BenA

RealHJ

Yes, you can. Data centers, that consume 10s if not 100s of times (depending on size) more power than an airport, do it all the time. It's not cheap. But that capital expense is part of doing business and the commonly accepted norm.

There is a much cheaper alternative option that could have been used if the airport is not deemed critical infrastructure (as is clearly the case in the eyes of the ATL operator and DL). Simply have two diverse power feeds powering two halves of each terminal. Then the worst that can happen, unless the power grid as a whole is down, is that 1/2 of each terminal is dark, while other 1/2 still functions. That way at least the situation could be managed (ground stop to stop new flights from coming in, but still get the pax already in the airport out safely within a several hour vs. day long delay; orderly evacuate and clear out the dark 1/2 of each terminal upon a power outage).

Only critical infrastructure like CBP should have power that can be switched, automatically or manually, between the two sources, in this very cost-efficient scenario that would cost very little more than all single source how it has been built now.

ethernal

I'd invest your time in worrying about other things that are far more controllable. Like heart disease or the dozens of other things that are above terrorism on the list (admittedly this chart is a bit misleading as it excludes domestic terrorism.. but that only ups it to about 1:30000).

CrazyEddie

Do you think people who think this way are unable to worry about heart disease?

ethernal

It's a matter of proportionality. If they spend about 10 minutes a day thinking about heart disease (that's generous) then they should spend about .14 seconds a day thinking about how to survive a terrorist attack.

Assuming it took some of the above posters two minutes to type up the posts, they've used up about 2-3 years of worry time.

edit - last I'll post on this as it's diverged a bit form the topic at hand (sorry about that)

CrazyEddie

Maybe the study doesn't factor in where people spend their time. And if it is based on data going back to 1975, maybe it's flawed or at least could be updated.

RatherBeOnATrain

WABE-FM in Atlanta is interviewing a pax who was stranded by the power outage. Her ATL departure was fully loaded and ready to go. Then the power went out and they were stranded at the gate because the jetway could not be moved back from the plane. (Doh!) After an hour of waiting, they were told to disembark.

Being unable to move the jetways probably explains why it took so long to unload planes!

jamiestr

Wow... Usually there are a bunch of armchair pilots here, but it's good to know there are also some armchair electricians! Even with the little information provided in the press releases, it's clear that Atlanta had A & B power, but that the fire was in or near the switching equipment.

I was stuck at the airport last night, and I will say Delta's response was impressive. Almost everyone I talked to was working double or triple overtime, including a lot of the headquarters employees. Can they improve, yes. But even with a power outage that was out of their control on a Sunday evening, I was pretty impressed.

3Cforme

Actually, no. People due to land at ATL are going to dozens of destinations never, ever served out of MEM or MCO. MSP and DTW each had twice MEM's flight count. DL could, if it wanted to, land aircraft where flights, facilities, and employees are available, by diverting to DTW and MSP. Yet they don't (not en masse). ATL serves about 80 more destinations than DTW or MSP. It can serve a lot more dest8nations because it has size. The idea of a ready-made relief hub lives on in the imagination of FTers and a.netters but the accountants kill it -- it is the textbook example of inflated overhead and poor asset utilization.

RealHJ

The point where A&B should converge is right at the equipment they are powering - until that point being diverse (e.g. coming from two opposite ends of the terminals), at the very last bit and not before that. Here it is clearly not a proper and normal A&B set, as the path diversity is missing - which renders it useless, if such is even in place.

No, DL should insist that the ATL airport operator designs, builds and run its facility as is commonly accepted and at the same level of power diversity and emergency preparedness as the competition (other airport operators) do.

Running generators for critical systems (airport ramp, lights, doors; exclude commercial merchants, SkyClubs, AC) is part of the bare minimum expectation in running an airport, and how other airports are run, as you would know if you read just some posts earlier in the thread.

Yes. Here DL/ATL operator have certainly not gotten the balance right. There is a commonly accepted way to do this. They have chosen the cheap way out by intermingling everything, that renders it useless (but probably it has been very useful to the engineer and contractor pockets who designed & built it). I mean even just having two diverse power paths each powering 1/2 of each terminal (or something along such lines), would have cost very little more, if anything - may be even less, over how it is now, but have provided at least some operational capacity.

Flaflyer

(bolding mine)

There is the problem. Even with two power sources, there is one transfer switch, and if it fails, power is cut. From what I gather the transfer switch was what failed in Atlanta. The power source (the electric company power plant) was always available, just not connected. However with proven Southern innovation, a couple of jumper cables and some NASCAR duct tape got the power reconnected.

If you want to claim "there should always be two of everything even transfer switches" then you better quit flying. Aircraft have multi layer windows, multiple computers, and multiple hydraulic systems for backup. At great expense, they even carry a spare pilot! However there are many things where there is only ONE item, and if it fails, you crash. For example, every current Delta plane has ONE wing. None carries a spare. Even if they did, the Delta mechanics say it is a real bytch switching them inflight. If the wing was strong enough to never break, it would be too heavy to fly. So it has a design safety margin, but it is breakable. FEAR the Single Wing. There is NO backup.

CalVol

I was asking the question that, ISTM, needs to be asked to actually make the comparison. We can't really determine the trade-off until we know how much is at risk. Best I can tell from the stories I've read indicates roughly half the schedule was cancelled. What will the total economic impact of this event be? Do those costs come anywhere near the costs associated with the kind of redundancies that would have prevented it?

Frankly, I don't know the answer to the question. My guess is there are a few people in FT world with some experience in either/or/both the costs associated with such backup systems and the negative economic impact associated with major business disruptions.

One thing SEEMS quite certain: they were totally unprepared for how to react.

Zorak

^ agree with the above, the only quibble being I would insert for completeness (maybe I am simply spelling out what you already intended) "Do these costs, multiplied by the probability/frequency of occurrence, come anywhere near the costs associated with the kind of redundancies that would have prevented it?"

(EDIT: and yes, I understand that it's not just a strict EV calculation and that in some circumstances it can be worth sacrificing EV for variance reduction and so on)

mysterym

I never said anything about sending the national guard in. I am merely pointing out that not having controls and mitigation in place for a power outage in such a location does put both life and property at risk. That is a fact. I am also stating my opinion that I would not want to be in a dark crowded space with many other people around. I like to be "in the know" and control of my destiny.

RealHJ

Yes. But the solution is simple, and very cheap. Just use a Manual Transfer Switch. It is purely mechanical (think a lever you pull down), so it has very low chance of failure. And certainly not all many 100s of them are going to fail at once. Downside is then the duty electricians would need to go and switch the MTSes to switch the load over. But, the upside is much lower capital expenditure upfront, and lower chance of failure. For something that is expected to happen very rarely if ever and where a downtime of some minutes to an hour or so is tolerable, that is the best set-up, and by far the most cost effective also. Only if you are switching frequently do you need ATS (or STS if you don't want power to even flicker and continue on). If the airport is deemed not critical or important infrastructure, just take the cheap option of MTSes between the A & B feeds, in front of each load or bunch of loads (e.g. in each gate area).

And again why not just power each end of the terminal from a different power source (A & B), that, in turn, come from a different transformer (ideally one transformer per terminal half, vs. one for the entire 1/2 of airport), and that, in turn, come from two different substations. That would provide no real redundancy, but at worst even if there is a major fire or the substation upstream blows up, still 1/2 of the airport would be functional and operational.