Performance evaluation of a new signal processing system design to improve CANDU SDS1 trip response during large break LOCA events

Performance of a recently developed signal processing system for CANDU (Canada Deuterium Uraniu) reactor shutdown system 1 (SDS1) is evaluated in this paper. The evaluation is carried out in MATLAB/Simulink software environment as well as with an existing power measurement and signal processing system. The new signal processing algorithm is obtained based on the synthesis of several first order low pass filters with different delayed time constants. Throughout this paper, a special attention has been paid to compare the new signal processing system with the existing one. The dynamic behavior of the new signal processing system in the practical large loss of coolant accidents (LLOCA) events has also been examined. Simulation results show that during the LLOCA event, the reactor trip time, as well as the peak power, is decreased remarkably. Through the simulation studies, it has convincingly demonstrated that the new signal processing system has significant advantages over the existing system in terms of the improved trip response and accommodation of the spurious trip immunity. This advantage will significantly enhance the safety margin, or will bring economical benefits to nuclear power plants.     

Testing the dynamics of shutdown systems instrumentation in reactor trip measurements

Periodic testing of the dynamics of the shutdown systems and their instrumentation is performed in the CANDU nuclear power plants of Ontario Power Generation (OPG) and Bruce Power. Measurements of in-core flux detector (ICFD) and ion chamber (I/C) signals responding to the insertion of shut-off rods (shutdown system No. 1, SDS1), or to the injection of neutron absorbing poison (shutdown system No.2, SDS2) are regularly carried out at the beginning of planned outages. A reactor trip is manually initiated at high power and the trip response signals of ICFDs and I/Cs are recorded by multi-channel high-speed high-resolution data acquisition systems set up temporarily at various locations in the station. The sampling of the seaprate data acquisition systems are synchronized through the headset communication systems of the station. A total of 120 station signals can be sampled simultaneously up to 2500 samples per second. The effective prompt fractions of the ICFDs are estimated from the measured trip response. Effectiveness and the timeline of the trip mechanism are assessed in the measurement as well. The measurement can identify ICFDs with abnormally slow response (under-prompt) or overshooting response (over-prompt) at the beginning of the outage. The time required for the signals to drop to predefined fractions of their pre-trip values (level crossing time) is plotted as a function of detector position and compared against safety requirements. The propagating effect of shut-off rod insertion or poison injection on the flux is monitored by the level crossing times of ICFDs and ion chambers.     

Probabilistic method for determining optimal ROP detector locations in the CANDU-6 reactor

In order to maximize the regional overpower protection (ROP) trip margin under any condition in the CANada Deuterium Uranium (CANDU) reactors, several methods have been devised and applied at the sites, such as steam generator cleaning and adjustor rod lock-in. However, the operating margins obtained from these techniques are calculated based on the current locations of the fixed in-core ROP detectors. There is a possibility for increasing the trip margin if insignificant ROP detectors are removed and new critical detectors are added at different positions that do not conflict with the current locations. In fact, AECL had proposed such a deterministic detector layout optimization (DLO) technique to minimize the number of ROP detectors in 1998. However, this deterministic approach could not propose the best or optimized three-channel solution for each shutdown system, which has the maximum trip setpoint. Recently, in order to overcome the defect of the current DLO method, KEPRI adopted a probabilistic approach to determine the ROP detector location and incorporated it in the ROP design code, ROVER-K. To verify the applicability of the new method, the optimal ROP shutdown system was obtained for an initial and aged core condition, respectively, based on only the current 58 detector locations. The results show that the total number of ROP detectors decreases from 58 to 46 or 49 and the small TSP gain is obtained simultaneously based on the re-arranged best three safety channels. Therefore, the new method can be used to select the best ROP detector locations and also to estimate the optimal ROP TSP for an aged CANDU reactor to recover the operation margin.     

Analysis of the Leibstadt power plant condensate and feedwater systems during selected operational transients

This paper presents a selection of plant analyses that were carried out by PSI in support of the Leibstadt Nuclear Power Plant (Swiss boiling water reactor). The analyses were performed as part of a collaboration between Leibstadt and PSI, to help resolve some operational problems that were experienced during the power uprate beginning in 1998. The issues under investigation were related to the behavior of the condensate and feedwater systems during transients initiated by a turbine trip, load rejection and a single feedwater pump trip, all of which increased the risk of an inadvertent reactor shutdown by reaching reactor pressure vessel water level limits. The possibility of a reactor shutdown was related to perturbations in the feedwater flow caused by transitory pump cavitation of the feedwater pumps, due to a rapid depressurization in the feedwater tank. In addition to a direct analysis of plant measurement provided by Leibstadt, steady-state and transient simulations of the events were performed at PSI using the system codes TRAC-BF1 and TRACE. Through a combination of the analysis of the plant measurements, the code simulations and an analysis of the whole plant behavior using the Leibstadt plant-simulator appropriate modifications of the plant hardware, control system and operational set points were proposed. The implementation and success of these changes were verified by a number of plant tests. Finally, the original designed plant capability not to shutdown during the aforementioned transients was demonstrated.     

Digital Computers in Candu Safety Systems Part I History and Concepts

For many years, digital computers have been used in CANDU (CANada Deuterium Uranium) reactors for direct digital control as well as control room functions such as alarm annunciation, data logging and the display of operating data on the control panels. However, until recently computers were not used in the special safety systems. This paper examines the increasing role computers are playing in CANDU safety systems, especially the two shutdown systems. The reasons for this strong trend toward increased use of computers are outlined and recent designs are described, with special emphasis on system concepts. A companion paper (Part II) describes implementation details for the safety system computer applications and summarizes the experience gained so far during development and operation of these systems.     

Digitial Computers in CANDU Safety Systems: Part II Implementation and Experience

Microcomputer based systems have recently been applied to CANDU safety systems. This paper describes three specific systems, a monitoring system, the programmable digital comparator (PDC) system, and a fully computerized shutdown system prototype. The system configurations, the hardware and the software used to implement these designs are discussed. Each of these systems uses commercial off-the-shelf hardware which has been modified and qualified when necessary to meet specific power plant requirements. The preliminary experience which has been gained with the systems already in operation also summarized.     

A model for estimation of reactor spurious shutdown rate considering maintenance human errors in reactor protection system of nuclear power plants

Modeling of spurious activations in safety instrumented systems has been studied for over a decade. The spurious activation of a plant protection system in nuclear power plants (NPPs) leads to increased electricity generation cost. An in-depth view on spurious activation of digital plant protection systems of NPPs for human errors in maintenance tasks is presented in this paper. A new model which considers human errors in maintenance and periodic tests to predict component failure rates is presented. The model has been applied to OPR-1000 reactor protection system for quantification of spurious trip frequency by fault-tree analysis. The major causes of spurious activation in a nuclear reactor protection system are identified. A set of case studies has been performed with the variation of magnitudes of human errors probability and maintenance strategies, in which, the human errors in maintenance are found to significantly influence reactor spurious trip frequency. This study is expected to provide a useful mean to designers as well as maintainers of the digital reactor protection system to improve plant availability and safety.     

System assessment of an FPGA-based RPS for ABWR nuclear power plant

The instrumentation and control (I&C) systems for the Lungmen nuclear power plant (LMNPP) are fully digitized based on microprocessor and software technology, and extensively utilize multiplexing networks. That is, undetectable software faults and common cause failures due to software errors may occur, and that will defeat the redundancy of a nuclear power plant (NPP). A diverse backup implementation for the digital I&C systems is an important means to defense against undetectable software faults.This paper presents system assessment of a quad-redundant reactor protection system (RPS) design for an Advanced Boiling Water Reactor (ABWR) by utilizing the field programmable gate array (FPGA) technology. The FPGA-based RPS has been assessed by using a full-scope engineering simulator for the LMNPP. Accident scenarios and abnormal conditions are inserted into the engineering simulator in order to activate the function of the FPGA-based RPS. In this study, conceptual design of the proposed quad-redundant FPGA-based RPS, including preliminary hardware architecture, software design and system assessment will be presented. The results demonstrate that the FPGA-based RPS system is a practical approach to implement a diverse backup for the digital I&C system of nuclear power plant applications.Also, the sensitivity study of probabilistic risk assessment (PRA) shows that RPS combined with ARI (Alternative Rod Insertion) contributes significant influence on the core damage frequency (CDF) calculation of LMNPP. The PRA sensitivity study is independent of the RPS technology.     

基于应用可编程逻辑器件的真空联锁保护控制系统

本文给出了国家同步辐射实验室光束线真空保护的控制原理,以及该光束线真空联锁保护控制的逻辑及要点.与此同时也给出了应用可编程逻辑器件(CPLD)来实现此真空联锁保护控制系统的方法,使真空联锁保护系统的响应速度与可靠性得到提高.     

数字电源控制模块的设计

为加速器高精度磁铁稳流电源设计了数字电源控制模块DPSCM，以硬开关拓扑结构的磁铁电源作为被控对象，实现电源的全数字化控制。DPSCM以现场可编程门阵列FPGA为控制部件，实现对高精度ADC和DAC的控制，由数字调节器产生高精度数字脉宽调制信号，并实现电源的逻辑控制和联锁保护功能。通过模拟负载测试了DPSCM的基本功能，并在数字电源样机上测试了DPSCM长期运行的可靠性及稳定性，样机电源连续运行72h，电流稳定度优于5×10^-5。     

Second trip system for NRU research reactor

For the past four decades, the NRU research reactor has played an important role at the Chalk River Laboratories, Atomic Energy of Canada Limited, serving as one of its major research and isotope production facilities. To ensure that it continues as an effective facility, compliant with the current safety standards, a comprehensive upgrade program is underway. Adding a second trip system (STS) is part of this upgrade program, aiming at improving the effectiveness and reliability of the overall shutdown function. This document describes the main features and basic principles of the STS.The STS is an independent, seismically qualified trip system, that guarantees reactor shutdown even if the existing trip system fails. It is designed based on 2 out of 3 general coincidence logic, with minimal interferences and changes to the existing system. In addition to the manual trip in the main control room, a remote manual trip is provided in the new Qualified Emergency Response Centre, which is also seismically qualified and always accessible. Thus, for any reason, if the main control room becomes uninhabitable, the reactor still can be manually shut down from this centre.     

Regional overpower protection analysis for CANDU 6 reactors using ROVER-F code

The regional overpower protection (ROP) systems (also known as neutron overpower protection (NOP)) for the CANDU^® 600 MW (CANDU 6) reactors are analyzed using the ROVER-F computer program, developed by Atomic Energy of Canada Limited (AECL). The objective of all such analyses is to ensure that the ROP systems protect the reactor from local overpowers in the fuel, which would reduce the safety margin to fuel dryout. For CANDU 6, the methodology utilized in the ROVER-F code has been used for redesigning the ROP systems and is currently used to perform periodic updates of the ROP trip setpoints of the operating CANDU 6 stations. This paper provides an overview of the methodology used in the ROVER-F code and highlights recent developments in the code.     

基于SOPC的核磁共振可编程脉冲源设计

研究了基于SOPC(可编程片上系统)技术的脉冲核磁共振(Nuclear Magnetic Resonance,以下简称NMR)设备中脉冲源的设计,充分利用了NiosⅡ软核处理器的强大处理能力以及FPGA设计的灵活性,构建了SOPC系统,阐述了系统总体设计方案,重点介绍了FPGA中可编程多脉冲产生逻辑组件的设计,给出了部分实现程序及任务逻辑仿真波形,最后在示波器上显示出清晰、稳定、质量好的多脉冲输出波形,有较高的应用价值。系统软硬件易于修改和升级,可满足不同应用场合对脉冲核磁共振中脉冲序列的各种要求。     

基于BEPC Ⅱ的数字束流位置探测器信号处理算法的FPGA实现

数字束流位置探测器(BPM)算法是数字BPM系统最核心的部分，其对束流位置测量的精度起决定作用。本文在完成数字BPM算法MATLAB模拟工作的基础上，将模拟优选出的数字BPM算法在自制的电子学硬件上进行FPGA实现。首先介绍了数字BPM算法的总体设计和实现方案；其次介绍了数字BPM算法各功能模块的设计原理及其在FPGA中的具体实现方法；最后在输入信号频率499.8 MHz、强度-10 dBm、BPM探头灵敏度系数23的条件下进行了实验室测试。实验结果显示：逐圈位置分辨达2.96 μm，快响应位置分辨达0.65 μm，闭轨位置分辨达0.33 μm，验证了本算法在束流位置测量中具有良好性能。     

新一代CANDU—9重水堆核电厂简介

ＣＡＮＤＵ－９是电功率为９００ＭＷ级的重水堆核电厂,其设计基于达灵顿和布鲁斯Ｂ多机组核电厂,并融入了一些最新的工程设计和研究成果,除了继续采用成熟的系统和部件外,在安全性,地可靠性和可维护性方面作了重要改进。ＣＡＮＤＵ－９综合考虑了安全审评和执照申请过程中发现的问题,产使其体现在安全设计理念中,特别是对慢化剂系统,端屏蔽冷却系统,系统和应急堆芯冷却系统进行了改进。     

Quantitative modeling of digital reactor protection system using Markov state-transition model

Recently, digital instrumentation and control systems have been increasingly installed for important safety functions in nuclear power plants such as the reactor protection system (RPS) and the actuation system of the engineered safety features. Since digital devices consist of not only electronic hardware but also software that can control microprocessors, the functions specific to digital equipment such as self-diagnostic functions have been becoming available. These functions were not realized with conventional electric components. On the other hand, it has been found that it is difficult to model the digital equipment reliability in probabilistic risk assessment (PRA) using conventional fault tree analysis technique. OECD/NEA CSNI Working Group of Risk Assessment (WGRisk) set up the task group DIGREL to develop the basis of reliability analysis method of the digital safety system and is now discussing about several issues including quantitative dynamic modeling. This paper shows that, taking account of the relationship among the RPS failures, demand after the initiating event, detection of the RPS fault by self-diagnostic or surveillance tests, repair of the RPS components and plant shutdown operation by the plant operators as a stochastic process, the anticipated transient without scram (ATWS) event can be modeled by the event logic fault tree and Markov state-transition diagrams assuming the hypothetical 1-out-of-2 digital RPS.     

Modelling the activity of I in the primary coolant of a CANDU reactor

A mathematical treatment has been developed to describe the activity levels of ¹²⁹I as a function of time in the primary heat transport system during constant power operation and for a reactor shutdown situation. The model accounts for a release of fission-product iodine from defective fuel rods and tramp uranium contamination on in-core surfaces. The physical transport constants of the model are derived from a coolant activity analysis of the short-lived radioiodine species. An estimate of 3×10⁻⁹ has been determined for the coolant activity ratio of ¹²⁹I/¹³¹I in a CANDU Nuclear Generating Station (NGS), which is in reasonable agreement with that observed in the primary coolant and for plant test resin columns from pressurized and boiling water reactor plants. The model has been further applied to a CANDU NGS, by fitting it to the observed short-lived iodine and long-lived cesium data, to yield a coolant activity ratio of ∼2×10⁻⁸ for ¹²⁹I/¹³⁷Cs. This ratio can be used to estimate the levels of ¹²⁹I in reactor waste based on a measurement of the activity of ¹³⁷Cs.     

Fault tree analysis of CANDU shut-down system

The CANDU shut-down system comprises electro-mechanical shutdown rods and liquid poison injection, each of which includes sensors, instrument channels and mechanical and fluidic subsystems. Published work so far has reported 10⁻⁵ for the unavailability of the reactor protection system. The basic component failure rate is assumed to be constant and to have a lognormal distribution. Mathematical models are developed to analyse the CANDU shutdown system, with aging of basic components. The Weibull distribution is used because it has a lower standard deviation. An unavailability of 10⁻⁴ is obtained. Time constraint on system operation and aging of components over a year do not significantly affect system unavailability.     

ADORE: Alternative detector layout optimization for ROP system of CANDU reactors

In CANDU^® reactor design, the regional overpower protection (ROP) systems protect the reactor against overpower in the fuel which could reduce the safety margin-to-dryout. Specifically for the CANDU^® 600 MW (CANDU 6) design, there are two ROP systems in the core, one for each fast-acting shutdown systems. Each ROP system includes a number of fast-responding, self-powered flux detectors suitably distributed throughout the core within vertical and horizontal assemblies. The placement of these ROP detectors is a challenging discrete optimization problem. The DLO (detector layout optimization) module of ROVER-F code was used to design the existing ROP detector layout of CANDU 6 reactors. In the past couple of years, a new methodology for designing the detector layout for the ROP system, called DETPLASA algorithm, has been developed. This method utilizes the simulated annealing (SA) technique to optimize the placement of the detectors in the core. This algorithm was developed to overcome the shortcoming of DLO method to produce a detector layout configuration when the size of the problem is large. An alternative method has been recently developed for solving the ROP detector placement problem. This method is called ADORE (Alternative Detector layout Optimization for REgional overpower protection system). Although technically any stochastic optimization technique can be utilized, presently this method utilizes the SA technique as its optimization engine. This paper presents an overview of ADORE methodology and provides some numerical results from its execution.     

The importance of uncertainties in the ROP detector layout design process for CANDU reactors

In CANDU® reactor design, the regional overpower protection (ROP) systems protect the reactor against overpower in the fuel which could reduce the safety margin-to-dryout. The increase in fuel power could be caused by a localized power peaking within the core (for example, as a result of a certain reactivity device configuration) or a general increase in the core power level during a slow-loss-of-regulation (SLOR) event. This overpower could lead to fuel sheath dryout. In the CANDU® 600 MW (CANDU 6) design, there are two ROP systems in the core, one for each fast-acting shutdown system. Each ROP system includes a number of fast-responding, self-powered flux detectors suitably distributed throughout the core within vertical and horizontal assemblies. A new methodology for designing the detector layout for the ROP system, called the DETPLASA algorithm, has been developed recently. This method utilizes the simulated annealing (SA) technique to optimize the placement of the detectors in the core. The evaluation of the trip setpoint (TSP) corresponding to each detector layout configuration (i.e., each history in the SA algorithm) is performed probabilistically using the ROVER-F code. In this evaluation, there are uncertainties related to both the detector components (i.e., related to the margin-to-trip) and to the fuel channel components (i.e., related to the margin-to-dryout). In this paper, the importance of these uncertainties on the outcome of the detector layout optimization process is evaluated. Some parametric studies have been performed to quantify the effect of uncertainties on the resulting detector layout. Two types of investigations have been performed. First, a given detector layout will be used to explicitly determine the effect of changing the uncertainty values. In this study, 343 sets of uncertainty values are used to produce the corresponding TSP values. The variation in the TSP values is analyzed. Second, three sets of uncertainty values (a subset of uncertainties from the first study) are used in independent DETPLASA executions. The resulting detector layout configurations will be examined to observe the effect of these uncertainties on the final design. Results from these investigations are presented in this paper.