There's a rooftop we recently discovered with a great view, and it's easily accessible for us. Obviously we'd like to go up there a bit more, particularly in summer with a couple of fold up chairs or something. I was thinking of getting someone in to look at it properly to assess the structural integrity of the roof, but I figured it might help posting on here first to get an early opinion from anyone willing to offer some advice.

The roof feels solid. As with any metal roof, there's obviously a few minor creeks here and there, but as a whole it does feel strong with 3 people up there. It's also reassuring to note that workers go up there quite often, and hence the walkways that have been setup on it so far. We want to set something up similar to the current walk ways - a slab of plywood for instance that will distribute the weight evenly across the vertical seams when standing or sitting on the roof.

The roof area is 6.25x8.50m and here are some photos we took: a b c

The advice anyone gives on here will not be taken with absolute authority - I understand there's only so much you can do with the limited data I've provided. All advice will be considered general and not relied upon, but whatever thoughts you have to share would be greatly appreciated. I'm particularly interested in comments of absolute rejection to us going up there, along with a reason of substance explaining why it's profusely dangerous for us.

Someone also talked about running the structure in the opposite direction to the support beams. Would this be better, from the point of view of safety?


  • $\begingroup$ Structure safety is one thing, and as @wasabi points out at his answer, living safety is a super-set of all technical safety concerns, including procedural allowables (i.e. having a small party, singing, dancing, jumping?). Dynamic loads is something that really might affect a lot: a person can jump up, and thus create ~3 times his weight. Worse case is, he could arguably jump horizontal and kill himself (falling from the top of the building). Good question. $\endgroup$ Aug 2, 2015 at 17:59
  • $\begingroup$ You don't say where this is, but since it's a flat roof there must be local codes as to how much snow load the roof has to be able to handle in the winter. A few people on a sheet of plywood isn't going to come close to the loading fro 3 feet of worst case slush across the whole roof. Just don't go up there and add your weight when there is already snow on the roof, and you should be safe. $\endgroup$ Aug 2, 2015 at 21:59
  • $\begingroup$ On a separate topic, once the owners find out you're up there, they'll probably lock the access to avoid liability issues. That roof doesn't have railings or a ledge, so clearly wasn't designed to have ordinary people hanging around on it. The owner might get into trouble if the right kind of inspector happens to see you up there. $\endgroup$ Aug 2, 2015 at 22:02

2 Answers 2


As you mention, this is a question that cannot possibly be answered in any definite way without an on-site inspection and, even then, it would be a (much more) educated guess.

That being said, in broad terms we can state that it's probably structurally fine to do so, so long as you don't overcrowd the roof. What exactly "overcrowd" means in practical terms I can't possibly say, but for an easy low-bar number of what is obviously acceptable, you can use the maximum number of workers you've seen up there at a time. This, however, is without a doubt highly conservative.

A better guess can be made if one knows the country and date (roughly) the building was constructed, from which one can check the relevant codes of the time for the design load of roofs. The current Brazilian codes, for instance, give three different design loads for roofs (prior to safety factors):

  • With access to the public: $3\text{kN/m}^2$ (around $4\text{pax/m}^2$)

  • Without access to the public: $2\text{kN/m}^2$ (around $2.5\text{pax/m}^2$)

  • Inaccessible: $0.5\text{kN/m}^2$ (less than $1\text{pax/m}^2$)

For an idea of what these population densities mean in practical terms, see here. $2.5\text{pax/m}^2$ is already quite a party.

In Brazil, such a roof would have to be designed for at least $2\text{kN/m}^2$, arguably (since you apparently had no issue getting up there) even $3\text{kN/m}^2$ (obviously, without express knowledge, you'd have to assume $2\text{kN/m}^2$). These values change from country to country and from time to time (revisions to codes may alter the values), so knowledge of the country and approximate time of construction is absolutely necessary to make any sort of educated guess.

Regarding aspects of "living safety" (no clue what the actual term would be), where there are fold up chairs there is alcohol, and given that this roof has no railings, that is arguably not the best of combinations...


In the UK roofs with access is 1.5 kN/m2. But the roof may not have been designed for access. It is safe to say however that the roof fabric will have been designed for access during construction.

Why not give an easily understood criterion such as no more than two people on the roof at any one time and to limit equipment carried to say 50 kg? At least this is clearer than an 'X' kN/m2 criterion to the layman.

If there were any long term works to be carried out or if there was any concern you should inspect from the underside, exposing structure from removal of any ceiling locally.

As an aside a report was done for Stanhope in the UK some years ago into insitutional over specifying of floor loads in the City of London (5 kN/m2 sometimes). One of the examples in the report showed a crowd standing in close formation with a quoted calculated load of 2.5 kN/m2.

  • 1
    $\begingroup$ A comment on your aside: paper has (roughly) a density of 1000kg per cubic metre. This provides a weight of 10kN per cubic metre. So in order to get 5kN per square metre, you'd need a stack of paper half a metre high. Does it sounds unreasonable to design an office floor for half a metre of paper? It doesn't to me! $\endgroup$
    – AndyT
    Aug 26, 2015 at 14:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.