Using existing fall arrest restraints for seismic restraint of small equipment

[skeletron]

I have a client that is wanting to install a 225lb heat pump above their unit. Metal deck, insulation, sheathing, membrane is the existing roof assembly.

I have been attempting to use a non-penetrating mount so that I don't have to deal with a phone call in a couple years because there is building envelope failure from 4 posts running down and attaching to the channels in their unit. There are a bunch of Unistrut style frames on the market that I want to attach to a wood framed platform (6x6 along the edges to accept screws from the metal frame above). The whole thing is to land on a big rubber mat to provide some friction resistance. With the added weight of the platform, I can't see this thing overturning or sliding under a realistic design load.

I am now realizing that the code won't allow me to use friction on its own to resist seismic loads. The platform will be almost adjacent to an existing fall arrest anchor. Can I attach a galvanized cable from the fall arrest anchor to the platform to provide that non-friction restraint? I need a sanity check.

Argument 1: Fall arrest can only be used as fall arrest alone.
Argument 2: It's a small unit with a nominal cable for seismic resistance. Fall arrest unlikely to be used during a seismic event. Maintenance schedule could be instituted so that the restraint cable is unhooked before and after window washing via the fall arrest anchor. Is this unreasonable?
Argument 3: The Code is the Code and the unit needs some sort of penetration to structure for seismic restraint.

With Argument 3, I think I would just go back to a steel system supported from the channels above the unit. It would require 4 penetrations through the roof, but ultimately provide the resistance required.

I'd be interested to know how others are handling this in their seismically active/inactive areas.

 

[driftLimiter]

I am currently trying to use a very similar approach for some roof mounted electrical equipment.

The telecom industry has long used these non-penetrating mounts from what I found.
The result was that the reviewer (AHJ) is requiring we test the friction capacity of the whole thing to verify that the shear resistance is sufficient.

For us it was worth pressing the AHJ on this option rather than penetrating a pretty much brand new roof.

I have seen many roof top units supported laterally by guy wires similar to what your proposing.
I think your arguments 1 and 2 are pretty strong for using them as a support.

If the code is requiring us to needlessly penetrate roofing systems, then the code should change.

 

[phamENG]

Not in a seismically active area, but I'm not sure I agree with argument 2. While you may be able to say that an earthquake occurring at the same time somebody using the fall arrest system is unlikely, I'd say that if those two coincide it would be a time when both could well be needed. What better than the building suddenly shaking beneath your feet to tip you off the roof?

I'm also very leery of specifying operational procedures to make my design work. Except in the best of industrial environments with detailed and scrupulously followed maintenance routines, they simply will not happen.

I'd go for the testing that driftLimiter suggested or find something else to tie off to.

 

[skeletron]

@DriftLimiter:
How are you approaching the testing phase? My seismic load is ~200lbs. I could see a winch test as a measured way to test that load. But practically speaking, the pushing force of a couple workers may hit that mark too.

 

[human909]

I think you have nailed it yourself with Argument 3. While I don't know your code and I logically agree with the practicality of your suggestion, I don't think it is readily justifiable.

No way would I be getting fall arrest users to unhook the restraint cable. That is just asking for somebody to call you out on things.

If you like to live close to the edge, then I'd run with your original idea as friction will likely keep the item in place for its service life and nobody will be the wiser. Seismic code be damned, this is just a heat pump. If there is a big enough seismic event to dislodge it then it would likely be the least of everybody's worries. (NOTE: I'm not advocating this attitude.)

 

[driftLimiter]

The test method is based on an ASTM test. The SEOC PV 1 document requires this in the context of roof ballasted solar arrays. So I called out the ASTM ref and the SEOC document.

My understanding is they can do it in a lab of the materials being tested are similar enough to the actual field conditions. Wet and dry measurements are taken.

I do agree with above points as well reg the guy wire approach. But as far as the likelihood of occurrence at the same time I think it's diddly squat. The load combinations in the code deal with this very principle. Statically it's a very low probability that the design earthquake happens at exactly the same moment some puts the design fall arrest on that particular anchor point. Not impossible but exceedingly unlikely.

If I used the existing fall arrests to anchor a unit. I would call it a permanent support no longer to be used for fall protection. The implications of that may or may not be prohibitive.