16-Civ-B5 Water Supply and Wastewater Treatment · May 2016

Question 4 of 5: UV Disinfection, Rapid Sand Filtration and the Settling Regimes

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

Question 4: UV Disinfection, Rapid Sand Filtration and the Settling Regimes (25 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. Representative design values are attached to each part so that the description is anchored to a computable result.

Find. The mechanism, design basis and limitations of UV disinfection; the construction, operating cycle and hydraulics of a rapid gravity sand filter; and the physical distinction between the three settling regimes, with the design parameter that each one implies.

Approach. Each part is answered by first identifying the physical mechanism, then reducing it to the one design equation an engineer would size a unit with, then evaluating that equation on the representative data.

(a) UV radiation based disinfection (8 marks)

1. Mechanism. Ultraviolet disinfection is physical, not chemical. Radiation in the germicidal band near 254 nm — the principal emission of a low-pressure mercury lamp — is absorbed by the pyrimidine bases of microbial DNA and RNA, and the absorbed energy forms covalent bonds between adjacent thymine or cytosine bases. These pyrimidine dimers distort the helix and block transcription and replication. The organism is not destroyed and not visibly changed; it is rendered incapable of reproducing, which is what “inactivation” means and why UV-treated water contains intact but harmless cells.
2. Design basis. The controlling variable is the delivered fluence, or dose, the product of the germicidal irradiance and the exposure time: $$D = I\,t$$ On the representative reactor, $$D = 12\ \mathrm{mW/cm^2}\times 3.5\ \mathrm{s} = \boxed{42\ \mathrm{mJ/cm^2}}$$ which meets the 40 mJ/cm² dose commonly adopted in Canadian practice as a general design target. Doses are not calculated from first principles in practice but assigned by biodosimetry: a full-scale reactor is challenged with a test organism, and the result is expressed as a reduction-equivalent dose in accordance with the USEPA Ultraviolet Disinfection Guidance Manual validation protocol. Sensitivity varies enormously by organism — roughly 12 mJ/cm² gives 3-log inactivation of Cryptosporidium and Giardia, whereas adenovirus needs about 186 mJ/cm².
3. Strengths. UV is exceptionally effective against the chlorine-resistant protozoa, which is its decisive advantage: the 3-log Cryptosporidium credit that costs 12 mJ/cm² is effectively unattainable with free chlorine. It forms no halogenated disinfection by-products, adds nothing to the water, does not affect taste, odour or pH, has a small footprint, and responds instantly with no contact tank.
4. Limitations, and how they are managed. UV leaves no residual, so a distribution system still requires chlorine or chloramine downstream — UV replaces primary disinfection, not secondary. Performance depends on ultraviolet transmittance, and design normally requires UVT above 85–90 per cent, so turbidity, colour, iron and humic matter must be removed first; particles also shield organisms directly. Quartz sleeves foul with hardness and iron deposits and need mechanical or chemical cleaning; lamp output declines over an 8000–12 000 hour life and an end-of-lamp-life factor is built into the design. Because the effect is photochemical rather than lethal, some bacteria can repair dimers by photoreactivation or dark repair, which is a reason to hold a chemical residual downstream. Finally, the process fails instantly on loss of power, so redundancy and alarms on intensity, flow and UVT are essential.

(b) Operation of a rapid sand filter (9 marks)

5. What the filter is for, and what it does not do alone. A rapid gravity filter is the final particle barrier of a conventional treatment train, polishing the settled water from typically 1–5 NTU to below 0.1 NTU. It only works on water that has been coagulated: an uncoagulated colloid passes through 0.5 mm sand as though the bed were not there, because removal is chiefly by attachment of destabilised particles to the media grains and not by straining.
6. Construction, from the top down. Figure 4.1 shows the standard arrangement: 1.0–1.5 m of supernatant water providing the driving head; wash-water troughs set above the expanded bed level; a dual-medium bed of coarse anthracite (effective size 1.0–1.2 mm, about 0.45 m deep) over finer silica sand (effective size 0.45–0.55 mm with a uniformity coefficient below 1.7, about 0.30 m deep); a graded gravel support; and a nozzled underdrain that collects filtrate and distributes wash water and air. The coarse-over-fine arrangement is deliberate — it uses the depth of the anthracite for storage and reserves the fine sand for polishing, giving longer runs than a single sand medium at the same headloss.
7. Removal mechanisms and the progress of a run. Particles are brought to the grain surfaces by interception, sedimentation within the pores and, for the smallest colloids, Brownian diffusion, and they are held there by van der Waals and electrostatic attraction once coagulation has neutralised their charge. As deposits accumulate the porosity falls and the headloss rises above the clean-bed value predicted by the Carman–Kozeny equation. A run ends on whichever of three criteria is reached first: terminal headloss, typically 2.4–3.0 m; turbidity breakthrough at the effluent; or a maximum run time set to prevent biological growth in the bed.
8. Filtration hydraulics on the representative cell. The plan area and the loading rate are $$A = 5\times 8 = 40\ \mathrm{m^2}, \qquad v = \frac{Q}{A} = \frac{240}{40} = \boxed{6.0\ \mathrm{m/h}}$$ comfortably inside the 5–15 m/h band of rapid filtration and two orders of magnitude above a slow sand filter. Rate control matters as much as rate: a constant-rate filter throttles an effluent valve that opens progressively as the bed clogs, so the flow — and therefore the plant's hydraulic balance — does not change during the run.
9. Backwash, and its cost in water. Cleaning reverses the flow. Air scour at 40–60 m/h first breaks the deposits loose, then wash water at 37–45 m/h fluidises the bed to 20–30 per cent expansion for 6–10 minutes, and the released solids spill into the troughs. On the representative cell, $$V_{\mathrm{bw}} = 40\ \mathrm{m^2}\times 40\ \mathrm{m/h}\times\frac{8}{60}\ \mathrm{h} = 213\ \mathrm{m^3\ per\ wash}$$ against a daily production of $240\times 24 = 5760\ \mathrm{m^3}$, so the wash consumes $$\frac{213}{5760}\times 100 = \boxed{3.70\ \text{per cent of production}}$$ which is why spent backwash is settled and recycled to the head of the plant rather than wasted. After a wash the bed is briefly less effective, so the first few minutes are normally filtered to waste, or the run is started at a reduced rate, to avoid the ripening turbidity spike reaching the clearwell.

(c) Discrete, flocculent and hindered settling (8 marks)

10. Type I — discrete settling. In a dilute suspension of particles that do not change size or shape, each particle settles independently at a terminal velocity fixed by the balance of gravity, buoyancy and drag. In the laminar regime this is Stokes' law: $$v_s = \frac{g(\rho_s-\rho)d^{2}}{18\mu}$$ For the representative 0.05 mm grain, $$v_s = \frac{9.81\times(2650-998.2)\times(5.0\times 10^{-5})^{2}}{18\times 1.002\times 10^{-3}} = \boxed{2.25\ \mathrm{mm/s} \;=\; 194\ \mathrm{m/d}}$$ The Reynolds number, $Re = \rho v_s d/\mu = 0.112$, is well below unity, so the Stokes assumption is valid. The design consequence is the overflow-rate concept: an ideal basin removes every particle whose settling velocity exceeds the critical velocity $v_c = Q/A_s$, and removes a fraction $v_s/v_c$ of the slower ones. Both statements are independent of depth, which is why grit chambers and presedimentation basins are sized on surface area.
11. Type II — flocculent settling. When particles are sticky — chemically coagulated floc, or the organic solids of raw sewage — they collide as they settle, coalesce, and grow. Because the settling velocity increases with the square of the diameter, a growing particle accelerates as it falls, and the longer the column of water it traverses the faster it becomes. Depth therefore matters: removal is a function of both detention time and depth, and no single settling velocity describes the suspension. There is no closed-form law, so the design is based on a settling-column test with sampling ports at several depths, from which iso-removal curves are drawn and scaled to the full-size basin with allowances of roughly 1.25 on detention time and 0.65 on overflow rate. Primary clarifiers and the upper region of a coagulation-sedimentation basin operate in this regime.
12. Type III — hindered or zone settling. Above a solids concentration of roughly 1000 mg/L — the condition in every secondary clarifier, where the mixed liquor enters at 2000–4000 mg/L — the particles are close enough that the water they displace must flow upward through the interstices between them. The particles then hold fixed positions relative to one another and the whole mass subsides as a blanket, with a sharp interface between clarified supernatant above and settling sludge below. The settling velocity is a property of the concentration, not of the individual particle, and it falls as the concentration rises. The batch curve on the right of Figure 4.2 shows the signature: a straight constant-velocity portion, whose slope is the zone-settling velocity, followed by a knee into compression. Below the hindered zone lies Type IV, compression settling, in which the particles are supported by the structure beneath them and consolidation proceeds slowly as water is squeezed out.
13. Why the distinction changes the design parameter. A Type I or Type II basin is sized on surface overflow rate alone. A Type III clarifier must satisfy two independent requirements: the overflow rate must be low enough for clarification, and the solids loading rate must be low enough for thickening, since the tank must also deliver a return sludge concentrated enough to maintain the reactor inventory. The thickening requirement usually governs, and it is evaluated by solids-flux or state-point analysis rather than by an overflow rate — which is exactly the calculation Question 5(b) requires.