Odour Control for Hydrogen Sulphide at Wastewater Works

Why does hydrogen sulphide dominate wastewater odour?

Municipal and industrial wastewater odour is a mixture, but hydrogen sulphide (H₂S) is almost always the governing compound. It is produced in large quantities, it is acutely toxic, it corrodes concrete and metals, and the human nose detects it at extraordinarily low concentrations. The other odorants of concern — organic sulphides and mercaptans (methyl mercaptan, dimethyl sulphide, dimethyl disulphide), reduced nitrogen species (ammonia, amines), and a range of volatile organic compounds (VOCs) such as short-chain fatty acids, aldehydes and skatole — usually travel with H₂S and are addressed by the same containment and treatment system.

H₂S is generated biologically. Under anaerobic conditions in rising mains, wet wells, primary tanks, sludge holding tanks and thickeners, sulphate-reducing bacteria (SRB, chiefly Desulfovibrio and related genera) use sulphate (SO₄²⁻) as a terminal electron acceptor, oxidising organic carbon or hydrogen and reducing sulphur to sulphide. The idealised stoichiometry for a simple organic substrate is:

The reason a tiny mass of sulphide creates a large nuisance is its volatility and the sensitivity of human olfaction. The 50% detection threshold for H₂S is around 0.5–5 µg/m³ (roughly 0.3–3 ppb), so a source releasing only grams per hour can trigger complaints kilometres downwind. This is why odour control is engineered to a mass-emission and dispersion basis, not merely to a “smells acceptable” judgement.

What controls whether sulphide escapes as gas?

Dissolved sulphide exists in equilibrium between three species — molecular H₂S(aq), bisulphide (HS⁻) and sulphide (S²⁻). Only the un-ionised, molecular form volatilises and smells. The first dissociation governs the pH dependence:

This equilibrium is the single most important lever in liquid-phase odour control. A drop in pH — caused by CO₂ stripping at a discharge point, by acidic industrial trade effluent, or by turbulence — shifts speciation towards molecular H₂S and drives it out of solution. That is why turbulent transitions (manhole drops, siphon breaks, weirs, pumped-main discharges into gravity sewers) are the notorious release points even when bulk sulphide concentrations are modest.

The rate of transfer from liquid to air follows two-film theory:

Corrosion and health: why H2S is more than a nuisance

Odour control is frequently justified on nuisance grounds, but the asset-protection case is often larger. Where H₂S accumulates in the headspace of sewers and structures, a two-stage biochemical process attacks the concrete. First, H₂S partitions into the thin condensate/moisture film on exposed crown and wall surfaces. There, aerobic sulphur-oxidising bacteria of the genus Acidithiobacillus (formerly Thiobacillus) oxidise it stepwise to sulphuric acid:

On the health side, H₂S is an acute chemical asphyxiant that inhibits cytochrome c oxidase. Its toxicological profile is unusually treacherous because olfactory fatigue sets in: the smell “disappears” at high concentration precisely when danger is greatest. UK workplace exposure is governed by HSE limits (long-term 5 ppm over 8 h; short-term 10 ppm over 15 min), while ~100 ppm is Immediately Dangerous to Life and Health and >700 ppm can cause rapid collapse. The table below anchors the design targets against these effect levels.

Odour, corrosion and confined-space safety therefore share a single root cause, and a coherent strategy tackles them together rather than as separate projects. For a broader treatment context, see our overview of industrial wastewater treatment processes and the sibling guide on industrial wastewater treatment.

Liquid-phase control: stopping H2S before it forms

The cheapest odour to treat is the one never generated. Liquid-phase (in-sewer or in-tank) dosing tackles sulphide at source and reduces the load on downstream vapour-phase plant. Four mechanisms are used, often in combination:

• Nitrate dosing (calcium or sodium nitrate). Supplying an alternative electron acceptor lets facultative heterotrophs respire nitrate in preference to sulphate, out-competing the SRB and also biologically re-oxidising existing sulphide. It suppresses generation without adding a solid.
• Iron salt dosing (ferric chloride FeCl₃, ferrous chloride FeCl₂, or ferric sulphate). Dissolved iron precipitates sulphide as highly insoluble iron sulphide (FeS, K_sp ~10⁻¹⁸), removing it from the volatile pool. Ferric additionally oxidises sulphide while being reduced to ferrous, which then precipitates more sulphide — a dual action.
• pH elevation (magnesium hydroxide or caustic). Raising pH shifts the H₂S/HS⁻ equilibrium toward non-volatile bisulphide (see the pK_a relation above). Slug-dosing to pH >12.5 also periodically inactivates the biofilm SRB.
• Oxygen or nitrate re-aeration to maintain a positive redox potential in the bulk liquid, preventing the onset of septicity in long rising mains.

Iron dosing is the workhorse for tanks and thickeners because it is robust, precipitates sulphide permanently, and the FeS leaves with the sludge. The governing stoichiometry for ferrous is one mole of iron per mole of sulphide; ferric is often dosed at a molar ratio above the redox stoichiometry to allow for competing hydroxide/phosphate precipitation:

Dosing is not free of consequence: iron salts consume alkalinity and add chloride/sulphate to the sludge and final effluent, so the trade-effluent and sludge-disposal implications must be checked. Where you want the flowsheet designed around these interactions, our team can help you design an integrated sulphide-control and dosing strategy.

Worked example 1: foul-air flow and H2S mass load

A covered primary tank and its associated inlet channel are to be ducted to an odour-control unit. Sizing begins with the extracted air volume and the mass of H₂S it carries — the two numbers that govern every downstream technology choice.

• Containment volume: the covered headspace and ducted structures enclose 250 m³.
• Ventilation rate: good practice for covered process tanks is 3–6 air changes per hour (ACH); take 5 ACH to hold a slight negative pressure and capture fugitive release.
• Extract flow: Q = 250 m³ × 5 h⁻¹ = 1,250 m³/h (≈ 0.35 m³/s).
• Headspace H₂S concentration: measured/estimated at 50 ppm_v. Converting to mass at 20 °C (molar volume 24.05 L/mol, H₂S molar mass 34.1 g/mol): 50 ppm_v = 50 × 34.1 / 24.05 = 70.9 mg/m³.
• Mass load: L = Q × C = 1,250 m³/h × 70.9 mg/m³ = 88,600 mg/h ≈ 2.1 kg H₂S per day.

So the odour unit must handle 1,250 m³/h of foul air carrying about 2.1 kg/d of H₂S, plus mercaptans and VOCs. If the discharge consent is 5 ppm_v at the stack (a common UK planning condition), the required removal efficiency is (50 − 5)/50 = 90%; if the target is odour-nuisance protection at 0.1 ppm_v, it rises to 99.8%, which pushes the design toward a two-stage system. This flow-and-load basis is what every technology in the next section is sized against.

Vapour-phase treatment: scrubbers, biofilters and carbon

The captured foul air is treated by one or more of three families of technology. Chemical scrubbers give the highest, most controllable removal in the smallest footprint but carry an ongoing chemical bill; biological systems (biofilters, biotrickling filters) have very low running costs but a larger footprint and slower response to shock loads; activated carbon is a compact polishing or standby technology. The comparison below frames the trade-offs.

A chemical scrubber for H₂S is typically a packed counter-current tower with recirculating liquor. A first stage of sodium hydroxide (caustic) absorbs and neutralises H₂S; a second stage of sodium hypochlorite (often with residual caustic) oxidises the captured sulphide and scavenges mercaptans; a third acid stage may be added to remove ammonia. The governing chemistry is:

The biotrickling filter and biofilter instead rely on the sulphur-oxidising bacteria we met in the corrosion mechanism — but here they are put to work deliberately on inert or organic packing. H₂S transfers from the gas into the biofilm water and is oxidised aerobically to sulphate; the acid produced is neutralised (by irrigation/trickling liquor in a BTF, or by the media buffering in a biofilter). Two parameters govern the biological design: the empty-bed contact time (EBCT = bed volume / gas flow) and the elimination capacity (EC = mass of pollutant removed per unit bed volume per hour). Activated carbon — usually caustic- or KOH-impregnated for H₂S — sits at the compact, instant-start end and is covered in depth in our guide to activated carbon adsorption. Because biological odour control is itself a microbial oxidation process closely related to digestion chemistry, it also connects usefully to anaerobic digestion fundamentals, where the same sulphur cycle appears as biogas H₂S.

Worked example 2: sizing a scrubber vs a biotrickling filter

Take the duty from Worked Example 1: Q = 1,250 m³/h of air carrying 2.1 kg/d of H₂S (= 0.0876 kg/h), to be reduced by 99%. We size both a chemical scrubber’s reagent demand and a biotrickling filter’s media volume so the trade-off is concrete.

(a) Chemical scrubber reagent demand. H₂S removed = 0.99 × 0.0876 = 0.0867 kg/h. Moles of H₂S = 86.7 g/h ÷ 34.1 g/mol = 2.54 mol/h.

• Caustic: 2 mol NaOH per mol H₂S → 5.09 mol/h × 40.0 g/mol = 204 g/h = 4.9 kg NaOH/day (as 100%). As 30% w/w solution (density ~1.33 kg/L) that is ~12.3 L/day.
• Hypochlorite: 4 mol NaOCl per mol H₂S → 10.2 mol/h × 74.4 g/mol = 756 g/h = 18.2 kg NaOCl/day (as 100%). As 14% available-chlorine solution that is on the order of 130 L/day — the dominant consumable and the reason scrubbers carry a real OPEX.

(b) Biotrickling filter media volume. Choose a design elimination capacity comfortably within demonstrated performance for H₂S on plastic random packing, EC = 30 g H₂S/m³·h (field values of 50–100+ are achievable once acclimated, so this is conservative).

An 8.3 s EBCT is on the tight side, so a designer would round the bed up (say 4–5 m³, EBCT ~12–14 s) for resilience to load peaks and to give margin for mercaptan/VOC removal, which need longer contact. The contrast is stark: the BTF consumes essentially only water, nutrients and a fan, whereas the scrubber consumes ~18 kg/d of hypochlorite indefinitely. For a steady, high-H₂S sewage-works duty the biological route usually wins on whole-life cost; for a small, intermittent or space-constrained source, or where absolute reliability against spikes is paramount, the scrubber (or a scrubber-then-carbon train) is preferred.

Worked example 3: iron dose for liquid-phase sulphide control

Now suppress the sulphide upstream so the vapour-phase unit sees a lighter load. A rising main delivers 40 L/s (144 m³/h) with a measured dissolved sulphide of 8 mg/L (as S). We dose ferrous chloride (FeCl₂) to precipitate FeS at a design molar ratio Fe:S of 1.5:1 (the 1:1 stoichiometry plus 50% for competing demand and a target residual <0.5 mg/L).

• Sulphide mass flow: 8 mg/L × 40 L/s = 320 mg/s × 3,600 = 1,152 g/h ≈ 27.6 kg S/day.
• Molar sulphide load: 1,152 g/h ÷ 32.1 g/mol = 35.9 mol/h.
• Iron required: 1.5 × 35.9 = 53.8 mol Fe/h × 55.8 g/mol = 3,004 g Fe/h = 72.1 kg Fe/day.
• As FeCl₂ product: FeCl₂ molar mass 126.8 g/mol, so pure salt = 53.8 mol/h × 126.8 = 6,822 g/h = 164 kg/day. Dosed as a typical ~30% w/w liquor that is ~546 kg/day of solution, roughly 400 L/day.

Two design checks follow. First, alkalinity: iron salts are acidic, so confirm the receiving water has buffer capacity or the pH will fall — counter-productively pushing residual sulphide toward volatile H₂S. Second, contact time: FeS precipitation is fast but needs a few minutes of turbulent mixing along the main to complete, so the dosing point is placed well upstream of the release point. In practice iron dosing at this main might cut the downstream headspace H₂S by 80–90%, shrinking the Example-2 scrubber or biofilter accordingly — the essence of an integrated, source-to-stack odour strategy.

Ventilation, containment and dispersion

No treatment plant works unless the foul air is first captured. Containment (covers on tanks and channels, sealed transfer points) plus a modest extraction rate holds the enclosed structures under slight negative pressure so odour is drawn to treatment rather than leaking to atmosphere. Design ventilation is usually specified as air changes per hour (3–6 ACH for process tanks; higher for buildings with occupancy) or as a face velocity across openings (typically ≥0.5 m/s inward) to guarantee capture. Over-ventilating merely inflates the treatment flow — and hence capital and running cost — so the rate is optimised, not maximised.

Finally, even fully treated air is discharged through a stack sized and located for atmospheric dispersion. The ground-level concentration downwind falls with stack height and exit momentum, following Gaussian-plume behaviour in which the peak ground-level concentration scales approximately with the emission rate divided by the square of an effective stack height. A well-designed odour-control scheme therefore layers four defences — source reduction (liquid-phase dosing), containment, vapour-phase treatment, and dispersion — so that no single element has to carry the full 0.3 ppb-threshold burden alone. For the wider design context, see the sibling guides on industrial wastewater treatment and anaerobic digestion fundamentals, where the same sulphur chemistry recurs.

How to design an H2S odour-control scheme

1. Characterise the source. Measure headspace and dissolved H2S, temperature, pH, sulphate and flow at each emission point. Identify turbulent release points where speciation shifts to volatile H2S.
2. Set the target. Fix the emission limit (e.g. stack H2S ppm, or odour-unit boundary concentration) from consent/planning and nuisance criteria, and back-calculate the required removal efficiency.
3. Reduce at source. Evaluate liquid-phase dosing (nitrate, iron salts, pH lift, re-aeration) to cut generation and lower the vapour-phase load. Size the iron/nitrate dose from the sulphide mass flow.
4. Contain and ventilate. Cover tanks and channels, seal transfer points, and extract at 3-6 ACH (or >=0.5 m/s face velocity) to hold negative pressure and capture the foul air.
5. Select and size vapour-phase treatment. Choose scrubber, biotrickling filter, biofilter or carbon on removal, footprint and whole-life cost. Size from the air flow and H2S mass load using EBCT/EC or reagent stoichiometry.
6. Disperse and verify. Discharge through an appropriately sized stack, then commission with H2S and olfactometry monitoring to confirm the boundary target is met.