Watering, can!

One of the things that attracted me most to our home when we first saw it was its established veggie garden, which included what looked like an extensive sprinkler system.   Soon after we moved in I set about learning how it worked.   There was no rainwater collection system and all the water came from the main (i.e. subscriber) supply.   The system itself was comprised of two zones, neither of which  worked particularly well.  In fact, there seemed to be some sort of cross-over -- a kind of short-circuit that allowed water from one to flow to the other --  between the two that caused neither to work particularly well on its own, but delivered water when both were turned on.   I tried to find this cross-over and concluded that it must have something to do with a siphon effect in the manifold, located at the main tap where the zones connected.  

The system had fallen into a bit of disrepair and I found most of the drip lines clogged or had heads that were missing, causing water to simply run out in a fast stream.  The main feed lines inside the garden were 13 mm black poly with smaller-diameter drip lines tapped into these which terminated in spray or drip heads.  I wanted to get something -- tomatoes at least -- growing during our first Summer and managed to clean out and re-head enough of the drips to get a few beds properly watered with the existing system.    I wound up removing many of the drip lines, plugging the 13 mm feeds, which helped to improve the pressure and flow to those lines that were still operational.   I knew, however, that this was only an interim solution and that at some point I'd probably have to re-engineer the whole system.

Part of this re-engineering would have to include a rainwater collection system.  We were still under Stage-2a water restrictions which limited the amount of water that we could use for the garden.    I installed such a system (see the Rainwater Tank page) last winter that included a 3400 liter tank and automatic pump, and diverted the line that feeds the veggie garden from the mains to the rainwater and pump.  Thus, I now have a supply of water restricted only by how much the tank(s) can hold and how often it rains.

Next, then, was the sprinkler system itself.  Our block is on a rather steep hill, the veggie garden being uphill from the house.  The collection tank is located under the house, so water must be pumped roughly uphill -- roughly 8 m -- to the sprinklers.   Even with the small (32 l/min) pump currently in use this is no problem and I get sufficient flow all the way to the top of the veggie patch.

The old system was basically a large loop, circumscribing the interior wall of the garden with one (or two?  three?) cross-over sections joining one side to the other.  The feeds entered at two separate points somewhere along the bottom of the main loop.   Branches then came out from the loop running along wires that were strung over each of the beds, suspended between metal fencing posts.   Sprinkler heads were then either directly attached to the branches, or on the ends of smaller tubes that hung down from the branches to put the heads closer to the ground.   Aside from the complexity of the system, its main drawback was the all-or-nothing nature of distribution.  Even if it was working perfectly, all the beds would be watered whether or not they actually needed watering.   I really wanted a system that allowed me to control not just whether a bed was watered, but how much water it received.   Some plants like a dryer soil, while others -- notably tomatoes -- prefer to be quite wet.

I decided to tear this all nearly all of the old system, keeping only the feed lines that came from the outside manifold to the garden itself.   In its place, I designed an inverted "U" system that would be supplied by just one of the two existing feeds, up at the top of the garden.  This feed entered the "U" precisely in the middle to help ensure both stems were supplied equally.   Each both stems run downhill, about a meter or so to the left and right, respectively, of center.    At each bed, I drop a branch down to ground level, with a valve near were the branch connects to the "U".   This allows me to control the amount of water for each bed.   The other end of this branch has an "L" joint connecting it to a length of 13 mm black poly, terminating in a stopper.   This section of tubing lays on the ground and I have sprinkler heads inserted into it that direct the water into the ground for watering.   This helps reduce loss due to evaporation, and it avoids watering the  foliage, thereby helping to reduce the risk of disease.

To finish off the system, I attached a timer valve at the manifold.  This allows me to turn the sprinkler on for a set amount of time in the morning, before leaving for work, and then forget about it.