In the year 2007, coal, natural gas, nuclear and petroleum supplied 48.5%, 21.6%, 19.4% and 1.6% respectively of the energy generated by the U.S. electric grid.  The balance was generated by  hydroelectric (5.8%) and other renewables like  wind, biomass and solar (2.5%) .  The average cost paid for the 4.16 billion MWH of electric energy used in 2007 was $0.091/kwh, accounting for about 3% of the U.S. economy that year (U.S. DoE Energy Information Agency Data) .

Driven mostly by a need to  limit  atmospheric carbon dioxide to 450 ppm (believed to correspond to a maximum temperature increase of 2° C over pre-industrial levels) and to combat climate change, the sources of our power, the technology behind its  generation and the costs we pay for using it are going to change dramatically.  The B-GC has previously discussed climate models that forecast temperature and sea level changes  as nations either proceed to emit as usual or curtail their emissions.  In this report, we discuss models of likely changes in our electric generation sources to achieve reduced emission goals.  Two source models are described.  One was created by the Electric Power Research Institute (EPRI) and the other by the International Energy Agency (IEA).   Both organizations emphasize that it should be possible to make the changes necessary to bring emissions down dramatically over the next 40 years.  EPRI does specify the need to invest in R&D to realize the energy efficiency savings, carbon capture and storage technology (CCS), plug-in vehicle (PEV) technology and energy transmission savings.

EPRI’s model, updated for 2009, is called MERGE (model for estimating the regional and global effects of greenhouse gas reductions), and it predicts the mix of likely, economically optimized energy sources necessary to comply with the Waxman-Markey targets for emission reductions: 17% by 2020, 42% by 2030 and 83% by 2050, all relative to our 2005 emissions.  Though the bill has not passed the U.S. Congress, the same emission reduction targets have been referred to by President Obama as the U.S. commitment at the Copenhagen climate change conference.  MERGE covers two possible outcomes – one called the Full Portfolio and the other the Limited Portfolio.  The Full Portfolio assumes that technologies like CCS (with 90% of emissions stored underground) and plug-in electric vehicles are technically possible and embraced by Americans.  It also assumes that new nuclear plants are approved in the future, though no new plants have been brought online over the last 30 years.  The Limited Portfolio puts some constraints on the optimum approach by assuming that CCS and PEVs are not deployed (for unspecified technical or economic reasons), and no new nuclear plants are added.   Based on strong interest in plug-in electric vehicles by all the major auto manufacturers and recent comments by American Electric Power on CCS plans, the B-GC believes something in between the Full and Limited Portfolios is likely.

The International Energy Agency recently issued its World Energy Outlook 2009.  The model is based on limiting the maximum CO₂ levels in the atmosphere to 450 ppm, and  it proposes possible changes in the energy industry and the timeframe for those changes that would be consistent with this  greenhouse gas emissions limitation .  With that wide a scope, the IEA report also covers petroleum-based transportation, which is not covered here.  Similarly, the IEA report, unlike this B-GC analysis, does not limit itself to U.S. emissions, but covers the changes to the energy industries throughout the world.  The 450 ppm limit used by IEA is similar to the emission reduction timetable for the Waxman-Markey bill used by EPRI, as that same limit was used as a guide in modifying the U.S. emissions in a manner consistent with the world peaking CO₂ levels at 450 ppm.   The IEA’s report differs with the EPRI report in that it is not an economic optimization, though it does consider the investments necessary to achieve the stated goals.  Of the two reports, the EPRI report goes into much more detail on technology issues and needs, and it also  predicts the impact on electricity prices.  Further, the EPRI report covers projections from 2005 to 2050, while the IEA goes from 2007 to 2030.  Also, while EPRI covers energy generation, IEA refers to power.  To enable a comparison with the EPRI models in terms of energy, the B-GC did have to make some assumptions to do the conversion, which may have added some uncertainty to the data in our graphs, particularly for the other renewables category.

The model results for total energy generation are shown in Figure 1.  The EPRI reference case is shown,  representing a  business as usual scenario without an increased pace of energy efficiency investments.  EPRI based this projection on the U.S. D.O.E Energy Information Agency 2008 forecasts.  In this case, energy generation and consumption rises to 6.5 trillion kwh/year, or a 62% increase, as the U.S. population increases 44%.  The difference between the EPRI reference case and the EPRI and IEA models equates to  expected savings as Americans learn to use energy more wisely.  EPRI’s Full Portfolio requires higher energy generation than the Limited Portfolio, mainly because the former assumes significant penetration of plug-in vehicles (100 Million PEVs on the road by 2030) as well as other conversions from direct fossil fuel energy to low carbon emission electric energy .  Realization of the carbon reduction goals under  the Limited Portfolio with its limited power generation options and its consequent higher costs (to be discussed at the end of the article), provides an additional financial incentive to reduce energy consumption. The nearly no-growth forecast lines of the Limited Portfolio and the IEA models suggest the great potential for energy efficiency savings from using electricity more wisely, using more energy efficient appliances and reducing transmission and distribution losses.  According to these models, the U.S. electric energy per capita can drop as much as the population increases – or about 44%.

Figure: 1

The next series of figures forecast what happens to our fossil fuel generating capacity.  Some differences in the models have to be noted here.  The EPRI model applies CCS only to coal, while the IEA model has CCS applied to coal and natural gas.  For that reason, the CCS graph shows both natural gas and coal systems.

Figure 2a shows how quickly our dependence on non-CCS coal, the highest emitter per kwh,  must end.  The rate of reduction in the IEA and EPRI Full Portfolio models are similar.  The EPRI Limited Portfolio drops much more quickly to compensate for the lack of reduced emissions from CCS.   In Figure 2b, our projected natural gas without CCS dependence is  similar in the IEA and EPRI Full Portfolio models, while the Limited Portfolio quickly shifts to natural gas from coal to meet its carbon reduction obligations.  Figure 2c, shows the rise of CCS in the two models that allow it. The IEA is a little less optimistic on the early adoption rate of CCS, but  shows it making a major contribution (about 15%) to our power generation by 2030.   According to EPRI’s Full Portfolio, the energy generated from coal is nearly as high in 2050 as it is today, thanks to CCS.

Figure: 2a

Figure: 2b

Figure: 2c

The models show nuclear power generation growing in the U.S. if allowed.  In Figure 3, the EPRI Full Portfolio shows the nuclear energy contribution increasing quickly in the next twenty years and then allows for retirement of some plants as renewable and CCS technologies  grow.  The IES model depends less upon nuclear, and the Limited Portfolio just tracks the retirement of nuclear plants in the U.S. as new plants are by definition not allowed.  In this case, by 2050 nuclear drops from supplying about 20% of our needs to just over 6%.

Figure: 3

Figure 4 shows the Hydropower contributions in the models and, as you would expect, it stays steady, with little significant difference between the models.

Figure: 4

The contributions from wind projected by these models are shown in Figure 5a. The three models are in pretty close agreement that, by 2020, wind energy will drive about 6% to 7% of our needs. Beyond that, the Limited Portfolio constraints push more wind turbines to be built so that it is contributing about 22% of our energy needs in 2050, and about 17% in the Full Portfolio model.

For the other renewables, EPRI broke the numbers down between biomass and solar. EPRI did not consider geothermal. IEA just lumped all the other possible sources into ‘other renewables’, and for reasons of conversion mentioned above, despite our best efforts, may not be accurately represented in Figure 5b. The models imply that the EPRI Full Portfolio does not grow other renewables significantly beyond 2020. In the Limited Portfolio, other renewables grow very rapidly. By 2050, biomass and solar contribute 0.87 trillion kwh/year and 0.51 trillion kwh/year, respectively. In this scenario, biomass is generating more energy in 2050 than nuclear does now, and solar comes on strong, contributing nearly as much as natural gas is today. 

Figure: 5a

Figure: 5b

By 2050, according to EPRI’s MERGE model,  substituting  higher cost means of generating electricity relative to today’s concentration on low cost, greenhouse gas emitting fossil fuels, will drive wholesale prices up 80% under  the Full Portfolio and 210% under  the Limited Portfolio.  Table I shows the predicted numbers and tries to put these large increases in some perspective.  The actual impact of these price increases is  mitigated somewhat by an increasing population  (up  44% to 439 M in 2050) and energy efficiency improvements, resulting in per capita cost increase projections of only 37% for the full portfolio and 96% for the Limited Portfolio.

Still, the forecasted cost increase for each American, double what it is now by 2050 for the Limited Portfolio, can be disconcerting.  As the EPRI reports do not detail the assumptions on costs sufficiently for us to assess their model’s pricing forecasts, prudence suggests that these price projections should not be taken too seriously.  The reference business as usual case could be understating the future price of coal, natural gas or nuclear.  Similarly, the model’s assumptions on innovation reducing the costs of low greenhouse gas technologies such as CCS, wind, solar or biomass, could be highly understated.  Further, the lack of consideration of geothermal energy, which currently has a relatively low cost, could lead to an overstatement of the cost consequences in the models.   Models for predicting pricing of energy do not have a good track record.

The review of these models, even with an awareness of the uncertainties associated with predicting the future, leads to some clear and important conclusions.  The main one is that, in the timeframe required to achieve these greenhouse gas reductions, it is possible to transform our electricity generation from one altering our atmosphere and threatening our planet with drastic climate changes to one maintaining our atmospheric greenhouse gases at a constant, non-threatening level.  Energy efficiency technologies will be a key factor in achieving these goals.  If CCS and plug-in vehicles do not achieve technical or financial acceptance, renewables will grow much more rapidly, becoming a larger part of the energy mix and requiring our grid  to overcome technical issues associated with the acceptance of these more intermittent sources.  The cost analysis also suggests that electric utilities will experience more rapid revenue growth over the next 40 years and become a larger part of the U.S. economy even though the energy they deliver will likely see little growth or even contraction through mid century.

Note:  EPRI did not respond to repeated requests for an interview with the B-GC on the MERGE model details.