Abstract
To solve the problems of surface deformation and destruction of buildings caused by urban mining and realise the coordinated development of mining cities, the solid backfilling method was used to extract coal resources beneath the buildings of Tangshan. Based on surface deformation monitoring data of the continuously operating reference station (CORS) system for the past 5 years, the surface deformation process caused by solid backfilling was analysed. The final results revealed a maximum surface subsidence of 66 mm in the T zone coal area and 31 mm in the F zone area. Furthermore, the surface control effects of the caving method and the solid backfilling method were compared and analysed, and it was shown that solid backfilling could meet the surface building setup requirements. Moreover, based on the probability integral method, the effects on surface deformation due to the surface length of the F zone, compression ratio, and coal pillar width were analysed, and the effects on the prediction results due to the subsidence factor, tangent of the major effective angle, and offset distance of the inflection point were studied. The results showed that the compression ratio is the main factor controlling the surface deformation and that it should be kept above 80% for solid backfilling of urban mines. The subsidence factor should be 0.82, and the tangent of the major effective angle should be 2.15 when the surface subsidence of solid backfilling is to be predicted. This paper provides a technical reference for the realisation of urban mining with solid backfilling.
1. Introduction
In recent years, the rapid development of China’s economy has increased the demand for coal resources. In 2019 alone, China’s coal production reached 3.85 billion tons, a year-on-year increase of 4% [1, 2]. Such large-scale coal resource development has brought exhaustion of conventional coal resources, and some mines have had to explore coal resources beneath buildings, railways, and water bodies, which directly affects and restricts the harmonious development of ecology and society around the mining area. China’s production of coal resources from beneath buildings, railways, and water bodies under unified allocation alone has reached 140 billion tons, with the amount of coal resources under buildings accounting for 70% of the total amount [3, 4]. When such coal resources are mined, the surface deformation causes buildings to be stretched, compressed, and bent, thus causing various degrees of damage and collapse hazards to the buildings. Mining is also accompanied by the discharge of a large amount of solid waste gangue, which pollutes the environment and encroaches on the land. All of these problems cause serious conflicts in the development of a mining city.
The coal industry has been actively exploring ways to coordinate the development of coal mining and mining cities and has done considerable research and practical work. Solid backfill mining [5, 6], which uses backfill to control rock movement and surface subsidence, has become the main technology to solve the conflicts between the development of the mine and the city. This technology has existed for some time and has been applied in more than 20 mines in China, effectively solving the problem of redundant resource extraction. However, the long-term practice has shown that even the solid backfill mining method can cause large-scale movement and deformation of the ground surface. Current research on this problem has focused mainly on how the fill body controls rock movement, monitoring of surface movement and deformation, and analysis of disasters caused by surface deformation.
Many scholars have researched these problems. Miao et al. [7–9], Zhang et al. [10–12], and Qiang et al. [13, 14] have advanced the equivalent mining height theory, key layer control theory, main roof control theory, and immediate roof control theory. These studies concluded that solid backfill mining can be regarded as extremely thin coal seam mining, allowing the surface subsidence to be controlled. The control targets for solid backfilling in the rock are the key layer, the main roof, and the immediate roof, with the first being the most difficult target and the last being the most precise target. Guo et al. [15–17], Cha et al. [18, 19], and Yao et al. [20, 21] proposed the use of the probability integral method to describe patterns of surface deformation, established a prediction model parameter system for surface subsidence with solid backfilling, and actually measured surface subsidence results using an electronic total station. Fei et al. [22] and Niu et al. [23] further analysed surface deformation characteristics using the analytic hierarchy process evaluation model and evaluation system.
The methods of surface deformation monitoring used in the aforementioned studies are not continuous and cannot show the surface deformation process caused by solid backfill mining because only the control effect of solid backfilling on the overlying rock layer was investigated. In this paper, taking the Tangshan mine as an example, 5 years of continuous surface subsidence data monitored by the continuously operating reference station (CORS) system were used to study the surface subsidence control effect of solid backfill mining. The effects on surface deformation due to surface length, compression ratio, and working surface arrangement were analysed. The effects of the subsidence factor, the tangent of a major effective angle, and offset distance of the inflection point on the prediction results were also studied. Finally, the ranges of the compression ratio control index and the surface subsidence prediction parameters for solid backfill mining in urban mines were obtained.
2. Research Area and Methods
2.1. Mining Geological Conditions
Tangshan mine, with a field area of 37.28 km2 and a mining area of 55 km2, is located in Lu’nan District, Tangshan City, Hebei Province. It is the only state-owned mega coal mine in China located in a city centre with convenient transportation access. Tangshan mine uses the progressive mining area development method of inclined shafts. At present, there are 7 vertical shafts, which divide the mine into 10 production areas. There are no large-scale faults in the Tangshan minefield, and the main coal seams are 5th, 8th, 9th, and 12th, with average thicknesses of 2.4, 3.7, 3.5, and 6.4 m, respectively. Among these, the 5th coal seam has an immediate roof consisting of 4.8 m thick fine sandstone and a bottom plate of 0.9 m thick mudstone, whereas the 9th coal seam has an immediate roof consisting of 0.7 m thick mudstone and a bottom plate of 1.0 m thick sandy mudstone. The geographical location and working face distribution of the Tangshan mine are shown in Figure 1.

(a)

(b)

(c)
As shown in Figure 2, the production area of Tangshan mine is located beneath the city. The main coal areas are the T zone and the F zone. The T zone has an area of 2.01 km2 and 30.366 million tons of coal, of which 1.427 million tons have been mined to date. There are 67 community and office buildings on the surface. The F zone has an area of 1.27 km2 and 34.413 million tons of coal, of which 34.26 million tons have been mined. Production and living facilities of the mine itself and a park exist at the surface. Figure 2 shows a comparison between the surface and subsurface structures. Figure 3 shows the arrangement of surface deformation observation points.


As examples, the T3292 working face and the F5001 working face of the Tangshan mine are shown schematically in Figure 4. The T3292 working face has a length of 90 m, is elevated at an angle, and was in operation from 1 March 2013 to 1 April 2017. The F5001 working face has a length of 66 m, is also elevated at an angle, and was in operation from 1 January 2016 to 1 September 2017.

The T3292 working face has two surface observation lines, that is, lines A and L (Figure 5), with 22 points online A and 8 points online L. The F5001 working face has one surface observation line, line F, with a total of 22 points. The F5002 working face has four surface observation lines, N, S, W, and E.

(a)

(b)
The test used the surface subsidence monitoring data obtained from 1 January 2015 to 1 December 2019 as a reference to analyse the surface subsidence after mining on the T3292 working face, F5001 working face, and F5002 working face. The monitoring data span more than 5 years and so can reflect the long-term surface deformation after solid backfill mining.
2.2. Solid Backfilling Control Method
The solid backfill mining method, which is carried out through integrated mechanised mining operations on working faces, is used for mining under urban buildings. In the solid backfill coal mining system, the gangue is transported from the surface feeding well and underground separation chamber to the working faces, is then filled into the goaf using key equipment such as the multihole bottom dump conveyor and hydraulic support system for solid backfilling, and is finally compacted by the ramming mechanism behind the hydraulic support system. From 2015 to 2019, the surface feeding well accumulated 604,600 tons of gangue, and the underground separation system accumulated 421,100 tons of gangue. The backfilling system, shown in the basic concept in Figure 4, is running well.
Inspection results of the surface buildings show that the main structures of the surface buildings above Tangshan mine are brick–concrete, brick–wood, and steel structures, which meet the national requirements for Class II–III protection. When mining the coal resources beneath the city, the buildings must be kept below the Class I damage level after mining, that is, horizontal deformation less than 2.0 mm/m, curvature less than 0.2 mm/m2, and inclination less than 3.0 mm/m.
2.3. Analysis Method of Surface Deformation
2.3.1. Monitoring Method
Monitoring the subsidence and deformation of the surface, buildings, and other structures is an important task in the process of mining beneath a city. The conventional monitoring methods used for the Tangshan mine include mainly theodolite tracking, total station monitoring, and water level monitoring. The main problems faced in the monitoring process are addressed below. The accuracy of the conventional monitoring methods is not very high. Moreover, the detection process is easily affected by rain, snow, and the mutual obstruction of the surface buildings. In addition, conventional monitoring systems are not capable of continuous monitoring.
The CORS system for intelligent monitoring of surface subsidence can carry out continuous and uninterrupted observation and achieve real-time collection, transmission, calculation, and analysis of data within the subsidence range. The CORS system applied in the Tangshan mine is equipped with China’s Beidou Satellite Navigation System, which is compatible with the United States’ GPS system and Russia’s GLONASS system, to improve the accuracy of surface subsidence deformation measurement. The CORS system basically consists of a data centre, a reference station, a data communication subsystem, and a user application subsystem. It achieves all-weather, fully automatic, and precise positioning through satellite positioning technology, the reference station, and the Internet. The base station and Class I and II measuring points of the Tangshan mine are shown in Figure 6.

2.3.2. Prediction Method
The probability integral method is used to estimate the surface movement and deformation. Its calculation principle is as follows.
The calculation of subsidence value at any point on the surface is shown in equation (1). The calculation of the inclination deformation value of any point on the surface is shown in equation (2). The calculation of curvature deformation value of any point on the surface is shown in equation (3). The calculation of horizontal movement value of any point on the surface is shown in equation (4). The calculation of horizontal deformation at any point on the surface is shown in equation (5).
where W0 = Me·q·cosα and r = H/tanβ. Me is the equivalent mining height of solid dense filling mining; q is the surface subsidence coefficient of solid dense filling mining; b is the horizontal moving coefficient; tanβ is the tangent of the main influence angle; and α is the dip angle of mining seam.
2.3.3. Test Scheme for the Cause of Surface Subsidence
Taking the F zone as an example, under the mining geological conditions of the Tangshan mine, it is considered that the surface subsidence above the F zone was influenced by the face length, the compression ratio of the working face, and the width of the coal pillar in the section. To analyse the causes of surface subsidence, 15 sets of test plans were designed, as shown in Table 1. Tests 1–5 analysed the effect of face length on surface deformation parameters, tests 6–10 analysed the effect of compression ratio on surface deformation parameters, and tests 11–15 analysed the effect of coal pillar width on surface deformation parameters.
The surface deformation prediction uses the probability integral method based on equivalent mining height. Due to the presence of a large number of primary fissures, joints, and strata in the subsiding rock, it is feasible to predict surface movement and deformation by the probability integral method of the random medium.
In comparison with the caving method, when the probability integral method is used to predict surface subsidence for solid backfill mining, prediction parameters such as the subsidence factor, tangent of a major effective angle, and offset distance of the inflection point change greatly. Therefore, these prediction parameters are considered to have a relatively large influence on the prediction results. Using a compression ratio of 0.8, 15 sets of test plans were designed to analyse the prediction results of surface deformation, and the plans are shown in Table 2. Tests 1–5 analysed the influence of the subsidence factor on the projected results, tests 6–10 analysed the influence of the tangent of the major effective angle on the prediction results, and tests 11–15 analysed the influence of the offset distance of the inflection point on the surface deformation parameters.
3. Results and Interpretation
3.1. Monitoring and Analysis of Filling Surface Deformation
3.1.1. Surface Monitoring Results in Area T
The T3292 working face was put into production in March 2013 and stopped being used in April 2017. Line A, measuring surface subsidence in the T zone, crosses the middle of the T3292 working face, and the projected positions of the selected measuring points below the working face are located at a distance of 340 m from the open-off cut of the working face. The surface subsidence data of each measuring point online A were obtained by processing the monitoring data, as shown in Figure 5(a). The surface subsidence data of each time period at the A9 measuring point are shown in Figure 5(b).
As shown in Figure 5(b), the surface subsidence increased slowly from the start of mining on the working face to June 2015, with the subsidence value lower than 25 mm. Between June 2017 and August 2017, after the working face was pushed 90 m beyond measuring point A9, the surface subsidence value increased significantly, with its maximum value reaching 41 mm at measuring point F16. In April 2017, the working face had been mined completely. By December 2019, the T3292 working face had reached full mining impact, and its maximum subsidence had increased to 66 mm. From the time of sudden subsidence to full mining impact, the subsidence value increased by approximately 37.8%.
The CORS monitoring system indicated a final maximum surface tilt deformation of 0.76 mm/m, a maximum curvature deformation of 0.064 mm/m2, and a maximum horizontal deformation of 50.6 mm, or 1.53 mm/m, which meets the surface building protection requirements.
3.1.2. Surface Monitoring Results in Area F
(1) F5001 Working Face. The F5001 working face was put into production in October 2016 and stopped being used in September 2017. Line F, measuring surface subsidence in the F zone, crosses the middle of the F5001 working face, and the projected positions of the selected measuring points under the working face are located at 400 m from the open-off cut of the working face. The monitoring data were processed to obtain the surface subsidence data of each measuring point online F, as shown in Figure 7(a). The surface subsidence data of each time period of measuring point F9 are shown in Figure 7(b).

(a)

(b)
As shown in Figure 7(b), the surface subsidence increased slowly, with subsidence values lower than 10 mm, from the start of the mining on the working face to May 2017. Between June 2017 and July 2017, after the working face had pushed through 66 m from measuring point F9, the surface subsidence increased significantly, with a maximum value of 19 mm at measuring point F16. In November 2017, the working face was mined completely, and by April 2018, the F5001 working face had reached full mining impact, and its maximum subsidence had increased to 31 mm. From sudden subsidence to full mining impact, the time elapsed was approximately 8 months, and the subsidence value had increased by approximately 38.7%.
The CORS monitoring system indicated a final maximum surface tilt deformation of 1.65 mm/m, a maximum curvature deformation of 0.094 mm/m2, and a maximum horizontal deformation of 16.1 mm, or 1.25 mm/m, which meets the surface building protection requirements.
(2) F5002 Working Face. The F5002 working face was put into production in November 2017 and stopped being used in December 2018. Line E, measuring surface subsidence in the F zone, crosses the middle of the F5002 working face, and the projected positions of the selected measuring points below the working face are located at 400 m from the open-off cut of the working face. The monitoring data were processed to obtain the surface subsidence data of each measuring point online E, as shown in Figure 8(a). The surface subsidence data of each time period of measuring point E3 are shown in Figure 8(b). There was almost no increase in surface subsidence from the start of mining on the working face to September 2018. Mining on the working face was fully stopped in December 2018, and by July 2019, the surface subsidence of the F5002 working face had reached 17 mm. By December 2019, the full mining impact had been reached, and the maximum increase in subsidence was 25 mm. From sudden subsidence to full mining impact, the time elapsed was approximately 5 months, and there was an increase in subsidence value of approximately 47.1%.

(a)

(b)
The CORS monitoring system indicated a final maximum surface tilt deformation of 0.29 mm/m, a maximum curvature deformation of 0.012 mm/m2, and a maximum horizontal deformation of 7.84 mm, or 0.81 mm/m, which meets the surface building protection requirements.
3.2. Analysis of Surface Deformation Monitoring by Caving Method
Observation of the surface movement and deformation caused by mining on the F5009 working face started in February 2018, and a total of 27 observations were made until August 2020. Line N, measuring the surface subsidence in the caving area, crosses the middle of the F5009 working face, and the projected positions of the selected measurement points below the working face are located 400 m from the open-off cut of the working face. The surface subsidence data of each measurement point online N are shown in Figure 9(a). The surface subsidence data of each time period of measuring point n4 are shown in Figure 9(b).

(a)

(b)
As shown in Figure 9(b), in comparison with solid backfill mining, the initial value of surface subsidence of the caving method was larger, reaching 60 mm. As mining on the working face continued, the value of surface subsidence increased gradually without any sudden increase. The maximum value of subsidence of measuring point n4 reached 210 mm by August 2020, when all of the mining operations on the working face had been completed.
The CORS monitoring system indicated a final maximum surface tilt deformation of 6.65 mm/m, a maximum curvature deformation of 0.59 mm/m2, and a maximum horizontal deformation of 168.1 mm, or 4.25 mm/m, which does not meet the surface building protection requirements.
3.3. Comparative Analysis on Control Effect of Different Mining Methods
A comparison of surface deformation parameters between the solid backfill mining method and the caving method is shown in Table 3.
In addition, the relationship between surface subsidence values and monitoring time was analysed for T3292, F5001, and F5002. The monitoring data were acquired at the time points of the goaf square period, end of mining, 3 months after mining, 1 year after mining, and 2 years after mining. The analysis results showed that the surface subsidence of backfill mining was significantly affected by monitoring time. Although the surface did not show large subsidence during the mining, surface subsidence increased significantly from 3 months to 1 year after mining. The time of 2 years after backfill mining can be regarded as the time when mining was completed, and the surface subsidence above the working face did not increase further at this time.
3.4. Determination of Surface Prediction Parameters in Tangshan Mine
According to the surface subsidence monitoring results of the Tangshan mine of previous years, the subsidence factor of the Tangshan mine with the caving method was determined to be between 0.55 and 0.85; the tangent of the major effective angle was between 1.92 and 2.40; and the offset distance of the inflection point was 0. Through parameter inversion of the surface subsidence in the F zone, the subsidence factor of backfill mining was determined to be 0.028, and the tangent of the major effective angle was 1.20, as shown in Table 4.
As shown in Table 4, the surface subsidence prediction parameters currently obtained from monitoring one to two working faces are not accurate, as the subsidence factor is extremely low. The prediction with these parameters is thus not representative of the surface subsidence after mining in the area. Therefore, by comparing the predicted results of surface subsidence, considering multiple factors in the F zone, the subsidence factor appropriate for the predicted subsidence with solid backfill mining of underground coal in the Tangshan mine was determined to be 0.82, and the tangent of the major effective angle is 2.15. At this time, the predicted result of surface subsidence is 152 mm, which can be regarded as the final amount of surface subsidence after the end of mining in the F zone.
4. Discussion
Surface subsidence is greatly affected by mining areas; it is necessary to analyse the influencing factors of surface subsidence. At the same time, the parameters of surface subsidence prediction will change accordingly. An analysis software was used for coal mining subsidence prediction, and the results are shown in Table 5. The surface subsidence, horizontal deformation, curvature, and tilt deformation under different influencing factors were analysed and compared. The test results obtained are shown in Figure 10.

(a)

(b)

(c)
As shown by the analysis in Figure 10, the subsidence, horizontal deformation, curvature, and tilt deformation increase gradually as the face length increases gradually. The subsidence, horizontal deformation, curvature, and tilt deformation decrease gradually as the compression ratio increases gradually, and the subsidence, horizontal deformation, curvature, and tilt deformation do not change much with an increase in coal pillar width.
In particular, when the compression ratio is 0.3, that is, the caving method is used to deal with the hollow area, the surface deformation exceeds the set damage level; thus, the compression ratio is the main influencing factor of surface deformation. From the predictive analysis of surface deformation, it is apparent that the compression ratio should be 0.6 or more to ensure the safety of surface buildings. In consideration of the influence due to early subsidence of the roof of the mining area, the compression ratio should be at least 0.8 in an actual backfilling operation.
An analysis software was used for coal mining subsidence prediction, and the results are shown in Table 6. The surface subsidence, horizontal deformation, curvature, and tilt deformation under different influencing factors were analysed and compared. The test results obtained are shown in Figure 11.

(a)

(b)

(c)
From the analysis shown in Figure 11, as the subsidence factor increases gradually, the subsidence increases gradually while the horizontal deformation, curvature, and tilt deformation tend to stabilise. As the tangent of the major effective angle increases, the subsidence, horizontal deformation, curvature, and tilt deformation increase gradually. As the offset distance of the inflection point increases, the subsidence, horizontal deformation, curvature, and tilt deformation do not change.
It can be seen that the subsidence factor is the main influencing factor of the surface subsidence prediction results and that the tangent of the major effective angle is the main factor affecting the prediction results of surface horizontal deformation and tilt.
5. Conclusion
(1)The CORS system was used to monitor the surface subsidence of coal mining under Tangshan city, which spanned more than 5 years and covered the T zone and the F zone, corresponding to a mining area of 150,000 m2. The monitoring results showed that the surface protection requirements of urban mining were met by solid backfill mining, which had gentler surface deformation than caving mining. The maximum subsidence in the T zone after 32 months of working face mining was 66 mm, and the maximum subsidence in the F zone after 6 months of working face mining was 31 mm.(2)Through the multifactor surface subsidence prediction based on the probability integral method, it was concluded that surface subsidence is strongly influenced by the working face length and the compression ratio. The surface deformation increases gradually as the face length increases and the compression ratio decreases. To ensure the safety of surface buildings, the compression ratio must be at least 0.8.(3)By comparing the actual measurement data in the field with the quality simulation data, the subsidence factor suitable for the Tangshan mine was finally determined to be 0.82, and the tangent of major effective angle was determined to be 2.15, at which time the predicted results are close to the subsidence results after extensive backfill mining in the mining area.Data Availability
The data used to support the findings of this study are included within the article.
Disclosure
This manuscript was presented as a preprint in research square (https://www.researchsquare.com/article/rs-221271/v1).
Conflicts of Interest
The authors declare that there are no conflicts of interest.
Authors’ Contributions
All authors contributed equally to this research.
Acknowledgments
The authors would like to thank the staff of Tangshan mine who worked hard to monitor and got precious measured data. All authors are grateful for the financial assistance provided by the Independent Research Project of the National Key Research and Development Program of China (2018YFC0604704).