Butler Space Grid system?

[OCellini]

Hi all. I am analyzing the roof of a building from the '60s/70's that is comprised of Butler's Space Grid System. I have found much older forum topics for this same roof, but Al Harrold at Butler has since retired, and Butler is not able (or willing) to supply any further info than what I've received from them (attached). I'm hoping someone can help me locate the information I need to analyze the roof's load capacity (we are installing solar panels on this roof). Thanks in advance!

Olivia

https://files.engineering.com/getfi...-4aca-8881-87a4ebdf4806&file=SpaceTrusses.pdf

 

[driftLimiter]

The detail you have received is much more than I would have expected. But you need more specifics about member thickness and profiles, steel grades, connection details, and any modifications made to the building since construction.

I think this one is going to require some feet on the ground(on a lift perhaps) with a drill and some calipers. Complicated by custom connection details at the nodes. I don't think its gonna be easy unfortunately. Perhaps some coupons sent off to the lab for material testing?

I'm curious what your plan of action is going to be, in my view this is not a simple task. I guess you could live on the prayer that someone here knows the specifics of material and section geometry that this company was using 50 years ago but would that really be enough to make an informed engineering conclusion?

 

[JedClampett]

Or three choices;
[ul]
[li]You could get the weight of the solar panels, divvy it up and see if it's less than the 20 psf roof occupancy load. In my experience, they're pretty light. Then assume that any place there's a solar panel is a place that can't have a 20 psf load.[/li]
[li]How about a load test? If you get a series of waterbeds and carefully fill them with water 6 inches deep (30 psf). Observe and measure deflections. If it deflects or fails, well, now you know. I'm assuming that you can test one sample section and extrapolate.[/li]
[li]Tell them no solar panels. They saved all this money 50 years ago, I bet that was fun.[/li]
[/ul]
We place great trust in calculations, which we should. But as driftLimiter points out, when you start piling assumption upon assumption (materials, thicknesses, etc.), the risk goes up. Add in normal construction mistakes, changes and just plain laziness and maybe calculations are not the way to approach this.

 

[SteelPE]

You could get the weight of the solar panels, divvy it up and see if it's less than the 20 psf roof occupancy load. In my experience, they're pretty light. Then assume that any place there's a solar panel is a place that can't have a 20 psf load.

Good idea, however, if the project is in snow country then this method likely will not work as (snow load) > (roof live load).

or fails

I'm not sure the client would like this outcome.

The only other method I know of revolves around the 5% limit for gravity members in the IEBC.... but this limit only equates to a few pounds per square foot and likely won't cover the weight of your new array.

Good luck and keep everyone in the loop as to how the project goes.

 

[Once20036]

Not applicable to the Butler space grid, but that latest ASCE says that you don't need to apply roof live load and solar loading concurrently.
You could check roof live load vs solar+snow load.
In low snow areas this could be an easy path forward. If that's not the case for you, I agree with the two previous responses.

 

[OCellini]

Thanks everyone for your comments. Unfortunately, as several of you pointed out, this project is in snow country (NY). It is difficult to minimize the high point loads that snow adds to panels, especially when the spacing of the supports do not line-up nicely with the joist spacing (they rarely do). I have been able to successfully use the 5% limit in the IEBC, as well as the elimination of roof live loads where there are panels on past solar projects- however, this typically involves solar systems that are attached, with no added ballasts.

@JedClampett: a load test is an interesting concept, but being able to successfully execute it with joists that span 30', is quite a challenge, especially at this time of year with snow in the forecast.

At this point, I will require that the solar support placements do not exceed the joist spacing, use attachments at areas that would require high ballasts, and then compare it to the original loading. With a low (1-story) building, not in a high wind zone, it may work. I'll keep you all posted.