wrote in message
...
A 200' version makes 82%...

But that's pretty big. How about this?

We collect all the shower water in a tank, with an infinite cold water
tank
next to it, then circulate the cold water through a coil in the shower
tank
until it all cools to the cold water temp... Then again, infinite tanks
are
hard to come by.


No, I don't think this is the way to go. When using warm water to heat cool
water, the smaller the temperature difference along each section, the better
as far as entropy goes.

That is one of the reasons why counter-flow hx are so good. The warm water
is never more than a few degrees above the 'cool' water. Where the 'warm'
water is the hottest, the 'cool' water has already been heated up to almost
the same temperature. And where the 'cool' water is the coldest, the 'warm'
water has already been cooled to almost the same temperature.

Compared to a parallel flow, where the 'warm' and 'cool' water start out
with vastly different temperatures and aproach each other over the length of
piping.

With your 'infinite' tank idea, you end up with very cold greywater, but the
freshwater is hardly warmed at all. Making the two tanks nearly equal, you
approach the limits for a parallel flow system (e.g. grey and fresh leave at
the same temperature, about (100+55)/2 = 77.5F.

I take it you've considered what the pressure drop would be with the flow
rates? Check the greywater side pressure drop carefully. Although you have
larger x-section, the *available* driving pressure for a shower drain is
just a few feet of water.

daestrom