A suite of laboratory measurements are being conducted on Geysers graywacke recovered from a drilled depth of 2599 meters in NEGU-17. The tests are being conducted to characterize the effect of pressure and fluid saturation on the seismic properties of the graywacke matrix. The measurements indicate that the graywacke is an unusual rock in many respects. Both compressional and shear velocities exhibit relatively little change with pressure. Water saturation causes a slight increase in the compressional velocity, quantitatively consistent with predictions from the Biot-Gassmann equations. Shear velocity decreases with water saturation by an amount greater than that predicted by the Biot-Gassmann equations. This decrease is attributed to chemomechanical weakening caused by the presence of water. Measurements of Q, from torsion experiments on room dry samples at seismic frequencies indicate unusually high Q, (~500). Water saturation decreases the shear modulus by 12 percent, again indicative of chemomechanical weakening. Q, is lower for the water saturated condition, but still relatively high for rock at low stress. Results of ultrasonic pulse propagation experiments on partially saturated samples are typical of low porosity rocks, being characterized by a monotonic decrease in compressional and shear velocity with decrease in saturation. An increase in shear velocity …

A large number of microearthquake seismograms have been recorded by a downhole, three-component seismic network deployed around the Coso, California geothermal reservoir. Shear-wave splitting induced by the alignment of cracks in the reservoir has been widely observed in the recordings. Over 100 events with body wave magnitude greater than 1.0 from microearthquakes recorded since March of 1992. have been processed. The results show that most events with paths within the critical angle that defines the shear-wave window, display clear shear-wave splitting, and the fast shear waves have predominant polarization directions for most stations. The rose diagrams of fast shear-wave particle motion suggest that there are three predominant fracture strikes (or directions of maximum horizontal stresses) in the Coso geothermal field: N 40°-60°E, N 0°-25°E, and N 25° - 35° W, which are consistent with photographically or magnetically mapped alignments on the surface. From the delay time of split shear waves, we estimate that the crack density in the most active geothermal reservoir area (above 3.00km depth) ranges between 0.030 and 0.055. values commonly found in other hydrocarbon or geothermal reservoirs.

A computer program has been developed to analyze multipass pressure-temperature-spinner surveys and summarize the data in graphical form on two plots: (1) an overlay of spinner passes along with a fluid velocity profile calculated from the spinner and (2) an overlay of pressure, pressure gradient, and temperature profiles from each pass. The program has been written using SmartWare II Software. Fluid velocity is calculated for each data point using a cross-plot of tool speed and spinner counts to account for changing flow conditions in the wellbore. The program has been used successfully to analyze spinner surveys run in geothermal wells with two-phase flashing flow.

The data from the chemical analysis of 68 samples from the 100-HR-3 Operable Unit Third Quarter 1993 Groundwater Sampling Investigation and their related quality assurance samples were reviewed and validated to verify that reported sample results were of sufficient quality to support decisions regarding remedial actions performed at the site. Sample analysis included inorganics and general chemical parameters. Fifty three samples were validated for radiochemical parameters.

The Fernald Environmental Management Project (FEMP), formerly the Feed Materials Production Center (FMPC), is a Department of Energy (DOE) site which produced high-quality uranium for military defense beginning in 1951. Production at the FEMP was halted in July 1989. Later that year, the facility was placed on the National Priorities List (NPL). The DOE is currently conducting a Remedial Investigation/Feasibility Study (RI/FS) and other response actions under the Amended Consent Agreement between the US Environmental Protection Agency (USEPA) and the DOE.

Variations in temperature and salinity in hypersaline liquid-dominated geothermal systems like the Salton Sea Geothermal System (SSGS) tend to be correlated such that liquid density is relatively constant in the system. The tendency toward small density variations may be due to connectivity with a surrounding regional aquifer at multiple depths in the stratigraphic column. We present numerical simulation results for natural convection in geothermal systems like the SSGS in hydraulic connection with a constant-density aquifer. Natural convection where there are two sources of buoyancy such as heat and salt, with different diffusivities, is called double-diffusive convection. Simulations of double-diffusive convection are carried out using our general-purpose reservoir simulator TOUGH2 with a newly developed twodimensional heat and brine transport module (T2DM) that includes Fickian solute dispersion. The model includes an accurate formulation for liquid density as a function of temperature and salinity. Our simulation results show many features that are consistent with observations of the SSGS, making conceptual models that involve hydraulic connectivity with a surrounding aqulfer appear plausible. The generality of our model makes the results broadly applicable to systems similar to the SSGS.

The effects of an adsorbing phase on the injection of liquid and eventual production of vapor from a low-porosity, vapor dominated geothermal reservoir was studied. The magnitude of delay caused by adsorption, diffusion partitioning, preferential partitioning, and permeability variation were compared. Results were then compared to measured tracer production data at the Geysers to determine the most likely delay mechanism for injected tracer at the Geysers. A one-dimensional numerical model describing vapor flow in a porous medium in the presence of a sorbing phase was used to investigate the delay of injected tracer caused solely by the sorbing phase. An analytical model was used to describe delay effects due to diffusion partitioning of tracer from the vapor phase into the liquid phase. Properties of steam and tracer used in Geysers tracer studies were compared to determine the effects of preferential partitioning. Finally, a streamline model of a tracer study was used to determine the magnitude of permeability delays possible using permeability values measured at the Geysers. It was concluded that adsorption alone has very little effect on the delay of injected tracer indicating that little recharge of the adsorbed mass occurs for a typical injection program at the Geysers. Diffusion …

Water adsorption in geothermal reservoir materials was investigated by transient flow technique using steam and COz gas. Theoretical and experimental results indicate that water adsorption exists in vapordominated type of reservoir, but experiments in the past have been limited to pure gases. The common presence of CO2, a non-condensible gas, in a geothermal reservoir necessitated a study of the effect of partial CO2 concentration on adsorption. Experimental laboratory work using a crushed Geysers rock sample at low pressure was carried out. Transient pressure exerted by steam pressure inside the sample was measured against time during a desorption process. It was found that the partial presence of CO2 did not significantly affect the adsorption of water.

Determination of true formation temperature from measured bottom hole temperature is important for geothermal reservoir evaluation after completion of well drilling. For estimation of equilibrium formation temperature, we studied non-linear least squares fitting method adapting the Middleton Model (Chiba et al., 1988). It was pointed out that this method was applicable as simple and relatively reliable method for estimation of the equilibrium formation temperature after drilling. As a next step, we are studying the estimation of equilibrium formation temperature from bottom hole temperature data measured by MWD (measurement while drilling system). In this study, we have evaluated availability of nonlinear least squares fitting method adapting curve fitting method and the numerical simulator (GEOTEMP2) for estimation of the equilibrium formation temperature while drilling.

Exploitation of a liquid-dominated geothermal system generally results in a transfer of mass that causes measurable changes in gravity. When the rate of mass transfer is controlled by the permeability of the reservoir rocks then analysis of measured gravity changes, using numerical reservoir simulation models, can yield values for reservoir properties. One such case is during the early stages of exploitation, during the formation and expansion of a 2-phase zone. Calculations using MULKOM models show that for Wairakei field the gravity changes associated with permeabilities of 50 and 100 md would be clearly distinguishable (> 50 microgal) in less than 2 years. A measured gravity change of -415 microgal between 1950 and 1961 suggests a permeability of 100 md for rocks in the upper part of the 2-phase zone. This value is consistent with those obtained from well tests.

According to analyses by the Department of Energy’s Energy Information Administration (EIA), geothermal electric power capacity could nearly quadruple over the next 20 years, and there is a tremendous potential for growth in the direct uses of geothermal energy. However, for a high rate of development to occur in either of these applications, the identified resource base must be expanded. To this end, the Department is supporting R&D efforts to 1) share with industry the costs and risks of evaluating promising new resource prospects with power potential; 2) reduce the costs of exploration to enhance industry’s cost-competitive posture; and 3) assess the location and characteristics of low-temperature resources. This paper describes DOE’s new cost-shared industry-coupled exploratory drilling program to be initiated with a solicitation by the Idaho National Engineering Laboratory, field manager of DOE’s reservoir technology activities. Proposals will be requested for drilling either core holes or full-size wells on prospects from which some information had already been gathered, such as surface geophysics or shallow heat flow. The paper also discusses the status of the project designed to demonstrate whether a geothermal reservoir can be identified and adequately evaluated to meet investment requirements with slimholes rather than the much more …

The Thelamork low-temperature geothermal system in N-Iceland has for the last decade been considered as a possible source of hot water for the Akureyri District Heating Service. A productive well was drilled in the summer of 1992 afer 10 years of geothermal research in the area Following that a feasibility study was performed in order to determine whether harnessing the geothermal gwtem for space heating would be economical. This study consisted of a nine month full scale production test along with partial reinjection and tracer tests. It also involved careful monitoring of production rates, water level changes and chemistry. Finally, the data collected were analyzed on the basis of simple reservoir models. The results of the analysis indicate that the system will sustain a production of 19-20 I/s, initially at 91 degrees C, for the next 10 years, given that 3 l/s will be reinjected. However, a cooling of 9-12 "C is predicted due to infiltration of colder groundwater and the reinjection The results also suggest that hamessing the geothermal system will be economical, despite the high cost of exploration and an 11 km insulated pipeline to Akureyri.

Quantitative flow modeling of fluids at elevated temperatures and pressures has generally been limited to consideration of either single-phase flow or two-phase flow at conditions below the critical point of water. In this paper, we introduce a version of the GEOTHER model that can simulate both multiphase flow and flow above the critical point, and demonstrate its capabilities by simulating flow in the vicinity of the critical point. GEOTHER2 is a multiphase, finite-difference model that simulates three-dimensional flow of pure water and heat at temperatures ranging from 0{degree} to 1,200{degree}C and pressures ranging from 0.5 to 10,000 bars. The governing equations are expressions of mass and energy conservation that are posed in terms of pressure and enthalpy. A series of one-dimensional experiments indicates that permeability is a pivotal factor in determining pressure-enthalpy/temperature trajectories near the critical point. At low permeabilities (<10{sup -18} m{sup 2}), heat transport by conduction dominates, and the trajectoq defines a constant thermal gradient. At higher permeabilities (>10{sup -16} m{sup 2}), advective heat transport dominates, and the pressure-enthalpy trajectory maintains a constant "flowing enthalpy".

The reservoir of the West Olkaria Geothermal Field is hosted within tuffs and the reservoir fluid is characterized by higher concentrations of reservoir CO{sub 2} (10,000-100,000 mg/kg) but lower chloride concentrations of about 200 mg/kg than the East and North East Fields. The West Field is in the outflow and main recharge area of the Olkaria geothermal system. Permeability is generally low in the West Field and its distribution is strongly controlled by the structures. Fault zones show higher permeability with wells drilled within the structures havin larger total mass outputs. However, N-S and NW-SE faults are mainly channels for cold water downflow into the reservoir. Well feeder zones occur mostly at lava-tuff contacts; within fractured lava flows and at the contacts of intrusives and host rocks.

A long-term flow test was carried out in the Fenton Hill HDR Phase-2 reservoir for 14 months during 1992-1993 to examine the potential for supplying thermal energy at a sustained rate as a commercial demonstration of HDR technology. The test was accomplished in several segments with changes in mean flowrate due to pumping conditions. Re-test estimates of the extractable heat content above a minimum useful temperature were based on physical evidence of the size of the Fenton Hill reservoir. A numerical model was used to estimate the extent of heat extracted during the individual flow segments from the database of measured production data during the test. For a reservoir volume of 6.5x10{sup 6}m{sup 3}, the total heat content above a minimum temperature of 150{degree} C was 1.5x10{sup 15}J. For the total test period at the three sustained mean flowrates, the integrated heat extracted was 0.088x10{sup 15}J, with no discernable temperature decline of the produced fluid. The fraction of energy extracted above the abandonment temperature was 5.9%. On the basis of a constant thermal energy extraction rate, the lifetime of the reservoir (without reservoir growth) to the abandonment temperature would be 13.3 years, in good agreement with the pre-test estimate of …

Disposal of hot separated brine by means of reinjection within the limits of the geothermal reservoir is, at present, a problem that remains to be solved. Possible thermal, as well as chemical contamination of the resources present key questions that have to be appropiately answered before a reinjection project is actually implemented in the field. This paper focusses on the basic heat-transfer process that takes place when a relatively cold brine is injected back into the naturally fractured hot geothermal reservoir after steam has been separated at the surface. The mathematical description of this process considers that rock matrix blocks behaves as uniformly distributed heat sources, meanwhile heat transfer between matrix blocks and the fluid contained in the fractures takes place under pseudo-steady state conditions with the main temperature drop occurring in the rock-matrix blocks interphase. Analytical solutions describing the thermal front speed of propagation are presented. Discussion on the effect of several variables affecting the thermal front speed of propagation is included, stressing the importance that a proper “in-situ” determination of the effective heat transfer area at the rock-fluid interphase has on the whole process. Solutions are also presented as a type-curve that can be practically used to estimate …

A study of the phase change phenomena in porous media using Christiansen filters continues in order to determine the experimental conditions most favorable to the use of this method. A calibration setup has been finished. Determination of the wavelength corresponding to the equality of the refractive indices varies with temperature is going to be carried out later. The dispersion curves of the solid and liquid phases constituting a transparent saturated porous medium generally have different slopes. It is thus impossible to achieve the equality of the refractive indices for all wavelengths simultaneously. For a given temperature, the dispersion curves intersect at a point corresponding to a single wavelength. From our investigation, we have found that we need to change the liquid phase material, ethyl salicylate (HOC{sub 6}H{sub 4}COOC{sub 2}H{sub 5}), that we have proposed before. This is because the boiling point of ethyl salicylate is too high for our purposes (about 233{degrees}C). Therefore, it is not a suitable material to do the phase change study in the Christiansen filters. A suitable liquid phase organic chemical material for our research must fit the criteria given in this paper.

Steam production data from wells surrounding Unit 13 injection well CA 956A-1 and Unit 16 injection well Barrows-1 were analyzed to estimate annual and cumulative recovery factors due to water injection into these wells. Production and injection data from SMUDGEO#1 and Bear Canyon wellfields were also analyzed to obtain recovery of the injected water in those wellfields. The results of this study may be useful in designing future injection projects in vapor dominated systems.

This paper presents a solution for the inverse problem to the flow of tracers in naturally fractured reservoirs. The models considered include linear flow in vertical fractures, radial flow in horizontal fractures, and cubic block matrix-fracture geometry. The Rosenbrock method for nonlinear regression used in this study, allowed the estimation of up to six parameters for the cubic block matrix fracture geometry. The nonlinear regression for the three cases was carefully tested against syntetical tracer concentration responses affected by random noise, with the objective of simulating as close as possible step injection field data. Results were obtained within 95 percent confidence limits. The sensitivity of the inverse problem solution on the main parameters that describe this flow problem was investigated. The main features of the nonlinear regression program used in this study are also discussed. The procedure of this study can be applied to interpret tracer tests in naturally fractured reservoirs, allowing the estimation of fracture and matrix parameters of practical interest (longitudinal fracture dispersivity alpha, matrix porosity phi2, fracture half-width w, matrix block size d, matrix diffusion coefficient D2 and the adsorption constant kd). The methodology of this work offers a practical alternative for tracer flow tests interpretation to …

TOUGH2, Lawrence Berkeley Laboratory's general purpose simulator for mass and heat flow and transport was enhanced with the addition of a set of preconditioned conjugate gradient solvers and ported to a PC. The code was applied to a number of large 3-D geothermal reservoir problems with up to 10,000 grid blocks. Four test problems were investigated. The first two involved a single-phase liquid system, and a two-phase system with regular Cartesian grids. The last two involved a two-phase field problem with irregular gridding with production from and injection into a single porosity reservoir, and a fractured reservoir. The code modifications to TOUGH2 and its setup in the PC environment are described. Algorithms suitable for solving large matrices that are generally non-symmetric and non-positive definite are reviewed. Computational work per time step and CPU time requirements are reported as function of problem size. The excessive execution time and storage requirements of the direct solver in TOUGH2 limits the size of manageable 3-D reservoir problems to a few hundred grid blocks. The conjugate gradient solvers significantly reduced the execution time and storage requirements making possible the execution of considerably larger problems (10,000+ grid blocks). It is concluded that the current PCs provide …

The behavior of water injection plumes in vapor-dominated reservoirs is examined. Stressing the similarity to water infiltration in heterogeneous soils, we suggest that everpresent heterogeneities in individual fractures and fracture networks will cause a lateral broadening of descending injection plumes. The process of lateral spreading of liquid phase is viewed in analogy to transverse dispersion in miscible displacement. To account for the postulated “phase dispersion” the conventional two-phase immiscible flow theory is extended by adding a Fickian-type dispersive term. The validity of the proposed phase dispersion model is explored by means of simulations with detailed resolution of small-scale heterogeneity. We also present an illustrative application to injection into a depleted vapor zone. It is concluded that phase dispersion effects will broaden descending injection plumes, with important consequences for pressure support and potential water breakthrough at neighboring production wells.

Compositions of coexisting liquid and vapor phases have been determined at temperatures from 250 to 350°C for brines containing NaCl and either HCI or NaOH by direct sampling of both phases from a static phase-equilibration apparatus. In these experiments, NaCl concentrations in the liquid phase ranged to 6.5 mol-kg{sup -1}, with corresponding vapor-phase NaCl concentrations varying strongly with temperature and brine composition. Acid or base was added to the brines to suppress unknown contributions of NaCl(aq) hydrolysis products to the observed volatilities. Thermodynamic partitioning constants for NaCl have been determined from the observed compositions of the coexisting phases combined with the known activity coefficients of NaCl(aq) in the liquid phase. An apparent dependence of the values of these partitioning constants on brine concentration is explained by considering the effect of decreasing pressure on the density of the vapor phase. Concentrations of HCI and NaCl in steam produced from various natural brines may be calculated as hnctions of temperature and brine composition based on these new results coupled with our previous determinations of the partitioning constants for HCl(aq). Application of these results to The Geysers will be discussed in terms of the composition of postulated brines which could be in equilibrium …

Economic optimisation of the Ohaaki Geothermal Field dual-flash system indicated the requirement to program for sliding High Pressure turbine inlet pressures and the de-rating of individual wells to Intermediate Pressure. A wellbore simulator was used to generate output curves up to 5 years into the future to enable 'what-if' modelling for maximum electrical generation under different scenarios. The key to predicting future output curves as a function of wellhead pressure was predicting two-phase well productivities as a function of field pressure and enthalpy trends. Using a wellbore simulator to generate inflow pressure curves from output test data and matching measured downhole data showed that the Duns and Ros flow correlation produced a linear response with a consistent relationship to static pressures for most wells. This was used to generate predicted output characteristic curves up to 1998, enabling the modelling of varying turbine inlet pressures.

Injectivity tests in wells with two permeable zones and internal flow is analyzed in order to include the usually severe thermal transient effects. A theoretical analysis is performed and a method devised to obtain information from the thermal transient, provided that temperature is measured simultaneously with pressure. The technique is illustrated with two real tests performed at Miravalles, Costa Rica. It allows to estimate total injectivity index as well as the injectivity index of each one of the two zones separately. Correct position of measuring tools and nature of spontaneous internal flow is also discussed.