Dataset of 15-minute values of active and reactive power consumption of 1000 households during single year

This article presents a comprehensive dataset comprising average 15-minute values of active and reactive energy consumption in 1000 anonymized households located in the Slovak Republic, a central European country, throughout the year 2016. The dataset provides a valuable resource for researchers and practitioners interested in analysing energy consumption patterns at the individual household level within the unique context of Central Europe. Privacy concerns are addressed through anonymization techniques, ensuring the dataset's compliance with ethical considerations and privacy regulations. However, ZIP code information is included for each household. Researchers can confidently analyze the data without compromising the households' confidentiality. The dataset offers significant opportunities for researchers to explore energy consumption patterns, develop targeted energy management strategies, and contribute to the advancement of sustainable energy practices.


Value of the Data
• This data represents a random statistical selection of electrical energy consumption of 10 0 0 households connected to LV network.Each household contains information about annual load shape curves of active and reactive power consumption with 15 min intervals.Typically, such data are inaccessible to the public.Also, the size of the statistical file represents a high added value to the dataset.Another added value of the dataset is information about the approximate geolocation of households.• The dataset can benefit academics, researchers or anyone interested in the field of power engineering, energy efficiency or power consumption economic analysis.Data are helpful for anyone who works with modelling of power systems, power quality analysis, economic analysis and risk management.• Data can be useful for modeling LV systems or to generate a probability models of power consumption.Such models can be further used for hosting capacities analysis of photovoltaic and electric vehicles integration into low voltage network.Size of dataset also allows to by useful in machine learning applications and studies.• Largest benefit of the data is that values of power consumption are in 15 minute time steps.This time interval is normally used for billing of consumption.Another benefit is the information about lagging and leading reactive power.• Loadshapes of active power consumption P and reactive power consumption Q can be used for typical load models in quasi -dynamic modelling used in softwares such as Matlab Simulink, OpenDSS, PSLF and others [1] .

Data Description
The published data contain information about active and reactive(leading and lagging) power electricity consumption from randomly selected 10 0 0 households in the Slovak Republic.Each household is a single point of consumption (PoC) and is connected to the low voltage (LV) network at a voltage level of 0.4 kV (phase to phase voltage).All PoC are anonymized, and the data do not contain any information about the consumers and their location.Each PoC has the identifier "meterID'', which has a value of 1-10 0 0 and is also indicated in the name of individual files [2] .
Electricity consumption is shown in 15-minute intervals throughout the year 2016.Each 15minute value is measured as an average active,P, and reactive,Q, power consumption in kW or kVar.
The dataset is published in raw form in .jsonformat.The folder in the repository has the following structure.The folder contains 10 0 0 .jsonfiles, each representing one PoC, one household.
There is also a file named "meter_info.csv"in the folder, which contains the ZIP code and reserved capacity of each PoC.
The following section describes the structure of .jsonfiles in detail.Each .jsonfile contains the following columns: • 'meterID' column: -identifier of point of consumption • 'day' column: -date information, day in a year • 'month' column: -date information, month in a year • 'year' -date information, year • 'lowConsumptionSum' -sum of the 15 minute values of active power consumption in current day between 0 0:0 0 -9:30 and 21:45 -23:45.This is power consumption in low price tariff.The following section describes the structure of meter_info.csvfile in detail.The file contains the following columns: • 'meterID' column: -identifier of point of consumption • 'ZIP' column: -ZIP code for each point of consumption • ' reservedCapacity' column: -reserved capacity for each point of consumption [kW] The following graphs shows the active power consumption, Fig. 1 , and lagging reactive power consumption, Fig. 2 , from the first to the 8th of January in PoC 104.

Experimental Design, Materials and Methods
Presented power consumption datasets were acquired using industrial direct single phase or three phase power energy meters, intelligent measuring systems (IMS).All IMS power meter were property of distribution system operator and are used for continuous measurement of consumption and supply of active energy and reactive energy.The basic interval for remote reading and processing of measured data is at least ones per day.Specific brands and models of IMS are not known.In practice, IMS of various manufacturers are used.All IMS power meters were in accordance with Measuring Instruments Directive (MID) 2014/32/EU [3] .The accuracy class for all used IMS power meters for measuring active and reactive power consumption was 1 in accordance with the IEC 62053-21:2003 [4] .DSO agreed to publish the approximate location of installed IMS devices (ZIP code level).Information on the reserved capacity of PoC was also provided.These are published as the maximum consumption allowed of PoC in KW.Further details on IMS devices and PoC parameters are not available for publication.
The data provided are in the form of .jsonfiles.We recommend using the simple Python script below for data processing.
Another option how to quickly work with data are online converters [5] .

Limitations
Not applicable.

Fig. 1 .
Fig. 1.Example of data -active power consumption from the first to the 8th of January in PoC 104.

Fig. 2 .
Fig. 2. Example of data -lagging reactive power consumption from the first to the 8th of January in PoC 104.

Table
SubjectElectrical and Electronic Engineering Specific subject area The dataset of average 15 minute values of active and reactive energy consumption of 10 0 0 individual anonymized households, however information about ZIP and reserved capacity are public.Data are collected from a mix of rural and urban households, during year 2016.