DIY ventilation heat exchanger

[pladijs]

Hi again,

It has been a long while since i last posted to this thread, but i have often returned to read your contributions. This is really insipring stuff!

In the mean time, I have been installing my own HRV, although (as I explained in a post here a long time ago), my main goal is actually de-humidifying in winter, as I have (or had) problems with condensation and I could smell mold developing.

I first started building a HX countercurrent core using a combination of plastic frames and aluminum foil (not plates). This turned into a horrible mess. I don't think I will finish that project as I won't get it sufficiently airtight and the airflows would mix. If i ever redo this, i'll take the double cross-flow core approach and coroplast, I think.

What i did do, however, was to drill holes in my walls and simply install a tube-in-a-tube duct system, to get some ventilation (and de-humidification) with minimal heat recovery. I reconned that if I ever feel the urge to make a second attempt at building a proper HX-core, I would need the tubing anyway, and i hoped the tubes might work as a minimalistic heat exchanger while solving my humidity problem.

The picture below shows my situation at home (first time i sketch something in google sketchup, quick and dirty job).

The red bar is where i installed a 160mm diameter PVC tube, slightly tilted to allow condenstion to drain, containing a 110mm flexible aluminum duct. The space between both tubes contains the stale air. The entry point for stale air is indicated on the right end of the red tube. The interior tube with clean air goes through an extra interior wall, into the kitchen. Those two rooms are separated by an old three-piece leaky wooden door (from the 1940's), the upper part (say 80cm) is permanently closed, providing a kind of trap for hot humid air coming from the shower which is integrated with the bedroom. The total heat exchange surface is really small, 0.11m*pi*5m ~= 1.7m2. But on the other hand, this aluminum duct surface is very irregular; and thin.

This is all quite minimalistic and low-cost, but it performs rather ok, i find; at least for my specific situation. But let me make some points:

- I have only a 60m2 appartment, its just me and my wife living there, and we both work during the day.

- My appartment is not the most airtight, I have old wooden floors, not the newest windows, there is bound to be quite a bit of leakage.

So i am quite content to ventilate moderatly. If i ventilate on a low volume all day long, even when absent, i get reasonable heat recovery on those slow airflows, in spite of the small surface of the tube.

If my wife and I are around, cooking, showering, breathing,... higher volumes are required and i have to live with some cold air flowing in (although heat from condensation seems to make up for a lot). Not that we ever feel a cold draft or so. To me, that is just the price to pay to get rid of that humidity problem, its not that i was planning on saving on heating costs or anything, quite on the contrary.

I wired everything up with four sensors (cheap DHT22), hooked them up to an arduino with a wifly shield, and programmed it to send all data to my router running dd-wrt. There, the data is stored, and i mess around in php, determining the return value to give back to the aruino to tell it how fast it has to spin the fans for some given sensor values. Having a router making the decisions on how fast the fans should spin has the advantage that I can tinker with the system in a friendlier langage of choice, no memory or time-keeping issues, datalogging. I do not have to reconnect the arduino every other day if i want to change something. I can do that from Spain (where I work), while this machine is running in Brussels (where i go every few weeks).
The fans are standard 120mm pc fans that push about 100m3/h without any pressure (but then these tubes are very large, there is hardly any pressure to overcome given this type of exchanger). The fans are 4 wire which means they can be controlled via a low-current pwm signal from the arduino.

I made a small website showing the aparatus in action. I will change the domain name in a few days to avoid being hacked (its just my badly configured router running that website).

pladijs.no-ip.org:8082/zip.php

What really surprises me is that it keeps the appartment quite dry, even with these small volumes of air. This makes me wonder (in retrospect!) about many of the systems i see here and planned to build myself: wouldn't running those have completely sucked all moist out of my house even at moderate speeds? Without any moisture-recuperation, isn't running small flows the only viable way? And given that, aren't relatively small units quite ok?

Obviously I am also simply quite jealous of many of the nice systems i see here! But from my experience low humidity really would become an issue for such larger machines. Of course one could keep a nice large system and have proper ventilation, using plants or humidifying otherwise... perhaps that comodo dragon could help keep moisture levels up too!

PS there are lots of issues left to do with my system: i have some stale are leaking into my fresh air entry outside, for example (which you can see in the graphs on the website). Also: the inner tube is too large, given the specifics of how i did things (the inner tube has no curves, the stale air has a narrow spot to pass where it enters)... i have to run the inward fan much slower, not to cheat on my efficiency, but even just to get the same change in temperature (excluding condensation), from which I would derive the same volume of air is flowing (comments ?)

An open issue for me is also how to control the unit. Simply by relative humidity? dewpoint? difference between inside/outside dewpoint? absolute humidity? relative humidity seems ok for a first approach, but say temperature drops at night, relative humidity increases, and the unit starts to speed up exactly when ventilation needs are low (and outside temperatures are coldest). Any ideas on this?