The dataset consists on two files. One, called “demand”, which describes the active power demand from 10 different households, in Watts units and hourly resolution. And other, called “generation”, which describes the solar generation considering the current installed PV (photovoltaic) power in each household, in Watts units and hourly resolution. The total peak power installed is 30.5 kWp (Kilowatt-Pico). Energy consumption profiles from file 'demand.csv' correspond to the same households than the energy production profiles from file 'generation.csv'. For example, 'C1' column in 'demand.csv' and 'G1' column in 'generation.csv' correspond to consumption and generation profiles from Household 1, respectively. The demand dataset was provided by the DSO (Distribution System Operator). No collection/generation work done by the researchers.The solar generation profile was download from PVGIS (Photovoltaic Geographical Information System) portal, considering the location and the installed peak power of each household. The study multiplies the data profiles (demand and generation) from the Dataset by different factors according to the scenario. Not other preprocessing than scaling the data. The date have been produced 2018-07-01.

In the study the authors talk about two scenarios. In the Scenario 1, the demand data is the same than the file 'demand.csv', in Scenario 2 the demand data is scaled by a factor of 14.In both scenarios, all generation profiles are scaled by a factor depending on the total installed PV power, considering than the original generation profiles correspond to a total peak power of 30.5 kWp.