Here is the spring on left, on the edge of a creek. This is what I developed in the fall using the same collector design as above.

In-line storage cistern for the gravity-feed spring water system. The manifold looks strange but is the best way I could think of to impliment continuous flow for the supply line without excess water entering and leaving the cistern. The two half-inch pipes are an anti-siphon vent for the overflow and a breather tube for the tank exit. The overflow and drain then share a single poly pipe that empties into a creek.

The tank is located around 40 meters up a steep hill. Limiting errosion was a priority in designing the installation. Nevertheless, terminating the drain assembly with another inline ball valve as a direct cleanout would have been better.

A simple float valve in the cistern regulates fill.

The stream water collector at the same house employed a design that is prone to silting up, which had occured. The assembly was a modified version of what I had encoutered at the house a few years ago: A PVC pipe passed through a simple rock and clay dam then collected water in a cavity behind that dam. In this case, the cavity was provided by a perforated bucket.

I prefer this design. Instead of a dam, a section of poly material is used as a bulwark to hold back incoming water and mount intakes via commodity poly bulkheads. I used the top of a 55 drum for this one.

The holes in what is now the bottom of the assembly are irrelevant. That area and the entire interface with the stream will be sealed with clay.

This is the site for the new collector before development.

Testing the fit after excavation.

The installation sealed with clay and partially filled with gravel.

The completed stream collector. The top tube is for overflow.

The spring box with overflow that doubles as cleanout when rotated.

Part of the tubing route.

There are about 100 meters of poly tubing connecting the springbox to the inline storage cistern, then another 100 connecting the storage cistern to the house. At the house, flow rate is considerable. I measured about 35L/minute.

The tubing where it exits the spring box. This is the only section that is in the creekbed.

Initially, I unspooled a full 100m roll of tubing down the creek, but had to reroute a few times to find a workable path to the cistern site. Even though the cistern is located at an elevation of at least a couple meters lower than the springbox, the terrain between the two installations is irregular.

Ideally the tubing would follow a constant >2% slope over its length. Failing that, it must never rise higher than the spingbox at any point. Finding that meant a very sweaty day of pulling the tubing through blackberries plus hand excavatian of unruly topographical features.

The spring box and spring together. The pipe between the spring and the springbox crosses the creek, which is not ideal. There has been much rock moving both last year and this year to prevent washouts and direct flow.

At the house end of the water system, the poly tubing is unceremoniously terminated to a PVC junction. Throught the system are a multitude of redundant but convenient shutoff valves. Shutoffs are essential for maintaining improvised water systems.

At the house, there are 30m of PVC supply lines feeding the house, 4 hose bib locations, as well as a future bathouse site. The house/shed came with a nicely-vented drain, but no supply plumbing. That meant some drywall work to terminate PVC and CPVC risers. The drain stub isn’t quite under where the sink will be located, but that doesn’t present any practical issues.

One of the many trenches is home to a hose bib supply line, sink greywater, and an upcycled section of huge diameter polypipe that is playing the role of conduit.

The conduit leads to these two 200w solar panels, which charge the battery inverter system.

I threw together the ground mount on the last day there. After researching pole and ground mounting systems, all expensive and dependant on parts not locally available, I put this together using maybe $25 of commodity materials. The frame is all cheap dimensional lumber, with the panels mounted using modified framing straps, screws and bolts.

I haven’t tested it it on a cloudy day yet. With sun, it charges back expected daily battery drain in less than 2 hours.