Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Advanced electrical systems: From megabyte to megawatt' taking place on Wednesday, 12 March 2014 at 09:00-10:30. The meet-the-authors will take place in the poster area.

Pinaki Mitra ABB, Sweden
Co-authors:
Tomas Larsson (1) F P
(1) ABB, Västerås, Sweden

Printer friendly version: printer.gif Print

Abstract

ABB's HVDC grid simulation center - A powerful tool for real-time research of DC grids

Introduction

With the present advent of renewable integration in the electric power systems many new challenges will appear. Some are already known and others are expected. Yet, there may be a number of so far unknown issues appearing in these new applications that have to be handled. These issues have to be treated before a full-scale installation.

Approach

Issues appearing in a full-scale wind-power installation will in most cases cause both extended time schedules and project delays as well as increased costs and budget over-runs. Generally speaking, the earlier issues can be resolved, the better both from a time and cost perspective. However, the earlier in time the work is done, more uncertainties are introduced. Among the largest uncertainties are how the electric grid will react to the implementation of remotely installed wind power, how will the receiving point(s) take care of disturbances and how shall the total system be operated in a coordinated way?
Traditionally, there are different ways to reduce the uncertainties. One very common is to simulate the systems and investigate how they react during energization, steady state operation, contingencies and faults. Most commonly, non real-time simulation is used, i.e. a ten second run will take more than ten seconds to run in a simulation environment. Advantages with this technology are modest hardware requirements and moderate costs, relatively fast setup and an easy-to-change environment as everything operates in software. Among the drawbacks are the long execution times and the lack of real-time operation that could be used to verify simulation-external equipment, for example control systems.
As an alternative, real-time simulation can be used. In such a setup, ten second really takes exactly ten seconds to perform. This has obvious advantages as everything around the simulation will see the simulation as if it could have been a real system. As such, the surrounding system can be tested, verified and debugged as if it had been in real life, and with the great advantage of doing it off-site with close access to all support functions needed for trouble-shooting. Compared to performing similar activities at site, this is a great advantage.


Main body of abstract

Because of advantages mentioned earlier, the current state-of-the-art procedure for control system factory testing in HVDC projects under delivery involves real-time testing. The set-up contains equipment to mimic the main circuit, i.e. the high-voltage part of the installation is substituted by a computerized real-time system working with massive parallelism. Here one can find both passive components such as transformers, reactors, capacitors, cables, arresters etc. Also active parts such as semi-conductor based converters are modeled in the real-time system. The latter has shorter time constants which are reflected in the real-time models where the converter is modeled in a shorter time step than for those components in the first group.
Now, having a mimicked main circuit, a real control system may be connected via its usual I/O, input/output, circuitry as if it had been in real life. As such, the differences are in most cases negligible and the results can be used as if it had been for a real full installation.
This opens up for possibilities to perform studies of both normal states such as energization, steady state operation and switching of the circuit. An even greater opportunity is with real-time simulation opened for protection testing and verification. Here faults can be introduced, which for a real plant would be extremely complicated to set up if it at all possible. Through the real-time setup, with a main circuit model and the real control/protection system, great confidence can be gained that it will fulfill its task also under the most severe faults.
With the ongoing advent of renewable integration solved with for instance HVDC point-to-point connections or complete DC grids, new and extended needs are raised for real-time simulations. With requirements on selective tripping of equipment on the DC side and fault ride-through abilities, a clear need for modeling of a DC breaker is established. With a DC breaker, possibilities are opened up to fully maintain the unaffected parts of a DC grid in operation with a minimum of disturbances. For a point-to-point HVDC connection, the DC line can be tripped and the converters can remain connected, after the fault has been cleared, and provide reactive power.
ABB has set up an HVDC grid simulation center to test and verify control algorithms and system solutions. The system comprises a massive real-time simulation capacity for the main circuit and a commercial MACH2 control system with I/O and HMI, Human Machine Interface.
Another task of the simulation center is to demonstrate ABBs HVDC Grid system solution to external instants. Here, coordinated control of a multitude of converter stations is shown and the results from both AC and DC faults are given.
The full paper will elaborate on the challenge of verifying the real-time model and how confidence of the system is created. Simulation system topology, capacity and possibilities will be discussed as well as differences between the current state-of-the-art real-time testing and extensions for HVDC grid simulation requirements. Also results from the impact from both AC and DC faults will be given.


Conclusion

For DC grids and modern point-to-point HVDC connections, new system requirements are raised for abilities to ride through faults and ensure that unaffected parts of the transmission will remain in operation under the fault and after it has been cleared. Compared to previous solutions, where a larger part of the system was tripped following a fault, the recent requirements are to keep as much as possible online and in operation. One specific requirement is to keep the converter stations in operation even if a DC line fault should occur. They will then go over into compensation of reactive power and contribute to the AC grid control and stability.
To handle these new requirements, an essential system component is the DC breaker and its control system. This has been developed by ABB and its model implemented in the ABB simulation system. Various operation cases and fault cases have been run and the results will be shown in the full paper. In the associated discussion, it will be shown that DC side fault currents will be driven both from the AC and DC sides. Unless interrupted these currents will rise very quickly and soon reach unacceptable levels. To conclude, a very quick response is required, which is enabled through the DC breaker.
Through real-time simulations of DC grids and involving new models and control of DC breakers, many of potentially time-consuming and costly issues can be handled in a systematical way. Also confidence will be gained on the coming full-scale system’s ability to manage normal operation as well as contingencies and faults.



Learning objectives
In ABB’s DC grid simulation center, the learning objectives are to verify and test new algorithms for a DC grid. This includes all aspects from energization through steady state operation to contingencies and fault handling with DC breakers.


References
To come later, as well as figures/photos