Underground Vertical Farms: LED Lighting Loads, Closed-Loop Water, and Payback Economics

Repurposing unused underground space for food production

A growing category of urban agriculture projects converts disused infrastructure — decommissioned rail tunnels, Cold War-era air raid shelters, unused parking structures, and mine workings — into controlled-environment farms. The appeal of going underground specifically, rather than building a conventional greenhouse, is thermal stability: underground spaces maintain a relatively constant year-round temperature close to the local mean annual air temperature, which cuts the heating and cooling load that dominates energy costs in surface greenhouses in cold or hot climates. The trade-off is that there is no natural daylight at all, so 100% of the light energy crops need has to come from artificial lighting, making LED energy the single largest recurring operating cost in most underground farm business models.

Sizing the LED lighting load

Plant growth lighting is specified in terms of photosynthetically active radiation (PAR) delivered as photosynthetic photon flux density, measured in micromoles of photons per square metre per second (µmol/m²/s) rather than in the lux or lumens used for human-vision lighting, because plants respond to photon count in specific wavelength bands, not perceived brightness. Leafy greens and microgreens — the crop types best suited to underground farms because they tolerate lower light intensity and short growth cycles — typically target 150-250 µmol/m²/s for 14-18 hours per day, a combination sometimes called the 'daily light integral.'

Converting a target PPFD to electrical power draw depends on the LED fixture's photon efficacy (µmol of photons produced per joule of electricity, with modern horticultural LEDs reaching roughly 2.5-3.0 µmol/J), giving a power density in the vicinity of 0.7-0.9 W/m² per µmol/m²/s of target intensity for reasonably efficient fixtures. For a 500 m² growing area targeting 180 µmol/m²/s for 16 hours/day, this works out to roughly 65-90 kW of installed lighting power and 1,000-1,400 kWh of energy per full lighting cycle — a substantial load that makes electricity tariff structure (off-peak rates, on-site renewables, waste-heat recovery from the LEDs themselves for space heating) a first-order factor in underground farm site selection and design, not a secondary consideration.

Closing the water and nutrient loop

Underground hydroponic and aeroponic systems use dramatically less water than field agriculture for the same crop mass — commonly cited figures for leafy greens are under 1 litre of water per kilogram of produce in a well-run recirculating system, versus tens to hundreds of litres per kilogram for open-field cultivation, because a closed system captures and reuses water that would otherwise be lost to soil infiltration and evaporation. Recirculation and dehumidification recovery — condensing the water vapour that transpiring plants release into the growing-room air, which is unavoidable in a sealed environment — can recover a majority of the water a crop cycle consumes, with recovery rates of 60-70% being a realistic operational target rather than a theoretical maximum.

Nutrient dosing follows a similar closed-loop logic: rather than soil-based fertilization, hydroponic systems dose a precise nutrient solution (nitrogen, phosphorus, potassium, and micronutrients) directly into the water stream, typically in the range of 0.1-0.15 kg of nutrient mix per kilogram of harvested crop for leafy greens, with sensors continuously monitoring electrical conductivity and pH to catch dosing drift before it affects yield. This precision is a genuine agronomic advantage over field farming, where nutrient runoff into waterways is a significant environmental cost that closed-loop systems largely eliminate.

The economics: high CAPEX, tight margins, long payback

Underground vertical farms carry substantially higher capital cost per unit of growing area than field agriculture or even conventional greenhouses, driven by lighting infrastructure, climate control, structural retrofitting of the underground space, and automation systems. A representative mid-size facility might carry CAPEX in the high hundreds of thousands to low millions of dollars for a few hundred square metres of growing area.

Payback period is calculated as CAPEX divided by annual profit, where annual profit is annual production volume × (revenue per kilogram − cost per kilogram). For an operation producing 18,000 kg/year at $4.10/kg revenue against $2.40/kg production cost, annual profit is 18,000 × $1.70 ≈ $30,600 — which against a $950,000 CAPEX gives a payback period of roughly 31 years, an uneconomic result that illustrates a real tension in this sector: the per-kilogram margins on commodity leafy greens are thin, and payback only becomes attractive at either a larger production scale (spreading fixed CAPEX over more output), a premium price point (specialty greens, herbs, or direct-to-restaurant sales rather than commodity produce), or with public subsidy support for the social value of local food security and reduced transport emissions, which several municipal sustainability programs now provide.

Where underground farming makes the most sense

Underground vertical farming is not a wholesale replacement for field agriculture — the energy cost of 100% artificial lighting means it will rarely be the cheapest way to grow bulk commodity crops. Its strongest use case is hyper-local production of high-value, short-shelf-life crops (microgreens, herbs, specialty lettuces) close to urban consumers, cutting transport time and cold-chain losses while reusing otherwise-derelict infrastructure that would cost money to demolish or maintain vacant anyway. Cities piloting these facilities typically frame them as a food-security and urban-resilience investment — a hedge against supply-chain disruption for fresh produce — as much as a standalone commercial venture, which is why many projects combine private operating revenue with public sustainability funding rather than expecting either to carry the project alone.