Abstract In response to increasing concerns over the environmental effects of the disposal of drill cuttings to the sea, a number of North Sea operators have begun to dispose of cuttings through injection into underground strata. Phillips Petroleum, as part of the Eko II redevelopment program for the Ekofisk field, conducted an evaluation of cuttings disposal options prior to selecting cuttings reinjection for all cuttings disposal from the new 2/4X wellhead platform. In support of the selection of the final alternative for reinjection of cuttings, reinjection of slurrified cuttings via a dedicated well into the mid-overburden overlying the northeastern flank of the reservoir, an extensive engineering study was performed. Within the study, a number of issues were evaluated and reviewed: the relevant experiences of other North Sea operators; the economic evalution of various mud systems and disposal alternatives; a complete review of the geologic structure of the overburden for determining the optimal location of reinjection; numerical simulations of fracture behavior and performance; well design considerations; final selection of the well location; contingency planning; and a subsidence risk assessment. The subsidence risk assessment, performed in cooperation with an outside consultant, evaluated the unique concerns of reinjection into the overburden overlying a compacting reservoir. Cuttings reinjection was initiated late in 1996 and by early 1998 more than 700,000 barrels of slurry had been reinjected into a single completion interval. Difficulties with the well have been relative minor, though the well has plugged-up several times, and the reinjection operation is considered successful. Building upon these results, plans are now being implemented for cuttings reinjection via dedicated wells at the Eldfisk field, another compacting reservoir lying to the southeast of Ekofisk. P. 95

Abstract In weak formations, significant pressure depletion may lead to compaction. Transmission of compaction through the rock layers overlying the depleted formation may, in turn, cause subsidence. At the Ekofisk field, a large producing oil and gas reservoir located in the Norwegian sector of the North Sea, subsidence of the seabed due to reservoir compaction led to extensive facilities modifications in the mid-1980's and has significantly contributed to the need for the on-going Ekofisk II field redevelopment project. While reservoir compaction is the source of the volume loss that leads to subsidence, transmission of compaction through the overburden strata often is a critical governing factor on its impact. Theoretical modelling work has shown, for example, the relationship between reservoir depth or width to the transmission of compaction to subsidence. Building upon the theoretical work and utilizing available field data, a numerical study was conducted to evaluate the impact that overburden behavior at Ekofisk field has on compaction and subsidence. For the study, available field data for validating the modelling of the overburden were reviewed, including reservoir compaction data, platform subsidence data from the center as well as the edges of the subsidence bowl, data from eight bathymetry surveys providing information on the extent and change of the subsidence bowl with time, and overburden core and log data. The results of the study shed light on the impact of overburden stiffness, the extent, nature, and development of reservoir compaction, and the compaction to subsidence transfer ratio (C/S ratio) for the field. P. 177

Abstract In order to get a realistic prediction of the in situ stress evolution, compaction and subsidence of the HP/MT Shearwater field, full 3-D finite element calculations were made with the program DIANA. An interface was made to generate 3-D meshes directly from seismic files, which led to a most realistic representation of the actual geometry. An extended and modified Cam Clay model was used with and without a Duvaut-Lions creep extension to describe the observed test behaviour. Detailed calculations were made of cross sections, including main faults, secondary fault systems, slip layers and wells. It was found that stress arching was very significant in the Shearwater field geometry and led to a reduction of the compaction strains within the reservoir till below the critical values. However, the stress arching also reduces the total stresses in shales, faults and slip layers in the rock directly abeve it. Assuming that the high initial pore pressure is preserved in the overburden. this may lead to very low normal effective stresses and relatively high shear stresses, and as a consequence to a reactivation of faults and slip layers at the end of field life. It was found that only secondary faults close to the main fault will be reactivated and that the slip in activated slip layers is minimal above the crest of the reservoir. Therefore a drilling window could be identified that will minimise the risk of damage to the completion. Further, zero effective stress or flill liquefaction of the overburden shales close to the reservoir was identified as a realistic problem in case of strong stress arching and high pore pressures; however because of the high strength of the shales it did not appear to be a problem in Shearwater. P. 53

Abstract A study was initiated to improve the geomechanical input to the full field compaction/subsidence prediction model for Valhall and wellbore stability analysis models used to plan long extended reach wells in the Valhall field. The results are also applicable for other geomechanical activities in the field, like well design, waste injection, hydraulic fracturing stimulation in the reservoir and the design of a potential waterflood. The study used an integrated approach to characterize the Tertiary formations above the chalk reservoirs. In-situ stresses (magnitude and direction), pore pressure and geomechanical deformation properties of the formations were determined. Results, observations and conclusions are presented. P. 193

Abstract Geologic, and historical well failure, production, and injection data were analyzed to guide development of three-dimensional geomechanical models of the Belridge Diatomite Field, California. The central premise of the numerical simulations is that spatial gradients in pore pressure induced by production and injection in a low permeability reservoir may perturb the local stresses and cause subsurface deformation sufficient to result in well failure. Time-dependent reservoir pressure fields that were calculated from three-dimensional finite difference reservoir simulations were input into three-dimensional nonlinear finite element geomechanical simulations. The reservoir models included nearly 100,000 gridblocks (100-200 wells), and covered nearly 20 years of production and injection. The geomechanical models were meshed from structure maps and contained more than 300,000 nodal points. Shear strain localization along weak bedding planes that causes casing dog-legs in the field was accommodated in the model by contact surfaces located immediately above the reservoir and at two locations in the overburden. The geomechanical simulations are partially validated by comparison of the predicted surface subsidence with field measurements, and by comparison of predicted deformation with observed casing damage. Further, simulations performed for two independently developed areas at South Belridge, Sections 33 and 29, corroborate their different well failure histories. The simulations suggest the three types of casing damage observed, and show that although water injection has mitigated surface subsidence, it can, under some circumstances, increase the lateral gradients in effective stress, that in turn can accelerate subsurface horizontal motions. Geomechanical simulation is an important reservoir management tool that can be used to identify optimal operating policies to mitigate casing damage for existing field developments, and also applied to incorporate the effect of well failure potential in economic analyses of alternative infilling plans and operating strategies. P. 403

Abstract The paper presents a field case based upon a reservoir operated by Saga Petroleum in the North Sea. The reservoir stress path was determined from a series of minifrac and fall off analyses and step rate tests on various appraisal tests and injection wells. The results show a strong heterogeneity of the stress path with a large difference between the heavily faulted Southern part of the field and the more quiet Central part of it. The stress path was further confirmed by a series of mud loss incident during drilling that are briefly summarised. Furthermore, in the case of two injectors, the step rate tests were performed after a significant re-pressurisation of the reservoir -i.e. 30 and 70 bars respectively. Such tests gave extremely low fracturation pressures compared to what was the reservoir stress path upon depressurisation. In fact, the fracturation pressures measured are consistent with the maximum depletion but not the reservoir pressure at the time of the test. In other words, the total minimum stress int the reservoir soems to be governed by the maximum depletion of the zone and not by the present level of pore pressure. Two wells form another North Sea reservoir will also be presented showing a similar behaviour upon repressurisation. Not that such an observation is consistent with one measurement performed on Ekofisk and published earlier on. These observations of heterogeneous and irreversible stress paths will be analysed and compared to other observations such as the behaviour of surface subsidence during aquifer depletion and repressurisation. P. 203

Abstract The combination of in situ effective stress and reservoir chalk mechanical weakness in the Valhall field has resulted in compaction and associated subsidence at the mudline. This paper summarizes recent finite element modeling of this phenomenon, emphasizing items of concern in a numerical simulation of this nature. The paper also presents new results from a simulation allowing inelastic overburden response. P. 377

Abstract Ekofisk, Valhall and other North Sea reservoirs are made of water sensitive rocks such as pure high porosity chalk. Production of these fields implying water-flooding leads in several cases to compaction and subsidence. In order to perform reliable predictions of compaction and subsidence when water-flooding such reservoirs, it is necessary to take into account not only the deformations that come from depletion in the compressible reservoir rocks but also to include additional deformations induced by water saturation (WS) increase. These additional deformations may eventually be determined by performing water-injection (WI) tests in laboratory. This paper presents additional tests that have been performed in laboratory in order to understand the mechanical behavior of chalk at the water-front. It shows the existence of a degree of WS above which chalks mechanically behaves as if it is fully saturated, while chalk with a lower WS looks stronger than if the degree is higher. Greater the WS is, smaller the consolidation pressure Pc0. Moreover, the deformations induced by water-flooding happen in two steps: – the first one happens as soon as chalk get into contact with water and is instantaneous. – the second is delayed and looks more like creep as it happens after chalk gets in contact with water. Observations and tests results are interpreted thanks to Piau's model. P. 505

Abstract Presented in this paper are the results and verification of field and wellbore scale large deformation, elasto-plastic, geomechanical finite element models of reservoir compaction and associated casing damage. The models were developed as part of a multidisciplinary team project to reduce the number of costly well failures in the diatomite reservoir of the South Belridge Field near Bakersfield, California. Reservoir compaction of high porosity diatomite rock induces localized shearing deformations on horizontal weak-rock layers and geologic unconformities. The localized shearing deformations result in casing damage or failure. Two-dimensional, field- scale finite element models were used to develop relationships between field operations, surface subsidence, and shear- induced casing damage. Pore pressures were computed for eighteen years of simulated production and water injection, using a three-dimensional reservoir simulator. The pore pressures were input to the two-dimensional geomechanical field scale model. Frictional contact surfaces were used to model localized shear deformations. To capture the complex casing-cement-rock interaction that governs casing damage and failure, three-dimensional models of a wellbore were constructed, including a frictional sliding surface to model localized shear deformation. Calculations were compared to field data for verification of the models. P. 429

ABSTRACT: Rock mechanics is a central component of many oil industry problems. The correct rock mechanics characterisation of formations in terms of behaviour is thus of paramount technical and economical importance. State-of-the-art characterisation relies upon core testing. After emphasizing the limits of such methods two field cases are used to show that the operational timing of the characterisation is of fundamental importance. To ensure the characterisation efficiency, it is suggested: to use data sources normally available during operation as part of a risk prevention policy; to develop new measuring methods compatible with field practice. For both areas, specific examples are given. 1. ROCK MECHANICS FOR PETROLEUM ENGINEERING TODAY As far as rock mechanics for petroleum engineering is concerned, the last decade can be considered as a "phase of intensive exploration". After the early days, when rock mechanics studies concentrated around hydraulic fracture design and pore compressibility assessment, the last ten years have seen the identification of most industrial aspects which have a rock mechanical content. A non-exhaustive list is presented in Table 1. Two of these contrasted illustrative aspects are discussed below. Perforation productivity has received only limited and recent attention from a rock mechanics point of view. It was first understood that the length of the perforation tunnel strongly affects its flow capacity and that such a length depends upon the strength of the reservoir rock. The physical nature of the permeability impairment around the perforation tunnel was subsequently identified and related to rock mechanics phenomena. From an industrial point of view the role of rock mechanics has thus been limited to feeding parameters into Semi-empirical models aimed at optimizing the perforation flow performance by selecting proper guns, charges, pressure conditions, etc. However, the margin for a larger rock mechanical involvement is quite large, particularly in the light of the first modelling studies published recently. This domain can be likened to "a field which has been discovered but remains to be appraised". Wellbore stability is a much investigated and mature topic in terms of the identification of the mechanisms responsible for the instabilities, of the observation of such instabilities in the laboratory on small scaled models and of modelling of the phenomena. The cost of wellbore instabilities has been delineated by most operators -e.g. 500 M$ per year for countries outside the ex communist block- and the topic is now widely held as a domain where major savings could be achieved by a proper application of rock mechanically based technologies. The numerous successful field cases published since 1985-e.g.- can be considered as prototypes of what a global industrial approach could be. This domain can be likened to "a field which has been appraised but remains to be developed economically". From this, it derives that the industrial challenge facing rock mechanics for the next decade is twofold:

Abstract Casing deformation as a consequence of reservoir compaction and surface subsidence has occurred in the Ekofisk Field. Casing deformation database results show that reservoir deformations occur throughout the field, while overburden deformations occur primarily on the field flanks. Deformation geometry shows compressive loading of casings in the reservoir resulting in axial buckling. In the overburden, deformation geometry suggests formation movement along near horizontal features. Identification of the deformation mechanisms will be used to refine mitigative techniques which minimize casing loading, thus enhancing well life. Introduction The Ekofisk Field is located in the south central North Sea. Ekofisk consists of three production platforms (2/4A, 2/4B, and 2/4C), two water injection platforms (2/4K and 2/4W), and other process support platforms. Pressure drawdown due to production results in compaction of the weak reservoir chalk with consequent surface subsidence. Subsidence of the seabed was first observed in 1984 and now measures 20 ft. at the Ekofisk Complex, located near the field's center. Casing deformations have been observed in wells from all platforms. Deformations are noted during drilling, workover, and wireline operations when obstructions in the casing or tubing prevent or impede the running of tubulars and wireline tools. The deformations result from the vertical and lateral displacement of formations as a consequence of compaction and subsidence. P. 507

Abstract This paper presents the development of the theory and of a computer program to predict subsidence and horizontal displacement due to oil and gas withdrawal. Geertsma's nucleus of strain model has been modified and combined with an influence function program, developed for the prediction of mining subsidence, to form a new technique for the prediction of ground movements over oil and gas fields. The modifications and the influence function theory arc described in detail, with particular emphasis being placed on the applicability of the developed theory to computer prediction. The influence function method allows complex reservoir geometry to be taken into account and both subsidence and associated horizontal ground movements may be calculated. Introduction When oil or gas is removed from a deep reservoir, the fluid pressure can reduce. Reduced reservoir pressure will result in rock shrinkage, and the reservoir will compact. Reservoir compaction may in turn cause subsidence at the surface as the overburden deforms under its own weight, which is illustrated in Fig. 1. It has been found that subsidence is practically insignificant for most oil and gas fields. Significant subsidence occurs only when one or several of the following conditions are present: When a significant reduction in reservoir pressure occurs during the production period. When the reservoir has considerable thickness. When weak and poorly consolidated reservoir rocks are present. When the reservoir has a considerable area extent compared with the reservoir depth (Geertsma, l973a). Considering the above qualifying conditions, it can be anticipated that only a few reservoirs cause severe subsidence problems, and that is indeed the case. However, under the right conditions significant and damaging subsidence effects can occur. Martin and Serdengecti (1984) have reported subsidence occurrence over oil and gas fields in several countries. For example, up to 9m of subsidence have been observed over the Wilmington oil field. It therefore becomes desirable to develop a prediction model to estimate reservoir compaction and surface subsidence in order to determine whether subsidence will be a significant problem for a particular oil or gas field. This paper presents the development of a computer program, formed by combination of a simplified nucleus of strain compaction model with a subsidence prediction influence function model. The resulting model was devised to calculate subsidence and horizontal displacement for a complex reservoir geometry. P. 621

Abstract A triaxial, downhole geophone was deployed within the Ekofisk oil reservoir for monitoring ambient microseismicity as a test to determine if microearthquake signals generated from discrete shear failure of the reservoir rock could be detected. The results of the test were positive. During 104 hours of monitoring, 572 discrete events were recorded which have been identified as shear-failure microearthquakes. Reservoir microseismicity was detected at large distances (1000 m) from the monitor borehole and at rates (> 5 events per hour) which may allow practical characterization of the reservoir rock and overburden deformation induced by reservoir pressure changes. Introduction The Ekofisk field. located in the Norwegian sector of the North Sea, produces from a chalk reservoir at approximately 3 km depth below the seafloor. The reservoir forms an elliptical dome with an areal coverage of 49 km 2 and a production interval approximately 305 m thick. Two fractured chalk formations make up the reservoir, the Ekofisk and Tor formations of Danian and Maastrichtian age, respectively. Average porosities are 32% in the Ekofisk formation and 30 percent in the Tor formation. 1 Small pore throats result in a low matrix permeability of about 1 md. 2 An extensive natural fracture system primarily controls the reservoir permeability and oil production. Fracture permeabilities as high as 150 md have been suggested from welltest data. 3

Abstract Water-injection is often implemented in oil reservoirs with mainly two goals. On one hand, it provides an efficient means of secondary production in water-wet fields by sweepage and pressure maintenance. On the other hand and precisely because of pressure maintenance it generally contributes to mitigate compaction and subsidence of highly compactable reservoirs. This last function is however severely put in question by the mechanical behaviour of North Sea chalk fields submitted to massive water-injection. From experimental results common to most labs we develop several models including a water-induced compaction, which allow to restore numerous field observations remained unexplained until now. Introduction The weakening effect of water on mechanical properties of chalk has long been recognized in the fields of Civil Engineering and Underground Works. In return the effect of sea-water on oil reservoir chalk is much more debated. Experimental findings vary a lot from one lab to another, probably because of core handling procedures. However the compaction/subsidence behaviour of North sea chalk reservoirs, which have experienced water injection, seems now to leave no doubt on a true effect of sea-water on high porosity chalk within in-situ conditions. The aim of this paper is to show how simple mechanistic reservoir models, in which the effect of water on chalk is introduced through an extra-compaction, can be put in relation with miscellaneous in-situ observations, which have remained without explanation until now. This paper is divided into five sections. We recall first some pieces of knowledge on chalk/water interaction gathered in the fields of Civil Engineering and Underground Works and summarize the main experimental results obtained in the case of oil reservoir chalk. P. 819

Abstract In-Situ reservoir compaction measurements have been performed in the Groningen field since 1974. The technique is based on regularly monitoring the distance between radio-active markers shot at regular distances across the reservoir. The application of this technique in the Groningen field is unique in the sense that it is normally used to monitor compaction in unconsolidated reservoirs, that compact much more than the Groningen reservoir. In practice, it has been proven difficult to obtain the required accuracy, 1 mm per 10 m. Until recently, the vertical rock compressibility could only be measured in three of the eight wells. The reservoir rock is so competent, that the target resolution is very close to the instrument resolution. Over the years, much effort has been put into improving the accuracy and reliability of the measurements, improvement on tools as well as on interpretation techniques. In 1993/1994, part of the surveys has been re-interpreted, using a new marker position algorithm, and a selection mechanism has been developed to identify statistically significantly deviating data. As a result, in all but one well could the vertical compressibility be obtained. The in-situ measured compressibilities match with the compressibility as inferred from the surface subsidence over the field. Introduction The Groningen gas field situated in the north- eastern, part of the Netherlands (fig. 1) was discovered in 1959. The producing horizon is the Slochteren sandstone (ROSL) at an average depth of 2900 m, that belongs to the Permian Rotliegend formation. The gross thickness increases from 70 m in the south to 240 m at the northern boundary. The porosity on average ranges from 10 to 20 percent, the permeability from 0.1 to several Darcies. The original reservoir pressure was 347 bar. Presently, after some 30 years of production, the reservoir pressure has dropped to 170 bar. It is anticipated that the field will be abandoned around the year 2050. At the start of production, in 1964, problems related to surface subsidence were not anticipated. Most other cases of subsidence are related to production of shallow (up to 2000 m), poorly consolidated reservoirs, whilst the Groningen reservoir consists mostly of well-consolidated sandstone. P. 535

Abstract Three aspects of reservoir compaction/subsidence simulation for Ekofisk are addressed: stress paths in the reservoir during depletion, impact of porosity heterogeneities in reservoir chalk on scaling laboratory compaction data to field behavior, and methods for treating the behavior of the overburden in a finite element simulator. Specially created finite element models were used for the first two issues. Results indicate that geometry has a major impact on stress path. Results of the scaling analysis indicate that porosity heterogeneities cause a block of chalk comprised of numerous cells with different porosities to appear strong when internal shear failure is not allowed, but to appear weak when it is allowed. The discussion of the overburden behavior concentrates on field observations, interpretations of these observations, and methods used to represent them in a simulator. 1 INTRODUCTION AND BACKGROUND 1.1 Brief History of Ekofisk Subsidence The late 1984 discovery of seafloor subsidence at the Ekofisk field in the Norwegian North Sea immediately precipitated an aggressive program to develop a solid understanding of the physical phenomena involved and to devise reliable methods to simulate these phenomena. 1–3 Subsidence at Ekofisk is due to compaction of the high-porosity chalk reservoir rock as a result of fluid withdrawal. When subsidence was discovered at Ekofisk, the depth of the subsidence bowl was about 3 m. Since then, the subsidence rate has ranged between about 25 and 40 cm/yr. The current (early 1994) bowl depth is about 6 m. 1.2 Numerical Modeling History Numerical methods implemented by Phillips to simulate the reservoir compaction and subsidence at Ekofisk have employed finite element (FE) computational methods. For most of this work, DYNAFLOW, a geomechanics code developed by Prevost 4 , has been the code of choice. Constitutive models for chalk were based on pore-collapse compaction 5 early in the program, and pore-collapse supplemented with shear-induced compaction 6,7 after about 1991. The latter mechanism was added because of two field observations. The first was that during the late 1980s, subsidence rates remained higher than predictions based solely on pore collapse, particularly during periods of minimal pressure decline. The second was that in-situ stresses have evolved in the reservoir during production in such a way that shear stresses become very large, large enough for localized shear fracturing and/or slippage on existing fractures to occur. The specific observation about large shear stresses is that the ratio of the change in effective horizontal stress to the change in effective vertical stress (called the stress path, K) is small at Ekofisk 8 (0.2) compared to the expected value (0.5). Just why K for Ekofisk deviates from the expected value, and in fact why K-values for other reservoirs also deviate from expectations (higher and lower) 9 , is the first of three issues considered in the present paper. The second issue is related to scaling the results of compaction tests conducted on small specimens of chalk to field behavior. In particular, the question is "how do porosity heterogeneities, which are more pronounced in large blocks (FE cell size) than in test specimens, affect model parameter values?" 1.1 Brief History of Ekofisk Subsidence The late 1984 discovery of seafloor subsidence at the Ekofisk field in the Norwegian North Sea immediately precipitated an aggressive program to develop a solid understanding of the physical phenomena involved and to devise reliable methods to simulate these phenomena. 1–3 Subsidence at Ekofisk is due to compaction of the high-porosity chalk reservoir rock as a result of fluid withdrawal. When subsidence was discovered at Ekofisk, the depth of the subsidence bowl was about 3 m. Since then, the subsidence rate has ranged between about 25 and 40 cm/yr. The current (early 1994) bowl depth is about 6 m. 1.2 Numerical Modeling History Numerical methods implemented by Phillips to simulate the reservoir compaction and subsidence at Ekofisk have employed finite element (FE) computational methods. For most of this work, DYNAFLOW, a geomechanics code developed by Prevost 4 , has been the code of choice. Constitutive models for chalk were based on pore-collapse compaction 5 early in the program, and pore-collapse supplemented with shear-induced compaction 6,7 after about 1991. The latter mechanism was added because of two field observations. The first was that during the late 1980s, subsidence rates remained higher than predictions based solely on pore collapse, particularly during periods of minimal pressure decline. The second was that in-situ stresses have evolved in the reservoir during production in such a way that shear stresses become very large, large enough for localized shear fracturing and/or slippage on existing fractures to occur. The specific observation about large shear stresses is that the ratio of the change in effective horizontal stress to the change in effective vertical stress (called the stress path, K) is small at Ekofisk 8 (0.2) compared to the expected value (0.5). Just why K for Ekofisk deviates from the expected value, and in fact why K-values for other reservoirs also deviate from expectations (higher and lower) 9 , is the first of three issues considered in the present paper. The second issue is related to scaling the results of compaction tests conducted on small specimens of chalk to field behavior. In particular, the question is "how do porosity heterogeneities, which are more pronounced in large blocks (FE cell size) than in test specimens, affect model parameter values?"

Abstract The Differential Global Positioning System (DGPS) has been adapted to the oil industry by the CEA with EAP. It allows a continuous, accurate (+ or -3 mm) and quasi-automatic monitoring of the subsidence above hydrocarbones field. The method is detailed and a simple case study explains how accurate subsidence data are able to precise the drainage pattern and help the production management. The Airborn Laser System to measure subsidence developed with IGN is now at experimental stage. Monitoring a great number of points, it will be suited to onshore fields and allow to map the subsidence and the drainage pattern changes. 1. INTRODUCTION Drop in pressure in reservoir leads to compaction. The effects gradually deform the overburden layers to finally affect the ground surface level and the installed structures. This induced movements are both vertical and horizontal (sagging and topling) and may be of various amplitude (from few centimeters to more than 10 meters). Constant research efforts are directed towards anticipation of such movements [1, 2]. While simulation tools made considerable progress concerning both theoretical approaches (i.e. within the poro-mechanical behavior of rocks) and numerical methods, the validation of this forecasted results needs high quality in-situ measurement data. Measuring subsidence methods have remained too long tied to ground settlements measuring and to optical leveling or bathymetric surveying using traditional topographical approaches [2]. This methods are ill suited to tracing temporal and spatial evolution and to the dynamics of the phenomenon (subsidence rate and his variations) [14, 16]. Over all his defects, they are discontinuous, poorly accurate, slow and for that raison, costly. Need are not only for episodic measurement but for actual monitoring methods, offering accuracy in the range of 1 cm and continuous follow-up; in addition unit cost have to decrease while the number of measures increases. Industrial constraints implie to limit needs for on site operating time and manpower.

Abstract For North Sea chalk fields, finite element models are used to evaluate the reservoir compaction and the associated seafloor subsidence. The present paper addresses aspects of the calibration of the stress strain laws to be used for the chalk and the overburden. Results of laboratory tests do not reflect the large scale deformation behaviour. The calibration is therefore mainly based on the back analysis of measured formation movements. Another topic is the coupling with reservoir simulation models, which provide the reservoir pressure. This coupling should be improved, if a local compaction cannot be derived from a simultaneous change in reservoir pressure at that location. Introduction The reservoir rock of North Sea chalk fields like the Ekofisk field locally shows porosities of more than 45%. The decrease of pore pressure in the course of production of oil or gas results in an increase of effective stress causing pronounced pore collapse. The compaction of the reservoir can reach several meters, and induce the failure of casings and a considerable seafloor subsidence. Finite element models are used to evaluate these displacements in numerical simulations. Such models are developed parallel to reservoir simulation models, from which the reservoir geometry, initial porosities and initial pressures as well as the time history of reservoir pressure are taken. Since 1988, the ISAMGEO model has been used in various studies for the Ekofisk the Valhall the Ekofisk and the Tyra fields. The large scale displacement field was evaluated for assessing the risk of casing failures in the overburden and for a forecast of compaction and subsidence. Most of the studies deal with the Ekofisk field, which is the largest of these fields and is on production since 1971. A great number of papers deal with compaction and subsidence at Ekofisk This paper focuses on aspects of the calibration process by matching observed formation movements and on aspects of the coupling with the reservoir simulation model which apply not only to Ekofisk. Typical ISAMGEO full field compaction models have about 30000 degrees of freedom. Additional subsidence models with refined overburden discretization have a similar size. The size of these models as well as the complexity of reservoir geometry and pressure history make it difficult to identify characteristic results. This paper therefore refers to the results of finite element calculations for which small idealized models are used. These small models may not be considered as a substitute for any quantitative compaction and subsidence analysis. P. 795

Abstract Geotechnical safety for a cavity field will be assessed on the basis of model computations. The statistical uncertainty of rock salt parameters as well as possible variations in the dimensions of a cavity field under design result in the necessity to carry out a parametric study. The purpose of the paper is to explain probabilistic modeling with respect to structural response computations, using sensitivity and uncertainty analysis. Probabilistic finite element computations have been carried out for a hexagonal cavity configuration, The results are discussed with respect to the variability of the following parameters and their interrelation: diameter, spacing, depth and height of the cavities, creep of the salt rock, and stiffness of the overburden. Introduction Salt deposits are being used not only as a source of minerals but increasingly also for the storage of primary energy reserves and for the permanent disposal of hazardous wastes. According to the favourable geomechanical properties of halite, namely steady state creep behavior, which allows the salt to close and to heal up any fissures or open gaps, a rock salt formation is functioning as a tight barrier. Moreover, rock salt allows the construction of large cavities without any particular support. However, the proper design of cavities for the different purposes requires: site specific investigations with respect to geological features, planning and determining the most favourable geometrical dimensions of the cavity configuration, e.g. diameter, spacing, height and depth. Besides that, the design of the cavities may demand for particular considerations for the different live stages concerning construction, operation and abandonment. The utilisation of salt cavities for the purpose of isolating hazardous wastes against the biosphere necessitates safety assessment with respect to long-term integrity of the host rock. P. 761