Lithium-ion batteries are gaining importance for a variety of applications due to their improving characteristics and decreasing price. An accurate knowledge of their aging is required for a successful use of these ESSs. The vast number of models that has been proposed to predict these phenomena raise doubts about the suitability of a model for a particular battery application. The performance of three models published for a Sanyo 18650 cylindrical cell in a self-consumption system are compared in this work. Measured photovoltaic production and home consumption with a sampling frequency of 15 minutes are used for this comparison. The different aging predictions calculated by these three models are analyzed, compared and discussed. These comparison is particularized for two management strategies. The first of them maximizes the self-consumption PV energy, while the second reduces the maximum power peak demanded from the grid. The authors would like to acknowledge the support of the Spanish State Research Agency (AEI) and FEDER-UE under grants DPI2013-42853-R, DPI2016-80641-R and DPI2016-80642-R; of Government of Navarra through research project PI038 INTEGRA-RENOVABLES; and the FPU Program of the Spanish Ministry of Education, Culture and Sport (FPU13/00542).