Continuous Roasting Operations and Quality Control

Reference ID: MET-74FD | Process Engineering Reference Sheets Calculation Guide

Introduction & Context

In continuous dry roasting operations, maintaining precise control over product quality and safety is critical. This calculation methodology provides a framework for modeling the transient thermal behavior of spherical food particles (such as nuts or legumes) within a forced-convection oven. By predicting the temperature evolution at both the core and the surface of the product, engineers can ensure that the core reaches the required lethality for pathogen reduction (e.g., Salmonella) while simultaneously managing the surface temperature to achieve the desired Maillard browning index. This model is essential for process validation, equipment sizing, and optimizing residence times in industrial roasting lines.

Methodology & Formulas

The thermal profile is determined using the one-term transient conduction solution for a sphere. The process begins by calculating the Biot number (Bi) to determine the appropriate heat transfer regime.

The Biot number is defined as:

\[ Bi = \frac{h \cdot R}{k} \]

For systems where Bi > 0.1, the one-term approximation for transient conduction is applied. The first eigenvalue λ₁ is obtained by solving the characteristic equation for a sphere:

\[ 1 - \lambda_{1} \cdot \cot(\lambda_{1}) = Bi \]

The Fourier number (τ) is calculated to track the dimensionless time:

\[ \tau = \frac{\alpha \cdot t}{R^{2}} \]

The core temperature (T_c) is derived from the dimensionless temperature ratio (θ_c):

\[ \theta_{c} = A_{1} \cdot \exp(-\lambda_{1}^{2} \cdot \tau) \] \[ T_{c} = T_{\infty} + \theta_{c} \cdot (T_{i} - T_{\infty}) \]

The surface temperature (T_s) accounts for the spatial temperature distribution within the sphere:

\[ \theta_{s} = \theta_{c} \cdot \frac{\sin(\lambda_{1})}{\lambda_{1}} \] \[ T_{s} = T_{\infty} + \theta_{s} \cdot (T_{i} - T_{\infty}) \]

Safety and quality are evaluated through numerical integration over time (t). Lethality is calculated using the Bigelow model, where D_T is the decimal reduction time at temperature T:

\[ D_{T} = D_{\text{ref}} \cdot 10^{\frac{T_{\text{ref}} - T_{c}}{z}} \] \[ \log\left(\frac{N}{N_{0}}\right) = -\int_{0}^{t} \frac{1}{D_{T_{c}}} \, dt \]

Browning development is modeled using the Arrhenius equation, where k_T is the reaction rate constant:

\[ k_{T} = A \cdot \exp\left(-\frac{E_{a}}{R_{\text{gas}} \cdot (T_{s} + 273.15)}\right) \] \[ C_{\text{total}} = \int_{0}^{t} k_{T_{s}} \, dt \]

Regime / Condition	Criteria	Action / Model
Lumped Capacitance	Bi ≤ 0.1	Use lumped capacitance model (uniform temperature).
One-Term Approximation	Bi > 0.1 and τ ≥ 0.2	Use one-term transient conduction solution.
Full Series Solution	Bi > 0.1 and τ < 0.2	Use infinite series or Heisler/Gröber charts.
Moisture Limit	Moisture > 10% (wet basis)	Model must include latent heat and mass transfer.

To ensure consistent moisture levels throughout the roasting process, process engineers should focus on the following parameters:

Maintain a stable mass flow rate of the raw material to prevent residence time fluctuations.
Monitor the exhaust air humidity sensors to adjust the burner output dynamically.
Calibrate the inlet air temperature control loop to compensate for ambient humidity changes.

Thermal degradation often manifests through specific physical and chemical changes that require immediate process adjustment:

A significant shift in the L*a*b* color space values measured by inline spectrophotometers.
An increase in volatile organic compound emissions detected by the exhaust gas analyzer.
A measurable decrease in bulk density indicating structural collapse of the product matrix.

Optimizing the air-to-product ratio is critical for achieving uniform heat transfer across the bed depth:

Adjust the fan frequency to maintain a constant static pressure within the roasting chamber.
Ensure the bed depth remains uniform across the conveyor belt to prevent air channeling.
Verify that the air velocity is sufficient to achieve fluidization without causing excessive product attrition.

Worked Example: Continuous Roasting of Almonds

Scenario: A continuous belt forced-convection oven is used to roast spherical almond kernels. The goal is to achieve a core pathogen reduction of at least 5 log cycles while controlling surface browning. The oven operates at a constant air temperature of 160 °C with a convective heat transfer coefficient of 40 W/m²·K. Almonds enter at 20 °C with radius 5 mm.

Knowns:

\( R = 0.005 \) m
\( T_{\infty} = 160.0 \) °C
\( h = 40.0 \) W/m²·K
\( T_{i} = 20.0 \) °C
\( k = 0.4 \) W/m·K
\( \alpha = 1.5 \times 10^{-7} \) m²/s
\( E_{a} = 120000.0 \) J/mol
\( A = 5.0 \times 10^{10} \) 1/s
\( D_{\text{ref}} = 50.0 \) min at \( T_{\text{ref}} = 80 \) °C
\( z = 15.0 \) °C
\( R_{\text{gas}} = 8.314 \) J/mol·K
Processing time \( t = 120.0 \) s
Integration steps \( N = 120 \)

Step-by-Step Calculation:

Biot Number Check
\( \text{Bi} = \dfrac{h \cdot R}{k} = \dfrac{40.0 \cdot 0.005}{0.4} = 0.5 \)
Since Bi = 0.5 > 0.1, the one-term transient conduction solution is required; lumped capacitance is not valid.
Eigenvalue Determination
For a sphere at Bi = 0.5, solving the characteristic equation \( 1 - \lambda_{1} \cot(\lambda_{1}) = 0.5 \) yields:
\( \lambda_{1} = 1.1656 \)
\( A_{1} = \dfrac{2[\sin(\lambda_{1}) - \lambda_{1} \cos(\lambda_{1})]}{\lambda_{1} - \sin(\lambda_{1}) \cos(\lambda_{1})} = 1.144 \)
The one-term approximation is valid if the Fourier number at the final time exceeds 0.2.
Core Temperature Profile at t = 120 s
Fourier number: \( \tau = \dfrac{\alpha \cdot t}{R^{2}} = \dfrac{1.5 \times 10^{-7} \cdot 120.0}{(0.005)^{2}} = 0.72 \)
Since \( \tau = 0.72 > 0.2 \), the one-term solution is valid.
Dimensionless center temperature: \( \theta_{c}(t) = A_{1} \exp(-\lambda_{1}^{2} \tau) = 1.144 \cdot \exp(-(1.1656)^{2} \cdot 0.72) = 0.430 \)
Core temperature: \( T_{c}(t) = T_{\infty} + \theta_{c}(t) \cdot (T_{i} - T_{\infty}) = 160.0 + 0.430 \cdot (20.0 - 160.0) = 99.80 \) °C.
Surface Temperature Profile at t = 120 s
Dimensionless surface temperature: \( \theta_{s}(t) = \theta_{c}(t) \cdot \dfrac{\sin(\lambda_{1})}{\lambda_{1}} = 0.430 \cdot \dfrac{\sin(1.1656)}{1.1656} = 0.339 \)
Surface temperature: \( T_{s}(t) = T_{\infty} + \theta_{s}(t) \cdot (T_{i} - T_{\infty}) = 160.0 + 0.339 \cdot (20.0 - 160.0) = 112.54 \) °C.
Integrated Lethality (Pathogen Reduction)
Trapezoidal integration is performed over \( N = 120 \) steps with \( dt = 1.0 \) s. At each time step, the core temperature \( T_{c} \) determines the instantaneous D-value:
\( D(T_{c}) = D_{\text{ref}} \cdot 10^{(T_{\text{ref}} - T_{c})/z} \)
At the final time step (\( t = 120 \) s, \( T_{c} = 99.80 \) °C):
\( D(T_{c}) = 50.0 \cdot 10^{(80.0 - 99.80)/15.0} = 2.40 \) min.
The cumulative lethality sum (trapezoidal integration in minutes) yields:
\( \log(N/N_{0}) = -\sum \dfrac{1}{D(T_{c})} \cdot \dfrac{dt}{60} = -0.083 \)
Thus, the pathogen reduction is 0.083 log cycles at 120 s.
Browning Index (Quality)
Trapezoidal integration is performed using surface temperature in Kelvin. At the final time step (\( T_{s} = 112.54 \) °C):
\( k(T_{s}) = A \cdot \exp\left(-\dfrac{E_{a}}{R_{\text{gas}} \cdot (T_{s} + 273.15)}\right) = 5.0 \times 10^{10} \cdot \exp\left(-\dfrac{120000.0}{8.314 \cdot (112.54 + 273.15)}\right) = 2.80 \times 10^{-6} \) 1/s.
Cumulative browning index: \( C(t) = \sum k(T_{s}) \cdot dt = 7.5 \times 10^{-5} \)

Final Answer:

Final core temperature: 99.80 °C
Final surface temperature: 112.54 °C
Log reduction at 120 s: 0.083 (does not meet target of 5 log reduction)
Browning index at 120 s: 7.5 × 10⁻⁵ (negligible browning)

"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

Continuous Roasting Operations and Quality Control

Introduction & Context

🚀 Skip the Manual Math!

Methodology & Formulas

Worked Example: Continuous Roasting of Almonds