Convex Storage Loss Modeling for Optimal Energy Management - Archive ouverte HAL
Conference Papers Year : 2021

Convex Storage Loss Modeling for Optimal Energy Management

Abstract

Models of energy storage systems used for optimal energy management commonly feature unsophisticated loss expressions (lossless or constant efficiency). Preserving the tractability (i.e. convexity) of the optimization motivates this simpleness, at the expense of physical realism. Relaxing the storage loss expression from an equality constraint to an inequality allows using more complex and physically realistic models while preserving convexity. Although this relaxation has been used quite casually for a decade in varied energy management applications (power systems dispatch, electric vehicles), only few articles studied this technique for itself. To increase the awareness of loss relaxation, along with its limitations, we depict a unified view of relaxed storage loss models. Also, we propose the "convex monomial loss model", one continuously parametrized family of nonlinear convex expressions which contains all classical loss models and unlocks new possibilities for loss dependency in storage power and energy level. This is one step to reunite piecewise linear and nonlinear models which are generally studied separately. Finally, we compare the effect of these varied loss models on the optimal charge/discharge profile of a storage in an elementary energy management application.
Fichier principal
Vignette du fichier
Haessig 2020 - Convex storage - untrimmed final.pdf (352.33 Ko) Télécharger le fichier
Convex_storage_slides.pdf (664.75 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Format Presentation
Comment Diapos de présentation à PowerTech

Dates and versions

hal-03032241 , version 1 (30-11-2020)
hal-03032241 , version 2 (14-04-2021)

Identifiers

Cite

Pierre Haessig. Convex Storage Loss Modeling for Optimal Energy Management. PowerTech 2021 (on line), Jun 2021, Madrid, Spain. ⟨10.1109/PowerTech46648.2021.9494786⟩. ⟨hal-03032241v2⟩
246 View
341 Download

Altmetric

Share

More