I’m out of superlatives – I used them all up in the title. But seriously – Enhanced Geothermal System (EGS) projects have had a really, really good summer. In this article, I summarize the results that have been recently presented by Fervo and FORGE.

Fervo

At their annual Tech Day and in a white paper posted this week (Norbeck et al., 2024), Fervo Energy provided their first update on Project Cape, a Utah project where they are developing 400 MWe of new production over the next two years. So far, fourteen wells have been drilled, and three of them have been stimulated. Last month, they performed a 30-day flow test with two injectors bounding one producer. These wells are roughly 450 ft apart laterally, and they are also offset vertically. Fervo’s configuration is still not quite a full-scale test – in the fully developed system, there will be a series of wells, alternating producer/injector, so that nearly all wells are producing/injecting from both directions.

The information below is drawn from the Fervo white paper (Norbeck et al., 2024) and from their presentations at their Tech Day. Here is an excellent writeup of the Tech Day from Mike Matson with BCG.

• The 14 wells at Project Cape were drilled with 100% success rate. Cost was 18% below budget. 80 fracturing stages were completed, and 95% of these stages placed the designed proppant volume.
• Average drilling rate along the lateral was 39.4 ft/hour in the first well. ROP improved incrementally, and by the fourteenth well, it had reached 64.2 ft/hour. For the final well, spud to total depth was 20 days.
• During the circulation test, the production rate of the well exceeded 95 kg/s. Fervo projects that at this production rate and temperature (383’F and still rising), the well will yield greater than 9.5 MWe gross when it is hooked up to a power plant (which is now being built). Keep in mind that there will be a few MWe of parasitic loss due to pumping requirements of the injectors. The injection WHP was held at 2000-2300 psi (while keeping BHP at least 1000 psi below the magnitude of Shmin). Prior to Fervo’s Project Red last year, the highest production rate ever achieved from an EGS project was 30 kg/s (at Soultz), after more than 50 years of effort and dozens of projects around the world (Norbeck et al., 2023). Fervo has now set flow rate records in each of their two production wells, and along with FORGE’s results (below), it is proven that very high flow rates are consistently reproducible.
• Fervo did not report injection rates, but from the flow configuration, I would expect the system experienced significant net fluid loss to the formation (ie, injection volume greater than production volume) This is likely because the two injectors are unbounded on the outside, with the producer on the inside. In a full-scale test, with an alternating array of wells, we would expect much higher recovery efficiency, because most of the injectors and producers would be bounded on both sides. As noted below, FORGE – using an over/under well configuration – reported a high fluid recovery.
• Induced seismic events during stimulation occasionally exceeded magnitude 2.0, reaching the ‘yellow’ range of the Fervo traffic light system. No events exceeded magnitude 3.0, reaching the ‘red’ range, which would have triggered a 24 hour pause in operations.
• Fervo is building modular, standardized ORC power plants, each with a capacity of 50 MWe. In conventional geothermal, wells must be widely spaced, and each well requires an average of around half a mile of pipe to transport water/steam to the power plant. Fervo’s designs allow them to perform ‘pad’ drilling, where many wells can be drilled in close proximity, with directional drilling used to space them out in the subsurface. As a result, they can place the power plants directly adjacent to the wells, reducing surface CAPEX by 22%.
• Fervo reports that as they grow, the ‘cost of capital’ will be a significant factor for their economic performance. They are finding that with compounding success, they are able to access cheaper forms of capital – such as debt, rather than equity – which is valuable for their long-term growth.
• At Fervo’s Project Red well, which was stimulated last year, the production well has been producing steadily and producing into a power plant – selling electricity onto the grid – for the past 10 months. No thermal decline has been observed.
• Currently, the laterals are 5000 ft. In shale, laterals are usually 10,000-15,000 ft. Fervo plans to drill longer, and larger diameter, laterals in the future, which will further increase production per well.
• Fervo reports that their CAPEX is now $5000/kWe. With economies of scale, they expect to reach \$3000/kWe within a few years. This is a baseload, zero-emission source of power. It is produced domestically. The resource is huge, with essentially unlimited ability to scale (although, it is not practical in all regions of the US). Considering all of that, these cost numbers are seriously heady stuff. There’s not a lot of ‘projection’ happening here – we are talking about real wells, which have already been drilled, and which are already producing.

This time next year, Fervo will have stimulated and circulated many more wells, and we’ll see the performance of these designs at larger scale.

 

FORGE

The FORGE project is operating at a site very close to the Fervo Project Cape wells.

First, some background on the operations prior to 2024. In 2022, FORGE performed three proppantless fracturing stages towards the toe of the 16(a) well – which is a highly deviated lateral (Xing et al., 2023). In 2023, a second well was drilled directly above the 16(a), and circulation tests were performed. The circulation tests showed only weak stimulation from the proppantless stimulations. The injection rate into 16(a) was low until they decided to inject with BHP above Shmin. This causes mechanical opening of fractures – essentially, a hydraulic fracturing injection – which is a procedure that enables high-rate injection into a well that would otherwise have low injectivity. Injecting above Shmin was acceptable for purposes of the test, but injecting above Shmin would not be realistic for practical long-term operation (at least, this is the conventional wisdom – Frash et al., 2024, have argued otherwise). Even with this very high- pressure injection forcing a high injection rate, the production rate of the 16(b) – only about 170 ft above – was very low, less than 0.5 kg/s (Xing et al., 2024).

In 2024, FORGE stimulated seven new stages in 16(a), along with restimulating the original three stages. In these stages, proppant was used, along with limited-entry plug and perf. This was a much more conventional ‘oil and gas’ style stimulation treatment, similar to the type of stimulation that Fervo has been using. The 16(b) well was also stimulated with several stages. The injection volumes in 16(b) were smaller than the 16(a) treatments; they were used primarily to ensure that the 16(a) fractures connected into the 16(b) perforation clusters. Each stage used a somewhat different design, with varying clusters per stage, cluster spacing, and fluid and proppant type. The design and results from the 2024 stimulations are summarized in the most recent FORGE annual report (Utah FORGE, 2024).

Overall, roughly 1000 ft of lateral have been stimulated along the 16(a). This is about 5x less than the Fervo wells (which themselves, are 2-3x shorter than typical laterals in shale).

In last week’s monthly FORGE seminar, John McLennan and Joe Moore summarized the results from a one-month flow test between the two wells, which was performed in August (McLennan and Moore, 2024). They maintained the injection wellhead pressure at 2800 psi, below Shmin (and moderately higher than used by Fervo in their neighboring test). Injection was performed at about 26 kg/s. The production rate steadily rose during the test, reaching a production rate around 23 kg/s within a few days, and with production rate reaching near-parity with the injection rate by the end of the 30 days.

The FORGE circulation rate is roughly 4x lower than reported by Fervo’s Project Cape producer. But it was achieved from 5x less stimulated lateral than the Project Cape wells, and 3x less stimulated lateral than the Project Red wells. Overall, the FORGE results corroborate the results from Fervo’s Project Cape and Project Red – that when we do multistage hydraulic fracturing with plug and perf and proppant, we can achieve high flow rates per well. The three projects achieved in the range of 0.015-0.03 kg/s per ft of lateral.

During the circulation, a spinner log was run in the 16(a) well (the injector). It showed outflow from the well along every stage (except two stages in the middle where it was not possible to place proppant, as discussed by Utah FORGE, 2024). Within stages, flow rate moderately uniform on a ‘per cluster’ basis.

The next steps for the FORGE project are being planned now. An immediate next step will be for the various parties involved in data collection to complete their analysis and report their results. FORGE performed an enormous amount of high-quality data collection. They placed fiber strain and temperature sensors in the 16(b) well; they had a high quality microseismic array; they ran tracers; and they even cored the 16(b) well while drilling through the SRV created by the first three stages of the 16(a). The field of geothermal reservoir engineering is going to learn a lot from this project. Stay tuned as the various investigators report their results over the next 6-12 months.

 

Were there any surprises?

At high level, everything performed as advertised. Personally, this is satisfying for me, because I’ve spent the past 10+ years telling anyone who would listen that multistage hydraulic fracturing with proppant for EGS is technically feasible, will deliver “dramatically improved economic performance relative to current designs” (Shiozawa and McClure, 2014), and “could transform the geothermal industry” (McClure, 2021). I’ve written papers explaining why these designs would be effective and why EGS needed to move away from designs based on ‘shear stimulating natural fractures.’ In many, perhaps most, historic EGS projects, the evidence shows that shear stimulation has been weak or non-existent, and in cases when it has worked, it has not yielded adequate reservoir characteristics, either for flow rate or for fracture spacing (McClure and Horne, 2014a, b; Li et al., 2016; Norbeck et al., 2018; McClure et al., 2022). Most concretely, the ResFrac simulator has been used by Fervo since 2018, and in particular, it was used to design the stimulation treatments and space the wells in Project Red and Project Cape. The wells are doing what we said they would do!

But yes, there have been some things that we did not anticipate. A preliminary analysis of the fiber and microseismic data from FORGE suggests that during some of the fracturing stages, flow probably diverted into a fault. Two of the stages at FORGE experienced high injection pressure and were unable to pump the designed volume of fluid and proppant.

For the Stanford Geothermal Workshop in February, we at ResFrac will be putting together an updated FORGE model, incorporating all this new data. A takeaway for me – we will probably need to incorporate at least one dipping natural fault zone into the model. This was not included in the FORGE model that we built prior to the stimulations (McClure et al., 2024).

 

What should we worry about next?

Five years ago, the top perceived risks for this technology were: (a) flow rate, (b) technical ability to drill and stimulate the wells, (c) induced seismicity, and (d) thermal drawdown, in that order. How are these risks perceived today?

• Today, flow rate has faded as a concern – it’s been consistently shown that EGS wells with multistage fracturing, proppant, and limited-entry completion can achieve high flow rate. In future tests, projects are likely to incrementally increase well spacing, in order to determine how apart the wells can be spaced without significantly degrading the connection. Eventually, they’ll find a breakover point where circulation rate diminishes. So far, it hasn’t been reached.
• Technical ability to drill and stimulate the wells – they’ve been hitting this out of the park. The teams at both FORGE and Fervo have dramatically improved on the state-of-the-art for drilling geothermal wells, consistently executed successful plug and perf stimulation, and generally executed on first-of-their-kind projects with high reliability. They’ve exceeded everyone’s expectations.
• Induced seismicity – results have been solid. Neither FORGE nor Fervo has reported a ‘red light’ event magnitude of 3.0, which would trigger a 24 hour cessation of activities. However, there have been several events in the ‘yellow’ range above 2.0. Keep in mind – these are conservative thresholds, it would require a much larger event than magnitude 3.0 to cause felt seismicity in the nearest town, Milford (Norbeck et al., 2023). As expected, seismicity has occurred during stimulations and has been limited during the circulations. Fervo and FORGE have operated in a very responsible manner. As other companies start to imitate their designs, it will be imperative that they follow best practices for induced seismicity hazard assessment and mitigation, as outlined by Majer et al. (2012). Regulatory agencies should make this a requirement.
• Thermal drawdown – it will take 10-15 years to know with certainty, but the early results are positive. Spinner logs confirm that flow is reasonably uniformly distributed across the laterals. This is the most important parameter that determines thermal longevity.

Conventional EGS designs – with a single stage and no proppant – have achieved much lower flow rates than the new designs. Also, with these older-style designs, flow localized into a small number of dominant natural fracture pathways (Baria et al., 2004; review from Li et al., 2016). Conversely, Fervo and FORGE are showing that when creating propped fractures through plug and perf completion, that process is able to defeat flow localization and create a much more pervasive, uniform distribution of fracture conductivity.

The FORGE wells are spaced quite close together – 170 ft. At that spacing, with 28 kg/s over 1000 ft of lateral – the production well will be likely to experience significant cooling within a few years. The Fervo Project Cape wells are spaced significantly further – 450 laterally and also some vertical separation. As long as flow is reasonably uniform, these wells will be productive for many years.

Nevertheless, thermal longevity remains – in my opinion – the biggest future risk for EGS. It will be a top priority for future R&D on EGS. It is inevitable that eventually, the wells will start to cool down. When that happens, technologies that improve uniformity and divert flow from cold pathways will have strong, direct impact on ROI.

There are at least two general strategies that could be adopted to reduce flow localization and improve thermal longevity. First, fluid additives could be pumped to adaptively plug colder short-circuit pathways. These concepts have been explored by investigators such as Hawkins et al. (2024). Second, devices may be placed in the wells themselves, designed to prevent excessive flow rate from any particular location, similar to how flow conformance devices are used in SAGD (Park et al., 2017). I’m itching for someone to do a ResFrac modeling studies on these topics, and so if you’re a grad student or researcher looking for a research topic – consider digging into one of these.

The last resort would be to pump a cement squeeze into a section of the well that hosts the short circuit. This would likely be effective, but would be invasive and expensive, and so we would like to avoid it if possible.

Simulations suggest that thermal stress changes from cooling could have a significant positive or negative effect on long-term longevity (McClure, 2023). This is an intriguing area that deserves more study. Especially, we need to research how this process may interact (positively or negative) with engineering solutions designed to improve flow uniformity.

 

References

Baria, R., Michelet, S., Baumgärtner, J., Dyer, B., Gerard, A., Nicholls, J., Hettkamp, T.,Teza, D., Soma, N., Asanuma, H., Garnish, J., Megel, T., 2004. Microseismic monitoring of the world’s largest potential HDR reservoir. Paper Presented at the Twenty-Ninth Workshop on Geothermal Reservoir Engineering, Stanford University.

Frash, Luke P., Meng Meng, and Bijay KC. 2024. Greenfield Reservoir Engineering for the Wattenberg Field with Comparison of Advanced, Enhanced, and Caged Geothermal Systems. Proceedings, 49^th Workshop on Geothermal Reservoir Engineering, Stanford, CA.

Hawkins, Adam J., Danni Tang, Aaron M. Baxter, Reeby Puthur, Daniel T. Korzukhin, Zach J. Zody, Bryan H. Abdulaziz, Patrick M. Fulton, Sarah Hormozi, Chris A. Alabi, Ulrich B. Wiesner and Jefferson W. Tester. 2024. Active Tracers for Hydraulic Control of Cooled Short Circuits: Bench-Scale Demonstration and Numerical Simulation. Proceedings, 49^th Workshop on Geothermal Reservoir Engineering, Stanford, CA.

Li, Tianyu, Sogo Shiozawa, and Mark W. McClure. 2026. Thermal breakthrough calculations to optimize design of a multiple-stage Enhanced Geothermal System. Geothermics 64, 455-465.

Majer, Ernie, James Nelson, Ann Robertson-Tait, Jean Savy, and Ivan Wong. 2012. Protocol for Addressing Induced Seismicity Associated with Enhanced Geothermal Systems. Prepared for the Geothermal Technologies Program, United States Department of Energy.

Matson, Mike. 2024. Fervo Energy Technology Day 2024: Entering the “the Geothermal Decade” with Next-Generation Geothermal Energy. LinkedIn Article.

McClure, Mark W., Roland N. Horne. 2014a. An Investigation of Stimulation Mechanism in Enhanced Geothermal Systems. International Journal for Rock Mechanics and Mining Sciences 72: 242-260.

McClure, Mark, and Roland Horne. 2014b. Characterizing hydraulic fracturing with a tendency-for-shear-stimulation test. SPE Reservoir Evaluation & Engineering 17, 233-243.

McClure, Mark. 2021. Why Multistage Stimulation Could Transform the Geothermal Industry. Guest Editorial, Journal of Petroleum Technology, October 2021.

McClure, Mark, Charles Kang, and Garrett Fowler. 2022. Optimization and Design of Next-Generation Geothermal Systems Created by Multistage Hydraulic Fracturing. SPE-209186-MS. Paper presented at the SPE Hydraulic Fracturing Technology Conference and Exhibition, The Woodlands, TX.

McClure, Mark. 2023. Thermoelastic fracturing and buoyancy-driven convection: Surprising sources of longevity for EGS circulation.

McClure, Mark W., Rohan Irvin, Kevin England, and John McLennan. 2024. Numerical Modeling of Hydraulic Stimulation and Long-Term Fluid Circulation at the Utah FORGE Project. Proceedings, 49^th Workshop on Geothermal Reservoir Engineering, Stanford, CA.

McLennan, John, and Joe Moore. 2024. FORGE Frontier Observatory for Research in Geothermal Energy. Utah FORGE Monthly Team Meeting Presentation. September 4, 2024.

Norbeck, Jack H., Mark W. McClure, and Roland N. Horne. 2018. Field observations at the Fenton Hill enhanced geothermal system test site support mixed-mechanism stimulation. Geothermics 74, 135-149.

Norbeck, Jack H., and Timothy M. Latimer. 2023. Commercial-scale demonstration of a first-of-a-kind Enhanced Geothermal System. EarthArXiv.

Norbeck, Jack Hunter, Christian Gradl, and Timothy Latimer. 2024. Deployment of Enhanced Geothermal System technology leads to rapid cost reductions and performance improvements. EarthArxiv.

Park, Sang-Yeop, Delon Saks, Venki Laksmanan. Anthony Singh, and Michael Ma. 2017. Flow Control Devices for SAGD Applications: Lessons Learned, Best Practices, and Suggested Design Improvements. Paper SPE-188149-MS presented at the SPE Thermal Well Integrity and Design Symposium, Banff, Alberta, Canada.

Shiozawa, Sogo, Mark McClure. 2014. EGS Designs with Horizontal Wells, Multiple Stages, and Proppant. Paper presented at the Thirty-Ninth Workshop on Geothermal Reservoir Engineering, Stanford, CA.

Utah FORGE. 2024. Phase 3B Year 2 Annual Report: Enhanced Geothermal System Testing and Development at the Milford, Utah FORGE Site. Prepared for the US Department of Energy Office of Energy Efficiency and Renewable Energy. June 1, 2024.

Xing, Pengju, Branko Damjanac, Zorica Radakovic-Guzina, Maurilio Torres, Aleta Finnila, Robert Podgorney, Joseph Moore, and John McLennan. 2023. Comparison of Modeling Results with Data Recorded During Field Stimulations at Utah FORGE Site. Proceedings, 48^th Workshop on Geothermal Reservoir Engineering, Stanford, CA.

Xing, Pengju, Kevin England, Joseph Moore, Robert Podgorney, and John McLennan. 2024. Analysis of circulation tests and well connections at Utah FORGE. Proceedings, 49^th Workshop on Geothermal Reservoir Engineering, Stanford, CA.