Demonstration project 3 Technical specifications towards offshore HVDC networks (DCGRID)

Main finding

  • CO2 emission and new DC technology costs have been detected as influent economic parameters in the economic drivers analysis.
  • Simple structures (such as backbones) are feasible from both technical and regulatory standpoints and can provide ancillary services (voltage support, frequency control, PSS) and Fault Ride-Through capability to the AC network.
  • The technical feasibility of an innovative DCCB was proven through successful medium-voltage power tests (at the time of writing, the assembly of the high-voltage DCCB demonstrator is nearly completed).
  • In addition to the control of the converters, a DC fault detection and selection algorithm was designed and successfully tested on the low-scale DC grid mock-up.

Project description

In order to integrate large amounts of offshore wind generation into the European power system from 2020 to 2030 and beyond, wind farms will need to be developed, connected to the existing grid, and operated at the required security level for the new grid facilities of the whole power system. This will probably happen first in the North Sea area where wind generation zones have been determined, some of them far from the shore.

Traditionally, the integration of wind power is managed nationally or between two neighbouring countries. This results in radial connections for existing wind farms, mostly in AC, but also through DC submarine cables in case of a long distance to shore. Over the next 10 years and beyond, larger amounts of wind power throughout Europe, generated by wind farms far out to sea, will probably require regional approaches, based on one or several multi-terminal networks, meshed or not. This will not only bring wind power to shore at the right place, but also help smooth out any variations, offer new interconnection capacities to the electricity market, and to provide ancillary services, compensating decommissioned conventional generation.

To tackle the complex task of designing, developing and operating new trans-national grid facilities, based on existing and future technologies, the third TWENTIES demonstration project, DCGRID, examined the most suitable DC technology (Voltage Source Converter - VSC) roughly split into three time scales:

  1. Short-term technical feasibility analysis (up to 2020).
  2. Medium-term technical feasibility analysis (2020-2030).
  3. Long-term economic analysis (2030 and beyond).

Results in detail

For the long-term analysis, a probabilistic methodology was developed to compare several possible future network topologies in the North Sea area, based on spe-cific economic and reliability criteria. According to its actual assumptions, the study has shown that none of the topologies under concern can be definitely sorted out, meaning that refining assumptions and criteria has to be pursued, stressing at the same time that the risk level in developing and financing future offshore grids remains high.

The medium-term analysis had two goals:

I. The development by ALSTOM Grid of a DC circuit breaker prototype which passed successful breaking tests witnessed by an independent observer.

II. The design and development of a low-scale real time mock-up to illustrate the control and protection algorithms of a DC grid, as well as the interconnection of the physical DC grid to a simulated AC grid.

Two live demonstrations were carried out on 21 March 2013 in Villeurbanne and on 3 April 2013 in Lille.

The short-term analysis was backed up with simulations verifying that a step by step development starting from several radial connections was feasible. The control of the H-shaped topology (see Figure 3) was analysed in detail, and extended to the 'backbone' (either tree-like or meshed versions) and 'meshed five terminal' topologies (Figure 4). It was concluded that methods of local voltage droop control are efficient in operating the AC/DC interconnected grids, enabling to accommodate for intermittent wind generation as well as for a large range of disturbances without any need for high-speed telecommunication (so far such communication is required for DC fault detection only).

This was illustrated, along with DC grid protection algorithms, by the small-scale demonstration in Lille on 3 April 2013.