Question 2 of 5: Chlorine chemistry and pH; water distribution system layouts

Question 2: Chlorine chemistry and pH; water distribution system layouts (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.

(a) Influence of pH on disinfection efficiency (12 marks)

Given. Free chlorine in water at 25 °C; the hypochlorous acid dissociation constant pK_a = 7.54; the hypochlorite ion is approximately two orders of magnitude weaker as a biocide than the undissociated acid.

Find. The chemistry that links pH to germicidal efficiency, with the distribution of free chlorine between its two forms evaluated across the pH range a Canadian distribution system actually operates in.

Approach. Write the hydrolysis and dissociation equilibria, express the fraction of the stronger species as a function of pH, evaluate it, and then translate the shift into required contact time.

Chlorine gas dissolved in water hydrolyses essentially completely at the pH and concentrations used in water treatment:

The hypochlorous acid so formed is a weak acid, and it is this second equilibrium that pH controls:

Free available chlorine is the sum of the two species, and rearranging the equilibrium expression gives the fraction present as hypochlorous acid directly from pH:

Substituting pK_a = 7.54, the fraction present as the strong species collapses across the ordinary operating band:

At pH 6 the residual is almost entirely hypochlorous acid; at pH 9 it is almost entirely hypochlorite. The ratio of the two extremes is

so raising the pH from 6 to 9 destroys twenty-nine thirtieths of the active disinfectant without removing a single milligram of chlorine from the water. The efficiency loss follows because HOCl is a small, electrically neutral molecule that diffuses through the lipid cell membrane and attacks intracellular enzyme systems, whereas the negatively charged hypochlorite ion is repelled by the similarly charged cell surface and must act at the outside of the cell. Reported germicidal potency ratios range from about 40:1 to 100:1 for bacteria; taking 80:1, the free-chlorine potency at pH 7 relative to pH 8 is

which is the practical statement of the effect: a plant that lets its clearwell pH drift from 7.0 to 8.0 must roughly triple its CT — by tripling the residual, tripling the contact time, or some combination — to hold the same log inactivation. This is precisely why the disinfection CT tables in Health Canada's Guidelines for Canadian Drinking Water Quality and the associated provincial guidance are tabulated against pH as well as temperature, and why the required CT for Giardia at pH 8.5 is roughly double that at pH 6.5 at the same temperature.

Two competing considerations stop a designer from simply acidifying. Lime softening leaves the water at pH 10 to 11 and it must be recarbonated before chlorination for the reason just given; but the finished water also has to be non-aggressive to the distribution network, so a target of roughly pH 7.5 to 8.0 with a positive Langelier index is usually adopted, and the loss of disinfecting power is bought back with a longer clearwell contact time. Where chloramines are used as the secondary disinfectant the pH preference reverses: monochloramine formation is favoured above pH 8, and dichloramine and the associated taste and odour problems appear below about pH 7, so a chloraminating utility deliberately holds the distribution pH high.

(b) Requirements of a distribution system; grid iron versus dead-end (13 marks)

An adequate distribution system must satisfy the following requirements simultaneously.

1. Quantity and pressure. It must deliver the maximum hour demand, and separately the maximum day demand plus fire flow, at a residual pressure that in Canadian municipal practice is normally not less than 275 kPa (40 psi) under peak-hour conditions and not less than 140 kPa (20 psi) at the hydrant under fire flow, while not exceeding roughly 700 kPa anywhere so that fittings and service lines are not overstressed.
2. Water quality preservation. The pipe materials, linings and appurtenances must not degrade the treated water; the layout must avoid stagnation so that a disinfectant residual is maintained throughout; and the system must be watertight, with valves and air relief arranged so that a depressurisation event cannot draw contaminated water in.
3. Reliability and redundancy. A single main break, or a length of main taken out for repair, should not cut supply to a large area; this requires looped mains and sufficient valving to isolate short sections.
4. Storage. Elevated or ground storage must be distributed so as to balance diurnal demand, provide fire-flow reserve and emergency supply, and stabilise pressures.
5. Fire protection. Hydrant spacing, main sizing (typically a 150 mm minimum in residential areas and 200 mm or more on hydrant runs) and looping must be sufficient for the fire flow determined for the land use served.
6. Operability and economy. Adequate valves, hydrants, blow-offs, meters and access for flushing, pigging and leak detection; a layout that minimises the sum of capital and pumping cost; and capacity for the design horizon's growth.

Grid iron (looped) system. Mains are laid out as interconnected loops so that every junction can be reached from at least two directions. Its advantages follow directly from that redundancy: water reaches a demand point by several paths, so head losses are lower and pressures more uniform for the same pipe sizes; a break or a planned shutdown can be isolated between valves with only a very small area losing supply; there is no stagnant water, so a disinfectant residual is maintained and taste, odour and nitrification problems are far less likely; and during a fire the hydrant draws on the combined capacity of several mains at once, which is usually decisive in sizing. The disadvantages are cost and complexity: more pipe length and many more valves and fittings for the same area served, and a network whose flow distribution is indeterminate and must be solved iteratively (Hardy Cross, or a modern network solver) rather than read off a simple branch calculation.

Dead-end (branched or tree) system. A single trunk main feeds sub-mains, which feed branches, which terminate. Its advantages are simplicity and economy: the least pipe length and the fewest valves for a given area, flows in every pipe are determinate and can be computed by hand, and it adapts naturally to ribbon-development along a road or a valley where a loop would serve nobody. Its disadvantages are the mirror image of the grid's advantages. Every consumer downstream of a break loses supply until the repair is complete; water stagnates at the many dead ends, where the chlorine residual decays, sediment accumulates and taste, odour and bacterial regrowth problems concentrate, so those ends must be flushed on a routine schedule, wasting water; and the fire flow available at any point is limited to what the single feeding main can carry, which usually forces larger pipe sizes than a loop would need and can still be inadequate.

In practice a municipality uses both. Trunk mains and the arterial framework of any developed area are looped, because reliability and fire flow govern there, while individual cul-de-sacs and low-density fringes are branched because looping them would buy nothing; good practice is to keep dead-end runs short, size them for flushing velocity, and loop them as soon as adjacent development makes the connection possible.