In high-density commercial CEA glasshouses cultivating C3 fruiting crops (such as high-wire tomatoes, cucumbers, and sweet peppers), ambient atmospheric carbon dioxide ($CO_2 \approx 420\text{ ppm}$) represents the single greatest biochemical bottleneck to maximum biomass synthesis. When supplemental horticultural LED arrays deliver 250–350 $\mu\text{mol}\cdot\text{m}^{-2}\cdot\text{s}^{-1}$ atop natural solar radiation, vigorous canopies can deplete greenhouse $CO_2$ concentrations down to 200–250 ppm within 45 minutes of vent closure, depressing net carbon assimilation by over 40%. However, blindly injecting pure liquid $CO_2$ or CHP flue gas at flat setpoints (e.g., static 1,000 ppm) leads to catastrophic financial waste during vent openings. By implementing the Farquhar-von Caemmerer-Berry (FvCB) biochemical model coupled with real-time ventilation mass-balance equations, growers across European production hubs (from Straelen to the Westland) dynamically match dosing rates to the marginal enzyme saturation of Ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCO).
1. Biochemical Foundations: The Farquhar-von Caemmerer-Berry (FvCB) Model
Net photosynthetic $CO_2$ assimilation rate ($A_{net}$, $\mu\text{mol } CO_2\cdot\text{m}^{-2}\cdot\text{s}^{-1}$) in C3 crops is governed by the minimum of two distinct enzymatic and thermodynamic limitation regimes:
A_{net} = \min(A_c, A_j) - R_d
Where:
- $A_c$: RuBisCO-limited carboxylation rate under low intercellular $CO_2$ ($C_i$).
- $A_j$: RuBP-regeneration limited assimilation rate driven by the photosynthetic electron transport rate ($J$) under high $C_i$ or low irradiance.
- $R_d$: Daytime mitochondrial respiration ($\mu\text{mol } CO_2\cdot\text{m}^{-2}\cdot\text{s}^{-1}$).
1.1. RuBisCO Carboxylation Kinetics ($A_c$)
Under ambient or depleted greenhouse atmospheres, RuBisCO operates far below substrate saturation. Carboxylation ($v_c$) competes directly with oxygenation ($v_o$), which initiates energetically wasteful photorespiration:
A_c = V_{cmax} \cdot \frac{C_i - \Gamma^*}{C_i + K_c \cdot \left(1 + \frac{O_i}{K_o}\right)}
Where:
- $V_{cmax}$: Maximum RuBisCO carboxylation capacity ($\approx 80-110\text{ }\mu\text{mol}\cdot\text{m}^{-2}\cdot\text{s}^{-1}$ at $25^\circ\text{C}$ for tomato).
- $C_i$: Sub-stomatal intercellular $CO_2$ partial pressure ($\mu\text{bar}$ or $\text{ppm}$).
- $O_i$: Intercellular oxygen partial pressure ($\approx 210\text{ mbar}$).
- $\Gamma^*$: $CO_2$ compensation point in the absence of day respiration ($\approx 42.7\text{ }\mu\text{bar}$ at $25^\circ\text{C}$).
- $K_c, K_o$: Michaelis-Menten kinetic constants for $CO_2$ and $O_2$ respectively.
2. Temperature Sensitivity and Enzyme Arrhenius Thermodynamics
The kinetic parameters ($V_{cmax}, K_c, K_o, \Gamma^*$) vary non-linearly with leaf temperature ($T_{leaf}$, in Kelvin) according to the Arrhenius temperature function:
f(T_k) = k_{25} \cdot \exp\left( \frac{E_a \cdot (T_k - 298.15)}{298.15 \cdot R \cdot T_k} \right)
Where $R = 8.314\text{ J}\cdot\text{mol}^{-1}\cdot\text{K}^{-1}$ is the universal gas constant, and $E_a$ is the activation energy ($E_{a,Kc} \approx 79.43\text{ kJ/mol}$, $E_{a,Ko} \approx 36.38\text{ kJ/mol}$). As canopy temperature elevates from $18^\circ\text{C}$ to $28^\circ\text{C}$, the affinity of RuBisCO for oxygen increases faster than for $CO_2$. Therefore, higher temperatures mandate elevated $CO_2$ setpoints to suppress runaway photorespiration.
| Leaf Temp (°C) | $\Gamma^*$ (ppm) | $K_c$ ($\mu\text{bar}$) | $K_o$ ($\text{mbar}$) | Optimal $C_i$ for 90% Saturation |
|---|---|---|---|---|
| 18°C | 31.4 | 235 | 220 | 680 ppm |
| 22°C | 37.2 | 328 | 255 | 820 ppm |
| 25°C | 42.7 | 404 | 278 | 950 ppm |
| 28°C | 49.6 | 502 | 304 | 1,120 ppm |
3. Stomatal Resistance and Canopy Mass Transfer
Enriching the bulk air to $C_a = 1,000\text{ ppm}$ does not directly translate to $C_i = 1,000\text{ ppm}$. Intercellular $CO_2$ is governed by stomatal conductance ($g_s$) and boundary layer conductance ($g_b$) via Fick's first law of diffusion:
A_{net} = g_{tc} \cdot (C_a - C_i) = \frac{1}{r_{bc} + r_{sc}} \cdot (C_a - C_i)
Where $g_{tc}$ is the total conductance to $CO_2$, and $r_{bc}, r_{sc}$ are boundary layer and stomatal resistances ($r = 1/g$). When vapor pressure deficit (VPD) rises above 1.5 kPa or root-zone electrical conductivity (EC) exceeds target osmotic thresholds, stomata partially close ($g_s \downarrow$). Under restricted $g_s$, elevating external $C_a$ becomes significantly more urgent to maintain intracellular diffusion gradients.
4. Dynamic Economic Dosing: Marginal Revenue vs. Ventilation Loss
In commercial Venlo facilities, $CO_2$ mass balance is a dynamic differential equation involving canopy assimilation, external dosing, and ventilation air exchange:
V_{gh} \cdot \frac{dC_a}{dt} = \Phi_{inj}(t) - LAI \cdot A_{net}(C_i, I, T) - ACH(t) \cdot V_{gh} \cdot (C_a - C_{ext})
Where $V_{gh}$ is greenhouse volume ($m^3/m^2$), $LAI$ is leaf area index ($m^2/m^2$), $\Phi_{inj}$ is the mass injection rate ($g\cdot m^{-2}\cdot h^{-1}$), and $ACH(t)$ is air changes per hour ($h^{-1}$).
The Marginal Cost-Benefit Formulation
The optimal $CO_2$ concentration ($C_a^*$) occurs when the Marginal Revenue of Photosynthesis ($MR_C$) precisely equals the Marginal Cost of Gas Consumption and Loss ($MC_C$):
MR_C = \frac{\partial A_{net}}{\partial C_a} \cdot \text{Biomass Conversion Efficiency} \cdot \text{Market Price of Produce}
MC_C = \left( \frac{\partial A_{net}}{\partial C_a} + ACH(t) \cdot V_{gh} \right) \cdot P_{CO2}
When greenhouse roof vents open beyond 15–20% on warm sunny afternoons ($ACH > 15\text{ h}^{-1}$), $MC_C$ skyrockets exponentially. Under these high-ventilation conditions, pushing $C_a$ beyond 500–550 ppm results in over 85% of purchased $CO_2$ escaping directly into the atmosphere with near-zero marginal crop return. Conversely, during winter when screens are deployed and vents are closed ($ACH < 0.5\text{ h}^{-1}$), maintaining 900–1,100 ppm delivers peak biomass accretion at minimal dosing expenditure.
5. Commercial Implementation Architecture
Modern climate control computers (e.g., Priva Connext, Hoogendoorn i-Grow, Ridder MultiMa) integrated with Inwoovation's predictive algorithms execute the following closed-loop hierarchy:
- Step 1: Irradiance Gating: If canopy PPFD is below the light compensation point ($< 50\text{ }\mu\text{mol}\cdot\text{m}^{-2}\cdot\text{s}^{-1}$), $CO_2$ dosing is completely inhibited ($A_{net} \le 0$).
- Step 2: Ventilation Tracking: Continuous recalculation of $ACH$ using leeward/windward vent position sensors and outdoor anemometer wind speed.
- Step 3: Real-Time FvCB Curve Fit: Continuous estimation of $V_{cmax}(T_{leaf})$ and $C_i$ to target the knee of the $A/C_i$ curve where RuBisCO transitions from $CO_2$-limited to RuBP-limited kinetics.
- Step 4: Heat Storage Buffer Synchronization: When using CHP flue gas, dosing is coordinated with the hot water thermal storage buffer tank to prevent heat dump waste.
6. Conclusion
Precision $CO_2$ enrichment is not a static setpoint game; it is an active biophysical balancing act between biochemical enzyme kinetics and thermodynamic fluid loss. By deploying model predictive control based on the Farquhar FvCB equation, commercial greenhouse operators increase seasonal fruit yields by 12–18% while simultaneously reducing pure $CO_2$ consumption costs by up to 35%.