The Sewer Flushing Siphon Chamber

A cistern siphon is started by a person lifting water over the crown. A flushing siphon must start with nobody present, and how it does that is the whole device. THE PRINCIPLE. A bell siphon has an inverted U — or a bell over a standpipe, which is the same thing — with its outlet lower than its inlet. As the chamber fills, water rises inside the bell and air is trapped above it. The trapped air is compressed as the level rises. When the water outside reaches the trigger height, the compressed air is forced out under the bell's skirt in a rush, water follows it over the crown, the leg fills, and the siphon primes itself. Once primed it runs until the chamber level drops to the skirt and air is admitted, which breaks it cleanly. BUILD IT. A 20 litre bucket as the chamber. Inside it, a standpipe of 32 millimetre pipe rising from an outlet in the base to a crown 250 millimetres up. Over that, a bell made from 75 millimetre pipe closed at the top, its open skirt ending 30 millimetres above the chamber floor. The bell must be a loose fit over the standpipe — a 20 millimetre annular gap — so the water has a path up and the trapped air has somewhere to be. The two dimensions that decide whether it works: the crown height sets the trigger level, and the skirt height sets the break level. The difference between them is the volume that gets discharged, and it must be the whole useful contents of the chamber or the flush is feeble. FEED IT SLOWLY. A tap dribbling at 200 millilitres per minute, or a peristaltic pump. Slow is the point — this device exists precisely to convert a slow, useless inflow into a fast, useful one, so feeding it fast proves nothing. NOW WATCH. The level rises, and for several minutes nothing happens. Then the level passes the crown, and there is a moment of gurgling as air escapes under the skirt. Then the siphon primes and the bucket empties in a few seconds with a strong solid flow. Then it stops dead, and refilling begins. THE COMMON FAILURE, AND IT IS WORTH CAUSING. If the annular gap is too large, air escapes gradually as the level rises instead of being held and released in one go, and the siphon dribbles rather than priming — it passes water at roughly the inflow rate and never surges. That is a snoring siphon, it is exactly what the device must not do, and the fix is to reduce the gap or lengthen the skirt so the air is genuinely trapped. Make it happen deliberately by propping the bell up a few millimetres, then fix it, so you know the symptom. MEASURE THE DISCHARGE. Time the flush and catch it in a graduated bucket. Divide volume by time for the mean discharge rate, and compare that with your 200 millilitres per minute inflow. A ratio of 50 to 200 is normal. That ratio is the entire value of the device: it does not create water, it concentrates it in time.

A flush is only useful if it scours. Build the sewer and prove it. THE TEST SEWER. Four metres of 75 millimetre clear pipe, laid at a shallow gradient — 1 in 100 is a realistic small sewer and it is deliberately too flat to be self-cleansing at low flow. Clear pipe, because the entire experiment is watching what happens to the solids. Support it every 500 millimetres so it does not sag; a sag becomes a silt trap and confounds the result. CHARGE IT WITH SOLIDS. Use a defined surrogate rather than anything real: coarse sand as the grit fraction, and soaked paper pellets or lentils as the organic fraction. Distribute a measured mass evenly along the invert. RUN THE BASELINE FIRST. Feed the pipe with your slow 200 millilitres per minute inflow directly, bypassing the siphon, for thirty minutes. Watch what happens. The water finds a shallow narrow stream in the bottom of the round pipe, moving far too slowly to lift anything. Some fine sand creeps. Nothing else moves. Photograph the invert and weigh what has arrived at the outlet — it will be almost nothing. That is the problem the whole subject exists to solve, and seeing four metres of solids sitting still under a running flow is a better argument than any calculation. NOW FIT THE SIPHON at the head and let it cycle. Each flush sends a wave down the pipe. Watch a single wave: it arrives as a distinct front, the depth jumps, the velocity is visibly high for a couple of seconds, and solids are picked up and carried. Then it passes and the pipe returns to a trickle. MEASURE THREE THINGS. The wave celerity — time the front between two marks a known distance apart. The peak depth, read against a scale taped to the pipe. And the mass of solids delivered to the outlet per flush, caught in a sieve and dried. RUN TEN CYCLES and plot cumulative solids recovered against flush number. It should rise steeply and then flatten as the easily moved material goes and the rest stays. That flattening is important and is the honest limit of the method. A flush wave scours what it can reach at the velocity it achieves. Material that has consolidated, or that sits beyond the distance the wave stays fast enough, does not move. The wave decays as it travels — it spreads out and slows — so a flushing siphon protects a limited length of sewer downstream, typically a few hundred metres at full scale. Beyond that you need another one. An egg-shaped sewer has no such limit. Its low-flow velocity is a property of every metre of the pipe, all the time, for as long as the sewer exists. That is what the extra capital cost buys, and now you have measured the thing it is buying.