ML20129E963
| ML20129E963 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 10/22/1996 |
| From: | James Knubel GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20129E967 | List: |
| References | |
| 6710-96-2336, NUDOCS 9610280217 | |
| Download: ML20129E963 (10) | |
Text
. - ~. -. - ~.-
.. ~
. -. - ~.... ~.
4 i
r 1
i l
GPU Nuclear,Inc,
(
Route 441 South i
NUCLEAR Post Office Box 480 Middletown, PA 17057-0480 Tel 717-944-7621
(
e (717) 948-8005 i
October 22, 1996 j
6710-96-2336 i
j U. S. Nuclear Regulatory Commission Attn: Document Control Desk l
Washington, D.C. 20555 Gentlemen:
.I t
s
Subject:
Three Mile Island Nuclear Station, Unit I (TMI-1)
Operating License No. DPR-50
[
]
Docket No. 50-289 l
l Response to the Request for Additional Information Related to Thermo-Lag i
Associated Ampacity Derating Issues i
The purpose of this letter is to provide a response to the NRC request for additional information dated July 5,1996 regarding Thermo-Lag associated ampacity derating issues. Included with the l
response to the fifteen questions is a package comprised of a calculation and supporting documentation. The calculation utilized an increased electrical raceway fire barrier system I
ampacity derating factor of 32% and comparison of the revised ampacity values against the expected load currents. The documentation supports the assessment of the ampacity of power cables in trays protected with Thermal-Lag fire barriers.
j We will be available to provide any clarification or additional information which may be necessary j
by your continuing review of the ampacity derating issue.
(\\
4 l
Sincerely,
/
'h 9610290217 961022 ADOCK 050002 9 J. Knubel j
= $DR Vice President and Director, TMI
}
WGH j
Attachments cc: Administrator, Region I
.TMI Senior Resident Inspector i
j Sr. Project Manager NRC, TMI 4
k
6710-96-2336 Page 1 of 9 Response to the Request for Additional Information Related to Thermo-Lag Associated Derating Issues Item 1 The TMI documentation regarding the Thermo-Lag fire barrier issue is not well organized. It consists of attachments that do not seem to support one another and contains incomplete references which makes the logic of the study difficult to follow. The licensee should resubmit a more coherent package ofinformation that documents all aspects of the analysis, identifies how the various documents are used to support the overall assessment, and includes the cited references required to support its analysis.
Response to Item 1 The references in the following paragraphs are identified in the attached calculation, C-1101-770-E420-018.
At TMI-1, derating of control, control power and instrumentation cables is not considered because neither type of cable carries any significant load near its ampacity rating. These words are adopted from the FS AR (.Ref. 3.1l) section 8.2.2.11.b. in the fifth sentence. However, in cases where power cables are routed with control power cables, then the control power cables are considered because they can effect each others' ampacity.
At TMI-l there are 480 VAC or 4160 VAC power circuits that are protected by three-hour rated barriers as per the FHAR (Ref. 3.1). For 4160 VAC power circuits, the one-hour barriers are installed over cabic trays or over cables that are in cable trays. For 480 VAC power circuits, fire barriers are installed over cable trays or over conduit. Testing performed for Texas Utilities (TU) showed one-hour fire barrier derating factors of 31.6%
for cable trays and 10.7 % maximum for conduit (Ref. 3.8). The one-hour fire intrier configuration at TU plants bounds the one-hour fire barrier configuration at TMI-l (Ref 3.13). Conduit is not included in this calculation because cable tray derating associated with the addition of Thermo-Lag is considered the bourding case. Table 2 shows the comparison between the cable trays and conduit.
Table 2: Cornparison of Tray and Conduit Ampacities with One-hour Thermo-Lag Derating l
{
AWG Tray Ampacity w/o Tray Ampacity w/
Conduit Ampacity w/c Conduit Ampacity w/
Thermo-lag (A)
Thermo-Lag (A)
Thenno-Lag (A)
Thermo-Lag (A)
- 8 41 28 36 32 1
350 MCM 285 194 253 225 l
l l
Values in columns two and four of Table 2 are taken from Ref. 3.14 which is based on l
Ref. 3.6 for a 90 C conductor rating operating in a 40 C ambient when installed with a total of 7 to 24 current carrying conductors inclusive. Columns three and five are calculated by multiplying the respective values in columns two and four by 0.680 and I
6710-96-2336 Page 2 of 9 0.890 representing derating factors of 32% and 11% respectively. When comparing columns three and five, it can be seen that the ampacity for cables routed in tray protected by a one-hour fire barrier is significantly less than that for cables in conduit. Thus, the analysis of ampacity of cables installed in trays protected by one-hour fire barriers is bounding when compared to that for cables in conduit.
Calculation # C-1101-770-E420-018, attached, provides a more coherent package of information regarding the Thermo-lag fire barrier ampacity derating issues. The calculation identifies all documents used, all assumptions made, and references required to support the analysis.
Item 2 The licensee's package included a "TSI Derating Study Test Plan." It is not clear that the tests described have been or will be performed. Regarding the tests plan, provide answers to the following questions:
Have the tests called for in the plan been performed?
If not, will they be performed in the future, and if so, when will the test results be made available for review?
If the tests have been performed, provide documentation of the results and describe how the test results i: ave been factored into the TM1 ampacity assessment.
Response to Item 2 The tests called for in the "TSI Derating Study Plan" were not performed. We do not plan to do any testing because our analysis, calculation # C-1101-770-E420-018, uses test results from TU/TVA configurations which bound our configuration.
j Item 3 The Cycle 6 cable sizing criterion does not indicate the basis of ampacity values included in each table. It appears that the licensee did not use the ICEA P-46-426 tables for ampacity values. The licensee should document the source ofits base ampacity values including any corrections to the base ampacities due to temperature, cable diameter variation, and number of conductors, etc.
_R_es.ponse to item 3 The Cycle 6 cable sizing criteria (Rev.1) page 2 paragraph 6 does give the basis for the base ampacity values by stating that " Specific cable manufacturer's free-air ampacity shall be used in lieu ofIPCEA P-46-426 table ampacities.. " The Cycle 6 cable sizing criteria
l a
6710-96-2336 i
Page 3 of 9 i
has ampacity tables for Kerite and also for Rockbestos manufactured cables. Calculation C-1101-770-E420-018 documents the source of the base ampacity values including any corrections to the base ampacities due to temperature, number of conductors in the tray, s
l and the effect of Thermo-Lag fire barrier material.
l Item 4 4
The base cable ampacities used by the licensee are not consistent with either NEC or ICEA P-46-426 ampacity tables. The ampacity-values should be reconciled or discrepancies between the licensee-cited values and those values documented in nationally j
recognized standards should be explained.
l Response to Item 4 j
The base cable ampacities used are not those given in the NEC or IPCEA P-46-426 2
because base ampacities from the manufacturer (Kerite) are used for original cables supplied in the late 1960's. Kerite has given free-air ampacity values for 90 C conductor a
temperature operating in a 40 *C ambient. Since ampacity values given by the i
manufacturer are taken to be the most accurate values available, reconciliation with NEC or IPCEA P-46-426 is not required.
I Item 5 l
)
It appears that the licensee used ICEA P-54-440 to determine cable tray ampacity. These calculations lack the documentation ofICEA tables from which the values are derived, method for calculating cable depth or fill, and corrections due to temperature, depth of fill and cable diameter, etc., to the base values.
Response to Item 5 ICEA P-54-440 was applied in some of the cases as an alternate verification of the ampacity values calculated by IPCEA P-46-426 methodology. Where ICEA P-54-440 was used, in some cases more margin was indicated and in other cases less margin was indicated. Our Updated FSAR is based on using the Cycle 6 cable sizing criteria (fiection 8.2.2.11 andfef. 8.6.2 of FS AR) which is based on IPCEA P-46-426. Thus we used IPCEA P-46-426 methodology for the attached calculation # C-1101-770-E420-018.
Calculation # C-1101-770-E420-018 uses the following: for ambient temperature, maximum HVAC design ambient temperature is used vice using 40 C across the board, for number of conductors, number of conductors is used vice number of cables, and for Thermo-lag a derating factor of 32% is used vice 28.04%. In general, the resulting ampacity values are lower and more conservative ampacity values than those calculated in the previously submitted analysis. For three cables (CG83, LP2 and LP6) the maximum expected load current exceeds the revised ampacity when using the IPCEA P-46-426 methodology. For a closer evaluation, the 1996 National Electric Code (NEC) Article
6710-96-2336 Page 4 of 9 318-11 methodology is applied to tray 590 which contains the three cables. The resulting revised ampacity for the three cables is greater than the maximum expected load current.
Item 6 The licensee needs to provide sufficient information regarding the physical and electrical characteristics including manufacturer, number of conductors, cable configuration (e.g.,
single conductor, three conductor, triplexed, shielded, insulated and jacketed cable, etc.),
cable outside diameter, voltage rating, conductor size, etc., of the Thermo-Lag protected
~
cables used at TMI.
Response to item 6 The requested information is provided in the attached calculation # C-1101-770-E420-018 Table 3.
Item 7 The documentation of the cable tray 590 temperature measurement experiment is considered inadequate. Test documentation should include the description of the test procedure, documentation of experimental methods and instrumentation, and at least a minimal demonstration of quality control over the experiments. This information should be provided if this, or other in-plant experiments, are to be credited.
Response to Item 7 We agree on further review that Attachment 2 of GPU Nuclear letter to NRC dated March 29,1995 page 9 and 10 does not represent adequate documentation for a credible test. The material submitted is no longer applicable since our current analysis, calculation
- C-1101-770-E420-018, is based on testing results from TU/TVA configurations which bound the TMI configuration and eliminate the need for further testing.
Item 8 The licensee has assumed an ampacity derating factor of 28.04 percent for a 1-hour Thermo-Lag cable tray fire barrier. This value does not reconcile with other Thermo-Lag ampacity derating test results using IEEE Standard Procedure P848, " Procedure for the Determination of the Ampacity Derating of Fire Protected Cables." The licensee should reconsider its analysis using more reasonable estimates of the ampacity derating impact of a Thermo-Lag fire barrier system or explain the basis for variation in parameters. (For example, Texas Utilities found a value of 32 percent for a nominal 1-hour Thermo-Lag 330-1 cable tray barrier system.)
)
=
1 6710-96-2336 Page 5 of 9 Response to Item 8 i
We have reconsidered our position on TSI derating in light of the latest information available and because the TU configuration bounds our configuration we are using the latest TU number; 32%.
Item 9 The licensee did not describe the fire barrier system for trays 551/553. A description of the fire barrier system for these trays should be provided.
Response to Item 9 The fire barrier for trays 551 and 553 is a 1-hour TSI wrapped fire barrier similar to all other trays wrapped with a 1-hour fire barrier, Item 10 The licensee's analysis appears to incorrectly interpret the cable tray multiple conductor j
derating correction to refer to a count of cables rather than conductors in the case of multi-conductor cables (see note 3 ofTMI Cycle 6 Tables X and XI). This appears in direct conflict with the guidance of the NEC Handbook which quite clearly indicates that the count should be based on the actual number of power cable conductors. The licensee should reconsider its analyses using the actual count of power cable conductors as the basis for multiple conductor derating.
Response to Item 10 We have reconsidered our position and calculation #C-1101-770-E420-018 determines ampacity values for cables in tray using the number of conductors sice using the number of cables resulting in lower, more conservative ampacity values.
Item 11 The licensee needs to provide a definitive technical basis to support the assessment of cable ampacity for those cables which are overloaded over the equipment life cycle. The licensee should indicate what measures will be taken to monitor for signs of accelerated age-related degradation.
Response to item 11 Based on the attached calculation no cables are overloaded over the equipment life cycle.
Consequently, measures for monitoring for signs of accelerated age-related degradation will not be required.
~
l i
6710-96-2336 Page 6 of 9 l
Item 12 i
The constant KVA loads will draw 11 percent more current at 90 percent of rated voltage available at its terminals. Additionally, some loads may operate at overload or at a service factor of 15 percent. Accordingly, the full load current (FLA) could be as high as 125 percent of FLA at nominal voltage. The licensee needs to address this aspect of system
)
operation in its analysis.
Response to Item 12 j
Full Load Current (FLA) for constant KVA loads are based on measured or calculated load values. For the circuits listed below,125% of the FLA is greater then the derated ampacity of the cable. The affect of reduced voltage was evaluated by either determining the FLA for the motor at the minimum expected grid voltage during normal operation, or determining the FLA at 90% of the motor nominal voltage. For other Thermo-Lag protected constant KVA load cables,125% of full load current is well within the derated cable ampacity. Since measured or calculated FLA were used operate at the limit of the motor service factor was not considered.
Circuit Component Nameplate De-rated Load Current Load Current Number FLA Ampacity
@ Nominal
@ Degraded Voltage Voltage LS6 NS-P-1B 140 amps 147 amps 144 amps" U) 146 ami2)
LP6 NS-P-1 A 140 amps 161 amps 141 amps
- 143 amps
- LS5 NS-P-lC 140 amps 147 amps 144 amps 146 amps
- m CG83 IC-P-1A 87 amps 105 amps 82.7 amps
- 91.9 amps
- CH61 IC-P-1B 87 amps 107 amps 82.7 amps 91.9 amps
- LP2 DC-P-1 A 120 amps 133 amps 103.2 amps
- 106 amps
- CL43 AH-E-7A 105 amps 89 amps 66 amps
- 75 amps
- CM43 AH-E-7B 105 amps 89 amps 63.5 amps
- 72 amps
- Note (1) Load Currents from TDR 995 (Ref. 3.9.2), Table I A-A and Table I A-B Note (2) NS-P-1B is modeled as not mnning in TDR 995 (Ref. 3.9.2), normal operation 2 of 3 pumps are running, used load current value for NS-P-lC.
Note (3) Load current values from Lotus Notes from Tom Akos to Dick Bensel dated 9/9/96 (Ref. 3.9.5). Currents are based on a grid voltage of 232 KV.
Note (4) Load current from TDR 836, Rev. 6 (Ref. 3.9.1), Table 1 A, Operating Load at 100% Power at Nominal Voltage and 100% Power at Degraded Bus Voltage. Degraded Bus Voltage values are calculated by applying a multiplier of 0.9 to the equipment rated voltage (460 VAC).
4
.., _ ~
.. -.. - ~. -. ~ -
i j
6710-96-2336 Page 7 of 9 i
Note (5) Shutdown during normal operation, ES load operates only for test 5g, j
load current at nominal voltage from TDR 836, Rev. 6 (Ref. 3.9.1j, Table IP, Operating Load LOCA at Nominal Bus Voltage. Load current at j
degraded voltage from TDR 995, Rev. 3 (Ref. 3.9.2), Table 4A, LOCA loads grid at 225.6 KV.
I Note (6) Reactor Building Purge Exhaust Fans, FLA values from Memo # MSS-l 86-079 (Ref. 3.9.4). One fan in operation during refueling outages or Reactor Building entries.. Purge is operated less than 500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br /> per year.
l Degraded Bus values based on FLA corrected to 90% of motor nominal voltage.
I Note (7) Load current values from Lotus Note from Tom Akos to Dick Bensel dated 9/11/96(Ref. 3.9.5). Currents are based on a grid voltage of 232 KV.
Item 13 The maximum load current of transformer (1000/1333 KVA) is based on breaker setting of 185 amperes. Breaker setting has a tolerance of +/- 10 percent. As a result the maximum current seen by the cable shall be 203.5 amps (185 + 10 percent). Provide justification why 185 amperes shall be selected instead of 203.5 amperes.
Response to Item 13 Using the breaker setting of 185 amps instead of the 203.5 amps is conservative, since the normal loading on the transformers is significantly less than the breaker setting. The normal loading for the affected cables is as follows:
Circuit No.
Transformer Measured Load
- Calc. Load MB11 IH Bus Transformer < 70 amps 176 amps
- MB13 1U Bus Transformer < 10 amps 56 amps
- MCl2 IM Bus Transformer < 50 amps 178 amps
- ME11 IT Bus Transformer 50 amps 70 amps *
.MDI 1 1R Bus Transformer 80 amps 69.4 amps
- Note (1) Based on 4 KV feeder breaker ammeter reading obtained on 9/03/96 with the plant at 100% power. The IH Bus and 1G Bus are supplied through a
)
shared breaker therefore the measured load is the total load for both 1
buses. The 1M and,1L Buses are supplied through a shared breaker therefore the measured load is the total load for both buses.
~
i
6710-96-2336 Page 8 of 9 Note (2) The IH Bus load of 176 amps is a calculated value based on cold weather space heating load (see Calculation No. C-1101-770-E420-018, Appendix 1). The IM Bus load of 178 amps is a calculated value based on cold weather space heating load (see Calculation No. C-1101-770-E420-018, Appendix 1).
7 Note (3) The 1U Bus load of 56 amps is based on Post Cooling Tower operation and operation of other connected loads (see Calc. No. C-1101-770-E420-018, Appendix 1).
Note (4) Load values from TDR 836, Rev. 6 (Ref. 3.9.1), Operating Load at 100%
Power.
Item 14 The licensee should consider either the load amperes flowing through cables based on breaker setting with a positive 10 percent tolerance (i.e.,110 percent of breaker setting) or the actual amperage of the load in its analysis.
Response to Item 14 The attached calculation considers the actual amperage of the connected load which is either based on field measurements or calculation. The calculation includes references for the source of the load current values. The loading for affected transformers under all l
anticipated operating conditions is less than the nominal trip setting for the breaker. The calculation therefore used the actual / anticipated load to evaluate the derated ampacity for these circuits.
4 Item 15 Circuits MA9 and MB9 for SR-P-3 A and SR-P-3B are both on tray 756. If both A and B pumps are on the same tray, how has the separation criteria been met. Circuits MB13, MEl1, and ME10 share both trays 751 and 756. The licensee should provide a discussion on how the analysis is consistent with the cable separation criteria as described in FSAR Section 8.2.2.12.
i s
I
6710-96-2336 Page 9 of 9 Response to Item 15 Circuits MA9 and MB9 for SR-P-3 A and SR-P-3B are Balance of Plant (BOP) non-Engineered Safeguards circuits, therefore separation criteria for redundant circuits does not apply. Circuits MBl1 and MB13 are feeder circuits to BOP 480 VAC Buses. Circuit MEIO for RR-P-1B is appropriately routed in Engineered Safeguards C1.annel "B" trays.
Plant circuit routing criteria allowed BOP circuits to be routed in Engineered Safeguard trays for one channel providing it never entered a tray associated with a redundant channel.