Nitrification Failure in Winter: SRT, Alkalinity or Biomass?
A 20 MLD activated sludge plant treats domestic sewage with influent BOD5 250 mg/L, TKN 42 mg/L, ammoniacal nitrogen 34 mg/L, alkalinity 210 mg/L as CaCO3 and pH 7.2. The aeration tank volume is 5000 m³, MLSS 3000 mg/L, SRT 5 days, dissolved oxygen 1.2 mg/L, and the wastewater temperature is 20 °C in summer and 14 °C in winter. Effluent limits are BOD 20 mg/L, TSS 30 mg/L and NH4-N 10 mg/L. At present the effluent ammoniacal nitrogen is 6 mg/L in summer but 19 mg/L in winter, when the effluent pH falls to about 6.2 and pin floc is occasionally observed. The aeration blowers run at 85% of rated capacity in summer. The operations team proposes raising the MLSS to 4500 mg/L to solve the winter problem, keeping everything else unchanged.
1. Compute the SRT required for nitrification at 14 °C and compare it with the 5-day SRT being operated.
Read the worked answer
Growth rate at 14 °C. Using the standard temperature correction for nitrifiers:
μ = 0.47 × exp[0.098(T − 15)] = 0.47 × exp(−0.098) = 0.47 × 0.9068 = 0.426 d⁻¹
With a decay coefficient b = 0.05 d⁻¹ and a safety factor of 2.5 on the net growth rate:
SRT_design = SF/(μ − b) = 2.5/(0.426 − 0.05) = 2.5/0.376 = 6.6 days
With a safety factor of 3 the requirement rises to about 7.9 days.
Comparison. The plant is operating at 5 days. The theoretical washout SRT at 14 °C is 1/(0.426 − 0.05) = 2.7 days, so 5 days gives a safety factor of only 1.9 - below the usual 2.5–3.
The DO makes it worse. Low dissolved oxygen reduces the effective growth rate: with a half-saturation constant of about 0.5 mg/L, DO 1.2 mg/L gives 1.2/(1.2 + 0.5) = 0.71. Effective μ = 0.426 × 0.71 = 0.30 d⁻¹, so the washout SRT becomes 1/(0.30 − 0.05) = 4.0 days. At an operating SRT of 5 days the safety factor is only 1.25. The nitrifier population is being held at the edge of washout all winter, and any shock load or aeration dip pushes it over.
This is exactly what the effluent data show: 6 mg/L NH4-N in summer at 20 °C, 19 mg/L in winter at 14 °C. The diagnosis is confirmed by the pin floc, which is typical of a low-SRT, low-DO sludge with poor floc formation.
So the first part of the answer is that the winter failure is an SRT problem - but not the only one.
2. Determine whether alkalinity is a limiting factor and quantify the consequence.
Read the worked answer
Stoichiometry. Nitrification consumes 7.14 mg of alkalinity as CaCO3 per mg of NH4-N oxidised.
- Nitrogen oxidised = 34 − 10 (target) = 24 mg/L at minimum; at the current 19 mg/L effluent only 15 mg/L is oxidised, yet the pH has already fallen to 6.2.
- Alkalinity demand at complete nitrification of 34 mg/L = 34 × 7.14 = 243 mg/L as CaCO3.
- Available alkalinity = 210 mg/L as CaCO3.
Even before allowing for the acidity produced by carbonaceous oxidation and for the alkalinity consumed in the primary stages, the demand exceeds the supply by about 33 mg/L. In practice the pH collapses once alkalinity falls below roughly 50 mg/L, because the system loses buffering capacity and free CO2 dominates; below pH 6.8 nitrification itself is inhibited, which is a self-reinforcing failure - the pH drop slows nitrification and the residual ammonia rises.
Consequence. The plant cannot achieve full nitrification on this sewage without either:
- dosing alkali - about 60–90 mg/L as CaCO3 (lime or soda ash) to leave 50 mg/L residual alkalinity, or
- providing an anoxic zone - each mg of nitrate denitrified returns about 3.57 mg of alkalinity, so denitrifying 20 mg/L of nitrate would recover around 71 mg/L as CaCO3 and roughly halve the dose. This also reduces aeration demand and waste sludge.
Why this matters for the proposal. Raising the MLSS without fixing alkalinity and DO simply pushes more ammonia through an unbuffered system; the pH falls further and the higher biomass is starved of oxygen and buffer at the same time.
3. Evaluate the proposal to raise MLSS to 4500 mg/L and recommend the correct control strategy.
Read the worked answer
Why the proposal is the wrong lever. SRT = (V·X)/(Qw·Xr + Qe·Xe). Holding SRT at 5 days while raising MLSS from 3000 to 4500 mg/L does not change the sludge age at all - it forces the waste sludge rate up by 50% to keep SRT at 5 days. The nitrifier population is governed by SRT, not by MLSS. The winter problem would be entirely unchanged, while three new problems appear:
- Clarifier solids loading. Solids loading rate = (Q + Qr)·X/A. A 50% rise in X raises it by 50%, and the pin floc already indicates a poor-settling sludge; a rising blanket and solids washout in the effluent are likely.
- Oxygen demand. Nitrification alone demands 4.57 mg O2 per mg N, i.e. 34 × 4.57 ≈ 155 mg/L of nitrogenous oxygen demand - about 3100 kg O2/d at 20 MLD, on top of the carbonaceous demand. Blowers already at 85% in summer cannot carry 50% more biomass through a winter with colder, more oxygen-soluble but slower-reacting mixed liquor.
- Low DO. At 1.2 mg/L the nitrifiers are already half-inhibited; more biomass at the same aeration power drives DO lower still.
Correct strategy.
- Control SRT, not MLSS: raise the SRT to 8–10 days by reducing wasting; MLSS will rise as a consequence, not as a target.
- Check the two hard constraints before raising SRT: clarifier solids loading (keep the blanket below about 0.3 of the tank depth) and blower capacity. If either fails, install a diffuser upgrade or a separate nitrification stage.
- Raise DO to 2.0 mg/L in the aerobic zone.
- Dose alkali or add an anoxic zone to protect alkalinity and pH.
- Verify with a winter profile: ammonia, nitrate, DO, pH and alkalinity across the tank, since the failure is seasonal and will not show in a summer sample.