vendredi 23 novembre 2012

The Potential Dangers of the Exploitation of Shale Gas and Oil – Analysis of the Geologic and Geotechnical Aspects

Translation from pages 173-185 of the final report of the Symposium of the Regional Council, Ile-de-France,  February 7, 2012, Paris.

Introductory note: The texts of the symposium were published originally on July, 2012, in France. Below is the translation of an extract made available in Quebec in response to the very highly publicized question of the exploitation of oil in the matrix rock of Anticosti Island.  I wrote this technical opinion in January 2012, at the request of the Scientific Council of the nine universities and research centres of the Ile-de-France, i.e. Paris and its surrounding area, in view of the possible exploitation of shale oil and gas in France. I was also invited to to make a presentation at the Symposium of the 12 of February, 2012 which united the representatives of the universities, the ministers involved, representatives of industry, and four North American experts retained by the Scientific Council. Normand Mousseau of the University of Montreal was the other expert from Canada; his well documented text can be read in the Scientific Council's Report pp.43-78.

The Potential Dangers of the Exploitation of Shale Gas and Oil – Analysis of the Geologic and Geotechnical Aspects

Technical opinion of Marc Durand, doct-ing in engineering geology, ENSG

Summary: 
The exploitation of unconventional hydrocarbon deposits presents grave problems of a technical nature. The irreversible modification of the permeability of the entire volume of a geological formation - tens of thousands of square kilometres, hundreds of meters in depth, employing tens of thousands [1] of horizontal drillings, using on the large scale, a new technology with unknown consequences,  is unprecedented in the mining industry. The geological process of methane migration to new fractures will continue over a geological time scale of thousands of years at least. The fraction that migrates during the very short period of commercial exploitation is only 20% of the volume of gas involved in the process. Capped wells at the end of the operation will not have a a technical lifespan on the geological scale; in this context they should have a carefully reviewed management protocol which absolutely cannot be the same as that previously used for conventional natural gas wells.

Introduction:
The exploitation of unconventional hydrocarbon deposits, among which are found shale gas and shale oil, poses a serious set of geotechnical problems -  which the expertise and best practices of the industry, or the rules of the game of stakeholders, are insufficient to handle. In this paper, we analyze more specifically geological and geotechnical issues as they relate to the possible exploration and exploitation of shale gas and oil.

The actual data are better suited to our analysis than the elements simply apprehended. As France chose in 2011 to allow time before authorizing the industry to go ahead, our report will use examples from North America for analysis: Haynesville and the Barnett shales of Texas, the Marcellus Shale of Pennsylvania and the Utica Shale of Quebec. Exploration and exploitation are already underway in these gas shales. But even in the USA where the industry really began around 2005, the effects and consequences in the medium and long term are not yet all measured and measurable. 

We will use the term shale as a synonym for clay shale. Whatever the term, these layers of sedimentary origin on the five continents contain, scattered organic material from their origins in marine basins. This organic material, transformed by temperature and pressure over the ages, produces the oil and gas known as thermogenic as opposed to contemporary gas (ex. biogenic swamp methane). The shales are numerous and widespread: it is estimated that there are more than one hundred distinct shale formations on the planet. Not all of these have exploitable hydrocarbons, but a great many countries do possess such shale deposits and will therefore be confronted sooner or later with the question of their possible exploitation.

In this report we make an analysis which has universal application to shales and is not limited to any one geologic formation. We have more concrete data on shale gas than on shale oil, due to the fact that the shale gas industry began in the USA with the need to substitute methane gas in the production of thermoelectric energy in the ageing coal power stations. But the analysis of the problems with shale gas also sheds substantial light on the extraction of shale oil, because in both cases comparable  techniques are used in similar geologic contexts.

The problem:
There are four essential elements which serve as a point of departure in the analysis of the problem of unconventional deposits:
1- The new applied techniques of unconventional exploitation of shales is unable to extract more than 20% (ref.2) of the oil and gas that they contain.
2- The exploitation irreversibly modifies the permeability of the whole volume of the deposit: without artificial fracturing the exploitation of the deposit is impossible.
3- The geologic process initiated by the fracturing will continue over geologic time, that is to say, over a period of time immeasurably longer than the lifetime of the structures built for the operation of the wells. 
4- It is impossible to return the rock matrix to its original state at the end of exploitation.

Why is it impossible to extract more than 20% of the gas present in the exploitation of shale wells and what are the consequences of that fact? Given the still limited knowledge available of the long term impacts of the technique of hydraulic fracturing in the lengths of horizontal drilling, we will analyze here the more obvious differences between this method and that of conventional deposit exploitation.

In conventional operation, the deposits are found in discrete geologic structures:  a formation or geologic structure of great porosity due to intergranular spaces and/or communicating natural fractures, the whole capped by a tight formation which covers the roof of the reservoir, as in fig.1.  Once it is located, by means of an actual geologic exploration, the wells reaching the reservoir are able to extract almost all (greater than 95%, ref.2) of the deposit.

In addition to its own pressure, the gas is pushed upwards by the water;  liquid hydrocarbons may possibly be present with the gas and water. It is important to note that the hydrocarbons have migrated very slowly from the rock matrix, that is a a sedimentary rock which may be shale. 

They have accumulated in the natural reservoir in a process which has taken thousands of years and more likely millions of years.  Why?  Because the permeability of the rock matrix of the shale type have a very low value (10exp-12 to 10exp-14m/s). On the other hand, once in the permeable rock, which constitutes the geologic pocket, they are then contained in a geologic layer whose permeability is several orders higher (greater than 10exp-6m/s).

Figure 1. Diagram of a conventional gas reservoir. The liquid hydrocarbons may be present between the water and gas.

In the strata where the gas has accumulated (the classic deposit shown in blue and pale green in fig.1) the porosity is significant (5 to 25 %) and the permeability is commonly a million times higher than in the rock matrix: in exploiting a natural gas deposit, the gas migrates easily towards the extraction well. This is why at the end of its useful life, the production of the well falls almost to zero. The reservoir is not 100% empty, but nearly. At this stage, the wells are abandoned, the sites restored and the property reverts to the state.

It is extremely dangerous to transpose this image to the case of the wells for shale gas; in this case, the fracturing is artificially created just before the extraction and the equilibrium is very far from being reached at the end of exploitation. In addition, the extensions are not limited to one localized deposit, but to a whole geologic layer which has been radically transformed.

The extraction is done by artificially fracturing the gas-bearing shale itself, the migration of gas takes place over a shorter distance than the long transit which created the classic deposit, but it is not an instantaneous process.  A few millimetres from the edge of a fracture, the gas escapes rather quickly (fig.2), but as the distance increases the more it takes geologic time for the gas to migrate into the newly fractured shale as it did in the migrations to the natural reservoirs. With a permeability of 10exp-12 cm/s for example, even under a gradient (i) high, the time required to travel only a few centimetres is counted in centuries and even millennia  (v = K . i) [2]. This is how it happens in the still intact parts of the shale between the fractures.

Figure 2. Migration mechanism of shale gas in shale surrounding new fractures; metric view of of shale at the end of exploitation (3-5 years?).

These phenomena on the millimetric and metric scale affect production rates of the wells, which therefore have exponential or hyperbolic decay curves as shown in figures 3, 4 and 5 for the Haynesville Shale, Marcellus Shale and Utica. The age of the wells in course of operation is counted in months or years in these shales,  and there remains some uncertainty as to the volume of gas which will ultimately be recuperated, which is denoted by '' EUR'' in figure 3. This figure shows five assumptions of production projections based on data obtained over 12 months. What is certain, however, is that the flow decreases significantly. For example, the well at St-Edouard (fig.5) is not worth more than 10% of its initial value after only 150 days. The threshold of unprofitability is reached in only a few years at this rate. The diagrams of of figures 3 and 6 are in semi-logarithmic mode, while the diagrams of figures 4 and 5 are in normal arithmetic mode. A purely exponential decay relationship is shown in Figure 3 (below) in the blue line.

Figure 3.  Normalized Haynesville Shale production rate decline based on differing hyperbolic exponents (ref.3, Aeberman, 2010).

Figure 4. Production decline curve in the Marcellus shale (ref.4, Johnson,2011).


Figure 5. Production decline curve in the Utica Shale, experimental well in Quebec.


Figure 6 below shows what happens at the end of operation when the flow is no longer commercially viable; the well is closed, sealed and reverts to the public ownership after a few years of production. The pressure is low at this time, but it then starts to rise at a rate which is a function of the rate of liberation of gas following the curve (the broken line):

Figure 6. What happens at the end of gas well exploitation.

As there is no zero flow in the evolution of the yield curve (more precisely no zero flow before a time equal to infinity), the restoration of  pressure is inevitable due to the presence of 80% of gas remaining in the shale at the time of cessation of production. This will be far more significant in horizontal wells with hydraulic fracturing than in other types of wells. There is nothing to stop the process once started.  It will continue over the centuries and millennia. And well caps [note 3] do not have this lifespan.

Figure 7. Degradation of well over time; compilation based on 15,000 conventional wells.

There is not much data specifically on new shale gas wells, but there is for conventional wells (fig.7- figure taken from ref.5).  For the new wells (age 0) it is 5% of the conventional wells which show problems of methane leaks.  In Quebec for 31 shale gas wells drilled since 2008 , the proportion of wells with leaks has been 19 out of the 31, more than 60% of the wells. The difference confirms what several researchers argue: that engineering problems, notably in the cementing of new wells with fracturing and horizontal extensions  of 1000 m or more, will show up more and more frequently and be of much more concern than in the case of conventional wells. Even with conventional wells, this question of the degradation of the wells is acute because, as shown in in the data of 15,000 conventional wells (fig.7), as the age of the well increases, the proportion of wells with problems quickly exceeds 50%.

The causes of the degradations and the leaks in the case of of conventional wells have been well analyzed by several authors including Maurice B. Dusseault (ref.6) and Wojtanowicz et al, 2001 (ref.7). The specific analysis for the new types of wells remains completely absent, but one first observation is already conclusive: the dynamic cycles repeated in the fracturing and the complexity of control of placing the tubing in the curved and horizontals of these wells, as well as the use of new chemical products, weaken the steel and grout and accelerate their ageing.

Figure 8. A 3D view summarizing the geological and geotechnical problems of a well at the end of useful life.

It would be surprising if the shale gas industry has invented in the last five years structures which are capable of surviving intact for millenia. Civil engineers have always wanted to have techniques that would give viaducts and bridges a life span greater than fifty years, without inspection and without maintenance like the new abandoned wells will be (fig. 8). Here, the gas industry, with the same materials, steel and cement, seeks to convince us that it has the recipe by which thousands of capped wells will eternally resist growing pressure. In fact, the hydrocarbon industry has never had to the obligation to plan so long-term. No state, no province in Canada, no country in the world currently has regulations adapted specifically for these new realities. The long history of classic gas deposit exploitation by conventional wells has put in place provisions essentially aimed at the security of the operation for the duration of the wells useful life i.e: the short period of exploitation.

It may seem incongruous to some oil promoters to see the problem posed in terms of lifespans of centuries and millennia.  But it isn't only shale gas that must be thought of in this way. The very long-term storage of radio-active waste is studied world-wide taking account of time frames of this order. In the sector very close to shale gas wells, the consulting firm Halliburton indicates this in it's documentation '' The post closure phase addresses post decommissioning – which has an extremely long time horizon of hundreds, if not thousands, of years...'' (ref.8).

These wells that Halliburton indicates need to be followed for millennia are wells at lower risk than the wells with horizontal extensions and hydraulic fracturing: they are vertical wells connected at the surface to CO2 storage tanks, less of a problem than methane. The gas industry is certainly not ignorant of the long term risks. However, they have never been placed under the obligation to take responsibility because past regulations and those currently in effect always transfer the ownership of the wells to the public domain once the production is finished. No regulation anywhere obliges them [4] and this obligation has never been included in their business plan.

We have not addressed in this text the more immediate and obvious consequences of the exploitation of shale gas in North America; the massive use of water for fracturing, the often secret chemical composition of the ''slickwater'', the chemical cocktail which modifies the water to optimize the fracturing, the occupation of agricultural land by a heavy industry, tanker trucks, drilling towers, flares, compressors, gas pipelines, etc. The disposal of about 40% of the water which returns to the surface as reflux, the radio-activity and the very high salinity of water from the deep formations which also finds new channels to the surface layers or which uses natural geologic faults, such as faults and fractures which are inevitably cut by the long horizontal drilling. All this has been abundantly set in relief and daily headlined by the newspapers in the USA. The cases of the contamination of the water tables by methane are the first problems to show up, because extremely mobile methane is the fastest to find its way to the surface through natural paths enlarged by fracturing. The slower migrations are also beginning to appear: chemical compounds of the fracturing fluids have been found in some wells (ref.9).

We recognize that all these questions are, in the near term, of first importance. If we have talked about them little in the analysis above, it is because other researchers have abundantly discussed them and some of these consequences relate specifically to hydraulic fracturing. Other techniques (liquified propane, CO2, compressed air, electric arc, etc.) may eventually be used and the list of consequences which would appear with that type of fracturing would be different. Whatever the technique used to artificially fracture the shale, the geologic process which we have analyzed can never be accelerated. We have presented an analysis which remains pertinent, even under differing technical alternatives, which are currently in trial or are coming later.

We have chosen to present a purely geologic and geotechnical look at the question, because a large part of the of the other problems associated with the exploitation of non-conventional deposits of shale gas and oil follow, and follow with even more intensity, from the geotechnic ''bugs'' of this industry which has been so precipitously launched, without the adequate studies that have previously been required. There have been previous studies, but they were directed to the optimization of the commercial production of the resource.

Conclusion and recommendations:

There are two important differences between shale gas and conventional gas deposits and these two differences alone provide the fundamental reasons to disregard the ill-considered idea to exploit shale gas by the currently proposed technique:

1- The technique of hydraulic fracturing artificially creates a network of interconnected fractures to which the gas starts to migrate. The technique initiates a process of gas flow in the deposit, as happened in conventional fields over  hundreds of thousands of years, but hydraulic fracturing can't possibly accelerate this geologic process. The construction of a well and its fracturing are accomplished in a few weeks;  the flow begins and continues on a geologic time scale (greater than 100,000 years). The amount of time before the wells are closed for unprofitability, represents only a infinitely small portion of this geologic time frame.

2- The drilling of wells and the fracturing of shale is a totally irreversible operation with no technical solution to restore the shale matrix to its original impermeable condition. The closed wells at the end of commercial operation become the potential conduits for the gas leaks. For these structures, like any structure made of steel and concrete, one must ask the fundamental question of their lifespan and about what will occur when their state of deterioration no longer prevents them from resisting the pressure of the gas. The pressure of gas in the reservoir will continue to grow slowly but continuously on the one hand,  while the deterioration of the wells will increase with time. These two phenomena will appear over time on the surface by the number, and amount of flow, of the leaks of methane. The management of this new type of underground works will cost colossal sums to the public treasury as the inefficient (20%) extraction technique leaves in place a very large portion of of the methane initially present in the deposit.

The exploitation of conventional deposits remains acceptable everywhere the environmental conditions and best practices are respected. But with regard to the unconventional deposits, France has found good reasons to ban hydraulic fracturing on their territory. Economic pressures will certainly continue to make it tempting to reverse or circumvent this decision. We recommend that the questions we raise in our analysis be included in the considerations which must ensue. It would be preferable not to tie the expression hydraulic fracturing to any prohibiting regulation.  Any fracturing technique will produce the same effects as we discussed in this text.

What needs to be a priority is to re-examine the current regulations in force. Already imperfect in its neglect of the long term management of conventional abandoned wells, it becomes completely inadequate for unconventional deposits. It is these obsolete regulations, as well as some specific exemptions granted to the shale gas industry in the face of the existing regulations which have permitted North America to start this industry. With rules which would eliminate the transfer of ownership back to the state at the end of life, with an obligation that the exploiter assume complete and permanent responsibility for the well, (for example a lease of 99 years with automatic renewal of 99 years in 99 years in case of persistent gas pressure in the well) this industry would perhaps never have emerged.

The business plans of the industry only have two steps: exploration  followed by  exploitation. We have published an analysis of the profitability for all of society when the long term is included in the parameters (ref. 10,  Durand 2011) The addition of this very long stage where the costs of the foreseeable consequences of this operation are supported by the state completely changes the short term vision in which the authorities have left North America confined.

References:
1- BAPE, 2011. Bureau d’Audiences Publiques sur l’Environnement, , 323 p.
2-  National Energy Board, Nov. 2009, 
3- Aeberman, 2010. Shale Gas-Abundance or Mirage? Why The Marcellus Shale Will Disappoint Expectations. The Oil Drum.
4- Johnson D W. 2011. Marcellus Shale Gas, présentation Enerplus Corp.
5- Brufatto et al 2003, From Mud to Cement—Building Gas Wells,
Oilfield Review , Sept 2003, pp 62-76.
6- Dusseault, 2000,  Why oilwells leaks : Cement behavior and long-term consequences SPE International Oil and Gas Conference and Exhibition in China held in Beijing, China, 7–10 November 2000.
7- Wojtanowicz et al, 2001, Diagnosis and remediation of sustained casing pressure Final Report US Dept of Interior, Mineral Management Service, 93p.

Notes on the text: 
[1]  For the exploitation of the Utica Shale deposit in the St Lawrence Valley between Montreal and Quebec City , 20,000 wells would be needed to cover the 10,000 square kilometres of corridor 2, the most propitious for the first phase of exploitation,(ref 1.- BAPE Commission, 2011).

[2]  V = K . i   this is the law of Darcy, which expresses in units of speed (m/s) the conditions which govern the flow of water in geologic formations. The analysis of  the flow of hydrocarbons in geologic formations uses more complex formulas, but we are using that of Darcy because it permits a simplified approach to the notions of flow and permeability. The parameter K is also referred to as hydraulic conductivity, synonymous with the permeability of Darcy.

[3]  The expression well cap  is used here because we have not found a designated term for this engineering structure. The term designates the structures, initially conceived as temporary, for a very precise engineering function: to extract the gas. At the end of the operation, it is summarily transformed to a totally opposite function: to stop the gas from escaping the well. In addition, this structure becomes a permanent and no longer temporary in its new function.

[4]  Some states and the province of Alberta, have put in place programs for orphan wells; these programs impose on the hydrocarbon industry subscription in a fund for inspections and sealing some of the wells. But these programs are at the very beginning of the inventory stage; they never constitute sufficient means to manage the new wells transferred to the public domain.