ML20195H873

Final Rept on Fort St Vrain 871002-03 Fire

ML20195H873
Person / Time
Site:	Fort Saint Vrain
Issue date:	01/14/1988
From:	Dender W, Gramling J PUBLIC SERVICE CO. OF COLORADO
To:
Shared Package
ML20195H847	List: ML20195H843 ML20195H873
References
TAC-66365, NUDOCS 8801200390
Download: ML20195H873 (53)

Text

FINAL REPORT ON THE FORT ST. VRAIN OCTOBER 2-3, 1987 FIRE January 14, 1988 Prepared by: .

Reviewed by:

- n.

T A'/, CwawtUum#,

Approved by: II<e <- _

lBR222883R S!85Ng7

I TABLE OF CONTENTS 1.0 Purpose / Summary...........................................,..... 1 2.0 Chronology of Event and Transient............................... 2 3.0 Hyd ra u l i c 0 '1 1 Sy s t e m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 3.1 Hydraulic 011 System Description ......................... 4 3.2 Pre sent Operating Condi tion s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3.2.1 Safety Evaluation ................................. 6 3.3 Hydraul i c 011 Fi re Root Cau se . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.3.1 011 Fi l te r Ca ni s te r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.3.2 Fl a n g e/0- R i n g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.3.3 Thermal Relief Valve /Ori fice Configuration. . . . . . . . .10 3.3.3.1 Flow Calculation......................... 10 3.3.3.2 Pressure Wave Testing ................... 10 3.3.3.3 Ori fi ce In spections . . . . . . . . . . . . . . . . . . . . . 11 3.3.4 Ignition Source ................................... 11 3.3.4.1 Hydraulic Oil Flammability Analysis ..... 11 3.3 4.2 Hot Surface Correlation ................. 11 3.3.5 Root Cau se Con cl usi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 l 3.4 Related Activities ....................................... 13 3.4.1 Revised Work Instructions ......................... 13 l 3.4.2 Handwheel Verification............................. 13 l

3.4.3 Plastic Handle Replacement......................... 13 3.4.4 Hydraulic 011 Storage Lockers ..................... 13 3.5 System Testing ........................................... 13 4

i I

V 4

4.0 Root Cause for ' C' Ci rculator Tri p . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 5.0 C o n t r o l R o o m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 5.1 HVAC System............................................... 16 5.1.1 System Description................................. 16 5.1.2 Root Cause for Smoke Ingress....................... 18 5-1.3

. Filter Testing / Replacement......................... 23 5.1.4 Differential Pressure Sensor Modification ......... 23 5.1.5 HVAC Testing....................................... 23 5.1.5.1 Building 10 Pressure vs.

Control Room Pressure ................... 23 5.1.5.2 Emergency Operating Procedures /

Surveillance Procedures ................. 24 5.2 Breathing Air System ..................................... 24 5.3 Electrical Contacts Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 6.0 Fire Detection and Annunciation System ......................... 27 6.1 System Description ....................................... 27 6.2 Equipment Inoperable Prior to Fi re . . . . . . . . . . . . . . . . . . . . . . . 28 6.3 Operations Order /FPOR Applicabili ty. . . . . . . . . . . . . . . . . . . . . . . 28 7.0 Fi re Model De t e rmi na ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 7.1 Methodology...............................................29

+ 7.2 FireMaps.................................................29 8.0 Equipment Damage Assessment / Corrective Actions ................. 34 8.1 Combustion Products Analyses ............................. 34 8.1.1 Clean-Up Methods................................... 34 8.1.2 Electrical Contacts Inspections ................... 34 8.2 Environmentally Qualified Equipment....................... 35 8.2.1 Components......................................... 35 8.2.2 Cable.............................................36

8.3 Concrete ................................................. 36 8.4. Structural Steel ......................................... 36 8.5 Conduit................................................... 37 8.6 Safety Va1ves............................................. 37 8.6.1 Hot Reheat Safety Va1ves........................... 37

'8.6.2 Main Steam Safety Valves........................... 39 8.6.2.1 Background............................... 39 8.6.2.2 Thermal Analysis......................... 40 8.6.2.3 Overpressurization Analysis ............. 40 8.7 Reactor Building Filter Testing........................... 41 9.0 Licensing Assessment ........................................... 42 9.1 Appendix A Determinations................................. 42 9.2 Appendix R Determinations................................. 42 9.3 Control Room Habitability................................. 43 9.4 Hydraulic 011 System Assessment........................... 43 9.5 Interruption of Forced Circulation (IOFC) Assessment . . . . . 44 10.0 Overall Testing Program......................................... 45 11.0 Long Term Enhancements ......................................... 46 12.0 Supporting Documents ........................................... 47 1

b n

I h

- -- -, , , - - - - - - . - - , . - - - , - , . - , - - - - , -~

1.0 PURPOSE /SUMMS,RY The purpose of this document is to provide information relative to the fire at the Fort St. Vrain (FSV) Nuclear Generating Station that occurred on October 2-3, 1987. This report summarizes all previous correspondence and is considered to be the final report.

The fire occurred after the loop II startup bypass block valve HV-2292 was stroked. A thermal relief valve would not reseat due to a missing orifice that allowed excess hydraulic oil flow to overwhelm the drain funnel system and contact the hot surface of valve HV-2292, which was of sufficient temperature for ignition. An Interruption of Forced Circulation (IOFC) was experienced during fire suppression efforts due in part to fire damaged controls, while attempting to isolate the hydraulic oil. Smoke entered the control room due to a design problem and equipment malfunctions, necessitating the use of the breathing air system until the smoke was evacuated. The Fire Brigade responded efficiently and extinguished the fire while the plant operators restored forced circulation.

Since the fire, necessary repairs and plant modifications have been effected, with extensive component and system validation and testing being completed. Recovery efforts were thoroughly monitored and reviewed by Unit ?d States Nuclear Regulatory Commission personnel. In a letter dated December 8, 1987 (G-87425), the NRC stated that PSC's short term fire recovery efforts, once completed, provide an acceptable basis for plant restart.

Longer term fire recovery evaluations and actions are ongoing to assure the continued safe operation of the Fort St. Vrain Nuclear Generating Station.

l l

l

?

l l

l

2.0 CHRON0 LOGY OF EVENT AND TRANSIENT On Friday, October 2, 1987, at 2350 hours0.0272 days <br />0.653 hours <br />0.00389 weeks <br />8.94175e-4 months <br />, the plant was being prepared to place the turbine on line. The eactor was at 27%

power and Loop I was on the main stea.7. bypass control. The Reactor Operator (RO) placed the Loop II main steam bypass control in Auto to close HV-2292 and noted a drop in hydraulic oil pressure which did not recover as expected. An Equipment Operator (EO) was dispatched to investigate. The E0 found oil flowing into catch basin on Level 5 of the turbine building and then proceeded to Level 6 where he discovered a hydraulic oil leak and a fire at HV-2292 and reported the situation to the control room. The fire was extinguished by the E0 with a dry chemical extinguisher but the fire reignited as the E0 joined the Fire Brigade.

Another E0 was dispatched to isolate the hydraulics to the Group 1 valves in Loop II of System 91, which includes HV-2292, the Loop II startup bypass block valve. *SV-2112, 'D' Circulator speed valve, and *HV-2254, Loop II reheat block valve are also in this group.

The R0 placed 'C' Circulator speed controller in manual and decreased 'D' Circulator speed while increasing 'C' Circulator speed to maintain steam temperatures in preparation for isolation of the hydraulic valve group.

About this time, the fire affected signal cables to RIS-73437-1 and

-2, falsely indicating high radiation in the building stack. The control room HVAC went into the high radiation mode taking minimum makeup from the turbine deck. Smoke from the fire began infiltrating the control room.

At 0006:22, 'C' Circulator tripped on fixed high speed. Plots show a substantial increase in cold reheat pressure at this time.

Apparently, fire affected the main steam bypass control valves causing one or more of them to go open, resulting in an increase in .___ -.

cold reheat pressure. This increase resulted in overspeeding 'C' Circulator since its speed control was in manual.

At 0008:06, Loop I received a shutdown signal from the PPS on indication of high activity in the hot reheat header due to fire damaged signal cables to the shine monitors. Both 'A' and 'B' Circulators were losing speed and at 0008:22, the core pressure drop went negative, (even though 'D' Circulator was self-turbining),resulting in an Interruption of Forced Circulation (10FC). Therefore, at 0008:27 the R0 inserted a manual Scram.

At 0015, the fire was reported to be extinguished.

• Note: These valves were erroneously numbered SV-2212 and HV-2253 in the preliminary report.

__.__ m __m. _ _ _ _ _ _ - _ _ . . _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ ..

B boiler feed pump was started and core cooling was re-established at 0023:54 with feedwater on loop II and 'C' Circulator er steam drive, thus terminating the 10FC.

At approximately 0030 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> on October 3, 1987, -the' Shift Supervisor declared an Alert and initiated the FSV Radiological Emergency Response Plan (RERP) according to EP-I due to the fire and subsequent damage to equipment.

'At 0032 hours3.703704e-4 days <br />0.00889 hours <br />5.291005e-5 weeks <br />1.2176e-5 months <br />, Loop I was recovered with 'B' Circulator providing primary coolant flow.

At 0817 hours0.00946 days <br />0.227 hours <br />0.00135 weeks <br />3.108685e-4 months <br /> the Alert classification was terminated and the plant entered the Recovery Phase, i

I L

l 1 ,

l i_ _

3.0 HYDRAULIC OIL SYSTEM 3.1 Hydraulic 011 System Description System 91, the hydraulic oil system, is designed with two separate loops, one for each secondary coolant loop. Each loop of System 91 was originally designed to have one pump supplying hydraulic oil to the four valve groups in that loop. There are two hydraulic oil supply headers that feed the four valve groups in each loop. In each line from a header to a valve group there is a flow control valve that limits flow through that valve to 6 gpm. The purpose of these flow control velves is to ensure that oil is supplied to the other three. unaffected valve groups in the event of an oil leak in one of the four groups.

When nitrogen header pressure exceeds oil header pressure by 250 psig, an alarm sounds in the control room. By checking the individual oil header pressure indicators in the control room, the operator can determine the affected header. Since the flow limiting valve on each oil inlet header limits header oil supply to a leak to 6 gpm, any leak greater than 6 gpm will indicate zero oil header pressure.

When oil header pressure is less than 75 psig greater than nitrogen header pressure, an alarm will sound in the control room indicating an accumulator is discharging or is not full of oil.

There are additional alarms in the control room for each of the following conditions:

1) Reservoir oil temperature greater than 115 F.

2) High differential pressure across the pump filters (bypass and high pressure pump discharge).

3) Low oil level in the low pressure oil tanks.

a) Backup storage tank low level - alarm in control room, b) 100 gallon level in the hydraulic oil l

reservoir - alarm in control room.

c) 50 gallons in the hydraulic oil reservoir -

I trip high pressure pump (s).

I r

! l l

3.2 Present Operating Conditions During initial operation of the plant, an increase in internal system leakage created a situation where the standby pump was cycling more often than desired. Over a period of' time, this situation of cycling and system pressure fluctuations could degrade the performance of this pump and other system components. For this reason, the second of the three pumps is operated in parallel with the first pump, and the third pump is available as necessary.

The present configuration has each of the running pumps feeding one of the two oil supply headers that supply the valve groups in one loop. With this configuration, each of the flow control valves limits oil flow through that valve to a maximum of 6 gpm, for a total maximum oil flow of 12 gpm per group. The purpose of the flow control valves remains the same, that is to ensure oil supply to the three unaffected valve groups in the loop in the event of an oil leak in one of the four groups. With the present configuration, there is 20 gpm (10 gpm per pump) available to all four valve groups. In the event of a leak, flow to the leak would be limited to 12 gpm (6 gpm per flow control valve) after the two accumulators associated with the affected group have discharged their contents. This leaves 8 gpm available to supply the three unaffected valve groups until the third pump is started, at which time 18 gpm of oil would be available.

The normal operation of two hydraulic oil pumps in each hydraulic oil loop does not conflict with the requirements of LCO 4.3.7 of the Technical Specifications. LCO 4.3.7 (the only LC0 associated with System 91) requires that at least two pumps be operable in each hydraulic oil loop, or that the reactor be shutdown within one hour and the non-affected secondary coolant loop be isolated. The basis for this LC0 states that each "hydraulic system will normally operate with two hydraulic fluid pumps and both hydraulic accumulators in service. The second hydraulic pump and accumulator is redundant." The redundancy afforded by the second pump is retained whether it is normally operating, with some of its supply recirculated to the backup storage tanks, or off and in a standby status. Experience has demonstrated that from a maintenance and equipment reliability standpoint, it is advantageous to have the second mmp operating. The third pump provides additional redundancy.

The situation where two hydraulic pumps are normally running does not change the purpose or function of the flow control valves. These valves were installed to ensure flow to the unaffected groups in the event of a leak in one group.

During the fire on October 2-3, 1987, these valves proved that they do perform their design function with two pumps running. Valves in groups of Loop II other than the one which experienced the leak were successfully stroked in the process of cooling down the reactor before the affected group was isolated. A secondary function of the flow control valves is to limit the flow of oil out of downstream piping or a component leak. With two pumps in operation instead of one, this flow is essentially limited to 12 gpm instead of 6 gpm, until the leak can be isolated. During the October 2-3, 1987 fire, the flow control valves functioned to restrict the hydraulic oil flow out of the open HV-2292 thermal relief valve to a maximum of 12 gpm until the header supplying the Group I hydraulic oil valves was isolated.

3.2.1 Safety Evaluation Based upon the above information, the operation of two hydraulic system pumps in each loop of System 91 supplying hydraulic oil to the four valve groups via two supply headers (in lieu of operation of only one pump in each loop supplying oil via one supply header as described in Section 9.11 of the FSAR) has been determined to not be an unreviewed safety question. The following evaluation supports this conclusion.

The probability of occurrence or the consequences of an accident or malfunction of equipment important to safety previously evaluated in the FSAR has not been increased because:

1. The probability of a leak causing an accident in the piping of System 91 is equally improbable with one or two pumps and supply headers operating, since System 91 is designed and installed in accordance with ANSI B31.1 and GA Spec 1-M-2. This piping was tested to 1.5 times the system design pressure of 3300 psig and is seismically qualified to withstand a Design Basis Earthquake.

2. The consequences of a fire accident in the area of the hydraulically operated steam valves in the turbine building are discussed in pages 4-38 through 4-40 of the Fire Hazards Analysis report (P-78182, dated 11/13/78, listed as Reference 2 in FSAR Section 9.12).

This Fire Hazards Analysis states the following: "The hydraulically operated main and reheat steam and feedwater valves and associated piping normally contain a relatively small quantity of hydraulic oil which does not represent a significant fire load for this fire area. However, in the event of a hydraulic oil leak, it is inherent in the system design that additional oil from the hydraulic power unit will be pumped to the area. This could result in up to 520 gallons of oil being burned, producing a fire load of 4500 Btu /Ft2 (4 minutes). The ability to isolate the oil supply to each hydraulically operated valve minimizes the impact of such a fire." This 520 gallons of hydraulic oil assumed to be released in the Fire Hazards Analysis envelopes the quantity of oil that would be supplied through a 12 gpm leak over a period of 15 minutes, a reasonable isolation time.

Therefore, the consequences of an accident have not been increased over those previously evaluated.

Likewise, pages 4-25 through 4-27 of this Fire Hazards Analysis conservatively assumes total combustion of 520 gallons of hydraulic oil supplied to a postulated leak in the hydraulic oil system in the Reactor Building PCRV and Auxiliary Equipment Area -

Elevations 4759 ft. and 4756 ft. Supply of 12 gpm of hydraulic oil to this leak instead of 6 gpm would not result in exceeding 520 gallons of hydraulic oil assumed to be burned. The consequences of a hydraulic oil fire in this area have not increased over those analyzed for the same reasons stated in the above paragraph.

The Fire Hazards Analysis, pages 4-28 through 4-31, analyzes complete combustion of the hydraulic oil contained in both System 91 hydraulic power units for the Reactor Building PCRV and Auxiliary Equipment Area Elevation 4740 ft-6 in. The consequences of this fire would be unchanged by the operation of either one or two pumps.

3. The probability of an equipment malfunction due to the operation of two hydraulic pumps (one pump in each supply header) during normal operation has been reduced because the second pump is now running continuously rather than cycling on and off to keep up with system demands, thereby increasing the life of the system's components.

4. The consequences of an equipment malfunction remain unchanged because the third pump is available to be started in the event of. failure of one of the two operating pumps.

The possibility of an accident or malfunction of a different type than any evaluated previously in the FSAR has not been ,

created because:

1. As shown on PI-91-1 (Loop I) and PI-91-2 (Loop II) and in Reference Design 50-91, each of the two loops of the hydraulic system have three hydraulic oil supply pumps and two supply headers. With this configuration, one pump can be aligned to supply either supply header. Two pumps can be aligned to supply both supply headers, one pump per header. The third punp can be aligned to supply either supply header, in the event that one of the other two pumps has failed. Each pump discharge is provided with a relief valve to return the pump's excess capacity to the hydraulic oil backup storage tank associated with its loop. Therefore, when necessary, system demand can be met with two pumps in operation supplying hydraulic oil to the system's users without overpressurizing any system components. In summary, there has been essentially no change in system function, capability or eouipment that could result in a different type of accident or malfunction.

2. A fire accident associated with the hydraulic oil system with two pumps, and two oil supply headers, in lieu of one pump and one oil supply header, supplying oil for operation of the hydraulically operated steam valves could double the rate of oil available to feed _ _ _ . _ _ _

a fire. The total amount of oil released with operation of two pumps would be less than that assumed in the Fire Hazards Analysis.

3. Operation of the hydraulic system with two pumps is no different than the operation discussed in the basis of the LCO 4.3.7, which states "The hydraulic system will normally operate with two hydraulic pumps and both hydraulic accumulators in service." The third pump is available for startup in the event of a failure of one of the two operating hydraulic pumps.

The margin of safety, as defined in the basis for any Technical specification or in the FSAR has not been reduced because:

-The margin of safety provided by the requirement for operability of two hydraulic oil pumps per loop in LC0 4.3.7 has not been changed. The sys'.am now routinely has tso hydraulic pumps operable and in operation and a third pump is available to be started in the event of an emergency.

3.3 Hrdraulic 011 Fire Root Cause As discussed in Section 2.0, Chronology of Event and Transient, the E0 dispatched to ascertain the cause of the low hydraulic pressure alarm stated that he saw oil flowing into the catch basin on Level 5 of the turbine building, he then proceeded to the area of HV-2292, Loop II startup bypass block valta, on Level 6 of the turbina building, because he knew that HV-2292 had been a historical source of hydraulic oil leakage. When he discovered the fire, he stated that he was on the opposite side of the hot reheat safety valves from HV-2292 and that oil was "flowing" into the fire. However, he could not detect its source.

3.3.1 011 Filter Canister ince the hydraulic oil filter canister had ruptured, PSC conducted cooprehensiva metallurgical evaluations to determine 'f the filter canister was the fuel source and initiator of the fire. This investigation was also directed by the following: the hot reheat safety valves wtre engulfed in the flames, the

canister was closest to these valves, and the discovery of pipe wrench marks on the canister indicating that perhaps past maintenance practices had compromised the canister integrity (although no failures had been reported at Fort St. Vrain).

However, the metallurgical examinations determined that the fiiter canister could not be the primary source of the fire, and that pipe wrench marks are not failure contributors (PSC Failure Analysis of the Filter Assembly Bowl from HV-2292, Laboratory Report No. 136, dated October 26,1987).

3.3.2 Flange /O-Ring The flange and o-ring external to the HV-2292 lagging was investigated to determine if it was the source of the fuel and initiator of the fire. Initial inspection by PSC found no evidence of o-ring extrusion. R. Beaufort of Paul Munroe ENERTECH, PSC's Hydraulic System Consultant, confirmed that extrusion would be visible, even in the crystallized state that the o-ring was in after the fire, and must be present to alicv leakage at this point. Therefore, the flange /o-ri.9 configuration was eliminated as the initiating fuei source.

.g.

s 3.3.3 Thermal Relief Valve / Orifice Configuration When PSC disassembled HV-2292, the thermal relief valve designed to compensate for hydraulic. oil thermal expansion was found to be degraded to the point that it could not reseat. Upon further- investigation, it was discovered that a .035" orifice designed to reduce relief valve flow and supply pressure during a relief was missing.

3.3.3.1 Flow Calculation Calculations were performed that indicated flow through the relief valve with the orifice in place was restricted to approximately 1.3 gpm at 3000 psig. Flow through the relief valve without the orifice in place was calculated to be approximately 16.4 gpm at 3000 psig.

3.3.3.2 Pressure Wave Testing The system normally operates at approximately 3000 psig and the relief is set at 3300 psig, with reset at 2900 psig.

, Controlled testing conducted on ti.e "sister" valve, HV-2293, indicated a closing pressure wave perking at approximately 3450 psig.

This t;st shows that the relief valve would experience sufficient pressure to actuate when the hydraulic valve was actuated until the pressure subsided.

Note: The thermal relief valve is not i.1 tended to _ _ _ _ _ _ _

reduce or control the effects or size of the pressure wave.

- s 3.3.3.3 Orifice Inspections Inspections performed on the remaining 5 valves that utilize the. orifice / thermal relief valve. configuration revealed that all orifices are in place with the exception of HV-2254, It_was found that on the cap side, there was no orifice or thermal relief valve installed: NCR 87-527 was issued. It has been determined that the relief valve / orifice is not necessary on this valve because when fluid is trapped on the cap side (HV-2254 closed), sufficient heat transfer through the piston to the cap side would not be. anticipated because all the heat would be dissipated in tne untrapped stem side fluid. Therefore, HV-2254 is considered fully operational.

3.3.4 Ignition Source 3.3.4.1 Hydraulic 011 Flammability Analysis Gulf Harmony 011 #68 is the hydraulic fluid used in the Fort St. Vrain hydraulic system.

This oil has an ignition point of 825'F when presented to a heat source in spray or mist form. However, when pr'esented to a heat i source in a laminar dispersion, the ignition point is - 515'F.

l 3.3.4.2 Hot Surface Correlation Startup bypass block valve HV-2292 closes l

when steam temperature reaches 760 F when j the hand switch is placed in the auto mode i and at 800*F when in manual mode. Since HV-2292 was in auto mode when it stroked to the i closed position at 2350 hours0.0272 days <br />0.653 hours <br />0.00389 weeks <br />8.94175e-4 months <br /> on October 2, 1987, as the turbine was being prepared to be placed on line, the steam temperature was approximately 760*F. The actual temperature of valve HV-2292 on October 2, 1987 is j unknown. GA Report A-13387 analyzed steam valves operating at 1000*F and determined that the exposed valve's bonnet area would be approximately 550 F. However, it is

clear that the surface contacted by the hydraulic oil (whether it was the valve or

, uninsulated piping) was of sufficient l

temperature for ignitiors.

l l

i i-

3.3.5 Root Cause-Conclusion The closing pressure wave is associated with the termination of EV-2292 valve motion and is sufficient to lift the thermal relief valve. The relief valve was unable to reseat since the header pressure, as seen by the relief valve, was not being reduced by the missing upstream orifice. The oil confinement system was unable to contain the flow volume and velocity,

, which coula have been as high as 17 gpm at 3000 psig without the orifice in place. This calculated flow rate does not take into account the 6 gpm flow limiting valve's function for each header, nor the accumulator volume. Operators reported that roughly 92 gallons of hydraulic oil were required to refill the hydraulic oil storage tank after the fire. On October 2, 1987, it is concluded that the leakage not contained by the drain funnel system contacted the hot surfaces and ignited.

The preceding statement is supported by the metallurgical analyses conducted on the hydraulic oil filter canisters. These evaluations concluded that the canister failure is secondary. Additionally, valve closure pressure is not sufficient to result in room temperature over pressurization failure.

The external flange /o-ring configuration is also not considered to be an initiator of the event since it did not show evidence of o-ring extrusion.

R. E. Beaufort of Paul Munroe ENERTECH, PSC's hydraulic system consultant, has substantiated PSC's conclusion in a letter to PSC dated October 27, 1987 _ _ _ .-- _ .

i as follows:

"My conclusion is that the .035" diameter orifice that was missing at the cap end port of HV-2292 resulted in the piston / seat destruction of the thermal relief valve at the cap end of HV-2292 valve actuator. This condition caused fluid to be forced tiirough the relief valve and related open drain

! piping when the valve actuator was selected "closed". The presence of fluid vapor on the adjacent steam line burned the filter

bowl causing it to lose strength and burst adding additional fuel to the fire."

i l

\

l l

l

~ n 3.4 Related Activities 3.4.1 Revised Work Instructions During investigation of the root cause of the hydraulic oil fire, pipe wrench marks were found on the hydraulic filter canisters. Pipe wrenches were used to overcome system pressure for removal of the canister. The' standard clearance points for System 91 have been updated to reflect the capability to vent system pressure through the existing 5-valve manifold.

Additionally, PSC has revised the maintenance procedure associated with the canisters to ban the use of pipe wrenches.

3.4.2 Handwheel Verification During the fire, a System 91 hydraulic group isolation valve handwheel was discovered missing. Therefore, PSC initiated a review of the Station Service Requests (SSR) to determine if any other handwheels had been identified as missing in the plant. Three additional valve handwheels in System s' were noted as missing.

Additionally, Ope-ations personnel fcund one more missing handwheel during tneir pre-startup system lineups. This represents a total of five missing handwheels in System 91: four have been replaced with the fifth awaiting parts (valve stem).

3.4.3 Plastic Valve Handle Replacement Plastic valve handles for locally mounted accumulators that actuate valves FV-2205 and FV-2206 for Appendix R purposes melted during the fire. Change Notice 2012, _ _.

Issue A, replaced these handles with metal handles.

3.4.4 Hydraulic 011 Storage Lockers NFPA 30/FM approved storage lockers have been located within the plant to reduce the fire hazards associated with the storage of cocbustible hydraulic and lubricating oils.

3.5 System Test's SR 5.3.5-A performs the annual calibration for the pressure indicators and low pressure alarms on the hydraulic oil accumulators' pressurizing gas and on the hydraulic power supply lines. This was last performed on April 16, 1987. SR 5.3,5-Q is the quarterly functional test for the pressure '

indicators and low pressure alarms on the hydraulic accumulators' pressurizing gas and on the hydraulic power lines. This test was last performed prior to the fire on September 17, 1987.

< r u

. t.

There is no calibration or function'al;; test presently

, conducted on,the 6 gpm flow limiting valves. However, PSC has concluded that the valves adequately performed their -

design function of limiting flow out of the damaged headers in order for the rest of the valves in the loop to stroke.

This was evidenced by the fact that all the valves not directly affected by the fire were able to correctly position themselves 'during the period of hydraulic fluid expulsion.

PSC satisfactorily re performed the quarterly functional testing prior to restart. Additionally,.PSC is procuring appropriate instrumentation that will facilitate flow limiting valve testing.

i I

l l

, l l

L -

4.0 ROOT CAUSE FOR 'C' CIRCULATOR TRIP At 0005:50 on October 3, 1987, 'C' Circulator Water Turbine Trip alerm was received with the circulator operating at approximately 8520 rpm.

At 0006:22, 'C' Circulator tripped on fixed speed high. Data logger printouts do not conclusisely validate this trip as an actual fixed high speed trip because there exists a 5 second interval between points. The last indicated speed at 0006:18.52 was 91 ' rpm. At 0006:23.52 the speed had dropped to 6666 rpm in a tripped condition.

PSC concluded that 'C' Circulate- trip was the result of an actual overspeed condition. Main steam bypass control was lost during the fire, as indicated by an increase in cold reheat pressure and a corresponding decrease in main st9am pressure. 'A' and 9' Circulators show an increase in speed cf 800 rpm in response to the cold reheat pressure increase. Their n spective controllers began to drive the speed valvas closed, w!'h controller output being reduced by 20%. 'C' Ci rci. l a to r control was in manual at 48*4 output, however, and coula act respond to the increase in pressure.

The increase in pressure caused the Circulator to exceed the trip setpoint of 10,710 - 10,800 rpm, and a trip resulted at 0006:22.

This is realistic considering the rapid response of this system to steam flow / pressure changes. Performance curves on coast dow times foi the Circulator also show that it is ciedible for the Circulator speed to decrease to 6666 rpm at 0006:23.52 after reaching the fixed high speed trip point at 0F.6 22.

Additionally, testing was conducted to verify the openability of PPS input parameters for th', type of trip.

The high speed PPS trips for 'C' Circulator have been testec The water turbine trips were verified to be set at 8620-8690 rpm, all within the acceptance criteria. The steam turbine trips were verified to be set at 10730-10800 rpm, all within the acceptance criteria.

The steam turbine speed indicaticn en 'C' Circulator, S1-2106, was tested and was within the acceptar.ce criteria.

The water turbine speed indication on 'C' Circulator, 31-2110, was '

tested and was within the acceptance criteria.  ;

Therefore, it is concluded that since all of the inciciting parameters for input into the PPS logir for a fixed high speed trip have been tested 4.nd proven to b'. capable of proper function, a steam spike caused by the loss of main steam cypass control as a result of the fire actually oversped 'C' Circulator.

Finally, 'C' Circulator was fun tionally tested and found fully operable.

4 5.0 -CONTRO.1, ROOM 5.1 HVAC System l

,- S 5.1.1 System Description i:

-System Design Intent - The control roem HVAC system is designed to provide controlled ventilation air to the cor. trol room and auxiliary equipment room under all plant operating conditions.

Normal System Operation - Normal operation of this system is divided into two modes - the refrigeration mode and the economy mode. (See Figure 1). The system operates in the economy mode when outside '

ambient temperature is at or below 53 F. In this mode, the free outside air i,as enough cooling capacity to satisfy all cooling requirements for the control room air handling unit. The system switches to the t refrigeration mode when the outside air can no longer satisfy all the necessary cooling requirements.

Control room supply fan (C-7504) supplies ventil6 tion ,

g air to the control rcom, operator's training room, end e' auxiliary equipment room via separate ducts to each of these rooms. ,

The control room return fan (C-7505) draws return air from the system and exhausts it to the control room return air plenum. Depending on pressure and temperature requirements, this return air is exhausted to atmosphere and/or recirculated to the air handling unit.

The control *oom toilet exhaust fan (C-7507) continually exnausts to the outdoors anytime the 3

control room return fan is on. Make-up air to the control room air handl W. unit is normally supplied from outside air through DV-75299/DV-75300. The access bay vent fan (C-7524) draws froin outside air '

i and/or control room return air, i Control room pressure is maintained slightly positive '

with respect to the turbine building to ensure that

any leakage is cutward. This pressure is maintsined during various modes of operation in the following manner

j 1) Defrigeration Mode: Pressure is maintained by throttling the outside and return air i dampers DV-75299 and DV-75300. ,

1 ,

I I

l

i I

b

• n.

2)- Economy Mode: Pressure is maintained by throttling the return air through either of two dampers - DV-75303 if the access bay fan (C-7524) is on or DV-75363 if this fan is off.

3) High Radiation Mode: Pressure is maintained by throttling the control room emergency fan

. inlet damper DV-75296.

Note: There is no pressure control during the purge mode other than that provided by supply and exhaust fan capacity differences.

Abnormal System Operation In the event or high radiation as detected by.any one of the four stack monitors (see Figure 2), the following actions occur:

1) Outside air damper (DV-75298) closes.

2) Control room HVAC changes to refrigeration mode.

~

3) Chilled water flow to the service building HVAC is isolated (FCV-7543 closes).

4) Control room emergency fan (C-7506) auto starts.

The control room HVAC in this high radiation inode (see Figure 3) now operates as during the normal refrigeration mode, except the make-up air is drawn _ . . _.

from the turbine building and passes through the

, emergency fan and filter (F-7502) before entering the i air handling unit. Control rcom pressure will now be l

maintained by thrcttling DV-75T96.

When sme t.e is decected in either the control room or auxiliart equipnnt room, vcntilation supply air is automatically directed through the control room l charcoal filter (F-7504) in :ddition to the normal j control room filter (F-7503).

The purge mode (see Figure 4) is initiated manually by j taking HS-75184 to purge. This action reopens the isolation dampers and positions the system dampers to supply 100% cutside air and exhaust 100'r, of the return j air to outdoors.

l L

! I

t .l Differential Pressure Control - The purpose of PDT-7556 is to provide the signal which maintains a positive pressure differential between the' control room and the turbine building, This information is -reflected in Engineering Evaluation EE-75-0003, Rev. A.

5.1.2 Root Cause for Smoke Ingress Operator Statements Operators in the control room at the time of the fire have indicated that the concentration of smoke that entered the control room was an eye irritant only.

The breathing air system was used periodically, as desired. However, at no time did the operators consider evacuating the control room. Operators noticed the smoke entering around the doors. After efforts to clear the atmosphere with the purge mode of the control room HVAC failed, doors were opened that allowed the Building 10 HVAC system to expell the '

smoke from '.he control room.

There was one entry into the auxiliary electric room ,

at 0047 hours5.439815e-4 days <br />0.0131 hours <br />7.771164e-5 weeks <br />1.78835e-5 months <br /> on October 3, 1987. At this time, the operator noticed no smoke in this area.

Conclusion The pressure differential control sensing line, which was located in the auxiliary electric room at the time -

of the fire, could not provide adequate differential pressure control between the turbine building and the ___

control room.

Additionally, a problem existed with damper DV-75324 that created a flow restriction-between the inlet and

outlet flows and resulted in a negligible differential pressure between the control room and the turbine i building, allowing smoke ingress in the high radiation mode.

, Placing the HVAC in purge mode caused greater negative pressure in the control room because a damper fciled to fully open, which resulted in increased smoke ingress under and around the doors.

h b

ll 1

, . ,_. __ _ m. , _ . . , . .__ _

& A PS nn '

l I I

• uV ,V vv IV i s M'***

Normal Econcery Mode Outside F

With .ccess Bay Fan On Outalde V -

C-7524 F-7502 DV-75298 Turbine  % w a ua Building < vm v w I p-75296 Dv-75297 i

(Modulating)

S-7504 DT-75303 j M-75363 Cold DT-75301 -752

• E

• N Deck (Modulating) j I j ,

F-7 W .

J '

r M; ' d1 8*E l operating

• j' D*Ch v _ meture wQ Dv-75 'l **"

(Modulating) '

so ,

Eit.chen C-7507 W outside C-7505

) Y To let DT-75330 V s CONTEDL RDOM DV 4 V x

DV-75'325 a =27 7 Wh -

09-75326 v w Atm. v W '

EQUIF W DV-75329 apon Dv-75328

_ - _ - _ m ____

RADIATION MONITOR CONTROL BLOCK DIAGRAM p/3 73]g. /_ A/S IA0l*/ WA /32 96I R/3 7324 2 A%$ 73437 2 LG4 132 97l

\J -

l114 2SX dAS WASTE' SY$ N SELECTOR (A-GAS WASTE SU4GE TAMX- C7Co6 ceN7'RM vtMr CCN740L EMEA6 F/LrdM s'AN (f 7$C1)

R/S 73 #

akI?? Wll0 gp A3v p /3'V 25X $7549*/ q 7349'3.

O pdc 7S54 Dv.?s19e toursice e rv ewro)

\ Accass anv ( Asove *27so1) f5V 73/9 W D V 7JSol, av 73 498, py734 9S D/1J 4H Acesss sAy f5V7W3 W Fe v7./M7643 at cc 14.4 44 ATf 7549-4

&R As 7Cr7.

I I I I Ris Ris A/s R/S 7J24 7Jtf 7324 7324 zSxa 2.93 25X4 2SXS

• C 7C27UEM C 7 S2 4 A Y.2:5 5 L 7GSS TUA0 u n9 M X Ain t V ! 847 Anst , c,7SSX & *CM?hGCA55 M/

• d 7.98 7168 C7525 KCM & %" f M LL'M 0/lEM S47 gy p;4g *p5y 7g/75

• C 75M WE@ .& 7S03' 72MB .p5V75/76 SHOP fM/A4 '

m Rgx

• C 75.30 W7R . c 1soix 72MJ 7AdPT M .c C 75 2 6 C M e#9 .p n Zose,ia j LM EM Fw Gg

'f.swMow##w *fsr 7S13

.m

• P3 Y 75236 A m --

A mm t

- a l l =

vv v'u

. i

v Outsidc-I 5.iy liigh Radiation Mode /( d i Outside w h a C-7524 r-7502 DV 75 8 Turbine v Building A

• DV-75296 DV-75297 (Modulating) 8-## -75363

,,73g3 DV-75303 D -752 9 (M dulating (Hodulating)

M , ,y fCold DV-75301 d' 'I '

D'ck (Modulating) c-7504 q 7,7 9 m - '

J L Wh M ;' 4h Hot j' Deck V' _

Return

1. I" DV E DV-75300

$ (Modulating) ,

See Note 1  ;

Kitc h 0-7507 C-7505 v & & M Outside Dg e T ilet DV-75330 M ' CONTROL 200M D 5 4 V

[ DV 7

^

• DV-75325  ;

1 tot h DV-75326 v m AUI. v

r. quit m DV-75329 ROOM DV-75328 Note: 1. Flow path through F-7504 initiated by smoke detection.

CDOTRCL RJCm H JAC "'**

outaide Pur*;e Mode as ConfIqured i ,

%bsequent to October 3.1967 Fire - -

Outeide y tu C-7524 r-7502 ev-7529e

~

g Dv-75296 DV-75297 (Modulating)

• See Note 1 8-i *

,,, ov-75303 / -75363 i

' j c,g, py_y33og 752 9 Madng) W u W ng) p 7 f Deck (Modulating) c-7504 - g ,_y g nu numa }( j ch 'a

~ s Retura

{yg vienum e

(Modulating) s See Note 2 III" C-7507 W Outside C-7505

[ To let DV-75330 x

Dv-75324 y amoi. m.

} Dv-75327 Dv-75325 v tog noom 3

Dv-75326 Er Y ov-75129 noose Dv-7532s Notes: 1. Flow path through F-7502 locked in by Radiation Monitor until manually turned of f.

2. Flow path through F-750t. Initiated by smoke detection.

5.1.3 Filter Testing / Replacement PSC has conducted testing on F-7502 (the control room HEPA and charcoal filters), which is the filter that will be included in the Technical Specification Upgrade Program. The testing was in accordance with '

Reg. Guide 1.52, according to the testing organization, Nuclear Containment Systems Corporation, and the results indicate that this filter arrangement i is acceptable.

PSC has replaced the 12 pre- and 12 post- filters for F-7503, although no testing was performed.

Replacement charcoal for F-7504 has been ordered.

However, F-7504 was not considered a start-up impairment since F-7504 does not function unless the

-control room fire detectors are actuated.

3.1.4 Differential Pressure Sensor Modification .

PSL has completed a modification per Change Notice No.

2713 that relocated the differential pressure sensor from the auxiliary electric room to the control room.

This was done to provide a more direct control of the pressure within the control room relative to the turbine building.

5.1.5 HVAC Testing

. The HVAC system was tested after preventive / corrective maintenance and sensor relocation had been performed.

This testing was monitored by the NRC Resident Inspector and proved that the system will maintain a positive pressure in the control room relative to the turbine building in all steady state modes.

i i 5.1.5.1 Building 10 pressure vs. Control Room Pressure The Building 10 HVAC system operates at a positive pressure due to the sizing of the air handling units, and is adjacent to the i control room. During trouble-shooting of the control room HVAC system, the potential l l

for air in-leakage to the control room from

Building 10 was identified. Dampers were adjusted that, when tested, allowed the control room HVAC to overcome the pressure

from the Building 10 HVAC system. PSC is  !

I continuing to investigate enhancements to j the control room HVAC system, t

I

5.1.5.2 Emergency Operating Procedures /

Surveillance Procedures '

PSC has reviewed the system operating procedures, abnormal operating procedures, surveillance procedures, and emergency procedures for the control room HVAC. These  ;

procedures were found to be adequate for the operation of the control room HVAC, .in accordance with the design intent.

5.2 Breathing Air System PSC has installed two additional breathing air system masks in the control room. This brings the total number of masks in the control room to five. Although there are a total of six ports in the manifolds, which allows some flexibility in staff location, the system will only support five users. In addition, Scott Air-Paks are available if necessary.

5.3 Electrical Contacts Inspection The Fort St. Vrain turbine building Fire of October 2-3, 1987 provided the source of certain combustion by prouucts which '

were pulled into the control room environment. There exists no quantitative measure of the amount or concentration of by-products pulled into this environment and either deposited or

. exhausted to the outside. Discussion with operating

! personnel has provided some visual measure of this  ;

concentration. This measure may be described by the following observations:  !

1. Smoke in the control room was described as a light ,

haze, greyish brown in color near the ceiling. The thickness of this layer was estimated as "down to the top of the control boards" (or approximately 24 inches). This is consistent with the smoke i

characteristics of Gulf Harmony 68.

l

2. The sensory effect was described as acrid and a clear l

, air irritant.

3. A black material (soot) formed a fan shaped deposit, ,

three (3) feet wide on the carpet at the center  !

joining of the double doors. A rectangular shaped area one (1) foot wide was deposited along the bottom  ;

of the west door.

l l

i l

l .. . . _ _ _ . ._

An investigation has been completed to determine, first, if there are any combustion by products visible on or in the control room electrical components and second, if any materials are found, could these materials be detrimental to contacts and other. exposed surfaces. A visual inspection of various main control room components and a series of swipe samples on November 11, 1987 provides the following:

1. Protected horizontal surfaces have an accumulation of dust particles which appear to be light grey to brown in color. (This material appears to be similar to materials found on office furniture remote to the control room and turbine building.)

NOTE: Nineteen (19) wipe samples were taken from horizontal surfaces in an "S" pattern of approximately twelve (12) inches long. These samples were examined using a ten (10) power optical comparator.

2. Material concentrations cover approximately 200 cm2 on the wipe samples. This concentration represents a surface area of approximately 30,000 cm'. Several wipe samples have black smears or spots 0.2 to 0.4 centimeters across. The area ratio of black material to the lighter materials is conservatively 6 to 1 million.

3. An inspection of relay contacts (to the extent visible and accessible) showed no visible foreign material.

All terminals and terminal board wiring inspected appeared to be free of foreign materials.

Control board mounted switches, controllers, indicators, and recorders are by virtue of physical construction protected against falling particulate. The various modules comprising the plant protective system are housed in enclosed NIM bins.

The safety related relays associated with this protective system are located in the lower portion of I-9310 and protected from falling or settling materials by the enclosed NIM bins mounted above.

CONCLUSION

1. The negative differential pressure condition of the control room during operation of the HVAC system in the purge mode resulted in the ingress of combustion by products.

2. Some part of these by products was in the form of a black particulate.

3. The major portion of this material was deposited on the carpet in the immediate vicinity of the south double doors and the west door.

i t

_ 4. The amount of material which settled on component surfaces is insignificant in terms of the total surface exposed. -

4. 5. The protective enclosure surrounding electrical and electronic equipment providos protection against the  ;

ingress of materials falling or' settling. -No credible i

-mechanism for forcing materials into component housings was identified. j

6. An inspection of component surfaces shows no visible ,

evidence of foreign. materials, visible surface discoloration or degradation.

i E

h h

i.

I

. 1 b

i h

I E

Y i

i L i

r k

t 1

1

6.0 FIRE DETECTION AND ANNUNCIATION SYSTEM 6.1 System Description The fire detection system is designed to quickly notify operations personnel of potential fires and to activate associated fire suppression systems if provided. Distinctive audible and visual alarms annunciate in the control room and locally when a fire detector is activated. The system complies with the intent of the following NFPA Codes as indicated in the PSC/FSV Fire Protection Program Plan:

72A (1979) local Protective Signaling Systems 720 (1979) Ir.stallation, Maintenance and Use of Proprietary Protective Signaling Systems 72E(1984) Automatic Fire Detectors The following plant areas housing safe shutdown equipment are provided with automatic fire detection:

Reactor building Turbine building Standby diesel generator rooms Auxiliary boiler room Turbine lube oil storage room Turbine lube oil reservoir room Control room Auxiliary electric room 480 volt switchgear room Building 10 Circulating water makeup pump house Service water pump house Technical support building Power for each detector is provided from a non-interruptible power supply and electric circuit supervision monitors for open circuit, closed circuit, or loss of control power conditions.

Exemptions have been granted by the NRC from 10 CFR Appendix R,Section III.G for the control room, reactor and turbine building fire detection systems, as noted in the NRC draft i FSV Fire Protection Safety Evaluation Report dated Decerober

! 18, 1987, l

l l

l 6.2 Equipment Inoperable Prior to Fire Prior to the fire, the following detectors were inoperable:

XD-45356 (photoelectric smoke detector on Level 2 of the reactor building), XD-45359 (linear beam detector on Level 2 of the reactor builai69), and XD-45372 (linear beam detector on Level 5 of the reactor building). PSC was working on these and other detection system problems prior to the fire and had scheduled a meeting with Gamewell-personnel in an effort to resolve these problems.

It has been determined that the audible fire protection control room annunciator was silenced due to frequent alarms associated with the detection system problems.

6.3 Operations Order /Fp0R Applicability Operations Order 87-14 was issued to -the control room operators after being approved by the Plant Operations Review Committee (PORC) on November 3, 1987. This implemented excerpts from the Fire Protection Operability Requirements for the minimum operable detectors in the interim period before the Fire Protection Program Plan became effective, and required fire watches when the minimum number of detectors were not operable. It also provided the operators with instructions as to removing a nuisance alarm from service without disabling the control room annunciator.

1 I

l i

l l

1 4

7.0 FIRE MODEL DETERMINATION 7.1 Methodology The fire model was developed by the PSC Fire Protection Engineer, a Fire Reconstructionist from Fay Engineering, and the Sargent.& Lundy Fire Protection Engineer'. The fire area was maintained intact until the forensic investigation was ,

completed. Physical inspection was performed to determine fire zones, temperatures and durations based on established melting, burning, or softening temperatures associated with the variety of materials found in the area and exposed to the fire environment. This inspection formed the initial fire curves.

.These curves were then revised based upon oil fire test data from tests done in Finland. Calculations based on post-fire hydraulic oil inventory and fire duration indicated that approximately one-fifth of the expelled oil actually burned.

This is consistent with plant personnel statements about a large pool of unburned hydraulic fluid on the floor below the fire area. Although a literature review was performed by Sargent & Lundy to determine if mathematical fire modeling 7 techniques could be utilized to predict a temperature profile of the fire, it was found that there are no existing mathematical fire modeling techniques for a "running liquid fire" such as occurred at Fort St. Vrain.

7.2 Fire Maps ,

The results of these investigations formed the basis for the development of detailed fire maps that were used to determine equipment repair and replacement efforts as represented on the following pages. ____ _

s i t t

i i

i p

]

- - w -=---

1 l 1 ,I i

a :

I! - <

i.

= .

-- m. . .e .-

O 1

• ii if-li

,Me, I;'L..si*

fi . .i. isii.s.t.!iiii!ilii!ii!'E!s! ,,.

, , . . . . . . t . . s ,.

II !!i 2 p xa-B s s aisisi!!!! !!!!!! iiiiiiii LIE liDi l 9 e 6

e N

' U

.J s

-e- ---------_---__- __----_ ,~ ~- ' f J

s n i i

g x u0 v~

'N- 3.,

---__---___-__--____-___, y\ ""E

- -. ~

g y

{ / f m ~--~ N $i WE l

s

---

__/--s_L____ q <

f 1

7 s ii s>

c

\ h /1h _ _/ _[ / r r.:l

"~

J w E c-II" Vs. / W" k '

- ---___ __(-_- D _ d,__/

\ /

N: gi O

\ / H '+

\ - ' @

N ' ,

s__- W

\

,_p hs N 4

4

? i e

i 1

i l

so s-32* e-O

-@- i i

T el

-@ cstareinco rino mra a :NIwT toeg seittmar i , , ,_ _, n. s 749

. <>, m

/[r '4[

8 l

/ i \ 8 a

e3: $

4

<.. dscyn tuon .,

,/ /

i 8

g\ l

' '5' =

-g ..'n l

l l \s \

i I

.ro sp 3errj, s fwr 8 s

\

\ ,

8 t

n = sag ion..

f,, i \ ,

I m 1;22=,,

' neer,s.g.ucx s i

' \ '

"" PJ?

// \\,i w "" ==.v-e- I/ i  !' N' N

-bzwr i \i

'z>a e

l . 3, ' E' X

• I i

' ys a .u.. m Z..

es,r,a.a

/

y , tu55 N 1,'4fI27s \

/ /'

3 3

- s.6e r$ N I  ; \' '

j,l j

-1

\k

<i, l

~+5

" ?=:~

l l js ,

, is .

\

\'s n

.,,, x :

8

/ ,/ l

' ' i ),I l a'a ib

_ _ _ _-.l _ ___ _ _' r.I 28 o

,',,,' _P g r" **

aMF as, 3

/ _ '$

/ /

J m,,.o .,-s..oume O rtoon rim ncvariou ais -s to sw-s-

s - r mn z or .

N 5 5 J  :

3

~

$ 1

@ 3 & -

g " ' ** * = u s o.

g . I i s a. he h

e- 1. . . i, I, & Ib, 1:. .s!.t.i rIII k '!E.!I $I IE$~i.c.!I

,,. 1,5$$!$.!$

.i. I I . n a,. $ ! 3il iO u-bh355iBs5iB!55$!!!$$ 555$5555 58 !k 9 Z$

9

/ __ _

/  %

[  %

N

/ m

\

[ h~_

• _e_.,-_ i

__ .~.__

________ \ w

[

~ - __

g a Y _i s

5 $'3 ~

"a "y

t' R O' Re; &

C.&,

1 Q

_ u; Aa- _ _ _ _r _c .{. _ _.- __ ._ _ _ _ . l pc a

4: a_

p

/

-y g-

/

- u w h

/ / [7 k

/ f

/ b,--_ -

d ru@Z

_p-_ ___- _- _ ____; g. 3-

_3 g> .

( ,

- ( - t in 1

( t - 88, 9~ \ e e c, ,

Ok Z

\$

v

! N. k --

IR  %

sN N ,w. 4 b

%  % s ' -z .,

3 g c. H ss _ _' _ _ _ _7_w s_ .gg e g. _ _ _%.._.s e

%a 3

_.fa-.!.a. _ Sg s/ 3

~ _$3

~.

\

W h0D rk

\ a'e3 -

a- @

1 '

J u

_,,. \ .e N I

's  %

% I

/

S 1

t i

2 ~

i al 3 3

~

-ill IM y

( )

• A d. .

.d. ~ ,s. g cs , y 5  ; *****

.~ c . .

g y . m .

El g,

Ii 1.i ,

ii* N '! ! ili~;c;i!i i i !ii.lli.!i i EE:s:

, ,[;.e,.l1:i.si

'!i O '

MIS 5EBisiBii5555555 !5555555

. .=..s. . ... . 1, .$. . . aa. Il , ne EG iiE55 3 9 4

,- s s 9

/ N

/ N N

I N

\

+4------ .------------- S p w l

. ~ .\ S.

3r um

.h

- 6 *- ~~

.- - - s #

' \ \ kG

- -y- -

' _ . s y, u$

- - - - ,. , - = x -

l"

/ f j' /~

o i

/ 62 3 y's h~

W' h / / sl i,:;)

I 6].J: h sz a W e5 tu y

. ;e....--

A ,$$ **

A z

---

.;. . ._-- ;g , .$.3a g>

~

r ik -

h .> I m a

[8

\s h t \(! a uEE 1

m s

N ,N.s'~  ! g5 4 y_----_

.= m ;3.. s;.- .

a

- ce yg N y ' i' E I

N 53 lj C '

N ' Eb x  ?'

V ' m

N u

+ ,- -

j 6

\ , --6 l

\ l

\

- N I 1

N l r

s

l

~ , ' ~ )

~ , '"** == sep ,,/  ;

4 e

t e i l

8.0 EQUIPMENT DAMAGE ASSESSMENT / CORRECTIVE ACTIONS 8.1 Combustion Products Analyses PSC initially performed analyses on the soot deposited by the fire. Chlorides were found to be less than 1% and sulfur was

~1ess than 2%.

Southwestern Laboratories performed re-analysis of the soot to determine leachable chlorides and sulfides. Leachable chlorides were found to be 951 ppm and the sulfide was evaluated as less than 200 ppm.

8.1.1 Clean-Up Methods Initial clean-up was performed using high pressure water with detergents added. Final clean-up for stainless steel components was accomplished by wiping down each component with demineralized water and alcohol.

8.1.2 -Electrical Contacts Inspections Control room electrical contacts were inspected and results are discussed in Section 5.3 of this report. '

Additionally, on November 17, 1987 an inspection was made of certain electrical components located in the Fort St. Vrain turbine building, The purpose of this inspection was to determine if any combustion by-products (or other foreign material) were evident on electrical components, circuit contacts and terminals.

A total of 10 wipe samples was taken. Each wipe sample was inspected under a 10 power optical comparator.

NOTE: For comparison purposes, base samples of dust from office furniture were taken.

These samples appear as light grey to dark brown under ten power magnification. A sample (#26) was taken from the exterior of a junction box located immediately west of the fire zone and at an elevation of 4821 feet. This junction box sample appears black in color.

All turbine . building samples taken from inside component enclosures with the exception of #22, appear to be grey or brown in color and free of the black material contained on sample #26. The components within the instrument air compressor auxiliary control panel (the source of sample #22) were inspected. A light film of grey material (dust) was noted, however, there were no visible deposits of materials on the relay contacts. The switch located on the front of-this panel (a SBM type) is structured such that material ingress around the exposed shaft is restricted.

As a result of these inspections, it was concluded that electrical component contacts and terminals within enclosures which are mounted within the turbine building and outside the fire zone show no visible deposits of black materials (scot).

8.2 Environmentally Qualified Equipment 8.2.1 Components Environmentally qualified components were evaluated from three perspectives: impact of peak temperatures on equipment in the fire area, the effect on qualified life of the materials in the fire area, and the ability of the equipment to withstand the effects of l

water spray. Components were analyzed on a case-by-l case basis. Equipment was investigated for internally I heat sensitive components and replacements made as necessary. Thermal Lag Analysis was performed for equipment to define exposure temperatures. Arrhenius calculations were then performed, assuming peak calculated temperatures for the fire duration.

Materials reviewed were: Buna-N, Polyurethane Varnish, Polyvinyl Chloride, Phenolic, Cross-Linked

! Polyolefin, Polyester Amide, Neoprene, Teflon, i Polyethylene, Viton, Kapton, Tefzel, Metal Film Resistors, Polythermaleze. Epoxy Glass, Epone, Polyvinylidene Fluoride, Hysol, Nylon 6/6, Glass Filled Melamine, Silicon Rubber, and Diallyl Phthalate. The qualified life as specified in the EQ binders was then reduced by this calculated time exposure at 120'F. Materials determined to be in need of replacement before the 1989 refueling outage were identified to construction for immediate removal and replacement.

This information is reflected in Engineering Evaluation EE-EQ-0065, Rev. A.

8.2.2 Cable Environmentally qualified cable was evaluated to address the thermal effects of the fire, the effects of fi re fighting water sprays and water cleanup, and to assess the resulting qualified life of the cable.

Temperatures at the cable insulation were determined using the fire maps, thermal lag techniques, and any other physical evidence on a case-by-case basis.

Based on information provided by the EQ binders, temperature ratings were compared to investigatory temperature profiles to determine the need for r2 placement. Arrhenius calculations were used to subtract the consumed life from the qualified life.

Replacements were performed as- necessary. This information is reflected in Engineering Evaluation EE-EQ-0066, Rev. A.

8.3 Concrete Concrete was evaluated both visually and with the Schmidt Hammer methodology. Minor surface cracks and minor surface dusting were observed.

Schmidt Hammer (qualitative compression strength) tests found the concrete compressive strength to be in excess of the 3500 psi specified. When struck by a small hatnmer, the concrete emitted a distinctive ringing sound, and no concrete spalling occurred, indicating that the concrete is satisfactory.

The details of these inspections are reflected in Engineering Evaluation EE-75-0002, Rev. A.

8.4 Structural Steel The structural steel in the fire zone was visually inspected with additional NDE performed on steel that was exposed to high temperature and possibly flame. The inspection, examination data and their evaluations are documented in Engineering Evaluation EE-75-0002, Rev. A. During inspection of a beam directly abcve the fire at HV-2292, both vertical and horizontal deformations of the beam were observed. The possible cause and the effect of the deformation on the structural integrity of the beam was analyzed. The conclusion is that the probable caus for the beam deformations was the fire, because a mechanism for operating loads to induce plastic deformation cannot be found. The strength of the beam was determined not to be degraded and it will continue to support its design loads. Beam straightening was not required as the relatively small deformations will not degrade the integrity of the piping, conduits and other items r apported by the beam.

The evaluations of the remainder of the structural steel members in the fire zone support the conclusion that these members were not damaged by the fire and the structural integrity of the turbine building has not been degraded.

8.5 Cobduit Microstructural examination of conduit revealed no evideace of high temperature damage to the carbon steel. Examina' ion of ~the galvanized layer at the 0.D. revealed several treas around the circumference of the conduit where slight raiting of the galvanized layer has occurred. The depth of the melting was approximately 0.002 inches, and is not considered detrimental to the life of the conduit. The slight melting of the galvanized coating indicates an exposure temperature of about 780 F, the melting point of zinc. This correlates with the fire curves developed for the affected fire area.

8.6 55fety Valves 3.6.1 Hot Reheat Safety Valves The section of the hot reheat line which was in the fire zone is the section which contains the six hot reheat safety valves. This section of the hot reheat line is seamed pipe and was shop fabricated at the Stearns-Roger Pipe Fabrication Plant. This shop fabrication included welding in the line connections which would ultimately be connected to the hot reheat saftty valves. When the shop fabricated section of the hot reheat line was completed and in place at Fort St. Vrain, the hot reheat safety valves were attached by welding in the field.

The sec. ion of the hot reheat line which was in the fire zone was insulated and covered. Since thermal shock would be minimal in this area, it was determined that the connection to the hot reheat safety valve, V-5226, which appeared to have been subjected to the greatest temperature, would be examined for possible fire related damage. This section of the hot reheat line, 8 1/4 L 52105-D6, would be the boundary where the insulation had been discontinued. Any damage which may have occurred from the fire, or the subsequent extinguishment, would most likely occur in this area.

Damage to the hot reheat line material as a result of the fire would be expected to be limited to material phase cransforaation or thermal cracking, (quench cracking). Cracking would be limited to the material's .,urface or possibly, slightly subsurface.

Defects of the material's surface and subse'. face would be observed by performing nondestructive eamination in. the form of flourescent magnetic. particle in;pection. Damage to the hot reheat line material, .

-by phase transformation, could be ruled c,ut as a result of the testing performed on the hot reheat safety valves. Thcse valves are attached to the hot reheat line but are not insulated which would result in their material being subjected to the most severe environment. The hot reheat safety valves yere examined by fluorescent magnetic particle inspection, hardness tests and nondestructive metallographic replication, and were found to have no apparent fire related damage. Therefore, the hot reheat line .las tested by fluorescent magnetic particle examination.

The hot reheat line, 8 1/4 L 52105-06, was stripped of its insulation to the connection weld of Line L5216 D-6, and cleaned by hand buffing to remove scale and surface debris. The entire surface was then fluorescent magnetic pc ticle inspected per Fort St.

Vrain Quality Control Inspection Manual, QCIM-24.

This type of inspection is very sensitive to any surface, and shallow subsurface, defects which would te expected if damage had occurred as a result of the rire. This inspection would have greater sensitivity to surface defects as compared to the original inspection which was applied to the material during criginal fabrication. This inspection found no rejectable defects on the material.

Based on the inspection performed on the hot reheat line, 8 1/4 L52105-06, it has b~en determined that the hot reheat line material is acceptable with no I deficiencies which could be attributed to the fire.

8.6.2 Main Steam Safety Valves 8.6.2.1 Background During start-up, main steam pressure is controlied at 1600 psig. Feedwater in the main steam piping is bypassed to the cold reheat header through the startup bypass flash tanks. The pressure control valve on this system is designed for flashing conditions. The steam enters the cold reheat header and the water is returned to the condenser. As reactor power is increased, the pressure in the mein steam header is increased to 2400 psig. The temperature of the feedwater in the main steam header will increase as reactor power is increased. When the feedwater in the main steam header reaches 662 F, the saturation temperature of water at .2400 psig, the . temperature will remain constant as the reactor power increases. During this time, there will be two phase flow in the main steam lines. As the reactor power increases, the quality of the steam improves and eventually begins to superheat. At approximately 100 F superheat, the main steam pressure control is transferred from the startup bypass flash tank to the main steam bypass flash tank. At this time HV-2292 is closed.

During the shutdown of the reactor, the main steam header remains on the bypass pressure control until the temperature falls below 800 F. At this time, the control is automaticaily transferred to the startup bypass control system. As the main steam temperature drops to the saturation temperature, t'e main steam header will begin to contain wet steam. As more heat is removed from the reactor, the main steam piping will go solid. This occurs shortly after a reactor scram. The time depends on the reactor power and the rate at which the heat is removed from the reactor following the scram. This is in accordance with the design of the system.

-39' i- __ _ _ .

8.6.2.2 Thermal Analysis Following- a reactor scram from full power, the main steam temperature peaks at 1030 F in 5 minutes. The temperature drops below 1000 F approximately 2 minutes after the peak. Wet steam is emitted from the superheater after 11 minutes and saturated steam after 14 minutes. The system sees a transient from 1030 F to 662 F in a period of 9 minutes. -

The temperature of the main steam piping dropped from 802 F to 615 F in 10 minutes following the scram from 27% power on 10/3/87. The temperature transient was not unusual, as the temperature transient was less severe than a scram from full power.

Therefore, it is concluded that the main steam safety valves did not experience an excessive temperature transient.

8.6.2.3 Overpressurization Analysis Gags are sometimes installed on main steam safety valves as necessary while going from a water system to steam due to excessive seat leakage caused by inadequate seatlag pressure. After steam pressure is sufficient to seat these valves, they are removed. Curing the fire and subsequent transient, valves V-2214 and V-2215 were gaggea. The remaining valve V-2216, with the lowest setpoint, remained operable.

1 I

l l

1

_ l

a :

3 This valve is capable of relieving 430,120 lb/hr of 1000 F, 2720 psig superheated steam. The t!me at which the maximum pressure at the economizer inlet -(which is well upstream of the steam genbrators) reached 2875 psig was tt- 20 minutes after

'B' Circulator steam trip was received..

This is the time when the relief valve would~

have lifted. The flow at 20 minutes was 0 lb/hr in Lcop II and 350,000 lb/hr in Loop I. Assuming the worst case where all the main steam bypass outlet valves that were open went closed, the one relief valve could have hand'ed the entire flow from the bciler feed pump. The outlet of the EES (Economizer-Evaporator-Superheater) probably consisted of heated water or steam near saturation. This relief valve is capable of relieving 6000 gpm as water or 430,120 lb/hr as superheated steani (equivalent to boiler feed pump flow of approximately 925 gpm).

The amount of saturated steam that can be relieved is between that of water and superheated steam. Therefore, no overpressurization of the steam generators or main steam lines occurred.

8.7 Reactor Building Filter Testing The reactor building filters were tested after the fire and found to be acceptable.

- --. c w

9.0 LICENSING ASSESSMENT 9.1 Appendix A Determinations Appendix A Fire _ Hazards Analyses evaluated the area of the fire as a high hazard possibility. Fire detecticn wasi enhanced with the addition of detectors above the hydraulically-operated valves. Suppression was enhanced with the addition of two 100' CO, hose reels.

The actual fire was enveloped by the existing Appendix A Fire

' Hazards Analyses and, therefore, demcastrated the adequacy of the analysis.

9.2 Appendix R Determinations The fire occurred in an area defined as a non-congested cable area, pSC Appendix R evaluations which are the subject of a draft NRC SER analyzed the effects of fire on shutdown /cooldown capability. It was realized that fi re in this area could affect the hydraulic and electric systems in both loops. There was a potential for an intarruption of forced circulation.

Locally actuated accumulators were specifically added to valves FV-2205 and FV-2206, Loop I and II feedwater flow control valves, to assure cooling water flow after an event of this nature as a result of these Appendix R evaluations.

The ader,uacy of the Appendix R analysis was demonstrated as the fire did eliminate electric control of these valves. FV-2206 was stroked to the open position with the local accumulator allowing feedwater flow to Loop II.

Initially, FV-2205 was thought to be damaged because it could ____ -_

not be stroked full open with the accumulator (it was in mid-position after the fire). However, a procedural error was found that, when eliminated, allowed the valve to stroke.

Therefore, both trains analyzed for use in Appendix R evaluations were functional.

Although the operators did not have to use the Appendix R trains, they were both operational and could have been used if normal ecoling modes could not be reestablished. This was verified by Operations persennel during the Alert. This I

event demonstrated that the Fort St. Vrair. approach to defense-in-depth was valid and enveloped by the Appendix R evaluations.

l l

l l

t t

i l

l

~ - -

9.3 Control Room Habitability The control room HVAC system functioned as designed when it switched to the radiation mode, which. takes minimum makeup.

from tne turbine building on receipt of a radiation alarm from the ventilation stack monitor. In this mode, the control room HEPA and charcoal . filters removed the smo'te 'that was being drawn in from the turbine building. This was-demonstrated in tnst the auxiliary electric room, which is served by the same HVAC system,-had no Halon discharge. The breathing air system worked as designed, and Scott air pak; were also available for the control room, had they been necessary.

The modifications performed on the control room HVAC system will assure that a positive control room pressure relative to the turbine building will be maintained, thus precluding recurrence of the smoke ingress that occurred on October 2-3, 1987.

9.4 Hydraulic Oil System Assessment An analysis was performed by GA in GA-A13887 to dttermine the readiness and adequacy of System 91, the hydraulic oil system. PSC reviewed the design bases for the rapid stroke times for valves served by System 91. Valves actuated by this system are required for normal operation and control.

Accident scenarios require fast acting valves to mitigate the consequences of an EES (evaporator-economizer-superheater) or hot reheat tube leak or a HELB (High Energy Line Break).

SLRDIS (Steam Line Rupture Detection and Isolation System) incorporated the current valve cloture times in the development of the temperature profiles used to determine environmental equipment qualification. From the GA analysis and the design bases review, 1c was concluded, in part, that selected valve stroke timot could be relaxed, but not sufficiently to justif.y elimination of System 91, since these valves required stroke times available only through the use of a hydraulic sy:tei,

L 9.5 Interruption of Forced Circulation (IOFC) As,essment An 10FC occurred on October 3, 1987, from 27% power for approximately 23 minutes.

The reactor core at Fort St. Vrain has been an'1vzed and ,

determined to be capable of withstanding an inter' Jp sion of forced circulation for 90 minutes after a reactor scram from 100% power, without significant fission product barrier degradation. For a permanent loss of forced circulation scenario where the reactor trip is postulated to occur while operating at or below 35% of rated power, analyses have shown that no significant fission product barrier degradation occurs if the PCRV liner cooling system is utilized to cool the core and the PCRV.

Therefore, the 10FC that occurred during the fire for approximately 23 minute- from 27% power is an analyzed condition within the licensing basis of FSV.

i 1

l

- e o . . - - -

10.0 OVER^LL TESTING PROGRAM PSC utilized a "start-up book" to assure overall system coerability prior to restart. This book identified all Technical Specificatioi, requirements, applicable tests, operational prerequisites, and 5 included specific authorization signat' ares for control of pre-startup and rise to power activities. Routine ;tartup operational requirements were, therefore, supplemented by .the enhanced administrative controls provided by the start-up book.

PSC's testing program verified Technical Specification operational compliance. This program assured that repairs and modifications a c tere satisfactorily completed, system function rtquiremeat; ,are adhered to, and Technical Specification surveillances were validated. Upon completion of the fire recovery rep?f r/ modification efforts, a combination of cold checkout tests, functional tests, surveillance tests, post maintenance tests, and specialized tests were performed to assure functional acceptability prior to returning the componefit or system to service. A cross-reference list of Technical Specification surveillance requirements to the tests performed, applicable to equipment repaired or modified,*,sas prepared and made available for NRC review prior to restart, l

11.0 LONG TERM ENHANCEMENTS PSC has committed to perform evaluations for each of the following areas ~ con 31dered as long term enhancement items for Fort St. Vrain.

I n' accordance with NRC letter .from Callan to Williams, dated

. December 8, 1987, PSC will provide results-of these evaluations to 4 tha NRC on or before March 10, 1988.

• Evaluate Replacement of 011 Filter Canisters: This item refers to the procurement of new filter canisters to replace ones in the hydraulic system that have pipe wrench-marks.

Evaluate Removal of Thermal Relief Valve: This item indicates a possible approach to reduce hydroalic oil leakage, Enhance Pre-Fire Plans- This item refers to upgrading fire bricade effectiveness.

• Evaluate Fire Detection System Enhancements: This item refers to a complete analysis of the present system to determine improvements.

Evaluate _S.uppression 1eeds fur Hvdr:ulic 011 Hazards: This item i: *

"iessci,3-iearned" approach to assess meuns of suppressing hydraulic oil fires beyond the Fire Mazard Analysis of this topic.

Evaluate Functional Testina of 6 GPM Flow Cor'.rol Valves:

This item refers to the procurement of appropriate

! instrumentation and subseqtient testing of the 6 gpm isnw control valves in System 91.

• Evaluate Hydraulic 011 Catch Basin Enhancements: This item L _ _ - - _

l refers to the evaluation of the hydraulic oil catch basins l for the possible inclusion of flame arrestors and a level alarm to detect abnormal system low flow leakage.

( Evaluate Hot Surfacas Adjacent to Hydraulic System: This

! item refers to the analy:is of hot surfaces (2 500 F) within a 10-foot sphere of hydraulic valves for possible shielding or shroudirg.

• Evaluate Building 10 vs. Control Room HVAC: This item refers to possible enhancements to the control rcom or Building 10 HVAC systems in the event of a fire in Building 10.

• Evaluate Replacement of F-/604 Charcoal Filters: This item refers to the change-out of control room filter cartridges which were not available prior to restart, r

f l k l

I

~

12.0 SUPPORTING DOCUMENTATION 1.' Preliminary Report on the Impact of the FSV October ?nd Fire, dated October 30, 1987.

2. NRC letter, Callan to Williams (G-87387), dated October 29, 1987.

3. NRC memorandum, Heitner to Calvo (G-87402), dated November 9, 1987.

4. NRC memorandum, Martin to Murley (G-87418), dated November 17, 1987.

5. PSC letter, Williams to Calvo (P-87414), dated November 23, 1987.

6. PSC letter, Williams to NRC (P-87419), dated December 3, 1987.

7. NRC memorandum, Crutchfield to Milhoan (G-87437), dated December 7, 1987.

8. NRC letter, Callan to Williams (G-87425), dated December 8, 1987.

9. PSC letter, Fuller to Gammill (P-78182), dated November 13,

.978.

10. Updated Final Safety Analysis Report. P9 vision 5.

11. Fire Protection Program Plan, Revision 0, issued December 15, 1987.

12 PSC Fai'ure Analysis of the Filter Assembly Bowl from HV-2292, Laboratory Report No. 136, dated October 26, 1987.

13. GA-A13887, FSV HTGR System 91 Hydraulic System Report of Engineering Review, dated April, 1976, revised June, 1976.

14. Paul-Munroe ENERTECH letter, Beaufort to Harmon(PPS 3951), dated October 27, 1987.

15. EE-75-003, Revision A, dated December 3, 1987.

16. Fay Engineering Report, dated October 26, 1987.

17. Sargent & Lundy Report, dated October 23, 1987.

18. Public Service Company of Colorado Chemistry Department Ar.alysis ( Andrew Howell), dated No < ember 16, 1987.

19. Southwestern Laboratories Report No. 87-2289, dated December 8, 1987.

. , , , , . . -. . - - - . _- .. ~ - . . . - . .-

5

20. EE-EQ-0065, Revision A, dated December 4, 1987.

.21. EE-EQ-0066, Revision A, dated December 7, 1987.

22. EE-75-0002, Revision A, dated November 25, 1987.

23. QAC-87-1186, dated October 27, 1987.

24. 0AC-87-1303, dated November 24,'1987.

25. QAC-87-1418, dated December 21, 1987.

F6.. NCS Corporation Report, dated October 19, 1987.

27. QAC-87-1142, dated October 20, 1987.

28. QAC-87-1349,' dated November 20, 1987.

29. NCR-87-514.

30. NCS Corporation Report, dates tested: October 21-26, 1987.

31. NRC letter, Heitner to Williams (G-87454), dated December 18, -

1987.

l 4

l .

, n,---w- - - ,,-c, -

ry- , ,.. ,,-g-- ,---.,, , , , _ , , , ., . , , - , - - , - - - , , , . , _ _ _ , . . , , y ,--.-gn,, , .,,.--,,-,new <g 4y - .. g y a q g--4

INDEPENDENT REYlEW TORM DOCUMENT NUMBER: P-88C16 COMMITMENT DUE DATE: Jan. 15. 1988

SUBJECT:

Final Fire Reoort INSTRUCTIONS: Please perform an independent review of the sections identified, indicate concurrence or changes and sign and dote.

MARK APPROPRIATE COLUMN RECOMMENDED SECTIONS CHANGES AS CONCUR SIGNATURE & DATE IDENTiflED J. Williams .o,12 gp q hg g D. Evans 2.0,10.0

( ,g g f/f y/g 3.1,3.2,3.2.1,3.3.2, P. Hermon 3.3.3,3.3.4,3.3.5, w% ,e,~l n 3.4.1,3.4.2,3.5 3.3.1,8.1,8.5,8.6.1 3.4.3,3.4.4,6.0,7.0 g l T. Johnson 4.0,8.6.2 gg7gg/3mw g3jgg, D. G1enn s'.3 '.

[pggg h) /jlQp l

l G. Lacasse 5.1.3,5.1.5,8.7 S>ce/dW fg3 g g,7 fly@ftwcMc2O gg -

N. Snyder 8.1.1 f_./4 _g I/

D. Brown 8.2,8.2.1,8.2.2

,ATM j- /l p/8Y rt< N . Em -

8. Gunnerson 8.3,8.4 4W # d a me // M#&

J. Selan 9.1,9.2,9.3,9.4,9.5 cevme C. s

/ /d87

., s, ht h~ / **

RETURN TO Jim Gramling/Will Dender by January 14, 1988 l