Mar-2009

Producing propylene glycol from biomass

Process and catalytic considerations for producing propylene glycol from glycerin-derived biomass sources are compared to current conventional sources, while economic factors and potential markets are discussed

Tony Pavone
SRI Consulting

Article Summary

Propylene glycol (PG) is a basic petrochemical used primarily in polyurethane polyols and  unsaturated polyester resins, and as an anti-freeze agent in aircraft de-icing applications, as well as an automotive coolant. Total annual global consumption is 1.5 million metric tonnes. Historically, PG has been made commercially by simply hydrating propylene oxide (PO). PO, in turn, has been made historically by the peroxidation of propylene using co-feeds of either ethylbenzene (EB) or isobutene (iC4).

When EB is used, PO is produced with styrene monomer co-product. When iC4 is used, PO is produced with tertiary butyl alcohol (or sometimes methyl tertiary butyl ether [MTBE]) co-product. An older technology produces PO from propylene via chlorohydrin chemistry. Recently, a new plant has been commissioned by SK Chemicals to avoid co-product formation by using hydrogen peroxide as the oxidising agent. Whichever historic route is taken, PG production economics are dominated by the cost of the basic feedstock propylene. Produced from either natural gas liquids or refinery naphtha, propylene prices have averaged $1000/mt over the past three years, and PG market prices have been $1500/mt during the same period. During the mid-2008 crude oil price run-up, with naphtha prices exceeding $1000/mt, market prices for both ethylene and propylene averaged approximately $1500/mt, driving PG market prices to $2000/mt.

Production of PG from glycerin
Most recently, several companies (Dow, Huntsman, ADM, Ashland/Cargill) have announced new commercial technology projects for producing PG directly from the glycerin that is a by-product of biodiesel production. Since biodiesel production has increased dramatically over the past five years, a glut of excess by-product glycerin has been produced that cannot be absorbed into conventional glycerin markets. Glycerin prices, once $1000/mt, have dropped by 50% for refined grades, while raw by-product glycerin is often used as boiler fuel.

Two organisations have patented their processes for converting glycerin to PG: Davy Process Technology in the UK and the University of Missouri (the Suppes process) using R&D funds from the US government.

Chemical structure of the relevant molecules
PG is a three-carbon molecule in which two of the carbon atoms are attached to hydroxyl (-OH) radicals. The third carbon atom is saturated with hydrogen. The carbon atoms attached to OH radicals are adjacent to each other. As a result, PG has its OH ions on one of the terminal carbon atoms and on its adjacent internal carbon atom.

A similarly structured, but more expensive molecule to make and buy, is 1,3-propane diol (PDO). Like PG, PDO is a three-carbon molecule with two of its carbon atoms connected to OH radicals. However, with PDO, the OH groups are both on the terminal carbon atoms, with the centre carbon atom saturated with hydrogen. Due to its structure, PDO can easily be polymerised with purified terephthalic acid to produce a polyester (poly tri methyl terephthalate) that performs much like nylon when fabricated as a fibre, but costs only half as much as nylon to manufacture.

The molecular structure of glycerin (mol wt: 92.1) is remarkably similar to PG and PDO. Glycerin is also a three-carbon atom molecule, but each of the carbon atoms is attached to an OH radical. As a result, glycerin is a solid at room temperature (both PG and PDO are liquids), and has a much higher boiling point than either PG or PDO.

From a chemical structure perspective, converting glycerin into PG or PDO simply requires substituting a hydrogen ion for one of glycerin’s OH radicals. This can be accomplished with conventional industrial chemistry (high-pressure, moderate-temperature hydrogenation at a low residence time over a base metal catalyst), or biologically using an enzyme catalyst (moderate temperature and ambient pressure and long residence time). If the substitution occurs on a terminal carbon atom, the product is PG. If the termination occurs on the internal carbon atom, the product is PDO.

Value proposition
The patent literature has taught for at least the past 50 years methods for converting glycerin to PG (albeit at low conversion rates). So why does the world have renewed interest in the chemistry? The basic reason is economics. As long as glycerin market prices are significantly higher than propylene monomer market prices, there is no compelling economic reason for making PG or PDO from glycerin. In this economic world, it costs more to make PG from glycerin than it costs to make PG from propylene.

Biodiesel has changed this economic world. Since 10% of the production of conventional biodiesel (methanol esterification of fatty acid) represents glycerin by-product, the conventional markets for glycerin are well satisfied by the conventional sources of glycerin (from soap and surfactant manufactur-ing). When there is an excess supply of any commodity to a market, the market price of the commodity usually drops precipitously.

If the market price of glycerin changes from initially being much more expensive than propylene to finally being much less expensive than propylene, it makes sense that industrial routes to PG from glycerin might be economically attractive.

Markets for glycerin
As shown in Table 1, the global demand for glycerin is approximately 600 kty. Historical and projected demand growth rates are modest (2–5%/year) and surely not high enough to absorb the enormous amount of new glycerin capacity expected from biodiesel by-product.

Where is all the glycerin from biodiesel going to go? One globally large producer of conventional glycerin, which produces on-purpose biodiesel, has intentionally decided to burn by-product glycerin in its steam boilers to avoid degrading the market price of glycerin.