Capstone: checking a plant upgrade design

Every lesson in this course has contributed a piece of two designs. In the last weeks of preliminary design, Martin asks Nadia to assemble them into design basis reports, the documents that record what each plant must achieve, the numbers each unit was sized on and the reasons for each decision, and to check the wastewater upgrade against them. A design basis report is the bridge between the process engineer's calculations and the detailed design by civil, structural, mechanical and electrical engineers. It is also what the regulator reviews and what the operators will consult years later when they wonder why a tank is the size it is. This capstone works through that check.

The water plant design basis

The water plant was settled earlier in the course. Its design basis is summarized below, with the lesson in which each number was developed.

Two items remain open for the water plant and are listed as such: the tracer test at commissioning that must confirm the clearwell baffling factor, and the GAC pilot that must confirm the bed life before carbon is purchased. Recording open items is as important as recording decisions.

The wastewater upgrade: what must change

The city plant's amended licence will require a total phosphorus of 1 mg/L as a 30-day rolling average and a total nitrogen of 15 mg/L, with its existing secondary limits, disinfection from May to September and the WSER limits. The design targets agreed with the city are 0.5 mg/L total phosphorus and 10 mg/L total nitrogen in winter at 12 °C. The process selected in lesson 10 is A2O, with a primary sludge fermenter to supply volatile fatty acids, a trim alum dose for reliability, and side-stream treatment of the digester centrate.

The check must answer four questions. Is there enough aeration volume for the longer SRT that nitrification needs? Can the secondary clarifiers handle the larger solids inventory? Is there enough oxygen and alkalinity? And what does the upgrade do to the solids train?

Volume for the longer SRT

Lesson 10 set the aerobic SRT at 10 days. With 30 % of the biological volume unaerated (anaerobic and anoxic zones), the total SRT is

The solids inventory follows from the sludge production, which at this SRT, including inert solids, is about 0.75 g TSS per g BOD₅ removed, plus the chemical solids from the trim alum dose. Dividing the inventory by the design MLSS of 3,200 mg/L gives the volume needed.

# Capstone design check: nutrient removal upgrade of the city wastewater plant
UNITS = "SI"
SI = UNITS == "SI"
Q = 30000.0; Qp = 2.2 * Q                       # m3/d
BOD, TSS, TKN = 150.0, 96.0, 40.5              # primary effluent, mg/L
Y_obs = 0.75                                   # g TSS produced per g BOD removed at long SRT, incl. inerts (typical)
SRT_aer, anox_frac = 10.0, 0.30
SRT_tot = SRT_aer / (1 - anox_frac)
MLSS = 3200.0
Px = Y_obs * Q * BOD / 1000                    # kg TSS/d (before chemical solids)
alum = 80.0 * 0.5                              # mg/L after adding biological P removal (half the alum)
Px_chem = alum * 0.45 * Q / 1000              # kg/d of chemical solids, about 0.45 g solids per g alum
V_need = (Px + Px_chem) * SRT_tot * 1000 / MLSS
V_have = 9000.0
print(f"total SRT {SRT_tot:.1f} d; solids production {Px+Px_chem:,.0f} kg/d; basin volume needed {V_need:,.0f} m3 vs existing {V_have:,.0f} m3" if SI else
      f"total SRT {SRT_tot:.1f} d; solids production {(Px+Px_chem)*2.2046:,.0f} lb/d; basin volume needed {V_need*264.17/1e6:.2f} MG vs existing {V_have*264.17/1e6:.2f} MG")
# oxygen: carbon plus nitrification minus credit for denitrification
N_nit = TKN - 7.0 - 1.0 - 1.5
N_den = 24.4
Px_bio = 1700.0                               # kg VSS/d of biomass wasted (lesson 9 model at long SRT)
O2 = (Q * BOD / 0.68 / 1000 - 1.42 * Px_bio) + 4.57 * Q * N_nit / 1000 - 2.86 * Q * N_den / 1000
print(f"oxygen demand about {O2:,.0f} kg/d average; with a 1.5 peaking factor {1.5*O2:,.0f} kg/d" if SI else
      f"oxygen demand about {O2*2.2046:,.0f} lb/d average; with a 1.5 peaking factor {1.5*O2*2.2046:,.0f} lb/d")
# alkalinity: 7.14 used per g N nitrified, 3.57 recovered per g N denitrified, alum uses 0.5 per mg
alk_in = 320.0
alk_out = alk_in - 7.14 * N_nit + 3.57 * N_den - 0.5 * alum
print(f"effluent alkalinity {alk_out:.0f} mg/L as CaCO3 ({'adequate' if alk_out >= 70 else 'supplement needed'})")
# secondary clarifiers with the higher MLSS and return rate
A = 1650.0; R = 0.67
SLR = (Qp + R * Q) * MLSS / 1000 / A / 24
checks = [
    ("aeration volume", V_need <= V_have * 1.0, f"{V_need:,.0f} m3 needed" if SI else f"{V_need*264.17/1e6:.2f} MG needed"),
    ("peak overflow rate", Qp / A <= 40.0, f"{Qp/A:.0f} m3/(m2 d)" if SI else f"{Qp/A*24.54:,.0f} gpd/ft2"),
    ("peak solids loading", SLR <= 7.0, f"{SLR:.1f} kg/(m2 h)" if SI else f"{SLR*0.2048*24:.0f} lb/(ft2 d)"),
    ("effluent alkalinity", alk_out >= 70, f"{alk_out:.0f} mg/L"),
]
for name, ok, val in checks:
    print(f"{name:20s} {val:22s} {'passes' if ok else 'fails: action needed'}")
# option: add a fourth clarifier and two new basins
V_new = V_need - V_have
print(f"new basin volume {V_new:,.0f} m3 -> two basins of {V_new/2:,.0f} m3" if SI else f"new basin volume {V_new*264.17/1e6:.2f} MG -> two basins")

total SRT 14.3 d; solids production 3,915 kg/d; basin volume needed 17,478 m3 vs existing 9,000 m3
oxygen demand about 6,360 kg/d average; with a 1.5 peaking factor 9,540 kg/d
effluent alkalinity 166 mg/L as CaCO3 (adequate)
aeration volume      17,478 m3 needed       fails: action needed
peak overflow rate   40 m3/(m2 d)           passes
peak solids loading  7.0 kg/(m2 h)          passes
effluent alkalinity  166 mg/L               passes
new basin volume 8,478 m3 -> two basins of 4,239 m3

The check shows that the existing 9,000 m³ of aeration tanks is about half of the 17,500 m³ needed. The upgrade therefore needs about 8,500 m³ of new biological volume, which Nadia lays out as two new basins of about 4,250 m³ alongside the existing ones, so that the plant becomes four parallel trains of roughly equal size, each divided into anaerobic, anoxic and aerobic zones.

An alternative is to keep the existing tanks and add IFAS carriers to the aerobic zones, which would provide the nitrifying biomass on media without raising the MLSS. Nadia's comparison finds IFAS about 15 % cheaper in capital cost but higher in aeration energy, because the carriers need more air to stay suspended and the biofilm needs higher dissolved oxygen, and it would leave the plant with no spare tank for maintenance. The city's operators prefer the new basins, and the redundancy settles the choice. The IFAS option is kept as a future measure if loads grow faster than forecast.

Oxygen and alkalinity

The average oxygen demand rises to about 6,400 kg/d, with carbon removal and nitrification partly offset by the credit for denitrification. With a peaking factor of 1.5 for diurnal and seasonal load peaks, the blowers and diffusers must deliver about 9,500 kg/d of oxygen under process conditions, roughly twice the present demand. The existing blowers are 40 years old; new high-speed turbo blowers with dissolved oxygen and ammonia-based aeration control will meet the demand at lower energy than the existing units use today.

The effluent alkalinity, after nitrification consumes 7.14 mg/L per mg of nitrogen, denitrification returns 3.57 and the alum consumes about 0.5 per mg dosed, comes out at about 166 mg/L as CaCO₃, well above the 50 to 70 mg/L needed for stable nitrification. The hard prairie water is an advantage here; a plant on soft water would need to add alkalinity.

Secondary clarifiers

With the MLSS held at 3,200 mg/L, the peak overflow rate stays at the top of the normal range and the peak solids loading reaches the 7 kg/(m²·h) limit with no margin. State point analysis with the plant's measured settling data confirms that the clarifiers would just cope at peak flow when the sludge settles well, but fail when the SVI rises in spring. The design therefore adds a fourth clarifier of the same size, which lowers the peak loadings by a quarter and provides redundancy. The A2O configuration's anaerobic selector should also improve settleability, which is a benefit, not a design basis.

The solids train and side streams

The upgrade changes the solids. Biological phosphorus removal raises the phosphorus content of the WAS; the longer SRT reduces biomass production but the chemical trim adds solids; the fermenter converts part of the primary sludge's carbon to fatty acids before digestion. Nadia's mass balance, using the methods of lesson 11, shows a net solids production close to today's, about 7,600 kg/d to the digesters, so the two existing digesters, once refurbished, have enough capacity at 20 days. The phosphorus released in digestion would form struvite and cycle back in the centrate, so the design adds a struvite recovery unit on the centrate ahead of the anammox side-stream reactor, which together remove about 85 % of the centrate's ammonia and much of its phosphorus.

Risks, uncertainties and the operators

A design check is not complete without its risks. Nadia's register lists the main ones: industrial loads from the food processor, which could grow; cold winters below the 12 °C design temperature during long spring melts; the reliability of biological phosphorus removal during wet weather, when dilute influent carries little volatile fatty acid; the PFAS content of biosolids, which could restrict land application; and the construction sequence, since the plant must keep meeting its current limits while tanks are taken out of service. For each, the register gives the consequence, the mitigation in the design and the monitoring that will show whether the risk is materializing.

Finally, the design is reviewed with the plant's operators. Their comments change several details: access platforms to the new mixers, sampling taps at each zone boundary, a bypass so that the fermenter can be taken off line, and on-line ammonia, nitrate and phosphate analyzers so that they can see the process respond. The operators will run the plant for decades after Willowbend has moved on; designs that ignore them perform worse than their calculations predict.

The report is sealed by Martin as the engineer responsible and submitted to Manitoba's Environmental Approvals Branch with the application for the amended licence. Nadia records the hours toward her own professional registration and starts, as most engineers in training do, on the next project.