Insp-struc:
For starters, you’ve already suggested a 10% increase (50 to 55lbs./sq.ft.) in the roof loading, and you should absolutely prove that that’s o.k. Furthermore, your new rooftop equipment does not add 5lbs./sq.ft. to the whole roof. It adds a fairly sizable load to a 20 or 30 sq.ft. area and then is initially carried by only a few members, which should be checked out. I’ve always understood the 5% load increase allowance to give you some permission to add a little load, after the fact, using some common sense, without having to redesign the whole structure. Our design loads and methods are generally not any more accurate than +/- 5%, we apply factors of safety to everything we do, we spec. min. strengths for all the bldg. materials, which are usually exceeded, and/so you can usually find 5% fat in most designs. We also know that loads, load paths and max. stresses can adjust and redistribute to some extent, but not if there is only one load path. This all has to be considered in the overall picture/context. Then it is up to the engineer to use some common sense, good judgement and experience in determining exactly where to draw the line. This should be interpreted with the same sort of logic (I guess, sort of the inverse logic) which allows us to start reducing some of the live load at some point, when it seems reasonable that every sq.ft., of a large enough contributing area, will not be fully loaded. I interpret this to mean that I can have a 5% overstress in most cases, members, connections, etc. and not have to strengthen these elements, even though that is not exactly the way it is stated. The code thought process was…, a 5% uniform load increase will cause a 5% stress increase in a simple beam or a column, period. It certainly doesn’t mean I can take 5% of the total cap’y. of a given element, or 5% of the max. allowable load on the element (a beam or a bar joist, for example) and apply it indiscriminately, as a point load, and not cause any problems. At the very least, you should be checking that this new load does not introduce any new local problems, excessive stresses, deflections, buckling, etc. in its application. Then, follow this out to the reaction points of the member and see that it has only changed things by a percentage point or two. The effect of the new load quickly peters out, and is less likely to be an issue, as you move further away from the point/area of loading. To stretch your example, only a little bit, you would/might take 5% of the weight of the fully loaded bldg. and apply that to a beam or bar joist, and think that met the letter of the bldg. code?