Why dairy wastewater is a special case for biological treatment
Dairy effluent is biologically treatable but punishing to operate. It carries high BOD and COD, a low C/N ratio that limits conventional nitrification-denitrification, and periodic fat, oil, and grease (FOG) plus clean-in-place (CIP) surges that hit the basin as shock loads. A 2023 IFAS-SBR study (S1) on real wastewater from a 300 m³/d Ukrainian dairy documented a C/N of only 3.0 to 3.5, which is the operating reason low-DO simultaneous nitrification-denitrification is often used in place of separate anoxic and aerobic basins.
The same study showed that suspended-growth SBRs without biofilm carriers plateaued at about 76 percent TN removal under that C/N. A 2013 lab study on simulated dairy prepared from 4 g/L milk powder (S5) reported a feed COD of 3900 mg/L and TKN of 113.18 mg/L, and concluded that flow and load equalisation ahead of the SBR is the single most effective protection against product-change and CIP slugs. Characterisation work in a US EPA HERO lab study (S4) measured milking wastewater at average BOD 1608 mg/L and soluble COD 1010 mg/L, with most phosphorus soluble and almost all suspended solids volatile, a profile that suits biological treatment but penalises any reactor with poor settleability. The practical consequence is that any dairy-duty biological step must be preceded by DAF pre-treatment for FOG and screening, and the basis-of-design must size for shock loads, not for the daily average. For nutrient-tight permits, the broader advanced nutrient removal guide shows how biological P and sidestream treatment fit around an SBR core.
Anatomy of a dairy-duty SBR cycle
The sequencing batch reactor runs fill, react, settle, decant, and idle in a single tank, and the way those phases are sequenced is what makes a dairy SBR work. S4's 8 h cycle, with 1 h anaerobic, 3 h aerobic, 2 h anoxic, 1 h aerobic, 50 min settle, and 10 min decant, is the most cited full-phase template in the dairy literature and should be treated as a starting point rather than a copy. A leaner 2016 SBR study (S3) used 30 min of anoxic stir followed by 3 h of aerobic react on simulated dairy at MLSS 2400 to 2600 mg/L and SRT 13 to 17 days, and still produced effluent COD 90 mg/L, NH4-N 9 mg/L, and PO4 6 mg/L. S1 ran a 6 h cycle at 20 °C and MLSS 3.5 g/L, with the aerobic phase held at DO 0.8 to 1.2 mg/L so that simultaneous nitrification-denitrification developed inside the biofilm rather than across separate basins. S5 showed that the 1 h fill and 1 h anoxic stir after fill are the two cycle adjustments that most affect COD and TKN removal, and that extending HRT beyond 24 h produced no statistically significant COD gain. For screening, an upstream rotary bar screen keeps rags and product solids from interfering with the decanter and the air-diffuser pattern, both of which drive cycle stability.
Design parameters the research actually supports
The numeric envelope below is what four primary dairy-SBR studies actually reported, and it is the only band a basis-of-design should be drafted from. HRT ranged from 15 to 30 h in the simulated-dairy study (S5), with 24 h the practical optimum, and 5.3 days in the EPA HERO lab study (S4) at the long-residence end. SRT was 13 to 17 days in S3 and 40 days in S4; the higher SRT in S4 is what enabled near-complete nitrification at lower mixed-liquor temperatures. MLSS was 2400 to 2600 mg/L in S3 and 3500 mg/L in S1, both inside the band that produced compliant effluent in their respective studies. DO in the aerobic phase was 2 to 3 mg/L in earlier aerobic-SBR work cited by S5 (Mohseni-Bandpi and Bazari), but S1 showed that simultaneous nitrification-denitrification remains effective down to DO 0.8 to 1.2 mg/L and cut aeration energy by 37.5 percent. Organic loading was fixed at 0.125 kg COD per kg MLSS per day in S4, beyond which COD removal collapsed. For chemical trim, an automatic chemical dosing skid handles phosphorus precipitation when the SBR effluent PO4 sits above the permit.
| Parameter | Range in the dairy-SBR studies | Source |
|---|---|---|
| HRT | 15–30 h (simulated dairy); 5.3 d (lab milking wastewater) | S5; S4 |
| SRT | 13–17 d; 40 d for near-complete nitrification | S3; S4 |
| MLSS | 2400–2600 mg/L; 3500 mg/L | S3; S1 |
| DO (aerobic phase) | 2–3 mg/L conventional; 0.8–1.2 mg/L for low-DO SND | S5; S1 |
| Organic loading | 0.125 kg COD / kg MLSS / d optimum | S4 |
| Cycle time | 6 h (low-DO IFAS); 8 h (full-phase template); 4–5 h (lean) | S1; S4; S3 |
| Temperature | 20 °C in the IFAS-SBR dataset | S1 |
Removal performance you can put in a discharge report
The headline numbers from the four studies line up well enough to put in a regulator-facing table. For COD, S3 reported effluent around 90 mg/L; S4 reported greater than 99 percent removal with effluent COD below 55 mg/L; S5 reported 96.7 percent at HRT 24 h on simulated dairy with feed COD near 3900 mg/L. For BOD, S4 showed BOD dropping below 25 mg/L at the optimal organic loading rate, comfortably below typical 30 mg/L discharge ceilings. For nitrogen, S4 achieved 96.7 percent TKN and 92.4 percent NH4-N removal; S1's IFAS-SBR reached 86.5 percent TN at C/N 3.0 to 3.5; S5's stand-alone aerobic SBR plateaued at 76.7 percent TKN, and that gap is the cost of skipping biofilm carriers at low C/N. For phosphorus, S4 hit 95.9 percent TP and 94.0 percent dissolved phosphorus, while S3's effluent PO4 sat around 6 mg/L, which is often above a 2 mg/L environmental limit and is the single most common reason dairies add a chemical precipitation or enhanced biological P step. S4 also measured 96.6 percent TSS removal with SVI consistently under 100 mL/g, indicating good settleability when FOG is removed upstream. The downstream solids train is usually a filter press for waste activated sludge, sized off the wasted-AS mass and the dryness target. For plants near a tight sewer, the food and beverage pretreatment compliance guide shows how pretreatment ceilings are set, which the SBR envelope must clear.
Where stand-alone SBR falls short on a real dairy line
Stand-alone SBR is rarely the end of the line for a dairy plant with a numeric discharge permit. TSS in SBR effluent is typically not low enough for direct reuse or for tight sewer limits, so MBR or DAF polishing is the usual fix. Effluent TP often misses stringent limits even when COD is compliant, and S3's 6 mg/L PO4 versus a 2 mg/L target is a typical gap that forces a coagulant dose or a biological-P upgrade. S1's data shows that at C/N below 4, a stand-alone suspended-growth SBR struggles to reach the 70 to 80 percent TN benchmark without biofilm carriers or external carbon. A 2026 microalgae-bacteria SBR study (S2) reached 99 percent COD removal with NH4-N below the detection limit in stage 2, but that process relies on photosynthetic oxygen supply, requires light, and is not yet a drop-in replacement for conventional SBR in a covered dairy plant. The practical conclusion is that an integrated MBR system downstream is the right answer when the limit is TSS or ammonia-driven toxicity, and a DAF polish is the right answer when the limit is FOG carryover or partial phosphorus removal.
Choosing between SBR, MBR, and an SBR+DAF train
The decision should be driven by the discharge or reuse target first, cycle parameters second, and reactor volume last. A stand-alone SBR has the lowest CAPEX and the smallest electrical scope, but it cannot typically meet TSS below about 20 to 30 mg/L or TP below about 2 mg/L without chemical polishing. Adding an MBR downstream of the SBR collapses footprint by combining the SBR with submerged PVDF membranes, and produces near-reuse effluent, at the cost of higher aeration energy and membrane maintenance. Adding a DAF downstream of the SBR is the most common retrofit path when the limit is TSS, FOG carryover, or partial phosphorus removal, because DAF handles the hydraulic surges that would blind an MBR. A side-by-side view of how these three trains compare for a dairy plant is shown below; the trade-offs are real and should be priced into the basis-of-design rather than discovered at commissioning.
| Criterion | Stand-alone SBR | SBR + MBR | SBR + DAF |
|---|---|---|---|
| Typical TSS in effluent | 20–30 mg/L | < 5 mg/L (near-reuse) | 10–20 mg/L |
| TP removal without coagulant | Often misses 2 mg/L target | Misses without chemical dose | Partial; coagulant usually required |
| Footprint | Largest of the three | Smallest (SBR + membrane cassette) | Intermediate (adds a flotation tank) |
| Aeration energy | Baseline; can drop 37.5% with low-DO SND (S1) | Higher (membrane scour) | Baseline for SBR plus saturator for DAF |
| Hydraulic surge tolerance | Limited by equalisation | Limited by membrane flux | Good (DAF handles shock loads) |
| Best-fit driver | Lowest CAPEX, modest limits | Reuse or tight TSS ceiling | FOG carryover, partial P, existing SBR retrofit |
For plants also weighing moving-bed biofilm carriers against an SBR core, the MBBR for beverage wastewater guide covers the hybrid configurations. Membrane module selection typically narrows to an MBR membrane bioreactor module sized on peak daily flow and the design flux.
Operating risks and how to derisk them
Filamentous bulking is the most common dairy SBR failure mode and is driven by high carbohydrate load and low F/M. The standard mitigations are upstream FOG removal with a DAF or a high-efficiency sedimentation tank plus a surface skimmer on the SBR. CIP and product-change slugs can drop dissolved oxygen to near zero inside a single cycle, and a feed-side equalisation tank of at least 6 to 8 hours of hydraulic retention is the lowest-cost protection. S4 reported SVI consistently under 100 mL/g when FOG and TSS were removed upstream, and that is the most useful daily health-check metric for an operator. Low-DO simultaneous nitrification-denitrification saves 37.5 percent of aeration energy (S1), but it depends on reliable DO probes and stable valve control, so instrumentation quality should be specified at procurement rather than left to commissioning. A chlorine dioxide generator on the polish loop gives operators a non-odor-forming residual for downstream disinfection when reuse or river-discharge limits apply.
Frequently Asked Questions
What HRT should a dairy SBR be designed around?
The 24 h HRT used in the simulated-dairy study (S5) is the practical optimum for suspended-growth dairy SBRs, and extending HRT to 30 h did not produce a statistically significant COD gain in that work. For lower-temperature or higher-FOG sites, the 5.3 d HRT used in the EPA HERO lab study (S4) is the long-residence benchmark. Request a bench- or pilot-scale kinetic run on the actual plant effluent before committing to a final HRT figure.
When does a stand-alone SBR justify adding an MBR?
Add an MBR downstream when the discharge or reuse target requires TSS below about 20 to 30 mg/L, when ammonia-based toxicity limits apply, or when the plant is targeting water reuse. S1's IFAS data shows that biofilm carriers can lift TN removal to 86.5 percent at C/N 3.0 to 3.5, but an MBR still gives the tightest TSS and a more stable effluent under hydraulic shock. Compare both trains on a 10-year OPEX basis, because membrane replacement and aeration energy dominate lifecycle cost.
What is the largest OPEX driver for a dairy SBR?
Aeration energy is the single largest controllable OPEX line in any aerobic dairy SBR, and the IFAS-SBR study (S1) measured a 37.5 percent reduction in aeration energy by holding DO at 0.8 to 1.2 mg/L instead of 2 to 3 mg/L. Sludge hauling and chemical cost for phosphorus precipitation are the next two lines, and both drop when upstream FOG removal and flow equalisation are sized correctly. Ask any supplier for an itemised kWh per m³ and a wasted-kg-AS per m³ figure before pricing the bid.
What should a buyer check when selecting an SBR or SBR+MBR supplier?
Require documented cycle parameters from at least one operating dairy reference at a comparable flow range, and ask for a written performance guarantee on COD, TKN, and TSS at the design HRT and MLSS. Verify that the supplier can deliver a working DO control loop, because low-DO simultaneous nitrification-denitrification (S1) is only as stable as its probe and valve train. Confirm membrane module lead time, the on-site service footprint, and the documented CIP protocol before signing; for compliance, check the supplier's familiarity with the local pretreatment rules covered in the food and beverage pretreatment compliance guide and the broader advanced nutrient removal guide. For equipment shortlisting, evaluate an integrated MBR system against a stand-alone SBR plus a DAF pre-treatment polish, and request a side-by-side mass balance at the design flow before any procurement commitment.
