Latest / The Manufacturing Podcast with Fexingo: Factories, Supply Chains, and Industrial Business / How a Japanese Factory Is Making Fish Feed from Insects
Transcript
- Lucas: If you've ever looked at the ingredients on a bag of salmon feed, you've probably seen fishmeal listed first. That's essentially ground-up wild-caught fish — anchovies, sardines, herring — fed to farmed salmon. Luna: Which feels a little backwards, right? We're farming fish by catching wild fish to feed them. Lucas: Exactly. It takes roughly three kilograms of wild fish to produce one kilogram of farmed salmon. That ratio is the central tension in aquaculture right now — and it's why a factory in northern Japan is growing black soldier fly larvae at industrial scale. Luna: I've seen the black soldier fly mentioned in agtech newsletters for years, but I've never heard of a factory actually producing insect feed at a meaningful volume. How big is this place? Lucas: The facility — it's operated by a company called Musca, which isn't a household name even in Japan — processes twenty-four metric tons of insect biomass per day. That's fresh larvae. After drying, that yields roughly six tons of protein-rich powder. Luna: So about six metric tons of insect protein per day. How does that compare to a typical fishmeal plant? Lucas: A small to medium fishmeal facility would turn out maybe fifty tons a day. So Musca isn't replacing fishmeal globally yet. But here's the interesting number: one ton of dried black soldier fly larvae replaces roughly one point two tons of wild-caught fishmeal in salmon feed formulations, based on digestible protein content. Luna: So you need less insect protein by weight to get the same nutritional result. That's a meaningful efficiency gain. Lucas: It is. And it's not just protein. The larvae contain lauric acid, a medium-chain fatty acid that's been shown to improve gut health in fish and reduce the need for antibiotics. That's a secondary selling point for salmon farmers under pressure to reduce antimicrobial use. Luna: I want to understand the actual production process. How do you 'farm' insects indoors at that scale? Lucas: Musca starts with adult black soldier flies living in mesh enclosures. They lay eggs on corrugated cardboard strips that are placed into the colony. Those eggs are collected and transferred to climate-controlled rearing trays filled with organic byproducts — mostly spent grain from a nearby sake brewery and vegetable trimmings from a local wholesale market. Luna: So the larvae are essentially upcycling waste that would otherwise go to landfill or incineration. Lucas: Exactly. The larvae eat the waste, grow rapidly over about fourteen days, then they're harvested before they pupate. They're either sold as whole dried larvae for poultry feed or processed into a defatted protein powder for aquaculture. The fat fraction, which is about thirty percent of the dry weight, is used in pet food and cosmetics. Luna: You said 'defatted' — so they press out the oil? Lucas: Yes, mechanical pressing, similar to how you'd extract oil from soybeans. The resulting meal has about sixty percent crude protein, which is comparable to high-grade fishmeal. And the oil itself is rich in lauric acid, so there are multiple revenue streams from the same biomass. Luna: What's the energy cost of running this facility? Indoor insect farming needs heating, ventilation, lighting. I've seen some critics argue that the carbon footprint of insect protein can be worse than fishmeal if the energy source is dirty. Lucas: Musca claims its carbon footprint per kilogram of protein is about sixty percent lower than fishmeal, but that depends heavily on the energy mix. The facility is in Fukushima prefecture, which draws power from a mix of hydro, solar, and some natural gas. If you built the same plant in a coal-heavy grid, the calculus would shift. Luna: Fair. And the economics? What does insect protein cost versus fishmeal? Lucas: Fishmeal currently trades around one thousand five hundred dollars per metric ton. Musca's insect meal is priced at roughly two thousand per ton — about a third more expensive. But they expect that gap to narrow as they scale. Their next facility, planned for 2028, is designed for forty metric tons per day, which should bring costs down significantly. Luna: A thirty percent premium is steep for a commodity input. Who's actually buying at that price? Lucas: A major Norwegian salmon producer — I can't name them due to the non-disclosure agreement, but they're one of the top five globally — recently signed a five-year offtake agreement for a portion of Musca's output. That's the kind of validation the whole industry needed. Luna: So one of the biggest salmon farmers in the world is betting that insect feed will be part of their supply chain long-term. Lucas: Exactly. And it's not just about price. Fishmeal supply is volatile — it's subject to El Niño events, fishing quotas, and geopolitical tensions in the waters off Peru and Chile where most of the world's fishmeal is produced. A stable, domestically produced protein source has strategic value beyond the per-kilogram cost. Luna: That makes sense. For a salmon company, predictability of supply might be worth paying a premium. Lucas: Right. And Japan's regulatory environment just made this easier. In April 2026, the Ministry of Agriculture revised its feed safety standards to explicitly include insect-derived proteins for aquaculture. That removed a legal grey area that had kept many Japanese feed mills from experimenting with insect meal. Luna: So before that, it wasn't exactly illegal, but it wasn't formally approved either. Lucas: Correct. The regulation was ambiguous. Now it's clear, and that opens the door for more factories like this one. There are already three other insect protein startups in Japan that have announced plans to build facilities in the next eighteen months. Luna: Let's talk about the waste input side. Musca uses spent grain from sake breweries and vegetable trimmings. Is that supply reliable? Lucas: They've signed long-term agreements with three local breweries and a wholesale produce market. The volume is consistent. One of the challenges they've faced is the seasonal variation in vegetable waste — more leafy greens in summer, more root vegetables in winter. The larvae don't seem to mind, but the nutritional profile of the resulting meal shifts slightly, and they've had to blend batches to maintain a consistent spec for their customers. Luna: Interesting. So the factory's output quality depends on what the local farms and breweries are throwing away that week. Lucas: Exactly. That's the trade-off of a circular model. You're dependent on someone else's waste stream. Musca has dealt with it by building a small buffer storage facility and using a blending algorithm to match target protein and fat levels. It's a lot like how a feed mill blends different grains to hit a nutritional spec. Luna: I've heard that black soldier fly larvae can also be fed on food waste from supermarkets and households. Why isn't Musca doing that? Lucas: Consistency. Household food waste is extremely variable and often contaminated with plastics and packaging. That adds a sorting step that raises costs. Industrial byproducts from breweries and produce markets are much more uniform. Musca made a deliberate decision to use low-cost but predictable waste streams rather than chasing the cheapest possible input. Luna: That's a smart operational constraint. It keeps the process simple. Lucas: And it keeps the output quality high enough to sell to salmon farmers, who are notoriously demanding about feed consistency. A salmon farmer will reject a batch if the protein level is even one percent below spec. Luna: Which brings us back to the economics. If you're selling at a premium, you have to deliver premium quality. Lucas: Exactly. And so far, Musca has managed that. The salmon producer that signed the offtake agreement did its own feeding trials for six months before committing. The results showed equivalent growth rates and feed conversion ratios compared to a standard fishmeal-based diet. Luna: So the science supports it. Now it's really a scaling problem. Lucas: It is. And scaling insect farming is genuinely hard. You're dealing with a living organism that has specific temperature and humidity requirements. A disease outbreak in a fly colony could shut down the entire facility. Musca has redundant breeding colonies in separate climate zones within the building to mitigate that risk. Luna: So it's biosecurity, essentially. Like a shrimp hatchery or a poultry barn. Lucas: Exactly. And the capital costs are significant. Musca raised about thirty million dollars in venture funding and government grants to build the current facility. The next one, at forty tons per day, is expected to cost around fifty million. That's real money for a company that's not yet profitable. Luna: But if the offtake agreement holds, and if more salmon producers follow, the path to profitability becomes clearer. Lucas: Right. And by the way, this is a good moment to mention something. If you find these deep dives into specific manufacturing operations valuable — whether it's insect feed, battery recycling, or mushroom leather — there's a small way you can help keep the show going. A handful of listeners already chip in a few dollars a month at buy me a coffee dot com slash fexingo, and that support literally covers the cost of producing episodes like this one. Luna: It's true. We don't run ads. The show exists because of that community. So if it's made a difference in how you think about industrial production, it's worth knowing that you can be part of keeping it ad-free. Lucas: Alright. Back to Musca. One more thing I found striking: the facility employs only twelve people per shift. The entire operation — from egg collection to drying and packaging — is highly automated. The only manual tasks are visual inspection of the larvae and cleaning the trays between batches. Luna: Twelve people per shift for a twenty-four-ton-per-day operation. That's an incredibly lean workforce. What's the skill set of those workers? Lucas: Most have backgrounds in agricultural engineering or food processing. They monitor sensors, adjust climate controls, and troubleshoot mechanical issues. The company invested heavily in sensors that measure larval weight gain automatically, so the team can optimize feeding rates in real time. Luna: So it's a data-driven insect farm. I love that phrase. Lucas: It's the future of protein, maybe. Not everyone is convinced. Some feed industry analysts argue that insect protein will always be a niche, limited by the complexity of scaling living organisms. But the numbers in Japan suggest something else. The total addressable market for fishmeal replacement in aquaculture is about five million metric tons per year globally. Even capturing one percent of that would support several facilities of Musca's size. Luna: One percent of five million tons is fifty thousand tons. That's a lot of insect protein. Lucas: It is. And if you look at the growth rate of aquaculture — it's expanding at about five percent per year — the demand for alternative proteins is only going to increase. Wild fish stocks aren't growing. So something has to fill the gap. Luna: And black soldier fly larvae are one of the most efficient converters of low-value organic waste into high-value protein we've found so far. Lucas: Exactly. Seven kilograms of waste yields roughly one kilogram of dried larvae. That's a conversion ratio that rivals soybeans, but without the land use. And the larvae can be grown vertically, indoors, year-round, in almost any climate. Luna: So the factory in northern Japan isn't just a proof of concept. It's a template. If the economics work there, they could work in other countries with similar waste streams and energy grids. Lucas: That's the bet. And it's a bet that several venture capital firms and at least one Norwegian salmon company are making real. Luna: I'll be watching the 2028 facility. If Musca hits forty tons a day, that's a different conversation entirely. Lucas: Same here. For now, it's one of the most interesting factory tours I've done in months.