Friday 15 May 2015

Leakier Than A Leaky Thing

Grrrr! Two afternoons pretty much wasted down in the van. I decided that I needed to get the slow sand filter system in and running as we were expecting rain and it would be a shame to waste it.

With a little hassle with delivery, I finally got all my guttering together and have fitted it to the van body. I'm not necessarily collecting water off the roof yet. If I decide to do that, I need to think about flashing the roof to the gutter. Right now, the gutter itself is my rainwater collector. With 18m of length and a width of about 114mm, that equates to a surface harvesting area of just over 2 square metres of collector - Enough to be getting on with.

I dragged the sand filter from the Toy Shed up to the van body, and with some heaving and grunting and lumps behind the ears, I got it located and set into position. A few modifications to change right handed to left handed plumbing and I was ready for a wet test.


Slow filter on the left - water tank on the right. From top to bottom, inlet pipe carrying rainwater in from gutter, overflow pipe for filter barrel and weir pipe (filter outlet to tank inlet).

It rained on Thursday, but not enough to get things rolling at any speed, so I hurried things along with a hosepipe. Stood on a stepladder outside, I could hear the water running into the tank through the guttering takeoff, and all seemed to be going well. Then, to  my dismay, I heard the unmistakable sound of water dripping onto the phenolic ply floor. Not good.

The collector pipe union with the outlet at the bottom of the filter butt was leaking like a leaky thing...so fast that the height of the water in the filter wasn't going up much. By the time the level reached the top of the weir pipe and started filling the clearwell tank, I figured I would have about 40 litres sloshing around on the loo floor. Despite my attempts to weld the union to the butt, I couldn't slow the leak down enough to conclude any testing, so I decided to drain the whole lot back out again (the tap on the side of the filter is there for that reason).

There is clearly some leakage involved with the water butts. My initial testing was carried out outdoors back in 2013 and a few leaks made no odds. Indoors, in the dry, this is going to be a real problem, so time to rethink the filter system. I think I can get away with the clearwell (storage tank). There is the slightest of weep around the copper tank connector, but nothing major to worry about.

Plan B, after a bit of discussion with Eco on the subject, is to make a thinner cylindrical filter out of PVC soil pipe and some solvent weld components and give that a go. There are two advantages to going down that road. Firstly, the filter unit is still about 1m tall (it's the height, not the volume that governs the working 'head pressure') but is only 110mm diameter so contains less gravel, sand and water. This makes for a much less heavy unit, and one that can be maintained much easier. Secondly, by using flexy couplings, it is possible to make the unit like a 'cartridge', so at end of run (3-6 months or so), the filter unit can be swapped for another and the first one taken down for maintenance and cleaning. If tests on the new Mk 2 system go according to plan, then I'm going to make a couple of spare units to allow for 'one on, one off and one in the wash'. Simples.

It may even be possible to double up the filter units, to increase the available water harvest if necessary. Two identical filter units running in parallel means the volume of water filtered is doubled.

Another advantage to using a much smaller diameter filter is that the working height will increase significantly for a much smaller volume of inlet water. In order to fill the butt filter from sand level to weir height the input volume needs to be some 60-70 litres. With a 110mm filter unit, the volume required is only 6 litres.

The new filter is made from a 110mm grey PVC soil pipe, a push fit access cap on the top end and a solvent weld blanking plug on the bottom. 21.5mm solvent-weld tank connectors will allow connection to the top and bottom of the filter housing using 3/4" nut flexy hoses (washing machine fill hoses). 


The lower outlet hose also becomes the weir pipe and will be connected to the threaded part of the tank connector on the clearwell. In order to carry out maintenance, this hose can be detatched from the clearwell inlet and drained into a bucket before the filter is then removed. 


The inlet pipe will again be flexy, allowing the top of the filter to be disconnected. With a screw-on access cap, the filter can be inspected and wet-harrowed in situ, or removed for complete overhaul and backflush as necessary.


If tests prove successful, I may even modify the inlet pipework to allow two filters to be connected. These can be operated individually or in tandem if necessary with a pair of inline valves (appliance valves). An individual filter can then be taken out if need be, leaving the other in place.


The collector pipe is a 20cm length of 21.5mm solvent-weld overflow pipe, drilled with an array of 3mm holes. This is welded to the inside of the lower tank connector and remains in place in the lower tube. The gravel needs to cover this tube, plus a few cm, and then there is a 40cm layer of sand over this as the main filter construction. A 20cm water level over the active 'schmutzdeck' provides the supernatent water for the filter unit. Oxygenation should be provided by the aeration of the inlet water as it flows from the gutter. The misshapen grey 'lump' on the end of the collector pipe is the dimple cut off the access plug during fitting the inlet connector. It has been heat-welded to the collector pipe with a soldering iron to prevent any of the filter media clogging up the tube.

All in all, a much better scheme than the original system. Lighter, faster, more compact and easier to maintain. To be fair, the prototype proved the principle a couple of years ago, so no harm - no foul.



No comments:

Post a Comment