Another Power Glitch

I've been eager for a proper power failure so I can capture data on the PriUPS system for real, instead of showing test data.  Sort of like shooting down a real enemy instead of a drone or test missile.  The local power company came very close a couple of months ago.  In that case, power remained off long enough to almost discharge the UPS battery to the point of allowing the Prius to power the house.  But, as you can see in the graphs for that incident, AC came back on just a bit too soon.

Two nights ago, I had another genuine power failure.  I was at work when I heard the fateful "AC Power Failure" in my own voice and I immediately looked at the telemetry to see what was going in.  This one was a lot more scary!  Here's the record:

Telemetry of a brief power failure at 19:47 EDST on Monday, 17 July 2006.  (Click the graph for a full-resolution version.)

Note that about 43 minutes into the graph, at 19:47 local time, the yellow line, representing the charge of the UPS batteries, drops significantly, which indicates that the UPS was providing power.  Although it's hard to see, the reason is that the blue line, the AC power input, had dropped out for a couple of 10-second intervals, and the house was without utility power.  I would normally wait several minutes before positioning the Prius for connection to the UPS because power drop-outs are typically brief, as this one turned out to be.  This 20-second power loss was no challenge to the UPS at all.

However, look at the blue line!  Right after the drop out, it was showing an AC input voltage to the house of about 265 volts.  This is 'way out of tolerance, and potentially damaging to electrical appliances.  Looking further along the blue line, you can see that the AC input voltage was, in effect, oscillating around its 235V nominal value.  In one case it reached 250V, again out of tolerance, although probably not dangerous for the minute or so it was at this level.  I have never seen AC power line voltage swings like this.  The timeline indicates that this was not a brief "surge" due to lightning, which my equipment probably wouldn't even measure due to its short duration.  Rather, this was an unusual and serious instability of the AC power system.

I thought it would be valuable to document this event.  Although it has nothing to do with the PriUPS per se, it points up the desirability of having a UPS for the house.  All the equipment connected to the UPS would be completely unaffected by this surge, while appliances connected to the AC line might well have been damaged.  Fortunately, nothing seems to have been destroyed this time.

I plan to refer the power company to this blogitem.  I'm going to get in touch with them to inquire as to just what happened here, and will post a followup if I manage to obtain an explanation.

I called the power company and they had no record of the 265V surge, although they were aware of the dropout.  I've been promised that they will look into it.  More info as it becomes available.