How to Size MBR for White Water: 2026 Engineering Specs & Step-by-Step Guide

Why White Water Breaks a Standard MBR Sizing

White water is clarified loop water from paper, board, or deinking lines, and it carries a fines fraction that municipal MBR flux curves never see. Typical influent runs SS 300–3,000 mg/L, COD 500–2,500 mg/L, and temperatures of 35–55 °C, with pH drifting between 5.5 and 7.5 across grade changes (Zhongsheng field data, 2026). The failure mode is mechanical, not biological: fine fibers, talc, and stickies lodge in 0.1–0.4 μm membrane pores, and transmembrane pressure climbs from a clean-membrane ~5 kPa baseline to 30+ kPa within 3–7 days when DAF or fine screening is skipped. A 2014 MDPI Water review confirmed that MF and loose UF membranes in MBRs are not size-exclusion barriers for the smallest particles and viruses, so any "0.1 μm rejection" claim is a marketing floor rather than a physical barrier (MDPI Water, 2014). For a fiber-bearing stream, that limit is exactly why pretreatment discipline decides whether the system runs for 5 years or 5 months. The five-step method below starts from that risk profile rather than a generic MBR spreadsheet, and the membrane module selected for the worked example is a DF-series PVDF flat-sheet cassette (80–225 m² per unit) because flat-sheet geometry tolerates fiber loading and air-scour recovery better than hollow fiber on this duty.

Step 1 — Characterize the White Water Stream

The first input to any defensible MBR sizing memo is a parameter set that survives a regulator's review. A single grab sample will under-size the system by 30–50% on a paper loop where peak-to-average flow ratios routinely hit 2–5× during grade changes, start-ups, and felt showers. The minimum sampling program is a 7-day composite covering at least one full production cycle, with discrete grabs on startup, steady-state, grade change, and shutdown so the engineer can bracket the worst case.

Engineers should also record the upstream fiber source — virgin pulp, OCC, mechanical pulp, or deinked — because OCC and deinking lines carry the stickies load that drives DAF chemistry, while virgin pulp lines run cleaner but hotter. A GX rotary bar screen with 1–3 mm openings is the standard first cut, removing 20–40% of TSS before the equalization tank.

Step 2 — Define the Discharge or Reuse Target

The effluent envelope drives whether the membrane area or the biological tank is the binding constraint. Two target regimes dominate 2026 factory white water projects. Discharge-to-sewer in China typically tracks GB 3544 limits for the paper industry: COD ≤500 mg/L and SS ≤400 mg/L on a daily-discharge basis, with pH 6–9 (per China GB 3544-2008, still the operative standard in 2026 enforcement). In-plant reuse for shower water, dilution, or chemical prep targets a tighter envelope: TSS ≤5 mg/L, turbidity ≤2 NTU, COD ≤50 mg/L — a band that the integrated MBR system (10–2,000 m³/day) is rated to deliver at the <1 μm filtration claim on the cassette spec.

The math diverges at this point. A discharge target is biology-rate-limited: the membrane has excess capacity and runs at the lower end of the 12–18 L/m²·h flux window, so tank volume and aeration dominate CAPEX. A reuse target is membrane-rate-limited: the membrane area must push flux to the upper end and absorb SS surges through the safety factor, which is why reuse MBRs typically demand 20–40% more cassette area than a discharge-only system on the same flow (Zhongsheng field data, 2026). Engineers who skip this step routinely over-size biology and under-size membranes, then watch TMP ramp up within weeks.

Step 3 — Size the Biological Tank (HRT, MLSS, F/M)

White water behaves differently from municipal sewage in the bioreactor: most of the COD is soluble BOD from dissolved organics, hemicelluloses, and starch, not particulate BOD, so the mixed liquor can run hotter (higher MLSS) without the settleability crisis that kills a conventional activated sludge tank. Standard 2026 design bands are HRT 6–10 h, MLSS 8,000–12,000 mg/L, and F/M 0.08–0.15 kg BOD/kg MLSS·day. Higher MLSS than municipal MBR (typically 6,000–10,000 mg/L) is feasible because the F/M window is comfortable on soluble BOD.

Worked example, 1,000 m³/day average flow: target MLSS 10,000 mg/L, HRT 8 h. Tank volume V = Q × HRT = 1,000 m³/day × (8/24) day = 333 m³. Split the tank 60% aerobic / 40% anoxic to strip color and reduce residual peroxide on post-bleach streams. With influent BOD 600 mg/L (mid-range) and MLSS 10,000 mg/L, the F/M ratio lands at 0.18 kg BOD/kg MLSS·day — at the top of the window — so a high-efficiency sedimentation tank or DAF clarifier upstream is what keeps the system inside the operating band during grade-change SS surges of 2–3× the average.

Engineers who skip the F/M check routinely discover that MLSS drifts above 14,000 mg/L on a deinking line, viscosity rises, and oxygen transfer efficiency falls off a cliff. A pre-clarifier protects the F/M ratio more than any aeration upgrade.

Step 4 — Size the Membrane Area (Flux, TMP, Modules)

Membrane area is the engineering payoff. The formula is: A (m²) = Q (L/h) ÷ J (L/m²·h), where design flux J sits in the 12–18 L/m²·h window on 0.1 μm PVDF flat sheet at 8,000–12,000 mg/L MLSS (Zhongsheng field data, 2026). Apply a 1.25–1.5 safety factor to absorb SS surges, scheduled chemical cleaning downtime (CIP every 30–60 days), and cold-weather viscosity correction if the loop runs at the low end of the temperature band.

Worked example, 1,000 m³/day: Q = 1,000,000 L/h ÷ 24 = 41,667 L/h. Design flux at 10,000 mg/L MLSS = 15 L/m²·h. Nominal area = 41,667 ÷ 15 = 2,778 m². Apply 1.35 safety factor → 3,750 m². With a DF-series PVDF flat-sheet module rated 80–225 m² per cassette (typical 32–135 m³/day per module on white water), the count lands at ~22 cassettes for the 1,000 m³/day base case, scaling linearly to ±20% with the SS surge profile.

Submerged flat-sheet versus sidestream hollow fiber: flat sheet wins on white water because the open channel geometry passes fibers without lodging them, the air-scour pattern is uniform across the panel, and chemical cleaning is a soak rather than a backwash cycle. Sidestream hollow fiber only earns its keep above ~5,000 m³/day on relatively clean streams. The 2014 MDPI Water finding that MF/UF in MBRs are not virus-size barriers (MDPI Water, 2014) also means virus log-removal credit cannot be claimed on the membrane alone — a downstream UV or chlorine step is required if the reuse spec touches a food-grade or process-water contact application.

Step 5 — Aeration, Sludge Wasting, and Pretreatment Selection

Aeration demand on white water has two components. Process oxygen: 0.3–0.5 kg O₂ per kg BOD removed at the F/M band above. Membrane scour air: 0.2–0.3 m³/m² membrane area·h, delivered as coarse-bubble diffusers directly under each cassette. The DF-series cassette's integrated aeration box combines both duties on a single blower train, which is what delivers the 60% footprint saving over a conventional activated-sludge-plus-clarifier line on space-constrained factory sites (Zhongsheng field data, 2026).

Sludge yield on white water runs 0.2–0.35 kg TSS per kg BOD removed at SRT 30–60 days — low because a fraction of the COD is already mineralized and the SRT is long enough to push the yield coefficient down. The wasted MLSS at 0.5–1.5% dry solids feeds directly to a plate-and-frame filter press for wasted MLSS dewatering, producing a 22–28% DS cake suitable for off-site incineration or landfill. Pretreatment selection is the part most often undersized: a GX rotary bar screen at 1–3 mm openings removes fiber and rag, then a ZSQ DAF unit for fiber and colloidal removal (4–300 m³/h) strips stickies, fines, and colloidal TSS that would otherwise cake the membrane surface. Skipping DAF is the single most common cause of premature membrane replacement on white water loops — the membrane area is sized for the pretreated stream, and the safety factor does not cover raw fiber carryover.

Worked Sizing Worksheet (1,000 m³/day Example)

Each row maps back to the worked math in Steps 3 and 4 so a procurement reviewer or a regulator can audit the trail in a single pass. The full envelope fits within the integrated MBR system (10–2,000 m³/day) product range without parallel trains.

Reuse Versus Discharge: When the MBR Pays Back

A discharge-only MBR caps flux on the low end because biology is the bottleneck and the membrane has slack capacity. A reuse MBR pushes membrane area up by 20–40% to meet the TSS ≤5 mg/L / turbidity ≤2 NTU envelope, and that premium pays back through purchased-water displacement at 60–85% recovery. The 2026 payback heuristic is: the reuse MBR CAPEX premium clears inside 18–36 months when fresh process water cost exceeds ~$1.50/m³ and inlet flow is steady above ~500 m³/day (Zhongsheng field data, 2026). Below that threshold — for example, a small OCC line paying $0.80/m³ for make-up water — a discharge-only MBR with a smaller cassette count is the economic answer. Where reuse spec demands conductivity or salt limits that MBR alone cannot deliver, pair the MBR with an RO polish step and reset the reuse envelope to 80–95% recovery at <50 μS/cm. For plants weighing a parallel comparison on automotive or coating wastewater, see the related guides on MBR sizing for paint booth curtain water and MBR sizing for e-coat UF reject — the procedure generalizes, but the pretreatment chemistry does not.

Frequently Asked Questions

What flux should I use for MBR sizing on white water?

Design flux on 0.1 μm PVDF flat sheet sits at 12–18 L/m²·h across the 8,000–12,000 mg/L MLSS operating window, with 15 L/m²·h as the mid-band value for a 1,000 m³/day paper-machine loop at 10,000 mg/L MLSS. Below 12 L/m²·h the system is over-sized for the biology; above 18 L/m²·h the TMP ramp becomes uneconomic without aggressive CIP.

Why is DAF pretreatment mandatory before MBR on white water?

DAF removes 70–90% of colloidal fines, stickies, and emulsified solids that would otherwise lodge in 0.1 μm pores and drive TMP from 5 kPa to 30+ kPa within days. A ZSQ DAF unit (4–300 m³/h) sized at 15–25 m³/h per m³ of membrane area is the standard envelope. Skipping DAF is the most common cause of premature membrane replacement on this duty.

Can I claim virus log-removal credit on the 0.1 μm membrane?

No. The 2014 MDPI Water review confirmed that MF/UF in MBRs are not size-exclusion barriers for the smallest viruses (MDPI Water, 2014). Any reuse spec that requires virus reduction needs a downstream UV or chlorine step; the membrane provides the TSS envelope, not the pathogen barrier.

What is the typical SRT and sludge yield for white water MBR?

SRT runs 30–60 days and sludge yield sits at 0.2–0.35 kg TSS per kg BOD removed — lower than municipal MBR because a fraction of the white water COD is already partially oxidized. The wasted MLSS at 0.5–1.5% DS feeds a plate-and-frame filter press producing a 22–28% DS cake.

How do I decide between flat-sheet and hollow-fiber MBR on white water?

Flat-sheet PVDF is the default below ~5,000 m³/day because the open channel passes fibers without plugging and the air-scour pattern is uniform. Hollow-fiber sidestream only earns its keep above that flow on relatively clean streams. For paper-machine loops with stickies and fines, the DF-series PVDF flat-sheet module (80–225 m² per cassette) is the engineering choice.

Further Reading

• bleach E-stage pretreatment before MBR