The Negawatt Revolution
-- Solving the CO2 Problem --

Keynote Address by Amory Lovins
at the
Green Energy Conference

Montreal 1989

Organized by CCNR


My 1976 article entitled "Energy Strategy: The Road Not Taken?"   which appeared in Foreign Affairs, suggested two ways in which the energy system could probably evolve over the next fifty years or so, using the United States as an example. If you divided by something like a factor of nine or ten, you would get Canada.

One way [Figure 1, below]   would be to continue to use depletable fuels inefficiently and to continue converting them into premium forms -- mainly electricity -- in ever larger, more complex, more centralized plants.

We have discovered many reasons why that doesn't work. It is too little, too late, and too expensive. In fact it is internally inconsistent; in order to pay for those very costly supply expansions, one would need to raise the price of energy to a level at which people would no longer demand enough to pay for the supply. Also, of course, this sort of future has very disagreeable side effects, and leads to a very fragile or brittle energy system that is easy to disrupt accidentally or deliberately.

Over the last ten years, we have taken quite a different course [Figure 2],   rather like what I have called the "Soft Energy Path". This results from redefining the energy problem. The relevant question is not simply where to get more energy, of any kind, from any source, at any price. Rather, it is a series of inter-linked questions. What do we want the energy for? What are the end uses we are trying to provide, such as comfort and light and torque? And how much energy, of what kind, at what scale, from what source, will meet each of those end use needs in the cheapest way?

This so-called "end-use, least-cost" approach has turned out to be a very powerful tool for understanding how people behave in a competitive energy service market where all ways to make or to save energy must compete with each other. In fact we have come quite close to realizing a pattern of evolution in which fossil fuels are being gradually squeezed out by a combination of energy efficiency (which provides growing services while reducing energy needs), and appropriate renewable energy sources (what I call the soft technologies).

We are now a quarter of the way through that quarter-century illustrative projection, and we are doing a little better than anticipated. It is interesting to note that renewable energy supplies are now right on that bottom line, 11-12 percent of total U.S. primary supply. Indeed, it is the fastest growing part -- except for efficiency, which has reduced total energy use to about 11-12 percent below my earlier projection [the top line in the graph, Figure 2].  

So we are not doing so badly despite a decade of what our Soviet colleagues aptly call the period of stagnation. Over the past decade my own country has gotten over seven times as much new energy from savings as from all net increases in energy supply combined. And of the new supply, more has come from renewables than from nonrenewables. The energy savings already achieved since the Arab oil embargo in 1973 have reduced the retail energy bill of the United States by 150 billion dollars per year. Think of that as one [U.S.] deficit. However, if we were now as efficient as our competitors in Europe and Japan, we would save an additional one and one-third deficits [i.e. $200 billion]. And if we choose the best energy buys for the rest of the century, we could get accumulated net savings by then of several times 10 to the 12th [i.e. several trillion] of today's dollars -- enough to pay off the whole [U.S.] national debt!

The Canadian progress, I think, has been probably parallel -- weaker in some areas, stronger in others. Certainly both our countries have a good deal to be pleased about in the efficiency progress so far, even though we have done it with really very modest improvements in the car fleet, plugging steam leaks, insulation, caulk guns, duct tape -- nothing very fancy or difficult.

To give you a notion of how much further there is to go, lest we congratulate ourselves prematurely, the energy now wasted in my own country each year costs more than the entire military budget of $10,000 a second. It costs about two deficits -- over $300 billion per year!

Energy and the Economy

That has a strong bearing on international competitiveness, as it does also for Canada. For example, the United States spends over twice as much of its wealth on energy as Japan does, by percentage. That gives a typical Japanese export an automatic cost advantage of 5 percent. But it is really much worse than that, because when you use energy inefficiently, you leverage enormous sums of capital into inefficient energy supply systems. Even today, in the twilight of the big steam plant era, the United States, for example, is spending about sixty billion dollars per year -- half of it in private investment, half in federal subsidy -- on expanding electric supply. That is about the same as our total investment in all durable goods manufacturing. It is not surprising that our durable manufacturers are uncompetitive in the world market; we are spending our money on the wrong things.

If the U.S. spent only enough on efficiency to keep up with growth and demand for electric services, plus the net retirement of generating capacity, we would have almost enough capital left in surplus to double our rate of investment in durable manufacturing industries.

And countries like Japan are not only more efficient than the U.S. or Canada, but are becoming more efficient much faster. In major industries recently analyzed, electric intensity per tonne is falling in Japan but rising in the U.S. (as also in Canada). Why? Because they are getting efficient faster than we are. So we have some catching up to do.

The Importance of Electrical Efficiency

Why do I concentrate on electricity? First, because it is by far the costliest form of energy. Each cent per kilowatt-hour is equivalent in heat content to oil at $17 dollars a barrel, roughly the world oil price. So the electricity we buy, even in Canada where it is quite cheap, is equivalent to heat at many times the world oil price. Therefore saving electricity is more financially rewarding than saving direct fuels. In addition, electricity has enormous capital leverage because central electric systems -- the whole systems -- are about 100 times as capital intensive as the traditional direct fuel systems (you know, Texas and Louisiana and Alberta oil and gas -- the sorts of things on which our economies were built). In fact a quarter of all the development capital in the world goes to electrification.

Also electricity has huge environmental leverage. Power plants burn a third of the fuel in the world. They account for a third of the CO2, therefore, released from the burning of fossil fuel. In my own country they release two thirds of the sulphur oxides and a third of the nitrogen oxides. What's more, every unit of electricity you save at the point of use saves typically three or four units of fuel, namely coal at the power plant. And in socialist or developing countries that ratio is more like five or six to one.

So you get the most environmental benefit from saving electricity, as well as the most financial benefit.

It is therefore paradoxical that almost all the energy savings so far achieved have not been in electricity but rather in oil (and to some extent in gas) because those are the things we were worried about. That is partly because electricity is much more heavily subsidized than direct fuels -- roughly eleven to one in the United States. I don't think we know what the number is in Canada. Of course, that sort of distortion makes the electricity seem disproportionately cheaper than it really is.

Four Revolutions in Electrical Efficiency

Fortunately, four revolutions have lately occurred that will let us reverse that misallocation of effort. We will henceforth be able to save at least as much in electricity as we save in fuel, thereby capturing the benefits that are waiting for us.

These revolutions pertain to:

  1. new technologies to wring more work from the electricity we already have;
  2. new ways to finance and deliver those technologies to the customers who need them;
  3. changes in the regulation of electric utilities;
  4. changes in the utilities' mission and culture.

Let us begin with the new technologies. Most of the best ways we have today to save electricity while providing unchanged or improved services were not on the market a year ago. You can now save twice as much electricity as could be saved five years ago, at only a third of the real cost. That is a six fold expansion in cost effective potential in five years, and almost a thirty fold expansion in ten years.

Let me give you some examples of what technologies I am talking about. All the ones I will show you are on the U.S. market. Many of them are on the Canadian market. Some of them may not be on the market yet, whether because of various kinds of protectionism, or inertia, or whatever.

To begin with, what is electricity used for? I am sorry I don't have a graph here of the Canadian end use structure, but it is not different in very important ways from the American one, which is like this [Figure 3].

About a fifth of U.S. electricity goes to lighting directly. However, that fraction becomes a quarter when you count the indirect space conditioning effects of the lighting. (In big buildings -- even in your climate -- the cooling system is often run all winter, and that is mainly to take away the heat of the lights. Otherwise people roast, because these big buildings are dominated by the internal heat gains, not by the heat lost through the shell.) Something like half of the electricity goes to motors. Some of it goes to cool space and food, but most of it is used in industry. About a sixth of the electricity goes to low temperature heating of space and water -- that fraction is probably a good deal higher in Quebec. About a twelfth goes to high temperature heat, a twentieth to electrolysis, a fiftieth to electronics, and so on...

Total Savings Potential from Lighting

Roughly half of U.S. (and I suspect half of Canadian) lighting energy is in fluorescent systems. There is a package of things you can do to them that will save 70 to 90 odd percent of their electricity, at an average cost of about half a U.S. cent per kilowatt-hour. The first thing you normally do is to install above the lamps a specially shaped, computer- designed reflector, made of a very shiny kind of metal that is precisely bent and installed in a certain shape. The reflector roughly doubles the optical efficiency of the fixture. That is, you can remove half the lamps, and still get the same amount of light underneath that you started with.

The ballast -- the device that controls the current to the lamps -- can be replaced with a remotely-dimmable high frequency electronic ballast. This is electronic Wonder bread: it saves electricity nine ways. The saving from the ballast is typically 50 to 90 odd percent depending on the control conditions.

Think of that as representing, in the U.S., approximately sixty (and in Canada, probably about six) Chernobyl-size power plants in savings potential. And the cost of the package, as we have shown by experiment, is only about half a cent per kilowatt-hour.

Another roughly 42 percent of U.S. lighting energy (in Canada the fraction might even be a little higher) goes to incandescent bulbs. Those we normally replace with compact fluorescents which miniaturize the features just described. There are integral compact fluorescent lamps in all sorts of sizes and shapes that have roughly quadrupled efficiency -- 11 watts replacing 40, 18 watts replacing 75, and so on. They last about thirteen times as long as a regular light bulb; therefore each one of them saves you not only three quarters of the electricity, but also a dozen replacement bulbs and trips up the ladder. That more than pays for them, even though these things are rather expensive.

Think of such a compact bulb, with 14 watts replacing 75, as a 61 negawatt power plant. By substituting 14 watts for 75 watts, you are sending 61 unused watts -- or negawatts -- back to Hydro, who can sell the electricity saved to someone else without having to make it all over again. It is much cheaper to save the electricity than to make it -- and not only in thermal stations. It is cheaper for society to use these bulbs than to operate a Hydro plant, even if building the dam were to cost nothing. Each bulb has a net cost of minus several cents per kilowatt- hour, and no dam can compete with that!

We have added up in considerable detail, based on measured cost and performance data, the potential for saving U.S. lighting energy. I think the Canadian potential will be essentially identical because we use the same hardware. It is about a 92 percent savings potential. There go (in the U.S.) about 50 more 1000-megawatt power stations -- and about 5 in Canada.

There are a lot of tricks we did not assume in that 92 percent savings figure. For example, here is a Fresnel-prism-grooved acrylic sheet from 3M. A firm called TIR Systems in Vancouver engineers this. You roll it up in a tube, put it in a plastic pipe and it is now a total internal reflection light pipe. You feed concentrated sunlight in one end, from which you can remove useful heat for other purposes. You feed in the other end a supplementary artificial light from an efficient source, dimmed with a photocell to give you only as much light as you need at the time. You run this along the ceiling. Wherever you want light you stick in an extractor film to counteract the prisms, and the light shines out. You're now using the same distribution system for daylight and artificial light. You can therefore get daylight way into the middle of a big building. And daylight is what people are designed for. The Japanese do the same with fiber optics. There are many other tricks of this sort.

Lighting improvements in the commercial sector (because of their negative costs and their net space conditioning effect) amount to roughly half of the total savings potential. It represents, in fact, about three quarters of the total electricity used by the commercial sector buildings served by Arkansas Power and Light Company. If you ran the same exercise for the commercial sector in Montréal it would yield a similarly interesting set of numbers.

This is especially so when you correct for the fact that the new more efficient chillers in big buildings can be made smaller because there's less heat for them to take away. You actually save more money by making them smaller than you have to spend to make them almost twice as efficient -- so here you can make money before you even start saving electricity. That's why I think commercial lighting is probably the biggest gold-mine in the whole economy.

In round numbers, you can save a quarter of U.S. electricity in lighting, a quarter in motors and a quarter in everything else. Let me say a bit about the motors.

Potential Savings in Electrical Motors

We've just completed what is, I think, the first systems analysis of motor systems. Motors turn out to use over half the electricity in either of our countries and in the world. In fact, motors use more primary energy, more fuel at the power station than we put into all highway vehicles, and yet we know very little about the stock of motors. We know a lot less about the stock of motors than we know about, say, moon rocks. Nonetheless we do have enough data pieced together from here and there to analyze what you can do with about 35 classes of measures applied to the motors, their choice and sizing and maintenance practices, four classes of electronic controls, improvements to the electric supply system and improvements to the mechanical drive train that transmits the power to the machine.

Just to make that last point a little more explicit, a great many motors transmit power with V-belts which have a tendency to stretch and slip. They need a lot of maintenance. They wear out. They need retensioning. They're generally a mess. They ought to be replaced with synchronous belts, which engage teeth in the sprockets so they don't slip. Also they don't stretch, because they have fiberglass or Kevlar bands inside them -- rather like in a radial tire -- so they last practically forever. They save typically 5 to 15 percent of the energy. They more than pay for them selves by the saved maintenance costs.

Even in the most classical areas -- Victorian technologies like gears -- we're still making remarkable progress just reinventing wheels. For example, a 42 to 1 speed-reducing gear which loses only four per cent of the energy going through it, to replace a worm gear that loses about half the energy that goes through it. There go another few power plants!

Now when you add up all of these improvements to drive power systems, you find a saving potential of roughly half, at an average cost well below half a U.S. cent per kilowatt-hour. Why is it so cheap? Well, because out of all the 35 classes of measures, you only need to pay for seven. The other 28 are free. For example, when you buy a new high efficiency motor, you will automatically correct in the process any damage done to the old one by bad rewinding (which is endemic, frying the core of the motor). You will correct inefficiencies resulting from gross oversizing, because you'll make the new motor the right size. You'll get a better match to different kinds of electronic controls. You'll get a better power factor. And so on. You get many benefits for one expenditure. This represents a total U.S. savings potential of about 80 to 190 big power plants, plus an operating cost net saving of many tens of billions of dollars per year. The average payback, at five cents a kilowatt-hour, is one-and-a-quarter years.

I'm not even counting here, by the way, the downstream savings in and beyond the machine being driven by these motors. For example, two-fifths of our motors run pumps. You can often double the efficiency of the pump by designing it and specifying it better. The pump then sends fluid through pipes. Suppose you make the pipes bigger, so there's less friction. Every unit of frictional energy you save in the pipe will save about nine units of fuel at the power plant. A lot of leverage. So you really ought to start at the far downstream end of whatever system you're adjusting and fix that part and then work back to and through the motor in order to get everything the right size.

The third quarter of the electricity you can save is in everything else; that is, building shells, appliances, miscellaneous equipment. The key to it, actually, turns out to be advanced windows, representing a U.S. savings potential of one Alaska. They insulate twice as well as triple glazing, but cost slightly less than triple glazing. Instead of a third layer of glass in the middle, it has only a 50-micron polyester film, with high-tech coatings on it that let visible light come through but reflect infrared in either direction. So it keeps you cooler in the summer and warmer in the winter. To make it insulate better it's filled with argon. Argon is almost one percent of the air you're breathing, so it's quite cheap, but it insulates one third better than air. Such a window, in most North American climates, will gain more heat than it loses, even in winter, facing in any direction, including North, because the diffuse skylight plus the light bouncing off the snow gives more energy coming into the house than the window loses.

Accumulated Electricity Savings

The next graph is a supply curve [Figure 4, below].   The vertical axis is the cost of saving electricity, and the horizontal axis is how much you can save. For convenience I have lumped together all the things you can do in each end use. At the far right is the total amount of electricity bought from the grid. The first rectangle represents the 120 gigawatts of lighting savings, the next signifies the 20 gigawatts or so from eight different savings measures in water heating, then come the drive power savings, and so on. By the time you get up to the space heating savings [all the way to the right]   you are saving about three quarters of the electricity in the country -- give or take about 10 percent of uncertainty.

The marginal cost of doing that is comparable to the cost of fueling a typical fossil fueled power plant and maintaining it. It is cheaper than running a typical nuclear plant. Moreover, the average cost of saving three-quarters of that enormous potential is only about 0.6 cents per kilowatt-hour. (The exact figure is a little uncertain, but it is certainly less than 1 cent.) Note also that you can accomplish all those savings indicated by the shaded rectangles at zero net cost, because the area below the axis [representing negative costs or savings]   equals the area above the axis [representing positive costs or expenditures].  

That is very good news for things like global warming and acid rain. Imagine being able to save half the electricity for free, and still get the same or better services! In lighting, for example, you get the same amount of light as before, with 8 percent as much energy over all -- but it looks better and you can see better. In drive power, you get the same amount of torque -- but it is more comfortable and more reliable. In the space conditioning case -- heating and cooling -- you get improved comfort. And so on. It is doing more with less.

The Implementation Revolution

But none of these technologies can do you any good unless there's a way to get hold of them, pay for them and get them properly installed. That brings me to the implementation revolution, which is just as important. And here I think the electric utilities, whether publicly or privately owned, have a very important role to play because after all they have a duty to serve whatever demand you place upon them. If you don't get efficient, they're going to have to build more power plants. Their payback horizon -- how fast they want their money back when they build a power plant -- is typically about 20 years. Yet when you invest in saving energy, you probably want your money back within two years, whether in your home or your business.

As long as you evaluate the ways to save electricity at roughly ten times the time value of money for which they evaluate the power plants which they would otherwise have to build, you'll buy too little efficiency, Hydro will buy too much capacity, and you'll continue to misallocate, in Canada, probably on the order of six billion dollars per year. It cripples your economy. The same goes for us in the U.S.

To put it another way, price alone is not an effective way to get you to buy as much efficiency as is worth while. Because the price, after all, is something the utility sets based on its very undemanding investment criteria. If you can avoid a tariff of 5 cents a kilowatt-hour by investing on your own to save electricity, and you have 10 times the discount rate the utility has, you're only going to buy efficiency costing about 0.5 cents per kilowatt-hour. The difference between 0.5 cents and 5 cents is called the payback gap. It comes from the roughly ten-fold difference between a 2-year and a 20-year payback horizon. We can argue about exactly what the numbers are, but I think the principle is clear. To allocate society's capital efficiently to meet its electrical service needs, requires that all ways to make or to save electricity be compared with each other on an equal footing, at the same discount rate.

One obvious way is to have the utility invest on your side of the meter to save electricity whenever it's cheaper to do that than for them to run or to build (as the case may be) capacity. There have been very exciting developments in how to do that. First, with what I'll call "old methods" -- not in any pejorative sense, but just to say that they are better under- stood and better established than the newer methods I'll describe in a moment. These old methods essentially market negawatts, through general information or specific technical information and design support, or concessionary loans which over half of US customers can get to save heat in the household.

It was found to be cheaper to give away that weatherization than to service and subsidize the loan. In fact, Southern California Edison Company has given away 600-odd thousand of the Japanese version of these energy-efficient light bulbs because it costs them less to do so than to operate existing power stations.

Rebates have been the backbone of most implementation programs. The utility pays the customer to use electricity more efficiently, either with a specific device or in any other way. Many utilities found that, instead of paying the customer directly, a better, cheaper result may be obtained by paying the retailer, the wholesaler, the installer, the specifier, or somebody else involved in your making the decision. So you pay whoever you have to to make the sale. You can pay people to scrap their old refrigerator rather than continuing to run it out in the garage. If there's a standard, you can pay people to beat that standard and thus elicit new improved technologies. Then you can raise the standard -- until you run out of cost-effectiveness.

Some utilities now lease (for example) energy efficient light bulbs for 20 cents per lamp per month on the electric bill, with free replacements. This encourages third party investors and energy service companies, which may in fact be owned by the utilities themselves.

Now these "old methods" are so effective that if all Americans saved electricity at the same speed and cost at which 10 million people in Southern California did save electricity by those means in the mid- eighties, then America's national forecast needs for long-run power supplies would go down by about 40 gigawatts per year. We could grow the economy by several percent a year, and shrink our absolute electric demand by several percent a year at the same time.

And the total cost for the utilities to do this, again based on what it actually cost in Southern California, would be about 11 percent of the cost of new power stations.

The Negawatt Marketplace

So far, so good. However, I think we have even better methods now, which don't just market negawatts, but create a market in negawatts. These methods, in other words, make saved electricity behave like a commodity -- like oil, and copper, and wheat, and sowbellies -- so that everybody can play. One way to do that is competitive bidding. After all, why offer somebody a rebate of $300 to save a kilowatt by, let's say, air-conditioner controls, when you can save the same kilowatt by planting trees to cool off the city for $10? In order to get the cheapest savings first, and to maximize competition between providers of savings and methods of saving, you need to let each possibility bid against all other opportunities. This started with the Central Maine Power Company, which offered cash grants to the factories it served to help them become more efficient. The company simply said, "We will offer the money to whichever firms offer to save the most electricity per dollar of grant." So they bid against each other.

This worked so well, it became a generalized option. If some company wants 30 more megawatts next year, it says, "Who out there would like to make or save or displace 30 megawatts, or some piece of it, by any means of your choice? At what price? We will take the low bids (which turn out to be energy savings)." At least eight states are now doing this and it's rapidly spreading nation-wide.

Of course, it's basic to any commodity that it has to be tradable across time and space. We already have had the first few deals signed in which utility A pays utility B to save electricity in B's territory and sell it back to A at a discount. Los Angeles is paying for electricity to be saved 1,000 miles away in eastern Montana and then sold back to Los Angeles. You can do the same thing between customers. If I, let's say, run a paper mill in Québec and I want cheap electricity, I ought to be able to come here and fix these ridiculous lights over our heads (they are all little cash cows waiting to be milked!) and then have Hydro-Québec sell my paper mill, the same amount of electricity I saved here -- wheel it back to me in effect -- but sell it to me at a discount so that I share the saved operating costs between myself and other customers. In that way Hydro would, in effect, declare open season on inefficiency. Anybody could go bounty hunting, and there would be a great deal of competition among Québec entrepreneurs who wanted to make a profit from this sort of very lucrative opportunity.

Now you can also do the same thing between countries. For example, Hydro-Québec wants to sell 450 megawatts of extremely expensive new hydro-power to Vermont for a cost which over time works out to about nine U.S. cents per kilowatt-hour. Vermont's not interested. That's too expensive. But Vermont has a brilliant counter-offer, namely let us come to Montréal, let us fix up your buildings (hiring local contractors to actually do the work) and we'll save 450 megawatts there. Then we'll buy it back from you, for three cents a kilowatt-hour. It will cost us maybe one cent to save each kilowatt-hour, and three cents to buy it back again. That's four cents, which is less than the nine cents you were offering. So we like that, but you at Hydro-Québec will make more money this way too, because the power that we save and buy back from you will be from an old dam you paid for twenty years ago. It's pure gravy. It's more profitable than your very expensive new dam at nine cents because, in effect, we'll have built you a cheaper dam in downtown Montréal, avoiding all the environmental and financial risks.

In other words, the power you have in Québec is much more valuable saved and resold than wasted at home. Notice that this amounts to pure arbitrage on the spread between the cost of making and saving electricity. It means that your most valuable asset in Québec, in terms of your electric system, is not the not-yet-flooded Cree hunting grounds, rather it's right here in this room, above our heads.

The Role of Regulation

When regulators set an allowed rate of return (or however else they go about it) what they are actually doing is setting up a tariff structure that says "We expect you'll sell this much electricity next year. If you do it at that price, you'll earn that much revenue, which is the proper amount for you to bring in." But if the utility manages to sell more electricity, than projected, its profits go up; if it sells less than projected, its profits go down. The utility is rewarded for selling more and penalized for selling less! That's what's done. Even if it's a non-profit utility, like many of yours, the same principle still applies. If it saves money for its customers [by reducing demand through energy efficiency]   all of the benefit goes to the customers, and none to the corporation.

In the U.S., our national association of regulatory utility commissioners (which has Canadian observers as well) has unanimously approved in committee, has also approved at the executive level and is now sending to the floor for very likely passage in November, a completely new principle of utility regulation: namely, decouple utilities' profits from their sales. That way, if they sell less than expected they're not in the hole, and if they sell more than expected they don't benefit from it. They're in different, then, to whether they sell more or less. If they do something smart to cut your bill, let them keep part of the saving as extra profit, or in some other way give them an exemplary reward for efficient behaviour, so they really have an incentive to be efficient.

The Culture Revolution

That's the implementation revolution. Let me just say a quick word now about the culture revolution within utilities. You notice that there isn't any demand for electricity for its own sake. What people want is the services it provides. The electricity is only a means to an end. It's what economists call an intermediate good. Nonetheless, most of our utilities, in both countries, have gotten into the habit of thinking they're in the kilowatt-hour business, so they should sell more. They've gotten into the terrible habit of looking at the top line instead of the bottom line, because for a century they've had sales and revenues going up together. For some reason, it's hard for them to get used to the idea that it's perfectly all right to sell less electricity, and so bring in less revenue, as long as costs go down more than revenues do. Now anybody who sells shoes knows that you can make money on margin instead of volume. But for some reason this is not a familiar concept in the utility business. It's time it were. Because their mission is not the sale of kilowatt-hours, it's the production -- and for a privately-owned utility the profitable production -- of customer satisfaction.

I think in an era of costly electricity and cheap efficiency, it's logical to expect that customers will sooner or later discover that they should buy less electricity and more efficiency. The only question is who's going to sell it to them, not whether it will happen. (With varying degrees of hassle, of course.) Therefore I think the only real choice that electric utilities face is between participation in the efficiency revolution, or obsolescence. A minority of U.S. utilities now understand this. They're trying to sell less electricity and more efficiency. And it's now becoming obvious that they're making more money, having more fun and attracting more talented people because they're much more interesting to work for. So this minority of utilities -- that see themselves not as commodity vendors, but as competitors in the energy service marketplace -- that minority is growing rather quickly. And I am very pleased to find some people in the Canadian utilities who are now thinking the same way and acting on it.

Economic Spinoffs

Finally I'd like to suggest some nice things that the negawatt revolution can do. First of all, for economic development. This applies to every little town in Québec. Consider the municipal utility in Osage, Iowa (population 4000). For nine years they've been helping people fix up their houses, to be more comfortable using less electricity and gas. This has saved the utility so much money that it was able to prepay all of its debt, build up a large surplus, and cut the tariff five times in five years. (In fact in real terms it went down by a third. It's now half as high as the typical level in Iowa.) This situation has attracted two new factories to town. Most importantly, the efficiency gains have kept over $1000 per household per year recirculating in the local economy -- money which previously went out of town and usually out of state, to buy utility inputs. But now that money is largely staying on Main Street, being spent and re spent locally, supporting local jobs and multipliers. That's the most powerful form of economic development we know.

Another example. I know you have some troubled industries in Québec. You might want to consider the example of the largest U.S. independent maker of rodwire and cable, South Wire. This firm had two smart engineers, who, over the course of six years, cut by 56 percent the company's total use of energy per kilo of product. It's hardly an exaggeration to say those two engineers may have saved 4000 jobs at ten plants in six states. Of course, many industrialists don't yet realize why this is important because in manufacturing they say "Oh, energy, that's only 2 or 3 percent of our cost of doing business." That's again the fallacy of looking at the top line instead of the bottom line, which is where the savings go to.

Environmental Spinoffs

Let's talk about the environment. One compact fluorescent lamp, over its life, will keep from putting into the air from coal plants about a ton of CO2 and eight kilos of SO2 and various other bad things. Or, if it's displacing nuclear generation, it will avoid making half a curie -- which is a lot! -- of long lived radioactive waste, and about two-fifths-of-a-ton high explosive equivalent of plutonium. Or, if it's displacing oil-fired electricity -- as I suppose it might in the Maritimes, or Long Island, or Hawaii, or developing countries -- one such lamp will save enough oil to run your family car 1000 miles or to run a super- efficient prototype car coast to coast round trip.

One lamp does all that! But far from paying extra for it, you find it makes you about 20 or 30 dollars richer, because that's the amount by which its savings in fuel, lamps and labor exceed its cost. And it will also defer hundreds of dollars of investment in utility capacity, money which can be more productively invested elsewhere in the economy.

That's a nice example of how you can abate many kinds of environmental problems, not by paying extra, but by making money on the deal. For example, acid rain. There are some Canadian coal-fired power plants that put out a lot of acid rain. Rather than raising people's electric bills to put diapers on those plants, suppose we help people get super-efficient lights, motors and appliances, so that Ontario Hydro, say, can burn less coal and emit less sulfur. But the main effect will be that Hydro will save a great deal of money, because efficiency is cheaper than coal!

Solving the CO2 Problem

Much the same principles apply to abating global warming, which is the last topic I'd like to address. Let me make clear, first, that it doesn't matter whether global warming is happening, or will happen, or what it would be like -- because we should do the same things about it, regardless! The scientific uncertainties are essentially irrelevant for policy. Here's why.

There are three main things we need to do to abate global warming. One of them -- and this will handle over half the problem -- is energy efficiency. But it costs less to save fuel than to burn it, so the cost of this abatement measure is strongly negative. Another thing we ought to do is promote sustainable farming and forestry practices. That will solve at least a quarter of the problem. But those are at least as profitable, in general, as what we do now -- often more so -- and there are other good reasons to do them anyway. Let's call that roughly a break-even situation. Finally, there's displacement of CFC's [chlorofluorocarbons].   That's about one sixth of the problem. But we need to do that anyway to protect the ozone layer, so its cost is irrelevant.

Now if the cost of abatement of global warming ranges from strongly negative, to roughly zero, to irrelevant, then why fiddle while coal burns?

Notice also that if abating global warming is going to be highly profitable, then the CFC Montréal protocol model is not appropriate for CO2. The Montréal protocol basically says that a bunch of countries will get together and do something costly and inconvenient in order to share a sacrifice for the common good. Now that may be an appropriate philosophy if indeed the substitutes cost more and work worse (and some of them probably do, although some may actually be a better deal than what we have now). On the other hand, since saving fuel is cheaper than burning it, we ought to do that not in a spirit of sacrifice, but largely through the market. Because people are going to make a lot of money out of it -- specifically, world-wide, about a trillion dollars a year, which equals the entire global military budget.

So rather than having a "treaty for mutual inconvenience", we ought to be helping nations, firms and individuals behave in their own economic self- interest. An important point that we all need to understand better is that the order of economic priority is also the order of environmental priority. That is, if we don't do the cheapest things first, we make global warming worse than if we did.

Nuclear Power and Global Warming

Let me give a specific example. My colleagues, Keepin and Kats, have looked at what it would take to abate some of the global warming trends using nuclear power. They said, "Suppose we try to displace all coal-fired electric generation with nuclear by 2025 throughout the world, assuming that nuclear plants cost one-third as much as they really do, and can be built twice as fast as they actually can in North America." In that case, [see  Figure 5  below]   if we assume three commonly used projections of global energy demand compared to today's, we would need this many 1000-megawatt nuclear plants [gesturing at Line 2 in the chart],  which means spending this many billion dollars per year [gesturing at Line 4]   -- again, assuming roughly one third of actual costs, and not counting any of the added fuel cycle and grid costs and so on.

Figure 5

Trying to displace all coal-fired power plants with nuclear plants by 2025

Assuming nuclear plants cost $1 per watt and can be built in 6 years
(~1/3 and ~1/2 of current US norms)

 LINE         ENERGY DEMAND PROJECTION:               HIGH       MEDIUM      LOW
 (1)  2025 world primary energy demand (1988 = 1.0)    3.6         2.1       1.1

 (2)  2025 nuclear gigawatts at 65 percent capacity  8,180       5,200     1,600
 (3)  average number of days for each GW installed     1.6         2.4       7.5
 (4)  average 1987 billion $ per year investment       229         151        49

 (5)  approximate percentage decrease by 2025         20-30      20-30        15

And when we get all through doing that we would only abate global warming by a rather modest amount as you can see [gesturing at Line 5 in the chart].   Why? Well, because only half of global warming is due to CO2, and only about four-fifths of that is from fossil fuel. (The rest is from deforestation and biological simplification in agriculture.) And only one third of the fossil fuel CO2 comes out of power plant smokestacks. So you're only dealing with a very small part of the problem when you displace coal-fired power plants.

Suppose you took this several trillion dollars in the right-hand column -- that's the only column that looks remotely feasible, and that's only because the demand, some of which you're trying to meet with nuclear, has been heavily pre-shrunk by strong efficiency improvements -- suppose you take the several trillion dollars that that modest nuclear contribution would cost you, and you put it instead into a little more efficiency, for example. Or for that matter, appropriate renewables. Where would that get you? That brings us to the opportunity costs of buying nuclear in the first place.

This graph [Figure 6]  shows what you might think of as cents-per-kilowatt-hour turned upside down -- in other words, how much coal-burning, or how many kilowatt-hours of coal-fired generation, you can displace per unit of money spent.

Now, nuclear power is expensive. We can argue about how expensive it is, but it won't displace much coal per dollar. In contrast, efficiency is cheap. We can argue about how cheap it is, but it's going to displace a lot more coal-burning per dollar. Therefore, whenever you spend a given dollar on nuclear instead of on efficiency, you're making global warming worse by releasing all this extra carbon into the air -- whatever's the difference in height between these two bars. A very simple argument. The same argument, by the way, applies to solar photovoltaics; the only difference is that they actually have a prospect of becoming cheap, and nuclear doesn't.

A book we did for the German government in 1981 entitled Least-Cost Energy: Solving the CO2 Problem sank without a trace, but we've just republished it. I think it was the world's first detailed low-energy scenario. We assumed a world of 8 billion people -- but completely industrialized, all with a 1975 West German standard of living. (Now this may be impossible for other reasons. It would make the World Product five times as gross as it was in 1975. Moreover, it assumes a ten-fold increase in economic activity in the developing countries.)

Nevertheless, assuming that -- just for the sake of argument, a sort of Los Angelization scenario -- if such a world used energy in a way that saved money, using 1980 prices and technologies, it would still require in total, only one-third as much energy as is used today. Moreover, each region could get as much as it would need of each kind of energy from renewable sources that were already available and long-run cost- effective in 1980.

Energy Efficiency in Poor Countries

This raises a very important point. If you look at the patterns of energy use in the world today, you find that the typical rich country is almost three times as energy efficient as the typical poor country. And yet we know how to make the typical rich country at least four times as energy efficient as it is now without coming anywhere near the limits of cost effectiveness, using present technologies & prices.

Now if you multiply those two factors, you find that a typical poor country ought to be able to grow its economy roughly ten-fold without increasing its energy use at all, if it leap-frogs over our mistakes and does it right the first time. I would suggest indeed that that is the only way it can afford to develop. Otherwise it will find itself in the dilemma of China, which decided some years ago that it was time for people to have refrigerators. So they built over 100 refrigerator factories. The fraction of Beijing households owning a refrigerator went from 2 to 62 percent in six years. Unfortunately, by not paying attention, China chose a very inefficient kind of refrigerator to build, and has therefore committed itself to spending several billion dollars it doesn't have to build power plants to run the refrigerators, thus creating power shortages in the name of development. India's doing the same thing.

If you give away light bulbs like these in a very poor country, like say Haiti, you can thereby increase disposable income by perhaps as much as one fifth, because so much of the sparse cash economy is going to electricity, mainly for lighting. If you give them to the North American equivalent of a Haitian -- let's say a North Carolina chicken farmer -- it will immediately increase his profit by a quarter. If you give them away in, say, Bombay, you will reduce by one-third the peak load that crashes the grid in the evenings. (If an industrialist wants to produce something in much of India, it's necessary to run bulky and expensive diesel generating equipment, because the grid is down a lot of the time.) So you see, you can achieve remarkable leverage for development by increasing the reliability of supply, and by avoiding heavy investments in expanded supply, through emphasizing efficiency.

Doing It Right

But "doing it right the first time" will mean that countries like ours will have to set a better example. We will have to export our most efficient technologies not our least efficient. Both the U.S. and Canada, I'm afraid, are often exporting technologies which are too obsolete to be sold in the home market. It's just like deregistered pesticides all over again. And developing countries, for their part, will have to become more sophisticated buyers and much more consistent and systematic users of efficient technologies. Because, increasingly, it's becoming obvious that about the only way to develop sustainably is to build efficiency into one's infrastructure the first time. Do it right from scratch.

Earlier, in the context of reducing CO2 emissions, I mentioned sustainable farming and forestry. Without going into all the reasons why it's a good idea, I'd like to emphasize that it would reverse the emission of carbon from the soil. In other words, it would take carbon out of the air and put it back in the soil where it belongs. If by switching to organic farming -- which has comparable yields and generally higher profits than chemical farming -- if by doing that, we were to build up the organic content of the soil (even several times slower than has been achieved in favorable conversions in the United States) then U.S. farmland would be able to take enough carbon out of the air to balance the carbon released by a moderately efficient U.S. car fleet burning fossil fuel gasoline.

That's a rather important contribution to abating global warming. So we should not just be stopping deforestation, not just planting trees (which is about a third best solution because it's a lot different from maintaining forests), but also rebuilding our biotic diversity below ground, where roughly half the standing biomass is. That, incidentally, will also greatly reduce emissions of other greenhouse gases such as methane and nitrous oxide.

A Cautionary Note

Let me conclude with a cautionary note. It's especially cautionary as I see the eager preparations being made to go ahead with the Darlington B nuclear station in Ontario, and the expanded James Bay hydroelectric project in Québec.

It is fashionable to say that we need to buy both supply and efficiency. We need balance. We need to do a bit of everything. (In practice, it's sort of like the classic recipe for elephant and rabbit stew: one elephant, one rabbit!) The trouble with that philosophy is that you may get neither, because they compete for the same resources. Even worse, you may get both efficiency and supply, and thus (as we've lately been doing) bankrupt the supply industries. Because to pay for costly supply ventures they need more demand, not less. If you get both supply and efficiency, and both succeed, then you get the worst of both worlds. You get the supply side costs without the revenue to pay for them.

It is rather sobering to remember that after two oil shocks, my government -- and, I think, your government -- still sought to boost supply while doing very little about demand. (Actually your government did a bit more than ours did in many respects.) Nonetheless, there was a gross over-emphasis on supply, and the policy landscape is littered with the wreckage: Project Independence, the nuclear power program, the western coal boom, the synthetic fuels corporation.... Remember those notions of enormous expansion in tar-sands? CANDU reactors all over the place? fast breeder reactors and so on? Well, all of those things failed, not only because they were often technically and politically infeasible, but most fundamentally because they couldn't compete in the market, which -- quietly, almost unnoticed, despite a lot of obstacles -- produced a gush of efficiency, and stuck the supply industries with unsaleable and costly surpluses.

There are many people still running energy policy in North America who would love to make the same mistake a third time. It is as if they had learned nothing and forgotten nothing, like an earlier ancien régime. 

Leadership and the People

I think with your help we're going to do better than that, because this group, I sense, understands something that Lao Tse taught us several millennia ago. He said

"Leaders are best when people scarcely know they exist, not so good when people obey and acclaim them, worst when people despise them. Fail to honor people, and they fail to honor you. But of good leaders who talk little, when their work is done, their task fulfilled, the people will all say, 'We did this ourselves.' "