Biorefineries and platform chemicals: The biorefinery concept and its types

2.1 Fractionate first, then convert: the corn wet mill already does it

The US Department of Energy defines a biorefinery as a plant where biomass is converted and extracted into a spectrum of valuable products, modelled on the petroleum refinery. Task 42 adds that it can be a concept, a plant or a cluster, and classifies any biorefinery by four features: platform (the core intermediate such as C6 sugars, syngas, lignin or oil), products, feedstock and processes.

Kamm and colleagues describe generations. A generation I plant, such as a dry-mill ethanol plant, takes grain and makes fixed amounts of ethanol, feed and carbon dioxide with almost no flexibility. A generation II plant, such as a corn wet mill, takes the same grain but can swing between products as demand changes. Generation III plants, using mixed feedstocks and many technologies, had not been built by 2006.

The corn wet mill is the working model. Corn is steeped in warm, slightly acidic water, then ground in stages to separate germ (oil), fibre, gluten (protein) and starch, the starch washed to 99.5% purity. Johnson gives the yield from a tonne of corn as 684 kg starch, 237 kg corn gluten feed, 45 kg gluten meal and 34 kg oil. Dale and Kim report that over 90% of the corn leaves as products worth more per kilogram than corn itself.

The starch stream is where flexibility lives. Converted to a 95 DE glucose syrup, it can be sold as dextrose, fermented to ethanol or other products, or isomerised to high-fructose syrup. Large mills swing output by as much as 50% between fructose syrup and ethanol, accepting some idle capacity as the price of flexibility.

• Kamm (2006), Vol 1 Ch 1 Biorefinery systems - an overview · pdf pp. 51-52, 60
• Task 42 (2012), Sections 3 and Figure 2 · pdf pp. 6-7
• Kamm (2006), Vol 1 Ch 15 The corn wet and dry milling industry · pdf pp. 373-379
• Kamm (2006), Vol 1 Ch 2 Biomass refining global impact · pdf p. 80
• Kamm (2006), Vol 1 Ch 3 Development of biorefineries · pdf pp. 102-103

2.2 Whole-crop and green biorefineries use every part of the plant

A whole-crop biorefinery takes cereals such as rye, wheat, triticale or maize and first separates grain from straw mechanically. The grain goes to starch, meal or glucose-based fermentation; the straw is lignocellulose and can go to a lignocellulosic line or to gasification for syngas. Kamm and colleagues list the starch options: plasticising it into polymers, chemically modifying it (for example carboxymethyl starch), cracking it by hydrogenation to sorbitol, glycols and glycerin, or fermenting it. They note the drawbacks: high raw material cost and costly processing.

A green biorefinery processes 'nature-wet' biomass: grass, lucerne (alfalfa), clover or immature cereals. The first step is gentle wet fractionation, typically pressing, which splits the crop into a fibre-rich press cake and a nutrient-rich green juice. Because the lignin-cellulose composite in green crops is less intractable than in straw, and the plant's own water can carry later biological steps, the chemistry is easier. The catch is perishability: crops must be processed fast or preserved as silage or by drying, and each method changes the contents.

Scale ranges widely. Kamm reports plants from a few tonnes of green crop per hour on farms to over 100 t per hour commercially. Kromus and colleagues suggest decentralised plants of about 35,000 t of raw material a year and central plants of 300,000 to 600,000 t. Raw material is not free: their cost table puts late-cut grass at 60 euro per tonne of dry matter, against 36 euro for straw, mainly because grass needs more field work per hectare.

• Kamm (2006), Vol 1 Ch 1 Biorefinery systems - an overview · pdf pp. 41, 58-62
• Kamm (2006), Vol 1 Ch 12 The green biorefinery concept (Kromus et al.) · pdf pp. 311-312
• Task 42 (2012), Section 4.6 Organic solutions platform · pdf p. 10

2.3 Lignocellulosic and two-platform biorefineries: cheap feed, hard chemistry

Kamm and colleagues expected the lignocellulosic feedstock (LCF) biorefinery to be the most successful of all, for two reasons: the raw material is abundant and cheap, and its products fit both existing petrochemical markets and new bio-based ones. They quote corn stover or straw at 30 dollars per tonne against 110 dollars for corn grain.

The LCF biorefinery splits its feed into three precursor streams. Hemicellulose gives mostly pentose sugars, the classic source of furfural and xylitol. Cellulose gives glucose for fermentation or for chemical routes to hydroxymethylfurfural and levulinic acid. Lignin, the phenolic polymer, is mostly burned. Kamm calls this unsatisfactory, because lignin is full of single-ring aromatics that would be valuable if they could be isolated economically.

The two-platform concept, from NREL and the DOE, organises the same idea by intermediate. The sugar platform uses biochemical steps to release and ferment sugars. The syngas platform gasifies biomass, or the residues of the sugar platform, to carbon monoxide and hydrogen and builds products back up from those small molecules. Gasification has the merit of accepting almost any feed, but it uses only the carbon and hydrogen and ignores the protein, oil and other structures nature has already made.

The DOE screening was blunt about syngas in 2004: only hydrogen and methanol were near-term prospects, because simple alcohols, mixed alcohols and Fischer-Tropsch liquids from biomass were not economically viable. Kent adds that biomass gasification experience in the US was limited to heat and power, demonstrated at 200 t/day (5 MW) in Vermont.

• Kamm (2006), Vol 1 Ch 1 Biorefinery systems - an overview · pdf pp. 38, 56-58, 63-64
• Werpy & Petersen (2004), Section 6 Syngas results · pdf p. 24
• Kent (2012), Ch 33 Biomass conversion · pdf pp. 1251-1252

2.4 The oleochemical biorefinery: the oldest and closest to petrochemistry

Oils and fats are triglycerides: one glycerol backbone carrying three fatty acid chains, which Clements likens to a comb with three long tines. Because the chains are hydrocarbon-like, oleochemistry slots into existing chemical plants more easily than sugar chemistry, and Clements calls it the largest contribution biomass makes to the chemical industry today.

The resource is large but mostly eaten. Task 42 puts world vegetable oil production in 2009-10 at close to 140 million tonnes, about 20% of it for non-food uses including biofuel. Wittcoff's figures for 2000 are starker: of 110.8 million tonnes of oils and fats, 88% was eaten, 6% went to animal feed and only 6% to the oleochemical industry.

There are three main ways to open a triglyceride. Saponification with caustic soda gives soap and glycerol. 'Splitting', continuous non-catalytic hydrolysis at high temperature and pressure, gives free fatty acids and glycerol and is the main route in the US and Europe. Alcoholysis, or transesterification, swaps glycerol for methanol to give fatty acid methyl esters (FAME), which are biodiesel and also the preferred feed for fatty alcohols. Wittcoff gives the hydrogenolysis of methyl esters to fatty alcohols as 250 to 300 C and 200 to 300 bar.

Task 42 counted 7.7 million tonnes of fatty acids and 2.0 million tonnes of fatty alcohols in 2009, mostly for surfactants in soaps, detergents and personal care. Every route also releases glycerol, roughly a tenth of the product mass, which is why glycerol became one of the DOE's twelve platform chemicals.

• Kamm (2006), Vol 1 Ch 5 The lignocellulosic biorefinery · pdf pp. 150-151
• Task 42 (2012), Section 4.4 Plant-based oil platform · pdf p. 9
• Wittcoff (2004), Ch 13 Fats and oils · pdf pp. 438, 442, 451
• Kent (2012), Ch 34 Animal and vegetable fats, oils and waxes · pdf pp. 1374, 1393-1395

See it in the table

Check your understanding

What makes a corn wet mill a 'generation II' biorefinery rather than generation I?

Why do Kamm and colleagues call the usual use of lignin 'unsatisfactory'?

In the oleochemical industry, which product appears as a by-product of splitting, saponification and transesterification alike?