Parratt & Associates Scoping Biorefineries: Temperate Biomass Value Chains

4.3 Commercial Australian lignocellulosic biorefineries development

Lignocellulosic biorefineries based on biochemical processing are the subject of intensive research and development activity internationally with several companies operating advanced pilot-scale facilities and a small number nearing commercial-scale refinery construction. In the temperate region of Australia only two companies currently appear to be actively engaged in commercial lignocellulosic biorefinery development. Firstly, Renewable Oil Corporation (ROC) is developing a thermochemical (pyrolysis) refinery based on commercial technology licensed from Dynamotive to operate on wood waste, with an initial capacity of 200 tonnes per day. The primary product is expected to be bio-oil, with char as a co-product. Secondly, Ethtec, a subsidiary of Wilmot Forests is currently constructing a pilot-scale biochemical biorefinery at Harwood in northern NSW to convert pine, bagasse and other forestry residues to fuel ethanol, with electricity from lignin combustion as a co-product.

Several collaborations are developing between US and European-based companies and Australia counterparts to transform biomass in Australia; however none are yet at the pilot or commercial scale. A number of barriers to entry are often cited for the lag in industry development in Australia. These will be addressed in subsequent chapters.

4.4 Commercial international lignocellulosic biorefinery development

Significant public and private investment is occurring internationally to develop lignocellulosic technologies and commercial operations. The Biofuels Digest currently lists 68 known advanced biofuels projects (Biofuels Digest, 2010). In addition the USA government has recently announced US$564M funding to support the development of integrated biorefineries (Biofuels Digest,2010). Concurrent technologies are being evaluated and researched. The current consensus appears to be that no single technology will replace the current petrochemical-based refineries. Some of the technologies supported by the Obama Administration include those that produce Ethanol as the major product (Bluefire, ICM), Butanol producers (Dupont-BP), ethylacetate /acetic acid producers (Zeachem), Syngas to bioethanol producers (Enerkem), and many others to produce feeder molecules for the chemical industry.

IEA Task 39 has recently reported on the status of 2 Generation biofuels demonstration facilities^nd. Although the report is focussed solely on biofuels production it provides both a guide and benchmark for the requirement for establishing facilities with other bioproducts as a focus. Overall 66 projects provided data on the current state of their operations. Two known operations in Australia were not identified, and presumably did not provide data (ROC and Ethtec). Of the 66 projects identified, 37 classified as using biochemical transformation, 23 as using thermochemical, and 6 a hybrid of both. Pilot facilities constituted 32 production facilities, 21 of which were classified as demonstration facilities and 13 as commercial facilities. The classification based on production is; pilot mostly less than 500 t/y, demonstration (mostly under 5000 t/y) and commercial (greater than 5,000 and up to 500,000 t/y).

The average investment, range and output for each type of facility for biochemical transformation are given in Table 4-2.

Table 4-2: Capital Investment and outputs from operational and planned facilities¹⁶⁰

Facility	Average	Range	Investment per Tonne
Pilot < 500 T/y	US$12.9 M	$3.5-46.2M	7.1 T per Million $
Demonstration < 5,000T/y	US$49.6M	$5.0-100.0M	77.9 T per Million $
Commercial < 500,000T/y	US$115.5M	$32-211.0M	1 558.2 T per Million $

The values in the table are derived from the report data. A number of companies did not provide any financials and were not included. For both the thermochemical and hybrid company’s facilities insufficient data were available to provide any meaningful comparisons. There sole exception related to investments in commercial-scale thermochemical transformation. The average investment was US$182.6M and an output of 1,040 T per million dollars invested. The benefits of scaling up are very obvious from this analysis. It should be stressed these are data based on a mix of sources and mostly provided by companies either at start up or in transition to full commercialisation. There is third party evidence to suggest that these figures are of the right order of magnitude¹⁶¹.

The top ten facilities reporting, only one of which is under construction, were broken down into 4 biochemical, 5 thermochemical and one hybrid. If all 66 facilities surveyed reach production capacity by 2016 total cumulative output will be approximately 1,700,000 t/y of biobased fuels from lignocellulose transformation.

4.5 Cost of infrastructure and operating Bio-Refineries

Building on the relatively simple analysis above a guide to the scale and costs of commercial bio-refineries is emerging.

1. Biochemical

The data in Table 4-2 suggest that a biorefinery plant producing 1Mt/y on approximately 2.5-3M t/y of green inputs would have infrastructure costs of $1.5B. Recently Gunns Pty Ltd proposed the development of large scale wood pulp mill in Tasmania. The estimated infrastructure costs for this mill are A$1.5-1.7B. This mill, effectively a biorefinery will process up to 2.5M t/y of green forest material to produce approximately 1M t/y of dry pulp. The estimated costs of the green inputs are approximately A$100/t¹⁶².

The National Renewable Energy Laboratories (NREL) in the US has suggested the following breakdown of cost components for cellulosic ethanol.¹⁶³

Figure 4-3 : NREL estimated costs for cellulosic biorefinery

Based on the NREL data and the cost analysis for Australian temperate biomass,¹⁶⁴ the breakdown of operational costs for the production of ethanol for a 1Mt/y output would be;

Table 4-3: Estimated costs for an Australian Temperate Biomass Biorefinery

Component	Cost by Area	Estimated A$
Feedstock	38%	250.0M
Pre-Treatment	18%	118.4M
Hydrolysis and fermentation	21 %	138.2M
Distillation and recovery	10%	65.8M
Waste Treatment	4%	26.8M
Other Process Costs	9%	59.2M
Total Costs	100%	658.4M

Current estimates on water use just for the pre-processing fermentation and distillation range between 4-7 litres of water per litre of ethanol produced¹⁶⁵. No costs for water access are included in the tables above. Gunns’ Tasmanian project report suggests that 23,000 litres of water would be required to produce dry pulp from 1000 tonne of forest biomass.

2. Thermochemical

There is little data or detailed information available on the production and costs of commercial scale thermochemical biorefineries. However, the US Department of Energy and the American Forest and Paper Association, reported in 2006 a two year study¹⁶⁶ on this topic. This report is probably the most comprehensive analysis available of the application of thermochemical transformation of woody biomass. The study was focussed on liquid fuels and chemicals via gasification of kraft processed black liquor solids and woody residues. The analysis was founded on the examination of the costs and benefits of 7 alternative biorefinery designs to account for a comparison in costs to the replacement of Tomlinson boilers in the aging paper and pulp mills of the USA.

The biorefineries modelled were intended to provide chemical recovery services and co-produce process steam for the mill, some electricity, and one of three liquid fuels (in addition to the usual pulp production). The liquid fuels considered were Fischer-Tropsch synthetic crude oil (which can refined into vehicle fuels at existing petroleum refineries), dimethylether (DME, a diesel fuel or LPG substitute), or an ethanol rich mixed-alcohol.

The re-designed pulp and paper mills would require a larger capital investment than simply replacing the boilers. However significant additional benefits would accrue;

• 
  higher energy efficiency;
  
• 
  lower air emissions;
  
• 
  more diverse products; and
  
• 
  attractive internal rates of return (IRR) on the investments.
  

The study showed under an assumption of a 25 year levelised world oil price of US$50/barrel the internal rate of return(IRR) lies between 14% and 18%. If the positive incremental environmental benefits are monetised the IRR rises as high as 35%. If the oil crude price is US$78/barrel the calculated IRR exceeds 45% in some cases when environmental benefits are monetised. These returns are based on retrofitting existing mills rather than on Greenfields developments. Depending on the design of the biorefinery the cost would be between US$250M and $470M.

The initial model throughput for a biorefinery under the above study produced the following;

• 
  1.43 M Tonne of biomass to produce;
  
• 
  1 Million Tonne bone-dry biomass to produce;
  
• 
  0.91 M Tonne of woodchips to produce;
  
• 
  0.42 M Tonne of pulp; and
  
• 
  0.40 M Tonne of solids concentration.
  

In the best thermodynamic model of a biorefinery 1549 barrels of Black Liquor Solids (BLS) were produced per day. This is the equivalent of 246,260 litres per day or 88.9ML per year. For the best BLS output design 756,267 litres per day or 276ML per year were produced. These figures are based on an average input per mill of 3500 t of biomass per day or approximately 2458 t of bone-dry biomass per day.

In Chapter 2 of this report the estimated the amount woody biomass currently available in Australia is 15.5M t/year. Under the USA DOE study the upper limit on production of BLS from Australian forest residues and pulp logs would be 4.5 billion litres per year. The best thermodynamically designed biorefinery would produce approximately 32.5% of this amount, as well as cellulose for other uses.

Comparisons between two alternate designs proposed by the DOE study suggest there are significant potential returns from increased scale. Comparisons made between two Fischer-Tropsch processes where the only difference was the scale of the synthesis plant (FTa and FTc in the study). FTc had three times the capacity of FTa in the synthesis plant, yet required only 41% more capital investment (US$330M versus US$465M). The increase in annual operating and maintenance costs for FTc were therefore more than offset. The FTc gave a 15% improvement in the IRR when compared with the FTa scenario.

In the conclusions to the study the authors stressed the advantages of ‘retrofitting’ this capability to the establishment of bio-refineries as Greenfields sites. They proposed that the models considered showed that an integrated pull mill requires much less biomass per unit of liquid fuel reasoning that BLS are charged against services provided to the mill (i.e. chemical recovery, process steam and power), not against the liquid fuel. In effect a cross-subsidisation. This comparison was made with NREL data on the production of ethanol and mixed alcohols. However, this assumes a simple biofuels model rather than a combined chemical and biofuels position. Production of higher value biobased products could call the study’s conclusions into question. The development of and growth in demand for platform chemicals is considered in the next chapter.

3. Comparing biochemical and thermochemical Biorefineries

The above analyses suggest comparisons are both difficult and open to critical interpretation based on the varying assumptions that have been made. Wright and Brown¹⁶⁷ undertook a comparative analysis of the capital and operational costs of a 568M litres/ year biofuel plant. The table below summarises their results.

Table 4-4: Capital and operating costs for a 568M L/yr petroleum equivalent plant (in 2005 dollars A$)

Fuel	Total Capital Costs $M	Capital Cost per unit of production: Per barrel per day equivalent	Operating costs ($ per Litre) Per litre petroleum equivalent
Grain ethanol	131	15,300	0.38
Cellulosic ethanol	889	89,400	0.55
Methanol	713	77,650	0.40
Hydrogen	639	69,400	0.33
Fischer-Tropsch	1,005	101,200	0.56

The conclusions from this analysis suggest that for a Greenfield site;

• 
  advanced biofuels have a very high comparative capital cost;
  
• 
  thermochemical production of hydrogen is the least capital-intensive.(However considerable downstream technical issues need to be resolved, e.g. storage and distribution); and
  
• 
  cellulosic ethanol and Fischer-Tropsch diesel are essentially the same.
  

A more recent analysis¹⁶⁸ looked at 14 different mature technology biomass refining scenarios. Whilst recognising the inherent difficulty of making side-by-side comparisons of performance and cost projections the authors provide a deeper insight into the potential efficiencies, environmental impacts and economics of biorefineries. The modelling and analysis examines thermochemical, biochemical and hybrid biorefineries as Greenfield developments. The outputs varied depending upon design and included ethanol, power, F-T fuels (syngas), protein, demethylether (DME), rankine power. The detailed study’s overall conclusions are;

• 
  mature cellulosic biomass refinery – especially when combined with thermochemical processing offers significant advantages;
  
  • 
    efficiencies equivalent to petroleum based fuels;
    
  • 
    avoiding substantial GHG emissions;
    
  • 
    require modest volumes of process water; and
    
  • 
    can achieve production costs competitive to petroleum production at about US$30/barrel of oil.
    
• 
  the best-performing scenarios involve the carbohydrate fraction being converted biologically and the lignin-rich residue being converted thermochemically.
  

This study can be linked to the ability to retrofit pulp paper mills, as discussed previously, and to use the cellulose stream as a basis for biochemical transformation in to fuels or other chemicals. In the case of Gunn’s pulp mill in Tasmania the anticipated price per MT of dry pulp was A$600/T. In this case an integrated biorefinery intending to divert the cellulose stream for either fuel or chemical production would need to achieve similar or better returns per tonne. Certainly with subsequent processing, prices well above this are achievable¹⁶⁹.

Significant R&D activities¹⁷⁰ need to be undertaken to realise the concepts of an integrated biorefinery. Some of these activities would include;

• 
  waste-heat recovery and exchange;
  
• 
  residue storage and handling;
  
• 
  scale compatibility between biochemical and thermochemical processing;
  
• 
  animal feed protein production as a co-product. Currently (Dupont send to land fall all waste solids from its 1,3 propanediol plants); and
  
• 
  reduction of process water requirements.
  

As an example of an integrated biorefinery focused on both biochemical and thermochemical transformation of biomass Zakzeski¹⁷¹ offered the following model for a lignocellulosic biorefinery focussed on the lignin stream.

Figure 4-4: Lignocellulosic biorefinery scheme with an emphasis on the lignin stream¹⁷².

Three strategies are considered. In the first strategy biomass is gasified or degraded by pyrolysis to a mixture of molecules that can be then used to produce chemicals similar to those produced by the petroleum industry. In the second strategy functional groups present on lignin yield aromatic compounds such as phenol, benzene, toluene and xylene. These platform chemicals (see Chapter 5) are converted using current petroleum-based technologies for the production of bulk and fine chemicals. The third strategy is best suited to fine chemical production with a high degree of functionality.

Studies have shown that the optimum scale for a thermochemical facility could be in excess of 1BL/yr¹⁷³. This is confirmed by the more recent work that compared alternatives feedstocks and operations to the scale of an oil refinery, large-scale ethanol plant and coal fired power station¹⁷⁴. The study puts at question the common proposals to restrict plants to within 100kms of the feedstock. Logistically, supplying feedstock by truck could support a facility producing up to 800ML ethanol per yr based on bioconversion. If supported by rail as the primary delivery mode the capacity increases to 2.3BL per yr. A similar result is possible when pellets are delivered via marine transport.

The analysis indicates logistics as a significant determinant for scale, location and technology. Green or high moisture content biomass is more suitable for biochemical conversion and yields of 300-340L of ethanol per bone dry tonne of feedstock can be expected. For thermochemical conversion of typically dry feedstocks yields could be expected to be 200L of Fischer-Tropsch liquids with energy content approximately 42% higher than ethanol.

Almost all bioenergy feedstock management studies have focused on local utilisation of the resource for bioenergy in a manner similar to that discussed above. However, the large-scale production and transport of woodchips from Australia, particularly in Victoria, Tasmania and NSW would suggest coastal-based facilities are practically and theoretically possible. Even though biomass is a distributed resource, without government intervention it is likely that economies of scale will cause a push toward larger, higher efficiency and lower installed costs-per-unit output plants. The parallel to the paper and pulp industry is highly relevant, new facilities are typically in excess of 1M air-dried tonne per yr of product (approximately 2.5M T/yr of green feedstock). This corresponds to the proposed Gunns’ Mill at Bell Bay.

Results for bio-oil production would suggest that F-T liquid facility of 990ML per yr could be supported by road transport. A 3.5BL per year facility could be supported by rail and 5.3BL facility per year by marine transport. The largest comparator facility in the study is the Aracruz pulp mill that receives feedstock by all three modes of transport¹⁷⁵.

The development of export ports for wood chips in Victoria for domestic and international markets and similar facilities in southern NSW would support consideration being given to the location of biorefineries at such locations. On the downside the slower growth rates for plantations in temperate climates could emphasise the need to develop greater forestry resources in northern Australia. In Brazil eucalyptus plantations grow in a cycle of 7-9 yr rotation and at much higher densities per ha than in temperate climates. This can reduce distance to market, reduce feedstock costs and give greater certainty of supply.

The development of biorefineries of the scale detailed above is not likely to occur in the short-term other than through a co-location with a pulp mill. Even then the technical considerations are yet to be resolved for processing the quantities of lignin and other bioproducts possible from such a facility. The history of the petrochemical industry would suggest economies of scale cause a move to larger more efficient plants capable of processing feedstocks from a variety of sources.

The concept of a biorefinery and the potential product streams arising from an integrated facility is currently undergoing dramatic evolution. There is also a growing consensus about the need to create a facility capable of both biochemical and thermochemical transformation of temperate biomass, which will provide the greatest possible flexibility and product diversity. The ability to mix and match feedstocks is of great significance – especially in the Australian situation, where there may be insufficient base material of any single variety in any one season.

The capital and operating costs for establishing cost-competitive biorefineries are far from trivial. The upfront investment required prior to reaching profitability can be substantial, with some estimates being in excess of A$1B. However, significant opportunities also exist for the establishment both of Greenfield sites and for retrofits of existing pulp mills that could provide rates of return that would justify this investment.

The overall success of any facility is determined by its ability to ensure secure access to feedstock, with the right pre-treatment technology, into a plant fully capable of utilising the right mix of technologies to deliver products that meet customer demand.

Access to technologies described above can be achieved by Australia through collaborations or licensing arrangements. Collaborations such as exist between Microbiogen and PureVision and ROC and Dynamotive provided evidence of willingness by USA based companies to explore utilisation of their technologies in Australia. Similarly, discussions with CIMV (France) and Inbicon in the Netherlands indicated access to technology and support would be possible. Clearly, there would be a price for access through licensing arrangements and technology transfer. The recent studies by Costkata to establish a MSW gasification plant in Victoria also indicate a willingness by international companies to invest in Australia given the right incentives by government. The Federal government’s current bioenergy programs could be expanded to include biobased (non fuel) products to provide financial leverage to attract international collaborations. The following table provides a spectrum of current companies that are producing or close to producing significant quantities of bio-based products. In addition to those mentioned previously discussions with Gevo and Dupont both expressed interest in partnering or establishing production facilities in Australia.

Table 4-5: Examples of companies currently producing or at proof of concept for bio-based products¹⁷⁶

Company	Product	Contact
Gevo	Isobutanol	www.gevo.com
Verdezyne	Adipic Acis	www.Verdezyne.com
Myriant Technologies	Succinic Acid	www.myriant.com
Opexbio	Acrylic acid	www.opxbio.com
Sythezyme	w-hydroxyfatty acids	www.sythezyme.com
Zeachem	Ethyl acetate	www.zeachem.com
genecor	Isoprene	www.genecor.com
Genomatica	1, 4 butanediol	www.genomatica.com
NatureWorks	Lactic Acid (140kT/a)	www.natureworksllc.com
Dupont	1,3 propanediol (45kT/a)	www.dupont.com
Braskem	Bioethylene (200kT/a)	www.braskem.com

A challenge with accessing international technology, as discussed with several research and development organisations (CSIRO, Bioenergy Australia, Monash University, Deakin University, University of Sydney) is that of the appropriateness of particular technologies to Australian biomass. International data, and small amounts of Australian research indicate variability in yields and sub-components of biomass following pre-treatment. Certainly, data provided by PureVision Inc, suggests that different biomasses will produce different quantities and quality of sub-components for different biomasses. The only feasible way to resolve these issues is through access to both pilot and demonstration size facilities to validate both quantity of pre-treatment material and the eventual quality of the downstream biobased products.

There is the potential for Australia to simply ‘re-invent the wheel’ in its R&D. However, by government facilitating collaborations and the development of an appropriate scale of facilities with a range of technologies these problems could be averted. Any such facility must be open to all R&D organisations and companies, similar to the NCRIS model however with attached operational and contestable funding to ensure the facilities are utilised by many not just a few.

Key Messages:

First-generation biorefineries have little scope for value-added co-products.

Thermochemically based second-generation biorefineries are approaching commercial scale. Biochemically based biorefineries are still moving through pilot to commercial scale development

Biochemically based biorefineries offer greater scope for multiple products.

Only 2 second-generation pilot-scale plants exist in Australia. There are 68 known advanced biofuels and bioproduct pilot facilities overseas.

Commercial integrated biorefineries offer significant economies of scale benefits. Improvements can be 2 orders of magnitude going from pilot to commercial scale.

Biomass used to produce bio-oil in thermochemically efficient operations as part of pulp mills could produce up to 4.5 Billion litres of transport fuel in Australia.

In the longer term an optimum size for a biorefinery will likely process in excess of 2.5M Tonne of biomass per year. In the shorter term early biorefineries can grow through applications of multiple product streams, development of specialty chemicals (e.g. Extractives from oils).

Access to and the development of demonstration facility with a range of pre-treatment technologies is essential to validate the application of Australian biomass to biobased products.