Biological or Chemical Phosphorus Removal: One Carbon Balance, Two Religions
Suggested readership: design teams setting a TP target for a tropical municipal plant and realising the effluent limit is really a carbon-management problem. Reading time: ~7 min. Companion tools: UCT / MUCT bio-P · Chemical P & carbon dosing
The same plant, two philosophies
Phosphorus leaves a wastewater plant exactly one way: in solids. The argument is only about which solids and how they are paid for.
Biological P removal (enhanced biological phosphorus removal, EBPR) engineers a microbial community that stores luxury polyphosphate. Run the carbon through an anaerobic zone first, keep nitrate out of it, and the biomass does the removal for free — except for the reactor volume and the carbon it consumes. On a 10,000 m³/d plant at 5 mg/L influent P (50 kg P/d), a UCT layout reaches 1.1 mg/L effluent TP biologically, removing 39 of the 50 kg P/d into waste sludge.
Chemical P removal precipitates the phosphorus with a metal salt — alum, PAC, ferric — and converts it to chemical sludge. It is brutally reliable and brutally expensive: the sludge is inorganic, bulky, and every kg of P removed carries several kg of metal hydroxide with it.
The two religions meet in one line of the mass balance: the carbon. EBPR needs readily biodegradable COD (VFAs) in the anaerobic zone. If you pre-settle the sewage (removing carbon) or let nitrate leak into the anaerobic zone (oxidising carbon), the biology starves and the metal salt bill arrives.
UCT in numbers
The UCT (and its MUCT variant) exists to solve one specific failure: in an A²/O layout, the return sludge carries nitrate into the anaerobic zone, where denitrifiers eat the VFA the phosphorus bacteria were waiting for. UCT sends the RAS to an anoxic zone first and recycles from there, so the anaerobic zone sees almost no nitrate — 7.3 kg NO₃-N/d instead of 48.5 on the same flowsheet.
The downstream effect is visible in the sizing:
- Aerobic SRT governs nitrification first: minimum nitrifier SRT at 30 °C is ~2.2 d, and the design lands at ~8.6 d with a safety factor of 4, giving an aerobic volume of 1,906 m³ (4.6 h).
- Anaerobic 1.5 h (625 m³), anoxic-1 1 h (417 m³), anoxic-2 2 h (833 m³) are then added on top: 3,781 m³ total, 9.1 h HRT, 17 d system SRT.
- Denitrification capacity from the internal recycle tops out at R/(1+R) — 67 % at R = 200 % — and the effluent TN lands near 9 mg/L.
Then the second bill arrives: nitrification destroys ~1,809 kg/d as CaCO₃ of alkalinity on this plant, and at 30 °C the influent alkalinity often is not enough to cover it. The design must add ~109 kg/d of alkalinity (or accept a pH slide that quietly slows the nitrifiers). That coupling — P removal, N removal and alkalinity on one balance — is why this is genuinely a design problem and not a tank-counting exercise.
Chemical P: the reliable fallback with a sludge price
Chemical dosing works even when the biology is exhausted, and tropical plants often need exactly that insurance. The engineering is simple: a molar metal-to-P ratio (roughly 1.5–2.5:1 for aluminium, allowing for the "residual envelope" — the P below the discharge target still has to be precipitated), dosed at the point of maximum mixing, with pH watched because metal hydroxide formation competes with metal phosphate formation.
Two numbers frame the cost. First, chemical sludge adds on the order of 5 % to biological sludge mass just from the hydroxide, before counting the P itself. Second, metal salts eat alkalinity too — the same alkalinity nitrification is also claiming. A plant that is simultaneously nitrification-limited and P-limited can spend more on alkalinity and sludge disposal than on the reagents themselves.
The carbon deficit: where bought-in COD appears
Both religions share a final failure mode. When the influent BOD₅/TN is low (say below 4), there is simply not enough carbon to denitrify all the nitrate the nitrifiers made — typically 18 mg/L NO₃-N that must go — and the calculator says you need about 92 mg/L of COD, or 920 kg/d, to close the gap.
Now the carbon-sourcing table starts to matter:
- Methanol, ~1.5 g COD per g product: 614 kg/d. The most dose-efficient of the three, but a hazardous liquid that needs acclimatised biomass.
- Acetic acid / sodium acetate: fast uptake, no acclimatisation, roughly 30–80 % more expensive per kg COD.
- Sugars / brewery or food waste: attractive where you have a reliable local source, but variable COD and a BOD load of their own.
Notice the irony available to a design team: an upstream primary clarifier that removed 27 % of the BOD is precisely what forces this 920 kg/d methanol purchase. The carbon you settle out in the primary tank is the carbon you buy back in the dosing day-tank.
The decision rule
- If the effluent TP target is above ~1.5 mg/L and the influent BOD₅/P ratio is above 20, design EBPR and protect the anaerobic zone from nitrate.
- If TP must be below ~0.5–1 mg/L, plan EBPR plus a trim chemical dose — biological for the bulk, chemical for the guarantee.
- If the carbon balance is structurally short (low BOD/TN), fix it at the flowsheet level first (skip primary, add a fermenter) before accepting a permanent carbon-dosing opex.