Investigation into support systems in SA collieries

A study of falls of ground in South African collieries by van der Merwe et al. (2000) concluded that the majority of falls of ground occur under supported roof. For this reason it was decided that roof support systems should be investigated for the purpose of obtaining an understanding of the fundamental mechanisms of roof support systems and developing guidelines and design methodologies for their improvement. To this end all of the currently available roof bolt support elements and related machinery were evaluated underground in three different rock types, namely sandstone, shale, and coal.
Roof bolts are available in many different forms. Full-column single-resin bolts, full-column slow- fast combination resin bolts, resin point anchors, and mechanical anchors are the most widely used support systems in South Africa.  
There are five important components of a bolting system, namely:  
  • Resin;
  • Bolt;
  • Hole;
  • Machinery/equipment; and  
  • Rock type
As part of this study, important parameters of these five components were investigated.
A detailed literature review showed that since the introduction of mechanical bolts in the 1940s a significant amount of research has been carried out on understanding the behaviour of roof bolts. Today, almost all coal mine roofs in South Africa use roof bolts for roof support.
In the early years, the design of roof bolt patterns was based on local experience and the judgement of mining personnel. The suspension mechanism was the most easily understood and most widely used roof bolting mechanism. However, significant advances have been made over the last 20 years in the development of resin anchors, tendon elements and installation hardware, advances which have resulted in an increased use of full column resin bolts.   
The design of roof bolt patterns has also improved, and four main rock reinforcement techniques have been developed: simple skin control, beam building, suspension and keying. 
The geology and the stress levels determine the mechanism required for a particular application.
Investigations into the causes of roof falls in South African collieries highlighted that the roof bolt densities were relatively low compared to those found in the USA, the UK, and Australia. It was concluded, therefore, that the main cause of falls of ground was the excessive bolt spacing and the skin failures between the bolts that this brought about.
The importance of tensioning roof bolts remains a subject of controversy. This report shows that the critical roof deformations in South African collieries are relatively small, therefore tensioned roof bolts may well be required to reduce roof deformations after the installation of support. Short encapsulation pull tests showed that pre-tensioning reduced the system stiffness, though the point was made that the testing procedure may not be well suited to evaluating tensioned bolts and therefore may have produced sub-standard results.
The selection of roof bolt type for different geological environments is well documented. However, the changing conditions underground must be determined and the design and the support system have to be modified accordingly. Therefore, widespread instrumentation and vigilant visual observations are important for ensuring safety and stability in coal mines.
While the effect of roof bolt diameter on support performance is well understood, controversy remains over the correct length of the roof bolts. Since skin failures (< 0.5 m thick) are more common in South Africa (Canbulat and Jack, 1998, van der Merwe and Madden, 2002) than larger roof falls, short roof bolts for skin control may be an effective support for a stiff system. The length of roof bolts, however, should be determined through in situ monitoring and assessment of the roof strata.   
Despite the fact that roof bolting has been the most researched aspect of coal mining, falls of ground remain the major cause of fatalities in South African coal mines. There is no commonly accepted design approach for underground coal mines. Roof bolts have been found to behave differently under different loading conditions, despite being tested in fully controlled environments in laboratories. The most important key to the design of roof support systems is a better understanding of roof behaviour in different geotechnical environments through continuous in situ monitoring.
A detailed investigation into the specifications of roofbolters that are currently being used indicated that the quality of installation of a support system is directly related to the performance of the equipment that is used to install the bolts. For this reason the performance of bolting equipment was investigated as part of this study in order that the range and relative importance of the various machine parameters could be ascertained. The study showed that there are no standards in South Africa for the parameters investigated (speeds, torque, and thrust). The variations in these parameters were found to be greater than previously believed.
The relationship between hole profile and speed, torque, and thrust was investigated. The following values for roofbolter parameters are recommended for optimally producing rough walled holes in South African coal mines:  
Spinning speed 450 rpm
Torque 240 Nm
Thrust 15 kN
The performance of roof bolts that are currently supplied to South African mines was also investigated. A series of short encapsulated pull tests in shale indicated that, on average, bond strengths obtained from the roof bolts supplied by Manufacturer “C” (referred to in the report) were approximately 18 per cent and 28 per cent greater than those obtained from the roof bolts supplied by Manufacturers “A” and “B”, respectively.  
To determine whether variations in the profile of bolts supplied by the different manufacturers could account for the differences in performance, the bolt-core diameters and rib diameters from different bolt manufacturers in South Africa were measured.
The parameters that determine the contact strength between bolt and resin are rib-height, spacing between the ribs, and the rib angle. An investigation was conducted into the dimensions of roof bolts that are used currently. The results showed insignificant differences between the parameters that determine the bolt profile of South African roof bolts. Owing to the physical similarity between the bolts studied, it was not possible to determine the influence of these parameters. On specifically manufactured or imported bolts that have a different configuration, it is recommended that a laboratory-testing programme be carried out to determine the effect of these parameters on the performance of roof bolts being used in South Africa.
The effect of rib angle was investigated and the results of a literature search showed that, as the rib angle increases away from normal to the bolt axis, so the pull-out load of the bolt decreases. It is therefore suggested that, in order to achieve relatively high pull-out loads, low rib angles on the bolts are required. This was confirmed by laboratory tests on different bolts with different rib angles in Australia (O’Brien, 2003). However, lowering the rib angle may result in poor resin mixing performances. It is therefore recommended that further work on the effect of bolt profile on rockbolt performance be carried out, with the aim of achieving failure on the rockbolt-resin interface. It is also recommended that the quality of resin mixing should be investigated for different rib angles in order that the most effective rib angles for roof bolts can be determined. Unfortunately, because rib configurations in South African bolt types are very similar and because testing took place in an underground environment (uncontrolled conditions), the effect of rib angle, rib height and thickness and spacing between the ribs could not be quantified. It is, therefore, suggested that these tests should be conducted in a controlled laboratory environment.
A conceptual model was developed to determine the effect of bolt profiles. This model indicated that maximum pull-out loads can be achieved between the resin and roof bolt when:  
  • The ribs are relatively high;
  • The distance between the ribs is relatively low; and
  • The ribs are relatively thick.  
An attempt to determine the effect of spinning parameters on resin characteristics showed that the gelling time decreases with an increase in free rotation speed. It is therefore suggested that the resin spinning times should be adjusted to improve resin performance.
A series of short encapsulated pull tests indicated that in the majority of pull tests, failure took place at the rock-resin interface, indicating that the rock failed before the resin shear strength had been reached. It is therefore suggested that the strength of resin currently being used in South Africa is adequate. However, the stiffness of the system of which resin is a part should be determined by short encapsulated pull tests.  
The conceptual model developed as part of this project was used to determine the effect of resin in the support system. It is concluded that the failure characteristics of a roof bolting system will be determined by the shear strength of bolt, resin, and rock.
  • The failure will take place at the resin-rock interface when the shear strength of the rock is lower than the resin (rock will fail).
  • The failure will take place at either the resin-rock or resin-bolt interface when the resin shear strength is the lowest in the system.
  • When the resin shear strength is the lowest in the system, the failure will be determined by the roughness of the hole and the bolt profile.   
The test results showed that the reinforcing system using bolts from all four manufacturers performed almost identically in sandstone, but performed in different ways in the other rock types. The bolts from Manufacturer “A” performed slightly better in coal and shale rock types than the bolts from other manufacturers.
The performance of resins that are currently being used in South African collieries was also investigated by means of short encapsulated pull tests. The results indicated that in sandstone the resin types from the two different manufacturers performed similarly. However, the strength of slow (5/10-minute) resins from both manufacturers was much lower than that of fast resins. The results also indicated that 15-second and 30-second resins from Manufacturer “A” achieved higher stiffnesses than resins from Manufacturer “B” in sandstone and coal. In shale, both resins from each of the manufacturers performed in a similar manner.
In order to investigate the effect of bit types, a series of short encapsulated pull tests were conducted. The results showed that the 2-prong bit outperformed the spade bit in sandstone and shale rock types. However, the annuli obtained from the 2-prong bit were always greater than the spade bit. It is thought that this is because 2-prong bits drilled a rougher hole profile. Both the stiffness and the maximum load obtained from the 2-prong bits was greater than for the spade bits. These findings suggest that 2-prong bits are more effective in collieries than spade bits are.
The effect of hole annulus was also investigated. The results show that an annulus between 2.8 mm 4.5 mm resulted in the most effective bond strengths. Another interesting point is that as the annulus drops below 2 mm, it appears to have a negative effect on the grip factors.
The effect of wet and dry drilling was also investigated by means of short encapsulated pull tests. The results showed that bond strengths and overall support stiffnesses are greater with the use of the wet drilling in all three resin types.   
Tensioned versus non-tensioned bolts is one of the most discussed topics in roof bolting. A number of papers have been published on this topic in Australia and the US. An additional 25 short encapsulated pull tests were conducted to determine the effect of tensioning on bond strength. The results showed that non-tensioned roof bolts achieved significantly higher bond strengths than the tensioned bolts in sandstone and shale roofs. Similarly, the overall support stiffness of non-tensioned roof bolts was significantly greater than that of the tensioned roof bolts. It is thought that, with relatively short bond length of 250 mm, the bonding could easily be damaged when the bolt is tensioned. It is therefore suggested that a new testing procedure should be developed for testing the performance of tensioned bolts.  
The effect of rock type on support performance was also investigated by means of a series of short encapsulated pull tests. The results from these tests highlight the very distinct differences between bolt system performances in different rock types. Sandstone was shown in the tests to produce significantly better results than shale and coal. From these results it can be concluded that rock type is one of the primary factors influencing the support system performance.  
A new support system design methodology has also been developed, on the basis of the roof softening concept. This concept highlighted that to maintain the stability of an underground opening, it is essential to keep the immediate roof-softening zone stable. Roof bolts in this zone force all the bolted layers to sag by the same amount; the layers within the bolting range thus act like a solid beam. Building such a beam is actually the ultimate goal of roof bolting where a beam building effect is the required mechanism.
In other SIMRAC projects, a total of 54 intersection and roadway sites were evaluated from mining depths of 32 m to 170 m, situated in significantly different geotechnical environments. The heights of roof softening at these sites were calculated. The results showed that for a 40 per cent increase in the span, taken across the diagonal of an intersection, relative to the roadway span, the magnitude of the displacement in the roof increased by a factor of four. The results also showed no evidence (in intersections and roadways) of a substantial increase in the height of bed separation. It was also found that the average height of roof softening measured at 54 sites in South African collieries was 1.07 m, which is less than the roof bolt lengths commonly used in South Africa. The new design methodology and above results indicated that on average almost all supported roofs will be stable in South Africa, if the support is properly installed.   
Support system stiffness, which can be calculated from in situ short encapsulated pull tests, has been found to be one of the most important parameters in the design and performance of support system. In order to achieve the maximum performance of support systems, the following support system stiffnesses have been recommended for different sizes of bolts. This stiffness would be determined from in situ short encapsulation pull tests.   
Bolt diameter Required Support Stiffness  for Non-tensioned bolts (kN/mm)
20 mm 60
18 mm 50
16 mm 40
This investigation recommended that an extensive study into the shear strength of full column resin bolts be undertaken.
An investigation into the quality control procedures of support systems was also conducted. Quality control procedures for compliance with the design, support elements and quality of installation are presented in the report. Recommendations for improving quality control measures and for developing testing procedures for bolt system components, installation quality and resin performance are provided. 
PDF icon SIM020205 FINAL REPORT.pdf2.71 MB