IN Brief:
- Protopia customer trials have used a 15,000-litre fermenter, with a 120,000-litre run planned later in 2026.
- The torula yeast ingredient contains at least 70% protein and is being developed around flexible fermentation feedstocks.
- A larger European production step is intended to support 2,000–10,000 tonnes of annual output before wider licensing.
dsm-firmenich is preparing to move Protopia, its torula yeast protein ingredient, from customer-trial batches into substantially larger fermentation runs as it tests whether the process can support commercial-scale production.
The company has already produced material in a 15,000-litre fermenter and plans to move to a 120,000-litre vessel later in 2026. The next stage is expected to use a larger European facility, most likely in Eastern Europe, with annual output targeted at roughly 2,000 to 10,000 tonnes. dsm-firmenich aims to make around 400 to 500 tonnes from the middle of 2027 to support market development and customer trials.
Protopia is based on torula yeast, with the whole microbial biomass harvested after fermentation. dsm-firmenich says the ingredient contains at least 70% protein, has a neutral taste and colour, and provides functional properties including gelation and emulsification. The company has optimised both the strain and process so it can use several feedstocks, including ethanol, methanol, acetate, sugar, and molasses.
Karim Kurmaly, director of single cell protein at dsm-firmenich, said the economics were approached from the production target backwards. “We started in reverse order: if this is the COGS, how do we get there?” The development work has focused on yield, productivity, feedstock flexibility, and a process that can be transferred to larger assets without relying on unusual downstream equipment.
Scaling fermentation without exotic hardware
dsm-firmenich has evaluated continuous fermentation but concluded that a semi-continuous process offers a better balance between cost and contamination risk. In a fully continuous process, a contamination event can compromise a larger volume of product and extend downtime, whereas semi-continuous operation creates more defined production boundaries.
The downstream route varies by end market. Aquaculture and pet-food applications can use a sequence of dewatering, pasteurisation, and drying. Food applications require additional preparation to keep the material pumpable before spray drying, while particle size can be adjusted to suit the intended formulation. Kurmaly’s point is that these stages can be handled with established industrial equipment rather than a highly specialised plant.
That distinction becomes more important as alternative proteins move from laboratory development into established food-processing environments. Fermentation capacity alone does not determine whether an ingredient is commercially viable. Separation, thermal treatment, drying, powder handling, hygiene control, and batch consistency all have to operate at compatible throughputs. A strain with a good laboratory yield can still become expensive if downstream processing consumes too much energy or creates bottlenecks.
The planned 120,000-litre run will test more than vessel size. Mixing, oxygen transfer, heat removal, contamination control, and downstream loading all become more demanding as fermentation volume rises. Behaviour that is stable at 15,000 litres cannot simply be assumed to remain unchanged at eight times that scale.
dsm-firmenich is also trying to avoid a model in which future output depends entirely on company-owned plants. Its longer-term plan is to license the strain and process technology to local production partners, with dsm-firmenich supplying know-how and technical support. Kurmaly said the company was working on licensing agreements in three countries.
Existing fermentation capacity could therefore become part of the commercial route. Kurmaly has pointed to Indian ethanol producers with underused assets as one potential fit, because plants already built around fermentation may offer a lower-capital route than constructing every Protopia facility from scratch. Any conversion would still depend on hygiene standards, feedstock economics, utilities, downstream equipment, and the specification required by the intended market.
Food applications present a different test from feed. dsm-firmenich says Protopia’s neutral colour and taste allow it to function as a primary protein source, while gelation and emulsification create additional formulation options. Those characteristics still have to hold across real products, where heat treatment, water binding, flavour systems, texture, shelf life, and interaction with other proteins determine whether an ingredient is practical.
Regulatory requirements will shape the pace of expansion by geography. The torula yeast itself has a history of food use in several markets, but the finished ingredient, production process, labelling, and intended claims still have to fit the rules that apply in each country. That makes the initial commercial volumes important because they give customers enough material to run factory trials rather than relying on bench-scale samples.
Single-cell protein has attracted substantial interest because microbes can produce concentrated protein without the same land requirements as conventional crops or livestock. Industrial economics remain the harder test. Fermentation plants need high utilisation, reliable feedstock supply, strong contamination control, and downstream systems capable of turning broth into a consistent, saleable ingredient.
Protopia is now approaching the scale at which those economics become easier to measure. If the 120,000-litre run performs as intended, the larger European step will shift the discussion from whether the strain can make protein to whether the entire process can make enough of it, consistently and at a cost customers will accept.
