4.3 Advanced biofuels: ethanol and gasoline replacements

Advanced biofuels from lignocellulosic feedstocks offer the opportunity to address some of the drawbacks of bioethanol products derived from food crops. Advanced biofuel feedstocks do not have to be grown on pasture or arable land. They do not, therefore, compete with food supplies. As a result, they also have the potential for much higher levels of production and very low GHG emissions. Although advanced biofuels are only just at the early stage of commercialisation, and costs are high, the cost reduction potential is good and higher than for conventional biofuels. However, the technology challenges facing advanced biofuels are significant and commercial production at large scale today incurs significant technical and commercial risk.

Ongoing R&D investment, funded by both public and private sources, is still essential to perfect different pathways and identify new promising production routes. However, the key immediate challenge is to gain experience with commercial-scale projects in each of the most promising pathways now that commercialisation is beginning. This will require major investment in new facilities that, if they are going to succeed commercially, will require appropriate risk reduction strategies from the commercial operators, but also from policy makers and regulators in order to help accelerate commercial deployment to meet the medium- to long-term sustain-ability goals for transport.

The two main production pathways for advanced biofu-els from lignocellulosic feedstocks are:

  • Biochemical routes: where enzymes and other micro-organisms are used to convert cellulose and hemicellulose components into sugars,24 these sugars can then be fermented into ethanol in a manner similar to conventional ethanol.
  • Thermochemical routes: these are processes which use pyrolysis/gasification technologies to convert the lignocellulosic feedstock into a synthesis gas and/or liquid biocrude from which a wide range of biofuels can be reformed. Gasoline can be produced from thermochemical routes, but the main products are likely to be biodiesel, bionaptha and jet kerosene so this route is discussed under biodiesel.

The biochemical production route for ethanol from lig-nocellulosic feedstocks requires four distinct steps:

  • Pre-treatment: this is designed to prepare the feedstock for further processing, and it needs to expose the cellulose and hemicellulose to subsequent enzymatic hydrolysis. Given the strong bonds in lignocellulose, this is challenging and expensive.
  • Hydrolysis to sugars: this could be done using enzymes or acids. However, enzymes appear to be the cheapest option for the fast and efficient conversion of cellulose to glucose and should provide better yields. Nevertheless, the enzymes required for lignocellulosic feedstocks are more complex and expensive than those needed for conventional ethanol. After hydrolysis, separation of solid lignin allows the liquid glucose and xylose to be fermented.
  • Fermentation: this is similar to conventional bio-fuels, except that yeasts and bacteria are required to convert the glucose and xylose into ethanol. This is more difficult, and more care needs to be taken to ensure nothing obstructs the fermentation.
  • Product recovery: this is similar to the process for conventional ethanol where hydrous ethanol is distilled into anhydrous ethanol.

Capital costs for an advanced biofuel ethanol plant

The installed costs of a commercial-scale advanced bio-fuel plant producing ethanol are only just emerging, and most data available up until recently has been based on engineering estimates.

The purchase of the equipment required for an advanced bioethanol plant using corn stover as a feedstock accounts for 55% of the total installed cost (Figure 4.11). Pre-treatment and the equipment for conditioning the pre-treated slurry prior to its passage to the sachirification and hydrolisation stage account for 8% of total installed costs. Sacchrification, fermentation and the equipment required for onsite enzyme production account for 11%. The boilers and turbines are the largest single equipment cost at 16% of the total, but allow the plant to meet its own process heat and electricity needs from the co-products and export electricity to the grid. This significantly improves the economics of production, although not to the same extent as for conventional sugar cane ethanol, where more surplus electricity can be produced. Wastewater treatment is also costly and very significant for advanced plants, which can require fve or more litres of water per litre of ethanol produced.

The cost of advanced bioethanol plants are estimated to be in the range of USD 1.82 to USD 2.5/litre/year of production capacity (APEC, 2010; Humbird, 2011; and Stephen, 2011) once deployed at scale, processes are de-bottlenecked and modular designs are rolled out. However, the first-of-a kind commercial plants currently being deployed, sometimes at relatively small-scale, appear to have much higher investment costs. Data for recently operational, under construction or advanced biofuels plants planned to be online by 2015 have capital costs in the range USD 1.5 to USD 4.6/litre/year of capacity (Figure 4.12). This is between twice and over six times more costly than conventional ethanol plants and reflects the more complicated pre-treatment and processing needs required to produce bioethanol from lignocellulosic feedstocks, but also that these plants are typically first-of-a-kind, unlike the mature technologies used in conventional ethanol plants.

Figure 4.11: Capital cost breakdown for biochemical production of bioethanol from corn stover

Source: Humbird, 2011

Figure 4.12: Capital costs for current or near future commercial-scale advanced ethanol plants

Sources: F.O. Licht, 2013 and Brown and Brown, 2013.

The indirect gasification of biomass to produce a syngas that can be synthesised into ethanol and other mixed alcohols (propanol, butanol, pentanol and hexanol) is estimated to be around 10% more capital intensive than for the enzymatic hydrolysis and fermentation route, once fully commercialised. However, it could yield lower overall costs per litre of ethanol produced (Dutta, 2011). This is in an interesting prospect for the future, but is less advanced that enzymatic hydrolysis to ethanol and the costs are therefore more speculative.

Feedstock costs of advanced biofuel plants producing ethanol

Lignocellulosic feedstocks can be agricultural or forestry residues (e.g. corn stover, bagasse, black liquor, hog fuel, forestry arisings and other wood processing wastes, etc) or dedicated crops (e.g. hardwood, softwood, switchgrass, poplar stems). Residues and wastes may be available at low or no cost at the site of production, but perhaps in quantities that would limit the scale of production to uneconomically low levels. Purchasing additional feedstock means higher costs but larger production scales. The planting and harvesting of dedicated feedstock crops can provide additional supply opportunities, but will typically be more expensive.

A key issue for advanced biofuel production is that biomass is very heterogeneous and the proportion of cellulose, hemicellulose, lignin, ash and even the microstructure of the plant at a cellular level varies by feedstock type (Stephen, 2011). Technologies and plant designs able to process a number of different feedstocks in a flexible way are therefore desirable. A multi-feedstock plant could buy the cheapest feedstocks on the market at a specific point of time throughout the year to complement any contracted volumes. However, multi-feedstock plants are more difficult to design and more costly to operate. Progress in overcoming these challenges would help improve the economics of advanced bioethanol plants.

As with conventional plants, the two drivers of the total feedstock cost per unit of ethanol are the ethanol yield per tonne and the price per tonne. In some cases, there are no feedstock costs because residues are available at the food/agricultural or wood processing site. In others, they amount to as much as USD 80/tonne because residues need to be collected from the production site, stored and transported to the ethanol plant. For instance, the collection, chipping and transport of logging residues in the United States was estimated to cost USD 30-35/wet tonne depending on assumptions made (APEC, 2010). The marginal cost of corn stover, taking into account collection, storage and transport is estimated to cost between USD 65 and USD 80/dry tonne (APEC, 2010 and Humbird, 2011).

Many dedicated lignocellulosic energy crops are also being explored for their economics and suitability for advanced ethanol production. In the United States, switchgrass (a native North American perennial grass) can be grown on poor soils and requires minimal fertiliser. Two types exist, with one more suited to semi-arid conditions and another to heavier soils and wetter climates. Yields for commercial operations might be expected to reach around 13.5 to 18 tonnes/hectare. The upper yield in ideal conditions may be as high as 22 tonnes/hectare (Garland, 2008). However, typical yields in less desirable growing conditions are likely to be 4.5 to 9 tonnes per hectare. With these yields, total costs could be USD 65 to USD 100/dry tonne.

Maximising ethanol yields from different feedstocks is crucial to the economics of lignocellulosic ethanol production. Table 4.3 provides details of the composition and potential technical yields from different feedstocks. This is shown for all sugars and for the C6 sugars alone. Actual yields for commercial plants are yet to be determined with certainty, but values of between 110 to 330 litres/dry tonne for agricultural residues and 125 to 300 litres/dry tonne for forest residues might be expected (IEA, 2009 and NREL, 2011). For instance, the yield from corn stover is estimated to be around 330 litres/dry tonne (NREL, 2011). Dedicated energy crops optimised for lignocelluosic ethanol production could achieve even higher yields. For instance, "energycane"25 could perhaps achieve yields of 375 litres/dry tonne (Alvarez, 2011).

The impact of feedstock costs and ethanol yield on feedstock costs per litre of gasoline equivalent is presented in Figure 4.13. For instance, an advanced ligno-cellulosic ethanol plant using corn stover at a cost of USD 70/dry tonne and yielding 330 litres/dry tonne would have feedstock costs of USD 0.32/lge. A yield of 330 litres/dry tonne is around 77% of the maximum theoretical yield from corn stover (Humbird, 2011). One important issue to consider is that higher yields are not necessarily more economic from an overall production cost basis. This is because there is a complex trade-off between process design, capital costs, operating costs and the impact on the yield from the feedstock (Table 4.3).

Table 4.2: Theoretical maximum ethanol yields from different lignocellulosic feedstocks

Source: Stephen, 2011.

Figure 4.13: Biofuel feedstock costs as a function of ethanol yield and biomass feedstock price

Table 4.3: Capital costs, ethanol yields and production, electricity generation and final ethanol cost for biochemical conversion of corn stover by pretreatment process

Source: Kabir Kazi, 2010.

Other operating costs for advanced biofuel plants producing ethanol

The other operating costs for advanced biofuels are currently estimated to be higher than for conventional biofuels, given the more complex process required for production and also the additional costs of the enzymes.

Chemicals, enzymes and other materials are projected to account for around two-thirds of the other operating costs in the case of corn stover ethanol (Figure 4.14). The costs of chemicals, enzymes and other materials alone are more than the total other operating costs for corn ethanol and more than twice those for sugar cane in Brazil. The significant solid lignin extracted from the feedstock makes the provision of process heat and electricity possible, with significant left over material for the production of electricity for export. Depending on the local electricity market, these exports can have signifi-cant value and help to reduce overall production costs.

For indirect gasification and mixed alcohol synthesis, operating costs are lower and dominated by the fixed operations and maintenance costs. This process uses around twice the electricity of hydrolysis and fermentation, leaving virtually no electricity surplus. However, indirect gasification and then synthesis of the syngas yields other mixed alcohols for sale in the ratio of around 1.1 litres of mixed alcohols for every one litre of ethanol produced. The estimated value of these co-products is highly dependent on the market for chemical feedstocks and oil.

Total production costs for advanced biofuels plants producing ethanol

The estimated total costs for advanced lignocellulosic biofuels are evolving constantly. This is because new R&D results come through and commercial experience, albeit limited today, brings new refinements to processes and a better understanding of the challenges and opportunities. Lignocellulosic bioethanol is in its infancy, as far as commercial deployment is concerned, with around 120 Ml of production capacity in the United States and around 70 Ml in Europe in 2012 (F.O. Licht, 2013). Uncertainty is therefore likely to remain about current costs and future cost reduction potential for some time.

Figure 4.14: Other operating cost breakdown for biochemical production of bioethanol from corn stover and from forestry biomass with indirect gasification and ethanol and mixed alcohols from synthesis

Sources: Humbird, 2011 and Dutta, 2011.

The often novel combination of processes means there is a high level of technological risk surrounding commercial scale lignocellulosic bioethanol production. This is true even if many of the individual processes are well known. In addition, given that until recently the estimated costs were also significantly higher than for conventional ethanol, there has been little investment in commercial-scale advanced bioethanol plants.

The cost of current advanced bioethanol from enzymatic hydrolysis is estimated to be between USD 1.04 and to USD 1.45/lge (BNEF, 2013; Humbird, 2011; IEA, 2011; Poet, 2011; and IRENA analysis) and is the pathway which is the closest to widespread commercial deployment. However, the investment cost data for operating, under-construction and plants that will be online by 2015 suggest that if the processes prove to be reliable, efficient and operate continuously at design capacity, then the lower end of that range could be around USD 0.75/lge (Figure 4.15). This is an estimate, as the actual data on the other operating costs and feedstock costs are not yet clear. The data in Figure 4.15 is therefore also based on the assumption that operating costs are a quarter higher than the long-term optimised level (Humbird, 2011) and that feedstock costs are USD 65/dry tonne for agricultural and forestry residues, and energy crops, while municipal solid wastes and sugar cane bagasse costs are half to a quarter of this. However, perhaps the key uncertainty still surrounds the efficiency of the process pathways and the ability to operate continuously (except for scheduled maintenance and estimated unplanned outages) at design capacity.

Recent advances in R&D and process integration suggest that costs for future plant using today's enzymatic hydrolysis technology could yield bioethanol costs of USD 0.7/lge for a feedstock cost of USD 30/dry tonne and USD 1/lge for feedstock costs of USD 100/tonne in 2020, once the process pathways are proven to be reliable, efficient and able to support continuous production. For this to be achieved a number of factors have to be in place: the learning experience from the initial commercial-scale plants needs to be incorporated into future designs and the scale of the market needs to grow to allow modular designs, rather than expensive, individually engineered first-of-a-kind plants (Figure 4.16).

Figure 4.15: Estimated current lignocellulosic bioethanol production costs

Note: Assumes a 10% cost of capital, a 20 year economic life, feedstock costs of USD 65/dry tonne for agricultural and forestry residues, and energy crops, and USD 15/dry tonne for municipal solid wastes and USD 15/tonne sugar cane bagasse; and that other operating costs are a quarter higher than the long-run costs (Humbird, 2011).

Sources: Figure 4.12; Humbird, 2011; IEA, 2011; and Poet, 2012.

Figure 4.16: Estimated current lignocellulosic bioethanol production costs and future cost with today's technologies

Note: Assumes a 10% cost of capital and a 20 year economic life for today's technologies.

Sources: Based on BNEF, 2013; Dutta, 2011; Humbird, 2011; IEA, 2011; and Poet, 2012.

Indirect gasification and mixed alcohol synthesis of the resulting syngas could eventually lead to even lower costs. With total production costs of between USD 0.6 and USD 0.9/lge for feedstock costs of USD 30 and 100/dry tonne respectively. However, this technology pathway is yet to be deployed commercially at scale and much needs to be done to prove the reliability, ef-ciency and plant availability of this process in real world circumstances.

24 The cellulose undergoes enzymatic hydrolysis to produce hexoses (also called C6 sugars) such as glucose. Pentoses (also called C5 sugars), mainly xylose, are produced from the hemicellulose.

25 A cross between commercial sugar cane and a species with higher fibre and lower sucrose contents.