Sunday, July 27, 2014

Renewables have higher ERoEI than fossil fuels

One the central claims of the peak oil/energy decline movement, is that renewable sources of power have extremely low ERoEITherefore, it is claimed, renewables are no substitute for fossil fuels, because they cannot provide enough “net energy” to power civilization. In support of this claim, energy decline adherents often post graphs like this one, showing that renewables (especially solar PV) have low ERoEI compared to fossil fuels. More recently, Hall and Prieto have published a book, Spain's photovoltaic revolution, in which they claim that the ERoEI of solar PV in Spain is only 2.45, which is far lower than the ERoEI of fossil fuels.

In fact, those claims are entirely wrong. Fossil fuels have ERoEI ratios which are very low, and renewables have ERoEI ratios which are generally comparable or higher. Although peak oilers reach a different conclusion, that is because they are carrying out the calculation incorrectly. They are ignoring or not including massive energy losses of fossil fuels, such as massive losses from combustion engines (generally 60% or more) as waste heat. Those losses provide no energy services to society, and should be counted as losses and not as "energy returns".

In other words, the notion that renewables have ERoEI ratios which are lower than fossil fuels, is simply mistaken. It arises from performing invalid, apples-to-oranges comparisons, or from not counting energy losses of fossil fuels.

Fossil fuels have very low ERoEI ratios

Take this graph as an example. It compares the ERoEI of solar PV for electrical power, against the ERoEI of coal and gas for heat. That comparison is invalid, because it’s an apples-to-oranges comparison. Thermal power plants (like coal-burning plants) waste approximately 2/3ds of their energy as waste heat. Waste heat is radiated out into the atmosphere from the power plant, and provides no energy services to society. This massive energy loss from fossil fuels is not counted in that graph of ERoEI, thereby artificially inflating the ERoEI of fossil fuels. If we subtract the energy losses from conversion of thermal energy to electricity, then the ERoEI of fossil fuels declines by approximately 2/3rds relative to solar PV. Conversely, we could also increase the ERoEI of solar PV by approximately 3x, thereby providing an energy quality correction. As a result, the ERoEI for thermal power plants which generate electricity is approximately 2/3rds lower than the graph indicates, or (conversely) the ERoEI of solar PV is approximately 3x higher.

It’s simply meaningless to compare the ERoEI of electricity generation from renewables, against the ERoEI of heat from fossil fuels, because heat is an extremely low-quality kind of energy which is far less capable of performing work. This is an elementary principle of thermodynamics. In order to convert heat to work, we must lose the vast majority of that heat as waste. For example, the vast majority of energy from fossil fuels is simply rejected as waste heat from power plants or internal combustion engines, and so shouldn’t be counted as an “energy return” in ERoEI calculations.

In general, the ERoEI of fossil fuels is extremely low. Natural gas may have an ERoEI of 10, but that falls to 5 when considering the massive waste heat losses emitted from natural gas turbines (generally less than half of the energy in gas is converted to electricity). Coal may have an ERoEI of 30, but that declines to 10 when considering that coal power plants lose approximately 2/3rds of the energy of the coal as waste heat.

The ERoEI of oil is particularly low because it's used in inefficient internal combustion engines inside of vehicles. Most car engines lose about 80% or more of the energy from gasoline, as waste heat, when you include both engine and transmission losses. As a result, the ERoEI of energy which actually turns the wheels of the car (rather than heating the outside atmosphere) is not 14.5 for oil, as commonly claimed, but only 2.9.

Renewable sources of energy do not suffer from those tremendous losses. Although renewables sources of energy do suffer from power grid losses, those losses are minor (usually less than 5%).

As a result, the ERoEI ratios of renewable sources of power are often much higher than their fossil fuel counterparts. Wind turbines have an ERoEI of 18, compared to 10 for coal or 5 for natural gas. Solar PV panels powering battery-electric cars have an ERoEI of about 7 (deducting grid losses and recharging heat losses), compared to 2.9 for oil in gasoline-powered cars.

Incidentally, the extremely low ERoEI of oil for driving cars and trucks (2.9), refutes the notion that an ERoEI less than 8 would lead to the collapse of industrial civilization. That claim is extremely common in energy decline circles, but it was pulled out of thin air and was wrong to begin with for several other reasons. In fact, modern industrial civilization has been growing for decades (especially China and Korea) with ERoEIs far lower than 8.

Hall and Prieto’s criticism

More recently, a book by Hall and Prieto, has become all the rage in energy decline circles. That book claims that the ERoEI of solar PV is grossly exaggerated. Hall and Prieto adjust the ERoEI of solar PV downwards, by adding all kinds of incidental energy costs. They add every incidental energy cost they can think of, like the energy costs of building fences around the solar farm, and so on. They even add energy costs for things like corporate management, security, taxes, fairs, exhibitions, notary public fees, accountants, and and so on (monetary costs are converted into energy by means of a formula). Sometimes, their estimates of those costs are absurdly high. According to Hall and Prieto, the ERoEI of solar PV is only 2.45 when all those things are added.

Once again, the calculation is incorrect, and the comparison is invalid. Hall and Prieto are adding every incidental energy cost to solar that they can think of. However, such energy costs are not included in the ERoEI calculations of fossil fuels. For example, the ERoEI of oil does not include the costs of security in the middle east, or the costs of pipelines, tankers, tanker trucks, road wear from tanker trucks, construction of gas stations, energy costs of driving to the gas station to refuel, the highway patrol, and countless other things. If those costs were counted, then ERoEI of oil (which is already low, at 2.9, when including waste heat losses) would only decline further.

It's necessary to perform an apples-to-apples comparison here. If we're going to add up every incidental energy cost of solar PV, then we must perform the same procedure for oil. Only then would we have a valid comparison.

If you carry out a detailed accounting procedure for both solar and oil, then the ERoEI of oil will be even lower in comparison, than it already was. The incidental costs of oil are almost certainly higher than those for PV. Whereas oil is a scarce substance which requires massive extraction and transportation costs, silicon is the most abundant mineral in the Earth’s crust (sand, rocks) and does not require expensive or elaborate techniques of extraction or transportation. Whereas oil comes from unstable regions and requires massive security and military costs, silicon requires only a few security cameras. Whereas oil is subject to ongoing transportation costs, silicon needs to be transported only once during the lifetime of the solar cells. In general, the incidental costs of oil are far higher than those for solar PV. As a result, if we include those incidental costs in both cases, the adjusted ERoEI of oil will be even lower in comparison than it already was.

Again, when you perform valid, apples-to-apples comparisons, the ERoEI of solar PV is higher than that of oil or natural gas. Oil for transportation in cars has an ERoEI of only 2.9 (because of waste heat losses), but that is before we include incidental costs such as security, infrastructure, and so on, so oil’s total ERoEI would only decline, and would likely be lower than 2.

Hall and Prieto’s analysis is mistaken in other ways. Their estimate of 2.45 for PV is certainly far too low. They include things like taxes and land leases, which are not energy costs, but which are just redistributions of money. Taxes provide services for society, so they should be counted as energy returns, not energy costs. If taxes in Europe on gasoline were counted as an energy cost, then the ERoEI of oil there would certainly fall to below 1. Also, Hall and Prieto include massive energy costs for premature retirement of solar cells because of rapidly advancing technology, but those cells won't be prematurely retired because they are paid for in advance and almost free to operate at that point, regardless of their efficiency compared to newer panels (newer panels would simply be added for future projects). Also, Prieto and Hall include things like administrative expenses, employees’ salaries, and so on, using a formula for converting dollars to energy which is far too high and is just wrong. You would obtain a far lower figure by 
converting salaries to energy using a more reasonable formula, of dividing the entire energy expenditure of a country by its entire GDP in order to obtain a conversion factor.

A correct calculation of the the ERoEI of solar PV including everything, would be more like 6, not 2.45. You can derive this figure by removing everything from Hall and Prieto’s analysis which is not an energy cost (such as taxes or land leases), and by using a more reasonable formula to convert monetary costs to energy.


In short. Renewables generally have higher ERoEI ratios than their fossil fuel counterparts. When you carry out a valid, apples-to-apples comparison, the ERoEI of renewables is generally better. This is because the ERoEI of fossil fuels is actually very poor--generally less than 5--when you correctly subtract the massive waste heat losses of combustion engines, and also subtract the massive incidental costs (such as security costs) of fossil fuels.

The only circumstance where fossil fuels have a higher ERoEI for renewables is when generating heat for smelting of ores or making cement or glass. That’s because such applications do not take place inside inefficient combustion engines, and so don't require subtracting the enormous waste heat losses of such engines. As a result, such applications still favor fossil fuels. Coal has a much higher ERoEI for this purpose than solar thermal plants, and (more importantly) is much cheaper. However, those uses are only a small fraction of total energy usage. Those uses will probably be the last energy uses which are converted from fossil fuels to other sources of energy, possibly more than 100 years from now.

Not that ERoEI matters much anyway. The whole idea is a mistake. What matters is the cost (in money) of net energy, for an energy source. If the cost of net energy is low, then the ERoEI is just totally unimportant. For example, if it were possible to build a 1 GW fusion power plant very easily out of duct tape for only $10, then it wouldn’t matter at all if it had an ERoEI of less than 2. We could just build more of them, and thereby produce the same amount of net energy as a higher-ERoEI (but more expensive) energy source. As long as an energy source has an ERoEI higher than 1, the ERoEI ceases to matter, and what matters is the total cost of net energy. This is discussed further here.

Monday, July 22, 2013

Being A Peak Oiler Means Never Having To Say You're Sorry

It has been about 15 years now since Colin Campbell and Jean Laherrere published their famous article in Scientific American, entitled "The end of cheap oil?" in which they predicted that oil supplies would soon peak and then decline. That article kicked off the modern peak oil movement1. So we can say that the peak oil movement is approaching its 15th birthday.

With that in mind, I think we should have a little retrospection. The peak oil movement1 is old enough now to reflect upon its history, its prior predictions, and its track record.

The peak oil movement has produced a long series of predictions, such as:

  • Production of all liquid hydrocarbons would peak in the mid-2000's or shortly thereafter (predicted by Campbell, Deffeyes, Laherrere, Aleklett, and may others)
  • Production of all liquid hydrocarbons would decline almost immediately thereafter, at about 2% per year (Campbell, Deffeyes, Laherrere, and many others)
  • Unconventional sources of oil, such as fracking, tar sands, and so on, would make little difference, and would not significantly delay the decline (Campbell, many others)
  • Oil production in the Kingdom of Saudi Arabia (KSA) would peak and then start declining in the mid-to-late 2000s. (Simmons, Stuart Saniford, WestTexas, and many others)
  • Natural gas production from all sources (including unconventional gas and fracking) would peak before 2010, and would immediately start declining rapidly (Simmons, Hughes, ASPO newsletter, ASPO Ireland, and many others)
  • Coal would peak around 2011 (Patzek)
  • Civilization would collapse (Savinar, Duncan, Heinberg, Kunstler, Orlov, Hansen, Ruppert, McPherson, and many, many others)
  • Global trade would end or be severely curtailed. Agriculture would re-localize. (Rubin, Kunstler, many others)
...and many other predictions, many times.

All of those predictions have been wrong. The peak oil movement has nearly a 100% failure rate of prediction, across many years, across an astonishing variety of domains.

What's more, many of the predictions of the peak oil movement, have not only been wrong, but drastically wrong. Predictions about a "natural gas cliff" in the late 2000's; predictions about civilization rapidly collapsing, and then reverting to a pre-industrial state; predictions that fracking would never work, or would amount to almost nothing; predictions that ocean shipping would soon end; and so on. Predictions like those were broadly shared within the peak oil movement. They were not only wrong, but drastically wrong.

Failed predictions are not the biggest problem of the peak oil movement, however. The biggest problem is how they respond when one of their predictions fails.

Every time a peak oil prediction fails, which is often, the prediction is immediately forgotten about within the peak oil movement, and is never mentioned again. There are no questions about why the prediction failed. Nobody ever asks, "what went wrong?" There is no retrospection. The theories are not modified in light of new evidence. Every time a prediction fails, peak oilers immediately forget about it and then just issue a new prediction; and when that fails too, which inevitably happens, it too is forgotten. Sometimes a few peak oilers will try vigorously to change the subject, when asked why a prediction failed, or will claim implausibly that the prediction had never been made. They do not, however, address the issue. Clearly, peak oilers are capable of forgetting.

The problem is, that nobody outside of the peak oil movement, forgets. People outside of the peak oil movement, are not part of the unspoken consensus to forget every failed prediction. They wonder why these predictions have failed. When they see the leaders of the peak oil movement simply dodging this issue, over and over again, it makes the movement appear like quackery. By forgetting their failures, the peak oil movement may comfort its members, but it further alienates itself from the wider public.

An example of this, is a recent post on The Oil Drum. The Oil Drum is shutting down at the end of this month, after 8 years of service. It's authors claim that the reason for it shutting down has nothing to do with the decline of readership after repeated failed predictions. No. Instead, one of the editors of The Oil Drum earnestly claims that they decided to shut it down because they've been so right that it would be pointless to continue repeating such correct things:

"The facts, in neither case, change, but the amount of new information while accumulating ... is often repetitive or confirmatory of earlier stories and thus harder to turn into interesting and exciting new material"

I don't feel that is an accurate recounting of the history of the peak oil movement. Nor is it an accurate recounting of the past content on The Oil Drum, or other peak oil websites. If I went back to 2007, I could find a fairly broad consensus, in peak oil circles, that industrial civilization was about to collapse and revert to a pre-industrial state. It's not clear to me, at all, that current stories are confirmatory of that. Even if I ignore the doomsday prophecies, and focus only upon the more sober elements of the peak oil movement, I come across graphs like this one (from Colin Campbell), showing that production of all liquid hydrocarbons should have declined by 25% by now.

This issue of the "memory hole" is very important, because of frequent claims by peak oilers that their movement is a science. Peak oilers frequently claim that their theories are scientific, that their predictions are based upon science, that their conclusions are derived in a straightforward fashion from the laws of thermodynamics, and that their theory is akin to climate science (people deny it because they're misled).

I must remind peak oil adherents that being scientific has nothing to do with just throwing around scientific-sounding terms. Nor does it have anything to do with mentioning the word "thermodynamics" frequently, or drawing tenuous analogies between themselves and climate scientists. Actual science requires falsifiable theories, something which the peak oil movement conspicuously lacks. In science, if a hypothesis fails in its predictions, completely, over and over again, then the hypothesis is wrong and must be modified or abandoned. You cannot just use the same hypothesis to issue another prediction (like, peak oil is perpetually five years in the future) because that method lacks any criterion of falsification. That method could simply be repeated, indefinitely, until peak oil actually occurs, whenever that will be.

It's not useful to retreat into saying "oil must peak some day" and call that a "fact". That is the method employed most recently by many peak oil writers. They say that peak oil is a "fact" because oil must peak some day. While true, that claim is obvious, extremely imprecise, and not a useful prediction. Even the major oil companies, such as Exxon, Shell, and Total, acknowledge that oil will peak some day, have always acknowledged it, and realized it before the peak oil movement ever came along. It's simply repeating a truism to say that oil must peak some day. It's simply a case of forgetting their earlier predictions, and then making their subsequent predictions less and less precise, until what they are saying is so general that it says almost nothing.

The question is whether the peak oil movement has been correct about any specific prediction, or anything which others denied. On the points upon which peak oilers differed from anyone else, they were mistaken. Although oil will certainly peak some day, and then decline, still, the methods from the peak oil movement for predicting when that will occur, how rapid the decline will be, and what will happen to civilization more generally, have been totally incorrect.

The peak oil movement is not science--quite the opposite. It does not contain any valid, specific, predictive hypotheses which were confirmed by subsequent evidence. It has not been honed in light of new evidence. It contains mistaken theories and methods, which have not been abandoned. Thus, the peak oil movement cannot be called science.

The peak oil movement is an ideology which is committed to a specific conclusion, no matter what. As such, the peak oil movement is more like an ideological group. It pretends to be scientific in order to comfort its members, but its pretensions to science are totally superficial. It lacks anything resembling a scientific method.

I realize these points must be very difficult for peak oilers to accept. Many peak oilers have devoted a considerable fraction of their lives and their resources, to the conclusion that oil would soon peak and civilization would soon falter. It must cause them significant suffering to consider that everything they thought about this topic, everything they spent so much time upon, has been wrong. That goes doubly for the many peak oilers who expressed very little uncertainty about these predictions.

Why, then, am I writing this? Do I want to embarrass peak oilers?

It has been fifteen years now since the peak oil movement began. That is long enough. It's time to acknowledge that there are serious problems with even the most fundamental tenets of peak oil theory2. I am not saying that the entire thing has been a waste. I'm sure many people learned a great deal about oil extraction, energy, and other topics, from reading The Oil Drum. However, it's time now to acknowledge that peak oil theory was wrong. It's time either to move on, or to make serious modifications to the theory, and to explain why its predictions have failed and why the newer theory avoids the prior mistakes.

There is a peak energy blog called "question everything", which has the following byline:

"When what is happening in your world doesn't make sense, when it doesn't conform to your beliefs about how things should work, it's time to ask hard questions."

Indeed, it's time to ask hard questions. It has been time to ask hard questions, for at least the last five years. So far, those hard questions have not been asked. Better late than never.


1. By "peak oil movement", I'm referring to the movement unintentionally started by Colin Campbell, Jean Laherrere, and so on. The movement was based upon the theory that global hydrocarbon production could be predicted using extrapolations of curves, such as creaming curves, Hubbert linearization, and so on. Using such curves, the movement predicted that global hydrocarbon production would imminently peak and then decline, which would cause severe disruption to industrial civilization, such as re-localization, and so on. The movement mostly congregated on websites such as "The Oil Drum", "", "Life After The Oil Crash", and so on. That's who I'm referring to, when I say "peak oil movement". I'm not referring to everybody who believes that oil will peak some day.

2. By "peak oil theory", I'm referring to the theory that global hydrocarbon production could be predicted using extrapolation techniques such as Hubbert linearization, creaming curves, and so on. I am referring to the theory that peak oil was very imminent, and subsequent declines would be fairly rapid. I am also referring to the related belief that imminent peak oil would cause major disruptions to industrial civilization, such as re-localization, and so on. Those things are what I'm referring to, when I say "peak oil theory". I'm not referring to the idea that oil will peak some day, which almost everyone believes.

Tuesday, December 25, 2012

New Year Predictions


In this article I will offer some of my own predictions for what the future holds, with regard to energy supplies. I have three reasons for doing this. First, I wish to provide an alternative set of predictions which does not commit any of the errors of the peak oil camp. Second, I wish to provide a set of predictions which does not assume either business as usual OR collapse, thereby avoiding the "either-or" thinking prevalent in peak oil circles which posits that either business as usual must continue (impossible) or we must collapse to a pre-industrial state. Third, I wish to counteract the notion (very common in peak oil circles) that economists or students of economics assume that the world is infinite or extends to infinity in all directions, or that new resources magically appear as a consequence of demand. By offering these predictions, I hope to provide material which hasn't been considered by peak oil adherents, because of their unfortunate tendency to ascribe magic-infinite beliefs to all those who do not accept their conclusions.

Unfortunately, I cannot give all my reasons for believing the predictions I'll put forth here. Each prediction would require an essay-length treatment. For example, my prediction regarding ocean shipping would require an explanation of the basic physics of ships, of the alternative fuels available, of the time required to build a new fleet of ships, of capital expenses, of the methods that ship buyers use to calculate the size and speed of new ships, of basic optimization problems, and so on.

Although I can't give a fully detailed explanation of my reasons (because I don't write peak oil blogs for a living), I still can give my reasons in overview, or explain briefly why I think something is the case. For example, I can explain that ship builders could easily build ships that are more than 5x as fuel-efficient as today, but it only becomes cost-effective to do so when fuel is more expensive than now; that ship builders pick the most cost-effective ship given their anticipations about the future prices of fuel; that we have more than enough time to transition the shipping fleet to more efficient ships; and that these things imply that ocean shipping will not increase by more than 15% in the long run because that is how much the costs would increase with a tripling of oil prices, and if oil prices increased by more than that then shipping companies would switch to alternative fuels like anhydrous ammonia which do not require any fossil fuels. This degree of explanation is possible here.

I will offer basic reasons along with each of my predictions. In some cases, I will leave out the basic explanation when I feel it's fairly obvious.

With that in mind, here are my predictions for the future.


Oil production will peak and start declining some time in the next 12 years (before 2025). I gather this figure from Jean Lahererre. Please note that prior predictions from peak oil pessimists, and from people using linearization methods, have been quite incorrect. Nevertheless, I will make a worst-case assumption here, and will assume that peak oilers finally get it right. Once the declines begin, oil will decline at less than 1% per year for the first decade or more.

Similarly, coal and natural gas will both peak before 2100 and will decline gradually thereafter.

A few years before the peak of oil, speculators will detect that oil production is nearing its peak. The speculators will immediately bid up the price of oil to over $200/barrel*. This could happen fairly rapidly, like within 2 years. Speculators will bid up the price of oil to what they believe the longer-term price will be. Speculators do this because they make money by anticipating things, thereby moving forward the date of price increases, and prompting the transition to alternatives long before any declines have actually occurred. I'll assume (for simplicity and convenience) that gasoline will cost $7/gal* after at the pump after the initial increase, in the USA, and will cost more in European countries where petrol taxes are higher. It's possible that the price of gasoline will briefly go higher than this.

Once the price of oil has increased, the world economy will accelerate its gradual shift from oil to other sources of energy for transportation. Also, the world economy will accelerate its gradual shift to more fuel-efficient kinds of transportation. Ships will become larger and slower, and thereby much more fuel-efficient. Trains will displace trucks to some extent, and new track will be laid. Cars will become more fuel efficient. Plug-in hybrids will become common. Eventually, transportation will be electrified. This is because the entire economy automatically and always transitions to the next-best alternative when something becomes too expensive.

The price of oil will eventually level off and will never go higher than $200/barrel* for long, because prices higher than that will cause car makers to switch to battery-electric drivetrains, thereby driving the price back down. Higher prices will also cause cargo carriers to switch fuels, electrify, and switch modes (like from trucks to trains), thereby driving the price back down.

Once oil has been declining for a few years, further declines will be accompanied, simultaneously, by improved efficiency, fuel switching, and greater electrification. The long-term price of oil will hover around the cost of its next-best alternative energy source and carrier (batteries in cars), continuously, until oil is nearly exhausted, 100+ years from now. There may be temporary periods of higher prices, however the price of oil will always ultimately revert to the cost of its next best alternative energy source and carrier.

In the long run, the economy will adjust in the best manner possible to much higher oil prices. This is not the same as saying that resources are infinite. It implies that people will drive plug-in hybrids and will end up paying modestly more for personal transportation. Cargo shipping will cost slightly more. That is all.

More detailed predictions

I wish to go into further detail about these predictions. However, from this point forward it will be necessary to divide my predictions into SHORT-RUN and LONG-RUN predictions. By SHORT-RUN I mean that period of time which is shorter than the replacement time for the auto fleet (like 15 years), plus the amount of time necessary to shift production lines to more efficient cars. By LONG-RUN I mean that period after all transportation sectors have turned over their fleet of engines, ships, cars, trains, etc.

It's necessary to divide predictions into short-run and long-run predictions, because there will be a period during which we transition to a more efficient transportation infrastructure. There will be a period during the transition, and a period after the transition, and the economy will look different during these two periods. After declines have begun, however, car manufacturers will anticipate further declines.

There will be a transition period because people and firms will not correctly anticipate the date of peak oil or the amount of price increases. This is because the exact date of peak oil is essentially uncertain, and also because consumers do not anticipate things, but essentially just respond to changing circumstances as they occur.

The short-run effects of high oil prices

In the short run, the abrupt increase in oil prices, caused by speculators, will trigger a nasty recession.

The world economy will eventually recover from the recession and continue growing even though oil prices will never return to their prior levels, and even though oil production will never again increase to its prior levels. Economic growth does not require increasing oil extraction.

At that point, car manufacturers will scramble to increase production of ultra-light cars, hybrids, plug-in hybrids, electrics, etc. Car manufacturers will take more than 5 years to shift their production to more fuel-efficient cars. Car manufacturers will "aim ahead" and will transition to producing cars which they think are appropriate for the longer run, given further gradual declines in oil supplies. This is because firms attempt to anticipate, just like with the Y2K bug. They may be imperfect at this, but they will anticipate further declines once declines have begun.

During the transition to more fuel-efficient cars, many people will be stuck with old gas guzzlers which are then obsolete. There will be an "interim period" of about 10 years during which many people drive old cars which get less than 30 mpg. Some of these people will need to curtail discretionary travel, like long-distance road trips, because fuel will be too expensive. Those people will need to drive somewhat less until they can replace their gas guzzlers with new and more fuel-efficient cars. Eventually (within 14 years) most of the car fleet will have shifted to much more fuel-efficient cars, and to plug-in cars. At that point, if there are further declines in oil supplies which are more rapid than car manufacturers had anticipated, then a few consumers will have to curtail discretionary longer-distance transportation (in other words, they will be limited to the range of their batteries some of the time).

The long run-effects of high oil prices (30 years after the peak)

In the long run (30 years after the peak) the entire transportation infrastructure will have been converted to more efficient modes of transportation, and to alternative energy sources.

In the long run, $7/gal* gasoline will make very little difference for first-world living standards. The economy will adjust to higher oil prices. Cars will require far less gasoline to travel a given distance, and will be able to travel a considerable distance without any gasoline at all. When this transition has occurred, the nastiest effects of peak oil will have passed.

In 30 years, personal transportation will be modestly different. A few more people will take public transportation. Most drivers will have plug-in hybrids that look like priuses.

The total cost of automobile transportation will be slightly higher than today. After the car fleet has transitioned to plug-in hybrids, the average person will pay about $50/month* more for driving a car than they do now, even with $7/gal* gasoline. This figure of $50/month increase is derived from the additional cost which prius-like plug-in hybrids impose, assuming that battery costs come down somewhat as a result of both mass manufacture and technological innovation, and that gasoline costs $7/gal* and people drive 1000 miles per month, on average. (People will drive 750 miles per month using batteries and 250 miles using gasoline. The cars will get 60 mpg when using gasoline, so gasoline costs will be $30/month, plus another $40/month for electricity, which is much cheaper than we pay for fuel now. However, the savings on fuel will be more than offset by a $6000 increase in the cost of the car).

Cargo transport by sea will be change a bit. Shipping companies will switch to ships which are much larger, and which travel at 2/3rds the speed of current ships. Shipping companies will do this because they are always replacing old ships with the most cost-effective new ships for what they anticipate fuel costs will be for the next 20 years. These decisions will cause a 60% reduction in fuel consumption per tonne-mile, while increasing the capital expenses of the shipping fleet by a modest amount. The ultimate result will be a very small increase (less than 15%) in shipping costs, per tonne-mile.

The figure for cost increases (~15%) is easily derived by looking at the additional capital cost of ships which use 1/3rd the fuel, and assuming 3x more expensive fuel and slightly higher operating expenses. You can look up a few basic figures and then solve a basic optimization problem using Calculus 1 to figure it out. I have not bothered to do this, however I can "eyeball estimate" what the approximate result would be. Shipping companies will figure it out since they know more about the topic than you and I, and they carry out optimization problems like that routinely. Their price of fuel will never increase by more than 3x in the long run since synthetic fuels are then much cheaper, thereby encouraging switching and driving down the price of oil, and this implies that total costs for ocean shipping will not increase beyond 15% in the long run.

Shipping companies may also switch fuels at some point, to natural gas or to coal (with steam turbines).

Cargo transport by land will change a bit. More cargo will be transported by rail, and less by truck, because rail uses about 1/4th the energy per tonne-mile. New rail will be laid. Trucks may have 3 or 4 trailers and may travel at slower speeds, thereby reducing fuel consumption. Other innovations are possible, such as trolley-trucks. Some rail ways will be electrified. Ultimately, costs of cargo transport by land will increase modestly, but definitely not more than 25% per tonne-mile.

The increases in shipping costs (both land and ocean) will cause negligible increases in the cost of transported goods. For example, a pair of shoes will cost about $0.15* more. Food will cost very slightly more.

Plastic will become more expensive and rarer. Food packagers will switch to other materials, such as glass bottles, aluminum cans, cardboard boxes, etc. Clothing and toy manufacturers will switch to other materials, such as silicone, rubber, canvas, and other materials not derived from fossil fuels.

Air travel will be considerably more expensive than today, perhaps more than 40% more expensive. This will be the most visible and dramatic long-term effect of declining oil supplies.

Competition from Chinese people for oil supplies will cause far more rapid declines in the supplies available to Westerners during the next 20 years than any geological constraints.

Ultimately, we have the technology available to us today, to reduce fuel consumption by more than half without any reduction in miles traveled per person, or or any reduction in cargo tonne-miles. We could simply switch car production to prius-like cars (thereby doubling mpg) and use more trains and larger ships. The market economy will optimize a solution which is at least as good as this.

The very long-run effects of fossil fuel depletion (After all fossil fuels have been exhausted, 100+ years from now)

People will ride in battery-electric cars. Energy for the cars will come from solar power plants, wind turbines, and nuclear power plants.

Electric power generation will shift to nuclear plants and renewable plants. This will happen smoothly and without any major disruption, since the entire fleet of power plants will be replaced several times in the interim, and power companies order new plants based upon anticipation of the cost of fuel over the next 20 years. Prices for electricity will probably be lower than today, because renewable energy costs are decreasing over time and already are not much higher than prices for fossil fuel plants.

Total energy production worldwide will be far higher than today, despite a concomitant decline of fossil fuels. This is because China and India are growing rapidly and ultimately do not require any fossil fuels to continue that growth.

Ships will use anhydrous ammonia for fuel, I would guess (speculative). The ammonia will be derived from wind power, taken from stranded wind resources. Overall, ocean shipping will cost slightly more than it does today.

Truck transportation will be rarer. There will be far more rail in the world, and far less truck transportation. Very large retail outlets (like Wal-Mart etc) might have their own rail terminals (obviously this is very speculative).

Housing will have better insulation and will make greater use of "passive heating" techniques.

The world economy will have gone through many recessions. Each time, the economy will recover, and will continue growing despite a near-total exhaustion of all fossil fuels.

The average temperature of the surface of the earth will be 4 degrees centigrade warmer.

In 100 years, most people in the world will have first-world living standards, despite the exhaustion of fossil fuels. Fossil fuels are not needed for economic growth. In fact, fossil fuels are not needed for any purpose. They are the cheapest fuels right now. That is all.

No Disruptions Will Ever Occur

Most importantly. There will NEVER BE ANY MAJOR DISRUPTION to industrial civilization as a result of declining supplies of fossil fuels. More specifically, there will never be any major disruption to food supplies, food transport, electricity production, or any other essential thing, in any industrialized country, as a result of declining fossil fuel supplies.**

The economy transitions to alternatives when it's appropriate to do so. There are alternatives for every use of fossil fuels, and the economy will use them when the time is right. The economy transitions very reliably, like clockwork. It is always transitioning, even right now, and it will continue to do so as fossil fuels decline.

Long-distance transport of goods will never end. Long-distance transport consumes less than 5% of all oil supplies now, so it wouldn't be sacrificed during the next 100 years even if oil were the only possible fuel. However, oil is not the only possible fuel, since ships and trains can easily use other fuels such as synthetic fuels or electrification (for trains). As a result, there is more than enough time to transition (100+ years), and more than enough alternatives to do so.

Industrial growth will continue despite the decline of fossil fuels. This is because fossil fuels are not needed for industrial growth. Growth could occur with any source of energy which returns more than was required to obtain the energy. For example, it is entirely possible to use a single solar thermal plant to smelt the ores, melt the glass, and manufacture the synthetic fuels needed to build five more solar thermal plants, and so on. As a result, industrial growth does not require any fossil fuels and has never required any. Fossil fuels were selected first because they were cheapest.

The economy has more than ten times as much time as required to transition to alternatives, while avoiding any disruption of essential services. In fact, the economy could reduce its consumption of oil by 90% within 10 years without any risk of collapse. Transitioning that quickly would impose severe restrictions on personal travel in the interim (such as fuel rationing) until new vehicles were built, but would pose no risk of collapse. There are two reasons the economy would avoid collapse in this case: 1) the economy sacrifices the most important things last; and 2) the economy starts transitioning right away to alternatives. Thus, even if we underwent a 90% reduction in oil supplies over ten years, there would be absolutely no risk of collapse, because 10% oil supplies are more than enough to provide essential services (~3% for tractors, food transport, public transport) and to provide for the remainder of the transition at the same time (building an electric transportation infrastructure).

I realize this point requires far more elaboration. Look at it this way: look up the amount of energy used by truly essential purposes in the US economy (i.e. food production and transport, etc). Could you devise a way to transition to trolley-buses and trolley-trucks throughout suburbia while continuing essential services? Try to think of ways to avert collapse with 10% oil supplies. What if you banished all personal auto transport, except to and from work in carpools? What if you diverted all construction resources to rapidly putting up wooden poles (like old telephone poles) with cables strung between them for trolley buses, throughout suburbia? If this were done in parallel everywhere, and you employed 25% of the population in doing only that, how long would it take to have working trolley-bus service almost everywhere? How much steel is made right now, and how much electrical cable could be built per year for the trolley buses? If all lumber were diverted to poles, how many could we make? Wouldn't it be possible to convert the densest suburban areas to electrical transport first, within a few years, thereby freeing up some of the 10% oil supplies for additional purposes almost right away? Couldn't we replace the wooden poles with sturdier structures once the new infrastructure was going? Also, could we build natural gas cars for rural residents within 10 years? How many cars do we build now, and how different are natural gas cars? How long would it take us to build synthetic fuel plants, at the same time as our trolley-bus system? Could that also be done with the 10% oil budget? The answers to these questions are fairly basic. We could build enough of an alternative transportation infrastructure within a few years. We could produce enough steel, poles, and trolley-buses within 3 years to transition enough of the economy to an alternative transportation infrastructure. Not to mention, we could start mass-manufacture of natural gas vehicles almost right away since they are nearly identical to the vehicles constructed now. These sudden changes would be at massive cost, and tremendous inconvenience, and would require the abandonment of much capital equipment; but it would avoid any possibility of collapse. Bear in mind that the petroleum shortage would start easing as soon as any fraction of our hypothetical alternative infrastructure was complete.

If you can figure out a way to avoid collapse, off the top of your head, the economy will do that or something better. The economy is like a giant learning machine. It optimizes, evolves, and finds solutions. That is how our economy got to this point. That is how the network of rail lines, mines, factories, suppliers, sub-contractors, fiber optic lines, retail stores, etc, developed in the first place and operates now. If the economy were so stupid that it cannot manage a simple optimization and reallocation problem, then the economy would collapse within a month, regardless of peak oil. The problems entailed by just operating the economy now, are vastly more complicated and difficult than the problems of adjusting to peak oil.

Again, there will not be any disruption of essential services in first-world countries due to peak oil. Anyone who claims there will be, is badly misunderstanding the speed with which the economy could adjust, how it adjusts, the alternatives available, and the likely pace of fossil fuel declines. Any serious analysis of these matters would not permit even the minutest chance of collapse due to fossil fuel declines.

Alternative Possibilities

Of course, I can't really predict the future. I'm relying upon certain assumptions while making predictions, and those assumptions may turn out to be incorrect. If some of my assumptions are incorrect, then obviously my predictions above would be incorrect to some extent also.

My major assumptions are as follows.

First, I am assuming that we don't suffer unpredictable disasters like a nuclear war, or an emergent disease which kills of 90% of the population, or a meteor strike which decimates the biosphere, or any other similar disaster. Things like that are non-linear and essentially impossible to anticipate. Obviously if any of them occurred, then we could have an interruption to civilization. I am only claiming that fossil fuel depletion will never cause any interruption.

Second, I have assumed that oil will soon peak and decline. In effect, I have graciously granted this point, and am willing to accept the prognostications of ASPO and Jean Laherrere about future oil production. I should point out, however, that their prior predictions have been overly pessimistic, and quite incorrect. Also, their predictions are controversial. There are petroleum analysts who believe we'll be able to exploit the massive resources of unconventional oil such as tar sands and shale. Were this to happen, then petroleum prices might inch up gradually over decades (in fits and spurts, to be sure) and the economy would face only a gradual transition to alternatives, without the irritating "transition period" described above. In this case, auto manufacturers would have more than enough time to transition production to more efficient cars, and there would never be any irritating "transition period". In this case, my predictions above would be too pessimistic, and we might never face anything more severe than slight recession and modestly higher prices as we transition to alternatives to oil.

Third, I have assumed that there will be no major technological developments in the ensuing years. To be more specific, I have assumed that batteries and other energy storage devices remain at their present level of development. Clearly, this may not be true. Someone may invent the better battery one day, and I have no way of predicting whether this will occur. If a better battery were invented, then the economy could transition off of oil before the peak even occurs, in which case peak oil would be a complete non-event of no interest to anybody except toy manufacturers and airlines, who would then face a century-long transition to other materials or fuels. Or, if the better battery were invented after peak oil, then we could return to "energy guzzling" cars again.


  • Oil may peak and decline before 2025.
  • If that happened, oil prices would increase fairly rapidly to over $200/barrel.
  • In the SHORT RUN, this would cause:
    • A nasty recession
    • An "interim" period, during which some vehicles are "gas guzzlers" relative to what is then required. This "interim period" will last until the car fleet can be transitioned to plug-in hybrids.
    • People who own these "gas guzzlers" will need to curtail discretionary travel, and drive less, until they can buy a more fuel-efficient car.
  • In the LONG RUN, peak oil will cause:
    • Very little difference.
    • People will drive prius-like plug-in hybrids
    • People will pay about $50/month* more for car transportation.
    • Slight (meaning barely perceptible) increases in the cost of goods due to increased shipping costs
    • Plastic will be rarer for food packaging, cases for electronics, and toys. Other materials (silicone, aluminum, paper, cardboard, glass) will become more common for these purposes.
  • In the VERY LONG RUN, the exhaustion of all fossil fuels will cause:
    • Very little difference in day-to-day living compared to continued fossil fuels. Obviously there will be major technological changes in that time, but most changes in day-to-day living won't be imposed by declining supplies of fossil fuels.
    • Energy and transportation prices will be slightly higher _per person_
    • Energy will come from renewable sources
  • There will never be any major disruption, to any essential services, in any industrialized country, over any time period, due to declining fossil fuel supplies
    • The economy adjusts rationally to declining fossil fuel supplies
    • We have more than 10x longer than would be required to adjust to declining fossil fuels
    • Peak oil poses absolutely no risk of collapse
  • If APSO's projections about oil supplies are incorrect, and we exploit the large amount of unconventional oil resources, then the transition away from oil could be very gradual and may never pose any difficulty more serious that mild recession and modestly higher prices.

* Throughout this article, the prices quoted are in 2012 US Dollars. Obviously they could be much higher in the future because of inflation.
** Of course there will be transient disruptions to electricity grids, etc. There will be brown outs and other things, which happen all the time and have always happened. I'm claiming there will be no sustained, severe interruption.
*** NOTE: I made several minor modifications to this article as of July, 2013, as follows: I took a more negative stance on the prior predictions of ASPO etc in light of recent events. I also clarified the section about the interim period, by adding the word "discretionary".

Wednesday, January 4, 2012

Renewables do not require a fossil fuel subsidy

Recently I was reading the comments at the excellent blog, do the math, when I found this claim:

"Solar and wind capturing devices are not alternative energy sources. They are extensions of the fossil fuel supply."

This claim is similar to the "fossil fuel subsidy" argument, which crops up very frequently in peak oil forums. Basically, the argument is that renewables are "subsidized" by fossil fuels, because renewables are built using energy from fossil fuels. For example, windmills require coal to smelt the iron ore, to extract aluminum from their oxides, and so on. So it could be said that windmills were "subsidized" by coal, and could not have existed independently.

The argument is incorrect. While it is true that the first generation of rewnewable plants would have a "coal subsidy", any subsequent power plants would have a "renewable subsidy". That is because we build everything using the prior energy source. Once renewables are established, we will use them to smelt ores, extract aluminum oxides, manufacture parts for renewable plants, and so on. Thus we do not have a permanent fossil fuel subsidy; instead, we have a fossil fuel ladder, which we use and then kick away.

The transition from coal "subsidy" to renewable "subsidy" will happen automatically, as a result of basic market mechanisms. When renewable electricity and heat are cheaper and more prevalent than coal, they will also be cheaper sources of energy to manufacture subsequent power plants.

Of course, the transition away from the coal "subsidy" will not happen all at once. What really will happen, is that the first renewable plant will have a 100% coal subsidy for its construction, then each additional renewable plant will have a declining coal subsidy and increasing renewable subsidy, until the coal subsidy reaches zero.

As an example, look at early industrialism. The energy for early industrialism came from British coal. British coal “subsidized” the subsequent energy sources, even those in the USA. Does that mean we can never transition away from British coal? Does that mean all power is subsidized by British coal? Why didn’t civilization collapse as British coal declined?

Here is another example. The initial metals for early industrialism were smelted using charcoal, from WOOD. Thus, coal had a “charcoal subsidy”. Does it still have a charcoal subsidy? Do all subsequent power sources also have a charcoal subsidy? What about the subsequent nuclear plants? Do they have a charcoal subsidy? No. Charcoal was the first step; it provided a subsidy once.

Of course there is also the issue of oil. Oil subsidizes renewables because oil is used to power the mining machinery. However, even oil is replaceable with renewables. We can substitute batteries, or can manufacture hydrocarbons using the fischer-tropsch process. At some point this will be cheaper than diesel fuel from oil, because diesel fuel will become more expensive and batteries less expensive.

In conclusion. Fossil fuels are not necessary for any purpose. They were a cheap and easy first step; that is all. Fossil fuels are like a ladder we used to climb upwards to industrial civilization, but now we could kick the ladder away.

Sunday, November 6, 2011

We are not running out of resources


Some doomer authors have claimed that we're running out of resources in general. Not just fossil fuels, but many resources necessary for civilization, like metals, fertilizer, and so on, will soon be exhausted, according to them.

A notable example of this view, is the book Limits to Growth, which is perhaps the greatest doomer classic of all time. That book contains repeated claims that we will soon exhaust "minerals", "aluminum", and various other resources in the fairly near future. It shows graphs of depleting aluminum supplies and depleting supplies of various other things as time goes on, with near-zero rates of extraction for various resources by mid-century.

In this article I will argue that we have virtually inexhaustible supplies of everything necessary for modern civilization and for a large population. I will argue that we aren't running out of any essential resources or minerals, and that we don't face peaks of anything irreplacable. In fact, I'll show that we have enough resources to provide a large population with a first-world standard of living for billions of years, until our Sun explodes.

Note that I'm not claiming we'll never run out of anything. Clearly we will eventually exhaust our supplies of fossil fuels and other things. What I am claiming, however, is that we'll never run out of anything essential. By "essential" I mean resources for which there are no obvious substitutes, and which are required for modern civilization to exist with a large population. For example, iron, aluminum, fertilizer, and adequate energy sources are essential and have no substitutes. On the other hand, fossil fuels are absolutely not essential for civilization, since they're used as a portable energy source and they have many (albeit more expensive) alternatives. Fossil fuels are a cheap convenience, that is all.

Obviously there is some question of whether civilization has enough time to transition to other sources of energy and other minerals as fossil fuels and rare earths are exhausted. I will not address that question in this article. All I am attempting to answer here is whether there are enough resources to support modern civilization for 10 billion humans, and for how long, using any substitutes possible, even if that means we must revert to electrified public transportation like trolleys, and must manufacture hydrocarbons for the niche uses (like tractors) which can't be electrified. I will address questions of how quickly we can transition (and whether it's quickly enough) in a subsequent article.

In order to estimate whether we'll run out of anything essential, we must determine three facts: 1) which elements and substances are essential for modern civilization, in other words, which elements are needed and have no substitutes; 2) how much of them exist in the Earth's crust, in a form which could be economically extracted; and 3) how much of them we're using on a yearly basis, in other words, how quickly we're depleting them. Once we have those three pieces of information, we can project how long we have until we exhaust essential resources and civilization must end.

Of course it matters a great deal what future consumption patterns will be like and how large the population will be. A larger and richer population will exhaust resources more quickly. A population which is exponentially larger could obviously deplete resources very quickly. For this article, I am assuming that the human population will eventually stabilize at a high level (10 billion) as demographers predict, and those 10 billion people will reach European standards of living. Obviously, if the population continues to grow indefinitely, then we will eventually hit some limitation which constrains the further growth of population. Note that hitting a limit of population would happen as we ran out of one resource (probably arable land) which serves as a bottleneck. It would not necessarily cause an industrial collapse or a reversion to medieval mode of life or anything similar. It would probably cause starvation in the countries which were then poorest and had the highest population growth rates.

Which substances do we need for civilization?

Obviously we need various things for modern civilization. We need objects of civilization, like buildings, bridges, trains, and so on. We also need energy in large quantities, in order to build to objects of civilization and to extract the minerals necessary for their construction. We also need food (and therefore fertilizer) sufficient for 10 billion people.

However, just saying that "we need buildings" doesn't answer any questions. What are buildings made out of? We must decompose buildings into their constituent substances, and find out how much we have. Furthermore, we must also decompose the other objects of civilization (trains, computers, etc), into their constitutent substances. We must also find out what fertilizer is made from, and so on. In other words, we must decompose civilization, and find out which minerals and substances are absolutely required to build and sustain it.

Remarkably, almost everyting in civilization is made out of a very small number of elements, which are already mined in massive quantities. For the most part, civilization is built out of: iron, aluminum, silicon, oxygen, calcium (for cement), nitrogen, hydrogen, potassium, and phosphorous. Those elements are the "macro elements" of civilization, and I'll refer to them as such from now on, because they are required in huge amounts and they have no substitutes. Those macro elements are remarkably diverse and are sufficient by themselves to manufacture almost everything we need, including: cement, buildings, bridges, skyscrapers, roads, industrial machinery, power plants, trains, transportation infrastructure, generators, power cables1, computers, glass, steel, plasticky substances, trucks, fuels, ships, houses, and almost everyting else. Also, those macro elements include the major constituents of fertilizer (nitrogen, potassium, phosphorous). Also, those macro substances include the major constituents of reinforced concrete (silicon, oxygen, calcium, iron, and a little carbon), and concrete is the most common material in modern civilization by a very wide margin. Also note that these macro elements are already traded in large quantities and are used for precisely those purposes, so I'm not being theoretical here. Note also, that it's entirely possible to make plasticky things out of silicon and oxygen; carbon is not necessary for this purpose (although it's slightly cheaper). Remarkably, most of civilization is "built" out of large amounts of very few minerals.

In addition to those macro elements, we need a few other elements, in small amounts. I'll refer to these as the "micro elements." We need carbon in small amounts because it's needed for steel, and for drug making. We also need trace minerals, in extremely small amounts, for fertilizer (zinc, magnesium, sulphur, chlorine, sodium, and a very few others). We also need Uranium in small amounts if we choose to generate our energy using nuclear power, and we need its byproducts for nuclear medicine and smoke detectors.

Of course, at present we also use other trace minerals, like rare earths (for magnets), lithium (for batteries), nickel (batteries), platinum (catalytic converters), mercury (thermostats), copper (luxury cookware, luxury stereo cables1), chromium (chrome-plated objects), gold (jewelry, rapper dentistry, luxury stereo connectors) and so on. All those things are either unnecessary or have obvious substitutes. We can make batteries out of zinc, magnets out of other metals, and flat computer monitors using OLEDs (chemical formula: Al(C9H6NO)3. Remarkably, we can now even make flat screens out of the macro elements listed above; flat screens and batteries are the commonest uses by far of minerals other than the macro elements). We can forgo chrome-plated objects, jewelry, and luxury cookware.

Note also that we don't require fossil fuels, either for transportation or for any other essential purpose. For transportation we can easily substitute the kinds of transportation which prevailed before cars were common: trolleys, steetcars, trains, buses, and so on; and those are electric or can easily be electrified. Or we could use plug-in cars, as those become cheaper. In the few cases where there is no obvious way to electrify something (like ships) we can easily use alternative fuels like woody biomass in steam boilers (the small amount of carbon already listed in "micro elements") or anhydrous ammonia, which is a combustible fuel and which can be easily manufactured from nitrogen and hydrogen (listed among the macro elements), provided we have energy.

In addition to the elements listed above, we also need energy. In some ways, energy is the master resource, insofar as it allows us to extract the minerals and elements found above. We need energy for many essential purposes in civilization: to mine the minerals, to extract the minerals from their oxides, to transport the minerals, to build the objects of civilization, to fix nitrgoen for fertilizer manufacture, to manufacture the fuels for the few forms of transportation which cannot be electrified, and to supply the basic needs of civilization like lighting, heating, air conditioning, computing, communication, and so on.

It's difficult to determine the minimum amount of energy required for civilization, so let's make a liberal guess and say that we need 30 MwH of thermal energy per person per year, for 10 billion people. (This number is much higher than present per-capita energy expenditures worldwide). As a result, we would need about 300 petawatt-hours of thermal energy worldwide.

Finally, we also need fresh water to grow plants, and space to grow them. Presumably we would need cropland, although we could use hydroponic agriculture on a large scale if it were really necessary. Already, crops are grown in large quantities in hydropnic greenhouses in some parts of the world. Soil is not actually a requirement for agriculture if you have fertilizer. However, let's ignore this substitute and assume that we'll need cropland.

So there we have it. We've completed our catalog of what's necessary for civilization: 8 macro elements, a few micro elements, 300 petawatt hours of energy per year, fresh water, and cropland.

How much do we have?

Now that we've listed the elements and substances which are absolutely required for civilization, let's take stock of how much of them are available, and how much we need.

First let's start with the "macro elements" listed in the prior section, which we need in huge quantities. They are: iron, aluminum, silicon, oxygen, calcium, nitrogen, hydrogen, potassium, and phosphorous. Happily, five of those elements (out of nine) are super-abundant in the earth's crust. Iron, aluminum, silicon, oxygen, calcium, and potassium combined make up about 94%2 of the earth's crust by volume and can be found in fairly high concentrations virtually anywhere you put your foot down. For the most part, the Earth's surface is made out of those elements. Of the remainder, hydrogen can be acquired from seawater using electrolysis, and nitrogen is the primary component of air (60%). Phosphate is the only one of the "macro elements" which is not super-abundant, in truly massive quantities, almost everywhere. Phosphate is required for fertilizer, and I'll speak more about it later.

Let me repeat this fact. All of the macro elements except phosphorous are available in essentially inexhaustible amounts, since the Earth is made almost entirely out of them. Furthermore, those macro elements are sufficient to make almost all the objects of civilization. As a result, we could cover the entire terrestrial surface of the Earth in a miles-deep layer of concrete, glass, power cables, fertilizer, industrial machinery, skyscrapers, computers, and plasticky crap3, and we still wouldn't have run out of minerals or come anywhere close to it.

What about the "micro elements" like zinc and magnesium (for fertilizer) and carbon? Those are also available in truly massive quantities relative to the amounts which we require. For example, magesium constitutes 2% of the Earth's crust, which is 2000x more abundant than oil ever was despite being needed only in milligram quantities per person.

Now let's deal with energy. Let's assume we need 30 MwH/person/year thermal for advanced civilization (as we said before). With ten billion people, we'll need 300 Petawatt-Hours per year of thermal energy. (Of course, thermal energy can be converted to electricity by solar thermal power plants). Let's calculate how many solar thermal panels we'd need to capture that amount. The surface of the earth is bombarded by 89 petawatts continuously. Of that, 25% ends up on land, of which 14% ends up on desert. So deserts worldwide are continuously bombarded with (89*.25*.14=) ~3 petawatts, which is (3*24*365) ~26,000 PwH per year. Which means we would need to cover (300/26000) 1.1% of the Earth's deserts to capture the amount of energy we need to sustain advanced civilization for 10 billion people. In other words, we could obtain almost 100x more energy than we require using only solar panels in deserts. Of course, solar thermal is not the only source of energy available to us. There are also many other sources of energy, like offshore windfarms and breeder reactors. There is also the possibility of nuclear fusion, which isn't ready yet but which conceivably could come online within the next few hundred years, and which would multiply the amount of energy available to us by a huge factor.

Of course, solar panels require mineral resources for their construction. However, we have more than enough resources to build the solar thermal panels necessary to capture that amount of energy. Solar thermal panels are made mostly out of glass, which is silicon and oxygen, which constitute over 60% of the Earth's crust and which are found in massive quantities in the sand beneath the solar panels.

All that remains are freshwater and arable land. Freshwater can easily be obtained from seawater using desalinization plants (of course desalinating would require energy, but the amounts of energy required per person for desalinization are modest, and we obviously have enough of it; see above). As far as arable land is concerned, we have enough already to feed 10 billion people if we increased agricultural productivity per hectare to first-world standards throughout Asia, by the application of fertilizer, of which we have virtually inexhaustible amounts, because it's made from the macro elements listed above. Of course, we have many other options for growing crops, like irrigating the massive unused land areas, by desalinating water and digging canals.

In short. We are not running out of anything essential, other than phosphorous, which I'll deal with later.

The next million years

From the previous section, we can see that we have vastly more resources than we require. We're not running out of anything essential in the foreseeable future. But what about the longer run? If we have millions of years' worth of minerals, won't we run out eventually, if only in the far future?

No. All of the "macro elements" needed in large quantities (aluminum, iron, silicon, oxygen, calcium, nitrogen, hydrogen, potassium) are inifinitely recyclable, and are not being used up at any rate. When we mine these minerals and then throw them away, we have not affected the amount of them available in the Earth's crust at all. They will gradually return to the oxidized state in which they were found originally. Then we can re-mine them from landfills and separate them again, provided we have energy (which we do).

Nor are the "macro elements" being "dispersed" by our mining them, as some doomer authors have claimed4. The macro elements are already found in high concentrations all around the globe, and so won't be dispersed in any meaningful way by our mining them. Even if we mined all of those macro elements everywhere on the Earth's surface, and then scattered them at random all around the globe, they would still constitute 91% of the Earth's crust and so would be in high enough concentration everywhere to warrant economical separation and re-mining.

Nor would we gradually exhaust our supplies of fresh water. When we desalinate ocean water and use it for crops, that water will eventually evaporate into the atmosphere, condense, and rain again into the oceans, where we can desalinate it again provided we have energy. Similarly with fixed nitrogen for fertilizer: we can separate nitrogen out of the air over and over, provided we have energy.

Of course, all of this recycling depends upon energy. Clearly, energy is the master resource, upon which everything else depends.

However we have 100x more energy than we need, just from solar alone. This flow of energy will be continuous (more or less) for billions of years. Thus, we do not face any shortage of energy, which means that we don't face any shortage of anything else essential either.

The sole exception is phosphorous. Phosphorous is needed for fertilizer and is a basic component of life, thus it has no substitutes. Phosphorous is mined from phosphate rocks which will be exhausted within 1,000 years. Phosphorous cannot be re-extracted or recycled from the environment, because it actually is being dispersed as we use it: it doesn't exist in high enough concentration everywhere for us to throw it away at random (actually to allow it to run-off into the ocean) and recover it later, since it subsequently will be much more thinly dispersed. At some point we will start to run out of phosphorous, and will have to be more careful with its use. At some point we'll need to recover phosphorous from sewage and from dead bodies, instead of throwing it away or burying it.

Other than phosphorous, we face no shortages of any essential substances or elements for the foreseeable future. There is no reason, right now, to conclude that we will "run out" of any essential elements within the next million years.

Brief diversion: an unusual resource

One element which deserves special mention is silicon. Silicon is a remarkable element, insofar as silicon atoms can be "chained together" (like carbon) to form a limitless array of complicated molecules with very diverse chemical properties. Silicon is a "master mineral" insofar as you can make almost anything out of it. Using silicon and oxygen, you can make glass, or metals, or conductive wires, or electrical insulators, or plasticky substances, or fleshy sex toys, or fiberglass, or gels, or caulks, or clothing, or building materials, or breast implants, or silly putty, or opaque substances, or transparent substances, or liquids, or combustible fuel, or turbines. Silicon can also be used to make computer chips, wires, fiber optics, and electronics. It's not the ideal material for many of these purposes, but it is a possible substitute for all of them.

Remarkably, silicon and oxygen are the most prevalent elements in the Earth's crust, by far, together constituting 60% of the Earth's crust by volume. Silicon and oxygen are the main constituents of dirt, sand, and rocks.

This fact deserves special consideration. We can re-arrange the atoms in sand and make a computer, including the case.5 Or, we could also make a train. Or a building, including insulation, wiring, and windows.

Silicon and oxygen therefore constitute the "ultimate backdrop" among mineral resources. They are the ultimate substitute, because they can take on so many different properties by re-arranging their molecular structure, and because they are available in such massive quantities. Silicon and oxygen are so common that they're usually the "dirt" which we throw away when we're trying to mine other things.

In fact, it would likely be possible to build all the necessary objects of civilization, except fertilizer, from silicon, oxygen, aluminum, water, and thermal energy6. Remarkably, even computers and high-tech equipment are made overhwlemingly from this substance.

The versatility and massive supplies of silicon should be considered every time we hear a doomer claim that "we are RUNNING OUT of x and there are no possible substitutes." Whatever "x" is, it most likely could be made from sand, which is mostly silicon. In the few cases where it cannot, we have many, many alternatives.


We will never just "run out" of essential resources. Instead, we'll eventually need to undergo a transition, from exhaustible resources, to inexhaustible ones. For example, we'll eventually need to move away from fossil fuels, to other sources of energy which are vastly more plentiful. We'll also need to transition from internal-combustion cars, to electrified transport. We'll also need to replace our very small usage of limited minerals (like cobalt) with obvious alternatives. After we have done so, we'll have enough resources and energy to provide 10 billion people with a 1st-world standard of living for the next few billion years.7

Whether we have the wisdom to make this transition before civilization collapses, is another topic which I'll address in a subsequent article. The point I'm making here is that we don't face any inevitable decline of civilization solely from exhaustion of essential resources. There is no mathematical law or curve which implies that civilization must end. There is no law of nature or ecology or thermodynamics which implies that we're about to run out of resources or energy. We face a gradual transition; that is all. Whether we have the wisdom to make that transition is another topic.

Of course, we will always require huge amounts of the "master resource"--energy. Energy is what allows us to extract and re-extract all these resources. Happily, we have vastly more energy than we require, and we will have vastly more for billions of years.

The only pressing shortage is phosphorous, for fertilizer. At some point fairly soon (within 1000 years), we'll need to stop wasting phosphorous. We'll need to start recycling sewage to recover phosphorous, rather than allowing it to dissolve or flow into oceans. Happily, we won't need to start recycling phosphorous until after everyone on this planet has already reached a 1st-world standard of living, and when the expense of recycling it will be quite tolerable.

When I started learning about these things, years ago, I was shocked to discover just how few minerals are required to make almost everything we need to sustain civilization. That's the magic of chemistry, I suppose. Already, almost everyting is made overwhelmingly out of iron, aluminum, silicon, oxygen, calcium, hydrogen, nitrogen, potassium, and phosphorous (plus carbon, but the carbon is not essential for almost anything we build). I was also quite amazed to find that precisely those minerals, which we require in large amounts, constitute 91% of the mass of the Earth's crust and are essentially what this planet is composed of. What a fortunate coincidence: we are living on a massive planet made of precisely what we need.


[1] Copper is not required for wires or cables--not even for computer cables. Aluminum is suitable for this purpose, although slightly worse. Already, most ethernet cables are made out of aluminum, not copper.


[3] I am not saying it's desirable or possible to cover the surface of the Earth with a miles-deep layer of plasticky crap. I'm saying we have enough mineral resources to do it.

[4] This claim of "dispersing elements" has repeatedly been made by the ecological economics school of thought, especially by Herman Daly. While this argument is true for some minerals, it does not apply to the macro elements I listed, since those macro elements are found in high concentations everywhere no matter where we scatter things, so they'll never need to be "filtered" out.

[5] Of course, it's not possible to make computer chips solely from silicon. We must also use boron to dope the silicon wafers. Boron must be added in concentrations of at least 1 part per 100 million. We have enough boron for this, since computer chips are very small and boron is not rare, and we require only 1 part per 100,000,000. Also, we would need monitors, which can be made using OLED technology. OLEDs are made from aluminum, nitrogen, hydrogen, and oxygen, which are all among the macro elements listed, plus a little carbon, which is among the micro elements listed. Remarkably, even flat screens can now be made from the macro-elements I listed. This is important because flat screens and batteries are two of the commonest uses of rare minerals.

[6] Except computers, which, as I pointed out in note #3, require trace amounts of the element boron, which we also have in massive quantities relative to our requirements.

[7] A more in-depth examination of our resource future can be found in the paper The Age of Substitutability, H. E. Goeller and Alvin M. Weinberg, Science, 191 (1976), pp 683-689. They take a different approach to explaining these issues from the approach I've taken, insofar as they're more mathematical. Also, they do not touch upon the many uses of silicon or dwell upon computers, since those were rarely used in 1976. Also, they assume far less energy is necessary for civilization than I have assumed.

Wednesday, October 26, 2011

We do not face rapid declines of oil

How much oil is there? 3 trillion barrels? 4 trillion? How do we know?

There is no single answer to that question, because the answer depends upon price. At $80/barrel, there is a certain amount of oil remaining. At $150/barrel, there is far more oil remaining. And so on. The reason is because we leave most oil in the ground, since it's uneconomical to extract at current prices. As prices increase, the amount of economically extractable oil increases also. So there is no single "amount" of oil remaining. The amount depends upon price.

The Supply of Oil

Most resources are distributed according to a "resource pyramid." A "resource pyramid" is divided into layers, with small amounts of easily-extracted resource at the top, and larger amounts of harder-to-extract resource in the middle, and vast amounts of difficult-to-extract resource at the bottom. At the very bottom of the pyramid is a very diffuse resource, which is availabile in massive amounts, but which costs a lot to extract because it requires so much energy and money to "filter" it out of some substrate.

Oil supplies are distributed according to a resource pyramid. At the top of the pyramid was a small amount of easily-accessible oil. When we started drilling for oil, it would sometimes "gush" out of the ground with great force after drilling only a shallow well. That oil was easily extracted and inexpensive, and it was all used up within a few years. After that, we had to move down the resource pyramid to more difficult (and far more plentiful) resources.

At the bottom of the oil pyramid are resources like shale oil and tar sands, which are very diffuse and difficult to extract. (In fact the lower strata of the resource pyramid aren't really oil at all but are hydrocarbons which are converted into oil using chemical processes, for example, coal-to-liquids). At the very bottom of the resource pyramid is manufactured hydrocarbons using (for example) nuclear power, gasification, and the fischer-tropsch chemical method to convert any carbon source into hydrocarbons. This technique could supply oil indefinitely at an EROI of about 5:1.


As oil becomes harder to extract, it also becomes more expensive, which opens up vast amounts of oil that were previously uneconomic. When we start to run out of oil, we take a step down the "resource pyramid" and start using the far larger (and more expensive) oil at the lower stratum. The result is that supply of oil increases, as price increases.

For this reason, oil production worldwide will not follow a bell curve or anything similar. Instead, oil will peak and then prices will gradually increase, thereby allowing us to extract previously uneconomic resources, and thereby preventing any decline. We'll know when we're transitioning to a lower stratum on the pyramid because prices will increase permanently by some amount. However this doesn't imply any kind of decline, because more oil may be available at that price.

Take the USA as an example. Oil production peaked in 1970 in the USA, as per Hubbert's prediction, and then declined for several years thereafter. Then something funny happened. In 1976, prices increased because of a cartel, and production started increasing again in the USA, years after it had peaked. When the cartel failed, and prices for oil started delining, production in the USA started declining again. Now, decades later, prices for oil are high again, and production in the USA is increasing again, decades after the peak! Clearly, production depends upon price and does not necessarily follow a bell-shaped curve.1

Of course, as oil prices increase, alternatives become relatively cheaper. Trains become cheaper than trucks, since trains use about 1/8th the energy per pound-mile. Hybrids become cheaper than convetional cars, since they cost only slightly more and use far less fuel. This trend will reduce demand, and further attenuate the decline.


Because of these facts, the very idea of "peak oil" as it's commonly understood (where we have a fixed amount of oil and it will follow a bell-shaped curve) is mistaken. Instead, we face a plateau or oil, or perhaps very gradual declines, for many years after the peak. Oil will not follow a bell curve or anything similar.

The doomer idea of applying a "bell curve" to worldwide production probably stems from their graphs of individual oil wells, which do in fact follow such a curve, roughly. But that doesn't mean that all oil wells in aggregate will follow a similar curve. It would be a tremendous mistake to assume that all oil wells will follow a bell-shaped curve because any one oil well has historically done so. Whereas an individual oil well doesn't increase prices as it depletes, the sum of all wells depleting will increase prices, and thereby increase supply and prevent subsequent decline.

As we deplete the cheaper oil, we will move down the resource pyramid, thereby increasing the price and the supply, and thereby attenuating any decline. We will never face an abrupt drop-off in supplies as a result of depleted resources, nor will we face anything like the 3% annual declines which doomers assume.2

If anything, we face a greater threat from China out-bidding us for oil than from declining production.

None of this implies a perfect future. We will face higher prices for oil in the future, particularly because of competition from developing nations, and this may force unwelcome changes to the American lifestyle, like driving hybrids instead of SUVs. However, we do not face rapid declines of oil, nor do we face any kind of interruption of industrial civilization.



2 I'm referring to abrupt declines caused by depletion of resources. Of course, it's possible there will be abrupt declines caused by war, disruption, severe economic downturn, etc. It's also possible there will be brief abrupt declines of a few percent based upon miscalculations because oil production does not follow a smooth curve. It's also possible there will be fairly rapid declines caused by peak demand. For example, it's possible that prices for oil would increase greatly, and the prices of batteries would decrease, to the point that plug-in hybrids became cheaper to operate than conventional cars. In that case there could be a rapid transition over the course of 15 years or so to alternative propulsion for cars, leading to a 5% annual decline or greater. That would mean that oil wasn't crucial anymore.

Saturday, October 8, 2011

Exponential growth

Doomer literature is replete with stories of exponential growth and then collapse. It's one of their favorite themes. Some doomer sources (like Albert Bartlett, or the 7th fold) focus on exponential growth as their primary topic. In fact Albert Bartlett has famously claimed (famously in doomer circles anyway) that the inability of most people to understand exponential growth is one of the great problems facing the world today.

The reason exponential growth is a topic in doomer circles, is because exponential growth of almost any important quantity will lead to disaster before very long. For example, exponential growth of population, by 4% per year, would lead to 6.5 million times as many people in the world in 400 years, necessitating a massive die-off. And exponential growth of energy usage, by a few percent per year, would lead to the oceans boiling in only a few centuries. In most cases, sustained exponential growth leads to die-off or collapse before very long. Just look at colonies of bacteria or yeast, which (as doomers frequently point out) grow exponentially until they suffocate in their own waste or exhaust their food supply.

There is only one big problem, with all this talk of exponential growth and then collapse. The problem is: no important quantities are growing exponentially.

Take energy consumption as an example. Energy consumption is not growing exponentially. In fact, energy consumption per capita isn't growing at all, not even linearly, and hasn't grown much for decades. This point should be obvious to anyone over age 40, by just looking around and then remembering how things used to be. Do you use exponentially more electricity than in 1970? Do you set your thermostat 3% higher every year? Do you use exponentially more gasoline than in 1970? In fact, you probably use less gasoline, if your car in 1970 was anything like my parents' cars, which got about 10 mpg.

In fact, energy usage per capita has been essentially flat for the last 5 years or so, and was growing only very slowly in the 2 decades before that, and not at an exponential rate.

Well, even if energy usage per capita isn't growing exponentially, isn't the world population growing exponentially? Don't we face exponentially more people, and so exponentially more resource usage even if resource usage per capita remains the same?

No, the population of the world is not growing exponentially. In fact, the population is growing, but at a declining rate. Already, the rate of population growth has reached 0% in many industrialized countries, and is declining rapidly in developing countries. Furthermore, there is every reason to believe that the rate of population growth in developing countries will also reach 0%, since they're basically following the same pattern which industrialized countries have laid down. As a result, most professional demographers believe that world population will level off at between 10 and 12 billion people.

Well, isn't the economy growing exponentially? We see figures like "3% growth this year" which seems to imply it's growing exponentially, right? And, since we need energy for economic growth, doesn't that mean that the rate of energy usage is also growing at 3%?

First, economic growth isn't the same thing as growth in energy consumption. (This mistake is very common). The term "economic growth" refers to growth in production of things you can buy. If someone invents a new drug, he has caused the economy to grow, even if manufacturing that drug takes less energy than manufacturing the older equivalent.

Second, the economy is not growing exponentially. The rate of economic growth is delining everywhere. As economies mature, their rates of growth decline. This has already happened in all first-world countries, which do not enjoy anyhwere near the rates of growth of (say) China or India.

Thus, essentially no important physical quantity is growing exponentially1. Population is not growing exponentially. Energy usage is not growing exponentially. Industrial output is not growing exponentially. The economy is not growing exponentially. Food production is not growing exponentially.

Since these quantities are not growing exponentially, it would be a basic mathematical error to use exponential functions to model them.

Both population and energy usage will grow and then level off before very long. There is ample historical precedent for this.


1 The term "exponential growth" refers to a mathematical function like this: f(t)=bt where b is constant. If b is not constant, then it's not an exponential function. None of the quantities listed in this article are growing according to an exponential function.