16-Civ-B5 Water Supply and Wastewater Treatment · December 2015

Question 3 of 7: Significance and Control of Ammonia in Treated Effluents

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Question 3: Significance and Control of Ammonia in Treated Effluents (15 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Given. A treated municipal effluent discharged to a surface water body, considered under three different receiving-water conditions; the illustrative effluent used below carries 25 mg/L of total ammonia nitrogen at pH 8.5 and 20 °C, a representative value for a well-run secondary plant that does not nitrify.

Find. Why ammonia matters in such a discharge, which analytical forms are reported, which form governs in each of the three receiving-water cases, and how pH controls ammonia toxicity and what can be done about it.

Why ammonia matters

Ammonia is significant in three separate ways, and confusing them is the commonest error on this question. First, it exerts an oxygen demand. Nitrifying bacteria in the receiving water oxidise ammonium to nitrate,

which consumes $2\times32/14 = 4.57$ g of oxygen for every gram of nitrogen oxidised. That nitrogenous oxygen demand is exerted downstream of the carbonaceous demand and is frequently the larger of the two: the illustrative 25 mg/L of ammonia nitrogen represents 114 mg/L of NOD, comfortably more than the residual cBOD of the same effluent. Second, ammonia is directly toxic to fish and invertebrates, and the toxic agent is the un-ionised molecule NH₃, not the ammonium ion. Third, ammonia is a bio-available nutrient that drives algal growth and, through the subsequent die-off and decay, a second and delayed oxygen demand. It also destroys any chlorine residual and complicates downstream disinfection.

Forms reported in effluent analysis

Four are routinely reported. Total ammonia nitrogen (TAN), sometimes written NH₃-N or NH₄⁺-N, is what the laboratory actually measures — the sum of the two species, because any analytical method shifts the equilibrium as it works. Un-ionised ammonia, NH₃, and ionised ammonium, NH₄⁺, are not measured separately but calculated from TAN with the pH and temperature of the sample. Total Kjeldahl nitrogen (TKN) is ammonia plus organic nitrogen, and the difference TKN minus TAN gives the organic nitrogen that will mineralise to ammonia later in the receiving water. Effluent permits are almost always written on TAN, because that is what can be measured and controlled at the outfall, while the water-quality objective in the receiving water is written on NH₃.

Which form governs in each receiving water

(a) High DO, endangered species sensitive to toxicity. The oxygen demand is not the binding constraint — the receiver has oxygen to spare — so the parameter that matters is the un-ionised ammonia, NH₃, evaluated at the worst-case pH and temperature of the receiving water after mixing. NH₃ is uncharged and lipid-soluble, so it diffuses across the gill epithelium; NH₄⁺ is charged and largely excluded. Inside the fish, ammonia disrupts the blood-to-water gradient the animal itself relies on to excrete its own metabolic ammonia, so the toxic effect is a failure of excretion. The CCME long-term guideline of 0.019 mg/L as un-ionised ammonia nitrogen is the number to design against; a discharge permit here would be written on TAN but back-calculated from that NH₃ objective at the summer pH and temperature.

(b) Low DO, no toxicity concerns. Now the binding constraint is oxygen, so the governing parameter is total ammonia nitrogen, since every milligram of nitrogen — ionised or not — is eventually nitrified and every milligram therefore draws 4.57 mg of oxygen from a receiver that has none to lose. Where organic nitrogen is appreciable, TKN is the more honest basis, because organic nitrogen mineralises to ammonia within the reach of interest and exerts the same demand a little further downstream. Speciation is irrelevant here: the nitrifiers consume ammonium, and the pH-dependent split between NH₃ and NH₄⁺ does not change the total oxygen bill.

(c) Both low DO and toxicity concerns. Both parameters must be satisfied, and the permit is set by whichever is more stringent — in practice almost always the un-ionised ammonia criterion, because the toxicity threshold is three orders of magnitude below the concentration at which oxygen depletion becomes the issue. The two constraints also interact in the worst possible way. Low dissolved oxygen is usually a symptom of a slow, warm, algal-productive reach; algal photosynthesis strips carbon dioxide during the day and drives the pH up, and both the higher temperature and the higher pH increase the un-ionised fraction. A fish already stressed by hypoxia must ventilate its gills harder, which increases its exposure to that ammonia. Design for the simultaneous worst case — low flow, high temperature, afternoon pH maximum — not for the annual averages, and the practical answer is nitrification at the plant, which removes both problems at once.

Effect of pH, quantified

Given. An effluent at 25 mg/L TAN and 20 °C, evaluated at pH 7.0 and pH 8.5, using the Emerson relation $\mathrm{p}K_a = 0.09018 + 2729.92/T$ with $T$ in kelvin.

Find. The un-ionised fraction at each pH and the ammonia nitrogen actually delivered in toxic form.

1. Evaluate the dissociation constant at the effluent temperature. $$\mathrm{p}K_a = 0.09018 + \frac{2729.92}{273.15+20} = 9.402$$
2. Apply the acid–base speciation relation. Since $\mathrm{NH_4^+ \rightleftharpoons NH_3 + H^+}$, $$f_{\mathrm{NH_3}} = \frac{1}{1+10^{\,\mathrm{p}K_a-\mathrm{pH}}}$$ which at pH 7.0 gives $f = 1/(1+10^{2.402}) = 0.00394$, and at pH 8.5 gives $f = 1/(1+10^{0.902}) = 0.1113$.
3. Convert to the concentration the fish actually sees. At pH 8.5 the toxic load is $$C_{\mathrm{NH_3}} = 0.1113\times 25\ \mathrm{mg/L} = \boxed{2.78\ \mathrm{mg/L\ as\ NH_3\text{-}N}}$$ against a CCME long-term guideline of 0.019 mg/L — a factor of 146 over. The same effluent at pH 7.0 would deliver 0.099 mg/L, still over the guideline but by a factor of five rather than 146.
4. Read off the sensitivity. One and a half pH units multiply the toxic fraction by 28.2, and a five-degree rise from 20 to 25 °C at fixed pH 8.5 raises it further from 11.1 % to 15.2 %. To meet 0.019 mg/L at pH 8.5 and 20 °C the effluent TAN would have to fall to 0.17 mg/L, which is below what even a well-run nitrifying plant achieves and shows that dilution in the receiving water is part of the answer, not an evasion of it.

How ammonia toxicity is controlled

Control acts on either the total ammonia or the fraction that is un-ionised. Biological nitrification at the plant is the primary and permanent measure: operating at a solids retention time long enough for the slow-growing nitrifiers, typically 10 to 15 days at 10 °C, converts ammonia to nitrate and eliminates both the toxicity and the nitrogenous oxygen demand. Adding an anoxic zone for denitrification recovers about half the alkalinity and part of the oxygen, and removes the nitrate as well. Breakpoint chlorination destroys ammonia chemically but costs 7.6 mg of chlorine per mg of nitrogen and generates chloride and by-products, so it is a polishing or emergency measure. Air stripping works only if the pH is first raised above 10.5, which is an expensive way to solve a problem best solved biologically. Directly, pH adjustment of the effluent to 7.0 or below before discharge cuts the un-ionised fraction by a factor of nearly thirty, and it is the fastest available intervention when a plant is temporarily out of nitrification. Finally, outfall design matters: a multi-port diffuser that achieves rapid initial dilution reduces the near-field NH₃ concentration, and equalising or timing the discharge away from the afternoon pH maximum exploits the same physics. Effluent limits should be set at the low receiving-water flow, conventionally the 7Q10, because the required effluent quality is nearly proportional to the dilution available.