The Shale Revolution has had a profound impact on U.S. NGL markets by vastly increasing production and by lowering NGL prices relative to the prices of crude oil and natural gas. That has been good news for the nation’s steam crackers, the petrochemical plants that have enjoyed low NGL feedstock prices since 2012. But NGL markets are in for some big changes as new U.S. steam crackers coming online over the next two years will be competing for supply with export markets, raising the specter of higher NGL prices—a good thing for NGL producers, but not so for petrochemical companies. How this plays out will be determined by the feedstock supply decisions petrochemical producers make as NGL prices respond to rapidly increasing demand. Today we begin a series on how steam cracker operators determine day-by-day which feedstocks are the most economic, and on the factors driving the value of ethylene feedstock prices.
NGL production, the pricing of NGL “purity” products (ethane, propane, normal butane, isobutane, and natural gasoline), and steam cracker economics are frequent topics in the RBN blogosphere. Nearly five years ago, when U.S. NGL production was just beginning to take off, we ran a series on feedstock economics (Let’s Get Cracking) that provided a primer on ethylene production and discussed this underlying principal of feedstock acquisition in the steam cracker industry: The best cracker feedstock is the one that will produce the highest ethylene margin possible, after deducting byproduct credits. Two years later, in the midst of the biggest run-up in NGL production in U.S. history, we took an even deeper dive into the NGL/steam-cracker world with our What’s Crackin’ With Steam Crackers Drill Down Report (available to RBN Backstage Pass members). In that report, we discussed the fact that the margin for producing ethylene with ethane (the lightest and most prolific NGL) had just hit an all-time high (~70 cents/lb—a record that still stands today) due to the combination of a low ethane price (~24 cents/gal) and a high price for ethylene (76 cents/lb). Well, a lot’s changed since then. NGL production volumes remain high and ethane prices are still on the low end, averaging only 19.5 cents/gal in 2016 (although just last week ethane prices hit a two-and-a-half-year high of 28 cents/gal). However, ethylene prices have tumbled (to about 25 cents/lb) and so has the margin for producing ethylene with ethane––that margin averaged ~19 cents/gal for 2016, and now stands at only ~15 cents/lb, down nearly 80% from the September 2014 pinnacle. Over the past year or so we’ve posted several blogs covering the changes affecting the NGL/steam cracker world, including Beyond Hypothermia (on how the falling price of crude oil and naphtha—an oil-based feedstock alternative, especially overseas—slashed worldwide ethylene prices), and our Drill Down Report on NGL/steam cracker infrastructure, It’s Not Supposed To Be That Way Drill Down/Part 1 and Part 2. In that report, we discussed our view that the three key assumptions behind the build-out of new U.S. ethylene production capacity and NGL export terminals—1) U.S. NGL production would continue to grow, 2) U.S. NGL prices would remain low, and 3) crude oil and naphtha would stay expensive—may no longer hold water. A year and a half have passed (a lifetime in “NGL-market years”) and the addition of new ethylene-production and NGL-export capacity is looming ever closer, so it’s high time to take a fresh look at the economics of steam cracker feedstock selection.
We’ll begin with a brief review of where things stand with NGL and steam crackers and how we got here. As mentioned above, U.S. NGL production has increased significantly during the Shale Era, rising from just over 2 MMb/d in 2008 to more than 3.7 MMb/d in 2016 (Figure 1).
The vast majority of NGLs is produced via the processing of raw natural gas—including “associated” gas produced with crude oil and condensate—at gas processing plants (blue bar segments), with most of the small balance (green bar segments) coming from propane produced at refineries. NGL production from gas processing plants has been growing quickly the past few years because an increasing share of gas production has been coming in the form of “associated” and “wet” gas (green layer in Figure 2), both of which typically are laden with large quantities of NGLs. A significant portion of this wet gas production has been in the Utica in eastern Ohio and the wet part of the Marcellus in western Pennsylvania and northern West Virginia. Associated gas production occurs in the Permian Basin, the Eagle Ford, the Bakken and other major crude-focused plays.
Figure 2; Source: RBN
In the Energy Information Administration’s (EIA) September 2016 numbers, gas-plant production of ethane averaged about 1.2 MMb/d, while propane production came in about the same 1.2 MMb/d. (The balance of the NGLs, totaling 1.1 MMb/d, are either normal butane, isobutane or natural gasoline.) Note that only a portion of the ethane in the raw gas stream fed into processing plants is separated out into liquid ethane for sale to steam crackers—gas producers and processors also have the option of leaving (or “rejecting”) at least some ethane into the pipeline-quality gas that emerges from the back-end (or “tailgate”) of their gas plants (see You Ain’t Seen Nethane Yet). The decision about how much ethane to reject into pipeline gas depends in part on the need to meet gas pipeline standards (too much ethane, too high a Btu/MMcf ratio), and in part on the price of natural gas vs. the price of purity ethane. If ethane is worth more if it’s removed from the raw gas stream with other NGLs and fractionated into a purity product (see Talking ‘Bout My F-f-fractionation), more ethane will be separated out and not be rejected. More on this later—ethane rejection is an important factor in the economics of steam cracker feedstock selection.
About half of U.S. NGL production is used as feedstock for steam crackers. That includes virtually all of the ethane that is not rejected into natural gas, as well as about one-quarter of total propane supply, 15% of normal butane and 10% of natural gasoline. (Isobutane, the other NGL, is used primarily in a refinery process called alkylation to make a high-octane, low-vapor-pressure motor gasoline blend component called alkylate. It’s also used in various specialty markets such as propellant for spray products and as a refrigerant, but in the U.S., is almost never used as a steam cracker feedstock.) As you can see in Figure 3, there has been a noteworthy shift the past few years in the mix of feedstocks used in U.S. ethylene plants, some of which are configured to “crack” only specific feedstocks and others of which are more flexible. Until the start of the Shale Era in 2008-09, naphtha and gas oil (heavier feedstocks produced at refineries; purple and aqua bar segments, respectively, at the top of the bars in Figure 3) together accounted for 30% or more of total feedstock volumes. Over the past eight years, however, their combined share has fallen sharply, and they have been largely replaced by lighter (and generally less-expensive) NGLs, especially ethane (blue bar segments) but also propane (red bar segments) and normal butane (green bar segments). Very small volumes of natural gasoline (whose characteristics are similar to naphtha) are used as well and are included in the purple segment along with naphtha.
Figure 3; Source: RBN, Jacobs Consultancy Hodson Report
As of 2016, ethane accounted for ~61% of total feedstock volumes in U.S. steam crackers (up from ~41% in 2005), while propane accounted for ~24%, normal butane ~8%, naphtha/natural gasoline ~6% and gas oil ~1%. Because each of these feedstocks has a different chemical structure, the amount of ethylene produced from, say, 1,000 barrels of ethane is considerably different than the amount of ethylene produced from a similar volume of propane or butane, etc. Each feedstock also produces different amounts of steam-cracking byproducts such as propylene, butadiene, benzene, fuel gas and hydrogen, each of which has a value of its own. The steam cracker feedstock selection model we will be discussing in this blog series—and testing against current prices and prospective price scenarios—will factor in not only the price of the various feedstocks, but the value of the ethylene and byproducts mix that each feedstock produces. Only by considering all these factors (the cost of the feedstock going in, and the value of products coming out) can we determine which feedstock (or combination of feedstocks) would yield the highest margin at any specific point in time.
Taking a fresh look at feedstock selection makes sense as 2017 begins because more than a dozen U.S. steam cracker projects that together would add more than 20 billion lb/year of new ethylene production capacity are either under construction or will be soon. As a group, these projects (most of them along the Gulf Coast in either Texas or Louisiana) could increase ethane demand by about 600 Mb/d by the early 2020s. Factor in the possibility of only modest growth in ethane production, continuing ethane takeaway constraints and rising ethane exports (overseas and to Canada) and, well, this could get very interesting.
In upcoming episodes in this series, we will discuss the feedstock selection model; the timing of incremental steam cracker demand for ethane, propane and other feedstocks; and the role of ethane and propane exports. Then we’ll assess what all that means for ethane and propane demand and prices, and use our model to consider what the highest-margin feedstocks of 2018, 2020 and 2022 might be.
“You’re the One That I Want,” performed by John Travolta and Olivia Newton-John for the film version of the musical Grease, was a #1 hit in the U.S., the UK and several other countries. It is one of the best-selling songs of all time, with worldwide sales estimated at 15 million.