On using pipes as header tanks and load levelers.

This page is about using pipes storage to improved the performance of a rainwater tank system which uses solar power for improving supply pressure. Relating to this is a bit about load leveling to reduce presure drop over long runs of thin pipe.

Using a water pipe as a header tank.
In many cases gravity fed water from rainwater tanks is problematic. The crux of the problem is that most rainwater tanks are filled from rain falling on the roof of the house, so the tank needs to be below the level of the roof gutter. Unless the roof is much higher than the places to water is used (ie shower head) there isn't enough pressure (static head) to get a decent flow rate. Also the pressure changes drastically with the level of the tank. There are many solutions including electric pressure pumps and header tanks. When I volunteered to install a header tank in the Nimbin rocks community house my first idea was to place a small (say 20 litre) tank in the ceiling space, this idea was rejected because the water would get hot. Then I concidered placing a tank on top of the roof, this might work better. Another idea was to use two tanks in tandem, one would be high and the other would be low. The top tank would supply pressure while the low tank which was cooler would supply the taps. I took this one step further by using one long skinny tank with a cool bit at the bottom. This is basically a pipe. The current plan (not tried yet oct-98) is to run a 6 meter 100mm pipe from the floor of the house thru the ceiling, thru the peak of the roof and a meter or so skywards. Water from the supply tank will be pumped into the bottom by a diaphram pump powered by the house's solar charged battery bank. The water supply to the kitchen and bathroom basin will come from the bottom of the pipe. Water to the solar hot water system may end up coming from the top of the pipe but I'm not sure yet. It's also possible a secondary header may be added for the shower cold water supply but it's unlikely to be needed. If the pipe is truely vertical all the weight will rest on the floor - this will be around 50 kilos. In the long term it is possible we will place a large header tank up the hill from the house. If this happens the pipe can be used for load leveling. In the short term - the pump will be controlled by a level dectector at the top of the pipe. In the long term - pump control may be given to a micro-controller which can detect fault conditions such as battery under voltage.

My level detector.
After spend too much time thinking about level detectors I came up with a very simple solution. I hung a 100 mm bottle full of water off the paddle of a micro-switch. The micro needs about 50 grams to pull it in. The bottle suspended in air weighs perhaps 130 grams easily switching the microswitch on. When the bottle is submerged in water it only weighs 30 grams or so and the micro turns off. The switch has some hysterisis built in so the water level would have to drop a bit before the pump turned on again. So my answer to the float switch is the sink switch. Note the wire or whatever is tying the sinker to the micro is always in tension. For the "FULL" level sensor in a properly functioning system the wire would be in tension even if we used a float - the float would have to weight more than 50 grams and the pump would turn off before it could start floating. However if we want to detect when the supply tank is empty (or any other level) then using a sinker has it's advantages - the micro-switch can remain high and dry above the high-tide mark and the sinker can hang at the end of a long line going to the bottom of the tank. For the pipe header I ended up making a float/sinker by filling an empty plastic 100 gram salt bottle with silicone. This ended up being neutrally bouyant and seems a good match for the switch I'm using and should continue to work even if the spring inside the micro weakens a bit over time.

What about using a big header tank?
Large header tanks placed up on a hill or tower are often used, sometimes on a large scale by water utilities. Using a tank that can hold days to weeks supply is handy if the system is manually control by running a pump once in a while to fill it. It is also handy in a solar electric system because rather than storing energy in batteries to run the pump on demand you can pump water when the energy is available (sunshine) and reduce or eleminate the expensive battery bank. There is also merit in pumping slowly as the frictional losses in the pipe increases dramatically with flow rate (approximately cubed). Ideally one would measure the actual level in the header tank (not just empty/full) and let a micro-controller decide how much power to use for water pumping versus battery charging etc.

Using a water pipe for load leveling.
Loading leveling is a common practice where the peak demand for something (water, power etc) is higher than the average demand. In the case of domestic water usage (where water supply is limited, ie tank water) there are generaly short periods of high demand with long periods of no demand in between.
For example one building at Nimbin Rocks is supplied with water via a 25mm pipe coming from a tank aproximately 100 meters away. The ground is level so there is only a fall of 400mm or so from the average water level in the tank to the tap at the other end. The pipe is also kinked in places. The end result was a mere dribble of water came out of the tap. The typical uses for the water is for drinking, cooking, washing dishes, cleaning teeth and filling solar showers (the 8 litre plastic bag variety). Most of the time less than a litre is used at a time. To improve the peak flow rate I simply used a piece of 100mm plastic pipe as a load leveling device. The 100mm pipe is aproximately 1.8 meters long and sits on the bench near the tap and extends upwards till it almost touchs the roof. It has to be long to cope with the full range of possible levels in the supply tank. A 19mm flexible polypipe runs from the bottom of the 100mm pipe to a T piece at the back of the tap. The large pipe fills slowly till it matches the level in the supply tank and empties rapidly when the tap is opened. Cheap and simple as it is - it increases the flow rate when filling the kettle by a factor of 10 or so. It was well worth the hour or so it took to make it.

Also see. http://www.watermagazine.com/
Hits since 29/oct/98 =

Back to Eddie's home page