Search

Calculations discussed in the Iceland Deep Drilling Project feasibility study in 2003 indicated that, for same volumetric flow rate of steam, a geothermal well producing from natural supercritical fluid would have the potential to generate power outputs an order of magnitude greater than from conventional high-temperature wells (240–340 °C). To reach supercritical hydrous fluid conditions in natural geothermal systems requires deep drilling to a minimum depth of some 3.5–5 km were temperature conditions can be expected to range between 400 and 600 °C in reasonably active high-temperature fields. Three geothermal fields in Iceland, Reykjanes, Hengill and Krafla, were selected as suitable locations for deep drilling to test this concept in search of natural supercritical geothermal fluid systems.

Drilling deeper in high-temperature geothermal systems by the IDDP is aimed at increasing the power output of shallower high-temperature geothermal fields by an order of magnitude without increasing their environmental footprints. The main thrust of the IDDP is to develop deep supercritical systems, but an unexpected encounter with a shallow body of magma demonstrated that very high power outputs are also possible from the contact zone of an intrusion. In the future it may be feasible to produce energy directly from magma. Favorable environments to implement these concepts are likely worldwide wherever active volcanoes and young volcanic rocks occur.

Preparation has begun for drilling the second deep IDDP well into the saline Reykjanes high-temperature field in SW-Iceland. The site selection for the IDDP-2 drillhole is under review and the prime candidate is essentially the same as the 1st priority site suggested for the Reykjanes field in 2003. More recent drillhole data and new MT surveys have amplified the justification for selecting that site. Deep drilling to 4–5 km depth is an important part of the HS Orka exploration strategy for enhanced power production, either by direct use of high energy steam, or by attempting to enhance the field performance by re-injecting geothermal fluid deep into very hot rocks. Pending on several decisions and development in Iceland, outside the control of the IDDP energy consortium, the IDDP-2 well might possibly be drilled to 4–5 km depth as early as 2014.

The Iceland Deep Drilling Project (IDDP) is a long-term program to improve the economics of geothermal energy by producing supercritical hydrous fluids from drillable depths. Producing supercritical fluids will require the drilling of wells and the sampling of fluids and rocks to depths of 3.5–5 km, and at temperatures of 450–600 °C. The IDDP plans to drill and test a series of such deep boreholes in the Krafla, Nesjavellir and Reykjanes geothermal fields in Iceland. Beneath these three developed high-temperature systems frequent seismic activity continues below 5 km, indicating that, even at supercritical temperatures, the rocks are brittle and therefore likely to be permeable, even where the temperature is assumed to exceed 550–650 °C. Temperature gradients are greater and fluid salinities smaller at Nesjavellir and Krafla than at Reykjanes. However, an active drilling program is underway at Reykjanes to expand the existing generating capacity and the field operator has offered to make available one of a number of 2.5 km deep wells to be the first to be deepened to 5 km by the IDDP. In addition to its potential economic significance, drilling deep at this location, on the landward extension of the Mid-Atlantic Ridge, is of great interest to the international science community. This paper examines the prospect of producing geothermal fluids from deep wells drilled into a reservoir at supercritical temperatures and pressures. Since fluids drawn from a depth of 4000–5000 m may prove to be chemically hostile, the wellbore and casing must be protected while the fluid properties are being evaluated. This will be achieved by extracting the fluids through a narrow retrievable liner called the “pipe”. Modelling indicates that if the wellhead enthalpy is to exceed that of conventionally produced geothermal steam, the reservoir temperature must be higher than 450 °C. A deep well producing 0.67 m 3 /s steam (∼2400 m 3 /h) from a reservoir with a temperature significantly above 450 °C could, under favourable conditions, yield enough high-enthalpy steam to generate 40–50 MW of electric power. This exceeds by an order of magnitude the power typically obtained from a conventional geothermal well in Iceland. The aim of the IDDP is to determine whether utilization of heat from such an unconventional geothermal resource at supercritical conditions will lead to increased productivity of wells at a competitive cost. If the IDDP is an economic success, this same approach could be applied in other high-temperature volcanic geothermal systems elsewhere, an important step in enhancing the geothermal industry worldwide.

Our detailed three-dimensional model of the natural flow regime of the Cerro Prieto geothermal field, before steam production began, is based on patterns of hydrothermal mineral zones and light stable isotopic ratios observed in rock samples from more than 50 deep wells, together with temperature gradients, wireline logs and other data. At the level so far penetrated by drilling, this hydrothermal system was heated by a thermal plume of water close to boiling, inclined at 45°, rising from the northeast and discharging to the west. To the east a zone of cold water recharge overlies the inclined thermal plume. Fission track annealing studies show that the reservoir reached 170°C only 10 4 years ago. Oxygen isotope exchange data indicate that a 12 km 3 volume of rock subsequently reacted with three times its volume of water hotter than 200°C. Averaged over the duration of the heating event this would require a flow velocity through a typical cross-section of the reservoir of about 6 m/year. The heat in storage in that part of the reservoir hotter than 200°C and shallower than 3 km depth is equivalent to that which would be released by the cooling of about 1 or 2 km 3 of basalt or gabbro magma. Although this is an extensional tectonic environment of ‘leaky’ transform faulting in which repeated intrusions of basalt magma are likely, for simplicity of computation we have modelled possible heat sources as simple two dimensional basalt intrusions of various sizes, shapes and locations. We have calculated a series of two-dimensional convective heat transfer models, with different heat sources and permeability distributions. The models which produce the best fit for the temperature distributions observed in the field today have in common a heat source which is a funnel-shaped basalt intrusion, 4 km wide at the top, emplaced at a depth of 5–6 km about 40,000–50,000 years ago. Although such forward modelling does not provide a unique solution of the thermal history, it does suggest that the igneous intrusion which supplied the heat to the Cerro Prieto system must be young, large and close. Modelling the hydrothermal system is useful in guiding geothermal resource evaluation; in turn we hope that modelling of the magmatic system will stimulate the search for the parent magma chamber by geophysical means, and eventually by deep drilling.

Vitrinite reflectance measured in immersion oil ( R o ) on kerogen extracted from hydrothermally altered mudstones in borehole M-84 at the Cerro Prieto geothermal field exhibit an increase in mean reflectance ( R o ) from 0.12 per cent at 0.24 km depth to 4.1 per cent at 1.7 km depth. Downhole temperatures measured over this interval increase from about 60° to 340°C. These R o data plotted against temperature fall along an exponential curve with a coefficient of determination of about 0.8. Other boreholes sampled in the field show similar relationships. A regression curve calculated for temperature and R o in borehole M-105 correctly predicts temperatures in other boreholes within the central portion of the geothermal system. The correlation between the reflectance values and logged temperature, together with consistent temperature estimates from fluid inclusion and oxygen isotope geothermometry, indicates that changes in R o are an accurate and sensitive recorder of the maximum temperature attained. Therefore, vitrinite reflectance can be used in this geothermal system to predict the undisturbed temperature in a geothermal borehole during drilling before it regains thermal equilibrium. Although existing theoretical functions which relate R o to temperature and duration of heating are inaccurate, empirical temperature- R o curves are still useful for geothermometry. A comparison of temperature- R o regression curves derived from nine boreholes within the Cerro Prieto system suggests that heating across the central portion of the field occurred penecontemporaneously, but varies near margins. Boreholes M-93 and M-94 appear to have cooled from their maximum temperatures, whereas M-3 and Prian-1 have only recently been heated. Comparison of the temperature- R o data from the Salton Sea, California, geothermal system indicates that the duration of heating has been longer there than at the Cerro Prieto field.

Rock samples from a total of 50 drillholes in the Cerro Prieto geothermal area have been analysed by stable isotopic techniques. In particular, calcite in cuttings and core samples from these wells has been analysed to observe systematic variations in carbon (δ 13C) and oxygen (δ 18O) isotopic compositions. Consistent behaviours of these isotopic systems are observed and provide a large amount of information on the processes of calcite-water isotopic equilibration, dissolution-precipitation and decarbonation occurring within and around the reservoir. Oxygen isotopic profiles of intergranular calcites in sandstones appear to be a reliable measure of the equilibrium temperature distribution in the field before production began. From these data, a detailed three-dimensional map has been developed, showing the equilibrium temperatures in the geothermal field. Of particular importance are many examples of temperature reversals at depth beneath the hot geothermal reservoir. Reversals of this sort provide strong constraints on the directions of fluid flow and the overall dynamics of the circulation system. Long-term natural flow directions within the geothermal system can often be inferred directly from the three-dimensional distribution of temperatures within the field. In addition, localized areas of inferred natural cold water inflow (recharge) or hot water outflow (discharge) can be observed by their effects on the calcite content of the rock and on the behaviour of carbon isotopic compositions of the calcite. Within the geothermal reservoir, a gradual disappearance of calcite, with increasing temperature, correlates generally with a slight decrease in δ13C of the calcite. Although this pattern is nearly identical to the isotopic changes occurring during decarbonation of calcite in skarn areas, it is probable that other processes involving dissolution by the brines or isotopic exchange with an external carbon reservoir are also locally important. A mass balance calculation has been performed using measured 18O depletion in subsurface whole-rock samples from Cerro Prieto wells and the inferred 18O enrichment of the geothermal brine. This calculation implies an overall water-rock volume ratio of approximately 3:1 during the history of the Cerro Prieto system.

Our ongoing studies of water-rock interaction at Cerro Prieto have now been extended to include samples from 40 wells. We have confirmed the regular sequence of progressive hydrothermal alteration zones previously described, and have mapped these alteration zones across the geothermal field. Our earlier work showed the relationships between hydrothermal mineralogy, temperature and permeability, in that alteration occurs at lower temperature in sandstone than in the less permeable shales. The effects of chemical parameters such as silica activity and differences in CO 2 fugacity have also become apparent when mineral assemblages are compared in sandstones from different wells at the same temperature. A rather complete picture of the shape of the reservoir and the nature of its boundaries is developing, and we have begun to identify patterns in the observed hydrothermal mineral zonations which are characteristic of different temperature gradients. We infer such different gradients to be indicative of different parts of the hydrothermal flow regime. In certain wells mineral zones are closely spaced, indicating steep thermal gradients, while in others they are much farther apart. We believe that patterns characteristic of recharge, discharge and upwelling zones as well as areas of primarily horizontal flow can be recognized. The geothermal circulation system at Cerro Prieto appears to be rather young and shows no indication of retrograde reactions due to cooling. The pattern of fluid flow does not seem to be significantly affected by faults, stratigraphic horizons or by the presence of a cap-rock. The mineral zones define a thermal dome which is apparently fed from the east and spreads westward.