AN OVERVIEW OF EMERGING PRACTICES AND MODELS USED IN COAL MINE METHANE RESOURCE ESTIMATION AND RESERVE EVALUATION
by Raymond C. Pilcher, Ronald C. Collings, & James S. Marshall
ABSTRACT
Establishing a reliable estimate of recoverable resources poses significant challenges in the initial stages of coal mine gas (CMG) project development. Gas resources often appear to be sizable and attractive for investment during the conceptual and planning stages of coal mine gas projects. During the pre-feasibility stage of development, the estimate of resource size is revised based on what is considered technically recoverable. Even if the results of the pre-feasibility studies are encouraging, additional information and data is required to further refine the estimate of producible gas, and determine the economic feasibility of the proposed project. After the gas recovery technology is chosen, the resource estimate is once again revised and the volume of economically recoverable resource is estimated. Unfortunately, at this point, it is often realized that the resource that was once thought to be sizable is significantly smaller and the potential for commercial development constrained by recovery economics and market availability for coal mine gas.
Although advanced statistically valid methods for estimating gas resources are used in the petroleum industry, they are less often used in the mining industry. Analyzing the impact of sampling density and variability in data is an important step in approaching coal mine gas resource estimation. Coal mine gas resources are often estimated based on single-value or averages of measurements of important parameters, such as gas content, coal density and thickness. Using frequency distributions developed for each of the parameters quantifies the risk associated with the resource estimate and therefore increases confidence in the estimate. PC based Monte Carlo simulators facilitate this type of analysis through simultaneous and random sampling of arrays of distributions, allowing the probability of occurrence of resource volumes to be determined.
Evaluation of recoverable coal mine gas reserves poses its own set of unique challenges. Reasonable valuation of reserves depends on a knowledge of recoverable gas volume, gas production rate through time and the impact of economic constraints on commerciality. Three dimensional multi-reservoir simulation using PC based software makes prediction of coal mine gas production rates and estimation of ultimate gas recovery possible during the early stages of project analysis and development. Furthermore, risks can be quantified through sensitivity analysis of input parameters, and subsequent economic decision-making based on simulation output can result in a higher likelihood of achieving economic goals.
1 INTRODUCTION
Companies involved in coal mining are focused primarily on one thing—mining coal in the most cost effective manner, without compromising the safety of the miners. Coal mining companies are continually faced with the need to develop safe practices that ensure the well being and productivity of their workforce. In gassy mining conditions, creating a safe work environment requires that coal mining companies develop practices that allow them to assess the amount of gas that will be liberated during the mining process, and determine the best way to remove the gas from the mine. No matter whether the gas is drained from the seam or adjoining strata in advance of mining or from the gob, the purpose is the same: remove enough gas from the mine so that the ventilation system can dilute the remaining gas that will be emitted into the mine to acceptable levels. Gas drainage systems are often not designed with the goal of optimizing gas recovery because of budget constraints and the overriding concerns of safety. Furthermore, for the same reasons, data available to the investigator for assessing the potential of developing a commercial coal mine methane resource estimate may be limited.
2 COAL MINE GAS RESOURCE ASSESSMENT AND ESTIMATION PRACTICES
2.1 Challenges of Coal Mine Methane Resource Assessment
Often, when a coal mining company or an outside developer decides to investigate the potential for developing coal mine methane resources, they discover that determining the size and value of the coal mine methane resource and reserves is challenging. The challenge arises when the project developers are faced with using data that was collected for the sole purpose of supplying information to decision-makers so that they may control the emissions of gas into active mine workings. The tonnage of coal mined, and the volume of gas that is liberated during mining for a given period are the data that are generally available. Developers may use the data to determine the specific emissions, the amount of gas liberated per ton of coal mined (expressed in cubic meters per ton of coal mined); and relative emissions, the volume of gas liberated over time (expressed in cubic meters per unit of time). These parameters may indicate the relative gas production potential of a mining property; however, the project developer will need to collect and analyze additional data to determine the commercial potential for developing the resource.
Successful development of a coalmine methane project requires a thorough understanding of the size and production potential of the gas resource. The coal mine methane resource comprises the volume of gas distributed throughout the coal and surrounding strata, often referred to as gas-inplace. An estimate of gas-in-place represents the upper boundary of the gas that is potentially recoverable; but 100% recovery of the gas-in-place is virtually impossible. Technically recoverable coal mine methane resources is the quantity of gas that is recoverable by utilizing proven modes of extraction while employing existing technology. The commercially extractable portion of the technically recoverable resources is the reserves. A developer’s estimate of reserves will vary depending on assumptions regarding the technology used for recovery and changes that may take place in future economic conditions. Figure 1 illustrates the relationship of reserves to the technically recoverable resources and ultimately the gas-in-place. This figure is helpful in conceptualizing the impact the selection of the technology used for recovery and variations in the estimated volume of reserves has on the volume of technically recoverable resources. Variations in the volume of reserves may be due to such factors as changes in sales price or costs associated with extracting the gas. The first task for the developer is to acquire a reasonable estimate for the gas-in-place so that he may determine the gross gas production potential of the property. Remembering that the gas-in-place estimate is the upper bound of the resource potential, this will be the point at which the developer can determine if the original project concept is even possible, given the magnitude of the resource.
Figure 1. Conceptual Diagram Of The Relation Between Resources And Reserves
2.2 Estimating Gas-in-Place
It may be difficult to take the specific and relative emissions and sparse data collected from the minable coal seams, such as gas content and thickness of the coal, and develop an assessment of the resources. There are many approaches to this task; but the following attempts to lay out an approach to collecting data, and analyzing and assessing the amount of gas that will be liberated from an active coal mine property.
The first step in estimating the gas-in-place (GIP) resource is to understand the limitations of the data and information that is available. Typically, the data available for estimating the GIP are:
- Thickness of the coal seam determined from outcrop data, borehole intercepts, and mining drivages;
- Gas content data determined from core or cuttings recovered from exploratory boreholes;
- Adsorption isotherms.
Many methods of estimating the GIP are acceptable, the most common of which are volumetric calculations. All methods of developing a volumetric calculation include at a minimum the following steps:
- Calculate the volume of the reservoir rock.
- Determine the distribution of gas throughout the reservoir. In the case of coal, much of the gas is adsorbed onto the matrix, so an investigator may make the simplifying assumption that the gas is distributed as a function of pressure, determined by an adsorption isotherm. Additionally, if gas occurs in the non-coal strata, the investigator must determine the distribution of porosity throughout those strata, as the gas may be stored in that pore space as well.
Estimating GIP is less labor intensive with the use of computer software applications. Integrated software applications use an areal grid and sampled data values, such as thickness, to fit a surface to the data using a mathematical algorithm. This algorithm assigns values to each of the grid points for the area under investigation. The investigator may use these values for further mathematical analysis, such as computing the volume of the reservoir.
Further, the investigator may calculate the GIP by simply using gas content data to determine the GIP. Gas content data may be multiplied by the volumetric values for each of the grid points resulting in a GIP value at each grid point. Integration of the resulting surface yields a GIP estimate. Although this estimate is usable, it does not account for the uncertainty associated with using single point estimates calculated from thickness and gas content data obtained from boreholes to compute intervening values at each grid point. In other words, assessing the likelihood of encountering a given gas quantity occurring at each one of the randomly sampled grid points. In practice, the developer does not know how reasonable it is to expect a resource of a given size to occur throughout the area of interest.
By using the same data set, the investigator can obtain an understanding of the likelihood of the occurrence of a given outcome by using Monte Carlo methods. As an example, Figure 2 shows a forecast distribution for the GIP of a coal seam. This example forecast predicts that there is an 80% probability that a 160-acre (64.75 hectares) tract will have between 1 and 6 million cubic feet (approximately 28,000 and 171,000 cubic meters of methane) of methane GIP resource.
Figure 2: Gas in-place forecast showing the probability of occurrence of each class of occurrence.
2.3 Estimating Technically Recoverable Resources
Only a portion of the original-gas-in-place will be recoverable through wellbores. The fraction of the total resource that is recoverable is called the recovery factor. Investigators can estimate the recovery factor based on experiences in similar geologic regions and mining conditions and utilizing a given type of technology for extraction. This type of estimate has the disadvantage of not incorporating unique aspects of the project area. Another way to predict the recovery factor is through computational fluid dynamics (CFD) modeling. This type of modeling uses numerical techniques to predict the gas emission rate into roadways or production into wellbores, and then accounts for the change in pressure (or flow potential) related to the removal of this gas. Figure 3 shows the predicted production rates for three vertical wells drilled into a longwall panel gob. To use a CFD model, the following steps need to be followed:
- A generalized geologic model is built of the targeted seam and surrounding strata.
- This model is then transformed to a digital grid that captures the significant features of the mine workings and the stratigraphic column affected by the mine workings.
- Values are then assigned to the parameters that control gas flow through the model, such as permeability (or conductivity) and porosity. This is done for each discreet object in the model, such as coal layer or rubble zone.
- The initial pressure of the area is specified. The initial pressure will have been established in the resource evaluation or original-gas-in-place calculation.
- The location and nature of pressure sinks are specified. These sinks can be either wellbores or ventilated mine entries, or both.
- The model predicts the gas emissions into the mine or production from wells through time.
- The cumulative gas produced at the end of the simulation divided by the OGIP is the
recovery factor.
Figure 3: Predicted Gob Well Production Rates from CFD Model
3 ESTIMATING ECONOMICALLY RECOVERABLE RESOURCES, RESERVES
There can be a significant range of uncertainty in the value of the parameters input to the model. However, these uncertainties may or may not have a significant impact on the recovery factor. The effect of each parameter's uncertainty can be quantified through sensitivity analysis. Figure 4 shows a "Tornado Diagram", which is used to evaluate the effect of the uncertainty of the value of a parameter on the recoverable resource. To construct a Tornado Diagram, all parameters affecting gas recovery in the model should be set at their best estimate, or most likely value. The model is run and the gas recovery is then used as the axis of the tornado. Next, one parameter is set at it's minimum expected value while leaving all others at their most likely values and the model is run again. This produces the minimum recovery value for that parameter's uncertainty. This is done again with the same parameter set at its highest expected value. The range of uncertainty in the gas recovery related to this parameter is indicated by the length of the bar. The greater the length of a bar relative to that of the others is an indicator of the significance of that parameter and the relative importance of minimizing its uncertainty. This type of analysis helps to focus additional data collection.
Figure 4: Sample Tornado Diagram Evaluating EUR
Once a development plan is proposed, the CFD model can help generate a reserve estimate. According to the Society of Petroleum Engineers and the World Petroleum Congress: "Proved reserves are those quantities of petroleum which, by analysis of geological and engineering data, can be estimated with reasonable certainty to be commercially recoverable, from a given date forward, from known reservoirs and under current economic conditions, operating methods and government regulations." In other words, the quantities of recoverable gas must be reasonably certain and the gas must be economically recoverable. The CFD model, based on the development plan relating to mining and well placement, is used to forecast gas rates. The predicted gas rates and associated revenue together with the capital and operating costs determine the economic limit of the project (the time when operating costs exceed revenue). This will provide the reserve estimate. Often, only a portion of the predicted reserves is classified as proved to ensure the "reasonable certainty" of recovery. Additional volumes are added to the proved category as actual production information is accumulated and the model predictions are validated.
Figure 5: Sample Tornado Diagram evaluating NPV
