I mentioned some time ago that I had a unventilated ceiling to insulate for my current renovation. It's on the top floor of a small apartment block. The existing structure, top to bottom, is:

- a tiled pitched roof attic that's subdivided into storage units used by various residents
- terracotta floor tiles in lime mortar
- thick timber planks
- joists at close centres that span between beams.

to which I'm adding insulation and a dense plasterboard ceiling.

Being rather tricky, I thought I'd better commission a WUFI analysis, and being optimistic I thought I'd kick off seeing what the results would look like with 300mm of hemp insulation batts. The results are now back and don't look pretty. With luck, the images will be below.

I'm now considering how far to change the spec for another run of WUFI. I know that some of you - Tom, maybe Tony, maybe others? - have played with WUFI in the past, so am hoping that you may have some thoughts on this.

An obvious track would be to cut the insulation. 200mm would be my ideal minimum (U = 0.21 W/m²K). In view of the results for 300mm, any opinions on this or an alternative thickness?

No doubt adding a smart VCL would make WUFI results look better, but in reality the stud partitions would pass through and bypass it (they're fixed to the structure). That's leaving aside the advisability of using VCLs in retrofit anyway (I think my scepticism is shared with Tom too?).

So, as I don't have endless funds to spend on WUFI models, what to try next?

A vapour variable membrane will certainly help, and there are some more options coming onto the market like the Partel Vara plus eco that reduce the plastic content.

My company tries to design without membranes wherever possible, but there are always edge cases like this that require them. They require careful detailing and sealing to work effectively.

Some smart membrane manufacturers offer free WUFI studies if you specify and use their membranes:

SIGA Majrex®
Partel VARA PLUS ECO
PRO CLIMA intello plus

I'd also recommend speaking to Chris or Andy at 'Back to Earth' as they have experience with tricky renovation jobs like this and understand the building physics involved.

Thanks for the feedback so far. Re the points made:

Posted By: sgt_wouldsThey require careful detailing and sealing to work effectively.

Yes, from having done airtightness I know that's not easy on a renovation. And stopping moisture is trickier than stopping air. Then the VCL would have to stay undamaged for the next few decades.

Posted By: sgt_wouldsSome smart membrane manufacturers offer free WUFI studies if you specify and use their membranes:

That's certainly interesting to know - I used Pro Clima Mento as part of the airtightness on this project.

Posted By: sgt_wouldsI'd also recommend speaking to Chris or Andy at 'Back to Earth'

And another interesting tip, thanks.

Posted By: djhWhat temperatures and humidities have you assumed above and below the ceiling? I see the times are in the past, so did you measure them?

The external temperatures are based on real ones from the local 2009-2020 weather data. Not sure exactly how WUFI uses them, but from the dates on the graph I'd guess that it's using the actual data from those years.

Internal temperature and humidity used are based on ASHRAE 160 - not something I've studied, and I know that there are other options, but presume that's a reasonable choice. From just having Googled it, looks like that uses a (minimum?) indoor temperature of 21°C, and that the indoor RH is related to the external RH, presumably from the weather data. Thinking about it, what I'm less sure about is what additional vapour load is added to account for people being in the apartment - for example whether that takes into account a specific number of people, or a flat load per m³, or whatever; maybe that needs looking at?

I've also got MVHR factored in at 0.42 ACH which, after calculating it in multiple ways according to the recommendations of various different organisations, seems a sensible minimum in this case (for 1 person occupancy).

Another thought since yesterday is whether there might be any benefit in tapping into the MVHR supply & return to add some minor ventilation to the space above the plasterboard. Never heard of anyone doing that, but it would (partially) decouple it from the internal moisture load. Hmm.

Posted By: djhwe used Intello+ in our roof and what's now called Solitex UM connect on the outside, under the standing seam. ...I have no idea whether it provides any tangible benefits.

I have seen a couple of WUFI studies comparing elements with and without smart VCLS; I may have a one filed away somewhere on my PC. From memory, they had the effect of smoothing the highs and lows, but weren't dropping the HR behind them radically. But the Solitex UM Connect on the outside would certainly be very helpful.

Posted By: djhisn't the space above the ceiling internal rather than external? In which case measuring that seems like the best approach too.

Yes, it is, but at the moment there is no ceiling, so I can't measure it yet. It may be worth putting some sensors up their when I install it though, if I can find something suitable. Anyone know of a university that may be interested?

Posted By: djhIf I'm reading the graph correctly it's saying the humidity under the ceiling is pretty much always between 70% and 75%? That seems way too high for the inside of a building even without an MVHR. With an MVHR I'd expect it to normally be between 40% and 50% and never higher than 60%. Something's weird.

Well spotted! That could be the key to the problem. I was reading 'point A' as being above the plasterboard, but checking the text it does say that it is 'a control point to represent the impact of the indoor environment' - which probably does mean that it is on the underside, in the room. If so, that does indeed look very wrong. I'll go back to the consultant and check that.

Posted By: djhFor calibrating MVHR, I use 30 m³/hr per person rather than an ACH figure.

I calculated it various ways (litres per person, by room and by volume using UK, French, German, Passivhaus and ACECB guidelines). The volume in this case is only around 90 to 100 m³, so using your 30 m³/h that would be roughly 0.33 ACH, compared to my 0.42, so we're not too far apart on that :)

Posted By: sgt_wouldsFrom your build-up, (hard to see the text) it looks like you have a service void between the plasterboard and insulation.

Actually only about 50% of the ceiling area has a void - the diagram was for modelling the 'worst case'.

Posted By: sgt_wouldsI know that in in Europe, some ventilated roofs (eaves and ridge ventilation but over 10m length) are required to have additional measures, such as connection to forced ventilation which may include MVHR. There must be some information 'out there' about effective strategies for force ventilation of narrow cavities.

That's interesting to know and would certainly be worth trying to track down. Otherwise I've been thinking that irrigation drip-pipes, or the equivalent in small diameter waste pipe, tapped into the ceiling plenums, might be worth experimenting with.

However hopefully it won't be necessary, as djh's spot on the internal RH is surely pointing to an error in the modelling. I've contacted the consultant so we will see.

Posted By: Mike1In fact, there's nothing in WUFI that can properly simulate an MVHR internal climate. Apparently the entire building would need to be simulated in WUFI Plus (€1,790.00 + VAT for a 1-year license) to achieve that.

Surely WUFI Passive must deal with an MVHR? I don't think there are many without one!

Posted By: Mike1I'm wondering whether any of you may have MVHR and logged temperature and humidity data, and may be willing to share it?

I have internal T&H data logged from last November. I've got three external readings since 2018 or so - one is my own in an outdoor cupboard; one is a neighbour's weather station, which goes offline for a day or more at a time sometimes, and the third is Wattisham airbase which is pretty reliable but six miles away.

Just to add that I think the internal T&H values and relationship will depend quite a bit on the materials in the house.

Duplicate post

Much appreciated - whisper just sent.

Posted By: djhThe main logging system is an OpenEnergyMonitor system < https://openenergymonitor.org/ >

I'll certainly take a look at that, as I've not progressed the thought process much beyond the sensors.

Just been taking a deeper look as sensors, and the BME280, which looks good and is well priced (if they're genuine). The BME680 seems less common and costs 3x as much, without a 3x jump in performance difference. I also found someone who'd done a comparison, in which the Bosch came out best too: http://www.kandrsmith.org/RJS/Misc/Hygrometers/calib_many_afterbake.html

However I then came across the SHT30-DIS-F (and some in the newer SHT4 range), which incorporates a PTFE film to protect it from contamination, which would be a plus. The sensors are cheap enough, but I haven't spotted anyone selling one on a PCB yet (other than the CN0537 £60 smoke detector board). I don't have time to play as well as doing a renovation, so the BME280 is looking most likely - thanks for pointing it out.

Posted By: Mike1However I then came across the SHT30-DIS-F (and some in the newer SHT4 range), which incorporates a PTFE film to protect it from contamination, which would be a plus.

Posted By: djhI'd forgotten the SHT4 range. Adafruit sell them on little PCBs

Thanks for the Adafruit tip - it helped me find some similar DFRobot SHT31-F sensors (SEN0332). The PTFE protective film seems worthwhile as they'll be buried in hemp insulation. I've ordered a couple to test, together with Adafruit LTC4311 range extenders, to overcome the distance limitations of I2C.