FORGE- Geology & Geothermal Details
Milford, which is 10 miles (16 km) from the proposed site, is incorporated as a city in Beaver County, and has a population of 1400 (Figure 1). It has good motel accommodation (with 24-hour diner), a supermarket, hardware, and a hospital. A major factor in Milford’s history is the Union Pacific (UP) Railroad, which passes through the town and has a siding complex and an office to facilitate freight train scheduling in this part of the state. The railroad offers the possibility of shipping drilling materials (e.g. pipe, proppant) by rail and then using truck transport for the final few miles to the FORGE site. Two miles north of Milford is the Milford Municipal Airport with a sealed runway 5000 feet long, and adequate for piston or turboprop, single- or twin-engine planes. The proposed site is 350 km (220 miles) south of Salt Lake City, and the drive time is about 3 hours. Longer-term, if the site is successful in creating power generation opportunities, the existing DC line from the IPP coal plant (Delta) to southern California (about 2 GW capacity) is nearby, and a new transmission line (Transwest Express) with 3 GW capacity is being permitted following a similar route to the IPP line. Although wind, natural gas, and solar are the dominant power sources being considered for these lines, the FORGE site has the potential to also prove GW-scale power potential. The PacifiCorp Energy 0.6 GW line cuts across the proposed site.
An area of about 25 km2 (10 miles2) has been identified as containing three possible reservoir options for the ultimate test site (Figure 2, labelled A, B, C). Within this area, granite and granitic gneiss occur between 2 and 4 km depth, with temperatures in the range of 175 – 225°C. This site is 4 – 8 km west of the Roosevelt Hydrothermal System (RHS) where power is being generated at PacifiCorp Energy’s Blundell facility, and the reservoir temperature ranges up to 265°C (Allis and Larsen, 2012).
The proposed FORGE site has been previously been suggested as a suitable for EGS tests (East, 1981, Shannon et al., 1983, Goff and Decker, 1983, Tosoya et al., 1988). A deep exploration well (Acord-1) drilled in 1979 to 3.8 km depth exists near the western edge of the proposed site. Here, centrally located near the middle of the Milford valley/basin, the granite surface is at 3.1 km depth, and the temperature at total depth is 230°C. Many logging runs were made in this unproductive well, providing excellent temperature information from the bottom hole measurements (Figure 3).
Figure 4 summarizes the thermal regime in other deep exploration wells drilled mostly in the late 1970s and early 1980s around the east side of Milford valley (locations, Figure 2). The deeper portions of all wells west of the Opal Fault Mound have conductive gradients of 40 – 65°C/km and indicate thermally conductive, low permeability host rock (granite). Figure 5 is a simplified west-east cross-section through the middle of the site area based on three deep wells north of the Roosevelt production zone, but intersecting the northern end of the Opal Mound Fault at well 12-35. The shape of the granite – basin fill interface is based on a 2-D gravity interpretation (discussed later).
Within the FORGE site, reservoir option A is next to Acord-1. This well was lightly plugged and suspended, and is available to be cleaned out and used for preliminary testing of tools without risking the more expensive, new holes nearby with horizontal sections. It is located between two wind farm turbine arrays, which present some possible constraints, but it is a viable and attractive option. Reservoir option B is 3 km to the east on the alluvial fan and is near the eastern edge of the proposed site area. Option B has the advantage that the granite surface will be shallower (about 1 km depth based on the gravity interpretation), the target zone for reservoir creation will be between 175 and 225°C at 2 to 3 km depth, and it lies at the eastern end of the wind farm array. Option C is 4 km south of these two sites, and based on gravity and thermal gradient wells, is expected to be very similar to Option A. Option C is south of the wind turbine array and is close to power and fiber-optic connections.
Over 40 thermal gradient wells were drilled by three exploration companies and by a DOE-funded research program at the University of Utah during the assessment of the RHS in the late 1970s. All the shallow thermal data were compiled by Professor David Chapman and his students, and they have been turned over to the UGS for this project. Figure 6 is a preliminary reassessment of these data showing the thermal regime at 200 m depth. Re-evaluation of the temperature data between 100 and 200 m depth in the original thermal gradient wells supports a thermal outflow zone at shallow depth. More importantly, these data indicate a deep thermal anomaly at least 14 km long in a north-south direction, and extending up to 10 km westward towards the center of Milford Valley, with the Acord-1 well near the western edge of the anomaly. A re-interpretation of the bottom hole temperatures in this well shows a conductive heat flow of 115 ± 20 mW/m2, a thermal gradient of 60°C/km over the uppermost two km of sedimentary and tuffaceous strata, with the main source of uncertainty in heat flow due to the assumed thermal conductivity of the host rocks (Figure 3; Allis et al., 2014). This heat flow is suspected to be close to the background heat flow in the Milford Valley, based on similar heat flow values in shallow wells west of Milford City (Shauntie Hills, Chapman et al., 1978).
The area with a temperature of at least 40°C at 200 m depth is about 100 km2. The proposed FORGE site occupies the northwest quadrant of this thermal feature, and our present conceptual model has about 90 km2 of this anomaly as low permeability. At 2 km depth the temperature decreases westwards from 190 – 200°C, to almost 150°C in center of the valley near Acord-1, and higher temperatures occur at greater depth beneath this region as shown in Figures 4 & 5. The heat flow within the proposed site ranges from 115 mW/m2 at sites A and C, to possibly as high as 180 mW/m2 at site B.
Ward et al., (1978) assessed from the very shallow thermal data (< 60 m depth) that the total heat output from the hydrothermal upflow was about 70 MWth. Becker and Blackwell (1983) modeled the possible deep fluid flow regime using the available geophysical data including the low velocity anomaly beneath RHS based on teleseismic P-wave delays (Robinson and Iyer, 1981). The “intensely anomalous region of low velocity and high attenuation extends” from about 5 km depth to the base of the crust and has been interpreted as caused by a small fraction of partial melt in a granite intrusion. This is centered west of RHS and has an area of about 100 km2. Anomalous He3/He4 in the spring fluids supports the partial melt interpretation (Kennedy and van Soest, 2007). The 2-D fluid flow models of Becker and Blackwell (1983) approximately fit the known geophysical data, with recharge occurring within the Mineral Mountains, upflow on the Opal Mound fault with the RHS, and shallow outflow in the sediments to the west. The large area of granitic rock beneath the Milford Valley is assumed to have a low permeability of 10-16 to 10-18 m2.
The extensive area of high temperatures at 3 km depth west of the RHS indicates substantial geothermal potential in the granite basement beneath the Milford Valley, as noted by Becker and Blackwell (1983). Figure 4 provides a guide to the stored heat beneath the valley. Assuming this cross-section is applicable to an 8 km length of the valley (i.e. the north-south extent of the 40°C contour in Figure 5), the area of hot rock between the location of well 82-33 in the east and the Acord-1 well in the west (about 8 km) exceeds 50 km2, and the volume between 2 and 4 km depth is more than 100 km3. There are various ways to calculate the power potential of this rock volume (MIT, 1986), but assuming the more conservative thermal recovery factors based on recent EGS projects (less than 2%; Grant and Garg, 2012), the scale of continuous power generation is about 300 MWe. Improvements in fracture network creation, one of the goals the FORGE project, could result in the Milford site having a power potential of more than 1 GW. Clearly, the stored of heat volume around the proposed FORGE site is massive if technologies can be developed for creating a fracture network in the granite and effectively sweeping out that heat for power generation.
Supporting Infrastructure and Data
The Milford Valley in west Beaver County has a rich history of renewable energy power developments (Figure 1). The first liquid-dominated geothermal plant in the U.S. was commissioned at Roosevelt Hot Springs in 1984 (Blundell, now Pacificorp, 36 MW) after much exploration drilling in the late 1970s. The proposed Milford EGS site benefits from the large volume of DOE- and industry-generated data and reports during this exploration period. Two other geothermal power plants have been subsequently built in the County: at Thermo Cyrq Energy has a 10 MW plant, and ENEL has a 23 MW (gross) plant near Cove Fort. Other nearby renewable power projects include Murphy-Brown (Circle Four) generating power from biogas collected in their hog farm facilities south of Milford; First Wind has 306 MW of capacity in North Milford Valley; and First Wind is also currently gaining permits for a further 3 x 80 MW of PV solar capacity west of the wind farm. The proposed Milford EGS field site has strong support from the local community, as shown in the letters of support from Beaver County and Milford High School (Appendix D).
The existing geothermal infrastructure includes of 40 thermal gradient wells (discussed above; data held by UGS), 15 deep geothermal exploration and development wells (logging data for many held by EGI and UGS), and a large collection of geoscientific reports generated by University of Utah Research Institute (UURI) during the late 1970s and 80s (summarized in Appendix B). The UGS holds the cuttings for the 3.8 km of Acord-1; EGS holds cuttings for 9-1, 52-21, 14-2, and 24-36. At least five groundwater wells are available for sampling or monitoring, and we expect additional wells in the area will be found. The University of Utah has been operating a 3-component, short period seismograph (NMU) 5 km from reservoir option B in the proposed FORGE site since 2012, and this will contribute important understanding of local baseline seismicity.
Water is available for the required drilling and testing at the proposed site. North Milford Valley is not closed to new appropriations of groundwater (see section below on permitting), and the UGS has received a water right for up to 50 acre-feet/year for use in this project during the next 10 years (total volume of water is 175 million gallons). Another 50 acre-feet of water will also be available from Murphy-Brown groundwater wells, the remainder will require an additional groundwater well, likely close to the final site option. Power for at least a project site office (residential load) close to the drilling site requires a line from where the main access road crosses the UPR tracks. This will require a 2 mile power line (Option C), or a 5 mile line (Options A and B). A fiber optic line allowing data transfer to the internet already extends to the First Wind operations and maintenance facility next to Option C (Figure 7a; link provided by South-Central Communications Inc.). In fact the fiber-optic line extends beside the road through the middle of section 11, which is Option C. Depending on the site option chosen, verbal estimates from representatives of Rocky Mountain Power and South-Central Communications indicate the cost of obtaining power and internet access will be in the range $20,000 to $80,000.
The existing developments in North Milford Valley mean there is already an extensive road system linking all three site options. These roads are largely graveled and capable of supporting large trucks and drill rigs. A 1 km length of two-track road into Acord-1 will likely need some improvement with gravel. Local quarries can provide and spread this gravel.
Other Site Characteristics
The three site options are on gently-sloping ground towards the west (Figure 7). The slope is greatest at Option B at 80 feet per half mile, or 3 degrees, which is too small to be an issue if this becomes the drilling site. The site is accessible at all times of the year.
The proposed site straddles lacustrine deposits near the center of the valley, and predominantly alluvial fan deposits in front of the Mineral Mountains. This part of the Mineral Mountains comprises Tertiary to Quaternary granite, diorite, rhyolite lava, and Precambrian gneisses also intruded by granite (Figure 8a). These rocks underlie the sedimentary section beneath the valley, and at the Acord-1 well are at 3.1 km depth. Immediately overlying the granite in Acord-1 is 560 m of andesite, almost 2 km of tuffs and tuffaceous sediments, and 600 m of claystone and surficial valley fill (Hintze and Davis, 2003; Figure 8b).
North Milford Valley has a 15 – 20 mgal negative isostatic anomaly oriented north-south, due to the effects of Neogene Basin and Range extension. Carter and Cook (1978) attempted a 2-D interpretation of the thickness of the sedimentary fill, assuming a uniform density contrast. This was prior to the drilling of Acord-1, so they under-predicted the depth to granite in the middle of the valley by over a kilometer. Our preliminary, simple modeling of the gravity data for this proposal using realistic density contrasts with increasing depth and calibrating the model on the known depth to granite in Acord-1 and 82-33, constrains the shape of the basin beneath the valley (Figure 9). There is a gently-dipping shelf on the granite surface westwards away from where it crops out near 82-33, and the eastern flank of the main basin lies between site options B and A (or C). This flank has a relatively steep slope and coincides with the location of the Lake Bonneville high-stand shoreline on the ground surface. We suspect there may be north-south faulting in the granite here, which is why the proposed sites avoid this zone. Seismic reflection imaging is a high priority for Phase 2. During the Phase 1 site characterization a 3-D model of the granite surface will be created to better delineate the shape of the basin. Two seismic reflection lines dating from the early 1980s were shot across the proposed site area (Seismic Exchange Inc. website). One line trends west east through site option C, and the second line runs northeast between site options A and C (along green shaded road in Figure 7a). Smith and Bruhn (1984) show a small reproduction of the east-west line, with reflections from basin fill visible. Unfortunately the data used by these authors is likely to have been lost, so early in Phase 2, the lines will be licensed from SEI to image the granite contact.
A low-level aeromagnetic survey over the region was flown for the University of Utah in 1978 with lines spaced one-quarter mile apart and draped at a nominal 1000 feet above the terrain (Carter and Cook, 1978). These authors removed the International Geophysical Reference Field to produce residual anomalies and then used a reduction-to-pole technique to remove bipolar effects. The resulting map is shown as Figure 10. The main magnetic highs in the Mineral Mountains are due to the Tertiary intrusions and older granitic gneisses. Quaternary rhyolite flows appear to be reversely magnetized. Carter and Cook (1978) compiled the density and magnetic properties of over 70 samples with varying lithologies, which will assist reinterpretation of both the gravity and magnetic anomaly maps.
The resistivity signature of the lacustrine sediments in the central Milford valley shows up in magnetotelluric soundings. A line of MT soundings contracted by the UGS in 2012 along the access road to the power plant shows the trend of increasing thickness of low resistivity (1 – 2 ohm-meters) claystone from the edge of the Mineral Mountains west towards the central portion of the valley (Figure 11). Basement and deeper crustal signatures may also be imaged by MT measurements. These will require full 3D inversion to account for diverse structural trends, but should yield information about potential rock failure directions and heat sources. Other industrialization nearby to the northwest and west of the proposed FORGE area (e.g., wind turbines) will require use of ultra-distant remote referencing in any new MT data campaigns, but this is standard procedure in such cases and has worked quite effectively for the project team in eliminating noise from other major industrial sources such as regional DC transmission systems.Resistivity
Seismicity recorded by the University of Utah Seismograph Network since 1981 is shown in Figure 12. This includes anthropogenic seismicity such as coal-mine induced events (clusters near the NE corner of the map), and quarry blasts (Cricket Mountains cluster and possibly also the cluster 10 km north of Milford. Since 1981 in this region there have been 327 events -0.2 < M < 3.9, and in the period 1850-2013 there were 14 Mw > 3 earthquakes including (the largest event) the 1908 Mw 4 Milford earthquake (Arabasz et al., 2014). The station NMU is 5 km from the proposed FORGE site and has had a 3-component short-period seismometer since 1989. The noise characteristics are excellent, and depending whether a few suitable template events (larger, well-recorded events) can be identified, the archive should allow identification down to M1 events for baseline studies. Preliminary inspection of the noise spectra over the last decade do not show significant changes in the 1 – 8 Hz range (generated by local earthquakes) due to development of the wind farm.Seismicity
Pressure and Stress
Several publications refer to the stress regime in the vicinity of the hydrothermal system (e.g. Yusas and Bruhn, 1979; Smith and Bruhn, 1984; Nielson et al., 1986.) Steep, high-angle normal faulting parallel to the range-fronts are commonly the most recent, dominant faults, but low-angle dip and listric normal faulting possibly associated with pre-existing Sevier-age thrusts may also be important beneath the Milford Valley. The Basin and Range extension direction is approximately west-east and is expected to be the minimum stress direction. Fluid pressures everywhere in valleys of western Utah are hydrostatic from near the ground surface (Figure 13; Allis, 2014). This is a characteristic of at least the eastern Great Basin and is attributed to widespread, range-bounding faults (e.g. Opal Mound Fault) with high permeability penetrating deep into the upper crust. Basement joints and fractures elsewhere in the extensional brittle crust may allow pressure communication on a 102 – 103 m length scale consistent with the similar length scale of the transition from strong to weak rock (JASON, 2014).
The groundwater west of the Mineral Mountains has a strong geothermal signature due to the outflow from RHS (Capuano and Cole, 1982; Ross et al., 1982; Moore and Nielsen 1994). Chloride and Boron are distinctive signatures of this outflow. Poor water quality is most likely the main reason the north Milford Valley has not been fully developed for agricultural water. Recently this was confirmed by First Wind, who drilled a groundwater well at their O&M site (next to Option C; Figure 7a) and found that the water quality was too poor for their site. The UGS sampled five groundwater wells near to the proposed FORGE site in preparation for this proposal to investigate the groundwater quality (Table 1).
The groundwater analyses confirm a strong geothermal signature with TDs values of ranging from about 2000 to 6000 mg/kg. A stock well 5 km north of site options A and B contains 430 μg/kg of arsenic, which is considerably above the EPA recommended limit for potable water of 10 μg/kg. This well is in the down-gradient drainage area of the Negro Mag Wash, the primary direction of the high-temperature hydrothermal outflow from the RHS. The chemical analyses in two wells with data from the early 1980s indicate water compositions largely unchanged from those measured prior to development of RHS. Poor groundwater quality is a positive environmental factor for the FORGE site because it diminishes concerns that somehow the project could contaminate potable groundwater, and it also removes concerns that the project will be consuming groundwater that might have been used for other purposes. Preliminary investigation of productive potential of these groundwater wells confirms excellent flow characteristics, so the project is optimistic that one well may be able to supply the all water needs for stimulation of the granite beneath the FORGE site. An analysis of the chemical reactivity of these ground waters when used as drilling fluid or during long term heat exchange testing will be part of Phase 2.
The project has agreements for surface access with the three main landowners (Murphy-Brown, Utah School Trust Lands – SITLA, and the BLM; all details in Appendix A). The primary landowner where the surface infrastructure will occur if this proposal is successful, Murphy-Brown, is a livestock production subsidiary of Smithfield Foods Inc., the world’s leading pork producer. All three FORGE site options identified above are on Murphy-Brown property. Murphy-Brown also owns the mineral lease rights to the subsurface resources beneath their lands. The agreement with Murphy-Brown for access to the surface for surveying and drilling, and to the subsurface for reservoir stimulation and heat exchange testing is contained in Appendix A. The agreement with SITLA also includes approval to drill underneath their land for the purposes of reservoir stimulation and heat exchange testing, if this is required.
In Utah, access to, and the use of, geothermal energy are authorized as water rights and administered by the State Engineer in Utah’s Division of Water Rights, DNR. The State Engineer has appropriated to the UGS 50 acre-feet/year of groundwater (16.3 million gallons/year) for this project during the next 10 years (Appendix A). Murphy-Brown has also offered an additional 50 acre-feet of water for this project from their appropriated but unused water rights if more than 50 acre-feet is required in a particular year. A change application may have to be filed with the State Engineer to allow use of the Murphy-Brown water, but this is not a difficult process. The project therefore has access to a total of 175 million gallons for the project, which is considered ample for the drilling, stimulation, and heat exchange testing. The water right from the State Engineer includes a permit for the deep production and injection wells with horizontal legs. The Utah Department of Environmental Quality (Division of Water Quality) has reviewed the solicitation, and after several discussions about our proposal, has decided to grant an “authorization-by-rule” as long as the project agrees to work with the Underground Injection Control program in developing a set of conditions (letter in Appendix A).
BLM have land parcels adjacent site options B and C. The agreement with BLM (Appendix A) allows access for non-invasive surveying as required in Phase 2B, and the ability to apply for a casual-use permit for drilling thermal gradient wells on their land during Phase 2B. Drilling under BLM land for the purposes of extracting heat as envisaged by this project would require a competitive permit, so the project will avoid this requirement.
First Wind is an existing user of Murphy-Brown land close to the proposed FORGE site options A and B. Option C is outside the area of influence of First Wind’s wind-shed. At site option A, a 10-acre work site can also be located outside their wind-influence envelope, but the location of a clean-out rig over Acord-1 is technically within their influence envelope. Discussions with First Wind have indicated that a request to temporarily put a rig over this well for the clean-out (or for drilling another nearby well) is reasonable and would be considered once the details on location and timing are known (copy of email exchange, Appendix D). Structures up to about 60 feet from ground level do not intrude the wind-shed, and this easily allows small rigs that might be needed to install instrumentation and other general FORGE site activities. First Wind has also requested no above-ground power lines nor ponds within the wind farm area. Both these constraints can be met. During the cleanout of Acord-1, the existing groundwater well and existing pond next to this well will be used. If site option A becomes the preferred choice, any ponds can be covered, or large frack tanks can be used for surface water storage.
As part of the NEPA approval process a cultural survey of the site will need to be carried out, with the results reviewed by Utah’s State History Preservation Office. SHPO does not have specific concerns about the proposed area, but has stated there are likely to be sites (which the project will be able to avoid) (letter in Appendix A). Due to the intense land use, the numerous wells drilled over the last three decades around the proposed FORGE site, and the recent wind-farm development which cleared NEPA, no difficulties with cultural clearances are expected.
A presentation of the proposed project was made to PacifiCorp Energy in Salt Lake City. They are the operator of the Blundell power plants several miles from the proposed EGS site. Their letter of support is attached in Appendix D. Other than being kept informed of progress, they are not interested in direct involvement in the project. Over the last few years they have allowed a considerable volume of data from the exploration wells at Roosevelt Hot Springs to be published (e.g. in Allis and Larsen, 2012). Data from the three wells west of the hydrothermal system (82-33, 9-1, and 52-21; Figures 2 and 6) are available for this project. In addition we have logs and cuttings from well 14-2 inside the production borefield – this well was the hottest well drilled in the RHS. Each of these wells has many geophysical logs, and penetrated granite a shallow depth. These logs should provide insights to the physical properties of the granite.
Other stakeholders indicating local support include Beaver County and Milford High School (letters in Appendix D). Beaver County has five different types of renewable energy power projects within a 50 mile radius, with several hundred MW of solar PV about to be commissioned. Renewable energy developed is a major driver of economic development within the county, so this project is a welcome addition to ongoing energy developments. Milford High School has for several years run courses for the students on renewable energy, and combined this with visits to the various sites. Some of their students have been able to gain summer work at these sites, and they run a renewable energy fair each year which attracts participation from the local community. There are opportunities for the proposed FORGE project to interact with both the High School and with the local community, and the National Energy Foundation has agreed to assist with educational initiatives.
The Utah Energy Office and the Governor’s Office are also excited about this proposal, and the widespread support from public and private sector agencies that has been shown for this project. A copy of the letter from the Governor to the Secretary of Energy is also included in Appendix D.
Permitting is not expected to be a problem at the proposed site, which is on private land, and may, if required, extend onto or under two nearby Utah State Trust Lands (SITLA) sections. The rights for access to the required water have already been obtained, and this includes permission to drill the deep wells. There has already been significant ground disturbance from early geothermal exploration and drilling in the region, and more recently from development of the wind farm, which achieved NEPA clearance. There are no significant wetlands, archeological sites, or sole source aquifers near the proposed site. The nearest population center is Milford about 16 km away. Anthropogenic seismicity from quarry blasts have historically occurred near Milford, and the daily passage of freight trains through the town likely cause far greater shaking (and noise) than potential fracturing events at the proposed FORGE site. The primary constraint is not interfering with the operations of the wind farm, which is easily manageable at reservoir site options B and C (east and south of the wind-shed respectively), and can also be achieved at option A. Demonstrating how geothermal developments and wind farm developments can be compatible is considered to be a valuable aspect of this project proposal.
MIT (2006). The future of geothermal energy: Impacts of enhanced geothermal systems in the 21st century. Massachusetts Institute of Technology, p. 372.
Grant M.A., and Garg, S.K., 2012. Recovery factors for EGS. Proc. 37th Workshop on Geothermal Reservoir Engineering, Stanford, California, p. 3.
JASON, 2014. Subsurface characterization letter report, September 2014, JSR-14-Task-013, p. 18.
Smith, R. B., and Arabasz, W. J., 1991, Seismicity of the Intermountain seismic belt, in Slemmons, D. B.,
Engdahl, E. R., Zoback, M. D., Zoback, M. L., and Blackwell, D., eds., Neotectonics of North America:
Geological Society of America, Decade Map Volume 1, p. 185-228.