Search

This article studies the energy management of distribution-level integrated electric-gas systems (D-IEGSs), considering their frequency support capability via providing the fast frequency reserve (FFR) service. The dynamics of the main grid frequency as well as frequency regulation behaviors of gas-fired turbines (GTs), power-to-gas (P2G) units, wind turbines (WTs) and Photovoltaic (PV) units in the D-IEGS are formulated, including the dead zone, limiting ranges, and the time constant of the governors. The impacts of frequency regulation behaviors of GTs and P2G units on the operation constraints of the D-IEGS are examined through a feasibility check sub-problem. Thus, the overall problem becomes a two-stage one, which is solved by the column-and-constraint generation (C&CG) algorithm. The variable-step difference and the binary variable reduction schemes are devised to decrease the computation burden. Simulation results on the IEGS with a 13-bus power network and a 6-node gas network verify the value of FFR provided by the D-IEGS and the necessity of modeling the operation constraints of the gas network.

With the high penetration level of renewable power generation, their impacts on the power system frequency dynamics need to be considered, especially the doubly-fed induction generator (DFIG) based wind turbines. By analogy with the synchronous generator, this article proposes the simplified transient model of DFIG wind turbines, which consists of the equivalent circuit model and the equivalent rotor motion model. It provides a viable and tractable way to analytically quantify the contribution of DFIG wind turbines to the post-disturbance frequency, both in the center of inertia (COI) frequency and frequency spatial variation cases. The unique frequency-supporting features of the DFIG wind turbine are discussed. The model implementation at the wind farm level is provided. The correctness of the proposed model is verified by simulation results of a weak power system. The comparison with the existing models on frequency dynamics analysis is discussed.

The post-disturbance rate of change of frequency (RoCoF) is a critical index for low-inertia power system frequency stability. This paper proposes a RoCoF constrained unit commitment model so as to enhance the frequency stability in operation time-scale, where the spatial difference in frequency dynamics is considered and the RoCoF constraint is enforced for each node. The maximum of node RoCoF is approximated by two terms, i.e., the initial post-disturbance node RoCoF and the maximum of the center of inertia (COI) RoCoF. Analytical expressions of the node initial RoCoF under typical disturbances are derived. Then, a tractable relationship between the node initial RoCoF and the working status of generators is obtained by the piecewise linearization approach. Simulation results carried out on the IEEE 39-bus system demonstrate the proposed model can limit the node RoCoF in an economic and accurate manner. The adaptability to renewable power generation and online application are also discussed.

To avoid DC overvoltage, the active power of photovoltaics in a DC access system (PV-DCAS) should be instantly cut down when AC voltage dips, and recover to its pre-fault state after fault clearance (post-fault). However, PVs may fail to switch from reducing active power mode to maximum power point tracking mode during the post-fault period, due to the miss of centralized recovery signal, resulting in the active power loss of PV-DCAS. Consequently, PV-DCAS cannot retain the expected power generation efficiency. This article proposes an active power recovery method for PV-DCAS based on local DC voltage. Based on the analysis of DC voltage during the recovery process, the stage characteristic of DC voltage variation is explored and revealed. A hierarchical criterion that combines the stage features of DC voltage variation and the dynamic response speed of PV is proposed. The proposed approach is validated via real-time digital simulation case studies for multiple AC grid fault scenarios.

The detailed model parameters are fundamental to fault characteristics analysis and protection design. Accurate acquisition of inverter controller parameters is essential to analyze the transient characteristics of Inverter Interface Renewable Energy Sources (IIRESs). However, it is challenging to identify the controller parameters using the three-phase current owing to the coupling relationship between the phase-locked loop (PLL) and the inner current loop. Given this, the fault characteristics of the IIRESs considering the non-linear coupling of different control loops is investigated. A fault dynamic response difference-based parameter identification method is proposed. The different time-length response of each controller is utilized for decoupling. The proposed method can identify the parameters of PLL and other controller parameters based on three-phase current and voltage. Simulation and field short-circuit test data are ultimately used to verify the efficacy of the proposed method.

Accurate identification of fault control parameters (FCPs) in the inverter-interfaced renewable energy source (IIRES) is crucial for simulating fault characteristics and designing protection. However, accurate identification can be challenging due to the correlation among FCPs. To address this issue, the correlation of parameters (COP) in the IIRES fault control process is comprehensively investigated in the s-domain and time-domain for the first time, and a multistage parameter identification method is proposed. This method partitions fault transients in different time sequences and control state spaces, allowing the COP to be decoupled and FCPs to be identified independently. The Mahalanobis distance-based objective function is used to further enhance the identification of low-sensitive parameters. Hardware-in-loop experiments and field short-circuit tests demonstrate that this method eliminates the influence of COP on parameter identification, and reliably identifies all FCPs involved during IIRES fault transients with an error of less than 9%.