Dark Mode
Mining Flotation Plant Process Data
Education & Learning Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This dataset provides real-world industrial data from a flotation plant, a critical part of a mining process. It aims to assist manufacturing plants in improving efficiency by enabling the prediction of impurity levels, specifically the percentage of silica, in the final ore concentrate. The ability to predict silica concentration hourly allows engineers to obtain early information, empowering them to take proactive corrective actions to reduce impurities and minimise ore waste in tailings, which also benefits the environment. This dataset offers a rare opportunity to explore actual manufacturing plant data for quality prediction.
Columns
- date: Date and timestamp, ranging from March 2017 to September 2017. Some entries are sampled every 20 seconds, while others are on an hourly basis.
- % Iron Feed: The percentage of iron present in the iron ore that is fed into the flotation cells.
- % Silica Feed: The percentage of silica (impurity) present in the iron ore that is fed into the flotation cells.
- Starch Flow: Starch (reagent) flow, measured in cubic metres per hour (m³/h).
- Amina Flow: Amina (reagent) flow, measured in cubic metres per hour (m³/h).
- Ore Pulp Flow: Ore pulp flow, measured in tonnes per hour (t/h).
- Ore Pulp pH: The pH scale of the ore pulp, ranging from 0 to 14.
- Ore Pulp Density: The density of the ore pulp, ranging from 1 to 3 kilograms per cubic centimetre (kg/cm³).
- Flotation Column 01 Air Flow: The air flow into flotation cell 01, measured in Normal cubic metres per hour (Nm³/h).
- Flotation Column 02 Air Flow: The air flow into flotation cell 02, measured in Nm³/h.
- Flotation Column 03 Air Flow: The air flow into flotation cell 03, measured in Nm³/h.
- Flotation Column 04 Air Flow: The air flow into flotation cell 04, measured in Nm³/h.
- Flotation Column 05 Air Flow: The air flow into flotation cell 05, measured in Nm³/h.
- Flotation Column 06 Air Flow: The air flow into flotation cell 06, measured in Nm³/h.
- Flotation Column 07 Air Flow: The air flow into flotation cell 07, measured in Nm³/h.
- Flotation Column 01 Level: The froth level in flotation cell 01, measured in millimetres (mm).
- Flotation Column 02 Level: The froth level in flotation cell 02, measured in mm.
- Flotation Column 03 Level: The froth level in flotation cell 03, measured in mm.
- Flotation Column 04 Level: The froth level in flotation cell 04, measured in mm.
- Flotation Column 05 Level: The froth level in flotation cell 05, measured in mm.
- Flotation Column 06 Level: The froth level in flotation cell 06, measured in mm.
- Flotation Column 07 Level: The froth level in flotation cell 07, measured in mm.
- % Iron Concentrate: The percentage of iron in the final ore concentrate, measured in the lab (0-100%).
- % Silica Concentrate: The percentage of silica in the final ore concentrate, measured in the lab (0-100%). This is the primary target variable for prediction.
Distribution
The dataset is typically provided as a CSV file. It has 24 columns and a size of 183.77 MB. The data includes approximately 737,000 valid records. The data spans from March 2017 to September 2017, with some variables sampled every 20 seconds and others on an hourly basis.
Usage
This dataset is ideal for:
- Predicting the percentage of silica concentrate every minute.
- Forecasting the percentage of silica in concentrate several hours in advance to enable proactive decision-making by engineers.
- Conducting research on quality prediction within froth flotation plants, utilising various machine learning techniques, including deep learning.
- Developing solutions to enhance the efficiency of manufacturing plants.
- Supporting environmental initiatives by reducing the amount of ore that becomes tailings.
Coverage
The dataset originates from a real-world industrial mining flotation plant. The temporal scope covers data recorded from March 2017 until September 2017. Specific geographic or demographic details are not provided as it pertains to industrial process data. Data availability varies, with some columns recorded every 20 seconds and others hourly.
License
CC0: Public Domain
Who Can Use It
- Engineers: To gain early insights into ore concentrate impurity levels, enabling them to implement timely corrective actions to reduce silica content and optimise operations. They can use the data to act in a predictive manner, thereby minimising the amount of iron lost to tailings.
- Data Scientists and Researchers: To develop and evaluate machine learning models for quality prediction in froth flotation processes, exploring both traditional machine learning and deep learning approaches. The dataset supports investigations into prediction accuracy and lead times.
- Manufacturing Plants: To leverage data-driven insights to boost overall efficiency and refine process control strategies.
- The Data Science Community: To engage with and help answer open questions regarding the predictability of silica concentration and the optimal prediction horizon for industrial applications.
Dataset Name Suggestions
- Mining Flotation Plant Process Data
- Ore Quality Prediction Dataset
- Industrial Froth Flotation Data
- Silica Impurity Prediction for Mining
- Real-World Mining Efficiency Data
Attributes
Original Data Source: Mining Flotation Plant Process Data
QUALITY