The power system analysis and simulations become more complicated recently due to the integration of distributed generators (DGs) such as photovoltaics and wind-power generation. In order to properly operate and design power systems with a large amount of DGs, numerical simulation with reliable models is necessary. In this paper, recent trends of simulation models for power system analysis are shown. The models are classified into generator models and network models and various examples of the modelling are introduced.
Flicker may be caused by “frequency feedback method with step injection of reactive power” installed in PCS which is generally used as grid-connected inverter in Japan. Therefore, in this paper, characteristics of the flicker were investigated by experiments with actual PCS in single phase circuit. Furthermore, effective means for suppressing the flicker was considered by taking the cause of occurrence of flicker into account. Moreover, it was clarified the differences among PCSs on flicker using multiple actual PCSs. In conclusion, gain of frequency feedback function should be reduced as one of effective means to avoid flicker.
This paper focuses on the continuous oscillation of reactive power caused by the frequency feedback method in the grid-connected inverters. The transfer functions of inverter and ac system are derived. Then the stability of the system is analyzed according to the transfer functions. The analysis clarified that the system becomes unstable if the product of line impedance, inverter capacity, and frequency feedback characteristic exceeds a specific threshold value. Then it results in the continuous oscillation of reactive power.
In power systems with a large amount of wind power plants, the supply and demand balance control is an important issue. In particular, uncertain output fluctuation depending on weather condition gives a large impact on frequency regulation. Moreover, the wind power fluctuation causes power flow issues such as loop power flow. Thereby, proper congestion management of transmission lines with ensuring the control reserve for frequency regulation is necessary to keep the system stable under the large fluctuation of renewable energy sources. On the other hand, recently, the balancing power has been traded in balancing market worldwide in order to promote the economic balancing control. It is planned that the balancing market will be newly established in the near future also in Japan. Hence, this paper presents an application of locational marginal price (LMP) to the balancing market and an optimal method to determine the LMP in the balancing market for tertiary control considering not only supply and demand balance but also power flow issue and secondary control reserve. The proposed method was tested in IEEE 39 bus system model and its effectiveness was shown.
Transient analysis is an indispensable tool for analyzing the voltage fluctuation of distribution systems according to the penetration of distributed generations with grid-connected inverters. Electromagnetic transient (EMT) analysis is suitable for the purpose because it enables detailed modeling of the inverters and accurate simulation of their dynamic behavior. However, the EMT analysis requires smaller calculation time step by a factor of 500 if the simulated power system includes the inverters. Simulation of the inverters is a bottleneck to speed up the EMT analysis. This paper proposes a novel average-value modeling method of the grid-connected inverters. The proposed inverter model operates in a larger calculation time step, and speeds up the EMT analysis of distribution systems containing the inverters. The maximum error in the output power was 8% compared with the result by the conventional model. Dynamic voltage simulations are demonstrated with a test case which includes tap changing transformers with a voltage controller and a photovoltaic generation facility. The proposed model reduces required calculation time by a factor of 754 compared with the conventional model whereas there is no significant difference in the simulated result.
High penetration of PV becomes reality. It is bidirectional power flow of forward power flow and reverse power flow in distribution bank. It is necessary to set up vector-LDC which can exert voltage control effect during reverse power flow in distribution bank and does not unreasonably limit the amount of possible PV integration. However, there have been no reports on such setting methods so far, and evaluation of better vector-LDC setting method that can increase PV integration amount was incomplete. Therefore, the paper leads difference of reverse power flow limit of three representative vector-LDC setting methods, and shows profit or loss of those. It is useful finding to increase PV integration amount under voltage control using vector-LDC. Finally, which vector-LDC setting method is used, it is shown that mixing plurality of pole transformer taps to adapt to forward power flow greatly decreases reverse power flow amount.
To use realistic wind power fluctuation data is important in simulating frequency change of power systems. It is well known that the smoothing effect must be taken into account when generating the artificial wind power output. Therefore, this paper proposed a new method to generate fluctuation waveform considering smoothing effect based on random number to fulfill the correlation among multiple sites and autoregression model. The effectiveness of the proposed method was tested by using practical wind speed data measured in Tohoku district.
Although output of individual RE (Renewable Energy) such as PV (Photovoltaic Generation) severely varies, total output of many REs locate with geographical diversity is considerably smoothed out. The phenomenon is well known and is called as (mutual) smoothing effect. To quantitatively and mathematically express it, a theory named “Transfer Hypothesis” using “Transfer Swing Period Tx” as an important parameter is presented. Since the theory regards the effect as a kind of low pass filter, the theory can be applied to not only spectrum on frequency axis but also time sequential data. However, whole view of factors that affect Tx is not made clear yet. Therefore, it is questionable that once identified Tx in the studied area can be always adequate for practical use. The paper verifies the adequacy. Seasonal factors such as month and week, impact by number of sites in the studied area, and geographical factors such as distance between neighboring sites and studied area's size are examined. It has been demonstrated that Tx of two sites is strongly depends on their distance, that number of sites does not affect Tx, and that Tx of the studied area is equal to that of the two sites apart 1/3 length of the studied area. Therefore it has been found that long time measurement on many sites is not necessary to identify Tx.
Because of the significant changes in environmental policies and electric power deregulation in the last decade, a lot of photovoltaic generations (PV) have been and will be installed into the power system in Japan and the ratio of PVs to other synchronous generators will be increased. As a countermeasure against the decrease in the rotational inertia in the whole power system, a virtual synchronous generator (VSG) model control of the PV has so far been proposed. However, the system stabilization effect of the VSG in large-scale power systems has been unclear. In this paper, a virtual step-out blocking method of VSG for improving the transient stability is proposed. In addition, the necessity of governor control of VSG in a large-scale power system is discussed. Finally, the rated kw and kwh capacities of the battery required for realizing the VSG-model control are evaluated.
In general distribution systems, it is necessary to order the nodes sequentially for faster computation of power flow. For the purpose of that, Longest Lateral (LL) method and Lower Node (LN) method have already been proposed to execute Forward Sweep power flow method based on Quadratic Equations (FSQE) which takes advantage of sequentially numbered nodes. The aim of this paper is to verify the difference of the node renumbering methods on the FSQE method. The numerical results show that the FSQE method using node renumbering methods reduces computation time compared to the Backward Forward Sweep (BFS) method based on Kirchhoff's law in large distribution systems. This paper also includes discussions of determinant of computation time on power flow methods by multiple regression analysis. It is suggested that FSQE-LL method is effective against a system including many nodes and large loads. Additionally, FSQE-LN method is also effective against a system which is reconfigured frequently.
Large integration of renewable energy sources (RESs), such as wind power and solar photovoltaic (PV) plants, into the power systems, impacts the system frequency stability. Normally, a wind farm (WF) and PV system do not provide frequency support because of the uncontrollability of the input energy. Moreover, overall system inertia will be reduced due to massive integration of RES because conventional generation units that provide reserve power need to be decreased. To overcome the problems of frequency stability as well as power system transient stability resulting from the insufficient inertia response, this paper proposes a new method to enhance the transient stability of the power system with RESs introduced, in which variable speed wind turbine with doubly fed induction generator (VSWT-DFIG) supplies its kinetic energy (KE) during generation outage to stabilize conventional synchronous generators (SGs). A suitable fuzzy logic based virtual inertia controller (VIC) is proposed to release the stored KE efficiently during transient period. This fuzzy logic controller (FLC) can continuously adjust the VIC gain depending upon the incoming wind speed. To verify the effectiveness of the proposed VIC, simulation analyses are performed on a multi-machine hybrid power system model composed of PV plant, VSWT-DFIG, fixed speed wind turbine with squirrel cage induction generator (FSWT-SCIG), and conventional SGs.
We propose the islanding detection method using harmonic current as active signal. The active signal can be injected in the suppressed direction of the harmonic voltage during grid interconnection. On the other hand, the active signal is injected in the synchronized direction of the harmonic voltage due to the loads during islanding operation. As the result, the harmonic voltage is increased and the islanding operation is detected within 0.2 seconds. We have developed the single phase inverter for grid-interconnection of low voltage with these functions and performed the experimental verification.