Hydraulic Ram Pump

Run a tap through a length of hose and shut it off fast. Feel the hose kick.

That thump is the momentum of the moving water column being stopped. In a plumbing system it is a nuisance to be designed out.

This pump is built to cause it, on purpose, dozens of times a minute. Understand that before assembling, because everything else follows from it.

Assemble from pipe fittings:

Drive pipe — long and rigid, feeding the body.
Waste valve — spring or weight loaded, open at rest.
Delivery valve — a one-way check valve into the air vessel.
Air vessel — a closed chamber above the delivery valve.
Delivery pipe — small bore, runs uphill.

Two moving parts total. That is the entire machine.

Look at the closed chamber above the delivery valve. It is full of trapped air.

Water is incompressible, so without the air the pressure spike would arrive at the delivery pipe as a hammer blow and burst it. The air cushion absorbs each spike and releases it as steady flow.

A ram whose air vessel has become waterlogged — the air gradually dissolves into the water — starts hammering, delivers less, and eventually cracks fittings. That is the commonest ram fault and it is a maintenance item, not a design flaw.

Admit water. Work the waste valve by hand a few times to start the cycle, then let it run.

Count the beats per minute. Adjust the waste valve weight or spring and count again.

Expect a trade-off: a heavier valve takes longer to slam, so it beats slower but each stroke moves more water. Lighter beats faster with less per stroke. There is an optimum and it depends on your drive pipe length — which is why every ram is tuned on site.

Measure over one minute: volume delivered, volume wasted, drive head and delivery head.

Compute delivered ÷ (delivered + wasted). Compare against (H_drive / H_delivery) × 0.6.

Most of the water leaves through the waste valve, and that is not a fault — it is the energy source. The wasted flow falling through the drive head is what lifts the delivered fraction. A ram is a machine for trading quantity against height.

Two inventors, and the distinction matters. John Whitehurst built a manually operated precursor in 1772 at Oulton, Cheshire, raising water about 4.9 m — a "pulsation engine" worked by hand-turning a tap. He did not patent it and the details are thin. Joseph Michel Montgolfier — of the ballooning brothers — made it self-acting in 1796 at his paper mill at Voiron, by replacing the hand tap with a valve the water itself operates. Matthew Boulton took out the British patent on his behalf in 1797. Whitehurst had the mechanism; Montgolfier removed the human being from it, which is what made it a machine.

The trail afterwards. Montgolfier's sons patented an improved version in Britain in 1816, and in 1820 Josiah Easton acquired that together with Whitehurst's design — which is how both lines ended up in one English manufacturing tradition.

Why it keeps being rediscovered. It needs no fuel, no electricity and no attention; it has two moving parts; it can be built from plumbing fittings; and installations have run for over a century. For hill villages with a stream above and a settlement below, nothing else does the job as cheaply. It is a standard appropriate-technology water supply today for exactly the reasons it worked in 1796.

Honest limits. It needs a continuous flow and a real drive head — usually at least a metre, and a drive pipe several times that in length. It wastes most of the water, so it is useless where the supply itself is scarce rather than merely low. And it is noisy: it beats, audibly, day and night, forever.