The shower heat reclamation plumbing from GFX was installed a long time ago, and now that we've moved in we have the opportunity to collect some data to see how much hot water this device can save. This is a small shower with a stone floor and a vertical heat reclamation unit in the shower drain plumbing.

Future analysis will look at data collected over many months, using sensors embedded in the plumbing. I was impatient to get some data, so this sample is derived using a Hobo temperature logger and some assumptions about shower head temperature. 

Here's what is supposed to happen:

1. Cold 60F water enters the house and passes through a heat exchanger, absorbing heat from water going down the shower drain.  Our initial guestimate was that it would heat up to around 75F, not cold but not really warm either.  This waetr would start at cold (60F) when initially turning on the shower, but how long would it take before the heat echanger kicked in?
2. This no-longer-cold water enters the shower mixing valve. The valve mixes it with domestic hot water (DHW) and delivers 100F shower water.
3. The thermostatic shower mixing valve adjusts to keep the shower temperature constant even though the temperature of the water going into it on the 'cold' side rises from 60F to something higher. 
4. As the cold side of the shower valve gets hotter, less and less DHW is needed, reducing the draw on the hot water system. 

The good news is that the above data gives an estimate of just over 20% less hot water from the hot water tank. That's a pretty good saving! We'll see what longer term data analysis shows. It seems to take about a minute for the heat exchanger to warm up, so savings would not be realized on really short showers.

In this shower it is possible to touch the pipes going into the shower mixing valve. It is very noticeable to the touch as the 'cold' pipe warms up, but the shower stays at the desired temperature. Its great - it's saving energy and I don't have to think about it.

The above data deliberately shows some of the challenges of creating a good experiment and also of designing and building things differently and the debugging challenges that ensue. First, I wanted to make sure the heat exchanger started off cold, so I started the shower cold to draw cold water into the heat exchanger - this also puts the 60F cold water temperature on the graph (see bottom dip). Next, it turned out that the custom designed hot water system was not correctly adjusted, so the hot water temperature was around 80F (obviously not good enough). I made the adjustment and the rise of DHW from 80F to 120F can be seen.  Once that was over, the experiment truly began.

The savings in hot water draw can be calculated as follows:

% savings = 100 * (1 - new_hot_water_draw / baseline_hot_water_draw)

where

baseline_hot_water_draw = (100F-60F)/(120F-60F) = 0.66

and 

new_hot_water_draw = (100F-cold)/(120F-cold)