Singapore Could Solve the Food Problem the Way It Solved Water
Share
In March 2019, Singapore announced "30 by 30": grow 30% of the nation's nutritional needs locally by 2030, without adding an inch to the roughly 1% of land already set aside for farming. It became the headline of the country's food security strategy, the thing every farm tour, ministry speech and school field trip mentioned in the same breath.
On 4 November 2025, at the Asia-Pacific Agri-Food Innovation Summit, Minister for Sustainability and the Environment Grace Fu announced it was being retired. In its place: a target of 20% for "fibre" (mostly leafy vegetables) and 30% for protein (seafood and eggs combined), pushed out to 2035. As of 2024, Singapore was sitting at 8% and 26% respectively on those two measures.
Five years, a pandemic that was supposed to prove the point, hundreds of millions of dollars in grants, and the target itself is what ended up changing.
Where the fresh produce actually came from before 2019
Before getting into who succeeded or failed, it's worth being clear about what Singapore's food system looked like going into 2019, because "30 by 30" wasn't launched into a vacuum. It landed on top of an earlier, and in some ways more decisive, round of the exact same choice: give up on local production where it's expensive to maintain, and import instead.
That choice was made once before, explicitly, with pork. By 1977, Singapore's Primary Production Department had actually reached self-sufficiency in pigs, chickens and eggs, with local farms producing over 1.2 million pigs, 27 million chickens and 508 million eggs a year, even as farmland had already shrunk from around 14,000 hectares in the 1960s to 8,400 hectares in the 1970s. Then in March 1984, Deputy Prime Minister Goh Keng Swee announced that pig farming would be phased out entirely. The reasoning reads almost identically to Grace Fu's remarks in 2025: the land was needed elsewhere, pig waste was polluting the reservoirs, and it would be cheaper in the long run to import pork than to keep subsidising local farms to produce it. By 1989, every pig farm on the island had closed, and Singapore has imported essentially all its pork ever since.
That decision set the template for what followed. Vegetable, fish and the remaining livestock farms were consolidated into a handful of agrotechnology parks, mostly in Lim Chu Kang, Sungei Tengah, Murai, Mandai, Nee Soon and Loyang, starting in the mid-1980s, each on 20-year land leases and mostly using conventional soil-based methods rather than anything resembling climate control. By 1987, official records counted 2,075 farms nationwide, but the total area under farming had already fallen to roughly 2,037 hectares, an average of under a hectare each. Some of these farms are still around: Kok Fah Technology Farm and Yili Farm, founded in 1996, and Bollywood Farms, founded in 2000, all predate "30 by 30" by close to two decades and still grow conventional Asian vegetables like kang kong, xiao bai cai and kailan using largely traditional methods rather than the climate-controlled indoor systems that came later.
By the time "30 by 30" was announced in March 2019, this older system, traditional soil farms plus the earliest wave of hydroponic ventures, was producing a small fraction of what the country actually ate. Sky Greens, one of the earliest commercial vertical farms, put local vegetable self-sufficiency at around 7% on its own website heading into the target. Almost everything else came from imports, overwhelmingly from two sources. Malaysia supplied roughly 40% of Singapore's vegetables, much of it grown in the cooler highland conditions around Cameron Highlands and trucked across the Causeway, a supply chain short enough that it didn't need refrigeration or climate control to stay cheap. China supplied a further 25 to 27%. The rest came from a longer tail of countries including Australia, Indonesia and Thailand, a diversification strategy that predates "30 by 30" by decades and has kept expanding alongside it, from around 140 source countries and regions two decades ago to 187 by 2024.
Eggs looked different even before 2019. Chew's Agriculture and N&N Agriculture, the two names that keep coming up in the "30 by 30" success story, were both already decades-old, commercially proven businesses well before the target existed. What changed after 2019 wasn't the existence of local egg farming, it was that these already-viable businesses got grants and support to automate and scale further, on top of a foundation that was already commercially sound. That's a meaningfully different starting point from vegetables and seafood, where most ventures chasing the "30 by 30" numbers were new entrants trying to build an entirely new capability, in a category, indoor climate-controlled produce, that had no real large-scale commercial track record in Singapore before the target existed.
The failures span every size and every technology
The natural instinct is to look for a single villain. A bad grant scheme. An overhyped technology. A founder who overpromised. The pattern just doesn't cooperate with that story, and it holds up even once you widen the lens past the two or three names that usually get quoted in this conversation. The full list, laid out below, runs from mega-farms and Temasek-backed ventures to a shrimp farm that was actually profitable and left anyway.
Singapore's agtech deal value fell almost 90% from its 2021 peak to just US$187 million in 2023, according to AgFunder data. That's a funding collapse that happened globally, but it landed on a local sector that already carried unusually high fixed costs relative to what it was producing. Different farms, different crops, different technologies, different founders, different owners, sometimes different nationalities of owner, and the same result. That's what a structural problem looks like. It isn't a run of bad luck.
The full roster, split by when each one actually started
Naming five or six failures makes the point, but it understates how consistent the pattern is once you line up more of the operators that have gone through Singapore's farming sector, and it also blurs a distinction worth keeping straight: some of these businesses predate "30 by 30" and were already operating before the target existed, while others were built specifically to chase it. That distinction matters, because it's a rough proxy for how much of each venture's cost base was locked in before the financing problems described later in this piece even applied. Some of the funding figures below are grant amounts, some are total private capital raised, and a few are asset costs rather than funding raised, so they're not directly comparable to each other. A few founding dates for the smaller or less-documented names weren't independently confirmable and are marked as such rather than guessed at.
Farms that predate 30 by 30
| Company | Segment | Founded | Capital involved | Status |
|---|---|---|---|---|
| ComCrop | Vegetables, rooftop hydroponics | 2011 (commercial from 2013) | Not fully disclosed | Still operating, though a meaningful share of harvest has gone to donation in past years rather than sale |
| Sky Greens (Sky Urban Solutions) | Vegetables, hydraulic vertical towers | 2011, farm opened 2012 | Not disclosed | Scaling down as of 2024, greenhouses reportedly dismantled |
| Sustenir Agriculture | Vegetables, climate-controlled indoor farm | 2013 | About US$33 million raised total, including a US$20 million 2024 acquisition deal | Still operating under new ownership (Future Food Foundry), after years of losses |
| Apollo Aquaculture Group | Seafood, multi-storey fish farm | Incorporated 2014; its S$65 million mega-facility was built later, largely during the 30 by 30 period | S$65 million facility cost; debt of roughly S$35 million by 2024 | Judicial management from 2022, ceased operations 2023, facility sold 2024 for S$3.5 million |
| Shiok Meats | Cultivated seafood (alternative protein) | 2018 | About US$21 million raised before merging | Merged into Umami Bioworks in March 2024 after funding difficulties and staff cuts |
| Chew's Agriculture / N&N Agriculture | Eggs | Decades-old operations, both well established before 2019 | Not disclosed | Still operating and expanding, part of the segment that actually hit its 30 by 30 number |
| Barramundi Asia | Seafood, sea bass farming | Founding date not independently confirmed; described in industry coverage as an established operation | Not disclosed | Stopped sea bass farming after a scale-drop disease outbreak |
| Universal Aquaculture | Seafood, shrimp farming (33 tonnes/year) | Founding date not independently confirmed; treated in reporting as an established, functioning farm rather than a new entrant | Not disclosed | Operationally profitable, relocated out of Singapore after its Tuas South lease expired in Nov 2023 |
| Edible Garden City | Vegetables, distributed rooftop and underused-space farming | Founded 2012 | Runs as a social enterprise rather than a pure commercial grower | Still operating, structured around social impact and training rather than competing purely on retail price |
Farms and ventures launched to chase the target
| Company | Segment | Founded / Singapore entity established | Capital involved | Status |
|---|---|---|---|---|
| I.F.F.I (Indoor Farm Factory Innovation) | Vegetables, indoor mega farm | Farm built and grant-funded from 2020 | At least S$2.9 million disbursed from the 30x30 Express grant | Holding company insolvent Nov 2023, farm vacated and closed May 2024 |
| VertiVegies | Vegetables, planned mega indoor farm | Company incorporated 2016; the mega-farm project itself was tendered and conceived as part of the 2019 push | S$40 million grant approved but never drawn down | Plans scrapped 2022, land returned to SFA, publicly confirmed 2024 |
| &ever / Kalera International / Growy Singapore | Vegetables, indoor mega farm at Changi | &ever Singapore entity set up in 2020 under the 30x30 Express grant | Acquired by Kalera for US$153 million in 2021; Kalera itself filed Chapter 11 in 2023 and sold the Singapore farm to Growy | Officially opened Nov 2024 under its third owner, entered provisional liquidation Nov 2025 |
| Singrow | Vegetables/seed genetics, hydroponic strawberries | 2019, an NUS spin-off | S$1.6 million seed round (2020), plus a US$4.5 million Series A (2025) | Still operating, expanding into China and other markets |
| GroGrace (Urban Farming Partners Singapore) | Vegetables, dry hydroponics with Dutch horticulture tech | Opened 2022 | Backed in part by SFA's Agriculture Productivity Fund | Announced on social media on 8 February 2026 it would fold by the 14th; founder Grace Lim later extended that to end of month while seeking a rescue partner, after the company had been defaulting on rent and utility payments |
| Umami Bioworks | Cultivated seafood (alternative protein) | Founded after Shiok Meats, absorbed it in the 2024 merger | Not fully disclosed | Operating as the combined entity post-merger |
| Eat Just | Cultivated meat | Singapore operations established during the 30 by 30 period | Not disclosed for the Singapore operation specifically | Suspended cultivated meat production in Singapore |
| Avant Protein | Cultivated protein | Founding date not independently confirmed | Not disclosed | Wound up its Singapore operations in early 2026, one of at least three cultivated-protein firms to scale back here |
| Artisan Green(s) | Vegetables | Founding date not independently confirmed | Not disclosed | Reported profitable since early 2022, one of the few in the category |
If you've come across other names that aren't listed, that's very likely because this is a subset rather than a full census. There's no single public registry of every farm that has opened, pivoted, or closed since 2019, so this list leans on what's been reported in the press, disclosed in grant records, or filed with ACRA, which understates smaller or quieter closures that never made the news. The cultivated meat and seafood names are worth including as their own cluster too, since they sit outside the vegetable and seafood farming discussed above but were explicitly part of the "30 by 30" alternative protein ambition Grace Fu referenced when she said the segment had been held back by high costs and, separately, by consumer acceptance that hasn't kept pace globally. It's the same demand-and-financing combination as the rest of the table, just wearing a biotech label instead of a hydroponics one.
A few things stand out once the roster is laid out this way rather than told as a handful of anecdotes.
The Growy/Changi story is the clearest single illustration of the point this piece is making. That farm has now been designed, built, and shut down under three different owners in Germany, the US and the Netherlands, with a Nasdaq-listed bankruptcy in between, and it still couldn't survive a year of actual operation once it finally opened in late 2024. If deep-pocketed multinational ownership and a change of parent company each time things went wrong couldn't fix it, the fault sits with the Singapore-specific economics of the site, not with any one operator's competence.
GroGrace's own trajectory closes out this pattern rather than breaking it. The farm posted on social media in February 2026 that it would fold within the week, then bought itself a few more weeks while its founder looked for a rescue partner, after falling behind on rent and utilities. By May 2026, a competing farm supplier was describing a wider run of holds, closures and downscaling across the Singapore indoor farming sector that year, and framing it as a broader mismatch between the operating model most of these farms were built on and what the market could actually support, not a problem unique to Singapore, but one playing out in Europe and the US too. That's the same story this piece keeps returning to, just with a fresher date on it.
Temasek's involvement across both Sustenir and Apollo Aquaculture tells a similar story from a different angle. This is Singapore's own sovereign wealth fund, with the deepest pockets and the longest time horizon of any investor on this list, backing two ventures in two different categories, and neither reached a stable, standalone profit without a rescue or a restructuring. Money clearly wasn't the constraint.
And the survivors don't cluster around a single obvious explanation either. Singrow survives partly by pivoting toward being a seed and genetics company that licenses technology elsewhere, rather than a pure grower competing on retail price. Artisan Green's profitability is the exception serious enough to be worth its own separate reporting rather than the rule. ComCrop has stayed open longest by keeping its footprint modest and hybrid rather than betting on a mega-farm. None of these are proof that a particular business model cracked the code. They're closer to evidence that survival has depended on avoiding exposure to the same demand and financing risk that sank almost everyone who scaled up quickly.
The pre/post split also sharpens the point made earlier about the pig farming precedent. The "predates 30 by 30" table is mostly small, slow-growing, or already-proven businesses that scaled gradually and organically, on cost structures they'd already worked out before any target existed. The "launched to chase it" table is almost entirely large, fast, capital-intensive bets, made in a compressed window because a 2030 deadline demanded speed rather than the kind of patient, decades-long build that got egg farming and the old soil-based vegetable farms to where they are. Singapore's food system has generally rewarded the first pattern and punished the second, which is itself a data point about how much the financing model, not the technology or the ambition, decided who made it.
The one segment that hit its number, and why it's the exception that proves it
Hen eggs are the one part of "30 by 30" that actually worked. Local production reached roughly 30 to 34% of consumption by 2024, and beansprouts got past 50%. SFA credits automation and stronger farm management at players like Chew's Agriculture and N&N Agriculture, which is true, but it undersells how different the starting conditions were for eggs versus everything else.
An egg doesn't need climate augmentation. A hen in a Singapore shed and a hen in a temperate country produce the same product, at the same quality, for the same use. Nobody has to be talked into believing a local egg tastes different, because it doesn't. The category was also already concentrated among a small number of established players who'd long since paid off their infrastructure, competing on efficiency rather than starting from zero.
Vegetables and seafood sit on the wrong side of that line on both counts, which lines up with what you'd already suspected going into this. Much of what Singapore households actually want to eat, temperate leafy greens, certain fruiting vegetables, doesn't grow here without the climate being manufactured: chillers, dehumidifiers, artificial lighting running around the clock, drawn from a tropical grid that isn't cheap. And once that climate-augmented produce lands on a shelf next to its imported equivalent, it typically costs 20 to 50% more, based on retail comparisons that have been reported locally, for a product a meaningful share of shoppers say tastes and looks worse, not better. Willingness-to-pay research on Singapore consumers backs this up too. The block isn't mainly income or headline price. It's that shoppers aren't convinced the local product earns its premium on freshness or quality, which are the two things it would need to win on to justify costing more.
Put a demand problem like that on top of a cost problem, and you get exactly the closures listed above. Ventures that were financially unsustainable almost from day one, not because any one of them was badly run, but because the whole category was set up to lose money at pretty much any scale.
The cost problem is a financing design, not a technology problem
This is where the second half of the structural issue comes in, and it's worth being precise about the mechanism, because "capex is high" undersells what's actually going on.
Singapore allocates farmland through competitive tenders. Under SFA's system, an operator bids on land price, either through a Fixed Price Tender, where the plot goes to whoever proposes the best production use at a set price, or a Concept and Price Tender, where a concept is judged first and the tendered sale price decides the winner among the concepts that pass. Either way, the bidding is for the land itself, upfront, before a single leaf has grown. The winning operator then carries the full risk of whether the market ever pays enough to earn that back, on top of the equipment, the electricity bill, and the climate-control system needed to grow something Singapore's climate won't grow on its own. Nobody on the other end of that deal has promised to buy what comes out of the farm.
Compare that with how Singapore built water security, which is a comparison worth taking seriously rather than treating as a slogan. PUB's desalination and NEWater plants are built under Design-Build-Own-Operate contracts. A private consortium designs, builds, owns and runs the plant, but the competitive tender isn't for the land or the asset price at all. It's for the tariff. Bidders compete on cents per cubic metre of water delivered, and PUB signs a Water Purchase Agreement locking in that price for 25 years, guaranteeing to buy the plant's output for the life of the contract. Singapore's fifth desalination plant was awarded at S$0.91 per cubic metre in 2017 on that basis. Its predecessor at Marina East locked in S$1.07867 for a 25-year term. The operator still takes construction risk and operating risk. It doesn't take demand risk, because the demand was contracted away on day one.
That's really the whole difference. Water security in Singapore never asks a private operator to also bet on whether Singaporeans will voluntarily choose desalinated water over some cheaper alternative at the tap. Food security, as it's currently financed, asks exactly that of every vegetable and seafood farm: win a land auction, sink capital into climate control, and then hope households choose your kale over the cheaper imported kale at the checkout. One model spreads demand risk out through a guaranteed offtake contract. The other dumps all of it onto the smallest, most capital-constrained players in the whole system.
The surplus has to go somewhere, and it isn't a shelf
There's a downstream consequence of that demand mismatch that's worth sitting with for a moment, because it shows up as cost twice over, not once.
A farm can't turn crops off the way a factory can pause a production line. Rent, energy and labour keep running on schedule whether or not a buyer is confirmed, so vegetables get harvested on the farm's timetable, not the market's. When retail demand softens, or a batch doesn't meet a supermarket's cosmetic standard, the produce doesn't quietly disappear from the cost sheet. It becomes waste that the farm already paid full climate-controlled production cost to grow.
At ComCrop, a rooftop farm in Woodlands, around 30% of produce was going to charity donation as of a few years back, down from about half the harvest the year before, according to reporting on the farm's own numbers. That's an improvement, but it's still close to a third of everything grown finding no buyer at all. Zoom out to the wholesale level and the picture is worse: at Pasir Panjang Wholesale Centre, sellers reportedly discard roughly a third of the vegetables that pass through on cosmetic grounds alone, and volunteer groups like SG Food Rescue can only manage to redirect about 1.5 tonnes of that a week. A joint study by the Singapore Environment Council and Deloitte put food loss at the local production stage alone at over 5,000 tonnes out of an estimated 150,000 tonnes lost across the whole food system.
None of that waste is random. It's the same demand problem described above, just showing up a second time, after the money has already been spent. A farm that can't sell what it grows at the premium it needs isn't just losing potential revenue. It's paying to grow food that then gets composted or given away, which makes the underlying unit economics even harder to fix than the price premium alone would suggest.
That's really a separate design problem in its own right, the question of why a meaningful share of what Singapore does manage to grow locally never reaches a plate, and what a supply chain built around that reality would need to look like differently. It's the kind of gap that deserves its own piece, the same way the blue bin did, and it's the natural next stop after this one.
The government is already reaching for this answer, just in different language
To be fair to SFA and MSE, the November 2025 announcement wasn't only a target reset. Grace Fu also said the government is studying a pilot multi-tenanted agrifood production facility, government-owned to start, offering farms plug-and-play space, shared utilities, and controlled-environment production, all meant to lower the start-up capital a farm needs and cut the business risk operators are carrying. Separately, SFA is redeveloping the Lim Chu Kang farming belt into what it's calling a high-tech agri-food cluster, with shared wastewater treatment, logistics and distribution infrastructure sited centrally instead of duplicated farm by farm.
Held up against the NEWater comparison, this is clearly a government-builds-shared-infrastructure model, arriving at more or less the same instinct that kicked off this piece. It just hasn't gone all the way there yet. A shared building with cheaper rent still leaves the demand side of the problem untouched. Farms in a multi-tenanted facility still have to sell climate-augmented vegetables to consumers who, on the evidence so far, aren't reliably willing to pay more for them. Lower capex solves half of what you flagged at the start. It doesn't solve the other half.
What closing that second half would actually take
A full NEWater-style model for food would need the piece current planning hasn't announced yet: something that functions like a Produce Purchase Agreement, not just a cheaper building to grow food in. That doesn't require Singapore to price-control a supermarket shelf. It could look like a public institutional offtake channel, schools, hospitals, SAF messes, town council contracts, guaranteeing a floor volume of local vegetables and seafood at an agreed price for the length of a farm's lease, the same way PUB's 25-year tariff takes the guesswork out of a desalination operator's business case. Farms would still compete, the way desalination consortiums do, on the price and efficiency of what they deliver. Just not on whether anyone will buy it at all.
Pair that with land tenders that price-compete on production cost per kilogram delivered into that offtake channel, rather than on the land parcel itself, and the risk moves to where Singapore has already shown it's willing to hold it for water: on the books of the public agency securing the resource, not on the books of whichever farm just won an auction. It's also a more direct answer for the shrimp farm that worked and left anyway than any grant, masterplan or shared facility could be on its own, because that farm's problem was never its production model. Nothing in the system was underwriting its market the way PUB underwrites a desalination plant's.
None of this means scrapping the multi-tenanted facility or the Lim Chu Kang cluster. Both go after the capex side of the same equation you started with, and both are worth finishing. What's missing is the other half of the NEWater model, the part that actually made it work: a government willing to be the guaranteed buyer, not just the landlord.