US BTU Usage

Last Update: Sun May 22 14:08:56 2016

US BTU Sources

Energy used in the US comes from a wide variety of sources. The major sources are petroleum, natural gas, coal, hydroelectric power, and nuclear power. Minor sources include wind, solar, geothermal, and wood. To keep things comparable to previous essays, I'll quote capacities in terms of the roughly 900,000 BTU per person that Americans consume. That also has the advantage of avoiding the mind numbingly large numbers with 12 or 15 zeros that turn up when talking about total US annual energy consumption.

I'm combining several topics here -- what technologies are used to produce American energy, where geographically the energy originates, and how much gaseous Carbon and miscellaneous pollutants are produced in generating the energy. The latter are quite approximate.


Roughly 40% of US energy (360,000 btu per person per day) comes from petroleum. Of that 360,000btupc, roughly a quarter (90,000btupc) comes from domestic oil wells. Another sixth (54,000btupc) comes from domestic natural gas wells that produce some liquids along with gas. The remaining 216000 btupc per person per day is imported. US petroleum imports oil primarily from Canada (20%), Saudi Arabia(10%), Mexico(10%), Venezuela(10%), and Nigeria(10%) with the remainder coming from a great many African, Middle Eastern and South American countries.

US petroleum production peaked in the late 1960s and declined until about 2010. It has risen a bit since 2010 due to production of oil from "tight shales" and production of Natural Gas Liquids. The latter are somewhat overcounted as their energy content is substantially less than that of comperable volumes of oil. The US is still the world's number 3 oil producer after Saudi Arabia and Russia, but further production declines are likely. The US is thought to have about 150 billion barrels of recoverable petroleum (About 6 or 7 years worth with no imports) still in the ground. Most of that oil is in the Arctic and in deep waters primarily in the Gulf of Mexico. Current (2010) Arctic and offshore production is around 700,000bpd in the Arctic and 2Mbpd from offshore wells. There are also very large amounts of "oil" locked up in oil shales for which there is currently no viable production technology.

About 70% of US petroleum (252000btupc) is used for transportation -- gasoline, diesel, jet aircraft fuel. 5% (18000 btupc) is used in commercial and residential applications -- primarily heating. 24% (86400btupc) is used in industry -- some of it as a feedstock for chemicals and plastics. It is not clear whether the energy needed for oil and metal refining is included in this 24%. A very small amount -- 1% (3600BTU) or a bit more is used to produce electricity.

Petroleum releases about 15.5 pounds (7kg) of Carbon Dioxide per 100,000 BTU produced. That works out to 56 pounds (25.4kg) per person per day of CO2 in the US from petroleum. That may be a few pounds high as it assumes that petroleum used as a chemical feedstock is burned rather than converted to plastics that may be sequestered in a landfill eventually.

Because oil and refined petroleum products are inexpensive and relatively compact, they are widely used across the planet. That results in a several long term problems for the United States. First, because domestic production can not come close to meeting domestic petroleum needs, the import of petroleum causes about half (40%) of the US's persistent Balance of Trade/Current Account deficit problem. Second, prices of petroleum including domestic product are set by world demand, not US demand. This can lead to great instability in price and availability of petroleum. Third, it is very unlikely that future worldwide demand for petroleum can be satisfied from future supplies. Sometime between 2005 (If you believe the pessimists) and 2030 (if you believe the moderates) world oil production will/has peak(ed) -- quite likely at levels well below world demand. This peak is predicted by the theoretical model proposed by M King Hubbert in the mid 20th Century that correctly predicted the 1970s peak in US oil production. But even if the moderates are correct, it appears that developing countries with large populations like China and India will increase their use of petroleum until the worldwide demand meets and exceeds the supply.

Natural Gas

About 23% of US energy comes from Natural Gas (207,000 btu per person per day). 85% (176000 btu per person per day) of the gas used in the US is produced domestically. About 15% is imported from Canada and Mexico via pipelines. A small amount is imported from various countries in the form of liquefied natural gas. Gas mixtures somewhat like natural gas can be produced synthetically -- primarily from coal ("town gas"), biologically or from certain unstable minerals in which Methane is combined with water. I have included these in the minor energy sources.

Recoverable US reserves of natural gas are poorly known but are thought to be at least 1 trillion cubic feet -- very roughly a 55 year supply at 2008 consumption rates. Recent developments in extracting gas from widespread black shale formations are causing the reserve estimates to roughly double.

Natural gas is used primarily for electrical generation (18%), heating/cooking (36%), and industry -- primarily as a feedstock (43%). A comparatively small number of vehicles run on compressed natural gas(3%) or Propane a slightly more complex molecule than Methane that is extracted from Natural Gas or Petroleum during processing.

Natural gas releases about 11.7 pounds (5.3kg) of Carbon Dioxide per 100,000 BTU. That works out to 24.2 pounds (11kg) per person per day of CO2 from natural gas. As with petroleum, the Carbon estimate assumes that all natural gas used in industry is burned which may not be the case when it is used as a feedstock.

Natural gas wells often produce hydrocarbons other than methane. These are generally removed before transport or during processing and sold separately. In production data, they may be called condensate and/or Natural Gas Liquids. Propane and butane are sold as liquified gases. The heavier compounds are pretty much a light crude oil and are handled as such. Short chain hydrocarbons (e.g. propane) predominate in NGLs which creates a minor problem in that NGLs are typically lumped with oil when aggregating production statistics, but the NGL products can not be pumped by the fuel systems in most transportation vehicles. As a result, about a third of US domestic "oil" production is materials that can not be put in vehicle gas tanks and can't easily be converted to transportation fuel. NGLs are not energy dense -- typically around 90,000 BTU per gallon vs 130,000 BTU per gallon for light crude oil. The difference is reflected in their pricing, but not in production statistics


About 22.5% of US energy comes from Coal (202,500 btu per person per day). Essentially all of the coal used in the US is produced domestically. in 2009, roughly 8% of US coal production was exported.

The US sits atop two of the world's major coal fields. Known US Reserves are estimated at 267.3 billion short tons which is 27% of the total known coal reserves of the planet. That is about a 250 year supply at current consumption rates. About 93% (197450btupc) of US coal usage is used for the generation of electricity. 5% (10859btupc) is used as a feedstock for industrial processes or to produce synthetic natural gas. 2% (4192btupc) is used as a raw material in iron production ("coking"). The efficiency of coal powered plants varies from 30% for older plants to between 50% and 60% for the most efficient modern plants.

Coal releases about 21 pounds (9.5kg) of Carbon Dioxide per 100,000 BTU. This is 80% more than natural gas and 35% more than petroleum for the same amount of energy. It is presumably due to the relatively higher percentage of Carbon-Hydrogen bonds and the absence of Carbon-Oxygen bonds in gas and oil. If precautions are not taken, coal also releases significant amounts of toxic metals (Lead, Cadmium, Mercury) as well as Sulfur containing gases when burned.

Because coal is a "dirty" fuel and is largely consumed in fixed power plants where bulky recovery devices can be accommodated fairly easily, serious attention is being paid to trapping Carbon Dioxide and various pollutants rather than venting them into the atmosphere. Currently, the technology to do this varies from immature (sulfates) to imaginary (CO2) depending on the substance to be sequestered. There is widespread support among liberals and some conservatives for "clean coal" -- something that we have only the vaguest idea how to achieve.

China is said to be using significant amounts of its rather limited coal supply as a feedstock for industrial chemical production (e.g. plastics) rather than as fuel. That could be done in the US as well. It is unclear whether such usage reduces the release of heavy metals and other pollutants from coal.


Hydroelectric facilities produce roughly 6% of US energy (54000 btu per person per day). Most hydroelectricity is produced domestically although a bit is imported from Canada.

As a practical matter, most high yield conventional hydroelectric sites in the US outside of Alaska are currently fully exploited although there is possibly some potential for additional generation from unconventional sources such as fixed turbines in fast flowing streams. There is said to be some potential for hydroelectric development at dams that were built for flood control and maintaining navigation depths rather than hydroelectric generation. In particular, there are a number of low head dams on the Mississippi River and its tributaries that might be able to generate useful amounts of energy thanks to the high water flows on those rivers.

The efficiency of hydroelectric plants is not constant because stream flows vary with seasonal affects and often seasonal irrigation needs. In the US, average annual efficiency is between 40% and 45%. Hydroelectric electricity is renewable, fairly inexpensive, and emits no Carbon Dioxide except for some from organic material trapped in the flooded area behind dams. Problems with hydroelectricity include environmental damage from flooding, alteration of natural riparian zone ecologies, interference with breeding of migratory fish such as salmon, and the dangers of flooding from dam failure. Silt accumulation in the artificial basins behind dams may be another issue.

Canadian hydroelectric resources are extensive and not as fully developed as in the US. The Canadian population is only about a tenth of the US population. It is conceivable that another percentage point or so of total US energy usage could be obtained from Canadian hydro power without reducing Canadian access to the resource.

If is sometimes asserted that US hydroelectric capacity could be doubled. While possibly true in some sense, it seems quite unlikely that it can or will happen in practice. It is probably more realistic to view hydroelectric power as a significant renewable resource that is exploited to the limits of current technology and is not likely to grow much in future years.


US nuclear power from 104 nuclear plants accounts for about 8% of US energy production = about 72000 BTU per person.

See the separate article on Thorium

There are for major issues with regard to nuclear power:

Even if one decides that the safety problems with competing energy sources are comparable to or greater than nuclear, one must consider the issue of finding sites for nuclear plants. Such sites must be tectonically stable (i.e. not on top of earthquake faults or volcanoes); must have abundant cooling water; must have minimum risk to nearby populations in the event of accidental release of radioactives; and must realistically allow evacuation of the neighborhood in the event of trouble. It is not clear how many such sites exist in the US. There are probably adequate sites -- especially East of the Mississippi River -- to accommodate substantial expansion in US nuclear power.

There is some question about the availability of Uranium fuel. Exploration for high grade Uranium deposits stopped dead in the 1980s when the market for the metal crashed. Since that time, Uranium for Nuclear Power has come from a few proven mines and from nuclear weapons decommissioned after the end of the cold war. These sources have a limited capacity to support future power generation. There is a short term problem with the availability of recycled fissionables from nuclear weapons because available supplies will be exhausted sometime around 2013 leaving a shortfall of about 40% in available nuclear fuel. On a longer timescale, there is perhaps a century's worth of Uranium available at current usage levels and even that will require exploration and development.

However, there are thought to be extensive Uranium deposits that have not yet been identified. In addition, there are vast low grade sources including large amounts of Uranium dissolved in sea water that are probably recoverable with current technology. There is little doubt that some of these sources can be used although the cost is a bit uncertain. Fortunately, the cost of Uranium is only a small part of the cost of nuclear power. Even if the costs of future Uranium are substantially underestimated, the cost of nuclear power will probably be only slightly affected.

There are many theoretical designs for Nuclear reactors that use lower grades of fuel than are presently used and/or are safer and/or that 'breed" additional fuel by transforming non-fissionable materials like U238 to fissionable forms. Although these present a paper solution to the problem of declining recycled material supplies, the high cost, long lead time, and public resistance to nuclear power make the construction of plants using these designs in a timely fashion extremely problematic. There are also a few proven designs that could be pressed into use such as small, liquid metal cooled reactors developed to power nuclear warships. These small, largely self contained, reactors generate perhaps 35-70 MW and would possibly be much easier to site and approve than full sized nuclear power plants. However, one must deal with the likelihood that the probability of significant nuclear accidents is probably more a function of the number of plants than the size of the plants. It may be difficult or impossible to meet reasonable safety expectations with a very large number of small facilities.

Nuclear power emits no Carbon Dioxide. Some CO2 is produced in the construction and decommissioning of nuclear power plants. The same is true of facilities like hydroelectric dams. Since the amount in question is presumably not all that large compared to the lifetime power output of a nuclear power plant, I have ignored it.

Minor Energy Sources


Thorium is an fairly abundant radioactive material that is potentially an alternative to conventional Uranium/Plutonium nuclear reactors. There is only one naturally occuring isotope of Thorium -- Th232. Although it is radioactive, its half life is on the order of 14 billion years and it is not a serious radiation hazard. Thorium is considerably more abundant than Uranium. Thorium can be altered to fissile Uranium-233 by activating it with subatomic particles. Because Thorium and Uranium have different chemical behavior, it is easier to concentrate U233 produced from radiating Thorium than it is to separate fissile U235 from non-fissile U238 both of which have essentially the same chemical behavior. Nonetheless, all currently extant nuclear reactors are Uranium based.

It is perfectly possible to build nuclear reactors operating on U233 bred in the reactor itself using Thorium "fuel" without separating/concentrating the fuel. Several experimental reactors have been built and designs exist for others. It is also theoretically possible to build nuclear weapons based on U233 from Thorium although it may be a bit harder than building Uranium and Plutonium weapons. India has detonated one small U-233 device.

It is frequently claimed that Thorium reactors would be safer than Uranium based reactors and would exhibit fewer nuclear waste problems. It should be noted that there are designs for Uranium based reactors (e.g. Pebble Bed) that are probably much safer than existing designs. The nuclear waste issue seems mostly to hinge around the fact that spent Thorium fuel is not a major radiation hazard whereas "spent" Uranium fuel is. But it should be pointed out that in practice, spent Uranium fuel is typically reprocessed to remove fissile isotopes, and the waste from reprocessing is much less of a problem after processing.

It's hard not to feel that Thorium fueled reactors are being oversold. But they may in fact turn out to be preferable to Uranium/Plutonium designs.


The US has long been one of the world leaders in the production of electricity from fluids heated by the Earth's internal heat. The US currently has about 3GW of geothermal generation capability in seven Western states -- mostly in California.

Geothermal power doesn't require fuel except possibly to run some pumps. Facilities tend to be expensive, but costs are competitive with other energy sources. Currently, only the lowest cost geothermal resources are being developed. Although no Carbon Dioxide is created by burning fuel, some may be vented from the depths along with the heated water. In some cases, the hot/superheated water may contain large amounts of dissolved minerals that can present a disposal problem.

Space heating directly from geothermal heated water is possible and has been done in the US occasionally on a small scale where suitable hot springs/wells happen to be handy. The largest current systems are in Boise, Idaho where geothermally heated water provides space heating for a number of downtown commercial buildings and the state capitol. However overall, geothermal space heating is not currently a significant contributor to US energy usage.

Conceptually, geothermal heat in volcanic areas could probably be used for industrial processes such as hydrocarbon refining, fertilizer production, etc. This is not currently done and would probably require many years of research and development before it could reduce US energy usage significantly.

Use of ground water as a source/sink of heat for heat pumps is practical most places where the ground isn't permanently frozen. (and where there is ground water at reasonable depth)/ I've chosen to count this as a more efficient form of Electric space heating/cooling and limit the term "geothermal" to the extraction of energy from water naturally heated by the Earth.

The current Geothermal contribution to US energy production might be as high as 8000 BTU per day, but is probably less.


Wind power has been widely used in the US for over a Century to pump water for agriculture. Technology to generate electricity from wind power was developed in the middle of the 20th Century. It is now possible to generate significant amounts of wind power at prices competitive with electric power from other sources.

Wind turbines generate no CO2 and essentially no pollution of any sort other than that required for their construction. In order to achieve high output and efficiency, current technology requires them to be very large. The diameter of a large turbine can approach 100meters (Roughly the length of an American football field).

A major problem with wind power is wind itself. Its velocity tends to vary dramatically over short time periods. Thus the output from wind turbines varies from moment to moment. Although the power grid has some flexibility, there are limits. Hooking a large number of wind turbines to a power grid without some intermediate buffer to smooth out the energy flow is turning out to be a substantial issue. There is no good proven design for the buffers, and no one has any reason to actually build the things. This may eventually limit wind power deployment at levels well below theoretical possibilities.

Another problem with wind power is low availability -- which is to say that over the course of a year a typical wind turbine farm rated at 1 gigawatt will produce much less total power (between 25 and 35% of the maximum output) than a nuclear plant with a similar rating. A second problem is that wind "fuel" must be used when it is available and can not readily be stockpiled for when it is needed. A third is that low wind power availability on any given day may apply over large regions, not just single sites. The fourth is the need for reserve non-wind capacity on the grid to make up for wind power when wind output is low. Another problem is that because of the huge size of efficient turbines, there is widespread opposition to deploying them in populated/scenic areas.

Yet another problem with wind is that even though the blades appear to be turning slowly, the ends of the blades are actually moving faster than cars on an expressway. This results in some fatalities amongst flying creatures. In particular, one wind farm in California is said to exact a substantial toll from raptors cruising for rodents and not expecting to be attacked by a turbine blade. There is considerable disagreement about how serious the problem overall is, but it is an issue that can always be hauled out and escalated by wind power opponents.

Wind turbines occasionally self destruct or fall down. Accident data can be found at]

Current turbines tend to be quite noisy which limits their deployability in populated areas and some rural areas.

Windpower in the US is currently around 25GW (equivalent to perhaps 8 modern nuclear power plants) and growing at 30% a year. That amounts to around 0.6% of US energy production (roughly 5400 BTU per person per day). Many critics of windpower agree that wind power can be expanded by considerably before its limitations become a problem, but they often feel that the strategies used to encourage wind power deployment are shortsighted. It seems likely that use of windpower will grow substantially for at least a decade at which point it will make substantial contributions to US energy production. It's possible that the eventual limitation in deploying wind generation will be the inability of the power grid to deliver large volumes of power generated in remote areas to distant urban areas for consumption.

Prior to the 20th Century wind was the primary source of motive power for shipping. Thus wind is less of a contributor to the nation's energy budget today than it was in say 1800. It is conceivable, but doubtful, that commercial sailing ships will return to the waters in large numbers someday. But only if hydrocarbon fuels become very expensive and nuclear is not an option.


Solar energy is more or less free and is non-polluting although some amount of Carbon Dioxide and other pollutants is likely to be released in the building of Solar energy collection devices. Solar energy really comes in several forms. Sunlight available at residential or industrial locations can be used for space heating, water heating or electricity generation. Alternatively, locations that are favored with large amounts of sunshine can be used for commercial power generation.

The principle problem with solar power is that power/heating can only be accomplished when the sun is shining. That can be as few as eight hours a day even on sunny days in Winter in Canada and the Northern US -- much less in Alaska and the Canadian Arctic in Winter. Cloud cover, precipitation, and dust reduce the power available. In addition, sun angles can be quite low requiring collectors to be tilted and reducing the collection area at a given site substantially since shading of one collector by another has to be avoided. Solar energy can be stored either by pumped storage or -- in the case of high temperature solar facilities -- storing molten salts. However, there will be additional losses.

Solar heating of domestic hot water has been practiced in the Southern US for nearly a century and became somewhat common in favored Southern areas after the 1970s petroleum disruptions. Hot water may sound trivial but 2.7% of US energy usage is devoted to heating domestic hot water (including heating swimming pools) and another 1% is used for commercial hot water heating. Together, that is 33750 btu per day per person per day. In Israel which has favorable climate and geography and has aggressively pursued solar hot water, 3% of the country's energy supply is provided by solar hot water heaters.

Sunlight can be used for space heating of residential and commercial property. Showcase installations achieve quite impressive results, but require very aggressive insulation and sophisticated air exchange technology that probably can not be retrofitted very effectively to the existing US housing base

Sunlight can also be used to generate electricity. This can be either local using roof mounted photoelectric cells or commercial using large arrays of photoelectric cells or boilers that use sunlight to heat a working fluid which is then used to drive generators. Domestic installations using full price commercial panels are currently cost effective only at off-grid locations and may require substantial lifestyle changes by users. Showcase installations typically do not pay representative prices for hardware, and often work out economically only by selling excess power during daylight to utilities at rates that surely are not sustainable for wide scale installations.

Large scale commercial electric generation has been accomplished at favored locations in the Southwestern US deserts and elsewhere. Projects are often announced, but only a few installations appear to be in actual daily use. One limiting factor is the need for large amounts of often scarce water as a working fluid for boilers and/or for mirror washing. Availability in optimal locations is typically about 20% of "sticker rating". Total "sticker ratings" for solar power in the US look to be around 1GW -- about 30% of a modern nuclear power plant. Total power produced from all solar electric sources in the US appears to be low -- perhaps a few thousand BTU per person per day.

Currently solar hot water is only a minor contributor to US energy supplies. At most a few thousand btu per day per person. It could probably be considerably expanded if inexpensive mass produced units were available that were easily integrated with existing water heating systems and were proof against damage from an unexpected freeze.

Solar space heating probably is a very small positive contributor to US energy usage even after subtraction the costs of air conditioning to offset unintentional overheating in Summer. And, of course, greenhouses use solar heating to extend growing seasons in cooler climates. It does not seem practical to estimate the size of these effects. It's small.

Solar electricity will possibly be a significant contributor to US energy usage in future years. Current contribution is small and there are no credible plans for aggressive installation of solar generation capacity.

Overall, solar contributes perhaps 0.5% of US hot water needs = .005*(.01+.024)*900000 = 150 BTU per person per day. Add a hundred or two BTU for space heating and perhaps 500 BTUpppd for existing solar electric power generation.

While it is conceptually possible to use solar heat directly rather than heat from electricity in manufacturing processes, in practice, there are usually a number of practical constraints that rule it out. There are few actual installations in place. see


Wood is a quite modest source of energy in the US. Several million primarily rural households heat at least partially with wood. A few commercial and public buildings in rural areas use wood for space heating. Although wood powered electric generation can be cost competitive with other sources, less than a dozen small (20-40MW) wood fueled plants are in commercial operation in North America -- fewer elsewhere. At a guess, wood provides perhaps 2% of the nation's space heating needs -- about 120 btu per person per day.

Wood burning and handling is often quite polluting in terms of particulates and toxic combustion products. Serious power plants can also exhibit fermentation and particulate problems from wood and ash storage and handling. But because wood is mostly burned in thinly populated rural areas and the wood burning is diffuse, the pollution is not a major concern. Although Carbon Dioxide is released when wood is burned, the carbon in the CO2 was previously sequestered by photosynthesis. Wood is a renewable resource except for the CO2 generated during wood harvesting, transportation, and handling.

The 25MW McNeil plant operated by Burlington (Vermont) Electric is said to consume a significant amount (perhaps 30%) of the easily available scrap wood in NorthWest Vermont. Clearly, there are real and modest limits to how much wood can sustainably be harvested for energy in the US and Canada. Probably wood burning could be expanded by a factor of four or perhaps even twenty before supply became a problem. However, it would still be only a very minor contributor to US energy supply. Some amount of energy -- mostly petroleum based -- is required to collect wood, process it, and deliver it to its point of use.


Biomass -- production of combustible liquids (or gases) from natural products is popular, but controversial. There are actually three aspects. The largest is the production of ethyl alcohol (Ethanol) by fermenting sugar rich plants -- typically waste sugar cane in the tropics. Corn in the US. Plant oils such as rapeseed can also be burnt in diesel engines -- typically after being used for cooking food. Finally, plants grown specifically as fuelstock can be grown and processed. There is a good summary of biomass crop energy yields at []

Biomass is renewable. Although Carbon Dioxide will be released when the fuel is consumed, it is CO2 that was captured recently from the atmosphere via photosynthesis. However, the large amounts of energy dependent fertilizer, cultivation, etc required to create Ethanol from corn -- and the impact of removing the corn from the food supply has made its use controversial.

Worse, ethanol (but not biodiesel) suffers from the problem that the fermentation product is a wine or beer that contains a relatively low percentage (10% give or take) of ethanol dissolved in water. For most (all?) applications, the ethanol needs to be extracted. In principle, that probably does not require a lot of energy. But in practice, the method used is distillation -- which does need a lot of energy -- typically provided by natural gas. As a result it turns out that production of corn ethanol in the US has a small positive or even negative net energy return depending on corn crop yields which vary from region to region.

Oils from various crops do not (usually) require distillation. But they often have undesirable characteristics for a fuel such as solidifying at low (or even room) temperature or forming varnishes via a phenomenon known as polymerization. Again, see

As of 2006, the US produced 712 trillion btu of ethanol and 32 trillion btu of biodiesel. However, that needs to be adjusted for the btus used to produce the energy. The latter numbers are highly controversial. I have assumed that fuel production requires 80% of the energy eventually released. The actual numbers could be lower or higher. On a per/person/per/day basis, the 2006 production was 1300btu pppd of Ethanol and 58 btu ppd of biodiesel.

There is currently no commercial production of biomass energy from cellulose fuelstocks such as switchgrass or from algae. Such production is considered to be an area for research and development.


The term Biogas is used for two different combustible gases. The more important of the two is a mixture of methane, water and carbon dioxide produced by waste materials as they decay. The other is a mixture of nitrogen, hydrogen and carbon monoxide somewhat similar to town gas. It is produced by heating wood.

Biogas is typically produced by landfills, sewage treatment plants and farms. It requires cleanup before being used for many applications, but in general can be used wherever natural gas is used. It has more or less the characteristics of natural gas albeit with fewer BTU per cubic foot. For the most part carbon dioxide released from the use of biogas is carbon dioxide that would have been released anyway were other disposal methods used. Recovered biogas produces about 0.014% of US energy -- 1260 BTU per person per day. Certainly biogas production/recovery can (and probably should) be increased, but it is unlikely to ever be a major contributor to US energy production simply because there is insufficient waste material.


Propane is a light (three carbon atom) gas produced as a byproduct of both natural gas and petroleum production. Unlike natural gas, Propane liquifies at a reasonably low pressure of about 11 atmospheres and boils at -44F (-42C). It is fairly widely used for home heating in areas where natural gas is not available, and for transportation. It is largely interchangeable with Compressed Natural Gas and is more widely available. Differences include minor changes in injection nozzles, a reduced risk of explosive container failure for Propane, increased risk of explosion from leaks (propane is heavier than air and will pool in low spots).

Propane is used to heat about 6.9 million homes, and is used to power about 190.000 on the road vehicles. It is also used to power a great many industrial vehicles from forklifts to locomotives that can not easily be driven to a gas station. About 90% of propane used in the US is domestic. The remainder is mostly from Canada. Propane yields only 91690 BTU per gallon and thus will deliver only 75% of the mileage per gallon compared to gasoline. Propane produces about 13.8 pounds (6.25kg) of CO2 per 100000BTU -- 20% more than natural gas. If I haven't made a mistake, Propane produces about 12560 BTU per person per day (=1.4% of US energy needs.)


Ocean waves are motions in water caused (almost entirely) by winds. Essentially, they amount to wind energy captured by the ocean and eventually lost to friction. The amount of energy present in ocean waves at any given time is quite large, but the technology to capture that energy and convert it to human usable power is pretty much non-existent. It has been estimated that the UK might be able to meet 15 to 25% of its current electricity needs from a fully developed wave power capability. Presumably similar figures would hold for the West Coast of the United States.

A wide variety of technologies have been proposed. A few have been prototyped. Several commercial plants have been announced. None are currently in operation. Wave power if/when it can be captured should emit no Carbon and require no fuel. The amount of power available at any capture site will vary somewhat predictably on a seasonal basis and quite unpredictably on a shorter term basis.

The principle problem with wave power other than the fact that we do not currently have much in the way of technology to capture it with is that it is diffuse. Capturing the wave power expended on 100km of coast line is likely to require a facility 100km long.

A secondary problem is that wave power generation equipment must be able to stand up to corrosion and occasional severe storms. There are environmental concerns including the effects (not necessarily detrimental) on commercial fisheries.

There is currently no regular production of wave power in the US.


Ocean tides are caused by the affects of the gravity of the moon and sun and the rotation of the Earth. They move very large amounts of water considerable distances in a highly predictable manner. There are usually two high and two low tides a "day". Basically, one can either stick turbines into the tidal flow or one can build a dam and generate power as water flows to the dammed side and then back. Both methods produce peaks and zeros of electric production every six hours in most locations because tidal flow reverses every twelve hours and 30 minutes.

"Sticker Ratings" for tidal power are the peak power ratings. They need to be halved to approximate average power. And obviously, they need buffering of some sort since the power output goes to zero at (predictable) times.

Tidal power produces no CO2 except for that produced during the construction of the facility. Neither does it require fuel. There are significant environmental concerns with regard to destruction/alteration of the local wetland/estuarine environment and/or the possible affects of large submerged turbines on marine life. Technical issues include low pressure heads and flow rates available at most potential sites. Because of the motion of the moon, the times of peak flow will move an hour ahead every day which means that in some parts of the week, tidal power will be unavailable at times of peak demand. On the other hand, the time of peak flow will be highly predictable. A final problem is that sites with good tidal flow often are also important shipping channels which may constrain the design of tidal power plants for those sites.

Only a handful of tidal power plants are currently producing power around the world and only two -- One in France and one in South Korea -- generate really significant power. Potential US sites include the East River in New York and San Francisco's Golden Gate. Two areas near the US with exceptionally high tides that may be exploitable are Canada's Bay of Fundy and the head of Mexico's Gulf of California/Sea Of Cortez.

Although the amount of energy in the Earth's tides is enormous, it is unclear how much is likely to actually be accessible. An engineering estimate of a facility at the Golden Gate estimated a sticker rating of 100MW and an averaged output of about 35MW. That is not a lot -- perhaps 5% -- of the power from a modern nuclear power plant. Barring major technological breakthroughs, it is unlikely that tidal power will be a significant contributor to US energy generation in the near future. That possibly could change as the century progresses. But for the time being, I will count tidal power as a very minor contributor to US energy needs.

See [] and []


Ocean Thermal Energy Conversion (OTEC) generates electricity driven by the temperature difference between cold deep ocean waters and warmer surface waters. Assuming that pumps and other overhead energy consumers are driven from the electricity generated, this is a non-carbon technology. Unfortunately it is not terribly efficient and is likely to be cost effective only where surface waters are warm and continental shelves are narrow putting deep cold water near shore. In the US, that combination would seem to exist only in Hawaii. The technology may be better suited to Japan, the Philippines and Indonesia than most of the US. Current contributions to US energy needs are zero.

Shale Oil

Oil shales are fine grained rocks containing hydrocarbons locked in the rock often as waxy long chain hydrocarbons that do not flow at room temperature. In the worst case, they can somewhat resemble dirty candle wax dispersed in the pores of bricks. They are a source of much confusion as there are a number of oil shale deposits throughout the world. Some are economically exploitable with current technologies. Some aren't. Depends on the rock and the nature of the hydrocarbons.

The US has vast deposits of oil shales in the Rocky Mountains which -- regrettably -- have defied numerous attempts to exploit them in the century since they were discovered. They are of interest because the total amount of energy stored is very large. The Green River oil shales are estimated to contain the equivalent of 3.3 trillion barrels of oil -- roughly twice the estimated proven oil reserves of the entire world in 2007.

In addition to the Green River shales, there are minor deposits of oil shale elsewhere in the US and Canada. A deposit of Ordovician black shale in Ontario supported a small amount of commercial oil production in the 19th Century. Similar rocks are found in the US.

The principle problems with oil shale are the difficulty of extracting the "oil" without using more energy than is made available in the final product, the problem of disposing of mountains of waste rock, the probable need for vast amounts of water for processing the shales; and the need to dispose of large amounts of contaminated water from mining and processing the rock. Most commercial production outside the US simply burns the oil shale to produce electricity rather than refining the hydrocarbons into liquid fuels. Given the vast US reserves of less troublesome coal, it seems unlikely that burning untreated oil shale would be a viable strategy in the US -- at least not any time soon.

The hydrocarbons in the Green River oil shales would appear to produce about 18 kg of Carbon per 100,000BTU, but that not terribly relevant as the material will probably be refined into gasoline or diesel fuel.

Oil shales are a tantalizing possibility, but there is no proven current technology to exploit the Green River formation shales and the likelihood that they will contribute to the US energy future in the short term is low. Their current contribution to the US btu budget is zero.

Tar Sands

Tar sands are sands and clays infiltrated with heavy oil that is too thick to pump. They are quite common worldwide, but uncommon in the United States. There are some US tar sand deposits -- mostly in Utah but in several other states as well. There vast deposits in Canada in the province of Alberta -- estimated at 1.7 trillion barrels of recoverable oil. US deposits potentially might yield 10-20 billion barrels of oil (1.5-3 years supply of petroleum for the US). Those in Alberta are about equal to the entire estimated world supply of conventional oil. It is entirely possible that later in the 21st Century, Canada and Venezuela which also has very large oil sand reserves will become the world's primary sources of petroleum.

Petroleum from oil sands is expensive -- around $70 a barrel in 2009 dollars, but there is clearly a lot of it. Production problems include the need for large amounts of water in arid regions; disposal of contaminated water; and disposal of large amounts of waste material.

Oil produced from tar sands is going to produce the same 7kg of CO2 per 100,000btu as oil from conventional sources when burned, but roughly 2 additional kg are emitted in extracting the oil. So, call it 9kg of CO2 per 100,000 btu. Although US tar sands are occasionally mined for road paving material and there has been some minor oil production historically, current US production is basically zero. OTOH, roughly 40% of Canadian oil production is from tar sands (aka "oil sands") and roughly 6% of US petroleum usage is oil imported from Canada. Additional oil is imported from Venezuela which also produces some petroleum from tar sands. Several percent of US petroleum usage is ultimately from tar sands. To avoid double counting that energy, let's consider US current oil sand energy to be zero.

Refinery Gains

Refinery gains refer to increases in volume of hydrocarbons when heavy hydrocarbons are refined. One barrel of heavy oil may produce 1.3 barrels of gasoline or diesel. There is no gain in energy. Rather the opposite in fact. But it is traditional to measure petroleum production in easily measured units of volume (barrels) rather than energy. Refinery gains can cause some confusion about actual production amounts.


Synthetic fuels are oil fuel mixtures -- kerosene, diesel, gasoline made from coal, biomass, or natural gas. The technologies to make synfuels has been around for nearly a century. Germany produced significant amounts of synfuels from coal during World War II. Currently, the only countries producing synfuels are South Africa (160,000bpd) which has considerable coal and very little oil and Malaysia which produces some diesel and lubricating oil from natural gas.

The technologies for synthetic fuel generation require considerable energy and tend to produce considerable CO2 over and above that which will be released when the hydrocarbon fuel produced is eventually burned. As a point of reference it currently takes about 200,000 btu worth of natural gas to produce 100,000 btu of diesel fuel.

Current (2009)US production of synfuels is very small and is limited to a handful of Research and Development facilities. I have considered it to be zero.

Pulping Liquor

Pulping liquor is a byproduct of some paper pulp manufacturing processes. It is a mixture of various "organic stuff" -- primarily wood lignin -- that is mixed with diesel fuel and used to fuel pulp plant boilers. As much as 12 billion gallons a year are burnt in the US. I found it difficult to get reliable estimates of the BTUs involved, but it appears that pulping liquor consumption makes a small, but significant, contribution to the US energy budget -- perhaps 5400 btupcpd = about 0.6% of US energy. About comparable to wind generated electricity in 2008. (I find it somewhat difficult to believe that the energy produced is really this high, but until I get time to look into the issue, we'll act as if the numbers are correct.).

Pulping liquor is a renewable resource. The CO2 released was extracted by the atmosphere by photosynthesis. It appears that the amount of resource available depends entirely on how much paper is produced domestically and is not expandable beyond the need for paper. Paper production in the US may be declining due to the replacement of printed newspapers and books with digital devices.

Methyl Clathrates

Methyl clathrates are marginally stable minerals containing methane gas in ice. They are found in sediments in polar regions as well as in oceans and some deep lakes at depths to about 500m where temperatures and pressures permit ice to exist in sea bottom sediments. When it was first recognized the clathrates can exist in the oceans, the volumes were thought to be very large -- many times other hydrocarbon reserves on Earth. However, as more knowledge has accumulated, it has become evident that clathrates are rarely present at depths below about 500m making them much rarer than was originally thought. Still, resources worldwide may be twice those of natural gas.

There are several major problems in clathrate exploitation. One is that the deposits are often small and widely dispersed. Another is the possible need to mine volatile and potentially explosive minerals under rather difficult temperature and pressure conditions. A third is the potential for self destruction of the deposit being mined through a mechanical chain reaction where escaping methane stimulates the escape of more methane -- a blowout. The liability from inadvertent destruction of a fuel resource, possibly creating a tsunami, and initiating the release of vast amounts of a greenhouse gas into the atmosphere might be enormous. Not an incentive to drill/mine.

It is conceivable that some defensive natural gas exploitation of clathrates may be undertaken if global warming threatens to cause some clathrate deposits to break down naturally. Better to burn the stuff than simply allow it to escape.

There is currently no identified commercial production from clathrate deposits in the US and Canada. It is conceivable that some natural gas in Arctic regions in both countries actually comes from decomposing clathrates, but the current lack of pipelines to bring Arctic gas to the temperate zones makes usable production zero.

Town Gas

Town Gas is a noxious mixture of Hydrogen, Methane, Carbon Monoxide, Carbon Dioxide and other gases. It is made by burning/distilling coal or heavy oils in the presence of water and controlled amounts of air. Its strong point is that it can be produced without mining the hydrocarbons and bringing them to the surface. Its weaknesses are that it is carbon rich, toxic, and not an especially efficient way to extract energy from fossil hydrocarbons. Town gas was widely used in the 19th Century in areas that had coal available but little oil. It was largely replaced in the second half of the 20th Century by natural gas.

There are many processes for the production of gas from other hydrocarbons. They yield different gas mixtures. Gas generation may be used in the future to extract energy from hydrocarbon deposits such as thin or deep coal beds that are not otherwise economically exploitable. A number of pilot projects are being built worldwide, but it is not currently a significant energy source either in the US or elsewhere.

Hydrogen Fusion

Hydrogen Fusion is a form of Nuclear Energy that obtains energy by combining two Hydrogen molecules (four Hydrogen atoms) to form one Helium atom plus energy. Hydrogen is very abundant. The amount of energy released by fusion is very large. Controlled Hydrogen fusion would provide virtually unlimited (but not necessarily cheap) energy.

Unfortunately, Hydrogen fusion is very difficult to control and exploit in small scale operations like a power plant. Despite decades of research the current situation is that Hydrogen fusion has been demonstrated in the laboratory, but never on a scale that produces enough energy to keep the reaction going, much less tap the energy produced as a power source. The next step is to be a multi-billion dollar multination research effort called ITER that may reach the "break-even" point around 2018. Even if ITER succeeds, an additional two decades or more will probably be required before the first Hydrogen Fusion power plant comes on line.

As a practical matter, the fuel for fusion power plants should they prove practical may be relatively rare isotopes of Hydrogen called Deuterium and Tritium. The supply of Deuterium is very large, but very diffuse. The costs of obtaining Deuterium fuel may be substantial and production may require substantial facilities. That is even more true of Tritium which basically is made via Nuclear chemistry. We will not run out of these Hydrogen "fuels", but obtaining them in the volumes needed for power generation may be neither cheap nor easy.

The danger of a nuclear fusion plant "melting down" is probably zero. But they can, and probably will, fail in fashions that can release radioactive material and/or raw energy -- they can blow up. Radioactive waste will be created, but not in the volumes produced by existing power plants using nuclear fission. The fuel can be used to make Hydrogen bombs, but only by countries that can build bomb triggers which currently requires a nuclear weapon. Over all, it appears that nuclear fission -- if it can be tamed -- will be a stable, reliable, safe, and minimally polluting power source with effectively unlimited fuel availability.

Stationary Turbines--Low Head Turbines

This section -- if I ever work it out will deal with hydroelectric generation from unconventional water sources. For the time being, let's just point out that household and very small community hydro electric power development is possible. Such systems do exist and are a viable alternative to solar and wind power in some rural locations. Low head generation for public power systems is addressed as part of the major section on hydroelectric power.

Osmotic Power

Osmotic power is a label for various techniques for exploiting the fact that water containing salts is denser than fresh water. This difference can be exploited to generate power when the less dense fresh water is mixed with saltier water. In the case where rivers are the source of the fresh water, the mixing would take place anyway, so the power is essentially free -- other than the possibly considerable cost of the generation facility.

Osmotic power is renewable and does not release significant carbon. It may however be disruptive to the environment where many species that currently live in the brackish water transition zone between fresh and salt water might be threatened.

The amount of power available is unknown with some sources claiming a potential for 1.6 to 1.7 Twh (over what time span?) This appears to work out to a "sticker" capacity of about 900GW (Note: These numbers do not appear to make sense), but it seems fairly unlikely that actual generation will reach anything like those levels. A different source quotes a potential continuous 100MW generation from 50m^3 fresh water flow. Currently there is a single 4kw pilot plant in operation in Norway.

Other Issues and Concerns


Energy Return On Energy Investment (EROEI) is just a number indicating how much Energy is required to extract Energy from some source. EROEI is related to costs of course. A high EROEI indicates that recovering energy from some source is cheap. A lower EROEI indicates higher costs. An EROEI of 1 indicates that it takes a btu to produce a btu from a given source making production kind of pointless. EREOI is important as it is a good idea to derate reserves by the energy cost of exploiting them. For example, a tar sand with an EROEI of 2 will need to produce two barrels of oil for each barrel delivered to a customer -- one barrel of product and one barrel to extract the product. So the amount of the apparent reserve should probably be halved.

Measuring EROEI isn't always easy. Some "popular" energy sources like corn Ethanol may actually have on EROEI near or even less than 1.0 when all energy involved in production and delivery are tabulated. Not surprisingly that true EROEI of such sources is subject to a lot of debate. Non-renewable resources with low EROEI are possibly best left to the future when the EROEI for the resource may be more favorable because of technology advancements.

It is important to keep in mind that from an efficiency viewpoint an EROEI of 100 is not twice as good as an EROEI of 50 because efficiency ((EROEI-1)/EROEI) is above 90% for EROEI values above 9 and efficiency losses only become really dramatic below EROEI 5.0. (efficiency 80%) See

Radioactive Waste

Radioactive waste is produced by all fossil fuels, by some industrial processes, by some medical procedures, by nuclear weapons production, and of course by nuclear power plants. Perhaps paradoxically, wastes from the nuclear power industry are better controlled and less likely to enter the environment than those associated with gas, oil, and especially coal.

There are a number of categories of nuclear waste which we can simplify to three. Low level waste is mildly radioactive. It is not very dangerous to those casually exposed, but too radioactive to expose workers and residents to on a daily basis. There is a lot of that. It needs to be sequestered but can be handled much like any toxic contaminant. High level waste falls into two categories. One is highly radioactive, but short half life, materials that will simply go away if the material is sequestered for a while. The other is a small number of longer half life isotopes that need to be isolated for centuries or millenia. A great deal of effort is being expended on ways to secure this stuff and make sure it stays secure. Fortunately there is not a lot of it.

Thanks to the diligent efforts of a lot of confused people, the US does not have adequate storage for either low or high level waste. The Yucca Mountain repository seems to be dead thanks to probably specious concerns about the geology and somewhat more realistic concerns about the safety of transport of material to the site. No alternative is under serious consideration.

Pumped Storage

Many power sources -- wind, tide, solar, stationary turbines in rivers deliver power when they want to rather than when grid managers would ideally want it delivered. Pumped storage stores electric power by pumping water to an elevated storage facility from which it will later be released to drive turbines. Pumped storage loses 15%-30% of the input power.

There are more than 30 facilities with 19.5GW of pumped storage capability in the US although not all are actually used. As currently used -- to pump water into storage at low off-peak rates in order to produce power later at higher peak rates -- these facilities appear to be a drain on energy resources rather than an asset. However, as more transient -- use it or lose it -- energy sources like wind and solar power come on line, pumped storage may become beneficial or possibly even necessary.

The principle problems with pumped storage are electrical losses in pumping the working fluid, more losses in generating power from the pumped fluid and evaporation/leakage of the pumped fluid. A specific problem for solar power is that the regions of the US with maximum solar potential are the Southwest desert where potable water is scarce and overcommitted. Conventional pumped storage is not particularly good for the environment as it results in continuously varying water levels at the storage facility.

Another issue with pumped storage is the rather impressive amount of water that must be pumped in order to store significant amounts of energy. It appears that storing one kilowatt hour of energy requires lifting one cubic meter of water 100 meters. (I'm a bit shakey on that number, but not on the message. Pumped storage requires pumping a LOT of water). It is not clear that adequate storage sites are available for significant pumped storage except in thinly populated areas with mountainous terrain. Such sites are difficult to identify in Europe, East Asia, or the Eastern US. One possible exception would be the North American Great Lakes where the surface areas are substantial and the elevation difference between the upper lakes and Lake Ontario is about 100 meters. Even there, the capacity would probably be limited by the tremendous number of (quite possibly meritorious) lawsuits engendered if lake levels were altered by more than a few centimeters.

One additional problem is that while pumped storage using existing reservoirs and generation facilities does store power, it may not increase baseline load capability if the generators are already being used to provide baseline load capability.

Other Storage

Molten salt storage is quite far along, as is Concentrated Solar Power (CSP). Another storage solution is to compress air, store it in old mines and natural caverns in the earth (as is done with natural gas) and release it to drive flywheels to generate electricity when needed. Ice created during low load periods is occasionally used to "store cold" for air conditioning. This does nothing for efficiency, but it does reduce maximum power grid loads somewhat. Large capacity batteries are coming along, as well. One of the California electricity utilities has posited that nickle metal hydride batteries from end-of-life Priuses would be fine in large banks to store electricity to ease peak load problems. Sodium-Sulfur batteries are capable of storing significant amounts of power. In April 2010, a 640MwH Sodium-Sulfur battery was brought on-line in Presidio, TX.

There are inefficiencies in "charging" and "discharging" energy storage facilities. In most cases, some energy leaks out during storage.

Transmission Lines

As a practical matter almost all electricity is generated someplace other than where it is used. Power is transmitted between locations using transmission lines -- often interconnected into a "power grid". Power is generally produced at fairly low voltages (less that 30,000 volts= 30KV), but is stepped up to hundreds of thousands of volts for long distance transmission. Raising voltages lowers currents resulting in less power loss for a given wire size. Traditionally Alternating Current (AC) has been used for transmission because it is easy to alter the voltage of AC. Direct Current (DC) transmission is occasionally used in modern systems even though voltage conversion is more troublesome for DC than AC. DC systems have lower peak voltages for the same power level and have other advantages that can overcome the greater difficulty of voltage translation.

Transmission systems are not 100% efficient. Losses occur in voltage translation devices, in the wires themselves, and also -- with AC -- because of "reactive" losses in dealing with the constant shifting of transmitted energy into and out of electrostatic and magnetic fields as well as the need in a transmission grid to combine AC power from sources that are not in phase ("reactive losses"). Reactive losses are not very intuitive, but they can be significant.

Numbers for power transmission and distribution losses seem not to be especially easy to come by and I suspect they may not be all that accurate. They vary with the length of the transmission line. For now, the best I can do is losses between 5% and 8% depending on the characteristics of the individual transmission arrangement.

It is worth pointing out that for practical purposes, North America has two power grids -- one including the East Coast to the Central Plains and one including the West Coast and reaching to and across the Rockies to the Rocky Mountain foothills. The interconnections between the two grids are currently minimal.

Base Load

There are two different definitions of 'base load' in common use:

There are two problem areas here:

Energy Availability (Sticker Ratings)

A point that is frequently overlooked is that some power sources are more equal than others. In particular the sticker ratings of renewable energy sources are generally maximum output values (i.e. Capacity ratings) that will be achieved only occasionally when the stars are aligned and the force is with us. A 1 megawatt wind power facility will almost certainly produce much less energy over the course of a year than a 1 megawatt hydroelectric, nuclear, or coal plant because the wind blows when it wants to not when the user wants it to.

This is handled by a power availability factor. Availability factors for base load facilities are typically around 70-90%. Availability factors (a.k.a Capacity factors) for wind and solar are much lower -- 20%-35%. One caution, availability can also be computed based on whether the facility is down for maintenance or otherwise unusable rather than power delivered divided by facility capacity. Regrettably, this number is also called an "availability factor." Don't confuse the two. Reality is that one needs 2 or 3 Megawatts of wind or solar "capacity" to match the day-in, day-out power generation of one megawatt of coal, gas or nuclear capacity.


Fracking is a process that recovers hydrocarbons (oil and gas) from rock formations from which natural flows are low to non-existant. There are a lot of those formations -- typically dark shales deposited in marine basins. In fracking, wells are driven into the target rocks then extended horizontally. The well is then pumped full of a fracking fluid which is subjected to great pressure in order to fracture the rocks adjacent to the well. The fracking fluid includes supended "sand" particles that hold the newly created fractures open even after the fraking fluid is removed. The hydrocarbons then flow through the now somewhat porous rock and can be recovered.

Fracking has made large quantities of natural gas and smaller quantities of oil available from deposits where production is otherwise uneconomic. The natural gas from fracking is quite cheap. The oil is fairly expensive -- around $70 a barrel.

Problems with fracking include high cost, short lifetime of fracked wells, disposal of possibly toxic fracking fluids, and the potential for contaminating aquifers. It is unclear whether all/most/any fracking fluids actually are toxic. The principle ingredients -- water, sand, salt, dilute citric and hydrochloric acids -- are not. There is a great deal of opposition to fracking typically from environmentalists who favor renewable energy sources that simply don't work very well and probably never will unless/until adequate energy storage technologies are developed.

Solar Electric Conversion Efficiencies

Although solar power is plentiful and free, it is rather diffuse. The theoretical maximum is around 1000 watts per square meter with the sun directly overhead on a cloudless day. However, the sun is never directly overhead outside the tropics. The actual power available from the sun is a complex function of length of day, sun angles, cloud cover, dust/water accumulation on the collector, ambient temperature and the conversion efficiency of the solar device.

Solar-electric conversion efficiencies are typically around 10-20% give or take a lot. But that is just for the solar cell. Additional corrections (all negative) are also needed for sun angle, cloud cover, dust on the collectors, etc, etc, etc.

There is little doubt that solar electric power generation can be practical in off-grid locations or in situations where the consumer is allowed to pump excess solar generated power back into the power grid at retail rates (which is clearly not economically viable on a large scale--who pays for power delivery infrastructure/maintenence?).

For a mature technology, it would appear that a 100 square meter building (about 1100 square feet) in the Northern US with its roof covered with solar cells and with efficient, reliable energy storage and voltage conversion devices might generate 40kwHours an a good Spring or Fall day. Since not all days are good days, 20kwHours a day might be a likely practical number. Enough to support a typical suburban house if one does not heat with electricity.

A power grid would still be required to support industry, high density housing, apartments, and periods of lousy weather. Presumably this need could be met by large scale solar energy facilities in the desert SouthWest. But a lot of R&D and infrastructure investment would be required.

Economic Issues

There appears to be an economic disconnection between renewable energy sources such as wind and and the cost of maintaining power flow to users. The problem is that wind farm economics typically assume that the power grid will be able to accept all of the energy produced whenever it is produced and that alternate sources/storage facilities can shut down when renewables are flowing and power back up to fill in when the wind is not blowing and the sun is not shining. Modeling seems to indicate that once renewables are widely used this is only true if the power grid and overall generation system includes considerable rarely used capacity. The problem becomes substantial and important as renewables contribute large portions of overall power since the hardware needed to provide the rarely needed extra capacity has to be paid for somehow. The result would appear to be that the delivered cost of large amounts of renewable energy is likely to be two times or more the cost of small amounts

Solar Power Effectively Unlimited

It is sometimes stated that humanity's entire energy needs for a year can be met by the energy that falls on North Africa in a single hour. This appears to be overstated by a factor of about 30, but it is true that the amount of energy that reaches the surface of the Earth is much higher (by a factor of about 6000) than the total energy from all sources currently used by the human race. The problem is that the solar energy is dispersed over an immense area. The effective cross section of the Earth is around 50,000,000 square miles. At 10% conversion efficiency, we would need multiple power plants covering 70000 square miles each to power the planet. ... and we'd have problems on cloudy days.

Saying that the energy available in an hour can power the planet for a year is somewhat akin to saying that the nuclear energy available in a ham sandwich can power the US for a week (A 'statistic' I made up). In point of fact we have no idea how to collect and utilize more than a minuscule fraction of solar power delivered to the surface just as we have no idea -- after seven decades of research -- how to industrially extract nuclear energy from a ham sandwich or anything else other than a few rare radioactive heavy metal isotopes and possibly -- in another few decades -- a couple of rare isotopes of Hydrogen.