Re: Kyoto, Driving our car (composite reply) [Part 2]
Dan Fabulich (daniel.fabulich@yale.edu)
Tue, 09 Dec 1997 16:41:53 -0500
>THE BEST - AND WORST - WAYS TO FORECAST FUTURE ENERGY AVAILABILITY
>
> The best way to forecast price trends is to study past price trends, if data are
>available and if there is no reason to believe that the future will be sharply different
>from the past. (The reasoning that supports this point of view is set forth at length in
>chapter 2.)
> For energy there are plenty of past price data available, as we have seen in figures
>11-2, 11-3, and 11-4. And there is no convincing reason to believe that the future will
>break completely from the past. Therefore, extrapolation of the trends in those figures
>is the most reasonable method of forecasting the future of energy supplies and costs, on
>the assumption that price has been close to cost in the past and will continue to be so
>in the future. This method of economic forecasting envisions progressively lower energy
>costs and less scarcity.
> Geologists and engineers, however, rely on technical rather than price-trend data in
>their forecasts of energy supplies. Because their forecasts have had so much influence on
>public affairs, we must analyze their methods and meanings.
> We must first dispose of the preposterous but commonly accepted notion that the energy
>situation can be predicted with the aid of "known reserves". This notion is an example
>of the use of misleading numbers simply because they are the only numbers available. We
>briefly considered the uselessness of this concept of "reserves" in chapter 2 with
>respect to mineral resources. Now let us discuss it with respect to oil.
> "Known reserves" means the total amount of oil in areas that have been prospected
>thoroughly, quantities that geologists are quite sure of. Individuals, firms, and
>governments create known reserves by searching for promising drilling areas long in
>advance of the moment when wells might be drilled - far enough ahead to allow preparation
>time, but not so far ahead that the investment in prospecting costs will not obtain a
>satisfactory return. The key idea here is that it costs money to produce information
>about known reserves. The quantity of known reserves at any moment tells us more about
>the expected profitability of oil wells than it does about the amount of oil in the
>ground. And the higher the cost of exploration, the lower will be the known reserves that
>it pays to create.
> "Known reserves" are much like the food we put into our cupboards at home. We stock
>enough groceries for a few weeks or days - not so much that we will be carrying a heavy
>unneeded inventory that bulges the cupboard and ties up an unnecessary amount of money in
>groceries, and not so little that we may run out if an unexpected event - a guest or a
>blizzard - should descend upon us. The amount of food in our cupboards tells little or
>nothing about the scarcity of food in our communities, because as a rule it does not
>reveal how much food is available in the retail stores. Similarly, the oil in the
>"cupboard" - the quantity of known reserves - tells us nothing about the quantities of
>oil that can be obtained in the long run at various extraction costs.
> This explains why the quantity of known reserves, as if by a miracle of coincidence,
>stays just a step ahead of demand, as seen in figure 11-5. An elderly man commented to me
>in the 1970s that, according to the news stories about known reserves, "we've been just
>about to run out of oil ever since I've been a boy." Yet most discussions of the oil and
>energy situation - among laymen and also among the most respected journalists - still
>focus on known reserves. Figure 11-6, taken from Newsweek, is typical. The graph
>apparently shows that the world's proven reserves have been declining, leading to the
>rhetorical threat above the picture "End of the oil? ... How much is left to find?"
> Figures 11-5 and 11-6
>
> Even more misleading is a graph of proven reserves in the U.S. alone, as in figure 11-
>7. As the U.S. turns to imports because they are cheaper than the home product, its
>proven reserves inevitably will fall. If one were to draw a graph of U.S. proven reserves
>of aluminum or gold, they also would appear tiny. So what?
> Figure 11-7
> A more "sophisticated" - and even more misleading - approach is to project present
>growth in demand, assuming the price will remain constant, and then compare that
>projection to known reserves, thereby indicating that demand will apparently outstrip
>supply very soon. This approach may be seen in figure 11-8. Even assuming that the
>growth in demand at present prices is reasonably estimated - and this would be difficult
>to do well - all that such a calculation would show is that price must rise in order to
>lower the demand and raise the supply until demand and supply meet. This basic economic
>way of looking at supply and demand is totally missing from figure 11-8.
>
> Figure 11-8
>
> Equally misleading is the assumption underlying figure 11-8 that there will be no
>developments in oil production or in other energy sources that will make future energy
>costs lower than they would be with the present state of technological knowledge.
>
>BETTER TECHNICAL FORECASTING METHODS
> If one insists on making a technical forecast of the energy supply - even though such
>a forecast is likely to be inferior to extrapolations of past economic trends - how
>should it best be done? That is, how might one make a sound material-technical forecast
>for oil and energy in the near term - say over the next ten or twenty years? (See
>chapter 2 for a general discussion of material-technical forecasts of resource supply.)
> During the next decade or two, increases in income and population in the U.S. and in
>the world may be assumed to be known. Therefore, they can be taken into account as data
>rather than treated as imponderables. In addition, forecasts of the production of energy
>in the near-term future utilize two other kinds of information: (1) engineering estimates
>of the cost of extracting fuel from such currently unexploited sources as shale oil and
>windpower with available technology, based on calculations of the engineering inputs
>required for each type of energy source, and (2) economic estimates of how many
>conventional new oil wells and coal mines and nuclear reactors will be developed at
>various prices higher and lower than the present energy prices, based on past data about
>the extent to which energy-producing firms respond to changes in market prices.
> Engineering estimates must play the dominant role in forecasts of the place of nuclear
>energy, shale oil, solar power, windpower, and other energy sources for which there are
>considerable uncertainties about technical processes and costs due to a lack of
>experience with these sources. But where an energy source is currently being employed
>sufficiently to produce a large body of data about the process of extraction and about
>producer behavior, as is true of the fossil fuels, empirical economic estimates of supply
>response to price changes should have the dominant role. The best overall energy
>forecast, therefore, would be a blend of both the economic and engineering approaches.
> There is great variety, however, in the estimates of engineers and scientists about
>the future costs of developing such energy sources as shale oil and nuclear power.
>Technologists also differ greatly in their guesses about the dangers to life from the
>various processes. And economists differ considerably in their estimates of the
>responsiveness of the energy industry to various price levels. For example, in 1977 the
>supply of natural gas became a very contentious political issue. These were some of the
>resulting supply estimates: 1) The predecessor agency of the Department of Energy made
>three production estimates within three months, varying by a factor of three! President
>Jimmy Carter offered an even lower estimate than the lowest of those three, that there
>was only "10 years supply ... at 1974 technology and 1974 prices". (2) The American Gas
>Association said that there is enough gas "to last between 1,000 and 2,500 years at
>current consumption." And the newspaper story continued that "Experts in ERDA [Energy
>Research and Development Administration] have been trying to tell the White House [this]
>too". The difference between this and the estimate in (1) above boggles the mind - 10
>years' supply versus a 1,000-2,500 years' supply! (3) A later "official" estimate, made
>in the midst of the congressional debate on energy in the same year 1977, by Dr. Vincent
>E. McKelvey, who was then director of the U.S. Geological Survey, was that "as much as
>... 3,000 to 4,000 times the amount of natural gas the United States will consume this
>year may be sealed in the geo-pressured zones underlying the Gulf Coast region". But this
>estimate ran contrary to what the Carter White House was saying, and within two months
>McKelvey was fired from his job as director - after six years as director and thirty-
>seven years at the Geological Survey, and after being nominated for the director's job by
>the National Academy of Sciences. As the Wall Street Journal put it, "Dr. McKelvey did
>not know enough to keep his mouth shut!" Such enormous variation can arise simply as a
>result of political fiddling with the figures.
> A more recent sober estimate by the "International Gas Union Committee on World Gas
>Supply and Demand estimates that even by the year 2000, the static lifetime of world gas
>reserves will be 112 years." - and that does not include future discoveries of gas, of
>course.
> With respect to still-undeveloped sources such as shale oil and artificial gas, the
>variation in estimates is greater yet.
>
> Why do estimates of supply response to price changes differ so widely? There are a
>host of reasons, including (a) vested interests - for example, the oil companies have a
>stake in lowgas prices paid to gas suppliers so that fewer gas wells will be drilled and
>more oil will be sold, and hence they want lower estimates of the responsiveness of
>natural gas supplies to changes in price; in contrast, gas companies have a stake in
>higher (unregulated) prices, and hence want higher estimates of gas supply
>responsiveness; (b) basic beliefs about the "finiteness" of potential supplies and about
>the likelihood of the human imagination to respond to needs with new developments; (c)
>differences in the scientific imaginations of the engineers and geologists making the
>estimates; and (d) professional differences among engineers and among economists due to
>differences in technical approaches.
>
> Every month, it seems, we read of new ways to get more energy. Item: Three-
>dimensional seismic exploration methods have produced large new oil discoveries at very
>low cost. In Nigeria and Oman, Shell "has found new oil reserves at costs of less than
>10 cents a barrel." Item: Lumps of methane hydrate on the ocean floor could constitute
>"a potential fuel reserve that may dwarf all the fossil fuel deposits on land combined."
> ***
>
> In my view, the data and theory continue to support a forecast made years ago by
>Herman Kahn and associates. "Energy costs as a whole are very likely to continue the
>historical downward trend indefinitely...Except for temporary fluctuations caused by bad
>luck or poor management, the world need not worry about energy shortages or costs in the
>future".
>
> WHAT ABOUT THE VERY LONG RUN?
>
> Chapter 3 alluded to the increase in efficiency in energy use over the decades and
>centuries. An analysis by William Baumol mentioned there shows that such increases in
>efficiency have huge effects. The key idea is that an improvement in productivity not
>only reduces resource use in the present but, even more important, also increases the
>future services of the entire stock of unused resources. This alone could mean that the
>future supply will never run out.
> This process may be seen in figure 11-9, where the amount of coal required to move a
>ton of freight by sea fell to about a tenth of its 1830 value by 1890. That is a greater
>proportion of increase in efficiency than was the increase in population over those
>years. The transition to oil represented an increase in economic efficiency (or it would
>not have taken place). And there is no reason why that process should not continue
>indefinitely, with ship surfaces getting smoother, and so on. Of course nuclear power
>can replace coal and oil entirely, which constitutes an increase in efficiency so great
>that it is beyond my powers to portray the entire process on a single graph based on
>physical units.
>
> Fig 11-9 from Lebergott, p. 419
> Much the same occurs with electricity in figure 11-9. A generator converts the heat
>from fuels or the power of falling water into electrical energy. One cannot extract more
>energy from a generator than one puts in, but in addition to increasing efficiency in
>generation, there is increasing efficiency in end-use products such as refrigeration,
>heaters, and appliances.
> The process is even more extraordinary with respect to the input of human energy (no
>matter how human energy is measured). A handful of humans can now move hundreds of
>thousands of tons of freight across an ocean in a single ship, many fewer people per ton
>than in centuries past. And measured in ton-miles per day, the increase in efficiency is
>even greater.
>
> THE NON-FINITENESS OF OIL
> You may wonder whether "non-renewable" energy resources such as oil, coal, and natural
>gas differ from the recyclable minerals in such a fashion that the non-finite arguments
>in earlier chapters do not apply. Eventually we'll burn all the coal and oil that
>powered these impressive advances, you may be thinking. But our energy supply also is
>non-finite, including oil as an important example. That was not a misprint. In chapter 3
>I showed that it is necessary to say how one would count the amount of a resource if one
>is to meaningfully say that the resource is finite. Therefore, let's consider the
>following sequence of difficulties with respect to counting the amount of oil. As with
>other resources, careful thinking leads to the conclusion that the potential amount of
>oil - and even more, the amount of the services that we now get from oil - is not finite.
> (1) The oil potential of a particular well may be measured, and hence it is limited
>(though it is interesting and relevant that as we develop new ways of extracting hard-to-
>get oil, the economic capacity of a well increases). But the number of wells that will
>eventually produce oil, and in what quantities, is not known or measurable at present and
>probably never will be, and hence is not meaningfully finite.
> (2) Even if we unrealistically assume that the number of potential wells in the earth
>might be surveyed completely and that we could arrive at a reasonable estimate of the oil
>that might be obtained with present technology (or even with technology that will be
>developed in the next 100 years), we still would have to reckon the future possibilities
>of shale oil and tar sands - a difficult task.
> (3) But let us assume that we could reckon the oil potential of shale and tar sands.
>We would then have to reckon the conversion of coal to oil. That, too, might be done, but
>the measurement is becoming increasingly loose, and hence less "finite" and "limited."
> (4) Then there is the oil that we might produce, not from fossils, but from new crops
>- palm oil, soybean oil, and so on. Clearly, there is no meaningful limit to this source
>except the sun's energy (land and water are not limits - see chapters 6 and 10). The
>notion of finiteness is making ever less sense as we proceed.
> (5) If we allow for the substitution of nuclear and solar power for oil - and this
>makes sense because what we really want are the services of oil and not oil itself - the
>notion of a limit is even less meaningful.
> (6) Of course the sun may eventually run down. But even if our sun were not as vast as
>it is, there may well be other suns elsewhere.
> The joke at the head of chapter 3 makes the point that whether there is an "ultimate"
>end to all this - that is, whether the energy supply really is "finite" after the sun and
>all the other planets have been exhausted - is a question so hypothetical that it should
>be compared with other metaphysical entertainments such as calculating the number of
>angels that can dance on the head of a pin. As long as we continue to draw energy from
>the sun, any conclusion about whether energy is "ultimately finite" or not has no bearing
>upon present policy decisions.
>
> About energy from the sun: The assertion that our resources are ultimately finite
>seems most relevant to energy but yet is actually more misleading with respect to energy
>than with respect to other resources. When people say that mineral resources are "finite"
>they are invariably referring to the earth as a bounded system - the "spaceship earth" to
>which we are apparently confined just as astronauts are confined to their spaceship. But
>the main source of our energy even now is the sun, no matter how you think of the matter.
>This goes far beyond the fact that the sun was the prior source of the energy locked into
>the oil and coal we use. The sun is also the source of the energy in the food we eat, and
>in the trees that we use for many purposes. In coming years, solar energy may be used to
>heat homes and water in many parts of the world. (As of 1965, much of Israel's hot water
>had been heated by solar devices for years, even when the price of oil was much lower
>than it is now, although I remember that the showers you got with this water were at best
>lukewarm unless you used a backup electrical system to boost the temperature.) If the
>prices of conventional energy supplies were to rise considerably higher than they now
>are, solar energy could be called on for much more of our needs, though this price rise
>seems unlikely given present technology. And even if the earth were sometime to run out
>of sources of energy for nuclear processes - a prospect so distant that it is a waste of
>time to talk about it - there are energy sources on other planets. Hence the notion that
>the supply of energy is finite because the earth's fossil fuels or even its nuclear fuels
>are limited is sheer nonsense. And this discussion has omitted consideration of any
>energy sources still to be discovered.
>
> CONCLUSIONS
>
> Energy differs from other resources because it is "used up," and cannot be recycled.
>Energy apparently trends toward exhaustion. It seems impossible to keep using energy and
>still never begin to run out - that is, never reach a point of increasing scarcity. But
>the long-run trends in energy prices, together with the explanatory theory of induced
>innovation, promise continually decreasing scarcity and cost - just the opposite of
>popular opinion. At worst, the cost ceiling provided by nuclear power guarantees that the
>cost of electrical power cannot rise far above present energy costs, political obstacles
>aside.
> The historical facts entirely contradict the commonsensical Malthusian theory that the
>more we use, the less there is left to use and hence the greater the scarcity. Through
>the centuries, the prices of energy -- coal, oil, and electricity -- have been
>decreasing rather than increasing, relative to the cost of labor and even relative to the
>price of consumer goods, just as with all other natural resources. And nuclear energy,
>which at present costs much the same as coal and oil, guarantees an inexhaustible supply
>of energy at declining cost as technology improves.
> In economic terms, this means that energy has been getting more available, rather than
>more scarce, as far back as we have data. This implies that the rate at which our stocks
>of resources increase, or the increasing efficiency of use over time, or a combination of
>the two forces, have overmatched the exhaustion of resources.
>
> Another way to look at the matter: energy has become less and less important as
>measured by its share of GNP. This is the same story as revealed by all other natural
>resources.
>
> The reason that the prices of energy and other natural resources decline even as we
>use more is the advance of technology. Nevertheless, just as with land and copper, there
>are other forces at play which make it possible for us to have increasing amounts of the
>services we need even as we boost the demands we make upon the supplies of those
>resources.
>
> One saving grace is improved techniques of use. Consider the steam engine, which when
>first invented operated at 1 percent efficiency. Engines nowadays operate perhaps thirty
>times more efficiently. That is, they use a thirtieth as much energy for the same
>result. The invention of the microwave oven immediately meant that only 10 percent as
>much energy was necessary to cook a meal as before. When someone finds a way to increase
>the efficiency of using a resource, the discovery not only increases the efficiency of
>the energy we use this year, but it also increases the effective stocks resources that
>are known or are as yet undiscovered. And this process could continue a long time,
>perhaps indefinitely.
> Also important are increases in energy supply. We learn how to dig deeper, pump
>faster. And we invent new sources of energy -- aside from coal, shale, oil, tar sands,
>and the like. We can also "grow" oil substitutes as long as there is sunlight to raise
>plants. (See chapter 6 on hydroponic farming using fresh water. And production of oil-
>seed crops that grow with salt water, which allows agriculture with irrigation of the
>desert, is now entering commercial development in Saudi Arabia. Also, nuclear fission
>power will be available at constant or declining costs practically forever.
> After our sun runs out of energy, there may be nuclear fusion, or some other suns to
>take care of our needs. We've got seven billion years to discover solutions to the
>theoretical problems that we have only been able to cook up in the past few centuries of
>progress in physics. It's reasonable to expect the supply of energy to continue becoming
>more available, forever.
> Another sort of summary may be found at the beginning of this chapter, starting with
>the fourth paragraph.
>
-YEAH BABY I CAN'T IMAGINE THAT YOU'D WANT ME TOO-