21st June 2019
Is the 10 times seam thickness rule all you need to know about risks from shallow coal?
The Coal Authority (TCA) has issued new guidance TGN01/2019 (1) on the potential effects of the collapse of deep pillar and stall working, based on events which occurred at a new housing development the Bayfield Estate in West Allotment North Tyneside (as reviewed by Tom Backhouse of Terrafirma in RICS Land Journal in December 2017) (2). In March 2017 a broad subsidence event related to coal mining beneath the site occurred resulting in, it is estimated by the Coal Authority, a total of 320-600mm of movement in the absence of any crownholes.
A Coal Authority investigation has concluded that the site was underlain by a coal seam between 2.7m and 4m thick which has been worked as “pillar and stall” (i.e. leaving intact coal pillars in place to support the mine roof). Significantly, they identified an unusually high level of voiding, broken or collapsed ground in the area of subsidence and an extraction rate of around 75%. As is often the case in Coal Measures deposits, the Coal Authority investigations suggest there was no competent roof rock to the mine (rock above the seam being described as “weak, laminated, weathered and fractured”).
The Coal Authority concludes that due to the amount of voiding the form of failure was a ‘generalised subsidence’ event, caused by stress redistribution – a ‘domino effect’ as individual pillars collapse and cause overstressing of adjacent pillars. This form of collapse is well-understood from pillar and stall mines in more competent materiel (e.g. limestone) but is unusual in pillar and stall coal mines where the effect of roof collapses and packing with mine waste normally limits the areas of open voids. Because generalised collapse, by definition, affects large areas of the mine it can generate subsidence effects at surface even where mines are at significant depth. Much as the effects from modern deep coal mines, these kinds of events can generate a broad subsidence trough but (as in this example) they may not result in crownholes.
In this case the Coal Authority has indicated that rock cover over seams ranged from 12.3 times and 9.7 times the seam thickness and have pointed out therefore that in some cases the usual “10 times seam thickness” threshold for generation of ground movement from shallow coal mines had not been exceeded. However firstly it must be remembered that this rule relates specifically to the development of crownholes (which it appears did not occur at this site). Secondly it should be noted that the expected approach in any mining risk assessment would have been to assume the maximum recorded thickness of any seams. Hence, where a seam up to 4m thick was known, a mining investigation to at least 44m depth would have been considered appropriate.
Perhaps more significant is whether or not the potential for shallow seams had been identified at the site prior to construction, which might then have informed the necessary investigations to fully assess the development risk. The Coal Authority rightly highlights the need for mining risk assessments to consider the potential for unrecorded workings. Terrafirma have previously concluded that an integrated review of available data would have identified the likely presence of unrecorded shallow workings at the site, including the possibility of adits from an adjacent quarry and a bell pit (see Figure 1). Hence a modern risk-based approach to mining data could have flagged-up the need for further investigations at this site.
Generalised subsidence is rare in pillar and stall coal workings, as evidenced by the unusual condition within the workings in Tyneside where large open voids remained despite the poor condition of the roof rock. In most cases a 10 times seam thickness rule will provide a reasonable assessment for the depth of coal workings investigations investigation provided this is used conservatively based on a suitable assessment of available information. As with all mines, where large open voids are located, the potential for generalised subsidence should always be considered, but the most significant risk mitigation remains an accurate conceptual model of mining hazards.