To make a dollar of real gross domestic product (GDP) in 2017, the US used 65 percent less oil than in 1975 (despite 1982–2008 stagnation in new autos’ efficiency), 66 percent less directly used natural gas (direct fuel or feedstock, not power-plant fuel), and 57 percent less total primary energy. Yet electric intensity — total electricity consumed per dollar of real GDP — fell by only 31 percent. That’s less than half the percentage savings in oil or gas, the economy’s main direct fuels (since 93 percent of US coal is burned to make electricity). So why is electric intensity going down only half as fast as total energy intensity, especially fuel intensity? The answer to this riddle is complex but important.
IT’S NOT ABOUT PRICE OR POTENTIAL
Slower electricity savings aren’t due to relative prices. Producing and delivering electricity takes huge capital investments; generating power from fuel loses about two-thirds of its energy; and the grid loses another 5 percent or so. For these three reasons, electricity is even costlier than oil. Its 2017 average US retail price is equivalent in heat content (without counting relative efficiency of use) to crude oil at $180/barrel, 2.4 times the average world price. Thus cutting electric intensity would seem to have a strong financial incentive — yet it lags far behind.
Nor is the cost-effective potential to save energy smaller for electricity than for oil and gas: their potentials are at least comparable. Some engineers miss this point by noting that over half of electricity runs motors, mostly big ones that are already around 90 percent efficient. But in fact, the way those motors are specified and used cuts their typical operating efficiency by at least half. Even bigger losses arise downstream in the equipment motors drive, such as inefficient air conditioners cooling inefficient buildings, or inefficient pumps whose effort (in pumping loops) is roughly 90 percent wasted on needless pipe friction. The biggest unseen part of these opportunities is in whole system design: for example, the most efficient new and retrofit US office buildings were over twice as efficient in 2015 as they were in 2010, using the same technologies but in more intelligent selections and combinations.
During 1986–92, Rocky Mountain Institute conducted a uniquely detailed assessment of potential electric end-use efficiency: Competitek’s six-volume, 2,509-page, 5,135-footnote The State of the Art series. It showed that full practical retrofit with about a thousand technologies could ultimately save three-fourths of 1986 US electricity, at an average technical cost equivalent to about 1.2¢/kWh. (All costs in this article are in constant 2013 $.)
Some who hadn’t read the analyses, or their later summaries in the Technology Atlas series by RMI’s spinoff E Source, thought those savings sounded extreme. Yet the utility industry’s Electric Power Research Institute concurrently found, and summarized in a joint article, a potential to save 39–59 percent of US electricity just in the 1990s, at an average technical cost around 3.3¢/kWh. Comparisons by Oak Ridge National Laboratory and myself found that simple methodological differences accounted for virtually the whole disparity in the savings’ quantity and cost.
“That so much electricity-saving potential remains on the table testifies not just to electric intensity’s painfully slow decline, but to the constant innovation that keeps new low-hanging fruit ripening faster than it can be harvested.”
The target kept moving: efficiency opportunities grew more than they were captured. By 2011, RMI’s Reinventing Fire synthesis, relying mainly on National Academies and Lawrence Berkeley National Lab analyses, found that three-fourths of US 2010 electricity use could be saved by 2050 (and more thereafter) at an average technical cost of roughly 0.64¢/kWh — half the late-1980s cost. That so much electricity-saving potential remains on the table testifies not just to electric intensity’s painfully slow decline, but to the constant innovation — in design, technology, finance, marketing, delivery, and business models — that keeps new low-hanging fruit ripening faster than it can be harvested.
Utilities’ programs to help customers save electricity are not optimally designed and have transaction costs (albeit very small ones if well designed), so they’ve lately cost an average of roughly 2–3¢ per saved kWh, as documented by the American Council for an Energy-Efficient Economy, Lawrence Berkeley National Laboratory, and E Source. But that’s still cheaper than just running the average US coal (~3.3¢/kWh) or nuclear (~3.9¢/kWh) power plant, even if building it cost nothing. Moreover, efficiency is already delivered, but delivering the average kWh from a central station to your meter costs an average of ~4.1¢ to pay for the grid’s costs and losses.
So if neither potential savings nor relative prices explain why the United States has so far saved electricity less than half as fast as oil and directly used gas, what could? At least nine reasons seem plausible.
PRICES, SUBSIDIES, AND BILLING
First comes pricing. Fuel prices change far faster and are far more volatile than electricity prices, making efficient fuel use seem more attractive and front-of-mind. Unlike fuels, retail electricity is often still priced at its embedded average cost, concealing the often-higher marginal cost of new supplies or less-efficient existing supplies. The same practice often conceals the far higher price of electricity at peak periods or seasons: most large businesses pay time-varying prices for electricity and fuels, and everyone pays gasoline and diesel prices that vary with market prices, but few households or small businesses pay such realtime electricity prices. Indeed, hot afternoons downtown can cost utilities dollars to deliver a kilowatt-hour that they sell for dimes or even for cents; they rarely charge their real cost of grid congestion, but cross-subsidize it from sales at other times or to other customers. In contrast, fuel prices typically reflect actual delivery costs, and fuels that cost more to haul to remote and rural areas are priced higher. For social equity reasons, rural electric cooperatives like the one I belong to were therefore built with federal financing to help equalize electricity prices between urban and rural areas. Co-ops sell 11 percent of US electricity to 80 percent of US counties.
“The reasons electric savings have lagged fuel savings all represent business opportunities that will gain more attention as their financial rewards and carbon leverage become more obvious.”
Prices are distorted by subsidies. When most of the US electricity system was built, and probably still today (though modern subsidy analyses are sparse and often deliberately biased), electricity was subsidized far more than fuels. Rick Heede’s detailed RMI analysis, summarized in The Wall Street Journal on September 17, 1985, found that electricity got 65 percent of fiscal-year 1984 federal energy subsidies while delivering only 13 percent of the energy, cutting electricity’s price by about one-fifth. Electricity was at least 48 times more subsidized per unit than energy efficiency — and if made in nuclear plants, 80 times, getting 34 percent of the subsidies to deliver 1.9 percent of the primary energy. No wonder utilities were investing about $1 per household per day to build power plants they didn’t need and couldn’t afford: their subsidies nearly equaled their investment. That wasn’t a free lunch; it was a lunch the taxpayers paid them to eat. The feast continues: nuclear subsidies expanded in 2005 rivaled or exceeded construction costs even after those had risen severalfold, and the last two new reactors now being built, if completed despite their builder’s bankruptcy, would get bigger operating subsidies than wind power. Even today, America is far from energy prices that tell the truth. Energy subsidies, especially to traditional giant power plants and their fuels, are so entrenched that taxpayers’ largesse keeps rising when it should be eliminated.
Then there are promotional tariffs. Some electric utilities wisely charge higher prices for greater usage (“inverted block rates”) to reflect their higher costs of meeting increased demand, but promotional practices seem more common. Many utilities still discount and cross-subsidize electricity for some uses and users to try to boost demand — notably for electric heating and for big, relatively steady loads like data centers. Some electric utilities’ marketers work harder to sell more electricity than their efficiency staffs work to help save it. That’s rare with fuels: filling stations charge the same per gallon whether you’re tanking up a Humvee or a Prius. And as structural shifts in the economy make the next kilowatt-hour less likely to go to manufacturing than to an air-conditioned, computer-intensive office complex, utilities gain more incentive to load costs onto such commercial buildings so they can cut prices to more price-sensitive customers like industry and households — maximizing their own sales, revenues, and (absent regulatory reform) profits.
The way electricity is billed makes a big difference too. Every time you fuel your auto, you receive a price signal, and you know where the fuel went. If autos refilled themselves and drivers were auto-billed afterward, they’d spend more on fuel. But that’s actually how we buy electricity. Your monthly-in-arrears electric bill isn’t itemized, so you can’t tell which device used how much, and you only “see” how much total electricity you consumed over the previous month. It’s as if the supermarket posted no prices, you took home your cartful of food and ate it, and only then you got a single unitemized bill for the past month’s shopping — so how could you tell that tuna was costly and kale was a bargain? In contrast, prepaid electricity (the same as filling your car before you drive) creates vigorous investments in efficiency and demand management. And the more information customers have on where their electricity goes, the more mindfully they tend to use it.
Electricity’s wholesale costs are more dominated by fixed than by variable costs, compared to fuels, where the commodity price dominates. This gives electricity providers a bigger incentive to promote and sustain high and steady demand to cover the fixed costs of paying off their huge long-term capital investments — especially if they’ve overbuilt, as many have, or if they believe traditional rate-of-return regulation rewards them for investing more capital.
REGULATORY AND MARKET FAILURES
Misdesigned regulation also gives many providers of electricity utterly perverse incentives. Except in the 16 states (with seven more pending) that now reward utilities for cutting customers’ bills, not for selling customers more electricity, utilities have a direct incentive to sell more electricity. Conversely, if they sell less, their mainly fixed costs must be spread over fewer units of electricity sold, making electricity prices rise and further encouraging efficient use — the “death spiral” I described in Foreign Affairs in 1976. But there’s a smarter alternative. Stagnating or falling sales make electric utilities, like gas utilities earlier, more motivated to seek state regulatory reform that makes a virtue of necessity by protecting their revenues through “decoupling” from sales volumes (and preferably also sharing savings with customers). These reforms, officially favored by Edison Electric Institute and the American Gas Association (gas is decoupled in 23 states with five pending), should further accelerate efficiency as it becomes utilities’ most profitable investment.
Saving electricity faces more and tougher structural obstacles than saving fuel. Devices that use electricity are more likely to be bought by a different party than will pay the energy bills, creating a “split incentive.” Buildings use nearly three-fourths of US electricity, roughly half each in commercial buildings and households. In rental properties, why should the landlord improve efficiency for the building when tenants pay the utility bills, why should the tenants improve a building they don’t own, and why should they even improve their own equipment if their electric bill is prorated on floorspace rather than submetered? In households, many appliances are bought by a developer, landlord, or public housing authority who doesn’t pay the energy bills, whereas an industrial boiler, heavy vehicle, fleet van, or personal auto is more likely to be chosen by its subsequent fuel-buyer. No wonder electricity use in buildings is less sensitive to price than in industry.
Further, many smaller electricity-using devices — and, despite widespread Energy Star labeling (a wildly cost-effective voluntary information program that the White House wants to cancel), some bigger ones too — still lack the efficiency labels or standards that most fuel-using devices display, so buyers can’t as easily judge their efficiency as they can read the mpg sticker on a car. And the basic causes of electricity’s inefficient use are often built into long-lived capital stocks, like building envelopes, that turn over slowly and are harder to fix than, say, buying a more efficient furnace or car on a faster replacement cycle.
WE’RE A DECADE PAST PEAK ELECTRICITY
Given all these obstacles to using electricity in a way that saves money, it’s not surprising that US electric intensity didn’t start falling consistently until 1994. Nobody knows why that was the year the tide turned, but turn it did, and now this long-delayed trend — an average drop of 1.5 percent per year — is solid and strengthening. US electric intensity fell in 21 of the past 24 years, all but two of which experienced real GDP growth. Simple trend-line analysis (see figures on p. 7) shows that GDP growth slowed, electric intensity fell at a comparable pace, and absolute electricity consumption fell at their combined rate. Consumption peaked in 2007 and fell in six of the past 10 years. Its decomposed trend line hit zero annual growth in 2009 (2006 per capita, before the recession) and continues to trend downward. In 2017, GDP grew 2.3 percent while electricity use fell 2.1 percent, so electric intensity fell by a record 4.3 percent. Yet official forecasts still show 0.6 percent annual growth to 2050.
Similar trends are now evident in most industrialized and some developing countries. The rest mainly see slow demand growth that is rapidly tipping their over-ordered power supplies from scarcity to glut, turning supposedly vital new plants — especially Chinese and Indian coal plants — into prestranded assets.
New US building standards that came into force in about half the states in 2012–13, expanding private and utility investment in efficiency ($7 billion in 2013 from utilities alone), and more and better efficiency vendors, hardware, finance, and design methods all seem bound to speed this trend. LED lighting alone will ultimately save close to an eighth of the world’s electricity. And while fossil fueled and nuclear electricity keeps costing ever more, efficiency (like renewables) keeps costing ever less because it improves faster than it depletes.
EFFICIENT USE CUTS MOST UTILITIES’ REVENUES, NOT THEIR COSTS
Some utilities still cling to shreds of hope that electric demand will magically rebound. They’ll probably be disappointed. A decade after peak electricity, US utilities urgently need business models robust against the “new normal” of stagnant or shrinking demand — a foundation of the next economy. RMI’s Reinventing Fire shows that even a complete switch to electric autos, and extensive electrification of heat applications too, will offset only about half the decline in electricity demand.
Beneath the complex causes of falling electric intensity are two simple insights. First, customers are figuring out that they’ll get better service at lower cost by using fewer electrons more productively, so that’s the mix they’ll buy — from their utility or from someone else.
Second, the reasons electric savings have lagged fuel savings all represent business opportunities that will gain more attention as their financial rewards and carbon leverage become more obvious. Efficiency’s enemies keep trying to block it. But ultimately the fourfold, and by then probably greater, gain in electric productivity, already costing a tenth the average retail price — less than just operating existing thermal power stations — will be captured. Its economic potential will not forever languish unused.
As that potential is realized, utilities that sell electrons will face disastrous declines in sales and revenues. They can survive only by financing or providing the services customers want, like hot showers and cold beer — a model Thomas Edison pioneered in the 1880s so more-efficient lamps would reduce the costs and increase the profits of his lighting-services business. But he was overruled in 1892 when New York Edison Company switched to selling kilowatt-hours. Ever since, utilities have sold electricity (except in street lighting) as a commodity, so customer efficiency cuts their revenues, not their costs.
That upside-down business model cannot long survive customers’ accelerating switch to buying negawatts whenever they’re cheaper than megawatts — which nowadays is virtually always. So if you can no longer deny or overcome the fundamental forces that are making your customers buy ever less of your product, best to sell or lease them what they want electricity for, aligning your interests with theirs.
Updated from first publication by Forbes on April 25, 2017, at https://www.forbes.com/sites/ amorylovins/2017/04/25/why-are-we-savingelectricity-only-half-as-fast-as-fuels/.