DC Grid Protection Aware Planning of Offshore HVDC Grids

Name: Jaykumar Dave

 Prof. dr. ir.  Dirk Van Hertem (promotor)
 De heer  Hakan Ergun (co-promotor)

The world economy has been heavily reliant on fossil fuels with the use of natural gas, oil and coal. This has added a significant amount of carbon dioxide (CO2) in the atmosphere in last few decades. The excessive amount of CO2 warms up the earth through the greenhouse effect. This has numerous consequences including the rise in the sea levels, extreme weather events, droughts, floods, forest fires and many other climate related disasters. The Paris Agreement is an important step to respond to the climate change, which sets the goal for all countries to keep the temperature rise well below 2 °C. As the fossil fuel based electricity generation forms a significant part of total greenhouse gas emission, the decarbonization of electricity supply is necessary to achieve this target. The wind and solar are the prime candidates to replace the conventional fossil fuel based power plants.

The wind farms can be placed either onshore or offshore which are both set to rise in order to meet Europe’s climate target. Specifically, the capacity of offshore wind farms are estimated to increase by 20 times until 2050. In the North Sea alone, the offshore wind capacity is foreseen to rise to 212 GW from the current capacity of 20 GW. At present, the offshore wind farms are connected to the shore individually, using submarine cable. However, a huge number of individual connections will be required to connect the above-mentioned capacity. Instead, an electricity grid can be formed within the sea that connects multiple windfarms at offshore before finally connecting them to shore. The electricity grid also has multiple economical and technical advantages over individual connections. The High Voltage Direct Current (DC) technology is more suited to perform this task than traditionally used High Voltage Alternating Current (AC) technology for onshore grids. This is because HVDC technology offers in the lower power losses and allows better control over power flows.

The planning of offshore HVDC grid offers several challenges over the traditional onshore HVAC grid planning. First, the layout of the grid is determined by solving a mathematical problem that is highly nonlinear and mixed integer. This type of problems is very difficult to solve by existing algorithms and is not solvable for large grids. Hence, the problem is simplified to obtain a solution. The widely used simplification makes assumptions about losses, reactive power and voltages. This simplification, called `DC’ approximations, leads to inaccuracies in the final layout. This thesis derives better simplifications of the planning problem and offers three different alternatives to `DC’ approximations. These alternatives provide better accuracy at reasonable computational speed.

Second challenge is related to the frequency stability. The frequency is a measure of balance between load and generation. If the balance is lost, via e.g. tripping of a load or power plant, the frequency deviates from its nominal value of 50 Hz. The electrical equipment are built to operate at the nominal frequency and may get damaged if operated at other frequencies. Hence, the power grid must be prepared against any possible unbalances to maintain the frequency to its nominal value. For the Nordic power system (Sweden, Norway, Denmark and Finland), the maximum power unbalance can happen due to loss of a nuclear power plant of 1.4 GW in Sweden. The Nordic power system is equipped with sufficient counter actions to maintain the frequency against loss of this plant. However, with the increasing amount of offshore wind farms, the future maximum unbalance could result from the offshore HVDC grids e.g. tripping of a DC line. The present planning methods deal with this issue by restricting the GW capacity of injected power from an offshore HVDC connections below the current maximum unbalance. This thesis lifts this restriction by using the DC grid protection which is capable of restoring the power injection quickly after a DC line fault. The thesis proposes novel mathematical models to include this aspect into operation and planning of HVDC grids. It is concluded that the full selective protection system that very quickly and selectively isolates the faulted line is the most economical.

https://lirias.kuleuven.be/handle/20.500.12942/695739

 Prof. dr. ir. Dirk Van Hertem (promotor)
 De heer Hakan Ergun (co-promotor)
 Prof. dr. ir. Paul Sas (chairman)
 De heer Willem Leterme (secretary)
 Prof. dr. ir. Erik Delarue
 Prof. dr. Michael Kleemann
 Prof. Keith Bell , University of Strathclyde (Glasgow)
 Prof. Jalal Kazempour , Technical University of Denmark
 Dr. Plet Cornelis , DNV