Can wind turbines self-replicate? (And how will we make beer?)

The ten biggest questions in energy & climate tech, Question 4

Current hypothesis: Yes! (And, with industrial heat pumps!)

In 2017, the same year I joined Energy Impact Partners, our founding utility LP Xcel Energy adopted “Steel for Fuel” as a slogan to highlight a major new investment in wind energy. Xcel’s service territory runs through the so-called “wind belt” of America, from Minnesota through Colorado, Texas, and New Mexico.¹   Xcel’s CEO at the time, Ben Fowke, explained the move to Utility Dive:

“If I were talking to you 10 years ago, I don’t think I’d be telling you that I think solar is competing with fossil. I wouldn’t tell you that wind is beating fossil. I am telling you that now.”

I love “Steel for Fuel”. I love it so much, I named this blog after the slogan.²

“Steel for Fuel” perfectly captures the essence of the energy transition. We’re substituting ongoing combustion of fossil fuel with big up-front investments in physical structures: steel wind turbine towers and solar mounting systems.  In financial terms, we’re trading opex for capex.  In environmental terms, we’re trading carbon emissions for much higher raw materials utilization.

In this way, “Steel for Fuel” also implicitly hints at one of the biggest hurdles that wind & solar power will need to overcome on their march to dominance in the new energy economy.  Renewables require tons of steel to deploy; but it’s not currently possible to use the form of fuel that they generate – electricity – to produce more steel. In other words, we’re currently able to trade steel for fuel, but we can’t yet trade that decarbonized fuel for steel…which happens to be the world’s #2 construction material by weight.  The same is also true for the world’s #1 construction material, cement, which we’ll need nearly as much of to construct wind turbine foundations as we need steel for wind turbine towers.

This isn’t an acute problem for wind or solar supply chains - at least, not the type of problem we’re seeing emerge in the EV sector. The steel & cement markets are both big enough to absorb new demand from the energy transition without flinching. But this clearly is a problem for long-term decarbonization. Globally speaking, steel manufacturing is responsible for a whopping 7-8% of greenhouse gas emissions.  Cement manufacturing is about the same. We need to close the loop!  Steel —> Fuel—> Steel, etc.

Self-replication is very difficult for wind turbines

Unfortunately, there are a few reasons why making wind turbines – hell, why making anything – is extremely difficult to do economically with clean electricity.  In some cases, it seems nearly impossible. Steel turns out to be a great case study. 

Steel manufacturing has historically been viewed as practically impossible to electrify, because fossil fuel plays a critical dual role in the process. Most obviously, it’s a source of energy to heat up a blast furnace to over 2,000 degrees Fahrenheit. But heat alone does not turn iron ore into steel. That’s because, like many metals, iron ore is mined from the earth as an oxide - which means that iron atoms are bound up with oxygen. 

Hence, making pure iron, which is essential to making steel, requires removing those oxygen atoms - a process referred to as chemical “reduction”.  Currently, burning fossil fuel in a blast furnace also satisfies this requirement, as carbon and hydrogen atoms ferry away oxygen.³ 

Electricity can readily substitute for coal or gas as a source of heat energy, as there are a number of ways to thermodynamically ‘downgrade’ electricity into heat.  In fact, about a quarter of steel demand today is met by recycling scrap steel in an electric arc furnace.  However, electricity alone can’t extract pure iron from virgin ore, because it can’t perform chemical reduction.⁴

And even if it could, electrifying steel manufacturing would be a very tough economic proposition.

Both coal & natural gas are frighteningly cheap, especially if all you want from them is thermal energy – meaning, all you want to do is heat something up.  In many parts of the world, one or both can be purchased and delivered to industrial sites for about $3-5 per ‘million british thermal units’ (mmbtu) of embodied energy.  The absolute cheapest wind & solar energy in the world can dip into that range, but more typical wind & solar power prices tend to be in the range of $7-15/mmbtu.  Also, those renewable prices aren’t accounting for the cost of electric transmission, or the cost of storage to manage the intermittency of the wind & sun. 

There are basically just three ways that new technology might allow clean electricity to overcome these disadvantages.

1. Use electrochemistry, which can wield electricity to drive a particular chemical reaction.

2. Use industrial-scale electric heat pumps, which can employ electricity to move heat, rather than generate it.

3. Use thermal energy storage to source really cheap renewable power whenever it’s available, at the lowest cost possible, and then dispatch that energy to fuel continuous industrial processes.

Let’s talk through these options, in this order.

The rise of electrochemistry

“Electrochemistry” is what it sounds like: a merging of electrical & chemical processes. Specifically, it’s the use of electrical potential differences to drive chemical reactions, or vice versa.  (For example, batteries, electrolyzers, and fuel cells all qualify as electrochemical devices.)

In some cases, electrochemical processes can avoid the need for lots of high temperature heat, making them more efficient than ‘brute force’ fossil fuel combustion. Even more importantly, electrochemistry can also simultaneously address the chemical requirements of certain industrial processes, like the need for chemical reduction of iron oxide to manufacture steel.

In the case of steel specifically, my firm EIP is an investor in an electrochemical approach; actually, I think it’s fair to call it the electrochemical approach, as there are no other credible alternatives that I’m aware of.  The process is called “molten oxide electrolysis”, and it’s being pioneered by a company called Boston Metal.

Rather than crudely converting electricity into heat to smelt metal ores, molten oxide electrolysis works by running electrical current through novel electrode materials to simultaneously liquify and reduce those ores.  One of the great benefits of this approach is its efficiency advantage over a blast furnace; because it consumes less energy overall, it can run economically on low-cost clean power.  Another advantage is that it’s much more modular than a giant blast furnace, which enables substantially lower risk investment in new facilities.

Cement production represents another great opportunity for electrochemistry because, like steel manufacturing, it can’t be entirely decarbonized by simply swapping out fossil fuel for electricity.  That’s because about half of the carbon emissions from cement manufacuring don’t come from fossil fuel combustion; instead they come from one of the raw material inputs to the process; calcium-carbonate, better known as limestone.  Heating up crushed limestone in a kiln causes calcium carbonate to give up its carbon, leaving us with lime (calcium oxide), which is a crucial ingredient in the world’s default cement formulation, called “Ordinary Portland Cement”.  The leftover carbon needs to go somewhere, of course, and currently “somewhere” is straight up into the atmosphere.  Hence, electrifying the heat that flows into cement kilns would only solve about half of their emissions problem.

Another EIP portfolio company, Sublime Systems, has developed an electrochemical approach to producing lime from a wide variety of calcium-bearing materials (limestone & others) at ambient temperature.  This lime can either substitute for the carbon-intensive lime in a conventional cement kiln, or it can be employed in another proprietary step to generate fully decarbonized cement with the same or better properties.  Much like Boston Metal, Sublime can do this economically because of the higher efficiency of their process relative to brute force heating.

There are other ways of utilizing electricity as a precision tool to create the right conditions for chemical reactions to occur, which aren’t strictly “electrochemical”. For example, Nitricity uses electricity to power the reaction of air and water to create nitrogen-based fertilizer. 6K uses electricity to manufacture a variety of specialty materials, including battery electrodes, with much higher efficiency and lower environmental footprint than conventional methods. (Note that both Nitricity and 6K are also EIP portfolio companies.)

So, if you’re reading this, and happen to be in college considering what to major in…Just saying: A joint electrical & chemical engineering program sounds really tough, but it would be extremely valuable for the energy transition.

Moving heat, not making heat

Much as I now love saying “electrochemistry”, it’s not a panacea.  First of all, electrochemical solutions don’t typically tend to make great sense for retrofitting existing facilities; they’re usually better suited for new build.  But of course, the world has already sunk trillions of dollars of capex into steel plants, cement plants, and all of the other plants in which we manufacture the materials of modern life.  Many were constructed in just the past two decades in China & India, and have at least a decade or two of life left in them.

Second, there are a whole host of industrial processes in which we really do just need to melt something, or boil something, or dry something, or fry something – in short, we really do just need heat.

If all you need is heat, whether to “re-fire” an existing facility or build a new one, then the best way to give electricity an efficiency advantage is a heat pump.  Heat pumps, as their name implies, don’t actually generate heat. They move it.  Specifically, the vast majority of heat pumps use electricity to drive a gas called a “refrigerant” (of which there are many options) through a cycle of evaporation and condensation.  When these gases evaporate, they aborb a bunch of heat; and when they condense, they release that heat. Because the universe is strange & wondrous, the amount of heat that this cycle can pump from one location to another is much higher than the amount of energy it takes to run the pump.

Heat pumps are already an extremely mature technology category – in fact they’re already one of the foundational building blocks of the modern world.  Mostly today, they’re used to move heat out of a place that we want to keep cool, like a refrigerator or an air conditioned home. Now, though, we’re also beginning to put them to work moving heat in the opposite direction to keep our buildings warm in the winter.  [Note: I’ll return to this topic in a subsequent post.]

The fundamental challenge for heat pump technology in industrial settings is that a heat pump’s efficiency is largely a function of the difference in temperature between its heat ‘source’ and heat ‘sink’ - a difference which is referred to as “lift”.  A heat pump powering a refrigerator, for example, might be called on to maintain a temperature inside the unit of about 35 degrees Fahrenheit. That requires ‘lifting’ heat to the temperature outside of the unit, which is probably a 75 degree kitchen.  Given this lift of 40 degrees, an average heat pump might be able to move around 4 times as much heat energy as the electrical energy input to the device.  This ratio is referred to as a “coefficient of performance”, or COP, of four.  (Sometimes it’s also referred to as an “efficiency of 400%”, nonsensical though that may sound.)

The lifts required for industrial applications are MUCH higher.  Most industrial processes that need heat energy need it to be at least hot enough to boil water, which is 212 degrees Fahrenheit; more often, the temperatures required are in the range of 300-400 degrees.  Gathering up all that heat from ambient air requires lifts of 150-350 degrees, which is up to ten times higher than the lift needed for refrigeration.  This is a lot harder than heating a home. It takes a lot more work, and requires specialized components built to operate at higher temperatures. 

But if one could develop a heat pump that would deliver heat in the range of 300-400 degrees, while maintaining a reasonably high COP, that solution could be applied to about two thirds of all the heat consumed in US manufacturing, which is about 7% of all US primary energy consumption.  Everything from dyeing clothes, to frying potato chips, to brewing beer requires heat delivered in that temperature range.

There’s two basic ways build an industrial heat pump with a reasonably high COP to address this massive opportunity.

1. Tap into “waste” heat. No industrial process is perfectly efficient.  For every 100 joules of 300 degree steam that flow into a process, 30 joules of 150 degree water might flow out the back end.  A heat pump could hypothetically tap into that otherwise wasted heat to achieve a lift of just 150 degrees, instead of a lift of about 250 degrees from the ambient air outside of the plant.  It’s quite a bit easier to build a heat pump capable of drawing energy from hot water than it is to build a heat pump capable of capturing energy from ambient air.

   Hence, there are already a number of waste heat focused products available in the market today, and they can achieve a major efficiency boost for some processses. Waste heat recovery isn’t a full decarbonization solution, of course, because heat that’s truly “wasted” constitutes a fraction of the energy flowing into any given process, and only a fraction of that fraction can be effectively recovered. However, it’s still a great idea as long as it’s not too expensive to set up.

   And there’s the rub. The cost of setting up a waste heat recovery system varies a lot from plant to plant. Sometimes there’s a hot water pipe flowing right by a boiler, where an industrial heat pump might be installed relatively easily. But more often than not, waste heat is only available at the opposite end of a facility from the primary production process. Moreover, waste heat isn’t necessarily available all the time; for some batch processes, the timing of waste heat availability doesn’t align well with the timing of primary heat demand.

2. Build something fundamentally new. The ideal solution would be an industrial-scale heat pump capable of producing 300-400 degree steam from the heat available in ambient outdoor air, with the highest coefficient of performance possible for such a high temperature lift.  This is exactly what EIP portfolio company AtmosZero (now freshly out of ‘stealth mode’) has been developing for over a year. The team at AtmosZero is building a heat pump engineered for full air-to-steam operation with a COP of about two, which is currently unheard of in the market. The solution can serve as a full replacement for any conventional boiler solution, and can also be engineered to support ultra-efficient waste heat recovery. At EIP, we’re super excited about the prospects for a complete electrification solution for a wide range of light industrial settings. And yes, they’ll be brewing beer.

Heat as renewable power storage

In regions with especially strong wind or solar resources, it’s possible to procure renewable power for $10-15/MWh.⁵ That equates to about $3-5/mmbtu, which I mentioned above is reasonably competitive with the cost of coal and natural gas in most places. However, industrial processes tend to need energy continuously, and of course renewable power is only available intermittently.

Fortunately, it turns out that one of the cheapest ways to store energy is to store it as heat. And also, the most efficient way to store and release energy is to do so as heat. I’ve written about this before in the context of EIP’s investment in the thermal storage technology developer Rondo Energy:

“Picture a giant insulated box filled with bricks that heat up to over 1,000 degrees Celsius. It’s an extraordinarily elegant approach, utilizing resistance heating elements similar to those you’d find in a toaster, plus an ultra-cheap class of thermally-dense brick that’s been employed for over a century in blast furnaces at steel mills. There’s nothing remotely toxic or exotic, relatively few moving parts, and as of March 2023 the company has commissioned the first of these systems in a real commercial application in California.

Electrical energy from solar panels or wind turbines is converted via resistance wires into high temperature heat. That heat can then be dispatched on-demand to fuel any industrial process which needs it (which is…most industrial processes). The round-trip efficiency of this cycle is startlingly high - upwards of 98% from solar or wind power input to heat output - because, of course, energy loss is heat.”

In my mind, the existence of Rondo’s proven, ultra-low-cost thermal storage solution ought to raise a big question in the mind of every industrial facility operator: Can I get my hands on some really cheap renewable power? And, shortly after, a second question: How close can I get it to my facility?

The answers to these questions will vary regionally, of course, but also according to some plant-specific characteristics. The facilities that are best positioned to take advantage of Rondo’s solution will be those that are large enough to procure energy at ‘utility-scale’ - in the tens of megawatts, at least - and with enough nearby land that they don’t need to pay an exorbitant cost for electric transmission. In some cases, there might be enough wind or solar power within reach of a plant that it can skip a costly interconnection to the electrical grid, and instead be routed directly through a Rondo unit to fuel an industrial process.

Today, of course, some facilities already have a lot of wind or solar power in their backyards. In some cases, that renewable power is already so bottlenecked by transmission constraints that facilities are presented with periods of very low, or even negative wholesale electricity prices. In fact, Rondo’s analysis of 2019 electricity prices in wind-heavy markets shows that electricity was already practically free if it could be purchased during the four lowest-cost hours of each day…and wind penetration has only grown in the three years since. The Rondo unit is specifically designed to turn as few as 4 hours of cheap, clean electricity supply into 24 hours of steady, high temperature heat.

In these cases, Rondo’s thermal storage solution already represents a game-changing option for decarbonization.

Steel—> Fuel —> Steel, etc.

In sum: I believe we can close this loop, and them some.  Whether electrochemistry, heat pumps, or heat storage is the right decarbonization solution for any given location will vary by region according to renewable resource quality, the state of the existing industrial base, and all kinds of site-specific variables. The optimal solution will almost surely include some of each of these options…in some cases, working in tandem within the same facility! For example, a facility operator might find that a combination of AtmosZero heat pumps and a Rondo thermal storage system might support the least-cost utilization of clean electricity.

Stepping back from the details, I’m confident that we will increasingly harness clean electrons to make steel, cement, beer, and the majority of all the other stuff of civilization. 

So yes, wind turbines will self-replicate.

Enjoy this post? Please check out the rest of the Ten Biggest Questions in Energy & Climate Tech

---

1. The wind belt happens to get a lot of sunshine as well.

2. Okay, my EIP colleague Shayle Kann was the one who suggested “Steel For Fuel”.

3. In most cases, the fossil fuel of choice for making steel is coal, in which case carbon is doing all of the reduction work. In some regions, natural gas has more recently gained traction, which when broken down into syngas (CO + H2) can also serve as a reducing agent.

4. There are also other important chemical reactions that need to occur to manufacture various steel alloys, but this is the important one for which electricity alone is not a viable substitute.

5. For example, in the US “wind belt” running from the Dakotas down through Texas (at least as long as the federal wind production tax credit holds up).