LNG Increasing volumes of the fuel are traded across the oceans in the form liquefied natural gas (LNG). Gasification entails cooling the gas down to -1620 C and reducing its volume to 1/600thof the gaseous state; this is done in several independent units (trains) that are typically about 300 m long (Linde 2010). Liquid gas is stored in superinsulated containers before it is transferred (via articulated pipes) to isothermal tanks on LNG tankers, the largest of which (Q-Max ships of Qatar) can carry 266,000 m3 of gas (Qatargas 2013). With 22.2 GJ/m3 of LNG that equals to almost 6 PJ of floating energy storage. Re-gasification takes place in seawater vaporizers. Most of the recently commissioned terminals have annual liquefaction capacities of 4-5 Mt, most of the new receiving terminals rate between 3-5 Mt/year (IGU 2013).
By the end of 2013 there were more than 40 liquefaction terminals in operation or under construction in 19 countries, while nearly 30 countries had more than 90 regasification terminals with the largest number (24) in Japan (Global LNG 2013). Further expansion is underway, with 10 liquefaction and nearly 20 regasification terminals under construction by the end of 2013. Both the gasification and re-gasification terminals make relatively small land claims (typically 30-120 ha) and, as with oil tankers, actual LNG loading and offloading is done on piers that often extend far offshore. Liquefaction capacity of 3 Mt/year and area of 80 ha would translate to throughput power density of 6,400 W/m2.
A few examples show power densities of some important LNG projects. Norwegian Snøhvit LNG plant, Europe’s first world-scale LNG project on Melkøya island near Hammerfest, has an annual capacity of 4.3 Mt LNG and its compact modular design occupies only about 70 ha (Nilsen 2012); that (with 53.6 GJ/t LNG) implies processing power density of about 10,000 W/m2. The first three liquefaction trains at Ras Laffan, the world’s largest LNG exporting facility, have annual capacity of 10 Mt and claim 3.7 km2 (Qatargas 2013a) resulting in throughput power density of about 4,600 W/m2. Australia’s Darwin (Wickham Point) LNG plant has annual capacity of 3.5 Mt and area of 60 ha, and hence power density of 9,900 W/m2.
Re-gasification plants have power densities as high, or considerably higher. America’s largest receiving LNG terminal, Cheniere’s Sabine Pass in Louisiana, has nameplate capacity of 19.5 Mt/year and hence the maximum power density of about 8,300 W/m2. Japan is the world’s largest importer of LNG, with more than 30 re-gasification terminals, many of them very compact with high throughput power densities. Higashi Niigata on the coast of the Sea of Japan (annual capacity of 8.45 Mt, area of 30 ha), has power density of nearly 48,000 W/m2.
Himeji and Himeji Joint LNG terminal have combined capacity of 12.67 Mt/year and they occupy a landfill area of about 60 ha on the northern shore of Seto Inland Sea (Google Earth 2014); this yields throughput power density of nearly 36,000 W/m2. Futtsu terminal on the southeastern shore of the Tōkyō Bay is the world’s largest regasification facility dedicated to supplying fuel for the world’s second largest gas-fired electricity-generating plant, TEPCO’s five-unit, 4.534 GW Futtsu station (TEPCO 2012): with annual capacity of 19.95 Mt and a compact area of 55 ha (containing 10 storage tanks holding up to 1.1 Mm3) the terminal’s throughput power density is about 60,000 W/m2.
Thermal electricity generation
Accounting for space claimed by thermal electricity generation, either by burning fossil fuels or by fissioning uranium, is a rather complex endeavor. Direct claims for plant sites are fairly small but indirect claims for energy inputs (including their transport), management of wastes and electricity transmission reduce the overall power densities.
Most of the world’s electricity (almost exactly 80% in 2010) is produced by thermal power plants, including a tiny share that comes from the just described geothermal stations. Development of thermal electricity generation began with the first Edisonian coal-fired stations of the early 1880s that relied on large steam engines and dynamos, and the standard set-up that prevails to this day –- combustion of coal in large boilers and electricity generation using steam turbogenerators invented by Charles A. Parsons (whose first patent was granted in 1884) -– became the norm by the beginning of the 20th century (Smil 2005). Subsequent progress of thermal generation was slower and fast growth of capacities and notable gains in performances resumed only after WW II (Smil 2006). At the same time, an increasing share of thermal electricity was produced by burning hydrocarbons, beginning with fuel oil, sometimes even just crude oil, and progressing to natural gas, with the latter fuel eventually destined for gas turbines rather than for large boilers.
And during the late 1950s came the commercialization of nuclear electricity generation based on highly exothermic fissioning of uranium to generate steam in a variety of reactors. Fuel is arrayed in rods and water can serve both as a moderator (to slow down neutrons) and as working medium, but some designs use gas for cooling. Pressurized water reactor (PWR) had eventually emerged as the dominant design: in 2011 the world had 435 operating reactors of which 270 were PWRs (IAEA 2012). Other relatively common choices have included British gas-cooled reactors and Canadian CANDU using natural, rather than enriched, uranium.
Rapid expansion of nuclear capacities came during the 1960s and 1970s, followed by virtual cessation of building in the US and Canada and limited additions in Europe (mostly in France, the country with the highest reliance on nuclear generation). Since the year 2000 most new additions have been in Asia (China, Japan, South Korea, India) but Fukushima disaster of March 2011 has changed even those prospects. All thermal plants share the key components: large halls with turbogenerators, cooling towers (in locations where there is no possibility for once-through cooling) and large transformer yards are seen at both coal-fired and nuclear stations.
But fossil-fueled stations require, obviously, fuel storage (coal yards, oil tanks) and all modern coal-fired plants have appropriate facilities to control air pollutants (capturing fly ash capture and desulfurizing flue gases). A complete account of land claimed by thermal generation must go beyond the confines of power plants and quantify the impacts of requisite fuel production, be it coal, natural gas or uranium, and its delivery to plants. Consequently, my account of power densities of thermal electricity generation is divided into three distinct categories that now dominate the industry: coal-fired plants, gas-fired installations, and nuclear power stations.
Coal-fired power plants
During the 20th century increasing shares of global electricity came from hydro power, burning of hydrocarbons and, since 1956, also from nuclear fission but coal has remained the single largest energy source for thermal generation. The fuel has been particularly prominent in the world’s two largest economies: in the US it still accounted for 56% in 1990 and 51% in the year 2000 and only the recent rise of gas-fired generation has brought is share to 43% in 2012 (USEIA 2014a). In China coal-fired power plants generated about 80% of electricity in 1975 and nearly two generations later the share was still 81% of the much expanded output in 2012 (WCA 2013).
Although the principal components and configurations of coal-fired power plants have not changed for decades, the stations have seen large increases in the size of boilers and turbogenerators (the two forming a generation unit) and in overall capacities (multi-unit stations, such 3 x 500 MW or 2 X 1 GW, have been the norm). This trend began after WW II and in North America and Europe it had levelled off mostly by the 1970s. Largest turbogenerators had eventually reached ratings in excess of 1 GW, and the largest plants, made-up of a series of units (boiler-turbogenerator assemblies), have capacities up to 5 GW.
Given the wide range of their designs and performances and different origins and qualities of their fuel -– the two extremes would be represented by mine-mouth stations located next to large surface coal mines and burning poor-quality lignite, and by coastal stations located near large ports receiving high-quality coal shipped in bulk carriers from other continents –- it is not surprising that there is no such thing as a standard coal-fired power plant whose spatial claim could be used to calculate typical power density. This is nothing new: during the 1960s and the 1970s, when many of today’s aging coal-burning plants were constructed, differences in land requirements for large stations in the US and in the UK were appreciable (OST 1969; CEGB 1971).
Principal structures Dominant structures that are shared by all coal-fired power plants –- tall boiler buildings, machine (turbogenerator) halls, electrostatics precipitators, tall stacks, maintenance and office buildings, switchyard and coal storage –- occupy a relatively small amount of space. Coal combustion in large boilers and expansion of highly compressed steam in turbines are two energy conversions that proceed at very high power densities. In all modern plants mixtures of air and pulverized coal (with most particles as fine as baking flour, or less than 75 μm in diameter) burn in a swirling vortex with flames reaching temperatures as high as 18000 C.
Modern boilers (steam generators) are tall prismatic structure structures with walls covered in stainless steel tubing that circulates water to be converted to pressurized steam (Teir 2002). Boiler supplying steam to a large 1.3-GW turbogenerator is about 52 m tall and its footprint is nearly 520 m2 (33.8 m x 15.5 m) which means that combustion power density will be between 6.5-7 MW/m2; smaller boilers in less efficient plants have power densities of 2-4 MW/m2. Large modern steam turbines are also very compact: for example, 250-MWe Siemens design is 19 m long and 8 m wide, Skoda 660-MW unit is about 32x13 m and Siemens 1,350-MW turbine is about 70 m long and 30 m wide (Siemens 2013a; Fiala 2010). Rectangular footprint of these turbines implies approximate operational power densities of 0.6-1.5 MWe/m2. The core structure of a coal-fired power plant –- boilers and turbogenerator halls –- thus claim only a small fraction of a station’s total area: for a 1.5-GW (3 x 500 MW) stations it can be just 1.2-1.6 ha, a larger (2-3 GW) station powerhouse can claim 2-3 ha, resulting in specific power densities of 105 We/m2.
Undesirable consequences of high-efficiency coal combustion include voluminous generation of fly ash and (where high-S coal is burned) of SO2. That is why all modern coal-fired stations must invest into expensive techniques to limit the emissions of these air pollutants. Structures housing these air pollution control facilities are sited immediately adjacent to boilers and tall chimneys. A typical linear configuration would proceed (left to right) from a coal bunker feeding a ball mill to pulverize coal, a boiler, an air heater, electrostatic precipitator, FGD unit and tall stack. In plants with capacities of 1.5-3 GW electrostatic precipitators used to capture fly ash will occupy 1-3 ha, flue gas desulfurization (FGD) units 1-3 ha. Including adjacent walkways and separation spaces will raise these totals by 50%.
Captured fly ash can be used to make cement, an appealing choice because it lowers energy cost of the material and avoids landfilling. In China two-thirds of all captured fly (total of nearly 400 Mt/year) ash have been recently used by cement industry (Lei 2011). In the US Virginia’s Ceratech blends 95% of fly ash and 5% of liquid ingredients to make a stronger concrete (Amato 2013). FGD removes SO2 by reacting with lime or limestone to produce CaSO4 that can be used in wallboard manufacture. al. 2010). Despite these efforts, large volumes of fly ash an FGD sludge are still deposited within many plant sites, claiming considerable amount of land during 35-40 years of typical power plant operation: for 2-3 GW plants these claims can add commonly to between 120-160 ha, that is disposal power density of 1,300-1,900 We/m2. For coal with high ash content the areas will be obviously larger, 200 ha for an Indian 1-GWe plant burning domestic raw coal with 40% ash, and 480 ha for a 4-GWe station burning domestic sorted coal with 34% ash (CEA 2007)
After the removal of fly ash combustion gases are led to stacks and discharged to the atmosphere. All thermal stations also must condense (cool) hot water discharged by steam turbines. This could be done by once-through cooling that withdraws water from streams, lakes or from the ocean with little additional space; spray ponds require about 400 m2/MWe (2,500 We/m2), ordinary cooling ponds need up to 5,000 m2/MWe (200 W/m2). Cooling towers allow water's continuing reuse; a complete water system for a 1-GWe plant (including water treatment and cooling) could occupy up to 20 ha for natural draft towers and only 10 ha for induced draft cooling (CEA 2007). Natural draft towers have inevitable evaporation losses but require no power to operate; their power densities (per m2 of tower base) are typically between 20,000-40,000 We/m2. Mechanical draft towers (counterflow or crossflows designs) use fans to move the air, and in dry cooling towers there is no evaporation as water circulates in closed pipe circuits; their compact size means that power densities of heat rejection can be on the order of 105 We/m2.
Mine-mouth stations can be supplied directly from adjacent surface or underground extraction but stations receiving deliveries by trains, barges or ships need on-site storage for 90-120 days of normal operation. Energy storage density of coal yard will obviously depend on coal quality and thickness of a coal pile (commonly 10-12 m). The area needed for storing 90 days-worth of coal for a 3 GW station will be 15-20 ha. The total are of a 3 GW station with adequate on-site coal storage and requisite ash and FGD sludge disposal will be thus between 350 and 500 ha, that is power density of up to 850 We/m2 for installed capacity and (with 75% load factor) and up to 650 We/m2 for actual generation.
Variable space requirements of coal mining and storage of captured pollutants will surpass the fixed space requirements of buildings, air pollution controls, cooling towers and coal storages that remain constant, or change only in minor ways if there is no major reconstruction or expansion. But reliable calculations of coal mining land claims, the single largest item on the list of space requirements, can be done only for the plants supplied by adjacent mines or for the stations that receive their fuel from a single distant place of origin. In reality, many plants receive fuel from different source, others change their suppliers during the decades of their operation (in Europe the switch has been commonly form original domestic coal to cheaper fuel imported from Australia, Indonesia, South Africa or the US). This makes it difficult, if not impossible, to impute any definite power densities to individual plants relying on mixed supplies.
In addition, there are significant differences in fixed land requirements (due to sprawling or compact plant design dictated by land availability) and in variable claims for storage of fly ash (ash content can range from less than 10% to more than 30%) and sulfates from FGD (some plants sell of all this waste to other industries, others have to have enough space to deposit it for 3-4 decades). Finally, there are major differences in land claims imposed to connecting the stations to existing high voltage (HV) grids. Some plants will require only short new links, others will necessitate upgrading of older lines or construction of new connectors.
Two realistic examplesObviously, there is no such thing as a typical land claim of a coal-fired power plant, only a surprisingly wide range of final outcomes. That is why I will present two realistic examples whose substantially different power densities reflect common differences in plant efficiency, coal quality, fuel delivery and the extent of environmental controls. The first case will quantify a high-density set-up: a plant with installed capacity of 1 GWe operating with high conversion efficiency of 40% will require 2.5 GWt as coal; its mine-mouth location means that coal can be supplied either by high-capacity conveyors or by short-haul trucking directly from the mine, obviating the need for a large storage yard; the plant burns good-quality bituminous coal with energy density of 24 GJ/t, ash content of 4% and sulfur content below 0.5%; it has access to nearby source of cooling water and hence can do without any massive cooling towers; and, finally, it operates with a high capacity factor of 80%. That station would generate annually about 7 TWh (about 25 PJ) and with 40% efficiency it would require about 63 PJ of coal.
Coal for a 1 GW plant 1 GW x 0.8 (capacity factor) = 800 MW
800 MW x 8,766 hours = 7.0 TWh
7.0 TWh x 3,600 = 25.2 PJ
25.2 PJ/0.4 = 63 PJ
I also assume that the plant’s bituminous coal (energy density 25 GJ/t, specific density of 1.4 t/m3) comes from a large surface mine whose main seam is 6 m thick and whose recovery rate is 95%; this means that under every square meter of the mine’s surface are 8 t of recoverable coal containing 200 GJ of energy. To operate the plant would require annual coal extraction from an area of 31.5 ha (315,000 m2). Adding 10% for access roads, buildings and parking would raise the total annual mine claim to almost 350,000 m2 and it would mean that the mining would proceed with power density of about 2,300 W/m2.
800 MW/350,000 m2 = 2,285 W/m2 = ~ 2,300 W/m2 In this optimal case I assume that, except for the initial cut to expose the seam, the overburden would not claim additional land but it would be re-deposited into the mined section as the extraction front moves forward as an advancing indentation in the landscape; eventual recultivation would leave behind a flat landscape approximately 6 m lower than it surroundings. In the absence of large coal stockyard (with just some hoppers holding enough fuel for a day’s operation) and with once-through cooling, the largest areas occupied by the plant itself will be its generating halls, maintenance and office buildings, parking lots, switchyard, and the pond used to deposit captured fly ash.
Generous allowance for plant structures would be on the order of 10,000 m2, switchyard would take no more than 50,000 m2, and because all captured fly ash would be used in nearby cement production there would be no need for any storage of the ash on the plant site. Even after adding, very generously, another 40,000 m2 for roads, walkways, parking lots and a green buffer zone, the site’s total would be 150,000 m2, and it would prorate to about 5,300 W/m2. After adding the coal-mining area, annual electricity generation at the rate of 800 MW would thus require about 500,000 m2 and the power density of the entire extraction-generation sequence would be about 1,600 W/m2 (800 MW/500,000 m2 = 1,600 W/m2).
The second case aggregates all factors that maximize the land claim and hence result in in a much lower overall power density. Again, it would be a 1 GWe plant, but an older one operating with only 35% efficiency and with a lower load factor (70%); it would be located far away from a coal mine and supplied either by a unit train peddling between the mine and the plant; it would burning low-quality sub-bituminous coal (18 GJ/t) that would be extracted mainly from a 6-m thick seam and that would contain 10% of ash and about 2% of sulfur, and it would require large towers to recirculate its cooling water. The plant would require an almost identical amount of coal energy but mining of that fuel would claim a much larger area and proceed with power density of only about 1,600 We/m2. Adding, once again, 10% to the account for land claims accompanying coal extraction lower that rate to about 1,500 W/m2.
Power density of a 1 GW coal-fired plant 1 GW x 0.7 (capacity factor) = 700 MW
The second plant would also occupy much more land because it would need off-loading facilities and a storage yard for its coal (large enough to supply the plant for up to 60 days), larger area for fly ash disposal and also a pond for storing slurry from flue gas desulfurization, and land to site cooling towers. Coal for 60 days of generation would amount to 575,000 t (about 410,000 m3) and when stored in a yard 10 deep it would occupy 41,000 m2 and up to 50,000 m2 including approaches and off-loading ramps. Captured fly ash (density of 1.8 t/m3 after compaction) deposited in a 5-m thick layer in a settling lagoon would claim annually about 38,500 m2 (nearly 4 ha), and FGD working with 85% efficiency would remove annually nearly 60,000 t S; when captured as CaSO4 to be deposited in a 5-m thick layer in a pond near the plant its storage would add annually nearly 22,000 m2.
Power densities of fly-ash and sulfate disposal
63.1 PJ/18 GJ/t = 3.5 Mt coal
3.5 Mt x 0.1% = 350,200 t ash x 0.99 (99% capture) = 346,500 t
346,500 t/1.8 t/m3 = 192,500 m3
192,500 m3/5 = 38,500 m2
3.5 Mt x 0.02% S = 70,000 t S
70,000 t S x 0.85 = 59,500 t S
59,500 t S x 4.25 = 252,900 t CaSO4
252,900 t/2.32 t/m3 = 109,000 m3
109,000 m3/5 = 21,800 m2
With, again, some 150,000 m2 for all plant structures (including cooling towers) and a switchyard, land required for the plant’s operation would add up to about 260,000 m2, implying power density of about 2,700 W/m2. Plant (260,000 m2) and coal extraction (480,000 m2) would thus require the grand total of 740,000 m2 and the overall power density of producing 700 MW of electricity would be about 950 W/m2, a claim more than a third more extensive than in the first case. These two realistic constructs set low 103 We/m2 as the right order of magnitude for power densities of large coal plants: roughly 2,500-5,000 We/m2 for the plant itself (including all structures and storages) and roughly 1,000-1,500 We/m2 for a compact plant and coal mining of good quality fuel.
Obviously, land claims would be greater for plants burning low-quality coal and requiring extensive coal- and fly ash-handling arrangements. A detailed report on land requirements by India’s new coal-fired plants (burning domestic fuel containing 15 GJ/t and up to 40% ash) illustrates these needs for a typical 1 GWe (2 x 500 MWe) station (CEA 2007). Plant buildings would occupy only 12 ha but its total area would be 240 ha (with coal handling and surrounding green belt accounting for about 60% if that total), but fly ash storage (18 m high, sufficient for 25 years) would alone claim 200 ha, raising the overall claim (excluding coal mining) to more than 500 ha and resulting in power density of about 150 We/m2 for generation at 75% of installed capacity. A 4 GWe Indian station would claim about 1,000 ha and have generation power density of about 300 We/m2,