Monday, September 27, 2010

EROI doesn't matter

EROI, or "energy return on investment," refers to the amount of energy we get in return for some amount of energy invested. It's a ratio of energy return for energy investment. For example, if we spend 1 unit of energy on building an offshore oil rig, and the rig yields 5 units of energy over its lifetime, then it has an EROI of 5:1, or just 5. In other words, it has "paid back" 5 units of energy for every unit of energy invested. As another example, assume I walk one mile and burn 200 calories, in order to acquire some stawberries which have 400 calories. In that case, the stawberries had an EROI of 2, because they yielded twice as much energy as was spent in acquiring them. Usually, EROI refers to industrial sources of energy generation, like coal-burning power plants, windmills, etc.

Various kinds of energy generation have different EROIs. For example, a typical coal-fired power plant has an EROI of 30, which means that it yields 30 times as much energy over its lifetime as it took to build it and supply it with coal. Other forms of energy, like wind and solar, have lower EROIs. Corn Ethanol is the worst, since it yields only about 20% more energy (EROI of 1.2) than was spent in manufacturing the fertilizer to grow the corn.

Over time, the average EROI for all energy sources has been falling steadily. It was about 100 in the early 20th century, and has fallen to about 30 now. It's continuing to fall. The reason is because we have "used up" the easiest sources of energy first, and must turn increasingly to lower and lower quality sources of energy thereafter. For example, we used up the shallow oil first, and now we must dig deep wells (at great energy expense) to get more oil. EROI is falling and will continue to fall, and probably will never again reach the heights experienced in the early 20th century.

There is an argument that declining EROI will cause the destruction of our civilization. The argument runs as follows. As EROI declines, we must spend more and more of the energy available to us, on generating more energy. As a result, the amount of energy left over, for our uses, is less and less. In other words, net energy (which is the amount of energy left over after subtracting energy investment) must decline as EROI declines. It's an inevitable physical fact. At some point, EROI will decline so far that we won't have enough energy left over to sustain our industrial civilization, at which point, civilization will collapse, and we'll revert to a medieval mode of life.

Or so goes the argument. This argument (which I'll call the "energy decline argument") was first advanced by Prof Charles Hall, who is a professor of ecology at SUNY. He first articulated the argument several decades ago. He also invented the concept of "EROI" at that time, and pointed out that EROI was precipitously declining. Since then, the energy decline argument has gained many adherents. It has achieved a widening influence. Recently, it has inspired an impressive number of books, articles, papers, websites, lectures, and so on, all claiming that energy flows will soon decline and that civlization must decline thereafter.

The problem with the energy decline argument is that it's totally wrong. It's wrong from beginning to end. It relies on incorrect implicit assumptions, and it reaches incorrect conclusions.

In this paper I will refute the energy decline argument by showing that its assumptions are wrong. I will show that EROI is unimportant and does not threaten our civilization. I will also show that energy is abundant and will grow over time.


The fundamental problem with the energy decline argument is this: it implicitly assumes that the rate of energy production is constant. In other words, it assumes that we have a constant number of power plants in the world and cannot build any more. If that were true, then the energy decline argument would also be true. If the amount of energy we generated were constant, then declining EROI would, in fact, imply declining net energy, because we would have to spend a larger fraction of the fixed amount of energy available to us in acquiring more of it. However, the assumption is false; in fact, the amount of energy we produce is not constant. We can build more power plants. As a result, we can easily compensate for declining EROI by just building a few more power plants to compensate for the EROI decline, while still generating the same net energy output. In other words, the assumption underlying the energy decline argument (namely, that enregy supply is constant) is wrong, and therefore the argument is false.

Let me provide a simple example. Suppose we have a 1000-megawatt coal-fired power plant with an EROI of 10; in other words, it produces 1000 megawatts continuously over its lifetime, and it consumes 1/10th that amount (100 megawatts) continuously over its lifetime. Now assume that we are running out of coal to supply that power plant, and must replace it with lower-EROI power plants. So we replace it with two 1000-megawatt solar thermal plants with EROIs of only 5 each. Despite a reduction of EROI by half, we still have increased total energy output by by almost 80%. (The high-EROI plant produced 900 (1000-100) megawatts of net energy, but the two low-EROI plants combined produce 1600 (1000-200)*2).

Each power plant with an EROI higher than 1 is an energy multiplier1. It multiplies the energy available to us, because it produces more energy than it consumes. With each power plant we build, we multiply the total amount of energy yet again, albeit by some small factor. From this fact, it's clear that we can produce any amount of energy we wish, by multiplying often enough (building more power plants), regardless of EROI. This reasoning follows from simple arithmetic: you can reach any number you wish, by multiplying by any factor higher than 1, provided you can multiply as often as you wish.

In fact, the energy available to us can grow exponentially with any EROI higher than 1. The reason is because the output of any plant can be used to build several other plants, each of which can then be used to build several more plants, and so on. For example, we can use the output of a single power plant, to smelt the iron ore, manufacture the components, and make the hydrocarbons necessary for the construction of several more power plants. In that way, we can increase the amount of energy available to us exponentially, with any EROI higher than 1. In fact, our society has already done that. We've already used the output of power plants to build more power plants. That's how we were able to increase the amount of energy generated in this country (USA) by a factor of ten in the six decades from 1920-1980, despite never having invested more than a small fraction of our energy in acquiring more energy.

Let me provide an example of what I mean. Suppose we build one power plant with a very low EROI of 8. We use the full output of that plant to build another 8 plants just like it and adjacent to it. Then we use the full output of the other 8 plants, to build an additional 64 power plants, and so on. After ten generations, we'll have about 1 billion power plants, without any outside investment of energy except what was needed to construct the first plant. Of course, we can't really build a billion power plants. At some point, we would reach the maximum theoretical amount of energy we could generate. However, that maximum amount is enormous and is more than 1,000 times higher than current worldwide energy production, so we won't reach it any time soon.

The only practical limits to energy generation are imposed by cost and demand. Cost and demand determine how much energy we can generate, not EROI. Cost and demand are the only reasons we don't quadruple our energy output in short order.

Nor does it matter if EROI declines. At present, we have an average EROI of about 30, which means that we spend only 3% of our energy in acquiring more energy. If our EROI fell by half, we could compensate for it by building ~3%2 more power plants, thereby keeping total energy output the same. (Assume 100 power plants, 3 of which are used to generate the energy needed to power the others. Now assume EROI falls by half, and the 100 power plants require twice the energy input for the same output. In that case, we would require 6 power plants (rather than 3) to supply the others. So we would then require 103 power plants (rather than just 100) to compensate for a 50% decline in EROI while keeping output constant).

Nor is there any reason to believe that EROI will decline any further in the future. The dramatic decline in EROI experienced during the 20th century was a one-time event which is now over. The reason is because there is exponentially more energy available at lower EROIs. At an EROI of 100, which is very high, there was a small amount of energy available, which was exhausted within a few years; but at an EROI of 15, which is much lower, the amount of energy available is practically limitless. As a result, the average EROI for the world will probably never decline below 15.

Nor should we bother to pursue higher EROIs. Higher EROIs don't necessarily lead to larger total energy production or lower cost. Larger total energy production would be achieved by pursuing the cheapest (in money) sources of energy, not the ones with the highest EROI. (By "cheapest" I mean the lowest-cost net energy). Cheaper energy leads to greater demand, which leads to more construction of power plants at that cost, which leads to higher total energy production and lower costs despite lower EROI.

The irrelevance of EROI is demonstrated by history. As doomers like to point out, EROI was about 100 in the early 20th century and has declined to about 30 now. During that time, we increased worldwide energy production by more than ten-fold, and increased per-capita income in the advanced countries by more than eight-fold. In this case, declining EROI didn't imply declining net energy (quite the opposite). Nor did it destroy the economy, nor constrain production. EROI made no difference.


We do not face declining energy flows. Quite the opposite, we have virtually unlimited amounts of energy available to us. We could increase the amount of energy we generate, practically without limit, subject only to cost and demand. Energy production can grow exponentially, which allows us to generate any amount of energy we require.

Declining EROI is not particularly worrisome. We can easily compensate for declining EROI by building a few additional power plants, thereby keeping energy output constant. For example, we could compensate for a 50% decline in EROI by building only about 3% more power plants.

Even if declining EROI were destructive, EROI is not declining much any more. EROI will probably never fall below an average of 15, over any time scale, because there are vast amounts energy available at that EROI.

EROI doesn't matter. It doesn't matter if it's increasing or decreasing. It doesn't matter that it decreased in the past. It doesn't cause decreasing energy flows, nor does it limit our energy generation, nor does it threaten our civilization. As long as EROI remains higher than 1, which it always will without any special effort on our part, it makes no difference. All that matters are cost (of net energy) and demand.

1 When we multiply energy, we're really only multiplying the amount of usable energy, like mechanical energy, electrical energy, chemical energy in food, and so on. We don't really increase or decrease the amount of energy in the universe. In fact, when we "generate" energy, we're really only converting energy, from a non-usable kind into a usable kind. Power plants multiply the amount of usable energy.

2 This figure is an estimate. In fact, we would need to increase the number of power plants by slightly more than 3% to compensate for a decline of EROI from 30 to 15.