Direct vs Indirect Refrigeration System Design

Reference ID: MET-8F94 | Process Engineering Reference Sheets Calculation Guide

Introduction & Context

Refrigeration system design is a critical component of process engineering, particularly in food, beverage, and chemical manufacturing. The choice between Direct Expansion (DX) and Indirect Refrigeration systems dictates the thermal efficiency, safety, and operational flexibility of a facility.

In a Direct Expansion system, the refrigerant evaporates directly within the process heat exchanger. This minimizes the number of heat transfer steps, resulting in higher suction pressures and improved compressor efficiency. Conversely, Indirect Refrigeration utilizes a secondary heat transfer fluid (such as a glycol-water mixture) to transport cooling capacity from a central chiller to remote process loads. While this introduces an additional temperature approach penalty, it is often preferred for its ability to centralize hazardous refrigerants (like ammonia), provide thermal storage, and allow for independent temperature control across multiple process stations.

Methodology & Formulas

The design process involves balancing heat loads, determining secondary fluid mass flow rates, and calculating the required heat transfer surface areas based on the Logarithmic Mean Temperature Difference (LMTD).

The total heat load is the sum of all individual process requirements:

\[ \dot{Q}_{\text{total}} = \sum \dot{Q}_{\text{station}} \]

The mass flow rate of the secondary fluid is determined by the total heat load, the specific heat capacity of the fluid, and the allowable temperature rise across the process loop:

\[ \dot{m}_{\text{glycol}} = \frac{\dot{Q}_{\text{total}}}{c_{p,\text{glycol}} \cdot \Delta T_{\text{glycol,rise}}} \]

The required refrigerant saturation temperature is constrained by the process outlet temperature and the two approach temperature differences:

\[ T_{\text{sat,ref}} = (T_{\text{product,out}} - \Delta T_{\text{process,app}}) - \Delta T_{\text{chiller,app}} \]

The overall heat transfer coefficient (U) accounts for the convective resistances of both fluids and the fouling resistance:

\[ U = \frac{1}{\frac{1}{h_{1}} + R_{\text{fouling}} + \frac{1}{h_{2}}} \]

The required heat transfer area is calculated using the LMTD method (counterflow configuration):

\[ \Delta T_{\text{lm}} = \frac{\Delta T_{1} - \Delta T_{2}}{\ln\left(\frac{\Delta T_{1}}{\Delta T_{2}}\right)} \] \[ A = \frac{\dot{Q}}{U \cdot \Delta T_{\text{lm}}} \]

Parameter	Condition / Threshold
Freezing Protection	\( T_{\text{glycol,in}} \geq T_{\text{freeze,glycol}} + 5.0 \, ^\circ\text{C} \)
Economic Approach	\( \Delta T_{\text{approach}} \geq 2.0 \, ^\circ\text{C} \)
Secondary Fluid Velocity	\( 1.0 \, \text{m/s} \leq v_{\text{glycol}} \leq 3.0 \, \text{m/s} \)
Boiling Regime	\( q''_{\text{boil}} < 50 \, \text{kW/m}^2 \) (to avoid film boiling)

Direct and indirect systems differ primarily in how the refrigerant interacts with the process load.

Direct systems circulate the primary refrigerant directly through the process heat exchanger, which maximizes thermal efficiency but increases the risk of product contamination.
Indirect systems utilize a secondary heat transfer fluid, such as glycol or brine, to bridge the gap between the refrigerant and the process, providing a safety buffer at the cost of additional pumping energy and heat transfer losses.

An indirect design is recommended when safety and containment are the highest priorities. Specifically, consider this approach when:

The process involves hazardous or toxic materials where a refrigerant leak could cause catastrophic contamination.
The facility requires strict compliance with environmental regulations regarding refrigerant charge limits in occupied spaces.
The system requires precise temperature control across multiple distributed process zones.

The total cost of ownership is influenced by both capital expenditure and ongoing operational efficiency.

Direct systems typically have lower initial capital costs due to fewer components and higher thermodynamic efficiency, leading to lower electricity consumption.
Indirect systems often incur higher upfront costs due to the secondary loop infrastructure, including pumps, expansion tanks, and heat exchangers, and they require higher energy input to overcome the secondary fluid pumping requirements and the lower suction pressure caused by the additional chiller approach temperature difference.

Worked Example: Design of an Indirect Refrigeration System for a Dairy Plant

This example demonstrates the design calculations for an indirect refrigeration system using a secondary glycol loop to cool two process stations at a dairy plant. A direct expansion (DX) alternative is also sized for comparison.

Scenario: A dairy plant requires chilled 30% propylene glycol/water (freezing point = -14.0 °C) to cool two stations: Station A (milk from 40.0 °C to 4.0 °C, cooling load 200.0 kW) and Station B (cream from 30.0 °C to 4.0 °C, cooling load 80.0 kW). Plate heat exchangers are used for process coolers and a flooded ammonia chiller. The target temperature approach for each exchanger is 4.0 °C. Assume a glycol temperature rise of 5.0 °C across the process loop.

Knowns:

Cooling load Station A, \( Q_{A} = 200.0 \, \text{kW} \)
Cooling load Station B, \( Q_{B} = 80.0 \, \text{kW} \)
Milk inlet temperature, \( T_{\text{milk,in}} = 40.0 \, \text{°C} \)
Milk outlet temperature, \( T_{\text{milk,out}} = 4.0 \, \text{°C} \)
Specific heat capacity of glycol, \( c_{p,\text{glycol}} = 3.6 \, \text{kJ/kg·K} \)
Temperature approach at process cooler, \( \Delta T_{\text{process}} = 4.0 \, \text{°C} \)
Temperature approach at chiller evaporator, \( \Delta T_{\text{chiller}} = 4.0 \, \text{°C} \)
Glycol temperature rise across process loop, \( \Delta T_{\text{glycol,rise}} = 5.0 \, \text{°C} \)
Fouling resistance, \( R_{\text{fouling}} = 0.0001 \, \text{m}^{2}\text{K/W} \)
Ammonia boiling heat transfer coefficient, \( h_{\text{NH3,boil}} = 4000.0 \, \text{W/m}^{2}\text{K} \)
Glycol heat transfer coefficient (chiller side), \( h_{\text{glycol,chiller}} = 3000.0 \, \text{W/m}^{2}\text{K} \)
Glycol heat transfer coefficient (process side), \( h_{\text{glycol,process}} = 3000.0 \, \text{W/m}^{2}\text{K} \)
Milk heat transfer coefficient, \( h_{\text{milk}} = 4000.0 \, \text{W/m}^{2}\text{K} \)
DX evaporator overall heat transfer coefficient, \( U_{\text{DX}} = 2000.0 \, \text{W/m}^{2}\text{K} \)
Glycol freezing point, \( T_{\text{freeze}} = -14.0 \, \text{°C} \)
Minimum economic approach, \( \Delta T_{\text{min}} = 2.0 \, \text{°C} \)

Step-by-Step Calculation:

Total cooling load: \[ Q_{\text{total}} = Q_{A} + Q_{B} = 200.0 + 80.0 = 280.0 \, \text{kW} \]
Glycol supply/return temperatures:
Glycol enters process coolers at: \[ T_{\text{glycol,in}} = T_{\text{milk,out}} - \Delta T_{\text{process}} = 4.0 - 4.0 = 0.0 \, \text{°C} \]
Glycol leaves process coolers at: \[ T_{\text{glycol,out}} = T_{\text{glycol,in}} + \Delta T_{\text{glycol,rise}} = 0.0 + 5.0 = 5.0 \, \text{°C} \]
Glycol mass flow rate: \[ \dot{m}_{\text{glycol}} = \frac{Q_{\text{total}}}{c_{p,\text{glycol}} \cdot \Delta T_{\text{glycol,rise}}} = \frac{280.0}{3.6 \times 5.0} = 15.56 \, \text{kg/s} \]
Refrigerant saturation temperature (chiller): (Based on leaving glycol temperature from chiller, which equals \( T_{\text{glycol,in}} \)) \[ T_{\text{sat}} = T_{\text{glycol,in}} - \Delta T_{\text{chiller}} = 0.0 - 4.0 = -4.0 \, \text{°C} \]
Overall heat transfer coefficients:
Process cooler (glycol-to-milk): \[ U_{\text{process}} = \frac{1}{\frac{1}{h_{\text{glycol,process}}} + R_{\text{fouling}} + \frac{1}{h_{\text{milk}}}} = \frac{1}{\frac{1}{3000.0} + 0.0001 + \frac{1}{4000.0}} = 1463.4 \, \text{W/m}^{2}\text{K} \]
Chiller evaporator (ammonia-to-glycol): \[ U_{\text{chiller}} = \frac{1}{\frac{1}{h_{\text{NH3,boil}}} + R_{\text{fouling}} + \frac{1}{h_{\text{glycol,chiller}}}} = \frac{1}{\frac{1}{4000.0} + 0.0001 + \frac{1}{3000.0}} = 1463.4 \, \text{W/m}^{2}\text{K} \]
Log-Mean Temperature Difference (LMTD) and area for Station A (counterflow):
Hot fluid (milk): 40.0 °C → 4.0 °C; Cold fluid (glycol): 0.0 °C → 5.0 °C \[ \Delta T_{1} = T_{\text{milk,in}} - T_{\text{glycol,out}} = 40.0 - 5.0 = 35.0 \, \text{°C} \] \[ \Delta T_{2} = T_{\text{milk,out}} - T_{\text{glycol,in}} = 4.0 - 0.0 = 4.0 \, \text{°C} \] \[ \text{LMTD}_{A} = \frac{\Delta T_{1} - \Delta T_{2}}{\ln\left(\frac{\Delta T_{1}}{\Delta T_{2}}\right)} = \frac{35.0 - 4.0}{\ln\left(\frac{35.0}{4.0}\right)} = 14.29 \, \text{°C} \] \[ A_{A} = \frac{Q_{A} \times 1000}{U_{\text{process}} \times \text{LMTD}_{A}} = \frac{200000}{1463.4 \times 14.29} = 9.56 \, \text{m}^{2} \]
LMTD and area for chiller (counterflow):
Hot fluid (glycol): 5.0 °C → 0.0 °C; Cold fluid (refrigerant): -4.0 °C (constant) \[ \Delta T_{1} = T_{\text{glycol,out}} - T_{\text{sat}} = 5.0 - (-4.0) = 9.0 \, \text{°C} \] \[ \Delta T_{2} = T_{\text{glycol,in}} - T_{\text{sat}} = 0.0 - (-4.0) = 4.0 \, \text{°C} \] \[ \text{LMTD}_{\text{ch}} = \frac{\Delta T_{1} - \Delta T_{2}}{\ln\left(\frac{\Delta T_{1}}{\Delta T_{2}}\right)} = \frac{9.0 - 4.0}{\ln\left(\frac{9.0}{4.0}\right)} = 6.166 \, \text{°C} \] \[ A_{\text{chiller}} = \frac{Q_{\text{total}} \times 1000}{U_{\text{chiller}} \times \text{LMTD}_{\text{ch}}} = \frac{280000}{1463.4 \times 6.166} = 31.03 \, \text{m}^{2} \]
Comparison: Direct Expansion (DX) for Station A:
For DX, \( T_{\text{sat,DX}} = T_{\text{milk,out}} - \Delta T_{\text{process}} = 4.0 - 4.0 = 0.0 \, \text{°C} \)
LMTD (milk: 40.0 °C → 4.0 °C; refrigerant at 0.0 °C): \[ \text{LMTD}_{\text{DX}} = \frac{(40.0 - 0.0) - (4.0 - 0.0)}{\ln\left(\frac{40.0}{4.0}\right)} = 15.63 \, \text{°C} \] \[ A_{\text{DX}} = \frac{Q_{A} \times 1000}{U_{\text{DX}} \times \text{LMTD}_{\text{DX}}} = \frac{200000}{2000.0 \times 15.63} = 6.396 \, \text{m}^{2} \]
Validity check: Freezing point margin
Minimum glycol temperature in system is \( T_{\text{glycol,in}} = 0.0 \, \text{°C} \).
Margin above freezing point: \( 0.0 - (-14.0) = 14.0 \, \text{°C} \), which exceeds the required 5.0 °C margin. No freezing risk.
The chosen approach temperatures (4.0 °C each) are above the minimum economic threshold of 2.0 °C. Design is valid.
Note: In detailed engineering, the secondary fluid velocity and boiling heat flux should also be verified against the bounds table. The velocity check requires pipe sizing data (not provided), while the DX heat flux (\(q'' \approx 200\,\text{kW}/6.396\,\text{m}^2 \approx 31.3\,\text{kW/m}^2\)) is well below the 50 kW/m² film-boiling limit.

Final Results Summary:

Total cooling load: \( Q_{\text{total}} = 280.0 \, \text{kW} \)
Glycol mass flow rate: \( \dot{m}_{\text{glycol}} = 15.56 \, \text{kg/s} \)
Glycol supply/return temperatures: 0.0 °C / 5.0 °C
Required refrigerant saturation temperature (indirect): \( T_{\text{sat}} = -4.0 \, \text{°C} \)
Process cooler area for Station A: \( A_{A} = 9.56 \, \text{m}^{2} \)
Chiller evaporator area: \( A_{\text{chiller}} = 31.03 \, \text{m}^{2} \)
DX alternative area for Station A: \( A_{\text{DX}} = 6.40 \, \text{m}^{2} \)

"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

Direct vs Indirect Refrigeration System Design

Introduction & Context

🚀 Skip the Manual Math!

Methodology & Formulas

Worked Example: Design of an Indirect Refrigeration System for a Dairy Plant