It was just last month that the Department of Transportation (DOT) said not to rush to judgment after seeing more cancellations and slightly fewer 3 hour delays in May. We needed more time, they said. Well apparently the DOT only thinks that’s the case if the numbers don’t help their cause. Now that June numbers are out, they’re jumping on them as proof of success. Slow down there, DOT. Your base level look at June numbers is way too superficial and smells simply of politics.

On the surface, things looked much better this year for long ground delays, and really, they were. There were three flights that sat on the runway for more than 3 hours during June 2010. In June 2009, that number was a whopping 278. This year, none sat for more than 4 hours. Last year, it was 42.

So did cancellations spike as a result? A little, but not much. Though the DOT would like you to think that both this year and last saw no change at 1.5 percent of flights canceled, that’s only thanks to rounding. In 2009, it was 1.48 percent. In 2010, it was 1.50 percent. So there was a very slight increase. In fact, three fewer flights were canceled this year but with 6,307 fewer flights operating.

Sounds like a tremendous success, right? Well, it’s good news for some, but it’s important to look at other factors year-over-year to see how we may have ended up in this place.

First, we have to remember that after last summer, the airlines did ramp up their efforts to reduce lengthy ground delays. So some of the reduction should be related to previous efforts and not simply the introduction of this rule. We’ll never know exactly what that is. But we can still dig in and see what happened last June. You’ll notice that the weather was significantly better, despite what many have said.

The Air Travel Consumer Report doesn’t give specifics on the flights that were delayed for three hours, but it does do it for the flights delayed more than 4 hours. So let’s focus on those. There were 42 of those, and most of them occurred during specific events. For example, 13 of them (11 at Dulles) happened in Washington on June 3, 2009. That’s the day that thunderstorms came and sat on top of Dulles from 8p to 10p.

So the weather was bad, but something had to be just as bad this year, right? Nope. In June, there were a couple days with quick passing storms, but the only day that saw sustained thunderstorms over a long period was June 28 when storms sat on Dulles from 145p to 3p, still a much shorter time period, especially when you’re looking at a 3 hour limit.

But the key is the time of day. I looked at flights scheduled on August 12, 2010 as a comparison. Not much should have changed between last year and this year. Between 145p and 3p, there are only 29 flights scheduled to arrive and 15 scheduled to depart. That’s an average of 23.2 arrivals per hour and 12 departures per hour during the time of the storm. Meanwhile, between 8p and 10p, 82 flights are scheduled to arrive and 32 scheduled to depart. Think about that, that’s 41 arrivals per hour or 16 departures. That’s a significant difference.

On top of that, the imbalance of arrivals to departures is huge at night. That’s because a lot of planes come in and stay the night before flying out in the morning. That means that the gates are full at night, so if planes don’t push back to depart on time, then arrivals have nowhere to go. That’s not usually the case during the day. And don’t underestimate the 2 hours of storms this year versus 1.25 hours last year. That’s a huge difference when 3 hours is your cutoff.

There were a couple other events we could look at in June 2009 as well, but I could paint a story for those too. I think the point is clear. Knowing that thunderstorms tend to be the biggest culprit for long ground delays during the summer, June was a luckier weather month in 2010 than it was in 2009.

That being said, even if we had the exact same weather this year, I would bet the performance would still have been better. Simply being more cognizant about the situation would have reduced the number of ground delays, but that’s going to have happened even if this rule didn’t exist.

The bottom line is that the DOT can’t have it both ways. If it wants to just use superficial results to advance its claim, go right ahead. But then don’t tell people who do more in-depth reviews that it’s too early to look at the results. This is a far more complicated issue than either the DOT or the pro-rule people want to admit.

Many of you have already heard that in its contract of carriage, Southwest has now decided that mechanical issues are outside the airline’s control. How do I know? Because I’ve received more email from readers on this issue than any other, I believe. It’s amazing how this has grabbed people’s attention. The reality of this, however, is not as dire as many are suggesting. Don’t get me wrong. I don’t like this move, but due to Southwest’s policies, this doesn’t change much.

Who cares if Southwest considers a mechanical problem under its control or not, right? You should, actually. Airlines make clear distinctions on how much they’ll help stranded customers depending upon whether it was due to circumstances within the airline’s control or not. Here’s a handy chart explaining what’s within airline control and what isn’t, traditionally.

It’s relatively straight forward. If it’s something that an airline can have control over, then it’s the airline’s responsibility. But what’s the difference for passengers? If something is within an airline’s control, then the airline will generally pay for hotels and meals while you’re waiting. The airline will also, in many cases, put you on another airline if available. If the event is outside an airline’s control, then you’re on your own. The airline will get you out when it has a seat available on its own flights, but that’s about it. You’re entitled to a full refund in both cases, assuming there’s a cancellation or excessive delay.

So why do I say that this isn’t as big of a deal here for Southwest? Southwest doesn’t put people on other airlines anyway. If you have a problem on Southwest, you’re waiting for the next seat on a Southwest flight or you’re taking your refund elsewhere. So it’s really just an issue of meals and hotels, not nearly as big of a deal but still important.

For all airlines, the contract of carriage is the binding document regarding air transportation, so this move in Southwest’s contract of carriage (PDF) is worrying, but Southwest also has its Customer Service Commitment (PDF) which outlines what it will do when things go wrong.

The Customer Service Commitment clearly states:

. . . if circumstances within our control, such as aircraft “swaps,” cause you to miss the last possible flight (or connection) of the day to your destination,
our Customer Service personnel have the authority to arrange for overnight lodging. We will find a hotel or motel as near to the airport as possible, and at no additional cost to you. We may also arrange for ground transportation to the overnight facility.

If the cause of your inconvenience is not within our means of control, we will do our best to assist you by securing a discounted rate at a hotel or motel at or near the airport.

With the underlying definition of “circumstances within our control” being clarified, it does make me wary. If a flight cancels because a plane breaks and there are no more flights that day, Southwest is now clearly not obligated to put you in a hotel for the night. Whether that holds up in practice or not remains to be seen, but it’s certainly a concern.

Southwest says that it simply clarified the definition of this in its contract of carriage but that it didn’t intend to change policies. That may be true, but from a legal perspective, that doesn’t appear to be the case.

[Updated on 7/28 @ 1234p: Southwest has listened to everyone and further clarified its contract of carriage to say mechanical difficulties from other entities. Smart move. Read the post on the Southwest blog at http://www.blogsouthwest.com/blog/southwest-airlines-addresses-misinterpretation-regarding-contract-carriage]

[Original photo via Flicker user swanksalot]

Back at the end of April, the rule enacting massive fines to airlines who keep people on an airplane on the ground for more than three hours went into effect. The Department of Transportation just released May’s results and the numbers show that more travelers were inconvenienced. Have you been reading articles about how ground delays are way down? Those aren’t looking at the whole picture. There were fewer ground delays than the previous year, but cancellations were up significantly.

On-time percentage in May 2010 was 79.9 percent across all reporting airlines. It was 80.5 percent in 2009, so it was comparable in that respect. Then we look at cancellations. In 2010, 1.24 percent of flights were canceled. Back in 2009? It was only 0.88 percent. Had the May 2009 rate held through in 2010, nearly 2,000 fewer flights would have been canceled in May of this year. If you assume an average of 100 people on a flight, you get almost 200,000 people who were inconvenienced this year that wouldn’t have been inconvenienced last year. Many have said that it was a small increase in cancellations. That seems pretty big to me.

Now, let’s look at the number of lengthy ground delays. In May 2009, there were 34 flights stuck on the ramp for more than 3 hours. This year, there were 5. If we stick with that 100 passenger per flight number, then we’re at 3,400 people who weren’t sitting on a plane on the ground for more than 3 hours this year. Remember, that compares to an increase of nearly 200,000 people who were inconvenienced by cancellations. Let me put this in a pretty picture:

Are we really making a fair comparison here? After all, there’s no way to directly attribute all the additional cancellations to this rule. There are differences in weather that could also cause large swings. But the increase in cancellations was spread out across airlines. Of the 18 reporting airlines, two-thirds reported an increase in the number of cancellations. Weather alone is not going to cause that to be spread across the country, though it can certainly count for some of it.

But that doesn’t really matter. There’s an even better way to look at this. If only 1.5 percent of those additional cancellations were due to the ground delay rule, then more people still would have been inconvenienced by higher cancellations than were saved from three hour delays. And that assumes that the reduction in long ground delays was entirely due to the new rule, something that’s highly unlikely.

But let’s not stop there. Lets look at the five long ground delays. Four of them were on United in Denver on May 26. That was the day that not only saw thunderstorms roll through the airport, but it also saw a tornado hit. Yeah, think the weather was bad? Maybe we can forgive four airplanes being stuck for that long.

The last one was Delta flight 2011 from Atlanta to Dallas/Ft Worth on May 28. Both airports saw thunderstorms that afternoon, but Atlanta had more than an inch and a half of rain. It was an awful day.

The Delta flight was stuck on the ground for 3 hours and 2 minutes. Yeah, that’s right. It went a mere two minutes over the limit. So I asked Delta what happened. Apparently, the airplane was in a “bad spot” on taxiway and there was lightning so the ramp workers couldn’t come out to meet the airplane. Had the airplane come back to the gate, it would have canceled and the passengers wouldn’t have gotten out of town for at least 10 to 12 hours.

That’s just one more example of when a 3 hour and 2 minute delay isn’t nearly as bad as the alternative. Of course, Kate Hanni, the director of FlyersRights.org thinks it’s a good idea to gloat. She has been quoted as saying “I hate to say I told you so, but I told you so.”

Despite Kate’s claim, there is nothing to gloat about. More people are being inconvenienced than before. So, Kate, I hate to say I told you so, but I told you so.

Tips to minimize the impact of a flight cancellation – The Detroit News
With the 3 hour ground delay rule likely causing an increase in cancellations, the Detroit News looked at what people should do. They came to Cranky Concierge for advice.

Virgin America’s First Quarter Loss Means More Work Ahead for Troubled Airline – BNET
The first quarter numbers are in, and once again, it’s a loss.

Subsidy reservations: Kansas taxpayers support airlines in Wichita, but KCI complains of turbulence – Kansas City Star
It’s an interesting situation in Wichita these days when it comes to airline subsidies, and the Kansas City Star dug in on the subject.

Facebook Fan Numbers Aren’t Nearly as Important as Airports May Think – BNET
There’s a lot of hype about Facebook and Twitter, but it shouldn’t be a race to collect fans. It’s all about how to deal with them.

Random Expansion: Virgin America Is Going to Regret Its New Flights to Mexico – BNET
My take on Virgin America’s move into Mexico. Wanna guess how I feel?

Open Skies Turns to Trains to Help Fill Airplanes – BNET
Open Skies has no feed on either end of its network, so it’s looking at trains. I like it.

Google Acquires ITA Software, and the Likes of Kayak and Bing Cry a Little – BNET
Personally, I think the acquisition of ITA by Google could be just awesome. But there are definitely going to be some unhappy campers.

One of the biggest complaints that I see when we start talking about long ground delays here on the blog is that nobody seems to have a good solution. We can all jump on bad solutions like the DOT rule we have now, but there are rarely better suggestions that are workable. I’ve found one, and it’s actually temporarily in place at JFK right now.

Some of you may know that I write the monthly newsletters for PASSUR Aerospace. PASSUR is a very cool little company that actually has its own private radar network at over 100 airports, primarily in the US and Canada. There are a million things that they do with this data, but as I was putting together the current newsletter, I found out about a temporary but truly awesome project going on at JFK during the runway construction that’s happening today. They’ve effectively created virtual queues so that airplanes don’t have to push off the gate until it’s closer to actual departure time.

Here’s how it works. During peak periods (operational for ten hours a day), a central command coordinates all departures and arrivals at the airport. A couple hours before scheduled departure time, each flight is assigned an actual departure time by the system. So maybe your flight was supposed to leave at 1p, but at 11a, the airline will be advised that the departure time will now be 2p. That means they can keep you at the gate and let you roam free. If they need that gate for an arriving flight, they can still board you and then push you out of the gate, but you’ll have a very clear picture of when you will actually take off. Expectations can be set appropriately, and they shouldn’t push you off the gate if a 3 hour delay is anticipated.

The result of this is that at any one time you end up with no more than eight airplanes waiting to take off (you need some kind of buffer in there) instead of 30 or more. This is good for passengers, but it’s also good for airlines. When they’re in that long line waiting for takeoff, they have to keep an engine running, sucking up gas as they wait. Now that won’t be an issue. In addition, if the winds shift and the airport needs to turn around and use different runways, there are fewer planes that need to be turned around so it can be done more quickly.

Awesome, right? Now this isn’t a fail-proof solution. When bad weather rolls in, that messes a lot of things up. A version of this system is actually used during severe weather events now at JFK for metering departures, but it doesn’t solve everything. If you have a ton of planes coming in on diversions or several planes landing while others can’t take off, you may need to free up gate space due to storms. Things can and will still go wrong, but this process is a big improvement.

So how has this been working? Well, during March 2010, they were able to handle just about the same amount of traffic as they handled in March 2009 but with one runway down for construction. Taxi-out times are virtually the same. And the average delay has actually gone down by about 5 minutes. Great stuff.

Unfortunately, this program is currently scheduled to end on June 30, when the big construction work on the runway is done. With any luck, the airport will see the benefit and decide to keep it around . . . and then hopefully other delay-prone airports will consider it as well.

[Original Photos via AZ DOT and abdallahh]