Analog Angle Article

Reinforcing the art of the estimate

A recent press release about a high-temperature-superconductor (HTS) power transmission line going into service boasted that it could carry 375 megawatts, enough to service up to 300,000 homes. The same week, a short item in the Wall Street Journal reported on a 900 MW line that would supply up to 900,000 homes.

I always wonder when I see numbers thrown around like popcorn, so I did a quick calculation. The first case works out to 1250 watts per household, or about 10 amps at 125 Vac; the second corresponds to 1 kW per household or 8 A each. Considering that most homes in the US run at least 100 A of ac service, all I can say is that these the customers in these homes know how to live small.or maybe that phrase “up to” is the key?

Perhaps the error is due to an overzealous PR person misapplying numbers, a simple calculation error, or power companies hoping to imply that their new lines are helping them an even-larger customer base. While the cause doesn't matter here, these examples show how casually numbers are used to add credibility, and how we as engineers should do a basic reality check when we see them.

Unfortunately, it's easier to be dazzled by numbers with implied precision than to stop, think, and use some rough estimates as a sanity check. We're so used to dealing with reasonably good numbers on credible data sheets that we forget that not all sources are quite so straightforward.

The Wall Street Journal has a regular column (“The Numbers Guy”) which looks at both the sources and extrapolations of some often-cited numbers. Among the more ridiculous are the annual “surveys” of how much driving people will do on vacations, or how much they will spend on Christmas-season gifts. These surveys are based on a very small sample, done by landline phone, of what people say they might do. Then they're extrapolated and promoted relentlessly to the media. It's hard to know what's worse: the polling method, or that it's people saying that they might do (and they all say they are going to diet and exercise, too), or extrapolation from such a small data set.

Before ubiquitous calculators and PCs, engineers would routinely do a rough estimate based on their feel for the situation, as a sanity check. That forced you to quickly check where your numbers were coming from, and where they were going to lead you. And it's still a good idea, before false precision leads to foolish conclusions.

-x x x-

0 comments on “Reinforcing the art of the estimate

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.