By Robert X. Cringely
bob@cringely.com External Link
When this column first appeared in 1997, the price of oil in 2007 dollars was about $20 per barrel. This week it is $122 per barrel -- more than six times as high. Many things besides the price of oil have changed in those 11 years. Thanks to Moore's Law a new PC today is typically 25-30 times as powerful as the machines we had in 1997 yet costs less than half as many dollars to buy. Consider inflation and the cost of computing is even lower. Our new PCs are substantially more energy-efficient on a per-MIPS basis, too. Though the bigger shift of attitude might simply be that now we care about how much power our computer is using versus simply not noticing a decade ago. Back then data center power consumption was inconsequential compared to the cost of Internet bandwidth. Today Google builds its data centers where the power is, not where the fiber or even the users are. Power is everything in modern computing and is becoming ever more important to our lives in general. The high cost of energy is starting to cause real pain in our society and real pain is, unfortunately, about the only incentive strong enough to make us change our ways.
I know a little bit about the oil business, which is at the root of our energy crisis today. At one point my job was to write about the international oil industry. I worked off and on in the Middle East back then and attended OPEC meetings in Vienna and Geneva during an earlier oil crisis when prices went for a moment to $43 per barrel and we all held our breath then, too. I was sitting in the lobby of OPEC headquarters in Vienna that day when Carlos "The Jackal" came in the door, told the sleepy off-duty Vienna cop providing security that he was from "the Palestinian delegation" and walked right into the big meeting, taking the oil ministers hostage.
Given that I know a little bit about the energy business, then, and I still have friends in it, here is what's going to happen over the next 2-3 years. The price of oil is going to come down substantially, but probably never to pre-9/11 price points. At least half of the current price for crude oil is driven by speculation and market manipulation as it was during the original oil crisis of 1973 (I have an interesting story about that in this week's links). But unlike '73, today our flexibility is less and our excess capacity is less, too. High prices will cut demand, spur exploration, and force governments to open new areas for exploration, but it is doubtful that we will EVER see oil prices under $60 per barrel again.
This is not all bad. Just as high oil prices spur exploration they also encourage conservation. With $2.50-$3 gas with us probably forever, we're finally starting to learn to do things somewhat differently, though it isn't at all clear to me whether these lessons will stay learned after prices subside somewhat.
Which brings me to the moral of this story -- the importance of platforms and standards, and when and how to abandon or change them. This applies to ANY capital-intensive technology, whether it is computing or energy or transportation.
The life expectancy of a car in the U.S. is about 10 years, during which it will pass through an average of three owners. I use cars as an example because they are an intrinsic part of any American energy crisis and we generally all own one. This automobile life expectancy means that any technology improvement has to trickle into the market because only about 10 percent of the total fleet is replaced each year with new cars -- the only cars that are likely to have the latest technology. So if some car company comes up with a way to get 100 miles per gallon, it is likely that no more than 10 percent of us will be getting mileage like that a year from now. The rest of us are stuck with old technology until we can afford to change: we're on the old platform.
Platform, in this automotive example, means some significantly different technology that offers real advantages though usually at a cost. Hybrids, diesels, electric cars, fuel cells, hydrogen, and ethanol cars are all examples of platforms.
If we want revolutionary change -- change where nearly everyone moves to the new platform in short order -- that is usually going to require heroic action on the part of government or the occasional mad scientist. If the mad scientist were able to offer a car that got 100 miles per gallon, was safer than the current standard, yet cost substantially LESS to buy, then maybe more of us would transition more than the traditional 10 percent replacement model suggests. Governments, on the other hand, could simply outlaw the old cars and force us to upgrade, though it still might not happen if we couldn't afford the new cars.
What's key here is the push and pull of platform change. We see this all the time in computing where somebody comes up with a clever new idea but for that idea to succeed we all have to get new computers. How likely is that? Well it depends on how great the improvement being offered. With computers I can tell you that the improvement has to be pretty darned substantial to inspire us all to jump. That's because Moore's Law is going to give us a 100 percent improvement anyway on our next PC without having to throw away any software or peripherals. So inspiring a revolution in computing generally requires a performance improvement of 10X or more.
There is a similar effect to be seen in the adoption of new energy technologies. That list of platforms I rattled off a few paragraphs back offer advantages but not without cost. Hybrids and diesels are the easiest to accept, but both are a bit more expensive. Electric, fuel cells, hydrogen, and ethanol are more expensive, too, but they also require infrastructure changes like finding new ways to manufacture, transport, store, and sell fuel. You won't go on a long road trip in your electric car until there are reliable places to plug in and recharge, for example.
This is our dilemma: we want to make radical energy improvements but these typically require expensive platform changes and platform changes can take a decade or more to happen.
A better solution would be to leave the platform alone and find a single variable that could be changed for everyone practically overnight.
Cars are the key to U.S. energy consumption. The dominant automotive platform here, whether you drive a truck, a car, or a motorcycle, relies on gasoline-fueled internal combustion engines. That's the platform we are unlikely to change quickly. So how do we leave that platform intact and unchanged, ask nobody to significantly sacrifice, yet still achieve the noble (and Nobel) goals of lower fuel consumption, lower greenhouse gas emissions, lower pollution levels, dramatically lower oil consumption, lower cost, and lower geopolitical vulnerability for our country? There's only one way I know to accomplish this: change the fuel.
This happened to a certain extent in Brazil during the '70s and '80s by embracing ethanol. But ethanol has less energy per gallon so fuel consumption goes up and mileage goes down. Ethanol can't be shipped in pipelines also used for oil. Cars have to be modified to run on it and even then there are issues about internal corrosion. Ethanol is far from perfect. What's needed is a replacement for gasoline that looks and feels and tastes just like gas to your car but isn't made from oil. Then the platform could remain completely unchanged yet my 1966 Thunderbird (and the world) could benefit starting with the very next tankful.
There is such a fuel, developed by a husband and wife team of scientists working in Indiana in cooperation with Purdue University. This new fuel, called SwiftFuel, is right now intended for airplanes, not cars, but the lessons are the same.
Piston-powered airplanes have a unique fuel problem. Their high-compression air-cooled engines require higher-octane fuel to avoid destructive engine knock. This higher octane is achieved through the use of tetraethyl lead as a fuel additive. Remember lead was outlawed from car gas in the U.S. more than 30 years ago to good effect: we all have significantly less of the toxic metal in our bodies than we used to. But lead is still used in aviation fuel, which accounts for an infinitesimal portion of total U.S. gasoline consumption. Lead is on its way out for aircraft use, too, with international treaties scheduling its demise in 2010.
If we aren't going to retire all the little airplanes in America -- force a total platform change -- we'll have to come up with a replacement for tetraethyl lead. The additive used most for this is ethanol added to gasoline to bump up the octane number. But ethanol does a number on seals and hoses typically used in aircraft to an extent that it is specifically prohibited by the Federal Aviation Administration from being used in certified aircraft. At the same time, U.S. energy policy is moving toward the mandated use of ethanol in ALL motor fuel, meaning there may be nothing available two years from now to fuel your Piper Cub.
Enter SwiftFuel, the Splenda of motor fuels because it is made from ethanol yet contains no ethanol. SwiftFuel is the invention of John and Mary Rusek from Swift Enterprises in Indiana. To your airplane SwiftFuel looks and tastes just like gasoline. It has an octane rating of 104 (higher than the 100 octane fuel it replaces) yet contains no lead or ethanol. SwiftFuel mixes with gasoline, can be stored in the same tanks as gasoline, and be shipped in the same pipelines as gasoline. It is made entirely from biomass, which means it has a net zero carbon footprint and does nothing to increase global warming. Its emission of other polluting byproducts of burning gasoline are significantly lower, too. SwiftFuel has more energy per gallon than gasoline so your airplane (or your car) will go 15-20 percent further on each gallon.
Oh, and based on an average $1.42 per gallon wholesale cost for the ethanol used as its feedstock, SwiftFuel costs $1.80 per gallon to produce, meaning that it ought to be able to sell for $3 per gallon or less no matter what happens in the Middle East.
Heck of a deal.
The ethanol used to make SwiftFuel can be any type, according to Mary Rusek, president of Swift Enterprises. The pilot plant they are building in Indiana will, interestingly, make ethanol from sorghum, not corn. The Ruseks claim that sorghum, which isn't a typical U.S. crop, can produce six times the ethanol per acre of corn, turning on its head the argument that ethanol production consumes more energy than it produces. China, the third largest producer of ethanol after Brazil and the U.S., is switching entirely to sorghum for its ethanol production.
The FAA is already testing SwiftFuel with the goal of approving it for use without modification in all aircraft, leaving the platform unchanged while improving its impact on almost any scale. Hopefully by the 2010 cutoff for tetraethyl lead SwiftFuel will replace the 1.8 million gallons of 100LL aviation fuel used every day.
"But what about cars?" I asked Mary Rusek. "We don't say much about that," she replied. "The aviation fuel market is tiny and has a real need we can fulfill so everyone wants us to succeed. Cars are different and we don't want to make any enemies."
I hope that SwiftFuel is a success. I hope it fulfills all Mary Rusek's claims. But if SwiftFuel doesn't succeed, I also hope that isn't because entrenched oil interests kill it. Yet I don't think many of us would be surprised if that is exactly what happens.
Original Article
