RBG-31583, Part 21 Rept Re Slow Closure of One Main Turbine Control Valve.Caused by Turbine Control Valve Assumed to Close Due to Unspecified Failure in Turbine Control Circuit.Standing Order Issued to Change Max Critical Power Ratio Limit

Part 21 Rept Re Slow Closure of One Main Turbine Control Valve.Caused by Turbine Control Valve Assumed to Close Due to Unspecified Failure in Turbine Control Circuit.Standing Order Issued to Change Max Critical Power Ratio Limit

ML20248J119
Person / Time
Site:	River Bend
Issue date:	10/02/1989
From:	Booker J GULF STATES UTILITIES CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
REF-PT21-89-168-000 PT21-89-168-000, RBG-31583, NUDOCS 8910130048
Download: ML20248J119 (3)

Text

};-g -(d!Y. ,

p t :; .p-

"' * ; Q..Q ?!? _ _ _-

y;j ,

.} ll l. . * >

. GULF ' STATES UTILITIES COMPANY r.E' <

RfVER MND STAflON POST OFFICE BOX 220; $7. FRANC!SVILLE. LOutSIANA 70775 '

AMA CODI t04 -- 635 6034 '346 8661.'

r .

V & 1 October'2,.1969 RBG File. os. h1583G9.5, G9.25.1.4 U. S. Nuclehr. Regulatory; Commitsion

. Document Control Desk?

Washington,10.C.:'20555

~

I'

. Gentlemen:- - 4

.c .

L ..

River Bend. Station' - Unit 1 LDocket No. 50-458-

~ P' lease find enclosed Gulf States Utilities') Informational Report regarding a recent discovery of'a condition at River 1 Bend Station - Unit 1. This report is1 being submitted.Eto provide information regarding the setting of the maximum combined flow limiter.1 q

LSincerely, l [ Y. *?'

. E.: Booker

-Manager-River' Bend' Oversight

.JEB/TF W PpJCE/ch

!, ~cc: U. S. Nuclear Regulatory Commission 611 Ryan Plaza Drive, Suite 1000 Arlington, TX'76011 l

l Senior Resident Inspector Post Office' Box 1051 I St. Fra'ncisville, LA 70775 INP0, Records Group . .

1100 Circle 75 Parkway,' Suite 1500 Atlanta,'GA 30339-3064

'f f

8910130048 891002-

.$fW' i

gDR ADOCK 0300o43g \

PDC

$ INFORMATIONAL REPORT REPORTED CONDITION By letter dated August 23, 1989, the General Electric Co. (GE) informed Gulf States Utilities Co. (GSU) of a condition potentially reportable under 10CFR Part 21, applicable to the River Bend Station (RBS), involving a slow closure of one main turbine control valve. This low probability event which was not previously considered results from a turbine control valve that GE assumes to close due to an unspecified failure in the turbine control circuit or in the servomechanism hardware. According to GE, if the valve closes in less than 2.3 seconds, a reactor scram is initiated as a result of high neutron flux and no safety limits are exceeded. However, if the valve closes in greater than 2.3 seconds, the reactor scram is initiated by high reactor pressure.

During this slow closure case, the minimum critical power ratio (MCPR) safety limit may be exceeded if the maximum combined flaw limiter is set for less than 113 percent of rated steam flow. The consequences of this postulated event is based on GE's assessment of a generic BWR/6 analysis. The analysis is not RBS specific nor has it been considered in the current RBS safety analysis.

INVESTIGATION Based on a generic BWR/6 analysis, General Electric has identified a low probability event which postulates a " smart" failure in which one turbine control valve (TCV) closes slowly. The other three TCV's open to maintain turbine inlet pressure. The turbine steam bypass valves will either open fully, or open partially to a position limited by the MCFL. Depending upon the capacity of the three operational TCV's and the setting of the maximum combined flow limiter (MCFL), a fast closure of the TCV causes a rapid pressurization of the reactor and correspondingly rapid neutron flux increase due to the positive void collapse feedback. However, a slow TCV closure causes a slow increase in reactor pressure with a correspondingly slow increase in neutron flux. In this case, the reactor scram would be delayed until the high pressure scram setpoint was reached. A large increase in surface heat flux would result under these conditions with a resulting large chanoe in critical pcwer ratio (CPR) before this transient is terminated by a reactor scram from a high pressure signal.

The MCFL setting is an adjustment made by the operator in the mair control room. The MCFL is an electronic control circuit that limits the steam flow demand signal that is transmitted to position the turbine control valves and the turbine bypass valves. The steam flow demand signal is proportional to the difference between the pressure regulator setpoint and the actual pressure sensed at the inlet to the turbine stop valves. A 30 psi difference is interpreted as 100 percent (rated) steam flow. Therefore, if a control valve should close due to the postulated failure, turbine inlet pressure would increase, thus increasing the apparent steam flow demand signal. When '

the demand signal exceeded the setting of the MCFL, turbine control valves or turbine bypass valves would stop opening, resulting in an uncompensated increase in reactor pressure.

I Page 1 of 2 ]

l 1

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ __.._..__J

y 7 . ,1 l' The original purpose of the McFL is to prevent an excessively- fast blowdown

-to the main condenser in the event of an upscale demand signal failure, particularly on plants with a large steam bypass capacity. An upscale demand failure (pressure regulator failure - open) is analyzed in Section 15.1.3 of-the RBS Updated Safety Analysis Report (USAR). This analysis is based on a L .MCFL setting of 130 percent, although the USAR indicates that the normal setting is 115 percent. However, GSU operating procedures currently require a setting of 109 percent. A review of startup test daira and operating records indicate that the MCFL has always been set at 109 or 110 percent at full power conditions since initial plant operations. ,

At the turbine control " Valves-Wide-0 pen" condition, the main turbine can pass a' steam flow of approximately 105 percent of rated with a turbine inlet pressure of 965 psia. With a nominal 10 percent bypass capacity available at RBS, the. actual total steam flow capacity with all turbine control and turbine bypass valves wide open is estimated to be approximately 115 percent ,

of rated flow capacity. Therefore, a MCFL setting of approximately 115

. percent or greater does'not provide a limiting function, since actual steam flow is limited by the capacity of these valves.

With a MCFL . setting of 109 percent of rated steam flow and assuming that the s plant is operating at the MCPR limit of 1.18, GE calculates that the MCPR safety limit would be exceeded by 0.02 if the postulated transient were to occur. This calculation was based upon a generic BWR/6 and is not RBS specific. A review of the operating history of RBS has shown that the margin to the MCPR operating limit has always been greater than 0.02. The smallest margin. occurred on September 9, 1989, when the margin to the MCPR operating limit .was 0.0377. Therefore, had this postulated event occurred at RBS, the MCPR safety limit would not have been exceeded.  !

CORRECTIVE ACTION I

A st'anding order was issued to change the MCPR operating limit from 1.18 to  !

1.20 until the MCFL could be reset. In the past, operations personnel have noted that the " Limiter-in-Control" light illuminated when the control knob for the MCFL was touched. Consequently, adjustment of the MCFL was not performed while the plant was operating because of the possibility for the limiter to inadvertently close the turbine control valves and initiate a j plant transient. During the outage of September 29-30, 1989, the MCFL was  !

changed to reflect the General Electric recommended setting of 115 percent.

As discussed below, GSU is performing a plant specific analysis to determine ,

the impact of this postulated event at RBS.

SAFETY ASSESSMENT Preliminary investigation indicates that the probability of the postulated event occurring is low. Since the actual margin to the MCPR operating limit at RBS has always been greater than 0.02, the MCPR safety limit would not have been exceeded even if this postulated event had occurred. This assessment is based on the GE analysis of a reactor similar to, but not identical to RBS. Because of the generic, bounding-type assumptions made by I GE, it is anticipated that the RBS specific analyses currently being performed by GSU will demonstrate an even lower impact of this postulated transient to RBS. These analyses are scheduled to be completed by December 22, 1989.

Page 2 of 2

- _ - _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ - - - _ _ _ _ - _ _ - _