Types of efficiency problems which can be addressed are:
- fouling and internal leak in heat exchangers
- leaking valves
- drift in sensors
The system is configured for use with steam turbine cycles and has been developed specifically with the multi-stage wet steam turbines typical in NPPs. In addition, its use of generic process components allows the modelling of a variety of other systems such as steam turbines in fossil fuelled plants as well as other auxiliary systems involving heat transfer by a water medium.
Since its first release in 2000 the TEMPO system has been used to model the turbine cycles of several NPPs. Installations include Forsmark 3, Loviisa I & II, Olkiluoto I & II with feasibility studies being carried out for numerous other NPPs.
The TEMPO system is based on the use of physical models as an approximation to the actual plant characteristics. These physical models can either use heat and mass balance constraints or can be extended to include other physical correlations. Models are constructed using the TEMPO Modeller, figure 1.
Figure 1. The TEMPO Modeller.
Examples of typical physical correlations required for turbine systems are those governing turbine efficiencies, turbine constants, heat transfer rates in heat exchangers. The type of model to use depends on the intended use. To support different uses TEMPO offers three different calculation modes.
- Data fitting mode
- Optimisation mode
- Simulation mode
Data Fitting Mode
Selected parameters in the physical model are fitted by data reconciliation so that calculated process states best match a given set of measurement data. In this mode the physical models are as much as possible based only on the principal of conservation of energy and mass. The TEMPO Modeller can be set in an on-line mode to automatically calculate this best fit when new measurement data is received.
The difference between measured data and values calculated by the physical model, measurement residuals, can be used to identify measurement or equipment errors, see figure 2.
Figure 2. Residual plot of data showing faulty measurements.
This requires more detail in the physical model. This detail can be obtained by tuning the physical characteristics to real plant data by using the fitting mode. The optimisation mode can then be used to see which settings of control parameters will give the maximum thermal efficiency.
This requires the highest degree of detail and consistency in the physical model. Physical characteristics need to be fitted for a wide range of plant operational states. The model can then be used to predict possible plant states with either new equipment or environmental conditions.
An important part of the TEMPO system is the analysis of results from the fitting mode calculations. For condition monitoring of turbine cycles the empirical analysis of a series of results can give better information about the plant state than that gained from just a single time point, see figure 3. Output data from the TEMPO Modeller can be stored in a Microsoft Access Database. This allows a variety of software to be used to analyse the data. The TEMPO system also contains its own software for this analysis, figure 4.
Figure 3. Time series of χ2 object function. Deviations from normal behaviour indicates faults.
Figure 4. The TEMPO Analyser.