Building the Next Refinery

Hurricanes Katrina and Rita have exposed many problems with our energy infrastructure, not the least of these being the concentration of our refining capacity in a region likely to be hit by more large hurricanes in the future. It took more than two decades of environmental regulations, permitting policies, industry consolidation, and corporate strategies to create this situation. This cannot be rectified overnight, and I'm not even sure we should try, given the high cost and the desirability of reducing our reliance on oil in the decades ahead. In the meantime, though, we can try to address the underlying capacity problem in a way that would mitigate the national impact of a future Class 4 or 5 storm in the Gulf Coast.

A few weeks ago I described how we got to this point, with so many refineries, pipelines, terminals, and production platforms lining the Gulf Coast. It wasn’t always this way, either. Thirty years ago, the refining industry looked very different, with a larger number of smaller refineries dispersed around the country, mostly located in proximity to oil fields, markets, or both. The 1970s energy crisis changed all that, helped along by the depletion of mature oil deposits in states like Kansas, Oklahoma and Pennsylvania.

When oil rose from $2/barrel to $10, then $30, efficiency trumped proximity. The big refineries got bigger, while many of the smaller ones became uneconomical and failed to attract the investments required to keep pace with increasingly stringent environmental regulations. After a couple of decades of this, just under half of the country’s refinery capacity is clustered into only five areas: Chicago, L.A., Philadelphia, San Francisco, and Houston/Beaumont/Port Arthur.

Even if we were to eliminate all the obstacles to building new refineries in this country, we are not going to reverse that concentration. US oil production is in steady decline, only partially offset by increasing oil sands production in Canada, so any new refinery would have to rely on a mostly offshore crude oil diet. The alternative is having someone else refine the oil and then importing the products. That might reduce some of our vulnerability to energy disruptions from weather and earthquakes, but it brings a host of other concerns. Besides, this is the strategy we’ve chosen by default.

Post-Katrina, there's been a lot of talk about building more refineries. The President mentioned it in a speech Monday. There's even a proposal for a mega-refinery at Cushing, OK (thanks to Mel at Engine of the Future for the info.) The location offers some advantages, at least for crude oil supply. Cushing is already the major oil pipeline hub in the Mid-continent, as well as the delivery point for the NYMEX WTI futures contract that gets so much attention. When Enbridge's Spearhead pipeline is completed, it will be able to receive the full range of Canadian crude, including the heaviest synthetic crude and bitumen. And it's hundreds of miles from the coast, so that any hurricane would have dissipated quite a bit before reaching Cushing. Unfortunately, it's in the heart of Tornado Alley.

If we're thinking about where to locate another refinery, we'd better think about how we're going to induce someone to invest in building it, unless we're suggesting the government go into the refining business--something that other countries' governments have wisely spent the last 20 years getting out of. The problem is that the companies that own refineries have little incentive to build another one, even ignoring the scale of investment required.

Oil refining is extremely capital-intensive. The last new refinery in which I was involved was the 135,000 barrel per day Star Refinery in Thailand. It cost $1.7 billion 1995 dollars. Until the last few years, however, refining has been a low-margin, low-return business. I don't see the major oil companies queuing up to build more in the US, even if the permits were handed to them on a platter. They've sold off dozens of refineries, largely because equity analysts kept telling shareholders what a crummy business segment this was, and that the majors should invest only in exploration and production of oil and gas, and a bit in marketing the products. Their boards of directors listened, and as a result, the largest refiner in the US today is not Exxon or Shell, but Valero, an independent that has bought and merged itself into the number one spot.

How keen would Valero be to build a new refinery? Perhaps more than most, but I suspect they'd start to hear concerns from their shareholder base. It's one thing to buy up existing capacity and say that you'll run it more efficiently. It's quite another to build expensive new capacity that by its mere existence will reduce the operating margin of every other plant in the markets it serves. In a way, the refining business is like the restaurant business, where the person who builds a brand new restaurant, with a beautiful new kitchen, typically goes broke. By the time the third owner has bought it, at pennies on the dollar for the kitchen hardware, he has a chance to make a profit.

I'm not sure a producing country would want to come in on this, either. Saudi Arabia already owns an interest in Shell's refining system in Texas and the Southeast, via a joint venture. Their appetite for more is doubtful. Citgo, another large refiner, is owned by the Venezuelan government, which recently expressed an interest in selling. If you look at the other sources of the oil imported into the US, there aren't many obvious choices for large investments in refining, other than possibly Russia. Lukoil has already bought some service stations in the Northeast and is selling under their own brand, so it's just possible they might want to build a refinery. How well would that go over in Washington?

You can see that building another refinery isn't going to be easy, unless the government is prepared to offer large inducements, including tax holidays, investment tax credits, and low-interest loans. It certainly won't happen quickly, either, and I haven't even mentioned the various interests that would line up to oppose such a development. While we do need more refining capacity, the 1-2% per year "capacity creep" that we've been able to rely on until recently might just be enough to get us by, if combined with sensible policies to reward companies for holding larger inventories, closer to markets. At the same time that we try to buffer the nation's petroleum product supply system from future shocks, though, we need to learn the lessons of over-concentration of infrastructure and plan for alternative energy on a more dispersed, less vulnerable basis. That's a subject for a future blog.