D.W.H.Su*,P.Zhang,H.Dougherty,M.Van Dyke,R.Kimutis

Center for Disease Control and Prevention (CDC),National Institute for Occupational Safety and Health (NIOSH),Pittsburgh Mining Research Division (PMRD),Pittsburgh,PA 15236,USA

Keywords:Longwall Mining Coal Gas Miner Safety Health

ABSTRACT This paper presents the results of a unique study conducted by the National Institute for Occupational Safety and Health (NIOSH) from 2016 to 2019 to evaluate the effects of longwall-induced subsurface deformations on shale gas well casing integrity and underground miner safety and health.At both deep-cover and shallow-cover instrumentation sites,surface subsidence measurements,subsurface inplace inclinometer measurements,and underground pillar pressure measurements were conducted as longwall panels were mined.Comparisons of the deep-cover and shallow-cover test site results with those from a similar study under medium cover reveal an interesting longwall-induced response scenario.Under shallow and medium covers,measured horizontal displacements within the abutment pillar are one order of magnitude higher than those measured under deep cover.On the other hand,measured vertical compressions under deep cover are one order of magnitude higher than those under shallow and medium covers.However,FLAC3D simulations of the casings indicate that,in all three cases,the P-110 production casings remain intact under longwall-induced deformations and compressions,which has serious implications for future mine design in areas where shale gas wells have been drilled ahead of mining.

1.Introduction

Due to a recent shale gas boom,approximately 1500 unconventional shale gas wells have been drilled through active and future coal reserves in Pennsylvania,West Virginia,and Ohio over the past 17 years.These shale gas wells have penetrated both current and future Pittsburgh seam coal reserves.These unconventional gas wells,whether tapped into the Marcellus or Utica formations,contain very high gas pressure.Strata deformations associated with underground longwall coal mining could induce high stresses and deformations in the shale gas well casings,which in certain situations could compromise the mechanical integrity of the production,intermediate,and coal protection casings.Damaged well casings could potentially introduce high-pressure,high-volume explosive gas into underground mine workings,particularly during an extended well shut-in period,which could jeopardize underground miners’ safety and health.

Fig.1.Surface,subsurface,and underground instrumentation layout at the medium cover study site.

To provide critical scientific data to the stakeholders,which include the Mine Safety and Health Administration (MSHA),the Pennsylvania Department of Environmental Protection (PADEP),the West Virginia Department of Natural Resources (WVDNR),the Ohio Department of Natural Resources (OHDNR),coal operators,and gas operators,NIOSH initiated a research project in 2016 to evaluate the effects of longwall-induced deformations on shale gas well casing stability under deep as well as shallow covers.The effects of longwall-induced subsurface deformations on shale gas well casing stability under deep cover were presented by NIOSH researchers in 2018 [1,2],and the results from a 2018 NIOSH research study focusing on the longwall-induced subsurface deformations and permeability changes under shallow cover,specifically near a major stream valley,were presented in 2019[3,4].This paper presents the comparisons among the deep cover,medium cover,and shallow cover studies.

2.Site description and geotechnical instrumentation

Geotechnical instrumentations were conducted at three test sites:deep cover,medium cover,and shallow cover test sites.The deep cover test site [1,2] was located over a southwestern Pennsylvania coal mine employing 457-m wide longwall panels to extract coal from the Pittsburgh seam,which lies 361 m below the surface.A three-entry longwall gate-road system of 21 m×45 m centers was employed at the mine.The medium cover test site [5] was also located over a southwestern Pennsylvania coal mine employing 457-m wide longwall panels to extract coal from the Pittsburgh seam,which lies 184 m below the surface.This study was carried out by the coal industry,gas industry,the Marcellus Shale Coalition,and the Pennsylvania Coal Association in 2013 and 2014.A three-entry longwall gate-road system of 18 m×38 m centers was employed at the mine.Fig.1 illustrates the layout of the longwall panels and the locations of the surface and subsurface instrumentation at the medium cover test site,which includes three subsurface inclinometer monitoring holes and one extensometer monitoring hole drilled over the center of the 38 m×83 m centers abutment pillar and eight borehole pressure cells installed into the 38 m×83 m centers abutment pillar.Four test wells were also installed.The test wells only had three casings,namely,surface,water,and coal protection casings.Pre-mining,post-first-panel mining,and post-second-panel mining 60-arm caliper logs were conducted in the four test wells.In addition to the subsurface and underground measurements,pre-and post-longwall surface subsidence measurements were conducted to corroborate surface responses with subsurface and underground responses.The shallow cover test site [3,4] was also located over a southwestern Pennsylvania coal mine employing 457-m wide longwall panels to extract coal from the Pittsburgh seam,which lies 146 m below the surface.A three-entry longwall gate-road system of 18 m×38 m centers was employed at the mine.

Fig.2.Measured and calculated surface subsidence at the shallow cover site.

Fig.3.Measured and calculated in-place inclinometer casing displacements at the deep,medium,and shallow cover test sites.

3.Site geology

Detailed overburden geology at the deep cover,medium cover,and shallow cover test sites down to the coal seam level were compiled by correlating a test site gamma log and a nearby core hole.The overburden geology at the three test sites contains many strong-to-weak rock interfaces,which have been demonstrated to have major influences on longwall-induced stresses and deformations from ground control research conducted over the past 30 years.In particular,the presence of a weak Uniontown Coal Zone above the A-,B-,C-,D-limestone sequence and the presence of clayey shale in the Sewickley Coal Zone sandwiched between the Benwood limestone and the Sewickley limestone indicate potential planes of weakness where substantial longwall-induced subsurface bedding plane movements may occur.

4.3D finite difference simulations

FLAC3Dfinite difference simulations [6] were conducted and analyzed to evaluate the effect of longwall excavations on the induced stresses and deformations within the gate-road abutment pillar.Detailed overburden geology,compiled at the individual test site,was the primary model input.Specifically,over 120 weak-tostrong rock interfaces were present and simulated in each of the three test site models,which employed over 400,000 zones.A hypoelastic longwall gob model with a maximum deformation of 25% was employed in the FLAC3Dmodels.The primary goal of the simulations was to duplicate measured surface subsidence,measured subsurface in-place inclinometer displacement,and measured underground coal pillar pressure increase.The FLAC3Dfinite difference program was selected since it has been calibrated and verified with field data from the Pittsburgh seam.

5.Results of geotechnical instrumentation and 3D numerical analyses

Fig.2 shows a measured surface subsidence profile over the supercritical panel at the shallow-cover test site.With supercritical panel width,longwall-induced surface subsidence is expected to reach the maximum possible value of 0.667×0.3 (mining height×depth of cover) from the headgate and tailgate panel edges.Fig.2 also shows the FLAC3D-calculated surface subsidence profile and illustrates good agreement between measured and calculated subsidence profiles.Typically,the measured and calculated subsidence profiles are in excellent agreement except near the panel edge,where a slight discrepancy might exist due to topography and under-represented planes of weakness in the overburden.Measured and calculated surface subsidence profiles under deep and medium covers,although not shown here due to page limitation,follow the same supercritical panel rule and exhibit the same excellent agreement.

Fig.4.Measured and computed pillar pressure increases at the deep cover,medium cover,and shallow cover test sites.

Fig.5.3D finite difference model of a gas well.

Fig.6.Longwall-induced von Mises stress in the production casing(grouted to the surface) at the deep cover,medium cover,and shallow cover test sites.

Fig.7.Maximum von Mises stress in the coal protection,intermediate,and protection casings versus distance to the projected longwall recovery line.

Fig.3a-3c illustrate measured in-place inclinometer casing displacements at the deep,medium,and shallow cover test sites.Fig.3a-3c also include computed in-place inclinometer displacements for comparison.Overall,FLAC3Dsimulations provided reasonable agreement with measured inclinometer displacements.At the medium cover test site,longwall-induced lateral displacements in the overburden were measured by a suite of 60-arm caliper logs.At the shallow cover test site,the inclinometer sensors were purposely oriented along the direction of longwall mining to capture the influence of the east-west regional horizontal stress under the shallow stream valley environment.Fig.3c also shows the FLAC3D-calculated inclinometer displacements perpendicular to and along the direction of longwall mining.The calculated maximum lateral displacements perpendicular to and along the direction of longwall mining occur at the 121-m depth horizon and are 0.168 and 0.018 m,respectively.The calculated lateral displacements in the direction of longwall mining are in general agreement with the measured in-place inclinometer displacements.Although not measured in this field study,the FLAC3D-calculated inclinometer displacements perpendicular to the direction of longwall mining are comparable to but slightly higher than those at the medium cover test site(Fig.3b)[5]and are one order of magnitude higher than those measured at the deep cover test site (Fig.3a)[1,2].This is expected since,under deep cover,for the same distance above the mining horizon,the weak-to-strong rock interfaces are subject to higher vertical loads under deep cover,which are expected to hinder lateral displacement.It is important to recognize that,while the FLAC3Dcalculates horizontal displacements along every individual plane of weakness,the in-place inclinometers have a 3-m gage length and thus average horizontal movement over the 3-m span.

Fig.4a-c show the measured pillar pressure increases at the three test sites.Fig.4a-c also show the comparison between measured and computed pillar pressure increases,which indicates excellent agreement between the measured and computed pillar pressure increases.

6.FLAC3D modeling of well casing stability

A few FLAC3Dfinite difference models,Fig.5a and b,were also constructed to evaluate the potential effects of longwall-induced stresses and deformations within the longwall abutment pillars on shale gas well casing integrity.

6.1.Longwall-induced casing stresses

It is well known that overburden depth and its associated geology have a significant influence on longwall-induced stresses and deformations.The calibrated FLAC3Dmodels were employed to evaluate the effect of longwall-induced stresses and deformations and their impact on gas well casings under deep,medium,and shallow cover conditions.Fig.6a-6c show that,at the three test sites and with the presence of major soft-to-hard rock interfaces,the longwall-induced deformations caused the production casings to deform.However,the resulting von Mises stresses in the production casings at the three test sites never exceeded the rated yield strength of the P-110 production casing,which is 758 MPa.

6.2.Gas well setback distance

The calibrated FLAC3Dmodel was also used to evaluate the effect of gas well setback distance on longwall-induced von Mises stress in the coal protection,intermediate,and production casings.Fig.7 shows the effect of setback distance on the induced von Mises stress in the production,intermediate,and coal protection casings of a hypothetical gas well drilled into a barrier pillar at a mining depth of 300 m.Fig.7 clearly indicates that when the projected longwall recovery line is at 30 m from the gas well,the induced von Mises stress in the J-55 coal protection and J-55 intermediate casings are still below their rated yield strength of 379 MPa,and the induced von Mises stress in the P-110 production casing is well below its rated yield strength of 758 MPa.It is important to note that the setback distance is expected to be different under different overburden depth and geologic conditions.

7.Discussion

The field instrumentation as well as the numerical modeling results presented in this paper are consistent with a few previous attempts to characterize unconventional subsurface movements along the bedding planes [1,2,5,7,8].Since about 1500 unconventional shale gas wells have been drilled recently ahead of mining in the Pittsburgh coalfield and because all casings of these wells are fully cemented to the surface,depending on the distance to the edge of future longwall extraction,potential casing deformation may be present.However,in all three case studies presented in this paper,the longwall-induced von Mises stress in the production casing stays below its rated yield strength.

It is important to note that longwall-induced von Mises stress in the casings depends not only on the longwall-induced lateral displacement,but also on the longwall-induced vertical compression or abutment pressure.Therefore,potential casing compromise is more likely to be present under very shallow cover with saturated overburden rocks under which the longwall-induced lateral displacement is large,or under deep cover under which the longwall-induced vertical compression is very high.

8.Conclusions

This paper confirms that under shallow and medium covers,longwall-induced subsurface deformations are one order of magnitude higher than those observed under deep cover.On the other hand,longwall-induced vertical compressions under deep cover are one order of magnitude higher than those under shallow and medium covers.However,in all three case studies presented in this paper,the longwall-induced von Mises stress in the production casing stays below its rated yield strength.These findings are expected to have certain impact on coal operators’ longwall pillar design decision.After all,a gas well’s setback distance,where the longwall-induced von Mises stress in the production,intermediate,and coal protection casings are below their rated yield strengths,is dependent on the overburden depth and geology and can be assessed on a site-specific basis.

Disclaimer

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health,Centers for Disease Control and Prevention.Mention of any company or product does not constitute endorsement by NIOSH.