Reactor simulator development facility for operating personnel training

Full text: A development facility of reactor simulators (FARSim) for operator training is presented. The simulator development facility can be divided into four main modules: the model manager (MM), the simulator human machine interface (SHMI), the instructor station (IS) and the simulation manager (SM). It is designed as a distributed system where each module takes charge of an specific simulator task that could run in the same computer or distributed in a computer network. Only the SHMI runs in a remote computer (specific hardware). The process distribution is configurable at the start of the simulation session. This type of process distribution makes it scalable. The main module is the Simulation Manager (SM) which is responsible of routing the messages between the other modules and managing the simulation. The Model Manager (MM) interfaces to the plant mathematical model (PMM). The Data Base Manager (DBM) handles the data base, in order to access and save necessary information during the simulation. The timing of the simulation is accomplished by the Clock Manager (CM) and the error and system messages are handled by the Logger. The Instructor Console (IC) is the process where the instructor commands the simulation. The SHMI is a process interface with the simulator SCADA (which can be identical to the plant SCADA) and is used by the trainee to observe the simulated plant output and to act upon it. The PMM source code, together with the necessary libraries is encapsulated into the MM to implement the model initialisation and the one-step simulation. The MM provides the initial conditions (ICs), and the model input, the PMM provides the plant output. Depending on the complexity of the model, it can be divided and the calculations can be distributed among different CPUs. The PMM source code generation is based on the Matlab-Simulink-Real Time Workshop simulation development environment. The models are developed on graphical windows interfaces based on component libraries, where they can be easily constructed, debugged, tested and compiled to C source code. A specific Simulink library with several thermo-hydraulic, nuclear and electrical components has been developed in order to construct the plant mathematical model. The parameterisation of these components (lengths, hydraulic areas, kinetic constants, equipment parameters) can be linked to inputs of reactor transient analysis codes (or any other plant design database), such as RELAP, in order to automate the model maintenance. Model limitations are included in the design database in order to implement out-of-scope messages to the instructor console. All these features together with several help and documentation facilities ease the model maintenance and modification process. The scope of the PMM is deep enough to represent correctly the tendency and timing of all plant variables relevant to the plant normal evolutions and to the malfunctions included in the design basis accident (DBA) list. For the DBA, model validation is performed against RELAP outputs. Typical normal evolutions included are: system enabling for start-up, approximation to critical state, critical state at zero power, manual holding of critical state, plant power operations, normal and fast reactor shutdown, recovery to rated power after a reactor trip. Typical malfunctions include: loss of electrical supply, excess reactivity insertion, loss of primary coolant accident, loss of flow accident, loss of heat sink, instrument/equipment/control loop failure, loss of shutdown cooling, failure or spurious activation of protection systems. The main functions that the instructor can perform using the IS are: load initial conditions (including different core lifecycles and plant states); start the simulation; pause the simulation; set the simulation mode on slow, fast or real time; restart the simulation; store snapshots (to be used as ICs); backtrack the simulation; replay the simulation; edit, define, save, load and activate malfunction sequences; and execute local actions upon trainee request, such as local variables inspection or equipment startup. The access to the local variables and equipment is through navigation on system hierarchies using plant tags. It also prints out-of-scope messages when the simulation runs out of the model limits. All these actions performed during a simulation session in the IS can be stored and further reloaded. The SHMI encapsulates the simulator SCADA, performing the communication of all the inputs/outputs with the SM and controlling the timing and initialisation of the different SCADA components (trend charts, control logics, mimics) to implement the different simulation velocities, the backtrack and the snapshot. If the same plant SCADA is used for the simulator, then this interface can be used to test changes on control logics or plant displays. This simulation facility is used to develop the RRR Reactor Training Simulator (RTS). The RTS is a full-scope and partial replica simulator. All the systems relevant to the plant normal evolutions and malfunctions listed in the DBA are included. Within these systems both the variables connected to the plant SCADA and the local variables are included, leading to several thousands input-output variables in the PMM. The same plant SCADA is used in the SHMI together with a touchscreen terminal to emulate the hard components of the main console (pushbuttons, switches). Some subsystems of the RRR RTS project are used to present the simulator features. (author)