by TigPil » Wed 03 Jan 2007, 17:56:57
The problem with predicting the growth of renewable energy is a lack of understanding about the infrastructure to manufacture the energy producing equipment. Since renewable energy has been relatively expensive and fossil fuels relatively cheap until a few years ago, there has been minimal incentive to invest in infrastructure for manufacturing the means of producing that energy. And when there was ramping demand for wind turbines and solar cells in the last few years, there have been shortfalls in manufacturing capacity or raw materials (e.g. refined silicon shortage).
There have been two scenarios presented above that work off these poor assumptions about infrastructure:
1) The first scenario is that renewable energy grows at historical trends based on the IEA data. The absolute growth in installed Gigawatts is wrong since it is simply the absolute amount of renewable generation capacity that could be built with the infrastructure available in recent years. Under any scenario of increasing need (caused by the realization of peak oil or simply by increased fossil fuel prices despite continued obliviousness) the infrastructure will increase from current levels. This makes linear estimations inaccurate since increases in the underlying infrastructure will represent significant non-linearities in the growth rate of renewables.
2) The second scenario takes the exponential growth rate and projects it into the future to get 60% renewable energy at a specific date. This is wrong because it bumps against the upper limit of investable infrastructure. The amount of infrastructure for generating renewable energy generation capcaity will be capped by the annual replacement requirements for existing infrastructure, as well as the potential amortization of the capital stock. This means that if we need to replace 3% of energy generation capacity per year, assuming a 33 year replacement cycle for our energy generation infrastructure, the capacity to produce such generation equipment will peak at that 3% replacement rate. Taking a pure exponential extrapolation means that growth exceeds that 3% cap for some of the later years of the model, making little economic sense. While it may be possible to reach some levels of overinvestment due to a business cycle peak or government subsidies, it is unlikely the overshoot would be more than a percentage of the cap itself, i.e. a 30% overshoot yields a production cap of 4%.
So in terms of renewable growth rates, the correct questions to ask are:
1) Assuming the occurence of peak oil (or a sustained price shock in fossil fuel prices even with the continued denial of peak oil status) how long would it take to build the infrastructure for building renewable generation capacity quickly? Could we get to the 3% cap in 5 years? If not then how long would it take? How much capacity would we have after a 1 year, 3 year, 5 year crash program? The actual renewable production growth is then a function dependent upon this step function in the growth of infrastructure for generating said capacity.
2) How long would it take to reach equilibrium between the decline in non-renewable consumption (initially oil and perhaps natural gas but eventually other non-renewables) and the increase in renewable capacity? Lets say that the answer to 1 above is that peak infrastructure capacity is a 3% replacement rate and it takes 10 years to reach this. If at that point the loss in oil production is also 3% per year then renewable generation would be able to replace the decline in oil output on an annual basis without further demand destruction.
3) What happends in the interim between the peak oil event and the end of the crash program for building out renewable infrstructure? This is largely dependent on how long the crash program takes and how much elasticity there is in energy consumption. This gets into the complex subject of demand destruction, which can be subdivided into several types.
The first type of demand destruction is constructive (lets call it conservation) in that it has no negative economic impact. For example, I choose to drive my SUV 10% less thus using 10% less gasoline which happens to be 10% more expensive. The net result has no negative impact on my economic activities nor any on those of the oil producers and distributors, domestic or foreign. At most it is an inconvenience for me but has no impact on monetized GDP. The second type of demand destruction is destructive to the economy (lets call it contraction or substitution). There can be several examples of this with different economic repercussions. One case is that I drive 25% less thus using 25% less gasoline which is still 10% more expensive. In this case I am depriving the oil distribution and production companies of their revenue and using the money elsewhere. The more likely scenario however is that I drive 25% less when the price of gas is 50% higher. In this case I am actually spending more on gas and am then forced to curtail my discretionary expenditures elsewhere.
Getting back to the renewables topic, the nature of the transition period that was asked about in 3 above is dependent upon the absolute sizes of the conservation, substitution and contraction windows. Again to provide a concrete example with some hypothetical numbers, lets say the US economy has a conservation potential of 10%, then at a 3% loss in non-renewable supply yields a conservation window of just over 3 years. If it takes 10 years to get the infrastructure for renewable generation equipment in place then we would still have about 7 years of substitution/contraction to deal with, at whatever rate GDP contracts for a 3% contraction in oil supply. These numbers are all hypothetical of course and I don't pretend to know what some of the actual figures may be with any certainly. At most I would venture to guess that the time to build infrastructure will be in excess of the conservation window so some contraction and substitution will be inevitable. Nor is it wholly realistic to treat the conservation window and the contraction window as distinct for a homogenous US economy. Both types of demand destruction will coexist and have varied effects on different regions and strata of the US economy.
So far I've said nothing about what these renewable sources are. Primarily I am talking about solar, either in the form of photovoltaics or concentrated solar power (CSP). It was mentioned earlier that photovoltaics are not yet a good investment as the cost per KwH is around $.20 or slightly more. In certain non-US markets (Germany and Japan) the cost of retail electricity is already higher so PV is a practical alternative. Peak oil would substantially increase electricity prices in the US (currently around $.08 per KwH), which would make a PV a better investment. Wind is obviously cheaper with prices as low as $.03 per KwH but the intermittency factor makes it difficult to scale wind past 20% of total electricity production. Domestic solar installations often include a battery system so are better able to deal with intermittency. CSP is another solar alternative which has reached production costs of around $.11 per KwH and when used to heat a molten salt compound is also better suited to deal with intermittency. Geothermal and tidal are limited to specific locations. Biofuels are limited by arable land and the need to produce food. While we currently produce a surplus in this respect, crop yields will drop with the decline in fossil fuel based agriculture and put significant pressure on all land currently under cultivation. Regardless of how we look at it there won't be enough arable land to produce substantial quantities of biofuel. Finally, coal and nuclear have exogenous costs that are not part of their current published generating costs and neither is renewable in any case and are at best stopgap measures.
The composite renewable picture would then be 15% hydro, 20% wind and 65% solar (split between 25% CSP and 40% PV for the sake of cost). This would mean an average generation cost of $.12(.05*.15 + .04 *.2 + .11 * .25 + .2 * .4). Retail costs would be higher of course since some of this would be subject to transmission costs.
Everything I've said has been about electricity generation. What about the transportation sector? I am amazed how the comments of the Hirsch report are so easily taken as gospel about the future of electric cars. Hirsch excludes electric cars as a mitigation option solely on the basis of consumer preferences:
$this->bbcode_second_pass_quote('', 'T')he effort was a failure because existing
batteries did not provide the vehicle range and performance that customers
demanded. In the future, electricity storage may improve enough to win
consumer acceptance of electric automobiles. In addition, extremely high
gasoline prices may cause some consumers to find electric automobiles more
acceptable, especially for around-town use. Such a shift in public preferences is
unpredictable, so electric vehicles cannot now be projected as a significant offset
to future gasoline use.
Will consumers continue to object to vehicle range and performance issues when the price of gasoline is over $5 per gallon, over $10? This seems like one of Hirsch's most erroneous conclusions about potential mitigation factors. Demand destruction of every type will result in increasing changes to consumer preferences, simply by curtailing economic and non-economic options. And here we have a technology that is know to work but is ignored simply on the basis of consumer preferences. If we can produce enough renewable electricity and the only economically reasonable option is for people to drive electric cars then people will switch to electric cars, regardless of the inconveniences involved.
