Researchers evaluate the stability of shale gas wells in longwall barrier pillars
by peter zhang, daniel su, and jun lu
Unconventional shale gas development in longwall mining regions has given rise to safety concerns in longwall mines. With the recent shale gas boom, approximately 1,500 shale gas wells have been drilled through current and future coal reserves in Pennsylvania, West Virginia and Ohio over the past 15 years. Longwall mining removes coal from underground in large blocks and causes the surface and subsurface to move as overburden strata above longwall panels settle to fill the mined void.
When gas wells are located in longwall pillars, the longwall-induced subsurface movement can influence their stability, inducing stresses and deformations in gas well casings in the coal pillars. If gas well casings are damaged or ruptured by excessive stresses and deformations, natural gas could leak into active longwall mines, potentially causing a fire or explosion in underground workings. For these reasons, unconventional shale gas wells in longwall pillars not only present safety concerns in longwall mines, but also cause safety and economic concerns for the gas companies.
To address this issue, the National Institute for Occupational Safety and Health (NIOSH) has been conducting research on gas well stability in longwall pillars to provide technical guidance for state and federal regulatory agencies as well as the coal and gas industry. Researchers have studied the critical factors through field experiments and developed numerical models to evaluate the stability of shale gas wells in longwall barrier pillars, as described in this article.
Review of Current Gas Well Pillar Regulation
The current gas well pillar regulation is the PA 1957 gas well pillar study (commonwealth of Pennsylvania, 1957). This study was completed by the Joint Coal and Gas Committee based on gas well failures caused by coal mining in the state of Pennsylvania prior to 1957. The study included 77 gas well failure cases that occurred over a 25-year span in room-and-pillar mines with full or partial pillar recovery in the Pittsburgh and Freeport coal seams. The mining depth in those mines ranged from 55 feet to 750 ft. The 1957 study provided guidelines for pillar sizes around gas wells under different overburden depths up to 750 ft, which became a gas well pillar regulation in Pennsylvania as well as for other states.
Because the technical guidelines developed in the 1957 study were based on data from room-and-pillar mining under shallow cover, they have been found to be inadequate for longwall gas well pillars, especially under deep cover. In fact, gas well casing failures have occurred in longwall chain pillars even though the chain pillar sizes met the requirements by the 1957 study. Although barrier pillars are usually larger than required by the 1957 study, there is still no guarantee that the gas wells are stable in all circumstances, and other factors have to be taken into consideration when evaluating stability.
Critical Factors Influencing Gas Well Stability in Barrier Pillars
The stability of gas wells in barrier pillars is mainly influenced by overburden depth, gas well location relative to the gob, overburden geology and floor stability. First, overburden depth determines how much abutment pressure could be induced over the barrier pillars. The greater the overburden depth, the larger the induced abutment pressure in the barrier pillars and thus the greater the stresses in the gas well casings. In this respect, the gas wells in barrier pillars under deep cover are potentially subjected to higher induced stresses in the casings near coal seams depending on how far the wells are away from the gob.
Overburden depth also influences where gas well failures could occur. Figure 1 shows locations of gas well failures as a result of retreat mining. Because no cases of gas well failures in barrier pillars could be found, the failure cases from the PA 1957 study as well as two failure cases in longwall chain pillars in the Pittsburgh seam are used to show potential failure locations along the vertical axis of a gas well. Based on these available cases, gas well failure can occur in three locations: in the coal seam, within about 100 ft of the roof strata, and within 40 ft of the immediate floor. The figure also indicates that at greater overburden depth, the failures are more likely to occur either in the coal seam or in the floor.
Although barrier pillars relatively large in size generally have no stability issues, the location of the gas wells in barrier pillars — i.e., the distance of gas wells to the edge of the gob — still has an effect on gas well stability. This effect is shown in Figure 2 using the failure cases from the 1957 study and the cases in the longwall chain pillars in the Pittsburgh seam. The case history demonstrates that the majority of failures occurred when the gas wells were located within about 50 ft horizontally from the gob edge, and few failures occurred up to 80 ft from the gob. This trend suggests that the possibility of failure diminishes greatly with the gas wells that are located farther away from the gob. However, with limited cases from longwall mining, it is still early to come to a conclusion that the gas wells in barrier pillars are safe if they are located beyond the range in which gas well failures have occurred.
Overburden geology, especially weak claystone layers and massive strong sandstone/limestone layers, also influences the stability of gas wells in barrier pillars. Claystone layers are common in the overburden of the Pittsburgh coal seam, and some claystone layers are moisture-sensitive and can become very weak when saturated with water. With longwall mining, large horizontal movements can occur at the claystone layers over barrier pillars due to its low modulus and low friction along the interfaces with other strong rocks. These movements, including vertical compaction and horizontal sliding, could induce significant stresses in gas well casings, potentially causing casing failure. Large horizontal movement up to 5.5 in. has been measured in the overburden strata about 55 ft from the longwall gob under shallow overburden depth of 604 ft in the Pittsburgh seam. This movement, which occurred near a stream valley and along the interfaces of weak claystone and strong sandstone layers, is likely to be associated with low friction and normal pressure along the interfaces under shallow overburden depth. Under deep cover, however, horizontal movement in the overburden over a barrier pillar would be small due to higher friction along the bedding planes. A small horizontal movement of 0.46 in. has been measured about 137 ft from the longwall gob under an overburden depth of 1,185 ft. Therefore, large horizontal displacement over barrier pillars is more likely to occur at weak claystone layers near a stream valley under shallow overburden depth, potentially inducing high shear stress in gas well casings.
Finally, claystone floor is also a concern for mining around gas wells under deep cover. Claystone is commonly present in the floor of the Pittsburgh seam and can become very weak if the floor is wet. A claystone floor, if weakened by water, can induce high vertical and shear stresses in gas well casings under deep cover. Recent experience has shown that gas well failures tend to occur in claystone floor as mining depth becomes greater.
Assessing Gas Well Stability in Barrier Pillars
To assess the stability of gas wells in barrier pillars, we must quantify subsurface movements and their effect on the gas well casings, which is made simpler by way of numerical modeling. NIOSH has developed numerical models that consider geologic and mining factors as well as the construction of the gas well casings. As an example of gas well pillar evaluation, a case is presented to demonstrate the effect of longwall-induced subsurface deformations on the gas well casings in barrier pillars. This case involves two Marcellus shale gas wells located within a barrier pillar between two longwall bleeders in the Pittsburgh coal seam. The gas wells remain intact after longwall mining without any safety issues.
A FLAC3D model — a proven numerical modeling software used for geotechnical analysis — was set up based on the geological and mining conditions near the gas wells. The model included sufficient details to simulate the mining sequence and longwall retreating as well as gas well casings. Shale gas wells are generally completed with five casings: surface, water protection, coal protection, intermediate and production. The coal protection casing is usually placed down to 250 ft below the coal seams. Longwall-induced subsurface movements transfer deformations to the gas well casings through back-filled cement. Because the modulus of steel is high, a small amount of subsurface movement will induce high stresses in the casings. In response to subsurface movements, the casings are likely to experience vertical compression, horizontal compression, and shear. The numerical model is capable of simulating longwall-induced subsurface movements in the overburden and calculating the resulting induced stresses in the gas well casings.
The first longwall panel was mined before the gas wells were drilled and installed. The gas wells were drilled within the center of a 145-ft-wide (rib-to-rib) barrier pillar. The bleeders for the second panel were developed later, and the second panel was mined about 350 ft away from the gas wells. The overburden depth at the gas well site was 850 ft. The average mining height of the two longwall panels was approximately 7 ft.
Figure 3 shows the predicted vertical displacement in the subsurface along the gas wells after both panels are mined. The maximum vertical displacement at the surface is predicted to be 0.75 in. after Panel I mining and 1.75 in. after Panel II mining. Since the gas wells were installed after the retreating of Panel I, mining of Panel II would induce about 1 in. of vertical displacement at the gas well site on the surface. The vertical displacement along the gas wells in the subsurface gradually reduces down to about 0.25 in. at the coal seam level after Panel II mining. Overall, the gas wells are shortened for about 0.75 in. between the surface and the coal seam.
Figure 4 shows the predicted horizontal displacement in the subsurface along the gas wells after both panels are mined. The maximum horizontal displacement at the surface is 1.25 in. after Panel I mining and -0.25 in. after Panel II mining. Importantly, the direction of the longwall-induced horizontal displacement would be toward the gob. Thus, after the first panel mining, the ground moves toward the first panel. However, the ground would move back toward the second panel after the second panel is mined. Since the gas wells were installed after the first panel mining, mining of the second panel would effectively induce about 1.5 in. of horizontal displacement at the gas well site on the surface. The horizontal movement reduces at deeper depth and diminishes to almost zero near the coal seam level.
Other techniques can also be used to assess gas well stability in barrier pillars, such as the von Mises yield criterion, which is commonly used to determine structural safety of engineering materials. For the same case described above, the authors applied this technique and found that high von Mises stress occurs at the weak claystone layers and also increases with depth. Based on this criterion, the casings will yield if the von Mises equivalent stress is greater than the yield strength of the steel.
In summary, the gas wells in barrier pillars are likely to be influenced or even damaged by longwall mining, but the influence is much less than that in longwall chain pillars. However, even if the risk of gas well failure in a barrier pillar is perceived to be low, a thorough assessment should still be performed in that any gas leakage from shale gas wells could pose a serious risk to underground mine workers. In many cases, the assessment is to determine the appropriate precautions that should be put into place during longwall retreating. Therefore, it is important to understand and quantify how the gas wells in longwall barrier pillars could be influenced by longwall mining and to make appropriate decisions on what measures should be taken to ensure safety for both longwall mining and gas production.
The findings and conclusions in this paper 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.
Peter Zhang and Daniel Su are senior service fellows with NIOSH. Jun Lu is a senior geotechnical engineer with CONSOL Energy.