Brattle Q&A: Energy Leaders & Innovators
Amory Lovins on the Big – and Hidden – Opportunities in Decarbonization

Ryan: Let’s start with the big picture. You’ve always had a great eye for the next “big thing” in clean energy. In your opinion, what is the single most important decarbonization opportunity that no one is talking about?

Amory: “Integrative design”: designing buildings, vehicles, factories, and equipment (and the larger systems that contain them – like cities, mobility, materials, and manufacturing) as whole systems for multiple benefits, not as piles of isolated parts for single benefits. This design shift can make the energy-efficiency resource severalfold bigger than commonly assumed, with the practical and profitable potential to raise global end-use energy efficiency ~5× by ~2060 or ~3× by 2040. That doesn’t include the ~3× upstream savings from shifting supply to primary renewable electricity.

The resulting astonishing product of potential upstream and downstream savings is also not an upper limit since we can keep refining the services needed to produce human happiness and satisfaction and find new ways to provide those services. The cost of integrative design’s efficiency also drops with volume and often exhibits increasing returns (much like solar and wind power), so the more efficiency you buy, the cheaper it gets.

Ryan: Could you give a few practical examples?

Amory: Sure. Let’s try one from each of the main energy-using sectors. Buildings use three-fourths of the US’s – and half of the world’s – electricity. My home (and office) is now harvesting its 82^nd passive-solar banana crop at over 7,000-foot elevation near Aspen, where temperatures used to approach -40F. But the building has no heating system. It was cheaper to build it that way in 1983 because we saved more on construction costs by omitting the heating system than we paid extra for the superinsulation, super windows, and ventilation heat recovery (which, together, eliminate the need for a heating system). This type of design helped inspire the European passive house movement. An analogous approach in a new Bangkok house cut air-conditioning energy by 90% with better comfort and normal construction costs. Nearly everyone on Earth lives in a climate between those of Bangkok and Aspen.

I drive a BMW i3 car, lightened ~300 kg by replacing metal structures with carbon-fiber composites. The supposedly unaffordable carbon fiber was paid for by needing fewer batteries to move less mass (mass causes two-thirds of the energy needed to propel a typical car). The [now-discontinued] i3 was also radically simpler to assemble, using two-thirds less capital and water and one-half less energy, space, and time, and with better working conditions. It quadrupled efficiency to 1.9 L/100 km or 124 MPGe without compromise and with many driver advantages. BMW made money on every one of the quarter-million units it sold from 2013–2022. And fewer batteries mean less charging time and infrastructure, less power supply and delivery, and lower emissions. EPRI found that just doubling road-vehicle efficiency through better physics (mass, drag, rolling resistance) before electrification could save the US $200 billion per year in charging and power investments. The barriers are not technical or economic but cultural. Similar opportunities apply to trucks, ships, and aircraft.

More than half of the world’s electricity runs motors. Half the motor torque turns fans and pumps. Making all their pipes and ducts fat, short, and straight rather than skinny, long, and crooked would save about half the world’s coal-fired electricity, or a fifth of all electricity, with typical paybacks below a year in retrofits and instant in newbuild. And if, while making the motors ~80–90% smaller, we also did 35 kinds of improvements to the whole drive system, that one-fifth potential saving of global electricity would rise to about one-third. Yet these approaches to fluid handling and drive power aren’t in any standard engineering text, government study, industry forecast, or climate model. Why not? Because they’re not a technology – they’re a design method, and few people yet think of design as a path to speed and scale.

Progress in both demand- and supply-side technology continues to be wonderful, and as a lifelong technologist, I admire, foster, and apply it. However, most energy forecasting tools and mindsets focus on fuels (commodities) rather than on the technologies that are swiftly displacing them. Technologies get faster and cheaper in ways fundamentally different than those observed for commodities, and often exponentially rather than linearly. Traditional models and year-old data can be hazardous to your financial health.

Another important frontier is using metals, water, and other materials far more productively. For example, about 15% of the world’s CO² comes from steel and cement, but at least half of both can be profitably saved just by better structural design.

Ryan: There seems to be a perpetual view that we’ve maxed out our potential to save energy. For example, someone recently told me that lighting standards have removed the last of the remaining low-hanging fruit for energy efficiency. Yet, for decades, we have continued to find new ways to improve in this regard. Looking ahead, what do you see as the biggest opportunities in this space?

Amory: The fruit isn’t low-hanging; it’s already fallen and is mushing up around our ankles, and it keeps growing back faster than we harvest it. The perpetual delusion you mention has contributed to the past decade’s relative neglect of efficient use, even though it remains the biggest, cheapest, fastest energy resource and a powerful partner with right-sized renewables. Saved energy has delivered about half the world’s past decarbonization – and can deliver at least half the world’s future decarbonization. That’s why both halves of the Glasgow Climate Pact – to deploy renewables three times faster and to improve efficiency two times faster – are essential. Triple-up now looks feasible; double-down is too, but it could and should move even faster if we pay attention, especially if integrative design can become widespread through images and memes via social media.

Lighting illustrates the depletion-of-efficiency misconception you describe. The best light sources have indeed spectacularly raised lux per watt, sometimes nearing theoretical limits, but if we do the right things in the right order, more efficient light sources and luminaires are actually only the fifth of seven steps. The first five steps improve the visual quality of the task, improve the geometry and cavity reflectance of the space, improve lighting quality by cutting veiling reflections and discomfort glare, optimize illuminance (lighting quantity), and harvest and distribute natural light. Only then comes higher luminous efficacy and better luminaire design. After that, the seventh step is better controls, maintenance, and training. But most retrofitters pursue this list in reverse order – “worst buys first” – so the luminaires don’t get fewer, the light doesn’t get prettier or more effective, and capital cost goes up rather than down. Thus, superefficient, beautiful, and effective lighting isn’t just about hardware; it’s equally or more about lighting design, which is often missing.

Another basic difference: Familiar resources like copper and oil are concentrated in depletable assemblages of atoms. But energy efficiency is instead an infinitely expandable assemblage of ideas, depleting only stupidity – a very abundant, if not expanding, resource.

Ryan: This all sounds like a no-brainer! What’s preventing this dramatic energy efficiency vision from being realized, and how do we address those barriers?

Amory: The ~60–80 market failures or institutional obstacles in buying efficiency are not the same as those that slow renewable deployment, so they diversify but don’t simplify the hassle portfolio. For example, we reward design professionals for what they spend, not what they save. Most states reward utilities for selling you more energy, not for cutting your bill. Incentives are split between builders and buyers, landlords and tenants. Fortunately, each barrier can be turned into a business opportunity – the alchemy, [industrial ecologist] Ray Anderson said, of turning stumbling blocks into stepping stones. But this isn’t easy, because institutional structures are complex and incentives are often perverse. For example, the commercial real estate value chain has about two dozen actors, each with stunning perfection penalized for efficiency and rewarded for inefficiency – and they speak different languages, use different metrics, and often don’t talk to each other. Plus, any actor you ignore can be a showstopper, so fixing this mess requires relentless patience and meticulous attention to detail.

Of the 20 or so scaling vectors we’ve identified, and hope soon to start testing and spreading, the hardest and slowest are in reforming pedagogy. Every day, millions of young engineers around the world are being mistaught – for example, to optimize thermal insulation against the present value of saved fuel or pipe diameter against the present value of saved pumping energy, both without including the avoidable sizing and cost of the heating or pumping equipment. Thus, my passive, net-positive house insulated about three times more than would save enough fuel to repay its extra cost – but more than offset the difference by eliminating the furnace, pumps, pipes, fans, ducts, wires, controls, and fuel-supply arrangements, so the total capital cost fell.

Ryan: Heat pumps are a cornerstone of building decarbonization initiatives due to their high efficiency and ability to be powered using clean electricity. However, consumer uptake has been slow. What would you like to see happening to accelerate progress in this area?

Amory: Heat pumps are spreading rapidly, and do need to spread even faster. I’d suggest educating tradespeople (many of whom disparage heat pumps through poor familiarity), ensuring that default to resistive heating is designed out in areas that can get cold snaps, integrating heat-pump with deep efficiency retrofits, sharing savings, financing on-bill, considering reforms – like Washington State’s – that redefine where and how savings are measured, internalizing or shadow-pricing the fair market value of fuel-price volatility (last I looked, methane gas’s Henry Hub price-volatility value approximated the gas price, implying ~2 times underpricing when comparing with nonvolatile competitors like efficiency and renewables), and using dynamic standards like Japan’s Top Runner Programme to shift the market to very efficient heat pumps. (The best air-to-water domestic heat pump produces 6–15 units of hot water per unit of electricity if the lift is 88–55C.) There’s more, but this shortlist illustrates the scope of the toolbox, mainly at the state level.

Ryan: I’ve enjoyed collaborating with your colleagues at RMI on the great work that they’re doing on virtual power plants (VPPs). A consistent theme of that research has been the cost-saving opportunities associated with more flexible energy consumption. Do aggressive decarbonization goals necessarily come with a big price tag, and what do you see as the important tools to ensure that the energy transition is affordable for all?

Amory: Thanks for that VPP work – indeed a very promising approach. With or without it, as the Oxford INET study showed, big and fast decarbonization should reduce both capital and levelized costs, not increase them. Contrary analyses are generally outdated or misframed – including some that, shamefully, forget to credit efficiency and renewables for the avoided costs of the traditional systems they displace. And of course, as you imply, negawatts and flexiwatts (timely use) are related but separate resources, often synergistic, and both severalfold larger than conventionally assumed. Buying them first by competing or comparing demand against supply-side options makes supply investments smaller, faster, simpler, and easier.

Lower costs don’t necessarily mean lower prices or wide, fair, ready availability. That’s why equity should be a foundation of designing policies and tools for delivery. Colleagues like Stanford’s Dr. Holmes Hummel know how to do this, but it’s too often an afterthought.

Ryan: Let’s talk about the 100 GW elephant in the room: data centers. That’s a lot of new load that will need to be served, but some hyperscalers have aggressive decarbonization goals. How will this shake out, and is it beneficial or detrimental to the energy transition?

Amory: Just as coal interests funded previous deceptive claims that IT will eat the grid so we need more coal power, nuclear and gas interests are now aggressively claiming that equally vital AI will eat the grid, so we need more gas and nuclear because somehow renewables (let alone efficiency) won’t suffice. However, AI doesn’t change the empirical fundamentals: efficiency is generally cheaper and faster than renewables, which are cheaper and faster than gas power, which is cheaper and faster than nuclear of any variety (especially the nonexistent varieties in which Big Tech has expressed nonbinding interest).

Counting not just levelized costs but also firming costs for equal reliability typically widens renewables’ advantage over thermal power, because thermal plants tend to have bigger, longer, more abrupt, and less predictable forced outages, so they need more and costlier backup (which can come from any combination of negawatts, flexiwatts, renewables, or fueled sources). And remember, there’s not just one carbon-free grid-balancing resource – big batteries – but 10, of which the batteries, though effective and profitable, are the costliest. They should all get to compete.

Thus, if the current hysteria based on extravagant projections of AI data centers’ electricity demand (and on some tech-company leaders’ personal nuclear enthusiasms) proves valid, a well-functioning market would continue to choose efficiency and renewables instead. Three reference points:

• ~90% of world capacity additions are renewable;
• In an average year, nuclear power adds ~1 GW/y of global capacity, net of retirements –the same capacity that renewables now add about every 12 hours, or the same output that renewables add every day or two; and
• Google has announced investments ~40× bigger in renewables than in nuclear.

Color me skeptical about AI. We don’t even know yet if it has a general business case. Will it secure broad popular markets (not just narrow technical applications, some of which will also save energy) that justify its high costs? Will its wide use survive differentiation by competitors offering faster, cheaper, more accurate pre-AI searches for ordinary users who may resist paying more for claimed AI advantages they don’t see or want? Even if so, then the many layers of fallacy in conventional electricity demand projections for data centers will keep emerging in literature and in markets. The last [pre-AI] data-center design I worked on found ~95% potential total electrical savings at about half the normal total capital cost. With good software and compilation, potential electrical savings could rise to 2–3 orders of magnitude. Nvidia press releases imply that the potential for AI system efficiency is probably much larger.

Ryan: Lastly, while I have you here, I have to ask. You’re famously fond of orangutans. What’s the backstory?

Amory: I do have a long and strong affinity with orangutans – perhaps I’m a born-again Bornean – so I help protect them and their habitat through the remarkable work of Dr. Willie Smits, and go hang out with my orange swinger buddies when I can. Last autumn, my wife Judy and I were in a little boat 130km upriver from a West Kalimantan roadhead, helping release a rehabilitated orphan into a magnificent jungle. Visitors to our passive-solar banana farm in Old Snowmass (near Aspen) may also have noticed the ~50 taxidermically challenged [Orlon] orangutans in the front hall. They’re on standby in case of a banana emergency, which is what happens if our six banana trees all fruit at once – a quarter-ton of ripe bananas.

Ryan: Amory, thanks so much for your time, insights, and inspiration!