I have come accross another company http://www.oasis-rainharvesting.co.uk They have some very interesting ideas about what makes for a good rainwater harvesting system.

For a collection filter I'm thinking of using a sand filter (there's probably a proper name for it). Anyway the water has to run through a tub with sand and gravel in it.

Then after the pump, a cheap cotton element house water filter.

Ah, sorry, I thought you were talking about rainwater.

A grey water system is used in the earthship. Its filter is a leg of a pair of tights tied on to the end of a 4" drain pipe. The water then goes into a tanked bed which contains various plants which thrive on grey water - the water that comes out of the end is clean and fresh enough to use to flush the toilet or water further plants.

It's the part of the design I was most skeptical of, but it turned out to be the only part which really worked properly.

Glad you like it! Just to clarify, in the setup I've seen, the tights are actually "loose", lying horizontal in a plastic channel or tub of some sort, which has quite a large diameter - so it can't expand to get stuck and you can access it by taking the top off that section. It didn't need doing very often though.

I should point out that the earthship is more of an office and has no bath or shower in it, which may make a difference.
If you need to scale the system up, you could use those "filter sacks" that are actually designed for wastewater treatment. I used those for waste veg oil filtering - one fits nicely into a 25l plastic can with the top cut off, then you just pour the stuff in the top and cleaned stuff runs out the bottom.

Does anyone know anything about ceramic water filters, really after something that could handle up to 300 litres a day.

We used to install 'sand filters' and 'DE' filters for spa baths. I never got too involved with them, but I do know that they needed 'back washing' to get the dirt out. Not sure where you could back wash to, it can smell badly (will never forget the interesting odour from the Sheffield YMCA's filter).

We also used to manufacture a UV filter, was a simple device but not sure how to calculate flow and mass rates, I am sure others can find that out. Has to be remembered that we had constant flow over them, so a bit different from a storage system.

BS8515 is not binding unless the building regs say you have to abide by it, is it? Part G says that guidance can be found in BS8515. Clearly it is relevant, and even recommended, but that's not the same as being a requirement.

I don't really buy this 'you can't store much water it'll go all manky' theory. This thread and the very similar http://www.greenbuildingforum.co.uk/newforum/comments.php?DiscussionID=1545 both have lots of examples of people storing quite large amounts of water, and some people drinking it, some people heating it and using it for bathing, and some of it not even in cool underground tanks, so clearly large tanks can work just fine. The 5% limit seems rather small to me.

We store 1000l in 5 waterbutts, open to the atmosphere with no filtering and over many years only one that had a dead mouse in it once went horrid. That water is used for watering plants. Surely if 3000l of underground storage is a lengionnaires risk then warm above ground waterbutts full of mank used for watering would be considered highly dangerous? Given the temps of underground tanks it seems to me that L is simply not an issue - it is a problem of warm water systems. This seems like the similar paranoia over solar hot water cyclinders (no real risk at all, just paranoia getting in the way of genuine efficiency).

A well-filtered, rodent/insect proofed, cool, dark tank should be able to store water in reasonable condition for a very long time (all year?).

Now there is a much more pertinent question to ask about the trade off between cost, tank size, system design and energy/water savings. Clearly there is such a thing as 'too big' at some point, because tanks and installation cost money, but I really don't believe the optimum is at the 1600l tank BS 8515 says is the largest I should fit (60m2 roof, 2 people). We ran out of water this summer just using it for watering, not any internal uses, so clearly >1000l is needed there. Seems to me that 2000l is a minimum, and about 3000l might be a sensible size. 41,000l/yr actually hits the main roof, suggesting 33,000l/yr might possibly be collectable. At a nominal 50l/per person per day and 2 people we have 36500l/yr usage, so you start to see how a bigger tank may not help - it's all about rainfall and usage frequency. But our average usage is about 7l/day, which is 5000l/yr so I'm pretty sure we would need no mains water at all except drinking if we could collect most of what fell.

Fortunately thanks to the library I don't have to spend 158 quid for the priviledge of reading the standard(!). Having read that it makes clear that compliance is not a requirement "As a code of practice, this British Standard takes the form of guidance
and recommendations. It should not be quoted as if it were a specification and particular care should be taken to ensure that claims of compliance are not misleading."

So we are all still free to fit rainwater systems that we think are apropriate - they just won't be to BS8515. Wastetech is being quite misleading in his claims that you _have_ to do this. Also almost all his posts include a link to oasis water treatment, without saying that he is associated with them. Are you Wastetech? (it's a useful website, but it's important to state affiliations if you have them - the forum rules require it, apart from anything else IIRC).

It also describes sizing a system to cater for stormwater control. In that case it's around 4500l for our roof. So in fact you can just size for storm runoff if you want a bigger tank _and_ to comply with BS8515