Why I have refused to build lightweight balconies, terraces or green roofs, and why I’m willing to change my mind. A working paper.

I have never looked into the stats behind balconies, terraces, and green roofs (defined as those that are framed in structural timber and steel), but I have always assumed that each of the above fits neatly into one of two categories in terms of watertightness - ones that have failed and ones that are about to fail. I have gained a lot of credibility with property buyers and homeowners by reliably predicting the failure of roof terraces.

I’ve always thought that there is a fundamental structural design flaw that is a problem rather than faulty workmanship. One thing I have suspected is that the combination of timber and steel is a contributing factor to the problem. How so? Well, timber shrinks in the heat as it dries and expands in the cold wet winter months - steel does exactly the opposite. Throw in a cement bed which just keeps shrinking and sometimes a dash of ply or even worse, particle board. What you have is shit sandwich of materials and a system slowly tearing itself apart. Then, as moisture starts getting in, it creates a feedback loop. This is consistent with observations I have made.

And even if the structure were miraculously stable, lightweight terraces are devilishly tricky to waterproof. Yes, the membrane application is easy enough - but it’s where you terminate the membrane and the penetrations including drainage, windows, balustrading, and many other things that make it difficult.

So let me tell you one thing; as a licensed builder - if you ask me to build a lightweight terrace roof garden or balcony above a livable space or even a garage - I won’t do it. I’ll help you redesign the terrace using concrete and structural steel posts, or I will decline the job - no exceptions. If you already have a leaky lightweight terrace, I will help you fix it - I’m not a complete asshole - but I will clearly tell you that it will be a recurring maintenance problem over time. And I will insist that if we fix it, we’ll use high-spec solid membrane - not liquid applied - and pedestal-installed pavers like those used in commercial roofs which can be readily lifted up to inspect the membrane every year.

As I have said - in my experience all lightweight green roofs, terraces and balconies will fail, it’s just a matter of how long.

Concrete on the other hand doesn’t tend to fail as long as it’s well vibrated, well engineered, well cured and has good cover (meaning that the reinforcement has a barrier of concrete between steel and the elements). But it gets better, concrete has amazing self healing properties where moisture that gets into a slab triggers more crystal growth closing any microvoids (is microvoids a real term or did I just make it up). In other words concrete has a tendency to be a waterproof element in and of itself.

I have been adamant that if you want a green roof, balcony, green terrace then it will have to be concrete.

But there’s a big price for concrete - both in terms of cost, carbon footprint and design implication for the entire structure which lead to more cost and more carbon. For example a suspended slab increases the size of supporting walls but also nudges the entire structure to be concrete. This is because it has never made sense to me to pour a part of floor in concrete and then build another part in lightweight - I have always figured that if you have formworkers on site you may as well keep going and speed up the construction program. 

So in a nutshell concrete has a tendency to increase the amount of concrete and steel beyond the terrace from bigger foundations (footings) to thicker walls to scope creep exacerbating the carbon footprint and cost.

But it gets worse in an era when we need to reduce our carbon footprint because concrete is not a great insulator either.

It’s a durable system that needs a rethink.

This is why - with some trepidation - I’m forced to rethink my concrete or nothing ultimatum.

Here are some ideas:

1. Waterproofing system is designed into the building. What does this mean; from my experience building designers and architects tend to draw terraces that will never work because they are impossible to waterproof efficiently, don’t have enough thickness to created adequate step downs and falls and generally don’t factor in waterproofing systems on the basis that the builder will work it out and if we pay enough for an expensive enough membrane it will all be fine. 

2. Make the building structural elements all steel (no timber) and the cladding elements all cementitious.

3. The use of connectors that are round steel elements - say 50-90mm diameter that separate the floor element from any walls enabling you to encase the entire structural element in membrane with round gussets for the posts

4. Waterproof the entire structure including walls

5. Delete Balconies - Balconies are a bit like designer shoes that look amazing but don’t quite fit your foot. They are nice to look at but you hardly ever use them and when you use them they hurt you.

6. Metal roofing and removable garden above separated by a void

So initially I was going to share this with a few engineers, waterproofers and industry experts, maybe do a bit more research but life is short and I thought I would share it and release it as a work in progress paper and add to it.

I welcome input from others in the industry.