It is very difficult to predict the emission of coal gas before the extraction, because it depends on various geological, geographical and operational factors. Gas content is a very important parameter for assessing gas emission in the coal seam during and after the extraction. Large amounts of gas released during the mining cause concern about adequate airflow for the ventilation and worker safety. Hence, the performance of the ventilation system is very important in an underground mine. In this paper, the gas content uncertainty in a coal seam is first investigated using the central data of 64 exploratory boreholes. After identifying the important coal seams in terms of gas emission, the variogram modeling for gas content was performed to define the distribution. Consecutive simulations were run for the random evaluation of gas content. Then, a method was proposed to predict gas emission based on the Monte Carlo random simulation method. In order to improve the reliability and precision of gas emission prediction, various factors affecting the gas emission were investigated and the main factors determining the gas emission were identified based on a sensitivity analysis on the mine data. This method produced relative and average errors of 2% and 0.57%, respectively. The results showed that the proposed model is accurate enough to determine the amount of emitted gas and ventilation. In addition, the predicted value was basically consistent with the actual value and the gas emission prediction method based on the uncertainty theory is reliable.
Research
Model of Prediction of Gas Emission based on Gas Content Uncertainty in Coal Mine using Geostatistical Simulation and Monte Carlo Method
https://doi.org/10.21203/rs.3.rs-607234/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Predictiongas emission
Uncertainty gas content
Monte-carlo simulation
Geostatistical simulation.
Gas emission forecasting is an important basis for building new mines and new ventilation plans in the mining panel and gas prevention and management. It plays an important role in reducing gas explosion accidents and ensuring safety in coal mine production (Zhang, 2014). Therefore, the prediction accuracy directly influences the economic-technological indicators of mine, especially in large mines (Harpalani and Chen, 1995). It is very essential to calculate the gas content of coal seams and predict the gas emission from the surrounding rocks in the coal mine. Gas emission has a complex system in which the exact amount of influential variables is not yet known, and a specific model cannot be determined to accurately predict gas emission (He et al., 2008).
As a basic safety practice, one must estimate the gas content of coal seam and proceed to predict gas emission into the mine. Gas may be discharged from gas sources or from the main seam before or during the extraction. The emission of large and irregular amounts of gas in the longwall mining method has provided the need for optimizing the methods used to predict this amount of gas and the ventilation required to dilute the gas.
In the 1950s, some methods and processes were presented to measure the gas content of coal seams, and the statistical methods were used to calculate and predict the gas content and gas emission in coal mines. In the 1980s, various methods were proposed to predict gas emission. Since then, the analog method, geological mathematical model, speed method and other methods of predicting gas emission have been proposed and used. All new methods and technologies provided a scientific basis for the mine design and reconstruction (Li, el at., 2014). Some new methods for predicting gas emission in coal seam are: multivariate linear regression (Shi and Wu, 2008), grey system theory (Wu, Tian, Song and et al., 2005), neural network (Zhou, Chang and Zhang, 2007), support vector machine (Fu and Shi, 2013), evidence theory (Cheng, Zhang and Xiaokun, 2012), chaos theory (Shi, Song, He and et al., 2006), fractal theory (He, Pan and Nie, 2006), and rough set theory (Shao, 2009), statistical method, plot headstream method and gas geological map method (Zeng, 2004; Zhu, 2012; Yang, et al., 2018).
Li (2014) used techniques based on self-organizing data method to predict gas emission in coal mine accurately. The results have shown that such methods can automatically analyze nonlinear relationships between gas emission and the effective factors. In the exploitation stages in this coal mine, numerous safety and disaster control factors have been considered in the framework of a comprehensive surveillance plan for effective control of gas-driven disasters. To simplify the analysis, the gas content, coal bed thickness, advance rate, coal production, gas content of the adjacent layer, layer thickness, and distance to adjacent layer were selected for gas emission prediction. Among these seven factors, the adjacent layer thickness exhibited a relatively weaker effect than the other factors. The modeling results were generally in agreement with real data (Li, el at., 2014).
An artificial neural network (ANN) learned on known experimental data was devised to predict the gas emission (Lei. et al, 2014). Gas content, coal bed depth and thickness, coal bed pitch, working face thickness, working face length, advance rate, recovery factor, gas content of the adjacent layer, adjacent layer thickness, layer spacing, petrology of the interbeddings, and exploitation intensity were considered as independent variables while gas emission was taken as the only dependent variable in this study. A total of 18 sets of data were analyzed using a Regression Neural Network (GRNN) and a Multilayer Feedfoward Neural Network, and the results showed that the GRNN could better predict the gas emission at an error of 0.50 (Lei. et al., 2014).
Numerous researchers have reported the use of Monte Carlo simulations for modeling methane gas emission, with the results compared to those of stochastic modeling to reflect the acceptability of the simulation results (Zhou, 2015). In these studies, gas methane emission prediction has been based on the data on the coal structure, coal thickness, coal quality, and gas content (Zhou, 2015).
A gas emission prediction model using a modified neural network algorithm was presented by Fu (2014) and historical data from a real mine were provided to test the model and analyze the results. The findings showed that, upon optimization with the so-called ant colony clustering algorithm, the Elman neural network model had its generalizability and prediction accuracy improved and provided for dynamic prediction of gas emission (Fu, et al., 2014).
Methane gas emission was predicted by a grey-gas geology method. The main factors considered in this study included gas content, coal bed depth, coal bed thickness, per-day coal production rate, geographic setting, and thickness of the adjacent layers, which were evaluated using the theory of gas geology. Results of the test showed that the model provided for high accuracy as its predictions were highly reliable (Wang, 2018).
Currently, researchers have conducted numerous studies to predict the emission of gas in coal. However, as coal gas is a complex phenomenon, the research to predict gas content and gas emissions has not yet reached a consensus, and no mature theory or accurate calculation method has been recognized so far.
Since the gas content is among the most important parameters affecting the accurate assessment and prediction of the coal gas emission, the uncertainty-based estimation of gas content tends to end up with more reliable outcomes. As such, one should evaluate the uncertainty along with the gas emission estimation. The review of literature shows that despite valuable research, no work has been done to predict gas emission based on the gas content uncertainty. In fact, all models have assumed that the parameters affecting the gas emission level are certain, while these parameters are uncertain.
In the present work, the in-place gas content and the associated uncertainty were evaluated in a coal mine. The respective analysis improved the design and led to the application of a system for capturing and control of methane. For this purpose, we used two methods, namely the kriging and Sequential Gaussian simulation, and their results were compared in terms of variance and distribution. The kriging method is recommended when the final objective function is to minimize a single prediction error or to develop a uniform exploratory map of an unknown attribute. The simulation, on the other hand, is advised for the cases where the main objective is to properly evaluate the confidence intervals or model some spatial continuity (Olea, 2009). However, there is no clear criterion for ruling out the simulation or kriging in the geostatistical method (Olea, 2009). The difference of simulation approach with the kriging method is the higher capability for reproducing the spatial continuity patterns and realistic uncertainty models. This difference is based on the inherent preference between the both methods. Accordingly, in the present work, the SGSIM was used as a preferable analysis technique. Then, six main factors are selected to predict methane gas emission, and an uncertainty model is constructed to predict and evaluate methane gas emission using the Monte Carlo simulation method. Seam gas content, coal seam thickness, production rate, advance rate, gas content in adjacent seams and thickness of adjacent seams are the considered independent variables, and methane emission level is the considered dependent variable. This study aims to combine two simulation concepts: geostatistical simulation, to capture gas content uncertainty, and Monte Carlo simulation, to predict coal gas emission based on gas content uncertainty. Therefore, the present article demonstrates an attempt to present this method and build a gas emission prediction model to help prevent and control gas-driven disasters while improving the control level significantly.
As a case study, a coal bed in the southwest of Tabas Coal Mine, Iran with a low-permeability coal seam was studied. On average, the main coal bed and the coal seam exhibited a thickness of 2.5 and 2.2 m, respectively. The mine was being advanced at 3.5 m/d leading to a mining intensity of some 725 t/d. Mean values of gas content in the coal seam and the adjacent bed were reportedly 8.52 and 7.69 m3/t, respectively. The examination of the stratigraphic column across the study area showed that the area was mainly composed of grey shales, siltstone, and sandstone with coal inter-beddings (Fig. 1).
Methane emission level is one of the major indicators in the design of mine ventilation system. Therefore, it is essential to accurately predict gas emission for mine design and production safety.Predicting gas emissions is a very complex process (Jenkings, 2008). This study was conducted based on the existing conditions in the coal mine located in Tabas region. Figure 2 shows the gas emission prediction trend.
The factors influencing the prediction of gas emission are as follows:
A) Gas content of coal seam and adjacent seams: Gas content is defined as the volume of released gas that is dispersed in the stope space or drift tunnel to extract one ton of coal, which is usually expressed in m3/ton (Skochinsky and Komarov, 1989). The accurate determination of gas content plays an important role in the final accuracy of gas emission prediction. As the gas content in the coal seams increases, the level of gas emission into the stopes also increases. In a nutshell, the gas content of coal seam is very important to determine the gas emission level.
The gas emission level increases with increasing the gas content. In addition, the gas emission in the stope exceeds the gas content in the coal seam, which demonstrates that the source of gas emission is not only from the coal seam, but also from the rocks around the coal seam. Figure 3shows the relationship between gas emission and coal mine gas content in Australian mines (Saghafi, Williams and Lama, 1997). As can be seen, the gas emission level is more than four times the coal gas content. Kissell et al. (1973) observed similar relations for US coal mines and showed that the gas emission from coal mines far exceeded (approximately seven times) the expected amount of coal gas content alone (Kissell, McCulloch, and Elder, 1973). This distinction is also related to the changes in Australian and US mining conditions and differences in geology.
In general, the amount of gas released from the main and adjacent seams per ton of coal mined is about 6 to 9 times the gas content. This difference is due to the emission of gas from adjacent seams, dead zones, and previous faces, and also depends on the passage of time. A significant portion of the gas that enters the underground mining environment originates from the surrounding coal seams. The number, distance from the main seam, and total thickness of the adjacent seams are critical.
B) Coal seam thickness: Thick coal seam contains more gas, thereby increasing the gas emission level (Mucho, et al, 2000).
C) Daily coal production: The amount of daily production has a certain effect on the gas emission level. Therefore, the higher the daily production, the higher the level of gas emission (Schatzel et al. 2006).
D) Thickness of adjacent seams: The presence of adjacent seams around the main coal seam affects the gas emission. As the thickness of adjacent seams increases, the emission level also increases. The study of the stratigraphic column of the area shows that the study area is mainly composed of grey shales, siltstone and sandstone with interlayers of coal seams (Figure 1).
The gas emission level varies significantly in a wide range of geological and mining conditions. During the extraction process, the methane gas released from the main coal seams and adjacent seams mixes with the ventilation air, and the insufficient air in the ventilation system causes the gas accumulation in the mine and, in certain cases, may lead to the sudden explosion. Therefore, the accurate prediction of the methane gas flow rate into the face and ultimately into the ventilation system is complicated due to the large number of variables involved in the gas emission sources.
The gas released in the underground face is defined as a combination of the amount of gas released from the extracted coal seam and the adjacent seams (Equation (1)) (Lunarzewski, 1998):
where:
Q(y): amount of gas emitted in mine per ton of extracted coal (m3/min)
Qm: amount of gas emitted from extracted coal seam (m3/min)
TA: thickness of adjacent seam (m)
GCA: gas content of adjacent seam (m3/ton)
TM: main coal seam thickness (m)
The amount of gas emitted from the main seam is a function of the gas content of main seam, coal seam thickness, and production and advance rate, which is expressed as Equation (2) (Lunarzewski, 1998):
where:
b: width of working face (m), which is considered 220 m in this study.
GC: gas content of main coal seam (m3/ton)
R: advance rate (m/day)
γ: relative density of coal, equal to 1.37
C0: emission coefficient of methane gas, which is considered equal to 0.13 (Lunarzewski, 1998).
Coal mines are among the mines in which ventilation is more important than other underground mines. The purpose of ventilation in coal mines is to dilute and remove coal gas by circulating the air inside the mine network. Therefore, it can be stated that coal mine fans are determined based on the required flow intensity of mine network, required flow intensity based on emitted methane gas, amount of methane gas released based on gas content, coal seam thickness, production rate, advance rate, adjacent seam thickness, and gas content of adjacent seams.
As a result, the airflow intensity for diluting the gases from the gas content is expressed as Equation (3) (Hartman and Howard, 1997):
where:
Q: Required airflow (m3/min)
Q(y): amount of gas emission in mine per ton of extracted coal (m3/min)
D: concentration of gas emission
D0: concentration of gas emission in air way inlet
Considering the factors affecting the gas emission across the working face of the mine and based on the previous literature, a set of parameters were selected for sensitivity analysis to calculate the correlation of gas emission to gas content (R1), coal bed thickness (R2), advance rate (R3), production rate (R4), gas content of adjacent layer (R5), and adjacent layer thickness (R6), with the results reported in Table 1.
Table 1 The calculation results of correlation degree
|
|
R1 |
R2 |
R3 |
R4 |
R5 |
R6 |
|
Degree of correlation |
0.75 |
0.70 |
0.60 |
0.73 |
0.69 |
0.65 |
Focusing on the sensitivity analysis results, the selected factors were found to affect the gas emission strongly, i.e. at correlation coefficients higher than 0.6. This showed that the selected factors were some of the most significant parameters controlling the gas emission. In the meantime, effects of the gas content, coal bed thickness, and production rate on the gas emission were powerful, so that one must account form them appropriately when building a model for predicting the gas emission. In the considered mine, the weakest effect on the gas emission was that of the advance rate.
In this section, a definition of the work steps is provided, which includes the Monte Carlo simulation method along with the geostatistical methods. Figure 4 shows the process of the research method in the form of a flowchart. This research aims to combine two simulation concepts, namely ground simulation and Monte Carlo simulation. In this method, the research is based on geostatistical simulation as input for the Monte Carlo simulation process.
The most important factors affecting methane emission based on the importance and the documents reviewed in previous studies are gas content, coal thickness, advance rate, production rate, adjacent seam gas content, and adjacent seam thickness. The influencing factors are as the independent variables and gas emission as the dependent variable. The in-situ gas content uncertainty is first calculated using the geostatistical simulation and enters the Monte Carlo simulation as an input function. The rest of the influential parameters are entered into the simulation process as constants. The Monte Carlo simulation is used to generate a list of random samples for independent analyses and a combination of input parameters and their effect on the output variable. The output of the Monte Carlo simulation is the prediction of the emitted gas and the air required for the ventilation.
4.1 Geostatistical simulation
In this work, the data from 64 exploratory boreholes was collected to predict the gas content of a coalbed. Before proceeding to the spatial modeling, one should check for the normality of the data.The mean, minimum, and maximum gas contents were measured 8.52, 0.44, and 20 m3/ton, respectively. Based on the sample data taken from different depths of a deposit, the relationship between depth-layer and gas content was calculated which can be used for determining the gas content from the depths where no samples were taken. The variation of gas content was also compared to X and Y axes to illustrate that no relationship exists between X and Y direction.
Figure. 5 illustratesthe variations between gas content and depth. According to Figure 4, the gas content variations in the small current range of Tabas increases relative to its depth as: y = 5E-05x2 + 0.036x - 3.9326. Given that the gas content data have a certain trend with respect to depth, therefore, the conventional kriging method cannot be used and the application of universal kriging method is recommended.
The means (x) and standard deviations (s) of gas content used to calculate the probability distribution of gas content are listed in Table 2.
Histogram of each attribute was further investigated for both the coal seams and non-coal stratum to check for their Gaussian behavior. According to Olea (2009) and Remy et al. (2009), normality (i.e., a condition under which distribution of the set of data points follows a particular pattern called Gaussian distribution function) is a requirement for applicability of certain geostatistical techniques. In this study, histogram analyses showed that such a condition did not hold true for the gas content data in all cases. Therefore, an automatic transformation scheme was adopted to bring the data distribution to normality (i.e., a Gaussian distribution with mean and variance of 1). Next, semi-variograms were fitted to the normalized data. With the transformation parameters at hand, the attributes were inverted back to the original domain upon the simulations. Examples of gas content histogram analysis and the normalization procedure are shown in Fig. 6.
Table 2Statistical summary of gascontent data
|
Gas content |
Mean m 3 /t |
Minimum m 3 /t |
Maximum m 3 /t |
Variance m 3 /t |
St. Dev. m 3 /t |
|
8.52 |
0.44 |
20 |
24.24 |
4.89 |
In order to achieve a good estimation and appropriately determine the simulation parameters, one should begin with the semivariograms analysis. The semivariograms analysis allows to investigate the correlation of data with distance. The knowledge of spatial correlation and the scope of monitoring this correlation beside the knowledge of data means are considered when evaluating the spatial distribution of a parameter and its associated uncertainties. This consideration is practiced through the stochastic or kriging techniqueslike SGSIM. By drawing different semivariograms along different directions, the geometrically isotropic zones can be identified. In this study, three theoretical semivariograms (i.e., spherical, exponential, and Gaussian) were used to fit to the experimental data. In order to select the best semivariograms model, the post-kriging cross-validation statistics were used. Based on the fitness evaluation function, the spherical model was found to provide the best fit to the measured gas content data (in m3 per ton of coal) (Fig 7). The Datamine software was utilized to fit the semivariogram.
As demonstrated in Figure 5, the data exhibited a seemingly strong north – south trend that was established by the increase in depth (in Z direction). Therefore, this trend was corrected before proceeding to the estimation.
Considering the presence of the trend in the data on the gas content of the coalbed, the universal kriging was adopted. The universal kriging is a methodology where the trending parameters are investigated iteratively to determine a semivariogram. Upon eliminating the trend from the data, the residual short-range surface variations were subjected to the static analysis. Since the trend was then automatically introduced before creating the final surface, the prediction would produce significant results. In this study, the data was propagated across 33×50×50 blocks (where 33 is the block size). Regarding the gas content, the estimates were made using the universal kriging, as implemented in the WINGSLIB software. Using this method for the gas content, a total of 159878 blocks are estimated. Figure 8(A) presents the results of cross-validation and prediction error for the spherical semivariogram model. Also, no correlation should exist between estimation error and estimated values. As shown in Figure 8(B), no correlation exists between estimation error and estimated values.
As indicated by this figure, the spherical model provided an acceptable estimation accuracy, making it the best model for fitting the coalbed gas content data (Olea, 2009). Figure 9 shows a two-dimensional map of the gas content.
Sequential Gaussian simulation (SGSIM) is a simulation method based on the semivariogram that takes advantage of Gaussian random functions (Remy et al., 2009; Gomez Hernandez and Cassiraga, 1994). The simulation or realization results are in the form of spatial patterns corresponding to the semivariograms and input data. The realization can be viewed as a numerical model of the probability distribution function for the attribute simulated in the space. The present paper generated 20 realizations for the considered gas content as the desired attribute using different numbers of grids to ensure that the data are random. These simulations were performed to analyze the uncertainties and distributions of different attributes across the study area. As far as the uncertainty assessment is concerned, the simulation outperforms the kriging (Olea, 2009), although the both SGSIM and universal kriging were applied in this work (in the WINGSLIB software). The realization actually takes the form of certain simulated maps representing an uncertain map at identical probability. Therefore, in each realization (i.e., simulated map), a grid represents the distributions of specific attributes. Such distribution can then be used for the statistical analysis of the data for their variance and the uncertainties associated in the form of probabilities.
4.2Monte Carlo simulation
Monte Carlo simulation is one of the most efficient methods used to analyse complex problems as a computational algorithm (Cheng, 2016). The first step in performing the Monte Carlo simulation is to determine the distribution function for the random variables. The Monte Carlo simulation is mostly used to describe a method to reflect uncertainties at the model input to uncertainties at the model output. Therefore, Monte Carlo is a kind of simulation that explicitly and quantitatively shows the uncertainty. The Monte Carlo simulation relies on the process of explicit representation of uncertainty by designating inputs as probability distributions. If the inputs describing a system are uncertain, then the prediction of the ongoing performance is definitely uncertain. This means that the result of any analysis based on the inputs represented by probability distributions is itself a probability distribution. Many parameters are considered for the safety and disaster control in the mine. Assuming probability distributions and distribution parameters related to random variables, the MCS simulation is first performed for a sample of 100 items. Then, for each parameter, the corresponding probability distribution is selected. In the next step, a random value is obtained for each distribution, and by repeating this process, the most probable solution is obtained as a distribution with a minimum and maximum value. Gas content is a very important parameter for assessing the methane emission through coal seams before and after the mining operations. Therefore, the gas content of a coal seam is one of the important parameters that should be considered for the ventilation purposes. In this study, six important factors for predicting the methane gas were investigated: gas content (m3/t), coal thickness (m), advance rate (m/d), production rate (t/d), gas content of adjacent seam (m3/t), and adjacent seam thickness (m). Gas content, coal seam thickness, advance rate, production rate, adjacent seam gas content, and adjacent seam thickness are uncertain or random variables that affect the air flow required in the face, and the methane emission is a dependent variable. The distribution function is obtained using the historical information based on experience or experiment. For gas content, a normal distribution function with a mean of 8.52 and standard deviation of 4.89 is assumed. Assuming the probability distribution and distribution parameters along with the random variables, the Monte Carlo simulation was first run 100 times. To obtain a good level of probability distribution, the number of Monte Carlo simulation samples should increase. In this study, the Monte Carlo simulations were run 1000 times. It is important to note here that only the gas content uncertainty obtained in the previous step (gas content enters the Monte Carlo simulation in the form of uncertainty) is considered on methane emission and the other parameters are constant.Therefore, by substituting the random values into Equation (1) and performing the evaluation, the system output changes are predictable. The characteristics of the parameters effective in predicting the gas emission are given in Table 3.
Table 3Influential parameters on gas emissions
|
No |
Gas emission (m3/min) |
Gas content (m3/t) |
Coal thickness (m) |
Advance rate (m/d) |
Production (t/d) |
Gas content adjacent layer (m3/t) |
Adjacent layer thickness ( m ) |
|
1 |
7.58 |
6.95 |
2.5 |
3.4 |
725 |
7.04 |
8.8 |
|
2 |
7.44 |
7.37 |
2.6 |
3.6 |
727 |
7.88 |
8.9 |
|
3 |
8.51 |
7.59 |
2.1 |
3.5 |
789 |
7.63 |
9.0 |
|
4 |
8.51 |
7.84 |
2.5 |
3.8 |
768 |
7.91 |
8.9 |
|
5 |
8.42 |
7.36 |
2.3 |
3.5 |
637 |
8.8 |
8.4 |
|
6 |
8.22 |
7.62 |
2.5 |
4.2 |
769 |
8.09 |
7.7 |
|
7 |
8.58 |
7.94 |
1.9 |
3.7 |
692 |
8.30 |
7.4 |
In this study, a combination of two simulation concepts, namely geostatistical simulation to capture in-situ gas content uncertainty, and Monte Carlo simulation, was used to predict coal gas emission based on the in-situ gas content uncertainty. Accordingly, the proper ventilation can be planned for the mine. Determining the amount (quantity) of gas content in coal mines is one of the important parts of coal extraction from deep underground. In a coal mine, the production system is a dynamic, large-scale, complicated system of components that are bound to one another in a time-space domain. The gas emission is nonlinear complex phenomenon (He, et al. 2008) that is affected by a large number of factors. The data affecting the gas emission are always independent of one another and, most of the time, exhibit degrees of nonlinearity that can be obtained upon monitoring the coal mines. Therefore, this requires proposing different methods for determining the amount of gas content at each point in the coal layer.
All the studies and results of modeling need some confirmation levels. This study compared the SGSim results with the actual data to develop the histogram and Q–Q plots from the hard data and the realization produced by SGSim and then calculate the elemental descriptive analysis. Figure 11 schematically demonstrates the analytic approach to evaluate the gas content of the coal bed. The statistical analysis of the data resulted from the simulation of coal attribute maps together with the hard data reported in Table 2 basically show the similarity between almost all basic statistical parameters. This confirms that the mean result of any simulation was similar in terms of range and the hard data range. Nevertheless, it should be noted that the data-derived variance was lower than the variance of the simulation results, as the point-average value for the simulated data on the map exhibited the scattering lower than those observed in the hard data aand independent realizations. Along with the comparison between the basic statistical parameters, which was presented before, the Q–Q plots were developed for each modeled gas content based on the hard data for the independent realizations. Q–Q plots are used to compare probability distributions. On a Q – Q plot, a straight line indicates that the compared distributions are identical, while the data on the two axes exhibit the very similar quantiles (Krishnamoorthy, 2006). The analysis of the Q – Q plots obtained in this work showed that the hard data were related to the realizations along acceptably linear trends. Q-Q plot (Fig. 10) compares the gas content data and simulation results. This figure reveals that the SGSim-derived results were almost linearly associated with the corresponding hard data from which those were produced, i.e., almost all data exhibited the same probability distribution.
Using the results of geostatistical simulation (in-situ gas content uncertainty) along with five other factors, the prediction of gas emission was modeled by the Monte Carlo simulation. As such, the Monte Carlo simulation method analyzes the relationship between gas emission and influencing factors and then predicts the methane emission. In addition, gas content, coal seam thickness, coal production rate, and adjacent seam gas content play a large role in gas emission, which should be addressed when constructing a model for the prediction of gas emission. Table 4 reports the gas emission at a relative error of 2%, implying a high gas emission prediction accuracy and finally, Table 5 lists the required amount of air by substituting into Eq. (3) according to the safety factor of 1.2 in Iranian coal mines.
A comparison between the predicted and measured gas emissions (Table 4 and Fig. 11) indicates that the model could closely predict the real data at an average error of 0.5 and a maximum error of 0.9. Showing the relationship between real data and prediction results, Fig. 12 proves the validity and reliability of the prediction model built based on the proposed system, introducing it as a reliable tool for predicting gas emission in a mine.
|
NO |
Gas emission measured (m3/min) |
Gas emission predicted (m3/min) |
Absolute error |
Relative error |
|---|---|---|---|---|
|
1 |
7.58 |
7.26 |
-0.32 |
-0.042 |
|
2 |
8.27 |
8.58 |
0.31 |
0.037 |
|
3 |
8.85 |
9.09 |
0.24 |
0.027 |
|
4 |
8.61 |
8.97 |
0.36 |
0.041 |
|
5 |
8.42 |
8.58 |
0.16 |
0.019 |
|
6 |
8.22 |
8.62 |
0.4 |
0.048 |
|
7 |
8.30 |
7.94 |
-0.36 |
-0.043 |
|
Predicted Gas emission (m3/min) |
Air flow Required (m3/min) |
|---|---|
|
7.24 |
1584 |
|
8.58 |
1930.90 |
|
9.09 |
1983.27 |
To successfully control the methane gas in the subsurface coal mines, it is required to know the gas emission zones including the gas content and associated uncertainty. The variograms were established along different directions on the normal data at each borehole to evaluate the spatial correlation for the gas content of coal layers. Sequential Gaussian simulation (SGSim) was further run across the studied area for obtaining the spatial distributions of the gas contents. Finally, the geostatistical modeling and simulation methods were found to be advantageous for the quantitative evaluation of gas content. Considering the spatial distribution of gas content and associated uncertainty, the geostatistical techniques and associated estimations can be utilized for improving the design of ventilation system and hence, labor safety. In this respect, the SGSim was found to outweigh the kriging for such applications, since its output results can be assessed irrespective of the mean and variance values to characterize the uncertainty. Various assumptions were analyzed to select the predicted methods for methane emission. In this paper, the amount of gas emitted in the underground mine was predicted using the Monte Carlo simulation method. This paper aimed to develop and understand the impact of extraction and geology on the emission and coal gas during the extraction. It is very difficult to produce an equation to accurately describe the factors affecting the coal gas emission, because many factors affect the methane emission which have complex properties. The Monte Carlo method analyzes the relationship between the factors affecting the coal gas emission and the outputs using known experimental data to generate a prediction model. The following techniques can be developed with these methods:
- Predict methane gas emission and sudden change in emission level.
- Improve methane emission control.
- The relative error of the proposed model is 2% and its absolute error is 0.57.
- Ability to predict the extent of methane emission zone, which in turn leads to the higher efficiency and the use of advanced technology.
The results also show that the proposed model is sufficiently acceptable and accurate.
Acknowledgement
On behalf of all authors, the corresponding author states that there is no conflict of interest.
- Cheng J (2016) Assessment of mine ventilation system reliability using random simulation method. Environmental Engineering and Management Journal, 15(4): 841-850.
- Cheng J, Zhang, H, Xiaokun U (2012) Combination Forecast of Gas Emission in Coal Mine Based on Evidence Theory China Safety Science Journal, 22(1): 106–111.
- Fu H, Shi D (2013) Study on Gas Emission Prediction Model Based on IGA-LSSVM. China Safety Science Journal, 23(10): 51–55.
- Fu H, et al. (2014) Gas emission dynamic prediction model of coal mine based on ACC-ENN algorithm. Meitan Xuebao, Journal of the China Coal Society 39(7): 1296-1301.
- Gómez-Hernández JJ, Cassiraga, EF (1994) Theory and practice of sequential simulation. In: Armstrong, M., Dowd, P.A. (Eds.), Geostatistical Simulation. Kluwer Academic Publishers, Dordrecht, pp. 111–124.
- Harpalani S, Chen G (1995). Estimation of changes in fracture porosity of coal with gas emission,Fuel, 74(10): 1491-1498.
- Hartman H L (1997) Mine ventilation and Air condition, Jhon Wiley & Sons, inc.
- He L, Shi SH, Song Yi (2008) Complexity and measurement of complex degree of gas gush in heading faces of coal mine. Journal of China Coal Society, 33(5): 547–550
- He J, Pan J, Nie B (2006) Research on Fractal Prediction of Severity of Gas Emission, China Safety Science Journal, 16(5): 22–25.
- Jenkins C (2008) Practices and pitfalls in estimating coalbed methane resources and reserves. In: Search and Discovery Article 80011 from Oral Presentation at AAPG Annual Convention, San Antonio, TX.
- Kissell F N, McCulloch C M, Elder C H (1973) The direct method of determining methane content of coal beds for ventilation design، Report of Investigations, 7767, US Bureau of Mines, Pittsburgh, PA.
- Krishnamoorthy K (2006) Handbook of Statistical Distributions with Applications. Chapman and Hall/CRC Press, Boca Raton, Florida. 346 pp.
- Lunarzewski L L.W (1998) Gas emission prediction and recovery in underground coal mines. International Journal of Coal Geology, 35(1-4), 117-14.
- Lei L (2014). Prediction of gas emission quantity using artificial neural networks. Journal of Chemical and Pharmaceutical Research, 6(6): 1653-1657.
- Li R, Shi Sh, Wu A, Luo W, Zhu h (2014) Research on prediction of gas emission based on self-organizing data mining in coal mines. International Symposium on Safety Science and Technology, Procedia Engineering 84, 779 – 785.
- Mucho T P, Diamond WP, Garcia F, Byars J D, Cario S L (2000) Implications of recent NIOSH tracer gas studies on bleeder and gob gas ventilation design، Proc. Annual Meeting of Society for Mining, Metallurgy and Exploration, Salt Lake City, UT
- Olea R A (2009) A Practical Primer on Geostatistics. U.S. Department ofthe Interior. U.S. Geological Survey, Open-File Report 2009-1103. 346 pp.
- Remy N, Boucher A, Wu J (2009) Applied Geostatistics with SGeMS, A User's Guide. Cambridge University Press, Cambridge, United Kingdom. 264 pp.
- Saghafi A, Williams DJ, Lama R D (1997) Worldwide methane emissions from underground coal mining، In: Ramani, R.V. (Ed.), Proceedings of the 6th International Mine Ventilation Congress. May, Pittsburgh, PA, pp. 17–22.
- Schatzel S J, Krog R B, Garcia F, Marshall J K (2006) Prediction of longwall methane emissions and associated consequences of increasing longwall face lengths: a case study in the Pittsburgh Coalbed Proceedings of 11th U.S./North American Mine Ventilation Symposium, pp. 375–382.
- Skochinsky A, Komarov V (1989) Mine ventilation, Mir Publishers.
- Shi Shi, Wu A (2008) Application of GM (1, 1) and line regression for prediction amount of mine gas emission in coal mines. Journal of China Coal Society, 33(4): 415–418.
- Shi Shi, Song Yi, He Li, et al (2006) Research on determination of chaotic characteristics of gas gush based on time series in excavation working face of coal min. Journal of China Coal Society, 31(6): 58–62.
- Shao L (2009). Disaster prediction of coal mine gas based on rough set theory. Journal of China Coal Society, 34(3): 371–375.
- Wang W, Peng L, Wang X (2018) Prediction of Coal Mine Gas Emission Quantity Based on Grey-Gas Geologic Method. Mathematical Problems in Engineering Volume 2018, Article ID 4397237, 7 pages. https://doi.org/10.1155/2018/4397237.
- Wu A, Tian Y, SONG Yi, et al (2005) Application of the grey system theory for predicting the amount of mine gas emission in coalmine. Journal of China Coal Society, 30(5): 589–595.
- Zhang Z M, Wu Y (2014) China coalmine gas-geologic laws and mapping, China University of mining and technology press.
- Zeng Y, Wu C (2004) Research on fuzzy fractal neural network for prediction of mine gas emission. Coal Science & Technology Magazine, vol. 32, no. 1, pp. 62–65.
- Zhou H, Chang W, Zhang B (2007) Different-source gas emission prediction model of working face based on BP artificial neural network and its application. Journal of China Coal Society, 32(5): 504–508
- Zhou F, Yao G, Tyson S (2015) Impact of geological modeling processes on spatial coalbed methane resource estimation. International Journal of Coal Geology. 146, 14-27.
- Zhu S, Wang Y, Wei L (2012) Gas monitoring warning signal identification based on time series similarity measure. journal of China University of Mining and Technology, vol. 41, no. 3, pp.474-480.
- Yang Y, Y Mu (2018) Research on time series characteristics of gas concentration at working face and application of them to early warning. China Safety Science Journal, vol. 3, pp. 120–125.