AAPG/GEODE Geothermal Subsurface Characterization: Geology & Geophysics Workshop. Summary & Review

   This was a very informative, free two-day online workshop. I have participated in many of AAPG Academy’s great webinars hosted by Susan Nash, and this one was great as always. There were seven individual presentations, nearly all presented by Ph.D. scientists. The summaries are from my notes. The bold parts are where I noted good slides that should go with this summary. When they become available, I will dig them up and add them to this post.  

 

Presentation 1 - Overview of Geothermal Systems and Basic Subsurface Characterization in Sedimentary Basins; Presented by Eric Stautberg, Colorado School of Mines

     Stautberg first went through EIA data about U.S. geothermal. We have 3.9 GW of total geothermal capacity, but only about 65% of it is currently utilizable. Seven states produce geothermal energy, but that is likely to expand in the future. He highly recommends the DOE’s GeoVision 2019 report. Conventional geothermal utilizes existing natural hydrothermal systems in the subsurface. Enhanced geothermal systems (EGS) require stimulation of the reservoir followed by the addition of water to create a hydrothermal system. Direct-use geothermal can be utilized alone or with geothermal power production via steam turbines. Blind geothermal refers to geothermal where there is no surface expression, such as hot springs. Super-hot geothermal, with temperatures above 400 degrees Celsius, is not currently viable, but research is ongoing. Sedimentary geothermal systems are beginning to be developed using deep brines, EGS, and other configurations. Geothermal energy storage uses are expected to increase, using depleted oil & gas fields. They can be used for piping hot water into buildings.

     Conventional geothermal and EGS may use igneous, metamorphic, or sedimentary rocks. Among sedimentary rocks, carbonates, siliciclastics, and possibly evaporites may be utilized. Where rocks are very hot, dry steam and flash steam plant designs may be used. The world’s largest geothermal power plant, Geysers Hydrothermal Power Plant in California, has a capacity of 630 MW.

     He notes that there are lots of analogies between geothermal systems and petroleum systems. Both require a fluid system in the rock. Both require drilling, casing, and other well systems.

     ESG requires drilling into hot dry rock and pumping in water. The FORGE projects, a collaboration of Fervo Energy, the U.S. DOE, and others, drilled into granitic basement. Binary cycle power plants are the most common geothermal configuration, where there is one injector well and one producer well. The Deep Corp Project in Saskatchewan, drilled into basal sandstone in the Williston Basin, has producers and injectors, and taps rock at 200-300° F temperature.

Geothermal wells use hot brine and working fluids. The working fluids make harvesting the energy more efficient. Most are open-loop wells, and now, in some configurations, fully cased closed-loop wells are connected together. Eavor Energy and a few other companies have been specializing in closed-loop geothermal. This has been known as advanced geothermal, but since the formation fluids do not contact the water being conveyed, that water is considered to be a working fluid. They heat by conduction, which transfers less heat and is thus less efficient. Some say conductive geothermal will never be economical, but it does have lower costs to construct and is easier and cheaper to operate than open-loop geothermal. Stautberg recommends the SMU Geothermal Laboratory as a good information source.

     Sedimentary Basins are being explored for geothermal, including different types – rift, passive margins, intra-cratonic basins, and foreland basins. Some are geopressured geothermal systems. Since there are lots of oil & gas wells in sedimentary basins, and some basins are drilled into hot reservoirs, they are being considered for repurposing. Hot water from oil wells can be used for heat or power. Converting depleted fields for storage or to create a geothermal network is being explored in California. NREL – slide on what it takes to make geothermal power. There is modeling that can calculate the amount of water needed, which is controlled by temperature. Required flow vs. temperature is the main consideration. One can then estimate the porosity, permeability, and fracture permeability needed. He notes that pressure is a double-edged sword. While it is good for moving the water, if it is too high, it can make it harder and more energy-intensive to inject, resulting in parasitic loads to the system.

     Reservoir temperature mapping is a major feature of geothermal exploration in sedimentary basins. Fortunately, there are great data sets with bottom-hole temperature (BHT) data collected routinely during geophysical well logging. That data can be used for isotherm mapping, maturity, basin modeling, geothermal gradients, engineering requirements, and other assessments. An important factor in BHT is ‘time since circulation’, or TSC. After a well finishes drilling, pulls out of the hole, and sets up logging tools, the drilling mud is not being circulated, so the more the time since circulation, the more accurate the BHT will be. BHT measures the temperature of mud at the bottom of the hole. Circulation in drilling extracts thermal energy. When hot fluids like drilling mud come up to the surface, they cool down. Mud type affects the rate of wellbore heat-up. Equilibrium temperature data, taken on other logs like Cement Bond Logs or cased Temperature logs, is closer to the true formation temperature since they are long past circulation. There are different BHT temperature correction data and methods, and different methods for different basins. Depth to 250° F is an important metric for power production threshold. Pressurized zones are targeted for high water volume rates. Halite salt diapirs along the Gulf Coast may have potential in geothermal. He thinks South Texas works for sedimentary geothermal. He also notes that we need better BHT correction methods for deeper data or for different areas and rocks. How wells were drilled and logged is important for BHT correction, as is geology such as uplift and high thermal conductivity salt diapirs. Temperature mapping for exploration and project ingenuity is needed.

 

Presentation 2 - Case Studies in Subsurface Characterization for Geothermal Exploration, Development, and Operations; Presented by Kellen Gunderson, Ph.D., Projeo Corporation

     Conventional geothermal requires buried magma plus local fracture networks. Rift systems typically have a magmatic source and convection cells along faults. Geothermal can be separated into magmatic vs. anamagmatic. Magmatic geothermal is mostly explored already. Only 2.5 GW of the total 3.9 GW capacity in the U.S. is producing. This is due in large part to the difficulties of maintaining production in geothermal wells, which are subject to mechanical failures and maintenance issues like metal precipitates, such as scale and corrosion. Gravity surveys and shallow field surveys are exploration tools. Gunderson notes that half of the geothermal fields in the U.S. were discovered accidentally, or blind. The geothermal conceptual model – good slide, including isotherms and thermal manifestations (high T readings). The three things needed to produce geothermal energy are heat, flow, and a balance between injection and production. Searching for convection cells involves noting where they would be expected, and as indicated by geothermal gradients above the normal gradient from magma or sometimes due to upwelling hot water. Plots of temperature vs. depth are used to look for isothermal areas.

     Some exploration methods use shallow temp surveys with 2-meter probes. Direct probe drilling uses a hydraulic hammer and can go to 100-150ft. It is truck-mounted and can drill six deeper holes per day. Twelve per day of the 2M probes is common. These methods are used to plan deeper exploratory drilling. Geologists are also looking for flow characteristics, mainly faults and fractures. Tensile fractures open up. 1 meter bit drop experienced when drilling into some faults. Geothermal does require high flow rates. He notes that 35,000 to 175000 Bbls per day are needed! = 1000-5000gpm. LIDAR data and high-resolution DEMs can be used to find surface fault expressions.

Gunderson notes eight categories of structural settings after Faulds and Hinz 2015. They found that subsidiary faults had better hydrothermal system development than main faults. Stepovers and fault intersections can be better than big faults. USGS has great LIDAR and digital elevation models (DEMs). Magnetotelluric surveys have long been used to image upwelling zones and are a good reconnaissance tool.

     Managing injectivity is crucial to geothermal success. Without injections, there is a T and P decline and lower power. Even with injectors, the T & P need to be balanced, but also need to be tweaked to do so. Temperature decline and flow decline both occur in wells. Highly fractured areas are harder to design where to put injectors and producers in conventional geothermal. Tracer tests can help understand flow paths. Cold water injection stimulation is commonly used to open up fractures when they are thought to be closing up. He recommends incorporating discrete fracture networks as in oil & gas. Play fairway mapping and machine learning can be used to develop risk maps. Oil & gas people are surprised by the lack of use of seismic for geothermal. It is, however, hard to map permeability with seismic. Cost is an issue, too. Chemical geothermometry is important. Measuring say amount of silica in hot springs can predict reservoir temperature at depth.

     Hot sedimentary aquifers are being tapped in the Rhine Basin, Netherlands, Williston Basin, etc. The Salton Sea play in California is a geothermal hot spot, literally and figuratively. In some places, faults feed sedimentary zones. Thus, production is via sedimentary aquifer but fed by convection along faults. Deep Earth in Saskatchewan is a project that taps a conduction-dominated temperature regime and basin, which is a commonly available resource. 5MW should come online in 2025 with plans to increase to 20MW.

     Higher-than-normal temperature gradients can indicate faults or convection. Normal temperature gradients = conduction area. In NW Utah horst rocks outcrop nearby, so geology can be better understood, especially fracture patterns and permeability. EGS project developers need to understand temperature, lithology, stress, and geochemistry. Temperature is always the main driving factor. Lithology requires brittle rock. There is a need to understand the fracture network and its extent to get connectivity. Stress is explored to keep fractures open. Geochemical rock-fluid interactions need to be understood. Precipitation minerals can be a risk for scaling. The FORGE project drilled to 230 °C temperatures in granites. Fervo is getting better than expected well flows. Stimulating parallel to the natural fracture system was employed so that the natural fractures work with the induced fracture system.

 

Presentation 3 – Lithologies, Reservoir Architecture, Fluid Flow & Thermal Behavior; Presented by John Holbrook, Ph.D., TCU

     Holbrook first notes that “ultimately we do plumbing,” referring to both oil & gas and geothermal. The typical petro workflow includes facies distribution models and, structural model, and the use of sequence stratigraphy used for facies mapping. There is heterogeneity in reservoirs such as changes within the rock that create flow baffles and flow barriers, which can be complex. We can see facies changes in outcrops. Connectivity is very important to successful geothermal. Models are also used to figure out why things went wrong with projects.

     He explores how a reservoir is built. The notion of migrating ripples on the shallow sea floor exemplifies that much deposition the rock record gets destroyed. Example: The Grand Canyon = only one-billionth of continuous deposition preserved.

     The Stratigraphic Machine:  What you are looking at is a statistical sampling of the past, not the past. Some preserve Milankovitch cycles (20,000-44,000 year cycles). Tidalites show yearly tides but there can be breaks in the records. Reservoir analysis is really about sampling spatial-temporal statistics = rock record, but not like a tape recorder. Turbidites are only preserved about every thousand years. The rock record is remarkably incomplete, with only a small fraction preserved. Sedimentation happens at different rates, not at all, or is eroded. There is hierarchy and completeness in the record, orders of hierarchy of sedimentary bodies. Time scales are important to know. A bunch of small pieces of pieces of pieces preserved in the rock record make reservoirs complex. They can exhibit Fractal qualities– big pieces look like smaller pieces. This can lead to mistakes in scale due to mistakenly identifying the scale similarity and the conundrum of scale = fractal effect, which occurs in channels, fluvial rocks, and fans = a different order of hierarchy. Valleys, belts, and channels can look the same but differ by scale. A complete section with no unconformities is typically just a millionth preserved. The Sadler Effect: thicker section = more time. Shorter section measured = less time has passed. Projecting modern into ancient: Look for modern analogues. Why do snapshot samples work?  Some may be very distinctive to certain interpretations. Fossils can confirm a marine environment, for example. Geologists must use deductive logic. Paleocurrents can be derived from cross-bedding sets to recreate the past. The Autogenic vs. the Allogenic: Autocyclicity is what actually makes the stratigraphy. Allocyclic = sea level change, tectonics = change from outside. Allogenic changes influence and alter autogenic processes. Sea level, climate, and tectonics are the biggest allogenic influences. Sea level changes are not different in modern times than in the deep geological past. Not all allogenic influences will generate signals in the sediments, but once in a while, they do. Allogenic changes can also be local, such as faults affecting meanders in the New Madrid fault zone area. Each “piece” in the pieces in pieces in pieces is a potential source of reservoir heterogeneity. This can really affect injector issues and connectivity issues. Injectivity and production must be connected via a useful path with the right flow. Temperature has heterogeneity, too. Heat depletes along flow paths. The way fluid runs through sediments generates heat, which is what is being mined. Geothermal project developers need to know how long to recharge after extraction. Flowing hot water faster when projects are turned on affects diagenetic processes and causes problems, including precipitates and blocked flow paths.

 

Presentation 4 - Fracture Detection, Mapping, and Characterization in Geothermal Reservoirs; Presented by Jon McKenna, Ph.D., MicroSeismic

     Microseismic can be used for geothermal as it is for oil & gas (and recently for CCS). Focal Mechanism Solutions: one wants to know strike, dip, and rake for stress fractures, which may be compressive or tensional. AutoMTI = automatic moment tensor inversion. Rake is in the direction of strike (strike-slip) or dip (dip-slip). 3D Fracture Propagation Analysis can be done with microseismic. Microseismic time/space propagation can be compared to AutoMTI to make sure active nodal planes are correct. 3D lineaments for both fracture propping and AutoMTI are plotted on a stereonet and compared. We want to know if fractures are opening more. Rigidity and k factor (modeled vol smaller than injected vol since not all events detectable, thus an educated fudge factor is employed) are important for volume changes measured from seismicity. Shear modulus can be derived from well logs in the zone where fractures are encountered. Aperture of fractures can be estimated by comparing to outcrops, measured w/ microseismic, then compared. In the example given, the dip-slips are opening more than the strike slips. Leakoff must be subtracted from the slurry injected volume. Microseismic permits evaluating interaction between wells, known as parent and child “frac hits.” Hydrocarbon and water production creates a pressure sink around that wellbore, which causes the child well to grow toward the parent well – ie, leaks off into the parent well, so it is lost. Fluid will move from high pressure to low pressure, leading to frac hits. One can draw an analogy between geothermal injection and producing wells; both are treated. The first well must be re-pressurized as needed. The need is to create adequate permeability for both wells. Stress shadows can be seen with microseismic, and the next stage after it will go deeper than where the stress shadow occurs. See slides. Finally, a stimulated reservoir volume (SRV) can be calculated. Colors show enhanced permeability magnitudes = open void volume.

     Two horizontal wells, one producer and one injector, are a common configuration for EGS. There is lower microseismicity and fewer fractures at the toe and more at the heel, typically. There is a greater amount of frac volume from toe to heel. Stresses can rotate slightly, and dip slip can change to strike slip. Stress rotation plus changes in stress magnitude occur.  Stress shadow impact on neighboring wells is time-dependent on stage lag time, as known from multiple well fracturing (zipper, etc). If you wait till the stress shadow goes away, then it may not effectively treat the shallower well since the energy will not go to the other well. The goal is to know where the proppant is going. There is a need to determine appropriate well spacing and stage spacing. A lower well is fracked first because it needs the stress shadow to keep propagation toward it from the shallower well treated afterwards. Simultaneous fracturing might be best, or to re-pressurize the first deeper well to keep it from receiving leak off. Avoid blowing out the heel and not stimulating the toe enough. Geothermal example from Nevada: fluid flow is primarily deep and along the fault system. Stress shadows are temporary and then return to virgin pressure.

 

Presentation 5 - Geological Considerations in Converting an Oil Well to a Geothermal Well; Presented by Steve Tedesco, Ph.D., Running Foxes Petroleum

     Tedesco first glosses over the history of geothermal. He notes that corrosion and metal precipitation are causing conventional geothermal to decline. There is also a need for adequate fracture networks to allow high enough water flow volumes. He notes that 22% of geothermal wells fail in the drilling stage.

      There are some ongoing oil & gas well projects using orphan wells to repurpose for geothermal. He mentions a project in Oklahoma that intends to provide geothermal energy for a school and an old Amoco field in Nevada, where wells are being converted to geothermal. 11 O&G wells in CA are being converted into GT energy storage. Some Austin Chalk wells are being converted in Texas. The Osage Nation's project to power greenhouses with geothermal is ongoing. Wyoming and North Dakota have projects. France has a project. Map shows pilots GT from O&G fields. Only a small section of orphan wells are applicable to conversion. In general, they need to be deeper and have casing integrity and adequate casing size (above 5.5” pipe). They also need to be near an end-user. Probably a lot less than 50,000 wells are applicable. Marginal wells may be applicable for conversion, especially if they were strong water producers.

     High-temperature geothermal >150C °C, med T 90- 150 °C, Low T 30-90 °C.

     There are very few orphaned horizontal wells, and no projects are known to convert them. Conversion requires injecting into the tubing and putting a heat exchanger in. Two vertical wells can be paired as an injector/producer as in a waterflood. Tight rock must be fracked, and most often ruins economics. A binary cycle is the most efficient way to do geothermal in low-to-medium temperatures. High-porosity wells are needed. The heat can be used for power generation (rarely) or direct use. Precipitates/scale, H2S, and corrosion must be monitored. Oil and gas output must be managed in converted wells. Seismic activity risk is applicable to the US West, though generally not in sedimentary basins. Site evaluation/feasibility is important. Existing data before 1980 is hard to find. Regulatory approval is a risk. Reconditioning wells is often needed, such as new casing and tubing. Newer wells are obviously best. Wells too far from transmission lines are not a good fit for power production. General Geologic and Engineering Criteria slide. One might need to run more logs and collect fluid samples. BHT > 70 deg C. GT Well Bore options slide. Subsequent slides – well bore checklist. States have variable rules and incentives. IRA provides 30% tax credits. States have them too, but vary. Mississippian Chat wells in Oklahoma are applicable because of their very high water production. In some cases, geothermal production could damage existing O&G production, so this must be considered.

 

 

Presentation 6 - Offshore Geothermal Energy; Presented by Joseph Batir, Ph.D., Teverra

     Site selection & flow rates are key parameters for evaluating offshore geothermal. Galveston Bay, near shore/offshore Texas, has the least regulatory red tape. One well evaluated had a very high maximum flow rate of 200,000 Bbls/day and a 219 °C temperature. 5-10% of wells have suitable temperature and flow rates. 40 gpm is the minimum flow rate considered. Thus, about 46 wells are suitable throughout the Gulf Coast. Oil wells may water out eventually, making them suitable for geothermal later. Reservoir temperatures vary between 270-350°F. F. 10,000-20,000 Bbls per day flow rate is preferred.

     Organic Rankine Cycles (ORCs) make the most sense for the conversion of offshore wells. The average power production potential is 80-300 kW, which would require 2 modular ORC systems. Larger wells with a maximum power production potential of 550-3000kW could use one larger optimized ORC.

     Economic models for power production indicate a 20-year payback at $2-$5 million/MW for installation. If a project can be economical without tax incentives, then it is doable. The cost of avoided electricity depends on location. Offshore power needs are in an islanded situation, and any geothermal power would offset natural gas power. Repurposing requires enough power production to be economical. Drilling new wells would have better power production, but would be vastly uneconomical. 50MW would be needed for economics, similar to the cost for well and offshore platform decommissioning. 50MW wells occur in California, Italy, Iceland, and other areas with hotspots. Thus, it would require super-hot rocks, which are not available offshore the Gulf Coast. Co-production could make it economical in the future. Maybe horizontal drilling will improve economics. Offshore geothermal is not yet viable due to economics, but may be later. A benefit would be extending the life of existing assets. Offshore energy storage could be a future use. Offshore geothermal proximal to offshore wind could be viable, as there is a use case for offshore geothermal for platform energy needs. Offshore wind does have a tie-back to the onshore grid. Geothermal can add some value to wind by providing some firm power. The economic models given do not assume production decline not fluid cooling, both of which are likely to occur over time. He notes that all geothermal systems require a high level of reservoir management, primarily pressure support.

 

Presentation 7 - AI for Candidate and Site Selection using Subsurface Data and Reservoir Characterization in a Map Augmented Framework; Presented by Patrick Ng, Shaleforce

     Risk assessment for geothermal is very important. Ng goes over Techno-economic analysis (TEA) for oil & gas: First, get data, seismic, and identify prospects. For geothermal, first do gradient mapping. He mentions giving AI 3 tasks: 1) First Play, explore ideas, de-risk prospect – ask Chat GPT questions, then more specific questions. Gut check. California benefited from a high GT gradient. Often, there are 2 injectors and 2 producers.

     NREL model Geophires is an advanced modeling tool for GT configuration modeling. Inputs include depth (say 12,000 ft/4km), 10-12,000ft often avoids deeper geopressured reservoirs. A project can generate breakeven electricity costs or breakeven costs in $ per MMBTU. In comparing to natural gas, one should compare it to natural gas with carbon capture, as that is likely to be a future trend and would make geothermal more competitive. De-risking = street due diligence. For geothermal TEA, user input is used to derive the predictive model response. Geothermal-GPT proceeds from user input to prompt template to large language model (LLM) to response. Retrofit opportunities for geothermal can utilize AI-enhanced well production analysis. Reservoir data analytics can involve using well EURs to estimate geothermal storage capacity. This is easier with oil, also a liquid, than with gas. Slide from Chapter 2 of Texas GT report.

     Geothermal heat is a very cost-effective decarbonization method for industries, and direct use aids economics. He mentions energy transition management strategies to 2035, where wells can be used for geothermal energy storage. Chat GPT can be trained for geothermal, along with data analysis improve geothermal success rates. Mispriced assets = misappraised risk. Initial production (IP) is a good proxy for return. See slide. Risk tolerance involves comparing geothermal to other energy forms, and in light of decarbonization scenarios. He thinks we can deliver an American geothermal Renaissance by 2030.