Steel thermal bridges: an engineer’s perspective Why they are a problem and how to avoid them

17 October 2022 by Jason Quinn

Last month’s blog post about thermal breaks for structural steel prompted some reader feedback, including from structural engineer Paula Hugens (director at eZED Ltd). Paula is well experienced with Passive House projects and she’s been railing at the egregious use of steel for years. 

So this month we revisit a piece she wrote in 2014. I love the opening line!  That sure nails it. First though, some remarks from me for the sake of context because fortunately some things have changed for the better in the past decade. There are now some well-educated clients who are already aware of the issues steel can create when it penetrates the thermal envelope; in general it’s now relatively easy to explain what the issues are.

I wish we could say we’re seeing a solid shift away from structural steel in both commercial and residential buildings but it’s still spotty. Some projects just design it out but there are still lots of drawings coming over our desk where either the Sustainable Engineering team or the client is forced to really lean on the structural engineer in order to make better choices (never Paula, just to be clear).  

One project that springs straight to mind involved the client finding and supplying laminated timber span tables to the engineer, to prove a steel beam could be replaced with a timber beam. The burden of research and education should not fall to clients. That’s why they hire experts. It’s up to us to be expert.

I also wish there was already  a strong swing away from steel because of concerns to reduce embodied carbon … but that’s not true yet. There are a few clients leading the way in this regard and I hope this will become far more commonplace in the next 12 months.

Finally, Paula (along with Jessica Eyers from Hiberna) continued more analysis for years after she wrote this initial article. The conclusion? The effects of steel thermal bridges are even worse than she had assumed back in 2014.

To this I’d add that it’s becoming clear that, in most buildings, the risk of mould created by these poor details is even higher than initially feared. That’s because of the low ventilation rates in Code-minimum buildings. That applies just as much to the bespoke expensive homes as it does to those churned out by group home builders. (Remember that awarded EnerPHit project in Piha was spurred by a correct prediction that the ceilings would be lined with mould.)

Over to Paula Hugens:

“A bunch of details are regarded as Acceptable Solutions in New Zealand when they are not even close to being satisfactory. I have no idea how they became so entrenched in the industry. Convincing building professionals that these familiar details are a problem is a huge battle. They’re convinced that change must involve additional cost. But what about the long term cost?

I want to cast the spotlight on some of the worst details we regularly come across and explain why they are such a problem. What I can’t fathom is that some of these details are being used on self-proclaimed “sustainable buildings”. As far as I am concerned, specifying any of these deserves an automatic fail. 

Some building consent authorities have worked out some of these problems themselves—but parts of the New Zealand Building Code are so permissive that consent authorities are left toothless. It also makes it hard for us to argue a point when we are told, “But that is how the detail is shown in E2”.

Structural steel ridge beams

As a structural engineer I really relate to this problem. In a structural sense there is no problem, you have a nice strong steel beam that’s going to hold up your roof. But have the consequences on the other building elements been considered? 

Steel ridge beams are a default solution for many architects as they are wanting to keep their roof structure as thin as possible plus create living spaces with epic proportions. I can’t altogether blame my esteemed peers for this as they are just responding to the brief they are given.

So what is the problem? 

It’s a massive thermal bridge right along the highest point of the building where the stack effects are most vigorous. Let me explain.

Below is a typical example of a ridge beam detail I am describing. The roofing and underlay etc has been omitted for clarity. The steel beam provides a perfect thermal bridge and being highly conductive the bottom flange surface temperature will be in equilibrium with the top flange. In other words the bottom flange surface temperature will be virtually the same as the outdoor temperature.  

A surface temperature below 12.6C will see relative humidity rise above 80%, the ideal combination for mould to form. Left unchecked this will lead to structural decay in the adjoining timber members. Surface condensation will form when the temperature drops below 9.6C. You don’t need surface condensation for mould to grow, just moisture laden air. Mould will happily grow on all of these surfaces, obtaining its food from air-borne particles. 

Now here comes the rub 

The stack effect during the heating season causes the warm air to rise up into this junction. Warmer air has a greater moisture carrying capacity and when it cools there is a greater moisture load. So the mechanism for condensation formation is increased, dramatically increasing the risk of mould.

If there are any recessed light fittings warm moist air will feed into the roof cavity and cause mayhem. They are like little chimneys!

Using common building methods and materials there is no real method of controlling the air infiltration into the connection. A perfect moisture-rich environment is created. 

Will an airtightness membrane save the day?

Let’s assume that an airtightness membrane is installed to the underside of the timber rafters. [In the past] the installer would be tempted to fix it directly to the underside of the steel beam; it is airtight after all. Installed this way, the membrane won’t help at all as the cold surface is still exposed to the rising warm moist air.

It may be reasoned to run the airtightness membrane right across the underside of the steel member; this would prevent moisture reaching the steel, wouldn’t it? Unfortunately this is not going to work. The problem has simply been transferred to the membrane as this is going to be chilled by its close contact with the steel. I would expect to see moisture beading on the membrane and dripping back down. The ceiling battens are likely to be saturated and start to decay.

What about a ridge vent?

What is a ridge vent going to do? A cold surface has a localised zone of high humidity. Ventilation won’t prevent this. The cavity would have to be blasted with air to have any meaningful effect. Far better to raise the surface temperature to avoid the localised increase in moisture vapour in the first place. 

How do you fix the problem? 

Easy, don’t use structural steel. Look at timber alternatives such as LVL, glulam etc. If really necessary use a steel flitch plate that is well embedded into the timber members but check the surface temperatures to make sure you have it well covered. Only in very, very cold climates will timber start to be a problem.

My philosophy is that steel members should either be in or out.  Never have them within the thermal envelope. On that basis, there are two alternatives if structural steel must be used. 

  1. Lower the steel beam so that it’s exposed below the ceiling line and kept warm. Alternatively you could create a coved ceiling line to disguise the steel. Make sure the insulation is running right across the top of the beam. Don’t use any silly steel fixing cleats to the top of the beam, these will be thermal bridges, so you won’t win.
  2. Don’t expose the steel to the indoor environment. Instead, run the insulation underneath the steel beam. This is probably best suited to a cove ceiling as you can’t be skimpy on the insulation here

Apply these principles not just to structural steel ridge beams but any structural steel members positioned within the thermal envelope, such as posts, lintels, rafters, studs, portal frames etc.”

Drawing and photo: Paula Hugens