Document Text Content
November 21, 2011
Topic: The quixotic search for energy solutions
Another Don Quixote Thanksgiving. Every year at Thanksgiving 1 , we look in-depth at an issue that affects markets and
portfolios. Last year, we examined the unraveling situation in Europe. Unfortunately, most concerns we expressed last year
have been borne out, and are getting much worse (I spent the weekend reading legal documents on a Eurozone break-up, just in
case). Like Don Quixote, Europe went on its journey for all the wrong reasons, adopting a half-pregnant monetary union to
support a political objective that had arguably already been achieved by 1955 2 . This year, a look at something just as worrying
in the long run as the fiscal problems of the West: the search for energy solutions. This journey has been fraught with
similarly quixotic dead ends, fairy tales and blunders ignoring economic (and thermodynamic) realities. This is important to us,
since energy cost and availability is central to how we think about growth, profits, stability and our portfolio investments.
As part of this effort, I made a pilgrimage to Manitoba to spend a day with Vaclav Smil. Vaclav is one of the world’s foremost
experts on energy, and has written over 30 books and 300 papers on the subject (he’s #49 on Foreign Policy’s list of the 100
most influential thinkers). Vaclav’s book “Energy Myths and Realities” should be required reading for politicians or regulators
impacting energy policy. We start with an unflinching look at these realities before turning to solutions, and some potentially
encouraging developments, which have less to do with how electricity is generated, and more to do with how it might be stored.
“A dream is a wish your heart makes” (Cinderella)
Over the last 50 years, a lot of proposed solutions have not panned out as expected. While the process of discovery and
invention always includes large doses of failure, energy policy is different than say, cell phones or VCRs, since more public
money, time and effort are spent on them. Hopes are raised, and as a result, less flashy but more reliable solutions are
sometimes postponed or avoided altogether. Here are a few memorable predictions of our energy future:
• 1945. Oak Ridge National Laboratory nuclear physicists Weinberg and Soodak predict that nuclear breeders will be man’s ultimate
energy source; a decade later, the chairman of the US Atomic Energy Commission predict it would be “too cheap to meter”
• 1973. “Let this be our national goal: At the end of this decade, in the year 1980, the United States will not be dependent on any other
country for the energy we need to provide our jobs, to heat our homes, and to keep our transportation moving.” Richard Nixon
• 1978. “Through modeling of supply and demand for over 200 US utilities it was projected that, by the year 2000, almost 60% of US
cars could be electrified, and that only 17% of the recharging power would come from petroleum.”
• 1979. An influential Harvard Business School study projects that by 2000, the US could satisfy 20% of its energy needs through solar
• 1980. Physicist Bent Sorenson predicts that 49% of America’s energy could come from renewable sources by the year 2005
• 1994. Hypercar Center established, whose lightweight material and design would yield 200 mpg cars with a 95% decline in pollution
• 1994. InterTechnology Corporation predicts that solar energy would supply 36% of America’s industrial process heat by 2000
• 1995. Energy consultant and physicist Alfred Cavallo projects that wind could have a capacity factor of 60%, which when combined
with compressed air storage, would rise to 70 – 95% 3
• 1999. US Department of Energy hopes to sequester 1 billion tonnes of carbon per year by 2025
• 2000. Fuel cell companies announce 250-kilowatt production plants that can fit into a conference room and produce energy at 10 cents
per kilowatt hour, with the goal of 6 cents by 2003
• 2008. “Today I challenge our nation to commit to producing 100% of our electricity from renewable energy and truly clean carbon-free
sources within 10 years. This goal is achievable, affordable and transformative.” Al Gore
• 2009. Gene scientist Craig Venter announces plans to develop next-generation biofuels from algae in a partnership with Exxon Mobil
How have things turned out? There are no commercial nuclear breeders on anyone’s horizon; global nuclear capacity is only
20% of the Atomic Energy Agency’s 1970 forecast; the Hypercar is nowhere to be seen; solar and wind make up a miniscule
portion of US electricity generation; wind capacity factors range from 20%-30%; the US is reliant for 50% of its oil from
foreign sources; 70% of US electricity generation comes from coal and natural gas; fuel cells haven’t worked as expected;
hybrids are 2% of US car sales; “clean coal” is mostly a blueprint; and Venter announced that his team failed to find naturally
occurring algae that can be converted into commercial-scale biofuel (they will now work with synthetic strains instead) 4 .
1 Some clients tell me it is helpful to have something to read this weekend, when/if family gatherings become unwieldy, or aggravating.
2 A few years ago, Swedish and Dutch politicians mobilizing support for the EU Constitution referred to “Yes” votes as necessary tribute to
the dead from the Second World War, and more urgently, to avoid the pre-war divisions which led to it. Conflict between European empires
existed for hundreds of years (1871-1914 was the only period of peace until 1945), so the idea of a united Europe would have seemed
appealing in 1945. However, conditions for securing a lasting peace within Western Europe were arguably already in place by 1954.
3 A 2005 paper from Stanford raised expectations further by estimating theoretical wind power at 72 TW, 30x global electricity production.
4 Algae are inefficient photosynthetic reactors (they do not consume CO 2 when the sun isn’t shining), and allocate only a tiny fraction of
captured solar energy into lipid production. A 2007 study by Krassen Dimitrov at the University of Queensland predicted GreenFuel’s
demise in advance, claiming that the company estimated its photosynthetic efficiency at almost double the maximum theoretical rate, and
could only be profitable at $800 per barrel of oil. Genetic improvements of plant life have historically focused on disease resistance and
modifying the split between production of “fruit vs. stem”; it is used less often to increase growth rates of biomass itself.
1
November 21, 2011
Topic: The quixotic search for energy solutions
Today’s US energy reality: electricity generation
Before exploring why some of these ideas did not pan out, let’s look at where the US is right now in electricity generation. The
table below shows each energy source; its installed capacity; the electricity this capacity generated in 2010 and percent of total
generation; its capacity factor; and its long-term levelized cost for new construction, estimated by the Energy Information
Agency. Capacity factors are important since they measure the intermittency of each source (capacity factor = actual generation
relative to potential maximum generation). Baseload natural gas plants can run at higher factors than 28%; this number reflects
the fact that many gas plants are used as “peaking” facilities to provide short-term energy during periods of elevated demand.
As stated above, fossil fuels dominate, followed by nuclear. Hydroelectric is next (efficient and cheap, but most large-scale
sites are already in use); followed by non-hydroelectric renewable energy, which across all categories makes up less than 5%, in
part due to their low capacity factors. Non-hydroelectric renewable energy is a similarly small component of the country’s
overall energy use, a broader category which includes transportation fuels 5 .
Energy
Information
Agency
Installed
base
2010 MW
Electricity
gen in 2010
mm MWh
% of
total
gen.
Implied
capacity
factor
EIA Levelized <---Levelized cost incorporates upfront and ongoing capital costs, cost of
cost 2016 capital, fuel and other operating costs, capacity factor and related power
per MWh transmission investments (in 2009 dollars) for new construction
Coal 316,800 1,847 45.4% 67% $95 - $110 Abundant and cheap, but with a substantial range of environmental problems
Natural gas 407,028 988 24.3% 28% $60 - $70 Capacity factors understate potential utilization
Nuclear 101,167 807 19.8% 91% $114 Efficient once built; very expensive to build (costs rising sharply in recent decades)
Hydro 78,825 260 6.4% 38% $86 Most viable sites already in use after incentives in the 1960s-1980s
Wind 39,135 95 2.3% 28% $97 Low capacity factor, maturing technology; cost more than doubles offshore
Biomass/wood 11,406 56 1.4% 56% $112 Expensive to aggregate and collect; high capital costs relative to energy density
Geothermal 2,405 18 0.4% 85% $102 Very expensive, except near areas with active geothermal reservoirs
Solar PV/CSP 941 1 0.0% 15% $210 - $312 Expensive, low capacity factors; this segment is commercial (non-res) installations
Energy Conversions 101
What went wrong with renewables? Theories generally fall into 3 buckets: (i) why bother, since there are plenty of fossil fuels;
(ii) renewable energy would have a larger share if it benefitted from the massive R&D put into things like nuclear; and (iii)
renewables have thermodynamic, structural and practical limitations that inhibit their ability to represent much larger shares of
electricity or transportation fuel production. While (i) and (ii) have some merit 6 , it is hard to escape (iii). Energy Conversions
101 is meant to show why, using examples 7 that I expanded from Vaclav’s narrative (unit equalities on p.8).
Question #1: How much more electricity would the US need if it switched
to electric cars?
Question #2: Do electric cars require less energy than gasoline powered
cars? If not, what might the other benefits of electric cars be?
200 watt hours per km for average electric car 4.4 MWh per electric car per year (see assumptions in #1)
20,000 km driven per car per year Now let's figure out the PRIMARY energy needed to make this electricity…
245,000,000 number of US passeneger cars 60% Efficiency loss of generation process (avg for US coal and nat gas generation)
980,000,000 MWh for US passeneger cars per year, all electric 10% Electricity transmission losses
980 TWh for US passenger cars per year, all electric 12.2 MWh of primary energy required per car per year
10% + Increase due to battery self-discharge 44,000 Megajoules of energy per electric car per year (3,600 MJ=1 MWh)
1,078 TWh for US passenger cars per year 2.2 Megajoules per car per year per km driven
4,325 TWh of US electricity production 15.9 km/liter for electric car when the primary energy (coal or gas used to generate
25% Incremental electricity need
electricity) is expressed in gasoline equivalents (35 MJ=1 L)
37.4 Primary energy requirement of electric car, expressed in miles per gallon
Implication: This is incremental generation, not capacity, since some
existing facilities could produce more. But it's still a huge increase in
generation, and the cost will depend on where you plan to get the
electricity from, and when. Gasoline is used on site; electricity is
generated offsite and then moved across what is perhaps the worst
electrical grid in the OECD. Note that we did not include transmission
losses here; if we did, generation requirements would be higher. This
also ignores electric car battery life issues (heat, cold, etc) and the rising
cost of rare earth metals needed for electric cars.
Implication: In other words, primary energy required to power electric cars is
not that different from high mpg gasoline cars, which exist already. Depending
on how electricity is generated, there could be some emissions benefits (but
not if coal is the primary source of electricity, as it is now). There would be
much less depedence on foreign oil, a US objective for decades. But some
benefits could also be obtained through a high mileage fleet, perhaps less of an
undertaking than switching to electric cars. If efficiency losses from electricity
conversion in coal, nuclear or gas plants were reduced from 60% to 50%, that
would help the thermodynamics of electric cars substantially; but that's a big if.
5 Domestically produced and imported biofuels make up around 14% of US liquid fuels consumption.
6 The nuclear industry was the recipient of 96 percent of all funds appropriated by Congress for energy R&D between 1945 and 1998.
7 These examples are of course assumption-dependent; I tried to be conservative. I am sure you will let me know if I wasn’t.
2
November 21, 2011
Topic: The quixotic search for energy solutions
Question #3: What if the world ends up relying on coal for the next 100 Question #4: What would be the reduction in gasoline needs if the entire U.S.
years, and seeks to prevent further increases in carbon emissions. How
corn harvest not already used for ethanol were repurposed for more ethanol?
large an undertaking is it to bury 15% of all CO 2 emissions?
160,000,000 US corn harvest, tonnes, 2010, not already used for ethanol
33.2 billion tonnes of CO 2 emissions (2010) 159,667,200,000 US corn harvest, kg, not already used for ethanol
5.0 Sequestration target, billions of tonnes 0.40 Conversion ratio, liters of ethanol per kg
Now let's shrink the CO 2 by compressing it before burying it…
63,866,880,000 Liters of converted ethanol
0.80 Compressed gas density, tonnes per cubic meter 67% Energy density of ethanol relative to gasoline
6.2 Volume of compressed CO 2 to bury, billions of cubic meters 42,699,571,200 Effective gasoline-equivalent savings (liters)
3.9 Amount of global crude oil extraction, billions of tonnes (2010) 521,845,394,389 Liters of total US gasoline consumption in 2010
0.85 Density of crude oil, tonnes per cubic meter 8% Reduction in gasoline needs by repurposing entire corn harvest
4.6 Volume of global crude oil extracted, bn cubic meters (2010)
Implication: Benefits of corn ethanol appear to be close to their maximum
production level. There is of course the issue of ethanol's "energy return on
investment" (EROI), for which estimates range from 0.8:1 to 1.6:1. Charles Hall
at SUNY ESF (originator of the EROI concept in the 1970's) published recent
EROIs for oil (10-20); Tar sands and Shale Oil (3-5); Nuclear (5-15) and Wind
(15-20, but that excludes the cost of back-up peaking plants). In that context,
the EROI for corn ethanol, which excludes the various layers of subsidies
involved, is well below the fully loaded economic benefits of other fuel sources.
Implication: Capturing a small portion of CO 2 emissions requires a
compression/transportation/storage industry whose throughput is greater
than the one used for oil extraction; and without the benefit that oil provides
as an energy input. Coal-fired plant capital costs could rise 40%-75% (as
per IPCC), and their electricity consumption could rise by 30%-40% for CCS
particulate removal and flue gas desulfurization. Unlikely in time to prevent
a further rise in CO 2 emissions; unexplored legal and NIMBY issues as well.
Question #5: What about cellulosic ethanol? And what about using spent
coffee grounds?
225,000,000 Tonnes of US corn stover, annual 2.3% Wind as a % of electricity generation
224,532,000,000 kg of US corn stover (using conversion factors from #4) 10 Growth factor
40% Amount that can be removed without destroying soil 23% Target wind generation
89,812,800,000 Stover removed 95,000,000 Existing wind generation, MWh, 2010
30% Efficiency losses (evaporation, transportation, etc) 950,000,000 Target MWh
62,868,960,000 Remaining dry stover of uniform condition for conversion 28% Wind capacity factor
0.34 Theoretical conversion ratio, liters of ethanol per kg of stover 387,312 Required incremental MW of wind
21,375,446,400 Liters of ethanol produced from stover 2 watts per meter squared required for wind farms
14,291,012,736 Gasoline-equivalent ethanol from stover (see #4) 193,656 square km of required area
2.74% Percent of gasoline needs reduced from conversion of stover And on the need for expensive HVDC transmission lines…
And another fun fact…..
30,099,000 US population living in prime wind and/or solar states
0.16% Percent of global diesel fuel production offset by somehow [AZ, OK, NE, WY, CO, ND, SD, KS, IA, MT, NM + Northern TX]
gathering all of the world's spent coffee grounds and then
309,350,000 US population
converting them into biodiesel
Implication: Apart from Brazilian sugarcane, which grows 365 days a year
and needs no irrigation or fertilizer (it self-fertilizes), biofuels are challenged
due to the cost of aggregation, low energy densities and high energy
extraction costs. For algae limitations, see note 3.
Question #6: How much area would be needed for a quarter of US
electricity generation to come from wind?
Implication: 194 thousand square km is about the entire area of Nebraska. It
would be a massive undertaking which requires, as stated earlier, hundreds of
billions of dollars for new transmission lines. To be clear, land under wind
turbines still have many practical economic uses. The larger issues are
transmission and intermittency, as described below.
As for wind, let’s put aside concerns about space requirements and transmission lines. Let’s also put aside problems of wind’s
reliance on rare earths like neodymium for its turbine magnets (neodymium prices quadrupled this year, and that’s with wind
still making up less than 3% of global electricity generation). Let’s also put aside debris (from birds/insects), ice storms and
other natural elements that reduce wind farm efficiency. The reason to put them aside: if wind were more reliable, like
hydropower, it could justify a lot more expense and effort. Unfortunately, wind is not that reliable. The first chart is the
“Mona Lisa” of wind unreliability, measured at one of California’s largest wind farms. The second is from the California
Independent System Operator, showing how wind power tends to be low when power demand is high (and vice-versa).
Day-to-day variability in wind generation in April 2005
Megawatts
700 Each day is a different color
600 Day 29
Day 9
500
California energy demand vs. total wind - summer 2006
Megawatts
Megawatts
1,200
38,000
California energy
demand (LHS)
1,000
34,000
800
400
Day 5 Day 26
300
Average
200
100
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hours
Source: Electric Power Research Institute. As measured in Tehachapi, CA
30,000
26,000
Total wind (RHS)
22,000
1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Hours
Source: California Independent System Operator , Integration of
Renewable Resources, November 2007.
600
400
200
0
3
November 21, 2011
Topic: The quixotic search for energy solutions
Wind should play an important role, but unless there is a high-voltage, high-capacity, high-density grid to accompany it (as in
Northern Europe), or electricity storage, the variability of wind means that co-located natural gas peaking plants are
needed as well. The cost of such natural gas plants are rarely factored into the all-in costs of wind, but perhaps they should be.
These exercises are important, since unfounded expectations might lead to suboptimal policy choices. One example: the
Keystone Pipeline extension, which the President has opted not to consider until after 2012. The US imports more oil from
Canada than from any other country. With the extension, the Keystone system would account for 13% of US petroleum
imports. The pipeline has been opposed on environmental grounds, but the extension itself would only add 1% to the entire
network of crude oil and refined product pipelines already criss-crossing the US. Moving petroleum products by rail or truck
instead is more expensive and riskier. If the US does not provide a market for the Alberta tar sands oil, it could end up on
tankers to China; and the US will end up importing more of its energy needs from the Persian Gulf and Venezuela. Could
misperceptions about wind, solar and biofuel 8 feasibility explain why some people are opposed to this extension? Unclear.
The art of the possible
Now let’s take a (desperately needed) look at some good news. Over the last 3 decades, the oil intensity of the developed world
has been falling, followed by non-OECD countries (see first chart). This is not meant to suggest that declining availability of
cheap crude oil isn’t a problem, since it is. There are lots of studies showing rapid declines in the production rate of existing
crude oil fields, and that the discovery of new fields is (a) not keeping up, and (b) are located where marginal costs of extraction
are considerably higher. No need to repeat them here. But oil’s importance to economic growth has been declining over time,
and there is no reason to believe that these improvements have completely run their course.
Oil intensity declining worldwide
Billions of barrels/real GDP (constant 2000 USD, trillions)
0.30%
0.25%
0.20%
0.15%
0.10%
Non-OECD
World
OECD
0.05%
1980 1983 1986 1989 1992 1995 1998 2001 2004 2007
Source: ISI Group, International Energy Agency, World Bank.
Actual and projected fuel economy for new passenger
vehicles by country, Miles per gallon
20
2002 2005 2008 2011 2014 2017
Source: The International Council on Clean Transportation, United
Nations Department of Economic and Social Affairs.
There is also room for reduced fuel consumption, although here’s another case where energy fairy tales might have postponed
smart policy choices. While waiting for a holy grail, the US left fuel efficiency standards unchanged from 1983 (light trucks)
and 1987 (cars) until 2010. Chrysler head Lee Iacocca said this in 1986 when Ford/GM lobbied the Reagan Administration to
lower (“CAFE”) fuel efficiency standards: "We are about to put up a tombstone that says, 'Here lies America's energy
policy'. CAFE protects American jobs. If CAFE is weakened now, come the next energy crunch, American car makers
will not be able to meet demand for fuel-efficient cars." Well, the rest of the world kept on truckin’ as he suggested, and
have more efficient fleets (see chart). If the US fleet were 30% more efficient, US gasoline consumption could fall by 40 billion
gallons per year (~1 billion barrels). For context, the US imports 0.36 billion barrels of crude per year from Venezuela, and
0.62 billion from the Persian Gulf. The US just increased fuel efficiency standards, but it will take time to make an impact.
Other possible good news includes ongoing research by Daimler Engine Research Labs on improving gasoline engines,
something the world should not give up on just yet. Prototypes with fewer cylinders and smaller displacement may yield a car
with both lower fuel consumption and lower emissions, eventually at fuel efficiencies greater than hybrids like the Prius. The
US Recovery Act included $100 million for Advanced Combustion Engine Research and Development; it could be money well
spent. One example the DoE is working on: semiconductors, powered by the heat exiting the car in its exhaust pipe, used to
create electricity and power the car’s accessories, which are usually powered by belts driven by the car’s engine.
55
50
45
40
35
30
25
Canada
S. Korea
Europe
Japan
China
United States
8 Here’s one view on biodesel from Giampetro (Barcelona) and Mayumi (Tokushima), authors of “The Biofuel Delusion” [2009]: “The
promise of biofuels as a replacement to fossil fuels is in fact a mirage that, if followed, risks leaving us short of power, short of food,
destroying biodiversity and doing as much damage to the climate as ever.”
4
November 21, 2011
Topic: The quixotic search for energy solutions
The other good news relates to the discovery of new natural gas reserves. US shale gas production is up 14-fold over the
last decade, and the EIA projects that by 2035, the US will no longer be a gas importer. Yes, the Energy Department recently
slashed estimates of gas in the Marcellus Basin from 410 trillion cubic feet to 84 trillion; this followed the latest survey by the
US Geological Survey, which last estimated the basin at 2 trillion cubic feet in 2002. However, the historical imprecision of
peak oil/gas estimates make it a difficult science. To be clear, shale gas production will be critical; EIA projections to 2035
assume that rising shale gas production will offset declines in almost every other gas category (see p. 7). Deep sea gas reserves
are a potential positive, but marginal costs may be an issue. As for shale gas exploration and radium (naturally occurring and
surfaced in sometimes dangerous concentrations), and fracking chemicals themselves, the cost of natural gas electricity appears
low enough to absorb costs related to wastewater collection and treatment. Eventually, replacements will be needed for fossil
fuels. What “art of the possible” solutions do is give the world more time to find them. In the meantime, many scientists
would prefer to put as much emphasis on efficiency as on new technologies. Examples include 95% efficient natural gas
furnaces, LED/fluorescent lighting and more insulation. The largest direct energy saver in a 2010 report by the Pacific
Northwest National Laboratory for the Department of Energy: deployment of diagnostic devices in residential and
commercial buildings to manage HVAC systems and lighting.
A potential game-changer: electricity storage that works, in commercial scale
What would potentially change the energy equation is storage. The world has been generating commercially available
electricity for over a hundred years, but as things stand now, the world has almost no electricity storage. The benefits of
electricity storage, if it could be implemented, are self-evident:
• increased cost-effectiveness of intermittent solar and wind power, and lower electricity costs, since electricity produced by
wind at night could be stored and sold during the day; and electricity produced during sunny days could be stored and sold
during cloudy spells. There are obvious tie-ins to the feasibility and cost of electric cars
• lower required peak production capacities of large urban power systems, by drawing on stored electricity reserves
• deferral or avoidance of costly upgrades to the transmission grid. As per the North American Electricity Reliability
Corporation, only 27% of grid upgrades relate to integrating renewable energy. Almost half are designed to improve overall
reliability, due to fluctuating loads (since the grid has to accommodate peak loads, and not just average ones)
• reduced consumption of fossil fuels which power most stand-by generators
Unfortunately, battery storage has moved along at a snail’s pace. Moore’s Law on doubling semiconductor capacity is
something of a distraction; technology improvements over 15-18 months are hard to find anywhere EXCEPT semiconductors.
Solar photovoltaic cell efficiency has doubled over 15-18 years; and battery storage has progressed even more slowly as it
relates to commercial-scale applications 9 (rather than lithium ion applications for cell phone and laptops). As a reminder,
electricity is simply defined as the movement of electrons, which can only be “stored” as potential energy, for example via large
height or chemical gradients (e.g., batteries).
The accompanying chart shows the existing state of
commercial-scale electricity storage; it’s all about
pumped hydro 10 , a process that uses cheaper electricity at
night to pump water uphill into a reservoir basin, and then
releases the water during the day to power a hydro-electric
generator. The other technologies are an afterthought, at
least right now. Note that more energy is expended in
pumping the energy uphill than is generated by releasing it
downhill; the economic value derives from much higher
electricity prices during the day. Around 10%-20% of the
potential pumped hydro energy is lost over time through
evaporation and conversion losses.
Pumped
Hydro
127,000 MW el
Over 99% of
total storage
capacity
Compressed Air Energy
Storage, 440 MW
Sodium-Sulfur Battery
316 MW
Lead-Acid Battery
~35 MW
Nickel-Cadmium
Battery, 27 MW
Flywheels
<25 MW
Lithium-ion Battery
~20 MW
Redox-flow Battery
<3 MW
Source: Fraunhofer Institute, EPRI, Electricity Storage Technology Options, 2010.
9 Companies like A123 produce commercial scale batteries, but they are primarily for grid-smoothing. A123’s lithium ion batteries are meant
to store energy for fractions of an hour, rather than for hours or days.
10 Most pumped hydro facilities are designed to run for 10 hours uninterrupted (before being empty). Assuming 127 GW of installed
capacity, that means that 1,270 GWh of electricity would be produced before their reservoirs ran dry. That amount of stored electricity is
0.0064% of annual global generation. That is a very small supply; inventory storage for crude oil is 10%-12% of annual production.
5
November 21, 2011
Topic: The quixotic search for energy solutions
There’s no room to go through the complexities of the storage technologies shown below. Here are a couple of generalizations:
• Less expensive options like pumped hydro and compressed air storage require favorable sites with the right geology, which
are rare in nature and expensive to build from scratch (and often not located near electricity demand centers), and in the case
of compressed air, require co-located gas turbines for compression
• Many battery-based technologies suffer from high upfront capital or operating costs; low energy storage volumes; delayed
response times; safety issues (such as zinc bromine); or short lives (limited number of recharge cycles)
I had a meeting a few weeks ago which was notable for its
optimism and enthusiasm. I met with the managers of Eos
Energy Storage, which is working on a zinc air battery solution
which aims to conquer all of the obstacles outlined in the second
bullet point above. If the Eos projections bear out, they will offer
battery storage at a capital cost of ~$160 per kWh, in the form of
a 1 MW battery that is the size of a 40 foot shipping container
(for 6 MWh of storage). As with the table on page 2, the concept
of “levelized cost” synthesizes upfront costs, financing costs,
useful life, fuel costs and ongoing maintenance expenses. Rather
than looking projections of capital costs per kWh, levelized cost
comparisons are more useful. As shown, Eos aims to be the
cheapest option that can be scaled, and flexibly and safely located
where needed. Note as well that they expect to be cheaper than
natural gas peaking plants. This is a relevant benchmark, since
most utilities rely on natural gas peaking plants to meet daily
peak load requirements and to compensate for intermittent
renewable generation of wind and solar. If storage works, the
need for lots of peaking facilities could disappear.
Eos has a prototype of its Zinc-Air technology that has run
around 2,000 cycles so far; we should all pray either for their
success, or for the success of similar efforts undertaken by their
competitors. Based on the outcome of energy dreams shown on
p.1, we should always be skeptical of breakthrough claims, given
the complexity of the challenge. Let’s hope for the best.
Here’s another look at the financial rewards to anyone who can figure this out. Note how demands on the Texas electricity
grid (ERCOT) are almost 100% inversely correlated with when the wind blows. Either ERCOT gets connected to the national
grid, storage solutions are invented, or a lot of wind energy continues to be underutilized. On the right, what happens when
70% of the grid’s transmission lines, transformers and circuit breakers are 25-30 years old: rising congestion problems,
signified by rising loading relief requests. Grid storage has the potential to alleviate some of this congestion.
Texas electricity demand vs. actual wind output
Megawatts
70,000
65,000
60,000
55,000
50,000
45,000
40,000
35,000
30,000
25,000
Demand
Wind output
Megawatts
20,000
0
8/1/11 8/2/11 8/3/11 8/4/11 8/5/11 8/6/11 8/7/11 8/8/11 8/9/11
Source: Electric Reliability Council of Texas.
7,400
5,920
4,440
2,960
1,480
90
80
70
60
50
40
30
20
10
The cost of electricity storage options
Range of levelized costs, $ per kWh
$0.60
$0.50
$0.40
$0.30
$0.20
$0.10
Pumped Hydro
CAES (Below ground)
CAES (Above ground)
Sodium-Sulfur
Transmission loading relief requests
Number of incidents, 2002 - 2008
Independent Coordinator of Transmission for Entergy
Midwest Independent Transmission System Operator
PJM Interconnection (Southeast/Midwest)
Tennessee Valley Authority
Southwest Power Pool
0
2002 2003 2004 2005 2006 2007 2008
Source: North American Electric Reliability Corporation.
Advanced Lead-Acid
Source: EPRI, Electricity Energy Storage Technology Options, 2010, Eos.
CAES: Compressed Air Energy Storage.
Zinc-Bromine
Vanadium Redox
Iron-Chromium Redox
Zinc-Air Redox
Proposed Zinc-Air Solution
Gas peaking plants
6
November 21, 2011
Topic: The quixotic search for energy solutions
A setback for nuclear, and some investment consequences
The saddest energy moment of the year was the failure of the Fukushima Dai-ichi nuclear power plant in March. Weaknesses of
the original design and actions taken in the immediate aftermath of a massive tsunami combined to produce a disaster: the latest
studies show emissions of radioactive cesium that are equal to half of the release from Chernobyl. The concept of nuclear power
is one of man’s greatest achievements, but generating it safely and in a cost-effective way (including decommissioning) makes
it a difficult undertaking. In some ways, nuclear’s goose was cooked by 1992, when the cost of building a 1 GW plant rose by
a factor of 5 (in real terms) from 1972. Before he died, father of the hydrogen bomb Edward Teller’s last paper argued that