From: http://www.ata.org.au/65watwel.htm
This pump
follows on from an idea taken out of earlier ReNew magazines, (issue 59,
page 60 and issues 60, page 8) and refers to a form of positive displacement
pump made from a single length of coiled polypipe. I have designed this pump
to be driven by a water wheel and suggest that they are a perfect match as
a water pumping unit.
For readers unfamiliar with the earlier descriptions of how this pump works, I will describe it briefly. The pipe is coiled in a vertical plane and is mounted on a horizontal axle. When the bottom quarter of the coil is immersed in water and the whole coil is rotated, an alternating sequence of air and water will be driven along the pipe towards the centre point of the coil. Additional successive coils of pipe lead to a cumulative increase in the pump's output pressure. When a reticulation pipe is connected to the end of the last inner coil, the water can be shifted to a higher point, such as a dam or tank on a hill. Add to this a set of paddle wheels and place it above a flowing river and you have one of the oldest and simplest forms of motor, driving one of the oldest and simplest forms of pump. The whole unit consists of one rotating part.
The working example in the photo is one which I fabricated for Jill Redwood, a dedicated environment campaigner living along the upper reaches of the Brodribb River in far east Gippsland. The 'spiral pump' was recently adapted to the water wheel to replace an old small piston pump, which was driven by a series of chains and cogs to give a 1:4 stepup ratio. Although this setup worked quite well, the pump was prone to mechanical failure and there was a possibility of pump oil leaking into the pristine Brodribb River. This new setup has several advantages; it is very environmentally friendly, it is made of basic materials(3/4" polypipe and a small amount of 25 x 25 x 3mm angle iron), it is relatively easy to make and it is not expensive.
One of the primary problems that I faced when assembling the coil of pipe on its spoked frame was to be sure of just how many coils, and at what diameter, would be sufficient to force water to the 16 metre head at this site. I suspect that there is a real lack of data on the performance of this type of pump. No doubt an opening for some research!
As Jill's property is a three hour drive from my workshop, I had to make a guess at what would work and hope for the best. The water wheel is around two metres in diameter, and as both the open end of the coil and the paddles need to dip into the stream, the spiral diameter needed to be this size also.
Three quarter inch polypipe can be coiled down to around 500mm in diameter before it starts to kink, and if close coils are laid over the angle spoked frame, then a total of around 40 coils can be made. I decided that that 40 coils would be excessive and instead made two sets of 20 coils which tee into a single outlet at the centre. Theoretically, this arrangement would pump twice as much water as a single coil which rotated at the same speed.
Well, as it turned out, 50 metres of 3/4" polypipe, coiled into 20 loops from 0.5m to 2.0m in diameter, is sufficient to pump water to a 16 metre head. But the paddle power (torque) of the water wheel was insufficient to drive both of the coils of pipe. We had to block one inlet end before the unit would spin at all.
My theory is that to pump water up to a head of 'X' metres, the length of coiled pipe needs to be around '3X' metres. I assume here that the size of the loops is not so important, but rather the total lenght of the pipe. Large loops are more effective at forcing water up than small loops, but consume more length of pipe. Many small loops may be just as effective as fewer large loops.
There is one aspect of this water wheel pump setup which throws a degree of complexity into it's design, and I will describe it only briefly. As with most creeks and river systems, there are times when flood waters may rise metres above the optimum level for a water wheel or pump. To avoid flood damage to this pump, I mounted the axle and bearings of the water wheel onto a three metre long boom made of 100mm square RHS (rectangular hollow section-square pipe), which can pivot at the opposite end to the water wheel. There is a height-adjustable support at the half-way point and a steel cable attached to the water wheel end. This allows for the whole unit to be winched up vertically at flood time, or for paddle height adjustments.
Herein lies the complication. In order for the water to get from the end of the smallest polypipe coil to the delivery pipe along the boom, the water needs to go through the centre of the shaft which supports the water wheel. The water, upon leaving the end of this axle, must pass through a rotating union.
A few additional pointers relating to construction are:
To attach the polypipe to the angle iron spokes, use 1.0mm stainless steel wire (available from apiarist's suppliers, or use cable ties instead). The end of the 3/4 inch polypipe which scoops up the water should be increased in diameter for the last loop. I used a one inch polypipe half loop, then a 11/4 inch for the last half loop. This allows for a greater volume to be scooped up each rotation.
As both water and air are pumped up the delivery line, it is best to send the pumped water directly to a storage tank or dam, otherwise a special air-bleed line is required.
There are some other variables which allow this design to pump effectively. These are: river water flow rate (speed), size of paddles, number of paddles, diameter of water wheel, diameter of coils, diameter of polypipe, number of coils, submergence of coil set, pipe mouth diameter and rise in delivery line. As this spiral pump was a direct replacement for a small standard piston pump, it was found that the volume of water pumped per day was much the same, so its efficiency of operation is very acceptable. The pumping rate at this site averages 2000 litres per day but varies according to the stream flow. Overall, a beautiful piece of alternative technology.
Interested: