MHRV Oppsie

An Eccentric Anomaly: Ed Davies's Blog

Over the last few days I've been writing a fairly detailed blog post about the design of the mechanical heat-recovery ventilation (MHRV) for my house. I got it all finished yesterday afternoon, did some shopping, then settled down to proof read it before uploading in the evening.

As I read it I thought of a point I should, perhaps, describe a bit more carefully so started scribbling a little diagram to help explain it. At this point I realised that my whole design (of the MHRV heat exchanger, not just the description in the blog post) was badly wrong.

Oops - feel a bit like this kitten. Back to the beginning - a post on the basics…


If a house is airtight enough not to be draughty (or, at least, over-ventilated) on a windy day it'll need some help with ventilation on a calm day. There are various ways to do that but the most obvious and easiest is mechanical ventilation: fan(s) to blow air into the house and/or to suck air out.

(Just blowing air in is probably a bad idea as it'll over-pressure the house and push warm damp internal air into any leaks in the walls, likely causing condensation. (This appears to be less of a problem in a house with trickle vents, says this GBF discussion but then who'd have trickle vents in a house that's supposed to be energy efficient?) You want to just suck or suck about the same or a bit more than you blow.)

There are other, non-mechanical, methods. For example, you can use the stack effect where the warmer house air creates a pressure differential between the bottom and top of the house. That needs some sort of controllable valve to regulate the flow rate so there isn't over ventilation. It also makes heat recovery a bit more awkward (though not impossible).

Anyway, despite the power requirements to run the fans, I'm planning to use mechanical ventilation as that seems more appropriate to a house occupied on one level (where, to the extent that warm air does rise into the loft and cathedral ceiling areas it tends to be a bit wasted).

Once you've got all your incoming and outgoing air nicely pumped through pipes it seems a shame to just dump warm air outside and pull in cold air so the typical cunning plan is to use the outgoing warm mucky air to heat the incoming cool fresh air using a counter-current heat exchanger. That's the “heat-recovery” bit.

Of course, MHRV boxes are available commercially but it seems to me that they're quite expensive for what they are so I'd like to make my own. I'll have to see if that works out to be sensible from a building regulations point of view (SAP ratings and all that) but the design is an interesting exercise anyway.

Probably one reason that commercial heat-recovery exchangers are relatively expensive is that they need to be compact for practical storage and shipping. A DIY exchanger can be constructed on site and so can be made larger, more robust and, hopefully, cheaper.

Exchanger Parameters

I'll use these variables for the temperatures of the airflows into and out of the exchanger:

To Fresh cool outdoor air into the exchanger to be warmed
Ts Warmed outdoor air from the exchanger supplied to the house
Ti Mucky warm wet air internal to the house fed to the exchanger to be used to warm the incoming air then got rid of
Te Mucky cool very-wet air exhausted from the exchanger to the outside

When there's no heat lost or gained through the outer walls of the exchanger the heat gained by the incoming air should be the same as that lost by the outgoing air. Further, when the airflows in each direction (through the exchanger) are the same, there's no condensation or evaporation happening (ie, no changes in latent heat) and approximating that the temperature and energy are proportional (ie, the specific heat capacity of the air doesn't change) the temperatures will balance:

Ts - To = Ti - Te

Exchanger Efficiency

The efficiency of the exchanger is the amount by which the energy of the supply air is increased over that of the outdoor air relative to the energy lost in the outgoing indoor air. Continuing with our simplifications, it's:

e = (Ts - To) / (Ti - To)

Note that the exhaust air temperature doesn't play in this game - that's gone and no longer of interest.

Exchanger ΔT

Obviously the temperature difference across the exchanger makes a big difference to the rate at which heat is transferred from the outgoing to the incoming air. In an idealized situation the temperature difference is uniform along the path of the exchanger. The warmest outgoing air from inside the house is used to add the last bit of warmth to the incoming air and as it cools on its way to be exhausted it meets progressively cooler incoming air going in the opposite direction.

ΔT = (Ti - Ts) = (Te - To)

Or, to put the first part a different way:

Ts = Ti - ΔT

Substituting for Ts in the efficiency equation:

e = (Ti - ΔT - To) / (Ti - To)
  = [(Ti - To) / (Ti - To)] - [ΔT / (Ti - To)]
  = 1 - (ΔT / (Ti - To))
ΔT = (1 - e) (Ti - To)

In other words, the greater the temperature difference across the exchanger the smaller the efficiency is. If the ΔT is one tenth of the difference between the indoor and outdoor temperatures (ΔT / (Ti - To) = 0.1) then the efficiency will be 0.9 (ie, 90%).

My Blunder

The mistake I made was to assume that the temperature difference across the exchanger interface was nearly equal to the difference between the indoor and outdoor temperature. In effect, I was, for some reason, thinking that:

ΔT = e (Ti - To) WRONG!!!!

Consequently, my designed size of the exchanger was about one tenth of that which would be needed. My intended material (twinwall polycarbonate) would have been fine at the designed size but gets a bit pricey when scaled properly. More thought needed.