What Makes Rack Wash Water Different from Municipal Sewage
Rack wash water — the combined discharge from bottle, crate, pallet, and conveyor washers in beverage, dairy, brewery, and food plants — is a fundamentally industrial stream wrapped in a municipal-sized volume envelope. Baseline COD after dilution sits at 50–500 mg/L, but during CIP (clean-in-place) cycles the same pipework can deliver 2,000 mg/L COD peaks over 10–20 minute windows. TSS runs 200–800 mg/L on a routine basis, composed of label glue, broken glass fragments, cardboard fiber from crate returns, and biofilm carryover from warm wash baths.
Three operational traits make this stream defeat conventional activated-sludge plants. First, intermittency: rack washers cycle 3–6 times per shift, so peak-hour flow routinely reaches 1.5–2.0× the 24-hour average. Second, surfactant and caustic spikes from CIP create foam events that collapse gravity clarifiers; MBR submerged membranes tolerate foam but require anti-foam dosing above ~50 mg/L surfactant to keep the biotank surface stable. Third, elevated temperature — 25–40 °C from hot CIP rinses — lowers oxygen solubility and must be reflected in aeration sizing with an α-factor of 0.85–0.95. The result is a stream that looks like sewage on a flow meter but fouls membranes and biomass like a true industrial effluent.
Step 1 — Characterize and Equalize the Flow
The first design move is converting a chaotic shift pattern into a stable feed the biotank can actually treat. Map the shift schedule first: rack washers cycle 3–6× per shift with 10–20 minute wash bursts, so the design flow must be the 95th-percentile hourly flow, not the 24-hour average. Apply an equalization factor of 1.5–2.0× to the daily average before sizing downstream equipment — a 200 m³/day average becomes a 300–400 m³/day peak design flow that the EQ tank must absorb.
Equalization tank HRT should be 8–12 hours minimum to flatten CIP peaks; extend to 18–24 hours if downstream processes are sensitive to temperature swings or if the plant operates intermittent shifts with overnight idle periods. Specify the tank as mixed, aerated, and level-controlled, with PLC-driven transfer pumping to the biotank at a steady rate. On the EQ tank inlet, install a rotary mechanical bar screen with a 3 mm aperture (or finer) — a GX rotary bar screen on the inlet strips rags, label fragments, and broken glass before they reach transfer pumps and clog the DAF or membrane cassette headers downstream. Skipping this screen is the most common cause of premature pump wear in rack-wash MBR installations.
Step 2 — Pretreat for FOG, Colloids, and Foam
Skipping DAF before an MBR is the single most common cause of premature membrane fouling on industrial streams. A DAF unit removes more than 90% of FOG, emulsified oils, and colloidal surfactants in 15–25 minutes of hydraulic residence, and surface loading should be designed at 5–10 m/h for oily industrial streams. For most rack-wash applications in the 4–300 m³/h band, a ZSQ DAF unit provides the right capacity envelope.
Float handling matters as much as float removal. Route DAF sludge to a plate-and-frame filter press for dewatering to 18–25% DS — this skips the digester entirely and reduces hauling cost, which is significant when the FOG fraction of the float is 5–15% of total sludge mass. If caustic or acid CIP is present in the wash stream, install pH adjustment (target 6.5–8.5) ahead of the biotank to protect nitrifiers and prevent membrane hydrolysis. An automatic chemical dosing system with closed-loop pH control is the standard approach. Without DAF, MBR membrane life typically drops 40–60% in rack-wash service — a 7-year membrane cassette becomes a 3–4 year cassette, and the OPEX delta pays for the DAF inside 18 months at most flow rates (Zhongsheng field data, 2026).
| Pretreatment Stage | Design Parameter | Typical Range | Purpose |
|---|---|---|---|
| Bar screen | Aperture | 2–3 mm | Rags, labels, glass |
| DAF | Surface loading | 5–10 m/h | FOG, emulsified oils, colloids |
| DAF | HRT | 15–25 min | Contact and rise time |
| pH adjustment | Target range | 6.5–8.5 | Protect biomass and membrane |
| Filter press | Cake DS | 18–25% | Sludge volume reduction |
Step 3 — Size the Biological Tank (MLSS, HRT, SRT)
Biotank sizing for a rack-wash MBR converges on three parameters: MLSS, HRT, and SRT. Target MLSS is 8,000–10,000 mg/L — higher than conventional activated sludge because MBR tolerates it, but capped below 12,000 mg/L to keep mixed-liquor viscosity and transmembrane pressure manageable. HRT runs 6–10 hours for rack wash at 8,000–10,000 mg/L MLSS, shorter than municipal MBR because the substrate is highly biodegradable. SRT should be 20–30 days to retain slow-growing nitrifiers and prevent washout of biomass acclimated to recurring surfactant pulses. The F/M ratio target is 0.08–0.15 kg BOD/kg MLSS·day for combined carbon removal and partial nitrification.
Worked example for a 200 m³/day plant: Q = 200 m³/day, S₀ = 600 mg/L BOD, MLSS = 9,000 mg/L, F/M = 0.15. Biotank volume = (Q × S₀) / (F/M × MLSS) = (200 × 0.6) / (0.15 × 9.0) = 88.9 m³, round to 90 m³. At an HRT of 10.8 hours that is well inside the 6–10 h design band. Dissolved-oxygen control should hold 2.0–3.0 mg/L in the biotank and 4.0–6.0 mg/L in the membrane tank — the elevated DO in the membrane tank supports simultaneous nitrification and membrane scouring from the same coarse-bubble diffuser grid.
| Parameter | Design Value | Operating Range | Note |
|---|---|---|---|
| MLSS | 9,000 mg/L | 8,000–10,000 | Cap below 12,000 for TMP |
| HRT | 10 h | 6–10 h | Shorter than municipal MBR |
| SRT | 25 d | 20–30 d | Retain nitrifiers |
| F/M | 0.12 | 0.08–0.15 | Carbon + partial nitrification |
| DO biotank | 2.5 mg/L | 2.0–3.0 | Aeration control loop |
| DO membrane tank | 5.0 mg/L | 4.0–6.0 | Scouring + nitrification |
Step 4 — Calculate Membrane Area and Flux
Design net flux for PVDF flat-sheet at 8,000–10,000 mg/L MLSS is 12–18 LMH; drop to 8–12 LMH if FOG carryover into the membrane tank exceeds 30 mg/L — the pretreatment chain in Step 2 is what keeps the design in the upper band. Continuing the worked example: Q = 200 m³/day = 8.33 m³/h; at 15 LMH net flux, required membrane area = (8.33 × 1000) / 15 = 555 m², round up to 560 m². Size 1.2× the calculated area (so 670 m² installed) so a single cassette can be isolated for cleaning or replacement without breaching the daily discharge permit.
Operating regime is 8 minutes filtration followed by 2 minutes relaxation, with no backflush on submerged flat-sheet systems; chemical cleaning (typically 1,000–2,000 mg/L NaOCl + 0.5–1.0% citric acid) runs every 30–90 days in normal operation. The 2026 DF-series PVDF flat-sheet module family — 0.1 µm pore, 80–225 m² per module, 32–135 m³/day per module — lets a designer pick module count directly from capacity. The flux-MLSS trade-off is the one table every reviewer will ask to see:
| MLSS (mg/L) | Net Flux Range (LMH) | Operating Implication |
|---|---|---|
| 6,000 | 20–25 | Light load, low viscosity |
| 8,000 | 16–20 | Standard rack-wash target |
| 10,000 | 12–16 | Heavier FOG carryover |
| 12,000 | 8–12 | TMP climb risk, avoid for design |
For a 670 m² installed area, the DF-series PVDF flat-sheet module cassette selection is 3× 225 m² units, or 4× 200 m² units, depending on skid layout and redundancy preference.
Step 5 — Aeration, Sludge Yield, and OPEX Anchors
Aeration has two distinct duties in a submerged MBR: biological oxygen delivery and membrane-tank scouring. Membrane-tank air demand is 0.3–0.5 Nm³ air per m² membrane area per hour, delivered as continuous coarse bubble from an integrated aeration box below each cassette. Biological air demand is 1.4–1.8 kg O₂/kg BOD removed, calculated with a rack-wash α-factor of 0.85 to account for the elevated temperature and surfactant content. For the 200 m³/day example at 600 mg/L BOD influent and 90% removal, biological air demand lands near 180–230 Nm³/h; add membrane scouring of 200–335 Nm³/h at 670 m² for a total blower duty of 380–565 Nm³/h.
Observed sludge yield at SRT 25 days is 0.25–0.35 kg TSS/kg BOD removed; route waste activated sludge to the plate-and-frame filter press, not a belt press — the higher DS target (18–25%) and enclosed operation suit a food-plant footprint. Total specific aeration demand (SAD) for the whole system is 8–12 Nm³ air per m³ permeate. A flat-sheet submerged MBR runs 10–20× lower energy than external cross-flow designs because the recirculation pump that hollow-fiber sidestream MBRs require is eliminated entirely.
Flat-Sheet vs. Hollow-Fiber MBR: Choosing the Right Module Type
The 2026 specification debate for industrial MBRs resolves around feed-quality variability. Flat-sheet PVDF modules tolerate FOG and grit better than hollow fibers, allow individual element replacement, hold a 0.1 µm pore, and consume less energy because no recirculation pump is required. Hollow-fiber cassettes offer higher packing density (lower footprint) and are favored for well-buffered municipal flows, but they foul faster from fibrous debris and grease carryover — exactly what a rack-wash stream delivers during a label flush or a CIP transition.
Decision rule: choose flat-sheet when upstream screening is coarser than 2 mm, FOG into the membrane tank exceeds 30 mg/L, or feed shows TSS spikes above 500 mg/L. Choose hollow-fiber only when the feed is pre-filtered to under 50 mg/L TSS and is steady — rarely the case in rack-wash service. For the 200 m³/day example at 670 m² installed, a flat-sheet configuration is 3× 225 m² DF-series cassettes, while an equivalent hollow-fiber train would be 4–5 modules of a different family, but the flat-sheet choice is the lower-risk specification for this stream. The two module types are not interchangeable in OPEX: a properly pretreated flat-sheet MBR runs chemical cleaning every 30–90 days; a hollow-fiber train on the same feed typically requires cleaning every 14–21 days, doubling chemical cost (Zhongsheng field data, 2026). For an integrated MBR system on a rack-wash stream, flat-sheet is the default 2026 specification.
| Criterion | Flat-Sheet PVDF | Hollow-Fiber |
|---|---|---|
| FOG tolerance | High (>30 mg/L into membrane tank) | Low (best <15 mg/L) |
| TSS spike tolerance | High (handles 500+ mg/L events) | Low (clogging risk) |
| Element replacement | Individual sheets | Whole cassette typical |
| Energy use | No recirculation pump | Recirculation pump required |
| Footprint | Larger per m² | Higher packing density |
| Best-fit stream | Industrial, variable feed | Municipal, steady feed |
Frequently Asked Questions
What flux is realistic for rack wash MBR in 2026?
12–18 LMH net flux is the realistic design band for PVDF flat-sheet at 8,000–10,000 mg/L MLSS. Treat 20 LMH as a ceiling, not a target — pushing past 18 LMH on a rack-wash stream invites rapid TMP climb and more frequent chemical cleaning.
Do I still need a DAF before the MBR if my rack wash is dilute?
Yes. FOG carryover is what kills membranes, not COD. A dilute stream (under 200 mg/L COD) with even 20–40 mg/L emulsified oil will foul a flat-sheet cassette in weeks without DAF. DAF removes the FOG fraction cheaply; the membrane is the wrong place to try.
How often will the membranes need chemical cleaning?
Every 30–90 days in normal operation with DAF pretreatment in place. If the DAF is bypassed or under-performing, cleaning intervals collapse to weekly — at that point the membrane life drops from 7 years to 3 years, which dominates the OPEX calculation.
Can MBR permeate go straight to RO for reuse?
Yes. MBR permeate at under 1 µm is excellent RO feed, with typical turbidity under 0.5 NTU and SDI₁₅ under 3. Install a 5 µm cartridge guard filter ahead of the RO high-pressure pump as standard practice; skipping the guard filter is the most common cause of RO fouling in reuse trains.
What is the typical CAPEX split for a 200 m³/day rack-wash MBR?
For a turnkey 200 m³/day system, bioreactor civil works typically accounts for 25%, membranes 35%, pretreatment (screens, DAF, dosing) 20%, automation and instrumentation 10%, and commissioning 10%. The membrane cassette share is the single largest line item, which is why pretreatment and SRT discipline pay back through extended membrane life rather than through reduced equipment cost. For comparison, the same numbers shift for sizing an MBR for white water, where the COD load is higher and the biotank share grows.
