On Purpose – What’s Driving New Propane Dehydrogenation Projects in North America?

Jun 24, 2019

Global demand for propylene is rising, but lighter crude slates at U.S. refineries and the use of more ethane at U.S. (and overseas) steam crackers has reduced propylene production from these plants. That has led to the development of more “on-purpose” propylene production facilities –– especially propane dehydrogenation (PDH) plants –– in both the U.S. and Canada. More than 2 million metric tons/year of new PDH capacity has come online in North America since 2010, another 1.6 MMtpa is under development, and propane/propylene economics may well support still more capacity being built by the mid-2020s. The U.S. and Canada are bound to retain their positions as exporters of propylene and propylene derivatives.

In a three-part series for the RBN Energy blog, Nexant's Randy Rabenhorst looks at “on-purpose” production of propylene by PDH plants and what the development of these facilities will mean for U.S., Canadian and overseas markets.

A Key Building Block

Propylene –– a particularly useful chemical building block –– is used in downstream petrochemical processes to make a number of important materials like acrylonitrile (used for acrylic fibers and coatings); propylene oxide (used for polyurethanes and other chemicals); oxo alcohols (used in PVC plasticizers and coatings); cumene (used to make epoxy resins and polycarbonate); and isopropyl alcohol (used as a solvent). About two-thirds of the propylene produced is used to make polypropylene –– one of the best-selling plastics, second only to  polyethylene. Polypropylene (or PP) is used extensively in automobiles and in the manufacture of packaging films, bottle caps, fiber ropes, as well as bicycle helmets and diapers. Global demand for propylene has been increasing at an average of about 5.2 MMtpa –– a 3.6% compound annual growth rate (CAGR) –– while North American demand is growing at a more modest average of 390,000 metric tons per year, for a CAGR of 2.2%. However, as we’ll get to in a moment, the traditional “co-product” sources of propylene supply have not kept up with demand –– a situation that has led to the development of on-purpose propylene capacity on both sides of the U.S.-Canada border.

Until a few years ago, propylene was mostly produced as a co-product of petroleum refining or the steam cracking of propane and naphtha. However, as Nexant discussed in On-Purpose Propylene in an Era of Uncertainty and more recent propylene market reports, propylene output from North American refineries has been falling due to lighter crude slates; since the advent of the Shale Era in the U.S., steam cracking has become more focused on lighter feedstocks (especially ethane), thereby reducing propylene co-product production. Further, investments in U.S. NGL export capacity (specialized terminals and ships for transporting refrigerated ethane and LPG overseas) have enabled the conversion to lighter steam-cracker feedstocks in other parts of the world (notably in India, Europe, South America and China), further limiting global propylene production. The remedy has been a focus on increasing propylene stocks with on-purpose production, and, as shown in Figure 1, on-purpose propylene technologies such as propane dehydrogenation (PDH; green bar segments), methanol-to-olefins (MTO, also known as methanol conversion; blue bar segments), and –– to a much lesser extent –– metathesis (orange bar segments) have dominated global propylene capacity additions over the past 10 years.

Global propylene capacity by process
Figure 1. Global Propylene Capacity by Process, 2005 to 2025 (MMtpa)

As the chart shows, steam cracking and refinery sources are historically the largest sources of propylene supply, but global propylene capacity additions from those processes slowed between 2010 and 2015. That pace looks to pick up somewhat between 2015 and 2020, led by new capacity in China, but it’s not expected to be enough to keep up with demand growth.

If we look specifically at recent years, steam cracking, with 55 MMtpa propylene capacity (or 46% of the total) in 2018, is still growing as a propylene source. But since cracking’s primary purpose is cost-effective ethylene production, the increased use of lighter feedstocks (especially ethane and propane) has slowed co-product propylene capacity growth to less than 2% CAGR. Refinery sources of propylene, with 39 MMtpa of capacity (or 32% global share) last year, arise naturally when complex refineries convert the bottom of the barrel (heavier, less valuable hydrocarbons) to gasoline and diesel using fluid catalytic cracking units. The volumes of refinery sources of propylene have grown slightly faster than steam cracker supply in recent years, but that growth pace is projected to slow in the future as demand for gasoline and diesel fuel stagnates due to greater fuel efficiency and vehicle electrification. Hence, the overall market share of propylene supply growth from both steam crackers and refineries has been waning, and North America and the rest of the world will increasingly need on-purpose propylene production to satisfy rising demand. In fact, global propylene capacity additions from PDH plants have accelerated, with capacity jumping from about 2.1 MMtpa (3% of the total) in 2005, to 13.6 MMtpa last year (11.4%). And MTO units have gone from zero in 2005 to 7.1 MMtpa in 2018 (6% of the total).

Global Economic Drivers for On-Purpose Propylene

On-purpose propylene expansion depends upon the availability of low-cost feedstock (either locally or imported), logistics to end-markets, and the cost of plant construction. For instance, MTO units (based on cheap coal to make low-cost methanol) have only been built in China, where coal is inexpensive, and methanol is oversupplied. The economic driver behind PDH units, in turn, is a large spread between propane and propylene prices. These plants have been built in many locations around the globe, including China, South Korea, Thailand, Europe, the Middle East –– and (as we’ll get to) the U.S. and Canada.

The solid lines and left axis in Figure 2 show the historic spread (in $/ton) between propylene (polymer grade) and propane prices in the four regions where PDH units are currently under construction: the Canadian province of Alberta (gray line), the U.S. Gulf Coast (USGC; blue line), Northwest Europe (green line) and Northeast Asia (orange line). For comparison, the chart also provides the price of Brent crude oil ($/bbl) as the dashed purple line, as measured on the right vertical axis.

Propane propylene price spread
Figure 2. Regional (Polymer-Grade) Propylene vs. Propane Spreads ($/ton)

The chart shows that propylene-propane price spreads have generally been the highest in Alberta, followed by the U.S. Gulf Coast and Europe. In the early days of the Shale Era, propane became oversupplied, somewhat stranded, and quite cheap in North America, leading to a wide price differential between propane and propylene. As propane export capacity grew along the Gulf Coast, the cost advantage diminished such that, by 2016, Europe and the Gulf Coast were largely at parity. From a cost-of-production point of view, Alberta is clearly the best place to build a PDH unit in 2019 –– the spread between propane and propylene currently tops $800/ton. But one must also consider the cost of logistics to end-markets and the cost of construction. As it turns out, Alberta is disadvantaged regarding the cost of transporting propylene to China (the largest source of propylene demand) and especially regarding the cost of construction, which is generally higher than most other places due to the province’s geographic isolation and unusually short construction season. 

Alberta’s consistent edge over the Gulf Coast, Northwest Europe and Northeast Asia on the price-spread front in the past four-plus years is driven at least in part by the 2014 reversal (and switch from propane to condensate) of Kinder Morgan’s Cochin Pipeline, which for many years allowed excess propane to move from Alberta to the U.S. Midwest and Ontario via pipe. After the reversal, the increase in the spread between Alberta and the Gulf Coast shown in Figure 2 is largely due to the higher cost for Alberta producers to ship propane by rail vs. lower-cost pipeline. With limited in-province propane demand, and increasing propane production in Alberta, this spread may continue for many years –– positive news for PDH economics in Alberta.

Figure 2 also shows the seasonality for the propylene-propane price spread in Alberta and the Gulf Coast for the period from 2010 through the first half of 2014. The price spread increased in the second quarter each year during this period as propane inventories built up and as the propane price dropped each year after the winter heating season. Seasonality became muted in Alberta and the Gulf Coast after the Cochin Pipeline was reversed and the price of crude oil dropped from more than $100/bbl in the third quarter of 2014 to less than $40/bbl in the first quarter of 2016. In contrast, the spread in Northeast Asia shows much less seasonality and has largely stayed in the range of $300-$550/ton throughout the 2010-14 period. 

PDH Projects in North America

The U.S. has historically remained fairly balanced in propylene production and consumption –– propylene that was produced from co-product sources was consumed domestically for derivative production, with little need for imports or exports. However, during 2010-14, as refineries and steam crackers shifted to lighter feedstocks, a significant amount of propylene capacity disappeared, as shown by the blue and gray bar segments below the zero mark in Figure 3. As a result, the U.S. began to import a portion of its propylene requirements. Canada, on the other hand, has limited propylene demand along with excess by-product propylene production from its steam crackers and refineries. So even though propylene production has fallen as Canadian crackers shifted to lighter feedstocks, Canada is still exporting over 75% of its production.

New North American Propylene Capacity by Process
Figure 3:  New North American Propylene Capacity by Process (in thousand tons/year)

Since 2010, PDH has accounted for the vast majority of new North American propylene capacity (green bar segments), with only modest amounts of new capacity resulting from new ethane steam cracker additions and refinery sources. In that time, three PDH units (all on the Gulf Coast) have commenced operations in the U.S.: Flint Hills (2010 startup, as Petrologistics), Dow Chemical (2015) and Enterprise Products Partners (2018). In addition, three additional PDH units have been approved for (or are under) construction:  Formosa Plastics (2021 start-up) in Texas; Inter Pipeline (2022) in Alberta; and Canada Kuwait Petrochemical Corporation (CKPC), which is slated to begin operation in Alberta in 2023. When fully operational, these six units will total almost 3.8 MMtpa of on-purpose PDH capacity, accounting for about 20% of North American propylene capacity in 2024. It is worth noting that the Flint Hills and Enterprise units are not integrated downstream and therefore sell propylene on merchant markets (reducing the propylene trade deficit). However, the two Canadian units will be integrated directly to new polypropylene units, while Dow and Formosa will use their PDH capacity to feed derivative units and displace merchant propylene purchases. The overall net impact will be continued exports of propylene from both Canada and the U.S.

In Market Analytics: Olefins - 2019, Nexant predicts 2,200 thousand tons (2.2 MMtpa) of additional new propylene capacity in 2023-25, most likely from three new, as-yet-unannounced PDH units. The size and location of these speculative PDH units are still uncertain. In 2013, Enterprise announced that it was considering a plan to build a second PDH unit, but it shelved those plans in 2015 as the propylene-propane spread collapsed. However, Enterprise recently announced that it is finalizing engineering and license arrangements for a second PDH unit and has also indicated that a third PDH unit is not out of the question. Enterprise already has substantial pipeline, storage, production, and terminal assets for propane and propylene, so building out its position in this area would not be a great surprise, even though it is not integrated downstream. In addition to Enterprise, there are a number of propylene derivative producers that may consider a PDH unit to improve their propylene cost position. Perhaps first among these is Dow Chemical, which is also said to be considering a second PDH unit, possibly using its own proprietary technology. While there have been no firm announcements, other companies with propylene deficits are said to be considering on-purpose propylene units. These include: LyondellBasell, Total, Braskem, Ascend and INEOS on the Gulf Coast, as well as Alpek in Mexico.

This blog discussed PDH projects in North America and the economic drivers for viable propylene production via PDH. Subsequent blogs in this series will analyze competitiveness and drivers for on-purpose propylene production via MTO and metathesis and compare them with the PDH process.