Steady State Operation and Control of Power Distribution Systems in the Presence of Distributed Generation

Licentiatavhandling, 2006

In this thesis, the impact of distributed generation (DG) on steady state operation and control of power distribution systems is investigated. Over the last few years, a number of factors have led to an increased interest in DG schemes. DG is gaining more and more attention worldwide as an alternative to large-scale central generating stations. A number of DG technologies are in a position to compete with central generating stations. There are also likely opportunities for renewable energy technologies in DG. Indeed, some renewable energy based DG technologies are not yet generally cost-competitive. However, technology development may lead to major innovative progress in materials, processes, designs and products, with higher efficiency and cost reduction opportunities. In electric power systems with large central generating stations, the electric power flows in one way direction: from generation stations to transmission systems, then to distribution systems and finally to the loads. Therefore, distribution systems were designed as radial systems; and many operation and control in distribution systems, such as voltage control and protection, are based on the assumption that distribution systems are radial. In a radial distribution feeder, voltage decreases towards the end of the feeder, as loads cause a voltage drop. However, it will be altered with the presence of DG. DG will increase the voltage at its connection point, which in turn will increase the voltage profile along the feeder. This increase may exceed the maximum allowed voltage when the DG power is high. One way to mitigate this overvoltage is when DG absorbs reactive power from the grid. This method is effective for mitigation of overvoltage-caused DG in low voltage (LV) feeders where the mean of voltage control is obtained from an off-load tap changer. However, if DG absorbs reactive power, feeder losses will increase. The maximum DG that can be integrated in a feeder (DG integration limit) is limited by maximum allowed voltage variation, conductor thermal ampacity and upstream transformer rating. The DG integration limit is usually defined based on maximum DG and minimum load scenario. However, when DG and load power fluctuate throughout the day, this scenario will lead to unnecessary restriction of DG integration. Minimum load and maximum DG may not happen at the same time. Stochastic assessment using Monte Carlo simulations will be more reliable to determine the DG integration limit in this circumstance. In medium voltage (MV) feeder, where the voltage control is normally achieved by using on-load tap changer (LTC) and capacitor banks; the mitigation of overvoltage-caused DG can be obtained by coordinating DG with the LTC and capacitors. The use of line drop compensation (LDC), which is present in most LTCs but often not used, can also mitigate the overvoltage. When the LDC is coordinated properly with DGs, LDC will even extend the DG integration limit. The DG integration limit in a MV feeder can also be extended by allowing DG to absorb reactive power as in an LV feeder, or by installing a voltage regulator (VR). However, if DG absorbs reactive power, it means that the reactive power should be generated somewhere else in the system, and VR installation means investment cost. The DG integration limit can also be extended by operating the MV feeders in a meshed system (closed-loop). The expense of this meshed operation is that the protection of the feeder is more complicated. The presence of DG will obviously increase residual voltage (dip magnitude) during a short circuit. However, depending on the location of the DG relative to the protection device (PD) and fault, DG may shorten or lengthen the duration of the short circuit, which directly correlates to dip duration. This is because, the location of the DG relative to the PD and fault defines whether DG will increase or decrease the short circuit current sensed by PD. However, PDs in distribution systems are normally overcurrent (OC) based PDs, which clear the fault in a certain time delay depending on the short circuit current sensed by them. Thus, though DG increases dip magnitude, further investigation on coordination of voltage dip and OC protection is needed to investigate whether the DG will prevent sensitive equipment from tripping, due to voltage dip, or not. Protection coordination in distribution systems can be affected by the increasing or decreasing short circuit current sensed by PDs. Certain corrective actions are then needed. However, when the DG is not expected to be in islanding operation; DG still has to be disconnected from distribution systems every time a fault occurs, even if all corrective actions have been implemented. Disconnecting all DGs every time a temporary fault occurs would make the system very unreliable. This is especially because most of the faults in overhead distribution systems are temporary. Thus, when the DG is not expected to be in islanding operation, a protection scheme that can keep DG on line to supply the load during the fault is necessary. The scheme should ensure that the OC PDs in on the feeder can clear the fault without loosing their proper coordination.