Technical details sometimes turn out to be just that – details. Sometimes, though, they turn out to be much more. A longstanding technical debate over how to think about natural gas liquids (NGLs) may be in the process of moving from something marginal to something that really matters.
NGLs are the light hydrocarbons other than methane that produced as part of natural gas; the most abundant ones are ethane, propane, and butane. Official statistical agencies tend to lump them together with crude oil when discussing long term trends. Thus, when the EIA or IEA say that world oil production has been rising consistently for the past decade, they are talking not only about crude oil, but about NGLs too. Similarly, the various cornucopians who have been heralding a coming age of oil abundance typically include NGLs in their totals.
This approach has been repeatedly challenged, particularly but not only by peak oil believers. It’s being questioned once again as talk of oil abundance, backed strongly by growing NGL supplies, heats up. Jim Hamilton, in a post last week, put the criticism this way:
“You can’t drive your car with ethane, and it’s currently almost exclusively used to make ethylene for the chemical industry. [Propane…] is currently used primarily for heating and petrochemicals. Ethane and propane between them account for 82% of the increase in U.S. NGL production since 2005…. The bottom line is that it doesn’t make sense to think of NGLs as something that you can simply add to traditional crude oil production.”
I’ve been thinking about this problem for a few months, and while there’s a lot to be said for Hamilton’s view, my sense is that it goes too far. How much so isn’t clear, and would benefit from some solid quantitative analysis.
There are two basic problems with the standard NGLs-aren’t-oil argument.
This first problem concerns what goes into a petroleum refinery. NGL skeptics love to point out that ethane, propane, and butane are of no use in cars and trucks. That’s true, but oil isn’t of any use in cars or trucks either. Drivers want mobility, and for now, that mostly means gasoline and diesel. Those fuels are produced by refineries. The right question to ask, then, is this: how does ethane, propane or butane compare to oil as a refinery feedstock? If they are interchangeable, then NGLs should be thought of as no different from crude oil. If NGLs are useless as refinery inputs, they should be treated as something different.
It turns out that ethane it pretty useless as a refinery feedstock, but propane and butane aren’t. Both (along with propylene and butylenes, products derived from propane and butane) are inputs into refinery alkylation units that produce high octane blending components for refined products. (They might be useful for something different – I’m whatever you call the opposite of a refining expert, and would love any feedback from people who know more about all this than I do.) What I don’t know is how much new propane or butane could be absorbed into gasoline and diesel manufacturing before the system became saturated. In the short run, the answer is probably “very little”, since alkylation units are expensive and take time to bring online. In the long run, that constraint weakens, which should change the equation. How much so is something that requires some real quantitative work.
The second problem comes when you look at the other end of the refinery, i.e. what comes out of it. One of the biggest products, after gasoline and diesel, is naphtha, a petrochemicals feedstock currently produced at a rate of about five million barrels a day globally. Ethane, propane (converted to propylene) and butane (converted to butylene) are all potential substitutes for naphtha in many industrial processes. Refineries, meanwhile, can install equipment that reduces the amount of naphtha produced while boosting output of other products, including gasoline and diesel for final use. (Yes, this is vague: like I wrote above, I’m sounding through this in the hopes that others will weigh in with more technical information.) Abundant NGLs, then, can shift refiners’ economic incentives toward producing more gasoline and diesel, even if the NGLs themselves never end up in anyone’s tank. Once again, the question isn’t whether this dynamic exists – it does – but how far it can go before the system is saturated.
Bottom line? It’s reasonable to treat small quantities of NGLs as equivalent to crude oil. But as NGL production grows relative to crude oil output, that equivalence should progressively break down. How much, how quickly, and where we are on that spectrum isn’t clear to me. It’s something, though, that some careful quantitative analysis should be able to help us understand.