Assessing the risks of cement casing failure at oil and gas wells in Pennsylvania.

Like other areas of the US, Pennsylvania has seen an increase in oil and gas production in recent years, driven by the expanded targeting of shale gas deposits using hydraulic fracturing (blasting water and chemicals into shale beds at high pressure to break up the shale and release any trapped gas) and directional drilling (drilling in any other direction than straight down; drill bores into shale gas beds often penetrate horizontally along the deposits for some distance). As in other areas this expansion has met with some concern, with residents of Pennsylvania particularly worried about the leakage of methane gas from wells into water supplies and the atmosphere. Methane is often cited as a more environmentally friendly fuel than coal, since burning it produces less of the greenhouse gas carbon dioxide, however methane is itself a powerful greenhouse gas, and if it leaks from wells into the atmosphere at even low levels then any environmental benefits of burning gas over coal are rapidly lost.

A drill rig at Roullette in Potter County, Pennsylvania, targeting the Marcellus Shale. Laurie Barr/Wikimedia Commons.

Raised methane levels have been found both in aquifers from which drinking water is drawn and in the atmosphere close to gas wells in Pennsylvania, strongly suggesting that there is a link between the methane levels and the drilling. One of the most likely ways in which this can happen is loss of integrity along the well bore due to failure of the cement casing, allowing methane to migrate along the borehole, into other strata or the atmosphere. Current regulations in Pennsylvania allow for low pressure leaks to be monitored and periodically bled off, but require that higher pressure leaks must be repaired, or if this is not possible the well must be permanently plugged; however such plugging, while likely to prevent further loss of gas from the wellhead into the atmosphere, may still allow subsurface migration of methane into other strata (including aquifers).

Clearly this makes the rate of failure in drill well cement casings in Pennsylvania a matter of great interest, however most previous studies of such well casings have concentrated on the structural integrity of offshore wells, with little direct relevance to the wells of Pennsylvania, other than the general observation that wells drilled during periods of rapid expansion in the industry are more likely to have problems, as are wells which are drilled in directions other than vertical. 

Data on the monitoring of structural integrity at wells in Pennsylvania is not publically available, however records Notices of Violations issued by the Pennsylvania Department of Environmental Protection are, and several studies on well integrity in the state have used these notices to try to estimate the rate of drill casing integrity failure there.

In a paper published in the Proceedings of the National Academy of Sciences of the USA on 30 June 2014, Anthony Ingraffea of the School of Civil and Environmental Engineering at Cornell University and Physicians, Scientists, and Engineers for Healthy Energy, Martin Wells of the Department of Statistical Sciences at Cornell University, Renee Santoro, also of Physicians, Scientists, and Engineers for Healthy Energy and Seth Shonkoff, again of Physicians, Scientists, and Engineers for Healthy Energy as well as the Department of Environmental Science, Policy, and Management at the University of California, Berkeley, describe the results of a study of the complete inspection records issued by the Pennsylvania Department of Environmental Protection for wells spudded (the spud date of a well is the day on which drilling begins) between 2000 and 2012.

Ingraffea et al. observe that Notices of Violation do not give a complete record of all leaks at well heads, as the inspection agency may not issue such a notice at a well where cement casing failure has been detected but remedial action is being taken by the drill operator, and that accessing the full inspection records for the wells should help to detect problems not identified by studies of Notices of Violation alone. They also not that problems with older wells are more likely to have been detected than problems with newer wells, as the older wells will have been inspected more times since their spud date, giving a better understanding of the full lifetime history of the well, and that many unconventional wells in Pennsylvania have been dug during a fairly recent boom, leading to a combination of two known high risk factors (wells dug during periods of rapid expansion within the industry and wells which deviate from the vertical) with shorter inspection records where problems are unlikely to have been detected.

Ingraffea et al. found that a total of 1.9% of all wells spudded in Pennsylvania between 2000 and 2012 suffered a loss of integrity. They further found that unconventional wells (i.e. wells targeting shale gas) were roughly six times as likely to suffer problems as conventional wells. The highest incidence of failure occurred in the northeast region of the state (Bradford, Cameron, Clinton, Lycoming, Potter, Sullivan, Susquehanna, Tioga, Wayne, and Wyoming counties) in wells targeting the Marcellus Shale spudded before 2009, with a failure rate of 9.84%. They further noted that wells drilled in the same area since 2009 had a failure rate of 9.18%, despite having a shorter inspection record (although wells spudded in 2012, which had an inspection record of less than 12 months when the study was carried out did have a significantly lower failure rate).

Map showing the extent and thickness of the Marcellus Shale beneath New York, Pennsylvania, Ohio and Virginia. Note the shale is thickest in the northeast. Marcellus Connection. 

The oldest unconventional well in the study was spudded in 2002, and unconventional wells remained a relatively small part of the industry in Pennsylvania until 2009. The rate of inspection also increased in 2009, with 76% of wells inspected in their first year prior to 2009, and 88.7% of wells inspected in their first year subsequently, which may partly account for the high rate of detected failure in post-2009 wells.

Wells in the northeastern counties were more prone to problems than wells in other areas, with 266 well failures recorded in the region, 52% of the total for the state. Within the northeastern region unconventional wells were 8.5 times more likely to have structural integrity problems than conventional wells. Unconventional wells in the northeastern counties had a 20% chance of problems within the first 3-4 years of operation, rising to 40% by year 7.

See also…

 Magnitude 3.4 Earthquake in southern Logan County, Oklahoma.

The United States Geological Survey recorded a Magnitude 3.4 Earthquake at a depth of 3.5 km in...

 Environment America reports on fracking in the USA.

Fracking, or Hydraulic Fracturing, is a process by which water, sand and chemicals are forced into buried sediments in order to shock them into releasing trapped...

 Secret fracking off the coast of California.

It emerged this week that hydraulic fracturing (fracking) has been in regular use in the oil fields of the Santa Barbara Channel off the coast of California since the late 1990s. There has been a moratorium on new drill...

Follow Sciency Thoughts on Facebook.

Drilling rig burning in southern Texas.

A drilling rig is on fire in Lavaca County in southern Texas following an explosion in the evening of Wednesday 28 August 2013. The rig, part of the Eagle Ford Shale development, owned by Nabors Industries and operated by EOG Resources of Houston, was involved in horizontal drilling for shale gas when it suffered a blowout, and consequently caught fire. Incidents of this kind typically occur when drilling operations unexpectedly encounter pockets of pressurized hydrocarbons, particularly gas. All workers at the site have reportedly been evacuated safely.

Fire burning at the drilling rig in Lavaca County on Wednesday 28 August 2013. Eagle Ford Shale.

Local fire crews have reportedly not been able to control the blaze, and have sort help from Wild Well Control, a company based in Houston who specialize in dealing with this sort of incident. It is not clear if the rig can be saved, and it is likely to burn for several days.

See also Gas rig on fire in Caspian Sea, Burning gas rig off the coast of Louisiana, North Sea oil rig partially evacuated following leak, Oil rig runs aground on Sitkalidak Island on New Years Day and Two feared dead after explosion on oil rig in Gulf of Mexico.

Follow Sciency Thoughts on Facebook.

The British Geological Survey reports on Shale Gasses in the Bowland-Hodder Unit.

Shale gasses have become an economically important source of hydrocarbons in the early twenty-first century, as other reserves have run low and exploration companies have begun to look for new supplies of energy. In conventional oil and gas reserves the hydrocarbons are usually trapped within pore spaces in discrete pocket, usually of sandstone or porous limestones, which can be drilled into to release the hydrocarbons. In shales the gasses are found throughout larger, essentially impermeable, units, and need to be released proactively. This can be done through a process known as Hydraulic Fracturing, or Fracking, in which water, sand and chemicals are blasted into the rock at high pressure, causing it to shatter and release the gas. This has proved to be highly lucrative in some parts of the world, notably the US where it has fueled a small hydrocarbons boom, but has also caused great alarm among environmentalists, who linked the process to small Earthquakes and pollution of important aquifers, as well as raising concerns that making large amounts of new hydrocarbons available will lead to further rises in atmospheric CO₂, with consequences for the global climate.

As in other countries, there is growing interest in developing shale gasses as a source of fuel in the UK, though the extent of potentially exploitable reserves is largely a matter of speculation. One of the geological units that has been of particular interest is the Bowland-Hodder Carboniferous Shales, which underlie much of northern England.  With this in mind the Department of Enregy and Climate Change and British Geological Survey have produced a report into possible gas reserves within the Bowland Shales, published on 27 June 2013. This report does not make any judgement on the ethics of shale gas exploitation, or wether it would be possible to extract these gasses at all, but simply to assess the extent of the reserves in order to facilitate further discussion. In the absence of such published information governments have sometimes naively (or possibly dishonestly) underestimated the value of mineral reserves within their territories and granted concessions to companies at far less than their actual worth.

Location of the DECC/BGS study area in central Britain, together with prospective areas for shale gas, currently licensed acreage and selected urban areas. Other shale gas and shale oil plays may exist. DECC/BGS (2013).

Natural gasses are formed from the 'cooking' of biological material (principally plant matter) in deeply buried sediments. When  biological material is sufficiently heated and pressurized, then volatile compounds are forced off as gas. In porous sediments this will typically escape from its point of origin, rising up until it becomes trapped at a high point in the porous rock, with impermeable sediments, such as clays, above. However when such material is trapped within clays or shales (finely laminated clays) it is unable to escape, and remains in situ. In order for gasses to develop in such sediments they need to be deeply buried, but they may subsequently be uplifted and found closer to the surface.

The Bowland Shales and underlying Hodder Mudstones were laid down in shallow marine basins in the early Carboniferous (347-318 million years ago), a time when Britain was close to the equator and sea-levels were rising and falling with the spread and shrinkage of ice caps. These deposits are not shale and/or mudstone throughout, rather there are layers of limestone within, formed as reefs which formed close to the shore spread back and forth with the rising and lowering level of the sea. These deposits have been deeply buried since the Carboniferous (which ended around 299 million years ago), and for the most part are still at some depth, though in places they have been uplifted and outcrop at the surface.

Shales from the Hodder Mudstone Formation outcropping on the flank of Ashnott High, Bowland Basin, Lancashire. Nicholas Riley in DECC/BGS (2013).

In order to assess the amount of gas trapped in these deposits, it was necessary to asses the amount of organic material within the sediment (estimated as typically 1-3%, but rising to 8% in places), as well as the thickness of the deposits. As these deposits are for the most part very deeply buried (the top of the unit is 4753 m bellow sea level at its deepest), they were only directly accessed by boreholes in a limited number of places, with much of the study being based upon geophysical data, principally from seismic studies and gravitational data. 

The depth of the top of the Bowland-Hodder Unit below sea-level. Note that in places this is a negative figure; the deposits are above sea-level but still underground. DECC/BGS (2013).

Previous estimates of the thickness of the Bowland-Hodder Unit have ranged from 2500 m to 4000 m. The DECC/BGS study estimates it to be around 3575 m at its thickness, thinning down to zero in places.  

The thickness of the Bowland Hodder Unit; it was not possible to measure map the unit in Derbyshire. DECC/BGS (2013).

However the volume of the unit cannot be correlated absolutely to the amount of gas it will contain. In order for organic material to have been converted to gas it will have to have been buried deeply enough for the process to have occurred, considered to be sediments that have been buried to a depth of 2900 m for the purposes of this study (even if they have subsequently been uplifted or exposed). This is referred to as the 'gas window', and can be mapped as a discrete horizon in itself.

The depth bellow modern ground level of the 'gas window' in the Bowland-Hodder Unit. Below this level the unit can be expected to contain shale gas, above it this is not likely to be the case. DECC/BGS (2013).

Thus the maps of the depth and thickness of the Bowland-Hodder unit do not give a reasonable estimate of the distribution of gas-bearing sediments in the study area. Furthermore sediments at depths of less than 5000 ft (1500 m) are not considered to be workable by hydraulic fractionation; the pressures in the rock are too low for the method to work.

Thickness and distribution of shales of the lower Bowland-Hodder unit that are within the gas window and at a depth greater than 5000 ft (1500 m). DECC/BGS (2013).

Thickness and distribution of shales of the upper Bowland-Hodder unit that are within the gas window and at a depth greater than 5000 ft (1500 m). DECC/BGS (2013).

The study produced three estimates of the total volume of gas in the rock (not the same as the total recoverable gas), referred to as the low, central and high estimates. The low estimate predicts a total volume of 4.6 trillion m³ of gas in the lower unit and 18.6 trillion m³ of gas in the upper unit, for a total of 23.3 trillion m³ of gas. The central estimate predicts a total volume of 7.5 trillion m³ of gas in the lower unit and 30.2 trillion m³ of gas in the upper unit, for a total of 37.6 trillion m³ of gas. The high estimate predicts a total volume of 12.7 trillion m³ of gas in the lower unit and 51.9 trillion m³ of gas in the upper unit, for a total of 64.6 trillion m³ of gas. 

See also Geological Society of London to host a public meeting on Shale Gas extraction, Report recommends Fracking should be allowed to continue in the UK, Protests against hydraulic fractionation across Bulgaria, Oklahoma shaken by Earthquake on Christmas Day and Earthquake in Texas attributed to fracking. 17 December 2011.

Follow Sciency Thoughts on Facebook.

Geological Society of London to host a public meeting on Shale Gas extraction.

Shale Gas is naturally occurring gas trapped within shale (sine grained sedimentary rocks, typically mostly clay) formations. This is harder, and more expensive, to extract than other forms of Natural Gas, but is becoming increasingly attractive to hydrocarbons companies as other sources of gas start to dwindle. Typically Shale Gas is extracted using a technique called Hydraulic Fracturing, or Fracking, which involves blasting water, sand and chemicals into shale beds at high pressure in order to fracture the rocks and release the Gas.

Diagram of a Hydraulic Fractioning operation. geology.com

Shale Gas extraction has expanded rapidly in the US in recent years, and is expected to produce half of all the Natural Gas extracted in the US by the year 2020. It has, however, proved to be extremely unpopular with environmental groups, both within the US and in other countries, to the extent that some countries have altogether banned the practice.

There are four principle objections to Shale Gas extraction. One of these is that Natural Gas is a hydrocarbon, and potentially contributes to Global Warming; this is no different to the objections to the extraction of Natural Gas from other sources, excepting that extracting the gas from shale significantly increases the available reserves. The remaining objections are with the Fracking process, and are therefore specific to Shale Gas extraction.

Firstly the process causes Earthquakes. This is not in dispute, though the scale of the quakes the process can cause is hotly disputed between environmentalists and the industry. An Earthquake is shaking in the ground, regardless of the source; a large truck driving past your house does not merely feel like its causing an Earthquake, it actually is. Blasting water, sand and chemicals into buried sediments with the intention of fracturing the rock will certainly cause Earthquakes (if it did not it would not work). 

Industry experts do not expect the process to produce quakes larger than a magnitude of 1 on the Richter Scale, but areas where Fracking occurs in the US have seen an unexpected increase in quake activity, with some quakes exceeding magnitude 3. Since the Richter Scale is logarithmic this represents quakes more than a hundred times as large as predicted, leading the industry to claim that any connection is impossible, but not able to provide an alternative explanation (in some cases this is further confused by the employment of lobbyists who do not understand the process and who will offer blanket denials for even the most minor of quakes). In the UK the process has been linked to two small quakes near an experimental Fracking operation at Preese Hall in Lancashire, leading to a halt in operations.

Secondly the process has been linked to the contamination of aquifers; the chemicals used in the process are potentially toxic, and people do not like the idea of these getting into drinking water. Again industry models do not predict that the chemicals could escape the targeted deposits into other strata, but the chemicals have been found in the aquifers. A report into the industry in the US was unable to confidently say that the process had caused the contamination, but only because the chemical containment at the surface was so poor that contamination from ground-level sources could not be ruled out.

Thirdly the process uses large amounts of water, a matter of some concern in more arid parts of the US, where the industry is suspected of the using water that could be used for other purposes, notably agriculture.

In the UK a report commissioned by the Department of Energy and Climate Change was published last month (April 2012), recommending that the process should be allowed in the UK, subject to very tight environmental regulation, bringing the process back into the public eye.

On 18 June 2012 the Geological Society of London is hosting a public meeting to discuss Shale Gas Extraction, at Burlington House in London. The meeting will not seek to address whether the process should be used in the UK, but will seek to explain the geological science behind the process, and whether it can be undertaken safely. The meeting is not aimed at geologists, but rather at elected representatives (politicians), local and central government officials, regulators, NGOs, representatives of other industries likely to be affected (such as water companies) and other interested parties.

The location of Burlington House.

The meeting will be addressed by Mike Stephenson of the British Geological Survey, who will discuss the nature and distribution of shale gas reserves in the UK, Richard Davies, of Durham University, who will discuss the Hydraulic Fractionation process, Peter Styles of Keele University (one of the authors of the Department of Energy and Climate Change report), who will discuss the safety of the process with regard to induced seismicity (i.e. causing Earthquakes). The meeting will also discuss the potential effects on groundwater, and the uses of water in the industry, as well as the regulatory framework for the industry in the UK, though the speakers on these subjects are yet to be confirmed.

The layout of Burlington House.

See also Report recommends Fracking should be allowed to continue in the UK, Protests against hydraulic fractionation across Bulgaria, Oklahoma shaken by Earthquake on Christmas Day, Earthquake in Texas attributed to fracking and Fracking linked to earthquakes in Lancashire, northeast England.

Follow Sciency Thoughts on Facebook.