BSFL Conversion Math: From Food Waste to Animal Protein

What Question the Conversion Math Answers

The conversion math page exists to answer a specific operational question: given a known feedstock intake volume, what can you expect to produce, and at what price, from each output stream? This is the calculation that determines whether a BSFL facility at a given scale is viable before any capital is committed.

The people asking this question are: operators evaluating whether to build or retrofit a BSF bay alongside an existing food processing operation, investors running due diligence on a BSFL company, feed buyers comparing BSFL protein meal to fishmeal and soy protein on cost-per-kilogram-of-protein, and agronomists evaluating frass as a biofertiliser input alongside conventional nitrogen sources.

This page does not repeat the biological detail that drives the conversion parameters. That is covered in the biology page. Here, the biology is an input, not the subject. The subject is the arithmetic of what a working facility produces, at what margins, from inputs that are frequently available at zero or negative cost.

The worked example uses the Protix facility in Bergen op Zoom, Netherlands, as the reference scale: 250 tonnes per day of food industry side streams processed, approximately 30 tonnes per day of dried BSFL protein meal output, as of 2022 operations (Protix company disclosures 2021-2023). This is a real facility with published throughput data. It is the largest BSFL operation in Europe and the benchmark for commercial-scale conversion math.

The Mechanism: How Bioconversion Produces Multiple Outputs

The BSFL bioconversion process is not a single-output system. This is the structural difference between BSFL and soy or fishmeal as protein sources. Soy produces one product: soy protein meal, with a defatted cake byproduct. Fishmeal produces one product: dried fish protein, with fish oil as a secondary output. BSFL bioconversion produces three marketable output streams from a single input: larvae (the protein and fat fraction), frass (the biofertiliser fraction), and chitin (extracted from the exoskeleton fraction). Each stream serves a different buyer at a different price point.

The input side is equally important. BSFL facilities do not buy their feedstock. They charge for accepting it. Pre-consumer food industry side streams, including brewery spent grain, bakery waste, fruit and vegetable processing offal, and supermarket perishable recall, are disposed of by their generators at a tipping fee. In Western Europe, these fees range from 20 to 80 EUR per tonne depending on waste category classification and regional regulation. The practical effect: the feedstock that is the primary input to protein production has a negative cost. The facility is paid to receive the input it converts into sellable protein.

This is the inversion that makes BSFL economics structurally different from soy or fishmeal. Both competing protein sources pay for their inputs. Soy operations buy soybean. Fishmeal operations pay for fishing vessel operating costs and wild fish purchasing. BSFL operations receive tipping income from their input. The three-stream output, the negative-cost feedstock, and the 14-day cycle combine to create a cost structure that conventional protein production cannot replicate through any efficiency improvement.

One caveat on the tipping fee income: it is only available for pre-consumer, segregated food industry side streams. Mixed municipal solid waste is not an acceptable feedstock for two reasons: contamination with plastics and metals would concentrate in the frass output, and the regulatory category for mixed municipal waste is different (and more costly) than food industry byproduct streams. The feedstock is the whole game. Facilities that built their business models on municipal waste streams rather than food industry contracts are among the failures in the European BSFL market from 2018 to 2022.

The Numbers: Worked Facility Calculation

Working from the Protix Bergen op Zoom benchmark: 250 tonnes per day of food industry side streams in, approximately 30 tonnes per day of dried protein meal out, plus 50-80 tonnes per day of frass (Protix disclosures 2021-2023). These figures allow calculation of the facility-level economics.

Input side

250 TPD feedstock at a tipping fee of 40 EUR per tonne (mid-range for European food industry byproduct): 10,000 EUR per day in tipping income, or approximately 3.65 million EUR per year before any protein or frass revenue. This is income from the waste intake alone, before a single kilogram of protein is sold.

Protein output

30 TPD dried protein meal at 2,100 EUR per tonne (mid-range for Q4 2023 European BSFL protein meal, per IFFO fishmeal price reports and European feed ingredient market data 2023): 63,000 EUR per day in protein revenue, or approximately 23 million EUR per year at full throughput. This is the primary revenue line.

Frass output

65 TPD frass (midpoint of the 50-80 range) at 250 EUR per tonne: 16,250 EUR per day, or approximately 5.9 million EUR per year. Frass is a significant secondary revenue line, not an afterthought. At the 250 TPD scale, frass revenue alone exceeds the tipping income.

Gross revenue of approximately 32.5 million EUR per year from a 250 TPD facility. The cost structure for such a facility includes: feedstock logistics (collection and transport, typically 15-30 EUR per tonne received), labour for colony management and harvest, facility energy (climate control is the dominant variable, roughly 0.8-1.2 EUR per kg of protein output depending on climate), processing (drying, separation, pelleting), and facility capex amortisation. Protix raised 50 million EUR in a Series D round in 2022 to expand, which implies a capital intensity consistent with 150,000-200,000 EUR per daily tonne of throughput capacity at that facility scale. Note: facility profitability at this scale is not fully public; the math above is a gross revenue calculation, not a net margin calculation.

Price comparisons require a protein equivalence adjustment. Soy protein concentrate at 900-1,200 EUR per tonne contains approximately 65-70 percent protein on a dry basis. BSFL protein meal at 1,800-2,400 EUR per tonne contains 45-55 percent protein. On a cost-per-kilogram-of-protein basis, the gap narrows considerably. At mid-range prices: soy delivers protein at approximately 1.54 EUR per kg of protein (1,050 EUR divided by 0.675 kg/kg). BSFL delivers protein at approximately 3.68 EUR per kg of protein (2,100 EUR divided by 0.57 kg/kg). BSFL protein is still more expensive per kilogram of protein at current scale and prices. The structural cost advantage (negative feedstock, lower water, faster cycle) has not yet translated to market parity, largely because BSFL protein volume is small relative to global soy and fishmeal supply. The case for BSFL is not commodity parity today; it is an improving cost structure trajectory against a soy market that is increasingly constrained by land and water, and a fishmeal market limited by ocean productivity.

Sources: BSFL conversion parameters from Diener et al. 2011, Waste Management Research. Soy water footprint from Mekonnen and Hoekstra 2012, Ecosystems. BSFL water figure from Oonincx et al. 2015, PLoS ONE. Pricing from IFFO fishmeal price reports and European feed ingredient market data Q4 2023.

The Practitioner View: Feedstock Contracts and Output Sales

The conversion math looks clean on a spreadsheet. The operational reality is that both sides of the equation, feedstock supply and product offtake, require contracts before a facility can be sized and financed. This is where the European BSFL failures of 2018-2022 concentrated. The facilities that struggled or collapsed had not locked in sufficient feedstock supply or protein meal offtake before committing to large-scale construction. The bioconversion was not the failure. The capital sequencing was.

Feedstock contracts for food industry side streams are typically 1-3 year agreements with a food manufacturer, distributor, or retailer. The contract specifies: waste category, delivery schedule, moisture and contamination tolerances, and the tipping fee. Most European commercial BSF operations prioritise brewery spent grain and bakery waste as their primary feedstocks because these have consistent composition, predictable moisture content (20-30 percent for brewery grain, versus 70-80 percent for fruit processing water), and are available in high volumes from urban industrial parks. Vegetable processing water is high in nitrogen but low in total solids, which reduces dry protein yield per tonne of input and complicates logistics.

Protein meal offtake contracts are typically sold to aquaculture feed compounders (Skretting, Cargill Aqua Nutrition, BioMar) or specialty poultry feed producers. EU Regulation (EU) 2021/1372, enacted in August 2021, authorised processed animal protein (PAP) from insect species including Hermetia illucens for use in poultry and pig compound feed. Prior to this regulation, the legal market for BSFL protein in the EU was limited to aquaculture and pet food. The August 2021 authorisation approximately tripled the potential market by volume.

Frass sales are typically handled through agricultural input distributors or direct to large organic farm operations. The market for frass as a biofertiliser is less developed than the protein meal market; most commercial BSFL operators treat frass revenue as a secondary income stream rather than a primary driver. The chitin extraction tier is earlier stage still: as of 2023-2024, only a small number of operators (primarily Ynsect and a few Asian producers) have commissioned chitin extraction at commercial scale. The price premium for pharmaceutical-grade chitin at 5,000-15,000 EUR per tonne (Grand View Research Chitosan Market Report 2023) is real, but the extraction process adds capital cost and operational complexity that smaller facilities defer to a second phase.

Where It Fits: Three Streams, One Loop

The conversion math is the economic argument for BSFL bioconversion as a system, not just as a protein source. The three-stream structure means a BSF facility operator can optimise across multiple markets simultaneously. In periods when protein meal prices are depressed (as they were briefly in 2020 during COVID-related aquaculture contraction), frass revenue provides a floor. When protein prices recover, the protein stream re-becomes the primary driver. When chitin extraction becomes commercially viable at smaller scale, it layers a third revenue line on top of the existing operation without changing the biological process.

The frass output connects directly to the composting and regenerative agriculture nitrogen stack. BSFL frass is compost by another name and enters the same market, but with higher nitrogen density (2-5 percent N versus 0.5-2 percent for finished food waste compost) and the additional chitin-mediated plant immune priming effect (Quilliam et al. 2020, Waste Management). An operator running a BSF bay adjacent to a market garden or organic farm creates a closed loop: food processing waste in, protein out for the local aquaculture or poultry operation, frass onto the growing beds. The math supports this at scales well below the Protix benchmark.

For a 5 TPD operation, the reference calculation: 5 tonnes of food industry waste per day produces approximately 1 tonne of wet larvae, yielding 300-350 kg of dried protein meal per day and 2,000-2,500 kg of frass per day. At a tipping fee of 40 EUR per tonne, tipping income is 200 EUR per day. At 2,100 EUR per tonne for protein meal and 250 EUR per tonne for frass: protein revenue is 630-735 EUR per day, frass revenue is 500-625 EUR per day. Gross daily revenue: approximately 1,330-1,560 EUR per day, or 485,000-570,000 EUR per year before operating costs. The break-even facility size for a well-run operation with secured feedstock contracts and local offtake agreements is closer to 5-10 TPD than the 50-100 TPD threshold cited in earlier market analyses. Modular facility designs available from 2023 onward have driven this threshold down considerably.

The poultry feed application and aquaculture fishmeal replacement detail how the protein meal output performs in the specific feed formulations buyers require. The conversion math in this page produces the raw numbers. Those pages translate raw numbers into inclusion rates, feed formulation ratios, and performance equivalence against the benchmarks the buyers are comparing against.