Open AccessStochastic optimisation with risk aversion for virtual power plant operations: a rolling horizon control
Author(s) -
Castillo Anya,
Flicker Jack,
Hansen Clifford W.,
Watson JeanPaul,
Johnson Jay
Publication year - 2019
Publication title -
iet generation, transmission and distribution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.92
H-Index - 110
eISSN - 1751-8695
pISSN - 1751-8687
DOI - 10.1049/iet-gtd.2018.5834
Subject(s) - virtual power plant , probabilistic logic , time horizon , computer science , schedule , operations research , grid , leverage (statistics) , stochastic optimization , risk aversion (psychology) , operator (biology) , stochastic programming , renewable energy , mathematical optimization , engineering , distributed generation , economics , finance , expected utility hypothesis , mathematics , mathematical economics , artificial intelligence , geometry , electrical engineering , repressor , chemistry , operating system , biochemistry , machine learning , transcription factor , gene
While the concept of aggregating and controlling renewable distributed energy resources (DERs) to provide grid services is not new, increasing policy support of DER market participation has driven research and development in algorithms to pool DERs for economically viable market participation. Sandia National Laboratories recently undertook a 3 year research programme to create the components of a real‐world virtual power plant (VPP) that can simultaneously participate in multiple markets. The authors’ research extends current state‐of‐the‐art rolling horizon control through the application of stochastic programming with risk aversion at various time resolutions. Their rolling horizon control consists of day‐ahead optimisation to produce an hourly aggregate schedule for the VPP operator and sub‐hourly optimisation for the real‐time dispatch of each VPP subresource. Both optimisation routines leverage a two‐stage stochastic programme with risk aversion and integrate the most up‐to‐date forecasts to generate probabilistic scenarios in real operating time. Their results demonstrate the benefits to the VPP operator of constructing a stochastic solution regardless of the weather. In more extreme weather, applying risk optimisation strategies can dramatically increase the financial viability of the VPP. The methodologies presented here can be further tailored for optimal control of any VPP asset fleet and its operational requirements.