RLOPF (risk-limiting optimal power flow) for systems with high penetration of wind power

Shin Yeu Lin*, Ai Chih Lin

*Corresponding author for this work

Research output: Contribution to journalJournal Article peer-review

10 Scopus citations

Abstract

In this paper, we formulate a RLOPF (risk-limiting optimal power flow) problem for systems with high penetration of wind power to address the issue of possibly violating the security constraints in power systems due to the volatility of wind power generations. To cope with the computational complexity of the proposed RLOPF problem, we propose a computationally efficient RLOPF algorithm assisted by the off-line constructed probability distribution models for bus voltage magnitudes and transmission line real power flows.We apply the proposed RLOPF algorithm to the RLOPF problems on a 26-bus power system for two cases of significantly different re-dispatching percentage share for non-renewable power generations to compensate the volatility of wind power generations. The test results reveal that the performance of all solutions obtained by the proposed RLOPF algorithm of various step-sizes in both cases meet the required probability 0.95 on satisfying the security constraints in the presence of variable wind power generations, and the CPU time consumption are mostly within 1s. We also test the performance of conventional OPF (optimal power flow) solution on both cases, and the resulted probability are all smaller than 0.783. These test results demonstrate the merit and the computational efficiency of the proposed RLOPF algorithm.

Original languageEnglish
Pages (from-to)49-61
Number of pages13
JournalEnergy
Volume71
DOIs
StatePublished - 15 07 2014

Keywords

  • OPF (optimal power flow)
  • Point estimation method
  • Probability distribution model
  • Risk-limiting operation
  • Security constraints
  • Wind power

Fingerprint

Dive into the research topics of 'RLOPF (risk-limiting optimal power flow) for systems with high penetration of wind power'. Together they form a unique fingerprint.

Cite this