How to Size ZLD for Potato Starch Water: 2026 Engineering Guide
Sizing a ZLD system for potato starch wastewater starts with feed characterization — expect 8-15 m³ of combined process and wash water per tonne of starch, with COD 8,000-25,000 mg/L, BOD/COD ≈ 0.55-0.65, and TSS 3,000-8,000 mg/L during campaign season. The train is then sized in four stages: DAF/UF pretreatment (≤95% recovery), RO concentration (70-85% recovery), MVR/forced-circulation evaporation (≥95% recovery), and crystallizer finishing. Civil works and tankage should be sized for the design-peak month, while phase-1 process modules match the annual average.
What Potato Starch Wastewater Actually Looks Like on the Flow Sheet
Combined effluent from potato starch processing typically totals 8-15 m³ per tonne of processed potatoes, encompassing fruit water, process wash water, and belt/centrifuge drainage. This flow rate fluctuates, with early-campaign days generally yielding cleaner, lower-load effluent compared to late-campaign days which see increased protein and saprophytic organism concentrations (industry-typical observation, 2026). The single biggest design risk for potato starch ZLD is the seasonal 3-5x swing in both flow and load between peak campaign operation and wash-down days.
Industry-typical parameter ranges for raw potato starch wastewater include:
- COD: 8,000-25,000 mg/L
- BOD₅: 4,500-14,000 mg/L (BOD/COD ratio typically ≈ 0.55-0.65)
- TSS: 3,000-8,000 mg/L
- Total Nitrogen: 200-600 mg/L
- Total Phosphorus: 80-200 mg/L
- pH: 4.5-6.5
- Temperature: 25-45 °C
A ZLD designer must treat two parallel sidestreams separately: high-COD protein fruit water (a strong candidate for direct evaporation/MVR due to its high organic content and potential for protein recovery) and lower-COD wash water (more suitable for RO recovery due to its lower fouling potential).
| Parameter | Typical Range (Potato Starch Wastewater) | Unit |
|---|---|---|
| Flow | 8-15 m³/tonne processed potatoes | m³/t |
| COD | 8,000-25,000 | mg/L |
| BOD₅ | 4,500-14,000 | mg/L |
| TSS | 3,000-8,000 | mg/L |
| Total Nitrogen (TN) | 200-600 | mg/L |
| Total Phosphorus (TP) | 80-200 | mg/L |
| pH | 4.5-6.5 | - |
| Temperature | 25-45 | °C |
The 5-Step Sizing Framework Adapted to a Starch Plant
A robust ZLD sizing method typically follows 5 steps: characterize the feed, define recovery goals, allocate recovery by process stage, size buffering and utilities, and validate operability under upset conditions (per S1, 2025-08). For potato starch operations, this framework guides the design process to address the specific challenges of seasonal, high-load effluent.
- Step 1 — Characterize the Feed: Conduct a 7-day composite sampling campaign during the peak campaign period to capture the highest flow and load conditions. This provides the critical baseline data for all downstream sizing (S1, 2025-08).
- Step 2 — Define Recovery Goals: Aim for 95-98% overall water recovery for the ZLD system. This typically breaks down to 100% recovery on the cleaner wash water stream and at least 90% recovery on the high-COD fruit water stream if an MVR system is integrated.
- Step 3 — Allocate Recovery by Process Stage: Distribute the recovery targets across the treatment train. Pretreatment stages (screens, DAF, UF) should achieve ≥95% water recovery. The RO concentration stage typically handles 70-85% recovery per pass. The final MVR/crystallizer stage is designed for ≥95% recovery on its feed stream.
- Step 4 — Size Buffering and Utilities: An equalization basin should be sized to handle at least 24 hours of peak flow, plus an additional 2 weeks of RO concentrate buffer capacity. Utilities, particularly steam for evaporators and electrical power for RO high-pressure pumps and MVR compressors, must be sized to meet peak demand.
- Step 5 — Validate Operability: Model and simulate a 2-6 week stabilization period for the ZLD system, especially during the initial startup or after significant process changes, to ensure it reaches normal recovery and stable operation (S1, 2025-08). Crucially, validate the system's performance during the campaign-to-washdown transition, which represents a significant shift in feed characteristics.
Pretreatment Sizing: Screens, DAF, and UF for Starch Water
Rotary bar screens, sized at 110-120% of peak hourly flow, are essential headworks equipment that protect downstream equipment from large solids, fibers, and potato peel fragments in starch wastewater. These screens, with apertures typically ≤2 mm, directly protect starch-recovery hydrocyclones and prevent fouling of pumps and heat exchangers (Zhongsheng field data, 2026). Following screening, an equalization basin, sized for at least 24 hours HRT, is critical to dampen the significant COD and flow swings inherent to potato starch campaigns, allowing for pH correction (typically lime or NaOH to pH 6.5-7.2) before further treatment.
A starch-water DAF for TSS and protein removal unit is typically sized for a surface loading rate of 20-40 m/h. With proper coagulant and flocculant dosing via an automatic chemical dosing system, DAF can achieve 85-95% TSS removal and significant reduction in colloidal starch and protein, which are major foulants for subsequent membrane stages (Zhongsheng field data, 2026). The UF stage, often specified with a 50-100 kDa or 0.1-0.2 µm pore size, is designed to operate at a flux of 40-60 LMH with a 90-95% recovery per pass. This stage effectively removes residual protein, colloidal starch, and any emulsified oils not captured by the DAF, providing a high-quality feed for RO.
| Pretreatment Stage | Sizing Parameter | Typical Value/Range | Purpose |
|---|---|---|---|
| Rotary Bar Screen | Aperture size | ≤2 mm | Gross solids removal, equipment protection |
| Rotary Bar Screen | Capacity | 110-120% of peak hourly flow | Handle peak hydraulic loads |
| Equalization Basin | Hydraulic Retention Time (HRT) | ≥24 hours | Buffer flow and load swings |
| pH Correction | Target pH | 6.5-7.2 | Optimize coagulation, protect membranes |
| DAF Unit | Surface Loading Rate | 20-40 m/h | TSS, colloidal starch, and protein removal |
| DAF Unit | TSS Removal Efficiency | 85-95% | Achieve target effluent quality |
| Ultrafiltration (UF) | Pore Size/Cut-off | 50-100 kDa or 0.1-0.2 µm | Remove residual protein, colloidal starch, oil |
| Ultrafiltration (UF) | Flux Rate | 40-60 LMH | Optimize membrane productivity |
| Ultrafiltration (UF) | Recovery per pass | 90-95% | Maximize water recovery |
RO Concentration Stage: Recovery Split and Membrane Area
Two-pass brackish-water RO (BWRO) on pretreated starch water typically achieves 70-85% water recovery per pass, concentrating the permeate from the UF stage to a concentrate TDS of 40,000-80,000 mg/L (Zhongsheng field data, 2026). The design flux for the first-pass RO elements is typically 15-20 LMH, necessitating careful pretreatment to meet the SDI feed requirement of ≤5. If the SDI is borderline or fluctuates, an RO feed multi-media guard filter should be incorporated as a safeguard.
The total membrane area (A) for a starch-line RO concentration rack can be calculated using the formula: A = Q_permeate ÷ (flux × 24). For example, a 300 m³/d line designed for 80% recovery from a 375 m³/d feed (Q_permeate = 300 m³/d) with a target flux of 17 LMH would require approximately:
A = 300 m³/d ÷ (17 L/m²·h × 24 h/d) = 300 m³/d ÷ (0.017 m³/m²·h × 24 h/d) ≈ 735 m² of active membrane area.
Anticipate 15-25% replacement per year on the first-pass RO elements, especially when feed COD after UF remains >500 mg/L, as residual organics contribute to fouling and reduce membrane lifespan (Zhongsheng field data, 2026). An antiscalant dosing system and a CIP (Clean-In-Place) skid must be sized to support once-per-week cleaning cycles during peak campaign operation.
| RO Parameter | Typical Value/Range | Unit | Notes |
|---|---|---|---|
| System Type | Two-pass BWRO | - | For robust salt rejection and permeate quality |
| Flux Rate (1st Pass) | 15-20 | LMH | Optimized for starch wastewater to minimize fouling |
| Water Recovery (per pass) | 70-85 | % | Achievable recovery on pretreated feed |
| Concentrate TDS | 40,000-80,000 | mg/L | Feed to subsequent evaporation stage |
| SDI Feed Requirement | ≤5 | - | Critical for membrane longevity |
| Membrane Replacement Rate | 15-25 | % per year | Higher end if post-UF COD >500 mg/L |
| CIP Frequency (Campaign) | Once per week | - | Essential for maintaining flux |
Evaporator and Crystallizer Sizing: MVR, TVR, or Forced Circulation
MVR (mechanical vapor recompression) is the default evaporation technology for starch ZLD when feed TDS exceeds 80,000 mg/L, offering specific steam/electric energy consumption of 0.25-0.35 t/t evaporated (industry-typical, 2026). This low energy footprint makes MVR highly attractive for large-scale ZLD applications, especially where energy costs are a concern. TVR (thermal vapor recompression) may be preferred in plants where cheap, low-pressure steam is readily available from an existing steam header, offering a simpler mechanical design but higher thermal energy consumption.
For viscous, scaling-prone starch concentrates, a forced-circulation evaporator design is often selected, sized for an evaporation rate of 8-12 kg/m²·h (Zhongsheng field data, 2026). The high circulation velocity minimizes fouling and allows for efficient heat transfer with slurries. The final stage involves a crystallizer, typically a DTB (Draft Tube Baffle) or Oslo design, sized to produce a saleable or landfillable solid from the MVR bottoms at 60-65% moisture content. The clean condensate from the evaporator is crucial for closing the reuse loop back to process wash water, often requiring additional condensate polishing (e.g., activated carbon followed by a polishing RO or ion exchange) to meet specific process water quality requirements (per S3, 2024-07).
Recovery Allocation Table and CAPEX/OPEX Envelope
A 200-500 m³/day starch ZLD line typically requires a CAPEX budget of USD 1.8-3.5M in 2026, depending on the specific feed characteristics, desired automation level, and site conditions (Zhongsheng field data, 2026). The project budget is generally split as: pretreatment (15-20%), RO (25-30%), MVR/crystallizer (40-50%), and civil works/automation (10-15%). The operational expenditure (OPEX) is roughly USD 1.2-2.0 per m³ treated, with the dominant factors being electrical duty for the MVR compressor and RO high-pressure pumps, and RO membrane replacement (Zhongsheng field data, 2026).
When designing, stack civil works and pipe corridors for a 50% capacity debottleneck in phase 2, allowing for future growth without major structural modifications (S1, 2025-08). This future-proofing aligns with an overall industrial water management framework for 2026.
| Stage | Target Recovery | Influent Load (Example @ 300 m³/d) | Effluent Load (Example @ 300 m³/d) | Equipment Sizing Rule | Key Cost Driver |
|---|---|---|---|---|---|
| Pretreatment (Screens, DAF, UF) | ≥95% | COD: 15,000 mg/L, TSS: 5,000 mg/L | COD: 500-1,500 mg/L, TSS: <50 mg/L | DAF: 20-40 m/h surface loading; UF: 40-60 LMH flux | DAF unit size, UF membrane area, chemical dosing |
| RO Concentration | 70-85% | COD: 500-1,500 mg/L, TDS: 5,000-15,000 mg/L | Permeate TDS: <500 mg/L; Concentrate TDS: 40,000-80,000 mg/L | RO: 15-20 LMH flux; A = Q_permeate ÷ (flux × 24) | Membrane area, high-pressure pumps, membrane replacement |
| MVR/Forced-Circulation Evaporation | ≥95% | TDS: 40,000-80,000 mg/L | Condensate TDS: <100 mg/L; Concentrate Solids: 60-65% moisture | MVR: 0.25-0.35 t/t evaporated; FC Evap: 8-12 kg/m²·h | Heat exchange area, compressor size, materials of construction |
| Crystallizer Finishing | Solid Waste | Concentrate Solids | Dry or semi-dry solid (60-65% moisture) | Volume to handle MVR bottoms; DTB or Oslo design | Crystallizer vessel, dewatering equipment |
Frequently Asked Questions
What is the typical water reuse rate for a potato starch ZLD system?
A well-designed potato starch ZLD system typically achieves 95-98% overall water reuse. This recovery is distributed across the treatment train, with pretreatment stages (DAF/UF) recovering ≥95%, the RO concentration stage achieving 70-85% per pass, and the MVR/crystallizer finishing stage recovering ≥95% of its feed. For more on food-industry washwater pretreatment before ZLD, consult our guides.
Can MVR handle protein-rich fruit water directly, or does it need pretreatment?
While MVR can handle high-TDS streams, direct feed of protein-rich fruit water without adequate pretreatment (typically UF or DAF) carries a significant risk of severe fouling, foaming, and scaling within the evaporator. Proteins can denature and precipitate, while colloidal starch can polymerize, leading to reduced heat transfer efficiency, increased cleaning frequency, and potential operational shutdowns. Pretreatment is a prerequisite for reliable MVR operation.
How much RO membrane replacement should be budgeted per year?
For potato starch ZLD systems, budgeting 15-25% replacement per year on the first-pass RO elements is a pragmatic approach. This higher-than-average replacement rate is primarily driven by the residual COD in the RO feed, even after UF. If the COD after UF consistently remains above 500 mg/L, expect to be at the upper end of this replacement range due to increased organic fouling.
Is ZLD for a small (<200 m³/d) starch plant economically justified in 2026?
For small starch plants (<200 m³/d), the economic justification for ZLD can be challenging. MVR evaporators, a core component of most ZLD systems, have a minimum economic scale, and their CAPEX can be disproportionately high for lower flow rates. While ZLD offers environmental benefits and avoids discharge costs, the CAPEX threshold of USD 1.8-3.5M for a 200-500 m³/d system implies that smaller plants might find it difficult to achieve a favorable return on investment unless local discharge regulations are extremely stringent or water scarcity costs are very high. You can compare this to ZLD sizing for paint booth curtain water, which has different feed characteristics but similar economic considerations at smaller scales.
How long does a starch ZLD line take to stabilize after startup?
A starch ZLD line typically requires a 2-6 week stabilization period after startup before it reaches normal recovery rates and consistent permeate quality (S1, 2025-08). This extended period is necessary due to the complex, high-COD nature of potato starch wastewater, which demands careful optimization of chemical dosing, membrane cleaning cycles, and evaporator operating parameters to achieve steady-state performance and mitigate fouling.
