Enhanced Geothermal Systems promise dispatchable zero carbon power

Geothermal energy is the Earth’s own internal heat. It’s a huge potential resource, but so far it’s seen only very limited use. Traditional geothermal power can only work where there are naturally existing hydrothermal systems that bring the heat of the interior to the surface. A new technique called enhanced (or engineered) geothermal systems (EGS) may make geothermal power much more widely available. If it can be scaled up commercially, EGS will enable us to create hydrothermal systems anywhere there’s hot rock not too deeply buried — which includes a large swath of Colorado. This is potentially significant in the context of creating a zero-carbon electrical system because like hydroelectricity, and unlike wind and solar, geothermal power can be dispatchable: you can turn it on and off at will. This makes it a great complement to intermittent renewable power, as it can be used to fill in the gaps then the wind’s not blowing or the sun’s not shining.  It remains to be seen whether it’s technically feasible, and if so at what price, and on what timeline, but it’s certainly worth investigating.

What is geothermal power?

There are three main applications of geothermal energy in use today. The first is direct extraction of heat from naturally occurring high temperature, near-surface hydrothermal systems, like the ones that create hot springs. This type of resource can be used to provide space heating and hot water, as is done widely in Iceland. Globally there’s roughly 100 GW worth of this type of direct thermal extraction in use. Hydrothermal systems can also be used to generate electricity with steam turbines, as is done at The Geysers power plant in Northern California, which has been feeding electricity into the grid since the 1960s. Today The Geysers puts out an average of about 1 GW of electricity, making it the biggest geothermal electric station in the world — it’s the same size as a large coal or nuclear power plant — capable of powering a city the size of San Francisco. However, high quality hydrothermal systems located close to major electricity demand centers are rare. As a result, in 2012 there were only about 11 GW of geothermal generating capacity installed worldwide.

The other common application of geothermal energy is more recent. Geothermal (or ground source) heat pumps use a little bit of electricity to move a lot of heat from one place to another. They can suck heat out of your home in the summer and dump it into the ground, and then pull the same heat back out of the ground in the winter to keep your house warm. Geothermal heat pumps work just about anywhere, and while they’re relatively expensive to set up today, they are one of the most energy efficient ways to do active heating and cooling. Heat pumps don’t generate electricity, they consume it, but you get much more heat out than you put electricity in, so if what you really want is heating or cooling, it can be a very good trade off.

However, geothermal heat pumps aren’t really using geothermal energy. The near-surface heat they’re extracting is ultimately solar in origin — a couple of meters underground, the Earth’s temperature is determined by the average annual surface temperature, which is a function of the local climate. If, on the other hand, you drill a few kilometers down, you find that the Earth is very hot just about everywhere. This is the Earth’s own energy, resulting mostly from the decay of radioactive nuclides throughout the planet’s rocky mantle. This heat diffuses up to the surface very, very slowly — on land, the total heat flux only averages 0.065 W/m^2, which is a tiny fraction of the solar power that falls on the same area — but within a few kilometers of the surface, an enormous amount of heat has accumulated over geologic time. It’s not a lack of heat that keeps geothermal energy from being widely utilized by humans, it’s the difficulty of bringing that heat to the surface cost effectively.

Natural hydrothermal systems only form where fluid can flow through the hot rock to the surface relatively easily — via the tiny spaces between mineral grains and through faults and fractures. If you don’t have hot rock near the surface, or there’s no fluid, or the fluid can’t flow, then you don’t get hot springs and geysers, and you can’t develop geothermal electricity. That combination of circumstances is rare, so people are trying to develop techniques for creating artificial hydrothermal systems in places with hot rock, but low permeability and a lack of natural hydrothermal fluids.

How could we build a hydrothermal system?

To create a commercially viable hydrothermal system from scratch, first you need to find a thermal reservoir — hot rock, preferably not too deeply buried. A well is sunk into the reservoir and cold water is pumped in to “stimulate” it. This means opening up a network of small fractures within the hot, dry rock that allows injected water to circulate and absorb heat. For commercial power production at the tens of MW scale, the fractured reservoir needs to be at least several cubic kilometers in volume. This may involve several rounds of drilling and stimulation at different depths. The creation of the reservoir can be monitored from the surface with a seismic network, which senses the tiny tremors and vibrations that occur as fractures open up in the reservoir deep below the surface.

Once the reservoir has been established, several production wells are sunk into its periphery. Water is pumped down the injection well, circulates through the hot rock, and is brought to the surface via the production wells. Depending on the temperature of the produced water, it can be used in several ways: very hot “dry” steam can be fed directly into a steam turbine. Pressurized water above about 180°C can be rapidly vaporized (“flashed”) to steam at surface pressures, and then fed into a turbine. Lower temperature fluids are generally used to flash a secondary working fluid with a lower boiling point than water, and that vaporized working fluid is then sent through the turbine, in what’s called a binary cycle power plant.  The hotter the resource, the more efficient the process is overall — the larger a proportion of the heat is ultimately turned into electricity.

Why is EGS potentially important?

Based on NREL’s Renewable Energy Futures 2050 study, and others, it seems clear that we can build a vastly lower carbon electrical grid using only technologies that are available today.  Investments in better regional transmission, energy efficiency, demand side management strategies, distributed generation, and a diverse portfolio of intermittent and forecastable renewable resources (both wind and solar, spread out geographically) can go a long way toward reducing our reliance on fossil fuel baseload (always on) and dispatchable (on demand) generation.  However, it appears likely that we will still need substantial dispatchable capacity (or large scale electricity storage) to fill in the occasional gaps in supply.  These generation resources would be operated like today’s load-following and peaker plants, except that instead of just following daily variations in electricity demand, they would also compensate for variability in supply from wind and solar resources.

The USGS estimated in 2008 that there’s about 500 GW worth of potential EGS electric capacity in the western US.  This is similar to the combined coal, natural gas, and storage capacity required in NREL’s 80-90% renewable energy 2050 portfolios (see the above charts).  This suggests that if we can commercialize EGS, we may be able to turn those 80-90% RE scenarios into something much closer to zero carbon power systems, without having to rely on the development of large scale electricity storage.

Where is EGS today?

EGS sounds great, but there’s a good reason NREL’s Renewable Energy Futures study left it out.  Though the concept has been around since the 1970s, when a pilot project was initiated at Los Alamos National Labs, today there’s only one true EGS facility in the world generating electricity.  It’s a government sponsored operation in Soultz, France that started in 1987 and that has just in the last few years started feeding about 1.5MW into the EU’s grid.  They now face the challenge of scaling the plant up to a more commercially relevant 25 MW, which should prove financially (if not technically) easier given a recently implemented €0.20/kWh French feed-in-tariff for this type of generation.  Another project called Habañero Power in central Australia’s Cooper Basin is privately funded, and managed by a company called Geodynamics.  It hasn’t started generating electricity, but it’s successfully brought substantial steam to the surface at a temperature of nearly 200°C. Unfortunately the Habañero wells are far removed from any electrical load centers, meaning substantial transmission would be required to commercialize the project. With Australia investing aggressively in rooftop solar and energy efficiency, their grid — not unlike Colorado’s — is currently over-supplied with electricity, so there will be little economic incentive for new generation facilities, absent a clear long-term commitment to de-carbonization and the early retirement of their fossil fueled generation.

In the US, a decade passed between the final shutdown of Los Alamos EGS project in 1996, and a major study of US geothermal resources spearheaded by MIT in 2006.  The DoE funded report was published as The Future of Geothermal Energy (PDF) in 2007, and mapped out the thermal resources in the US at depths from 3-10 km. The authors suggested that with a modest commitment of R&D funds ($1 billion over 15 years) it ought to be possible to develop ~100 GW of EGS based electricity by 2050. The DoE subsequently did its own independent assessment of the technologies required to scale EGS up commercially (see the Evaluation of Enhanced Geothermal Systems Technology, 2008) and examined the assumptions underlying the MIT study.  DoE agreed with the MIT study’s assessment of the overall thermal resources, but was skeptical about the ease with which they might be commercially developed.  The Geothermal Technologies Program (GTP) subsequently adopted their own high level goals for EGS: of demonstrating the creation of a 5MW reservoir by 2020, and lowering the levelized cost of electricity (LCOE) for EGS to $0.06/kWh by 2030.

Since 2009 the program has funded technological development and EGS dmonstration projects across the US, looking to adapt existing technology from the oil and gas and traditional geothermal industries, to develop new techniques where our current abilities are lacking, and to gain valuable field experience (see their 2009-2015 Multi-Year Research, Development and Demonstration Plan).  The most prominent US EGS demonstration projects are one in central Oregon at a site called Newberry Volcano (which has a fairly active blog) run by AltaRock, and another in a portion of The Geysers geothermal province that lacks natural hydrothermal activity because the rock is dry and impermeable, run by Calpine, the company that operates The Geysers (see their 2012 GTP status report — DoE provides details online for all the GTP grants).

The two main classes of technological issues that need to be addressed before EGS can be commercialized seem to be understanding the creation and connectivity of reservoirs, and reliable long-term downhole operations at very high temperatures.  Both of those areas have seen a great deal of technological progress as the oil and gas industry has pursued unconventional hydrocarbon reserves into deeper, hotter, less permeable formations.  The oil services giant Schlumberger has a geothermal consulting practice, and is interested in migrating their considerable knowledge and technology base into the commercialization of EGS (see their 2008 white paper Improving the Economics of Geothermal Development for some of the parallels).

The oil and gas industry has made great strides in the last couple of decades on similar problems, and those advances have led to the commercialization of hydrocarbon bearing shales and tight sands, ultra-deep reservoirs, and the thermal extraction of high viscosity heavy crudes.  The sudden economic viability of these large “unconventional” reserves is terrifying to anyone fighting for a stable climate.  It’s also a testament to the power of well directed R&D efforts.  We can apply the same ingenuity to commercializing EGS if we want to.

No Free Lunch

Any energy source utilized at the scale of hundreds of gigawatts — or terawatts globally — is going to have side effects.  EGS is no different.

Most readers will have noticed a striking similarity between the EGS development process and hydraulic fracturing (or fracking) which has proliferated across North America in the last 5 years as a way to extract oil and natural gas from shale and tight sand formations, sending gas prices into bargain basement territory.  In many cases, the fluids used in fracturing for EGS are much simpler and less toxic than the ones that have become widespread in the gas industry, but that’s probably due in part to the fact that we don’t really know what works best in the context of EGS.  Going forward, both applications of the fracturing technology will need to be held to the same stringent environmental standards to protect our groundwater resources.  At the same time, EGS doesn’t have to worry about fugitive methane emissions, the energy it’s ultimately producing is virtually carbon free, and can integrate well with traditional renewables like wind and solar.

All hydraulic reservoir stimulation creates microseismic events as fractures open up in the rock — tiny tremors detectible only with sophisticated sensors — but in some circumstances the additional fluid pressure is also capable of activating existing faults under stress, creating larger seismic events that can be felt by people at the surface — essentially precipitating an earthquake that was already waiting to happen.  This happened at an EGS project near Basel, Switzerland and also at The Geysers in northern California as they attempted to increase the flow rates within their existing geothermal wells.  Both seismic events were small, but citizens were understandably spooked.  The projects were discontinued, and the DoE worked to develop a protocol for dealing with induced seismicity.

The big up front investments required to develop EGS also make it likely that these generation resources would be controlled either by investor owned utilities, or larger public power organizations and independent power producers, which many advocates of increased distributed generation may not be pleased by.

There are possible environmental impacts associated with EGS, and significant technical hurdles to be overcome before the technique is ready for broad commercial deployment.  However, because of it’s potential scale and dispatchability it’s worth investing in solving those problems now, so it can be ready to fill in some of the last fossil fueled niches in our energy economy tomorrow.  Hopefully we’ll have the foresight to support it until private capital is ready to take over and scale it up.

Further Reading:

• US DoE Enhanced Geothermal Systems Program
• Geothermal Resource Assessment Fact Sheet (USGS, 2008)
• Evaluation of Enhanced Geothermal Systems Technology (US DoE, 2008)
• EGS Basics Fact Sheet (US DoE, 2012)
• The Future of Geothermal Energy (MIT, 2007)
• The Soultz-sous-Forêts’ Enhanced Geothermal System (InTech, 2012)
• Status of the EGS project in the NW Geysers Hydrothermal Field (Calpine, 2012)
• GTP 2009-2015 Multi-Year Research, Development and Demonstration Plan (US DoE, 2009)