Rainwater harvesting seems like an obvious thing to do to me - capturing the rain that is inevitably going to fall on top of our buildings seems like a even more of a no-brainer than capturing the sunshine that reaches our roofs. Ironically enough, the logistical difficulties we're having on this project due to road closure are because United Utilities are having to put in an enlarged sewer due to the original Victorian drain being unable to cope with today's increased water usage (as well as its poor state of repair); if every house on the street harvested its rainwater instead of dumping it down the drains, that would yield a reduction in wastewater volume of at least a third.
Rainwater harvesting systems are difficult to retrofit in anything other than a whole-house refurb, because you have to have a large underground tank (rainfall has seasonal variation of course, so you need to store a large volume to benefit from the system) and two sets of pipes running around the house, to keep drinking water and rainwater supplies separate. But if you are starting from scratch with your plumbing, rainwater harvesting seems to be a no-brainer.
There has been some criticism of the significant embodied energy in the rainwater tanks not being fully taken into account in assessing the environmental benefit of rainwater harvesing systems, compared to energy saving from reduced drinking water consumption. On this project, we're very aware that there is also significant embodied energy in the drainage systems that take rainwater away from buildings, if it isn't harvested. And for this particular project, because we've chosen to install a fire sprinkler system, we are going to use one shared tank for both the sprinklers and WC flushing. So it could be argued that, as the tank is required anyway for fire safety reasons, only the additional plastic in the increased sidewall area to accommodate the extra water for WC flushing would be included in a rainwater harvesting embodied energy calculation (by the way, we're not doing any quantitive embodied energy analysis in this project - not because it's not important, but due to lack of capacity).
The difficulty, for a mid-terraced house such as this, is that it's only really practical to capture rainwater from one half of the roof (the rear, presumably); a two-tank system and running large diameter pipes through the house were both quickly ruled out as not viable. For this reason, we're only using the rainwater for WC flushes - the calculations we've done are that it won't necessarily provide 100% of WC flush water, so it wuold be pointless to start plumbing in other things eg washing machine.
The chosen system is available "off the shelf" from rainwaterharvesting.co.uk, who supply the whole system in kit form. The things I like about their offering, compared to others I've seen are:
1. There is a header tank in the attic, rather than supplying appliances directly from the underground tank. This means that the supply is more reliable, and the electricity usage from the pump (already very low, we're promised) is significantly reduced, because pumps work more efficiently if run continously rather than stop-start.
2. The Rain Director control unit deals with switching over to mains water in the event that there's no rainwater left, and can cycle the water to stop it going stale if, for example, people are away on holiday.
3. The tank is shallow-depth, only requires a hole 1m deep.
I spoke to the supplier at Ecobuild, and again more recently on the phone. They are going to supply an adaptor to allow us to fit an upstsand pipe to the inlet of the submerged pump, so that the inlet will be physically part-way up the tank, thus guaranteeing that the WC flushes can't "steal" the water that is reserved for the fire sprinkler system (the sprinkler system will use a separate suction lift pump).