Deliverables
The developments already undertaken by MeSy and Socomine on the completion of the well GPK2 in 1999 were pursued and led to the development of packers with a larger diameter with cupper-nickel shell which is inflated and expanded through high-pressure cement injection. The existing model already tested in 1999, operational in an 81/2" diameter wells, was modified to be able to seal 9 5/8" diameter casings in wells drilled with a diameter of 12 1/4¿.
The main problem of this development was to balance the larger hydraulic forces associated with the increase of hole size and packer diameter and still maintain the packer OD at the level of the OD of the casing couplings. Eight prototypes were successfully tested in the workshops of the MeSy company, co-owner of the patent for this equipment. The tests showed that the shells start to expand below 10 MPa and that the internal overpressure while setting the element should not exceed 16 MPa. Two packers of this type were installed without any obvious problem at the bottom of the 9 5/8" casing in the well GPK3.
Two further packers were used with the same for the completion of well GPK4. Unfortunately, in this case the situation was more complicated as it was not possible to fully complete the cementation programme as the injection pressure started to rise before the full cement volume could be pumped into the annulus. Some 10 m3 (of 40 m3 total) were left inside the casing, string. It appears that the casing elongated while lifting the cement in the annulus and the casing shoe sat down on the 8 1/2¿ hole reduction thus preventing the circulation of the final 8.79 m3 of cement. At this point, the pressure increased to above 200 bars immediately and the packers were set. With the premature setting of the packers, approximately 10 m3 of cement were left inside the casing,, some 8 m3 more than planned.
The original plan was to bring the top of the cement in the annulus up to 3960 meters. Approximately 31 m3 of cement were circulated around the casing before the packers were set, this placed the top of the cement around 4130 meters according to the Schlumberger logging measurements.
This cement top was later verified through logging operations. There were no indications that the packers had not been set.
Cementing heights of the order of 500 m above the internal casing shoe require complex operations, involving risks of premature setting of the cement (due to the duration of the operations, the length of the casing that the cement must pass through before arriving at its destination) and the high temperature at such depths. In case of failure, these operations risk to compromise, the future use of expensive wells. The use of a cement which sets as progressively as possible is therefore a first desirable condition.
Furthermore, the very high chloride content in the brines that impregnate the terrain and the need to use cements at Soultz that are capable of resisting stimulation with fluids containing hydrochloric acid make the resistance of the cement to chlorides a second desirable condition. To satisfy these two conditions, it has been decided to use once again a type of cement known as "HMR" (High Magnesium Resistant), which was already studied, then used in 1999 by SOCOMINE for the completion of well GPK2 after tests carried out simultaneously by the laboratories of the French Petroleum Institute (IFP) and the company Dowell Schlumberger. These tests, however, revealed the very great sensitivity of the "setting time" of this product to the retardant agent used and concluded that it was absolutely necessary to carry out preliminary tests on the blends of pre-mixed cements.
The basic composition of the HMR cements used can be described as follows: - 30% Portland cement + 70% blast furnace ash ¡ú HOZ cement -72 % HOZ cement +28% Fly ash HMR cement - Various additives (salt, retardant, antifoam) HMR cement therefore barely contains more than 20% Portland cement and it has a very low porosity due to the filling of the pores by the very fines particles of fly ash (which gives it the required resistance to chlorides). Cement densities may be adapted to the in situ requirement through the use of high strength glass perls. In this way the cement densities may be dropped considerably (up to 30%) still maintaining the compressive strength. However, the use of such high strength glass perls is rather cost intensive. After the two operations carried out (on GPK3 and GPK4) as part of this programme, it seems that we can confirm that this type of cement is quite adequate for the completion of wells drilled for EGS operations in media similar to the hydrothermalized granite massif at Soultz-sous-Forts. However, it is recommended that very complete tests must be carried out on site on the HMR + retardant mix as late as possible before it is used.
The measurements made by Schlumberger to explore the network of natural fractures and the petrography in the terrain crossed by wells GPK3 and GPK4 have shown that the techniques used [oriented 6-arm calliper, HNGS), UBI)] are now perfectly operational in the conditions of the wells drilled at Soultz. Various adaptations of the downhole probes used to record the very low level microseismic activity generated by the hydraulic stimulations operations have led to improvement in the lifespan of such equipment placed in somewhat difficult observation conditions (temperatures of the order of 125ºC and depths of 1500m).
The probes available today, whether they are equipped with accelerometers or geophones, have been used throughout the duration of the operations necessary to the development of the hydraulic performances of well GPK4 (September 2004, then February - March 2005).
Nevertheless, the reliability of this equipment must be improved, even if it was possible to compensate in the meantime this leak of reliability by the number of deep wells equipped.
The scientific partner CNRS / EOST for its part calculated the locations of the events using the three-dimensional distributions of the sonic speeds considered as valid according to the evolution of the physical properties of the medium during stimulation and based on a tomographic study, using the recordings of 2250 events detected at the surface (magnitudes >-0.5). An example was published (obtained by this method) of the evolution over time of the speeds at a depth of 4600m in successive sets of 250 events. This method allowed sufficiently precise locations using only the EOST's surface recordings for the relations between the spatial distribution of the events and the major known tectonic structures in the substratum to become clear. An exemple was published showing the relationship between the major fault crossed by GPK3 (whose location, angle of inclination and azimuth are supplied by UBI logging), which absorbs 80% of the flow rate injected and the position of the first set of 250 events located in this way.
The modelling of geothermal wells: The evaluation of the behaviour during experimentation of a geothermal reservoir requires knowledge from the surface at least of the temperatures and pressures at the top of the well�reservoir intersection.
On the contrary, the forecasting of the performances that can be used on the surface of a deep reservoir requires that we be able, based on characteristics of this reservoir at depth (productivity, temperature), to estimate the flow rates and temperatures on the surface of the fluid produced according to the operating conditions set by the user.
Measurements taken at the bottom of the well are often hazardous (high risk of equipment breaking down), costly, sometimes doubtful and in some cases even impossible with conventional equipments (case of production by submersible pump).
It is therefore particularly interesting to be able to calculate the data for the bottom of the well using measurements made on the surface. It is with this aim in mind that a model of a well has been developed, an example of whose application at Soultz during the stimulation of well GPK3 Modelling of the impact of the variations in fluid density Virtually all traditional numerical models fail to take account (or do so in a very rudimentary way) of the variations in the density of the fluids circulating underground.
Such variations have two causes (often combined): - mixtures of fresh water/geothermal brine In the case of Soultz, this is a major point due to the abundance of extremely salty thermal water, which is mixed with fresh water for injection in the hydraulic tests and therefore complicates interpretation considerably. - water/rock heat transfers
These will cause considerable variations in temperature and therefore in the density of the fluid and may be intense enough to generate in the reservoir zone exploited (by injection of cold water and production of hot water) some real convection currents.
Such currents will be capable, on top of the risks involved in interpreting hydraulic tests without taking these phenomena into account, of modifying long-term the distribution of the internal permeabilities in the zone exploited by dissolution�scaling reactions.
They may also have a considerable influence on the "thermal life" of the exchangers by developing their exchanges with the natural medium around them. Modelling of the hydromechanical impacts of the hydraulic tests (microseismicity): The aim of this type of modelling is to provide a contribution to the debate on the effectiveness of the hydraulic stimulation operations. The relevant models are still in the development stage.
The next stage (in progress) is essentially oriented towards the comparison observation vs. modelling of the "moment magnitudes"; however, they enable us now to simulate the microseismic activity generated by hydraulic stimulation operations at Soultz. Modelling of water-rock chemical reactions Reservoir conditions at Soultz (i.e. high temperature/pressure and concentrated brine) have involved the development of a new numerical code, FRACHEM, which couples thermal, hydraulic and chemical processes. FRACHEM is adapted from existing codes and its development was possible through the collaboration between ETH-Zürich, CHYN and CNRS teams.
These research works have been published as several PhD theses and articles. FRACHEM code, through the investigations of water-rock interactions, highlights the important reactivity of carbonates (calcite and dolomite) compared to other minerals: carbonated species, more soluble at low temperature, tend to dissolve near the injection well before re-precipitating farther in the fracture. Only a few silicates and aluminosilicates precipitate near the injection well but in low concentration. Chemical reactions induce changes on physical properties of the reservoir and in particular on the permeability.
Its evolution follows the same tendency than the one of carbonates with a permeability increase near the injection well and a low decrease in the fractured zone. This high reactivity of the carbonates is at the origin of "soft" acidification tests carried out at Soultz which results give encouraging prospects.
I:Drilling: The drilling rig was operated by ENEL, a partner of the EEIG's, under the supervision of Southern International Inc. which was also acting as �maître d�oeuvre� for the work and shared "de facto" this task with the EEIG "EMC". The latter provided the necessary logistic support. SHELL, also a partner in the EEIG, for its part played a consultancy role in the operations.
The depth and the trajectory of the wells are in line or very close with the project's objectives. I.1.GPK3:The drilling operations on well GPK3 were carried from 25 June 2002 to 13 November 2002. The total time taken was a little less than 5 months.
Overall, the work took place without any major incidents except for a slight anomaly which occurred when installing the internal casing. The casing shoe is not resting on a restriction in the well diameter as planned, but got stuck about 25 m above the objective. As a result, the casing is anchored to the granite by the packers and the column of cement above them.
The total cost of the drilling work came to just over 5 750 000 Euros. I.2.GPK4: The drilling operations on well GPK4 were carried from 29 August 2003 to 15 April 2004. The total time taken was a little more than 7.5 months. This was about 60% more than the time required to drill GPK3. This was mainly due to the fact, that GPK4 had to be drilled with a horizontal deviation of 1200 m from the wellhead in order to achieve the 600m separation at 5000m depth.
Two technical incidents occurred during the drilling operations and gave rise to insurance pay-outs. Due to the instability of the open hole section, the 9-5/8" internal casing had to be put into place before the final depth of the well was reached, thus protecting the already drilled hole. The total cost of the well came to 8 250 000 Euros, including a little bit more than 1 000 000 Euros which were covered by the project's insurers. II: Wells trajectories II.1.Definition of the well trajectory: It was decided that the separation between the injection and a production wells will be in the range of 600m. In both cases, GPK3 and GPK4, an elliptical area of 150 to 200m to the SSE of the bottom of the previous well has been defined as the target area. The situation in Soultz, in the Rhine Valley, is in so far specific as the direction of the maximum horizontal compressional stress and the strike direction of joints are close. This obviously apparently facilitates the positioning of wells as the fluid flow will be dominated by these two factors.
However, microseismicity showed that at the location in Soultz two different stress controlled fracture opening mechanisms can be observed: shearing and tensile opening. Thus a limited risk to create sub-parallel flow paths always remains.
Consequently, not knowing which opening mechanism will be dominating in the next stimulation experiment, in the case of Soultz, the well targets were positioned in a compromise location in between shear stress and maximum compression, thus theoretically allowing to link fractures through both opening mechanisms. II.2:Directional drilling: Several problems were observed during the directional drilling of the deviated, deep wells in granite.
Mainly the failures of the MWD system (measurement while drilling) caused severe difficulties. One drilling incident occurred as the drill pipe got stuck due to wellbore breakouts while a wireline survey was performed.This incident clearly shown the risk associated with these wireline surveys at great depths. MWD failures cannot be related to the failure of a single component. In fact a whole series of tool components was involved. They appear to be caused by a combination of mechanical and temperature related problems.The rough drilling conditions in granite cause problems with downhole vibrations. The high circulation temperatures tend to weaken the electronic components, even the electronic boards themselves. The combination of these two impacts results in �random� tool failures. Directional wells in granite cause excessive drill string and reamer wear. III Special equipment III.1Annular sealing package: A special equipment is required at the well head to allow the production casing to grow and shrinking according to the well temperature. The casing is sliding inside a lubricated packing which at the same time isolates and seals the outer casing annulus annulus.
The new equipment operational up to temperatures of the order of 200ºC and pressures of 35 MPa has been designed and installed on the three well heads. Its behaviour will be assessed during the next phase of the project. III.2 Drilling techniques: A new generation of drilling bits showed, during the drilling of the wells GPK3 and GPK4, performances notably higher than those observed during the previous drilling of well GPK2. The lifespan of the bits was often between 70 and 80 hours, which indicates a level of performance comparable to what can be achieved when drilling sedimentary terrain.
Comprehension of the mechanisms, which, at Soultz, could generate the biggest microseismic events, is based on a faulting system proposed by the BRGM and the EOST. From this system it appears that it is shearing in the natural fissures "en echelon" internal to the large local "faults" (mapped by BRGM ) that are probably the tectonic source of the shearing plans determined as being the focal mechanisms mapped later by the EOST using its surface network. Such a system suggests that until now the maximum magnitudes of the seisms induced by the hydraulic stimulations at Soultz were limited by this phenomenon [which might be the cause of the shift observed for the stronger magnitudes. However, it cannot be asserted that this mechanism constitutes an absolute guarantee of safety rendering impossible a shearing along the main fault plane involving much larger surfaces capable, if they slip, of generating much more energetic and therefore more dangerous seisms.
Searching for OpenAIRE data...
There was an error trying to search data from OpenAIRE
No results available