By Stephen Rodgers
Lake Plan View, Waterfall Seal Detail Version1, Waterfall Seal Detail Version 2 ~10.4 GIFs
As the engineering for the sump became more complicated and less physical, my digging efforts had been transferred 2 metres up the garden, to the lake. At the edge furthest from the kitchen the level of the soil had been built up to nearly 60cms above the patio level. I wanted the water level in the the lake to be just below the patio, I wanted a waterfall and along the rest of that face I wanted a 'cliff' or rockface effect. All of this is pretty conventional aquatic gardening, and in my last pond I had had a problem concealing the polythene edges, so in this one I resolved to conceal all the edges by supplying other, more attractive edges.
The lake is all straight sided with the deepest point along the edge of the patio. The waterfall takes care of some of the edges, and for the others I used concrete gravel boards, set in cement on top of and inside the polythene. The plastic supplies the water seal, the concrete supplies the aesthetic edge.
These gravel boards can be found in the same section of the builders merchants as the slotted fence posts. To build a fence you are supposed to use a line of posts, with a 'gravel board' dropped down the slots to form a bottom edge, the wooden fence panel then sits in the the slots, on top of the gravel board, without being in contact with the soil. The boards come in various designs, some plain, some more ornate. I chose quite an ornate model.
The waterfall was built onto a brick foundation, in rather pretty 'mini bricks', which are identical to real bricks, but half the dimensions. I wanted the water to fall into as deep water as possible, to give the fish something to play with, so it was built on the edge of the foundation, with a small fillet of mortar right at the base supplying a curved edge for the polythene to sit on.
To date, all my bricklaying had been with old used bricks. The mini bricks were new and very dry - if you apply mortar to dry bricks it dries and becomes like sand. Soak all new bricks before use.
Splashing quite a lot of water around when bricklaying is quite a good idea - provided you have a good wire brush to hand. Just before the mortar sets hard this can be used to remove mortar splashes and generally clean up the face of the wall.
With the wall built, the hole lined with newspaper and the polythene liner in place but not trimmed, I used the pump for the first time 'in anger' to temporarily fill the lake. This allowed the liner to take up the shape of the hole. I then pumped most of the water back into the sump, put a lot of concrete onto the polythene at edges B, F & G, and set the gravel boards into place.
When that was set, I drained the pool completely and threw a 5cm layer of coarse concrete into the bottom. When, in turn, that was set I was able to stand on it and set the edge of the trimmed liner into concrete, carefully insuring that the polythene had concrete on both sides of it, to provide a waterproof seal. A lot of care was taken to produce a beautifully sculpted and smooth fillet, with the waterfall set to fall into at least 10cm deep water.
The other complication was the overflow drain which takes the water from the lake, and feeds it back into the sump. I had set a piece of timber 1cm * 5cms along edge A, with a 15cm notch cut 2cms deep towards one end. I then nailed the liner (above the waterline) to this edge, pulled the excess liner into the notch, and sculpted a drain to connect the two bodies of water together. This all lives under one paving slab.
It looked beautiful. The only polythene edges in view would be buried when the garden was finally dug over and the rockery was constructed, and the waterfall stood majestically over it, it only remained to be tested. To test the lake itself I simply filled it and left it for two or three days without detecting a drop in level. Then I started pumping from the sump straight into the lake, the lake overflowed, through its drain and into the sump, I went to bed.
The following morning, soundings were taken which revealed that I had lost 900 litres in 9 hours. It couldn't be a puncture in the lake, it had to be the drain. I thought about it for the rest of the week (having stopped pumping), and in the following weekend I rebuilt the drain, and applied a lot of bathroom sealer all along edge A, all over the nails, and particularly around the notch. Pumping resumed, and testing revealed the same losses. I decided that the leak was all along edge A, as some of the nail heads (by now covered in sealer) were below the waterline. I bought another piece of timber, treated it, and applied almost a whole tube of sealer along it. I then screwed it, through the polythene, to the other board, using about twenty screws to supply an even pressure. Squelching, popping noises were heard and bubbles and beads of sealant exuded all along its length. It didn't look very nice, but I could see an unbroken line of sealant along the length of edge A - it simply couldn't leak there.
In despair I drained the pool and examined it in minutest detail. It was then that I realised that I had fallen victim to the 'out of sight is out of mind' school of engineering. At the base of the falls I had cut the polythene below the waterline, but encased it totally (I thought) in concrete. In fact NOTHING sticks reliably to polythene, the concrete must have shrunk in setting, and the water was running under the concrete and over the edge of the polythene.
With great reluctance the hammer and chisel were produced, and days of work were destroyed as the gravel boards and the lovely concrete fillet were removed, along with the sealant covered timber. The pool was then relined, leaving much more polythene spare at the base of the falls, the gravel boards were reset, and the drain was rebuilt. In fact this process was completed twice, because this river engineer's hand slipped, cutting the polythene too low. You didn't want to know that.
This time the base of the falls looked rather different. The bottom course of mini bricks, which used to be half in the water, is now obscured by a fairly thick fillet of concrete which rises above the water line. This is still there to hide the edge of the polythene, and keep it in place, but it is no longer expected to supply a waterproof seal. On the positive side, the water still falls into deep water, as it tends to fly out quite a long way from the wall.
'Out of sight is out of mind' engineering is also still prevalent in the drain from the sump to the outside drains, where I am sure a leak must occur. I have no way of measuring this leak, but I do not really mind. The drain only operates when the system is totally full, so provided that most of the water is finding its way to the drains, it is all right by me.
With the sump, pump, lake and drain all established, the pump was now left turning continuously. In a small way, the river had started to flow.
Return to Index