CraigB, I think you misunderstand me. First of all, this is not a stress issue. It's purely an allowable load issue. I'm not suggesting at all that the idle pump line be allowed to overstress and "hand wave" it away. As you well know, it tends to be difficult to approach allowable stresses for the piping system when you are limited by the pump nozzle allowables.

My point is that pump allowables are not based on the stress capabilities of the material of the pump, but on limiting distortion of the pump case and deflection of the pump shaft. Long term exposure to excessive load on a running pump leads to premature seal failure.

Nor am I saying that you just ignore the allowables when looking at pump loads for the pump idle scenarios. Quite frankly, it's a rare thing that I run into the need to consider going over for an idle pump. More often I find that the idle pump ends up with lower loads than when it is running.

My point is that, particularly when you have three or more pumps, and you been tearing your hair out for days working on a routing that satisfies all the running pumps - and you are still a bit over on a component or two for a pump in an idle case - that it's time to stop. Too much flexibility creates issue of it's own.

Regarding MoverZ - I think your example is a bit of a red herring. Particularly with a decent sized API pump, you can still have quite a bit of load on the idle pump nozzle at the allowable limits. As such, you are still likely to need a come-along to get the pipe back into alignment if you pull the pump away and it's no longer is place to hold the load. Unless you want to set a criteria that the loads on the idle pump are minimized while the other pumps are running (and make all us stress guys go bald in the process ) so that they can be easily taken and out and reinstalled, I don't see that changing.