(This topic was presented at the Conference of Electric Power Supply Industry, CEPSI held in Kuala Lumpur, Malaysia on September 18-20, 2018. For the full presentation, please see this link.)
One of the tasks transmission planners, analysts, and system operation engineers to evaluate stability and resiliency of electric power system is to perform computer simulation of dynamic models and representation of power system components. The computer simulation may be part of an overall transmission planning process (5-10 years), evaluation of new generator interconnections, and assessment of system performance. Typically, the study interest for transient and voltage stability simulations falls on the range of 0.1-10 Hz where the said range of phenomena is sufficient to model and evaluate the dynamic response of power system components. With the influx of renewable energy sources on electric power system, the way traditional stability studies are performed would be affected physically (in terms of assessing the performance of the power system with renewables) and mathematically (in terms of algorithms and dynamic models being implemented).
The dynamic characteristics of renewables such as faster control response, independent active and reactive power control, and limited thermal capacity for transient reactive power support sets it apart from the traditional responses of conventional machines. The resulting effect of the dynamic characteristics of renewables are as follows:
- Increase in upper frequency of study interest. From Figure 1 below, traditional stability studies fall on the range of 0.1-10 Hz (or 0.1-10 s to a couple of minutes). The addition of renewables would make the frequency of study interest increase due to very fast switching times of semiconductor devices such as insulated gate bipolar transistors (IGBT) and thyristors which make up the converter/inverter unit of most renewables.
- Further reduction in integration time steps. An increase in frequency of study interest translates to much faster phenomena like those exhibited by electromagnetic transient events. To cover very fast processes of inverter controls, a smaller integration time step used in dynamic simulation is needed. Say for instance we have a decaying current response of a R-C circuit shown in Figure 2. To capture the linearized response will require at least two points (or time step) in the same duration of time constant.
To figure out how the impact of addition renewables affect electric power systems, a sample active power and voltage response curves illustrated in Figures 3 and 4, respectively were obtained from an offline positive-sequence dynamic simulation of power system in Southeast Asia as part of interconnection study requirements.
A number of spurious high-frequency spikes were observed in the plots of active power and voltage responses. As mentioned earlier, the addition of renewables may have a physical impact on the power system dynamic performance. The high-frequency responses may be due to difficulty of phase-locked loop algorithms (PLL) of solar PV inverter finding the required active power and voltage set-point following the clearing of fault.
To prepare dynamic databases used for dynamic simulation with the addition of renewables, some tasks for system operators and utility owners are suggested to be undertaken:
- Practical approach in modeling renewables. Solar PV, Wind Farms, and Battery Energy Storage Systems (BESS) usually covers a wide geographical area for their installation. An aggregate modeling of these facilities in power flow and dynamic studies is needed to reduce computational burden without great loss in accuracy of the simulation results.Generic library models for renewables are sufficient to represent dynamics of inverter controls and associated equipment. The Renewable Energy Modeling Task Force (REMTF) of the Western Electricity Coordinating Council (WECC) developed these generic models and became available in most commercial power flow and transient stability programs such as PTI’s PSS/E, GE PSLF, and PowerWorld Simulator. Figures 5 to 7 shows generic dynamic models for solar PV, wind turbine generator (WTG), and BESS.
The use of user-written models specific for manufacturer of solar PV inverter, BESS, and WTG should be limited as these sometimes introduce numerical instability and initialization issues.
- Review dynamic models of existing conventional units and associated controls. An up-to-date dynamic model and parameters of generators, exciters, governors, and power system stabilizers (PSS) is required to have a confidence using such models when performing dynamic simulations. To obtain the parameters representing dynamics of a piece of equipment, offline tests may be required and/or inspection of technical data sheets and plant operating instruction manuals.
- Set-up Model Validation Process. The role of synchronous machine becomes critical and accurate modeling is required as more renewables are being added in the grid. The model validation process may be started from obtaining actual responses of synchronous machines during system event (via disturbance record or synchrophasors) and compare these responses to offline simulations. Inconsistencies of synchronous machine responses obtained from simulation maybe refined by tuning dynamic model parameters of the machine itself, exciters, and governors.
- Review time constants and integration time steps. Appropriate selection of time steps is necessary to improve the fidelity of dynamic database used for power system stability studies. Time constants of generators and exciters should be inspected on a timely basis as the power system is continuously evolving especially with the integration of renewables.