ML20198H202
| ML20198H202 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 05/21/1986 |
| From: | Cooney M PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC |
| To: | Ebneter S NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| Shared Package | |
| ML20198H207 | List: |
| References | |
| NUDOCS 8605300155 | |
| Download: ML20198H202 (11) | |
Text
.
PHILADELPHIA ELECTRIC COMPANY 23ol MARKET STREET P.O. BOX 8699 PHILADELPHIA. PA.191o1 (215) 841 5020
~"U.'..",*..".'.S* ". m.,
06..,
May 21, 1986 Docket Nos. 50-277 50-278 Inspection Report:
50-277/85-41 Mr. Stewart D.
Ebneter, Director Division of Reactor Safety l
U.S. Nuclear Regulatory Commission Region I 631 Park Avenue King of Prussia, PA 19406
SUBJECT:
Peach Bottom Atomic Power Station Fire In Radwaite Building Dear Mr. Ebneter Your letter dated January 3, 1986 forwarded the report for the special safety inspection (50-277/85-41) conducted on November 13-14, 1985 at Peach Bottom Atomic Power Station, concerning the fire which occurred in the Radwaste Building on November 10, 1985.
In response to your request for a copy of a final report on the cause of the fire, Event Report No. 7 is attached to this letter.
This report was prepared by our Nuclear Safety Section and represents our considered judgement as to the most probable cause of the fire.
Although we have concluded that the cause of the damage to the electrical cables was the ignition of combustible below the cable tray, we have not determined precisely the ignition source.
Our Security Division engaged a consultant who investigated the cause of the fire and concluded that it was electrical in nature.
However, Philadelphia Electric Company's Electrical Engineering Division cable specialists have concluded that the evidence does not support the consultant's conclusion.
Addendum 1 to Event Report No. 7 is the consultant's report, and Addendum 2 is Philadelphia Electric Company's Electrical Engineering Division's comments on the consultant's report.
fok b
f o\\
, e-
Mr. Stewart D. Ebnator May 21, 1986 Page 2 The fire in the Radwaste Building did not affect the operation of reactor safety systems; however, there were some control room alarms and electro-hydraulic control pump runbacks and recirculation pump runbacks due to the event.
The damega was restricted to the divers' equipment cage and a cable tray located above the cage.
If you have any questions or require additional information, do not hesitate to contact us.
Ver-ruly yours, j?'lVW)
Attachments cc:
T. P. Johnson, Resident Site Inspector i
PHILADELPHIA ELECTRIC COMPANY NUCLEAR SAFETY SECTION PEACH BOTTOM ATOMIC POWER STATION EVENT REPORT NO. 7 l
CABLE TRAY FIRE IN RADWASTE BUILDING ON NOVEMBER 10, 1985 i
1.0 DESCRIPTION
OF EVENT i
On November 10, 1985, Unit 2 was in operation at 100%
1 power, and Unit 3 was in a refueling outage.
At 1602 hours0.0185 days <br />0.445 hours <br />0.00265 weeks <br />6.09561e-4 months <br /> a Burns Guard on fire watch patrol called the control room and reported smoke on the 165 ft. level in the radwaste building.
A smoke detector alarm window lit for VENT AREA CSR/SWGR EQUIPMENT SMOKE DETECTORS A4 ELEVATION 165'.
The Fire and Damage Team was activated and told to report to the 165 ft. elevation in the radwaste building.
The Fire and Damage Team could not find the source on elevation 165 ft. but observed heavy smoke building up on the ceiling.
The smoke forced the Team out of the 165 ft. elevation.
At 1606 hours0.0186 days <br />0.446 hours <br />0.00266 weeks <br />6.11083e-4 months <br /> an operator in the control room noticed that instrument air header pressure was decreasing after "A and C air compressor trouble" alarms annunciated.
At 1607 hours0.0186 days <br />0.446 hours <br />0.00266 weeks <br />6.114635e-4 months <br />, the following alarms and events took place:
"EHC electrical malfunction" alarm, "A & C battery ground" 1
alarm, " static inverter trouble" alarm, recirculation pump i
runback to 60% and an EHC runback.
At 1611 hours0.0186 days <br />0.448 hours <br />0.00266 weeks <br />6.129855e-4 months <br /> a STA, who had gone to the' cable spreading room to investigate the EHC electrical malfunction alarm, reported that three of the DC power supply indicators for EHC indicated low voltage but voltage was recovering.
He also detected smoke in the cable spreading room and reported this to the Control room.
At 1617 hours0.0187 days <br />0.449 hours <br />0.00267 weeks <br />6.152685e-4 months <br />, the Fire and Damage Team reported to the control room that there was fire in a cable tray on elevation 150 in the radwaste building.
The fire was extremely hard to find due to all the smoke.
The team members heard a very loud roaring noise, like steam or air blowing, and saw molten material falling from the cable
[
tray down to a cage underneath the cable tray.
The cables e
on the underside of the cable tray directly above the cage were observed to be glowing.
They saw material on the i
floor in the cage on fire with yellow and orange flame, and i
the flames were a couple of feet high.
The fire in the cage did not seem to be concentrated in any one place; it i
was spread out over a couple of feet.
The fire in the cable tray was extinguished with carbon dioxide.
Some arcing of cables were observed during extinguishment.
The i
f I-
Page 2 of 9 fire in the cage was suppressed but not extinguished with
. carbon dioxide and dry chemical extinguishers.
A water hose line was deployed from an elevation below 150 up through the stair tower to the fire.
The fire was then rapidly extinguished with water.
The loud roaring noise was coming from a ruptured air line underneath the cable tray.
Air from this ruptured air line had fanned the fire and prevented the Fire and Damage Team from getting the e
fire extinguishing material up on to the cable tray.
During the fire fighting, one of the members of the Fire and Damage Team valved out the air line.
The ruptured line was a service air line, just a few inches below the cable tray.
The line split along its length; the split was approximately four inches long.
At 1626 hours0.0188 days <br />0.452 hours <br />0.00269 weeks <br />6.18693e-4 months <br />, the Fire and Damage Team reported the fire was out.
2.0 GENERAL CLASSIFICATION OF APPARENT CAUSE 1
2.1 Fire of unknown origin.
Five separate investigations t
to determine the origin of the fire were conducted.
d These investigations were done by 1) plant staff, 2) the PECo Fire Marshall, 3) a representative from the PECo Fire School, 4) a cable expert from Electrical Engineering Division, and 5) an outside consultant.
The conclusions from these five investigations do not point to a single cause of the fire.
2.1.1 Based on the type of cable used, the results of exposing this type of cable to various fire tests, the fact that the burned area on i
top of the cables in the tray was smaller than the area on the bottom, and the fact that there was only moderate soot accumulation on surfaces above the fire damaged cable tray, the fire most likely did not originate in the cable tray.
(See Section 6.9) 2.1.2 Based on a burn test of plastic bags similar to those used to contain material in the cage, the combustion characteristics of this bagged material, the observation that the flames on the floor were a couple of feet high, and a post fire observation that the burn pattern in the cage appeared to be low, there appears to have been insufficient combustible material to cause the flames to reach the cables 7-8 feet above the floor.
(See Section 6.5)
Page 3 of 9 1
2.1.3 The facts in the above two paragraphs tend to j
support the postulation that material on top of the cage burned; however, there was no evidence of debris on top of the cage.
Although there is no evidence that material was on top of the cage just prior to the fire, a photograph of the area, taken as part of the laser video project last summer, shows e
a board wrapped in plastic bags on top of the cage.
The board is shown in the photograph to be at the end of the cage opposite the end under the burned cable tray.
The plant J
security director recalls seeing this board above the cage several hours after the fire.
The laser video photographs weren't shot at an angle that permits observation of another board on top of the cage and under the cable tray.
(See Section 6.7) 2.1.4 The origin of the fire could have been in the cage if a rapidly-burning, high-BTU-content material had been present.
This material could have been an accelerant, i.e.,
a flamable liquid, or a material entirely consumed during the fire.- The outside consultant did not find any evidence that an accelerant was used.
A significant amount of BTU's had to be released next to the filter cylinders because the lead shielding around the filter cylinders melted, puddled and subsequently solidified and entrapped the debris left on the cage floor.
The photographs taken as part of the laser video project show an orange colored life e
preserver, wrapped in clear plastic, hanging on hooks on the wall at the back of the cage just below the cable tray.
These orange colored life preservers are used at other locations at the plant, e.g.,
at the outer intake screens.
These orange colored life preservers contain a plastic foam.
According to the manufacturer (Atlantic Pacific Co.)
the foam is polyvinylchloride, the cover is nylon, and the life preserver will burn if exposed to flame.
The presence of a life preserver on the back wall in the cage, shortly prior to or after the fire could not be confirmed; however, if the lite preserver were present and exposed to flame, it could i
have burned and caused flames to reach up to the cable tray in the area where most of the damage occurred.
(See Section 3.4).
2.1.5 The fire could have been purposely set or accidentally set through carelessness by
Paga 4 of 9
)
i someone smoking.
The area near the cage is i
not a heavily traveled area, and, although smoking is prohibited in the area, cigarette butts have been found in this area.
A match or cigarette butt tossed onto material on top of the cage, if it was present, or onto the material on the floor of the cage could have l
}
caused the fire.
If the material in the cage j
~
caught fire first, the flames had to j
initially reach high enough to ignite material on top of the cage, if it was present, or reach high enough to burn the life preserver if it was present.
Plant and corporate security personnel questioned people who were at the plant when the fire occurred but they were unable to determine, j
through questioning, if the fire was
[
purposely set.
i 3.0 IMMEDIATE CORRECTIVE ACTION k
i 3.1 Shift supervision began notifying senior plant i
management shortly after the event.
At 1815 hours0.021 days <br />0.504 hours <br />0.003 weeks <br />6.906075e-4 months <br />, senior plant management were on site and received a t
status report.
Unit 2 was at 75% power and the repair of the air header was in progress.
Investigation revealed that there were no ECCS or RPS cables involved in the cable tray that burned.
i 3.2 At approximately 1845 hours0.0214 days <br />0.513 hours <br />0.00305 weeks <br />7.020225e-4 months <br />, the Operations Engineer i
and others surveyed the burned cable tray.
At this time, the service air header had been cut, a new valve had been installed and service air was l
reestablished to the radwaste building.
With HP assistance, the Operations Engineer entered the cage i
and observed the following.
"The damage across the bottom of the cable tray was fairly uniform and all the way across the cable tray; portions of the side rails of the cable tray were melted; a cross bar of I
the cable tray was gone; insulation on many of the cables was damaged to the extent that the conductors l
were exposed; and at least one conductor was completely severed."
The Operations Engineer looked for but could not find one cable that seemed to be significantly worse than the others, and he could not
(
find the center of where the fire might have been.
3.3 Plant staff determined that a cable ground developed during the fire in the tray and caused a voltage dip in the uninterruptable power supply.
This dip in voltage deenergized relays in the EHC and recirculation pump runback logic which caused EHC and j
recirculation pump runbacks.
i I
l
Paga 5 of 9 3.4 On the evening of November 10, 1985, the PECo Fire Marshall, his assistant, and a PECo photographer arrived on site and observed the fire area.
The Fire Marshall interviewed several individuals involved in fighting the fire.
The Fire Marshall, his assistant, a representative from claims-Security, a PECo photographer, the Operations Engineer, HP personnel and several other plant staff members reexamined the fire area in an attempt to identify the cause of the fire.
The Fire Marshall reported the following.
"The fire occurred in the area of the cage directly
^
under a 90 degree bend in the cable tray.
The fire damage in the cable tray was confined to the 90 i
degree bend.
The contents within the cage included four metal filter cylinders, without filters, each surrounded with lead shielding and placed in seven plastic bags, two garden hoses placed separately in plastic bags, two plastic buckets containing tools, and one small pump with an electric motor.
None of this equipment was energized or plugged in, nor were there any chemicals in the area.
All materials in the cage were lifted except the filter cylinders.
The cylinders were contaminated.
The burned area in the cable tray was found to be smaller on top of the tray than on the bottom of the tray.
There was no evidence of any cable that was overheated, nor was there any evidence of a catastrophic fault.
- However, since some of the cables were still energized, they were not probed during this investigation."
3.5 Plant security personnel determined that about 80 individuals were in the power block that day and thus could have been in the radwaste building.
3.6 By 0130 hours0.0015 days <br />0.0361 hours <br />2.149471e-4 weeks <br />4.9465e-5 months <br /> on November 11, 1985, plant staff had identified the loads that were lost, which were the most important, and which might need temporary feeds to support plant operations while the damaged cables were blocked and replaced.
3.7 Unit 2 was brought back to full powr at 0245 hours0.00284 days <br />0.0681 hours <br />4.050926e-4 weeks <br />9.32225e-5 months <br /> on November 11, 1985.
4.0 SAFETY EVALUATION 4.1 This event resulted in a minor operational transient on Unit 2.
4.2 The fire was confined to the cable tray and contents of a cage in the radwaste building.
There was no uncontrolled release of radiation.
Pcga 6 of 9 4.3 According to the Fire Marshall, the fire in the materials on the floor of the cage had to be put out with water because it was a Class A fire and carbon dioxide and Dry Chemical extinguishers are for Class B & C fires.
4.4 The fire damaged cable contained fire retardant insulation and jackets.
This type of cable has been 5
subject to extensive testing, and the testing results demonstrate that this type of cable does not self ignite from sustained overcurrents.
The insulation b
and jacket will burn when exposed to flames.
This type of cable is used throughout Peach Bottom plant.
4.5 A fire in a nuclear plant has the potential to seriously affect the capability to shutdown the reactor and maintain it in a safe shutdown conditicn.
Modifications have been made at both Peach Bottom units to provide additional assurance that the units can be brought to a safe shutdown condition in the event of a serious fire that affects the Reactor Protection System, the Emergency Core Cooling Systems (HPCI, CS and LPCI) and the Residual Heat Removal System.
The cables involved in this event were not part of these systems nor were they safety related.
5.0 SIMILAR EVENTS None 6.0 SUBSEQUENT CORRECTIVE ACTION TAKEN 6.1 By 1200 on November 11, 1985, all permits needed to cut and replace the damaged cables were issued.
One half of the 34 cables in the tray were control and signal cables.
The other half were power cables.
6.2 At 1430 hours0.0166 days <br />0.397 hours <br />0.00236 weeks <br />5.44115e-4 months <br /> on November 11, 1985, the Unit 2 reactor building elevator and penthouse were inspected.
The power cable for this elevator runs through the damaged cable tray.
One of the two contactors in the elevator controls had contacts open and burned all across the face, which is approximately an inch square.
The Unit 2 reactor building elevator power cable is the largest cable in the damaged cable tray.
The cable was significantly damaged; however, the cable did not appear to have failed nor did the conductor look like it melted or severed.
A Field Engineer subsequently inspected the Unit 2 reactor building elevator motors and controls and indicated that the equipment had not drawn a high enough current to cause a cable failure.
He felt
Pcga 7 of 9 that the burned contactors were the result of normal degradation over time.
6.3 At 0830 hours0.00961 days <br />0.231 hours <br />0.00137 weeks <br />3.15815e-4 months <br /> on November 13, 1985, Electrical Construction completed the repair of all 34 cables in the damaged cable tray.
By 1800 hours0.0208 days <br />0.5 hours <br />0.00298 weeks <br />6.849e-4 months <br />, the essential equipment in the radwaste billding was ready for return to service with the exception of the radwaste discharge monitor and isolat ion valve.
These two items were returned to service the next day.
After equipment was returned to service, there were no e
electrical trips of the equipment.
6.4 On November 13, 1985, representatives from Electrical Engineering Division examined the damaged cables and cable tray sections which had been removed.
The following observations were made during this examination.
6.4.1 The insulation and jacket of the cables were charred along a stretch approximately four feet long, and the conductor was exposed for a distance between twelve and eighteen inches.
6.4.2 Only a minor melting of the bare copper conductors was evident, primarily on the #14 conductors.
6.4.3 Cable insulation and jacket beyond the charred section appeared in good condition.
Also, there were no signs of cable damage beyond the charred section and current overload due to a fault.
6.4.4 The aluminum tray rungs and portions of the tray rails were melted away in the vicinity of the 90 degree cable tray bend.
6.5 On November 13, 1985, the Fire Marshall, a representative from Claims-Security and the Security Director at Peach Bottom performed an unofficial test burn of plastic bags at Peach Bottom to determine their burning characteristics.
They found that the bags do not produce a fast flash fire.
Instead, the fire progresses slowly until the bags melt down and form a puddle of molten plastic with flames of one to t
two feet in height.
This characteristic is consistent with the burn pattern found in the cage.
The bags alone did not appear to produce the heat necessary to burn cable 6 to 7 feet above them.
6.6 On November 13, 1985, Otis Elevator representatives performed an inspection of the Unit 2 reactor
Pcgo 8 of 9 building elevator motor and controls and found no problems.
6.7 On November 20, 1985, a representative from PECo's Fire School, upon request of the Director of Security, investigated the fire by interviewing personnel at the plant, reviewing pictures taken the day of the fire and examining the fire scene and debris.
This representative was assisted by another individual who had been associated with the Fire School prior to his retirement.
The conclusions made by these two investigators are 1) that the initial fire started on top of the metal cage, 2) the burning debris fell onto the plastic and other material inside the cage and ignited it, and 3) the fire on the top of the cage burned the cables in the cable tray and caused the failure of the air line.
These conclusions are based on the following rationale.
"The material that was on the floor would not be sufficient to cause the heat damage that was observed on the top of the cage near the rear wall.
The plastic bags and material inside would burn with a much lower intensity than would be required to heat the top of the cage.
The air line, while it would have failed at a lower temperature due to its internal pressure, would not have had the concentrated heat on it from a fire on the cage floor.
A fire starting in the cable tray would have heated the air line on the upper portion not the lower portion where it failed."
The report further states that the air from the ruptured air line cleared any remaining debris from the top of the cage.
6.8 Electrical Engineering Division determined that the last time work was performed in the damaged cable
(
tray was in early 1984 when several #14 cables were installed to support use of the high integrity containers.
These added cables are 120 volt AC control cables.
Electrical Engineering further determined that there is no historical record to determine if splices were made in cable; however, there is no apparent reason for splices to have been made and there is no indication of the power cables having been spliced based on visual inspection.
6.9 The damaged cables removed from the damaged cable tray were shipped to PECo's corporate offices and then to Forensic Technologies International for analysis.
This action was taken by the Assistant Director of Security.
The cables were not shipped to Forensic Technologies Inc. until January, 1986.
The report, which is attached as an addendum, concludes that the fire was accidental and electrical in nature.
Electrical Engineering Division personnel
P g3 9 of 9 have reviewed this report and strongly disagree with the report's conclusions.
Electrical Engineering believes that "the nature and testing of the fire retardant cross-linked polyetheylene cable insulation and fire retardant neoprene cable jacket rule out an accidental ignition of a fire within the cable tray.
Results of extensive testing... clearly demonstrate that these cables in trays of various fill do not self-ignite from sustained overcurrents."
A report from Electrical Engineering Division is attached as an addendum.
7.0 RECOMMENDATIONS FOR ADDITIONAL CORRECTIVE ACTION 7.1 If in the future a fire of unknown origin occurs, the materials involved in the fire should be quickly removed and analyzed for evidence of an accelerant.
The delay of two months that occurred after this event was too long.
Addendum 1 - Report from Fcrensic Technologies International Addendum 2 - Electrical Engineering Division's Analysis of Addendum i e
,