By Jan H. Schut
It’s already mid January 2011. How much longer before the biodegradable biopolymer PHA (polyhydroxyalkanoate), which Metabolix Inc. in Cambridge, Mass. (www.metabolix.com) has been developing since 1992, becomes commercial? For the January blog, I combed public filings for Metabolix, a technology spinoff from Massachusetts Institute of Technology, and its joint venture partner in developing PHA, agro-products giant Archer Daniels Midland Co., Decatur, Ill. (www.adm.com), for clues.
PHA biopolymers range from ABS-like engineering resins to sticky adhesives, all with backyard compostability. They’re grown inside the cells of bacteria in a batch fermentation process that goes back to the 70s and 80s, when Imperial Chemical Industries (ICI) first developed Biopol PHA from naturally occurring bacteria. ICI’s patents were first bought by Monsanto and ultimately acquired by Metabolix in 2002.
Metabolix co-founders, Oliver Peoples and Anthony Sinskey at MIT identified key genes in the bacteria that could be engineered to make them store more PHA than naturally occurring microbes. PHA can also be produced with different properties by changing the bacteria and especially changing how the bacteria is fed, creating rigid PHA molecules with a short side chain and more elastomeric ones with a longer side chain.
Metabolix and ADM have partnered on PHA development since 2004. ADM has since built and started up the world’s largest microbial fermentation plant for PHA in Clinton, Iowa, called Clinton 1, with nameplate capacity of 110 million lb/year. ADM made test PHA product there in December 2009 and started the plant up in March 2010.
At that time, Metabolix expected to be fully commercial in late 2010. They still aren’t. Commercial production is clearly defined in the joint venture as requiring “the sale to third parties of at least one million pounds of Mirel [PHA] manufactured at the [Clinton plant]. Qualifying sales must meet certain criteria, including a minimum order size, product must be accepted by the customers in accordance with the terms of their contracts, and payment must be received.” On Jan. 12, 2011, Metabolix told analysts at a Piper Jaffray conference in New York it expects to reach the million lb threshhold by the middle of this year and to sell out the Clinton plant by mid 2013. Why the delay?
Metabolix has produced pre-commercial Mirel PHA by fermentation since 2006, first on a toll basis at a 10-ton/month pilot plant at National Ford Chemical Co., Fort Mill, S.C. (www.nationfordchem.com), then at a pilot plant in Lowell, Mass., leased in March 2007. Compounding and formulation were done at U. Mass., Lowell, combining different types of PHA to get different grades for injection molding, blown and cast film, sheet and thermoforming.
Precommercial sales after 2007 were steady, but small. Metabolix reported them as R&D revenue, which totaled $147,000 in 2007; rose to $229,000 in 2008; fell to $152,000 in 2009; and rose sharply to $212,000 in the first six months of 2010. Metabolix won’t disclose precommercial sales volumes, but it prices PHA at $2.25-$2.75/lb. That works out to about 58,800 lb sold in 2007; 91,600 lb in 2008; 60,800 lb in 2009; and 84,800 lb in the first half of 2010 for a total of about 300,000 lb in nearly four years.
Metabolix named eight early customer test programs. In 2007, Target Corp., Minneapolis, Minn. (www.target.com), tested biodegradable PHA gift cards. In 2008, Heritage Plastics Inc., Picayune, Miss. (www.heritage-plastics.com), tested PHA and commercialized it in 2010 as an ingredient in one formulation for BioTuf can liners. Ball Horticultural Co., West Chicago, Ill. (www.ballhort.com), tested it in horticultural pots. Labcon North America Inc., Petaluma, Calif. (www.labcon.com), tested it for trays for medical pipets. And the Sanford/Paper Mate division of Newell Rubbermaid Inc. in Oakbrook, Ill. (www.newellrubbermaid.com) signed a supply agreement for an injection molding grade for a biodegradable ball point pen and mechanical pencil.
In March 2009, Bioverse Inc., Pipestone, Minn. (www.bioverse.com), which makes environmentally friendly algae control products, agreed to buy injection molding grades for “a biodegradable version of its AquaSphere PRO pond treatment for golf courses.” And in October 2009, Pharmafilter BV in Amsterdam, the Netherlands (www.pharmafilter.nl), chose PHA for disposable hospital products.” Pre-commercial sales all came from the pilot plant.
Metabolix also makes PHA for sister company Tepha Inc. in Lexington, Mass. (www.tepha.com), a spin off from Metabolix in 2006 by Metabolix co-founder Simon Williams. Tepha has exclusive rights to PHA for in vivo applications and uses it for absorbable sutures, cardiovascular devices and drug delivery systems. Sales to Tepha totaled $157,000 in 2007; $120,000 in 2008; and $120,000 in 2009, all from the pilot plant. Metabolix then “reduced Mirel pilot plant manufacturing activities as a result of commencing operations at the commercial facility.” After June 30, 2010, pre-commercial sales from the pilot plant ceased completely.
WHAT’S HOLDING MIREL UP NOW?
Now PHA production is all at Clinton 1. It’s hard to tell how often that plant runs, but it ran enough last summer to annoy the neighbors. PHA production apparently gives off a strong, unpleasant plastic smell. Neighbors of the Clinton PHA plant filed a suit in U.S. District Court in Davenport, Iowa, last September, complaining about “’deafening’ noises, noxious fumes, vibrations that shake items off the walls of nearby homes and cause structural damage, intense light, dust that covers their homes and property,” and other complaints.
As might be expected, ADM is also working to make PHA production more efficient, especially converting the plant to a more efficient way to separate the polymer from the bacterial waste. Microbial PHA production leaves 10% to 20% biomass residue from bacteria cells, which has to be removed from the plastic. ADM intends to burn this residue along with unusable corn seed to replace up to 20% of the energy for a new coal-fired co-gen plant that ADM built in Clinton to supply steam and power to both its wet corn and PHA plants. (ADM nationwide is under pressure from the Environmental Protection Agency over air pollution, so burning 20% biomass is one way to improve its carbon footprint.) So far, ADM hasn’t begun to burn the biomass waste from PHA.
Sales are slow partly because Telles LLC in Lowell, Mass. (www.mirelplastics.com), the 50/50 ADM/ Metabolix joint venture to sell Mirel PHA, is being extremely careful about who gets material. You no longer have to sign a confidentiality agreement the way you did in the pilot plant phase, but a processor still can’t pick up the phone and order 1,000 lb. PHA comes in 10 grades, requires specific processing conditions, and has a reverse temperature profile, so Telles’s position is that it needs to work closely with new customers.
Old customers agree.“PHA molds very well provided you know what you are doing,” says Bret Marschand, senior development manager at Paper Mate, which injection molds the Mirel PHA pen and pencil. “Bottom line, Mirel has a learning curve, and Metabolix was very supportive of us getting up to speed with production.” Costomers also speak highly of Telles’s reliability as a supplier.
The process of selecting appropriate potential customers, getting them up to speed, waiting for market testing of a new product, and finally getting an order takes 9 to 15 months, Metabolix says. Some like Target don’t make it. Target had a successful launch, but Metabolix couldn’t meet their volume requirements from the pilot plant, Metabolix CEO Richard Eno told analysts. Eno gave an idea of attrition in his latest earnings call when he said that Telles continued to move “a large number of potential customers through the product development process” in 3Q 2010 with material from Clinton, but also said that Telles got only three small orders from new customers in that quarter.
Delay could hurt Mirel, which faces competitors who are already in production. Tianan Biologic Material Co. in Ningbo, China (www.tianan-enmat.com) makes PHBV (polyhydroxybutyrate-valerate); Kaneka Corp. in Osaka, Japan (www.kaneka.co.jp) makes PHBH (poly 3-hydroxybutyrate-co-3hydroxyhexanoate); and Tianjin Green Bio-Science Co., Tianjin, China (www.tjgreenbio.com) makes PHA with support from DSM Venturing. Green Bio started a commercial plant ahead of ADM in late 2009 with nameplate of 22 million lb/year. Meredian Inc., Bainbridge, Ga. (www.meredianpha.com), which bought Nodax PHA technology from Procter & Gamble Co., is another potential competitor.
My suspicion is that there are two holdups: a problem with the purification step for PHA and slow commercial orders. ADM has never said how big “the initial phase” of the Clinton plant is, but normal engineering practice for a new process would be to go from a pilot plant to a scale up plant of something like 3 million lb/year before building more. This is especially doable since PHA is a batch process. That could explain why Tellis seems to pick only small test applications. But it doesn’t explain why Tellis still hasn’t sold even a million lb. That has to be just the slowness of commercial orders–the simple fact of life for all biopolymers, that while everyone wants to do a trial, it’s hard to get customers to convert a product and give a purchase order.
Meantime, ADM gets a peculiar tax windfall out of its PHA plant. By integrating the PHA plant with ADM’s new co-gen plant in Clinton, the co-gen plant falls under state tax assessment, not municipal. ADM’s tax bill reportedly drops from $4 million/year (after 10 years) to only $800,000/year. That tax break could repay the cost of the PHA plant faster than selling PHA. The clearest sign, however, that ADM is serious about PHA in the long run is it laid out the plant with space to be quadrupled.
Mirel Potential Applications Site: http://www.mirelplastics.com/discover/applications.html>
Heritage Plastics Biotuf Site: http://www.heritage-plastics.com/en/products_biotuf.htm
Metabolix’s latest 10-Q report to the SEC: http://ir.metabolix.com/secfiling.cfm?filingID=1437749-10-3778