metadata

dataset_name: hourly-load-curve-ons
tags:
  - energy
  - electricity
  - power-grid
  - time-series
  - hourly
  - Brazil
  - smart-grid
  - ons
license: cc-by-4.0
language:
  - en
  - pt
pretty_name: Hourly Load Curve - ONS (Brazil)
task_categories:
  - time-series-forecasting
size_categories:
  - 10M<n<100M
dataset_info:
  features:
    - name: id_subsistema
      dtype: string
    - name: nom_subsistema
      dtype: string
    - name: din_instante
      dtype: timestamp[ns]
    - name: val_cargaenergiahomwmed
      dtype: float64
  splits:
    - name: train
      num_bytes: 5790000
      num_examples: 185280
  download_size: 2304250
  dataset_size: 5790000
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

🇧🇷 Hourly Load Curve - ONS (Brazil)

This dataset contains hourly electricity load data for Brazil, published by the ONS - National Electric System Operator. It spans from the year 2000 to the present (currently 2025), with continuous updates.

📌 Description

The data represents the hourly electricity demand profile across the Brazilian National Interconnected System (SIN). It is especially suitable for:

Electricity load forecasting
Energy demand pattern analysis
Time series and machine learning modeling
Smart grid simulation and optimization
Academic research and planning studies

🗃️ Dataset Structure

The data is provided in .csv, .xlsx, and .parquet formats and organized by year (2000–2025). Each row corresponds to a specific hour of a year and includes:

Date and time
Total load of the SIN
Load per subsystem (SE/CO, South, North, Northeast)

🧪 Example Usage

import pandas as pd

# Load one year of hourly data
df = pd.read_parquet("CurvaCarga-2023.parquet")
print(df.head())