Whey Demineralization Process Design

Reference ID: MET-725B | Process Engineering Reference Sheets Calculation Guide

Introduction & Context

Whey demineralization is a critical unit operation in the dairy industry, primarily used to produce high-value ingredients such as demineralized whey powder for infant formula and specialized nutritional products. The process involves the removal of inorganic salts (ash) from whey, which would otherwise impart an undesirable salty taste and interfere with the functional properties of the protein.

The two-column ion exchange system, consisting of a strong acid cation (SAC) exchanger followed by a strong base anion (SBA) exchanger, is the industry standard for achieving high-efficiency mineral removal. This calculation blueprint provides the engineering framework to size these columns based on mass balance, resin capacity, and cycle time requirements, ensuring the system meets specific product ash targets through controlled flow splitting and regeneration cycles.

Methodology & Formulas

The design methodology relies on a stoichiometric mass balance of the mineral content within the whey. For a bypass configuration where a fraction of the feed is sent through the columns and completely demineralized, then blended with the untreated bypass stream, the fraction of feed that must be treated to achieve a target product ash is determined by:

\[ f_{\text{col}} = \frac{X_{\text{feed}} - X_{\text{prod}}}{X_{\text{feed}} - X_{\text{prod,col}}} \]

where \(X_{\text{prod,col}}\) is the ash mass fraction leaving the column (typically taken as 0 for complete demineralization in this simplified model). The flow rate directed to the columns (\(\dot{V}_{\text{col}}\)) is then:

\[ \dot{V}_{\text{col}} = \dot{V}_{\text{total}} \cdot f_{\text{col}} \]

The mass removal rate of minerals (\(\dot{m}_{\text{rem}}\)) is calculated based on the whey density (\(\rho\)) and the difference between feed and column product ash:

\[ \dot{m}_{\text{rem}} = \left( X_{\text{feed}} - X_{\text{prod,col}} \right) \cdot \dot{V}_{\text{col}} \cdot \rho \]

The equivalent removal rate (\(\dot{E}_{\text{rem}}\)) is derived using the average equivalent weight of the whey minerals (\(EW_{\text{avg}}\)):

\[ \dot{E}_{\text{rem}} = \frac{\dot{m}_{\text{rem}}}{EW_{\text{avg}}} \]

The total equivalents to be removed per service cycle (\(E_{\text{cycle}}\)) are determined by the cycle time (\(t_{\text{cycle}}\)):

\[ E_{\text{cycle}} = \dot{E}_{\text{rem}} \cdot t_{\text{cycle}} \]

The required resin volumes for the cation (\(V_{\text{cat}}\)) and anion (\(V_{\text{an}}\)) columns are calculated using their respective usable capacities (\(C_{\text{cat}}\) and \(C_{\text{an}}\)):

\[ V_{\text{cat}} = \frac{E_{\text{cycle}}}{C_{\text{cat}}}, \qquad V_{\text{an}} = \frac{E_{\text{cycle}}}{C_{\text{an}}} \]

The superficial velocity (\(v\)) is verified against the cross-sectional area (\(A\)) of the columns to ensure proper contact time:

\[ A = \frac{V_{\text{resin}}}{H_{\text{bed}}}, \qquad v = \frac{\dot{V}_{\text{col}}}{A} \]

Note: Ensure unit consistency – \(V_{\text{resin}}\) in m³ and \(H_{\text{bed}}\) in m, or apply the conversion 1000 L/m³.

Regeneration volumes for the acid (\(V_{\text{HCl}}\)) and base (\(V_{\text{NaOH}}\)) are calculated based on the regeneration levels (\(RL_{\text{cat}}\) and \(RL_{\text{an}}\)) and the chemical properties of the regenerants:

\[ V_{\text{HCl}} = \frac{E_{\text{cycle}} \cdot RL_{\text{cat}} \cdot EW_{\text{HCl}}}{w_{\text{HCl}} \cdot \rho_{\text{HCl}}}, \qquad V_{\text{NaOH}} = \frac{E_{\text{cycle}} \cdot RL_{\text{an}} \cdot EW_{\text{NaOH}}}{w_{\text{NaOH}} \cdot \rho_{\text{NaOH}}} \]

Parameter	Regime / Constraint	Typical Range
Cation Capacity	Strong Acid (H⁺ form)	1.4 – 1.8 eq/L
Anion Capacity	Strong Base (OH⁻ form)	0.8 – 1.2 eq/L
Superficial Velocity	Operational Flow	5 – 20 m/h
Bed Depth	Standard Design	1.0 – 2.0 m
Regeneration Level	Cation (Acid)	2.0 – 3.0 eq/eq
Regeneration Level	Anion (Base)	1.5 – 2.0 eq/eq

To ensure optimal performance and longevity of the demineralization system, process engineers must evaluate the following criteria:

Selectivity for monovalent versus divalent ions to achieve target ash reduction levels.
Physical stability of the resin beads to withstand frequent regeneration cycles and osmotic shock.
Compatibility with the specific pH profile of the whey feed stream.
Regeneration efficiency and the associated chemical consumption costs.

Maintaining strict sanitary conditions is critical when processing dairy streams. Recommended strategies include:

Operating the system at chilled temperatures, typically below 10 degrees Celsius, to inhibit bacterial proliferation.
Implementing automated Clean-in-Place (CIP) protocols using validated sanitizing agents.
Ensuring the system design minimizes dead legs and stagnant zones where biofilms can develop.
Monitoring the total plate count of the permeate and retentate streams at regular intervals.

The total solids content of the whey feed significantly affects the mass transfer kinetics and pressure drop across the resin bed. Key considerations include:

Higher feed concentrations increase the viscosity, which may lead to higher pressure drops and potential channeling within the column.
Optimal demineralization is generally achieved at lower solids concentrations, though this must be balanced against the energy costs of subsequent evaporation.
High concentrations can lead to increased fouling rates, requiring more frequent backwashing and regeneration cycles.

The regeneration cycle should be determined by monitoring the breakthrough curve of specific ions rather than relying solely on throughput volume. Engineers should:

Track the conductivity of the effluent to identify the point of ion leakage.
Analyze the composition of the effluent to ensure that the target demineralization percentage is maintained throughout the run.
Adjust the cycle time based on the specific mineral load of the incoming whey batch to prevent resin exhaustion.

Worked Example: Whey Demineralization by Two-Column Ion Exchange

A pilot plant must reduce the ash content of whey from 8% to 4% by weight using a two-column ion exchange system (cation exchanger in H⁺ form followed by anion exchanger in OH⁻ form). The feed flow is 100 L/h. To achieve exactly 50% removal without complex partial loading, a bypass strategy is adopted: 50 L/h is sent through the columns (where it is completely demineralized to 0% ash) and the remaining 50 L/h bypasses the columns. The two streams are blended to yield the target 4% ash. The system operates on an 8-hour service cycle.

Known Design Parameters

Total feed flow rate: \( \dot{V}_{\text{total}} = 100.0 \, \text{L/h} \)
Whey density: \( \rho = 1.03 \, \text{kg/L} \)
Feed ash mass fraction: \( X_{\text{feed}} = 0.08 \, \text{kg/kg} \)
Product ash mass fraction (after blending): \( X_{\text{prod}} = 0.04 \, \text{kg/kg} \)
Service cycle time: \( t_{\text{cycle}} = 8.0 \, \text{h} \)
Average equivalent weight of whey minerals: \( EW_{\text{avg}} = 0.06 \, \text{kg/eq} \)
Cation resin usable capacity: \( C_{\text{cat}} = 1.6 \, \text{eq/L} \)
Anion resin usable capacity: \( C_{\text{an}} = 1.0 \, \text{eq/L} \)
Bed height: \( H_{\text{bed}} = 1.5 \, \text{m} \)
Cation regeneration level: \( RL_{\text{cat}} = 2.5 \, \text{eq/eq} \)
Anion regeneration level: \( RL_{\text{an}} = 2.0 \, \text{eq/eq} \)
HCl concentration (36% w/w): \( w_{\text{HCl}} = 0.36 \), density: \( \rho_{\text{HCl}} = 1.18 \, \text{kg/L} \), equivalent weight: \( EW_{\text{HCl}} = 0.0365 \, \text{kg/eq} \)
NaOH concentration (50% w/w): \( w_{\text{NaOH}} = 0.50 \), density: \( \rho_{\text{NaOH}} = 1.53 \, \text{kg/L} \), equivalent weight: \( EW_{\text{NaOH}} = 0.04 \, \text{kg/eq} \)

Step-by-Step Calculation

Determine column feed flow. Since the bypass fraction is 50%, the flow through the columns is: \[ \dot{V}_{\text{col}} = \dot{V}_{\text{total}} \cdot 0.5 = 100.0 \cdot 0.5 = 50.0 \, \text{L/h} \] Note: The bypass derivation confirms \(f_{\text{col}} = (0.08-0.04)/(0.08-0.0) = 0.5\).
Mass removal rate of minerals. The columns remove all minerals from the treated stream: \[ \dot{m}_{\text{rem}} = (X_{\text{feed}} - 0.0) \cdot \dot{V}_{\text{col}} \cdot \rho = 0.08 \cdot 50.0 \cdot 1.03 = 4.12 \, \text{kg/h} \]
Equivalent removal rate. Using the average equivalent weight: \[ \dot{E}_{\text{rem}} = \frac{\dot{m}_{\text{rem}}}{EW_{\text{avg}}} = \frac{4.12}{0.06} = 68.667 \, \text{eq/h} \]
Equivalents removed per cycle. Over the 8-hour service cycle: \[ E_{\text{cycle}} = \dot{E}_{\text{rem}} \cdot t_{\text{cycle}} = 68.667 \cdot 8.0 = 549.333 \, \text{eq} \]
Resin volumes. Required volume for each column: \[ V_{\text{cat}} = \frac{E_{\text{cycle}}}{C_{\text{cat}}} = \frac{549.333}{1.6} = 343.333 \, \text{L} \] \[ V_{\text{an}} = \frac{E_{\text{cycle}}}{C_{\text{an}}} = \frac{549.333}{1.0} = 549.333 \, \text{L} \]
Column cross-sectional area. For a bed height of 1.5 m, the area (converting L to m³ with 1000 L/m³) is: \[ A_{\text{cat}} = \frac{V_{\text{cat}}}{H_{\text{bed}} \cdot 1000} = \frac{343.333}{1.5 \cdot 1000} = 0.229 \, \text{m}^2 \] \[ A_{\text{an}} = \frac{V_{\text{an}}}{H_{\text{bed}} \cdot 1000} = \frac{549.333}{1.5 \cdot 1000} = 0.366 \, \text{m}^2 \]
Superficial velocity check. Flow rate through columns is 50.0 L/h = 0.05 m³/h. \[ v_{\text{cat}} = \frac{0.05}{A_{\text{cat}}} = \frac{0.05}{0.229} = 0.218 \, \text{m/h} \] \[ v_{\text{an}} = \frac{0.05}{A_{\text{an}}} = \frac{0.05}{0.366} = 0.137 \, \text{m/h} \] These values are well below the typical recommended range of 5–20 m/h, indicating that the columns are oversized for this flow. In practice, one would reduce the bed height or increase the flow per column, but for this simplified design the volumes are accepted as a first estimate.
Regeneration requirements.
Cation column (HCl): \[ m_{\text{HCl}} = (E_{\text{cycle}} \cdot RL_{\text{cat}}) \cdot EW_{\text{HCl}} = (549.333 \cdot 2.5) \cdot 0.0365 = 50.127 \, \text{kg} \] \[ V_{\text{HCl}} = \frac{m_{\text{HCl}}}{w_{\text{HCl}} \cdot \rho_{\text{HCl}}} = \frac{50.127}{0.36 \cdot 1.18} = 118.001 \, \text{L} \]
Anion column (NaOH): \[ m_{\text{NaOH}} = (E_{\text{cycle}} \cdot RL_{\text{an}}) \cdot EW_{\text{NaOH}} = (549.333 \cdot 2.0) \cdot 0.04 = 43.947 \, \text{kg} \] \[ V_{\text{NaOH}} = \frac{m_{\text{NaOH}}}{w_{\text{NaOH}} \cdot \rho_{\text{NaOH}}} = \frac{43.947}{0.5 \cdot 1.53} = 57.447 \, \text{L} \]

Final Design Summary

Cation column volume: 343 L
Anion column volume: 549 L
Service cycle: 8 h
Regeneration chemicals per cycle:
- 36% HCl: 118 L (50.1 kg pure HCl)
- 50% NaOH: 57 L (43.9 kg pure NaOH)
Note: The calculated superficial velocities (0.218 m/h for cation, 0.137 m/h for anion) are far below the empirical optimum of 5–20 m/h. The volumes above are suitable only as a preliminary sizing; an economic design would adjust bed height or column configuration to achieve higher velocity.

"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