Petrochemicals: Time To Stop Investing – For Ever?
SEE LAST PAGE OF THIS REPORT Graham Copley / Nick Lipinski
FOR IMPORTANT DISCLOSURES 203.901.1629/203.989.0412
April 16, 2018
Petrochemicals: Time To Stop Investing – For Ever?
- Independent chemical companies should put large capital investment decisions on hold immediately and wait to get a better view of how refiners and oil majors are going to cope with expectations for lower gasoline and diesel demand, possibly starting as soon as 2022.
- A broad refining move from gasoline to chemicals and chemical feedstock could cause significant petrochemical oversupply from 2022/23, increasing through 2030, well within the investment horizon of any chemical company looking at what to build next and where.
- A 10% drop in gasoline/diesel demand by 2030 could swamp the chemicals markets if refineries redirect to chemicals and chemical feedstock.
- This move could supply 100% of propylene and aromatics growth and most ethylene growth – minimal additional ethylene from shale gas needed – no additional propane dehydrogenation (PDH).
- A defensive move by refiners or integrated oil companies looking to protect/generate captive oil/refining demand would have an economic justification that could likely not be matched by stand-alone investment by traditional non-integrated chemical companies.
- Possible Conclusions:
- Aromatics in oversupply – good for buyers of benzene and xylenes – TSE, 1COV, OLN, ASIX, HUN, polyester producers in general.
- Propylene in oversupply – PDH plants closed – good for Braskem, HUN, OLN, 1COV, EMN (if the company sells its ethylene plants), Arkema.
- Ethylene – probably a flatter cost curve – otherwise too many variables; likely better to be an ethylene buyer than seller at the margin – WLK.
- Inorganics – chlorine, TiO2, etc. largely unaffected.
- Strategic options will be very different company by company.
Source: BP Review of World Energy – SSR Models
The changes coming to the automotive world have far reaching implications for many industries, with as many scenario models as there are those interested or invested in the topic. While some of the models appear to be sophisticated and drive all sorts of demand conclusions for a variety of fuels and materials, everyone struggles with the same problem: what are the right input assumptions. Today the range of possible assumptions is so wide that you can generate almost any 10-12 year set of conclusions that you could imagine. Here we deal with just one aspect – falling demand for gasoline and diesel, whenever that starts and whatever the decline rate may be.
Aramco’s recent comments about plans for the Motiva refinery and for a new integrated refining/petrochemical initiative in India show one of the possible outcomes and is a very risky outcome for the basic chemicals industry.
- Aramco is interested in securing oil demand – these projects would consume as much as 20% of Aramco’s current crude production on an equivalent basis.
- Plus, there is a Second Sadara project in the planning stage.
- Shell is also in the news this week discussing the possibility of crude demand falling by only 1% per year post 2020 in part because of shift towards chemicals.
- If Shell is discussing a 1% decline in crude demand per year – be sure that the company is planning for something more severe.
- Aramco’s Motiva (until recently a JV with Shell) is talking about adding significant chemical capacity to the existing refining footprint including both aromatics and possibly a new ethylene unit.
- The proposed project in India would consume 1.2 million barrels a day and produce 18 million tons a year of chemicals and plastics.
All refiners and oil producers, we suspect, are already planning for two things that they fear the most: peak oil demand and declining demand for refined products. With that in mind they have two objectives, protect their oil demand and/or refinery throughputs.
Oil demand may peak, but chemical demand is likely to continue to grow. Reconfigured, refineries can make a lot of chemicals – ensuring throughput and crude oil consumption. The world processes over 4.5 billion tons of crude oil a year, much on some very sophisticated equipment designed to upgrade as much crude as possible to the more valuable products – today gasoline, jet fuel and diesel. Roughly 60% of global refining output is for gasoline and diesel, more biased to gasoline in the developed world and diesel in the emerging economies.
- 60% of 4.5 billion tons of crude is 2.7 billion tons
- 10% of this is 270 million tons – or 5.5-6.0 million barrels per day
Refiners are already significant suppliers of chemicals, with most benzene, toluene and xylenes coming from reformers, most of which sit in refineries (there are some dedicated units operated by chemical companies in Asia). The FCC unit, common to most refineries, is already a major supplier of propylene. The unit also produces ethylene, but the concentrations are small and it is generally consumed as a fuel. These facilities are optimized for gasoline, as the most valuable product, but catalyst choice and temperature can maximize propylene and ethylene output with no major capital investment. Significant capital investment would be needed to extract the ethylene from a largely methane and ethane stream.
Exhibit 2 shows the output from an average US refinery. Even though refineries produce more than 50% of the propylene in the US and more than half of the aromatics, these account for a very small part of the barrel, and are picked up in “hydrocarbon gas liquids” and “other oil for Chemical use”. Outside the US the “naphtha for chemicals use” fraction would be significantly larger as so much US ethylene is produced from NGLs (including condensate) rather than refinery based naphtha. The US refiner is likely more exposed to declining gasoline and diesel demand than refiners outside the US because of the current refinery demand profile – however, the US already has significant ethylene and derivative surpluses.
Global demand for ethylene, propylene, benzene, toluene and xylenes is around 400 million tons a year. If we see refineries reconfigure to increase the chemical proportion of their output by 2030 to equal an additional 6% of throughput, we could add 60% to supply – mostly propylene and aromatics – hence the differences in Exhibit 1. However, the ethylene increase would require dedicated (and expensive) investment in ethylene capacity – effectively to consume lower or falling value refinery streams. To achieve the step up in ethylene production suggested in Exhibit 1, and shown on an oil equivalent basis in Exhibit 3, we would need to see more than $300 billion of capital spending over the next 12 years. While this seems like a high number, with Aramco talking about $44 billion for the India project alone, it is no stretch to see how we would get there if other integrated oil companies or refiners (most likely in JV with traditional chemical companies) follow suit.
It may be better for new Dow to look at someone like Marathon or Valero as a possible partner for the next US ethylene investment rather than chase more ethane – but clearly the next question is economics which we will discuss later, as it may be better for new Dow to look at something more comprehensive strategically.
Note: this is a 2022-2030 problem, but any traditional chemical producer considering new capital spending today should stop all planning and make friends with a refiner or oil major. We could:
- See non-traditional players enter the market – Motiva is part of Aramco, but other refiners have the same problems – Valero, Marathon and Tesoro in the US for instance.
- See big oil/refining buy chemicals for demand – new DOW, LYB and Ineos would be more obvious targets, WLK less so because of the ethane bias to its ethylene production.
- It might make much more sense for LYB to consider a large integrated ethylene investment at the Houston refinery than to chase high priced acquisitions. However, today, such an announcement would likely be just as unpopular with investors as the Schulman deal was.
We recognize that we are scratching the surface of an almost unquantifiable set of questions today. This is the first in what will be a series of pieces that address the potential outcomes. We are more than happy to address specific client requests on a bespoke basis.
Source: BP Review of World Energy – SSR Models
Too Much Hydrocarbon
What is summarized above sounds both logical and scary, but at the end of the day economics are going to play a big part – we still have cheap, shale based, NGLs in the US and an ethane cracker is a much cheaper and less complex undertaking than a fully flexible unit, which needs to be integrated with a refinery, and run as a part of the optimization of that refinery to make a reasonable return.
Plus, from a logistical and potentially economic perspective, the US does not need more ethylene capacity, as what has been recently built and is planned to be built in the US already far exceeds local demand growth. There are so many unknowns as we look at possible scenarios from 2022 to 2030, but at least one would have a lowering of crude oil based ethylene costs versus US NGLs, either because the refinery fractions drop in value or because crude oil prices decline. Any flattening of the ethylene cost curve, which as shown in Exhibit 4 is very supportive of US shale today, could leave the US with a lower competitive edge, and a huge surplus.
Source: IHS, Woodmac and SSR Analysis
Refining 101 and What It Could Mean for Chemicals
Exhibit 5 shows a crude (pun intended) diagram of a modern refinery. There is already a chemical interface and potential as follows:
- An alkylation unit consumes refinery produced propylene and/or butylenes – these may be worth more extracting for the chemical industry than blending on site in the future.
- Catalytic reforming creates benzene, toluene, xylenes and other aromatic compounds – these have high octane, but are not a favorite with the environmental lobby as benzene has health concerns.
- These units can be optimized to make more or less benzene and other aromatics depending on catalyst operating conditions – refiners would need to invest in separation facilities or expand existing separation facilities to increase the conversion to chemicals and away from gasoline – Motiva is discussing this specifically.
- Catalytic Cracking (FCC) is already a significant source of propylene – accounting for roughly 30% of global production (50% of US production). These units are run to make gasoline components, but propylene and ethylene are by-products.
- The units tend to make olefins (unsaturated hydrocarbons) and there is often a hydrotreating plant between the FCC and the gasoline blending step to increase the octane value – the process creates a lot of hydrogen but uses it again on the gasoline component fraction.
- FCC propylene is generally a 70% stream – the balance being propane. It can be used on site in alkylation – used as a dilute stream to make cumene and isopropanol, or purified to make other propylene derivatives.
- FCC ethylene is produced in much smaller volumes and is generally used as a fuel given the expense of extracting it from a larger stream of ethane. Less than 2% of world ethylene is currently extracted from FCC streams.
- But… FCC catalysts and temperatures can be varied to produce more or less propylene/ethylene – historically a side bar on optimizing for the best mix of fuels for gasoline and heavier oils, but potentially more of a driver if gasoline becomes less valuable. To extract the ethylene, refiners would need a cryogenic compression train and a tall separation tower. The compressor would be the big-ticket item.
- FCC catalyst manufacturers are already working on higher light olefin producing catalysts either at the behest of refiners or in anticipation of a possible change in requirements.
- Hydrocracking produces no olefins as the hydrogen is added to minimize the production of unsaturated hydrocarbons as the chains are broken down. It is unclear whether these units could be converted economically into conventional ethylene units – they could certainly crack the feed, but all of the separation equipment is missing, and again the compression step is expensive.
- If hydrocracking were reduced in the US and FCCs run for maximum light olefins, we would change the hydrogen balance in the US Gulf which could negatively impact the industrial gas companies – like everything else – post 2022.
- A more conventional route is simply to build a flexible ethylene unit – capable of consuming any or all of the streams outlined in green in Exhibit 5 – this is likely what Motiva is hinting at. These units are expensive – much more so than the ethane only units we have seen in the US recently – but they give a refiner many degrees of freedom when it comes to optimizing stream values. The ethylene plant, run optimally, can become the “trash can” for the site, consuming the least valuable streams of refined products or intermediate streams on any given day.
- Few companies do this well, in our view, with ExxonMobil the poster child for how to add the most value.
- These facilities produce a lot of propylene as well as significant volumes of butadiene, butylenes, aromatics and components that flow back into gasoline – as well as a lot of hydrogen.
- The refinery based propylene and aromatics sources – plus the propylene and aromatics that come from a flexible ethylene plant – drive the conclusion that propylene and aromatics could be more oversupplied than ethylene.
Source: BP Review of World Energy – SSR Models
©2018, SSR LLC, 225 High Ridge Road, Stamford, CT 06905. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein. The views and other information provided are subject to change without notice. This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results. Sources: Capital IQ, Bloomberg, Government Publications.