I agree with Tim Acland. It ...

I agree with Tim Acland. It must be started with local specific conditions and available materials, if there is no infrastructure. Then start to build on site like disposable tranches from household toward the main channel (a bigger tranche) and a reservoir (a big hole in the ground). Have this first instance flow control, then it might be imaged next step as the use of the gray water, considering the biochemical charge and the costs (maybe for some irrigation is less costed than for human use, even if not potable). Of course the tranches and holes could be reinforced, if they are in good position and suitable for primary use/installation. Afterwards, all the technical solutions mentioned here might be considered at least like design ideas for local hand manufactured pieces of installation or, if enough money available, engineering infrastructure works and standard installations procured and mounted in.