Clean Disruption of Energy and Transportation: How Silicon Valley Will Make Oil, Nuclear, Natural Gas, Coal, Electric Utilities and Conventional Cars Obsolete by 2030 by Tony Seba

Clean Disruption of Energy and Transportation: How Silicon Valley Will Make Oil, Nuclear, Natural Gas, Coal, Electric Utilities and Conventional Cars Obsolete by 2030 by Tony Seba.  The following notes and references are from the kindle edition.

Transitions in technology markets can be swift. It took twelve years to reach 50 million laptops, seven years to reach 50 million smartphones, but only two years to reach 50 million tablets.2351

The percentage of YouTube traffic that comes from mobile Internet went from 6 percent in 2011 to 25 percent in 2012 to 40 percent in 2013.3 Similarly solar, the electric vehicle, and the autonomous vehicle started out as different sets of products and markets, but their symbiosis will complement and accelerate one another’s technological development and adoption in the marketplace.354

Lidar data maps resulted in a “higher prediction of solar PV yield and a 10.8-percent reduction in costs.”5 Lidar can also be used to measure the speed, angle, and intensity of the wind, data that managers can use to improve the planning and operation of wind power plants.379

Because the self-driving car is basically a mobile computer, it will also benefit from improvements in existing Silicon Valley computing and communications technologies: data storage, computers, operating systems and applications software, communications, and graphics accelerators. “Electric vehicles are the natural platform for autonomous cars,” said Takeshi Mitamura, director of the Nissan Research Center – Silicon Valley, speaking from his office in Sunnyvale. 384

People don’t want to just consume content; they want to create and share it. Companies that enable people to participate in the generation and dissemination of content have been amply rewarded. Witness the rise of Facebook, Twitter, and Linkedin.398

Solar is causing energy production to be pushed to the edges (customer site) from the center (large, centralized, hub-and-spoke power plants). The nodes are getting smaller, more modular, more connected, and more intelligent.417

Distributed solar generation, the electric vehicle and the autonomous vehicle are information products. As such, they are governed by information economics and increasing returns.425

Despite all the talk of abundance and a “golden age of energy,” fracked wells may deplete by 60 to 70 percent the first year alone.8 The industry has started calling this depletion phenomenon the “Red Queen Syndrome” (after the Red Queen in Through The Looking Glass, who tells Alice “it takes all the running you can do just to stay in place”). Because of Red Queen Syndrome, you need to frack millions of new wells just to keep up with existing production. This is not just a ‘fracking” phenomenon. Production from traditional wells declines by half in about two years, after which the wells drip on for a few more years. Extraction economics is about decreasing returns: The more you pump, the less each well produces. The more you pump, the less the neighboring well gets. The more you pump, the more each unit of energy will cost in the future. Solar, electric vehicles and the clean disruption are about increasing returns. Solar photovoltaic (PV) panels have a learning curve of 22 percent. PV production costs have dropped by 22 percent with every doubling of the infrastructure. The more demand there is in the market, the less your neighbor pays for her panels, and the more your neighbor benefits. Every time a solar power plant is built in Germany, Californians benefit from lower costs when the next solar power plant is built. Every solar panel sold in Australia cuts the cost of the next solar panel in South Africa. Lower costs benefit all new solar customers. Every large solar power plant in the desert benefits not only the people who buy its power, but everyone who buys solar power in the future. The higher the demand for solar PV, the lower the cost of solar for everyone, everywhere. Your neighbor benefits, the warehouse owner in Australia benefits, and future buyers of solar benefit from lower costs. All this enables more growth in the solar marketplace, which, because of the solar learning curve, further pushes down costs. This mutually beneficial arrangement is the opposite of extraction industries like oil and gas. When China’s demand for oil surged in the last decade, world prices for oil went up by a factor of ten. The higher the demand for oil in Beijing, the higher gasoline prices are in Palo Alto and Sydney. This is not just a theoretical framework. Solar PV has improved its cost basis by more than five thousand times relative to oil since 1970 (see Chapter 7). By 2020, as the market for solar expands, solar will improve its cost basis relative to oil by twelve thousand times (see Chapter 7).435

Network effects explain why the value of a network increases exponentially even when adoption increases linearly. Network effects are the reason AT&T so thoroughly dominated telephony in the U.S. for a century; they explain why Microsoft Windows has generated so much cash for three decades and why Apple’s iOS and Google’s Android platforms have become so valuable. Network effects are a winner-take-all proposition; after a technology platform such as Windows, Android, or TCP/IP wins in a market with network effects, it’s extremely difficult for others to compete in that market. Network effects apply to the market for autonomous vehicles (AVs). As the value of the autonomous vehicles marketplace increases exponentially (not linearly), the market grows. The more autonomous cars on the road, the more each one benefits from other autonomous cars on the road (see Chapter 5). For this reason, the returns for companies that win in the autonomous vehicle market will grow with each additional AV in the market.461

Energy and transportation as we know it today will be history by 2030.526

“A great many people think they’re thinking when they are really rearranging their prejudices.” – Aldous Huxley. “First they ignore you, then they laugh at you, then they fight you, then you win.” – Mahatma Gandhi. “If you evaluated rooftop solar a year ago, or even three months ago, you are way out of date.” – David Crane, CEO, NRG Energy. On February 1, 2013, El Paso Electric agreed to purchase power from First Solar’s 50 MW Macho Springs project for 5.79 ¢/kWh. “That’s less than half the 12.8 ¢/kWh from typical new coal plants,” according to models compiled by Bloomberg.12 Solar costs are going down so fast, solar is already becoming the lowest cost power provider to utilities and to retail commercial and residential consumers. This is true in Australia, the United States, Germany, Spain, and many other markets around the world. In the United States, new solar capacity has grown from 435 MW in 2009 to 4,751 MW in 2013 for a compound annual growth rate of 82 percent (see Figure 1.1).13 Solar represented 29 percent of all new generation capacity in 2013, up from 10 percent in 2012 and 4 percent in 2010.14528

On a clear afternoon on May 25, 2012, solar in Germany generated 22GW, which represented a third of the entire country’s power needs.15 The world record for high penetration of solar was broken on May 25, 2012. The following afternoon, solar generated fully 50 percent of Germany’s power, breaking the previous day’s world record. One out of every two electrons flowing through Germany’s grid was sunshine just a microsecond before. Astonishing as these penetration numbers were, they have now become commonplace. In Germany, a country with half the sunshine (solar insolation) as the United States, solar energy drove German wholesale electricity costs down by more than 40 percent in 2013 compared with 2008.16 This represented more than €5 billion ($6.7 billion) in cost savings for the German economy.17 Solar also pushed volatility down significantly (see Figure 1.2). Figure 1.2— Wholesale power costs in Germany, 2008, 2012, and 2013. Higher solar penetration has lowered both the price of energy and the volatility of energy prices. (Source: Meikle Capital).18 The combination of wind and solar is just as powerful. Shortly before noon on October 3, 2013, solar and wind together provided 59.1 percent of all the electric energy in Germany.19 Exactly one month later, on November 3, 2013, wind generated more than 100 percent of the power demand in Denmark.20 You read that right: One hundred percent of the electricity in Denmark was kinetic energy blowing in the wind less than a second before. Halfway around the world, Australia has set a few world records of its own: one million solar installations in about four years.21 This represented a market share of about 11 percent of all residential power consumers in the country.543

The Australian Energy Market Operator (AEMO) expects that, by 2020, 97 percent of all new energy generation additions to the grid will be wind or solar.31 This is the shape of grids to come. China, the world’s largest manufacturer of solar photovoltaic panels, has quickly turned into the world’s largest consumer of solar products. After tripling its solar PV demand in 2013, China set a goal of installing 14 GW of solar in 2014. In other words, China is expected to install in one year as much solar as the United States has installed in its history. China’s high rate of installation is not uncommon in exponentially growing markets. The United States installed more solar in 2013 than it did in all years prior to the end of 2011.596

Australians went from hardly any residential solar to one million solar homes in about four years. About 2.6 million Australians, 11 percent of the population, now have rooftop solar.766

Once it hit critical mass, it took a little more than a decade (1908-1921) for the auto industry to completely disrupt the horse-based transportation business. We should expect the same adoption curve with solar energy. What’s holding it back?975

Participatory finance and participatory energy went hand in hand in Denmark. By 2005, over 150,000 families owned turbines or were members of turbine cooperatives, and these cooperatives provided 75 percent of all wind power in Denmark. At that point the private sector started catching on to the importance of distributed wind power.87 By 2008, Denmark’s wind power capacity had increased such that wind now provides 19.1 percent of the country’s electricity (see Figure 2.2).1095

Dan Rosen, Mosaic’s co-founder and president, told me that solar is a trillion-dollar opportunity that will be served best by a peer-to-peer market. When people invest directly in a solar project, they benefit through the interest rate they earn; the users of the solar power plant, meanwhile, benefit from lower energy bills.1158

Most of the projects listed on the Mosaic website provide yields to investors in the 4.5- to 5.75-percent range; two outliers yield 7 percent. For individual investors who are getting less than 1 percent on their bank accounts or certificates of deposit (CDs), a return over 4 percent is considerable. So far, Mosaic projects have 100-percent on-time payments. There is no reason for the credit card-type returns that banks have so far demanded of solar project developers. Financial services companies and energy companies have huge appetites for outsize returns on capital. A peer-to-peer technology platform like Mosaic removes financial services companies and energy companies from the playing field. When small investors participate in a peer-to-peer platform like Mosaic, both sides win; individual investors benefit from a stable, long-term cash flow with a return that was previously available only to the large players in energy, while solar users benefit from the cheaper, more stable energy flow “For every 100 basis points (1 percent) we reduce in the total cost of capital, the cost of solar electricity drops by 1 ¢/kW to 2 ¢/kWh,” reported Dan Rosen, president of Mosaic. “As solar approaches and beats grid parity around the nation, a 1-percent savings in the cost of capital makes a huge difference.”1176

“It’s astounding how inefficient the solar finance and development process is,” said Rosen. “We’re trying to bring standardization to solar finance. Our goal is to make solar loans be like auto loans: you apply online and get instant approval for your solar project.” Mosaic wants to build an Internet cloud-based platform where developers of solar projects of any size can post their projects and investors of any size can invest in them. The projects and investors can be located anywhere. Said Rosen, “The small investor could put a few hundred dollars and the pension fund can put a few million. Community-oriented investors can find and invest in projects in their own neighborhoods while investment managers can build a portfolio of several projects in different markets.” About a hundred years ago, GMAC created auto loans; it has made hundreds of billions of dollars since. Along the way GMAC helped turn the early-stage automotive industry into a multi-trillion dollar powerhouse. Mosaic’s ambitions are to do with solar what GMAC did with autos by using participatory finance. “Americans have five trillion dollars in their IRAs,” said Rosen, “What we want to do is give every American the opportunity to say, ‘I have a solar farm in my IRA account.’”1188

According to Warren Buffett’s “Market Wisdom”: “I like businesses I can understand.” “I want to know what a business will look like ten years from now. If I can’t see them where they will be ten years from now, I don’t buy them.” “We don’t have huge returns, but we don’t lose our money either.”101 In other words, Warren Buffett likes to buy easy-to-understand, boring businesses that can generate good cash for decades and, in the worst case scenario, don’t lose the money he put into them.1222

Like a home mortgage, the cost of capital has become the most expensive line item in the construction of a solar power plant. The lower you can push down the interest rate, the lower the cost of solar electricity will be. As the installed cost of solar keeps decreasing, the cost of solar electricity will approximate the cost of capital. This is good news because the cost of capital for the U.S. government is less than 1 percent.1238

In April 2013, Hannon Armstrong Sustainable Infrastructure Capital, Inc. went public on the New York Stock Exchange (NYSE: HASI). It sold 13.33 million shares and raised $155.4 million.105 The IPO was otherwise unremarkable except for one important fact: HASI was a real estate investment trust (REIT) focusing on clean energy investments. A real estate investment trust (REIT) is a legal structure by which a company can invest in, own, and operate income-producing real estate.106 REITs were created by the U.S. Congress in 1960 to give investors the opportunity to invest in real estate in the same way they can invest in liquid market securities such as stocks and bonds.1071281

“REITs have a market valuation of $630 billion and give average dividends of 5 percent,” according to Dan Reicher, a professor at Stanford University who is executive director of Stanford’s Steyer-Taylor Center for Energy Policy and Finance. “Clean-energy REITs would have access to hundreds of billions of private investor dollars at far lower capital costs than they are currently getting.”1288

Earlier in this chapter, I explained that one requirement of PACE financing is that the asset be attached to the property. Rooftop solar PV plants are attached to the property (they are attached to apartment buildings, warehouses, and homes). Solar qualifies for PACE financing but the IRS has not qualified solar as a REIT-able investment. Can investors convince the IRS that solar, because it is attached to property, can qualify for REIT investments? A company can request, and the IRS can issue, a “private letter ruling” that details whether a type of infrastructure qualifies for REIT investments. Hannon Armstrong got such a “private letter ruling” from the IRS. In his testimony to Congress in October 2013, Dr. Reicher of Stanford mentioned how a company like FirstWind told him that they pay up to 14-percent cost of capital for raising tax equity.108 To put that into context, Hannon Armstrong’s dividend yield was 3.19 percent, according to Morningstar.109 Because Hannon distributed 100 percent of its earnings, its dividend yield is effectively its cost of capital.1295

Making clean energy assets REIT-able could cut the cost of solar (and wind) electricity as much as a third, according to a letter that 35 members of Congress sent to President Obama in December 2012. The letter read in part: Small tweaks to the tax code could attract billions of dollars in private sector investment to renewable energy deployment, reduce the cost of renewable electricity by up to one third, and dramatically broaden the base of eligible investors.111 It actually doesn’t take an act of Congress to make solar and wind REIT-able, according to Dr. Reicher of Stanford. All it takes is an administrative “revenue ruling” by the IRS. Hannon Armstrong may have opened the door for the IRS to move from a “private letter ruling” to a “revenue ruling.” The congressional letter to President Obama also urged him to open up another type of legal structure to clean energy: master limited partnerships.1310

As a business structure, the master limited partnership has the tax advantages of a limited partnership, but its stock can be traded like corporate stock. Unlike typical “C” corporations, which have to pay corporate taxes, MLPs pay no corporate tax. This gives them a big advantage. Their net earnings pass through to the shareholders as dividends. (See Figure 2-5.)1328

MLPs for clean energy projects. He believes clean energy MLPs could substantially increase the number of investors and decrease the cost of capital for clean energy projects.116 According to Reicher, MLPs have a market capitalization of $440 billion and pay on average a dividend of 6 percent. This compares to the 10- to 20-percent “credit card-like” cost of capital that clean energy companies have to pay for tax equity. To understand the difference that MLPs can make, start by considering why it’s hard to raise equity capital. The main incentive program for solar in the U.S. is the investment tax credit (ITC). The ITC is a 30-percent tax credit for solar systems on residential and commercial properties.117 Oil and gas companies also receive investment tax credits. For example, they receive the foreign tax credit deduction and the deduction for intangible drilling costs. Respectively, these two tax deductions will save the five largest oil companies $2 and $7.5 billion over the next decade, according to the Congressional Joint Committee on Taxation.118 Individual projects cannot directly use the ITC until they start producing profits, which can take several years. By contrast, large fossil fuel energy companies already have huge profits and they can offset these profits right away by using investment tax credits. Solar companies in the United States raise equity capital by attracting investors who can use investment tax credits. This is called “tax equity.” There are several issues with using tax equity to raise capital for a project. The first issue is that this is a highly illiquid market. In any given year only ten to twenty investors nation-wide have the appetite for the billions of dollars of tax equity needed by solar developers. The second issue is that the lack of competition allows these few equity investors to charge “credit card” rates for their capital. If the oil and gas industry only had access to the limited number of investors that the government allows solar and wind to access, the industry could never have developed the millions of wells and thousands of miles of pipelines it did over the last decade. Master limited partnerships would give solar and wind companies the opportunity to directly access millions of investors through public capital markets. Liquidity alone would cut the cost of capital for clean energy projects.1338

The Master Limited Partnership Parity Act was introduced in Congress by Senators Chris Coons (Democrat of Maryland) and Jerry Moran (Republican of Kansas). In 2013 it was amended to broaden the scope of energy projects that it covers. The bill has been “referred to Committee on Finance” and is still waiting for a hearing in Congress.119 The Joint Committee on Taxation studied the financial impact of extending the Master Limited Partnership Parity Act to clean energy. It concluded, “The MLP Parity Act is a bargain.” The Committee said the Act could result in $10 billion in clean energy investments “right away.”120 Moreover, the Act would cost taxpayers $307 million over five years and $1.3 billion over ten years compared with a forecasted taxpayer cost of $6.7 billion over ten years for existing fossil-fuel MLPs.1362

Mosaic is also building an Internet platform that it hopes will democratize energy finance. Individuals will be able to invest directly in, and profit from, the multi-trillion dollar solar infrastructure. In November 2013, SolarCity offered the first residential solar securitization deal in history. The deal opened the door to a more liquid residential solar finance market. The SolarCity deal was for $54.4 million in solar asset-backed notes at a 4.8-percent interest rate.1211382

The following month, Hannon Armstrong, the first investment fund to successfully go public as clean energy REIT, announced that it sold $100 million of asset-backed sustainable yield bonds™ (HASI SYBs) with an even lower yield: 2.79 percent. Kristian Hanelt of San Francisco-based Clean Power Finance (CPF) expects the solar finance market to become a robust capital market by 2016. Clean Power Finance is a financial services and software company that manages a half-billion dollars of residential solar project financing capital.122 The CPF business model takes advantage of Internet cloud-based software tools. Using these tools, investors and lenders can access and invest in residential solar projects managed by solar installers who also use the CPF website.1386

Chapter 3: Electricity 2.0. Distributed, Participatory Energy and the Disruption of Power Utilities “You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.” – Buckminster Fuller “Cellular phones will absolutely not replace local wire systems.” – Martin Cooper, co-inventor of the first handheld mobile phone (1981) “Change before you have to.” – Jack Welch, former CEO, General Electric1397

Palo Alto has a 100-percent clean energy mandate and is on track to buy 30-percent clean energy by 2015 and 48-percent clean energy by 2017. These numbers don’t include solar that homeowners and businesses have installed on their rooftops. Not only will Palo Alto be powered with inexpensive and 100-percent clean electricity, the low costs it pays for electricity are guaranteed for the next 20 to 25 years. As solar technology improves, the market scales and financing costs decrease. Solar generation costs are decreasing quickly. The distributed nature of solar makes the disruption of the existing utility business model inevitable. This disruption will happen faster than existing command-and-control energy companies expect. Existing energy companies are missing the big picture. Every single aspect of solar is distributed: technology innovation, design and development, finance, installation, and maintenance. Some pundits expect many decades to pass before the market adopts solar energy, but solar markets, because of their distributed nature, can turn on a dime. Most utilities have responded to the distributed clean energy disruption by hiring lobbyists, lawyers, and accountants to justify charging higher rates and new fees. What these “captains” of the energy industry are doing is tantamount to raising the price of food on the Titanic. Higher rates and new fees may increase their short-term cash flow but they won’t prevent the inevitable disruption.1413

Furthermore, the 11-percent penetration rate is an average number for Australia. In the state of South Australia, a full 20 percent of homes have solar rooftops. Some neighborhoods have a solar penetration of 90 percent, according to Mike Swanton of Energex, a Queensland utility.133 What happens to a power utility when users start generating their own solar energy? 1. Demand for utility energy drops. As users generate their own energy, they buy less from the utility. 2. Competition increases. The utility competes with myriad solar installers. 3. Utility revenues drop. As demand drops and competition increases, the utilities make less money. 4. Utility margins drop disproportionately. Solar generates the most energy during peak pricing billing cycles, which undercuts the power utility’s highest margin products. Australian electricity prices have gone up 50 percent over the last five years, from about 25 ¢/kWh to 38 ¢/kWh.134 The price increases occurred despite the fact that Australia is a major coal and natural gas producer. In 2103, solar was already as low as 12 ¢/kWh — and dropping.1439

At the retail level, distributed solar generation is disruptive to the conventional utility business model because it destroys its most lucrative revenue stream: peak pricing. Utilities have historically made a disproportionate amount of income from something called peak prices. Arizona Public Service, for example, may charge just 5 ¢/kWh during off-peak hours, but nearly five times as much (24.4 ¢/kWh) during on-peak hours, and nearly ten times as much (49.4 ¢/kWh) during the “super-peak” hours of June, July, and August, when Arizona is at its hottest1456

Power plant operators bid at their marginal cost, that is, at the cash cost to produce their next unit of energy (MWh). The marginal cost is mainly determined by the cost of fuel. The marginal cost of solar (and wind) is zero. Because the cost of sunshine fuel is zero, the cost of producing the next unit (MWh) of solar energy is also zero. Solar (and wind) can always clear competitive markets because they can bid at a marginal cost of zero and can therefore always clear the market and always sell for more than the marginal cost. This is not the case with operators of fossil fuel and nuclear power plants. Their marginal costs are set by the high and increasing cost of fuel. Figure 3.4 —Wholesale Markets Clearing Price auctions.1495

of power, and their retail business. The monopoly position that utilities enjoy has allowed them to be inefficient and still make guaranteed above-market returns on their capital. However, as electricity markets have become more competitive and as independent generators have been allowed to enter the market, the conventional utility business model has shown its inefficiencies. When there is a high penetration of solar, conventional utilities — utilities that own large fossil or nuclear generators and sell to retail end users — see their margins being squeezed on both the retail and the wholesale sides. Both wholesale and distributed solar generation are already disrupting the conventional energy companies’ century-old business model. It’s not just residential users who are generating their own solar energy. Commercial customers are doing that, too.1513

It occurred to me that designing a wind or solar energy power plant is very much like an information technology programming project. In both cases, highly skilled individuals collaborate over the Internet using open data, big data analysis, open source technology, and knowledge that others had previously created and shared with the world. The wind data for the Republic of Georgia plant was publicly available from NASA and the U.S. Department of Energy. Google had built Google Earth and the tools to create maps and the designs to take advantage of these maps. The wind simulation software was originally an open source program; dozens, maybe thousands of programmers and wind energy engineers had given millions of hours to build this tool and make it available to others. Anyone who wanted to improve this software could do so. The simulation exercise might even be run over a network of computers in several countries to take advantage of downtime in many people’s computer usage. I imagine our Spanish wind assessment engineers used computers in their American office that were idle after the American team went home. Or they may have used a cloud service like Amazon’s Web Services (also a standards-based, openly available platform). All the elements that had contributed to the growth of the Internet itself were there. Electricity 2.0 is both an information technology and an energy infrastructure. As such, it is governed by information economics. Like computing, solar and wind are based on economies of increasing returns.1748

The solar industry involves highly skilled individuals collaborating over the Internet who build using open data, big data analysis, and open-source technology. They take advantage of knowledge that others previously created and shared with the world. Google built Google Earth and the tools to create maps and the designs to take advantage of it. The solar insolation data was based on publicly available information from NASA and the U.S. Department of Energy. Individuals and companies around the world are contributing content, technology, and skills to improve these tools. Sungevity takes this technology mash-up and adds its unique technology skills and intellectual property. Here you have the Silicon Valley ethos in a nutshell. Silicon Valley bits have blended with solar electrons to create an open, scalable Internet-based infrastructure. The economics of this bits-and-electrons Internet infrastructure is based on increasing returns against which the extraction-based, command-and-control, atom-based energy industry can simply not compete.1790

The Edison Electric Institute tells its members that it isn’t necessary for utilities to change their business model. Instead, they should go to the public utility commission and ask for more money from ratepayers: So, while the telecom example is a tale of responding to the threat of obsolescence, the near-term challenge to the electric sector is providing the proper tariff design to allow for equitable recovery of revenue requirements to address the pace of non-economic sector disruption.169 Translation: “Don’t worry, just raise prices!” Some of the “immediate” and “long-term” actions it recommends include: A monthly customer service charge. Charging ratepayers a fee to help the utility invest in new equipment. Charge a fee to customers for leaving the utility. In an era of increasing customer choice and distributed clean energy at lower prices than conventional sources, the electric utilities are being told to raise the barriers to consumption. The Edison Electric Institute promotes the “head in the sand” strategy. Just wait for your Kodak moment! It will come soon enough! In the meantime, utilities in Europe, where clean energy adoption is leading the rest of the world, are already feeling the pain of disruption. Since its peak in 2008, the top twenty European electric utilities have lost half their market valuation, dropping from a trillion Euros ($1.3 trillion) to half a trillion Euros ($650 billion) in market capitalization.170 German-based utility giant E.ON is an example of the disruption of utilities in Europe (see Figure 3.9). As of October 2013, its American Depositary Receipt (ADR) stock had dropped by more than half to below 15 from its highs near 30 in 2009 and 2010.1804

A drop in a utility’s stock price means that the equity value drops, which implies that capital investment will drop. The deterioration in credit ratings means that the interest rates that a utility has to pay on the power plant mortgage goes up. This combination of less equity and higher interest payments delivers a one-two punch increase in the cost of capital. The higher cost of capital means that fewer projects will be NPV (net-present-value) positive so fewer conventional power plants will get built. It also means that those plants that do get built will produce more expensive power because their interest rate payments are higher. This expensive power cannot compete with ever-decreasing cost of solar (and wind). Conventional utilities are already falling into a vicious cycle. As they become less competitive, they lose customers to solar and find it harder to raise capital at low rates. They become even less competitive.1822

Conventional power utilities will soon be hit by several disruptive waves. Each wave will carve a large hole in the utilities’ century-old business model. Networks of sensors, machine learning, and connected devices allow for the distributed generation of power and customer-centric energy management. All this technology will hit utilities were it hurts most by Lowering the clearing price of wholesale competitive markets Flattening the peak premium pricing in retail markets Lowering demand because of increased end-user self-generation A critical mass of customers going net-zero This one-two punch of decreasing revenues and dramatically lower earnings would scare most industrialized organizations, but not power utilities. Utilities are so deeply entrenched in the regulatory process, they don’t have to reorganize to meet advances in technology, changes in consumer preferences, and market changes.1830

Storing electricity at the point of use provides consumers with several benefits. At a minimum, someone who stores the equivalent of a few hours of daily demand can purchase electricity (from a solar rooftop or from the grid) when electricity costs are low and use it when electricity costs are high. Storing four hours of power, for example, would help consumers avoid the high cost of power during peak usage cycles. This can make a difference of hundreds of dollars per month during the summer months when usage is at its peak.1870

Utah’s Rocky Mountain Power (RMP) recently asked its public utility commission to approve a minimum monthly bill of $15, a customer service fee of $8 per month, and a monthly solar fee of $4.25. That’s a minimum monthly bill of $27.25 plus actual usage.174 Fees like this make on-site storage more financially viable. At $600 per kWh it would cost a user $18 per month to have four hours of on-site storage to store excess solar generation. By 2020 this number is expected to fall to $7.70 per month. Not only will solar generate cheaper energy than utilities, it will make more financial sense for users to store energy on-site rather than share it with the grid. By 2025 it will cost just about $12.30 per month to have 20 kWh of onsite storage. For the average American residential user 20 kWh represents two thirds of daily consumption. Fifty or sixty million American homeowners will essentially be able to generate all their energy using solar; they will be able to store what they don’t use at the moment of generation for the rest of the day and night. For just $18.30 per month they will store enough energy to essentially get off the grid. By inventing new fees, raising existing fees, and raising energy prices, the utilities are increasing their short-term cash flow at the expense of their very survival.1890

By increasing the price of their service at a time when the cost of distributed solar generation is dropping dramatically and the cost of on-site storage is becoming competitive, utilities are actually helping to accelerate the adoption of these technologies. Utilities are helping their own Kodak moment to arrive faster.1900

competitive. A report by investment bank CreditSuisse points out that the number of nuclear plant outage days has increased significantly, necessitating higher capital costs for repairs and upgrades (see Chapter 6.)1942

disruption fighting to keep their cozy monopolies. Utilities in California and other states and countries know that the best way to increase their cash flow is the good-old fashioned way — in the closed-door meetings of state legislatures, public utility commissions, and regulatory bodies.1976

Having learned how the cell phone swiftly made the landline phone system obsolete, the power utilities want to have it both ways: they want to milk their obsolete infrastructure through their monopoly power and at the same time use the regulatory system to profit from the transition to a distributed power infrastructure by taxing it.1994

“I do not believe the introduction of motor-cars will ever affect the riding of horses.” – Scott-Montague, UK MP, 1903. “The next 20 years of technology change will be the equivalent of the last 100 years.” – Ray Kurzweil. “The best way to predict the future is to invent it” – Alan Kay. On November 11, 2013, Motor Trend magazine announced the winner of Motor Trend’s 2013 Car of the Year award. It was the Tesla Model S.1882006

Model S achieved the highest safety rating of any car ever tested by the National Highway Traffic Safety Administration (NHTSA).195 Then Consumer Reports called the Tesla Model S the best car it ever tested.196 As I walked through the SF Auto Show on December 2, 2013, I thought of how little the industry had changed in a century.2028

Market disruption will soon make the internal combustion engine (gasoline or diesel) a thing of the past. Even worse from the standpoint of gasoline and diesel cars, the EV is not just a disruptive technology; the whole business model that the auto industry has built over the past century will be obliterated. The San Francisco Auto Show of 2030 will be quite different from the 2013 version. Even today’s impressive BMW 950 and Audi R8 will be the equivalent of beautifully designed horse carriages. There are nine reasons why the electric vehicle is disruptive. 1. The Electric Motor Is Five Times More Energy Efficient Of the major energy users in the U.S., the transportation sector is the most wasteful. A full 79 percent of petroleum energy used in transportation goes up in smoke (see Figure 4.2). On average, only 21 percent of the gasoline or diesel (both of which are petroleum-based) that is pumped into an internal combustion engine turns into useful energy.2045

When you consider both city and highway driving for an average car in the U.S., only 17 to 21 percent of the source gasoline energy is used to actually move the wheels, according to the U.S. Department of Energy (see Figure 4.2). A century of knowledge gained from building billions of cars and investing hundreds of billions of dollars in research and development has given us an internal combustion engine vehicle with an energy efficiency of about 21 percent.2056

the electric vehicle does not have: radiators, pistons, an exhaust, a crankshaft, a clutch, pumps, and myriad other thermal engine necessities that waste so much energy (see Figure 4.2). An electric motor has an energy conversion efficiency of up to 99.99 percent.200 Tesla’s first-generation electric vehicle, the Roadster, had an overall drive efficiency of 88 percent.201 This figure is four to five times the energy conversion efficiency of an average American gasoline car. The electric motor produces not just a smoother ride, but a far more energy-efficient ride.2067

Because the internal combustion engine is a heat engine, you need to increase the temperature of the engine to achieve higher energy conversion. Even coal and nuclear power plants, with their far higher temperatures, waste two-thirds of their potential energy. To give an idea of how energy-efficient electric vehicles are, the Tesla Model S electric engine is three times more energy-efficient than even large multi-billion dollar nuclear or coal plants! 2. The Electric Vehicle Is Ten times Cheaper to Charge2074

It costs $15,000, or $3,000 per year, to fill up a Jeep Liberty over five years, according to Consumer Reports.202 That’s assuming you drive 12,000 miles per year. To drive 12,000 miles in a Tesla Roadster, it costs $313.2079

Electric motors are four to five times more energy-efficient than gasoline vehicles. Gasoline is two to three times more expensive per energy unit than electricity.2086

The Electric Vehicle Is Ten Times Cheaper to Maintain Conventional cars are supposed to have an oil change every three to five thousand miles. There’s no need for that with an electric vehicle (EV). But it’s not just oil changes that make the EV superior. Because the EV is powered by an electric motor, it doesn’t need any of the parts that have to do with combustion: no spark plugs, starters/alternators, fuel injector, combustion chamber, pistons, piston crown or cylinders, filters, or exhaust. The electric vehicle has no crankshaft or timing belt or catalytic converter. Because an electric vehicle has fewer parts, it is not nearly as needy as the internal combustion engine car. Fewer parts need to be loaded on the chassis (see Figure 4.3). Data is limited, but it’s safe to say that electric vehicles need 90-percent fewer repairs and maintenance work than gasoline/diesel engines do. Consequently, an EV driver spends 90-percent less money on repairs during the lifetime of the car.2093

The Electric Vehicle Will Disrupt the Gasoline Car Aftermarket In 2010 there were 257,576 light vehicle repair shops in the United States.204 These shops broke down as follows: 3,978 department store locations performing services; 16,800 vehicle dealer repair locations; 77,674 general repair shops. These shops perform all kinds of maintenance and repairs on the nearly 250 million light vehicles on American roads. Total car manufacturer aftermarket revenues in 2010 were $83 billion; revenues are expected to grow to $98 billion in 2017, according to Frost & Sullivan.205 Aftermarket costs include repair parts for things like carburetors, spark plugs, starters/alternators, filters, and exhaust components. But guess what? The electric vehicle doesn’t have any of these parts. Remember those engine oil changes every three to five thousand miles? There’s no need for that with an EV. Ultimately it’s not just technology that destroys industries; it’s the fact that the disrupting companies have business models with which the incumbents cannot compete. Remember Kodak? They didn’t just lose sales of film to digital camera owners. They lost the whole aftermarket for developing film: equipment, paper, and chemicals. Aftermarket income is an essential component of the conventional auto industry business model. Car manufacturers make an additional $25.8 billion selling tools and equipment to the aftermarket value chain.2103

In New York in 1891, Nikola Tesla publicly demonstrated the world’s first wireless transmission of energy by electrostatic induction.207 Tesla went on to invent many of the key technologies and concepts that underlie our electric power infrastructure. Fast-forward 120 years and you’ll find electric buses in Italy being recharged at bus stops while passengers get on and off (see Figure 4.4).208 The electric vehicle industry has adapted induction power transfer (IPT) so cars can be charged wirelessly without a “typical” charging infrastructure.2121

Wireless charging essentially untethers electric vehicles. Because they are replenished many times a day, they can use a smaller battery. A bus with a 240 kWh battery operating 18 hours per day can go the same distance as one with 120 kWh that is recharged using inductive charging, according to Conductix-Wampfler, the company that built the inductive chargers used in Italy. Using a smaller battery cuts the cost of running the bus by $100,000 or so (at current battery prices).2130

The EV Has a Modular Design Architecture The standard architecture for cars has consisted of a single engine that powers two wheels or four wheels via a combination of a transmission, a differential, and drive shafts. The first generation of electric vehicles used this single-motor architecture. However, as electric automakers have developed confidence in their engineering prowess, they have started to take advantage of something that the electric motor has but the fossil-fuel engine cannot compete with: modularity. The Tesla Model X and the Audi eTron Allroad will have two electric motors, a rear drive unit to power the rear wheels and a front drive unit to power the front wheels (see Figure 4.5). Like an Intel processor with multicore architecture, this modularity allows for increased power, more design flexibility, and more traction control. It also makes the EV more secure. Electric motors last far longer than internal combustion engines. However, should one of the two motors fail, an electric car does not have to stop running. The second motor can still power the car. And you don’t have to stop at two motors. You can have four electric motors, one driving each wheel. Figure 4.5—Tesla X dual electric motor (front drive unit and rear drive unit).2139

The electric vehicle product development process is shortened; it resembles the computer industry’s ultra-fast development cycles. Tesla and the other EV companies will be on a product development cycle at the exponential speeds of Moore’s Law; Detroit will be on conventional linear speeds.2169

Electric Vehicles Can Contribute to Grid Storage and other Services Imagine being paid to own a car. A recent report by the California Public Utilities Commission concluded that EV owners could be paid up to $100 per month for providing electricity to the electric grid infrastructure.2172201

Being paid, say, $100 per month for performing grid services would change the economics of car ownership.219 A car would go from an expense-only item to an income-producing investment.2221

Cell phones never really matched the cost of landline phones yet they managed to disrupt the landline telephony industry. Last I checked, a landline telephone costs $10 or $20, while the Apple iPhone costs $600. The cell phone is a reinvention of the old conventional phone, not just a substitution. The cell phone offers value (mobility, for example) that the landline phone can’t match; the smartphone has turned into the center of our social lives. The innovation that enabled cell phones to disrupt the landline phone industry in the U.S. was a business model innovation: the customer committed to a two-year service contract and, in exchange, the carrier financed the phone over that period of time with no down-payment required. The innovation that enabled the internal combustion engine car to disrupt horse carriages was also a business model innovation: car loans. By the time cars dropped in price to match the horse carriage, the disruption was pretty much over. Similarly, by the time the electric vehicle matches the capital cost of the gasoline car (sometime before 2030), the disruption of the gasoline car industry will be pretty much over. When a consumer asks “How much car can I afford?” she is asking what the monthly payment is that fits her budget. Electric vehicles just need to match or be close to the monthly payment of a gasoline car.2232

Free Maintenance Electric vehicle (EV) companies could make the demise of the internal combustion engine industry happen even faster by offering free maintenance. EV maintenance costs are an order of magnitude lower than internal combustion engine costs. An electric motor can last for decades, but a heat engine breaks down time and again. The EV doesn’t have hundreds of parts — carburetors, spark plugs, starters/alternators, filters, and exhaust components — that need constant care. Imagine an EV company that offered free maintenance for five years or 60,000 miles. This would be another disruptive move that ICE companies couldn’t match. The aftermarket business is actually a big cash cow for car manufacturers today. Turning a major revenue line item like the aftermarket business into an expense line item would probably send most car manufacturers into bankruptcy. Earlier in this chapter I explained nine reasons why the electric vehicle is disruptive. As we speak, two guys or gals in a Silicon Valley garage may be at work on new disruptive business models that Detroit won’t be able to compete with.2282

In the summer of 2010, I gave a keynote speech in Dickinson, North Dakota, where I predicted that gasoline cars would be obsolete by 2030. (The video is here:

When Li-on reaches $100/kWh, the batteries for an EV with a range of 200+ miles will cost about $5,000. Assuming that the battery is about a third of the price of an EV, you’ll be able to purchase an entry-level equivalent of a Tesla Model S for $15,000. By comparison, the average cost for a new vehicle in the U.S. in 2013 was $31,252.227 Even the low-end vehicles, such as the Huyndai and Kia brands, sold for an average $22,418.2314

The gasoline industry will simply not be able to compete with a Tesla Model S-quality vehicle that retails for $15,000 or $20,000. Not Kia, not GM, not Toyota, not BMW. The gasoline car industry will be in trouble when storage batteries reach $100/kWh. At that point, the disruption will be long over. My initial projections in 2010 pointed to the industry reaching $100/kWh in 2028. Now it looks like it will happen even sooner. My 2010 prediction was in the “ballpark,” but electric battery costs are going down a little faster than I predicted in 2010. As is usually the case in technology markets, the virtuous cycle of innovation, competition, and scale is pushing Li-ion battery costs down a bit more quickly than anticipated. “The cost of manufacturing Li-on batteries has dropped from $1,000-$1,200 to $600 in just three years [from 2009 to 2012],” said the secretary of the U.S. Department of Energy in 2012.230 I was not far off. My original prediction said that, by 2014, EV-Li-on batteries would be around $600/kWh. Today, EV batteries are in the neighborhood of $500/kWh. My New Prediction for the End of Gasoline Cars by 2030 Tesla’s car batteries are composed of thousands of small Li-on cells similar in size to the Li-on cells in laptop computers. It makes sense that the cost curve for EV Li-on batteries should approximate that of laptop computer Li-on batteries. How much further will the cost of Li-on batteries fall? We can look to a recent historic precedent to guide our thinking in making this prediction: the cost of Lion batteries to power laptop computers, smartphones, and tablets. According to Deutsche Bank, laptop computer battery costs fell from $2,000 to $250 over a period of about fifteen years. This represents a cost improvement rate of 14 percent per year.231 Figure 4.10—Electric vehicle Li-ion battery cost projection in $/kWh. (Tony Seba) Based on this yearly cost improvement rate of 16 percent (see Figure 4.10), my 2014 cost projection for Li-on batteries is $498 per kWh in 2014. This fits the data better. I love the scientific method. Evidence-based data should drive hypotheses, not the other way around. It makes sense that Li-on battery costs are falling faster than the historical precedent. Never has investment in electricity storage been so high. At least three multi-trillion dollar industries are now investing billions to come up with better batteries: electronics, automotive, and energy. Apple, Samsung, and Google are as interested in batteries as Tesla, SolarCity, and General Electric. Tesla recently announced a $5 billion next-generation battery factory (dubbed “GigaFactory”) in the U.S. This investment would singlehandedly double the world’s Li-on battery manufacturing capacity. The factory will open around 2017 and produce enough batteries for 500,000 cars per year by 2020. Tesla projects it will sell about 35,000 cars in 2014, so the factory would represent a minimum car unit growth of fourteen times in six years.232 Panasonic, the Japanese electronics giant, was said to be in talks to invest $1 billion in Tesla’s “GigaFactory.”233 SolarCity has been using Tesla’s batteries as part of its solar panel installations.234 These batteries allow solar customers to store solar power as well as purchase grid electricity when it’s cheap and use the electricity when it’s expensive. In fact, the factory itself will get 100 percent of its energy from solar and wind power generated near the factory. Presumably, the factory will make the batteries that will store the solar and wind energy that it will use to manufacture more batteries.2323

The 16-percent battery cost improvement rate does not even take into account the possibility of major breakthroughs. Tesla management is talking about a “step change in battery technology within five to ten years that would enable 500-1,000 miles of range and full-charging within seconds.”2352356

Academic institutions around the world, including my alma maters Stanford University and the Massachusetts Institute of Technology, have made energy one their highest research and development priorities. Their work has already started to bear fruit. MIT Professor Donald Sadoway has focused on the development of liquid metal batteries for grid storage using cheap and widely available materials.236 Ambri, the first successful spinoff of his lab at MIT, quickly raised $15 million in venture capital from Bill Gates and others; Ambri has already been named one of the “50 Disruptive Companies” of 2013 by MIT Technology Review.237 Stanford Professor Yi Cui’s group is using nanotechnology and cheap materials such as silicon and sulfur to build batteries from the ground up using carbon nanotubes, graphene, and other advanced materials. Early results point to order-of-magnitude improvements in cost as well as energy densities for energy storage.238 An example is a new configuration for Redox Flow batteries that could decrease their production cost to about $45/kWh. My new electric vehicle Li-ion battery cost projections (see Figure 4.10) fit the recent past (2010–2013) better. Assuming the next dozen years are also governed by the 16-percent yearly improvement in the cost of Li-on batteries, the automotive industry is in for a quick transformation (see Table 4.1). Table 4.1— Electric vehicle Li-ion battery cost projection in $/kWh. (Tony Seba) Essentially, Table 4.1 points to EV-batteries reaching $100/kWh in the 2023 timeframe. By 2025, the cost falls to $73/kWh This looks to be an aggressive price reduction projection. Tesla, however, is already ahead of the curve. While Tesla does not disclose its costs, we can deduce the approximate value of its batteries from published prices.2359

Tesla is already in the disruption sweet spot. The highest ranked affordable midsize SUV in America in 2013 was the 2014 Buick Enclave, according to US News & World Report.241 The Enclave retails for $38,698 to $47,742. Tesla’s sports-utility vehicle, the Model X, to be launched in 2015, is expected to retail for $35,000 to $40,000, placing it squarely in the American mainstream.242 The base Model X will have a 60 kWh battery and a range of 265 miles. The Model X SUV, however, “will have the performance of a Porsche 911 Carrera,” according to Tesla CEO Elon Musk. This means that Tesla will offer a $100,000 performance car-quality SUV for the price of a $40,000 “affordable” SUV. The Buick Enclave does not stand a chance. Neither does any other SUV in that price category. Not even Porsche. If Tesla could manufacture millions of cars per year, the mass migration to electric vehicles would start with the Model X. However, despite their technology prowess, design capabilities and market success, Tesla has put a premium on manufacturing quality cars. Tesla’s devotion to quality has limited its ability to quickly scale to producing millions of cars per year.2386

market will probably reach the $350/kWh level in 2016 or 2017. This means that the mass migration to electric vehicles will start in 2016 or 2017. It also means that Tesla has a two-year battery cost advantage over the rest of the market. The Mainstream Mass Migration to Electric Vehicles My projections indicate that Li-on batteries will reach $200/kWh by 2020. (see Table 4.1). My projections aren’t far from the current consensus. EV batteries will reach $200/kWh to $250/kWh by 2020, according to Anand Sankaran, executive technology leader for energy storage and high-voltage systems at Ford Motor Company.243 Consulting company McKinsey pointed to $200/kWh by 2020 as far back as July 2012;244 Navigant points to costs as low as $180/kWh by 2020 When Li-on reaches $200/kWh, the batteries for an EV with a range of 200+ miles will cost about $10,000. Pushing battery costs down makes the cost of the battery a smaller portion of the cost of making the car. The original Roadster battery cost about half the cost of the car, but Tesla is pushing down battery prices. Currently the cost of the battery in a Tesla is about a quarter of the cost of the car.2452397

Assume for a moment that the battery is a more conservative one-third of the cost of an EV. At a one-third cost, when Li-on batteries reach $200/kWh, you will be able to purchase the equivalent of a 200-mile range Tesla Model S for about $30,000. In fact, Tesla’s next-generation vehicle, the Model E, currently planned for 2017, is expected to cost around $35,000. Again, the average cost for a new vehicle in the U.S. in 2013 was $31,252.246 This “average” vehicle (Toyota, Ford, GM, Honda, Nissan) will cost about the same as a Tesla EV “with the performance of a Porsche 911 Carrera”; with a sticker price around $30,000. Gasoline cars will not stand a chance. Did I mention that EVs cost 90-percent less to fuel and maintain?2407

“The electric vehicle is a re-invention of the automobile, not just a substitution,” said Masato Inoue, chief product designer for the Nissan Leaf. A Porsche cannot compete with the Model X because it will cost more than twice as much as the Model X but deliver an equivalent performance. The Enclave will cost the same as the Model X but will deliver a fraction of its performance. The electric vehicle changes the basis of competition in the transportation industry (see Figure 4.11). Of all the reasons why the electric vehicle is disruptive, this is probably the most powerful one. Figure 4.11—Electric vehicles disrupt the basis of competition in the automotive industry. (Source: Tony Seba) Neither the high-end or low-end gasoline car stands a chance against electric vehicles once EVs are in the same price range. When EV-quality batteries reach $100/kWh, the internal combustion engine industry will be toast. At that point it won’t make financial sense to own a gasoline or diesel vehicle no matter the cost of petroleum. Figure 4.12—Projected cost of an electric vehicle with a 200-mile range. (Source: Tony Seba) My projections point to the industry reaching the $100/kWh price level by 2024 or 2025 (see Figure 4.12). Starting in 2025, it will make no financial sense to purchase a new gasoline car in any market. At this point most of the high-end and mainstream ends of the market will have transitioned to electric vehicles. Even assuming that my predictions are off by five years or that it takes an extra five years to build out the manufacturing infrastructure and transition to the new EV world order, gasoline cars will be the 21st century equivalent of the horse carriages by 2030. There may still be millions of older gasoline cars and trucks on the road. Ten-to twenty-year-old cars are still on the road today. We may even see niche markets like Cuba where 50-year old cars are the norm.2425

essentially no internal combustion engine cars will be produced after 2030. Oil will also be obsolete by then. Oil will likely be cheaper in 2030 than it is today (see Chapter 8). Still, as the gasoline car industry starts imploding around 2025, the ICE car aftermarket will collapse with it. Electric cars do not need that much maintenance and do not wear out as much. There will be fewer gas stations, repair shops, and used-parts stores to cater to ICE car owners. As ICE car companies go under or move to making electric vehicles, they will stop making parts for their older cars.2442

it will be progressively harder to get parts for older ICE cars. The cost of operating and maintaining a gasoline car will necessarily go up until used ICE car ownership becomes too expensive for all but the most devoted gasoline car owners. So there’s the answer to former General Motors CEO Dan Akerson’s question of how Tesla could disrupt the “established business model” of the gasoline car industry. The electric vehicle will disrupt the gasoline car and make it obsolete by 2030, maybe by 2025. GM thought it killed the electric car in the 1990s, but the electric car may well end up killing GM and all its internal combustion engine siblings as soon as 2025.2447

“This ‘telephone’ has too many shortcomings to be seriously considered as a means of communication.” – William Orton, president of Western Union, in 1876. “The significant problems we face today cannot be solved with the same level of thinking that created them.” – Albert Einstein. “You have to let it all go: fear, doubt, disbelief. Free your mind.” -Morpheus, The Matrix.2457

Zipcar claimed more than 760,000 members and $270 million in revenues in 2012.247 According to the company, each of its cars replaced fifteen cars on the road.248 Under this 15-to-1 share-to-own ratio, Zipcar’s 10,000 cars canceled 150,000 potential car sales. To put it another way, Zipcar’s car-sharing model may have prevented car manufacturers from selling 150,000 cars. I’m not sure how many big auto executives have lost sleep over the share-to-own ratio of Zipcars to conventional cars, but they soon will lose sleep.2470

After houses, the automobile is most Americans’ largest asset, yet Americans only use their cars about two hours per day. That figure represents less than 10 percent of capacity utilization for what is a pricy asset. We pay hundreds of dollars every month for car loans, insurance, parking, repairs, gasoline, and maintenance — all for an asset that is idle and unused 90 percent of the time. Is there a way to make money from cars during this “car downtime”? Since we only use cars a couple of hours per day, there is plenty of spare capacity for people to rent cars. After all, if homeowners are happy to share their most valuable financial asset, shouldn’t car owners share their second most valuable asset too? A number of new car-sharing services with a number of slightly different business models have sprung up in San Francisco. Lyft aims squarely at the taxi market. It allows people to use their cars in their spare time as taxi cabs. Everything, from requesting a car to making payments to leaving ratings of drivers, is done with a smartphone app. Uber started by connecting limousine drivers to potential customers on-demand. The company added an eBay-like auction-pricing model to match limo demand to limo supply. The result is that a peak-demand ride can be much more expensive than a standard taxi ride. I rode an Uber car to a New Year’s Eve party in San Francisco’s North Beach and paid $50 for a ride that would have cost $15 or $20 in a normal taxi. The driver told me that after midnight rides were expected to cost more than $100. Lyft and Uber are market-making intermediaries in the personal transportation market. They help make this market more efficient by connecting sellers with spare capacity to buyers who would otherwise not be able to use this spare capacity. These companies have already transformed the taxi market in San Francisco and are expanding globally at a breakneck speed.2488

Another peer-to-peer service called aims to make better use of the world’s cars. Instead of leaving your car idle for hours every day or days at a time, GetAround asks you to rent it to a neighbor or someone who lives nearby. This service is closer to the Zipcar model in that the consumer can rent a car by the hour or by the day. Unlike Zipcar, however, GetAround does not have to buy and maintain a fleet of cars. The company just connects buyers and sellers and pockets a percentage of the transaction.2512

Who owns the autonomous car, a car-sharing company like Zipcar or an individual who rents it out while she’s at work, is irrelevant. You know there will be a car to drive you anywhere at anytime. You don’t even need to live in a high-density area. The autonomous vehicle “mobility on demand” business model would expand the transportation market. Think of the millions of people with disabilities, children, and the elderly who can’t drive cars. They would have cars to drive them to school, the park, the doctor, or to the homes of family or friends. Parents who are pressed for time will not need to drive their children to school every morning or their parents to the doctor. The blind could go to a restaurant across town. All this could be done without a licensed driver or a private car. Gasoline Cars: The Ultimate Waste Machine During rush2526

The gasoline car is the ultimate waste machine. The car has brought us waste in at least six different dimensions: 1. Waste of lives 2. Waste of time 3. Waste of space 4. Waste of energy 5. Waste of money The autonomous car is a game-changing product in terms of minimizing every single waste item on this list. Waste of Lives The number of deaths caused by car accidents is a human tragedy of unspeakable proportions. In the U.S. alone, six million car crashes caused 32,788 deaths in 2010. It is estimated that 93 percent of those deaths were caused by human error. In 2009, 2.3 million adult drivers and passengers ended up in a hospital emergency room.257 To put highway deaths in context, 58,220 Americans died in the Vietnam War (1956-1975).258 During those same years, 757,538 people (thirteen times as many) died in motor vehicle accidents in the United States.259 Globally 1.24 million people died of traffic-related accidents in 2010, according to the World Health Organization.260 Almost half of those deaths were pedestrians, cyclists, or bikers.261 Additionally, 20 to 50 million people suffer non-fatal injuries because of traffic accidents annually. Worldwide, death by traffic accident is the leading cause of death for people between the ages of 15 and 29; it is the second leading cause of death for children between the ages of five and 14. More children age five to 14 die in auto accidents than die from malaria, tuberculosis, or measles.262 Clearly, humans are not great drivers. We are easily distracted. We drink and eat while driving. We text and speak on the phone. We reach for radio dials and the glove box. We put on make-up, talk to our fellow passengers, try to reason with kids in the backseat, and daydream — sometimes all at the same time. We’re also at the mercy of our physical limits. How well we can see, our reaction times, and even our sleep patterns determine how well we can drive a car. Mathematical models based on Center for Disease Control data imply that drowsy drivers might be involved in 15 to 33 percent of all fatal crashes in the United States, according to Harvard University Professor Sendil Mullainathan.263 Autonomous cars can be superior drivers in many ways. They have a 360-degree “visual range.” Computers don’t get distracted. They can “see” at night and are not at the mercy of sleep patterns. Their attention doesn’t waver because they’re texting or talking on the phone. They don’t drink, travel at inadequate speeds, or daydream. Autonomous driving technology is not perfect yet, but autonomous cars may already drive better than most humans. “Autonomous cars are 6,500 times better at detecting danger than humans,” said Masato Inoue, chief product designer of the Nissan Leaf. The Google self-driving car has gone 500,000 miles without causing an accident. It has, however, been rear-ended — by a human driver. Autonomous cars are getting exponentially better because their technological components — vision, sensing, processing, and machine learning — are getting exponentially faster, cheaper, and better. The “artificial intelligence” software that drives autonomous cars is also improving. The amount of data a self-driving car can access is increasing exponentially; the computing platform on which it runs is also improving exponentially. When a human being learns a lesson, he may or may not share it with others. And even if he shares it, whether others can take heed of the lesson is questionable. A lesson doesn’t really sink in unless you actually experience it. People tend to make the same mistakes again and again. This is one reason why there are so many traffic accidents. By contrast, the Google car gathers more than 1GB of data per second!264 To give you an idea of how much data this is, the iPhone 5 has 16GB to 32GB of data storage capacity. The Google car would fill the data storage capacity of Apple’s latest smartphone in 32 seconds or less. Like every other computing platform, data generation is growing exponentially. It won’t be long before an autonomous…2549

Any accident anywhere will make every autonomous car everywhere a better driver. In information economics this is called network effects. The value of the network increases exponentially as the network acquires more information. Autonomous cars’ ability to learn will improve exponentially, which will soon make them smarter, better, faster, and safer than the best human drivers.2594

Autonomous cars will soon save more than one million lives a year. This alone is potentially revolutionary. Waste of Space If an archeologist from outer space came to study earth, she would rightly conclude that cars are the dominant life form on the planet. More urban space is dedicated to the car than the human being. In North American cities, roads and parking lots respectively account for 30 and 60 percent of the total surface.265 This does not include driveways and garages. Highways are also a massive waste of space. Automobile driving occupies ten to a hundred times more road space than other forms of transportation. For instance, it takes 200 m2 (2,152 square feet) of road space per car passenger versus 30 m2 (323 sqft) for arterial driving, 2 m2 (21.5 sqft) per public transportation passenger, and 3 m2 (32.3 sqft) for walking.266 It’s easy to think there aren’t enough highways when you’re going down the road at 60 miles per hour (100 Km/h) and traffic is moving smoothly, but vehicles only use 5 percent of the road surface, according to UC Berkeley Professor Steven Shladover.267 This means that, at best, 95 percent of the highway surface is not being used at any given time. This space is wasted because, at highway speeds, cars need 120 feet (40 m) to 150 feet (50 m) of space in front for safety purposes. They also need lanes that are twice their width.2599

automated vehicles require 25 percent less space for merging and lane changing.268 Cars equipped with adaptive cruise control (ACC) can improve highway capacity by about 40 percent. Using both ACC and inter-vehicle communications can boost highway capacity by an astonishing 273 percent, according to research done at Columbia University.269 In other words, autonomous vehicles could end congestion on highways by increasing highway capacity by 3.7 times. After the self-driving car disruption, we will have to decide what to do with all that unused highway space. Waste of Time Traffic congestion cost Americans $121 billion in 2012; this figure is expected to grow to $199 billion in 2020, according to TTI’s Urban Mobility Report.270 Congestion costs Americans 4.8 billion hours of time, 1.9 billion gallons of wasted fuel, and $101 billion in combined delay and fuel costs each year.271 While I have mostly looked into cars, anyone who has seen a FedEx or UPS truck double-park knows that trucks also waste time, space, and energy. In 2004, before the explosion in truck deliveries caused by the growth of online commerce, delivery trucks caused an estimated 1 million hours of vehicle delays.272 A study found that trucks double-parked the equivalent of seven hours per day, turning road space into their own parking places and worsening already-congested cities during peak daytime hours.273 Time wasted on the road is also stressful. MIT Professor Carlo Ratti, who developed the road frustration index to quantify the impact of traffic on mental health, concluded that the stress of city driving is as high as that of skydiving.274 By decreasing congestion, self-driving cars will also dramatically decrease commuting time. Furthermore, because self-driving cars don’t need us to direct them, time we waste today in our cars will be turned into productive time. Some may choose to surf the web (or whatever has disrupted the web by 2030) or sleep in their cars. Either way, time not driving is time added to our lives. Self-driving cars will also save us all the time we waste parking and looking for parking places. They will drop us off and go on their merry way to self-park or pick up the next ride. Waste of Energy In congested urban areas, 40 percent of all gasoline usage is wasted looking for a place to park the car, according to the MIT Media Lab.275 Congestion costs Americans 1.9 billion gallons of wasted fuel and $101 billion in combined delay and fuel costs. That’s $713 per year for each commuter.276 The first energy-saving feature of the autonomous vehicle is very mundane: self-parking. When they do need to park, autonomous vehicles (AVs) are more precise at parking; they can squeeze into smaller parking places. Some parking places the average driver would think too small to fit a car will be suitable for autonomous vehicles. To find a place to park, the AV can communicate with sensor-based parking places (or even with other vehicles) in a radius several blocks wide; the AV can then go directly to the parking place without having to drive around and search. Reducing wind drag is another energy-saving feature. Because they can sense the presence of other cars better, AVs can drive closer to one another, which reduces wind drag. This reduction in wind drag would cut fuel usage by an estimated 20 to 30 percent, according to the Rocky Mountain Institute.277 Waste of Money An American minivan owner who drives 10,000 miles per year spends an average of $8,161 in annual car costs, according to the American Automobile Association (AAA).278 This is a relatively large sum when you consider that the median wage in America in 2011 was $26,684.279 Car costs are after-tax expenditures; they absorb more than a third of the average American’s income and cost 81.6 cents per mile to drive. Global monetary losses due to traffic-related injuries cost $518 billion per year; they cost governments between 1 and 3 percent of their gross national product, according to the World Health Organization.280 Traffic congestion costs Americans $101 billion in…2613

Andy Palmer, executive vice president of Nissan.283 BMW and Mercedes have also pledged to have an autonomous vehicle ready by 2020. Andy Palmer also said self-driving technology would be available across the entire Nissan portfolio within two vehicle lifecycles after its first self-driving car.284 He re-iterated that the company was committed to “zero fatality and zero emissions.”2661

Mercedes-Benz “cross traffic assist” technology helps the driver avoid rear-end collisions as well as cross-traffic collisions that occur, for example, at junctions in the road. The car’s stereo cameras and system of short-, medium-, and long-range radars generate visual data. The car processes this data to determine whether cross-traffic (from bicycles to trucks) poses a collision risk. If a collision is imminent, the car not only warns the driver, it applies the brakes to bring the car to a full stop.288 The BMW X5 offers “traffic jam assistant” technology by which the car drives itself in dense traffic at speeds of up to 25 mph (40 km/h). In other words, the BMW can become a self-driving car in traffic jams.289 Add up these tasks and you have a semi-autonomous car.2694

Several cars already have traffic jam assistants by which the driver can give control to the car (level 3). Fully automated campus shuttles (level 3) where there is no driver at all (but the car travels only on pre-specified paths) are already in operation.2702

LIDAR makes up nearly half the cost of a Google car. For the car to drop in price, LIDAR technology will have to drop first. Airborne LIDAR technology, measured by pulse repetition frequency (PRF), has been improving by roughly 100 percent every two years.293 This represents an improvement rate of 41 percent per year, which is similar to Moore’s Law. If this trend continues, LIDAR technology will drop from $70,000 in 2012 to a more manageable $4,481 in 2020 (see Figure 5.4). Google being an information technology company with a view to the future, I assume that much of the equipment in a Google car consists of computers, communications, sensors, optics, and other advanced technologies. Most of these technologies are improving exponentially. Assuming that Moore’s Law applies to all this equipment, the technology components that cost Google $150,000 in 2012 would cost $9,691 by 2020 and drop even further to $3,425 in 2023 (see Table 5.2). From a cost perspective, it makes sense that Nissan, BMW, and Mercedes-Benz have all announced self-driving cars by 2020.2715

Technological change can happen faster than anticipated. In late 2013, Google announced that its next-generation self-driving car would use a smaller LIDAR sensor with at least twice the technological performance as before. This LIDAR sensor would cost just $10,000, or one seventh the price of the previous version.294 It appears the technology cost curve is accelerating even faster than predicted. This is a point I emphasize in my market disruption class at Stanford: Acceleration is accelerating! I met a Silicon Valley startup CEO who claims to have developed car-quality LIDAR sensor equipment that will sell for $1,000. LIDAR is just one of several technologies for vehicle visualization. Machines can also use high-definition video technologies to scan and understand the environment. Semiconductor companies are racing to develop computing and sensor hardware and software that can read input from cameras. This technology will make cars more autonomous. Fujitsu, for example, has announced the “world’s first 360° wraparound view system with approaching object detection” (see Figure 5.5).295 According to Fujitsu, its MB86R24 chip comes equipped with six HD input channels (video cameras) and three display output channels; the chip incorporates “approaching object detection functionality” whereby drivers are notified when people and objects such as bicycles are approaching. Fujitsu’s “360° wraparound view system” allows drivers to check their entire surroundings in 3D from any angle. The Fujitsu system costs just 5000 Japanese Yen, about $50, according to a company spokesperson.2726

it’s not just technology that is disruptive. It’s the business model that the technology enables, namely the car-as-a-service business model. Imagine you can hail a car anytime from anywhere and have the car show up at your doorstep within minutes. Companies like Zipcar, Uber, and Lyft provide versions of this service today. Imagine being picked up by a self-driving car instead of a human-driven car. Try this thought exercise: Assume that every vehicle has autonomous technology capabilities and that every car owner makes his or her car available to a company under a car-as-a-service contract.2785

Think of how the Internet disruption and the cell phone disruption overlapped and complemented each other. They eventually combined to become “mobile Internet.” However, “electric vehicles are the natural platform for autonomous cars,” according to Takeshi Mitamura of Nissan’s Silicon Valley Research Center. While the electric vehicle disrupts the gasoline car industry, the self-driving car will disrupt it as well and deal the final death blow to whatever remains of the gasoline car industry. Two industries will be disrupted. The auto industry will shrink massively and the oil industry will either disappear as a supplier to the automotive market (the EV scenario) or shrink massively (autonomous-use-ICE scenario) and eventually disappear altogether (autonomous-EV scenario) as an automotive energy source. Either way, it doesn’t look good for the oil industry. Even if only a relatively cautious share-to-own ratio of 5-to-1 applies after the autonomous car disruption, the automobile market will still severely shrink to at most 20 to 30 million cars per year. In other words, even assuming that the electric vehicle doesn’t disrupt the gasoline car, the autonomous car disruption will cause demand for gasoline to decrease dramatically – maybe by 80 percent. Business Model Innovation Most people think of market disruption in terms of “disruptive technologies.” Many times, however, the source of disruption is not a new technology per se but an innovative business model made possible by the new technology. Consider how Skype disrupted the long-distance telephone market. Many companies had access to the voice over internet protocol (VOIP) technology. It was Skype with its innovative business model that revolutionized the industry. Car companies have used the same business model for a hundred years. It goes something like this: We make the car, you buy the car, we fix the car until it breaks; repeat every few years. The most radical business model innovation in the auto industry was probably the introduction of car financing by GMAC around 1917. This single innovation helped the car market go from 8 percent of ownership to 80 percent ownership. (See Chapter 2.) In terms of business models, the auto industry has made very little progress over the last century, but new business models made possible by new technologies are starting to change that. After a car becomes more like a mobile computer-on-wheels, the rules of the game will change dramatically. The automobile will be another product category that falls under Moore’s Law. Technology improvements can then happen exponentially. Rather than recall a million cars to repair a defect, auto companies will download new software to the cars over WiFi. It is not a stretch is to imagine that the auto industry as we know it will not exist in a decade or two. Will electric vehicles disrupt the internal combustion engine auto industry? Will a software platform eat Detroit? It seems to me that several disruption waves are coming: Electric vehicles, software-enabled vehicles and, ultimately autonomous vehicles. Moreover, car sharing will radicalize how we use cars. Combine all this with innovative business models and the gasoline car industry is truly kaput. It’s not a matter of if but how and when.2808

“There is no technical demarcation between the military and civilian reactor and there never was one. – Los Alamos National Laboratory Report LA8969MS, UC-16. “If the world should blow itself up, the last audible voice would be that of an expert saying it can’t happen.” – Peter Ustinov. “Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius — and a lot of courage — to move in the opposite direction.” – Albert Einstein. On April 26, 1986, Chernobyl reactor number four blew up. The explosion caused the biggest industrial catastrophe of the 20th Century. Clouds with four hundred times more radioactive material than that produced by the atomic bombing of Hiroshima blew across Europe and Asia.306 Radiation levels were so high, they set off alarms at the Forsmark Nuclear Plant in Sweden, located 1,100 Km (660 miles) away from Chernobyl. Soviet leadership and the world learned of the disaster from scientists who measured radioactive clouds in Sweden.307 On May 7 and then on May 26, France’s “Central Protection Service against Ionizing Radiation” (SCPRI) broadcast its measurements of the radioactive fallout in France. Not to worry, the SCPRI said, the radiation was rather modest, ranging from 500 Becquerels per square meter (Bq/m2) in the eastern part of France to 25 Bq/m2 in the Brittany area, in the country’s northwest.308 But according to Le Monde, those numbers were not quite right: In 2005, a measure from the Institute of Radioprotection and Nuclear Security (IRSN), a successor of SCPRI, that pieced together the fallout from May 1986 showed a far different picture: the deposits of Cesium 137 alone exceeded 20,000 Bq/m2 in certain areas (Alsace, the region around Nice, southern Corsica), with some points in excess of 40,000 Bq/m2.309 The radiation numbers that the French government released in 1986 were made up. In France, the real Chernobyl cloud fallout measurements were about a thousand times larger than what the French government had told its citizens. The real radiation measurement only came to light because the successor organization to the SCPRI was sued in 2001 by the French Association of Thyroid Disease Sufferers (AFMT). This organization accused the government of deliberately falsifying information and failing to take the minimum sanitary measures that neighboring European countries took (for example, banning certain foods). The government of France knowingly hurt millions of its citizens to protect the nuclear power industry. After the Fukushima Dai’ichi nuclear disaster in February 2011, the Japanese government similarly misinformed its citizens by minimizing the extent of the damage.2857

The disruption brought about by the Internet, cell phone, and personal computer gave citizens the power to create, collect, and publish information. These technologies enabled the rise of an inclusive participatory culture.2888

Today’s technology-enabled participatory culture is the opposite of the closed, secretive, hierarchical culture that characterizes the nuclear power industry. One week after the March 11, 2011, Fukushima Dai’ichi nuclear meltdowns, a not-for-profit organization called Safecast published its first website and started a sensor network for collecting and sharing radiation measurements. Safecast volunteers soon started taking radiation measurements in Fukushima. Later, they took measurements throughout Japan and, later still, in the rest of the world. Using an open source microcontroller platform called Arduino and radiation Geiger counters from International Alert, Safecast built small, mobile Geiger counters that cost less than $1,000.310 Safecast called the device “bGeigie,” short for “bento Geiger,” because it looked like a Japanese bento box. To fund its operations and equipment, the organization raised $35,000 through Kickstarter, the crowd-funding website.2891

Safecast now collects more radiation data at a finer scale than the Japanese government itself. Instead of the government’s single Geiger counter per city, Safecast takes nuclear radiation measurements at 50 to 100 meter (150 to 300 feet) resolution every five seconds. It uploads the data every day as zero-license, open-domain content. Anyone can use the data without copyright or financial limitations. On its weather pages, Internet giant Yahoo! Japan has a link that displays radiation information from the Safecast sensor network. Safecast has uploaded more than 10 million data points, a number that is growing exponentially. It has developed a new version of its Geiger kit sells for $450 to Safecaster volunteers around the globe (see Figure 6.1).312 In economics, “regulatory capture” refers to what happens when a state regulatory agency that was created to act in the public interest works instead to advance the commercial or special interests in the industry it is supposed to regulate.313 In other words, the government agency protects the industry at the expense of the public. Regulatory capture encourages companies to pollute, cut health and safety corners, and take financial and technical risks safe in the knowledge that citizens and taxpayers will bear the costs. Open data can be considered political when it shines a light on regulatory capture and secrecy. When asked if Safecast is an anti-nuclear organization, co-founder Sean Bonner answered, “Safecast is not anti-nuclear or pronuclear; we are pro-data. Data is apolitical.”2901

When the nuclear industry tells people about the cost of nuclear power it doesn’t include the costs of decommissioning (cleaning up) nuclear power plants. Decommissioning is a never-ending source of cash from taxpayers to the nuclear industry.2923

In June 2013, soon after the decision to permanently close the San Onofre nuclear power plant in California, its owner and operator, Southern California Edison, started transferring nearly $5 billion dollars in failed repair and decommissioning costs to rate payers. “The traditional way, of course, is all these costs are passed through to the customers, to the rate payers in the form of the rates,” said Ted Craver, chairman and CEO of Edison International.315 After profiting for decades, nuclear plant operators pack up and turn over the costs of cleaning up their mess to taxpayers and rate payers. What if energy regulators in the United Kingdom were looking after the citizens they’re supposed to serve instead of the nuclear industry? How much solar would the US$110 billion cost of cleaning up Sellafield buy the British people?2930

In 2012, electricity demand in the UK was 35.8 GW on average.317 The peak demand was 57.5 GW. With these figures in mind, the cost to the taxpayer of decommissioning a single nuclear site (Sellafield) would be equal to the total cost of installing unsubsidized solar that would generate 190 percent of the average electricity demand and 117 percent of the peak power capacity in the UK. Knowing this, you would think that UK regulators would stop any nuclear development and turn to solar (and wind).2941

Nuclear is already the most expensive method of generating electricity. An unsubsidized peak watt of solar or wind costs less than $2 in Germany and Australia. Why would the British buy into nuclear at more than ten times the capacity cost? The UK is not a sunny place, but Germany, a country with a similar climate to the UK, generates solar power for less than what the British government wants to pay for nuclear generation.2951

The 16 GW nuclear expansion that the UK government proposed is 66 times the size of the generation capacity at Sellafield. How many hundreds of billions of dollars will it cost to reprocess the fuel, clean up, and decommission those reactors in forty years? Furthermore, the $407 billion in nuclear subsidies doesn’t include the taxpayer cost of insuring the industry against a nuclear meltdown. The UK is a small country. A Chernobyl or Fukushima-type disaster would have a catastrophic effect on the whole country, cost several trillion dollars, and take countless lives. So why would a power utility even consider building a nuclear power plant? Three words: government protection and subsidies. Or four words: all gain, no pain.2960

The nuclear industry has been characterized by failure to deliver: cost overruns, delays in construction, and lack of safety. The nuclear industry cannot compete with other methods of energy production. Nuclear reactors today are about ten times more expensive to build than they were in the early 1970s. The costs keep going up. The nuclear industry may be the only major industry in the world with a negative learning curve. By contrast, solar PV has improved its costs by 154 times 1970. Solar has improved its cost position relative to nuclear by 1,540 times since. 19702972

The learning curve says that the more you produce a good or service, the better you become at producing it, so you can make it faster and cheaper.324 Engineers have measured learning curves for many industries. Learning curves help engineers quantify product cost curves as production scales in the foreseeable future. For instance, if the learning curve in shipbuilding is 20 percent and the first ship costs $100, as production doubles, the next batch of ships will cost $80 ($100 * [1-0.2]). As production doubles again, the next batch will cost $64 ($80 * [1-0.2]), and so on. You can usually find the learning curves for different industries in engineering books and manuals. The Federation of American Scientists offers an online calculator where you can plug in the learning curve and calculate future costs.325 Jonathan Koomey of Stanford University has plotted the actual costs of building nuclear power plants in the United States since 1970 and the results are revealing (see Figure 6.2).326 As the industry has gained experience building reactors, nuclear power plants have become more expensive.2978

The time it takes to build nuclear power plants has also lengthened. The industry that promised energy that would be too cheap to meter produces energy that is too expensive to compete. Vermont Law School Professor Mark Cooper has done an in-depth analysis of the nuclear industry in the United States and France. He found, among other things, that the time it takes to build a nuclear power plant has lengthened considerably2989

cost and construction time to build a nuclear reactor have both risen. The nuclear industry is unique among major industries in that it has what climate expert Joe Romm (a Ph.D. in nuclear physics from MIT) has insightfully called a “negative learning curve.” The more experience it gains building its product, the more expensive the product has become and the longer it has taken to build. And the negative learning curve is not a small one. Looking at the data, you can see that, since the 1970s, the industry has increased costs by about ten times and delivery times by about four times.2998

No new nuclear plants broke ground after Vogtle and its peers were commissioned in the 1980s because nuclear power is economically uncompetitive. It survives because the U.S. government gave the energy industry the keys to the taxpayer treasure chest in 2005. In 2005, the U.S. Congress approved $18.5 billion in new loan guarantees for the nuclear industry, according to the Nuclear Energy Institute, an industry lobby group.330 The 2005 Energy Act authorized the U.S. Department of Energy to guarantee up to 80 percent of nuclear project costs. It also provided extra insurance of $2 billion for cost overruns and another $1 billion in production tax credits spread over the first eight years of a reactor’s life.331 After the 2005 Energy Act, the industry loudly predicted a nuclear rebirth. Hoping to emulate the high market penetration of the nuclear industry in France, the nuclear industry borrowed a French word (renaissance) to describe its rebirth and promptly went to work. In 2006, Georgia Power, a Southern Company subsidiary, announced it would build two new 1.1 GW reactors at Vogtle, the Vogtle 3&4. The U.S. Nuclear Regulatory Commission gave its approval and the project broke ground in April 2009. Georgia Power estimated the cost to build the two reactors at $14 billion. They were to start operation in 2016 and 2017 respectively. In 2009, the Georgia Senate approved Senate Bill 31, which allows Georgia Power to collect up to $2 billion from rate payers to finance the new Vogtle nuclear plants.332 Rate payers would finance the nuclear power plants while they were being built. On February 6, 2010, the Obama Administration offered a federal loan guarantee of $8.33 billion for the construction of the reactors.333 Construction of the Vogtle 3 reactor officially began on March 12, 2013 with the pouring of concrete for the nuclear island. The story of Georgia Power and its Vogtle 3&4 nuclear plants illustrates how the nuclear industry uses taxpayer money to finance nuclear power plants. Look how Georgia Power took advantage of its cozy relationship with regulators and policy-makers: Federal loan guarantee: $8.3 billion. Financing from rate payers: $2 billion. Production tax credit: $1 billion. Assuming that the Vogtle 3&4 nuclear plants come in under budget and actually produce energy, $11.3 billion of the $14 billion cost of building the plants will have come directly from taxpayers. But don’t expect these projects to be finished on time or under budget. Remember that the original Vogtle 1&2 reactors were thirteen times over budget. Vogtle 3&4 have barely broken ground and are already two years late and up to two billion dollars over budget. The project is now expected to cost up to $16.5 billion and open in 2018 and 2019.334 Is anyone surprised that a nuclear project is late and over budget? The industry is pathological about over-promising and under-delivering. Every nuclear plant built in the United States has been late and over budget, or else has been canceled.3004

According to the Congressional Budget Office, the 75 nuclear plants built between 1966 and 1986 were three times more expensive than their builders originally estimated.335 Moreover, of the 253 nuclear plants that were originally ordered between 1953 and 2008, 121 (48 percent) were canceled before completion.336 Of the 132 plants that were built, 21 were permanently shut down due to reliability or cost problems, while another 27% completely failed for a year or more at least once, according to Rocky Mountain Institute energy expert Amory Lovins.337 “Many nuclear plants operate profitably now because they were sold to current operators for less than their actual cost.”338 The Washington Public Power Supply System Service (WPPSS), now Northwest Energy, ordered five nuclear power plants in the early 1970s. Delays and cost overruns caused the utility to cancel two of five, halt construction on two mores, and default on $2.25 billion, then the largest municipal bond default in history,339 Only one of the original five nuclear plants, the Columbia Generation Station, is working today.340 A Forbes magazine cover story from February 11, 1985 titled “Nuclear Follies” said, “The failure of the U.S. nuclear power program ranks as the largest managerial disaster in business history, a disaster on a monumental scale … only the biased can now think that the money has been well spent. It is a defeat for the U.S. consumer and for the competitiveness of U.S. industry.” Is Georgia Power shamed or worried about cost overruns? Not at all. The 2005 Energy Act conveniently set aside $2 billion of taxpayer money to account for cost over runs. For Vogtle, the treasure chest of taxpayer subsidies keep adding up: Federal loan guarantee: $8.3 billion. Financing from rate payers: $2 billion. Production tax credit: $1 billion. Cost-overrun protection: $2 billion. Forgive me if all this reminds me of Wall Street’s “Mickey Mouse” loans that brought about the Great Crash of 2008 and pulled the world economy down with it. Nuclear power is an industry that has proven that it can deliver an ever more expensive product and take longer to produce it. Nuclear is already an uncompetitive industry with a negative learning curve. If energy were based on market forces, the nuclear industry would have been out of business long ago. The only way nuclear can stay afloat is to take subsidies from the taxpayer. Sadly, some governments are delivering on those subsidies. The Obama Administration’s 2012 budget called for tripling the nuclear loan guarantee program from $18.5 billion to $54.5 billion.3413029

the most expensive subsidy may be yet to come: nuclear insurance. In the U.S., nuclear insurance is called the Price-Anderson Nuclear Industries Indemnity Act. Whatever the failings of the nuclear industry have been so far, bailing out — in other words, insuring — the nuclear industry is one area where failure could bankrupt not just a company or an industry, but an entire nation. In 2009, there was one accident for every 1.4 million Western-built jet aircraft flights, according to the International Air Transport Association.342 Based on those figures, you have a 0.00007 percent chance of being in an aircraft accident. On the other hand, 1.5 percent of all nuclear reactors ever built have melted down, according to Stanford Professor Marc Jacobson.343 The likelihood of a nuclear plant reactor having a meltdown is almost 1 million times greater than the chances of your next aircraft flight crashing. Now imagine you were told that the flight you and your family are about to board has a 1.5-percent chance of blowing up. Would you get on board? The Fukushima nuclear disaster has reminded us once again that nuclear power is not safe.3054

But the fact that taxpayers insure the nuclear industry is not a Japanese thing. It’s a nuclear regulatory thing. If there is nuclear power plant in your country, then you, too, are in the nuclear insurance business. Do you know how much your liability is?3072

In the United States, Congress has decided that the taxpayer is liable for nuclear disasters. It’s the law of the land. It’s called the Price-Anderson Nuclear Industries Indemnity Act. Congress passed the Price-Anderson Act in 1957 in an effort to protect the nascent civilian nuclear industry. In 1957, the private insurance industry did not have enough data to accurately price insurance premiums for a nuclear power plant. But the nuclear industry has grown and matured since 1957. Today, more than 430 reactors in 31 countries provide 370 GW of nuclear power capacity, according to the World Nuclear Association.345 The nuclear industry has achieved high penetration in countries such as France, Japan, Russia, and the United States. France has 59 reactors that generate about 75 percent of the country’s electricity.346 Before the Fukushima disaster, Japan had 50 reactors that generated about 30 percent of the country’s electricity.347 The U.S. has about 100 reactors that produce about 19 percent of the country’s electricity.3483074

Private insurers have enough data to quantify the safety of nuclear plants. They have enough data to create an insurance product for nuclear power plants, right? Yes, they have enough data but no, they won’t insure nuclear. Not a single insurance company has stepped forward to cover the full costs of a nuclear disaster. Private insurance companies insure buildings such as the new Freedom Tower (which was built after the World Trade Center terrorist attacks). Private insurance companies insure against the risk of hurricanes and airplane accidents. But no private insurance company will insure the full costs of nuclear.3084

What premium would insurance companies charge to insure nuclear power plants? The German government (where taxpayers also insure the nuclear industry) commissioned a study to answer that question. The April 2011 report concluded that, for a private insurance company to insure a nuclear plant, the insurance premium would be in the 0.139 €/kWh (19.9 ¢/kWh) to 2.36 €/kWh ($3.39 ¢/kWh) range.349 To put those figures in context, the city of Palo Alto has a 25-year power purchase agreement in which it pays 6.9 ¢/kWh for solar.350 So the total price that Palo Alto pays for each unit of solar energy (6.9 ¢/kWh) is about one third of the lower estimate of the insurance premium that a nuclear producer would have to pay for each unit of energy (19.9 ¢/kWh). Here’s another way to look at it: The solar independent power producer (IPP) that supplies Palo Alto has to pay capital costs, installation costs, management costs, insurance costs, operation and maintenance costs, taxes, permit costs, and other costs. After all that it has to make a small profit from the energy it sells to Palo Alto for 6.9 ¢/kWh. All those costs put together, plus the developer profits, are about one third of the insurance premium that a nuclear power plant would have to pay for each energy unit (kWh). And that’s for the low-end nuclear insurance premium estimate (19.9 ¢/kWh). If you take the high-end estimate ($3.39 ¢/kWh), the total cost of solar electricity is fifty times lower than the premium to insure nuclear.3089

The German report also concluded that the expected damage value of a nuclear disaster would be €5.756 trillion Euro ($8.27 trillion). Germany’s gross domestic product (GDP) in 2012 was $3.4 trillion, according to the World Bank. The cost of a nuclear disaster in Germany, according to the report, would be about 2.4 times the size of the German economy.351 In order words, a single nuclear disaster could bankrupt the largest economy in Europe and the fifth largest economy in the world. Nuclear is not just prohibitively expensive, it could also bankrupt a whole country. And the smaller the country’s economy, the swifter the country would collapse. Russia’s GDP is now about $2 trillion, but in the late 1980s, it was closer to $500 billion.352 Michael Gorbachev said that the 1986 Chernobyl nuclear disaster “was perhaps the real reason the Soviet Union collapsed five years later.”3533106

The Nuclear Death Spiral Faced with such daunting data, Germans decided after the Fukushima disaster to shut down eight nuclear reactors immediately and shut down their whole nuclear industry by 2022. Meanwhile, they accelerated what already was the most ambitious clean energy program in the world, a program based on solar and wind generation, energy efficiency, and electric vehicles. Most countries in Europe have also accelerated the nuclear phase-out process that they started after the Chernobyl nuclear disaster in 1986. Italy held a referendum in June 2011 in which 95 percent of voters resoundingly rejected their prime minister’s push for a new nuclear energy program.354 Global Data expects 150 of Europe’s 186 nuclear power plants (80 percent) to shut down by 2030.355 All of Japan’s fifty nuclear plants have shut down. Japanese citizens gathered eight million signatures against the government’s plan to restart its nuclear power plants.3117

Four more nuclear power plants were closed in 2013: Vermont Yankee and Wisconsin Kewaunee because they could not compete in wholesale markets; Crystal River in Florida because of structural damage; and San Onofre in California for a number of reasons, including structural damage, failed repairs, and safety concerns.358 The existing fleet of nuclear power plants in the U.S. is getting older, more inefficient, more expensive to operate and maintain, and increasingly less competitive. A report by investment bank Credit Suisse points out that the number of nuclear plant outage days has been increasing significantly (see Figure 6.4). This increase in outage days has increased the costs for repairs and upgrades. The San Onofre nuclear plant, for example, closed in 2011 so its malfunctioning steam generators could be replaced. After spending $670 to make the repairs, the steam generators were deemed unrepairable. San Onofre has not reopened and is scheduled to be decommissioned. Operations and maintenance (O&M) costs have gone up by 4.8 percent annually; fuel costs rose by 9.1 percent per year during the years 2007–2011; the fully loaded costs are expected to continue to rise by 5 percent per year in the foreseeable future, according to a Credit Suisse report.359 An analysis by Mark Cooper of the Vermont Law School lists 38 nuclear plants that are “at risk of closure” for purely economic reasons, of which ten face “particularly intense challenges.”360 Nearly all of these plants are among the 47 nuclear power plants that have to compete daily in the open wholesale markets in the U.S. 3613130

On December 16, 2013, USEC (formerly the Uranium Enrichment Corporation), the only American uranium enrichment company in the business, announced it would file for bankruptcy.365 USEC had received $257 million in aid3155

As solar and wind increase penetration nationally, nuclear is toast. New nuclear plants like Vogtle, assuming they will eventually be built and commissioned, will not be able to compete in an open market. These “new nuclear power plants” are expected to produce electricity at an uncompetitive cost between 25 ¢/kWh and 30 ¢/kWh.367 The average retail price of electricity to residential customers in the U.S. in September 2013 was 12.5 ¢/kWh.368 So the expected cost of “new nuclear,” when and if the new plants ever get built, would be twice the retail price of today’s retail electricity. Add the cost of transmission, distribution and utility overhead, and nuclear could sell for three times today’s retail prices. By contrast, solar costs are going down dramatically. First Solar’s 50 MW Macho Springs project will sell solar to El Paso Electric for 5.79 ¢/kWh.369 That’s about five times less than “new nuclear.” When the Chicago-based utility Exelon announced it was scrapping its nuclear reactor project in Victoria County, Texas, the company said it was all about the economics: “…economic and market conditions make construction of new merchant nuclear power plants uneconomical now and for the foreseeable future,” the company said.370 John Roe, CEO of Exelon, a company with a portfolio at the time that was 93-percent nuclear, dismissed the so-called renaissance altogether. “Don’t kid yourself that [nuclear] is economic. Building out nuclear capacity would require as much as $300 billion in federal loan guarantees and other subsidies.”3713162

Disrupting the Nuclear Zombie Nuclear is a zombie — not quite alive but not dead either. The real danger from zombies is that they want to suck the life out of the living. Nuclear depends on taxpayer subsidies and always has. The nuclear lobby has succeeded in positioning nuclear as a “clean” alternative to fossil fuels and getting even more subsidies than before. You can fool most of the people most of the time, but the lethal realities of Fukushima and the brutal market reality of an uncompetitive nuclear industry are far too difficult to hide in open societies and open energy markets. However, in the course of building a large market in the U.S. and Europe, a critical mass of engineers, academics, and suppliers coalesced around the nuclear industry. As the industry shrinks, these engineers and academics will leave for better jobs elsewhere; new university talent will not seek careers in a dying industry. Suppliers will go out of business or shift to other industries. Having fewer suppliers focused on nuclear will mean even higher costs, longer delays,3184

Solar and wind keep growing their market share, increasing quality, and cutting costs. As solar and wind decrease costs and beat nuclear in the retail and wholesale electricity markets on a daily (and nightly) basis, more nuclear plants will have to shut down for purely economic reasons. I mentioned earlier in this chapter that solar has improved its cost position relative to nuclear by 1,540 times. Solar costs are expected to drop another two thirds by 2020.3198

The end of nuclear will mean the end of a popular deception — that the “civilian nuclear” industry as a viable industry. We will have to pay to clean up the nuclear mistake for generations to come in places like Sellafield, Chernobyl, and Fukushima. But make no mistake, nuclear is already obsolete. The nuclear industry is imploding because it’s too expensive, too dangerous, and too dirty. Let this zombie go before it does more irreversible damage to the living.3215

“When the wind of change blows, some build walls, others build windmills.” – Chinese Proverb. “I’d put my money on the sun and solar energy. What a source of power! I hope we don’t have to wait until oil and coal run out before we tackle that. – Thomas Alva Edison, 1931. “The Stone Age didn’t end for lack of stone, and the oil age will end long before the world runs out of oil.” – Sheik Ahmed Yaki Zamani, former Saudi Arabia Oil Minister.3220

Saudi Arabia had better hurry up. As enlightened as its plan to produce 41GW of solar is, it may not have another twenty years of oil income to fund its solar plans. Solar has cut its costs exponentially relative to oil; the solar cost advantage will only improve over the next few years.3240

In reality, solar PV is 154 times cheaper now than it was in 1970 (costs went from $100/W down to 65 ¢/W), while oil is 35 times more expensive (oil went from $3.18/barrel to $110/barrel) (see Figure 7.1). Put these numbers together and you find that solar has improved its cost basis by 5,355 times relative to oil since 1970. If you think that an industry can compete with a technology that has improved its cost position relative to yours by more than five thousand times, clearly you’re in denial about the impending disruption.3247

In an industry that is about to be disrupted, a company has three choices: 1. Get out. That is, sell at high prices while you can. 2. Invest in the disrupting industry. 3. Die.3253

oil companies are mostly lifting oil at $20 and selling it at $100. That’s five times production costs. Not a bad gig if you can get it. When oil prices return to, say, 1990s levels in the $20–$30 per barrel range, the following will happen: 1. Only low-cost producers will survive. Only highly productive fields with breakeven points below $15– $20 will be developed. 2. The majority of environmentally catastrophic developments (which happen to be massively expensive) will be stranded. Canadian Oil Sands with its $65/bbl costs, offshore developments with their $60–$70 breakeven points, and deep Arctic developments with as yet unknown financial and environmental costs will be put on hold. Rational investors will never come back to these projects. Ever. Prime Minister Stephen Harper of Canada3300

India is not alone. “One of the energy industry’s best-kept dirty little secrets is the $300 to $400 billion per year world government subsidy of diesel fuel,” according to André-Jacques Auberton-Hervé, CEO of the electronic semiconductor manufacturer Soitec.390 Government officials usually justify energy subsidies by claiming that they help the poor, but the evidence shows that subsidies help the rich, not the poor. According to the IMF, the richest 20 percent of households in low- and middle-income countries net six times more in total fuel product subsidies than the poorest 20 percent of households.391 One billion Indians have access to a cell phone, but only 366 million have access to a toilet, according to the United Nations.392 The country’s leadership has to admit that it is not very good at providing infrastructure to its people even when doing so is rather simple. The toilet-vs-cell phone numbers prove that setting up a bit-based distributed infrastructure is easier than setting up an atom-based infrastructure consisting of water and sewerage pipelines, centralized processing plants, and command-and-control management.3344

Assume that each Indian citizen gets about 100 solar Watts. A family of three would get 300 Watts, or about one solar panel; a family of five would get 500 Watts. India gets on average about five hours of sunshine every day. This means each family of five would get about 2.5 kWh per day, enough to charge a couple of cell phones and to run a computer, a television set, several LED light bulbs, a table fan, and a coffee pot.393 The 2.5 kWh per day figure is more than most grid-connected Indians consume. In 2005, 45 percent of Indians had no access to the grid at all; 33 percent had access to the grid but consumed less than 50 kWh per month (1.6 kWh per day); and 11 percent had access to the grid but consumed only 50 kWh to 100 kWh per month (1.6 kWh to 3.3 kWh per day). Eleven percent of Indians had access to the grid and consumed more than 100kWh per month (3.3 kWh per day).394 The cost of solar PV today is about $0.65 per Watt. That’s just for the panel. Adding the costs of inverters, cables, installation, and so on, to the cost is about $2/Watt, which is the total installed cost per Watt for residential solar in Germany. Add a small battery for nighttime usage and you have a solar home system for about $3/W. To provide solar electricity to 100 million people, India would need about $30 billion, less than what the government spends to subsidize diesel and other fossil fuels today. In other words, if the Indian government diverted the amount it spends to subsidize fossil fuels to home-based distributed solar generation, the country could bring electricity to 500 million people in just five years. The diesel and kerosene industries would be totally disrupted. There would be no need for new coal plants, hydroelectric plants, or transmission lines. Solarizing the bottom 500 million in India would not only cost less than subsidizing fossil fuels, it would end the unnecessary carnage caused by indoor air pollution. According to the World Health Organization, three to four-hundred thousand people in India die of indoor air pollution and carbon monoxide poisoning every year because of biomass burning and the use of chullahs.395 For the 1,411 people in tiny Tokelau or 500 million people in India, energy poverty can be ended quickly. Solar is already cheaper than diesel, kerosene, and many times more valuable than firewood. When governments wake up and stop subsidizing fossil fuels, the disruption of petroleum as an energy source will be inexpensive and swift.3356

What the sunlight-to-power efficiency is depends on the technology used. At the low end of the market, thin film photovoltaic converts around 12 percent of solar energy to power, polycrystalline PV converts around 16 percent, and monocrystalline converts close to 20 percent. Concentrating solar power (CSP) technologies may double these numbers. Concentrating photovoltaics (CPV) has set records near 36 percent, while thermal CSP with combined heat and power (CHP) can reach an efficiency of 75 to 80 percent.4003406

About a thousand square miles of solar are needed to power every single electric car-mile driven every year in America. A solar plant the size of King Ranch in Texas with its 1,289 square miles could generate all of America’s electric vehicle power with 40-percent extra electricity to spare.3416

about 143,000 square miles from the U.S. government — to meet just a third of America’s transportation needs. Multiply 143,000 square miles by three and you get the total number of square miles the oil and gas industry needs to power every single vehicle-mile in America: about 400,000 square miles. To power just one third of the gasoline car-miles in America, oil uses 143 times the surface area that solar would need to power all electric car-miles. Here’s another way to look at it: The combination of solar and electric vehicles used land 400 times more efficiently for energy production than the combination of oil production and gasoline vehicle usage.3423

When a technological convergence (solar and EVs) is 400 times more resource-efficient than incumbent technologies (oil and internal combustion engine cars in this case), it’s time to pay attention. The solar and EV technology conversion is bound to be disruptive, especially when you remember that, in this case, the resources are as valuable as land and water. Of course, building a solar power plant that is 874 square miles is not feasible. Nor is it productive. The disruptive potential of solar technology lies not just in its low cost but in its distributed nature. It’s better to generate most of that power close to the cars themselves from residential and commercial rooftops, malls, big-box stores, parking lots, landfills, and so on. Wal-Mart is expected to cover 218 square miles in 2015.402 Wal-Mart alone could power one fourth of all electric car-miles in the United States. All Wal-Mart would need to do is cover its rooftops with solar panels and its parking lots with solar canopies. Leaks, Spills, and Contamination Needless to say, oil drilling leaks and spills damage more land and water than the numbers reveal. The 2010 BP Gulf Oil disaster damaged tens of thousands of square miles beyond its oil wells. As of June 2010, the U.S. National Oceanic and Atmospheric Administration (NOAA) Fisheries Services closed the 80,000 square miles around the wells to commercial fishing. The BP Oil spill alone was eighty (80) times larger than the surface area that solar plants would need to power every electric vehicle car-mile in America. Meantime, no one has ever heard of a solar spill. Oil is not just dirty and expensive. Oil is a land and water hog. The convergence of electric vehicles and solar would be 400 times more land-efficient than oil and wouldn’t threaten the kind of pollution that oil is famous for. Summary: The End of Oil Oil is obsolete. The oil age will end by 2030. Electric vehicles, autonomous cars and solar energy will disrupt the oil industry. Most of the multi-trillion dollar investments in the oil industry will soon be stranded.3428

Wal-Mart is expected to cover 218 square miles in 2015.402 Wal-Mart alone could power one fourth of all electric car-miles in the United States. All Wal-Mart would need to do is cover its rooftops with solar panels and its parking lots with solar canopies.3434

The 2010 BP Gulf Oil disaster damaged tens of thousands of square miles beyond its oil wells. As of June 2010, the U.S. National Oceanic and Atmospheric Administration (NOAA) Fisheries Services closed the 80,000 square miles around the wells to commercial fishing. The BP Oil spill alone was eighty (80) times larger than the surface area that solar plants would need to power every electric vehicle car-mile in America. Meantime, no one has ever heard of a solar spill. Oil is not just dirty and expensive. Oil is a land and water hog. The convergence of electric vehicles and solar would be 400 times more land-efficient than oil and wouldn’t threaten the kind of pollution that oil is famous for.3438

Electricity storage is going down so fast that, by 2020, solar PV with grid storage will be cheaper than petroleum anywhere at any scale. The disruption of petroleum is already fast underway. The electric vehicle will disrupt the oil industry by 2030 — maybe before. The increase in oil demand over the next few years from energy-intensive growing economies like China and India will mask the fact that oil is on its way to quick obsolescence. Solar will disrupt oil as a power generation source and as an automotive power source.3453

“In a time of universal deceit, telling the truth is revolutionary.” – George Orwell.3464

About a century before the San Bruno explosion, in 1906, an earthquake and subsequent fire destroyed the then most important cultural, financial, and trading center on the west coast of the United States: San Francisco. About 25,000 buildings were destroyed, three thousand people died, and as many as 300,000 people (out of a total population of about 400,000) were left homeless.405 While the magnitude 7.9 quake was devastating, it is estimated that 90 percent of the destruction in San Francisco was due to fires caused by ruptured gas mains.3473

The first-ever reported survey of pipeline leaks in the U.S. found 3,356 leaks along 785 miles of Boston roads.414 That translates to 4.3 leaks per mile. Extrapolating this ratio, the United States may have 1.3 million natural gas leaks across 305,000 miles of natural gas pipelines. The same research team that investigated Boston, led by Prof. Robert Jackson of Duke University, recently found 5,893 leaks in the Washington, DC natural gas pipeline system.415 Some gas manholes had methane concentrations as high as 500 thousand parts per million — ten times higher than the threshold at which explosions occur. “The average density of leaks we mapped in the two cities were comparable, but the average methane concentrations are higher in Washington,” said Nathan G. Phillips, a research team member and professor at Boston University’s Department of Earth and Environment. Some of the leaks were comparable to the amount of natural gas used by two to seven homes; a dozen of the leaks pose a risk of explosion. Methane, a greenhouse gas, is 72 times more potent (over twenty years) than CO2.416 Even a one-percent leakage rate would negate the fact that methane emits 50-percent less CO2 than coal when burned in an efficient power plant. Even if you take the longer view and consider that methane is just 25 times worse than coal over a hundred years, a 3-percent leakage rate would negate its warming benefits versus coal.3503

gas is cheap mostly at the wholesale level. Distributing gas is expensive, even within a single country. In the U.S., the cost of exporting gas can be higher than the cost of extraction; exporting costs could wipe out any advantages from low domestic costs.3564

the oil and gas industry in the U.S. is exempt from many environmental laws, including: Clean Air Act Clean Water Act Safe Drinking Water Act National Environmental Policy Act Resource Conservation and Recovery Act Emergency Planning and Community Right-to-Know Act Comprehensive Environmental Response, Compensation, and Liability Act (Superfund)437 Essentially, when it comes pollution caused by extraction, the oil and gas industry is either above the law or writing the laws. The oil and gas industry has the right to pollute water, land, and air almost at will. When regulatory capture happens to the extent it has in the United States, disasters like the British Petroleum Gulf Oil spill are inevitable. We have forgotten the words of Abraham Lincoln’s Gettysburg Address: “The government of the people, by the people, for the people, shall not perish from the Earth.” Instead we have “the government of the energy industry, for the energy industry, by the energy industry, shall have the right to destroy the earth.” These pollution costs are very real but are paid by the taxpayers, not by the industry that causes pollution. The industry doesn’t have to disclose the names any of the hundreds of toxic chemicals that it pumps into the ground (and water) every time it “fracks” a well. We now know they pump radioactive radium out of the ground. I wonder if the industry pumps radioactive uranium or plutonium into those well holes. American citizens, even those who live in the areas being fracked, don’t have the right to know.3618

To date fracking has been performed more than one million times in the United States. In 2009, there were 493,000 active natural gas wells in the U.S.439 It is estimated that, in 90 percent of these wells, fracking was used to get more gas flowing. In the state of Pennsylvania alone there are 150,000 abandoned oil and gas wells.440 Imagine how much water would be conserved if solar PV (or wind) generated the power that is now generated as a result of fracking. If we used solar PV (or wind) to generate the daily energy needs in the U.S., solar (or wind) would need about 11,000 m3. That’s about 2.9 million gallons of water. To power the entire country with solar and wind would require the same amount of water as fracking a single natural gas well. Solar is literally a million times more water-efficient than gas. In terms of water usage, gas simply cannot compete with solar or wind. When you take into account the pollution that fracking generates when the “fracked” water is dumped into streams, solar and wind look even better (see Figure 8.8). It is illegal to discharge this polluted water, but in the social media era in which we live it has been difficult for the industry to hide a practice that is embedded in its DNA. A search for “fracking illegal dumping of wastewater” gets more than 50,000 results on Google. Results include Bloomberg’s “Exxon Charged with Illegally Dumping Waste in Pennsylvania”441 and CBS’s “Oil Company Caught Illegally Dumping Fracking Discharge in Central Valley.”442 The latter news story showed a video recorded by a farmer in Kern County, California. It showed an unconcerned drilling crew dumping illegal fracked wastewater in a river. Figure 8.8—Shale gas production technique and possible environmental hazards.3634

The oil and gas industry promises a “golden age,” but the “golden age” requires massive, multi-trillion dollar investments. It requires untold amounts of sand, water, and who-knows-what chemicals. It requires society to bear frightening and burdensome environmental costs. And the end result is at best 1.37-percent annual growth. Would any business executive invest $15 trillion of her company’s money to achieve one percent growth? A look at the International Energy Agency report explains why the answer may be “yes” if the executive worked in the oil and gas industry. The report shows that the fossil fuel industry received annual subsidies of $523 billion in 2011.447 Over the period 2010–2035, this would amount to $13.5 trillion, about 90 percent of the $15 trillion that the oil and gas industry will need to invest in that period. The oil executive’s spreadsheets work after all: Taxpayers finance 90 percent of the capital investment for drilling oil and gas; the drilling takes place mainly on public lands using publicly owned water; the drillers are exempt from any damage to the air, water, and soil. Taxpayers bear the risks and costs and the industry takes trillions in profits. This formula generates unbelievable returns for the oil and gas industry. Natural gas, the new “magic fossil fuel,” is a bridge to nowhere. It’s a destructive, resource-inefficient, financially unviable source of energy. Government protection, exemptions from the rule of law, plus trillions of dollars in taxpayer subsidies keep natural gas flowing.3660

Squandering Water Resources with Biofuels Water is energy and energy is water. Energy is needed to pump, clean, and transport water. Water is used to mine, wash, and generate energy. With agricultural biofuels, water is used to “grow” energy. The entire thermal energy industry relies on massive amounts of water. About 15 percent of the world’s freshwater withdrawal goes for energy, according to the World Bank.452 In a world where nearly half the population will live in areas of “high water stress,” demands on the fresh water supply by the thermal energy industry will be especially taxing. It takes 13,676 gallons of water to produce a single gallon of biodiesel from soybeans, according to WaterFootprint.453 It would take 820,560 gallons of water to produce enough soybean biodiesel to fill up a 60-gallon biodiesel bus. To put this in context, an Olympic-size swimming pool holds approximately 660,000 gallons of water.454 To fill its tank with soybean biofuel, a “clean and renewable biodiesel” bus like the one in Figure 9.1 would use enough water to fill an Olympic swimming pool. Figure 9.1—“This bus runs on biodiesel fuel,” but the fuel isn’t clean or renewable. (Photo: Tony Seba) How many gallons of water are needed to fill up an SUV with corn ethanol? I put this question to Prof. David Pimentel of Cornell University. He has studied biofuels for more than three decades and a foremost authority on water and biofuels.455 “It takes 1,700 gallons of water to produce one gallon of ethanol,” Prof. Pimentel answered. “Assuming a 30-gallon gasoline tank in a SUV, we calculate 51,000 gallons of water for the corn ethanol fill up.” According to the United States Geological Survey, the per-capita residential consumption of water in the U.S. in 2005 was about 99 gallons per day, down slightly from 101 gallons per day in 1995. Daily water usage ranged from 51 gallons per day in Maine to 189 gallons per day in Nevada.456 On the basis of these numbers, each time someone fills up an SUV with corn ethanol, he or she uses as much water as the average American residential user consumes in 16 months! All forms of energy generation use water in the production process. However, the amount of water that is used varies by orders of magnitude. IBM’s “Carbon Disclosure Project” gives an indication of the differences (see Table 9.1). To generate 1 MWh of energy (about what an American home uses each month): Solar PV and wind use negligible amounts of water: 0.1 liters or less than half a cup. Natural gas uses ten thousand times more water than solar to generate the same 1 MWh of energy. Nuclear and coal use about twice as much as gas. Oil uses twice what coal uses (and forty thousand times the water solar uses). Hydroelectric power uses 680,000 times more water than solar. Biofuels use an astonishing 1.78 million times more water than solar to generate the same amount of energy.   Total Water consumed per megawatt-hour (m3/MWh) Water required for US daily energy production (millions of m3) Solar 0.0001 0.011 Wind 0.0001 0.011 Gas 1.0000 11.000 Coal 2.0000 22.000 Nuclear 2.5000 27.500 Oil 4.0000 44.000 Hydropower 68.0000 748.000 Biofuel (1st-gen) 178.0000 1198.000 Table 9.1—Total water consumed per MWh for different sources of energy.3694

It would take 1.2 billion cubic meters (m3) of water for biofuels to produce all the daily energy needs in the U.S. China and India together consume about 2.4 billion m3 of water per year.458 That is, all the water for drinking, agricultural irrigation, power plants, and factories that 2.5 billion people in China and India need per year is roughly equal to the water biofuels require to produce two days’ worth of energy in the U.S.3734

The whole world consumes about 9 billion m3 of water per year.459 To produce about one week of America’s energy needs, biofuels would suck up all the freshwater that our whole planet consumes in a year. It would take just one week of American biofuel energy production to turn earth into a planetary Sahara desert. Compare that with solar PV or wind. To generate the daily energy needs in the U.S., solar (or wind) would need about 11,000 m3. That’s about 2.9 million gallons — less than five swimming pools. Fracking a single well takes more water than that. It takes up to 4 million gallons of water to drill and fracture a single natural gas well using hydraulic fracturing methods.460 It just doesn’t make any sense to use a resource as valuable as water the way most of our energy sources do. In an era of record temperatures, a rising population, and growing water needs for drinking and food production, biofuels cannot be part of a viable energy strategy. Agricultural biofuels are a true environmental catastrophe in the making.3738

Ogallala Aquifer.461 It also provides drinking water for 82 percent of the people who live within its boundaries.3755

Altogether, the Ogallala is probably being depleted at a rate of about 9.7 million acre-feet (12 km3 or 420,000 million cubic feet) annually, which is equivalent to eighteen Colorado Rivers pouring out to the sea every year.4623758

The Ogallala Aquifer may dry up in our lifetimes. Some estimates give it twenty-five years.463 By that time, the agricultural biofuels industry may have succeeded in turning the breadbasket of the world into a massive desert.3761

of this exercise, we will examine jatropha. According to WaterFootprint, 19,924 gallons of water are needed to produce one gallon of biodiesel from jatropha. Based on that number, 43.8 million gallons (166 million liters) of water were needed to produce enough biofuel to power half of the Gulfstream G450’s flight from New Jersey to Paris. To put that number in perspective, producing 2,201 gallons of biodiesel from jatropha requires an amount of water equal to the daily water consumption of 442,956 Americans. In other words, producing enough biofuel from jatropha for a transatlantic flight requires the amount of water needed daily by the people of Atlanta (420,003).465 A flight from New Jersey to Paris in a small Cessna airplane that carries 14 to 19 people half-filled with biofuels requires an amount of water equivalent to the daily residential water consumption of a city the size of Atlanta. This is an insane amount of water. I showed these figures to Prof. Arjen Y. Hoekstra of the University of Twente (Netherlands). The professor is the creator of Water Footprint, and one of the world’s foremost authorities on water management. “Your numbers are correct,” he wrote in an email. “They show that we need to carefully look at how we use limited resources like freshwater. Those who want to generate sustainable energy should look at solar and wind. If you want to do biofuels then you should start with bio waste, not agriculture.”3769

In the end, the solar-to-ethanol conversion ratio for sugarcane is just 0.13 percent, according to Scientific American.4763826

Compare that with the average solar panel’s conversion rate of 16 percent. The average solar panel is 123 times more efficient in converting sunshine into usable energy. Furthermore, solar panels don’t need fertilizers, water, pesticides, or extra energy to convert sunshine into electricity. Concentrating photovoltaic (CPV) can turn about 40 percent of the sunshine into electricity, which makes it more than three hundred times more efficient than sugarcane ethanol biofuel. Other solar technologies, such as concentrating solar with combined heat and power (CHP), can convert more than 72 percent of the sunshine into usable energy.478 Solar CHP can convert sunshine 550 times more efficiently than sugarcane can. Solar is anywhere from 123 to 550 times more efficient than biofuels. Take any land in the world and PV technology will be at least a hundred times more efficient in turning sunshine into energy. PV will do it without using valuable land or water, let alone fertilizers or toxic pesticides that contaminate and create persistent runoffs.3829

The oil lobby will probably battle biofuels on environmental grounds. They will likely say that, when you take into account the whole production lifecycle (including fertilizers, energy, transportation, manufacturing, and so on): Oil generates less CO2 and other greenhouse gases than biofuels. Oil uses orders of magnitude less water than biofuels. Oil doesn’t use prime agricultural land for energy production. Oil doesn’t need as many taxpayer subsidies. Two obsolete forms of liquid transportation energy that lost in the marketplace will fight each other in the alternative universe of politics. It will be like watching the jukebox and the 8-track tape industry lobby the government for a piece of the taxpayer pocketbook. The oil lobby will take the high environmental ground against agricultural biofuels. May we live in interesting times!3861

“It is not the lack of inventive ideas that set the boundaries for economic development but rather powerful social and economic interests promoting the technological status quo.” – Joseph Schumpeter. “I am become Death, the destroyer of worlds.” – Robert Oppenheimer. “Countries choose to live in the energy dark ages not because the sun fails to shine but because they refuse to see it.” – Tony Seba (paraphrasing Michener).3869

lending standards will be more stringent. As a consequence, the cost of capital for coal power will go up. Finance 101 tells you that, as the cost of capital goes up, two things will happen in the coal industry: Fewer coal plants will be built because they will not be financially viable. Many projects whose net present value (NPV) was positive due to artificially low capital costs will have a negative NPV and will be turned down. Coal projects that do get built will produce more expensive electricity. The electricity will be more expensive because the plants will have to pay a higher interest rate on their bank loans.3893

The coal power plant building boom started in the mid-1950s and peaked in the early 1980s, according to the U.S. Energy Information Agency (see Figure 10.1). In the 1970s, utilities started turning to the promised land of nuclear plants, which had a peak-commissioning era in the 1980s3923

Most coal-fired plants in the U.S. are near or past retirement age. A typical coal-fired power plant lasts forty years. More than 540 GW (51 percent) of total generating capacity in the U.S. is older than thirty years, according to the Energy Information Agency (see Figure 10.3). More than 74 percent of all coal-fired capacity was older than thirty years. All this generation will have to be replaced over the next ten to twenty years. In fact, much of this generation is way past retirement age (about forty years) and is on maintenance life support.3957

China is the world’s second largest subsidizer of energy on a post-tax basis, with annual subsidies amounting to $279 billion. The United States is the world’s top subsidizer at $502 billion, according to an International Monetary Fund report.504 Domestic fuel subsidies in India reached 2 percent of GDP in 2011–2012. But you won’t find these subsidies in the government’s budget. “Fuel subsidies have been financed through a number of channels, including off-budget sources,” according to the IMF report. As long as the governments of China and India still support coal and provide regulatory protection and financial support to the coal industry, coal will prosper in the two most populous countries in the world. Governments apart from China and India have multinational financial organizations with their own agendas. Japan’s Bank of International Cooperation has provided more than $10 billion to fund coal projects.505 These countries are certainly not funding coal on the basis of credible business plans. Fossil-fuel costs are extremely volatile in the short and medium terms and historically go up in in the long term. Oil prices rose by more than 14 times in less than a decade (between 1999 and 2008). Natural gas prices are similarly volatile. Coal is no exception to this “expensive and volatile” rule.3994

Coal prices have gone up by 5.8 times since 1970. In the meantime, the cost of solar PV has dropped by 154 times. Solar has improved its cost position relative to coal by nearly nine hundred times since 1970. Coal prices have gone up despite production gains and increased market share since the 1970s. The decreased prices in the period 1980–2000 (see Figure 10.4) correspond roughly to productivity gains and massive layoffs in the industry. The coal mining industry increased its production by 21 percent from 1980 to 2000 while letting go of 69 percent of its workforce. The number of coal mine employees fell from 228,569 in 1980 to 71,522 in 2000 (see Figure 10.5).507 Since 2000, coal prices have regained the upward trend as productivity gains have hit a plateau and even reversed course.4010

science of combining central planning and market-based entrepreneurship. If it can be manufactured, China will manufacture it. If it can be built, China will build it. There’s seemingly nothing that it can’t do. Or is there? The one missing element in this equation is one of the most precious elements on earth: water. Cheap and abundant water has been as important to building civilizations as cheap and abundant energy. We need water to generate energy and we need energy to pump, transport, clean, and process water. Both water and energy are needed to grow food. Lack of water is a limiting factor for fossil fuels and nuclear energy. About 15 percent of the world’s freshwater withdrawal goes for energy, according to the World Bank.510 Nearly half the world’s population will live in areas of “high water stress” affecting energy and food insecurity. China is already in a water crisis. China has 20 percent of the world’s population but only 7 percent of its freshwater. The fast growth of its industry and population has caused the country to draw unsustainably on its rivers and aquifers. Since the 1950s, China has lost 27,000 of its 50,000 rivers.511 The numbers that speak to China’s water crisis are telling: 400 out of 600 cities, including 30 of the largest 32 cities in China, face water shortages to varying degrees. Ninety percent of city groundwater sources are contaminated; 70 percent of rivers and lakes are polluted. Three-hundred million people in China don’t have access to safe drinking water.512 Despite the current crisis, total water use is expected to increase from 599 billion m3 (158 trillion gallons) per year in 2010 to 670 billion m3 (177 trillion gallons) per year in 2020.513 This water will not feed the country’s increasingly wealthy population. Agriculture is expected to decrease water withdrawals from 62 percent of total national freshwater in 2010 to 54 percent in 2020. Where is this valuable freshwater going? To quench the coal industry’s insatiable thirst. China, the world’s largest consumer of coal, is already feeling the stress caused by coal’s unquenchable thirst for freshwater. The coal industry drinks 138 billion m3 (36.5 trillion gallons) per year, or 23 percent of the nation’s freshwater. This number is expected to grow to 188 billion m3 (49.7 trillion gallons) by 2020, raising the coal industry’s consumption of China’s water to 28 percent.514 Here’s another way to look at how the government in China protects the coal industry. From 2010 to 2020, China will increase its freshwater consumption by 71.9 billion m3 (19 trillion gallons). Of that increase, 49.9 billion m3 (13.2 trillion gallons) will go to the coal sector. That is, 69 percent of the increase in valuable freshwater in China will go to the coal industry. In China, coal truly is king, and like any king worth its salt, it doesn’t have to pay for its drinking habit, clean after itself, or pay any of its expenses, however extravagant they might be. The Chinese government support for coal is all the more vexing because coal is found in the most water-stressed parts of the country: the north and northwest. Sixty percent of new proposed coal plants in China are concentrated in just six provinces (Inner Mongolia, Shaanxi, Gansu, Ningxia, Shanxi, and Hebei) that together account for just 5 percent of the freshwater in China, according to the World Resources Institute.515 Within those six provinces, 60 percent of coal plants are further concentrated in areas of “high or extremely high water stress”4031

China has taken 8.5 million hectares (21 million acres) of farmland out of production since 1998.519 The country also suffers from desertification on a grand scale. According to China’s State Forestry Administration, 27 percent of the country (2.6 million square km, or 1 million square miles) now suffers from desertification.520 Soil erosion impacts the lives of four-hundred million people and causes economic losses of $10 billion in China, according to the United Nations Convention to Combat Desertification (see Figure 10.7).521 Figure 10.7—Soil degradation in China. (Source: United Nations Convention to Combat Desertification) 522 China is literally sucking country the country dry to feed its coal industry. Wen Jiabao, China’s premier from 2003 to 2013, said that water shortages threaten “the very survival of the nation.”523 Can China’s government afford to keep sacrificing its people for the benefit of its coal industry? Death by Coal On October 22, 2013, the smog in northeastern China was so heavy the government closed roads, schools, and a major airport.524 Visibility was as less than sixty feet (20 meters) in Harbin, a city of more than ten million people. All expressways in Heilongjiang province were closed.4077

Even short-term exposure to PM2.5 at elevated concentrations can cause heart disease.4096

Air pollution in China is a human catastrophe. Outdoor air pollution caused 1.2 million premature deaths in China, according to The Lancet, a leading medical journal.527 Within China, the differences in life expectancy with respect to air pollution are stark. Life expectancy in northern China was 5.5 years less than southern China due to heart and lung disease caused by coal burning, according to a recent study published by the Proceedings of the National Academy of Sciences (PNAS).528 Much of the conversation about coal has to do with the damaging long-term effects of climate change, but coal already is one of the chief causes of death and disease around the world. Coal is not cheap. We are paying for it with extra trips to the hospital, loss of lives, lower economic output, and loss of quality of life. The coal industry gets the profits and people pay for the costs. India has a growing coal-induced human catastrophe of its own. Outdoor air pollution caused 600,000 deaths in India in 2010.4100

If the additional 519 GW of coal power capacity is brought online in India, the country will quadruple its coal capacity infrastructure using lignite, the lowest-quality coal. Lignite generates ten times the particulate matter of anthracite coal. The human death toll in India because of coal pollution could be an order of magnitude higher than it is today, as many as six million deaths per year. That the Indian government knowingly supports this scheme is beyond rational comprehension. Moreover, the water needs of the coal industry would devastate a country already ravaged by water mismanagement and desertification. Indian reports on soil degradation have increased by a factor of six, according to the United Nations Convention to Combat Desertification.533 Despite the millions of deaths caused by coal air pollution, government agencies that propose increases in coal production do so because, they say, “coal is cheap.” Is it really?4113

from 2001 to 2013, 5,281 American soldiers died in combat.535 During those thirteen years, more than 312,000 Americans died because of coal pollution. Coal costs the U.S. $500 billion per year in health, economic, and environmental damage, according to a Harvard University report.536 That is, each man, woman, and child in America pays more than $1,600 per year for damages caused by the mining, transporting, burning, and disposing of coal. That’s a massive tax. If the coal industry paid for the external damages it causes, it would have to pay us 26.89 ¢/kWh.537 In other words, U.S. taxpayers are subsidizing the coal industry to the tune of 26.89 ¢/kWh. The coal industry would not exist if there were a free market for energy and this industry was not protected by the government. In a free market, companies would not be able to unload the costs of pollution onto the people with the blessing and protection of their government. When is the destruction of human life going to catch up with coal? When are government regulators going to stop aiding and abetting the coal industry in the killing of millions of human beings?4123

misinformation to protect the coal industry. However, when unsubsidized solar becomes cheaper than subsidized coal, it will be hard for politicians and regulators to weave a convincing narrative about the benefits of coal. It will be hard for citizens to give their lives and turn over their wallets to support the coal industry when a cheaper, cleaner alternative exists.4144

on any increase in costs to the rate payers. Utilities love capital-intensive power plants like coal and nuclear with rising fuel costs because these power plants increase the utilities’ income stream every year. Solar is disruptive to that business model. Solar prices go down, not up. Solar can be installed on the customer’s rooftop without any help from the utility. Solar needs no fuel to be mined, processed, transported, burned, or dumped back into the ground.4148

the second quarter of 2013, Cloud Peak Energy made more money playing with financial derivatives than selling coal.538 Energy flow is cash flow. Energy volatility is also cash flow. Coal companies will make money coming and going. But they can’t guarantee you the same price for twenty years like solar and wind can.4153