The unprecedented time span of concern and consequent need to predict future geological evolution, coupled with the potential consequences of failure, dictate that geological characterizations prepared for nuclear waste repository evaluation be immensely complex. The complexity of such investigations can, however, be used to optimize the knowledge achieved by a synergistic interaction between the individual geological disciplines. Additionally, we view the use of "back analysis" of geological structures as an essential tool of the characterisation process, for example for the assessment of constitutive relationships applicable during long periods of time. Considering a sedimentary basin as an example of a geologic system to be characterised, the role of the stratigraphy/lithology and structural geology/tectonics disciplines are reviewed. Comments are included on the role of vertical crustal movements in intraplate environments, as a directly measurable and therefore readily appreciable facet of the evolution of geologic systems. The significance of vertical crustal movement upon strain energy changes is noted. The roles of state of stress and geomorphology/palaeoclimate are introduced and used to illustrate the use of interaction between disciplines to optimize the geological characterisation. INTRODUCTION Emplacement in a geological system is a favored solution to the alternative proposals for isolation of high level nuclear waste. Selection of a suitable site, repository design and safety assessment depend upon a characterisation of the geological system within which the waste is to be stored. It must be assumed that the quality and relevance of geological characterisations prepared for repository development will therefore be directly reflected in the subsequent performance of the repository. In general, all activities involved in the isolation of waste in the earth's crust depend upon the geological characterizations prepared for this purpose. Because of the potential consequences of failure of a waste isolation facility, it is assumed that the investment in geological exploration will be substantial and thus the evaluation of geological conditions will be detailed. This in itself means that such investigations will be complex. Additionally, however, the comprehension of the geological system which must be achieved by such investigations is unprecedented. The lack of precedence arises largely as a result of the vastly increased time scale of concern. Previous investigations for potentially hazardous facilities have been related to engineering lives of generally less than 10 2 years; the nature of the nuclear waste to be isolated is such that time spans of the order of 10 6 years are of concern. Thus, the geologist is rather abruptly faced with the change of perspective determined by a time scale which is increased by approximately 10 4 years. This means that the geologist must consider evolution of the geologic system during the life of the waste repository. Other reasons which influence the level of the characterisation to be achieved, and the difficulties in achieving such, include the need to minimize disturbance of the rock mass, to account for groundwater flow at a regional scale and to account for thermomechanical effects. However, it is the time scale of concern and the consequence of failure which primarily determine the complexity of the geologic characterisation, and the need to optimize investigation.

Rainfall variability is a major constraint to human life in arid and semi-arid areas. The variations can be seasonal as well as interannual. When the average precipitation is very low, fluctuations in rainfall, both as droughts and as floodings, can have severe consequences for a country. Many developing countries are situated in the arid and semi-arid belt on the Earth, e.g. in the Sahel belt of Africa. The rainfalls in these regions mostly appear as showers. Due to the dry soil most of the rainwater will go into surface run off and evaporate. Lakes and other natural basins for the collection of rainwater are very rare. To overcome this unequal distribution of rain it is necessary to find some kind of storage, natural or artificial, in which the rainwater, infiltrated into the ground, can be stored to be used during the dry season. The usual artificial type of storage is surface dams. However, they are very unfavourable in dry areas due to intensive evaporation and sedimentation. Underground storages in soil or rock do not have these disadvantages. On the other hand it is a problem to find necessary storage capacity in these media for the infiltrating water. Deep layers of alluvials have been used since long as aquifers. The storativity of these deposits is dependent on the extent of the strata and its porosity. Sedimentary rock, e.g. sandstones and limestones, have a natural porosity. They mostly are good aquifers. Hard rocks on the other hand have a secondary porosity, a fracture porosity, which can be waterbearing. Due to the tectonic type of the fractures, shear or tensile, they have quite different storage capacities. The authors show how special tensile fractures systems have a remarkable storativity. A tectonic model of this type of fracture is demonstrated. The rain water, which infiltrates the rock via the soil, is drained downstream the fracture system by gravity. A method is shown how to stop this drainage by means of an underground dam. This is achieved by an injection screen transvers to the main direction of the fractures. The authors show how the ground water storage in the subterranean dam and how the necessary leakage downstream are controlled by a simulation model. In areas, where drinking water is very costly, the storativity of the tensile fracture system can be increased considerably by the construction of an underground rock storage in close connection to the fractures. A preliminary model of this arrangement is discussed. INTRODUCTION Strong climatic fluctuations, seasonal as well as interannual are characteristic for the arid and the semi-arid areas of the Earth. In large parts of Africa the mean deviation in annual rainfall is more than 25 % of the average. This means that rainfall of less than half the average or more than twice the average is a real risk in arid and semi-arid areas (Morales, 1977). The rainfalls of this type of climate mostly fall as showers. Due to the dry soil most of the rainwater goes into surface run off and finally evaporates.

Nations confronted with the safe disposal of high-level radioactive wastes expect to solve that problem by using subsurface repositories in deep (<1,000 meters) geologic formations. Principal rock types being considered include domal and bedded salt, granite, argillaceous sediments, basalt, volcanic tuff, and anhydrite. Geologic and hydrologic characteristics of these rocks are being investigated on a regional scale as well as at several specific sites. Thermal, geochemical, and rock-mechanical properties are also being studied by in-situ and laboratory methods. Subsurface disposal of radioactive wastes has gained increased acceptance from twelve years of favorable experience at the converted salt mine at Asse, Federal Republic of Germany. That nation has also selected the Gorleben salt dome in which to develop the first repository for high-level wastes. Considerable research has been conducted on bedded salt at the WIPP site in the United States, on the granite at Stripa, Sweden under a joint Swedish-American program, and on Tertiary clays at Mol, Belgium, and Trisaia, Italy. Collective developments from fourteen national programs are summarized and include additional applications of "subsurface space" for the disposal of radioactive wastes other than high-level. INTRODUCTION During the decade of the 1970's there developed within those nations confronted with the permanent disposition of radioactive wastes an increased commitment toward safely utilizing subsurface space to solve this international disposal problem. Where intermediate- and high- level wastes, including spent fuel, are involved, the technologic approach that has clearly emerged as the most widely supported, especially by the technical-scientific community, and is the most extensively pursued by every national waste-disposal program involves the engineered containment of solidified wastes within subsurface repositories. Although low-level wastes, including some transuranic elements, continue to be emplaced by several nations in shallow trenches and lined boreholes at land-burial sites, interest has also increased in using shallow-depth, bedrock repositories or converting abandoned subsurface mines to handle this category of wastes. The only non-repository disposal schemes currently being followed or investieated are injection of liquid wastes through controlled boreholes into deep rock units, and placement of containerized wastes into the soft sediment of the deep-ocean sea floor. For more than ten years, the USSR has injected low- and intermediate-level wastes into confined sandstone aquifers some 1,500 meters deep below a site near Dimitrovgrad. Soviet activities in the design of surface engineered storage however seem to indicate a decidedly lower potential for deep-well injection relative to the disposal of high level wastes. Slurry-cement injection of intermediate-level wastes combined with hydraulic fracturing accomplished beneath the United States government facility at Oak Ridge, Tennessee during the 1960's is presently not practiced. Research into true seabed disposal, as opposed to conventional ocean dumping, is very preliminary, and only three nations, Japan, United Kingdom, and United States, have seriously investigated this approach. Seabed disposal faces both formidable technical problems as well as numerous international policy issues. In the opinion of the authors, these collective obstacles could well preclude this method from becoming a viable disposal option for any of the more hazardous radioactive wastes.

The storage and transport of water-borne heat directly in rock caverns and tunnels is today a relatively untested technique which is judged to offer economic and technical advantages when applied on a large scale. The storage of hot water however, causes a change in the chemical composition of the storage water and gives weathering products on the exposed minerals of the cavern wall. These exolution and weathering processes are very complex and it is at present necessary to determine the ion exolution rate for each rock type at a certain specific water quality. INTRODUCTION A research project was initiated in the early 1970's by the Swedish Rock Mechanics Research Foundation (Befo) in order to solve these problems of large scale hot water (l500C) storages. This project was divided into various specific subjects dealing with the rock mechanic due to heating, the behaviour of concrete structures in hot water, the chemistry of water-rock interaction and the corrosion - precipitation in heat exchanger systems. The present results of water-rock interaction when storing hot water in crystalline rocks show changes in the natural chemical equilibrium between the rock and ground water which previously prevailed, chiefly as a consequense of increasing temperature and pressure. For most rock forming minerals, except the carbonate group, there is an increase of solubility due to increasing pressure and temperature. This causes a change in the chemical composition of the storage water and gives a precipitation on the exposed minerals of the cavern wall. This precipitate, however, does not prevent further ion exolution from the hidden mineral phases. It has also been proved that there is an equilibrium between the newly formed weathering material and the water phase. The ion content of the storage water reaches after a period of time a stable level. It is evident that the exolution processes are complex and even if it is possible to get data for a single mineral system, it is impossible to combine the results from tests of single mineral systems forming a certain rock type and thereby get a good estimation of ion exolution. Studies have been carried out on the mineral composition of the granites used in the weathering tests without getting any correlation between the solid phase and the water phase composition. Therefore, it is now necessary to determine the ion exolution rate for each rock type at a certain specific water quality. PREINVESTIGATIONS AND LABORATORY METHODS. The laboratory tests in this investigation have been run at 1500C and 0.6 mPa in small teflon lined autoclaves with a total volume of 200 cm3 and a water content of 75 cm3. (Fig. 1.) The water in the autoclave entirely covers the rock slices. All the reaction water is replaced after the first week and after that every fourth week as long as the test goes on. The first replacement is due to the initial rapid reactions. This procedure leads to an increasing ion exolution process for the succeeding four week periods.

A 14.5 km long tunnel in bedrock constitutes an essential part of the heat transmission line by which district heat will be transmitted starting in 1982 from Naantali Power Plant to consumers in the region. This report examines the financial and technical factors which led to the choice of the tunnel alternative, as well as the tunnel design including investigations and studies related to it. The report examines also the environmental harm and risk factors caused by rock excavation and the use of tunnel in a population center. The concluding part of the report deals with the time schedule, mode of implementation and cost of project. BACKGROUND Imatran Voima Oy's Naantali Power Plant, 3 × 133 MW, constructed in 1958–1978 is situated on the border of a densily populated region comprising two coastal towns Naantali and Raisio and one coastal city Turku. There are 200 000 residents in this region. Until recent years the buildings have taken care of their own heating with their own boilers with the exception of small area heating systems. Steep escalation in the prices of oil - the fuel primarily used - has made the transfer to district heat and to concentrated production of heat more economical in large production units using cheaper fuel. After 1982 the major part of the district heat required by the Turku area will be produced with coal at Naantali Power Plant. Because of this the condensing units with a capacity of 133 MW will be altered to produce district heat. Owing to improved efficiency one unit will produce in addition to 90 MW of electric energy also about 170 MW of district heat. However, the construction of the heat transmission system with its stand-by and peak heat centres form the major part of the project as far as costs are concerned. In order to take care of the construction of the district heat line and heat transmission to consumers at a maximum distance of 25km, the municipalities in the region and Imatran Voima Oy have established a limited company by name of Turun Seudun Kauko Lampö Oy. Alternatives The terrain in the Turku region on the shore of the Baltic Sea comprises rocky hills and between them valleys filled with soft clays. It was found that the varying and difficult foundation conditions would increase the cost of an ordinary district heat channel to be sunken in the ground owing to great need for piling and supports in the excavation. Likewise placing large approx. 1.4 × 2.4 m DN 800 district heat culvert in the narrow street areas of the old city would have caused expensive alteration works in the existing municipal pipelines and cables as well as great difficulties in traffic arrangements. It would have been impossible to avoid lowering of the ground water level in sections where the ground water level is high. Even crossing of water courses, with railways and highways would have been cost-increasing details.

The reasons to store sand in in-rock bunkers are described. A case history on the planning and excavation of the Malmi sand bunkers is presented. A complete state of stress-measurement was carried out and the whole structure was analyzed with a finite-element method. The results of these investigations verified assumptions based on previous experiences. The construction was carried out by me ns of long-hole drilling from above simultaneously with tunnelling from below. INTRODUCTION Sanding is a well-known and widely used method of improving trafficability and traffic safety on snowy and icy roads. Under heavy traffic sand tends to fly up from the road surface and that's why salt is mixed with the sand. The salt solution forms a membrane around the sand grain. This membrane melts a hole in the ice and snow and the sand grain will stay on the road. However, when much salt is used, it causes increased corrosion in cars. In order to make sanding as efficient as possible, the sand used must be non-frozen and readily available. The sanding must be done within a short period after a snowfall, or when the roads are wet and the temperature below 0°C. The need for speed is the main reason for storing the sand near the place where it is needed.' Sand can be stored in various ways: in heaps, in sand-boxes, in sheds and in special bunkers built for this purpose. When the sand is stored in heaps, pre-salting is necessary to prevent freezing. This can, however, have negative environmental effects, e.g. damage to vegetation and pollution of ground water. Sand-boxes, being small in volume, are useful only at places where the need is quite local, e.g. at railway crossings. The sheds are usually light wooden structures, in which sand can freeze if salt is not added. The special bunkers are built on or under the surface. In bunkers built on the surface the sand must be warmed up to keep it non-frozen. This, of course, involves the use of energy and is not economical. The underground bunkers are mostly blasted out of rock. Some typical bunker shapes are shown in Fig. 1. In these in-rock bunkers the sand does not freeze. Furthermore, filling and discharging is easy, thanks to gravitation. Since they are constructed under the surface, the bunkers can be large enough to fulfill regional sand requirements for a long time, although they can be refilled in the middle of winter, if necessary. By using underground space for stores, valuable areas are saved for houses and parks. Only slight changes, i.e. the roads to and from the bunkers, can be seen after construction. SAND BUNKERS IN HELSINKI In Helsinki there are some 900 km of roads which need sanding in winter. To satisfy the increasing demand for sand it has been necessary to store sand in several places in the city area. The main stores are in-rock bunkers, from which the sand is distributed.

A geotechnical investigation was conducted to confirm the feasibility of using the Central Rock Limestone Mine, Lexington, Kentucky, U.S.A. for crude oil storage as part of the Strategic Petroleum Reserve. Evaluation of containment potential was based on the criteria that there should be a static head of water above the mine greater than the maximum allowable operating pressure in the vapor space in the storage cavern, and that there should be an hydraulic gradient toward the mine sufficient to prevent escape of vapors. In the investigation, emphasis was placed on evaluating the groundwater conditions around the mine by means of piezometers installed in boreholes drilled from the surface. INTRODUCTION As part of the United States Department of Energy Strategic Petroleum Reserve (SPR) program, a number of existing, conventional room and pillar mines have been evaluated to establish suitability as crude oil storage caverns. These include two salt mines, Weeks Island and Cote Blanche in Louisiana, and two limestone mines, Ironton in Ohio and Central Rock in Kentucky. This paper presents the results of the geotechnical certification program conducted for the Central Rock Limestone Mine which has a potential storage capacity of approximately 14.0 million barrels (2.2 million m 3 ). Emphasis is given to the procedures used to demonstrate that the existing groundwater conditions around the mine are such as to prevent product leakage. CERTIFICATION CRITERIA Prior to the SPR program, there had been widespread experience in the United States with the storage of LPG's and natural gas in solution caverns in salt and in purpose-built mined caverns in a variety of rock types. Natural gas had been stored also in a converted coal mine, but there had been no experience with underground storage of crude oil either in solution caverns, purpose-built mined caverns or converted mines. This lack of experience meant that there were no established criteria for certifying a mine as suitable for storage. The solution caverns included in the SPR have been certified by means of hydrostatic testing, a straightforward technique for a cavern that is always full of a fluid. This technique could be considered for a converted mine using either water or a gas, but it would be expensive due to the extensive measures required to seal the mine, and possibly environmentally unacceptable due to problems with disposal of the test fluid. It was thus determined that the converted mines in the SPR should be certified by indirect means on the basis of thorough geotechnical investigations. The general criteria established were that the mine should be sufficiently stable as to allow safe and satisfactory operation for the projected life of the facilities, that there should be no significant leakage of stored product, and that there should be no adverse reaction between the stored product and the host rock. In the case of an existing mine, it is relatively straightforward to assess stability conditions on the basis of in-mine inspection, supplemented by appropriate quantitative analyses. The potential for reaction between the host rock and the stored product can be evaluated with appropriate laboratory tests and is also a relatively straightforward criterion to satisfy.

In underground air-raid shelters there are located many functions for civilian use, for instance storage and sports. In Turku, Finland, an underground shelter is being built in which two ice-hockey rinks will be situated in adjacent halls. The unusually long spans of the halls (32 m) presume thorough site investigations and accurate calculations. The construction site is composed of firm crystalline rock. Excavation and strengthening work will take place in many different stages. CIVILIAN USE OF UNDERGROUND BOMB SHELTERS IN FINLAND According to Finnish legislation new buildings of a volume of at least 3000 m 3 shall be provided with air-raid shelters. The air-raid shelter can be built either in the basement of each building or, alternatively, regional shelters can be built with spaces provided for several buildings, blocks or a whole district. As the Finnish bedrock generally is of good quality~ in almost every place there is rock available, the large regional shelters are almost always built in rock. In the rock shelters the space reserved for one person is 1.1...1.2 m 2 . Usually the rock shelters have plenty of hall space which can be used for civil purposes. The designing and constructing of rock shelters should always be programmed so that the intended civilian use is known already at the initial stage of the designing process, thus making it possible to design the spaces, technical equipment, passages and entrance structures above ground purposefully. from the point of view of the use as well in crisis as in normal times. When the constructing of the rock shelters started in the early sixties in Finland, the shelters were designed to be used in normal times as simple storage rooms or car parks. When the first shelters were completed it was discovered and generally accepted that the shelters could, with small additional expenses, be equipped for more exacting use and activities. In the rock shelters of today the civilian use is more versatile than earlier. Halls for gymnastics and playing ball, running tracks for athletics, places for jumping and throwing have been built in the shelters and in two towns there are public swimming pools in the rock shelters. In addition to that, provided in the rock shelters. bowling alleys will be built. rinks will be built. hobby rooms for young people and rooms for musical training are In the shelters being designed 'at 'present, cinemas, small, theatres, In the rock shelter of the Varissuo -residential area in -Turku two ice THE ROCK SHELTER IN VARISSUO The building of the Varissuo residential area in Turku was started in 1978 and by now about 60% of the dwelling houses have been built. The total area of the dwelling houses is about 320,000 m2, excluding schools, nursery schools, business buildings etc. The topography of the area varies and there are large exposures of rock. It is possible to build the regional rock shelter for the whole residential area, as the area is being built centralized and rapidly by one and the same builder, the population of the area is sufficient and the bedrock is suitable for the purpose.

For the past 20 years almost all new storage plants in Finland, Norway and Sweden installed for crude oil and refined products have been constructed in mined, unlined caverns. This has proved to be the cheapest and safest way. However, the promotion of this technology in other countries has been more complicated than expected. Many obstacles have appeared e.g. less favorable geophysical conditions, fears about environmental damage, laws etc. not related to underground storage. Another major obstacle is the present uncertain state of the petroleum world market. The various obstacles we have encountered are discussed in this paper. BACKGROUND The method of storing crude oil and refined products in mined, unlined caverns is, in Finland, Norway and Sweden, the cheapest and safest way provided that the quantities are not too small. This is confirmed by the fact that during the two last decades nearly all new storage facilities - both commercial and strategic - have been constructed underground. More than 100 plants with a total storage capacity of more than 25 million cubic meters (160 million barrels) have been completed. With this experience and knowledge of all the advantages, considerable efforts have been made to introduce this technology in several other countries. But what was simple and easy in the three countries mentioned above has proved to be more complicated and has taken much longer than expected. We are surprised and somewhat disappointed that this technique has not been generally accepted in other countries and we have difficulties in understanding the reasons for this. Based on the experience of several pre-investigations and feasibility studies executed in South and Central Europe, North America, Africa and Asia the different causes of the obstacles met with are discussed here. GEOPHYSICAL CONDITIONS The geophysical conditions in our countries are in general favorable for subsurface constructions. Most of the bedrock consists of very competent granite or gneiss. Decomposed and weathered rock was removed during the last big ice age which means that it is easy to reach the bedrock without any major excavation of overburden. However, it should not be assumed that all rock in these countries is of good quality; now and then we also encounter bad rock conditions. Special methods and equipment have been developed to be used in such cases and the projects can still be completed on a sound economic basis for the owner. The ground water situation is also favorable and it is usually easy to find a stable groundwater table, the level of which is secured and proved by proximity to the sea or lakes. The lack of active seismic zones is also important. It is quite obvious that it is not possible to find places all over the world with this favorable combination of conditions. However, there are several places with similar, good conditions in many areas of the world. Simple estimates also show that there are considerable economic advantages to be gained in constructing storage facilities even in less favorable rock masses; such plants are still able to compete on an economic basis with steel tank storage systems.

The purpose of this paper is to review a number of design proposals and to summarize and assess the current state of the art in underground nuclear power plant design. In addition Some conclusions is drawn and a model of a safe nuclear power plant underground is suggested. 1. INTRODUCTION Underground siting of nuclear power stations has drawn increasing attention in various parts of the World. European countries took the early lead in this field because they have experience in the design and construction of underground hydroelectric stations and industrial plants, and because they have near-surface rock formations which are ideally suited for hosting large underground caverns. The world's first underground nuclear facilities were built in Norway (Halden) and was commissioned in 1959. Recent feasibility studies in a number of countries have proposed conceptual designs for full scale nuclear power stations sited in either mined-rock cavities or in shallow excavations covered with backfill (pit siting). These investigations, primarily in European and North American countries, have revealed a number of potential benefits of underground siting: improved containment of radionuclides under extreme accident conditions, increased station protection and security, reduction in seismic loading on nuclear equipment and structures. These benefits, however, could be realized only in combination with higher station construction costs and certain operational penalties. Because the above investigations were based on a number of different basic reactor designs and were geared to different national regulatory standards and different geological environments, a wide variety of underground design concepts have evolved. The purpose of this paper is to review a number of these design proposals and to summarize and assess the current state of the art in underground nuclear plant design. In addition, the paper attempts to weigh the advantages and disadvantages of this unique siting concept, as identified in the various investigations, and to draw some conclusions to assist energy planners in making decisions on the complex issue. 2.0 RECENT FEASIBILITY INVESTIGATIONS This section reviews the more recent and prominent feasibility studies in several North American and European countries. It is impossible, in the allocated space, to do more than highlight the major design features, economic implications, and general conclusions which were derived by these investigators and agencies. 2.1 United States The early feasibility studies of underground nuclear plant siting in the United States were carried out as joint efforts by the Aerospace Corporations/California Institute of Technology (1972) and by United Engineers/Acres American (1974). Both of these studies were preliminary first cut attempts at providing a broad but cursary evaluation of underground siting. Conceptual plant designs were not developed beyond the stage of a general plant arrangement and thus evaluations of the plant safety and economics were very qualitative. Nevertheless, the tone of these early reports was clearly optimistic and prompted other agencies to take a closer, more detailed look at the concept. In 1977 Sandia Laboratories, under contract with the US Nuclear Regulatory Commission, carried out a technical assessment of the potential benefits and penalties associated with underground siting of nuclear power plants.

The following perspective was presented for discussion at Rockstore ''80. It is intended to address, in very general terms, the following three questions: Upon what experience can we base geological investigations for subsurface nuclear waste isolation facilities? What are the additional requirements to be met, for which our previous experience must be extended? What approaches can be adopted to meet these new requirements? INTRODUCTORY COMMENTS Geological characterisations provide input to all the subsequent phases of repository development, including performance assessment. Consequently, the quality and accuracy of safety assessments can directly reflect the quality and accuracy of geological characterisations. Before presenting this perspective, it is appropriate to explain the meaning of some terms to be used in this discussion. The word "system" is not used in the traditional geological sense, but in the more universal sense to denote a body of interacting and interdependent parts. The disposal medium is viewed as a part of the system to be employed, not to be treated in isolation from the geological system of which it is only a part. An additional term, characterisation, is used in the recognition that we cannot expect to define all the parts of a system as complex as the geological system within which disposal of waste is to be carried out. The overall objective must be to gain a knowledge of those processes pertinent to containment within the system which is sufficient to rationalise the potential for waste/system interaction. AN HISTORICAL PERSPECTIVE A new class of industrial development has appeared within the last 30 to 40 years. This class of engineered structures can be identified on the basis of the consequences of operational failure. For these structures, such failure is apparently of potentially far greater severity, in terms of public health, than is the case for other industrial development. In recognition of this fact, a far greater level of effort and care in siting, design, and construction is warranted. Examples of this class of structures are nuclear power plants and such chemical facilities as LNG plants. Commonly, the design lives of such facilities are approximately 40 years. Regulation of such development has been established at a national level. The relevant national and other regulatory bodies have sought a much more stringent degree of performance assessment than previously known. Because of the potential impact which such a facility can have in terms of escape of hazardous material in the event of operational failure, a knowledge of the natural environment of the facility is required. The process of safety assessment is thus dependent upon a wide range of scientific disciplines, not simply upon that which governs the industrial process of concern, such as chemical or nuclear engineering. Yet it would be unreasonable to assume that man has an equivalent level of understanding of each technical discipline of relevance. To assume that the levels of knowledge of nuclear physics and of geologic processes were similarly advanced at the onset of nuclear power development, for example, would be absurd.

INTRODUCTION A cooperative project in the management of radioactive waste storage is currently in progress in the Stripa Mine, located in central Sweden, Fig. 1. The formal agreement between the participants - the Swedish Nuclear Fuel Safety Project (KBS) organized within the Swedish Nuclear Fuel Supply Company (SKBF) and the Lawrence Berkeley Laboratory (LBL), University of California sponsored by the US Department of Energy (DOE) - was written in July 1977 and the work was initiated shortly afterwards. The main tasks in the project are heater experiments and hydrogeological investigations in the Stripa granite. Analysis of instrumentation behaviour and mathematical modelling related to the different experiments constitute an extensive part of the program. Additional support are geophysical measurements and determination of material properties. Following the Swedish-American work in Stripa is a multilateral project - the Stripa Project - concerning the same subject as previously. Economical and technical support are provided by Canada, Finland, Sweden, Switzerland and the United States. The Organization for Economic Cooperation and Development (OECD), acting on the basis of the radioactive waste management program, has undertaken to promote the cooperation. Sufficient interest has been established to proceed under the auspices of the Nuclear Energy Agency (NEA) and preliminary proposals have been developed. ORGANIZATION The organization plan for the Stripa Project is shown in Fig. 2. The Swedish SKBF/KBS organization, in charge of the operation of the Stripa Mine, will act as the host organization for the project. Experimental programs will be recommended by the participating countries and reviewed ~y technical subgroups (TSG's). The technical subgroups cover the fields of hydrology and chemical transport, rock mechanics and engineered barriers. The TSG''s will consist of members from the participating countries and each task will be assigned to a principal investigator. Supervision of the experimental program will be entrusted to a Joint Technical Committee (JTC) composed of representatives of participating countries that are providing financial support to the project. Full membership in the Joint Technical Committee will be contingent on commitment of at least $400.000 (US) per year unless otherwise agreed by the committee. The project is planned to last 4 years. EXISTING TECHNICAL PROPOSALS Below a summary of existing experimental programs within each technical field is given. The programs are preliminary and will be reviewed and commented by the representatives of the different countries in the technical subgroups. The Joint Technical Committee will in each case be responsible for the financial management and the budgeting of the project. Hydrogeological Investigations Objective. The purpose of the project 18 to design and test methods and instrumente for geologic, hydraulic and geophysic studies in hori~ontal or subhorizontal boreholes to a length of 300 m. where the purpose is to determine hydraulic characteristics and interactions between fractures in the bedrock, see Fig. 3. The project aleo aims at gaining further knowledge on the chemical and physical conditions in the groundwater at great depth in crystalline bedrock by deepening an existing vertical borehole to a depth of 1000 m. In addition, the influence of drift excavation on the hYdraulic conditions will be studied.

Storage caverns in salt in sizes up to 1 million m' have been made available in the past primarily in saltdomes. Planning is executed using calculation methods derived from mining experiences and laboratory investigations on salt core material. In the meantime the construction of storage caverns even in stratified salt formations is becoming a necessity, as the dissemination of such salt deposits is far larger than that of the saltdomes. The computation procedure used for dimensioning and assessing the stability of saltdome caverns are now only conditionally applicable. Examples of realized and proposed projects show the development of the stability assessment for caverns in stratified salt. Stress rearrangements connected with the construction and operation of storage caverns in the rock material are evaluated using theoretic simulation models and are more realistically investigated using numerical computation - such as the finite elements method. INTRODUCTION The installation of caverns in salt formations for the underground storage of hydrocarbons has become increasingly widespread since the beginning of the 50's. Significant for the decision to construct an underground storage were its low investment costs compared with the conventional tank storage. For the seasonal balance between rate resp. production and consumption of primary energy relatively large storage units were necessary. For this purpose conventional tank storages whose economicability lies especially in high turnover rates are less advantageous. As long as solely economic criteria are decisive, it is only too understandable that for the selection of the most suitable storage formation the salt masses developed into mighty saltdomes penetrating right into the proximity of the surface can be particularly considered. The solution- mining process has long since been mastered, although a definite geometrical form could not first of all be established due to the lack of suitable measurement procedures for checking the cavity growth. Now such measuring procedures exist. Very soon, however, it became all too apparent that it was necessary to calculate the creep strength of caverns in order to ensure their serviceability for many years in the case of everincreasing individual volumes by appropriate planning of the cavern configuration as well as of their distribution resp. arrangement. The dimensioning of safety pillars was attributed decisive significance. Individual cases became known, whereby caverns suffered considerable volume losses after a short operation time due to unusual convergence. One was faced with the task of dimensioning and arranging the planned cavities among one another in such a way as to be able to achieve maximum effective volumes on the one hand and on the other hand to ensure that the convergence measured by the expected lifespan of the cavern remained within a tolerable size order. A further condition was the optimal utilization of the salt deposit, be it of the deposit itself or even of the available, usable area above ground. In the meantime cavern sizes of 0.5 - 1.0 × 10 6 m 3 have been established in numerous storage facilities or are in construction and caverns of 1.6 × 10 6 m 3 are in the planning stage. The dimensions of an existing 585 000 m 3 cavern are reproduced in a vertical section in Fig. 1.

Efficient utilization of solar energy and industrial waste heat requires heat storage facilities. Such facilities can be provided in groundwater aquifers. Two aquifer heat storage projects currently being carried out in Sweden are presented. INTRODUCTION Efficient utilization of solar energy and industrial waste heat presupposes the availability of short-term and seasonal heat storage facilities. Storage in natural groundwater aquifers is at present being studied in Sweden as one possibility. Various kinds of geological formation can be used, primarily in sedimentary rocks or Quaternary deposits. An ideal storage facility provides natural boundaries against water and heat flow from the facility. Methods are being developed to control this flow. Heat storage in aquifers could be used for internal heat storage in factories or as part of a district heating system. Low temperature storage can be used in combination with heat pumps. In this paper, two projects conserning heat storage in connection with district heating are presented. EXPERIMENTAL PROJECT ON STORAGE OF SOLAR ENERGY IN A SEDIMENTARY ROCK AQUIFER Background A pilot low energy housing project is being planned in Landskrona, in the south of Sweden. The basic concept of the project offers a range of possible energy-saving and energy-collecting installations. One example is a glass-covered street or arcade which functions as a giant solar collector for nine one-family houses, see Fig. 1. Aquifer storage system The low temperature heating system (40oC) will be built around heat-pumps which use groundwater (+5 to 20oC) for heat storage and extraction. The groundwater will be pumped up and injected by means of a two-well system, consisting of one "warm" well and one "cold" well. The system is primarily planned for short-term storage, for example, from sunny days to cloudy days. Experiments will be performed to find out if it is also possible to use the system for long-term storage, i.e. from summer to winter. The system is schematically illustrated in Fig. 2. On sunny days, the heat-pumps use warm air from the arcade to heat the cold groundwater being pumped up from the cold well. The temperature of the groundwater is raised from min +50C to max +20°C. It is then injected through the warm well into the aquifer. As a result, the temperature in the arcade will be kept at a comfortable level. On cloudy days, heated water will be taken from the warm well and used for heating. The whole pumping-system is closed and the groundwater is never in contact with the air. The maximum water flow has been calculated to be 200 m3/day. There is a suitable aquifer in the sedimentary rock about 30 m below the ground surface. The rock is more than 500 m thick and consists of Cretaceous limestone with banks of flintstone. The Quaternary layers covering the aquifer are mainly clay and clayey till. The aquifer is fracturous as well as porous and it is very extensive in area. The experimental siteis located near the sea and the natural hydraulic grandient is very small.

A description is given of the studies and field investigations undertaken to develop the design of caverns to be excavated in hard rock for large central energy storage facilities. Two schemes are considered including a 1000 MW capacity compressed air energy storage (CAES) and a 2000 MW capacity underground pumped hydro (UPH) requiring the excavation of approximately 900,000 yd (690,000 m3) and 8,000,000 yd 3 (6,100,000 m) respectively. Material presented includes the site selection methodology, site investigations, the design approach including construct ion and geotechnical considerations, and the construct ion cost estimates. Geotechnical considerations discussed include the selection of the cavern arrangement and cross sect ion through evaluation of precedent experience, empirical assessment of the rock mass and preliminary strength/stress analysis. INTRODUCTION This paper has been prepared to present the findings of a major study, sponsored and financed jointly by the U.S. Department of Energy and the Electric Power Research Institute (EPRI), into the technical, economic and environmental feasibility of underground energy storage. The program, which was undertaken under the general direction of the Potomac Electric Power Company of Washington, D.C., was concerned with two energy storage concepts, underground pumped hydro (UPH) storage and compressed air energy storage (CAES). Both of these concepts call for the provision of large caverns excavated in rock at depths ranging up to 5000 ft (1500 m) below ground level. Energy storage is required by electric power utilities to allow surplus power available at times of low power demand (usually at night or during the weekend) to be stored for subsequent use during periods of high power demand - usually during weekdays. This function, currently filled by "conventional" surface-located hydro pumped plants, not only promotes the most efficient use of "base load" coal and nuclear plants, but also displaces the high fuel cost gas turbines often used to meet peak loads. Although both the CAES and UPH concepts utilize rock caverns, their basic mode of operation is entirely different. For CAES, surplus electrical energy from the power system is fed to a motor to drive an air compressor; the compressed air is stored in the underground cavern until times of peak demand when it is released and fired with oil fuel to drive a gas turbine and hence generate power, now using the motor as a generator. Typically, the most economic pressures at which to store the air lie between 600 and 1100 psi (4.8 to 7.6 Mpa). At 1000 psi (6.9 Mpa) the volume of air required to provide 10,000 MW hours of storage (1000 MW for 10 hrs) amounts to approximately 900,000 m (690,000m 3 ). For UPH, the operating concept is essentially the same as for "conventional" surface-located pumped storage: to store energy, water is pumped from a lower reservoir, here located at depth below ground surface in a rock cavern, to an upper reservoir located at ground level. To release energy the water is returned to the lower reservoir through a power plant located at or near the lower reservoir.

Preliminary geotechnical investigations are described in view of selecting a site for disposal of radioactive wastes in geological formations. INTRODUCTION The peaceful use of nuclear energy for industrial electricity production has been ruled out in Austria by the popular referendum of November 5, 1978. However, prior to that date extensive work on radioactive waste management and particularly on waste disposal has been carried out in Austria by various organizations. The work reported here refers to projects which were executed by the Austrian Electricity Generating Board. In connection with granting the operational licence for Austria's first 722 MWe boiling water reactor at Zwentendorf the utility has been requested by the licencing authorities to satisfactorily demonstrate, that is to prove, the Capability to close the nuclear fuel cycle. Thus, according to the "polluter pays principle" the Austrian electricity utilities carried out extensive projects for the disposal of radioactive wastes which included both non-site specific as well as site specific geotechnical investigations. GENERAL The principle geotechnical efforts were mainly concentrated on the northern parts of Austria being one of the geologically most suitable areas for radioactive waste disposal-which even preliminary geotechnical surveys had already shown. LANDSAT-satellite imagery of the chosen area has been studied and evaluated. More detailed magnetic measurements were carried out with distances of 50 m between the single stations. The rock itself being granodiorite turned out to be coarse-grained with porphyric texture. Lately, acoustic emission analyses were conducted to investigate rock-mechanic properties in view of establishing thermal load criteria as a consequence of high level waste disposal. SITE INVESTIGATIONS The geophysical site investigations (Holzer, 1979) included amongst others, magnetic measurements, hydrological surveys and a morphological analysis of the environment. The Bohemian Massif in which the detailed site investigations were performed consists dominantly of medium- to high-grade metamorphic suites which in Hercynian times were intruded by widespread granitic to granodiorite plutons. The general seismotectonic situation is such that the central part can be considered as seismically inactive and tectonically stable. According to the utilitie's reprocessing contract with COGEMA, conditioned vitrified high-level waste of about 3–4 m 3 /a depending on the fission product concentration and resulting from the operation of the first nuclear power plant should have eventually been deposited. It was intended to dispose of high-level, medium-level and low-level radioactive waste at one site (Held, 1978). The different emplacement levels are connected by two shafts, A and B. Disposal of the low-level' and medium-level wastes was planned at a depth of about 100 - 130 m, whereas the ∝ -bearing medium level and high-level waste was to be emplaced at a depth of about 500 m. It is evident that the site evaluation for a waste repository is a multivariable problem. The choice of the weights associated with the individual site parameters is one of the most difficult tasks. Moreover, it is also import to realize that besides the safety related aspects for site selection other aspects may also be important such as socio-political, economic and other factors.

Geological characterizations for nuclear waste repositories are without precedent in terms of scope and intended utilization. Furthermore, the time scale of concern is vast in comparison to those with which previous geological characterizations has been concerned. Geological evolution is to be expected during this time scale consequently, evaluation of future changes must be undertaken. The basis for such evaluation can only be obtained from the geological record. Examples of past changes are outlined and an approach to the use of past changes for evaluation of future change is noted. INTRODUCTION The main objective of geologic investigations for nuclear waste repository development is the acquisition and analysis of appropriate data to characterize'' a complex geological/geotechnical system. Input derived from this characterization is required by all activities comprising the process of development of a geologic repository for ultimate disposal. and, permanent isolation of nuclear waste. A geologic nuclear waste repository must satisfy two broad' objectives: isolate the waste from the biosphere for a sufficient time such that release to the biosphere poses- no hazard and, achieve compatibility of the mode of waste storage wi.th any retrieval criteria. ''1''fre, adequacy of the approach adopted to satisfy these objectives must be demonstrated to the general public by means of the licensing procedures. Although these Procedures are not well defined, and in most licensing processes non-technical judgments often prevail, the concept adopted as a guideline for preparation of a licensing application may be to show that, for It given issue: 1. we adequately understand it. 2. We can adequately measure/evaluate it, and 3. We can estimate its long-term effects. The term "adequately" is obviously the key to acceptance of any technical knowledge. Adequate is used to mean a level of understanding, measurement, or evaluation capability sufficient to gain acceptance by the scientific community The characterizations required are without precedent in term's of scope and intended utilization (design and licensing of a repository). Moreover, the time scales which are applicable to repository performance are vast in comparison to those with which previous geological characterizations have been concerned. THE REQUIREMENTS OF GEOLOGIC CHARACTERIZATIONS FOR WASTE REPOSITORY DEVELOPMENT Background Prior to the introduction of nuclear technology, geological characterizations intended for immediate technological consumption were relatively simple. The information provided was typically utilized in mineral/petroleum and civil engineering industries. The characterizations seldom involved a wide spectrum of earth-science disciplines and/or a large amount of effort and funds. The attention was focused exclusively on definition of the geologic status quo. This limited participation of earth-science technology was justified by the relatively minor socio-environmental consequences and social commitments resulting and/or associated with the activities of the sponsoring industries. The emergence of nuclear technology, together with its benefits, potential hazards, and by-products, dramatically changed the sophistication of the geological characterizations required. Input involving an in-depth characterization, from a wide spectrum of earth-science disciplines, became an integral and obligatory part of any application for a license to construct and operate a commercial nuclear facility (usually a nuclear power plant).

Storage of heated water in rock caverns is judged to be an interesting alternative for solving the storage problems connected to large-scale utilization of solar energy and surplus industrial heat. For verifying and demonstrating the technique, a test plant with a storage volume of 15 000 m 3 will be constructed, in which a comprehensive research programme will be carried through. The effects of various time/temperature cycles will be studied. Main items on the research programme are: heat losses, rock stability, heat exchange and material questions, water layer studies, and influence on the environment. INTRODUCTION The use of solar and wind energy as well as surplus or waste heat from industrial processes, power production, sewage water etc is judged to be an important way of reducing the very strong dependency on oil for heating purposes in Sweden. The increasing price on oil and the risk of disturbances of the supply together with the political decision on the limitation of use of nuclear power have lead to very strong efforts in Sweden to develop new techniques for the use of new energy sources and to reduce the need for energy. As the main part of solar and waste heat is normally available only during summer, this technique nessecitates seasonal storage. Extremely inexpensive methods will have to be used to make seasonal storage of low temperature energy economical. For large scale storage there are today few other choices available than storage in water or rock and soil. Since the "energy density" of such ·magazines is low, huge volumes will be needed to cover even a small part of the energy requirement for heating in Sweden. In particular storage of heated water in rock caverns or storage of heat direct in penetrated rock masses are judged to be most interesting due to common occurrence of suitable rock formations in Sweden and the possibilities this entails for optional localization of the storage close to the producer or user of the heat. The storage of heated oil in large rock caverns used in Sweden for many years has shown us that the use of rock caverns offers storage at low costs and that the energy consumption for heating the oil is small compared to the above ground alternatives. It was therefore natural to choose the same concept for the first heat storage facility in Sweden. The conditions for storing heated water in rock caverns have been studied in several R&D-projects in the 1970's. These investigations indicate that a large-scale storage in rock caverns can be advantageous from both technical, economical and environmental points of view compared to storage above ground. So far the technique has not been tested on a large scale. Research and demonstration plant For verifying the results of earlier investigations and for demonstrating the technique under realistic circumstances, full scale tests are necessary. For this purpose a test cavern, equipped with a comprehensive instrumentation for research, is now being designed.

The paper presents a conceptual design for a four unit (850 MWe each) underground CANDU nuclear power station constructed in a series of large caverns excavated in Precambrian rock. Recent results of a geotechnical test program are described and the corresponding in-situ and laboratory measurements are considered in assessing design of rock cavern support systems, cavern liner requirements, and rock cavern stability under both thermal and seismic loads. The overall safety of the underground station design is also discussed. In this context the predictions of a preliminary analysis of the station accident containment capability suggest that post-accident temperatures and pressures in the rock caverns are below acceptable levels. Operational concerns for the underground CANDU design are also briefly considered. Finally, preliminary station capital cost estimates, based on a construction schedule extended by sixteen months, reveal a penalty of 31–36% relative to an equivalent surface facility. INTRODUCTION Ontario Hydro, Canada's largest power utility with a total installed capacity of 24,000 MWe of which roughly 22% is nuclear generation, has recently completed a two-year investigation of the feasibility, both technical and economic, of constructing an underground CANDU nuclear power station (Oberth and coworkers, 1979). CANDU is a unique Canadian pressure tube reactor design using natural uranium fuel and heavy water as both the coolant and moderator. The investigation, performed by in-house engineering staff with the aid of geotechnical consultants, concentrated on deep rock siting in which reactor units are constructed in caverns excavated in Precambrian rock. This paper presents a conceptual underground design for the rock-sited 4 × 850 MWe CANDU power station, and evaluates the safety, operational, and economic implications of the design. Previous investigations of underground siting in Europe (Lindbo, 1978; Kröger and coworkers, 1976) and in the USA (Allensworth and coworkers, 1977; California Energy commission, 1978) revealed possible benefits such as improved containment, increased station protection and security, and significant reductions in seismic ground motion experienced in underground rock caverns. This paper examines these potential benefits in the context of the CANDU reactor design and Ontario's unique geological situation. The Ontario Hydro study examined two methods of constructing an underground nuclear power station: rock cavern excavation and cut-and-cover burial. Each method has its own set of siting constraints, technical and economic implications, and environmental and safety factors. This paper concentrates on the first of these two underground siting options in which the reactor components are erected within large underground rock caverns with the rock face serving as both the reactor building superstructure and the containment boundary. GEOTECHNICAL INVESTIGATION PROGRAM A number of areas in Ontario were examined for local geological conditions and other general siting considerations. While no specific site was selected for the underground nuclear station design study, the economic benefits of siting close to a large urban load centre provided an incentive to focus on areas near Toronto. These economic benefits include reduced transmission costs and the potential use of waste heat for district heating purposes.

A cross-hole high-frequency acoustic investigation of a granitic rock mass subjected to sustained heating is reported. Compressional and shear-wave velocity measurements along four different paths between four vertical boreholes made prior to turning on the heater, during 398 days of heating and after the heater was turned off correlated well with the presence of fracture zones, in which the fractures were closed by thermal expansion of the rock upon heating. When the rock mass cooled, the velocity measurements indicated a greater intensity of fracturing than had existed prior to heating. Laboratory compressional and shear-wave velocity measurements have been made on intact rock specimens obtained from the site and subjected to axial stress. When used to interpret the increases in velocities measured in the field upon heating the rock mass, increases in horizontal normal stresses to between 30 and 40 MPa were inferred. Increases of these magnitudes agree with stress measurements made by the other techniques. The ratio of measured compressional to shear-wave velocity appears to provide a sensitive measure of the fraction of crack porosity containing water or gas. INTRODUCTION One of the more promising methods developed in the past few years for geotechnical site investigation and the characterization of rock masses is the higher-frequency acoustic wave technique. The high frequencies employed permit detection of discontinuities and the outlining of zones having different physical properties between boreholes or behind surface boundaries in much more detail than the conventional low-frequency seismic methods. Price, Malone and Knill (1970), McCann, Grainger and McCann (1975) and Auld (1977) describe the use of acoustic measurements between boreholes for geotechnical purposes. Price and colleagues employed the results of their study to determine the optimum rock-bolt pattern to stabilize a rock mass. McCann and colleagues used the between-hole technique to delineate interfaces between homogeneous media, to detect localized, irregular features and to estimate the degree of fracturing in the rock mass. Auld used between-hole acoustic measurements to determine the elastic properties of the rock mass. Acoustic techniques employed within a borehole have been described by Geyer and Myung (1971), Myung and Baltosser (1972) and by King and colleagues (1975, 1978). The application of acoustic borehole logs in detecting fractures, for rock classification and in determining the in situ elastic properties of rock have been discussed by these workers and by Carroll (1966, 1969) and Coon and Merritt (1970). In this paper are described the results of a research project involving cross-hole acoustic measurements in a fractured granite rock mass subjected to thermal stresses. The acoustic research project is itself part of a comprehensive rock mechanics and geophysics research programme associated with large-scale heater tests in an abandoned iron-ore mine in centralSweden, as described by Witherspoon and colleagues (1979).