With 2012 over, it’s fun to look back on performance statistics to see what stands out. For me, it’s SFO and its lagging on time performance that once again sticks out like a sore thumb. I’ve long felt that it wouldn’t be fair to restrict traffic at SFO just because things go downhill on bad weather days, but I’m reconsidering my position. Assuming Virgin America stays in business, delays are only going to increase and a regulatory solution may the only short term option that will work. But I’m getting ahead of myself. Let’s start with the problem at hand.
For those who aren’t aware, San Francisco has a runway problem that can make life miserable for travelers when the weather gets bad. Let’s look at a visual to make some sense of this.
As you can see, SFO has two sets of parallel runways which are perpendicular to each other. That in itself isn’t a terrible design for places where the winds can shift since you generally want to go into the wind when you can. The standard operation at SFO has most departures using runways 1L and 1R. Arrivals along with some long haul and Hawai’i departures usually use runways 28L and 28R.
This works great when the weather is good with up to 60 arrivals per hour. The problem is when the weather isn’t good. Those parallel runways are separated by a mere 750 feet. Nobody would ever design an airport that way today because when the weather goes downhill, you lose the ability to land on both runways at the same time. What defines bad weather? It’s poor visibility that matters. If you’ve landed at SFO, you know that you can get pretty close to an airplane next to you in good weather. When you lose that visibility, it kills the arrival rate.
This might not be a huge issue for some places, but SFO is extremely fog-prone, to put it mildly. So any time low clouds roll in, the airport gets snarled. Does it happen a lot? Oh yeah.
According to Doug Yakel at SFO, the airport can operate with its full 60 arrival-per-hour capacity about two-thirds of the time. Of course a lot of those hours may be in the middle of the night when demand is less, so it may not be a perfect way to measure how often capacity constraints really cause problems. When the weather gets bad, arrival rates are traditionally cut in half or even more if a runway shift is needed. Thanks to something called PRM/SOIA, SFO can sometimes land around 35 airplanes per hour in poor visibility. Starting next summer, that rate will be even easier to achieve without PRM/SOIA thanks to FAA procedure changes, but it won’t change that fact that operations are severely curtailed in bad weather compared to the regular 60-per-hour rate.
Where does this leave us? While people usually associate the New York airports with being the most delayed, SFO is right up there and quite often worse. For the 12 months ending November 2012, the Department of Transportation (DOT) shows SFO with about 71 percent of arrivals occurring within 14 minutes of schedule (the DOT definition of on time). Of the 29 largest airports in the US, only Newark (70 percent) and LaGuardia (77 percent) came even close. Nobody else was below 80 percent. It’s the same story with cancellations. At 2.25 percent of flights canceled, only Newark (3.07), LaGuardia (3.36), and Washington/National (2.43) had higher numbers.
But these numbers only tell a small piece of the tale. The bigger issue is how this impacts small communities that rely on SFO service.
The Real Pain is Felt in Small Cities
When the weather gets bad, airlines have to decide how to slow down their operations. In general, the flights that impact the fewest, least important people are delayed the most or even canceled if needed. That means regional flights to smaller cities get hit hardest. (Yes, I’m mostly talking about the big hub carrier here – United.)
This chart uses data from masFlight and shows the number of flights arriving within 14 minutes of schedule at SFO from cities in the general region for all of 2012.
As you can see, there are a couple of cities that do ok but then there are plenty that get hit hard. For that reason, people in this area know to build in pretty hefty connecting times if they want to go anywhere, but on time performance is only a part of the problem.
This chart shows the percent of flights that were canceled going from these cities into SFO in 2012.
It’s bad, really bad. Monterey sees more than 10 percent of all flights canceled. At least the people there have the ability to drive to San Jose or SFO without too much trouble. But good luck doing that from Crescent City or many of these other ones that see more than 5 percent of flights canceled.
I would say that I can only imagine how difficult this is for people living in those areas, but I’ve dealt with it first hand. We had plenty of calls at Cranky Concierge from people stranded in the area over the holidays. We even found out that gate agents at one airport were suggesting people call us because they couldn’t find options!
So look, the problem is a big one. How do we fix it? Well there are plenty of ways. If we were building a brand new airport, we’d smack the person who designed this one and instead ask them to do this:
With this kind of spread between runways, SFO could handle the normal number of operations whether in clouds or sun. But to do that now is a very expensive proposition. And there are plenty of environmentalists in the Bay Area who would have a fit.
So what else can we do? Well, as usual, technology should be the solution. The reason that those airplanes can’t land at the same rate in clouds as they can in sun is because current technology isn’t accurate enough to guarantee that they’re actually far enough apart. I say “current” technology, but I really mean the technology that is being used in air traffic control today. There’s nothing current about that.
The FAA finally got Congress to approve funding for the so-called NextGen air traffic control system which, if done right, should have far better accuracy. There’s no reason that airplanes shouldn’t be able to land the same in poor weather as they do in clear weather. We just need better technology to make that happen.
Will we get it? Eventually, yeah. But it’s going to be years and years before that happens. So what do we do in the meantime?
I hate to say it, but I think it might be time to start doling out slots. The FAA has already started moving in that direction by designating SFO as a Level 2 airport, but that is one step below actually restricting access.
There are real problems with slots, of course. First, if airlines have to cancel flights, it’s likely those smaller nearby cities that lose out most. But at least the flights that remain will actually go. Also, it’s hard to know the exact number that should be handed out, because arrival rates can swing so much depending upon the weather. But we need to find a way to cancel fewer flights. That probably means having fewer flights to begin with.
Travelers may love the idea of more on time flights, but there is a price to be paid as well. Basic economics explains that if we restrict the number of flights, prices are going to rise to get supply and demand back into an equilibrium.
So what would you rather have? Lower fares or more on-time flights? I’ve always leaned more toward lower fares, but I’m now leaning the other way more and more. Something has to give.