05000266/LER-2007-003
| Event date: | |
|---|---|
| Report date: | |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications 10 CFR 50.73(a)(2)(v)(D), Loss of Safety Function - Mitigate the Consequences of an Accident |
| 2662007003R00 - NRC Website | |
Event Description:
Between April 4 and April 7, 2007, refueling outage calibrations were completed on six Unit 1 steam pressure instrument channels. The as-found values for all six pressure compensation modules identified time constants which were slightly outside the Technical Specification (TS) limits, and thus were inoperable.
The procedures used in calibration of these modules have historically included the development of graphical traces. The graphs were developed using a method called "Step Input." A different method, called "Ramp and Hold," combined with additional electronic noise filtering in test equipment, was used for the first time for Unit 1 calibrations during April 2007 to develop the graphical traces. Upon recognition of the Unit 1 time constants being outside of TS limits also having possible applicability to Unit 2, additional calibrations were planned. On May 24, 2007, calibrations were completed on all six Unit 2 steam pressure instrument channels in a sequential manner. The as-found values for all six pressure compensation module time channel was sequentially calibrated with adjustments to time constants made such that the channel was returned to operable status during the calibration process. The first Unit 2 use of the "Ramp and Hold" methodology was during the 2006 refueling outage.
Event Analysis and Safety Significance:
The TS requirement for lag is =/ modules exceeded the TS lead value; the lowest being at 11.7 seconds.
A review of the safety analyses was conducted to determine whether the low steam line pressure safety injection setpoint (more particularly, the lead/lag characteristics of the setpoint) had been credited in mitigating the effects of analyzed transients and accidents. Three analyses were identified that potentially credited the setpoint.
Rupture of a Steam Line (MSLB) - Core Response. This analysis evaluated the Peak Clad Temperature (PCT) and the margin to Departure from Nuclear Boiling (DNB) resulting from a postulated limiting main steam line break (MSLB). While this analysis credited the low steam line pressure safety injection as the primary trip, the assumed analytical setpoint was 350 psig and included a 1.5 second delay for sensing. The anticipatory lead/lag functions were not credited. Based on this, the deviations in the lead/lag module settings would not have had an adverse impact on the safety function credited with mitigating the analyzed transient.
Rupture of a Steam Line (MSLB) - Containment Pressure Response. This analysis evaluated the containment pressure and temperature responses to a postulated MSLB using computerized numerical methods. While the input "deck" for the computer runs included a Safety Injection (SI) system actuation setpoint of 515 psia for steam generator pressure, it was specifically noted in the description of input parameters that this setpoint was not a factor in the outcome of the analyses. Rather, the analyses concluded that the containment hi-hi pressure SI was received before the low steam pressure SI setpoint could be reached, specifically because the lead/lag settings prevented processing the low steam pressure SI signal more rapidly. As a result, the low steam pressure SI was not credited with mitigating the containment pressure transient.
FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) Point Beach Nuclear Plant 05000266 YEAR Rupture of a Steam Line (MSLB) - Mass & Energy Releases for MSLBs Outside of Containment.. This analysis developed the limiting mass and energy releases caused by a postulated High Energy Line Break (HELB) located outside of containment. The output of the analysis is used to determine the severity of potential harsh environments resulting from a HELB. As such, the first order considerations are the mass flow rate of steam emanating from a postulated break and the duration of the break flow before it is isolated.
Because safety injection alone does not precipitate main steam isolation, and the main steam isolation signal is independent of the steam line low pressure signal, minimal variations in the lead/lag module settings would not have had a significant impact on the duration and mass flow rates of postulated steam line breaks evaluated by this analysis. Additionally, the HELB-related EQ parameters in the primary auxiliary building and the turbine building do not protect plant safety limits related to a fission product barrier.
Summary: In each of the three steam line break analyses that cited the low steam line pressure safety injection, the variations in the settings of the lead/lag functions in the signal would not have been consequential should an actual event have occurred.
A review of TS Tables 3.3.1-1 and 3.3.2-1, RPS and ESFAS respectively revealed no other occurrences of specific lead/lag time constant limits. TRM 2.1, the Core Operating Limit Report (COLR) for Unit 1 and TRM 2.1 for Unit 2, and TRM 2.2, Pressure Temperature Limits Report for Units 1 and 2 were reviewed. No occurrences of specific instrumentation lead/lag time constant applicability was found in the TRMs.
Cause:
The cause of the event was human error. The first occurrence was during the development of the calibration procedures prior to plant initial operation. The individual(s) developing the procedures failed to identify the difference between actual and ideal graphical displays of circuit output. Current engineering evaluations identified the inappropriate graphical display curves. The curves were not ideal because a resistor in the circuit is wired such that it can not be removed from the circuit during graph production. The resistor prevents establishing the lead function at zero. A lead function not at zero causes a slight difference in the graphical output that is not easily distinguished from an ideal exponential function. The test procedure does not require a comparison of the developed actual curve output versus ideal curve output. Rather, the actual curve is used to determine mathematical values subsequently used in doing calibrations. Circuit simulators were used to develop ideal graphs during the engineering evaluations done to determine the cause for exceeding the TS requirements. It was only under these close examinations, performed to explain the out-of-tolerance values, that the difference in graphs was discovered. Calibration procedure results back into the 1970s were reviewed. The reviews indicated the presence of the resistor in the circuit. It was therefore concluded the condition has existed since initial operation of both units.
Human error also prevented the identification of the problem in 2006 during the Unit 2 refueling calibrations.
Electronic noise is present in the circuit and is subsequently present in the graphical displays produced during testing. Technicians did not accurately select the mid-point of the electronic noise band during performance of the calibrations. Point selection techniques are addressed during technician training and are considered as "skill of the craft," and are not proceduralized. The levels of electronic noise were recognized after the Unit 2 calibrations were completed and were addressed by the use of additional filtering, implemented via test equipment software programming. Improved filter constants were employed during the FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) Unit 1 calibrations performed during April 2007. The resultant improved accuracy facilitated problem identification and understanding of the Unit 1 results.
Corrective Action:
Previous calibrations had used a "Step Input" Method to generate graphical curves. A "Ramp and Hold" method was used for the first time during each unit's last refueling outage. The "Ramp and Hold" method effectively removes the effect of the R-12 resistor in the circuit, thus producing correct curves. The "Ramp and Hold" method has been incorporated into each unit's testing procedure for use in steam pressure channel calibrations. In addition test equipment software programming changes have been made. The refueling outages. The reprogramming is internal to the MTE, improving the electronic noise filtering which improves the accuracy and usability of graphical display outputs. Coaching and feedback discussions on point-selection techniques and "Ramp and Hold" methodology have taken place between Instrumentation and Control (I&C) technicians and supervisors.
System and Component Description:
The system that was affected is the Engineered Safety Features Actuation System (ESFAS) Instrumentation; and specifically, the main steam line pressure (low) instrument loop to generate a safety injection signal. The ESFAS system monitors plant conditions that require Engineered Safety Features (ESF) equipment actuation and automatically initiates ESF equipment to mitigate plant accidents. The Safety Injection — Steam Line Pressure — Low function consists of three channels on each steam line. The three channels satisfy protective requirements with a two-out-of-three logic. The function is anticipatory in nature and includes a lead-lag ratio.
Previous Occurrences: A review of recent LERs (past three years) was performed. No events or conditions involving the same underlying concern or reason as this event were identified.
Failed Components Identified: None.
Additional Information: None.