A novel approach for disaggregating photovoltaic generation in distribution networks
Distribution system operators usually lack information regarding the installed photovoltaic power in their low voltage distribution networks due to its private nature. This, together with the increasing amount of private photovoltaic installations, is compromising the optimal operation of these netw...
| Autores: | , , , , , |
|---|---|
| Formato: | artículo |
| Fecha de publicación: | 2026 |
| País: | España |
| Recursos: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:dnet:upcommonspor::a4659cdd5694d774811258c85e491b19 |
| Acesso em linha: | https://hdl.handle.net/2117/461992 https://dx.doi.org/10.1016/j.epsr.2026.113207 |
| Access Level: | acceso abierto |
| Palavra-chave: | Photovoltaic power LV distribution networks PV power disagregation Grid planning |
| Resumo: | Distribution system operators usually lack information regarding the installed photovoltaic power in their low voltage distribution networks due to its private nature. This, together with the increasing amount of private photovoltaic installations, is compromising the optimal operation of these networks. The recent blackout in Spain highlighted the potential consequences of such unmonitored distributed generation, underscoring the need for better visibility and control of photovoltaic contributions to the grid. For planning the infrastructure and the operation of these networks, it is necessary to know the amount of photovoltaic power generation they are hosting. This short communication introduces a simple and efficient methodology for disaggregating photovoltaic power generation in low voltage networks. Considering that small-scale photovoltaic installations typically do not inject reactive power, this methodology takes advantage of the existing correlation between the signed power factor and solar irradiance. The methodology is tested in multiple real Secondary Substations from a Spanish DSO, proving its effectiveness in accurately decomposing photovoltaic generation profiles.The results indicate that the nature of the consumption profiles conditions which method outperforms the other. Despite this uncertainty, both methods provide less than 10% MAPE for penetrations beyond 150%(with respect to the secondary substation peak load) with an R2 higher than 0.9 across all secondary substations. |
|---|