DIY ventilation heat exchanger

[kostas]

Originally Posted by pladijs

But one of the reasons I am a bit afraid of going for a counter flow version is that it would be harder to clean....

Sorry to answer you now, have been quite busy at work lately!
About cleaning, I think the best thing to do is installing at least a basic polyester filter, a specific HRV/ERV or furnace filter would be better. If you frame the plates like I did (forget the plywood caps, I'll change that with two metal sheets per core) you can wash the whole HX by immersing it to a tank of warm water and some mild soap. For a coplete sanitation add some specific detergent and rinse.

Originally Posted by pladijs

Kostas, hope you are reading this, is it me, or is there a difference in the height of your cores in the inital pictures you sent and the latest ones? seems like they shrunk? is there a reason for that? Could just be my eyes of course.

Yes, I shrunk them by accident using my atom shrinker!
Just kidding, the small cores you see are the first ones I built to do some testings to see the aforementioned differences between the various materials and configurations. Once rejected the PP distanced version and confirmed the equality between PP and aluminum sheets I decided to use all the material I had and made up the two big cores I'm currently using in the HRV.

[ham789]

Thanks for the info.
The whole metrology problem is why I don't chase 100% efficiency.
The goal is to subtract two equal numbers. But, as you approach
that equality, the measurement error in the numbers gets dramatically
worse. And air flow measurement is much worse. 20% error in the
flows is a computational disaster.

My toy HRV has no ducts, so it's nearly impossible for me to get
accurate measurements under true operating conditions.
Current inside/outside differential is only 14F. Makes no sense to
attempt any measurements.

If I understand correctly, you have one symmetrical core of aluminum
and another that alternates coroplastic channels and wide-open
areas. How much difference that makes to the relative air flows
is back pressure and fan dependent. I tweaked mine for zero
inside/outside pressure differential. The fan speeds are significantly
different.

I like the direct measurement approach. Measure what's important
to you. For me, the bottom line is operating cost.
If I had a ducted system like yours, here's what I'd do.

I'd stick a heating element in the plenum inside the box at port4.
I'd poke a 3/4" sense port in ducts 3 and 4 far enough away that
turbulence has stabilized the air temperature.
If you stick a velocity meter in the hole, you'll get a wide range
of velocities depending on how far you stick it in. But if you
go for the peak somewhere in the middle, you should be able
to adjust the fans so that the air speed is equal in ports 3 and 4.

Now, heat up the heating element until the temperatures are
equal at the sense ports in 3 and 4.
The amount of heat necessary to do that is a direct measurement
of what it costs to operate.

To be clear, I don't propose you heat the air. It's just a
temporary measurement technique.

Humidity measurement is another problem area. I have a decent
humidity probe, but the settling time is about an hour. I have to keep
the dew point above 43F or so or I get sinus problems. So my HRV
dumps a lot of water.

[Fornax]

Some pictures from my build: Here

Some even come with comments. The whole album is in reverse since the forum thinks it is a splendid idea to put the latest picture first, so do start with the last one and then click on previous.

There are more pictures on my camera, those will follow soonish.

[Fornax]

Originally Posted by kostas

T3 (hair dryer): 49°
T4 (HRV exit air): 45,7°
T2 (hair dryer outlet): 38,5
T1: (ambient air inlet): 21°

PP solid core efficiency: 88,2%

That is the efficiency for 1 of the 2 channels, the formula we are using to calculate that (perceived efficiency) is:
100% * (T4-T1) / (T3-T1) == 88,2%

For the other channel we would get:
100% * (T4-T2) / (T3-T2) == 68,6% <-- wrong formula
100% * (T3-T2) / (T3-T1) == 37,5% <-- correct

It appears something was not symmetrical during your testing and it could be the mass of air through both channels.
Warmer air is lighter, especially when one stream is 294 Kelvin and the other is 322 Kelvin (9,5% more density for the colder air.) but that alone doesn't account for the 28% difference. Were all the ductwork and the fans identical?

With this unballance you do indeed get nice and warm air inside but you are exhausting more air out than re-enters, and the exhausted air is warmer than it should be. On top of that the surplus of exhausted air will be made up by air entering your house via other ways, unheated.

-----

In your test the distanced PP core performed poorly. Due to it's construction the two channels in that one will definitely have a different resistance. If your fans cannot handle this different airpressure then you will get weird and unbalanced results, like 45%-ish efficiency in 1 channel.

Kostas, could it be you overlooked something?
I was as surprised as Ham789 was when you reported that the distanced PP performed so poorly when compared to the stacked PP. We feel it should be the other way around.

If you still have that core and the other testing-equipment could you in some way repeat the test and report T1 - T4 (including T2) and then rotate the core 90 degrees and repeat the test?

[kostas]

I think the problem with those measurements were the fans I used. The models were similar (looked identical but had no tecnical label on them) and over 10 years old, so I guess the air flows were unbalanced.
There were no ducts except for a piece of pipe (120mm diameter, 1mt long) that i had to use for the hair dryer to prevent the polystyrene casing from melting down.
As for the distanced PP core's performance I think the issue is the lack of small channels in the distanced flow which allows more air to pass through exchanging smaller amounts of temperature.
Sorry but I no longer have this HX as I used the material to built up the bigger ones I currently use.

@ ham789: thanks for your suggestions. I'm still waiting for that anemometer to arrive from the bay... I'll try to measure air flows as soon as possible.

[theworldtrekker]

I'm too am surprised at the low results from what is getting called the distanced PP core. One question that pops to mind is has anybody tried adding some air turbulators (on an aircraft's wing they are vortex generators) to the open channels? I know in some solar hot air collectors gains are observed on extracting the heat from the absorber sheet if a few (~10-20) turbulators are added to the back of a 4'x8' collector. My understanding is that in reducing laminar airflow the temperature gradient of the air at the heat exchanger interface skin is reduced. For these scale heat exchangers running course grit sand paper over the PP might suffice. Although that might increase dust buildup across the entire surface negating any positive change. These should also help to balance the air flow/pressure restriction between the two heat exchanger channels as well.

[ham789]

My Toy HRV runs the air straight thru the coroplast
channels, but the open layers have side-exit at each end.
That should give some turbulence to the flow.

A relatively easy experiment is to cut some coroplast strips just
wide enough to be mechanically stable and insert them
off-angle into the open channels. That oughta add significant
turbulence without reducing the surface area much.

Poke a pair of thermocouples down into the one channel
and see what difference it makes.

A/C has some machine-cut black coroplast squares.
Maybe we can convince him to do the experiment.

FWIW, some recent construction plans for HRV have a block
down the center of the open channels and a reversal space
at the end. They run the air down one half and back up the
other. That gives some temperature gradient that has some
of the benefits of a full counter-flow exchanger. And it's
easier to route the ports.

I have a very crude thermal imager that was made for firemen
to find cold bodies in a flaming building. It won't focus up
close, so may not yield useful results, but next cold spell,
I'm gonna break it out and see if I can learn anything
about the thermal profile from the outside of the exchanger.

[rellaw]

Hi,
I see lots of good info on here to help me save the heat from the bathroom go out the window.
I need to vent the steam out now three girls now take for ever in there.
Looking at all the options open to me.
I have a flue from a gas warm air/water heater that I would like to incorporate into the system to boost the heat entering the rooms. (just thinking about it).
I have a 2.3Kw solar system on my south west roof to offset the power of the motors.

Thinking of using the excess solar power to preheat my water to my immersion tank - but that's another project.

[kostas]

I was working on the operational part of my DIYHRV.
In this hypothesis I used Fornax's Comair motors and tried to figure how the various apparatus I'd like to use generally fit and combine.
Here is a first draft:


You can also see a small design concept video I uploaded on the Tube:

http://www.youtube.com/watch?v=LrzLT...ature=youtu.be

Here is the parts description

1.Fresh air entrance (from roof)
2.Filters chamber (insect metal screen and fine dust or furnace filter). I was also planning to install a small UVC insecticidal lamp here, but it's still a maybe.
3. Pre heating exchanger (low power, hooked on my propane home heater by a separate circuit)
4. Dual coe HX
5. Condense drain metal sink
6. Post heating exchanger (as n.3 but with more power, like 2Kw or so)
7. Humidifier system (don't know how yet, maybe a free dropping water system)
8. Inlet motor
9. In house plenum
10. To the inlet ducts (living room and bedrooms)
11. Stale air plenum and filters (from bathrooms, kitchen and corridor)
12. Exhaust motor
13. Air exit (to the roof).

Ducts n.1 and 13 will be obviously much more distant from each other (3mts)
Another option would be to intersect them (1 to 13) by a motorized valve for the recycling mode.

In winter I'm planning to pre and post heat the incoming air using my heater as long as it takes to change all the air in the house. Pre heating is mandatory to prevent frosting and the post heating fills the temperature gap between the HX outlet and the house ambient.
After that the system would switch to the recycling mode for an equal distribution of the internal heat (no pre and post heating).
Same work in the summer, when it's hot and humid outside and with the A/C on in the house. (obviously heating is off).
If i'ts cooler and less humid outside then motor n.12 goes off and lets only fresh cool air enter the house.

What do you guys think?

[ham789]

Very nice work. What tools did you use to make the drawings?

I think you have a fundamental air flow problem.
You want the ports exposed to the outside to be
on the same end of the exchanger. If the air doesn't
flow in opposite directions, you're giving up the efficiency
improvement due to dual cores. You want the greatest
possible temperature difference everywhere in the system.
Then drive that temperature difference toward zero across
the board.

I'll say the same thing differently.
If you have a counter flow heat exchanger, you do just that.
With a long path, you can approach 100%.
If you reverse one air flow so they are in the same direction,
your output temperatures will be half way between the two
temps at the other end. You can't do any better than 50%.

On the outside end, you want the outgoing air to flow upward
through the wide-open channels so the water can condense
and drain back into the warm area instead of toward the
freezing end. This won't work so well if you put the air thru
smaller coroplast channels because of surface tension and you
don't have enough air pressure to blow the water uphill.
If all of your channels are small, you'll want the air to flow
downward to blow out the water...but I think that will have
more freezing issues. Would be interesting to experiment
with the plate spacing and surface tension to see what the
minimum effective spacing might be.

A counter-flow exchanger is intuitive because you can predict
the temperature at various places in the core.
For cross-flow, it's not so simple.
In the output end of the system, the coldest part of the core
is at the corner where the coldest incoming air is. Since there
is a differential in the average temperatures at the ports,
we know that the other corner has to be warmer.

The objective is to not freeze up. The warmer air exiting
is waste. And it's higher than the colder corner of the exchanger.
We need some way to get that heat transferred back into
the cold corner. That might be done with a temperature
controlled path from the hot part back into the cold part.
If you allow ALL the air to flow back, the system temperature
will rise to the interior air temperature. So, there's some amount of flow
that keeps the coldest corner of the core above freezing.
This also affects where you put the fans so the pressures
are right to make the air go where you want it.

If you live in an area with moderate cold, that's the simplest
and most efficient thing I can think of. You trade fresh air
for defrosting. Do the math to see if it works in your climate.

If you have to add heat, that's also the place to add it.
Any more heat than the amount to keep the output port
minimum temperature above freezing is wasted.
But it's worth a lot of careful thought to verify that.

I'd be interested in opposing views on optimum defrosting.
I really have little more than intuition to support my claims.