Retort Loading Density Optimization

Reference ID: MET-291C | Process Engineering Reference Sheets Calculation Guide

Introduction & Context

Retort loading density optimization is a critical process engineering task in thermal food preservation. It involves determining the spatial arrangement of containers within a batch retort to maximize throughput while ensuring uniform heat distribution. In a saturated steam retort, the primary objective is to maintain sufficient steam circulation to prevent cold spots, which could lead to under‑processing and food safety risks. This calculation is typically used during the design phase of crate loading patterns or when transitioning to higher‑density packaging configurations.

Methodology & Formulas

The optimization process relies on three primary physical constraints: the void fraction of the packed bed, the condensate film thickness on the container surface, and the pressure drop across the crate assembly.

1. Geometric Void Fraction

The void fraction (ε) represents the ratio of open space to total volume within the crate. It is calculated from the container diameter (D) and the centre‑to‑centre pitch (P) for a square‑pitch arrangement:

\[ \varepsilon = 1 - \frac{\pi \cdot D^{2}}{4 \cdot P^{2}} \]

2. Condensate Film Thickness

To prevent condensate bridging between adjacent containers, the gap (g) must exceed twice the maximum Nusselt film thickness (δ_max). The film thickness is determined by the physical properties of the condensate and the temperature differential (ΔT = T_sat − T_s):

\[ \delta_{max} = \left[ \frac{4 \cdot \mu_{f} \cdot k_{f} \cdot H \cdot \Delta T}{g \cdot \rho_{f} \cdot (\rho_{f} - \rho_{g}) \cdot h'_{fg}} \right]^{0.25} \]

where the modified latent heat of vaporization (h′_fg) accounts for sub‑cooling in the condensate film:

\[ h'_{fg} = h_{fg} + 0.68 \cdot Cp_{f} \cdot \Delta T \]

3. Pressure Drop (Ergun Equation)

The pressure drop (ΔP) across the packed bed of containers is calculated using the Ergun equation, which accounts for both viscous and inertial losses. The superficial velocity (V_s) is derived from the mass flow rate of the condensing steam (ṁ_v):

\[ V_{s} = \frac{\dot{m}_{v}}{\rho_{g} \cdot A_{crate}} \]

The pressure drop per unit bed depth is then:

\[ \frac{\Delta P}{L} = \frac{150 \cdot \mu_{g} \cdot V_{s} \cdot (1 - \varepsilon)^{2}}{\varepsilon^{3} \cdot D^{2}} + \frac{1.75 \cdot \rho_{g} \cdot V_{s}^{2} \cdot (1 - \varepsilon)}{\varepsilon^{3} \cdot D} \]

Regime/Constraint	Condition/Threshold
Void Fraction (ε)	\(0.3 < \varepsilon < 0.85\)
Condensate Bridging	\(g \geq 2 \cdot \delta_{max}\)
Pressure Drop (ΔP)	\(\Delta P \leq \Delta P_{max}\)
Flow Regime (Reynolds Number Re_p)	\(Re_{p} = \frac{\rho_{g} \cdot V_{s} \cdot D}{\mu_{g}}\)

To establish the optimal loading density, process engineers must conduct heat penetration studies that account for the specific thermal properties of the product. Follow these steps:

Perform a cold spot analysis using calibrated thermocouples to identify the slowest heating zone within the dense pack.
Validate that the come‑up time and lethality values (F₀) remain consistent with your established process authority filing.
Ensure that the spacing between containers allows for adequate steam or water circulation to prevent localised cold spots.

Overloading the retort can lead to significant safety and quality failures. Key risks include:

Inadequate heat distribution resulting in under‑processed product and potential microbial survival.
Increased mechanical stress on container seals due to restricted expansion during the heating cycle.
Extended come‑up times that may negatively impact the texture and nutritional profile of heat‑sensitive products.

Dividers and trays are critical for maintaining uniform flow patterns within the retort. When optimising your loading configuration, consider the following:

Ensure that tray perforations are not obstructed by the container footprint, which would impede the flow of the heating medium.
Use standardised stacking patterns to ensure that the pressure drop across the load remains within the design specifications of the retort circulation system.
Verify that the material of the dividers does not act as an insulator, which could alter the heat transfer rate to the containers in the centre of the stack.

Worked Example: Retort Loading Density Optimization

A batch saturated steam retort operates at 121.0 °C and 2.05 bar(a). A crate of internal dimensions 0.8 m × 1.2 m × 0.2 m is loaded with 50 cans (size 211×400, \(D_{can} = 0.068\) m, \(H_{can} = 0.106\) m) arranged in a single square‑pitch layer. The gap between adjacent can walls is set to 0.05 m. The objective is to verify whether this loading density satisfies the key constraints for steam circulation and heat transfer, and whether the gap could be reduced to improve throughput.

Known data (input parameters and their units):

\(D_{can} = 0.068\) m (can diameter)
\(H_{can} = 0.106\) m (can height)
\(g = 0.05\) m (space between can outer walls)
\(N_{\text{cans per crate}} = 50\) (number of cans per crate)
\(Crate_{W} = 0.8\) m, \(Crate_{L} = 1.2\) m, \(Crate_{H} = 0.2\) m (crate internal dimensions)
\(T_{sat} = 121.0 \, ^{\circ}\mathrm{C}\) (saturated steam temperature)
\(T_{s} = 20.0 \, ^{\circ}\mathrm{C}\) (initial can surface temperature)
\(P_{\text{retort}} = 2.05\) bar(a)
\(h_{fg} = 2\,200\,000\) J/kg (latent heat of condensation)
\(\rho_g = 1.15\) kg/m³ (steam density at 121 °C)
\(\mu_g = 1.3 \times 10^{-5}\) Pa·s (steam dynamic viscosity)
\(\rho_f = 998.0\) kg/m³ (condensate density)
\(\mu_f = 8.9 \times 10^{-4}\) Pa·s (condensate dynamic viscosity)
\(k_f = 0.606\) W/(m·K) (condensate thermal conductivity)
\(Cp_f = 4184\) J/(kg·K) (condensate specific heat)
\(g = 9.81\) m/s² (gravitational acceleration)
\(\Delta P_{\text{max}} = 0.05\) bar (maximum allowable pressure drop)
\(g_{\text{min, practical}} = 0.005\) m (minimum practical gap)

Step‑by‑Step Calculation

Compute the centre‑to‑centre pitch \(P\): \[ P = D_{can} + g = 0.068 + 0.05 = 0.118 \ \text{m} \]
Calculate the void fraction \(\varepsilon\): \[ \varepsilon = 1 - \frac{\pi D_{can}^2}{4 P^2} = 1 - \frac{\pi (0.068)^2}{4 (0.118)^2} = 0.739 \] This value (0.739) satisfies \(0.3 < \varepsilon < 0.85\); the Ergun packed‑bed correlation is valid.
Check the condensate film thickness constraint:
Compute the modified latent heat: \[ h'_{fg} = h_{fg} + 0.68\, Cp_f \, \Delta T = 2\,200\,000 + 0.68 \times 4184 \times 101.0 = 2\,487\,357.12 \ \text{J/kg} \] Maximum Nusselt film thickness: \[ \delta_{\max} = \left[ \frac{4 \mu_f k_f H_{can} (T_{sat} - T_s)}{g \rho_f (\rho_f - \rho_g) h'_{fg}} \right]^{0.25} = \left[ \frac{4 \times 8.9 \times 10^{-4} \times 0.606 \times 0.106 \times 101.0}{9.81 \times 998.0 \times (998.0 - 1.15) \times 2\,487\,357.12} \right]^{0.25} = 1.76 \times 10^{-4} \ \text{m} \] The anti‑bridging condition \(g \ge 2 \delta_{\max}\) gives \(0.05 \ge 3.52 \times 10^{-4}\) m, which is satisfied by a large margin.
Compute the heat load per crate:
Can surface area: \[ A_{can} = \pi D_{can} H_{can} = \pi \times 0.068 \times 0.106 = 0.0226 \ \text{m}^2 \ (\text{to 3 significant figures}) \] Using a typical condensation heat transfer coefficient of 10,000 W/(m²·K), the heat load per crate is: \[ Q_{\text{crate}} = N_{\text{cans}} \times 10\,000 \times A_{can} \times \Delta T = 50 \times 10\,000 \times (\pi \times 0.068 \times 0.106) \times 101.0 = 1\,143\,552 \ \text{W} \]
Determine the steam mass flow rate: \[ \dot{m}_{\text{vapor}} = \frac{Q_{\text{crate}}}{h_{fg}} = \frac{1\,143\,552}{2\,200\,000} = 0.520 \ \text{kg/s} \]
Calculate the superficial velocity \(V_s\): \[ A_{\text{crate}} = Crate_W \times Crate_L = 0.8 \times 1.2 = 0.96 \ \text{m}^2 \] \[ V_s = \frac{\dot{m}_{\text{vapor}}}{\rho_g A_{\text{crate}}} = \frac{0.520}{1.15 \times 0.96} = 0.471 \ \text{m/s} \]
Evaluate the particle Reynolds number \(Re_p\): \[ Re_p = \frac{\rho_g V_s D_{can}}{\mu_g} = \frac{1.15 \times 0.471 \times 0.068}{1.3 \times 10^{-5}} = 2.83 \times 10^{3} \] Since \(Re_p > 1000\), the flow is in the inertial regime; the full Ergun equation is applicable.
Apply the Ergun equation to obtain the pressure drop per unit bed height: \[ \frac{\Delta P}{L} = \frac{150 \mu_g V_s (1-\varepsilon)^2}{\varepsilon^3 D_{can}^2} + \frac{1.75 \rho_g V_s^2 (1-\varepsilon)}{\varepsilon^3 D_{can}} \] The two contributing terms are: \[ \text{Term}_1 = \frac{150 \times 1.3 \times 10^{-5} \times 0.471 \times (1-0.739)^2}{0.739^3 \times (0.068)^2} = 0.033 \] \[ \text{Term}_2 = \frac{1.75 \times 1.15 \times (0.471)^2 \times (1-0.739)}{0.739^3 \times 0.068} = 4.24 \] Total pressure drop over the crate height \(Crate_H = 0.2\) m: \[ \Delta P = (\text{Term}_1 + \text{Term}_2) \times Crate_H = (0.033 + 4.24) \times 0.2 = 0.855 \ \text{Pa} \] Convert to bar (\(1\ \text{bar} = 100\,000\ \text{Pa}\)): \[ \Delta P_{\text{bar}} = \frac{0.855}{100\,000} = 8.55 \times 10^{-6} \ \text{bar} \]
Compare the pressure drop with the maximum allowable limit:
\(\Delta P_{\text{bar}} = 8.55 \times 10^{-6}\) bar \( \ll \Delta P_{\text{max}} = 0.05\) bar. The pressure drop is well within the acceptable range.

Final Answer

The design with a 5 cm gap yields a void fraction of 0.739, a condensate film check that is easily satisfied, and a negligible pressure drop of approximately \(8.6 \times 10^{-6}\) bar. All constraints are met, but the very low pressure drop indicates that the spacing is over‑designed for steam circulation. The gap could be reduced to well below 2 cm (e.g., to the practical minimum of 0.5 cm) to increase throughput while still satisfying both the condensate bridging and pressure drop limits. For the current configuration, the loading density is acceptable but not optimal for maximising can count per crate.

"Un projet n'est jamais trop grand s'il est bien conçu." — André Citroën

"La difficulté attire l'homme de caractère, car c'est en l'étreignant qu'il se réalise." — Charles de Gaulle