Doctor of Philosophy
School of Electrical, Computer and Telecommunications Engineering
Perera, Devinda, Contributions to the understanding of harmonics, flicker and voltage unbalance management in future electricity distribution networks, Doctor of Philosophy thesis, School of Electrical, Computer and Telecommunications Engineering, University of Wollongong, 2014. http://ro.uow.edu.au/theses/4174
The connection of electromagnetically disturbing installations, including new forms of generating systems, to power distribution networks continue to increase. As a result, managing the network electromagnetic compatibility (EMC) within stipulated limits has become a major challenge to distribution network service providers (DNSPs). Therefore, the availability of a set of guidelines and recommendations based on well researched engineering practices would facilitate DNSPs in the provision of adequate supply quality to customers connected to the distribution networks. Accordingly, the International Electrotechnical Committee (IEC) has released a series of standards and technical reports to address the management of EMC in distribution networks.
Despite the availability of these standards and technical reports, there remain inadequacies in these documents as well as gaps in the existing knowledge in relation to management of EMC, where further renements are required. These inadequacies and gaps must be addressed in order for DNSPs to properly manage EMC in distribution networks.
The low frequency EMC concerns addressed in this Thesis include harmonics, voltage fluctuations and flicker, and voltage unbalance (VU). The focus of this Thesis is to investigate the management of these key EMC issues in future distribution networks.
The recently published IEC Technical Report IEC 61000-3-14, provides power quality (PQ) emission limits for large disturbing installations connected to low voltage (LV) distribution networks. Noting that harmonic voltages in networks arise due to both large installations and smaller installations (whose harmonic current emission levels are governed by equipment standards), IEC 61000-3-14 proposes an additional factor, which is referred to as a `reduction factor'. This factor represents the fraction of the contribution to global emission allowance from the harmonic current emissions of smaller installations. IEC 61000-3-14 recommends that DNSPs determine these reduction factors considering the prevailing system conditions in their networks. The analyses presented in this Thesis show that the establishment of a universal value (network independent) for the reduction factor is not advisable, as it depends on a number of variables which are unique to the distribution network under consideration. This difficulty undermines the applicability of the IEC approach in relation to practical LV networks.
Furthermore, a number of methodologies that exist in the current technical literature in relation to the assessment of harmonic current emission limits for disturbing installations connected to the public LV network are closely examined, emphasising the strengths and weaknesses of each approach. A comparative study using test LV distribution networks is conducted. The study shows that, though underlying philosophies and data requirements for each of the investigated methodologies vary significantly they provide emission limits for each individual installation which are not too dissimilar.
In relation to VU, application of the IEC methodologies may or may not lead to a conservative emission allocation, especially in complex radial networks. An alternative VU emission allocation methodology based on the constrained bus voltage (CBV) method is proposed in this Thesis. The theoretical bases for the formulation of the CBV methodology are presented together with several application examples. The CBV methodology is shown to be superior in comparison to the VU allocation methodology presented in IEC Technical Reports, as the former enables the network VU absorption capacity to be fully utilised.
Compared to both IEC and CBV methodologies, the novel VU emission allocation procedure presented in this Thesis that is based on the concept of voltage droop (VD), provides a simplistic, less computationally and data intensive technique.
In addition to harmonics and VU, voltage fluctuations and flicker are expected to become a major concern for DNSPs due to integration of intermittent and fluctuating renewable energy generators (REGs) to the distribution networks where little or no knowledge exists in relation to the contributions made by such sources to the network. The impacts of REGs with dierent control modes (i.e. power factor control operation, voltage control mode and reactive power dispatch mode) on voltage or no knowledge exists in relation to the contributions made by such sources to the network. The impacts of REGs with dierent control modes (i.e. power factor control operation, voltage control mode and reactive power dispatch mode) on voltage fluctuations and flcker are examined in this Thesis. In addition, the attenuation characteristics of distribution system loads and their impact on flicker levels in the distribution network are also investigated. The results demonstrate that flicker emission characteristics of REGs are influenced negatively by the reactive power control strategy employed and the flicker attenuation characteristics are influenced by the various load types connected to the distribution feeder. This is an aspect which has not been recognised in existing literature. The outcomes of the study emphasises the need for adequate planning by DNSPs in relation to voltage flctuations and flicker, before connecting REGs to the distribution network that have the capability to control the power factor or voltage.
This Thesis also examines the PQ characteristics of small single-phase photovoltaic inverter (PVI) systems, with the objective of establishing realistic information relevant to their PQ impacts on the distribution network, as such knowledge is not widely available and will be useful for DNSPs in managing their networks. This is achieved by conducting controlled laboratory experiments using an experimental test setup based on IEC recommendations from which the harmonics and flicker emission levels have been generally observed to be within stipulated limits, however, some PVIs exceeded the emission limits for even harmonics.