When all the radar screens went dark and all the radios went silent for 45 minutes, Northern California airspace became uncontrolled and nobody knew what to do about it.
August 24, 1995
FAA tiger teams are still trying to sort out exactly what happened
in the wake of an unprecedented, 45-minute power outage that shut
down all radar and radio communications at Oakland Center on Wednesday
morning, August 9th.
All ARTCCs are equipped with three independent sources of power
for critical equipment. Each includes an emergency diesel generator,
plus a bank of batteries capable of operating the center for 30
minutes. Consequently, a complete power failure was considered
by FAA planners to be virtually impossible. But it happened anyway.
How It Happened
Although the outage is still under investigation, the probable
cause is now pretty clear. One of the three power sources was
down for testing and maintenance at the time of the episode. The
second power source failed unexpectedly. When technicians attempted
to bring the third power source on-line, a faulty circuit board
in Critical Power Panel failed, preventing power from being restored.
The power-failure was widely misreported in the popular press
as being related to the FAA's chronic problems with computers
and related ATC equipment. A spokesman for NATCA, the controller's
union, wasted no time in blaming the problem on obsolete equipment
and reduced staffing at the Center.
Although Oakland Center is only the 16th busiest ARTCC in the
country, it's responsible for the largest block of airspace of
any ATC facility 18 million square miles extending from San
Luis Obispo to the Oregon border, and including most Pacific oceanic
The failure occurred, at 7:13 a.m. local time during the morning
"departure push" period. Controllers estimate that 60
to 80 aircraft were under ZOA control when the power died. All
radar screens went dark and all radio communications were cut
off. Lights and telephones were unaffected. It took 45 minutes
to restore radio and the backup DARC (direct access radar channel)
radar system and more than an hour before the NAS computerized
radar was restored.
What Do We Do Now, Boss?
Controllers were stunned as their screens went dark, their headsets
silent, and the perpetual whirr of equipment cooling fans ceased.
Once controller described it as "the loudest silence I'd
Another controller told AVweb that there was "panic
on everyone's face" as the controllers realized that they
had been rendered deaf, dumb, and blind.
"We don't train for a situation like this," the controller
It took a minute or two for controllers to realize that the failure
was facility-wide. Most turned to their supervisors for guidance,
and the supes turned to their supes. But there's no book procedure
for handling a situation like this. So in the end, most controllers
wound up improvising.
Controllers at adjoining Los Angeles, Salt Lake City, and Seattle
ARTCCs and at various TRACONs were asked to take over surveillance
and control of all airspace within their radar coverages, and
to divert aircraft under their control bound for destinations
in Northern California. Control towers at San Francisco, Oakland,
San Jose, Sacramento, and other airports in the area were instructed
to hold all IFR departures on the ground.
The most difficult problem was getting notification to airborne
flight crews. In one case, said controller Mike Seko, "we
had the Napa Airport tower telling planes that Oakland Center
had lost everything, and to switch to emergency frequency."
But most aircraft who were airborne on Oakland Center frequencies
were left in a state of lost-comm unless they figured out what
was happening and switched to another ARTCC or TRACON facility.
Meantime, In the Cockpit...
Flight crews apparently did a fair amount of improvising of their
own. Some pilots squawked VFR and proceeded on that basis. Many
others continued to follow their last clearance, and some climbed
into or descended out of Class A airspace without a clearance.
A lot of ASRS forms were undoubtedly filed that day. All indications
are that the FAA is not pursuing any actions against the pilots
involved in this mess.
In retrospect, one of the biggest problems was that nobody believed
that a prolonged outage like this was possible. Controllers and
supervisors alike operated on the assumption that their radios
and radar would come back "any moment now." The same
was true of controllers at adjacent facilities like Bay TRACON
that were paralyzed by the Center's blackout.
It's impossible to say how many losses of separation and TCAS
resolution advisories occurred during the hour-long episode. Some
near-midair reports will undoubtedly be filed, but the vast majority
of loss-of-separation situations will probably go unreported.
After power was restored and the primary radar system returned
to normal operation, extensive air traffic delays, diversions,
and flight cancellations persisted for many hours at Bay Area
airports, particularly airline departures from San Francisco International.
Every controller involved had a war story to tell: Controller
Jonathan Smith told AVweb, "My shift started moments
after power was restored. I will never forget the atmosphere upon
my arrival. Talking to my fellow controllers was much like talking
to accident victims in various states of shock. All were hyper-active,
some were exhibiting bravado or gallows humor, others repeating
incidents and experiences. All were immensely proud of their having
survived a controller's worst nightmare."
It will be months before we know the full aftermath of this incident.
Clearly, changes will be made in the way critical power feeds
are wired and maintenance is done. Contingency plans will be rewritten
and controllers will receive training in how to execute them.
The FAA may lobby Congress for funding to provide redundant links
to radio and radar sites so that when one ARTCC has a catastrophic
failure, adjacent facilities can take control of the airspace.
Meantime, you might want to brush up on your non-radar and no-radio