In the OSTI Collections: Geothermal Energy

Dr. Watson computer sleuthing scientist.

Article Acknowledgement:

Dr. William N. Watson, Physicist

DOE Office of Scientific and Technical Information

View Past "In the OSTI Collections" Articles











During 2013, the combined power input to all of our transportational, industrial, commercial, and domestic machinery averaged about 24 trillion watts.[IEA]  For a long time now, the primary fuels for all of this have been coal, oil, and natural gas, with most of the other power coming from biofuels, nuclear fuels, the flow of rivers, wind, heat from the earth’s interior, and sunlight.  These energy resources differ in size, potential output rate, and accessibility, with little correlation among these parameters.  For instance, the total power of sunlight reaching the earth’s surface, and the worldwide heat outflow from the earth’s interior, are 29,000 trillion watts and 47 trillion watts, respectively.[Wikipedia]  The geothermal heat flux alone would be more than enough to run all our present machinery, and the solar influx would be more than 1,000 times enough, but turning either form of energy into some other form requires a suitable conversion device. 


Solar power can be converted directly into electricity by photovoltaic devices, or it can be converted into mechanical work by concentrating it onto the high-temperature input of a heat engine.  The latter method’s maximum possible efficiency is fundamentally limited by the ambient temperature’s ratio to the engine’s input temperature, which equals the minimum possible fraction of heat that the engine won’t convert into useful work; an input temperature no higher above absolute zero[Wikipedia] than the temperature of the surrounding air would leave 100% of the solar energy unconverted.  Likewise, heat engines can also convert geothermal energy into work if the geothermal heat flux reaches the engine at a higher temperature than the ambient temperature at the earth’s surface. 


The fact that rock deep underground is much hotter than rock at the earth’s surface provides a potentially useful temperature difference.  If water can permeate this rock, the rock can heat it; if the heated water can circulate through the rock to a heat engine at the surface, the engine will have the higher-temperature input it needs to turn some of the heat into work.  The big “if” here is whether the rock is permeable enough for the water to transport enough energy to make the conversion worthwhile.  As one report to a workshop on geothermal reservoir engineering states,


“The primary goal in geothermal exploration is to find a location with high temperatures and sufficient flow rate to allow for economic recovery of the heat.  While elevated temperatures in the sub-surface are not uncommon, the high permeability and flow rate required for economic prospect is more rare.”—“Estimating Subsurface Permeability with 3D Seismic Attributes:  A Neural Net Approach”[SciTech Connect], p. 1.


The same point is made by a set of slides from the National Renewable Energy Laboratory, “Design Requirements for Commercial Sedimentary Geothermal Projects”[SciTech Connect].  These slides also point out that people in the petroleum industry have relevant expertise for finding, developing, and improving the performance of sites from which geothermal energy can be economically extracted. 


Various ways to characterize any given site’s geothermal-power potential are described in numerous reports of investigations sponsored by the Department of Energy.  Other reports describe investigations of particular sites’ potential, or show how geothermal energy has actually been extracted in specific projects. 




Using seismic data


One obvious way to estimate how easily water can flow through a rock formation is to examine the formation by drilling through it.  However, as the aforementioned report “Estimating Subsurface Permeability with 3D Seismic Attributes:  A Neural Net Approach”notes, this method is problematic.   


“Unfortunately, drilling is expensive and time-consuming, and often several wells may be necessary to adequately delineate a field.  Therefore, finding a cost-effective method of estimating permeability at depth using geophysics would reduce costs of exploration and reduce risk, leading to lower costs of produced power and additional prospects.”—P. 1, op. cit.[SciTech Connect] 


The paper goes on to discuss researchers’ efforts to infer the structure of subsurface rock from information that includes the way the rock reflects artificially-generated seismic waves[Wikipedia]—basically sound waves in which the rock elastically vibrates back and forth or side to side as the wave itself moves forward[Wikipedia; Wikipedia].  Data from the Walker Ranch area near a producing geothermal site in Idaho, which associates the rock’s permeability to water and other attributes, was supplied as “training” input to a computer-based artificial neural network[Wikipedia] to estimate how much the permeability depends on the other attributes.  This turned up some unobvious relationships among the rock attributes, but didn’t result in a clear relationship that matched the geology and effectively predicted the rock’s permeability.  The paper concludes with a summary of findings and things to try next: 


“Achieving a high correlation for a training [data] set is possible, but tends to yield a large number of false positives when applied to real data.  It is clearly possible to identify zones of faults and complexity, but not all faults are permeable.  The addition of mineralogical data did not improve results.  Structurally derived data tended to influence results, but not always as an improvement.  Lithological data did improve results significantly.  Given the inherent risks, it proved not possible to locate a new [w]ell based on these data alone. 


“… either/or 1) an attribute with a more robust link to permeability is needed, or 2) the model space needs to be better zoned or segmented and modeled separately such that attributes that work in part of the reservoir are not applied to other areas where they may be less relevant, or worse, falsely represent the underlying reservoir characteristic, or 3) the seismic data may contain inaccuracies or misrepresentations, which ultimately misleads the modeling efforts.  Future efforts will focus on incorporating new attributes (elements of #1), with #2 above, to yield improved results for prioritized target areas.  The quality of the seismic data appears good, and we have no direct evidence that is cause for lack of modeling success.  Indeed, it is likely that the modeling would not perform as well as it does on known data points if the seismic data lacked a reasonable and consistent tie to geologic reality. 


“Examination of microearthquake locations and 3D seismic data show that the earthquakes correlate areas of more complexity but are not obviously associated with clearly defined faults”


A particular difficulty can occur with determining the structure of subsurface fracture zones from the way seismic waves travel through them.  Where fractures are mainly aligned with a single orientation, the speeds of seismic waves that travel through them depend on the waves’ own direction of motion instead of being the same for every direction.  A new mathematical method, described in “Anisotropic seismic-waveform inversion: Application to a seismic velocity model from Eleven-Mile Canyon in Nevada”[SciTech Connect], can handle arbitrary differences in how a solid’s elasticity differs for different orientations of the same stress, such as those associated with seismic waves traveling parallel or perpendicular to a rock’s main fracture direction.  The method’s developers applied it to data that included seismic and well-log information taken at Eleven-Mile Canyon in Nevada’s southern Dixie Valley, a geothermal exploration site, and found that the method had potential to reconstruct elastic parameters of subsurface rock for imaging and characterizing fracture zones. 


Some geothermal systems work by natural water flow, while others—called “enhanced geothermal systems”[Wikipedia]—bring heat from hot dry rock underground by pumping water to it from the surface to be heated and return.  One goal of a project reported in “Seismic Fracture Characterization Methodologies for Enhanced Geothermal Systems”[SciTech Connect] was to determine what seismic measurements on the surface and in boreholes would be more useful for characterizing potential pathways for the water in such systems, like that at Brady’s Hot Springs in Nevada, with their “volcanic cover, highly altered rocks, severe structure, extreme near surface velocity contrasts and lack of distinct velocity contrasts at depth”.  Other goals were to test borehole tools that could operate at depths below one kilometer at temperatures over 200° C and to develop new data-processing methods and apply new mathematical algorithms for imaging. 


Figure 1.  Simulated snapshots of a seismic wave 0.25 second after its initiation (left) as it crosses a major fault and 0.50 second after initiation (right) as it scatters from the offset beds associated with subsurface faults.  The wave field is severely disrupted by the fault structure.  Colors indicate compressional speeds of seismic waves as shown by the color bar to the right of each snapshot, vertical magenta lines represent vertical seismic profile wells, and yellow dot represents the source of the seismic wave.  (After “Seismic Fracture Characterization Methodologies for Enhanced Geothermal Systems”[SciTech Connect], pp. 23, 25, and 26.) 


Much as was found in the Walker Ranch project just described, seismic reflection data was only useful when combined with other extensive geological and geophysical data.  Mathematical modeling suggests that surface seismic reflection data was unlikely to give any information about the extent, thickness, and filling of faults and fractures; indeed, no such information was found in the researchers’ actual surface data.  On the other hand, mathematical models do indicate that sensors located at depth in wells would be the most effective seismic tools for this purpose.  While useful data was obtained from the high-temperature borehole sensors, wells at Brady’s Hot Springs were too few and far between for detailed fault and fracture characterization, and precluded use of advanced algorithms.  Mathematical modeling also suggested that in other geothermal sites, where seismic shear waves are not strongly attenuated near the surface as they are at Brady’s Hot Springs, such waves should be useful for characterizing faults and fractures. 



Other site characterization techniques


Measurements of things other than artificial microearthquakes have also been studied for clues to the underground structure of geothermal reservoirs.  “Fracture Network and Fluid Flow Imaging for Enhanced Geothermal Systems Applications from Multi-Dimensional Electrical Resistivity Structure”[SciTech Connect] describes University of Utah researchers’ algorithm for inferring how the resistivity of reservoir rock to electric current varies underground from measurements of electromagnetic fields at the surface.  Rock, and spaces in and around rock through which water flows or could flow, will each conduct electricity differently, so determining where the electric resistivity changes belowground should show where water can flow and bring heat to heat engines at the surface.  The algorithm, a further development of a different research group’s public-domain program for three-dimensional modeling, was tested on data taken from Mount St. Helens and on an artificial data set made publicly available for researchers to test algorithms with.  The improvements substantially exceeded the researchers’ objectives.  The report includes copies of four journal articles and a conference paper describing progress with the algorithm’s development. 


Resistivity and seismic data are only two of the information sources used as input to the algorithm whose development is described in “Time-lapse Joint Inversion of Geophysical Data and its Applications to Geothermal Prospecting - GEODE”[SciTech Connect], other sources being electromagnetic and gravitational data.  Both of these latter types are important clues to a reservoir’s structure.  Electromagnetic data provides information about underground resistivity to the electric currents induced by natural variations in the earth’s magnetic field due to lightning and to solar activity.[Wikipedia]  Variations in the strength and direction of gravitational attraction to underground rock from point to point provide information about the rock’s density variations.  The significance of such geophysical data was demonstrated by how its presence at the Jersey Valley geothermal site near Reno, Nevada revealed subsurface structure that surface features alone did not.  The next highly important issue to address after the site’s structure, according to the report, is information about the system’s fluid flow, which is to be addressed with the addition of data taken in May 2015. 


Extending another algorithm to account for significant chemical reactions in geothermal systems is the purpose of a different project, as described in a report entitled “Recovery Act: An Integrated Experimental and Numerical Study: Developing a Reaction Transport Model that Couples Chemical Reactions of Mineral Dissolution/Precipitation with Spatial and Temporal Flow Variations”[SciTech Connect].  The work involved compiling consistent thermodynamic data to accurately describe multiphase fluid interactions with solids—not easy to do since existing databases for aqueous solutions[Wikipedia] and for solids showed inconsistencies, particularly when adding data about carbon dioxide dissolved in water.  This matters since heat extraction from a geothermal reservoir might be enhanced by injecting carbon dioxide into the brine that flows through it at a pressure and temperature high enough to render the carbon dioxide’s gaseous and liquid states indistinct[Wikipedia], though such injection might instead be detrimental if it dissolves the reservoir rock or clogs its pore spaces.  Lab studies of dissolution and precipitation at high temperature and pressure under carbon-dioxide reservoir conditions, combined with observation of particles entrained in multiphase fluid in typical flow paths[Wikipedia], informed the construction of reservoir-scale mathematical models that could realistically account for chemical reactions between supercritical carbon dioxide and reservoir rocks whose pore and fracture shapes vary throughout the reservoir and over time. 


Most of these reports on site characterization techniques focus on data processing, though the work at Brady’s Hot Springs did involve testing tools to operate at great depths at high temperatures.  The patent “Encapsulated microsensors for reservoir interrogation”[DOepatents], on the other hand, is entirely about tools that can be injected into reservoir fluid to determine the fluid’s condition.  The patent’s background section describes how information about where reservoir fluids go or conditions along their paths is often obtained from radioactive or chemical tracers which are easily distinguished from the fluids they’re inserted into, but notes that few tracers are suitable for conditions in geothermal, oil, or gas wells, and that little is known about the stability or behavior of the suitable tracers in specific environmental conditions.  Accordingly, the patent describes how encapsulated microsensors can be injected into reservoir fluid to record data about the reservoir or even transmit such data from within it. 

Figure 2.  “… the encapsulated microsensor 104 is fed into an injection well 112 using fluid from a fluid source 106 … subsequently travels down the injection well 112 … continues into the reservoir 118 … [and] may be configured to detect and/or record one or more conditions of the fluidic medium 132 in the reservoir 118 … [which] may include, but is not limited to, a flow path(s), a temperature, a pressure, a density, a sweep efficiency, a fluid conductivity, a thermal conductivity, a chemical composition, a pH, a turbidity, types of fluids and/or analytes at given depths, a velocity, and other such conditions as would be understood by one having skill in the art upon reading the present disclosure.”   (From the detailed description section of the patent “Encapsulated microsensors for reservoir interrogation”[DOepatents].) 


Inferring a site’s internal structure through a combination of data processing and direct experiment was the subject of a presentation to the 50th US Rock Mechanics/Geomechanics Symposium in 2016.  “Revisiting Fenton Hill Phase I reservoir creation and stimulation mechanisms through the GTO code comparison effort”[SciTech Connect] shows how new things can still be learned from present-day computer analyses of data from experiments done in 1974.  One of the first experimental geothermal power sites, Fenton Hill in New Mexico[Wikipedia], was operated intermittently between 1974 and 1995.  The symposium presentation describes a recent study in which two different computer codes, based on somewhat different hypotheses, were compared to see how closely their mathematical models matched the actual behavior of one well in a series of early experiments with techniques for stimulating geothermal-well productivity by injecting either water or (in the last experiment modeled) a different material for propping open induced fractures[Wikipedia].  The comparison was made to gain insight into exactly what happened inside the well when the water was injected and vented in the first experiments, though given the limited amount of experimental data, the fact that some hypotheses match this data does not by itself rule out other hypotheses.  The exercise proved that the injections may well have opened natural fractures, and that the much higher water-recovery ratio in the last experiment of the series may have resulted from a combination of the injected fluid’s higher viscosity, its lower volume, and its reduction of the surrounding rock’s impedance to fluid flow. 






Although some of the projects described in the reports just mentioned involved the use of data from particular sites, their main purpose was to develop techniques applicable to any prospective geothermal site.  Many other projects, however, have been focused more on whether and how particular sites might be developed as geothermal energy sources. 


Earlier studies have suggested that the volcanic lava field Harrat Rahat in Saudi Arabia might be a productive area for geothermal energy.  “Passive seismic investigation of Harrat Rahat”[SciTech Connect] proposed low-cost identification of prospective sites by combining geologic maps with passive seismic data—observations of natural earth tremors rather than artificial ones like those discussed earlier.  The report also shows cross-correlations of seismic readings between different pairs of seismometers from among a set of 14 spread over a roughly 120-kilometer by 160-kilometer area in and around Harrat Rahat, which further analysis is expected to turn into a three-dimensional map of how fast seismic waves of different frequencies travel between different points throughout the lava field.  Productive geothermal sites should show up as anomalous areas on the map indicating where shallow magma[Wikipedia] chambers lie near faults. 


A somewhat larger area in the United States, Idaho’s Eastern Snake River Plain, is already known as one of six high-grade enhanced geothermal system regions.  But the usual methods of characterizing a region’s geothermal potential are hindered for the Eastern Snake River Plain by its also being the site of one of the nation’s most productive aquifers[Wikipedia], whose fast-moving cold groundwater efficiently intercepts heat before it gets near the surface so that relatively little heat output is observed there.  Nonetheless, a characterization is possible precisely because of the aquifer’s heavy use.  There are numerous wells throughout the plain, some in agricultural areas, and others used to monitor the groundwater near the site of Idaho National Laboratory.  A report from this lab entitled “Wellbore and groundwater temperature distribution Eastern Snake River Plain, Idaho: Implications for groundwater flow and geothermal potential”[SciTech Connect] describes how groundwater temperature data from 250 wells, including 9 deep wells that penetrate the aquifer, was combined with other data to trace aquifer flow patterns, determine the depth to which groundwater flow controls temperature in the plain, and identify both cold plumes associated with recharge from tributary valleys and adjacent uplands and warm zones associated with geothermal input to the aquifer—particularly several zones hot enough to warrant further exploration. 


Figure 3.  Groundwater flow pathways in the Eastern Snake River Plain, Idaho, deduced from ratios of uranium-234[Wikipedia] to uranium-238[Wikipedia].  “High ratios characteristic of mountain recharge source regions extend far into the aquifer in preferential flow zones.  Low isotope ratios characterize surface recharge areas … .”  (From “Wellbore and groundwater temperature distribution Eastern Snake River Plain, Idaho: Implications for groundwater flow and geothermal potential”[SciTech Connect], p. 17.) 


Various aspects of the Astor Pass Geothermal Field, Northern Pyramid Lake, Nevada are described in a poster and two conference papers.  The poster[SciTech Connect], prepared for a Fall Meeting of the American Geophysical Union, displays several kinds of data about fractures in three completed wells and describes ongoing characterization of the fractures: 


“Statistical analysis of fracture orientation, densities, and spacing obtained from borehole imaging logs is used to determine stress orientation and to generate a statistically equivalent Discrete Fracture Network (DFN) model.  Fractures at depth are compared to fracture data collected in nearby outcrops of the same lithologic stratigraphy.  Fracture geometry and density is correlated to mechanically discrete layers within the stratigraphy to test whether variations in fracturing can be attributed to variations in Young’s modulus [i.e., how much the rock is compressed or stretched per unit of pressure or tension—wnw][Wikipedia].  Correlation of fracture geometry and densities with spinner flowmeter logs and distributed temperature sensor records are made in an effort to identify potential flowing fracture zones intersecting the borehole.  Mean fracture aperture is obtained from open fracture counts and reservoir-scale transmissivity values (computed from a 30 day pump test) in the absence of readily available aperture data.  The goal of this thorough fracture characterization is to create a physically relevant model which may be coupled with a multipurpose fluid flow and thermal simulator for the investigation of geothermal reservoir behavior, particularly at the borehole scale.” 


One of the conference papers, “A Preliminary Geochemical Description of the Geothermal Reservoir at Astor Pass, Northern Pyramid Lake, Nevada”[SciTech Connect], describes the chemical content of three fluid samples from one well.  About 1600 milligrams of solids are dissolved in each liter of fluid, about half the concentration found in the Needle Rocks area six kilometers south of Astor Pass.  Most of the content is sodium, chlorine, and sulfur, which are not found in the reservoir rocks, and are “probably derived from dissolution of sodium chloride and sulfate minerals by precipitation recharging through areas that were once the bed of Lake Lahontan”, a prehistoric lake[Wikipedia] that included the present-day Pyramid Lake[Wikipedia] but covered roughly 4 ½ times as much land.  According to the paper, measurements of the concentrations of tritium (hydrogen-3), helium-3, helium-4, carbon-13, and carbon-14 in the fluid indicate that most of the helium in the reservoir is derived from the earth’s crust, with possibly about 3.3% derived from the earth’s mantle, and that the reservoir fluid was recharged sometime between 60 and 1500 years ago, though the 1500-year limit “is suspect due to assumptions that may be poorly met”.  The helium content indicates that the system “is most certainly heated by deep circulation of groundwater”. 


The other conference paper, “Geothermal resource characterization and evaluation at Astor Pass, Nevada”[SciTech Connect], deals with an even wider range of information:  fracture logs, flow spinner logs, and temperature profiles along boreholes, reservoir pressure and temperature responses during a 30-day reservoir test, structural analyses, seismic profiles, and shallow temperature and electromagnetic-field surveys.  From this information, the authors arrived at the following conceptual model of the reservoir, whose description (below, with links added) indicates what is involved in evaluating a site for geothermal power production: 


The Astor Pass geothermal reservoir is a small, low-enthalpy system with relatively isothermal temperatures of 95°C. The shallow temperature and magnetotelluric[Wikipedia] surveys outline a circular to elliptical convective plume of geothermal fluid circulation. Well bore hydraulic testing with a spinner tool and temperature probe indicates that the granodiorite[Wikipedia] basement rocks[Wikipedia] have insufficient permeability to serve as a laterally extensive reservoir, suggesting that a discrete fault or fault intersection (the geometry of Astor Pass tufa tower[Wikipedia] suggests a fault intersection) is responsible for providing a pathway for upwelling of geothermal fluids through the basement and into the geothermal reservoir located in the overlying Tertiary[Wikipedia] volcanics of the Lower Pyramid Lake sequence.


There is a strong correlation between tufa formations and geothermal activity on the Pyramid Lake Paiute Reservation (Coolbaugh et al., 2006; Eisses et al., 2011). In addition to the Needles and Astor Pass, a hot spring discharges from Pyramid Rock, the largest massive tufa tower in Nevada and the namesake of the Pyramid Lake area. Eisses et al. (2011) conducted seismic Compressed High Intensity Radar Pulse (CHIRP) surveys to image fault structures at the base of Pyramid Lake. Washouts and formation of tufa in the processed CHIRP images along normal or dip-slip fault structures, the same structures that formed the tufa features at Astor Pass and the Needles, indicate that these structures are permeable and conduct hot fluid.


The high fracture permeability and large-scale connectivity of the basaltic[Wikipedia], basaltic-andesite[Wikipedia] and rhyolite[Wikipedia] reservoir rocks of the Pyramid Lake sequence support findings from the structural geologic analysis that Astor Pass is in a region of enhanced extension. The reservoir rocks offer little resistance to the upward migration of geothermal fluids. Upon encountering cooler, near-surface temperatures, the geothermal fluids cool, and migrate downward along the outline of the convective cell due to density contrasts. The long duration of the reservoir hydraulic test and the high resolution monitoring of fracture inflow locations and borehole temperatures eliminate the possibility of preferential flow of geothermal waters exceeding 100°C. These features would have been intersected as the cone of depression (which spread to a radial distance of at least 1 km as evidenced by drawdown in the water well), and hotter inflow temperatures would have been detected during the temperature surveys and high resolution DTS[Wikipedia] measurements.






While the preceding reports are oriented toward extracting geothermal energy at multimegawatt power-station rates, other recent reports deal with setting up geothermal power systems for single buildings—in particular, replacing decades-old systems with geothermally-powered ones.  “Final Report Roberts Wesleyan College Geothermal Energy Project CPD No. 84.09”[SciTech Connect] is a brief description of a project in Upstate New York to convert the college’s Carpenter Hall, an academic building of 36,000 square feet (roughly 3,300 m2), from radiator heating and window air conditioning to a HVAC system using heat pumps powered by five grouped series of eight geothermal wells apiece next to the building.  The new system has cost roughly $12,000 more per year to operate above the previous average annual cost of about $27,000, but also provides more air conditioning than the earlier system did.  Another single-building project, which replaced several systems in Colorado’s state capitol building, actually reduced energy costs while adding cooling to the building’s Senate and House chambers for the first time.[SciTech Connect]  The capitol’s new system, whose design was chosen to be minimally disruptive as well as economical, is powered from two geothermal wells in the building’s adjacent parking circle.  The more extensive report on this project describes design and construction problems solved and lessons learned in considerable detail. 


While single buildings may be near the smallest important scale for geothermal power, single utility-scale power stations aren’t the largest.  Small- and large-scale electrical grids now include geothermal power sources as one among many types that affect their functioning.  Geothermal energy is particularly important in certain regions, such as the Big Island of Hawaii, where it provides about 20% of the electricity to the island’s power grid.  As such, it is a significant (though not dominant) consideration in a report on “Development of a Hydrogen Energy System as a Grid Frequency Management Tool”[SciTech Connect].  State policy requires displacing the use of fossil fuels to generate electricity with “renewable” (practically nondepleting) sources, which include geothermal energy.  While geothermal sources can feed a power grid at a very steady rate, other such sources, like sunlight, wind, and ocean heat, provide a much more irregular input.  This presents a problem, since some electrically powered machines depend on steady alternating-current frequency to function properly. 


The report deals with a series of experiments with a possible solution to this problem:  feeding input energy that fluctuates above immediate demand to water electrolyzers to make hydrogen, which can be used as a fuel, while the extra input lasts.  When the demand goes back up or the energy input fluctuates back down, the electrolyzers would reduce their hydrogen production.  Geothermal energy’s importance for grid frequency management is its lack of significant fluctuation, which lessens the demands on electrolyzers connected to the power grid.  The electrolyzing system’s controllers functioned much better in the project’s later experiments than in the earlier ones, and the experimenters concluded that “the electrolyzer is a valuable grid frequency management tool capable of controlling intermittent renewable sources of energy for grid frequency management applications”.   The electrolyzer didn’t respond as quickly to input changes as a battery system used in the experiments, but it was useful in managing slow changes.  The experimenters propose designing a system that mixes both electrolyzers and batteries to regulate the grid at minimum cost, and note that “controls work is well underway in the management of multiple renewables and loads on a microgrid”. 


The report also mentions another role for geothermal power in an economy that uses hydrogen as fuel:  since hydrogen can be electrolyzed from water using any kind of power source, geothermal energy can directly power off-grid electrolyzers to produce even more hydrogen.  The uses described for geothermal energy, hydrogen, and other practically nondepleting power sources would advance Hawaii towards its policy goal of generating all its grid electricity from such sources by 2045.  Currently Hawaii generates half its grid power from them today, as it continues to strive toward meeting as much in-State demand for fuels as is feasible from indigenous fuel sources, for instance by using nondepleting power sources like geothermal energy to electrolyze hydrogen.  









Reports available through OSTI’s SciTech Connect








  • Recovery Act: An Integrated Experimental and Numerical Study: Developing a Reaction Transport Model that Couples Chemical Reactions of Mineral Dissolution/Precipitation with Spatial and Temporal Flow Variations. [Metadata]
    Regents of the University of Minnesota











Patent available through DOepatents


Other references





Last updated on Thursday 10 November 2016