Lightning & Its Effects on the Florida Power & Light - - PowerPoint PPT Presentation

Lightning & Its Effects on the Florida Power & Light Distribution System & Other Related Issues

LARRY VOGT

Lightning – Its Effects on the Florida Power & Light Distribution System and related issues – Larry Vogt

Note: Actual interruption data from 2007

2

On average, there are more than 300,000 lightning strikes on the FPL service territory every year. Lines in areas with high ‘Isokeraunic’ levels (#

• f strikes/sq. km) could

experience 2-3 direct strikes per mile

Flashes per sq. km

Lightning is a major contributor to customer interruptions and equipment failures on the FPL Distribution System.

3 YTD - Customers Interrupted by Major Categories - All Types

374,387 43,728 192,922 260,953 264,158 110,276 149,325 173,550 546,161 169,731 166,050

• 300,000
• 200,000
• 100,000

100,000 200,000 300,000 400,000 500,000 600,000 700,000

EQUIP Weather Accidents Unknown VEG OTHER Wire Cable Imp Process Animals Request

YTD 6/06 YTD 6/07 GAP

Weather, including lightning, is the second highest cause of customer interruptions (CC 01 – Lightning) Note: Many of the equipment related interruptions may also have been initiated by lightning

4

Current Situation

Monthly Correlation - Lightning Strikes vs Storm Interruptions (Codes 01 & 02)

500 1000 1500 2000 2500 3000 3500 4000 4500 5000 20000 40000 60000 80000 100000 120000 140000

Lightning Strikes

Storm Interruptions

An obvious correlation exists between lightning strikes and customer service interruptions

Lightning – Typical lightning can deliver

currents ranging from 10,000 - 200,000

• amps. Lightning currents also induce

voltages exceeding one million volts. When lightning hits a distribution line, a high-energy traveling wave is created on either side of the strike point, as the lightning charge travels to the nearest

• ground. FPL’s distribution system has

insulation levels between 95 kV and 150 kV BIL. While a surge arrester can clamp voltages at 40 to 50 kV, a flashover can still

• ccur before the wave reaches the

arrester.

5

Modified Vertical Framing

Surge Arresters: FPL has over 1.5 million arresters installed on its distribution system, and still experiences lightning flash-overs on all typical distribution framing, Better Framing for overhead lines: The FPL distribution system contains approximately 30% cross-arm construction; 25% triangular; 25% Modified vertical; 15% Vertical; and, 5% Vertical with overhead ground wire. Modified vertical framing has been the standard for over 30 years.

 After a lightning storm, crews typically find

• bvious lightning related damage … shattered

insulators, blown fuses, ruptured transformers,

• etc. Lightning, however, causes many

problems that don’t show up right away. Compromised transformer insulation, cracked insulators, pitted contacts, damaged blocks on arresters, etc…..these could show up as equipment failures, much later on.

6

7

Analysis - Operation and Limitations of Lightning Arresters

• Modern Lightning Arresters are designed to have a high

resistance at normal line voltage, essentially acting like an insulator.

• Under surge conditions, as the voltage rises on the

distribution line, the resistance of the arrester drops. This allows surge current to be diverted through the arrester to ground.

LS Energy # of LS

Not to scale. For demonstration purposes only.

Most Lightning Strikes

• Arresters are designed to dissipate a

limited amount of energy.

• If an arrester is subject to energy levels

higher than its rating, the internal metal

• xide blocks short, causing the

disconnector and ground terminal to separate from the body of the arrester.

Arrester energy limitation To Ground To line

Disconnector Direct strokes in this area

Estimated Population on the FPL system – 1,545,000 Porcelain – 390,000 Polymer – 1,155,000 Expected life: 20-25 Years Usage (2007): 83,000 Failures: 7000 annually Porcelain – 4000 Polymer - 3000

FEEDER INTERRUPTIONS (with exclusions) ARRESTER - 12MOE

10 20 30 40 50 60 70 Sep-05 Dec-05 Mar-06 Jun-06 Sep-06 Dec-06 Mar-07 Jun-07 #

• f

I N T E R R U P T I O N S

8 There are approximately 50 feeder interruptions related to arresters, annually

• Lightning - While arresters can handle some strokes effectively, their energy dissipation capacity is sometimes

exceeded.

• Moisture Intrusion - In older porcelain arresters, the top seal often fails with age, allowing a great deal of moisture
• inside. Polymer arresters also have moisture intrusion.

Arresters typically fail for one of two reasons:

ARRE STE R FAILURE MODE S (12 MOE)

causing FE

E DE R INTE RRUPTIONS (with exclusions)

20 13 11 6 40% 66% 88% 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 Framing - Gnd lead Too close Gnd/prim. lead into phase Tracked Bracket

Failure Modes # of FDR INTERRUPTIONS

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

N=50

9

.

The manner in which the arrester was originally installed, accounts for 66% of the feeder interruptions. (1st two bars on chart)

• Ground lead too close - The ground

lead is formed too close to arrester base (An oversized or stiff lead makes matters worse)

• Ground or Primary lead too long -

The ground lead was left too long from the last staple on the pole.

• Tracked bracket –

This is the result of degradation of the glass-filled polyester bracket.

1

Analysis - Scenarios where a Surge Arrester may be identified as the cause of a Feeder Interruption

Scenario 1: Ground lead does not allow arrester ground terminal to separate

• 1. High-energy lightning strike hits line

and exceeds arrester energy ratings, causing disconnector operation.

• 2. Ground lead is too stiff at the bottom
• f the arrester. Often the case in 3 Ph

triangular installations, where one ground lead is used for all arresters. The non-failed arresters keeps the ground lead in place.

1 1

Analysis - Scenarios where a Surge Arrester may be identified as the cause of a Feeder Interruption

Scenario 2: Arrester ground lead is too long and makes contact with primary conductor

1. Ground lead was too long from the last staple on the pole (Note: Long leads also reduce the degree of protection). 2. High-energy lightning strike hits line and exceeds arrester energy rating, causing the ground terminal stud to separate from body of arrester. 3. Ground lead makes contact with another phase (phase- to-ground fault).

1 2

Analysis - Scenarios where a Surge Arrester may be identified as the cause of a Feeder Interruption

Scenario 3: Tracking on Mounting Brackets

1. Ground lead is attached according to Construction Standards. 2. High-energy lightning strike hits line and exceeds arrester energy rating, causing the ground terminal to separate from the arrester body. 3. Arrester is not replaced or removed promptly and remains attached to the primary. 4. With the ground lead removed, full line voltage is now present across the bracket. 5. A combination of time and contamination leads to tracking across the bracket. When the leakage current increases to a significant level over time, an interruption occurs (phase-to- ground fault).

13

1 4

Analysis - Scenarios where a Surge Arrester may be identified as the cause of a Feeder Interruption

Scenario 4: Porcelain Arrester shatters; primary Line lead makes contact with another phase

1. Line lead is installed correctly, but may have been too long. 2. Water leakage shorts blocks, and gas builds up, causing the arrester to shatter. All distribution arresters installed prior to 1987 were porcelain. Because of seal problems, some tend to get water inside. If this happens, they may fail violently, causing the primary lead to contact another phase or ground. We are addressing this problem by removing all porcelain arresters after they have been de-energized. 3. When the arrester shatters, the line lead propels into the air and makes contact with another primary line or ground, causing either a phase-to-phase or phase-to-ground fault.

LA

1 5

Countermeasures- Surge Arresters

1 6

• Modified vertical framing (our present

standard) flashes over 90 % of the time for direct strikes to the line. Note: Direct strikes range from 10% to 20% of all strikes depending on line location. i.e. Rural areas have highest exposure, while buildings and trees provide more shielding in urban areas.

• There is an average of 5 return strokes per

strike, with continuing current flowing between return strokes. The energy created from the continuing current is a major factor in arrester failure.

Analysis – Framing & Lightning study

Average 5 return strokes per strike. Arrester energy exceeds 84 KJ Continuing currents flow between each return stroke and damage

• arresters. These currents

are low frequency and average a few hundred amps.

To better understand the capability of our distribution line framing, we joined forces with University of Florida and their Triggered Lightning Experiments at Camp Blanding. Our conclusions follow:

 With direct strikes, arresters are called to dissipate above 84 kilo-Joules of energy,

yet, they are designed to handle only 20-30 kJoules. A distribution class arrester will fail more than 50% of the time if its rated energy dissipation capacity is exceeded .

 80-90% of lightning strikes are indirect or nearby strikes. The arresters see far less

energy, and consequently almost always survive. Hence, Distribution class arresters provide adequate protection for indirect strikes.

 Overhead ground wire framing provides the best protection for direct strikes,

and combined with arresters (for indirect strikes), gives the best overall protection for our lines.

1 7

At Camp Blanding we also found that:

1 8

Causes of lightning related interruptions



Feeders, with standard framing (modified vertical), flash over 90 % of the time, when lighting strikes the line directly. Other framing standards without OHGW have similar performance



There were no specific standards addressing arrester installation, which took the failure mode of MOV blocks into consideration.



Flash-over may occur even if arresters do not fail.

1 9

Countermeasures - Framing

5

1 8 Arrester Stations eight spans apart

Line Vertical (OHGW) & Arresters Vertical (OHGW) Modified Vertical 0.225 (88.9 %) 0.692 (65.9 %) 0.258 (87.3 %) 2.012 1.963 1.963 2.028 2.028

Total expected # of Interruptions

***

Total Expected #

• f Flashes

**

** Expected Number of Flashes per mile per Year *** Total Number of Expected Interruptions per mile per Year (Lightning Performance Improvement Over the Modified Vertical Framing)

VERTICAL FRAMING WITH OVERHEAD GROUND WIRE & ARRESTERS (Note: this study was done on wool poles, and does not take into account the concrete poles used in the hardening effort)

Avian Line (larger insulators and cover

• ver terminals)

2

Implementation

6



Calstrom Feeder 5961 Toledo Blade. Two miles of line behind recloser



Adams Feeder 8461 Treasure Coast. Four miles of line behind recloser



Peacock Feeder 11662 Treasure Coast. Two miles of line behind recloser



Sabal Feeder 8765 Treasure Coast. Two miles of line. No recloser.

2 1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Performance

7



Sabal Feeder 8765 Treasure Coast. Two miles of line. No recloser.



Special lightning activity polygon created for this feeder section



Lightning activity within polygon monitored using GLADS



Feeder interruptions monitored using SCADA. Interruptions correlated to lightning activity within polygon

• 1. Date/Time: Mon Jul 2 17:43:41 EDT 2007
• 2. Date/Time: Mon Jul 2 18:17:28 EDT 2007
• 3. Date/Time: Sun Jul 15 15:29:19 EDT 2007
• 4. Date/Time: Mon Jul 2 17:42:49 EDT 2007
• 5. Date/Time: Sun Jul 1 16:25:39 EDT 2007

6.Date/Time: Mon Jul 2 17:47:08 EDT 2007

• 7. Date/Time: Mon Jul 2 17:48:26 EDT 2007
• 8. Date/Time: Sun Jul 15 15:39:38 EDT 2007
• 9. Date/Time: Mon Jul 2 18:30:07 EDT 2007
• 10. Date/Time: Mon Jul 2 18:09:06 EDT 2007
• 11. Date/Time: Tue Jul 3 19:56:03 EDT 2007
• 12. Date/Time: Wed Jul 4 15:37:40 EDT 2007
• 13. Date/Time: Sun Jul 15 15:48:53 EDT 2007
• 14. Date/Time: Sun Jul 15 15:48:53 EDT 2007
• 15. Date/Time: Sun Jul 15 15:27:55 EDT 2007
• 16. Date/Time: Sun Jul 15 15:41:38 EDT 2007

Lightning Activity July 1 – 15 2007

No interruptions reported

• Obviously there is a substantial cost to reframing existing lines.

Because of the multiple hurricanes in 2004 through 2006, however, FPL funded an extensive Hardening effort. Since this involved replacing many older wood poles, these framing changes were incorporated into that program for a minimal extra cost.

2 2



Best protection for direct strikes is achieved with OHGW



Best protection for nearby strikes is achieved with arresters, but they need to be installed correctly



Combine OHGW and arresters for over all best protection



FPL now uses Vertical framing w/OHGW and arresters on non-urban construction, where exposure to direct strikes is high, and 23 KV construction is required.

23