2th International Conference on Structural Mechanics in Reactor Technology (SMiRT 2) Espoo, Finland, August 9-1, 29 SMiRT 2-Division 5, Paper 197 A Failure Mode Evaluation of a 8V MCC in Nuclear Power Plants at the Seismic Events Min Kyu Kim a, In Kil Choi b a Senior Researcher, Integrated Safety Assessment Division, Korea Atomic Energy Research Institute, Daejeon, Korea, minkyu@kaeri.re.kr b Principal Researcher, Korea Atomic Energy Research Institute, Daejeon, Korea Keywords: electric cabinet system, Nuclear Power Plant, 8V Motor Control Center (MCC), shaking table test, seismic capacity 1 ABSTRACT In this study, a seismic behavior of electric cabinet system in Nuclear Power Plants was evaluated by shaking table. A 8V Motor Control Centers (MCCs) was selected for shaking table test and a real MCC cabinet for Korea Nuclear Power Plant site was rented by manufactured company. For the shaking table tests, two kinds of seismic input motions were used which were a US NRC Reg. guide 1.6 design spectrum and a UHS spectrum. Especially, the UHS input motion was selected for an evaluation of high frequency seismic effect. PGA levels for shaking table test were scheduled by.2g to 5.g for the assessment of real seismic capacity of electric cabinet system. For the evaluation of structural seismic response, three directional accelerations were measured at three points of outside on the cabinet system and also that of the incabinet response amplification, accelerations were measure at two points which mounted in electric equipments such as relay. Through this shaking table test, structural and functional failure mode which didn t recognized yet of MCC cabinet system according to the seismic event was evaluated. 2 INTRODUCTION A 8V MCC Cabinet is one of major equipment system in Nuclear Power Plant. For the shaking table test, a real MCC cabinet was rented from the manufacturing company. For the evaluation of a failure mode for Motor Control Centers (MCCs), a shaking table test was performed. For the shaking table test, two kinds of seismic input motions were used. One is an artificial seismic input motion based on the NRC Reg. guide 1.6 design spectrum and the other is also an artificial seismic motion based on the Korean Nuclear Power Plant site specific Uniform Hazard Spectrum (UHS). The UHS motion was selected for an evaluation of a High frequency effect of the electric equipment in a NPP. PGA levels for shaking table test were scheduled by.2g to 5.g but the test was stopped at about the 2.5g level because of the chattering of the relay systems. The shaking table tests were performed with a one dimensional shaking which was a front to back direction (horizontal) and a vertical direction. Functional and structural failure modes were also evaluated by this shaking table test. For the evaluation of a relay chattering, the electric signal was measured at several points. There are two kinds of ground fault relays and thermal relays are installed in the MCC. It is impossible to measure the electric signal of all relays, only some of the relays and electric equipments were considered. In the case of the US NRC spectrum, 8V AC power was supplied so it can measure the signals of the equipments related to a power system like a power transformer. But in case of UHS spectrum, 8V AC power wasn t supplied because of the safety of experiment. Therefore, only the signals from the relays were measured. For the measurement of a relay system, an arbitrary input power was supplied to the MCC. Also, for the evaluation of structural failure modes, in-cabinet responses and response amplifications of MCC, acceleration responses were measured at major points of cabinet. Through this test, several kinds of functional failure modes can be founded and the chattering effect of several relays in the MCCs can be certified. As a result, it can be recognized that the 8V MCC has a 1
sufficient seismicity as a SSE level earthquake. But in the case of a higher level earthquake motion, a chattering happened for both seismic motions, moreover both a horizontal and vertical shaking cause a relay chattering. 3 SHAKING TABLE TEST 3.1 Input Seismic Motion For the shaking table test, two kinds of seismic input motions were used. One is an artificial seismic input motion based on the NRC Reg. guide 1.6 design spectrum (US NRC, 1973) and the other is also an artificial seismic motion based on the Korean Nuclear Power Plant site specific Uniform Hazard Spectrum (UHS) (Choi et al., 2). The UHS motion was selected for an evaluation of a High frequency effect of the electric equipment in a NPP. The response spectrums of the target seismic input motions are shown in Figure 3. As shown in the Figure 3, a resonant frequency range of the NRC design earthquake and uniform hazard spectrum is 2-1 Hz and 1-3Hz, respectively. (a) US NRC Design Spectrum (b) Uniform Hazard Spectrum Figure 1. Seismic Motion for Shaking Table Test 3.2 Overview of Shaking Table Test For the shaking table test, a real MCC cabinet was rented from the manufacturing company. A figure and drawing are shown in Figure 2. The MCC cabinet was fixed as jig plate by using welding which have performed as same way of NPP and Jig plate was connected on the shaking table by bolt. PGA levels for shaking table test were scheduled by.2g to 5.g but the test was stopped at about the 2.5g level because of the chattering of the relay systems. The shaking table tests were performed with a one dimensional shaking which was a front to back direction (horizontal) and a vertical direction. Figure 2. An Overview of MCC 3.3 Measurement System For the measurement of acceleration response about MCC cabinet system, two kind measurement systems were applied. At first, for the measurement for structural acceleration response of the cabinet system, the 2
accelerations were measured at the outside of cabinet. The measurement points were pointed out at the figure 3(a). As shown in the figure 3(a), three directional accelerations were measured at the three position of side panel of MCC cabinet. The other acceleration measurement is for the evaluation of in-cabinet response acceleration response (ICRS) of MCC cabinet. That because a failure of electric cabinet system according to the seismic event is a functional failure of electric equipment which mounted on the panel of cabinet system caused by shaking of inside panel, door or back wall significantly. For the measurement of in-cabinet acceleration response four positions which mounted major electric equipments were selected. Of the chosen measuring place showed in figure 3(b). (a) Structural Response (b) incabinet response Figure. Acceleration Measurement System RESULTS OF THE SHAKING TABLE TEST.1 Modal Test for Resonant Frequency Evaluation For the assessment of dynamic characteristics of a target MCC cabinet system, the random acceleration input motions were used for shaking table test. A longitudinal and transverse directional structural response according to a random vibration was showed in Figure. As shown in the Figure, resonance frequency of longitudinal and transverse direction is 72Hz and 12Hz, respectively. Angle(deg) 2 15 1 5.5 Frequency (Hz) 2 6 8 1 Angle(deg) 2 15 1 5.1 Frequency (Hz) 2 6 8 1..8 Amplitude.3.2 Amplitude.6..1.2. 2 6 8 1 Frequency (Hz). 2 6 8 1 Frequency (Hz) (a) Longitudinal Direction (b) Transverse Direction Figure. Structural Response according to a Random Vibration.2 Structural Response Evaluation For the evaluation of structural response, acceleration responses which measured at the side of MCC cabinet were compared. At first, a zero period acceleration (ZPA) response of each measuring point is shown in Figure 5 according to the input seismic motion. As shown in the Figure 5, an acceleration response amplified according to the location of cabinet. It can be recognized that the more amplify according to the location high position. For the clearly distinguish about an amplification ratio, the amplification ratio of MCC cabinet system is shown in Figure 6. As shown in Figure 6, in the case of NRC design earthquake motion, the 3
acceleration response is amplified with about maximum two times. In the case of UHS earthquake motion, it is amplified over acceleration response this maximum treble till a quadruple. These results are shown that an electric equipment which have high frequency characteristic should more acceleration amplify than structures. (a) NRC design earthquake motion (b) UHS earthquake motion Figure 5. Peak acceleration response according to the seismic input motion and measurement position (a) NRC design earthquake motion (b) UHS earthquake motion Figure 6. Amplification of acceleration response according to the seismic input motion and measuring position The acceleration response spectrum of structural is shown in Figure 7 and 8 according to the input seismic motion. In the case of the design earthquake and UHS earthquake, the response spectrum about the input earthquake of the acceleration response spectrum at the cabinet upper, middle and platform vibrator were shown. As shown in the Figure 7 and 8 the resonant frequency characteristics are well showed up according to the upper position of cabinet. As shown in Figure 7 and 8, acceleration amplification is not much clearly defined at the under 1 Hz region, but an acceleration is much amplified vicinity of resonant frequency region of MCC cabinet which is 12Hz. And the acceleration amplification effect is more clearly defined in case of UHS earthquake because of the resonant frequency of input seismic motion. 1.6 12 1.2.8. 8.1 1 1 1.1 1 1 1 (a).2g (b) 1.2g Figure 7. Acceleration Response Spectrum of NRC design earthquake motion according to a target input maximum earthquake motion at the top position
2 3 2 1 16 12 8.1 1 1 1.1 1 1 1 (a).2g (b) 1.2g Figure 8. Acceleration Response Spectrum of UHS earthquake motion according to a target input maximum earthquake motion at the top position.3 Incabinet Response Assessment For the evaluation of in-cabinet acceleration amplification effects, peak acceleration responses at the two points (E1, A2) which were shown in Figure were shown in Figure 9 according to the input seismic motion. As shown in the Figure 9, the amplification effects of in-cabinet responses are little different compare to that of outside responses. In the case of low peak acceleration experiments, the amplification effects are similar to that of outside response, but the acceleration amplification characteristics are changed as increase of peak acceleration. Especially, in case of UHS earthquake even the response of high position in-cabinet responses is inversed according to the input seismic motion at the 2.g and 3.g level acceleration shaking table test. These results might come from a nonlinear effect of cabinet unit. As shown in Figure 1, this MCC cabinet is constructed as combination of separate unit system, so the separate units can behave nonlinearly after long time usage or experience of extreme vibration. This is the reason why it is very difficult to estimate seismic behavior of electric cabinet using a numerical analysis. (a) NRC design earthquake (b) UHS earthquake Figure 9. In-cabinet acceleration response according to the input seismic motion For the comparing more precisely about amplification effect, amplification ratio which calculated peak response acceleration response divide as peak acceleration of input seismic motion is presented in Figure 1. As shown in Figure 1, in case of NRC design earthquake motion more 1 times amplification can be found at the upper position (E1) of cabinet. Otherwise, in case of UHS seismic motion at most 6 times amplification we can recognize. This kind of the result of being contrary to the prediction that it is the result of being contrary to the amplification feature of the seismic motions observed in the structure side and the high-frequency seismic motion will reach the affect that it is more disadvantageous to the electricity cabinet to a pre existence. In case of lower position (A2) of cabinet, almost 2 times amplification effects can be found of both seismic motions. It is determined that the resonant frequency characteristic of a structure is not reflected since the location of the installed accelerometer is low. It can be concluded that the electric equipments which plugged in the cabinet system in Nuclear Power Plants should be considered as frequency characteristics of possibly generated seismic motion an also location of that equipments will be mounted in. 5
8 Figure 1. Amplification factors according to the input seismic motion at the In-cabinet position For the investigation of in-cabinet acceleration response characteristics, acceleration response spectrums which measured at the point E1 and A2 are presented in Figure 11 and 12 according to the input seismic motion. As shown in Figure 11, the spectral acceleration is significantly amplified at the resonance frequency region compare to that of structural response which measured at the outside of cabinet. In case of UHS earthquake motion shown in Figure 12, spectral acceleration was amplified more significantly also. But in case of NRC design earthquake motion, spectral response at the high frequency region near 3- Hz is peculiarly amplified, so as a result the peak acceleration responses were presented very much higher. That is, in the case of UHS, since the spectral response was more amplified than that of NRC design earthquake motion but the spectral response is excessively amplified in the high-frequency region in the case of the NRC earthquake and that causes the amplification of ZPA, the seismic response is more estimated to show up in an in-cabinet. The response amplification effects can be remarkably revealed in the UHS earthquake, that is the high-frequency earthquake, but the maximum acceleration response was evaluated very high because the amplification phenomenon of the high-frequency region in the case of the NRC earthquake. It was estimated to be different from at in case of comparing the maximum acceleration response but for the high-frequency earthquake, amplification is generated in high frequency region through the response spectrum comparison. And in case of NRC design earthquake, it is determined to be the part in which the amplification phenomenon generated in the high-frequency area should not overlook. 1.6 1.2.8. 6 2.1 1 1 1 (a) E1.1 1 1 1 (b) A2 Figure 11. Acceleration response at the in-cabinet position according to an NRC earthquake 1.6 12 1.2.8. 8.1 1 1 1 (a) E1.1 1 1 1 (b) A2 Figure 12. Acceleration response at the in-cabinet position according to UHS earthquake 6
5 FAILURE MODE EVALUATION 5.1 Functional Failure The functional failure of electric cabinet system in NPPs was summarized as a relay chattering (KEPCO, 1992). For the evaluation of a relay chattering, the electric signal was measured at several points. There are two kinds of ground fault relays, and thermal relays are installed in the MCC. It is impossible to measure the electric signal of all relays, only some of the relays and electric equipments were considered. In the case of the US NRC spectrum, 8V AC power was supplied so it can measure the signals of the equipments related to a power system like a power transformer. But in case of UHS spectrum, 8V AC power wasn t supplied because of the safety of experiment. Therefore, only the signals from the relays were measured. For the measurement of a relay system, an arbitrary input power was supplied to the MCC. The shaking table test was scheduled as.2g to 5.g but a severe chattering signal was observed in the 1.2g test so the functional test according to the NRC design spectrum was stopped. The electric signals measured from the shaking table test by the NRC design spectrum are shown in Figure 13. As shown in the Figure 3, in the case of the SSE level of an earthquake at.2g test, there are no chattering events observed, but in the case of the 1.2g shaking, a chattering is observed both in the horizontal and vertical direction. But in this case a chattering is not related to a relay but to a secondary power transformer system. (a) 1.2g Horizontal Figure 13. Electric Signals from NRC Spectrum (b)1.2g Vertical In the case of the UHS spectrum, the shaking table test was performed at.2g to 3.g. The measurement results are shown in Figure. As shown in Figure, measured voltage is about 6.6V. In case of UHS earthquake, 8V AC power was not supplied to the MCC but for the measurement of relay chattering 6.6V power was supplied to the relay system arbitrarily. As shown in Figure, there was no chattering happened in case of.2g shaking. But in case of 3.g shaking it can be clearly observed chattering effects. In this case chattering happened both horizontal and vertical shaking. (a) 3.g Horizontal (b) 3.g Vertical Figure 1. Electric Signals from UHS Spectrum 7
5.2 Structural Failure The structural failure of electric cabinet system in NPPs was summarized as failure of a supporting system like anchor bolts or welding points (KEPCO, 1992). But during this test we can find shear failure of side panel of MCC cabinet system during the PGA 2.5g level shaking table test. The reason is the welding stiffness of this test is harder than that of the MCC cabinet. 6 CONCLUSION A shaking table test for the evaluation of a functional failure of electric cabinet systems in a NPP was implemented. The 8V MCC was selected for the shaking table test. A US NRC design response spectrum and a Uniform Hazard Spectrum were selected for the input seismic motion. As a result, it can be recognized that the 8V MCC has a sufficient seismicity as a SSE level earthquake. But in the case of a higher level earthquake motion, a chattering happened for both seismic motions, moreover both a horizontal and vertical shaking cause a relay chattering. In the case of structural failure, the shear failure of side panel of electric cabinet system can be found during this test. Acknowledgements. This work was supported by Nuclear Research & Development Program of the Korea Science and Engineering Foundation (KOSEF) grant funded by the Korean government (MEST). (grant code: M27233-8M23-31). REFERENCES Choi,I.K et al, (2) Development of a Uniform Hazard Spectrum for a Soil Site by Considering the Site Soil Condition, 2 KNS-spring conference. Korea Electric Power Corporation (KEPCO), "Ulchin Unit 3& Final Probabilistic Safety Assessment Report," 1992. US NRC Regulatory Guide 1.6. Design Response Spectra for Seismic Design of Nuclear Power Plants, 1973. 8