Issues Magazine

Converting Biomass to Ethanol Fuel

By Tony Vancov

Tony Vancov reports on an alliance formed in NSW to investigate the establishment of a biofuels industry using novel feedstocks.

Global warming, a forecast decline in world reserves of crude oil, growing demand for petroleum products, inability to protect supply lines from international political intrigues and record crude oil prices (US$145 per barrel in June 2008) all ensure that the current frenzied efforts in biofuel R&D continues and remains in the public eye. Among these fuels, ethanol from renewable feedstocks is regarded as an ideal supplement and credible replacement fuel.

However, ethanol currently produced from sugar (Brazil) and from corn starch (USA) is unsustainable because it clashes with food production for human use and, more importantly, does not significantly diminish greenhouse gas emissions, as reported by Alexander Farrell of the University of California, Berkeley (Science 2006). These shortcomings could be addressed by producing ethanol from lignocellulose feedstocks, such as agricultural and forest waste residues as well as dedicated energy crops. In the same paper, Farrell and colleagues projected that greenhouse gas savings approaching 90% may be realised from cellulosic ethanol compared with petrol usage.

The US Department of Energy projects that by 2020 twice as much ethanol fuel will be produced from lignocellulosic raw materials than from sugar and starch-based crops, for which there will be more direct food-based competition. Internationally, significant research is underway. For example, the US has recently committed over US$1 billion to lignocellulosic research and conversion technology. Pilot-scale studies are being carried out in a number of countries in response to the need for technological “breakthroughs”, for example in pre-treatment, enzymes and recombinant microorganisms, and more accurate cost estimates for ethanol production from lignocellulosics.

The economic and environmental drivers of the global expansion of biofuels are relevant also to Australia, together with the additional issue of our rapidly increasing oil imports, which rose from approximately A$5 billion in 2004–05 to A$12 billion in 2005–06. Thus, Australia needs to invest selectively to capture the outcomes of international research efforts. Given our particular climate and potential to utilise wood-based products (such as waste wood, wood and crop residues and energy crops), there is a significant need to understand and apply this technology in an Australian context.

Economic, energy and environmental characteristics of different inputs, as well as the means for their production, vary widely. Optimising biofuel inputs involves difficult questions covering many aspects of primary production and transport, and will require region-specific design. Integrated bioprocessing and engineering strategies for conversion of agricultural/forestry biomass to ethanol and other valued-added products are vital if lignocellulosic biomass to ethanol is to be economically practicable.

Developed through the Primary Industries Innovation Centre (an alliance between the University of New England and NSW Department of Primary Industries), a pilot-scale project investigating technical issues associated with setting up a biofuels industry using novel feedstocks was initiated and funded by a NSW Climate Action Grant. It draws on the work of the federal government’s Biofuels Taskforce Report 2005, the developing NSW Biofuels Strategy and the NSW E10 Taskforce. It will help deliver technologies and assist the NSW Government’s foreshadowed E10 mandate (in which all transportation fuels are to contain at least 10% biofuel) and deliver other benefits.

Specifically, the major technical objective of the project is to first identify potential ligno-cellulosic feedstocks and provide an assessment of the availability of biomass, the sustainability of supply and economic viability. Second is the evaluation of existing and new bioprocessing strategies (including appropriate microbial agents) for the fermentation of lignocellulosic feedstocks to biofuels (ethanol) and other potentially valuable products.

Characteristics of Lignocellulose

Lignocellulose forms the structural framework of plant cell walls and comprises three polymers:

  • Cellulose: a linear polymer composed of 100 to 10,000 linked glucose units. In their native state, cellulose molecules form fibres largely composed of compact crystalline domains separated by amorphous regions. Hydrogen bonding between cellulose layers accounts for crystalline cellulose’s resistance to degradation.
  • Hemicellulose: a polymer with more than one type of subunit, predominantly pentose (C5) sugars (xylose, arabinose) and a smaller amount of hexose (C6) sugars (glucose, mannose).
  • Lignin: an aromatic (six-carbon ring) polymer with high molecular weight and calorific (heat-release) value, and the second most profuse renewable carbon source on Earth.

Inside plant cell walls, cellulose fibres are embedded in a matrix composed of lignin and hemicellulose.

Potential Lignocellulosic Feedstocks

Adrian Bugg and colleagues have estimated Australia’s potential agricultural biomass waste to be 65 million dry tonnes per year (Bioenergy Atlas of Australia, 2002). Residues such as bagasse (sugarcane pulp) and cereal stover (leaves and stalks), including wheat, sorghum barley, rice and rye, are among the feedstocks that can be used to generate bioethanol.

Researchers at CSIRO Sustainable Ecosystems estimate that 25% of crop residues (~15 million dry tonnes) could be made available for ethanol conversion after accounting for ground cover (soil management practices) and livestock feed. Another advantage of using crop residues is that farmers are familiar with their management and already have the capacity to harvest, store and transport them. Given these qualities, logic dictates that agricultural waste residues are included in our lignocellulosics to ethanol assessment trials.

Trees potentially offer better feedstock options for biofuel production than agricultural crops because of their higher calorific content. Moreover, trees potentially possess a ligno-cellulosic energy conversion factor of 16 (compared with 1 and 8 for corn and sugarcane, respectively), and can be grown on marginal land, thereby minimising encroachment on food crop terrain. In Australia, potential forest biomass waste was estimated by Paul Fung and colleagues (Biomass and Bioenergy 2002) to be seven million dry tonnes per year, and is predominantly in the form of sawdust, thinnings (foliage), wood chips and shaving residues. Plantations already provide 25% of the world’s wood fibre supply, and in Australia about two-thirds of the 26 million cubic metres of logs currently harvested each year in Australia are grown in plantations. However, the development of high-output plantations for new and traditional forest products must be balanced with carbon mitigation objectives while ensuring and maintaining important ecosystem performances such as clean water and biodiversity.

In due course, dedicated biofuel crops such as the Poaceae (including switchgrass and perennial reeds and grasses) and short-rotation coppice species (eucalypt, poplar and Salix) will become popular. Undoubtedly, they will be unique to a region’s climate and soil, produce more biomass per unit area and possess increased resistance to drought, cold and/or salt stresses. Genetic engineering and conventional breeding programs are also being employed to tailor plant biomass with lower lignin and higher polysaccharide content, thereby facilitating processing and boosting sugar and ethanol yields.

Bioprocessing Strategies

Depending on the raw material feedstock, there are four to five steps in the ethanol bioconversion process. The first step involves size reduction and pretreatment, which makes hydrolysis of the lignocellulosic material possible. Constituent sugars are then released by enzymatic or acid hydrolysis and subsequently separated from the residual lignin. The lignin portion is either used to fuel the process, due to its high calorific content, and/or is chemically altered to generate value-added products. The sugar solution is then fermented by yeast or bacteria and the resulting ethanol stream concentrated via distillation and molecular filtering.

Pretreatment

Efficient utilisation of lignocellulosic biomass requires pretreatment to liberate cellulose from its lignin seal and disrupt its crystalline structure before effective enzymatic hydrolysis can take place. This is by far the most costly step of the process, strongly influencing the success and feasibility of both prior and subsequent operations. For example, pretreatment impacts on the success and viability of process variables such as enzyme loadings, power use/generation, sugar and subsequent ethanol concentration/yield and waste treatment demands. In addition, pretreatment processes liberate natural biomass inhibitors and generate toxic degradation products, which inhibit subsequent hydrolysis and fermentation processes. Steps to remove these inhibitors create extra waste and raise the cost of processing.

Most pretreatment options are based on either physical or chemical approaches, and some incorporate both to increase efficiency. Contemporary pretreatment techniques include thermo-mechanical techniques such as grinding and milling to aid handling processing, acid treatment (dilute or concentrated sulfuric acid), alkali treatment (sodium hydroxide, ammonia, alkaline peroxide) and autohydrolysis (steam pressure, steam explosion, liquid hot water).

The project team at Wollongbar Agricultural Institute have exploited and optimised bench-scale pretreatment platforms based on pH and temperature. Initially, pretreatment trials have focused on deploying dilute acid and/or alkali treatment at moderate temperatures and holding times.

The dilute-acid treatment predominantly targets and solubilises the hemicellulose fraction (lignin to a lesser degree) and disrupts the crystalline structure of cellulose fibrils. The resulting liquefied portion, termed hydrolysate, is typically composed of pentose and hexose sugars and may account for up to 60% of the total extractable sugar yield from lignocellulosic biomass. To date, most of the data reveals that crop straw residues yield hydrolysates with higher sugar content per weight than, say, woody type feedstocks.

In contrast, alkali-based treatments aim at removing the lignin fraction from lignocellulose whilst decrystallising the cellulose microfibrils. Moreover, alkali pretreatment does not dissolve the hemicellulose component, nor does it release any sugars of note into the hydrolysate. Coincidently, alkali hydrolysate resembles crude oil in appearance and texture but not in odour.

Enzymatic Saccharification

The remaining solid pulp is then subjected to enzymatic (cellulase) hydrolysis, or saccharification, to release free sugars. Like the pretreatment work, our R&D efforts are focused on examining parameters of cellulase saccharification for individual pretreated feedstock pulps. Although the maximum cellulase activity of most commercial cellulase preparations is around 50°C at a pH of 4.0–5.0, we have found that the optimal conditions vary according to type of biomass feedstock, pretreatment process and source and amount of enzyme mixture used. To date, our data shows that alkali-pretreated pulp samples are more susceptible to saccharification (higher sugar yield) at significantly lower enzyme loads. This indicates that alkali conditions are ideal for delignification and minimising the release and generation of toxic inhibitors. To date, the trend for sugar conversion efficiency of biomass feedstock under these mild pretreatment conditions is sorghum stover > wheat straw > bagasse > hardwood shavings > pine shavings.

Fermentation

The resulting sugar streams derived from biomass are composed of hexoses (glucose) and pentoses (xylose), and they contain inhibitory compound. Most wild-type Saccharomyces cerevisiae yeast strains can and do ferment hexose to ethanol on a commercial basis, but they cannot metabolise xylose nor tolerate biomass inhibitory compounds. Consequently, research efforts have focused on developing efficient xylose-fermenting microorganisms.

The first approach has concentrated on introducing pentose metabolic pathways into natural ethanologens such as S. cerevisiae and the bacterium Zymomonas mobilis. The second approach is to improve ethanol yields by genetically engineering ethanologenic genes from Z. mobilis into pentose-fermenting micro-organisms such as Escherichia coli and Klebsiella oxytoca. Although both tactics have demonstrated a measure of success, productivity yields from mixed sugars derived from biomass have failed to reach commercial industrial targets.

The inability of ethanologenic strains to tolerate toxic by-products of lignocellulosic biomass conversion limits their commercial potential. Our research has shown that exposure to gradually increasing levels of each inhibitor following mutagenesis can lead to the isolation of Z. mobilis strains capable of tolerating relatively high levels of these compounds.

Biochemical and genetic studies of selected mutants are currently underway in our laboratories to determine mechanisms of inhibitor detoxification and tolerance. Information gained will be applied to engineer strains and optimise the fermentation process. Ethanologens with improved stress tolerance will lower the cost of producing ethanol fuel from hydrolysates of native Australian lignocellulosic biomass.

Challenges

The challenges of producing fuel ethanol from lignocellulosic biomass hinge on making technological breakthroughs in key process fields and slashing infrastructure costs. To make processes more cost- and energy-efficient, advancements in pretreatment, enzymatic hydrolysis and fermentation are required. These improvements include tailoring pretreatment according to the feed­stock’s chemical and structural composition (thereby minimising chemical and energy inputs) and creating enzymes with greater specific activity and inhibitor tolerance, which will undoubtedly serve to increase reaction rates and sugar conversions with much less enzyme. Like­wise, cofermentation of glucose and xylose without prior detoxification will lead to higher ethanol concentrations at reduced costs. Finally, inexpensive supplies of readily available biomass to feed the process may require the development of purpose-built infrastructure for harvesting, transporting, storing and refining biomass.