clock menu more-arrow no yes mobile

Filed under:

The Earth itself could provide carbon-free heat for buildings

The world needs clean heat, and geothermal energy has it.

geothermal district heating
Geothermal district heating.

The heat stored in the Earth’s crust, known as geothermal energy, is carbon-free and effectively inexhaustible. There’s enough of it to run all of civilization for generations, if it could be cost-effectively tapped.

Tapping it turns out to be no small feat, but efforts have ramped up recently due to new urgency by the climate crisis and the search for low-carbon alternatives to fossil fuels.

The cutting edge technological developments in the field (including, yes, lasers) are devoted to drilling deeper and deeper, into hotter and hotter rock. Heat anywhere from 302°F (150°C) up to 703°F (373°C), where water enters its “supercritical” phase and above, can be used to profitably generate electricity.

But electricity is only half of the geothermal story. Well before humans generated electricity with it, they used geothermal heat directly, to bathe, cook, and heat buildings, among other things. Geothermal direct heat is still used today in industry, agriculture, and for buildings, but only a tiny fraction of its potential has been unlocked.

When it comes to direct use of heat, geothermal resources don’t need to be quite so hot. It doesn’t require 300°F to heat the air in your home to 68°F. Just about anything 50°F or above (which is available just 10 feet down or so) can be used for something, whether drying grain, running a greenhouse, melting ice on airport runways, or heating commercial buildings.

A diagram showing the uses for geothermal direct heat.
Direct uses of geothermal heat.

Geothermal heat is accessible almost everywhere and useful in a wide range of applications. The US Department of Energy has a research program devoted to these “low-temperature and co-produced resources.”

But the most important application, in my mind, is the use of low-temperature geothermal resources for large-scale heating and cooling of buildings.

Heating and cooling buildings isn’t as sexy as electricity in the energy world these days, but it is important, representing just over 12 percent of US greenhouse gas emissions and a larger proportion of emissions in cities, many of which have aggressive decarbonization goals. To achieve those goals, they need to figure out carbon-free heating, and geothermal is one of the best (out of very few) options.

In this post, we’ll dive into the other half of geothermal: heat. First we’ll take a look at the market and the need for low-carbon heat. Then we’ll look at the technologies and companies involved, and wrap up by considering how government might help accelerate the development of geothermal solutions.

It’s hot, or at least warm, stuff!

Decarbonization means an improved competitive landscape for geothermal heat

Cities across the world are setting aggressive decarbonization goals, pledging to zero out their direct carbon emissions by 2050. The first three challenges facing a decarbonizing city are electricity supply, transportation, and heating and cooling of buildings. The pathways to decarbonization of electricity and transportation, while extremely challenging, are at least fairly well understood: renewable energy, electric vehicles, and good urban design that minimizes the need for cars.

For most cities, though, heat is a big unanswered question.

Oil and natural gas furnaces will need to be phased out, which means cities will need an extraordinary amount of low-carbon heat to compensate. And low-carbon options are much more limited in heat than in electricity.

Some furnaces can run on biomethane, other biofuels, hydrogen, or hydrogen-derived fuels, but in a mostly electrified world, low-carbon liquid fuels are likely to be used for high-value applications in industry and transportation — not heating your living room.

A chart showing that most heat is used for space and water heating.
Most heat is used for space and water heating.

That leaves geothermal district heating or, at an individual building level, electrical options like electric resistance heating or heat pumps. In heat pumps, it’s either air-source (exchanging heat with the outside air) or ground-source (exchanging heat with the earth). The latter is far more efficient. And geothermal district heating is the most efficient of all.

In a decarbonizing world, it is these — the other low-carbon heating options — that will eventually comprise the competitors in the heating and cooling space. It’s a competition some decarbonizing cities, like Boston, are already grappling with. Boston will have trouble building lots of new electrical infrastructure to heat buildings with electricity, so it is leaning toward geothermal.

So what exactly are the technologies that can provide heat from the Earth? There are two basic categories. Let’s start by looking at the smaller side.

Ground-source heat pumps are the most efficient source of individual building heat

It’s a bit of a fudge to include ground-source heat pumps (GSHPs) here, because technically they do not make use of geothermal energy. They make use of stored solar energy, from sunlight striking the Earth’s surface. It’s only when you get much deeper, or in active volcanic areas, where you get into heat from the planet’s core. If you want to be precise, GSHPs harvest solar heat stored in the shallow earth.

I don’t think this terminological issue matters all that much though — it’s heat in the earth!

Anywhere from 10 to 1,000 feet beneath the surface, the temperature is a steady 54°F, year-round, everywhere in the country. GSHPs take advantage of that fact to heat and cool buildings. When the air is colder than 54°F, they draw heat from the earth; when it’s hotter than 54°F, they dump heat into the earth.

A GSHP consists of two parts. The first is the “ground loop” pipe buried beneath the ground with water circulating through it. Via conduction, the water draws heat from (or returns heat to) the earth, so the more pipe surface area there is, the more efficient the system. That’s why there are often several loops of pipe in the overall ground loop. The rule of thumb is one loop equals one ton of capacity, which equals about 12,000 BTU per hour. An average US home will need 2 to 3 tons of capacity, thus two to three loops (or one very deep one).

The second part is the heat pump itself, which sits inside, connected to the ground loop, exchanging heat with the water via a vapor compression refrigerant cycle (not unlike the way your refrigerator exchanges heat with the surrounding air). In the winter, it takes heat out of the circulating water and puts it into the air, thus warming the building; in the summer, it takes heat out of the air and puts it into the water, thus cooling the building.

A diagram showing how ground-source heat pumps (GSHP) heat and cool buildings.
GSHPs heat and cool buildings.

You can think of a GSHP as two linked heat transfers. Via the ground loop, the water exchanges heat with the earth; via the heat pump, the water exchanges heat with the indoor air.

Because ground temperature is basically the same 10 or 1,000 feet below the ground, somewhat counterintuitively, the depth of the ground loop doesn’t matter much. What matters is the square footage of pipe exposed to earth. Installers use either long horizontal loops or deep vertical loops depending on the project. (Most projects these days are “closed loop,” meaning no fluids are exchanged with the ground, but in the right circumstances, an “open loop” system that works directly with water heated by the earth can work.)


A GSHP is not generating heat, like an oil or gas furnace, but harvesting heat from the ground. The water does not circulate itself, of course; it requires electricity to run a GSHP. But in terms of units of heat out per units of energy in — what they call, in the business, Coefficient of Performance (COP) — it is the single most efficient way to heat a building.

An oil or gas furnace has a COP less than 1; one unit of energy input produces about 0.7 to 0.9 units of heat. Electric resistance heating (baseboard heaters, wall heaters, space heaters) have a COP of 1. Air source heat pumps (ASHPs), which draw heat from the outside air rather than the earth, vary somewhat with the temperature of the air, but generally can reach a COP of 3. GSHPs, depending on the climate, can get to 4, or as high as 6. (They work better in extreme climates, with a high temperature differential between air and earth, than in temperate climates.)

In the best circumstances, GSHPs are 600 percent efficient. Nothing else, except a district heating system serving multiple buildings, can match that efficiency.

GSHPs are an old technology — they first popped up in the US around 1940 — with well known benefits and drawbacks. On the benefits side, the system runs quietly, operating costs are low, maintenance costs are low, there are no indoor pollutant emissions or GHGs, and it lasts a long time. (Heat pumps inside can last 25 years; ground loops can last 50 years or longer.) It is a nice thing indeed to already have a GSHP installed.

Unfortunately, it has also been expensive as hell to install one. They typically run from $20,000 up to $50,000 in upfront costs (quite a bit more than your $1,000 natural gas furnace) and installing them has typically involved extensive drilling and excavation that can last for weeks (quite a bit more than the 1-2 day turnaround for a gas furnace or ASHP). These limitations have made them impractical for most homeowners.

At least at the moment, when it comes to a remodel, it’s a real question whether GSHPs are worth the additional cost over and above ASHPs, which have improved enough to work in almost any climate. If an ASHP isn’t enough for a given building, it’s generally cheaper to reduce heating needs through insulation and efficiency than it is to buy a bigger system.

For new builds, though, “geothermal is a no brainer,” says Adam Santry, the president of Allied Well Drilling. “You don’t need any [tax] credits. Rolling [a GSHP] into your mortgage, you are cash flow positive that first month.” The savings on heat are greater than the loan payment on the GSHP, right off the jump.

“Yes, there’s an upfront cost,” says Alan Skouby, a 40-year veteran of the industry now with GeoPro, Inc. “But it will pay for itself in relatively short order, and once it’s paid for, it’s a money printer.”

GSHPs face a problem that all sorts of clean-energy technologies early in their cost and development curves have encountered: Though they pay off in the long run, the substantial upfront investment often deters customers. The two key strategies for growth, then, are reducing those upfront costs and spreading them out over time through clever financing.

One new company is currently attempting to do both, focusing on the residential market.

Dandelion is trying to make ground-source heat pumps easy

The secretive X Lab at Alphabet (Google’s parent company) has been working away on clean energy problems, spinning off companies as it goes. One of them, formed in 2018, is called Dandelion, and it is directly attacking the problems that have held GSHPs back.

Dandelion’s team “didn’t grow up in this industry, they grew up in the solar industry,” says Skouby. “They’re coming to all of this with a fresh perspective.”

Typically an HVAC contractor can install a furnace or an ASHP themselves, harvesting all the profits and tax credits. For a GSHP job, they have to find a drilling subcontractor and split the profits — more hassle for less money. They also frequently have furnaces in stock that they need to move, and might need to special order a GSHP. The incentives don’t line up.

One of Dandelion’s key moves has been to vertically integrate, to pull all those links in the supply chain into one organization. The people who find customers, assess properties, drill ground loops, and install heat pumps all work for Dandelion, so they can coordinate efficiently.

Vertical integration also means Dandelion can order custom-built, high-quality equipment. “Because they’ve got a game plan to reach much bigger scale,” Skouby says, “they can leverage that and buy down cost. Nobody else has been willing to do that.”

For instance, the company designed its own heat pump. “We looked at what was taking installers a lot of time,” says Kathy Hannun, Dandelion’s founder and president, “and every time, there was an opportunity to take those things and just build them into the heat pump.” There’s less on-site assembly required and it has a smaller form factor than comparable heat pumps. It is also covered in sensors, which provide real-time information on how it performs in the field, something the industry has lacked. It’s also cheaper than its competitors.

The company has ordered purpose-built drills, smaller than typical geothermal drills and able to fit into tighter spaces. Similarly, they have optimized piping, grouting, and other components. The strategy is more like a solar startup’s: Invest big early on to drive down costs and begin scaling up; trust that scale will pay back the investment.

Drilling vertical ground loops — 4 to 6-inch holes around 500 feet deep — Dandelion has substantially cut down on the time and disruption of installation, from weeks or months to one week. The company has got the upfront, delivered cost of a system down to $18,000 from $25,000.

Just as importantly, it has devised a financing model to overcome the upfront cost barrier. It loans the cost of the system to customers, who pay nothing upfront. Instead, they repay the loan at a fixed monthly rate that is lower than their previous heating and cooling costs. They save money from day one.

“They’re targeting the type of customer our industry needs,” says Santry, “medium- to lower-income people that this was not available to.”

The loans are still attached to the homeowner, though. What the industry needs, says Hannun, is a model like rooftop solar’s, with “third-party ownership models where, if you’re a homeowner and you don’t plan on living in your house forever, you can put no money down — just buy solar power, essentially, instead of buying normal electricity.” This kind of “solar as a service” model could work just as well with “heat as a service.”

Dandelion is taking off in New York, where some localities like Westchester County have banned gas in new buildings, and there are millions of people heating with expensive propane and fuel oil furnaces (against which a GSHP will pay itself off in five years). “When they see that they can get renewable energy for less than they’re paying for fuel oil,” says Hannun, “it’s very compelling.” The company recently expanded to Connecticut.

“I think they’re going to be successful, because the scope they’re projecting is attracting a lot of utility types that have the financial wherewithal to help drive what they’re doing,” says Skouby, “or get behind them on an exclusivity arrangement, which they wouldn’t be willing to do with a local contractor.”

New York also has substantial incentives for low-carbon heat, which will likely be needed anywhere GSHPs must compete with natural gas. But the company is learning as it goes and sees plenty of room to bring down costs “across the board,” says Hannun. And of course, in a carbon-constrained world, natural gas will be fazed out.

So that’s the smaller geothermal heat technology. Now let’s look at the bigger stuff.

Low-temperature geothermal can heat multiple buildings for cheap

In my previous post on geothermal, I described how a traditional geothermal system works. One well, the production well, taps into hot water trapped in underground aquifers; the water comes up, the heat is extracted, and the water is cooled and returned to the earth via a second well, the injection well.

A diagram showing how a geothermal system works.
Geothermal power.

To access the high heats needed to generate electricity, such systems typically must be sited in specialized (and relatively rare) areas near volcanic activity, where there is extremely hot water trapped in porous rock underground.

But saline aquifers containing warm water — not hot enough for electricity, but plenty hot enough for direct heat — are practically ubiquitous, in the US and elsewhere.

Geothermal systems that tap into warm (sub-300°F) water can be used as a source of heat for a district heating system, i.e., a single connected system of hot water loops that heats multiple buildings.

District heating found one of its very first expressions in the US — Boise, Idaho, has used geothermal to heat buildings since 1890 and heats its downtown with it to this day — but it is far more popular and advanced in Europe, especially Iceland (though China is, in this as in all things, scaling up quickly). Paris, Munich, and Reykjavik are all known for their extensive district heating systems.

A diagram showing a geothermal district heating system.
An example of a geothermal district heating system.

In the US, district heating has never quite caught on, but it is a frequent feature of college campuses. As part of its decarbonization goals, Princeton University is shifting from a natural gas steam system to geothermal. The Oregon Institute of Technology, Carleton College in Minnesota, and Ball State University in Indiana (among others) already heat with geothermal district heat.

Once the upfront capital costs are paid off, geothermal district heat is dirt cheap, for decades or even centuries. (The world’s oldest working geothermal district heating system, in Chaudes-Aigues, France, has been going since the 14th century.) But the upfront costs remain daunting.

There are some new technological developments in the space. The Department of Energy is studying deep direct use (DDU) geothermal systems, which go deeper to find suitably warm temperatures in almost any geography and use them as large-scale heating sources for campuses, military installations, hospital complexes, or residential developments. “Large-scale, fully integrated DDU geothermal systems have not been realized in the United States,” the DOE writes, “although efforts of this type are increasingly popular in Europe and elsewhere.”

Some of these DDU efforts are using “closed loop” systems (not unlike GSHPs) that don’t exchange fluids with the earth at all, thus eliminating any possibility of groundwater pollution. The Canadian company Eavor (covered in my previous post) is working on closed-loop systems that can, in addition to going deep for electricity-level heat, be used for lower-temperature systems that harvest heat for buildings.

A diagram of Eavor’s closed-loop deep geothermal system.
Eavor’s closed-loop deep geothermal system.

Some DDU systems, if they tap high enough heat, can “co-produce” electricity and heat, thus blurring the line with geothermal power systems.

The fact is, though, when it comes to shallow saline aquifers, the oil and gas industry already knows its way around. “The low hanging fruit [for geothermal heat] is our sedimentary basins, between two and three kilometers depth,” says Marit Brommer, who runs the International Geothermal Association but started out her career as an oil and gas engineer, “and they have been mapped extensively because of our oil and gas runs. We know their temperatures extremely well — and we found more water than oil in those reservoirs, by the way.”

“We have a lot better tools now [than in previous decades] — better drilling technology, much better geophysical logging capability, better seismic reflection imaging,” says Jeff Tester, a professor of sustainable energy systems and principal scientist for Cornell University’s Earth Source Heat project. “We know so much more about how to find permeability and fluids in the rock.” Drilling at that depth, avoiding pollution or seismic disruption, is something oil and gas has been working on for decades.

Geothermal district heating is a no-brainer for anyone building new housing developments, campuses, or industrial clusters. It represents low, stable heating costs (rather than the fluctuating costs of oil and gas) for generations.

Forward-thinking cities like Munich (which is seeking to reduce greenhouse gas emissions 50 percent by 2030) have begun to think of geothermal loops as part of city infrastructure, to be installed and maintained alongside water and sewer lines, so that any new building or development can simply connect to the main line through a utility, like other basic services.

The larger such a system grows, the more its unit costs fall. And it’s a local resource that generates local jobs; it is not dependent on imports or global markets. It gives cities some measure of independence.

A diagram showing a city’s heating system. Engie

Again, the barrier is the upfront costs. A decent-sized geothermal district heating can run $25 million, says Brommer, and though, “on average, it takes you about a quarter of your life cycle in order to get rid of your [debt] burdens,” the capital costs are often enough to scare off developers and municipalities.

Costs will come down with scale and knowledge-sharing. “What we need is multiple companies who work in multiple countries in similar subservice settings, that understand the drilling requirements and the service needs,” Brommer says, “meaning that the lessons learned in country one can be applied to reduce costs in countries two, three, and four.”

But that kind of learning requires growth. Just as with GSHPs, the trick is finding tools to bring down upfront costs and spread them out over time.

Geothermal costs more upfront, but less overall. Government could help with that.

Accelerating the development of geothermal electricity is mostly about technology research and demonstration, but when it comes to geothermal heat — both GSHPs and larger solutions like DDU — the primary need is for the kind of public policy pull that draws demonstrated technologies into a broader market.

That means incentives like grants, tax credits, or feed-in tariffs (heating tariffs, in this case) to bring down the upfront costs. At the city or county level, it means regulatory reform to reduce the costs of permitting, siting, and constructing systems. But perhaps most importantly, it involves financing mechanisms.

Remember, a geothermal district heating system or a GSHP is already a better value than their competitors over the lifetime of the system. They just face the awkward problem that almost all of the costs are stacked upfront, while the benefits accrue over time. It is the timing of the costs and benefits that poses the challenge.

That is the kind of problem financing mechanisms, which move costs and benefits around in time, can solve. The 30-year fixed rate mortgage was invented in the 1930s to spread the large upfront costs of a house out over decades, thus opening home ownership to millions of Americans. Dandelion’s fledgling financing model, which requires no upfront money from the customer, could do the same thing for GSHPs if it can be scaled (and attached to the property rather than the owner).

The government can help by offering low-interest, long-term loans for low-carbon heating systems, or backing such loans if banks or other private institutions offer them. Those loans can help soften the substantial front-end risks of exploring for new resources.

“Iceland addressed this risk in the 1960s with the establishment of a National Energy Fund, which offers loans to fund the initial cost of drilling and exploration,” says Tester. “If the initial drilling stage is unsuccessful, the loan defaults to the state; if the drilling is successful, the loan will be paid as planned.” It is the single most powerful policy tool for expanding geothermal in Iceland, he says.

A geothermal-heated greenhouse in Iceland.
A geothermal-heated greenhouse in Iceland.

Along with financing, new models of ownership and service delivery are needed. “The challenge for the energy transition is that oil and gas companies are unlikely to be operating heat,” says Brommer. “There is a need for smaller intermediary operating companies that understand what it takes to mine heat and can sell it as a service to utility companies. That’s the way forward.” Such intermediaries could even be owned by local communities, along the lines of the popular “community solar” model.

There is plenty of room for innovation around geothermal heat — in technology, but especially in policy and financing. But the US will need to get serious about the investments, policies, and regulations necessary to scale it up to the necessary size.

A large investment of time, money, and policy attention in geothermal heat could help create jobs in almost every US zip code. The DOE’s comprehensive 2019 Geovisions study found that “technology improvements could enable more than 17,500 geothermal district-heating installations nationwide, and 28 million U.S. households could realize cost-effective heating and cooling solutions through the use of geothermal heat pumps.” That number of geothermal systems would require over 50 times the number of wells dug by the entire US oil and gas industry — a bonanza of skilled trade jobs.

Geothermal heat could help towns and cities achieve a measure of energy independence, giving them a reliable source of heating and cooling that never changes price and requires no imports. It could help put retired, laid-off, or just bored oil and gas engineers to work; Dandelion recently hired Jeremy Smith, a 20-year oil and gas veteran, as their new VP of drilling.

But most of all, it could help solve the riddle of how to rapidly decarbonize the heating and cooling of buildings, a problem that has not gotten nearly enough attention and is not exactly awash in solutions. Geothermal is such a solution, right beneath our feet. We just need to get digging.

Sign up for the newsletter Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.