Determining the Quality of Mine Gushing and Mixed Water Using Coupled AHP and Fuzzy Comprehensive Evaluation Methods

This study focused on analysis of the chemical characteristics of mine waters. The aim of this study is to correlate the degree of different ionic components in mine water and the influence of their convergence using a combination of the three-scale AHP and fuzzy evaluation methods for the comprehensive evaluation of water quality. Ion chromatography (ICS 1100) has been used to analyze the content of the water sample while portable pH/EC/TDS/Temperature meters (SX 811 and SX 813) were used to test physical-chemical parameters. The results of this study show that chemistry of in No.11 gushing mine is dominated by HCO3-Na and HCO3-Ca, and had a pH between 7.1 and 8.00, belonging to neutral or slightly alkaline water. In addition, water were found to have the hardness between 18 mg/L and 542.5 mg/L. Results also show that the TDS of the roof sandstone and goaves water are higher than Cambrian limestone water, while the turbidity of the mixed water is 20 NTU in the sump, again higher than in other samples such as Cambrian limestone water. Total dissolved solids and the total hardness of Cambrian limestone groundwater mainly depend on the content of K + Na, Ca, { } 1 2 , , , j B b b b =  and 4 SO − . Thus, chemical composition changes remarkably after mine water mixing. Results showed that the coal roof sandstone water is class V while that in the sump is class III, and the Cambrian limestone groundwater is class I. In gushing, the quality of water can vary greatly; thus, water from the coal face roof sandstone and the Cambrian limestone should be stored and treated separately before being utilized. *First author. How to cite this paper: Wang, J.Z., Zhao, W., Wang, X.Y., Hersi, N.A.M., Zhang, P.Q. and Sang, X.Y. (2018) Determining the Quality of Mine Gushing and Mixed Water Using Coupled AHP and Fuzzy Comprehensive Evaluation Methods. Journal of Water Resource and Protection, 10, 1185-1197. https://doi.org/10.4236/jwarp.2018.1012070 Received: November 27, 2018 Accepted: December 26, 2018 Published: December 29, 2018 Copyright © 2018 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access


Introduction
In the course of coal mining, a large amount of mine water needs to be drained to ensure the safety of underground production.In China's northern coalfields, the annual discharge of mine water is about 1.787 billion m 3 and the average discharge of mine water per ton of coal is about 1.29 m 3 , but the average utilization rate is less than 25% [1] [2].Due to the impact of coal mining activities, the water discharged can contain acidic substances, heavy metals, organic compounds, radioactive elements, bacteria, and other harmful substances that seriously pollute rivers and soils.Because this process also wastes huge volumes of water [3] [4] [5], appropriate strategies for treatment and reuse are of great significance to coal mining ecological and environmental protection.
The key to effectively utilizing and managing mine water is understanding its quality [6].A number of mathematical models have been proposed to comprehensively evaluate water quality, including the analytic hierarchy process (AHP) [7], entropy weight analysis [8], principal component analysis [9], multivariate statistical methods [10], fuzzy comprehensive evaluation methods [11], and artificial neural networks [12].These methods compare measured data with water quality standards to enable comprehensive evaluation.Singh [13] compared water chemistry indexes of mine water with corresponding standard values to determine their effectiveness but did not comprehensively evaluate the influence of various indicators on mine water quality.In later work, Sun [14] applied the fuzzy comprehensive evaluation method to a comprehensive evaluation of coal mine water quality in the arid area of western Chongqing; the results of this study are problematic, however, because this complex method requires the weight determination of each water sample separately.Similarly, Liu [15] used the gray clustering method to evaluate environmental water quality in the Dongsheng coal field, Inner Mongolia.This method applies threshold values to determine the weight value of each index, and therefore does not take into account changes in factors within the same level and their influence on water quality evaluation.Most recently, Gao [16] utilized a SPA-ITFN(Set Pair Analysis-Interval Triangular Fuzzy Numbers) coupling model for coal mine groundwater quality evaluation and showed that when the value for a measured factor is larger, weight will also increase; these results are therefore of importance for water quality evaluation.
Mine water gushing contact with coal and rock formations, A series of physical, chemical and biological reactions occur, coupled with the impact of mixed with the production of waste water, which dissolved the chemical composition becomes very complicated [17].Due to the limitations of the existing discriminant models, the evaluation results of mine drainage water quality often do not accord with the actual situation, which limits its scientific utilization.It is necessary to find a discriminant model that can be realized automatically and the evaluation results conform to the reality.
This paper evaluates mine water gushing from Pingdingshan No. 11 coal mine.A quality evaluation was conducted using AHP (Analytic Hierarchy Process) in tandem with the fuzzy comprehensive evaluation method, and the characteristics of mine water chemistry, ion composition, and confluence were analyzed.The AHP simplifies the process of determining the weight of the fuzzy comprehensive evaluation method and avoids human factor interference; this approach therefore makes the evaluation of mine water quality more scientific and provides a more reliable evaluation.The results of this paper provide a reference method for the automatic and reasonable evaluation of mine drainage water quality.

Overview of the Study Area
Pingdingshan No. 11 mine is located in the southwestern margin of this mining area, within the transitional zone between northern subtropical and warm temperate monsoonal climate.The annual average temperature is 15˚C and annual average rainfall is 747 mm.About 70 percent of annual rainfall occurs between July and September.
The water source within the Ji group of No. 11 mine is Cambrian and Carboniferous limestone, Permian sandstone, and old goaf water.No. 11 mine is located within the Cambrian limestone recharge area to the south of the Pingdingshan mining area; where the recharge water source is mainly groundwater, surface water, and shallow aquifer groundwater.As this recharge source is abundant and relatively close, this results in a large volume of water inflow during the mining process, more than 500 m 3 /h on average.It is therefore of paramount importance to evaluate the water quality of mine and mixed water to enable the rational use of valuable resources.

Water Sample Collection
The samples used in this study from No. 11 mine include mixed water from the sump (denoted C1, C2, and C3), drilling into Cambrian limestone (H1, H2, and H3), roof sandstone fissures water (D1) and goaf water (K1).The positions of these sampling points are shown (Figure 1 and Table 1); All samples were analyzed at the Henan Geological Environment Monitoring Institute.Simple analyses were performed on samples C2 and C3 and the remainder were subjected to full analysis.The simple analysis includes the common ions, and the full analysis includes all ions, etc.

Chemical Characteristics of Water
Test results of conventional hydrochemical components are shown Table 1.A piper diagram was constructed based on the six major ions was shown as Figure 2 [3].Thus, four water sample groups within this analysis were characterized as HCO 3 -Na (D1, K1, C1, and C3), three were HCO 3 -Ca (H1, H3, and C2), and one was HCO 3 -Mg (H2).The data presented in Table 1 shows that because the pH of all samples is between 7.1 and 8.0, this is neutral or weak alkaline water.In addition, the turbidity of mixed water in the central water sump (C1) is 20, while the remainder of samples are less than 1; this is because the central water sump contains a large volume of coal dust, which leads to an increase in turbidity.The TDS values for samples D1 and K1 were 2777.83mg/L and 1156.65 mg/L, respectively, and so these are classified as brackish water (1000 mg/L < TDS < 3000 mg/L), while the rest of these samples have TDS values less than 1000 mg/L and so are classified as freshwater (TDS < 1000 mg/L) [18] [19].The reason is that the D1 and K1 samples were taken from coal roof and goaf, respectively.The total hardness of these samples falls between 18 mg/L and 542.5 mg/L (average: 309 mg/L), within the range of hard water.

Analysis of Ion Composition
A correlation analysis was performed to measure the relationship between two groundwater variables and to demonstrate source consistency and variability.
According to mathematical statistics theory, when r > 0.9, it indicates that there is a significant correlation between the two variables; when 0.6 < r ≤ 0.9, it indicates moderatet correlation between the two variables; When 0.4 < r ≤ 0.6, it indicates low level correlation between the two variables, similar, When r ≤ 0.46, it indicates little correlation between the two variables [20] [21].
Pearson correlation coefficients for Cambrian limestone groundwater were calculated using the software SPSS (Table 3).These data show that the correlation coefficients of TDS and total hardness and K + + Na + are 0.985 and 0.988, respectively.The correlation coefficients of TDS and Ca 2+ reached 1. And, the correlation coefficients of TDS and total hardness and 3 HCO − are 0.998 and 0.992, respectively.In addition, the correlation coefficients of TDS and total hardness and 2 4 SO − are 0.999 and 1, respectively.This indicates that TDS and total hardness, K + + Na + , Ca 2+ , SO − are significantly correlated.In other words, The concentration of TDS and total hardness in Ordovician limestone groundwater is mainly determined by the content of K + + Na + , Ca 2+ ,

The Influence of Confluence on Mine Water Quality
The data presented in Figure 1 show that mine water flows into a sump in the Ji SO − and 3 HCO − , the difference between the calculated value and the measured value of most ions is small.In addition, the trend of the connection between the simulated value and the measured value is consistent.This shows that the software simulation results of mine water confluence can be used to estimate the content of ion in mine water.
From Table 4, we know that values of TDS changed from 667.16 mg/L (H2) to 795.3 mg/L when Cambrian groundwater (H2) was mixed with water from the Ji 4 water sump (C3) and the goaf (K1).Thus, TDS increased from 683.3 mg/L (H3) to 1375 mg/L while water quality changed from freshwater to brackish when the Cambrian limestone groundwater (H3) was mixed with that from the Ji 2 water sump (C2) as well as 16,150 roof water from the working face (D1).
In addition, TDS fell to 1071 mg/L and total hardness increased to 305.9 mg/L when this was mixed with water from H2 + C3 + K1 and Cambrian limestone groundwater (H1) as the latter was combined with sake.The above results comprehensively show that the mine water with different ion concentrations will change its ion concentration due to the confluence and affect the overall quality of the water, ultimately.

Establishment of a Fuzzy Relationship Matrix
A set of evaluation factors were created on the basis of the chemical composition of test results from sampling points.Thus, on this basis, those with greater impact on the water quality of chemical indicators were selected, as follows: U = {Cl − , NO − , F − , Ba, TDS, Total hardness, Turbidity}.Chinese Groundwater Quality Standards (GB/T 14848-93) have established a set of five evaluation factors V = {I, II, III, IV, V} [22]; Table 5 shows selected standard values for groundwater chemical grade classifications.
Take water sample H1 is as an example to state evaluation process.According to the comparison of the evaluation factor matrices and the evaluation rating matrices, the linear membership function is used to calculate the membership of each pollution factor [23], the applied formula as follows: ) ( ) where x is the measured value; i is evaluation rank vaule, μ is the membership of the evaluation factor.Thus, the corresponding Fuzzy relationship matrix is calculated as follows: 1 0.883 0.117 0.000 0.000 0.000 0.000 0.300 0.700 0.000 0.000 1.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000 1.000 0.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 0.034 0.966 0.000 0.000 0.000 0.000 0.07 0.925 0.000 1.000 0.000 0.000 0.000 0.000 And the membership matrix of other water samples are calculated as in sample H1.

Weight Determination Using AHP
The AHP method [24] was initially used to compare each element and to establish a comparison matrix, as follows: Thus, in this matrix is more important than and are just as important The following formula is then used to build an optimal transfer matrix, , ik jk q q are the element in the comparison matrix Q.
( ) ( ) Next, after an optimal transfer matrix is established, a judgment matrix is calculated using the formula ( ) , as follows: 1.000 0.717 1.000 0.368 0.295 1.560 0.329 0.513 0.895 Finally, in order to determine the weight of water chemical indicators, the following formula is applied [25]:

Figure 2 .
Figure 2. Piper chart of water samples.

2 and
Ji 4 mining areas.This flow then raises mine water into the −593 m west tunnel through the pump house, flows to the central water sump, and then into the ground.Results from the software Aq•QA used to study the water chemical ions in samples are shown in Table 4. Simulation calculation value of a mixture of various sources of water (H2 + C3 + K1 + H3 + C2 + D1 + H1), based on Aq•QA software and measured value are shown in Table 4 and Figure 3. Obviously, except for2 4

Figure 3 .
Figure 3.The comparison between the Aq•QA software simulation values and measured values.

Table 1 .
Conventional water chemical composition of mine water.
Figure 1.Distribution map of water samples collected from No. 11 mine.

Table 2 .
Chemical composition of trace water in mine water (mg/L).

Table 3 .
Pearson correlation coefficients for Cambrian limestone groundwater.

Table 4 .
The major ion changes under different water samples mixing (mg/L).

Table 5 .
Water quality grading standards based on evaluation factors (mg/L).