The Sedimentation Tank

Make a rectangular tank from clear sheet, roughly a metre long, two hundred millimetres wide and three hundred deep. Clear is essential — every interesting result in this blueprint is a flow pattern, and you cannot see a flow pattern through a wall. The inlet is the part that decides whether the tank works. Water arriving as a jet will shoot across the tank and out of the outlet in a fraction of the design time, carrying its particles with it, and the rest of the tank will sit there doing nothing at all. So the inlet must dissipate that momentum: feed into a full-width chamber behind a baffle with the water passing through a perforated plate or under a submerged wall, so it enters the tank body slowly and evenly across the whole cross-section. The outlet has the same problem in reverse. Draw off through a single pipe and you create a suction that pulls flow preferentially from nearby. Real tanks use a long weir across the full width, sometimes a V-notch weir, so the withdrawal is spread out. Make yours a full-width weir plate. Fit a sludge zone at the bottom — a hopper or simply a flat floor with a drain — and keep it well below the outlet. Settled solids that get re-entrained have wasted the entire process. Mark the tank at ten equal stations along its length so you can describe where things happen.

This is the measurement that separates the design from the reality, and it is the most valuable thing in the blueprint. Run the tank at a steady, measured flow until conditions settle. Calculate the theoretical residence time: tank volume divided by flow. Write it down. Now inject a slug of dye at the inlet, as fast and as compactly as you can, and record the dye concentration leaving the outlet against time. A photodiode and an LED across the outlet pipe gives you a curve; a series of samples in numbered jars gives you the same thing more slowly. What comes out is the residence time distribution, and it will not be a spike at the theoretical time. Expect dye to appear at the outlet far earlier than theory says — that is short-circuiting, water crossing the tank on a fast path. Expect a long tail — that is dead space, water trapped in corners and eddies, leaving late. The ratio of the time to first appearance against the theoretical time is a direct measure of how badly the tank is short-circuiting, and anything below about a third is poor. Then go and fix it. Add or improve the inlet baffle. Lengthen the outlet weir. Add intermediate baffles to lengthen the flow path. Re-run the dye test after each change and watch the curve tighten up. Most real tanks achieve considerably less than their theoretical residence time, and knowing your number is the difference between a tank you designed and a tank you understand.

Build a plate pack and drop it in. Cut rectangles of thin sheet to fit the tank cross-section, space them fifty millimetres apart with strips of the same material, and set the whole pack at about fifty-five degrees to the horizontal. The angle is a compromise you can test. Shallower plates settle better because the sludge has less distance to fall, but below about fifty degrees the sludge stops sliding off and the pack blinds with accumulated solids. Try forty-five, fifty-five and sixty-five degrees and see for yourself where sliding fails — that failure is the reason the angle is always in the fifties. Run the same clay suspension through the tank with and without the pack, at the same flow, and measure outlet turbidity. Then find the flow at which each configuration reaches the same outlet quality. The ratio of those two flows is the real gain, measured on your bench, and it should be several times. Where this stops. You have built a clarifier and measured its hydraulic efficiency, which is genuine process engineering and transfers directly to any separation by density — brewing, casting slurries, machining coolant, aquaculture, rainwater harvesting. What it is not is drinking water treatment. Settling removes what settles, and what settles is not what makes people ill. Bacteria and viruses are far too small to be caught by gravity in any practical tank, which is precisely why every process after this one in the batch exists. Clear water from this tank is clarified, not treated. Use it for irrigation, for washing, for process water, or as feed to a filter — and never assume the clarity means anything more than clarity.