Campus Energy Dashboard

Problem Description

Energy efficiency and sustainability are two topics gaining more and more attention, but do you know how much energy the Bucknell campus uses? What about the engineering building alone? Or the residence halls? Are there ways we can save energy? If we want to limit our energy consumption, we surely need GOOD ways to monitor our usage. Anyone who has ever has gone over their 500 minutes per month cell phone plan or used more than their prescribed number of text messages knows the importance of keeping tabs on consumption. This concept is sorely lacking when applied to our energy. Even if you wanted to, there is currently no easy way to monitor minute-by-minute energy consumption.

Goals

The goal of this project is to create a campus-scale energy dashboard. The format of the dashboard is up to the team but this could be a kiosk installed in the lobby of each building, information displayed on LCD screens already installed in many buildings (e.g., Dana), a smartphone app that anyone could access,  a dynamic webpage (e.g.,energy.bucknell.edu), or any combination of these. The team would have to coordinate with facilities to automate collection of this data from the existing meters and investigate how to display the information in a way anyone can understand. The basic information should include electric, hot water, and chilled water consumption (in the appropriate energy units) as well as any on-campus generation (along with fuel consumption) and imported energy from the grid. This data should then be compared to historical norms or targets set by facilities.

Criteria

Basic questions

• Communicating with various energy meters
• No common protocol — may need protocol conversion layers
• Efficient collection and storage of fine-grained energy data: energy sampled up to 1 Hz but need to visualize with variable fidelity (eg daily snapshots) — how do we make this efficient?
• Processing of energy data for efficient display/searching
• Display the data in an intuitive way

Advanced questions

• Detect meter errors from correlated meters: whole building meter with metered sub-circuits

• Fault detection / safety analysis: 200A rated line at or near capacity; power quality issues

• Basic non-intrusive load monitoring (detect big loads switching on/off)

Skills required

• Web services and low-level TCP/IP networking

• Understand basic serial communication protocols

• Large database design and optimization

• Interactive graphics/GUI design

Resources

• You might want to take a look at this web resource that showcased what is going on with electric power generation on campus.

Point of Contact

• Prof. Alan Marchiori, Dept. of Computer Science, Bucknell University