Statistics

on

Frequency Departure from 60 Hz in

The Eastern Interconnection

Priority-based Control Engineering (PCE)Tel: 614 799 0300 --- email: pce@pce-co.com

PCE

Priority-based ControlEngineering

PCE-2

Information Presented in the Bar-Charts

• The pictures displayed in section 1 show the number of times that frequency in the Eastern Interconnection in an average 365-day period in the studied interval remained higher or lower than 60 Hz by the indicated threshold for more than the number of minutes shown on the horizontal axis. See slides 4 and 5 for examples of reading the bar-charts.

• The pictures displayed in section 2 are created using the same method as those in section 1, except the fact that they only consider events where frequency was less than 60 Hz.

• Section 3 describes the data we used to create the bar-charts. Double-click on the charts to view the Excel file used to create that chart.

PCE-3

Section 1(Frequency is Higher or Lower than 60 Hz)

PCE-4

Example 1.

The height of the multi-colored bar at minute 0 indicates that the frequency was above or below 60 Hz for at least one 6-second sample by:

• 30 mHz in about 55,000 occasions, turquoise bar in slide 11.

• 50 mHz in about 3,000 occasions, yellow bar in slide 12.

• 55 mHz in about 1,450 occasions, purple bar in slide 12.

• 60 mHz in about 650 occasions, blue bar in slide 12.

Example 1.5 .

The height of the multi-colored bar at minute 1 indicates that thefrequency was above or below 60 Hz for at least 10 consecutive6-second samples, i.e. 1 minute, by:

• 30 mHz on about 7000 occasions, turquoise bar in slide 11 Onceevery 3 hours

• 50 mHz on about 325 occasions, yellow bar in slide 12 Onceevery 3 days

• 55 mHz on about 140 occasions, purple bar in slide 12 Once aweek

• 60 mHz on about 55 occasions, blue bar in slide 13 Twice amonth

• 65 mHz on about 20 (?) occasions Once a month.

• 70 mHz on about 7 (?) occasions Once a season

• 75 mHz on about 2 (?) occasions Once a year

Example 1.5a.

• 30* mHz on about 3600 occasions, turquoise bar in slide 16 Once every 6 hours

• 50 mHz on about 200 occasions, yellow bar in slide 17 Once every 4 ½ days

• 55 mHz on about 83 occasions, purple bar in slide 17 3 times a month

• 60 mHz on about 40 occasions, blue bar in slide 17 3 times every 2 months

• 65 mHz on about 15 (?) occasions Once every 2 months

• 70 mHz on about 6 (?) occasions Once a season

• 75 mHz on about 2 (?) occasions Once a year

The last graph in this PCE PowerPoint demonstrates that an imperceptible tiny fraction of these occasionsconsists of events approaching 15 minutes in length, let alone 30 minutes in length.

The height of the multi-colored bar at minute 1 indicates that the frequency was above 60 Hzfor at least 10 consecutive 6-second samples, i.e. 1 minute, by

* When time-error correction (-20 mHz offset) is invoked, the underfrequency FTL becomes -30mHz from scheduled frequency and the number of FTL-exceedences (occasions) is morefrequent than the average # of exceedences for a given FTL setting. When time errorcorrection is not being invoked, the number of exceedences is less frequent than the average# of exceedences. Time-error correction is invoked every day, for at least 4 hours a day andpractically always to correct fast (over-frequency) time-error according to NERC's Joe Emde,without regular seasonal variation over the years. Total over- & under-frequency error doesnot capture the immediate effect of time-error correction on the number of errors because theamount of underfrequency error introduced far exceeds the amount of overfrequency errorreduced. (The Draft Standard calls for setting the FTL relative to scheduled frequency ratherthan to 60 Hz so that time-error correction is not counteracted by a BAAL mechanism that, ifset relative to 60 Hz, treats the time-error correction the same as an unscheduled error!)

Example 2a.

The height of the multi-colored bar at minute 5 indicates that thefrequency was below 60 Hz for at least 50 consecutive6-second samples, i.e. 5 minutes, by:

• 30 mHz on about 250 occasions, turquoise bar in slide 17 Onceevery 5 days

• 50 mHz on about 12 occasions, yellow bar in slide 18 Onceevery 2 ½ months

• 55 mHz on about 5 occasions, purple bar in slide 18 Once aseason

• 60 mHz on about 2 occasions, blue bar in slide 18 Once a year

PCE-8

Example 2.

The height of the multi-colored bar at minute 5 indicates that the frequency was above or below 60 Hz for at least 50 consecutive 6-second samples, i.e. 5 minutes, by:

• 30 mHz in about 460 occasions, turquoise bar in slide 12.

• 50 mHz in about 20 occasions, yellow bar in slide 13.

• 55 mHz in about 8 occasions, purple bar in slide 13.

• 60 mHz in about 3 occasions, blue bar in slide 13.

Example1.5b.

The height of the multi-colored bar at minute 1 indicates that thefrequency was above 60 Hz for at least 10 consecutive6-second samples, i.e. 1 minute, by:

• 30 mHz on about 3400 occasions Once every 5 hours

• 50 mHz on about 125 occasions Once a week

• 55 mHz on about 37 occasions Once a month

• 60 mHz on about 15 occasions Once every 2 months

• 65 mHz on about 5 (?) occasions Once a season

• 70 mHz on about 1 (?) occasions Once a year

• 75 mHz on about 0 (?) occasions Once every few years

The last graph in this PCE PowerPoint demonstrates that an imperceptible tiny fraction of these occasionsconsists of events approaching 15 minutes in length, let alone 30 minutes in length.

Example 2b.

The height of the multi-colored bar at minute 5 indicates that thefrequency was above 60 Hz for at least 50 consecutive6-second samples, i.e. 5 minutes, by:

• 30 mHz on about 210 occasions Once every 5 days

• 50 mHz on about 8 occasions Once every 4 months

• 55 mHz on about 3 occasions Once a year

• 60 mHz on about 1 occasion Once every 2 ½ years

PCE-11

PCE-12

PCE-13

PCE-14

Loga

rithm

ic S

cale

PCE-15

Section 2(Frequency is Lower than 60 Hz)

PCE-16

Negative Events Only

PCE-17

Negative Events Only

PCE-18

Negative Events Only

PCE-19

Negative Events OnlyLo

garit

hmic

Sca

le

PCE-20

Section 3

PCE-21

Frequency Data

• Frequency source was NB 6-second frequency data, provided by Elmer Bourque, covering 2002, 2003, and more than 10 months of 2004.

• Excludes 5/11/2002, 5/12/2002, 5/13/2002, 8/14/2003, and 8/15/2003, and a few short periods. In these periods– data was not reliable

– data failed a reasonability check

– operation was not typical (due to situations such as islanding)

• Values are normalized to find the expected values for a 365-day year.