A Framework to Analyze the Stochastic Harmonics and Resonance of Wind Energy Grid Interconnection

Youngho Cho,Young-Do Choy,Choongman Lee,Sang-Ho Park,Eduard Muljadi,Yong Kang,Gi-Gab Yoon,Kyeon Hur

doi:10.3390/en9090700

Abstract

This paper addresses a modeling and analysis methodology for investigating the stochastic harmonics and resonance concerns of wind power plants (WPPs). Wideband harmonics from modern wind turbines (WTs) are observed to be stochastic, associated with real power production, and they may adversely interact with the grid impedance and cause unexpected harmonic resonance, if not comprehensively addressed in the planning and commissioning of the WPPs. These issues should become more critical as wind penetration levels increase. We thus propose a planning study framework comprising the following functional steps: First, the best fitted probability density functions (PDFs) of the harmonic components of interest in the frequency domain are determined. In operations planning, maximum likelihood estimations (MLEs) followed by a chi-square test are used once field measurements or manufacturers’ data are available. Second, harmonic currents from the WPP are represented by randomly-generating harmonic components based on their PDFs (frequency spectrum) and then synthesized for time domain simulations via inverse Fourier transform. Finally, we conduct a comprehensive assessment by including the impacts of feeder configurations, harmonic filters and the variability of parameters. We demonstrate the efficacy of the proposed study approach for a 100-MW offshore WPP consisting of 20 units of 5-MW full converter turbines, a realistic benchmark system adapted from a WPP under development in Korea and discuss lessons learned through this research.

Highlights

Integrating high penetrations of variable renewable sources into the electric power grid presents a range of unprecedented grid operations and planning challenges requiring new study approaches and tools so that the impacts of a proposed wind power plant (WPP) can be properly assessed prior to interconnection [1,2,3,4,5]
A base case model of the WPP is built based on the following scenarios: The standard deviation of each harmonic is 10% of its average value, and the phase angle is uniformly distributed with its interval size of 2π
Two largest harmonic components at the 4.5th and 12.5th order result from the two resonance points of the WPP with power conditioning system (PCS) filter

Summary

Introduction

Integrating high penetrations of variable renewable sources into the electric power grid presents a range of unprecedented grid operations and planning challenges requiring new study approaches and tools so that the impacts of a proposed wind power plant (WPP) can be properly assessed prior to interconnection [1,2,3,4,5]. WPP are stochastic due to the inherent intermittency of the wind and wide control bandwidth of the power electronics interfaces [6,7]. These wideband stochastic harmonic emissions from modern converter-based WTs may adversely interact with the grid via vast underground or submarine cable systems, leading to many series and parallel resonance points that significantly change. Recent analyses of harmonic emissions from WTs by type interestingly claim that we may statistically characterize harmonic components with certain probabilistic distributions independent of the wind power generation (i.e., operating condition) in contrast with the fundamental current, which should be proportional to the wind power [7,8,15]. It is well understood that the resonance points vary with changes in the system operating conditions; for example, the number of operating

Results

Discussion

Conclusion