The aim of this thesis is to (1) enhance understanding of mechanisms that are important for the protection of high-voltage direct-current (HVDC) grids, and (2) suggest possible technical solutions. To that end, digital technologies were used both in simulation, as well as in a laboratory environment.

Literature studies were carried out on fault detection algorithms and substation communication leading to a recommendation to use a combination of single- and double-ended algorithms for fault detection, as well as to use the EtherCAT protocol for substation communication.

A limitation of simulation studies are possible parameter uncertainties. For that reason, large protection margins by design are important. The simulation part of this thesis includes, firstly, a study concerning protection margins showing the detrimental effect of not being able to share information in a multi-vendor context. Secondly, a new method is presented for holistic protection system design taking into account a large variety of parameters and making sure that no hardware or software constraints are violated. For this, genetic optimization was found to be the most suitable technique. It is found that the holistic method is particularly useful for complex optimization problems, such as HVDC grids with different DC circuit breaker opening times and no converter blocking. In one test case, the genetic optimization resulted in a 71% decrease of total inductor size compared to the initial dimensioning provided by an engineer.

Due to the destructive nature of faults, HVDC protection can obviously not be systematically tested full-scale or even in a laboratory environment. Still, real-time testing using real controllers or protection devices is useful because it is more realistic than offline, electromagnetic transient simulations. In this thesis, an intelligent electronic device (IED) prototype for HVDC grid protection was developed, providing a crucial device for subsequent studies on IED type testing and HVDC protection system testing, both of which were conducted outside of this PhD work. A test of the IED prototype with actual fault recordings from an operational HVDC link further increased confidence in HVDC protection, because the successful testing is based on both a real protection IED, and a real fault recording, and not a simulation that could be subject to inaccuracies.

Finally, based on the need to share information during design of a multivendor HVDC protection system, as well as control-related problems reported from the field, a proposal for open-source HVDC control and protection is put forward, aiming to enhance vendor-interoperability.