Ammonia removal from wastewater is accomplished through four established routes: biological nitrification (ammonia oxidized to nitrite then nitrate by Nitrosomonas and Nitrobacter, typically 85-95 percent removal), air stripping at pH above 10.5, breakpoint chlorination, and ion exchange or membrane processes for polishing. The dominant method at municipal plants is two-stage biological nitrification, which reliably reaches effluent total ammonia nitrogen below 1-3 mg/L when dissolved oxygen stays above 2 mg/L, alkalinity is sufficient at roughly 7.1 mg CaCO3 per mg N, and water temperature holds above 15 degrees Celsius. Sizing, cost, and failure modes all follow from those three variables.
Why Ammonia Is Hard to Remove and What Drives Toxicity
Ammonia in water exists as two species in equilibrium: the ionized ammonium ion (NH4+) and un-ionized free ammonia (NH3). The split is governed almost entirely by pH and temperature. At pH 7 and 25 degrees Celsius, roughly 99.4 percent sits as NH4+; at pH 9.5 the free ammonia fraction climbs above 50 percent. This single fact explains most of the confusion in the field, because the toxic species and the strippable species are the same one.
Free ammonia is the hazardous form. It crosses fish gill membranes and disrupts ion regulation and oxygen uptake at concentrations as low as 0.02 mg/L un-ionized NH3 for sensitive coldwater species. Total ammonia nitrogen (TAN), the number most permits regulate, therefore obscures real risk: an effluent reading of 5 mg/L TAN is mildly concerning at pH 7 but genuinely dangerous at pH 8.5 and 20 degrees Celsius, where the free fraction approaches 15 percent.
In plain terms: the same ammonia number means different things depending on pH and temperature, so treat the free ammonia fraction as the true hazard.
In the environment, ammonia is also an oxygen sink. Nitrification consumes about 4.3 mg of oxygen per mg of ammonia nitrogen oxidized, so a plant discharging 20 mg/L TAN imposes a substantial oxygen demand on the receiving river long after the discharge pipe.
The Nitrification Mechanism, Step by Step
Biological nitrification is a two-step sequential oxidation carried out by two distinct guilds of bacteria that share little except a need for oxygen.
The first step converts ammonia to nitrite, performed by ammonia-oxidizing bacteria (AOB) such as Nitrosomonas and, increasingly recognized, ammonia-oxidizing archaea. The second step converts nitrite to nitrate, performed by nitrite-oxidizing bacteria (NOB) such as Nitrobacter and Nitrospira. Both steps are aerobic and autotrophic, meaning the organisms fix carbon dioxide and gain little energy per mole, which makes them slow growers.
That slow growth is the central engineering constraint. AOB doubling times run 8-24 hours at 20 degrees Celsius. NOB are similar or slower. Because biomass accumulates so gradually, a nitrifying system needs a solids retention time (sludge age) long enough to prevent washout, typically 8-15 days at 15 degrees Celsius and 4-6 days at 25 degrees Celsius. Drop below the critical SRT and the population declines faster than it reproduces, and nitrification collapses within days.
The stoichiometry matters for operations. Complete oxidation of 1 mg of ammonia nitrogen to nitrate consumes about 4.57 mg oxygen and 7.14 mg alkalinity as CaCO3. A plant nitrifying 25 mg/L TAN therefore destroys roughly 178 mg/L of alkalinity and demands over 100 mg/L of oxygen, which is why nitrifying basins are frequently oxygen-limited and why pH drifts downward as the reaction proceeds.
Quantified Context: Effluent Standards and Typical Performance
Ammonia limits vary widely by jurisdiction and receiving water, and the numbers below are representative rather than universal.
| Treatment approach | Typical effluent TAN | Capital intensity | Operating cost driver | Best fit |
|---|---|---|---|---|
| Conventional activated sludge with nitrification | 1-5 mg/L | Moderate | Aeration energy, alkalinity | Large municipal plants |
| Sequencing batch reactor (SBR) | 1-3 mg/L | Moderate | Aeration, cycle timing | Small-to-mid plants |
| Moving bed biofilm reactor (MBBR) | 2-8 mg/L | Moderate-high | Media, aeration | Upgrades with limited footprint |
| Air stripping tower | 5-30 mg/L (as NH3) | High | pH adjustment, energy, scaling | High-strength industrial streams |
| Breakpoint chlorination | Below 1 mg/L | Low capital | Chlorine demand, byproducts | Polishing, emergency control |
| Ion exchange (zeolite) | Below 1 mg/L | Moderate | Regeneration brine | Low-flow polishing |
| Membrane bioreactor (MBR) | Below 1 mg/L | High | Membrane replacement, energy | Tight limits, reuse |
The performance difference between conventional nitrification and membrane or ion-exchange polishing is roughly one order of magnitude in effluent concentration. That gap drives most upgrade decisions, because a permit tightening from 5 to 1 mg/L TAN cannot be met by conventional secondary treatment alone in cold climates.
Energy demand deserves mention. Aeration for nitrification commonly represents 45-60 percent of a municipal plant's total electricity use, so the biological route, while effective, is the dominant operating cost.
Temperature, Alkalinity, and Dissolved Oxygen: The Three Failure Levers
Nitrification rates roughly double for every 10 degrees Celsius rise up to about 30 degrees, then decline above 35 degrees as enzymes denature. Below 10 degrees the rate falls sharply, and many plants that meet limits in summer fail in winter without adjustment. Operators respond by extending sludge age and increasing aeration, which raises cost exactly when energy demand for heating is also high.
Alkalinity is the second lever. Because the reaction consumes 7.1 mg CaCO3 per mg N, a wastewater with 200 mg/L influent alkalinity can nitrify roughly 28 mg/L TAN before pH crashes. Below pH 6.5, AOB activity stalls and the process becomes self-limiting. In plain terms: soft water cannot nitrify heavily loaded waste without lime or caustic addition.
Dissolved oxygen is the third. AOB need roughly 1.5-2 mg/L DO to function near maximum rate, and NOB can be inhibited at low DO, which causes nitrite accumulation, a condition called nitritation. Some operators exploit this deliberately in shortcut nitrogen removal processes that save aeration energy by stopping at nitrite and then denitrifying it.
A Counterintuitive Detail: Biofilm Beats Suspended Growth in the Cold
A commonly held belief is that cold temperatures make biological ammonia removal impractical and force chemical alternatives. Field experience contradicts this in a specific, useful way. Biofilm systems, including MBBR and trickling filters, retain slow-growing nitrifiers within a protected matrix, so their effective sludge age is far longer than the hydraulic retention time suggests.
A biofilm carrier with a protected internal surface can hold nitrifying biomass with an apparent sludge age of 20-40 days even when the suspended reactor SRT is only 3-5 days. The consequence is that biofilm systems often out-nitrify conventional suspended-growth plants at 8-12 degrees Celsius, precisely where conventional systems struggle. The trade-off is higher aeration demand and media cost, and the risk of media clogging if pretreatment is inadequate.
The professional caveat here: biofilm nitrification performance is harder to model than suspended growth, and design margins of 20-30 percent are prudent when sizing for winter peaks. Published rate constants vary widely, so pilot testing at the actual wastewater remains the most reliable predictor.
Where Chemical and Physical Alternatives Fit
Chemical and physical processes are not competitors to biology in most cases; they are complements used where biology cannot go.
Air stripping uses the pH-temperature equilibrium directly. Raising pH to 10.5-11.5 converts nearly all TAN to volatile NH3, which is stripped in a packed tower with countercurrent air. Removal of 90-95 percent is achievable, but the process suffers from calcium carbonate scaling on the packing at high pH, which fouls the tower and requires periodic acid cleaning. It suits high-strength industrial streams such as landfill leachate, where ammonia loads are far above municipal levels.
Breakpoint chlorination oxidizes ammonia to nitrogen gas and is fast and reliable for polishing to very low concentrations. The stoichiometry requires about 7.6 mg chlorine per mg ammonia nitrogen, and the "breakpoint" is the dosage at which residual chlorine suddenly appears. Beyond that point, excess chlorine forms chloramines and then free chlorine, which can create disinfection byproducts such as trihalomethanes. Because of the byproduct risk, it is used sparingly and mostly for emergency control or final polishing.
Ion exchange with natural zeolites such as clinoptilolite selectively captures ammonium ions. Regeneration requires sodium chloride brine, producing a concentrated ammonia-rich waste stream that must itself be treated, so ion exchange shifts the problem rather than destroying it. It is best reserved for low-flow, tight-limit polishing.
A Realistic Timeline: What to Expect During a Nitrification Upset
The following reflects a typical municipal activated sludge system; the exact figures will vary with temperature and biomass inventory.
| Time after upset | Observation |
|---|---|
| Hour 0-6 | Aeration blower trip or toxic slug enters; DO rises as oxygen demand falls |
| Hour 6-24 | Ammonia breaks through in effluent, rising from below 1 to above 10 mg/L |
| Day 1-3 | Nitrite spikes first; biomass population declines as growth no longer exceeds washout |
| Day 3-7 | Effluent ammonia stabilizes at influent-like levels; operators begin alkalinity and DO corrections |
| Day 7-14 | Re-seeding of nitrifiers begins; ammonia starts a slow decline |
| Day 14-28 | Full recovery possible if temperature and SRT support regrowth |
| Beyond 28 days | Cold-weather recovery can stretch to 6-8 weeks without supplemental seeding or bioaugmentation |
The non-obvious lesson in this table is the recovery asymmetry: a system takes days to lose nitrification and weeks to regain it. That is why buffering capacity, backup blowers, and toxin monitoring matter more than peak efficiency.
What to Actually Do: Practical Actions for Operators and Designers
Positive, specific steps rather than warnings:
- Maintain a solids retention time at least 2.5 times the minimum required for the coldest expected month, not the average.
- Monitor free ammonia, not just TAN, using the pH and temperature partition to calculate the un-ionized fraction before comparing to toxicity limits.
- Feed supplementary alkalinity (lime or sodium bicarbonate) before pH drops below 7.0; recovery after a pH crash is slower than prevention.
- Use a nitrite slope on daily effluent data as an early warning: rising nitrite almost always precedes rising ammonia.
- Consider MBBR or bioaugmentation for winter capacity rather than accepting annual permit exceedances.
- Reserve chemical processes such as breakpoint chlorination for polishing and emergency control, and budget for their byproduct and reagent costs.
Red flags that warrant immediate operator action:
- Effluent ammonia rising by more than 2 mg/L in 24 hours with no load change
- DO above 4 mg/L in the aeration basin with declining ammonia removal (indicates biomass loss)
- pH below 6.5 in the nitrifying zone
- Nitrite above 1 mg/L in final effluent
- Any observed sheen or scum in the secondary clarifier coinciding with ammonia breakthrough
Limitations and What Remains Uncertain
No treatment train eliminates ammonia; every route converts it to another form or moves it to another stream. Nitrification converts it to nitrate, a drinking-water contaminant regulated at 10 mg/L nitrate-nitrogen in the United States. Denitrification then converts nitrate to nitrogen gas, which is the only true destruction pathway, but it requires an anoxic zone and a carbon source. Air stripping releases ammonia to the atmosphere, where it returns as deposition. Ion exchange concentrates it for disposal.
Design equations and rate constants for nitrification published across references vary by a factor of two or more, especially in the 10-15 degrees Celsius range where most winter failures occur. No single design number should be trusted without site-specific pilot or full-scale data. For high-strength industrial wastewater, the uncertainty widens further because trace toxics, salinity, and inhibitory organics can suppress nitrifier activity unpredictably. The most defensible engineering position is to design for the measured worst case, keep a chemical backup available, and treat the biological system as a living asset that needs redundancy rather than a fixed-capacity machine.