Third-Floor Showers Weak and Coliform in the Tank Water

Pressure: this is a head-and-friction problem, not a tank-height problem

• Head available at a third-floor shower: tank water level about 13 m above ground (11 m roof plus 2 m tank), minus the shower outlet at roughly 9.5 m, gives about 3.5 m of static head, i.e. about 35 kPa - before any loss.
• Losses: a single 20 mm line with six elbows serving eight fixtures is badly undersized; at any simultaneous draw-off the velocity and friction loss climb steeply, and the half-closed gate valve adds a deliberate extra loss.
• That is why the ground floor is fine (about 13 m of head there) and the third floor is not. The residual at the top outlet falls below the 1 to 2 m needed for a decent shower.

Judgement on the proposals

• Raising the tank by 1 m: adds only about 10 kPa (0.1 bar). Useful at the margin, but it does not fix a 20 mm line, adds dead load to the roof (needs a structural check), and costs more than it returns.
• Booster pump on the third floor: does restore pressure, but it treats the symptom, will run dry whenever the tank is low, needs its own electrical point and dry-run protection, and does nothing for the bacteriological problem. A pump pushing into a line shared with the lower floors can also over-pressurise them.

What actually solves the pressure

1. Open the gate valve fully; a throttled valve is not a pump-protection device.
2. Replace the 20 mm third-floor run with 32 mm, cut the elbows to two, and give each fixture a 15 mm branch off a manifold.
3. Only if residual pressure is still under about 1 m at the top outlet, add a small hydro-pneumatic booster set with a pressure vessel, or a break tank and pump.

Quality is a separate failure and no amount of pressure work will fix it - see next.

Pathways, ranked by likelihood

1. No air gap on the overflow. The overflow runs straight into the soil drain. Whenever the supply pressure drops or the line is emptied, drain water can be siphoned back into the tank. This is the most direct route to faecal coliforms and explains the sudden outbreak.
2. No vent. A tank that cannot breathe draws air in through the only opening it has - the overflow, i.e. through the drain. It also allows stagnation and mosquito entry.
3. Sump too close to the soak pit. At 2.5 m, and with a soak pit serving 30 users, lateral seepage through the soil into a cracked or unsealed sump is very plausible. Guidance for such separation is of the order of 15 m; 2.5 m is not defensible.
4. Nil free chlorine. There is no disinfection barrier, so any ingress survives to the tap. Turbidity of 8 NTU also shields bacteria from chlorine and indicates the ingress is ongoing.

Corrections, in priority order

• Re-route the overflow to discharge freely above the drain with an air gap of at least twice the pipe diameter (minimum 25 mm), and never pipe it into the drain.
• Fit a vent pipe with a mosquito-proof screen, taken above roof level and away from the overflow.
• Relocate or reconstruct the sump at least 15 m from the soak pit, with a sealed RCC lining and a tight cover; or relocate the soak pit.
• Clean and disinfect the tank and sump (scrub, drain, then 1 percent bleaching solution contact), then chlorinate the supply. Rough dosing with bleaching powder of about 33 percent available chlorine: around 3 g per 1000 L gives roughly 1 mg/L dose, so a 5000 L tank needs about 15 g; leave 30 minutes contact and check residual.
• Provide safe drinking water immediately - boiling or a certified filter - and tell residents why.

Why this order: the air gap and vent stop the ingress at source and cost almost nothing; relocation removes the reservoir; chlorination is the barrier that covers whatever remains.

Before the repairs are signed off

• Pressure test the altered distribution line (typically 1.5 times working pressure for a short duration) and flush it; check residual flow at the worst-case third-floor shower, not at a ground-floor tap.
• Visual check of the air gap, the vent screen, the tank cover and the sump sealing.

Water quality

• Bacteriological: coliform count by membrane filtration or multiple-tube method on a sample taken from the tank outlet and from a third-floor tap, after disinfection and after the lines have been flushed. Target: nil coliform per 100 ml.
• Chlorine residual: DPD test at the farthest tap; a residual of about 0.2 to 0.5 mg/L after contact time is the working range - nil chlorine means the barrier has failed again.
• Turbidity and pH: turbidity should fall well below 5 NTU (it was 8); pH around 6.5 to 8.5 so that chlorine works.
• Repeat the full set after two weeks: one clean sample can be luck, a clean series is a fix.

Records and routine

• Tank cleaning and disinfection log, twice a year, with date and person.
• Chlorine residual checked weekly at the far tap.
• Annual bacteriological test at a recognised laboratory, results kept on file.

Judgement to state plainly: the pressure complaint and the illness have the same root cause - a system designed and maintained without regard to the plumbing rules. Fixing only the pump would have left the residents drinking drain water at higher pressure. Report the coliform finding to the institution's management in writing; a hostel with 24 residents and 12 illness reports is a public-health event, not a maintenance ticket.