Thesis
Simulation platform for a virtual power plant
Honours, Murdoch University
2022
Abstract
This project uses DIgSILENT’s PowerFactory software to develop a platform that simulates a Virtual Power Plant (VPP) and its controller. The VPP controls the dispatch of Distributed Energy Resources (DER) in a grid based on the Horizon Power electricity network in Carnarvon, Western Australia.
Control is undertaken at the transmission level (22 kV) where an 8 MW bank of diesel generators supplies a distributed load of 3.6 MW. PV systems at the distribution level are represented as an aggregated VPP of 3 PV systems with a peak output of 3 MW. The power output of the aggregated PV system is modelled using a series of irradiance data taken from Murdoch University’s skycamera system over a 10-hour period from 7 am to 5 pm with a resolution of 30 seconds. The simulation is thus quasi-dynamic, enabling the PV power output to be calculated and control actions to be taken at each time step. At each time step the PV output is subtracted from the network load, leaving a net load that remains to be supplied by the diesel generators. The generator dispatch scheduling is then controlled using the DIgSILENT Programming Language (DPL). Two DPL scripts (DPL1 and DPL2) implement the control actions.
The main method of control is by entering minimum and maximum setpoints into DPL1 to govern the diesel generator spinning reserve. Generators are switched in and out of service before they can breach a setpoint. This technique is designed to keep enough spinning reserve online to deal with rapid drops in PV power caused by clouding.
In the second control application (DPL2), skycamera nowcasting is used to provide adjustable setpoint spinning reserve control. In this case, the normal operating setpoints that protect the system during cloudy weather can be adjusted downward to a lower minimum and maximum when clear skies, with no reduction in irradiance, are predicted.
Spinning reserve setpoint control and nowcasting spinning reserve setpoint control is also implemented by incorporating PowerFactory’s VPP element into DPL scripts, named VPP1 and VPP2. Using this method, diesel generators are combined into another VPP and dispatched to meet an active power target that is programmed to include a minimum spinning reserve.
The VPP element is also used to demonstrate ramping of the aggregated PV system. The final control script (VPP3) uses nowcasting to proactively ramp the PV output down in order to slow the rate of any predicted drop in PV output. This control mechanism is offered as a demonstration, but due to time constraints, has not been fully incorporated into the VPP simulation.
Details
- Title
- Simulation platform for a virtual power plant
- Authors/Creators
- Malcolm J McGill
- Contributors
- Martina Calais (Supervisor) - Murdoch University, Centre for Water, Energy and WasteR. Samu (Supervisor) - Murdoch University
- Awarding Institution
- Murdoch University; Honours
- Identifiers
- 991005548670007891
- Murdoch Affiliation
- College of Science, Health, Engineering and Education; School of Engineering and Energy
- Resource Type
- Thesis
Metrics
239 File views/ downloads
423 Record Views