Seems like my stale-air exit sensor is dying, returning NAN quite regularly. its been hanging there for months in a condensing environment. I just hope its my bad soldering, or a drop of water causing a short, and not the sensor itself.

I also see random jumps in temperature and humidity at the fresh-air exit sensor. I thought they were due to mixing-in of stale outgoing air at the entry of the fresh-air tube, outside. Although some mixing is likely going on there, I compared my graphs to the observations on wind speed and direction from a nearby official weather station. Weer: Weerrapporten van Uccle WMO (6447)

It is quite obvious to me know that even with some moderate wind out of a south-west direction (which is exactly opposite of the fresh-air entry and flow direction), air is being sucked out of that fresh air entry tube in the wrong direction, the flow reverses, and the sensor (which is say at 20cm from the end of the fresh air tube inside my house) picks up stale hot air from inside. My fans just cannot take much counter-pressure, but obviously especially when run slowly, as I do 24/7. Does this mean i have to adjust my fan-speeds according to observed windspeeds and directions? :D The funny thing is, with this arduino-dd wrt combo that would actually be rather easy to do. But perhaps it would be more efficient to make my front-door more airtight, to avoid wind getting in and building up pressure.

All in all, I must say that re-visiting this site has made me feel like picking up my project again, attachging a larger DIY HRV to this minimalistic HRV-tube system of mine, with some more powerfull fans, and decent efficiency so I do not have to run them to slowly, which would also solve this problem with reverse flows. But I have some questions I would like to hear your opinions about:

- you think it is necessary for the core to be cleaned, ever? Part of what made my alu-foil counterflow HX-core build flaky, was that I want to keep the different frames separable, not glueing them together. I see a lot of glueing here, but then even a glued coroplast core would be washable (not so with my breakable alu frames).

If i would pick up glue, and fix everything permanently, I might be able to use the frames I already made (half of the numbers i needed, my move to spain was sudden), but boy that was a lot of work to make those. I took the approach of Fornax, but with foil, not plates, also used very thin frames (2.5mm? IIRC), and no "baffles". Although I seriously considered that, I was afraid they would tear the foil. Also, the general dimensions of my box are similar to the one by Fornax. But (now it becomes difficult to describe...) I thought that this design would be too simple and decided to to split my core and have a connection at one end of the box steering the flows from the bottom into the upper half of the frames. Im not sure why i did that (quite possibly I got inspired by something I read here, but then I forgot). But a configuration with all entry and exit on one side of the box better suits my particular situation, and I thought for some reason that it would be better to have the air moving a longer distance at a higher speed, rather than having slow speeds in a "single core". I you see what I mean. Does that make any sense at all? I can still drop the idea of splitting the core. So far I have only made the box and half of the frames, not the connections.

- Related to cleaning the core: what kind of filters do you use, if any at all? do you think G3 should be enough (or even too much?). Would a F7 filter bring any real benefit in your opinion? I live in Brussels, a densely populated area, lots of traffic)... but I have the feeling a filter wont help much. If i install one, I was thinking of ordering a "pocket filter"; total overkill for such a small system, but then I hope I would earn that back easily by having less presure to overcome.

The ice problems I will get with this unit: I will just stop the incoming air flow every X minutes when the incoming temperature drops below a certain treshold.

Apologies for my long posts!

This also painfully reveals how little heat recovery is in what i should perhaps better refer to as my de-humidifyer/ventilator. I think i would get around 40% at the moderate speeds which would just suffice to keep my house dry; but such flows are actually insufficient when we are with two in the house, and active. No way to tell though, without actually measuring the airflows. Does anyone here has any experience at attempting to measure that? I was thinking that in a non-condensing environment, and equal DeltaT for incoming and outgoing would imply equal airflows... does that make sense?

As far as my short experience allows I can advise you to glue less possible. Separate parts means easier cleaning and replacement in case something breaks.
As for the HXs, direct flow is the best, simple and efficient with low pressure drop. I think the one you're talking about which is similar to one seen in an old article of Popular Mechanics is more complex and gives less performance. Do as the real HRV on the market, they are quite simple.
About the filters, it depends on what kind of pressure your fan can give. More dense and more pressure you need to pass. I'd stick to a normal furnace filter, like this MERV 10 AC and Furnace Filters (Better) - 4 in. thick (2 Pack)

Kostas,

I do like the simplicity of the double cross-flow core approach. I regret the ton of work that i put into that counter-flow alu-foil experiment of mine; good thing it didn't cost too much.

I would be very (very!!) ;) interested in learning more on how you actually did build your cores.

What material did you use? People here talk about coroplast, is that compareable to "polycarbonate panels"? Did you leave spaces between the "coroplast" layers? (kind of like here http://ecorenovator.org/forum/26367-post226.html, to me that sounds like a very good way to do it). Are you still convinced that the coroplast core performs rather compareable to the alu one? Do you have an idea of your efficiency at different airflows?

If you do not glue, how do you prevent leakage? just by pressing the frames together tightly? Would you do anything differently in your setup? Do you think it is possible to do this with alu foil, rather than "solid" plates? I do not have such plates lying around, they are damn expensive to buy as far as I can tell.

Many many questions!

PS something unrelated i have wanted to ask for a while, regarding http://ecorenovator.org/forum/26465-post245.html , wouldn't it be better to heat the air after the HR? if you have a really efficient HR, say 95%, with pre-heating, you will never get to 100%, with post heating you could. And isn't the heat transer most efficient when the temperature difference is the largest?

Yes, actually it does. Per square cm/inch the amount of heat to be moved is low.

For thin materials the most limiting part is transfering the heat from the air to the material and on the other side again from the material to the air again. Heattransfer through the thin layer of either aluminium or plastic is only a minor factor here, even with aluminium conducting heat a few hundred times better than coroplast.
If your material is thin enough like 0,5mm (1/50") you can leave it out of calculations altogether for simplicity.


You either have to use flexible materials, like the weatherstripping I used, so you get away with mildly pressing things together, which works well for me.
Or you need very precise fitting parts and then still press them together more firmly.
I guess stripes of coroplast glued on 1 side could fit the 1st category if they appear to be fairly straight. Most types of wood will probably be too sturdy and need more pressure (also these wouldn't handle moist too well in the long term.).


No.
Unless your 'plates' are separated by at least a cm or 1/2" the foil would deform too easily due to pressure differences.
Aluminium is too soft to use very thin sheets/foils, I even had some problems in my build with 0,8mm thick plates (1/32"). Actualy steel or plastic would be a better material for compact HX-cores.

Fornax, thanks a lot for your reply. your points seems quite valid! If only I would have asked for some feedback before starting on my aluminum foil project.

On a different topic, I have read some of the previous posts here on fans, and often was surprised about the sheer power and size of them. Perhaps it is because of the small size of my appartment, or because I feel happy with somewhat lower replacement rates in my not-so-airtight place, but I was thinking of buying something much smaller, such as

SANYO DENKI Web CATALOGE

Or perhaps Pabst has something similar. I really want one of those 4 wire fans, though, to combine with the arduino. I am actually quite illiterate on how AC/DC, ECM and the like compare with fancs, so please tell me if this is a bad thing to do... but looking at how much air it pushes, even against some pressure, at such low wattage... seems hard to beat. Am I missing something here?

Mind you: this is a special low-power consumption series. But they seem to have a rather high life expectancy as well. This low power series does not seem to come in a 120mm variant, but perhaps that is just telling such large fans are not economical from a power consumption point of view? no idea really.

When I try to estimate the pressure drop that I expect, I get rather low numbers. Confession: I actually bought one of those pocket filters already, if you look at the documentation of a typical one (not that I bought this type), you get perhaps 10-20Pa or so (see the graph on the last page). http://www.volzfilters.com/en/pdfneu...filtration.pdf

Any idea what pressure drop to expect (at say 100m3/h or about 50CF/M), from the heat exchanger itself, say a coroplast double cross-flow type? And are there some quick rules of thumb on the pressure drop to expect per meter of duct? Luckily those ducts are completely straight, they will go directly into the HRV.

Im writing long posts again. In essense I am just looking for some feedback on my guestimate of about 20-30 Pa resistance in my small system, and on this choise of fan. Please stop me if this choice is absurd; it specifics are quite out of line with what I have seen suggested here before, so I ask rather than being sorry.

Also, I will have to install this outside my house, ... no space for it inside, nor the will to open up my lowered ceiling. Also the noise would bother me I think, as it would have to end up in our bedroom. So its going to end up in a well isolated box outside, directly attached to my outer wall. Crazy?

Your thoughts are greatly appreciated!

bunchostuff
 

I'll try to express some of my experience relevant to recent posts.
Be aware that I'm a firm believer in the 80-20 rule
and the concept of diminishing returns. It's fun to discuss
100% efficiency, but you can get a LOT of return for relatively
little investment. I try to apply as much cleverness as I can
muster to a minimalistic approach.
My requirements are lower than many people would accept.

Filters:
I use the fiberglass from the cheapest A/C filters on mine.
Just enough to keep out the (larger) bugs.
I did build an electrostatic filter for the inside. I recently
acquired an air particle counter. Will be interesting to see
the numbers when hay-fever season arrives.
Filtration is very expensive in terms of cost of moving air.
Use as little as possible.

Post heating the air:
Air coming in is colder than ambient. But it gets reheated
by whatever you use for a primary heat source. If you
can add heat more efficiently than that, you should work
on your primary heat source. I can imagine a situation
where it makes one room too cold and might require
remediation.

Cleaning:
I think it's critical to be able to clean your HRV. The thing
is wet and dirty. Perfect environment for stuff to grow.
I decommission mine in summer and soak it in bleach.
Since it's completely self-contained, I have no duct work
to worry about.

Bent metal:
I experimented using aluminum flashing. I bent a thin
lip on each side. I intended to stack them up and seal
the stack with tape along the lips. Looked good on paper,
but I discovered that the slightest imperfection in the bend
warped the whole thing way beyond usable. I'd need a much
better bending brake.
Then I decided I'd use flat flashing and aluminum bars to separate them.
But the cost of the bar stock and my lack of confidence that
I could drill hundreds of holes that all lined up killed that idea.

Then, there's the jelly roll approach.
Imagine two long strips of flashing. Glue a rubber hose
down each edge between the strips and another on the top
side. Roll the thing up using the rubber hose as spacer and seal.
If you cut the hose short on the inside and the other side
on the outside, you can make the flows exit opposite sides
of the roll. Or just make end caps with spiral grooves
to hold it in place. Stick a fan on each end cap.
Since you have only one channel, the spacing has to be
large, so the thing gets big really fast. There'll be lots
of resistance due to the changing direction, but that may
improve the boundary layer problem???
I never got around to trying it.

If coroplast is just as good, seems silly not to use it.

Fans:
I'm not convinced that super-duper fans are much benefit.
For a given air flow, pressure, duct size, etc.
there's an optimum blade/fan design. Just do that and be done
with it. Just turn the thing off when not needed and run
at optimum when it is.
Sophisticated fan controls help when you have a wide
range of operating conditions...don't do that...80-20.

Fan Balance:
I gave up on trying to balance the fans on my small coroplast
unit.
It's poked thru the wall and both fans are on the inside.
I have a decent hot-wire anemometer, but cleaning up the
air flow to get a repeatable reading has been problematic.
I have a differential pressure gauge with 2Pa resolution.
I just tweak the fans for zero inside/outside pressure differential. The slightest breeze causes more differential
than that, so I just pick what looks like average. Might not
be optimal efficiency, but I have only one variable to
adjust.

I tried poking a hole and using the anemometer to set the
air flow thru the hole to zero. Much more sensitive than the
pressure measurement, but when you're already swamped
by noise, better resolution isn't much help.

Another experiment I did was to modify a cheapo HVAC
pressure sensor. It's a UEI EM150 I picked up at the junk
dealer. Has 25Pa resolution. I cranked the gain up by a
factor of 10. Was 10x more sensitive, but the op-amps
were too crummy to keep it centered in the dynamic range.
If I didn't alredy have a better unit, I'd build a better amplifier for it.

You'd think that equal flows would be maximum efficiency.
In a symmetrical unit, maybe. Not so sure about the stuff
we throw together. Mine has narrow coroplast channels
on the output side and wide-open-spaces on the input side.
The idea of equalizing temperature
differences at the inside/outside ports sounds rational,
but there are many variables in the real world. It's hard
to measure precisely.
Just turning on the range vent hood can produce a 5Pa
pressure differential that completely swamps my HRV.

Leakage:
I just used a soldering iron to stitch-weld my coroplast
plates together. I made no attempt to seal the entire joint.
Would be a problem with air pressure, but my pressure is so
low that I don't worry about it. I'm pushing air out
and sucking it in, so heat leakage comes right back in.
80-20 in action ;-)


My problem statement was: House stuffy in winter.
$20 worth of fans and coroplast cured the problem.
And I get back about 75% of the energy lost.
I like discussing the subject, but I'm not strongly motivated
to improve things.

While I'm rambling...
I did an experiment using heat-seal tubing and popsicle sticks. Stack 'em up and the air pressure inflates them.
Totally unmanageable in practice. Might work better with
full plastic frames that could be bolted together.
I attached a picture of the gasbag HRV for fun.

Hi all,

here are some construction details of what I did since last july or so..:)

For first thing I made up the two HXs, one with coroplast sheets (which is a 3,3 mm thick Polypropylene extruded profile) and one with 2,7mm aluminum sheets cut out of used printing plates (a printer gave them up for free).
I didn't have any clue about which was the best, so I decided to try both of them.

Here is the process:

For the coroplast core I had two options: stack them altogether alternating direction by 90° OR spacing them by using coroplast sticks. I also decided to try both so I cut some 300 sticks using a cutting mask I made for the purpose:

http://ecorenovator.org/forum/attach...8&d=1359285863

http://ecorenovator.org/forum/attach...9&d=1359285930

http://ecorenovator.org/forum/attach...0&d=1359285930



http://ecorenovator.org/forum/attach...1&d=1359285930

http://ecorenovator.org/forum/attach...2&d=1359285930

Assembling: I first tried silicon gluing the sticks on the one face of the aluminum sheet, two on the sides and on in the middle to prevent bending. That proved to be rather painstaking, but also when the glue dried it curved the sheets outwards. So I changed method and used some kilometers of double sided tape. That was much easier and faster to do.
(in the first image you can see a gluing mask I made in order to keep the same distance in each sheet and prevent spilling)

http://ecorenovator.org/forum/attach...3&d=1359286448

http://ecorenovator.org/forum/attach...4&d=1359286448

I also taped the same way the coroplast sheets with the sticks:

http://ecorenovator.org/forum/attach...5&d=1359286741

http://ecorenovator.org/forum/attach...6&d=1359286741

http://ecorenovator.org/forum/attach...7&d=1359286741

(to be continued as I reached the 10 images limit)

(continues from previous post).

I made up a frame to easily stack and hold inside the sheets for each HX using two plywood 10mm thick sheets as top and bottom and 4 angular aluminum profiles for the 4 sides. Placed the sheets inside the frame and closed it with 8 small screws.

http://ecorenovator.org/forum/attach...8&d=1359287635

http://ecorenovator.org/forum/attach...9&d=1359287635

http://ecorenovator.org/forum/attach...0&d=1359287635

With these three kinds of HXs I made up a small bax out of Polystyrene sheets to do the first benchmark. I made the 4 120mm diameter holes and placed two 12V PC fans (arctic cooling 120 m3/h)

http://ecorenovator.org/forum/attach...1&d=1359287823

http://ecorenovator.org/forum/attach...2&d=1359287823

http://ecorenovator.org/forum/attach...3&d=1359287823

I tried these configurations:

SOLID PP CORE
DISTANCED PP CORE
ALUMINUM CORE

SOLID PP DUAL CORE
ALUMINUM DUAL CORE

The distanced PP core was a total looser, did not even reached a 40% efficiency

Solid PP and aluminum core have equal efficiency, reached 98% with low flow (PC fans with no relevant pressure...)

Dual core rates quite the same, but allows much more air to pass trough.
The last dual core photo was a combination of PP and aluminum cores, same results (but that was quite obvious..)

I will post more precise data as soon as I find them.

Very nice work.
I'm a little confused

I tried these configurations:

SOLID PP CORE
DISTANCED PP CORE
ALUMINUM CORE

Which is the Solid PP core and which is distanced????

SOLID PP DUAL CORE
ALUMINUM DUAL CORE

The distanced PP core was a total looser, did not even reached a 40% efficiency

How are you measuring efficiency?
Assuming no condensation and equal air mass flows, I've decided that
efficiency is
1-((stale Tout-Fresh Tin)/(inside temp - outside temp)).

For a cross flow exchanger, if the thermal conductivity of the barrier
is infinite, you can take out the metal and let the air mix. That
says the best you can do is 50% efficiency with the internal temperature
midway between inside and outside temperatures.

As a practical matter, you probably can do a little better
than 50% because there is a thermal gradient across
the core. Stated a different way, I think that a barrier
with less than perfect conductivity does increase the
efficiency by allowing a thermal gradient.

Solid PP and aluminum core have equal efficiency, reached 98% with low flow (PC fans with no relevant pressure...)

98% sounds unreasonable for any decent air flow. I'm too sleepy to
do the math, but I'd expect a series dual core cross flow exchanger
to be something like 75% efficient.

If you imagine a series of many cross flow exchangers, the limiting
case becomes a counter-flow exchanger. At any point along the length
the incoming and outgoing air is at the same temperature, but
that temperature varies along the length. Theoretical maximum
efficiency can approach 100%.

Dual core rates quite the same, but allows much more air to pass trough.
The last dual core photo was a combination of PP and aluminum cores, same results (but that was quite obvious..)

I'm confused by that paragraph.
Two equal cores in series should have twice the resistance
to air flow. If e is the efficiency of one core, I'd expect
n cores in series to be something like 1-(1-e)^n...but it's
too early in the morning to try to prove that.

What am I missing?