Skip to content

Dynamically Powering Data Centers with Renewable Energy

Notifications You must be signed in to change notification settings

kevokevol/vgreen

Repository files navigation

Dynamically Powering Data Centers with Renewable Energy

This project visualizes the power usage of worldwide data centers, along with their relative CO2 outputs.

Data centers are set to consume 20% of the world's electricity by 2025. However, the energy used to power these centers are mainly from fossil fuels. The challenge with sustainably powering these data centers is that renewable energy sources, such as wind and solar, have highly unstable outputs, leading their capacity to be rated lower than their maximum output, and wasted output. Rong et al. identified a major challenge in determining the energy profile of data centers as a major challenge.

Our project seeks to solve that challenge. Together with technology that allows companies to selectively reduce power consumption at data centers, our visualization helps data center administrators identify which data centers are currently experiencing a surge in renewable energy, and dynamically reallocate energy needs to data centers using greener energy at the time.

The frontend is created using React and utilizes WebGL Globe and react-vis. The backend is created using Python3 and MySQL. A dump of the MySQL database at the time of our demo is located at data/dump.sql. There is a hardware component: Our worldwide power output is represented by a solar panel connected to our database through an Arduino. The circuit diagram for this can be found here.

Power station data is obtained from the US Energy Information Administration, and their longitude and latitude are then obtained using the Google Maps Geolocation API. THe MySQL database is hosted on AWS. We were not able to obtain data center locations, so we approximated them using VMWare's global office locations. CO2 footprint of each data center is calculated using a weighted average of carbon emissions from the three closest power centers weighted on their relative power output, multiplied by how much power is drawn by that data center.

This is the 2019 VMWare Borathon submission of Kevin Li, Andrew Chen, Aubhro Sengupta, James Kim, and Dan Kwon.

Live demo at be viewed here.

Startup Instructions

Clone the repo

Run npm install

Run npm run start

About

Dynamically Powering Data Centers with Renewable Energy

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published