The propylene oxide problem

06 July 2017 15:06 Source:ICIS Chemical Business

Propylene oxide (PO) is in strong demand, especially for polyurethanes applications, but tight control on the technology is limiting the building of new plants.

This could change in the near future as the Chinese launch new production processes. PO is one of the fastest growing uses of propylene, with production up 3.5%/year from 2010-2016. ICIS forecasts this level of growth to continue for the future, driven by demand for polyols, which is rising nearly 4% a year over the next 10 years.

Polyols, when combined with isocyanates like methyl di-p phenylene isocyanate (MDI) or toluene di-isocyanate (TDI), make polyurethane foams. These can be in the form of flexible foams for furniture, mattresses, car seats etc, or rigid foams, which are mainly used for insulation. Non-foam polyurethane is also growing in importance for applications like coatings, binders, as well as replacing thermoplastics in some applications like car bumpers, and electrical items.

Polyurethanes are extremely versatile, and can be tailored to the application. The product is formed on site and this means low cost tooling and short lead teams.


Growth in emerging markets is driven by flexible foam for use in mattresses as modern furniture takes share from more traditional materials. Polyols use increased by more than 8%/year from 2010-2016 in areas like Africa and south and southeast Asia, according to ICIS.

However, in developed markets, growth is also good, driven by better insulation in housing and white goods as well as the development of new applications.

Polyurethane insulation has the highest thermal resistance for a given thickness and lowest thermal conductivity. Polyols are relatively easy to make. Production is usually local and fragmented. Building a new polyether polyols unit is relatively easy and low cost. However, availability of raw material PO is more problematic.

ICIS estimates the global operating rate for PO at 87% in 2017, compared to only 70% for polyether polyols. Given the forecast growth rates, we need to build more PO units to satisfy demand. But planned plants are few and far between with technology limitations strong.

There are three generations of PO plant technology:

■ First generation: Propylene chlorohydrin route. These plants convert propylene to propylene chlorohydrin which is then de-chlorinated. This process is problematic. It is highly polluting in terms of waste calcium chloride. It also needs significant amounts of chlorine.

■ Second generation: Newer plants oxidise propylene with an organic peroxide. These include tert-butyl hydroperoxide, and ethylbenzene hydroperoxide. By-products include t-butyl alcohol (TBA) for making methyl tert-butyl ethyl (MTBE), or ethylbenzene for styrene production. This makes the investment decision more complex.

■ Third generation: In 2006 Sumitomo developed a method using oxidation by cumene hydroperoxide. The by-product cumyl alcohol can be recycled back to cumene. This was followed in 2008 by technology oxidising propylene with hydrogen peroxide (HPPO), with water the only by-product. This was simultaneously developed by Degussa-Uhde (now Evonik) and BASF/Dow Chemical.


The environmental and economic problems with the first generation chlorohydrin technology are not insignificant. Most of these plants were built in the 1960s and 1970s and the only new chlorohydrin plants to have been built since 1990 are in China. Even the construction of these has slowed in recent years. Since 2010, at least 175,000 tonne/year of chlorohydrin capacity has closed in China.

So what about the second option – plants that produce either styrene or MTBE as a by-product? The propylene oxide/styrene monomer (POSM) technology is widely available with Japanese, Spanish, Russians and Chinese producers developing version of the technology originally developed by Lyondell (ARCO), and Shell. Fourteen POSM plants now exist globally. The problem with POSM plants is that styrene demand is not growing as fast as PO: consumption of styrene increased by only 1.7%/year from 2010-2016, whereas demand was up by 3.7%/year according to ICIS data.

In areas like India or China, where styrene demand is still growing and the country is short of product, building a POSM unit makes sense. In the US, which is long on styrene, we see no potential for new POSM units. In January this year, Shell pulled out of its Saudi joint venture SADAF. This move may lessen the chance of a Middle Eastern POSM unit – at least in the near term.

So what about a PO/TBA plant? The technology for this is held fairly tightly by LyondellBasell, which bought technology holders ARCO chemical and Huntsman. There are only five plants globally: two in the US, two in Europe and one in China. A sixth in China will open imminently: a 240,000 tonne/year unit in Nanjing, jointly owned by Huntsman and Sinopec. Another new plant is under consideration in the US, according to LyondellBasell. These PO/TBA plants actually produce around twice as much TBA as PO – so the market for MTBE has to be as much, or more of, a concern than PO.

Sumitomo’s technology has not taken off to the extent the company hoped. There are only two plants globally – the original one in Japan and one in Saudi Arabia, launched in 2009 by PetroRabigh, a Saudi Aramco/ Sumitomo joint venture. Another two are planned to open in Asia, with licences from Sumitomo. One will open in 2018 in South Korea, operated by S-Oil. The second is in Thailand, operated by PTT Global Chemical and due to start in 2019.


Industry sources say this process is complex and costly, and has been overshadowed by the newer, more efficient HPPO process. HPPO is a relatively simple process with few by-products. As such, it is seen as the most modern, cost efficient process. Degussa-Uhde claim the first plant – built in 2008 by SK Chemicals in Korea. This was very shortly followed by a BASF/Dow plant in Antwerp also in 2008.

More plants came from BASF/Dow – a plant in Thailand in 2011, and a Dow joint venture Sadara Chemicals opening in Saudi Arabia in 2017.

Degussa-Uhde (now owned by Evonik) sold the technology under licence to Jishen Chemical industries in Jilin, China (from 2014). However, a couple of Chinese players are working on their own HPPO technologies and this could be a game changer for the PO market.China is promoting the HPPO route of PO production.

If Chinese players are successful, this could lead to a new wave of PO plants and a loosening of supply. We also see some new TBA and POSM plants – particularly in areas like China and India where more MTBE and styrene are needed. ■

Rhian O’Connor is a Senior Analyst in ICIS Consulting covering the styrene and propylene oxide chains, including the ICIS supply/demand database plus single client work.

By Rhian O'Connor