Karl,

Thanks a lot. I have passed this information to Grant Masson
who worked here on a related problem.
He will get in touch with you regarding this technology.

Vince





Karl Tomlinson@ENRON_DEVELOPMENT
07/24/2000 06:00 PM
To: Vince J Kaminski@ECT
cc:  
Subject: Outage Tracker Option And Background


Vince,

To follow up on the idea of a means for effectively tracking instantanious 
plant faliure utilising either system frequency or connection point voltage.

The system frequency phase shift across a network will probably be the best 
option as this would allow alll significant deviations to be tracked along 
with relative network performance from a few points.

I am currently chasing NEMMCO (AU system operator) to get hold of a few weeks 
of 4 second metering data for the whole system to see if there is enough 
measurement consistency to prove one of the ideas.

The idea follows along the lines (literally !) that when a unit fails it will 
introduce a shock into the system and reduce system frequency, which is then 
reacted to by frequency control services offered in by generators. The drop 
in frequency is noticable across the whole network and as one option may 
change the phase shift across the whole grid. The phase shift across the 
network is constantly changing due to loads and power factor correctrion 
devices switching on and off, however a unit failure may be distinct.

The second option relates to how a unit fails, whereby if a circuit breaker 
is involved as in the case of uncontrolled shutdowns, then the outage will 
cause an RF pulse that should propogate across some of the network. 
Transformers will attenuate the pulse, however it should be detectable many 
miles away from the fault location. Measure the arrival time at several point 
on the network, work out the shift and backtrack on the network. Part of the 
solution is already proven in lightning trackers. The solution may be made 
more simple by detecting the exact point on the sinewave that the unit failed 
(i.e. measuring the three phases at 1MHz and then sending the data through a 
DSP). This solution is more complex, however this should allow unit failure 
to be pinpointed to a station. The timing base for each of the nodes may be 
sourced from GPS timing, self locating at the same time !

Will send you some links and documents for potentail hardware and setups.


Karl.