Slow Decay, Part 1

I’m not talking about buildings or any other physical object. I’m talking about paper-based human-created constructs: building codes. (I’ll be adding more on other paper constructs tomorrow.)

I’m doing a little research and study right now, so I had the opportunity to look at the codes we use. There’s the New York City Building Code, the International Building Code that is the basis of but not identical to the NYC code, the steel code, the concrete code, the masonry code, the wood code, and the loading standard. There are plenty of others that we don’t use on a regular basis, such as the light-gage steel code. Of course, there’s no such thing as “the building code” or “the concrete code” because all of these codes are updated more or less regularly. We have shelves devoted to codes of various ages.

There are two justifications for the agencies and non-profits that create the codes to update them: we discover flaws or gaps in the codes as time goes on, and research provides us with better information. An example of new research is that the formulas that represent the capacity of steel columns have changed numerous times since the steel code was created in the 1920s, as researchers try to fit a mathematical model to the observed empirical behavior. A clear example of a gap being addressed is found by comparing the egress provisions of the 1901 and 1916 versions of the NYC code, and seeing the effect of the Triangle fire. So the changes are not frivolous, although every designer I know tends to lag a little behind.

As I’ve said before, when codes change every existing building is in danger of “not meeting code,” which is why codes have grandfathering provisions for extant buildings. But there’s another effect, not about the buildings themselves so much as about how engineers and architects work. We have shortcuts and design aids, tools that we develop ourselves or find in a book that help us take the general provisions of the codes and turn them into specific designs for individual buildings.

I can only speak for myself, but a lot of the engineering portion of my mental furniture is design and analysis methods rather than the codes themselves; since the methods are based on the code provisions, I’ve memorized material one remove from the source. The codes are easily referenced, while the shortcuts (especially those I’ve developed myself) tend to be less well documented. When a code changes, all of that one-remove material has to be updated, which takes some time.

If I design using the 8th edition of the AISC Steel Manual – the version that I learned in school some thirty years ago, and which honestly I remember in the most detail – most of my work is will be okay under the current (14th) edition of the manual. But some work will be wrong. If I use the 13th edition, newly everything will be okay; if I use the 5th, most of the work will be wrong. “Wrong” isn’t the end of the world: most of the wrong designs from the 5th edition will be overly conservative, which is unfortunate but not dangerous. But there are some areas (to be technical for a moment, column effective lengths are a good example) where the old code is unconservative and newer research has improved safety.

The old codes are gradually decaying, becoming less accurate and in some areas less safe as time goes on. They are immensely useful to help us understand the design of old buildings, but any new structure we design as part of a renovation or repair uses the current code. As the codes decay, the design aids and other paraphernalia erected on top of them decays as well.