Generally speaking, power transformers are very efficient devices. However, even low percentage of inefficiency (i.e. losses) can result in considerable absolute loss magnitudes, which is why additional emphasis is placed on this aspect of power transformer design. Consequently, abundant resources are invested in research of various losses in power transformers, resulting in continuous development of tools, methods and algorithms and an ever increasing expert knowledge base. On the other hand, instrument transformers are not considered to be a substantial source of losses, which is why loss analyses are generally simplified and primarily controlled by testing. With the emergence of Power Voltage Transformers, which are a novel conceptual cross-over between instrument and power transformers, with a primary task of power delivery, the topic of losses becomes increasingly important. This dissertation is based on open-core transformers, which exhibit several specialties with regard to loss determination and classification. These specialties are not covered adequately, if at all, by relevant literature. This is why the aim is to propose, describe and verify loss classification of open core units for no load and short-circuit conditions, as well as for operation under load. 2. Basic considerations of losses in transformers with closed core Losses of conventional closed-core transformers can be separated into two dsitinct groups; no-load losses and load losses. The former appear when the unit is placed under voltage with windings open-circuited, while the latter appear when one of the windings is short-circuited. No-load losses are almost completely confined to the transformer core and can be devided into hysteresis losses, eddy current losses and excess losses. There is a multitude of availble approaches based on analytical, numerical or empirical foundations, specialized for determination of core losses. Load losses consist of DC losses in the windings and additional losses. When a transformer is short-circuited, only leakage flux is present within the active part, and that flux causes additional losses in the windings and structural components, such as tie plates, core clamping systams, tank walls, etc. Typically, additional losses are determined by means of numerical calculation, such as finite element analisys. 3. Power Voltage Transformers Power Voltage Transformers are single-phase units used for direct transformation of power from high, transmission level voltage (between 72.5 and 550 kV) to low voltage (under 1 kV). Via that direct energy transfer, the distribution grid is bypassed entirely. This eliminates the necessity for the intermediate transformation of power, consequently removing the related costs of primary and secondary equipment as well as the costs of the corresponding infrastructure. Such units are primarily instrument transformers by design philosophy and technology of production, meaning they are currently defined by standards applicable to instrument transformers. However, they are used to provide auxiliary power to substations, remote consumers or communities. Power voltage transformers are an expanding market niche, and are available with paper-oil and gas insulation. This dissertation focuses on a transformer concept which is based on the open core. 4. Simplified model and key advantages of the open core concept Open core is a single limb core whose magnetic flux path is closed through the surrounding non-magnetic material. It can be assumed that the entire reactive power used for magnetizing the magnetic circuit of the open core is spent on the air gap. Furthermore, using analytical formulations developed previously, it is possible to substitute the open core with a simplified model. The magnetizing current necessary to magnetize the open core is of similar magnitude as the rated load current, and therefore cannot be neglected. Consequently, this means that actual losses during transformer operation are not an arithmetic sum of load and no-load losses, as the current flowing through the primary winding is a vectorial sum of magnetizing and load currents. Furthermore, the vastly predominant contributor to additional losses in both the windings and the structural components is the main magnetic flux which spreads in all directions in vicinity of the core. Due to a typically low rated flux density, the effects of leakage flux can be neglected for open core transformers. These are the reasons why the actual classification of losses in open core units is drastically different than conventional classification, explained in chapter 2. Open core power voltage transformers exhibit several benefits in regard to operational safety and reliability. Each of these benefits is briefly explained in the dissertation. In short, those benefits are: ferroresonance immunity, inherent explosion safety, inrush current decrease, insusceptibility to overvoltages of various type and origin and possibility of simultaneous AC and DC excitation. 5. Classification of losses in open-core power voltage transformers This chapter outlines classifications of losses in open-core units for three separate situations; no-load conditions, short-circuit conditions and continuous transformer operation under load. Each loss type will be determined by methodology presented in chapters 6 - 8, while the adequacy of the classification itself is verified by measurements on ten actual power voltage transformer units, as presented in chapter 9. All relevant parameters of units used in the analyses are presented in appendix A. 6. Losses in the open core Determining losses in the open core is a complex task, which can require lengthy, computationally intensive numerical simulations. For that reason, a novel method was developed which combines loss measurement data with simple numerical post processing. The objective of the method is to obtain referent specific loss curve for every core design. The proposed test circuit was developed to enable measurement on individual open cores. Measured data serve as input values for determining the average flux density in open core for every measurement point. The method was applied to several configurations which included scaled-down core models. These models were used to measure and verify the magnitudes of flux density vector components, as well as to corroborate the entire methodology. Multiple actual size open cores were used to analyse the influence of characteristical geometric quantitates, core material and magnetization on open core losses in order to determine different trends of open core behaviour in regard to losses. 7. Winding losses This chapter provides a detailed analysis of losses in the windings due to magnetizing current, load currents and resultant primary current. Furthermore, additional losses in the secondary winding, including losses due to skin effect and circulating currents are also investigated. Every loss group has its own set of key parameters. The influence of all these parameters is also analysed and quantified. Analytical approach is used to determine the losses in the primary winding, while additional losses in the secondary winding are primarily determined by numerical calculation, and verified on several models which represent actual transformer geometry. 8. Losses in structural components Losses in various structural components are determined using numerical modelling, and corroborated with measurements on different models, each adjusted to represent losses in a certain component as accurately as possible. Furthermore, losses in every component are analysed in more detail in order to obtain the influence of geometry, material and positioning on additional losses in those structural components. For components that exhibit loss magnitudes which are not accurately measurable, loss distribution was verified using temperature measurements. 9. Verification All losses which were explained and predetermined in chapters 6 - 8, are implemented into loss classification models, disclosed in chapter 5. The acquired loss values are compared to measured data, obtained through tests on a total of ten actual power voltage transformer units. Error margins of the proposed methodology were below 10% for every considered situation. The chapter also discusses the implementation of the developed methodology into actual transformer production. 10. Conclusion This dissertation establishes and verifies loss classification of open core units under no-load conditions, short-circuit conditions and continuous transformer operation under load. The methodology for determining various types of losses characteristical to the open core concept is presented in detail and verified on different specialized models, as well as on actual transformer units. All characteristical types of losses are accompanied with analyses of influencing parameters, thus establishing foundations and rules for designing open core units with an aim of controlling and consequently decreasing total losses in them.