This website uses non-intrusive cookies to improve your user experience. You can visit our cookie privacy page for more information.

Safety and health in mines research advisory board

Annual Review 1999



Ground control research continues to be a major part of the portfolio reviewed by SMRAB. The majority of the work was carried out by Rock Mechanics Technology Ltd (RMT) but HSL and the University of Nottingham have also worked on certain aspects. Progress made during 1999 is described under various topic headings below. In some cases, the work programmes described drew on more than one project.

During 1999 four research projects related to mine safety were completed:

The first three were funded mainly by the ECSC with contributions from HSE, mining companies and manufacturers. The fourth was funded by HSE.

RMT was also awarded a SIMRAC (South African Safety in Mines Research Advisory Committee) research project entitled "Assessing and evaluating acoustic techniques to test roof conditions in coal mines" which is completed and the final report is being prepared.

Ongoing ECSC funded ground control research projects by RMT are:

HSL and the University of Nottingham are also partners in the last of these.

Siting and Support of Roadways in the Vicinty of the Old Waste

This project addressed the key recommendations of the public inquiry into the fall of ground at Bilsthorpe Colliery regarding the need for further research to improve understanding of ground behaviour close to wastes. The Project took four years to complete and included an extensive literature review, many measurements of rock behaviour and rock stress in the vicinity of wastes and a major programme of numerical modelling. One important aspect recognised early in the research was that the term "skin to skin" working can be applied to many different mining practices. It was concluded from the literature review that, although skin to skin longwall working in its broadest sense has been, and still is, in use throughout the world, the practice employed at Bilsthorpe of driving a roadway directly alongside an existing waste, which had not been contained by a pack, with no intervening pillar and almost no standing support, was extremely uncommon. The literature review did not find any documentation of major falls in identical circumstances, as such circumstances were very rare, but it did identify many instances of falls where similar mechanisms and geometries were involved, most commonly where longwalls had retreated into predriven roadways.

The underground measurements were undertaken for a number of distinct objectives,

Measurements were made at many UK collieries with particularly intensive programmes at collieries which continued to employ small pillars between retreat longwalls, following the Bilsthorpe fall.

Numerical modelling was initially applied to the sites where the extensive field measurements had been made and this provided further insight into yield pillar behaviour and design. Towards the end of the project, the understanding that had been gained was applied to a re-appraisal of the Bilsthorpe fall, which included reanalysis of powered support monitoring data from the adjacent longwall and numerical modelling using two modelling packages, FLAC and UDEC. Direct simulation of the fall was only possible by significantly reducing the strength properties of the joints defining the potentially unstable block above PG45s Supply Gate roof. The specific modifications to the model necessary to simulate the fall tended to confirm that it only occurred due a combination of many detrimental and unusual factors all combining at a common location and time. Three scenarios were determined for possible failure mechanisms to account for the fall. That considered most likely did not involve significant departure from the interpretation presented in the original HSE investigation report but increased understanding of the likely processes which contributed to the fall.

The main recommendations of the Project were as follows:

Ground Control at Small Mines

The aim of this HSE funded project was to study small mine working methods and support practices, to apply improved knowledge of rock mechanics to the small mine situation and to compile draft guidance on ground control for small mines based on these findings. Provision of this advice would assist small mines to improve ground control safety and comply with the new Ground Control Regulations which came into force on 1st December 1999.

The project included studies of a mine with fully mechanised shortwall production, four mines using partially mechanised shortwall and bord and pillar production and four mines using unmechanised bord or stall working. The research confirmed that mining and support conditions in small mines differ in important ways from those commonly found in larger deep mines. These differences tend to increase ground control risks and help to explain the disproportionate accident rates experienced. These differences were seen as being related to the prevailing relatively shallow, low stress conditions in the mines in three ways. Rock joints and other discontinuities may be weathered and/or open allowing movement of blocks to occur. Horizontal confining stresses are reduced, giving greater potential for movement or failure on pre-existing weaknesses such as joints, bedding disturbance and faulting. Water ingress and/or weathering may affect the competence of weaker roof rocks.

Simple risk assessment procedures in the form of checklists were developed for the range of ground control failure mechanisms identified at the mines visited. The assessment consists of a short list of risk factors for each failure type together with suggested actions to reduce risk. The intention is that individual small mine operators should tailor the checklists to their specific conditions and situation.

Computer modelling techniques were used to compare methods of stall working, advancing using stone packs and retreating leaving pillars. This indicated that both methods are suited to high rock strength and/or low stress conditions. The potential for roof failure with these methods increases with working depth. However the modelling indicated that the retreating stall method was likely to be applicable at greater depth than the advancing stall method.

Pillar sizes for room and pillar working were investigated at three sites and a recommended method of pillar sizing for small mines established.

The application of a prototype device, known as the acoustic energy meter (AEM), for detecting unsound roof was investigated. It was found to be very effective in identifying poor immediate roof conditions caused by slabbing or other discontinuities. The maximum slab thickness positively identified as giving high readings was 0.25 to 0.3m. It was not ascertained if the device could be effective in identifying thicker slabs or loose blocks. These results were very promising for the applicability of the acoustic energy meter for small mines (see separate section on AEM).

In general it was concluded that the main role of typical small mine support systems is to provide protection from falls of immediate roof. The only protection against large roof falls which would easily overwhelm these support systems is observation combined with understanding of potential failure mechanisms. This underlines the need for a risk assessment based approach. The research indicated that the importance of observation also implied that support systems which conceal the condition of the surrounding rock (such as use of corrugated sheeting) should be discouraged. Operators should be encouraged to consider the use of higher capacity support such as linked steel beams and legs or hydraulic props in higher risk situations.

As an Appendix to the main project report, a draft handbook was produced which provides practical advice and tools to assist small mine operators in reducing accidents from falls of ground. The handbook includes risk assessment and pillar design tables. It has been published as HSE Contract Research Report 264/2000.

Remote Reading Telltale System

RMT's new remote reading telltale system spent 1999 going through the UK and European Intrinsic Safety approval process. This is expected early in 2000 and an underground demonstration of the system is planned to follow at an RJB Mining (UK) Ltd colliery. US MSHA approval is also imminent.

The system will allow up to 400 dual height telltales to be connected to a surface PC in a low cost mine -wide monitoring system. The surface PC will display current readings, generate action level alarms on A and B displacements and rates of movement and archive data for offline plotting using Telltale for Windows. The transducers have been designed to be low cost and rugged and will be visually almost identical to current dual height water diverting telltales, retaining their features for underground recognition of roof movement above and below the bolted height. The system allows up to 100 telltales to be strung together along each of four separate roadways in a simple "daisy chain" configuration using a twin core cable. As each roadway advances, new telltales can be added, with the trailing cable crimped to the leading cable from the last telltale. At the other end of each chain of telltales the cable is connected to an underground interrogation and communications unit. This sends the data to the surface PC, using a standard telephone cable and provides local interrogation and diagnostic facilities.

The delay in approvals has allowed further development of the user interface software which now includes automatic recognition of new telltales added to the system. During the year a technical paper was published on the use of early telltale data to predict long term roof displacement trends. Such analysis will be made much more practical and useful when used in conjunction with a remote reading telltale system.

Strain  Gauged Flexible Bolts (SGF)

A major focus of work during the year has been to progress the strain gauged flexible bolt from a relatively unreliable prototype to an instrument which can be manufactured and installed with confidence. This required a major redesign of the original prototype involving re-routing of the strain gauge wires, redesign of the end fixing/spinning-in system and incorporation of temperature compensation. The SGF bolt is now delivered to the colliery with an end protection system fitted, which remains in place throughout underground transportation and installation, preventing damage to the delicate cable terminations. The first of the new design bolts has been installed at Middleton limestone mine for 10 months and was still operating well under light loading at the end of the year. A second 4m SGF was installed in moving ground at a colliery in October 1999. By the end of the year, this was indicating considerable loading above the height of the standard 2.4m AT rockbolts, with a maximum mean strain of over 4000 microstrain and a maximum strain difference of over 6000 microstrain, both at 3.6m into the roof. In order to improve understanding of the reinforcing action of flexible bolts, it is necessary to begin to compile data from SGFs installed in different geotechnical settings. This is considered important as flexible bolts are being used in increasing numbers in UK coal mines as an integral part of the roof support system with little current data on their in-situ loading patterns.

Acoustic Energy Meter (AEM)

This instrument was originally developed for locating loose roof in gypsum mines and was used with great success for assessing back grouting behind segmental concrete linings during the construction of Asfordby mine. The apparatus comprises a transducer which is in contact with the surface of the structure to be assessed, a method of striking the surface close to the transducer and a detector connected to the transducer. The detector provides an output which is a function of the reverberation time of the surface and thus of the degree of "hollowness" or "looseness" in the locality of the transducer. It is an electronic equivalent of "sounding" the roof. As stated above the instrument was tested on mine roofs during the Small Mines Project with very promising results. However the instrument has not been submitted for UK Intrinsic Safety Approval and RMT are unlikely to do so unless a much larger UK or European coal mining market than is currently forecast can be envisaged. The instrument can be used underground under special Managers' Rules. It is believed that the instrument has major potential for improving safety in UK civil engineering tunnelling projects and for routine monitoring of tunnel lining competence in existing tunnels.

Also this year, considerable work has been undertaken in evaluating the instrument for use in South African mines. A SIMRAC research project has been completed which has concluded that the instrument would be a valuable safety tool for South African coal mine supervisors. (Telltales are not commonly used in South Africa; roof bolting densities are lower and cut out distances are larger than would be typical in UK mines.) This has led to development of a pocket sized Intrinsically Safe instrument. This has been submitted for local South African IS approval, which is expected by Spring 2000. Instruments have already been purchased by a number of South African coal, platinum and gold mines.

In-situ Integrity Testing of Rock Reinforcement

A Project on the use of ultrasonics to determine the integrity of in-situ rockbolts was completed. Research is continuing on a radio frequency method for determining the integrity of stranded tendons in-situ. The main conclusion of the ultrasonics research was that it is possible, using guided ultrasonic waves, to obtain reflections from the far end of resin encapsulated rockbolts, installed in a coal mine roof, provided the rockbolts are reasonably straight. Reflected signals from the far ends of rockbolts have been obtained from 3m long fully encapsulated bolts with flat reflecting ends, and from 2.4m long fully encapsulated bolts with chamfered reflecting ends. There is some variability in the quality of signal response from bolt to bolt. Factors influencing this have been found to be:

Other features which have also been detectable using guided ultrasonic waves include:

Some of the successes achieved during this project were made possible through the use of laboratory test equipment not compatible with use underground in coal mines. This equipment included a number of features that improved signal quality thereby increasing the scope of testing. In particular variable tuning of the centre frequency of the output signal from the amplifier, toneburst excitation of the piezo-electric transducers, high and low pass frequency filtering, a good signal to noise ratio and signal averaging.

Although these features were not available on the portable prototype instrument used during the project for underground tests, the portable instrument was found to be relatively successful when used in conjunction with purpose made probe transducers and useful results were obtained from surveys carried out underground in coal mines. With this equipment, near field defects (within 1m) and end reflections from straight bolts were detectable. This equipment offers a test capability which will be useful to mine operators wishing to detect such defects.

Future improvements in ultrasonic systems for testing rockbolts in coal mines are expected to lie in the development of a portable instrument which has the facilities which were available on the laboratory test equipment used during the project. A portable version of this equipment is currently being developed for use in the oil industry for pipe testing and a version capable of testing rockbolts may be available later in 2000.

The principle of measuring standing radio wave resonances in two parallel in-situ stranded tendons is being investigated and has been patented with a view to developing a method which can be used to determine the average length of the two tendons and thus their in-situ integrity. Initially, very promising results were obtained in the laboratory and in a test limestone mine. Also trials at Dinorwig Pump storage station in slate have been successful. Less good results were obtained in a coal mine roof, however. Further investigations and trials have indicated that the use of a combination of "balun" transformers and inline resistance could solve the problem in coal measure strata. Further trials are planned for early 2000.

Fall of Ground Risk Assessment

In addition to the work on risk assessment for small mines described above, research on development of a risk assessment methodology to reduce falls of ground accidents has continued with particular emphasis on sites at risk from rib falls. Development of a gate road risk assessment system covering hazards of potential falls from both roof and sides for a specific site with very weak sides was recently completed. This is the most complex risk assessment yet developed by RMT as it covers six potential failure mechanisms, two each for the hazards of roof failure above the bolts, roof failure within the bolted height and rib falls. A full risk assessment survey of the gates will soon be undertaken at the site prior to longwall retreat. This will be the longest retreat of a face with rockbolted gates to date at the mine. A section of one of these gates is supported by a combination of rockbolts and flexible bolts.

The more simple risk assessment methodology, developed to assess increased risk of roofbolt failure or overloading due to excessive roof shear, was enhanced to include increased stress in tailgate drivages caused by retreat of an adjacent panel. This procedure has now been applied to the gate roads of six successive panels at an RJB Mining (UK) colliery with great success. Current practice is for the mine to undertake the underground survey in the gateroads of a new panel prior to commencement of face retreat, using the specially designed booking sheets. The survey results are then supplied to RMT for application of the risk assessment algorithms and a report on the risk assessment returned to the mine. This ensures an independent analysis whilst minimising the cost to the mine.

Laboratory Testing of Support Systems and Reinforcement Consumables

RMT have continued to develop the laboratory short encapsulation pull test as a laboratory test for rock reinforcement systems which can provide a measure of the actual performance capabilities of such systems in realistic conditions. It is envisaged that this test will supersede the double embedment test as the main laboratory test of reinforcement system performance in a revised BS7861. The double embedment test has major limitations in that it measures the best possible performance of a reinforcement system in highly idealised conditions. Recommendations to this effect are being prepared for the Rockbolting Research Liaison Committee, which is currently considering a possible revision of the British Standard. HSL is also contributing to this exercise with work on fracture toughness and other metallurgical properties of rockbolt steels.

RMT's work during the year has involved developing short encapsulation pull tests in various rock materials ranging from sandstone to coal. A database of test results todate is currently being prepared in order to assist with ensuring that sufficient data is available in time for setting objective threshold criteria in a revised British Standard.

During the year laboratory short encapsulation pull tests have been undertaken on double birdcaged cable bolts (in coal and sandstone) and also with double nutcage cables (in sandstone only). In both media the double birdcage cables reached their full rated load without bond failure (500mm embedment length). The double nut cage cables showed a much lower bond stiffness with bond failure at 51 Tonnes.

Two laboratory short encapsulation pull tests in sandstone were performed to examine the effects of a pretension load of 20 Tonnes on a cable bolt. In the first test, as soon as the grout had been pumped into the pre-tensioned assembly, the tendon was subjected to a slow strain rate. In the second test, after curing for 3 days, the assembly was subjected to axial load until failure.

The tests showed that loading after pretensioning results in stretching of the free length until the load equals the pretension. The system's stiffness is effectively infinite until additional loads are generated within the tendon which exceed the pretension load. The stiffness then equals the unpretensioned stiffness, but at a substantially higher loading. Further investigations of the actions and effectiveness of pretensioned systems are required to gain an understanding of their applicability and relevance to UK mining conditions.

An extensive programme of laboratory compressive strength tests was conducted on a number of variations in design of Link-N-Lock cribs using a range of timber from different sources. This is a system which has been designed to replace traditional four and nine point cribs in a range of mining applications. The manufacturers claim that, the design of the Link-N-Lock system, which gives a much greater surface area contact between each layer than with conventional cribs, increases timber efficiency and maximises strength and stability of the unit. The aim of the work was to ascertain base data on the Link-n-Lock system for evaluation against other cribbing systems and to assist the manufacturer in reducing the cost to the industry by using lower priced locally grown timber if possible. The work indicated that reducing the width of the notch from 98mm to 92mm allowed an English Beech crib constructed with 76mm wide members to develop a good loading characteristic, comparable with an English Ash crib constructed from 89mm wide members and with similar overall dimensions.

Numerical modelling of Support Systems

Computer modelling of the behaviour of coal ribs has generally been less satisfactory than can be achieved for the roof, due to the greater complexity of material and reinforcement behaviour. This is particularly important as all recent fall of ground fatalities have been associated with rib falls. Work has therefore been undertaken to improve the ability to model rib behaviour by improving the representation of coal properties. The main modifications have been:

The updated models give a more realistic rib extensometer response. Further underground measurements should help to confirm that the mechanisms being simulated in the models are those occurring underground. It is hoped that this will allow effective comparison of alternative rib support systems. This will complement previous and planned work on testing of rib consumables.

Numerical modelling and stress measurement was undertaken to assess the potential for increasing bolt row spacing for a new gateroad in a new, deeper area of a mine with a thick, weak coal seam. It was concluded that the modelling indicated a bolt row spacing of 0.8m (rather than the previously planned 0.65m) was likely to be suitable for drivage but additional support may be required in the gate during retreat of the panel. However as the spacing of rows of bolts increases, the reduced confinement provided between the bolts may allow increased deformation of the roof coal in these zones. This may be the effective limit on increasing row spacing rather than large scale instability of the gate. There are many approximations involved with examining the effects of changing bolt row spacing using a two dimensional model. This is an area that requires the development of suitable three dimensional modelling techniques and expertise which will be addressed under the new ECSC modelling project.

Work has commenced in collaboration with Exeter University on examination of the applicability of three dimensional computer modelling to reinforced gate roadways. Work to date has comprised a review of typical rock mass properties, the development of a function to produce an elastic brittle plastic failure criterion and initial investigation of a stepped excavation sequence. Use of elastic and elastic-plastic modelling has correctly simulated the observed detrimental effects of horizontal stress direction on roadway stability. Further interrogation of the elastic analysis has provided useful information regarding stress distribution effects around a face end. Analysis of principal stress direction cosine data has been used to assess the implications of fracture orientation in the immediate vicinity of the face-end. This analysis confirms observed fractures by researchers in Australia.

Numerical Modelling for Mine Design

A major aspect of the new ECSC numerical modelling project is development of modelling capabilities for mine design. This includes modelling of multiple seam interaction and caving mechanisms. Several potential study sites have been identified where the application of numerical modelling would aid mine design and which are expected to provide suitable example problems. Work has commenced at two sites.

Neural Networks Analysis of Rockbolt Data

HSL is also participating in the ECSC research project 'Targeted Rockbolting II' with this study. It is anticipated that by analysing information collected from rockbolted roadways (such as extensiometer, strain gauge data and associated geophysical/geomechanical data) a neural computing technique can be found that will complement current methodology to give a better estimate of the stability of the supported roof. The principal aim of the research is to evaluate the effectiveness of using neural computing techniques in the analysis of rock bolt data.

Work so far has established that the data obtained from straingauges, roadway roof telltales and geomechanical data may be sufficient to enable meaningful predictions of roadway stability to be made. Current work is concerned with the development of computer programmes to process raw data from these sources in order to generate neural networks.

Surface subsidence studies were identified as promising sources of geomechanical data. In order to obtain indications of the importance of geomechanical factors in determining roadway stability, a neural network was generated to predict surface subsidence using roadway geometrical data and geomechanical data. Good correlation between neural network prediction and observed subsidence was obtained that did not require the use of geomechanical data. The implications of this finding for roadway stability predictions are being considered.

Forward to

Updated 2009-05-22