SFIA hired Charles River Associates and John F. Brown Company to study delays - some two years after stating they needed new runways to solve the delay problem. The final report, Reducing weather-related delays and cancellations at San Francisco International Airport, April 2000, is an interesting read. Does it conclude new runways are required? Of course it does. Who paid for it?

But, if you chose to read the numbers rather than the text, you will find some interesting facts. The graph below comes from Appendix C.3, page 5. (Here's the PDF of the page.) There are some problems with the data: One, there are about 600 arrivals per day at SFIA. Only about two thirds were analyzed. The international flights are not included and, according to John Costas, Chief of Staff at SFIA, are rarely delayed. The smaller airlines and General Aviation are not covered. I assume General Aviation is never late since it doesn't fly on a schedule.

Two, the report assumes a plane one minute late is delayed. The FAA standard of 15 minutes is far more reasonable.

The report made the calculations to produce the data by assuming that, on a good weather day, the delays were due to reasons other than weather at SFIA. They called that the "baseline delay." About 150 planes a day are late due to reasons out of SFIA's control. These reasons include mechanical problems, crew availability, flight scheduling, en route weather, weather at other airports in the flight's itinerary, and air traffic congestion.

Then, they calculated the delays for bad weather days at SFIA and subtracted out the baseline delays. That result shows that only 35 planes a day were delayed due to weather. And it's only those planes that would be aided by new runways ... or new technology. That's less than 10% of the analyzed flights and less than 6% of the total flights. SFIA wants to say averages don't matter and they need to design for the worst case. Aren't you glad they aren't designing our freeway system? Just imagine, a Bay Bridge with 50 lanes. (Remember though, it is an average.)

By the way, departures aren't really a problem. Planes can take off in worse weather than they can land in parallel. That's why the report only covers arrivals.

From: Reducing weather-related delays and cancellations at San Francisco International Airport

Charles River Associates • John F. Brown Company

April 2000

Appendix C.3, page 5

One of the primary reasons for the expansion into the Bay, according to SFIA, is the "horrendous delays". Consider this. Delays and airports go together like congestion and freeways. All major airports in the U.S. suffer delays regardless of runway configuration. There are different ways to measure delays. The FAA likes to measure percent of flights delayed more than 15 minutes. This graph shows the top airports in the US by that measure. Note that SFIA is not in the top ten.

Another way to measure delay is by percent of flights arriving on-time. FAA Bay TRACON(the air traffic controllers) personnel told me that the reason for 4300 feet of separation between runways was that LAX's runway separation is 4300 feet and thus it was known that it would work. LAX's percent of on-time arrivals is 76% according to the Chronicle. SFIA has an on-time arrival rate of 71%. Separating runways doesn't fix the delay problem. I might add that, in a response to my question asking how many US airports actually had 4300 feet between runways, John Martin, Director of SFIA, said, "All of them."

Do we really want to destroy an irreplaceable resource on the outside chance of reducing delays by a few percent?

Air Traffic Control Solutions may minimize or solve the problem.
See the table listing different technologies.
As an example, Simultaneous Offset Instrument Approach(SOIA) is currently being installed at SFIA. It is expected to be operational in 2001
SOIA would increase the Arrival Acceptance Rate to 45 planes per hour according to Airport Consultants(P&D Aviation). See the chart for an indication of the impact this would have at SFIA