EFTA00618036.pdf

Eye on the Market I November 21.2011 J.P.Morgan Topic: The quixotic search for energy solutions Another Don Quixote Thanksgiving. Every year at Thanksgiving', 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 19552. 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 ow 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 convened into commercial-scale biofuel (they will now work with synthetic strains instead)°. Some clients tell me it is helpful to have something to read this weekend, wherilif 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 CO2 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. EFTA00618036  Eye on the Market I November 21.2011 J.P.Morgan 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'. Energy Installed Electricity % of Implied EIA Levehzed c-Levelixed cost Incorporates upfront and ongoing capital costs, cost of Wormation base gen in 2010 total capacity cost 2016 capital, fuel and other operating costs, capacity factor and related power Agency 2010 MW mm MWn gen. facbr per MUM transmission investments (In 2009 dollars) for new construction Coal 316,800 1,847 45.4% 67% $95 - $110 Abundant and cheap, but with a substanhal range of enuronmental problems Natural gas 407,028 988 24.3% 28% $60 - $70 Capacityfactors understate potenhal uhlizabon Nuclear 101,167 807 19.8% 91% $114 Efficient once built; tery 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 Biomasshvood 11,406 56 1.4% 56% $112 Expensive to aggregate and collect; high capital costs relative to energydensity Geothermal 2,405 18 0.4% 85% $102 Veryexpensive, except near areas with active geothermal reservoirs Solar PWCSP 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 merit6, it is hard to escape (iii). Energy Conversions 101 is meant to show why, using examples' 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 Question it2: Do electric cars require less energy than gasoline powered to electric cars? cars? It not, what might the other benefits of electric cars be? 200 wall hours per Iola average electic car 4.4 UM per electic car per year (see assurrptons n 20.000 km&von per car per year NOW101Shjur0 our The PRIMARY energy neededle make this eleetneiy... 245.000.000 number ol US passenger cars 60% Money loss of contralto process (avg be US coal and nat gas generabn) 980.000.000 MTh br US passenger camper year. at °tactic 10% Becht* tansrrissico lassos 980 TWn br US passenger cars per year. all elate 122 Wet of plenary energy requred per car per year 10% + Increase due b batery seldischarge 44.000 Megapules ol energy per electric car per year (3.600 MJ=l SINN) 1.078 TWn br US passenger cars per year 2.2 Megapubs per car per year per let driven 4325 TWn ol US Setter pi-ado:fon 15.9 Smiler br dean car when he eatery energy (coal or gas used b genera? 25% Incremental eledrtty need eteeti:itr) ?expressed m gascrine equ'realents (35 M.61 L) 37.4 Primary energy requirementof electic car. expressed n rites per galon Implication: This is incremental generation. not capacity. since some Implication: In other words, primary energy required to power electric cars is existing facilities could produce more. But its still a huge increase in not that different from high mpg gasoline cars, which exist already. Depending generation, and the cost will depend on where you plan to get the on how electricity is generated, there could be some emissions benefits (but electricity from. and when. Gasoline is used on site: elecincity is not if coal Is the primary source ol electdchy, as it is now). There would be generated off site and then moved across what is perhaps the worst much less depedence on foreign oil, a US objective for decades. But some electrical grid in the OECD. Note that we did not include transmission benefits could also be obtained through a high mileage fleet. perhaps less ol an losses here: if we did, generation requirements would be higher. This undertaking than switching to electric cars. If efficiency losses f rom electricity also ignores electric car battery Zile issues (heal. cold. etc) and the rising conversion in coal, nuclear or gas plants were reduced from 60% to 50%. that cost of rare earth metals needed for electric cars. 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 EFTA00618037  Eye on the Market I November 21, 2011 IP Morgan Topic: The quixotic search for energy solutions Question 43: What if the world ends up relying on coal for the next 100 Question 84: What would be the reduction in gasoline needs if the entire U.S. years. and seeks to prevent further increases in carbon emissions. How com harvest not already used for ethanol were repurposed for more ethanol? large an undertaking is it to bury 15% of all CO2 emissions? 160,000,000 US corn honest. tomes. 2010. not already used for ethanol 33.2 billion tonnes of CO2 emissions (2010) 159,667,200,000 US corn hanest. kg. not already used for ethanol 5.0 Sequestration target. billions of tonnes 0.40 Conversion ratio, liters of ethanol per kg Now letS shrink The CO2 by compressing ii before burying ii... 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 CO2 to bury. billions of cubic meters 42,699,571,200 Effective gasoline-equivalent savings (liters) c 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 4.6 Volume of global crude oil extracted. bn cubic meters (2010) Implication: Capturing a small portion of OO2emissions requires a 8% Reduction in gasoline needs by repurposing entire corn honest Implication: Benefits of corn ethanol appear to be close to their maximum production level. Them is of course the issue of ethanors "energy return on compression/transportation/storage industry whose throughput is greater investment" (EROI), for which estimates range from 0.8:1 to 1.6:1. Charles Hall than the one used for oil extraction: and without the benefit that oil provides at SUNY ESF (originator of the EROI concept in the 1970's) published recent as an energy input. Coal-fired plant capital costs could rise 40%45% (as EROIs boron (1020): Tar sands and Shale Oil (85); Nuclear (5.15) and Wind per IPCC). and their electricity consumption could rise by 30%-40% for CCS (1520. but that excludes the cost of backup peaking plants). In that context. particulate removal and flue gas desull urization. Unlikely in time to prevent the EROI for com ethanol, which excludes the various layers of subsidies a further rise in CO2 emissions: unexplored legal and NIMBY issues as well. involved, is well below the fully loaded economic benef its of other fuel sources. Question #5: What about cellulosic ethanol? And what about using spent Question 86: How much area would be needed for a quarter of US collect grounds? electricity generation to come I rom wind? 225.000.000 Tonnes of US com stover. annual 2.3% Wind as a % of electricity generation 224.532.000.000 kg of US corn stover (using conversion factors from 44) 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 44) 193,656 square km of required area 2.74% Percent of gasoline needs reduced from conversion of stover And or?the need for ewensive HVDC transmission lines... And another fun lea 30.049,000 US population living in prime wind and/or solar states 0.16% Percent of global diesel fuel production offset by somehow IAZ OK. NE. WY, CO. ND, SD, KS. IA. MT, NM + Northern 13() gathering all of the world's spent coffee grounds and then 309,350,000 US population converting them into biolesel Implication: 194 thousand square km is about the entire area of Nebraska. It Implication: Apart from Brazilian sugarcane. which grows 365 days a year would be a massive undertaking which requires, as stated earlier. hundreds of and needs no irrigation or fertilizer (it self -fertilizes), biof uels are challenged billions of dollars for new transmission lines. To be clear, land under wind due to the cost of aggregation. low energy densities and high energy turbines still have many practical economic uses. The larger issues are extraction costs For atom limitations see note 3 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 California energy demand vs. total wind - summer 2006 Megawatts Megawatts Megawatts - 1.200 700 Each day is a different color 38,000 - California energy 600 Day 29 demand (LHS) • 1.000 Day, 34,000 • • 800 500 400 - 600 ay26 30,000 - 300 - 400 Total wind (RHS) 200 26,000 . - 200 100 1 22,000 0 0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324 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 Hours Source: California Independent System Operator. Integration of Source: Electric Power Research Institute. As measured in Tehachapi. CA Renewable Resources. No vember 2007. 3 EFTA00618038  Eye on the Market I November 21, 2011 J.P.Morgan 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 biotite 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 Actual and projected fuel economy for new passenger Billions of barrelskeal GDP (constan t 2000 USD, trillions) vehicles by country, Milesper gallon 0.30% 55 Europe ,••••• 0 25% Non-OECD 50 a is s sf...Japan 45 China 0. 0.20% - 40 ......... „,. 35 0.15% World S. Korea ...wee"' 30 anada 25 United States 0.10% OECD 20 0.05% 2002 2005 2008 2011 2014 2017 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Source: The International Council on Clean Transportation, United Source:151Group, International Enemy Agency, World Bank. 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 Prins. 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. Here's one view on biodesel from Giampetro (Barcelona) and Mayumi (Tokushima), authors of "The BiofuelDelusion" 12009]: "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 EFTA00618039  Eye on the Market I November 21, 2011 J.P.Morgan 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 applications9 (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). Compressed Air Energy The accompanying chart shows the existing state of Storage, 440 MW commercial-scale electricity storage; it's all about Sodium-Sulfur Battery pumped hydro10, a process that uses cheaper electricity at 316 MW Pumped night to pump water uphill into a reservoir basin, and then Hydro • Lead-Acid Battery releases the water during the day to power a hydro-electric 127,000 MWe, -35 MW generator. The other technologies are an afterthought, at • Nickel-Cadmium least right now. Note that more energy is expended in Over 99% of Battery, 27 MW pumping the energy uphill than is generated by releasing it total storage Flywheels downhill; the economic value derives from much higher <25 MW capacity electricity prices during the day. Around 10%-20% of the Llthlum-Ion Battery potential pumped hydro energy is lost over time through -20 MW evaporation and conversion losses. Redox-flow Battery <3 MW Source: Fraunholer Inslihrte. 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 forfractions of an hour, rather than for hours or days. 19 Most pumped hydro facilities am 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 EFTA00618040  Eye on the Market I November 21. 2011 J.P.Morgan 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 The cost of electricity storage options optimism and enthusiasm. I met with the managers of Eos Range of levelized costs, $ per kWh Energy Storage, which is working on a zinc air battery solution $0.60 which aims to conquer all of the obstacles outlined in the second bullet point above. If the Eos projections bear out, they will offer $0.50 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 $0.40 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 $0.30 ■ comparisons are more useful. As shown, Eos aims to be the cheapest option that can be scaled, and flexibly and safely located $0.20 M where needed. Note as well that they expect to be cheaper than natural gas peaking plants. This is a relevant benchmark, since EN . most utilities rely on natural gas peaking plants to meet daily $0.10 2 . -,.. 1 i . T —3 Gas peaking plants Advanced Lead-Add peak load requirements and to compensate for intermittent Vanadium Redo renewable generation of wind and solar. If storage works, the -o 2 = 2. W: E need for lots of peaking facilities could disappear. • to 2 > cg -Q m • JD -o o = a. m a co Eos has a prototype of its Zinc-Air technology that has run m m around 2,000 cycles so far; we should all pray either for their ta w 0 o a 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 Source: EPRI. Electricity Enorgy Storage Technobgy Options, 2010, Eos. CAES: Compressed Air Energy Storage. 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. Ether 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 Transmission loading relief requests Megawatts Megawatts Number of incidents.2002 -2008 70,000 7.400 90 65,000 Demand tto - •Independent Coordinator of Transmission for En t erg y 60,000 55,000 t\" 5.920 70 • Midwest Independent Transmission System Operator 60 • • PJM Interconnection (Southeast/Midwest) 50,000 4.440 1 45,000 50 - • Tennesse° Valley Authority • Southwest Power Pool 40,000 2,960 40- 35,000 30 • 30,000 V j Jtl 1,480 20 • 25,000 Wind outpu V V 10 - 20,000 c 0 8/1/11 8/2/11 813111 8/4/11 8/5/11 8)6/11 8/7/11 &SIM 0 2002 2003 2004 2005 2006 2007 2008 Source: Electric Fteliability Coundl of Texas. Source: North American Electric ReliablityCcuporation. 6 EFTA00618041  Eye on the Market I November 21.2011 J.P.Morgan 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 nuclear power plants (molten salt reactors, specifically) do not belong on the surface of the earth, and belong underground instead, to deal with the clean-up and failure, if it happened. And that's from one of nuclear power's greatest supporters. From a broader perspective, the era of cheap oil appears to be over. As shown below in the first chart, almost the entire future increase in oil supplies projected by the EIA are based on unconventional supplies (tar sands, deep-sea drilling, enhanced oil recovery, oil shale, etc.), with the word "unconventional" being shorthand for "more expensive". As for natural gas, as shown in the second chart, EIA projections assume that rising U.S. shale gas production, with all its uncertainties in terms of associated costs, will offset declines in almost every other category. It is hard to precisely quantify the speed bump on growth that this creates for the world, and as things stand right now, deleveraging of household, corporate and sovereign balance sheets in the US and Europe is a much bigger risk for financial markets. As I write this, Italy has entered an acute state of distress; its credit default swap spreads are now at levels which prompted bank runs in Greece and Ireland. In the long run, as we outline return expectations for the future, uncertainties related to energy availability are yet another reason why price-to- earnings multiples may remain well below their historic avenges. A break in the chain of unfulfilled promises from breakthrough technologies will be needed to alter this view. Global liquids production US dry gas production Million barrels per day Trillion cubic feet per year 120 History I Projections 30 Hbtory 100 25 1% OPEC Conventional 40% Net imports 80 11% 45% 39% 14% Shale gas 60 Unconventional 12% 15 20% 8% 5% 9% Non-associated onshore 40 10 Non-associated offshore 8% Non-OPEC Conventional 48% 28% Tig