Non-Absolute Zero: The Strange Case Of Thermal Bridging

Passivhaus is a tried-and-trusted, reliable system of low-energy housing developed by the Passivhaus Institut in Germany, a non-profit commercial company. I’ve mentioned it on these pages before, in connection with Dún Laoghaire-Rathdown county council mandating it as – effectively – compulsory for all new buildings constructed within their functional area, which will be required by planning permission. In that piece I hopefully made it clear that I was questioning Dún Laoghaire-Rathdown mandating a proprietary system, rather than claiming there was anything wrong with Passivhaus as such.

There are some things wrong with Passivhaus as a system, of course. Before I get a torrent of complaints from Passiv advocates – who, if I may so, can be a slightly zealous bunch – I mean this in the sense that there’s no such thing as a perfect system. The theory’s simple enough; it’s a pass-fail standard, a house is certified or it isn’t. There are no renewables; it works by limiting heat loss. The house is heavily insulated, thoroughly draft-proofed, and a mechanical system uses heat recovery to minimise the heat loss through ventilation. Passivhaus clearly performs at high levels, and reliably delivers low-energy housing with good ventilation.

So that’s what’s good about it. On the downside… it’s perversely easier for large houses to comply than small ones. The emphasis on mechanical ventilation makes sense in central Europe, but is a little strange in a mild climate like Ireland’s. There’s no cost-benefit analysis between the cost of insulation versus the cost of renewables, so insulation might not always be the cheapest solution but has to be used anyway. And there’s also the question of the type of living it encourages, which maybe concerns me most of all but is rather too nebulous to discuss here.

However, one of my biggest concerns with Passivhaus is its use of language. Part of its appeal appears to simply be how it describes itself; even the now-standard translation “Passive House” isn’t really accurate. It uses a word (“passive”) which suggests something heated completely passively – by solar gain, for example – even though this doesn’t actually apply to Passivhaus in a meaningful way. It’s been stated elsewhere that Passivhauses are often stated to have no heating system, even though this is simply not the case¹. “Passive House” doesn’t formally claim to have no heating system – it just has a name that heavily implies it does. In much the same way, when asked about DLR’s decision to mandate Passivhaus, Dan Boyle (the former Green Party senator) claimed on Twitter that the term “Passive House” could be used generically, like “Hoover” or “Photoshop,” to mean carbon neutral housing. But Passivhaus isn’t carbon-neutral.

I’m not suggesting any serious mendacity on the part of Passivhaus here, but sometimes it’s easy to be reminded of the Laboratoires Garnier sketch by Mitchell & Webb… “The invention of the word Nutrisse, which sounds like nutrition but doesn’t guarantee it, is some of the best work you’ve ever done!”

This ran through my mind when reading this builder’s-eye view by Paul Doran in Passive House+. It’s an interesting article which I’ve no doubt is written with integrity. However this issue of language crops up repeatedly in the piece – and it’s particularly problematic when comparing Passivhaus to the Building Regulations.

This passage in particular makes for interesting reading: -

[Eliminating cold bridges] was vital for the passive house standard, which is unequivocal in accepting no thermal bridges. Yet how important was it for Part L which specifies thermal bridging should be reduced, ‘“insofar as reasonably practical”. To me this appeared to be a ‘get out of jail free card’ for having to eliminate cold bridges.

This is… damning, on the surface. Thermal bridging is a huge factor in how much heat a property uses. One of the main boasts of Passivhaus is that it provides a construction free of thermal bridges, so it’s reasonable to ask why Part L doesn’t?

To give a concise explanation of thermal bridging: heat doesn’t behave in a simple way, escaping neatly through every external wall, floor and roof at right angles. It tends to go through the lines of least resistance – joints between one construction and the next where the insulation will narrow, or just where the geometry of the building allows more heat to escape (such as corners). In a way, it helps to imagine that heat escapes through a building like water. Anyway, these lines of least resistance are called “thermal bridges,” and the extra that escapes through them is called thermal bridging. To make the magnitude of the thermal bridging issue clear; Part L default values assume that roughly a third of total heat loss for a new house will come through thermal bridges. In reality it’s unlikely to be quite that high, but even so you can see why Passivhaus – “free of thermal bridging” – seems like a no-brainer.

To highlight just how much of an issue thermal bridging can be, the image below shows a (highly) simplified rising wall detail using a strip foundation with PIR insulation beneath the slab. This is standard construction in Ireland, and the coloured view on the right makes it clear that heat is pouring through out junction.The Ψ-value² of 0.130W/mK means that, if there’s 20 degrees C between inside and outside, heat is escaping through every metre of that junction at a rate of 2.6W. Added up around the whole house, that’ll be roughly the equivalent of leaving an old-fashioned incandescent light bulb on 24 hours a day.



This detail lets a lot of heat escape, although it is possible to improve it. Adding a lightweight, aerated concrete block (these are cheap and easy to buy) cuts down on heat loss enormously. However, the coloured image makes it clear that even that detail is letting extra heat escape: -

Not surprisingly, Paul Doran wasn’t impressed. So instead he looked elsewhere: -

I received a quote for the Viking House passive slab from Airpacks which appeared to be very reasonable. It gave a U-value as low as .09, which was more than sufficient for PHPP. It also eliminated all cold bridges at the floor-to-wall junction. This was vital for the passive house standard…

Viking make their details available on their website, but here’s a simplified colour detail of a slab meeting a full-fill cavity wall: -

The benefits are obvious. Because there isn’t a strip foundation, it’s possible to have a continuous line of insulation. The insulation here is expanded polystyrene – this doesn’t resist heat as much as PIR, so there’s a greater thickness of it – and this is cleverly used to wrap around the split foundation. It’s a neat system and I may well use it myself one day (although I find the shallow foundation a little worrying – I’d need that clearing up).

So this is what Passivhaus means by thermal free of thermal bridges, right…?

Not only is some heat escaping, it’s performing fractionally worse than the traditional strip-foundation with an aerated block. How can this be called “free of thermal bridges”?

And here is where the issue of language crops up. What Passivhaus calls a “thermal bridge” isn’t what the Building Regulations, here or in the UK, define as a thermal bridge. Without wanting to get too technical, the Building Regulations calculate thermal bridges as an “extra over” based on a theoretical heat escape from the inside of the walls, Passivhaus measures the theoretical figure from the outside of the walls. Using the outside of the walls as a basis will tend to overestimate the heat loss to begin with – so the “extra” lost through thermal bridges is lower. Hence you can claim that your construction has no thermal bridging at all, even though – all over the UK and Ireland – only Passivhaus defines a thermal bridge this way, and it’s a frankly illogical definition when computer software shows you the extra heat flowing out through the junction.

To be clear, neither one of these methods of calculation are inherently better as such. The two means of calculating a thermal bridge will give the same figures eventually, they’re just different ways of skinning a cat – it’s just that one gives a much nicer soundbite than the other. Passivhaus isn’t free of thermal bridges as the term is commonly understood, it is free of something different that it just happens to call a “thermal bridge.” In fact, “limiting thermal bridging as far as is practicable,” as the wording in Part L is written, is exactly what Passivhaus also does.

Let’s be absolutely clear: I don’t want this to be understood as me trying to discredit Viking’s system. It performs fine. The article claims it’s cheaper and – as it’s got a 100mm floor slab instead of the standard 150mm, and also uses far less material below the ground – this seems believable to me. However, getting back to that article, Paul Doran initially rejected a traditional strip foundation because of Passivhaus’s claim to build with no thermal bridges at all. And this isn’t really true. In fact, use a strip foundation with EPS instead of PIR below the slab, and you can outperform it further still: -

In other words, this misleading use of language is an issue. Passive House – as it’s generally (and inaccurately) termed here – is becoming seen as a gold standard, partly because it performs well, but also because it cleverly uses terms that promote its brand. More worryingly, this has succeeded with those in power, as we can see by Dún Laoghaire-Rathdown mandating it and the Green Party, among others, enthusiastically touting it.

This isn’t surprising. It’s simply what brands do. It’s pointless and counterproductive to blame Passivhaus for trying to sell itself as a system… but it’s also important to analyse the marketing language and clarify what it really means. What the Passive House+ article inadvertently shows is a well-intentioned, conscientious, capable builder who appears to have been misled by the marketing speak of “no thermal bridges”. That’s not an admirable or desirable thing.

The deeper issue here is the state of Part L of the Building Regulations. Let’s set aside the fact that Part L has a requirement for renewables which doesn’t make any sense; not worry that its Approved Thermal Details are in some instances questionable, and in some instances non-existent; ignore those other bits of supporting information to be found on the DECLG website with no real context, making them easy to misunderstand. The core point is that if the government standards are a document that’s extremely difficult to understand (even for trained designers, let alone for a conscientious contractor or a curious self-builder), if there’s almost no support for those coming to grips with it, if there’s no certification proving that people are competent to use it… then what you create is a vacuum in knowledge in the construction industry. Insulation is a new science. There’s no “folk memory” to rely on, no knowledge from the past that can be absorbed and passed on. This stuff has to be taught, and commercial organisations shouldn’t be doing it.

Paul Doran’s still built a good house, his client has a good product, and everybody’s happy. However it’s worrying that an article in Passive House+ can suggest that a Passivhaus can be built more cheaply than a regulations-compliant house, and nobody even challenges this claim – even though it requires some pretty selective data to stand up.

Until some real work is done to make the Energy-Efficiency regulations more understood, this knowledge gap will widen and these sort of marketing-led claims will go unchallenged – and, even if you accept that Passivhaus is a positive or at least benign organisation, this is still an issue for concern. After all, if Passivhaus can get away with this, then so can anybody else. The problem here is that, with a lack of knowledge in the industry and nobody in government making serious efforts to remedy that, marketing jargon can become confused with actual data – and nobody benefits from that.

1. A more detailed list of articles referring to Passivhaus as being “a house without a heating system” can be found in this article, which also discusses some of the limitations of the system.
2. A ψ-value is a measurement of how much extra heat is flowing through a junction. They vary according to the U-Values of the elements around them. The values in this piece are derived using LBNL Therm, assuming an 8x12m slab internally. PIR insulation was assumed to have a conductivity of 0.022W/mK; EPS, 0.038W/mK; XPS, 0.029W/mK; blown bead insulation, 0.033W/mK. Concrete blockwork is assumed to be medium-density. Plaster/render has been omitted for simplicity.

About Mike Morris

A British / Irish architect who practiced in Ireland and Ontario, now working in the U.K.
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>