The MFP trips and they get a scram. water level declines to the low level scram setpoint. The RFP, HPIC and RCIC starts up and feeds the vessel. Vessel increases so fast it goes to the high level trip. The RFP, HPIC and RCIC trip. This is not control of vessel level. This happened in all these LERs. This is call banging around vessel level.
These systems aren't tuned for each other. You are not suppose to get to a low level trip then the high level trip. All these guys got control functions aiming to get level in the mid level or slightly higher. Why doesn't it work as designed?
This unprofessional vessel control is distracting the operators from the big picture.
Enclosed is Licensee Event Report (LER) 2014-005,
"Loss of Feedwater Results in Automatic Reactor Protection System Actuation".
On November 7, 2014, at 0847 hours, the reactor protection system (RPS) automatically actuated due to a loss of feedwater flow to the reactor pressure vessel (RPV). There were no complications during the shutdown as all control rods fully inserted and pressure was maintained by normal means. The high pressure core spray (HPCS) and the reactor core isolation cooling (RCIC) systems actuated based on a valid low reactor water level initiation and injected to restore RPV water level.
RPV water level continued to decrease to the Level 2 setpoint (130 inches above TAF) when the RCIC and HPCS systems started and injected into the RPV. Balance of plant isolation occurred with isolation of all required valves. Both reactor recirculation [AD] pumps tripped as designed. The division 3 EDG, which supplies emergency electrical power to the HPCS system started but, as designed, did not load onto the bus. The MFP started as designed on a RFP trip signal. At approximately 0850 hours, the HPCS and RCIC system injections terminated on a Level 8 setpoint (219 inches above the TAF) as designed. The lowest RPV water level reached during the event was 77.2 inches above the TAF. RPS was reset at 0915 hours. Mode 4, Cold Shutdown was entered at 1752 hours.
CAUSE OF EVENT
The RPS scram was caused by an invalid feedwater runback signal from the division 1 RRCS. A recorder was installed for additional monitoring purposes and identified signals being injected from the RRCS self-test system (STS) feature into the DFWCS. Data analysis determined that the voltage perturbations correlated to the STS within RRCS. The voltage perturbations had amplitudes of - 66 VDC with pulse durations of - 1 msec. These pulses would repeat in a repetitive pattern between 5 to 7 pulses with noted frequencies varying as short as 130 - 230 msecs. The patterns would occur for a period of - 10 seconds on 2 minute intervals. This signal has a large enough amplitude for actuating the input on the field bus module (FBM); however, the DFWCS software has a 1 scan (200 msec) delay feature to prevent the actuation. A DFWCS runback signal can occur when a signal is in for greater than 200 msecs or these 1 msec pulses align exactly at 200 msec apart. The root cause was determined to be a latent design flaw in the upgrade design package of the DFWCS modification in 2005. Due to implementing the new digital upgrade, the interface between RRCS and DFWCS involving the runback signal was altered. The original design used interposing relays as the interface between the RRCS and the feedwater control system. The digital upgrade changed the design interface and removed the interposing relays tying the output of RRCS directly into DFWCS.
Enclosed is Licensee Event Report (LER) 2014-004,
"Loss of Feedwater Results in Automatic Reactor Protection System Actuation".
On October 20, 2014, at 0217 hours, the reactor protection system (RPS) automatically actuated due to a loss of feedwater flow to the reactor pressure vessel (RPV). There were no complications during the shutdown as all. control rods fully inserted and pressure was maintained by normal means. The High Pressure Core Spray (HPCS) and the Reactor Core Isolation Cooling (RCIC) systems actuated based on a valid low reactor water level initiation and injected to restore RPV water level.
RPV water level continued to decrease to the Level 2 setpoint (130 inches above TAF) where the RCIC and HPCS systems started and injected into the RPV. Containment isolation occurred with isolation of all required valves. Both Reactor Recirculation [AD] pumps tripped as designed. The Division 3 EDG, which supplies emergency electrical power to the HPCS system started but, as designed, did not load onto the bus. At approximately 0221 hours, the HPCS and RCIC systems and the MFP stopped injecting when the Level 8 setpoint (219 inches above the TAF) was reached. The lowest RPV water level reached during the event was 87.1 inches above the TAF. RPS was reset at 0240 hours. Mode 4, Cold Shutdown was entered at 2323 hours, when the average reactor coolant temperature decreased to 200 degrees Fahrenheit.
CAUSE OF EVENT
The RPS scram was caused by an electrical transient in the balance-of-plant (BOP) 120 volt AC Uninterruptable Power Supply (UPS) system [EJ]. At the time of the event plant operators were in the process of shifting the BOP static transfer switch [ASU] to its alternate source for maintenance on the BOP Inverter. The transient was caused by a degraded static transfer switch component. Alternate supply voltage was available but a static transfer failure resulted in a loss of power to the UPS system loads. During the subsequent investigation, it was found that the static transfer switch's alternate power silicon controlled rectifiers (SCRs), were not firing due to an issue from the sensing and transfer card [ECBD]. Without the alternate SCRs firing, no voltage would be provided from the alternate source. Laboratory analysis determined that the card had a degraded logic chip. A NAND gate used in the logic chip was degraded. The degraded NAND gate caused a voltage drop resulting in 6.5V at the input to the downstream logic. This was lower than the expected 15V and failed to generate an "on" signal to the downstream logic. This prevented a firing signal being sent to the alternate source's SCRs. Analysis determined the degradation to be the result of a manufacturing defect. The control logic for the DFWCS is one of the electrical loads serviced by the UPS. Among the loads was an input signal to the RFP availability logic. Disruption of the DFWCS power due to the electrical transient affected the feedwater system causing the control circuit to believe it was not available and drove the output to zero. As a result, feedwater flow was lost to the RPV and the RPS actuated, as designed, when RPV Level 3 was reached.
Enclosed is Licensee Event Report (LER) 2013-001,
"Loss of Feedwater Results in Automatic Reactor Protection System Actuation."
On January 22, 2013, at 0332 hours, the reactor protection system (RPS) automatically actuated due to a loss of feedwater flow to the reactor pressure vessel (RPV). There were no complications during the shutdown as all control rods fully inserted and pressure was maintained by normal means. The High Pressure Core Spray (HPCS) and the Reactor Core Isolation Cooling (RCIC) systems actuated based on a valid reactor water level initiation and injected to restore RPV water level.
The cause of the event was failure of a balance-of-plant inverter/static transfer switch, which provides electrical power to the digital feedwater control system. A circuit card in the static transfer switch degraded, which affected operation of the inverter. The electrical loads serviced by the inverter/static transfer switch were placed on an alternate power source. This alignment will continue until permanent repairs are made which are currently scheduled for the next refueling outage.
RPV water level continued to decrease and when it reached the Level 2 setpoint (i.e., 130 inches above the TAF), the RCIC and HPCS systems started and injected into the RPV. Both RFPs and the main turbine tripped. Containment isolation occurred with isolation of all required valves. Both Reactor Recirculation [AD] pumps tripped as designed. The Division 3 EDG, which supplies emergency electrical power to the HPCS system started, but did not load onto the bus, as designed. The MFP started as designed when the RFPs tripped. At approximately 0335 hours, the HPCS and RCIC systems and the MFP stopped injecting when the Level 8 setpoint (i.e., 219 inches above the TAF) was reached. The lowest RPV water level reached during the event was 79.8 inches above the TAF. RPS was reset at 0413 hours. Mode 4, Cold Shutdown was entered at 2036 hours when the average reactor coolant temperature decreased to less than 200 degrees Fahrenheit.
CAUSE OF EVENT
The RPS scram was caused by an electrical transient in the balance-of-plant (BOP) 120 volt AC Uninterruptable Power Supply (UPS) system [EJ]. The transient was caused by a degraded static transfer switch component [ASU] coincident with a failed DC to AC inverter [INVT]. The static transfer switch did not seamlessly transfer the loads to the alternate source. The inverter was found on the alternate source with the fail light illuminated and its protective fuse actuated. The control logic for the Digital Feedwater Control system (DFWCS) is one of the electrical loads serviced by the UPS. Disruption of the DFWCS logic due to the electrical transient affected the feedwater system by driving the RFP controllers to minimum flow with no start signal being sent to the MFP per design. As a result, feedwater flow was lost to the RPV and the RPS actuated, as designed,when RPV Level 3 was reached.
No comments:
Post a Comment