Data Analytics and Visualization for Large Oil and Gas Datasets
This project is funded by an NSERC Engage grant in partnership with Evaluate Energy / CanOils.
Data collection, analysis, and visualization are complex problems compounded by the size of the datasets involved. Our industry partner collects data from over 770,000 oil and gas wells across Alberta, British Columbia, and Saskatchewan. Each well has over 300 attributes, making it a quite large dataset. Each well is updated monthly, and our partners need to be able to understand trends within all of this data quickly in order to make time-sensitive recommendations to their clients.
In this project, we will create tool support for data analytics and visualization to allow our partners to view, understand, and share large datasets. This tool will allow users to understand which wells are performing best and what factors lead to this performance by mining the dataset for patterns.
One of the research questions we will evaluate during this process is whether high and ultra-high resolution displays help teams and individuals to work better with data visualizations. To understand this, we will compare visualizations on standard monitors against large 4k TVs against the 35 megapixel, 4.88 meters by 1.85 meters visualizations wall in the Taylor Family Digital Library at the University of Calgary. Finding techniques for interacting with such large displays will also form a key challenge.
Overview showing heatmap of production for one well type.
All well types shown at once. The amount of data quickly becomes overwhelming.
Detailed view of the same wells.
Progress to Date
Most work to date has dealt with data formatting, experimenting with potential cloud-based systems, and experimental visualizations using existing tools.