Until recently, not many people were aware of this important bit of Peak Oil wisdom, but because of the successful efforts of activists in the Post Carbon Institute and elsewhere, many people now know it; everybody should:
When a barrel of oil costs more energy (to find, extract, refine, deliver, and sell) than the energy in the barrel of oil itself, rational actors will choose to leave it in the ground.
Having heard this formulation a hundred times, and having repeated it to various readers and audiences, I want to speculate a little about oil and "rational actors."
Money is a representation of energy, not the other way around. The energy--as oil, or sunshine, or labor measured in food calories, etc--is physically real; the money's just a paper symbol charged with socially constructed power. Huge amounts of energy are chemically inherent in petroleum, but unextracted oil becomes worthless when its EROEI (energy return on energy investment) approaches zero.
It does this because every variable in the game tends to deteriorate as time goes on: at the start (c. 1853, when the first commercial well opened in Poland), the Earth is untapped, riddled with great pockets of delicious pressurized oil lying just under its skin. Rockefeller shows up, sticks a straw in the ground, and a gusher blows 100 feet into the air. This amazingly easy oil is of the highest quality, because while the heavy sulfurous gunk sinks deep down, the light sweet crude rises toward us nifty surface dwellers. The easy oil is the better oil, so latecomers have the indignity of expending huge new levels of effort and resources just to get their hands on inferior petroleum. That's the heavy stuff, the stuff you prefer to use for asphalt. The stuff Venezuela can only sell to the Chinese, because everyone else with lungs is scared to burn it.
On any timescale relevant to human affairs, the planet's natural endowment of petroleum does not grow at all. Nearly all of it was formed in two special, geologically exotic eras that occurred roughly 150 m.y.a. and 90 m.y.a. (million years ago). Liquid hydrocarbons occur in only a finite but unknown number of spots around the globe; of course, once we started searching for them, the number of undiscovered ones began shrinking toward zero. The more we searched for oil fields (which they call "exploration"), the better we got at finding them (which they call "discovery"). Discovery peaked in 1964. In other words, in no year before or since have we discovered more oil than we discovered in 1964. Except for some non-linear fluctuations, we have discovered less and less oil every year since 1964.
Once you discover oil, you have to extract it (which they call "production"). Production seems to have peaked around 2008, plus or minus about six years, depending on whose numbers you find persuasive. The date of the Peak is something we can only really know in retrospect, though by then our minds may well be focused on huntin' up our next rack o' possum ribs, not reminiscing about Saudi Aramco's bygone century of make-believe field survey data.
As with discovery, practice makes perfect: the more we extracted, the better we got at extraction. Over the past century and a half, a lot of brilliant petroleum geologists invented scads of amazing instruments, techniques, mathematical models, software, and geochemical expertise, for the purpose of sucking out an ever-greater percentage of the oil in a given deposit... then going back for more (which they call "secondary recovery"), often with a new approach, like injecting the field with a million tons of seawater daily, so the oil will float to the top... then giving it one last shot (which they call "tertiary recovery"), where they send down a rhesus monkey on a rope with a bucket and a spoon. The point is: the better we got at extraction, the more quickly we depleted the planet's limited supply.
Okay, fine. So the viability of an energy source depends not on the amount of money required to exploit it, but on the amount of energy returned per unit of energy invested. But is money totally irrelevant to the question above? The question was, whether people will deliberately leave oil in the ground.
Well, the law forbids corporations from behaving like decent members of the human community, since their sole purpose is to generate profit. They don't ultimately care whether the profit comes from the financial side (selling stock to investors) or the commercial side (actually selling a product to customers). Corporations, like the mentally compartmentalized individuals who run them, are only "rational actors" within the tragically narrow frame of reference that defines them. As Chomsky pointed out in an interview I used to play over and over when I was a teenager, "If you're the Chairman of the Board and you start behaving as if human beings are more important than profit, you find very quickly that you're not Chairman of the Board anymore."
When a barrel of oil costs more energy (to find, extract, refine, deliver, and sell) than the energy in the barrel of oil itself, rational actors will choose to leave it in the ground, lest they lose more energy than they gain. Unless, of course, someone else takes the losses for them, as happened in the 2008 bailout of the big banks, when the taxpayers covered the banks' losses (and then some). I can imagine a scenario in which a young Dubya-like one-percenter goes after oil long after the other companies have given up doing so, because for him, somehow, the public foots the bill.
Suppose a Mafioso is given an oil company for a birthday present. Suppose all his operating costs shrink to near zero because his "people" have "ways" of reducing costs. He could afford to extract oil that others had left in the ground because it cost them too much money to extract. But unlike them, this guy could buy up depleted old fields for a song, sell the thousand defunct rusty old derricks for scrap metal, and get ta drillin'. He starts injecting seawater, firing up the triple-zoom robo-snake and the other hi-tech gizmos, running the pumps, paying the employees, etc., and then sells whatever heavy sour crude he can for profit. The energy he's investing per barrel could be double the energy he's extracting per barrel of this mediocre but useful goo, and he might just keep on going anyway, because it barely cost him anything (neither money, nor the energy for which money is a symbolic place-holder).
Imagine a network of one thousand permaculture-based eco-villages, powered by a solar-and-wind operation where a million men and women work every day to keep things going smoothly. One day the energy gets diverted, somehow stolen by a gangster. Only the villages are in a physical position to make efficient use of the energy they collect, since they live on site; the thief, by contrast, can't steal it without wasting half of it in messy processes of storage and transmission. But he has no interest in efficiency, because this is stolen energy that cost him relatively little (hire some paramilitary contractors; bribe a few cops n' Congressmen). He's like the little rascal who runs home from Old Man Hayseed's well, sloshing his open bucket full of stolen water; he doesn't care that it's nearly empty when he gets home, because he didn't get caught and he did get some free water.
Anyway, back to our story. The energy thief dumps his ill-gotten gains into a tertiary recovery petroleum operation and sells the meager results for a fabulously high price because it's become so rare. When a unit of energy requires a month's labor from a million hungry villagers, plenty of people will prefer black market petroleum if they can get it.
People already have sometimes chosen to leave petroleum in the ground for reasons other than inadequate EROEI. They do it to drive up the market price; or to save some for later; or to placate people who want some field to go offline for a while; or if the military hasn't yet got the place under adequate logistical control for Halliburton's comfort; or if President Putin wants to remind a freezing former Soviet satellite that he's in charge. In my little fable, where things are more simple and extreme, it's easier to see the perspective-dependent nature of "energy accounting"; it's one thing in the two million hands of a community with long-term goals, but quite another in the two hands of a tweaky aristocrat who knows he's above the law.