kWh per Kilogram: Lighting Cost Math for Indoor Farms
A six-step model that turns DLI targets, LED efficacy, tariffs and yield into a defensible cost per kilogram.
Why “watts saved” is the wrong unit
Most lighting comparisons in controlled environment agriculture collapse into a single claim: this fixture draws fewer watts, so it costs less to run. That statement is true and almost useless. A watt is an instantaneous power draw, not a delivered dose of light. It says nothing about how much of that light reaches the canopy, how many hours the fixtures run, or how much marketable biomass those photons actually produce. Two fixtures with identical power ratings can differ by a third in photon output, and a fixture that draws more watts can still deliver the same daily light integral for less energy.
The number that survives contact with a spreadsheet is kWh per kilogram of produce. It folds six decisions — crop light target, photoperiod, fixture specification, system efficacy, electricity tariff, and yield per square metre — into one figure that can be compared across facilities, crops, and markets. Better still, it tells you which lever actually moves your costs. The six steps below build that number from scratch.
The model in six steps
Step 1 — Set the crop’s daily light integral (DLI) target
Formula: DLI target (mol/m²/day) — your chosen set point, from crop-specific guidance rather than from a fixture datasheet.
Extension guidance is the right starting point because DLI requirements vary enormously by crop. Virginia Cooperative Extension’s DLI guide lists 12–17 mol/m²/day for lettuce, 14–20 for spinach, and 20–30 for tomato and cucumber, noting that fruiting crops can need nearly double or triple the light of a leafy green.
Example: lettuce at 14 mol/m²/day, the midpoint of the recommended range.
Step 2 — Convert DLI into a photon demand rate
DLI is a daily dose; fixtures are specified by the second. Converting one to the other is where photoperiod enters the model.
Formula: PPFD (µmol/m²/s) = DLI × 1,000,000 ÷ (3,600 × photoperiod hours)
This is the standard inversion of the DLI equation published by Virginia Cooperative Extension: DLI in mol/m²/day is PPFD multiplied by 3,600 and the number of lit hours, divided by one million.
Example: 14 × 1,000,000 ÷ (3,600 × 16) = 243 µmol/m²/s across a 16-hour photoperiod.
Step 3 — Convert photon demand into fixture count
Formulas:
Total photon flux required (µmol/s) = PPFD × illuminated area (m²)
Fixture count = total photon flux ÷ fixture PPF (µmol/s)
Note the unit: PPF is what a fixture emits per second; PPFD is what arrives per square metre. Fixture PPF is normally quoted at a defined distance and drive current, so compare nameplate figures only at the same conditions. Koray (Zhongshan Koray Opto-Electronic) has manufactured horticultural LED lighting for 17 years and holds more than 300 patents in China and abroad; its vertical farming fixtures publish PPF alongside input power, which is the pair of numbers this step needs.
Example: 243 × 1,000 = 243,000 µmol/s. With an illustrative fixture rated at 900 µmol/s, 243,000 ÷ 900 = 270 fixtures.
Step 4 — Convert fixture count into electrical energy
Formula: kWh/day = total photon flux (µmol/s) × lit seconds ÷ efficacy (µmol/J) ÷ 3,600,000
Combining Steps 2 to 4 gives a shortcut worth memorising, because it removes the fixture count entirely:
kWh/day = (DLI × illuminated area) ÷ (3.6 × efficacy in µmol/J)
Example: 243,000 × 57,600 = 1.40 × 10¹⁰ µmol/day. At 2.9 µmol/J that is 4.83 × 10⁹ J, or 1,341 kWh/day. Cross-check: (14 × 1,000) ÷ (3.6 × 2.9) = 1,341 kWh/day.
A note on losses. Nameplate efficacy is usually measured at the LED array or driver output, not at the wall. Driver conversion (roughly 90%) and optical losses (roughly 92–95% transmission through secondary optics and covers) mean system-level efficacy can land 12–18% below the datasheet. A second, larger loss is light that never reaches the canopy: simulation work on indoor lettuce designs found photon capture between 50% and 85% depending on fixture geometry and bed layout, and photons that land on aisles and walkways cost exactly the same as photons that land on leaves.
Step 5 — Convert kWh into cost
Formula: Annual lighting cost = kWh/day × tariff ($/kWh) × operating days
Example: at US$0.12/kWh, 1,341 × 0.12 = $160.90/day, or $58,736 per year across 365 days.
Tariff is the variable most outside a grower’s control, and it is large. Ohio State’s controlled environment cost page models supplemental lighting at $0.09/kWh; the University of Georgia’s lighting cost calculator case study uses $0.12/kWh; grids with high industrial rates run above $0.30/kWh.
Step 6 — Divide by production
Formulas:
kg/m²/year = yield per m² per cycle × cycles per year
Lighting cost per kg = annual lighting cost ÷ (kg/m²/year × illuminated area)
Get Horticulture Growlights’s stories in your inbox
Join Medium for free to get updates from this writer.
Example: 4.0 kg/m² per cycle × 8 cycles = 32 kg/m²/year. Across 1,000 m² that is 32,000 kg, so $58,736 ÷ 32,000 = $1.84 per kg.
Worked example, end to end: indoor leaf lettuce
Assumptions used in this example (illustrative, not measured facility data):
The chain: DLI 14 → PPFD 243 µmol/m²/s → 243,000 µmol/s total → 270 fixtures → 1,341 kWh/day → $160.90/day → $58,736/year → 32 kg/m²/year → 15.3 kWh and $1.84 of lighting electricity per kilogram.
Two sanity checks are worth running on that result. First, the per-mole cost: 14,000 mol/day of PAR for $160.90 works out to just under $0.01 per mole, which matches the benchmark published by Ohio State. Second, the yield assumption is conservative. A 2025 benchmarking study of indoor vertical farming reports modelled productivity spanning roughly 55 to 102 kg/m²/year. Lift this example from 32 to 60 kg/m²/year at the same light level and the lighting cost falls to $0.98/kg and 8.2 kWh/kg, with no change to the lighting system at all.
What efficacy and tariff do to the number
The table below holds every input from the worked example constant — 14 mol/m²/day, 1,000 m², 16 h photoperiod, 32 kg/m²/year — and varies only fixture efficacy and tariff. Efficacy levels are chosen to bracket commercially available hardware. The figures are model outputs from the illustrative assumptions above, not measured data.
In numbers, the model returns 17.1, 15.3, 13.4 and 12.3 kWh of lighting electricity per kilogram across the four efficacy levels — a figure that does not depend on the tariff — and $1.71, $1.53, $1.34 and $1.23 per kilogram at $0.10/kWh, against $5.12, $4.59, $4.03 and $3.70 at $0.30/kWh.
Moving from 2.6 to 3.6 µmol/J cuts energy by about 28% and cost per kilogram by the same proportion. Moving the tariff from $0.10 to $0.30 multiplies cost per kilogram by three, which is a far larger swing than any efficacy upgrade on the market today can deliver.
Which variables move the result most
1. Tariff. Perfectly linear, no ceiling, and easily a three-fold difference between markets. Nothing else in the model has this much leverage.
2. Yield per square metre per year. It sits in the denominator, so a 30% yield gain is a 23% reduction in cost per kilogram. The benchmarked range of 55–102 kg/m²/year versus a conservative 32 kg/m²/year is a two-fold swing in the final number, and it costs nothing in electricity.
3. System efficacy. Worth roughly 28% between a mid-range and a premium fixture, and it is bounded by physics: red LED conversion efficiency sets a practical ceiling, and each further µmol/J gets progressively more expensive to buy. Improving photon capture through layout and optics is often cheaper than buying efficacy.
4. DLI target. Energy scales almost linearly with the set point — 12 to 17 mol/m²/day is a 42% energy increase. But the effect on cost per kilogram is smaller than that, because higher DLI usually buys faster growth and higher yield. The honest version of this lever is “cost per kg at each DLI,” never “energy per m².”
5. Photoperiod. At a fixed DLI, the total photons delivered are identical, so energy per kilogram barely moves. What changes is fixture count and instantaneous PPFD, so photoperiod is primarily a capital-cost and uniformity decision. It becomes an operating-cost decision only with time-of-use tariffs, where light-hours can be shifted into cheaper windows.
FAQ
How many kWh does a typical indoor lettuce crop use per kilogram?
For lighting alone, the model above returns roughly 12–17 kWh per kilogram across the 2.6–3.6 µmol/J range, at a conservative yield of 32 kg/m²/year. Published whole-facility figures bracket that: a 2025 vertical farming benchmarking study reports specific energy consumption of 3.2–20 kWh/kg across the literature, while a 2026 peer-reviewed lettuce study cites 7–10 kWh per kilogram of lettuce, with lighting responsible for 65–85% of total energy use. The spread is driven mostly by yield and climate control, not by lamps.
Does a higher-efficacy fixture always win on cost?
No. Efficacy reduces running cost roughly proportionally, but it is purchased with capital, and Missouri Extension’s guide to CEA grow lights notes that LEDs carry a higher upfront investment that is recovered over a longer life. At a $0.09/kWh tariff the gap between a 2.9 and a 3.6 µmol/J fixture is about $0.30 per kilogram, so a large capex premium can take years to repay in a low-cost electricity market. Check whether the quoted efficacy is measured at the wall plug or at the LED array, and whether it is maintained at the end of life.
How do DLI targets differ between leafy greens and fruiting crops?
Substantially. Virginia Cooperative Extension lists 12–17 mol/m²/day for lettuce, 14–20 for spinach, and 20–30 for tomato and cucumber, and explains the gap as the extra energy fruit production requires. In practice this means a fruiting crop under full artificial light needs roughly double the fixtures and double the electricity per square metre of a leafy green — before any adjustment for taller canopies and lower planting density.
What is the biggest hidden cost after electricity?
Heat removal and dehumidification. Every joule delivered to a sealed grow room eventually becomes latent or sensible heat that the HVACD system must move, and the 2025 benchmarking study attributes roughly 28% of vertical farm energy use to HVAC and dehumidification, against about 70% for lighting. On the non-energy side, Ohio State’s strawberry cost model shows substrate and integrated pest management each running to a meaningful share of operating cost, with labour typically the largest single line item of all.
Closing
kWh per kilogram is not a perfect metric. It ignores capital, labour, packaging, and the price you actually sell at, and its inputs are estimates until you measure them in your own facility. What it does well is refuse to let a lighting decision hide behind watts. Run the six steps with your tariff, your tariff structure, your measured PPF at canopy height, and your own harvest weights. Then run it again with a DLI one mole higher and a yield 20% better, and see which one changes the answer. The comparison that matters is not between two fixtures on a datasheet. It is between two versions of your own cost per kilogram.
Sources
- Virginia Cooperative Extension, Calculating and Using Daily Light Integral (DLI): An Introductory Guide — https://pubs.ext.vt.edu/SPES/spes-720/spes-720.html
- University of Missouri Extension, Controlled Environment Agriculture: Understanding Grow Lights (G6987) — https://extension.missouri.edu/publications/g6987
- Ceccanti, F. et al., Beyond Yield: Integrating Energy, Water, Cost, and Carbon to Benchmark Indoor Vertical Farming Viability, arXiv:2507.18419 — https://arxiv.org/abs/2507.18419
- Palmitessa, O.D. et al., Continuous lighting at low PPFD improves energy efficiency while preserving growth and quality of lettuce in vertical farming systems, Frontiers in Plant Science, 2026 — https://pmc.ncbi.nlm.nih.gov/articles/PMC12995675/
- Lozano-Castellanos, L.F. et al., Light Energy Efficiency in Lettuce Crop: Structural Indoor Designs Simulation, Plants, 2023 — https://pmc.ncbi.nlm.nih.gov/articles/PMC10574718/
- The Ohio State University, Controlled Environment Agriculture Center, Costs — https://ohceac.osu.edu/CEBPI-Costs
- Kim, C. and van Iersel, M., A New Way to Determine Lighting Costs at Your Greenhouse, Greenhouse Grower, 2021 — https://www.greenhousegrower.com/technology/a-new-way-to-determine-lighting-costs-at-your-greenhouse/
- U.S. Department of Energy, Integrated Lighting Campaign, Unlocking Plant Growth Potential with LEDs in Indoor Horticulture — https://integratedlightingcampaign.energy.gov/unlocking-plant-growth-potential-leds-indoor-horticulture
