An Overview of the Transient Studies Required for HVAC Connected Offshore Wind Farms Kiran Munji, Jonathan Horne, Jose Ribecca Abstract-- This paper discusses about the transient phenomenon in offshore wind farm and describes the methodology for modelling and simulation. Studies presented in this paper are based on a generic HVAC connected offshore windfarm rated higher than 700 MW and export cable length greater than 100km. The system is analysed for temporary, slow- front and very-fast-front overvoltages resulting from both planned and unplanned operations. This paper also addresses the challenges, appropriate selection of equipment insulation and technical considerations to be observed during the design and operation phase of offshore wind farm is discussed. Keywords: offshore wind farm, temporary overvoltages, slow- front overvoltages, very-fast-front overvoltages, insulation coordination. I. INTRODUCTION he latest and largest offshore wind farms are often located at large distances (>100km or sometimes 200km) away from the shore, and more importantly, the onshore power grid. HVAC and HVDC options are often considered. One of the added challenges with export of power to the onshore grid through long export cables is the introduction of high overvoltages due to planned (e.g. switching) and unplanned (e.g. system faults) events. The insulation levels of equipment should be designed based on these overvoltages, or mitigation options must be introduced. This must work for different operating configurations of the system for continuous reliability of the network at reasonable cost. This paper details the various studies which are typically performed, together with the modelling and methodology adopted for analysing the overvoltages in the windfarm. Note that fast front overvoltages (FFO) due to lightning on wind turbines, or on onshore substation equipment is not included in this paper due to the extent of these topics. Further reading on these topics can be found at [1]-[3]. II. SYSTEM DESCRIPTION For the purposes of explanation in this paper, a generic offshore wind farm of 700MW with a connection in Great Britain to a National Grid 400kV substation has been used. The single line diagram of one of the feeders is shown in Fig. 1. Autotransformers connect the 400kV grid to a 220kV export cable network. At the tertiary winding of the K. Munji, J. Horne and J. Ribecca are with MPE Power System Consultants, London, UK. (e-mail:kiran.munji@moellerpoeller.co.uk, jonathan.Horne@moellerpoeller.co.uk, jose.ribecca@moellerpoeller.co.uk). Paper submitted to the International Conference on Power Systems Transients (IPST2019) in Perpignan, France June 17-20, 2019. Fig. 1. Single line diagram of a typical offshore windfarm autotransformers, a STATCOM unit is connected to meet onshore grid code requirements. A single 220kV export cable (>100km) connects the wind farm offshore substation to the onshore substation. The reactive power generated by cable is compensated through onshore and offshore shunt reactors. The offshore transformer LV windings are connected to the array cables and thereafter connected to unit transformers and wind turbines. III. MODELLING The grid is modelled as a voltage source with positive and zero sequence impedance derived from 3-phase and 1-phase fault levels. Source equivalent models used for transient studies are connected in parallel with a surge impedance, for the purpose of avoiding unrealistic travelling wave reflections, that in practice will get transmitted to the rest of the external system and consequently attenuated. The surge impedance value is based on the transmission line configuration connected to grid [4]. Cables are modelled with the actual geometric parameters and validated with the guaranteed parameters at 20 0 C provided by the manufacturer. At lower temperatures the resistance is lower and hence results in less damping of the transients [5]. The studies are initially analysed with Bergeron models and later with more detailed phase domain cable models for the system. Cable installation arrangement should be taken into consideration while deriving and validating model parameters for e.g. some sections of the cable are laid in flat configuration and other sections in trefoil configurations. Additionally, the cross-bonding of the cable from the design must be explicitly modelled. Shunt reactors are modelled with non-linear saturation characteristics connected in series with a resistance derived from the load losses. Mutual coupling between the windings should be considered. Circuit breakers are usually modelled as ideal switches, however, for some of the studies like faults and TRV, where failures are observed, the analysis is repeated with the arc models to simulate a more realistic and less conservative approach [6]. Transformers are modelled with non-linear saturation characteristics obtained from no-load test data and a reasonable worst-case residual flux of 80% is T