Online State Estimation of the Integrated Electricity and Gas System Based on the Gaseous Circuit Method

• Published in: IEEE transactions on smart grid Vol. 14; no. 5; pp. 3465 - 3481
• Main Authors: Yin, Guanxiong, Cetenovic, Dragan, Levi, Victor, Sun, Hongbin, Terzija, Vladimir
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.09.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Circuits; Constrained WLS state estimation; Constraint modelling; Electric power systems; Electricity; Frequency analysis; Frequency measurement; Gas dynamics; Gas pipelines; Gas sampling; gaseous circuit method; Indexes; integrated electricity and gas system; Mathematical models; Natural gas; Parameter sensitivity; Partial differential equations; Pipelines; Power flow; State estimation; Time measurement; Voltage measurement; Yard waste
• ISSN: 1949-3053; 1949-3061
• DOI: 10.1109/TSG.2022.3231407

In this paper, an online state estimation (SE) scheme of the integrated electricity and gas system (IEGS) is developed. The online SE scheme handles different time resolutions of the electrical power system (EPS) and natural gas system (NGS) by determining the SE model to be applied. When the combined state estimation of the two systems is executed, the proposed Unified State Estimation (USE) model is applied. The USE model is based on the weighted least squares (WLS) estimation within the pre-specified electricity and gas sampling windows. The EPS is modeled via AC power flow because (fast) electromechanical transients die out between two SCADA sampling points. Gas dynamics in pipelines of a natural gas system (NGS) are much slower and can be encapsulated by successive measurements in time. It is proposed to convert the governing model in the form of partial differential equations (PDEs) into algebraic equations in the frequency domain giving the gaseous circuit method (GCM). Time-space coupling constraints are incorporated into the model by considering temporal EPS quantities and NGS frequency components in the time domain. The proposed USE model can adapt to both the NGS's steady and dynamic operating conditions by considering different frequency components. To analyze the sensitivity to USE parameters and demonstrate the effectiveness and scalability of the proposed model, case studies are developed using one small IEGS and one larger IEGS obtained by integrating the IEEE 39-bus EPS and 20-node Belgium NGS.