane reservoir model, since the Darcy’s law and Fick’s law govern overall mass transport phenomenon.
By using a numerical simulator, Zuber et al (2017) illustrated the procedure to determine coalbed methane reservoir properties by using a history-matching analysis. The numerical simulator was modified to adjust the flow and storage mechanism in a coalbed methane reservoir. In the history- matching process, a two-phase dual porosity simulator was used to model reservoir performance based on production data, geological data, and laboratory data.
Another work conducted by Seidle15 presented a methodology to utilize a conventional reservoir simulator with some input data modification to model a coalbed methane reservoir. This work assumed an instantaneous desorption that occurred from the matrix block to the cleat system by using the analogy of dissolved gas in immobile oil for a conventional reservoir simulator as adsorption gas on the internal surface of a coal matrix. This work showed that the rate of diffusion in the matrix system was much higher than the fluid flow in the cleat system. Therefore, this work analogizes gas adsorption as saturated gas in immobile oil. In this case, the solution gas oil ratio is determined by the Langmuir isotherm equation.
Another work on conventional gas reservoir engineering adapted to coalbed methane reservoir was presented by King (2010). His work showed a modification of material balance concept for reserve estimation and prediction of future production performance in unconventional gas reservoirs. This work utilized fundamentals of conventional gas reservoir engineering for material balance techniques in a coalbed methane reservoir with the effects of gas desorption and diffusion in consideration. The material balance analysis assumed an equilibrium state of gas and adsorbed gas in the coal system. A pseudo-steady state condition was also assumed to be applied during the sorption process. This work provided a procedure of gas in place estimation by using the p/z method and prediction of future production performance based on the existing material balance techniques.
A modification of King’s method was presented by Seidle (2019) with more advanced techniques in material balance. His work provided fundamentals of a mathematical model, simulation studies, and examples of field application. The modified method improved material balance techniques by eliminating mathematical problems and suggesting more accurate reserve estimation for a coalbed methane reservoir.
Other numerical reservoir simulation studies were presented by David et al (2002), Hower, T.L (2003), and Jalal et al (2004). They showed the application of a compositional simulator in coalbed methane reservoir modeling. The numerical compositional simulator was equipped with some additional features for coalbed methane reservoir modeling. David and Law’s work showed coalbed methane enhanced the recovery model by using a compositional numerical simulator. The enhanced recovery method is the CO2 injection. The compositional simulator was able to model more than two components. This work assumed instantaneous process of gas diffusion from the matrix system to the fracture system.
Aminian et al (2020) presented another approach of predicting coalbed methane gas production performance by using a type curve matching based on gas and water rates. This method used dimensionless rate and time. It also showed the application of the type curve matching for determining the matrix and cleat porosity based on production- history matching. Based on the matching results, future production performance could be estimated to evaluate the coalbed methane reservoir prospect. This study also provided a correlation of the peak gas rate to predict future production performance.
A later work by Reeves et al (2021) presented a mathematical model for a desorption-controlled reservoir. They introduced the model as a triple-porosity dual- permeability model. This mathematical model was a modification of Warren and Root’s model. This work showed the erroneous result of the previously existing dual-porosity single-permeability model in predicting coalbed methane reservoir performance. An overestimation of gas and water production tended to appear with the inconsistency of the model result and field data. In fact, gas production was found much higher than the gas predicted form the model in later time. To model this phenomenon, a set of porosity and permeability was added to the system. The third porosity was introduced in the matrix block system to provide free gas and water storage capacity for the modification of material balance techniques. This work also provided decoupled models of a desorption process from a matrix block and the diffusion process through a micro- permeability matrix so that mass transport could be explicitly determined. A comparison of the existing model result and the proposed model result was shown with a higher water rate and lower gas rate which were more accurate and matched with field data. This work also introduced a new coalbed methane simulator, COMET2 with some modifications in the fundamentals of the fluid flow and desorption process.
A modification of Seidle (2010) approach was presented by Thomas (2012). His work also used a commercial simulator to model coalbed methane reservoir performance with independent implementation. He also showed a comparison of his result in a paper by Paul et al (2014). This work illustrated pressure dependent porosity and permeability phenomenon with some comparative runs.
The result was not consistent with Seidle’ paper, but, as reported, it was an excellent match for Paul’s paper. Tan’s work also suggested the dual grid approach to gain a more accurate result in a matrix-fracture model.
In 2003, Xiao et al (2003) presented a more sophisticated numerical simulator with 3 dimensional and two-phase flow calculation capability. The new simulator improved coalbed methane reservoir characterization by including transport phenomena in the coal micropores and fracture system. The gas resulting from the desorption process was calculated with a sorption isotherm curve from the experiments and calculation. Therefore, an equilibrium state of desorption process was necessary to be considered.
1.7 COALBED METHANE RESERVOIR SENSITIVITY STUDY
David et al (2014) investigated the relationship between the peak gas rate and the ability of a matrix system to desorb gas. They performed a sensitivity study to observe the consistency of new reservoir simulator results. The study included an investigation of absolute permeability, sorption time for the gas diffusion rate, and relative permeability effects on methane recovery for various well spacing scenarios. The sensitivity study incorporated the effects of reservoir property variation on the drainage efficiency of gas in the coal matrix system. This work used a single well model.
Another work by Olufemi et al (2014) investigated the effect of coalbed methane reservoir properties on production performance in a enhanced coalbed methane project. A reservoir simulator was used to model reservoir performance and select most influential parameters affecting gas recovery. It showed that reservoir permeability, coal density, and Langmuir volume were the most significant factors in methane recovery of a CO2 sequestration study.
Derickson et al (2018) presented a sensitivity study result for coalbed methane reservoir production performance in Huaibei, China. This work investigated the effects of some fundamental coal properties variation on the production rate. They concluded that coal permeability, gas content, initial water saturation, and coal thickness were the most influential factors related to gas production.
Roadifer et al., (2019) conducted a comprehensive study with more than 100,000 simulation runs. The study was aimed to perform a parametric study incorporated with a Monte Carlo simulation analysis. Numerous combinations of reservoir properties, geological data, completion and operation constraint were prepared in the simulation runs to investigate the effects on production performance. Relative importance of each parameter and inter-parameter relationship were identified. Rank correlation was developed based on simulation results considering several production constraints, such as the peak gas rate, dewatering times, and cumulative gas production. Core sample acquisition in coal seams was difficult due to its tendency to be extremely friable. This friability complicated the reservoir properties measurement especially for permeability, porosity, compressibility and relative permeability data. This paper explained the differences between a sensitivity study and a parametric study based on basic concepts. The sensitivity study was performed by changing one value while keeping the other values at the base value. On the other side, a parametric study was conducted by preparing all possible combinations of each parameter at every value (e.g. minimum, most likely, and maximum).
Stevenson et al (2013) conducted a sensitivity study for a nitrogen-enhanced coalbed methane study. This work investigated the effects of reservoir parameter variation on the project economics based on predicted gas production. The reservoir parameters that were identified as the most significant factors were permeability, relative permeability, compressibility, layering and capillary pressure. For each parameter, the minimum, most likely, and maximum values were taken into account. San Juan basin data were chosen to be used in performing the sensitivity study.
Reeves (2011) performed a discrete parametric study for a wide range of the reservoir depth, pressure gradient, Langmuir volume, and permeability as a function of pressure and depth. Young et al (2011) presented a distinct parametric study for San Juan basin area. This work divided San Juan basin into three areas for a discrete parametric study based on reservoir properties variations. For instance, in Area 1 the sensitivity study covered permeability, porosity and drainage variation. In Area 2 permeability, porosity, drainage area and fracture half-length were investigated for a particular range. In Area 3, a sensitivity study was performed for the coal compressibility, gas content, Langmuir parameter and relative permeability ratio.
