NovaSolver›Agricultural Greenhouse Energy Balance Simulator Back
Agricultural Engineering
Agricultural Greenhouse Energy Balance Simulator
A tool for assembling the heat balance of an agricultural greenhouse. Change the floor area, cover material, outdoor temperature, solar radiation, ventilation rate and crop transpiration to see solar gain, conduction loss, ventilation loss and transpiration cooling — along with the net heating/cooling load and annual heating energy — update in real time.
Sun rays (yellow) pass through the cover, conduction heat (red) escapes outward through the skin, ventilation (blue arrows) leaves through the roof vent, and the crop canopy (green) cools the air by transpiration. Colour reflects the net load sign (red=heating, blue=cooling).
Heat balance — gains vs losses
Heating load vs outdoor temperature
Theory & Key Formulas
$$Q_{net} = Q_{cond} + Q_{vent} - Q_{solar} + Q_{transp},\quad Q_{vent} = 0.34 \cdot V \cdot n \cdot \Delta T$$
V = room volume (m³), n = air changes per hour (1/h), ΔT = indoor-outdoor temperature difference (K), U = cover overall heat transfer coefficient (W/m²K). Positive Q_net is heating load, negative is cooling load.
$$Q_{cond} = U \cdot A_{cover} \cdot \Delta T, \qquad Q_{solar} = I \cdot A \cdot \tau$$
Conduction loss and solar gain. A_cover is the cover surface area (approximated as floor area × 2.5), I is solar irradiance (W/m²), τ is the effective transmissivity.
Latent cooling from transpiration (q_tr in W/m²) and an order-of-magnitude annual heating energy (MWh) integrating 90 winter days. Adjust the coefficient to local climate.
Greenhouse energy balance — solar, transmission, transpiration and HVAC load
🙋
A greenhouse is just a box made of glass or plastic that traps sunlight, right? Is the sun really enough to keep it warm in winter?
🎓
That's the picture most people have, but in a real production greenhouse the balance between "heat in from the sun", "heat lost through the cover" and "heat lost by ventilation" is what really matters. On a winter morning with outdoor 0–5 °C and an indoor setpoint of 20 °C, the cover U-value dominates the heating load. Try sliding the outdoor temperature from 5 °C down to -5 °C on the left and watch the conduction loss jump.
🙋
Wow, the conduction and ventilation losses really climb. And switching to "double poly" almost halves them.
🎓
Exactly — that's the appeal of double-skin (air-inflated) covers. Glass and single poly sit at U = 6–8 W/m²K, double poly drops to 4.5, twin-wall acrylic to 4.0. Cold-climate Dutch-style and Japanese tomato/strawberry growers combine double covers with thermal screens (a foil curtain pulled across at night) to chop the heating bill by 30–40%. The trade-off is a small drop in transmissivity, so it's a winter-yield vs heating-cost decision.
🙋
I'm surprised "transpiration cooling" lands on the cooling side. So crops are not a heat source — they actually cool the air?
🎓
Right. A mature tomato or cucumber canopy on a clear day absorbs the equivalent of 100–200 W/m² as latent heat through transpiration. Per unit volume that's not trivial — roughly half of the incoming solar can be cancelled out. So daytime cooling in a summer greenhouse is a team play: vents wide open, shade screen, fogging (high-pressure mist) and the crop's own transpiration. At night when transpiration drops, the humidity/dehumidification balance changes too.
🙋
So "annual heating MWh" — what does that actually cost in money for a small grower?
🎓
At the default (1000 m² glass, 90-day winter) you're looking at about 160 MWh. In heavy fuel oil that's roughly 16 kL, around US$10–15k per season, with 40+ tonnes of CO₂. That's why government schemes in Japan and the EU subsidise "heat pump + double curtain + biomass" packages that halve the heating bill. NASA's Veggie and Advanced Plant Habitat (space), Plantagon and Mirai's vertical farms (urban), and the Dutch Westland glasshouse cluster are all running the same kind of optimisation between light, heat and CO₂.
FAQ
The net heat load is Q_net = Q_cond + Q_vent - Q_solar + Q_transp. Q_cond = U·A_cover·ΔT is the conduction loss through the cover, Q_vent = 0.34·V·n·ΔT is the ventilation loss (V: room volume m³, n: air changes per hour), Q_solar = I·A·τ is the solar gain, and Q_transp is the latent cooling from crop transpiration. A positive Q_net means heating is required, a negative value means cooling/ventilation. This tool computes each term in real time from floor area, cover material, outdoor temperature, solar radiation, ventilation rate and transpiration load.
Typical U-values are: glass 6.0 W/m²K, single-layer polyethylene 8.0, double polyethylene 4.5, and acrylic 4.0. Going from single to double poly cuts conduction loss by about 44%, and the heating bill drops in roughly the same proportion. The trade-off is transmissivity (τ): glass ≈ 0.85, double poly ≈ 0.70, so winter solar gain falls slightly. In cold climates double-skin air-inflated houses with thermal screens are the norm; in mild climates a single poly skin maximises light intake.
Minimum winter ventilation for CO₂ and humidity control is 0.5–1.0 air changes per hour, while summer peaks reach 30–60 ACH. Ventilation loss scales linearly with n, so during heating you keep n as low as possible; in summer all roof and side vents go fully open to dump heat by outdoor air. This tool covers 0.1–30 1/h: raising n increases the ventilation loss (kW) and shifts the net load. If you do CO₂ enrichment, more ventilation flushes the CO₂ out, so dehumidifying heat pumps are increasingly used to keep n low.
Crops release water vapour through transpiration, and the latent heat of vaporisation (~2,450 kJ/kg) is absorbed from the surrounding air, dropping the room temperature. A mature tomato or cucumber canopy on a sunny day absorbs the equivalent of 80–200 W/m². Combined with high-pressure fogging (pad-and-fan), this can suppress summer peak temperatures by 5–10 K. Of course humidity rises too, so the balance between transpiration cooling, ventilation and dehumidification is the core of greenhouse HVAC design.
Real-world applications
Large-scale production greenhouses (Dutch-style, Japanese): 1–10 ha glass and double-poly houses growing tomato, sweet pepper, cucumber and strawberry year-round. Heat balance models like this tool are used at design time to size the heating boiler, decide how many thermal screens to fit and how much vent area to provide; at the operations stage the same model — driven by tomorrow's weather forecast — predicts the next-day fuel consumption. Dutch Westland tomato growers and Hokkaido/Aichi tomato houses in Japan all operate on this logic.
Plant factories and vertical farms (NASA Veggie / Plantagon / Mirai): Fully closed plant factories replace the sun with LEDs supplying photosynthetically active radiation, while temperature, humidity and CO₂ are tightly controlled. About 70% of the LED electricity ends up as heat, so re-mapping this tool's "solar gain" to LED input power gives you a first cut at HVAC sizing and energy use. NASA's Veggie unit on the ISS, Sweden's Plantagon and Japan's Mirai lettuce factories are well-known examples.
FAO and JICA greenhouse projects in developing countries: FAO promotes simple net-houses and low-tech greenhouses for arid and tropical regions. The baseline design is evaporative cooling (pad-and-fan), shade nets and natural ventilation, holding summer temperatures 5–10 K below ambient. Running this tool's "solar → transpiration → ventilation" chain with regional parameters lets you quickly compare greenhouse specs against local cost and electricity constraints.
Carbon-neutral agriculture assessment: Agricultural greenhouses are a large share of farm fuel consumption, and visualising A-heavy oil, LPG and electricity usage together with CO₂ emissions is now standard practice. Multiplying this tool's "annual heating MWh" by a fuel emission factor gives a quick CO₂ estimate, useful for sizing heat-pump retrofits, biomass boilers and ground-source loops. Government smart-agriculture pilots in Japan use the same kind of model as a screening tool.
Common pitfalls and caveats
The biggest trap is to over-estimate solar gain. The Q_solar = I·A·τ in this tool uses a single, constant transmissivity. Real greenhouses see (1) higher reflection at low sun angles morning and evening, (2) shading from neighbouring houses and structural members, (3) τ falling 10–20% from condensation and dust, and (4) hour-by-hour swings in I from clouds and snow. As a design rule, take the peak solar value at about 70% as the "effective" gain to stay on the safe side. Expecting "the sun will take care of it" on a winter morning leads to heating lag and chilling injury to the crop.
Second, do not assume that setting ventilation n to zero makes heating free. The tool will indeed show ventilation loss disappearing, but in a real house you immediately get (a) leaf condensation and grey mould, (b) CO₂ drawdown to ~200 ppm collapsing photosynthesis, and (c) humidity high enough to stop transpiration and starve the plant of nutrients. A minimum of 0.5–1.0 ACH is non-negotiable — you pay fuel for it because the crop needs it. Modern designs increasingly use dehumidifying heat pumps to hold n low while still managing moisture.
Finally, do not convert the annual MWh directly into your electricity bill. This tool's annual heating is the upper bound from running the average load 24/7 for 90 days. In practice (i) daytime solar shuts heating off, (ii) thermal screens cut night losses by 30–40%, and (iii) boiler/heat-pump efficiency (COP) ranges 0.9–4.0, so actual fuel use is typically 50–70% of the calculated value. On the other hand, a cold-wave morning can spike the instantaneous load to 2–3× the rated value, so size the boiler for peak load and pin down annual consumption with a proper time-step simulation (DesignBuilder, Hortinergy, etc.). Use this tool's numbers for first-pass comparison only, and re-evaluate with regional weather data before committing.
How to Use
Enter floor area in m² (typical range 100–500 m² for commercial operations) and select cover material (polycarbonate, glass, or polyethylene film) from the dropdown
Set outdoor temperature in °C, indoor setpoint in °C (typically 18–22°C for vegetables, 25–28°C for tropical crops), and solar radiation in W/m² based on season and latitude
The simulator calculates four energy flows: solar gain through the cover, conduction losses through walls and roof, ventilation losses through air exchange, and transpiration cooling from plant evapotranspiration, then outputs net heat load and annual heating demand in MWh
Worked Example
A 250 m² tomato greenhouse in southern Spain uses polycarbonate twin-wall cover (U-value 3.8 W/m²K). With outdoor temperature 8°C, indoor setpoint 20°C, and solar radiation 650 W/m² on a winter morning: solar gain = 156 kW, conduction loss = 23 kW, ventilation loss = 8 kW, transpiration cooling = 12 kW. Net heat load = 23 – 156 + 8 + 12 = −113 kW (surplus). Annual heating demand for 4,200 degree-days ≈ 180 MWh natural gas equivalent.
Practical Notes
Polycarbonate (U ≈ 3.8 W/m²K) outperforms polyethylene film (U ≈ 5.2 W/m²K) in winter heating but increases capital cost by 60–80%; payback occurs within 3–4 heating seasons in northern Europe
Transpiration cooling peaks at 15–20 W/m² during high-humidity, high-LAI (leaf area index) growth stages; account for 20–30% reduction in heating load during full crop cover versus bare soil
Ventilation losses dominate on calm days; install roof vents and side louvers covering 15–20% of floor area to avoid excess humidity and fungal disease above 85% RH
Use actual regional TMY (Typical Meteorological Year) solar radiation data rather than theoretical values to capture cloud cover, albedo, and seasonal variation within ±8%