One of the main arguments in favor of ethanol production in the U.S. is that it supports the goals of energy independence by getting us off of foreign oil. After all, we could just tell the entire Mideast to take a hike while we grow our own fuel. In fact, there have been some truly grandiose claims made around this theme. Of course if we are making more ethanol, we are importing less oil as a result. Right? Maybe not. Has anyone actually taken a good look?
A couple of years ago, I looked at total gasoline consumption in an essay called The Mythical Ethanol Threat. My conclusion from that was that despite the rapid ramp up of ethanol, there was no apparent drop in gasoline demand. In fact, gasoline demand (which was corrected for ethanol content by backing that out) actually grew at a steady pace even as ethanol was ramping up sharply. But a couple of years have passed, and some comments following my last essay got me curious: Has U.S. ethanol production actually impacted petroleum imports?
From 2002 through 2007, ethanol production in the U.S. more than tripled: From 2.1 billion gallons per year to 6.5 billion gallons per year. (Source – RFA: Historic U.S. Fuel Ethanol Production). Yet total net petroleum imports (oil, gasoline, diesel, etc.) increased over that time period by 2.1 million barrels per day – from 10.2 million bpd in 2002 to 12.3 million bpd in 2007. (Source – EIA: Weekly U.S. Total Crude Oil and Petroleum Products Net Imports). So what does this mean?
I wasn’t going to jump to a hasty conclusion, so I started to dig. I started with several hypotheses. Perhaps U.S. oil production had fallen by 2.1 million barrels per day over that period of time, and the increase in imports were merely to compensate for that. So I checked. No, domestic production did fall over that period of time, but only by 682,000 barrels per day. Domestic production fell from 5.746 million bpd in 2002 to 5.064 million bpd in 2007 (Source – EIA: U.S. Field Production of Crude Oil). But one could allocate that much of the 2.1 million barrel per day import increase to the lower U.S. production.
Had demand growth accounted for the additional 1.4 million barrel per day increase in imports? Yes, in fact petroleum demand did grow (partially rebounding from the 9/11 attacks that reduced demand) from 19.8 million barrels per day in 2002 to 20.7 million barrels per day in 2007. (Source – EIA: U.S. Product Supplied of Crude Oil and Petroleum Products.) So of the remaining 1.4 million barrels per day of the increase in imports, 900,000 could be explained away as being due to an increase in demand. That still leaves a real increase in petroleum imports of 500,000 barrels per day – despite a tripling of ethanol production.
So how to explain this discrepancy? How can petroleum imports rise above and beyond the total increase in demand plus the drop in domestic production? There are two possibilities that I can think of. If the product in storage increased from 2002 to 2007, that can explain part of it. And we did in fact put a lot of oil in the Strategic Petroleum Reserve during those years (but not enough to account for 500,000 barrels per day).
Another portion can be allocated to declining energy returns as oil becomes heavier, and as we switch to lower energy return options like ethanol. For instance, as the quality of crude oil worsens – higher sulfur and lower gravity – it takes more energy inputs to refine it. Likewise as sulfur standards for clean products tighten; energy inputs increase and the net energy falls. This can result in some cannibalization of the oil. In a case with light, sweet crude you may end up with 9 BTUs of net products for 10 BTUs of petroleum inputs. As the crude gets heavier, the net BTUs may drop to 7 because of the need for higher energy inputs for processing. This can explain more of the discrepancy.*
The same is true of ethanol. It does take some liquid petroleum to grow corn and process ethanol, and as ethanol ramps up some of the petroleum imports will now be required in the ethanol industry. This is similar to the case of light, sweet crude gradually becoming heavier, more sour crude. You may have to increase the imports just to net out the same amount of fuel.
But one thing is pretty clear. Our petroleum imports have not fallen as ethanol has ramped up. So it is really hard to make a strong case based on the data that increased ethanol production is reducing our dependence on foreign oil. One reason for this is something I have talked about before, and that is scale. In 2007, our oil demand was 20.7 million barrels per day. When the lower energy content of ethanol is factored in, the 6.5 billion gallons of ethanol produced in 2007 is only worth 0.26 million barrels per day – just over 1% of our total petroleum consumption.** Factor in that some petroleum (and other fossil fuels as well) was used in the manufacture of the ethanol, and the net contribution falls even further.
Factor in all of the fossil fuel inputs that can also be used as fuels (diesel, natural gas, gasoline) and the total net contribution of ethanol toward our petroleum consumption ends up at under 0.5% (and that includes the energy credit from by-products). This relatively low contribution is another likely reason that there is no obvious impact on our imports from ethanol: The contribution may be simply too small to measure.
In closing, this more than anything explains why I often come out against our ethanol policy. It is being presented as a bigger solution than I think it can ever be – and yet we are throwing a lot of taxpayer money at it. That doesn’t mean that I am against ethanol. If you read a post like this, you might come to that conclusion. But I think ethanol is a fine fuel, and if we had a more efficient way to produce large amounts of it, I would happily support that. I strongly support attempts to get the fossil fuel inputs out of ethanol production. In fact, in my current job I keep a very close watch on ethanol developments – ready to jump in if I see one that I think has major long-term potential.
I also believe – as stated in my essay on Biofuel Niches – that corn ethanol may work out well in specific situations. For instance, it may never provide more than around 1% of net U.S. petroleum needs, but it may be able to supply a fair fraction of the needs in the Midwest. But then I also think that a local solution for Iowa – if it must be subsidized – should be subsidized by the taxpayers of Iowa. If the fuel is produced and consumed in Iowa, and the jobs are created in Iowa, then Iowa should support it. Try to scale it across the U.S., and again I think the net contribution will be lost in the noise – and money from taxpayers outside the Midwest won’t be well-utilized. In the latter case you essentially have a transfer of wealth from taxpayers across the nation into the Midwest.
I actually wanted to be wrong about my initial suspicions as I worked through this, because I don’t like the idea that there has been no measurable impact on imports from our massive ethanol ramp-up. But maybe a reader can spot a mistake that will change the overall conclusion.
In this exercise, I used data available from the Energy Information Administration website. I used annual averages to dampen out any noise. I looked at net petroleum imports, which includes those destined for the Strategic Petroleum Reserve (SPR). The reason for using net imports is that this subtracts out the imports that simply went into increased exports. For example, our exports of fuel oil have increased over the past few years, so the imports that ended up being fuel oil exports are excluded.
I only considered data from 2002 through 2007 for two reasons. First, the ethanol ramp-up was pretty steep over those years. An impact should be noticeable as ethanol production tripled. Second, the end of 2007 approximately defines the beginning of the current recession. Imports definitely fell during 2008, but overall consumption fell even more. So inclusion of 2008 would make it more difficult to separate out cause and effect, especially considering the speed at which demand fell. But it will be interesting as we come out of the recession – and as ethanol continues to scale up – whether we eventually see a sustained drop in net petroleum imports.
* While it can explain some of the phenomenon, it can’t explain a whole lot, because most of the energy used to remove the sulfur from oil is derived from natural gas. Some may be cannibalized from fuel gas produced as the oil is refined, and in that case it would show up as an incremental increase in the barrel inputs into a refinery to produce the same amount of net products. That could translate into higher imports in order to keep production steady.
** A barrel of oil contains around 5.8 million BTUs of energy. It takes approximately 500,000 BTUs to process that barrel into finished products, for a net energy content of finished products of 5.3 million BTUs, or 126,000 BTUs per gallon. Ethanol contains 76,000 BTUs per gallon, so one gallon of ethanol is worth 76,000/126,000 = 0.6 gallons of oil.