DISPATCHER TRAINING SIMULATOR

Role of the dispatchers that have to take decisions under stress pressure is central in the operation of the European Transmission Network (ETN) and is one of the major components of the security of the overall system. Recent instabilities on the European and USA power system have shown that human factor can have extreme consequences.

Dispatcher Training Simulator (DTS) is a key tool of the integration of the operation of the ETN. It is crucial to train system operators on multilateral cross border operational issues and events that might lead to large-scale blackouts if not timely tackled. This requirement is reinforced by the fact that there is an increasing number of electronic devices impacting significantly the electromechanical behaviours in the system. The present state of the art of DTS consists of using a replica of a single control centre linked to a quasi steady state power system model. This is no longer satisfactory. The PEGASE project has developed a real time simulation prototype meeting the requirements of the project: capable of simulating the entire ETN with high fidelity, whatever the electromechanical behaviour of the system, up to the complete blackout and the subsequent restoration. This simulation engine capitalizes on the deep knowledge of the algorithms gained along the project.

ACHIEVEMENT 1: NEW GENERATION OF GROUND-BREAKING ALGORITHMS



Developing a computation engine suitable for a European DTS is real challenge. Tests of an existing industrial DTS engine on so large system showed that the resulting simulations were 20 times slower than real time.
To achieve the required performances, new algorithms were needed. Different approaches were considered:

1. Fine grain parallelization approach for the function and Jacobean evaluation. It allows exploiting the new computer architecture presenting a constant increase of shared memory cores.

2. Advanced direct and iterative linear algebra algorithms dedicated to power system.

3. Domain decomposition methods allowing exploiting the new parallel computers characterized by a reduction of their clock frequency counterbalanced by a significant increase of the number of cores:
  • Schwarz method coupled with advanced preconditioning techniques to exploit shared memory architectures;
  • Waveform Relaxation algorithm to exploit distributed computation architectures.

4. Localization technique which allows a reduction in the computation burden by skipping the resolution of already converged components. These techniques correspond to the group of approaches common to all time domain simulations type presented in the Achievement 1 of Time Domain Simulation.
The DTS simulation engine uses the optimum mix of these techniques which must be coupled with a high attention to performance in the implementation. The DTS uses a fix step-size (typically 10 or 20 ms), as it allows to be faster during transients than the variable step size; while the variable step size approach is faster on average.
As compared to the algorithm included in the full accuracy prototype, the precision of the numerical process is just slightly reduced to the benefit of the robustness of the integration method whereas the purpose of the 2 prototypes is different.


ACHIEVEMENT 2: PROTOTYPES DEMONSTRATED ON PAN-EUROPEAN SYSTEMS

Based on the research results of the PEGASE project a DTS simulation engine prototype has been developed and implemented in an existing DTS framework.

This prototype includes:

  • The first real time dynamic simulation engine allowing the simulation with good accuracy of a European size power system. Real time is respected for the simulation of the whole ETN. The produced results are of very good accuracy (validated by comparison with the results produced by the EUROSTAG reference variable step-size software). The following graphs represented the performances reached by the prototype during a scenario where the ETN splits in 3 systems;
  • The possibility for the user to define the controllers using block-diagram. This feature allows the integration of all models developed in the frame of PEGASE into the DTS, including power electronics models. This functionality is also very important for a DTS used in production environment where the models must be easy to maintain and/or improve to follow the evolution of the real system;
  • A modern communication protocol. The integration of the OPC/UA communication protocol allows to easily connect the DTS engine to various modern MMI and/or SCADA;
  • The compatibility with the Full Accuracy prototype. All models handled by the DTS prototype can be exported to the Full Accuracy Prototype. It allows to validate solution of the DTS and provide the trainer with greater flexibility to set up complex scenarios.
    This prototype has been validated by the TSOs present in the PEGASE consortium for different scenarios on a realistic test model of the ETN. The model provided by ENTSOE as equivalent to the ETN:
  • is representative: implementing a typical set of detailed substation topology and associated protections obtained from TSO partners;
  • represents the same level of complexity as the real ETN;
  • has a European size: 24000 busbars, 12500 lines, 3800 generators, 125000 variables;
  • is able to reproduce typical phenomena of the European grid;


With the collaboration of the TSOs, the prototype has been:

  • Connected to a replica of the control centre of a consortium member for interactive demonstration;
  • Demonstrated through European-wide, extreme conditions scenarios:
  • Splitting of the ETN in 3 islands: Cascade of Line tripping splitting the system in two (1) and then in three islands (2) and consecutive resynchronization (3) using ENTSO/e POLICY 5 as shown in the figure below. The scenario illustrates some key features: detailed substation topology handling, simulation of frequency deviation in islanded parts of the system, conditions for resynchronization through synchro-couplers, dynamical behaviour of the ETN (4), etc.
  • Blackout of a whole country, blackstart, re-energization and resynchronization, after islanding of a TSO through overload protection and the action of under-speed relays, the scenario demonstrate a blackstart procedure, the re-energization of the TSO network and the reconnection of the country with the rest of the ETN;
  • Unexpected power shortage due to incorrect wind power forecast, combined with lack of reserve from conventional production, leading to overall power shortage and frequency drop and dangerous situation that should be handled by the dispatchers using corrective measures to avoid automatic load shedding.

ACHIEVEMENT 3: PAVING THE WAY FOR FUTURE RESEARCH AND FOR INDUSTRIALIZATION

PEGASE project brings a major breakthrough in real time dynamic simulation. At the end of the project, a fully validated prototype able to simulate in real time the entire ETN system is operational. It has been connected to a copy of the ELIA SCADA/ MMI system and uses the static and dynamical data of the Belgian TSO together with a very complex large scale model of the rest of the European interconnected system.

ETN DTS architecture has been designed. Both the prototype and the architecture allow large scale demonstration of the solution and its added value. To get closer to a real European DTS and to validate the DTS infrastructure proposed in this document, the next natural step is to build a demonstrator for the ETN DTS. This demonstrator project should focus on:

  • Data issues: static and dynamic data must be gathered from different TSOs in order to be merged in the simulator. The availability of models is driven in the TSOs by the needs of analysis by other applications addressing also the system dynamics (e.g. Dynamic Security Assessment, offline simulations...). These data and models will be developed and aggregated to assess the overall security of the future ETN on dynamic tools and, of course, can be shared with the DTS application. In case real data are not available for some of the TSOs, generic data and/or models can be used instead. A common information model has to be agreed between all the involved parties;
  • Model validation issues: the data gathered for the model are supposed to be duly validated by the different TSOs. Anyway, the merged model needs to be globally validated. Using a DTS for training operators on high end scenarios, checking common operational procedures, and testing advanced control tools will consolidate the quality of the model used in other applications such as static and dynamic security assessment or study tools;
  • Visualization: Advanced Man-Machine Interfaces have to be designed to help operators to understand the state of the system and to have their attention drawn towards system-wide phenomena. These interfaces should be developed also for daily operation, the DTS can be used to develop and validate them;
  • Legal issues: data gathering will induce several confidentiality questions linked to commercial sensitive data which have to be addressed. Modification of the legal framework for technical data sharing;
  • Logistic issues: training of operators is currently organized in TSO or in specialized companies. Joining forces in a common European Training Centre enable economy of scale. Alternatively, independent entities could use technology developed in PEGASE to setup multi TSOs training services.







PUBLICATIONS

    The following non-exhaustive list of publications illustrates the numerous scientific advances performed during the PEGASE project:

  • D. Fabozzi, T. van Cutsem, “Assessing the Proximity of Time Evolutions through Dynamic Time Warping” Proc. IET Generation, Transmission & Distribution (2011), 5(12), pp. 1268-1276.


  • D. Fabozzi, T. Van Cutsem , “Localization and latency concepts applied to time simulation of large power systems”, Proc. IREP Symposium on Bulk Power System Dynamics and Control - VIII, Buzios (Brazil), 1-6 Aug. 2010.


  • V. Savcenco, B. Haut, E. Jan W. ter Maten, R. M.M. Mattheij, “Time domain simulation of power systems with different time scales”, Proceedings of Scientific Computing in Electrical Engineering (SCEE) Conference, Toulouse, France, September 19-24, 2010.


  • V. Savcenco, B. Haut, “Multirate Integration of a European Power System Network Model”, Proceedings of 8th International Conference of Numerical Analysis and Applied Mathematics, Volume 1281, pp. 2037-2040, 2010.


  • B. Haut, V. Savcenco, P. Panciatici, “A multirate approach for time domain simulation of very large power systems”, HICSS 45 Proceedings, pp 2125-2132, 2012.


  • F. Pruvost, P. Laurent-Gengoux, F. Magoulès, F.X-Bouchez, “Speed-up the Computing efficiency of waveform relaxation for Power system Transient Stability”, SC’11, Seattle, WA, USA, November 13, 2011.


  • F. Pruvost, T. Cadeau, P. Laurent, F. Magoulès, F.-X. Bouchez, B. Haut, “Numerical Accelerations for Power Systems Transient Stability Simulations”, Proceedings of the 17th PSCC conference, Stockholm, Sweden, 22-26 August 2011.


  • F. Pruvost, T. Cadeau, P. Laurent, F. Magoulès, F.-X. Bouchez, B. Haut, “Numerical Accelerations for Power Systems Transient Stability Simulations”, Proceedings of the 17th PSCC conference, Stockholm, Sweden, 22-26 August 2011.


  • D. Fabozzi, A.S. Chieh, P. Panciatici, T. Van Cutsem, “On simplified handling of state events in time-domain simulation”, Proceedings of the 17th Power System Computation Conference (PSCC), Stockholm, Sweden, Aug. 2011.


  • F.-X. Bouchez, B. Haut, L. Platbrood, K. Karoui, “HPC for power systems in the framework of the PEGASE project”, Proceedings of IEEE PES General Meeting 2012, San Diego, CA, USA, 22-26 July 2012..


DOWNLOADS