The Ndrreksr Enge wind farm in Denmark is replacing 77 old-style turbines with 13 large new ones. At 2.3 MW peak, these few modern turbines can produce twice the power of the old ones. Since winds are steadier higher off the ground, the average power over the year will be four times larger. Germany is planning to add 25,000 MW by 2020 by repowering their wind farms with the new turbines.13 Why is it worth the trouble? Not only is the first generation of turbines getting old, but wind is stronger and steadier at higher altitudes. Doubling the height of the turbine will increase the wind velocity by 10%. Since the power varies as the cube of the velocity, this 10% translates of a 34% increase in wind power. The trend is to build fewer very tall towers with very long blades.

These new turbines are huge. The largest so far is Enercon’s E-126, shown in Fig. 3.7. Its rotor diameter is 126 m (413 feet) and its total height is 198 m (650 feet)! This is like the length of two football fields stretching up into the sky. Compared to this, the height of the Statue of Liberty is only 93 m (305 feet); of the Washington Monument, 169 m (554 feet); and of the Eiffel Tower, 324 m (1,063 feet). Those who have been up the Eiffel Tower can testify to the winds up there! Unlike these other structures, the turbine cannot be built step-by-step from the ground up. The blades and the nacelle (the housing holding the blades and the generator) have to be preas­sembled and lifted up by cranes. The cranes themselves are so tall that they have to be assembled with smaller cranes. It is a very dangerous operation: the slightest wind can cause everything to come crashing down.

Each blade is 200 feet (60 m) long, and the blades catch so much wind that they have to turn at only five revolutions per minute, or once every 12 seconds. No birds

Fig. 3.7 The Enercon

E-126 turbine and cranes

(http://www. metaefficient.

com/news/new-record-

worlds-largest-wind-turbine-

7-megawatts. html,

February 2008) would be so slow as to be struck. This turbine is rated at 6 MW but is expected to produce more than 7 MW peak power. Calculations like the ones we did above show that a single E-126 turbine can power 5,000 European households or 1,776 American households. Of course, the power cannot go directly to houses because wind power varies. The power is fed into the electrical grid as a small fraction of the power there, and the wind replaces only some of the fossil fuel or nuclear energy that the grid has to supply.

Figure 3.8 shows the interior of the nacelle of a smaller turbine made by Germany’s Siemans. Inside the nacelle, there are motors and controls that change the pitch of the blades as the wind varies, generators that convert the rotating motion into electricity, and a gearbox that connects the rotor to the generator. These nacelles can be the size of a conference room. Figure 3.9 shows an offshore array of 5-MW turbines made by REpower of Germany. A closeup of the nacelle can be seen in Fig. 3.10. Since the turbine is not easily accessible, the nacelle is made to accommodate workers lowered to the platform from a helicopter.

Before we talk about energy, let us be sure we know what it is. If you turn on a 100-W light bulb, it will use up 100 W of energy, right? Not exactly! Watts measure the rate at which energy is used, which is called power. Energy is something we can store, and power is how fast we use it up. A toaster takes about 1,000 W, or 1 kW, of electricity to run. If we turn it on for an hour, it will consume 1 kWh of energy. A 200-W light bulb left on for 10 h would use up 2,000 Wh, or 2 kWh of energy. On a more personal note, suppose you ate a 200-calorie hamburger (a small one). That’s energy which you store. Suppose it takes you 2 h of exercise to burn off that energy, then you are using up 100 C/h, which is the average power you put out during the workout. What confuses most people is that the well-known electrical unit, the watt, is a unit of power, not energy. You have to multiply by time to get energy.

To compound the confusion, articles about the energy crisis do not use the same units for energy. There are British thermal units (BTUs), terawatt-years, millions of barrels of oil equivalent (MBOE), megatonnes of coal equivalent, and so forth. In this book, we convert all the data to metric units; namely, watts and joules and their multiples. The conversion factors among the most common units are given in Box 2.1.

Box 2.1 (continued)___________________________________________

A terawatt-year is 32 million terajoules, since there are that many seconds in a year. A large power plant generates about 1 GW of power, and thus a GW-year of energy per year. A terawatt-year is the annual output of 1,000 power plants. Since 1 BTU is about 1 kJ, a million BTU (MBtu) is about a billion joules or about 1 GJ. This size unit is used for partial energies. A Quad is a quadrillion (1015) BTU or a billion MBtus, a unit appropriate for worldwide production. It is equal to 172 MBOEs, a unit often used in magazine articles as well as technical journals. We shall convert all graphs to Quads and MBtu’s here, the saving grace being that they are close to the modern metric units.

There is more oil and gas to be found if you are willing to endure conditions in freezing, inhospitable places. Russia owns a lot of property where no one wants to go. North of Japan lies Sakhalin Island, where they used to send prisoners. The deposits there contain 14 billion barrels of oil and 2.7 trillion cubic meters of gas.21 Shell and Royal Dutch want to build an 850 km (500 mile) long pipeline to carry LNG to the USA. This is still being contested by Russia. The third largest gas field in the world is at Shtokman, in the Barents Sea near Murmansk, the largest city north of the Arctic Circle. That reserve contains 3.2 trillion cubic meters of gas, compared with 177 trillion in proven reserves. Western companies are bidding to get a part of this. But the cold conditions require the latest equip­ment and hard-learned techniques. There are icebergs, and shore is 550 km (340 miles) away. A pipeline on the bottom could be scraped by icebergs. Worse yet, antifreeze (glycol) has to be added to prevent the gas from reacting with the water that comes with it to form gas hydrates (more on this later), which can clog up the pipe. The water and glycol have to be separated out later. These arctic mining techniques are being tested in the Snohvit gas field in Norway. It may take $3 trillion to exploit these reserves.22 It is clear that Russia can afford to do this only with foreign investment.

The Arctic north of Canada contains oil and gas fields made more accessible by the shrinking ice cap and the opening of the Northwest Passage. The US Geological Survey estimated that between 25% (some say 10%) of the world’s “undiscovered oil reserves” could lie in the Arctic.23 This is, of course, an oxymoron. How would you know how much is undiscovered? One deposit was estimated to contain 31 bil­lion BOE in gas, enough to supply the US for four years. The problem here is a political one: no one knows who owns these deposits. North of Canada is also north of Russia, and the Russians planted a Russian flag at the North Pole. Stay tuned.

Here, we show in detail the present-day peak seen in Figs. 1.4 and 1.5, followed by projections of the temperature rise in the future as computed by climate modeling using the extensive observational database illustrated in the previous section. Figure 1.9 shows the temperature variation over the past 1,000 years as deduced by various methods (proxies). There is considerable disagreement up to about 1850, but with better data since then, all the proxies agree on the most recent temperature rise. The agreement is quite amazing since the range of the entire graph is only 1.8°C.

The decreasing uncertainly is seen more clearly in Fig. 1.10, where the weighted global average temperature deviations are shown with error bars for the calculated standard deviation.

Figure 1.11 shows the data from 1850 to the present for the northern and southern hemispheres and their average. The North has higher recent temperatures, probably because there is more industry; but other than that, the histories are similar, showing that the trend is truly global. The error bars on each point are significant: they indi­cate that there is only a 5% chance that the true value lies outside those ranges. It is quite clear from this that the earth’s temperature has risen from about -0.3 to +0.6°C (relative to 1980) since the preindustrial period.

The question is now whether the temperature increase is anthropogenic or not. Climate modelers have calculated the natural forcings and those caused by man, as shown in Fig. 1.3. Remember that these forcings depended on the “parameters” that the modelers chose to find the average over fine-scale variations in space or time.

Their projections, shown in Fig. 1.12, all agree up to year 2000 by design. The param­eters had been chosen so that the twentieth century data were correctly predicted by the models from the data from the century before that data. This is how the models are calibrated. The models can predict the future as long as the parameters do not change. Nonetheless, different models give different results for the future, and there

is a large range of uncertainty. The lowest curve in Fig. 1.12 is what would happen if the GHG level were held constant at the 2,000 levels with no further emissions. The temperature will not go down because the CO2 in the atmosphere stays there for hundreds of years. The three models shown predict a temperature rise of 1.8-3.6°C

by the year 2100. The 2007 IPCC report [6] gives the results of six scenarios ranging from optimistic to pessimistic. The most optimistic scenario predicts a temperature rise of 1.1-2.9°C in the next 100 years, and the most pessimistic one is a rise of 2.4-6.4°C. The range given for each model represents the 66% probability level.

I have chosen graphs which give an idea of the uncertainties in both the data and the models because the IPCC report has been challenged by individual scientists who have arrived at different conclusions.4 Though the ICPP’s Working Group 1 had input from over 600 scientists, only a fraction were involved with any one problem, and arguments are bound to arise. Nonetheless, it seems clear that GHG emissions will be harmful to some extent in the future, and these can be suppressed by replacing fossil fuels with other energy sources. There is no need to argue.

In Europe, the emphasis is on offshore turbines because of the lack of space and objections to the noise and aesthetics of onshore wind farms. It is more expensive to build towers in the sea, and there are problems with storms, icebergs, and salt water, raising the cost of operation and maintenance. However, the wind can be steadier and stronger at low altitudes so that the towers do not have to be quite so high. Denmark had the most installed offshore wind power as of 2005 (Fig. 3.11) and has led in the development of the technology.

As Fig. 3.12 shows, there are different ways to mount the towers in the sea depending on the depth of the water. If the installation is kilometers offshore, the turbines have to be floated and tethered to the bottom. This is much harder than for floating oil rigs because the towers have to be kept from turning, leaning, or tipping over. Except for experimental trials, no floating turbines have yet been installed, though Germany envisions placing them as far as 40 km offshore.9 In September 2009, Vestas Wind Systems of Denmark announced its V112-3.0MW turbine specifically designed for offshore use.14 This turbine incorporates new technology for increased efficiency, reduced noise, and resistance to the severe conditions, including a heating system to keep the parts from freezing. The power curve for the V112 is shown in Fig. 3.13. The turbine cuts in at a wind speed of

Denmark

53%

Fig. 3.12 Methods for installing offshore turbines (Energy from Offshore Wind, US National Renewable Energy Laboratory, NREL/CP 500-39450, February 2006. Engineering Challenges for Floating Offshore Wind Turbines, NREL/CP 500-38776, September 2007)

3 m/s and achieves its maximum output at 12 m/s. It can maintain this output up to 25 m/s. The steep dependence of power on wind velocity can be seen from this curve.

The consumption of energy in the world and in the USA is shown in Fig. 2.1. For the world, the total of 472 Quads is dominated by oil, with all fossil fuels accounting for 79% of the total. For the USA, the total of 71 Quads is dominated by coal, with fossil fuels accounting for 86% of the total. Renewable energy, mainly from wind, solar, and biomass (wood and waste), amounted in 2006 to only 1.3% of the total in the world and 5.5% in the USA.

500 450 400 350 300

СЛ

1 250

a

200 150 100 50 0

The growth of the world’s energy consumption over the last 36 years is shown in Fig. 2.2, organized by source. The total dominance of fossil fuels is evident. The contribution of renewable sources is only the thickness of the black line at the top. The dashed lines show that the rate of increase of total annual energy was rather steady from 1970 to 2002 at about six Quads per year. However, the rate seems to have increased since 2002 to about 16 Quads per year.

Figure 2.3 shows the fraction of the world’s resources that the USA consumes. We can see at a glance that the USA, with less than 5% of the world’s population, consumes 22% of its energy. It is noteworthy that most of this energy, 15% of the
total, is produced within the USA, as shown by the middle bar in the graph. The rest is imported. The USA is relatively rich in fossil deposits, and this explains why it has been lagging in the race to develop alternative sources. Countries like France, Germany, and Japan are more dependent on imports and have taken the lead in developing fossil alternatives.

Far below sagebrush country where mule deer and sage-grouse roam in Colorado, Utah, and Wyoming, there lie layers of organic marlstone bearing oil. The USA is reported to have two of the world’s 2.6 trillion barrels of shale oil locked in the rock there. Of this, 800 billion barrels are deemed recoverable, 2/3 as oil and 1/3 as gas. By comparison, the proven oil and gas reserves of the Middle East total 1.2 trillion BOE. But to get it out, one has to essentially boil rock. Rather than digging up 200 million tons of rock per year to get a million barrels of oil a day, it is less destructive to get the oil out in situ, by drilling rather than digging. In western Colorado, Shell Oil has drilled 1,000-foot deep holes to test the feasibility of this process, which works as follows. Three holes a few feet apart are drilled into the shale. In two of them, electric heaters, like toaster wire, are inserted in pipes to heat the rock to some 700°F (370°C). It takes months or years for the rock to reach this temperature, and it has to be kept there for the life of the well, say 10 years. Fortunately, earth is a good insulator. The gas and oil are boiled out of the rock and can then be pumped out conventionally in the third pipe and sent in a pipeline to a processing plant. Mentioning electricity used for heat should raise your hackles because electricity is much more efficient for mechanical work than for heating. That is why your microwave or toaster runs on 1,000 W while a large window fan uses only 100 W. In situ mining uses electricity generated by a conven­tional power plant that loses 69% of the fossil fuel energy that it consumes, as can be seen above in Fig. 2.9a.

Will mining shale oil produce net energy? It is marginal. Let us do a back-of-the-envelope calculation to see if shale mining can be in the right ballpark. One ton (2,000 lbs) of shale will yield 25 gallons of oil.24 It is easier to use metric units: 1 ton is about 0.91 metric tons (tonnes). At 42 gallons per barrel, we get 0.65 bbls of oil per tonne of shale. If we were to heat water, that would take 1 C/g/°C, so 1 tonne (a million grams) would take a million calories per degree centigrade. One calorie is about 4 J, so a million calories is 4,000 kJ. It is easier to heat rock, however. The specific heat of rock is only about 0.2, so now we only need 800 kJ per tonne per degree centigrade. We have to heat it by 700°F, which is about 380°C. To heat one tonne of rock by 700°C then requires 800 x 380 or about 300,000 kJ. From Box 2.1, we see that 1 kJ equals 1.6 x 10-7 BOE, so 3 x 10 5 kJ equals about 0.05 BOE. But we get only 0.65 bbls from 1 tonne of rock, so 1 bbl of shale oil requires 0.08 BOE of electrical energy for heating alone. If the elec­trical plant is 30% efficient, 1 bbl of shale oil needs 0.25 barrels of real oil for heating. To this we have to add the energy to run the refining plant. Using micro­waves to heat, as proposed by Raytheon,25 would be even less efficient.

In addition to all this, it is planned to build a “freeze wall” to keep oily liquids from seeping into the groundwater. This would be a wall of existing rock and water 1,800 feet deep and 20 feet thick surrounding the drill sites. By drilling more pipes, a cold ammonia solution is circulated to keep the wall at freezing temperature. Since refrigeration is even more inefficient than heating, this could double the electrical cost, and we would get only two barrels of shale oil for each barrel of oil equivalent in, say, coal used to generate the electricity. There would be an advantage in that oil is a liquid and much more valuable than coal for transportation. Destruction of the environment is a high price to pay for this marginal fossil resource.

Estonia provides an example of what happens if shale is strip-mined.26 There, shale oil provides 70-90% of the electricity. Shale is crushed to 6-10 mm size (about 0.5 in.) and burned in boilers topped by 250-m high chimneys. The ash and pollutants are blown up the chimneys, and the large particles of shale are re-burned when they fall back down. CO2 emission is 10 million tonnes a year. Only after new boiler technology from Foster-Wheeler of the USA was adopted did the SO2 and N2O emission fall below acceptable levels. Solid slag is piled up 100 m high. Five million tonnes of ash are produced annually. This is pumped with waste water into a huge lake formed by a surrounding levee 30 m high. This blue-green lake looks nice but is a toxic stew containing potassium, zinc, sulfates, and hydroxides.26

Consequences of a global temperature increase have provided fodder for journalists always looking for a new angle. We have all read about recent hurricanes, floods, droughts, and heat waves, as well as the dangers posed to coral, birds, and other species of wildlife. The connection to global warming is circumstantial and conjec­tural at best, but the connection with local warming can be established with more certainty. Some phenomena can be modeled quite successfully; the most certain of these is sea-level rise.

Sea level has been rising at the rate of 3 mm (1/8 in.) per year, which would amount to an inch in eight years, or about foot a century even if the rate does not increase. Low-lying places like the Netherlands, Indonesia, and Bangladesh would be the first to feel the loss of hundreds of square miles of land area. There is some evidence that the rise seen in Fig. 1.13 has accelerated since the onset of industri­alization. Most of this can be attributed to global warming.

The three main causes are thermal expansion of water as it is heated, the melting of glaciers that have slid into the sea, and the melting of ice sheets on land. The contributions from each of these sources are shown in Fig. 1.14. The bottom part of each column is the sea-rise rate averaged over the past 42 years, while the recent average is given by the total height of the column. The rate-of-rise scale is in mil­limeters per year (mm/year). The four columns add up to the 3 mm (1/8 in.) figure quoted above. In each case, it is clear that the rate of rise has accelerated. The breakdown into the four effects required computer modeling, since the water from melting glaciers, for instance, cannot be measured directly. However, the sum of the calculated effects can be shown to agree quite closely with the sea-level rise actu­ally measured. This gives us confidence in the accuracy of modeling procedures.

Icebergs that are already floating will not change sea level as they melt because the part that is underwater (85-90% of the iceberg, depending on the temperature and salinity of the seawater) occupies exactly the volume that the iceberg will fill when it melts.5 Glaciers, ice caps, and ice sheets that are on land, however, are a different story. As land ice melts, it not only adds water to the oceans, but it also wets the ground under glaciers, making them slide into the ocean faster. Glaciers are melting at the rate of two cubic miles per week,6 and the shrinking of glaciers

Fig. 1.14 Contributions to sea-level rise by glaciers, thermal expansion, and ice sheets in Antarctica and Greenland. The lower part of each column is the 42-year average rate; the most recent 10-year average is the height of the entire column. Data from Intergovernmental Panel on Climate Change [6] over the past decade can be seen in many photographs. This is direct evidence of rising temperatures, but the unseen feedback effect is more treacherous. Ice has a high albedo, reflecting sunlight efficiently. As it melts, ground is uncovered, and this absorbs more sunlight, causing higher temperatures. As permafrost in Greenland is defrosted, exposed vegetation can rot, giving off CO2 and methane. Although the total forcing from albedo change is negative, as seen in Fig. 1.3, it is the local heat­ing where ice cover is disappearing that causes the runaway effect.

Permanent ice covers only 10% of land surface and 7% of oceans, which is why the catastrophic changes in glaciers that we can see is not the main cause of sea — level rise. As seen in Fig. 1.14, the main effect is simply the expansion of water when it is warmed. Not all consequences of ice melt are negative. Ice over the North Pole is definitely getting thinner, as directly measured by submarines there.7 The long sought-after Northwest Passage is becoming a reality. Trees growing on newly exposed ground can absorb CO2. The negative aspects, however, are dominant. If all the snow and ice on Greenland and Antarctica were to melt, the sea level would rise by 7 and 57 m, respectively [6]. This has happened in geologic eras, and the earth has undergone hot and cold periods before, even in human history; but what is new here is that it is happening extremely fast, before mankind can slowly adapt to the changes as it did previously.

The picture of the V-112 in Fig. 3.14 shows that the blades of modern turbines have been designed with special shapes to maximize efficiency at all wind speeds and to minimize turbulence.15 Such shapes are also seen in newer airplanes (Fig. 3.15). Blades may evolve further to incorporate scalloped edges, as these have been found to reduce drag on a humpback whale’s flippers.16 As the wind speed varies, each blade’s pitch is changed with a motor to capture the most energy. In very strong winds, the blades are feathered as in airplanes. The fiberglass blades are much thinner than on windmills and there are only three of them per rotor. This design is driven by cost.17 More blades will not only be too expensive but will also require sturdier towers to support in strong winds. The blades are so long that even at only 5 rpm, the tip of a 200-feet (60-m) blade travels at 170 miles per hour (75 m/s).

The diameter of the rotors is very large because these catch more wind when the speed is low. This is explained in Fig. 3.16,18 which is drawn for a situation when the average wind speed is 7.5 m/s (17 m/h). The smooth, peaked curve at the left shows how often each wind speed occurs. The speeds are on the bottom scale. Notice that most of the time, the speed is between 2 and 12 m/s. The rightmost of

the rising curves shows the turbine’s output power for a 50-m diameter rotor. The power is limited by the size of the generator. The curve labeled 50 m-3.0 MW, therefore, rises as the wind speed increases but stops rising and stays flat when the curve reaches 3 MW (at around 16 m/s). Increasing the generator’s capability to

4 or 5 MW permits capturing the energy of the strongest winds, as shown by the uppermost curves on the right. However, these occur only a small part of the time. If, instead, we keep the generator at 3 MW and increase the rotor diameter, we get the colored curves labeled 70, 90, 120, and 150 m. These rotor diameters utilize the slow wind speeds more efficiently, even though the 3-MW generator cuts the curves off when the available power reaches 3 MW.

Even larger rotors would capture more of the slow winds under the peaked curve, but then the towers would have to be even taller than the monsters that we now have. High hub heights also contribute to a turbine’s efficiency. Winds are stronger away from the ground, where the trees, grass, hills, and structures impart a drag. This was a rather technical discussion, but it shows why it pays to tear down old turbines and replace them with fewer large ones.

Estimating the energy the world will need in the future is risky business. We have to depend on computer simulations, as we did for climate change. Some of these models are the same ones used in Chap. 1, and they differ widely in the assumptions made in each scenario. Results up to 2030 are shown in Fig. 2.4. The middle bar in each group is the reference scenario, in which policies and laws remain unchanged. The low and high bars in each group are the minimum and maximum predictions across all scenarios. As expected, uncertainty increases with time, and so does the range of predictions. For the case of high economic growth, we see that the present consumption of some 470 Quads will grow to 760 Quads by 2030. By the end of the century, the level will be above 1,200 Quads. The problem is obvious: this doubling and then tripling of energy demand will occur while oil and gas reserves are being completely depleted.

a

73

(0

3

О