State of Charge Control of Second Life Batteries Integrated by a Modular Multilevel Converter

  • Violeta Rivera
  • , Carlos Smart
  • , Claudio Burgos-Mellado
  • , Yeiner Arias-Esquivel
  • , Miguel Torres*
  • , Marcos Orchard
  • , Roberto Cardenas-Dobson
  • , Diego Muñoz-Carpintero
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Modular multilevel converters (MMCs) have been proposed as a suitable power electronics topology for the integration of second-life batteries (SLBs) sourced from electric vehicles (EVs). The inherent parameter variations among the SLBs located in different clusters of the MMC require the implementation of control strategies to regulate battery discharge rates and prevent battery damage from overcharging or overdischarging. In this article, a decoupled modeling approach based on the Σ∆αβ0 transform is proposed for the SLB-MMC, and the resulting model is used to develop a decoupled nonlinear control strategy for state of charge (SoC) equalization of SLBs integrated in the MMC. The benefits of using the proposed controller are demonstrated using simulation work, moreover, the proposed control strategy is experimentally validated using an experimental rig where 18 SLBs retired from electrical scooters with capacities between 9 and 13 Ah are integrated in a three-phase MMC composed of 18 sub-modules (SMs). The experimental results demonstrate good SoC balancing performance between phases and clusters of the MMC in three scenarios comprising a discharge test, a cycling test and a charge test where reactive power is provided to the grid.

Original languageEnglish
JournalIEEE Transactions on Industrial Electronics
DOIs
StateAccepted/In press - 2026

Bibliographical note

Publisher Copyright:
© 1982-2012 IEEE.

Keywords

  • AC battery
  • energy storage
  • multilevel converters
  • second-life batteries (SLBs)

Fingerprint

Dive into the research topics of 'State of Charge Control of Second Life Batteries Integrated by a Modular Multilevel Converter'. Together they form a unique fingerprint.

Cite this