Site Power
06Aug22
Monitoring the site power
the plots
plot
description
programs used for
monitoring
Site power
documentation
history:
141009: switched to new 13kV line ao -> 129-> ao
170607: installed new main breaker in substation.
all measurements in time order:
180421: another slow transfer from generator
to commercial.
180419: transfer switch after island wide
blackout.
170608: power outage. Voltage
Phase->gnd to 13Kv.
Monitoring the site power
A programmable breaker (ABB eVD4) was
installed in the power substation in mar13. This
breakersmeasures the power, voltage, currents .. at the input
to the substation. The information is queried and written to
disc once a minute. At 7:10 am every morning plots are made of
the power usage from the last 7 days. Once a month plots are
made for the entire month:
The plots:
Power substation plots
for the last 7 days (.ps) (.pdf):
Plot description:
The plots are made once a day (at 7:10 am).
The data is sampled once a minute. The frames in
each plot are:
- Frame 1: power (Kilowatts)
- white: apparent power (Kvolts*Iamps) at input
- red: active power. This is the power used to do work.
- The plot is -Active power. The convention is an
inductive load is negative, a capacitive load is
positive
- blue: minus the Reactive power
- Frame 2: Power factor:
- Active Power/Apparent power. It is negative since the
load is inductive.
- Frame 3: Current (amps)
- white: phaseA, red phaseB, blue phaseC
- solid line (what breaker calls instantaneous power)
- This looks like it has a threshold for measurement.
When the current changes by a certain amount a new value
is recorded.
- dashed line: 1 minute demand current provided by
breaker. Not sure whether this is a max or average.
- Frame 4: Phase to Phase voltage (Kilovolts)
- white phaseAtoB, red phaseBtoC, blue phaseCtoA
- Frame 5: Phase to ground Voltage (Kilovolts)
- white: phaseA, red phaseB, blue phaseC.
- Notes on the measurements
- The values reported by the browser screen
differ from these values by 1 or 2 %.
- The browser shows two types of values:
_INST_ and _DB_. I think the _DB_ values have been
corrected.
- The analog input configuration section shows
some correction factors (.1 to 1.2%) for some
measurements (errors in a/d ?).
- I tried applying these corrections to the
_INST_ data and got values that were closer but not
exactly the same as those reported on the measurement
screen of the browser.
- In the end, the values shown in the plots have
not had any corrections made to them. they are the
_INST_ values from the breaker.
Programs used for the
monitoring:
- /share/megs/phil/svn/aosoft/src/sitepwrlog.c
- Program to read the breaker and output the data
- Writes to /share/phildat/sitepwr/sitepwr_yyyymmdd.dat
- This program runs on aserv12. It gets restarted in /etc/rc.local
- /share/megs/phil/svn/aosoft/src/sitepwrlog &
- Making the plots:
- /share/megs/phil/x101/sitepwr:
- sitepwrdaily.sc script run via phil's
crontab on megs3 to make the daily plots
- sitepwrdaiily.pro idl code to make the plots
Site power Documentation:
Power issues:
180421: Another slow transfer from
generator to commercial.
On 21apr18 there was a short power dip around
5am and the site switched to generator (the rented one). The
commercial power came back with a minute or 2. The transfer switch
put the site back on commercial power 2.7 hours later .
The plots show the
reading from the site main breaker for 21apr18 AM (.ps) (.pdf)
- Top: Phase to Phase voltage at input to breaker vs hour.
- black,red, green colors are the 3 phases.
- The dashed purple line is when the breaker closed and the
site moved back to commercial power.
- The dashed blue line is 1/2 hour before the switch. This is
the normal delay for the transfer switch to move from
generator back to commercial. The commercial power needs to
remain stable for this 1/2 hour.
- Middle: Current through the breaker.
- Each color is a different phase
- The purple line shows when the breaker closed and
current started flowing through the breaker.
- Bottom: Power through breaker vs time
- black is KVA, red is active (resistive) power and green is
reactive power.
- prior the transfer switch closing the breaker, there was
around 60KV of reactive power recorded by the breaker (even
when no current was flowing through the switch?). This is the
same seen on 19apr18
Summary:
- Commercial power went off at 5 am on 21apr18 and returned 1 or
2 minutes later.
- AO switched back to commercial power (automatically) at 7:41am
21apr18. a bout 2.7 hours later
- For the transfer back to commercial power, the input voltage
at the breaker needs to be stable for 1/2 hour. The monitor
program reads the breaker values once a minute. It is possible
that there were a number of short glitches that the monitor
program missed (but i'm skeptical).
- If you look 1/2 hour before the transfer generator ->
commercial, phaseA of the phToPh voltages had dropped down to
around 14.3KV (and then stayed below this value)
- I wonder if the transfer switch requires the phaseTophase
voltages to be less than 14.3 KV before the 30 minute timer
starts?
- Without a load, the commercial power at the input to the
breaker tends to run above this value.
processing: x101/180421/pwrswitch.pro
180419: Site transfer switch after island
wide blackout.
On 18apr18 the power to the entire island went
down around 10:20 am. The FEMA generator automatically switched in
and ran the site.
The power in the local neighborhood returned around 12:30 am
19apr18. The AO transfer switch returned the site to commercial
power around 8:22 19apr18.
The local neighborhood and the ao power come
from different lines. But i still thought it was a bit strange
that AO waited 8 hours before it got commercial power back. I
plotted up the data from the site main breaker to see what had
occurred.
The plot shows the reading from the site main
breaker for 19apr18 AM (.ps) (.pdf)
- Top: Phase to Phase voltage at input to breaker vs hour.
- black,red, green colors are the 3 phases.
- The dashed blue line is when the input voltage to the
breaker returned for the final time (before the transfer)
- Middle: Current through the breaker.
- Each color is a different phase
- The blue line shows when the voltage returned, the purple
line is when the breaker closed and current started flowing
through the breaker.'
- The delay from Voltage return to breaker closing was 30
minutes. This is the normal delay time for the transfer
switch.
- Bottom: Power through breaker vs time
- black is KVA, red is active (resistive) power and green is
reactive power.
- prior the transfer switch closing the breaker, there was
around 60KV of reactive power recorded by the breaker (even
when no current was flowing through the switch?).
Summary:
- Commercial power went off at 10:20 am on 18apr18 and returned
intermittently around 1:30 am 19apr18.
- AO switched back to commercial power (automatically) at 8:22
am 19apr18.
- The delay from voltage at switch input to transfer was 30
minutes.
- During the early morning there were stretches of greater
than 2 hours when there was voltage at the breaker input, but
the site did not transfer to commercial power.
- The phase to phase voltages were up to 14.7KV during these 2
hours when no switch occurred. It must be that there is a
lower and an upper limit before the transfer switch will move
the site to commercial power.
- I don't quite understand why there is reactive power when no
current is flowing?
processing: x101/180419/pwrswitch.pro
170608: Power outage. PhToGnd voltage->
13KV. (top)
The main breaker in the substation was
replaced on 07jun16. On 08jun16 the following occurred.
- 08jun17:
- 15:33 I restarted the monitoring of the breaker
- 15:34 power outage..
- 09jun17: 22:00 back to commercial power.
The plots show the recorded data for
this period (.ps) (.pdf):
- Data is sampled once a minute.
- The voltage data is sampled at the input to the breaker.
- Page 1: 08jun17 15:30 to 24:00
- Top frame: Phase to Phase voltages.
- Black Phase A to PhaseB voltage
- red Phase B to PhaseC voltage
- green: phase C to Phase A voltage
- 2nd Frame: Phase to Ground voltages. (we must have a Y
connection).
- 3rd Frame: Current vs hour of day,. A blowup at the power
outage.
- We only had 1 complete sample before things started to
change.
- After 2 minutes, the current had gone to 0.. The site must
have switched to the generators.
- Note. this gets done at 480Volts, after the
13kV has gone through the 13kV to 480V transformers.
- The breaker did not open during this period (all of the
voltage protection in the breaker is currently disabled).
- Bottom frame:
- A blowup of the phase to ground voltages at the power
outage.
- Page 2: 09jun17 00:00 to 12:00
- The site was on the generators the entire time.
- The substation breaker and the transformers 13kV ->
xx were still energized, but they supposedly did not have any
load.
- Top: phase to phase voltages
- bottom: phase To ground voltages.
- 9:30 enoc opened the breaker (manually).
- The phaseA to Ground voltage went from 12KV to 0 volts.
- The other 2 phases remained unchanged.
Issues:
- Phase A voltage to ground did the following:
- 7.8 kV
- Load starts to drop:
- PhtoGndV -> 13KV for 7 minutes
- the load was 0 after about 2 minutes
- PhToGndV 13Kv -> 3.5Kv.. stayed there for about 15
minutes,
- PhToGndV 3.5Kv -> 12KV. stayed there till 09:30 the
next day
- 09:30 main breaker manually opened. PhToGndV 12Kv
-> 0V
- The breaker is a 13KV breaker. This is for the phase to phase
voltages.
- We were putting 13KV between phase A and the common connection
of the Y.
- I wonder if the breaker is setup to handle this voltage?
- When the breaker was opened, Phase A voltage to ground went
from 12KV to 0v.
- The 12KV must have been coming from the AO side of the
Breaker.. since opening the breaker, the input voltage went to
0.
- could the 430 recloser have been closed, and the 430
transformer (13kV -> 4160) was still connected to the
outside line?
- Do we have some other problem with the components on the AO
side of the breaker for Phase A?
- It would probably be a good idea for us to get the
substation programmed to open when we have voltage differences
like this.
- It could then protect things on the AO side of the breaker.
- When doing this, we need to keep in mind:
- Would every power outage cause the breaker to open? Will
it reclose automatically?
- What would happen during a lightning strike?
- If the breaker does open, then our transfer switch won't
be able to switch back to public power until the breaker is
closed.
Processing: x101/170608/spwr.pro
<-
page up
home_~phil