ML17345A998: Difference between revisions

From kanterella
Jump to navigation Jump to search
(Created page by program invented by StriderTol)
(Created page by program invented by StriderTol)
Line 60: Line 60:


I
I
                          '
.122- 272F /112173/ml  s FIGURE 1 DATA LINK INTERCONNECTION DTE                                                          DCE Pin  g    Function                  function Pin        g AA                  1        GHD                      GHD            1      AA BA                  2        TXD                      TXD            2      BA BB                  3        RXD                      RXD            3      BB CA                  4        RTS                      RTS                    CA CB                  5        CTS                      CTS            5      CB CC                  6        DSR        Fiber Opti c  DSR            6      CC AB                  7        SIG  GND      Cabl e      SIG  GHD        7      AB CD                  8        DTR                      DTR            8      CD CF                  20        Carrier                  Carrier        20      CF Detect-                  Detect gSPDS  Serial    Fiber Optic                      ~
.122- 272F /112173/ml  s FIGURE 1 DATA LINK INTERCONNECTION DTE                                                          DCE Pin  g    Function                  function Pin        g AA                  1        GHD                      GHD            1      AA BA                  2        TXD                      TXD            2      BA BB                  3        RXD                      RXD            3      BB CA                  4        RTS                      RTS                    CA CB                  5        CTS                      CTS            5      CB CC                  6        DSR        Fiber Opti c  DSR            6      CC AB                  7        SIG  GND      Cabl e      SIG  GHD        7      AB CD                  8        DTR                      DTR            8      CD CF                  20        Carrier                  Carrier        20      CF Detect-                  Detect gSPDS  Serial    Fiber Optic                      ~
Fi ber Opti c          SAS Line Adapter          Modem                                Modem              Cotmuni-cati on Multi -,
Fi ber Opti c          SAS Line Adapter          Modem                                Modem              Cotmuni-cati on Multi -,
Line 93: Line 92:
5.1.2        Interface Comands The  internal  'gSPDS  data link interface cards shall accept and implement as    a  mininum the following processor commands:
5.1.2        Interface Comands The  internal  'gSPDS  data link interface cards shall accept and implement as    a  mininum the following processor commands:
: a. Separate    Interrupt Enable/Disable/Disarm      Commands  for both 0        b.
: a. Separate    Interrupt Enable/Disable/Disarm      Commands  for both 0        b.
Transmit and Receive, Data Terminal Ready
Transmit and Receive, Data Terminal Ready (CD),
                                                .
(CD),
: c. Request    to  Send  (CA)'.
: c. Request    to  Send  (CA)'.
                                                                                             ~ I'
                                                                                             ~ I'
Line 105: Line 102:
   /2F/ f12173/ml 5 I
   /2F/ f12173/ml 5 I
l
l
                           . PREFERRED  ALARM STATUS BYTE CONFIGURATION LSB  1-    HI            (High Limit Alarm) 2-    LO            (Low  Limit
                           . PREFERRED  ALARM STATUS BYTE CONFIGURATION LSB  1-    HI            (High Limit Alarm) 2-    LO            (Low  Limit Alarm)'Failed 3    FAIL                    Sensor) 4  - BAD'          (Bad Data  -              Out of  Range)
                -
Alarm)'Failed 3    FAIL                    Sensor) 4  - BAD'          (Bad Data  -              Out of  Range)
                 - SUSPCT        (Suspect Data )
                 - SUSPCT        (Suspect Data )
6  - QSPTRB        (gSPDS  trouble) 7  - SET TO  1'To      Avoid Confusion with                  GS)
6  - QSPTRB        (gSPDS  trouble) 7  - SET TO  1'To      Avoid Confusion with                  GS)
Line 203: Line 198:
C ~
C ~
1 C2- 272F /112 173/ml s NOTES TO CROSS      REFERENCE TABLE    (Continued)
1 C2- 272F /112 173/ml s NOTES TO CROSS      REFERENCE TABLE    (Continued)
,
Byte 42 Bit  1      Reactor  Vessel level 7 (Coolant/No Coolant)
Byte 42 Bit  1      Reactor  Vessel level 7 (Coolant/No Coolant)
Bit  2      Reactor Vessel Level 8 (Coolant/No Coolant)
Bit  2      Reactor Vessel Level 8 (Coolant/No Coolant)
Line 277: Line 271:
     " ICE-272F/112173/mls I
     " ICE-272F/112173/mls I
CROSS  REFERENCE TABLE CHANNEL B  (Continued)
CROSS  REFERENCE TABLE CHANNEL B  (Continued)
V SSAGE                                                                                VALUE NUMBER                  POINT ID              DESCRIPTION                    (GIVEN IN RANGE)      UNITS DT1B          Differential  HJTC Temperature Level    1                  -2268  to  +2268      oF DT2B          0ifferential  HJTC Temperature Level    2                  -2268  to  +2268 37                      DT38          Differential  HJTC Temperature Level    3                  -2268 to +2268        oF 38                      OT48          Differential  HJTC Temperature  Level 4                  -2268  to  +2268 39                      DT5B                          HJTC    'ifferential Temperature  Level  5                  -2268  to  +2268 40                      OT6B          Differential  HJTC Temperature  Level  6                  -2268  to  +2268      oF OT7B          Differential  HJTC Temperature  Level  7                  -2268  to  +2268      oF 42                      DTSB          Differential  HJTC Temperature  Level 8                  -2268 to +2268        oF
V SSAGE                                                                                VALUE NUMBER                  POINT ID              DESCRIPTION                    (GIVEN IN RANGE)      UNITS DT1B          Differential  HJTC Temperature Level    1                  -2268  to  +2268      oF DT2B          0ifferential  HJTC Temperature Level    2                  -2268  to  +2268 37                      DT38          Differential  HJTC Temperature Level    3                  -2268 to +2268        oF 38                      OT48          Differential  HJTC Temperature  Level 4                  -2268  to  +2268 39                      DT5B                          HJTC    'ifferential Temperature  Level  5                  -2268  to  +2268 40                      OT6B          Differential  HJTC Temperature  Level  6                  -2268  to  +2268      oF OT7B          Differential  HJTC Temperature  Level  7                  -2268  to  +2268      oF 42                      DTSB          Differential  HJTC Temperature  Level 8                  -2268 to +2268        oF 43                    "
          "
43                    "
PC18          Heater Power Control Signal 1                                0  to  100 44                      PC2B          Heater Power Control, Signal 2                                0  to 100 45                      Q1HIB          CET Highest Temp Quad-1                32 to 2300 46                    , Q1HIOB          CET Highest Temp IO (Quad-1)            0 to 10 47                      Q1NHIB        CET Next Hi ghest Temperature Quad-1                      32  to  2300 GET'ext Hi ghest t
PC18          Heater Power Control Signal 1                                0  to  100 44                      PC2B          Heater Power Control, Signal 2                                0  to 100 45                      Q1HIB          CET Highest Temp Quad-1                32 to 2300 46                    , Q1HIOB          CET Highest Temp IO (Quad-1)            0 to 10 47                      Q1NHIB        CET Next Hi ghest Temperature Quad-1                      32  to  2300 GET'ext Hi ghest t
48              .      Q1~2u~
48              .      Q1~2u~
Line 301: Line 293:
Pa ge  28 of 32 oF  * ~ i
Pa ge  28 of 32 oF  * ~ i


      '
1CE-272F/112172/m1s l ~ l' CROSS  REFERENCE  TABLE CHANNEL 8  (Continued)
1CE-272F/112172/m1s l ~ l' CROSS  REFERENCE  TABLE CHANNEL 8  (Continued)
MESSAGE                                                              VALUE NUMBER          POINT ID            OESCRIPTION              (GI VEN IN RANGE)        UNITS 74              CET58 F3    Core  Exit  Temperature F3      32  to  2300            oF 75              CET108 HB    Core  Exit  Temperature H8      32  to  2300 76              CET78 F9    Core  Exit  Temperature F9      32  to  2300 77              CET208 EB    Core  Exit  Temperature E8      32  to  2300 78              CET28 810    Core  Exit  Temperature 810    32  to  2300 79              CET18 85    Core  Exit  Temperature 85      32  to  2300 80              CET158 K11 Core    Exit  Temperature    K11  32  to  2300            'F 81              CET118 H15 Core    Exit  Temperature    H15  32  to  2300            oF 82              CET228 H13 Core    Exit  Temperature H13,    32  to  2300 83              CET218 H9    Core  Exit  Temperature H9      32  to  2300 84              CET48 E14    Core  Exit  Temperature E14    32  to  2300 85              CET38 E12    Core  Exit  Temperature E12    32  to  2300 oF'F 86              DUMMY 2B    Dummy Ya,lue 87            TMARAB        Loop A  RCS Temp  Sat. Margin  -2100 to 700
MESSAGE                                                              VALUE NUMBER          POINT ID            OESCRIPTION              (GI VEN IN RANGE)        UNITS 74              CET58 F3    Core  Exit  Temperature F3      32  to  2300            oF 75              CET108 HB    Core  Exit  Temperature H8      32  to  2300 76              CET78 F9    Core  Exit  Temperature F9      32  to  2300 77              CET208 EB    Core  Exit  Temperature E8      32  to  2300 78              CET28 810    Core  Exit  Temperature 810    32  to  2300 79              CET18 85    Core  Exit  Temperature 85      32  to  2300 80              CET158 K11 Core    Exit  Temperature    K11  32  to  2300            'F 81              CET118 H15 Core    Exit  Temperature    H15  32  to  2300            oF 82              CET228 H13 Core    Exit  Temperature H13,    32  to  2300 83              CET218 H9    Core  Exit  Temperature H9      32  to  2300 84              CET48 E14    Core  Exit  Temperature E14    32  to  2300 85              CET38 E12    Core  Exit  Temperature E12    32  to  2300 oF'F 86              DUMMY 2B    Dummy Ya,lue 87            TMARAB        Loop A  RCS Temp  Sat. Margin  -2100 to 700
Line 346: Line 337:


I
I
'
   'J
   'J


Line 440: Line 430:
   ~
   ~
l      I
l      I
    ,


   ~
   ~
Line 463: Line 452:


I 3.
I 3.
    -
Turkey Point Operating History and Plans Since startup (Unit 3    = 1972>  Unit  0 = 1973),    both units have operated on annual cycles (with one exception for each plant). Half of these cycles used conventional fuel management (fresh fuel on the periphery) and the other half used "standard" low-leakage fuel management.
Turkey Point Operating History and Plans Since startup (Unit 3    = 1972>  Unit  0 = 1973),    both units have operated on annual cycles (with one exception for each plant). Half of these cycles used conventional fuel management (fresh fuel on the periphery) and the other half used "standard" low-leakage fuel management.
As of january 1, 1983, Unit 0 has accumulated          6.37 (Figure 3.1) EFPY and Unit  3 slightly less. Subsequent  to steam generator replacement at both units, 18 month operating cycles are planned. Annual cycles will be used only when contingencies necessitate    it. Figure  3.2 illustrates this schedule. Unit 3  Cycle 8  started up in April 1982 and is an eighteen month cycle. As of this date (February 8, 1983), an annual Cycle      9  for Unit  0  is intended because of schedular constraints.
As of january 1, 1983, Unit 0 has accumulated          6.37 (Figure 3.1) EFPY and Unit  3 slightly less. Subsequent  to steam generator replacement at both units, 18 month operating cycles are planned. Annual cycles will be used only when contingencies necessitate    it. Figure  3.2 illustrates this schedule. Unit 3  Cycle 8  started up in April 1982 and is an eighteen month cycle. As of this date (February 8, 1983), an annual Cycle      9  for Unit  0  is intended because of schedular constraints.
Line 486: Line 474:
Lt lIT 3
Lt lIT 3
'YCLE  8        CYCLE 9                    CYCLE 10                -
'YCLE  8        CYCLE 9                    CYCLE 10                -
CYCLE  ll
CYCLE  ll CYCg 9                  CYCLE    10              CYCL  11          CYCLE 12 1983                        1985                1986              1987      1988 FIGURE    3,2
                                                                -
CYCg 9                  CYCLE    10              CYCL  11          CYCLE 12 1983                        1985                1986              1987      1988 FIGURE    3,2


FLORIDA POWER  R LIGHT CO, TURKEY POINT DESIGN BASIS PERIPHERAL POWER DISTRIBUTION P~)      ,93    ,77 1,12    ,80
FLORIDA POWER  R LIGHT CO, TURKEY POINT DESIGN BASIS PERIPHERAL POWER DISTRIBUTION P~)      ,93    ,77 1,12    ,80
Line 557: Line 543:
were full length. The small axial extent of needed flux reduction, however, indicates that part length poison rods can be just as effective with a lesser decrease in overall radial power. Part length BPs would, therefore, assist in mitigating the loss in operating margin for a given level of flux reduction.
were full length. The small axial extent of needed flux reduction, however, indicates that part length poison rods can be just as effective with a lesser decrease in overall radial power. Part length BPs would, therefore, assist in mitigating the loss in operating margin for a given level of flux reduction.
The impact of the near term design changes on the axial power shapes is illustrated in Fig. 5.5. The use of'spent fuel on the core flats has      a large advantage    compared    to the generic power shape by shifting the powers upwards, away from the critical weld in addition to the expected reduction in axial peaking. This factor results in about a 1096 decrease    in critical weld flux in addition to the decrease in radial power.
The impact of the near term design changes on the axial power shapes is illustrated in Fig. 5.5. The use of'spent fuel on the core flats has      a large advantage    compared    to the generic power shape by shifting the powers upwards, away from the critical weld in addition to the expected reduction in axial peaking. This factor results in about a 1096 decrease    in critical weld flux in addition to the decrease in radial power.
-
   ~  Combining the radial powers and the axial shapes        results in the powers plotted in Fig. 5.6. The expected impact of implementing these changes is given in Fig. 5.7. The design changes planned for Cycle    9  of Unit 3 and  Cycle 10  of Unit  0 correlate with Curve C on Fig. 5.7 which indicates that the screening criterion would be reached in August 2000.        Assuming no further changes, dummy assemblies could be used beginning in 2001 to reach licensed lifctime.
   ~  Combining the radial powers and the axial shapes        results in the powers plotted in Fig. 5.6. The expected impact of implementing these changes is given in Fig. 5.7. The design changes planned for Cycle    9  of Unit 3 and  Cycle 10  of Unit  0 correlate with Curve C on Fig. 5.7 which indicates that the screening criterion would be reached in August 2000.        Assuming no further changes, dummy assemblies could be used beginning in 2001 to reach licensed lifctime.
These changes,    however, are not without penalty.      Increases  in hot spot peaking (~F) and radial channel peaking (Fz H) are expected.        In addition, compared to designs without these changes, core reactivity is lost. In future cycles, this will be recovered by increasing the amount of U-235    load/  in the core. These penalties are summarized      in Table 5.1. Table 5-2 lists the expected RTNDT values associated with the near term design changes.
These changes,    however, are not without penalty.      Increases  in hot spot peaking (~F) and radial channel peaking (Fz H) are expected.        In addition, compared to designs without these changes, core reactivity is lost. In future cycles, this will be recovered by increasing the amount of U-235    load/  in the core. These penalties are summarized      in Table 5.1. Table 5-2 lists the expected RTNDT values associated with the near term design changes.
Line 617: Line 602:


I-I ORIM POWER      .", LICiIIT CO.
I-I ORIM POWER      .", LICiIIT CO.
TURKEY POVIT tlat'I! T    i!-
TURKEY POVIT tlat'I! T    i!-
VESSEL    I LL>E!!CF VS. VESSEL LIFF A            B          C            D 8/95        V2000      8t2(6}        2035 rr r
VESSEL    I LL>E!!CF VS. VESSEL LIFF A            B          C            D 8/95        V2000      8t2(6}        2035 rr r
Line 720: Line 704:
   ~
   ~
     ) "
     ) "
          !
~      )
~      )



Revision as of 22:27, 3 February 2020

Rev 0 to Interface Design Requirements for Qualified Safety Parameter Display Sys/Safety Assessment Sys Data Communications for Turkey Point Units 3 & 4.
ML17345A998
Person / Time
Site: Turkey Point  NextEra Energy icon.png
Issue date: 01/21/1983
From: Earles J, Feeney M, Foster R
ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY
To:
Shared Package
ML17345A993 List:
References
16081-ICE-3111, 16081-ICE-3111-R, 16081-ICE-3111-R00, GL-82-28, NUDOCS 8303160313
Download: ML17345A998 (150)


Text

~ ~ ss "

I CE- 27 2F /11217 3/ml s

/d1.C'A nn C 4 W. G INTERFACE DESIGN REQUIREMENTS FOR QSPDS/SAS DATA COtOUNICATIOHS FOR FLORIDA POWER AND LIGHT COMPANY TURKEY POINT UNITS HO. 3 AHD 4 REQUIREMENT NUMBER 16081-ICE-3111, REVISION 00

.Huclear Power Systems COMBUSTION ENGINEERING, INC.

Minds or,. Connecti cut Prepared by Date ar es (Microprocessor roducts Independent Review by Date Mi croprocessor Products Approved by k Cg. Date / 20-gg R. G. Foster Supervisor, Microprocessor Products )

Approved by Date =

3.d

. Pucak Manager, nstrumentation Systems Desi gn)

Approved by ates Project Manager 0 ~Mi&

This document is the property of Combustion Engineering, Inc. (C-E),

Windsor, Connecticut and it is to be used only for the purposes of the agreement with C-E pursuant to which it is furnished.

~ .

'I'ssue Date ':

~

':I'/21/83.'303i603i3

'I

' ~ ~ s c 8303i0 Page 1 of 32 PDR ADQCK 05000250 P PDR

.iL 5-4/zw/ L I 6 I /4/m1s RECORD OF REYISIONS PAGES PREPARED IHDEP END EHTLY NO. INVOLYED

'Y.

REYIEWED BY APPROYALS 00, 1/21/83 All 0. M. Earles M. H. Feeney R. G. Foster J. L. Pucak T. P. Gates

~ .,'. L

~

~

Requirement t/o. 16081-ICE-3111 Revision 00 Page 2 of 32

I

~I

'ICE-272F/112173/ml s TABLE OF CONTENTS Section Ho. Title'Pa e Mo.

1.0 PURPOSE 2.0 SCOPE 3.0 APPLICABLE REFERENCES 5 EHGIHEERIHG DOCUMENTS 5 3.2 CODES 5 3.3 STANDARDS 5 4.0 FUNCTIONAL DESIGN REQUIREMENTS 6 4.1 INFORMATION TRANSFER REQUIREMENTS 6 4.2 DATA TRANSFER RATE 6 4.3 ELECTRICAL DESIGN REQUIREMENTS 7 5.0 OP ERATIONAL REQUI REMENTS 10 5.1 INTERFACE CONTROL 10 5.2 COl@UH I GATI ON PROTOCOL 12 6.0 0 IAGNOSTI C TEST REQUIREMENTS LIST OF TABLES Table No. Title Paoe No.

1 CROSS REFERENCE TABLE - CHANNEL A 17 2 CROSS REFERENCE TABLE - CHANNEL B 25 s ' ~ ~ '  %, ~

,, ~ (

Requirement No. 16081-I CE-3111 Revi si on 00 Page 3 of 32

272F / f 12173/ml s k; I

'PURPOSE is document provides the criteria governing the digital interfaces between the gualified Safety Parameter Display System (QSPDS) and the Safety Assessment System (SAS) for Florida Power and Light Company's Turkey Point Units No. 3 and 4.

The. interface design requirements presented herein are tntended to def ine both the functional and operati onal r equi rements for data comnunications between gSPDS and SAS. Hardware and software requirements are established to complete the specification and design of the interface.

SCDPE The gSPDS/SAS interface shall consist of full duplex digital data links between the two gSPDS processors and the SAS processor.

~>> ~ ~

ai rement lfo. 16081- ICE-3111 Revision 00 Page 4 of 32

I

.122- 272F /112173/ml s FIGURE 1 DATA LINK INTERCONNECTION DTE DCE Pin g Function function Pin g AA 1 GHD GHD 1 AA BA 2 TXD TXD 2 BA BB 3 RXD RXD 3 BB CA 4 RTS RTS CA CB 5 CTS CTS 5 CB CC 6 DSR Fiber Opti c DSR 6 CC AB 7 SIG GND Cabl e SIG GHD 7 AB CD 8 DTR DTR 8 CD CF 20 Carrier Carrier 20 CF Detect- Detect gSPDS Serial Fiber Optic ~

Fi ber Opti c SAS Line Adapter Modem Modem Cotmuni-cati on Multi -,

ple xor gSPDS Cabinet The RXD to TXD, CTS to RTS and DSR to DTR interchanges are done by the modems. Therefore from computer to modem no interchange is required, and there is a one-t'o-one connection as shown above. The above configuration diagram assumes that the OSPDS and SAS computers are configured as Data Terminal Equipment (DTE).

~ J, ~

Requi rement No'. 16081- ICE-3111 Revision 00 Page 8 of 32

~ 1CE 272F/112173/ml s

~ gs I I

Des i onati'on .

Def i ni ti on

. AA Overall Shield (Prot ctive Ground)

AB . Si gnal Ground BA Transmit Data (TXD)

BB Recei ve Data (RXD)

CA Request.to Send (RTS,)

CB Clear to Send (CTS)

CC Data Set Ready (DSR)

CD Data Terminal Ready (DTR)

CF Carrier On The interconnection of these signals is shown in Figure 1. Signal characteristics are defined by the EIA Standard RS-232-C (Reference 3.3.1).

0 I

~

r ~

s Requir'ement Ho. 16081-ICE-3111 Revi si on 00 Page 9 of 32

I II

ICE-272F/112173/ssl s

~ s

'5.0 OPERATIONAL REQUIREMENTS 5.1 INTERFACE CONTROL s

A.l.l A~d There shall'e two consecutive device addresses for each of the

(}SPDS/SAS data links; one for receive and one or transmit. Each address shall have separate interrupt control logic associated with it.

5.1.2 Interface Comands The internal 'gSPDS data link interface cards shall accept and implement as a mininum the following processor commands:

a. Separate Interrupt Enable/Disable/Disarm Commands for both 0 b.

Transmit and Receive, Data Terminal Ready (CD),

c. Request to Send (CA)'.

~ I'

~

" ~ ~ ~ " ~

. - ~

d Requi rement No. 16081-1l.'E-3111 Revision 00 Page 10 of 32

/2F/ f12173/ml 5 I

l

. PREFERRED ALARM STATUS BYTE CONFIGURATION LSB 1- HI (High Limit Alarm) 2- LO (Low Limit Alarm)'Failed 3 FAIL Sensor) 4 - BAD' (Bad Data - Out of Range)

- SUSPCT (Suspect Data )

6 - QSPTRB (gSPDS trouble) 7 - SET TO 1'To Avoid Confusion with GS)

MSB 8 - PARITY (Odd Parity)

Exp lanati ons:

Failed Sensor- Equipment associated with the sensor P

has fai led.

Bad Data- Sensor input is outside the valid range for the sensor..

Suspect Data- Calculated results which were affected/revised due to bad data or failed sensor being present.

The convention "1" = alarm/failed condition and "P" =

normal /operational condition will be employed.

SIGNAL VALUE Signal value can be any number represented by 1 to 8 ASCII characters.

Ex: 2000.2 is represented by 6 ASCII characters including the decimal point.

dRaa~ '-

iremen No. 16081-ICE-3111 Revision 00 Page 14 of 32

I ~

0

l ~

I ICE-272F/112173/mls

(

CROSS REFERENCE TABLE CHANNEL A (Continued)

MESSAGE VALUE NUMBER POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS 49 Q2HIA CET Highest Temp Quad-2 32 to 2300 oF 50 Q2HI DA CET Highest Temp ID (Quad-2) 0 to 10 51 Q2NHIA CET Next Highest Temperature Quad-2 32 to 2300 52 Q2NIDA CET Next Highest Temperature ID (Quad-2) 0 to'10 53 Q3HIA CET Highest Temp Quad-3 32 to 2300 54 Q3HIDA CET'i ghest Temp I D (Quad-3) 0 to 10 55 Q3NHIA CET Next Highest Temperature Quad-3 32 to 2300 56 Q3NIDA CET Next Highest Temperature ID (Quad-3) Oto 10 57 Q4HIA CET Highest Temp Quad-4 32 to 2300 oF 58 Q4HIDA CET Highest Temp ID (Quad-4) 0 to 10 59 Q4NHI A CET Next Highest Temperature Quad-4 32 to 2300 60 Q4NID A CET Next Highest Temperature ID (Quad-4) Oto10 61 CET26A P7 Core Exit Temperature P7 32 to 2300 oF 62(6) CET3A E7 Core Exit Temperature E7 (Nl 1 ) (Nl1 ) 32 to 2300 oF 63 CET25A N10 Core Exit Temperature N10 32 to 2300 oF 64 CET24A N8 Core Exit Temperature N8 32 to 2300 'F 65 CET20A L6 Core Exit Temperature L6 32 to 2300 66 CET7A K8 Core Exit Temperature K8 32 to 2300 'F 67 CET23A M3 Core Exit Temperature M3 32 to 2300 68 CET18A H5 Core Exit Temperature H5 32 to 2300 oF 69 CET17A H3 Core Exit Temperature H3 32 to 2300 oF

.CET14A=G2'Core Exit: Tempe'r'ature G2:::.'."' .32,to 2300:

~

.1 ~ '70, 71 '.:'" . CET2A E'4': Core Exit Temperaturo4. ' 32 to 2300 oF Requi remend No'. 16081-ICE-3111 Revision 00 Page 20 of 32

I

.I CE-272F/ I 12173/ml s

- PREFERRED ALARM. STATUS BYTE CONFI GURATION LSB 1- HI (High Limit Alarm) 2- LO (Low Limit Alarm) 3>> FAIL (Failed Sensor )-

4 - BAD (Bad Data - Out of Range) 5 - SUSPCT (Suspect Data )

6 - gSPTRB ((}SPDS t roub le )

7 - SET TO 1 (To Avoid Confusion with GS)

MSB 8 - PARITY (Odd Pat ity)

Exp lanati ons:

Failed Sensor- Equipment associated with the sensor has fai led.

Bad Data- Sensor input is outside the valid range for the sensor..

Suspect Data- Calculated results which were affected/revised due to bad data or failed sensor being pr'esent.

The convention "1" = alarm/failed condition and "P" =

normal /operational, condition will be employed.

SIGNAL YALUE Signal value can be any number represented by 1 to 8 ASCII characters.

Ex: 2000.2 is represented by 6 ASCII characters including the decimal

, point.

Requirement No. 16081- ICE-3111 Revision 00 Page 14 of 32

.I CE-272F/ I 12173/ml s

. GROUP SEPARATOR'roup Separator {GS) is sent to the SAS to indicate the end of message packet. An acknowledge (ACK) or no acknowledge (NAK) ASCII character .

is sent to gSPDS by the SAS after every message packet. If an ACK is not received by the gSPDS, the message packet is retransmitted up to a maximum of two (2) times before declaring and tagging the'ata link as failed. The gSPDS will consider parity, framing, and overrun errors as NAKs in that the last data link transmission will be repeated following the above protocol.

5.2.2 Messa e Block Format Message block consists of the message packets. Approximately every 1 to 2 seconds,, gSPDS transmits the entire Message Block to the SAS.

The Message Block,has the following format.

STX Message Packet Messa ge ETX CHK EOT Packet N The Message Block starts with start of text (STX) charac er, followed by message packets,and ending with End of Text character (ETX),

checksum {CHK, which is an Exclusive Or of all the data bytes between ETX and STX excluding the control characters GS) and End of Transmi ssi on { EOT) character.

1

'I

~ ~ ~

Requir'ement No. 16081-ICE-3111 Revision 00 Page 15 of 32

I 'l Pt 0

': 1CE-272F/112173/ml s A

7 6.0 ... DIAGNOSTIC TEST REQUIREMENTS The QSPOS/SAS data link diagnostic checks shall be responsible for detecting serious failure of the data link hardware. This shall be accomplished by checking the status of the data link hardware and checking the number of HAKs (or incorrect responses) received consecutively from the SAS. If more than 3 NAKs {or incorrect responses) are received consecutively the data link betwe n QSPDS and SAS is tagged as failed and the error condition is alarmed on the pIasma display unit. Mhen a failed data link is detected the transmission is stopped by the QSPDS for the present scan cycle. 'The transmission of data from the QSPDS to the SAS is restarted the next scan cycle. If a NAK/ACK is not recei ved within 3 seconds after a message packet is sent, the data link is tagged as failed and alarmed on the plasma display unit. The QSPDS tries to establish corrnunication again with SAS the next scan cycle. The QSPOS continuously searches for the operation of the data link every 3 seconds until the link becomes operational. The QSPDS will consider parity, framing, and overrun errors as HAKs in that the last data link transmission will be repeated following the above protocol.

Requi rement No. 16081- ICE-3111 Revision 00 Page 16 of 32

'I

1 CE-272F/112173/ml s CROSS REFERENCE TABLE CHANNEL A (Continued)

MESSAGE VALUE NUMBER POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS 49 Q2HIA CET Highest Temp Quad-2 32 to 2300 oF 50 Q2HIDA CET Highest Temp ID (Quad-2) Oto 10 51 QZNHIA CET Next Highest Temperature Quad-2 'i 32 to 2300 oF 52 Q2N IDA CET Next Highest Temperature ID (Quad-2) 0 to 10 53 Q3HI A CET Highest Temp Quad-3 32 to 2300 oF 54 Q3HIDA CET Hi ghest Temp ID (Quad-3) Oto 10 55 Q3NHIA CET Next Highest Temperature Quad-3 32 to 2300 oF 56 Q3N IDA CET Next Highest Temperature ID (Quad-3) Oto 10 57 Q4HIA CET Highest Temp Quad-4 32 to 2300 oF 58 Q4HIDA CET Highest Temp ID (Quad-4) 0 to 10 59 Q4NHIA CET Next Highest Temperature Quad-4 32 to 2300 oF 60 Q4NID A CET Next Hi ghest Temperature ID (Quad-4) Oto 10 61 CET26A P7 Core Exit Temperature P7 32;,to 2300 oF 62(6) CET3A E7 Core Exit Temperature E7 (N11) (N11) to 2300 63 CET25A N10 Core Exi t Temperature . N10 32 to 2300 64 CET24A N8 Core Exit Temperature N8 32 to 2300 'F 65 66 CET20A L6 CET7A K8 Core Core Exit Exit Temperature L6 Temperature K8

'2 32 32 to to 2300 2300 oF oF l

67 CET23A M3 Core Exit Temperature M3 32 to 2300 68 CET18A H5 Core Exit Temperature H5 32 to 2300 oF 69 CET17A H3 Core Exit Temperature H3 32 to 2300 a'70 CET14A: G2 'Corse:Exit:,Temperatur.e 'GZ "..:..'."' 32 to 2300:

71 CET2A E'O': Core Exit Temper'ature'4. 3'2 to 2300 Requirement No. 16081-ICE-3111 Revision 00 Page 20 of 32

I

" I C"=-272F/I 12173/m1 s CROSS REFERENCE TABLE CHANNEL A (Continued)

MESSAGE VALUE NUMBER POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS 72 CET10A 03 Core Exit Temperature 03 32 to 2300 73 CET15A G8 Core Exit Temperature G8 32 to 2300 74 CET12A E10 Core Exit Temperature E10 32 to 2300 75 CET11A 05 Core Exit Temperature 05 32 to 2300 oF 76 CET9A C12 Core Exit Temperature C12 32 to 2300 77 CET8A C8 Core Exit Temperature C8 32 to 2300 78 CET1A A8 Core Exit Temperature A8 32 to 2300 oF 79 CET22A L14 Core Exit Temperature L14 32 to 2300 oF 80 CET21A L12 Core Exit Temperature L12 32 to 2300 81 CET6A 012 Core Exit Temperature J12 32 to 2300 82 CET5A J10- Core Exit Temperature JIO 32 to 2300 oF 83 CET19A Hll Core Exit Temperature Hll 32 to 2300 oF 84 CET16A G15 Core Exit Temperature G15 32 to 2300 0 F'F 85 CET13A F13 Core Exit Temperature F13 32 to 2300 86 CET4A Fll Core Exit Temperature Fll 32 to 2300 oF Sat. Margin -2100 to 700 OF 87 THARAA Loop A RCS Temp 88 PHARAA Loop A RCS Press Sat. Margin -3000 to 3000 PSI to 700( ) OF 89 TMARBA Loop B RCS Temp Sat. Margin -2100 90 PMARBA Loop B RCS Press Sat. Margin -3000 to 3000 PSI to 700( OF 91 THARCA Loop C RCS Temp Sat. Margin -21OO 92 PMARCA Loop C RCS Press Sat. Margin PSI 93 (3) Reactor Vessel Leve1 1 0<1(3) Coolant through 8 Status Message / No Packet Cool ant Requi rement'o. 16081-ICE-3111 Revision 00 Page 21 oi 32

I

~ '

1CE-272F/112173/ml s 1

I NOTES TO.CROSS REFERENCE TABLE (1) + sign indicates subcooling.

- sign indicates superheat.

(2) Message Number. is a 2 ASCII character nurser."a It varies from 00 through Sge (3) Reactor Vessel Level 1 throu h 8 Status Messa e

~Data B te Odd Parity Bit MSB LSB 8 7 6 5 4 3' 1

~Bte al Bit 1 Reactor Vessel Level 1 (Coolant/Ho Coolant)

Bit 2 Reactor Vessel Level 2 (Coolant/No Coolant)

Bit 3 Reactor Vessel Level 3 (Coolant/No Coolant )

Bit 4 Reactor Vessel Level 4 (Coolant/Ho Coolant)

Bit 5 Reactor Vessel Level 5 (Coolant/Ho Coolant )

Bit 6 Reactor Vessel Level 6 (Coolant/Ho Coolant)

Bit 7 Set.to "1"

~ ~

v v

Requirement Ho. 16081-ICE-3111 Revision 00 Page 22 of 32

C ~

1 C2- 272F /112 173/ml s NOTES TO CROSS REFERENCE TABLE (Continued)

Byte 42 Bit 1 Reactor Vessel level 7 (Coolant/No Coolant)

Bit 2 Reactor Vessel Level 8 (Coolant/No Coolant)

Bit 3 Set to "P" by gSPDS - should be ignored by SAS computer.

Bit 4 Set to "9" by gSPDS - should be icnored by SAS computer.

Bit 5 Set to "P" by gSPDS - should be ignored by SAS computer.

Bit 6 Set to "P" by gSPDS - should be ignored by SAS computer.

Bit 7 Set.to "1" P indi cates presence of coolant.

1 indi cates absence of coolant or no coolant.

Bit 7 of these data bytes will be set to "1" (as shown) to avoid confusion which may arise by the SAS deciphering this byte as a 'group separator'.

(4) For durmrry values, the integer -format will be,erssployed. 'An example is:

gP. Integer format is detailed in. note 5.a.

(5) Format of Anglo Values a) Integer type: The field width is the size of the maximum range of the value plus 1 for a sign.,Positive values have a blank in the sign position, negative values have a minus sign in the sign position. The numeric field is .leading zero suppressed, replaced by blanks. If the value is zero, the right most position will contain a zero.

Example: If saturation margin, range +700 to -2100 F, is 50'F the transmitted data is 5+50. If it is -10'F, the transmitted data is -+10. If it is zero, the transmitted data is ++0.

r t

";. where g

~ ~ '..'"

is ASCII. space.:

. ~ . ':': '..'.'

(blank-);:.';.-:: .

-:.'..':;::. - ".*.'.

Requirement Ho. 16081-ICE-3111 Revision 00 Page 23 of 32

I (I

0

1CE-272F/112173/ml s I

r NOTES TO CROSS REFERENCE TABLE (Continued) b) Exponential Format: Above a value of 10 and below a value of 1000, the integer format {described above) will be used. For the other values, the field width is 8 characters as follows:

a.aaa+bb, where a.aaa is the fractional part of the value and

+bb is the exponential part.

(Note: no sign information is transmitted since the data is always ti posi ve. )

For example: 1.23$ power is transmitted as 0.123 + 01.

6) CET E7 is for Turkey Point Unit No. 3.

CET Nll is for Turkey Point. Unit No. 4.

I

~... ', ',;, ~ ~ ", r e'I ~ ~

e Requirement No. 16081-1CE-3111 Revi si on 00 Page 24 of 32

1

" ICE-272E/I12173/mIs

(

" CROSS REFERENCE 'TABLE CHANNEL B MESSAGE VALUE(5)

NUMBER< ) POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS 00 THOT1B Hot Leg Temp Loop A 0-750 'F Ol THOT2B Hot leg Temp Loop B 0-750 02 TCOLD1B Cold Leg Temp Loop A 0-750 03 TCOLD2B Cold Leg Temp Loop B 0-750 ,,oF 04 PRESSB Pressurizer Pressure 0-3000 PSIA 05 THOT38 , Hot Leg Temp Loop C 0-750 oF 06 'TCOLD3B Cold Leg Temp Loop C 0-750 07 THEADB Upper Head Temp 32-2300 08 TRCETB Representative Core Exit Temperature 32-2300 09 TMARHEADB -

Upper Head Temperature Satu rati on Margin -2100 to 700( ) 'F PMARHEADB Upper Head Pressure Saturation Margin -3000 to 3000(') PSI TMARRCSB Minimum RCS Tempera ure Saturation Margin -2100 to 700{')

12 PMARRCSB Minimum RCS Pressure Saturation Margin -3000~o 3000( ) PSI 13 TMARCETB Core Exit Temperature (CET)

Saturation Margin -2100 to 700( ) 'F 14 PMARCETB CET Pressure Saturation Margin -3000 to 3000( ) PSI TMARURB, RCS/Upper Head Temp Saturation Margin -2100 to 700((1)) D oF 16 RLEVHB Reactor Vessel Level - Head 0 to 100

, 17 RLEVPB Reactor Vessel Level - Outlet Plenum 0 to 100 18 DUMMY 1B Dungy Value 19 TU1B, Unheated HJTC Temperature r

.'2, .', .

',""-..',Le'v'el'.. 1";: .:,:..-:-..: ':.: '.32 to..2300

~ r

... Ot,.

i, ~

Requi re'ment No'. 16081-ICE-3111 Revision 00 Page 25 o 32

4 f

+ e

- 1CE-272F/112173/mls I

CROSS REFERENCE TABLE CHANNEL B MESSAGE VALUE NUMBER POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS 20 TU2B Unheated HJTC Temperature Level 2 32 to 2300 oF TU3B Unheated HJTC Temperature Level 3 32 to 2300 22 TU4B Unheated HJTC Temperature Level 4 32 to 2300 'F 23 TU5B Unheated HJTC Temperature Level 5 32 to 2300 24 TU6B 'nheated HJTC Temperature Level 6 32 to 2300 'F Unheated HJTC Temperature

'Level 7 32 to 2300 oF TU8B Unheated HJTC Temperature M

Level 8 32 to 2300 oF 27 TH1B Heated HJTC Temperature Level 1 32 to 2300 'F 28 TH2B Heated HJTC Temperature .

Level 2 32 to 2300 29 TH3B Heated HJTC Temperature Level 3 32 to 2300 'F 30 TH4B Heated HJTC Temperature Level 4 32 to 2300 oF e

TH5B Heated HJTC Temperature Level 5 32 to 2300 oF 32 THGB Heated HJTC Temperature

, Level 6 32 to 2300 33 TH7B Heated HJTC Temperature Level 7 32 to 2300 34 THBB Heated HJTC Temperature

~ .

":.;Lev,el: 8 ">:.::; ":.:;.":,:.'.". 32

/

ta.2300 ., oF Requirement No. 16081-ICE-3111 Revisi on'0 Paae 26 of 32

P l

I

" ICE-272F/112173/mls I

CROSS REFERENCE TABLE CHANNEL B (Continued)

V SSAGE VALUE NUMBER POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS DT1B Differential HJTC Temperature Level 1 -2268 to +2268 oF DT2B 0ifferential HJTC Temperature Level 2 -2268 to +2268 37 DT38 Differential HJTC Temperature Level 3 -2268 to +2268 oF 38 OT48 Differential HJTC Temperature Level 4 -2268 to +2268 39 DT5B HJTC 'ifferential Temperature Level 5 -2268 to +2268 40 OT6B Differential HJTC Temperature Level 6 -2268 to +2268 oF OT7B Differential HJTC Temperature Level 7 -2268 to +2268 oF 42 DTSB Differential HJTC Temperature Level 8 -2268 to +2268 oF 43 "

PC18 Heater Power Control Signal 1 0 to 100 44 PC2B Heater Power Control, Signal 2 0 to 100 45 Q1HIB CET Highest Temp Quad-1 32 to 2300 46 , Q1HIOB CET Highest Temp IO (Quad-1) 0 to 10 47 Q1NHIB CET Next Hi ghest Temperature Quad-1 32 to 2300 GET'ext Hi ghest t

48 . Q1~2u~

Temperature ID (Quad-1) 0 to 10 49 'ZHIB -

CET Highest Temp Quad-2

." "-.: '"Q2RIDB-'.:: 'ET:":Hi 32 to 2300 oF

.50:: ='"-::, ghest. Temp.,ID "(Quad-2) 0 to 10 -:."

Requi rement'o. 16081- ICE-3111 Revision 00 Page 27 of 32

~ k F

' 1'CE-272F/112173/ml s CROSS REFFRENCE TABLE CHANNEL 8 (Conti,nued)

HESSAGE VALUE NUHBER POINT ID DESCRIPTION (GIVEN IN RANGE) UNITS Q2NHI 8 CET Next Highest Temperature Quad-2 32 to 2300 Q2N ID 8 CET Next Highest Temperature ID (Quad-2) Oto 10 53 Q3HI 8 CET Highest Temp Quad-3 32 to 2300 54 Q3HID8 CET Hi ghest Temp ID (Quad-3) 0 to 10 55 Q3NHI 8 CET Next Highest Temperature Quad-3 32 to 2300 oF Q3NID 8 CET Next Hi ghest Temperature ID (Quad-3) . 0 to,10 57 Q4HI,B CET Highest Temp Quad-4 32 to 2300 'F 58 Q4HID8 CET Highest Temp ID (Quad-4) 0 to 10 59 Q4NHI 8 CET Next Highest Temperature Quad-4 to 2300 60 Q4NIDA CET Next Hi ghest Temperature ID (Quad-4) Oto 10 61 CET198 Core Exit Temperature R7 32 to 2300 oF

'F 62 PB Core Exit Temperature P8 R7'ET188 32 to 2300 'F CET178 N6 Core Exit Temperature N6 32 to 2300 'F 64 CET258 N4" Core Exit Temperature N4 32 to 2300 oF 65 CET248 Hll Core Exit Temperature Nll 32 to 2300 66 CET168 Mg Core Exit Temperature big 32 to 2300 oF CET238 LB Core Exit Temperature LB 32 to 2300 oF 67 68 CET148 K5 Core Exit Temperature K5 -'2 32 to 2300 oF 69 CET138 K3 Core Exit Temperature K3 32 to 2300 70 CET128 J2 Core Exit Temperature 32 32 to 2300 oF t 71 72 73, .'~'., ~ ~:

A CET98 G6 CETBB Gl CET6B-: F5 ...

Requirement No. 16081-ICE-3111 Core Core Exit Temperature Exit Temperature Cor.e;Exi't:Tempera.ture G6 Gl

.F,5.;

Revision 00 32 to 2300 32 to 2300

'32, to 2300':;,::.

~ I ~

Pa ge 28 of 32 oF * ~ i

1CE-272F/112172/m1s l ~ l' CROSS REFERENCE TABLE CHANNEL 8 (Continued)

MESSAGE VALUE NUMBER POINT ID OESCRIPTION (GI VEN IN RANGE) UNITS 74 CET58 F3 Core Exit Temperature F3 32 to 2300 oF 75 CET108 HB Core Exit Temperature H8 32 to 2300 76 CET78 F9 Core Exit Temperature F9 32 to 2300 77 CET208 EB Core Exit Temperature E8 32 to 2300 78 CET28 810 Core Exit Temperature 810 32 to 2300 79 CET18 85 Core Exit Temperature 85 32 to 2300 80 CET158 K11 Core Exit Temperature K11 32 to 2300 'F 81 CET118 H15 Core Exit Temperature H15 32 to 2300 oF 82 CET228 H13 Core Exit Temperature H13, 32 to 2300 83 CET218 H9 Core Exit Temperature H9 32 to 2300 84 CET48 E14 Core Exit Temperature E14 32 to 2300 85 CET38 E12 Core Exit Temperature E12 32 to 2300 oF'F 86 DUMMY 2B Dummy Ya,lue 87 TMARAB Loop A RCS Temp Sat. Margin -2100 to 700

-3000 to 3000( )

88 PMARAB Loop A RCS Press Sat. Margin PSI 89 TMARBB Loop B RCS Temp Sat. Margin QF 90 PMARBB Loop 8 RCS Press Sat. Margin .0.( ) PSI 91 TMARCB Loop C RCS Temp Sat. Margin -2100 t 700( ) OF 92 PMARCB Loop C RCS Press Sat. Margin, -3000 to 3000 PSI 93 (3) Reactor Vessel Level 1 O>1(3) Coolant through 8 Status Message / No Packet Coolant R equi rement N o. }5081- I CE-3111 Revision 00 Page 29 of 32

I 0

i.i ICE-272F/112173/el S

'NOTES TO CROSS'REFERENCE TABLE:.

(1) + sign indicates subcooling.

- sign indicates superheat.

(2) Message Number is a 2 ASCII character number. It varies from 00 through 89.

(3) Reactor Vessel Level 1 through 8 Status Hessa e

~Data B te Odd Parity Bit MSB LSB 8 7 6 4 3 2 1

~Bte,'1 Bit 1 Reactor Vessel Level 1 (Coolant/Ho Coolant)

Bit 2 Reactor Vessel Level 2 (Coolant/Ho Coolant)

Bit 3 Reactor Vessel Level 3 (Coolant/No Coolant)

Bit 4 Reactor Vessel Level 4 (Coolant/No Coolant)

Bit 5 Reactor Vessel Level 5 (Cool ant/Ho Coolant )

Bit 6 Reactor Vessel Level 6 (Coolant/No Coolant)

~ Bit 7 Set to "1" Requirement No. 16081- ICE-3111 Revision 00 Page 30 of 32

'I I h

4 I

'ICE-272F/I12173/ml s I

NOTES TO CROSS REFERENCE TABLE '(Continued)

Byte 42 Bit 1 Reactor Vessel Level 7 (Coolant/Ho. Coolant )

Bit 2 Reactor Vessel Level 8 (Coolant/Ho Coolant)

Bit 3 Set to "P" by QSPDS - should be icnored by SAS.computer.

Bit 4 Set to "P" by QSPDS - should be ignored by SAS computer.

Bit, 5 Set to "P" by QSPDS - should be ignored by SAS computer.

Bit 6 Set to "P" by QSPDS - should be i gnored by SAS computer.

Bit 7 Set to "1" P indicates presence of coolant.

1 indicates absence of coolant or no coolant.

Bit 7 of these data bytes will be set to "1" (as shown) to avoid confusion which may arise by the SAS deciphering this byte as a 'group separator '.

(4) For dunmy values, the integer format will be employed. An example is:

PP. Integer format is detailed in note-5.a.

(5) Format of Anglo Values a) Integer type: The field width is the size of the maximum range of the value plus 1 for a.sign. Positive values have a blank'in the sign position, negative. values have a minus sign in the sign position. The numeric field is leading zero suppressed, replaced by blanks. If the value is zero, the right nest position will -contain a zero.

Example: 'f saturation margin, range +700 to -2100'F, is 50 F the transmitted data is Q550. If it is -10'.F, the transmitted data is -$ /10. If it, is zero, the transmitted data is )+$ 0,

'.: ','; where P:.is "ASI:sp,ace,'(b'1ank):

Requi rement No. 16081-ICE-3111 Revision 00 Page 31 of 32

I

'J

(

t P

NOTES TO CROSS REFERENCE TABLE (Continued) 4 b) Exponential Format: Above a value of 10 and below a value of 1000, the integer format (described abave) will be used.'or the other values, the field width is 8 characters as follows:

a.aaaWb, where a.aaa is the fractional part of the value and

+bb i s the exp onent i a 1 pa rt.

(Note: no sign information is transmitted since the data is always positi ve. )

For example: 1.23" power is transmitted as 0.123 + 01.

0 Requirement No. 16081- ICE-3111 Revi s i on 00 Page 32 of 32

~ t I

'}

ATTACHMENT A PRESENT PLANT STATUS

( 3 f

CRIT1CAL YiATER1AL 1HTERYiEDE ATE TO LONER SHELL HELD

, RT),DT . )'TP5 265 PTPS 264 OF DATE PTP UN1TS NILL EXCEED SCR=Et<1NG CR1TER1A (lJS!hG 1ST 8 CYCLE AVERAGE) t:Ti,.DT RATE. OF 1hCREASE

/ F/EFPY

4~ I I I

TURKEY POINT. UNlTS 3 a 0 BASIS FOR RT ~ CALCLJLATION

= RTo + ~RT + 2~ TERf"l RT<m

-RT ~

0 F 0.

~

~RT- GUTHRIE 2< TERN S9o F X Cu 0,32K 5 N> 0,57K CAPACITY FACTOR 80K PTP 5 EFPY 6,3 PTP. 5 FLUENCE 1 x 1019 N/cH2 PTP 0 EFPY 6,35 PTP v FLOENCE 1,02 x 101 Nlc,;2

L ~ I I ~

'C I

ATTACHMENT B VESSEL FLUX REDUCTION PROGRAM

Table of Contents

1. Purpose and Objective
2. Dimension of Flux Reduction Requirement
3. Turkey Point Operating History and Plans Flux Reduction Achieved to Date
5. Near-term Flux Reduction Plans
6. Long-term Flux Reduction Plans
7. Schedule

~

~ i > '

l I

1. Pur ose and Ob ective The Turkey Point nuclear units are the most economical power plants owned by Florida Power R Light. As such, these units are good candidates for extending their operating lifetime beyond current license life-The present objective of the flux reduction program is to reduce the fast neutron flux at the vessel surface sufficiently to allow operation to at least the licensed lifetime. To achieve this objective, changes to core designs are anticipated to substantially reduce vessel flux. Fuel management analyses are underway and quantitative vessel fluence analyses are planned to determine the best means of reaching the sufficient flux reduction condition.

a ~ s >

~ s I

0 2. Dimension of Flux Reduction Re uirements The Turkey Point pressure vessels have only circumferential welds with a screening criteria of 300oF RTNDT. This corresponds to a limiting fluence in the most limiting weld of 1.85 x 1019 n/cm2. The last reviewed submittal (August 31, 1982) quantified the radially dependent flux level in the critical weld as depicted in Figure 2.1 for the "8 Cycle Average."

The time dimension of the flux reduction requirement is defined by the need to reach licensed lifetime (year 2007) and the potential desire to reach a later year such as 2015. This implies 19.2 effective-full-power-years (EFPY) and 25.6 EFPY of further operation, respectively, beyond 3anuary 1983. The fluence to date is about 1 x 10 n/cm for both units after 6.37 EFPY of operation. Table 2.1 provides a summary of the current status of both units.

The axial spatial dependence of needed flux reduction can be seen by referring to the axial cross-section of the vessel illustrated in Fig. 2.2. The limiting weld is about five feet above the bottom of the active core and is about 16" from the nearest assemblies in the core at the N-S and E-W axes.

The fluence in the base vessel material will not be limiting compared to the weld because of its considerably lower copper content. These factors lead to the need to reduce the source of fast neutrons from the core to an area extending about lYi feet above and below the weld elevation.

The radial dimension of the required flux reduction is presented in Fig. 2.3.

To reach currently licensed lifetime, some flux reduction must occurmver an angle of about +15 about the core axes, Referring to a radial cross-section of the vessel, Fig. 2.0, flux reduction to a length of 00" of the weld about

~ ~ ~

~

I I

each of the axes is necessary. Visual inspection of assembly placement reveals that all twelve assemblies on the core "flats" must reduce their

~

source of fast neutrons. The same inspection leads to the observation that no other assemblies are nearly as important to the needed flux reduction.

Assemblies near the core diagnonals and at the core edge could even be allowed to increase their source substantially.

FLORIDA POKIER R LIGHT CO, TURKEY POIHT CURRE;"tT STATUS (1/1/83)

JFPF~<C (1019 N/cN2) jlrZI (oF)

UNIT 3 1,00 263 Ut<IT 0 1,02 269 SCREENI JG CRITERIA 1,85 300 TABLE 2,1

FLORIIN PQ'KP 'IGHT CO.

TUm POIm WIT i~

FAST FLUX vs AZINHQL AViLE 11 10 6

2 l

GP.jERIC 0"-SIGi<

10 ". CYCLE AVE.

7 6

~* 5 u

9 10 0 5 10 . 15 20 25 33 55 4 45 59 AZlNP,lAL A%)LE '( D-:C)PKS )

FI6URE 2 1

l I I 1 I

f.LQRIDA P9!'lE.' LIFljT C",

TURlE'OI!JT RE;ACTOR VESSEL %NIAL CROSS SECTIO<

i<0 "LE SHELL T3 INTERMEDIATE SV. LL 'lELD ACTIVE CORE I.'JTE:".:.E9IATE S lCLL T~

LO;.'E!', SHELL 'LD LP.!E:", SHELL 'FIELD TO DL'TCl/f'1nl) i'(ELB FIGURE 2,2

~ I ~

~

l I

~

~ k l

FLORIDA PONER R LIGHT CO, TUPt,'EY POINT UNIT 4 101 REDUCTION FACTOR YS, ANGLE "9

8 7

6 5

CO I

10 9

8 015 u

Cl> 007 4

10-1 0 5 10 15 20 25 50 55 40 45 50 AZINUTHAL ANGLE (DEGREE)

FIGURE 2,3

FL%IM f9KR a LIGHT CO.

TUMY POINT fKACTOR CORE CROSS SECTI9"'l FIGURE 2 0

~ ~ ~

I I ~

~

I

I 3.

Turkey Point Operating History and Plans Since startup (Unit 3 = 1972> Unit 0 = 1973), both units have operated on annual cycles (with one exception for each plant). Half of these cycles used conventional fuel management (fresh fuel on the periphery) and the other half used "standard" low-leakage fuel management.

As of january 1, 1983, Unit 0 has accumulated 6.37 (Figure 3.1) EFPY and Unit 3 slightly less. Subsequent to steam generator replacement at both units, 18 month operating cycles are planned. Annual cycles will be used only when contingencies necessitate it. Figure 3.2 illustrates this schedule. Unit 3 Cycle 8 started up in April 1982 and is an eighteen month cycle. As of this date (February 8, 1983), an annual Cycle 9 for Unit 0 is intended because of schedular constraints.

Use of 18 month cycles and a planning basis 91% capacity factor between refueling results in approximately an 8096 total capacity factor. This factor is used in any discussions of EFPY and calendar dates.

This historical operation of the Turkey Point nuclear units along with "generic" core radial and axial power distributions were previously used to quantify the vessel fluence. The generic power distribution places the axial power peak at the critical weld. The generic radial power distribution is given in Fig. 3.3.

The "8 Cycle Average" Turkey Point specific calculation of fluenck used a revised radial power distribution, also provided for Unit 0 in Fig. 3.3. Shown in Fig. 3.0 is the "8 Cycle Average" radial power distribution illustrating the

~

~

equivalence of Units 3 and 0.

- The axial power specific to Turkey Point has not been accounted for, however. inspection of the actual axial powers on the core flats leads to the estimate of a 0% lower accumulated fluence at the critical weld than the results provided in August 1982. This reduction in fluence to date is referred to in subsequent discussions of needed flux reductions.

I

~.

FLORIDA POt'fER G LIGI/T CO, TURKEY POINT UNIT ACCur1ULATED BUR'tUP VS, YEAR 6,37 EFPY 1/V83 7S 76 77 78 79 83 81 82 83 Sl

%8 FIGURE 3.1

I I ~

FLORIDA POWER F LI6llT C~.

TUf'VEY POINT OPERATI',1G SCIIEIjUt E (AS OF JANUARY 1, 1983)

Lt lIT 3

'YCLE 8 CYCLE 9 CYCLE 10 -

CYCLE ll CYCg 9 CYCLE 10 CYCL 11 CYCLE 12 1983 1985 1986 1987 1988 FIGURE 3,2

FLORIDA POWER R LIGHT CO, TURKEY POINT DESIGN BASIS PERIPHERAL POWER DISTRIBUTION P~) ,93 ,77 1,12 ,80

,85

,92 TURKEY POIi"JT UNIT 0 8 CYCLE AYERAGE PERIPHERAL POWER DISTRIBUTION

,73 ,62

,56 1,10 .64 1 02 e

FIGURE 3,3

FLORIDA POWER a LIGHT CO, TURKEY POINT UNIT 3 8 CYCLE AVERAGE PERIPHERAL POWER DISTRIBUTION

,75 ,60 1 16 .98 ,52 TURKEY POINT UNIT 0 8 CYCLE AVERAGE PERIPHERAL POWER DISTRIBUTION

,73,62 1,17 . 95,56 1,10 ,64 1,02 FIGURE 3.A

' I k

0. Flux Reductions Achieved To Date In late 1981, the Pressurized Thermal Shock (PTS) issue became a serious

- concern with respect to fuel management because of the vessel flux limitations which would be a part of PTS. Flux reduction for the next reloads for each Unit were given attention even though quantitative flux targets were not yet known. Attempting to err on the side of prudence, the Unit 0 Cycle 9 reload was specified in March 1982 with a "modified low-leakage" loading pattern. At that time, the planned startup of Cycle 9 was 3une 1983.

The annual Cycle 9 "backup" design was also set with the same approach and achieves greater flux reduction than the eighteen month cycle presented in this section. Similarly, in 3uly 1982 the Unit 3 Cycle 9 design used "modified low-leakage" (planned startup December 1983).

The "modified low-leakage" is feasible within existing operating margins.

The predicted radial power distributions (cycle average) for Cycle 9 of both Units is provided in Fig. 0.1. It is anticipated that these designs provide almost a factor of two reduction over the "8 Cycle Average." The impact of this reduction in light of the now known target fluence, is illustrated in Fig.

0.2.

If the Turkey Point Units had operated since initial criticality with conventional fuel management, stainless steel dummy assemblies would need to be implemented now in order to stay below the RTgDT screening criteria at licensed lifetime. The drop-dead date for dummy assemblies based on the "8 Cycle Average" flux level is 1986. With modified low-leakage,>dummy assemblies would not be necessary until 1990. These projections assume that use of stainless steel dummy assemblies in all twelve core flat positions

I I ~

~

I

achieve a factor eight flux reduction relative to the generic power distributions.

I I I ~

FLORIDA POHER 8 LIGHT CO, TURKEY POINT UNIT 5 CYCLE 9 PERIPHERAL POWER DISTRIBUTION

.50 ,42 1,16 .98 ,48 1,15 ,46 TURKEY POI.AT UNIT 4 CYCLE 9 PERIPHERAL POktER DISTRIBUTION

,41 ,42 1,12 ,92 .42

,40 FIGURE 4 1

~ I I ~

1.IORIBA PONEtl ~ IGIIT CO, TURKEY POt"iT tl'1.'-.T t!

VESSEL FLUE!!CE VS, VESSEL LIFE A B C E 11/86 Hl89 ill% 2007-

/ r r~

/'990 A GENERIC HESIGl'1 B 8 CYCLE AVG.

C- CYCLE-9 D

NEEDED E DUNNY ASSENBLIES t.o 20 30 I=;FPY FIGURE 4.2

~

~

I

5. Near Term Flux Reduction Plans In the second-half of 1982, with the establishment of the screening criteria,

. the limiting fluence became known and flux reduction became more urgent.

Because materials were already in process for the next reloads, further modifications to the Cycle 9 designs were evaluated which did not entail change to the fuel loading. Time constraints limited changes to the Unit 0 Cycle 9 design to those which fell within existing operating margins.

As will be seen in subsequent sections of this report, increases in operating margin are required for Unit 3 in time to allow more extensive changes in its Cycle 9 design. The annual Cycle 9 Unit 0 design now has no time to be changed but has a radial power of 0.32 on the core flats which is about the same as modifications to the 18 month cycle could have achieved. As a general point, annual cycles can achieve lower vessel flux levels because of the greater inherent operating margin to LOCA and DNB limits. The lower number of feed assemblies increases the designers flexibility in shifting power away from the core flats.

The switch to the annual Unit 0 Cycle 9 has caused the Cycle 10 reload to start the design process now. This design assumes increased operating margins and will implement flux reduction features described in this section.

Cycle 10 is now planned to start in May 1980 and will be an 18 month cycle.

A portion of the design flexibility associated with annual cycles can be obtained by moving to higher assembly discharge burnups (fe r feed t

assemblies). Achievement of high burnups and NRC approval of 'the high burnup topicals submitted by the fuel vendors in 1982 is seen as a high

priority with respect to flux reduction.

The Unit 0 18 month Cycle 9 design was used for the near-term flux reduction fuel management studies. Conclusions resulting from these studies are generally applicable to any 18 month Turkey Point cycle-Figure 5.1 summarizes the anticipated current magnitude of flux reduction.

The previous Cycle 9 design, and using equivalent core designs in the future, would cause the screening criteria to be reached in August 1995. Switching to dummy assemblies would be needed eight years from now if no other actions were to be taken. Translating these limitations to flux, Fig. 5.2 illustrates the flux levels versus azimuthal angle which cannot be exceeded (on the average) to avoid reaching the screening criterion. These flux limits assume the 096 reduction in historical flux level due to the corrected axial shape.

Even with increases in operating margin, the time required to implement exotic assembly designs or materials constrain the near term solutions to "off-the-shelf" materials and standard assembly designs. The options considered for near term implementation on the core flats were spent fuel (lowest reactivity), fresh full or part length burnable absorbers, part length control rods installed on burnable poison spiders, and assemblies containing natural or depleted uranium.

The radial power impact of the two most simple changes compar to the previous Ccyle 9 design are provided in Figs. 5.3 and 5.0. The case of low reactivity fuel and burnable poisons is anticipated to achieve the majority of

~

~

l

s s

needed flux reductions. The burnable poisons (Fig. 5.0) used in the study s

were full length. The small axial extent of needed flux reduction, however, indicates that part length poison rods can be just as effective with a lesser decrease in overall radial power. Part length BPs would, therefore, assist in mitigating the loss in operating margin for a given level of flux reduction.

The impact of the near term design changes on the axial power shapes is illustrated in Fig. 5.5. The use of'spent fuel on the core flats has a large advantage compared to the generic power shape by shifting the powers upwards, away from the critical weld in addition to the expected reduction in axial peaking. This factor results in about a 1096 decrease in critical weld flux in addition to the decrease in radial power.

~ Combining the radial powers and the axial shapes results in the powers plotted in Fig. 5.6. The expected impact of implementing these changes is given in Fig. 5.7. The design changes planned for Cycle 9 of Unit 3 and Cycle 10 of Unit 0 correlate with Curve C on Fig. 5.7 which indicates that the screening criterion would be reached in August 2000. Assuming no further changes, dummy assemblies could be used beginning in 2001 to reach licensed lifctime.

These changes, however, are not without penalty. Increases in hot spot peaking (~F) and radial channel peaking (Fz H) are expected. In addition, compared to designs without these changes, core reactivity is lost. In future cycles, this will be recovered by increasing the amount of U-235 load/ in the core. These penalties are summarized in Table 5.1. Table 5-2 lists the expected RTNDT values associated with the near term design changes.

I Florida Power R Light intends to implement the most effective of these design changes. Near-term approvals, however, of topicals, technical specification changes and licensing analyses are required by third quarter 1983 for the following items.

High-burnup topical Enrichment limit on fuel storage Analyses for higher F~H operating limit Analyses for higher LOCA (Fq) operating limit.

I I I-i ORIANA PO>tEr:,.', Lir~iT CO, TURKEY PnI~tT tl'I.'-, f ~!

VESSEt r-LuE,lCE VS. VESSE Lrr-E A 8 C 11/% 8/95 2007. 2035

~'991 rr A- CYCLE-9 8 A MITH IX AXIAL C

HEEDED D

DUNNY ASSENBLIES 20 50 FFP~

FI60RE 5.1

FLORIN PolKR 'I.GHT CO, TURZ( POIt'tt lF:)IT 0 FAST FLUX vs AZINTNLA""6IE 7

6 5

2007 2015 2

G F."P,IC e ,-: KSIGN 8 CYCLE A'ItE, 7

6 NIAL CORRECTIG'l 9

19 0 5 10 35 20 Z 30  % A0 I 45 ZIW,W AWHILE ( DEGREES )

FIGURE 5,2

TURKEY POINT UNIT 0 CYCLE 9 PERIPHERAL POWER DISTRIBUTION

,01 .42 1,12 ,92 ,02 1,00 ,00

,83 CASE A HIGHLY BURNT ASSENBLIES

.29 ,27 1,12 ,90 .02 1.10 ,l2

,88.

FIGURE 5 3

FLORIDA POMER R LIGHT CO, TURKEY POINT UNIT 4 CYCLE 9 PERIPHERAL P01'(ER DISTRIBUTION

,41 ,42 1,12 ,92 ,42 1,04 ,40

,83 CASE B HIGHLY BURNT ASSEtSLIES + BPS 23 .21

,90 ,42 1,11 .42

,89 FIGURE 5 4

\

I

FLORIDA PO)CER 8 LIGHT CO.

TURKEY POINT UNIT 0 PERIPHERAL AXIAL POHER Sl]APE TOP D

BOTTOf]

,25 .5 75 RELATIVE POHER (HORf'lALIZED TO 1)

A - GENERIC O' ACTUAL 8 CYCLE AVERAGE C

- SPENT FUEL D

SPENT FUEL + PLBP FIGURE, 5,.5,

I l

FLORIDA PO'HER 8 LIGHT CO, TURKEY POINT UH I T PERIPHERAL ASSEf'SLY POMERS TOP l

I l

I F /

BOTT Ot1

,25 .5 ;75 1,25 RELATIYE PO!HER A SPE(<T FUEL + BP D 8 CYCLE AVG, B SPENT FUEL E - GENERIC CYCLE 9 DESIGN<

(

C F SPENT FUEL

+ PLBP FIGURE 5,6

I I

I-I ORIM POWER .", LICiIIT CO.

TURKEY POVIT tlat'I! T i!-

VESSEL I LL>E!!CF VS. VESSEL LIFF A B C D 8/95 V2000 8t2(6} 2035 rr r

r ~Vms r

r rr C'l 199lwc- r r rr CQ n r A 4K AXIAL LxJ B SPENT FUEL LU

/

/ C SPENT FUEL + BP

/ D NEEDED

// E DtjNNY. ASSE%LIES

//

/

/

//

/

20 50 FFPY FIGURE 5.7

I I I

NEAR-TERN FLUX REDUCTION PERIPHERAL REDUCTION CYCLE PEAKIN6 FLUKE LEJ6jj. EAQM 6ENERIC 1s0 ,76 8 CYCLE AV6. .76 1.0 f<EEDED (2015) ,17 4.5 (2007) .21 3,4 PTP 4 CYCLE 9 ,45 1.7 SPENT FUEL ,30 -2.5 6 DAYS +2K BURNABLE POISONS ,27 2,8 6 DAYS +2%

~T'w SPENT R BP's .23 12 DAYS +4X

" AT CRITICAL MELD AXIAL PLANE TABLE 5.1

FLORIDA POMER 8 LIGHT CO, FLUX REDUCTION OPTIONS (CU = .32, i~I = .57)

-OPT IOfi'T NDT RT NDT DATE RT NDT = 300 oF aL2W RZKL5 GE!'IERI C 376 396 11/86 8 CYCLE NE; 374 11/89 7/og>>

CYCLE 9 DESIGN'I 325 8/95 SPENT FUEL- 312 322 1/2000 SPE)JT FUEL + BP 304 313 8/20OA STAIiILESS STEEL 286 293 9/2025 "I"ICLUDES AXIAL CORRECT IO.'l TABLE 5 2

1

~

I I

6. Lon Term Flux Reduction Plans The long term flux reduction actions have several purposes. These are o Reduce vessel flux further than the near term actions o Increase the flexibility in means to accomplish flux reduction at the lowest cost o Quantify for NRC review all flux reductions The long term options currently envisioned are summarized in Table 6.1. The most flexibility and lowest cost is expected to come from concentrating on axial zoning of fuel although the manufacturing problems associated with this have not yet been identified.

Quantification of flux reduction is expected to proceed in several steps using the DOT 0.3 computer code.

o Historical cycle specific flux levels using actual radial and axial powers for both units through Cycle 8.

o Near term cycle flux levels to establish expected date of reaching screening criteria.

o Axial and radial adjoirit calculations using various materials in the long term options to establish guidelines to be used for future reload design.

Though not yet filled in for other long term options, Table 6.2 does provide the expected peaking factor impact of the dummy assembly option. The expected increase in fuel cy'cle cost of dummy assemblies is very large as is the original cost of implementation. Therefore, very high motivaticp exists to avoid dummy assemblies in view of the high confidence that they will not be necessary.

FLUX REDUCTION OPTIONS PERIPHERAL POISONS BURNABLE ABSORBERS PART LEiNGTH BURilABLE ABSORHERS HAFNIUH HIGH BUfU'3UP ASSEfSLIES NATURAL OR DEPLETED URANIUf'1 PARTIAL FUEL ASSEf%LIES NON-FUEL ASSEf1BLIES AXIAL OR RADIALLY ZONED ASSEf'SLIES TABLE 6,1

LONG-TERtl FLUX REDUCTIOr<

(2015)

PERIPHERAL REDljCTION CYCLE- PEAKING LBSIH E8GXR 8 CYCLE AVG, ,76 1,0 NEEDED 0,16 STAINLESS STEEL II 0 ] )II 6,3. +10%

NATURAL U DEPLETED U

, i'NATURAL + HP

" AT CRITICAL HELD ELEYATION PARTI AL ASf'I.

7. Schedule The following time table provides the currently envisioned actions for the FPL flux reduction program for the Turkey Point nuclear units.

Date Milestone 1978 1mplement low-leakage core designs March 1982 Set modified low-leakage designs Fall 1982 Near term design change fuel management evaluation Spring ) 983 Finalize Unit 3 Cycle 9 and Unit 0 Cycle 10 Design changes.

Obtain DOT 0.3 Code at FPL Load modified low-leakage core in Unit 0 Cycle 9 (annual).

Fall 1983 Perform long range flux reduction fuel management studies.

Submit FPL lattice physics topical Establish DOT model for Turkey Point Winter 1983-0 Evaluate fluence using DOT Submit PDQ model topical Load Unit 3 Cycle 9 with near-term flux reduction changes.

Have fuel vendor assess fuel assembly designs needed for long-term flux reductions.

Spring 1984 Set Unit 3 Cycle 10 design Load Unit 0 Cycle 10 Submit historical fluence calculations

ATTACHMENT C ASSESSMENT OF SAFETY MARGINS

I Assessment'of Safet Mar ins Xntrodoction The core configurations aimed at reducing fluence described previously involve a reduction in the power of the periph-eral assemblies'his leads to an increase in peak heat flux in other regions of the core which translates into an increase in the radial nuclear peaking factor and a commen-surate increase in the hot spot total peaking factor.

This discussion will focus on how the higher peaking factors can be accommodated without exceeding the core design safety limits, and without reducing reactor power from the current level o f 22 00 MWth.

Table','Assessment of - Saf et Mar ins at Turke Point There are four basic safety 3.imits associated with the design and operation of a reactor core. The total pea'king factor,'q, has to be maintained below the is determined by the requirement. that duringFq a limit, LOCA, the which peak clad temperature must be maintained below 2200 F.

The enthalpy rise factor, F>H, which is closely related to the radial peaking factor has to be maintained below its limit which is set so that during anticipated transient of low and moderate frequency there will be no departure from nuc3.cate boiling (DHB) in the core and therefore no fuel damage.

For low probability accidents DNB is permitted, but the extent of fuel damage must be limited so as to assure maintenance of a eoolable core geometry and radiation dose rates within limits specified in 10CFR100.

Maximum reactor coolant system pressure during transients must be limited so that the stresses in the pressure vessel and piping stay below the ASME code limits.

An assessment of the available operating and design margin for each one of these parameters shows that there is substantial margin to fuel damage at Turkey Point so as not to present a concern when the nuclear peaking factors are increased. The effect of higher peaking factors on coolant pressure is negligible so that pressure need not be considered further. The concern therefore need to be focused on the availability of F~ and Fq margin when low fluence core configurations are implemented.

Fi ure 1; .'esign 'ar i n 'nd 'a fet 'Limit Here are depicted actors which must be considered the operating and design margins available. 3:t is in'valuating possible that the current Technical Specification limit for the peaking factors could be substantially below the safety

limit thus providing design margin which can be. utilized" to raise the Tech Spec limit. To accomplish this usually requires new analytical methods which reduce the magnitude of the uncertainties, either through more sophisticated calculational methods or by factoring in new data that became available since the previous safety analysis was performed. I The expected peaking factors (nuclear peaking plus cal-culational and measurement uncertainties) for the low fluence core configurations will increased and therefore the Tech Spec limits need to be raised.

Table 2;'Projected F Har in at Turke Point This table compares the expected enthalpy rise peaking factor,'~ for the various low fluence core designs with the -corresponding Tech Spec limit and suggests ways in which the F>H Tech Spec limit can be increased to accommodate the increased nuclear F>H. The values shown in this and the following table are projections only,'ased on previous generic sensitivity studies, and must be confirmed by plant specific calculations after the design has been finalized.

The table shows the F>g values for the present low leakage core design typified by Turkey Point 4, Cycle 9 and three stages of contemplated fluence reduction designs:

near term flux reduction measures,'uch as those contemplated for Turkey Point 3, Cycle 9; long term lux. re-duction schemes, such as placing, natural or depleted uranium fuel on the flats; and replacing outer assemblies with dummy stainless steel assemblies. The F<H for the present low leakage design is quite close to the current Tech Spec limit of 1.55,'hich is also the generic limit for all current Westinghouse fuel. The nuclear F~ is expected to increase by 4,' or 10%,'espectively for the designs with lower fluence. The table indicates that for Turkey Point 3; Cycle 9 the available DHB margin identi-fied in the Westinghouse Rod Bow Topical Report (WCAP-8691),

already approved by the NRC, can be utilized. For further flux reduction the Westinghouse improved Thermal Design Procedure (iTDP); which is based on a new DHB correlation (WRB-1) and on statistical combination of uncertainties must be implemented. This methodology has been generically approved for Westinghouse fuel, but the uncertainties and sensitivities must be qualified on a plant specific basis.

From this table it can be concluded that with the implementation of the improved Thermal Design Procedure there will be sufficient F~H margin to accommodate, any of the contemplated low fluence core designs.

I I

Table 3", Pro 'ected'F Mar in't Turke 'Point This table compares the expected total peaking actor, Pq, for the low fluence core designs with corresponding Tech Spec limits and proposes ways to minimize or accommodate the increase in Fq ~ The increase in hot channel peaking inherent in the flux reduction designs has a dual effect on Fq margin. It raises the hot spot nuclear peaking Pq and simultaneously lower the allowable Fq as calculated by t'e LOCA analysis. To counteract these effects new methodol-ogies must, be applied. One is BART (Best estimate Analysis Ref lood Transient), submitted by Westinghouse to the HRC in 1980 (WCAP-9561) and expected to be approved by the HRC in 1983. BART utilizes more favorable heat transfer coeffic-ients and axial profiles during the reflood phase of a LOCA calculation'. Another new methodology is BASH (Best estimate Analysis System Hydraulics) representing a advanced reflood model. BASH is to be submitted to the HRC still more in 1983 but NRC review will probably not be completed 1985-86. Each of these new LOCA models is expected to till increase the allowable Pq by about 0.1. To obtain additional margin the nuclear(expected) Fq can be reduced with axially zoned burnable poison. rods with the active portion of the rods near the mid plane.

The conclusion from this table is that with HRC approval and implementation of the BART methodology and axially zoned burnable poison t'e low fluence core. designs under consider-ation will have the required Fq margin. To implement dummy stainless steel assemblies would require approval and implementation of the BASH methodology.

Conclusion; Assessment of Safet 'Margins

l. It can be concluded that sufficient design margin exists at Turkey Point to implement low fluence core loadings at.

the current po~er level of 2200 MW"h wit'hout exceeding safety limits, provided HRC approval of the'ART LOCA methodology (already reviewed by Sandia for the NRC) is received in time for Turkey Point 3, Cycle 9 startup in December 1983.

2. To implement long term flux improvements would require approval of t'e Improved Thermal Design Procedure (already generically approved) . To implement a core with dummy assemblies would require additional NRC approval of the BASH LOCA methodology, which can not be expected before 1985-86.'.

Relief from the rules or criteria of regulations, such as those of Appendix K of 10CFR50 is not needed.

I I

t I

ATTACHMENT D TRANSIENT ANALYSES

PLANT SPECIFIC ANALYSIS TURKEY POINT PLANT SCOPE AND SCHEDULE FPL is currently considering a plant specific analysis for the Turkey Point Plant. The intent of such an analysis would be to identify the dominant sequences of events which could lead to pressurized thermal shock of the reactor vessel. The results of this analysis would be used in the evaluation of modifications to plant systems, equipment and/or procedures. In addition, the analysis would support the continued operation of the Turkey Point nuclear units past the date when they exceed the RTNDT screening criteria.

The current analysis schedule conservatively assumes that Turkey Point units vill exceed the screening criteria in late 1989. Based on FPL's ongoing flux reduction program, the required submittal date is not expected until the mid-1990's..

As stated earlier in this report, the vessel flux evaluation to be completed by the summer of 1983 will bette&define F the analysis schedule.

ANALYSIS DEVELOPHENT PLAN FPL has considered a number of different approaches to the Turkey Point plant specific analysis. The most promising general approach identified to date is similar ro that taken by Westinghouse in their thermal shock probabilistic risk assessment (PRA)*. Cooldown sequences are identified by constructing event trees for the major transient classes. trees are further t

and LOCA The event resolved and quantified by developing fault trees for the systems and THERP diagrams for operator actions, The cooldown sequences are then passed through a thermal analysis screening. Using'conservative criteria, the sequenc'es are .

~ J y

I

PAGE TWO t identified as potential crack initiators or non-initiators.

frequency potential initiators mechanics analysis to more At present, there are no are then sub)ect to clearly define the thermal established acceptance a

The high detailed fracture shock scenarios.

criteria for this type of analysis. FPL recognizes this is aq ongoing NRC effort and is willing to assist the staff in developing such criteria.

  • Summa of Evaluations Related to Reactor Vessel Inte ritv, Westinghouse Electric Corporation, May 1982 DEPARTMENTAL RESPONSIBILITIES The analysis described in the previous section requires coordinating the efforts of a number of disciplines. Responsibility for the overall effort lies with FPL's Nuclear Energy Department. The tasks of the analysis have been assigned as follows: 1) Fuel Resources Department thermal/hydraulic analyses and fluence calculations; 2) Nuclear Energy Department - vessel material properties and; 3) Power Plant Engineering Department - probabilistic risk assessment and fracture mechanics.

ACTION TO DATE In planning the plant specific analysis, FPL engineers have reviewed much of the available literature on the thermal shock sub)ect. In particular, a detailed comparison of the generic plant described in the Westinghouse thermal shock PRA to the Turkey Point plant was made. A number of significant differences were identified such as RWST temperature and High Pressure Safety ln)ection System performance characteristics. Based on this comparison, FPL concludes that the Turkey Point units would respond more favorably to the cooldown sequences identified than the generic Westinghouse plant.

~

) "

~ )

~ ~

ph g

TRANSIENT ANALYSES FEBRUARY 19S3

s

" J

Introduction Florida Power and Light has been actively pursuing the resolution of pressurized thermal shock concern both on a generic and on a plant specific basis. In mid 1981 when Rancho Seco overcooling transient highlighted this concern, the issue was given top priority by the analysis subcommittees of the Westinghouse and Combustion Engineering Owners The Westinghouse Owners 'roup -(WOG) 'roups.

evaluated bounding over-cooling transients for all of their plants and concluded that in the near term all plants would operate safely. The analyses were documented in a report WCAP-10019 and were submitted to the Nuclear Regulatory Commission in December 1981. A plant specific evaluation of Turkey Point Units 3 and 4 submitted to the NRC in January 1982 concluded safe plant operation for the end of design plant life for bounding overcooling transients.

Through dialogues with the NRC staff it was recognized that the overcooling transients resulting from multiple component failures need to be evaluated to completely address the pressurized thermal shock concern ~ A generic study, prepared through t'e Westinghouse Owners'roup and submitted to the NRC in May 1982, concluded that high probability overcooling transients resulting from multiple component failures would not cause flaw initiation in any Westinghouse plant over the next three year period. In mid the formation of an FPL Task Committee for the 1982,'ith resolution of PTS issue,'n in-house investigation of small breaks was initiated to. explore the benefits of plant modifications and operating procedure the longer term, dominant overcooling transients changes'n identified by Turkey Point probabilistic risk assessment will require further evaluation.

Januar "1982 'ubmittal The plant specific submittal included calculations for the bounding overcooling transients initiated by large and small breaks in the primary and secondary systems. Plant specific thermal/hydraulic analyses were used as input for large break fracture mechanics calculations while the generic small break thermal/-

hydraulic analyses for three loop Westinghouse plants provided input for small break calculationsf Stress analysis and fracture mechanics evaluations were perfygmed based on an end of life weld fluence of 6.3 X 10 nvt which 0

corresponds to an end of life RTND of 407 F. Operator action was assumed only for the large steam line break for isolating the

'I

~ l

supply of auxiliary feedwater to the faulted steam generator within ten minutes. In case of a small pri-mary break, a two inch break in the hot leg resulted in loop stagnation and therefore', no credit was taken for the mixing of safety injection with the primary fluid. Based on warm prestressing it was concluded that all cracks would arrest within three quarters of the vessel wall.

Anal 'es in Pro ress In mid 1982 when the FPL/PTS task was decided to investigate higher force was formed, probability small it breaks further to generate plant specific thermal/-

hydraulic transients and to assess the effects of plant modifications and operating procedure changes.

An analysis of a two inch small brea'k loss-of-coolant in the hot leg concurrent with loss of offsite power which trips reactor coolant pumps is in progress.

Minimum decay heat, maximum safety injection flow, maximum auxiliary feed water flow,'inimum safety injection temperature and minimum auxiliary feed water temperature are assumed. The break size considered produces primary loop s agnation,'hus minimizing the safety injection mixing and maximizing the reactor vessel cooldown.

Another analysis currently in progress is the small steam line break from zero reactor power initiated by a stuck open steam safety valve concurrent with loss of offsite power. Initial conditions and sequence of events are chosen such as to maximize cooldown.

Sensitivity studies which would provide an assessment of ways possible for minimizing the cooldown 'will be performed to evaluate the effects of safety injection temperature, auxiliary feed water flow rate, steam relief valve isolation and operator action. It is desirable that for high probability overcooling transients, the downcomer fluid temperature be maintained above the end of life. RT> T. With the implementation oZ reduced flux coreVesigns,'he end of life RTNDT is estimated to lie between 300 F and 330oF.

The system transient analysis is performed with the RETRAH computer code developed by the Electric Power Research Institute. FPL has contracted with Energy Incorporated to conduct an independent check( of the Turkey Point model. A topical report on Ufe RETRAN code has been submitted to the NRC for review by. the utility RETRM Users'roup.

Anal ses'Being Considered'- 'Hear 'Term is considering carrying the transient'analyses small breaks further to evaluate mixing of safety for FPL injection, thermal and pressure stresses in the reactor vessel and crack growth. Since the end of life RT~DT is expected to lie between 300 and 330oF, a.t is desirable to demonstrate that the flaws would not initiate for high probability small breaks and for others, the cracks would arrest in less than three quarters of vessel thickness without having to depend on warm prestressing.

A dialogue has been established with EPRI to acquire their computer codes for performing mixing, stress and fracture analyses. EPRI is at present performing pressurized thermal shock analyses or Robinson-2, Calvert Cliffs and TMI-1 using the COMMIX code for mixing, the ABAQUS code for stress analysis and the PTS-1 code for fracture mechanics analysis.

Lon 'Term'-'PTS'Anal ses Long term PTS analyses would address dominant events identified by Turkey Point probabilistic risk assess-ment. The overcooling events which have cooldown rates higher than 100 F/hr and which result in downcomer water temperature below the end of life RTHDT would be considered potential flaw ini i-aters. These transients would be further investigated for crack initiation and arrest using fracture mechanics codes. Analysis results from probability events would then be evaluated to assess plant modifications and operating procedure changes to prevent crack initiation. Low probability events would be investigated for crack arrest. The long term effort would aim to demonstrate that the plants could operate safely at, the end of life with an RTHDT 300oF Conclusion

'The analyses submitted to the HRC thus far have demonstrated that probable overcooling transients would not initiate flaw propagation for the next few years. The analyses have further demonstrated that flaws would be arrested for the end of plant design life. The near term and the long term analyses would provide an evaluation of beneficial plant mpdifi-cations and operating procedure changes in Pase the end of life RT>DT exceeds the screening limit of 300 F.

PRESSU.',>7EO THERf'lAL SHOCK; TUN;EV POINT uf/ITS 5 a ~

PLAi/T SPECIFIC AHALYSES o DECENBER 19'1 GEf<ERI C EVALUATION NCAP 1.001a o JAHUARY 1982 PLAUDIT SPECIFIC o HAY 1982 GEHERIC PRA o NEAR TERN EYALUATIOH OF SNALL BREAI(S o PLANET MOD I F I CAT I OltS o OPERATING PROCEDURES o LONG TERN EVALUATE DONI NA,"lT TRANS I E,")TS o PRA

SUNPlARY OF PLANT SPECIFIC ANALYSES SUBNITTED IH JAt UARY lo.82 EVENTS ANALYZED o LARGE LOCA o SHALL LOCA (GENERIC TPANSIEHT) o LARGE SLB o SYiALL SLB (GENERIC TRAI'!SIENT)

ASS Ut'lPT I OHS o EOL RT = 007oF z

o 5/0 T CRACI( ARREST o l<ARN PRESTRESS IHG o NO YiIXItlG, SHALL LOCA o 10 YiIHUTE OPERATOR ACTIOH, LARGE SLB CON CLUS I Oil o CRACI( ARREST FOR EOL

1 ANALYSES IH PROGPESS EVENTS BEING ANALYZED o SMALL LOCA (STAGl'tA:"T LOOP) o SNALL SLB OBJECTIVES o PLAUDIT SPECIFIC TRAt'<SIEHTS o PLANT NOD IF I CAT IOfl EVALUATIO:"I o RHST TEMPERATURE o AUXILIARY FEED! fATER o BLOCK VALVE Oi< ATMOSPHERIC DUMP o OPERATING PROCEDURES EVALUATION f'iETHODS o RETRAff MODEL FOR TURKEY POINT

AflALYSES BEI HG CONSIDERED OBJECTIVE o SHOl( If"lPROVEYiEf<T OVER JAiiUARY 1 82 SUB!'iITTAL o PREVENT CRACK If<ITIATIOf'f o CRACK ARREST 0 - j./2 T o CRACK ARREST WITHOUT UPS EVENTS o SflALL LOCA (STAGNANT LOOP) o SHALL SLB ASS UNPT I ON o EOL RT 300 - 360oF CALCULAT=I0,'iS o ,'lIXIHG OF SI o STRESS ANALYSIS o FRACTURE NECHA>< I CS HETHODS o S IHPLE hI X I '$G NODEL/COf"lYIIX o ABAQUS o PTS-1

l.'

I

LON6 TERN - PTS ANALYSES OBJECTIVE o EVALUATE DONIHAiNT PTS TRANSIENTS IDENTIFIED BY TURKEY POIf<T PRA DESIRED 6OAL o 13EtlOtlSTRATE SAFE PLANT OPERATIOH AT EOL RT5500 F

L 1 J ATTACHMENT E SURVEILLANCE PROGRAM

l 0, ~

VESSEL INSPECTION The ultrasonic weld examinations performed on the Turkey Point Units 3 ana 4 reactor pressure vessels utilized 0', 45'nd 60'ngle beam techniques. All examinations were performed in accordance with the requirements of the ASME BSPV Code Section XI, Appendix I of the 1974 edition with addenda through the summer of 1975 plus, the requirenents of the USNRC regulatory guide 1.150 were cl osely adhered to. Contact examination techniques were conducted on the vessel i nteri or clad surfaces.

The 0 degree straight beam exami nation was relied upon to detect flaws oriented essentially parallel to the surfa'ce and to monitor sound transmission efficiency.

The 45 degree angle beam technique was modified to a full vee technique in order to monitor the area directly under the cladding. Sensitivity for this examination area utilized a two inch by .140 inch notch (2g code notch),

The 60 degree angle beam technique was relied upon to complement the 45 degree beam in the detection of flaws oriented essentially perpendicular to the surface of the vessel.

In addition, during the Unit 4 examination, a dual 70 degree refracted longitudinal team technique was employed to complement the 45 degree beam in the detection and/or evaluation of flaws located at the clad interface and the area beneath the clad for a distance of one inch.

During the examination of both units, the vessel girth welds joining the upper shell-to-intermediate shell and intermediate shell-to-lower shell courses were covered 100 percent. There are no existing aXial welds in either vessel.

I The Unit 3 examination exhibited no recordable indications.

The Unit 4 examination exhibited indications oriented at the vessel outside surface which were attributed to probable surface anomalies. Cladding indications were detected with the 45 degree beam, but not confirmed by the 70 degree technique and thus attributed to cladding irregularity. These indications are not indicative of flaws in the base material or in the clad-base materi al interface.

45'EAR SURFACE EXAM. (clad area + 1 inch)

Re erence Leve = notch response (.140 x 2")

Recording Level = 5(5 of reference (notch) 70'EAR SURFACE EXAM (Clad interface + 1 inch)

Reference Level = 1/16" dia. SD hole DAC curve Recording Level - 5(C of reference BALANCE OF EXAMINATION VOLUME Reference Level = .312" dia. SD hole DAC curve Recording Level = 2(C of reference

~

y "'

(

4

The Turkey Point Surveillance Program has six capsules remaining only two of which contain weld material. This" leaves a relatively small sample of critical material to be managed over plant li fe. The Unit 3 weld material is representative of both critical welds in Units 3 and 4 in that it contains the same weld wi".e number and flux lot as both critical welds in Units 3 and 4.

The flux lot number in Unit 4 capsule is different than those found in the critical welds.

It is FP51 's plan to ranove a capsule at the canpletion of Cycle 1O which is sometime in 1986. At the present time we are considering integrating our l

surveillance program so the capsule removed may be either fran Unit 3 or 4 but not both.

Some other options which are being considered are:

Changing a lagging capsule to a leading position.

Removing a capsule and inserting it into a test reactor to end of life

. fl uence.

Reconstituting charpy samples to either more fully develop Energy Temperature curves at existing radiation levels or create additional capsules.

Modi fying existing WOL samples to obtain better fracture toughness informati on.

FPSL is continuing a search for archival materials and archival materials i n fo rmati on.

l

'~i

~" L)

~

~~