Cache Money
Team Members
- Matt Dickerson
- Brennen Direnzo
- Cole Duclos
- Nathaniel Frank
- Andrew Fruge
- Tim Furlong
- Lars Hager
Problem Identification
Photovoltaic solar panels have become increasingly used throughout the country; in some places up to 10% of the area's power (and even up to 40% in remote places such as Maui) comes from solar. The problem with solar panels is that cloud cover can dramatically decrease the amount of power produced by these PV arrays within a matter of seconds. With the increasing numbers of PV arrays being used, power companies need to be able to accurately forecast the cloud coverage and predict the amount of energy that will be produced from these solar panels. Solar forecasting will help companies prepare and store enough energy for these periods of cloud cover.
Project Concept
Our goal is to create a long lasting and fairly inexpensive sensor that will take images of the clouds, measure wind speeds/directions, and send this data to a server (via internet) every couple of minutes. This server will be programmed to create a time lapse in order to predict what areas will be covered by clouds up to 2 or 3 hours in advance as well as calculate surface solar flux and the amount of energy that will be harnessed over the course of these future few hours along with the associated predicted error. This way, power companies will know if they will need to be prepared to rely on power from other sources in the upcoming hours.
This idea can be accomplished by using low cost and reliable sensors and cameras spread around a large area. Ideally, something small that can be attached to the PV arrays is already assimilated into the grid. A possible choice may be a basic smart phone (with any necessary attachments/sensors to be built) because they incorporate sensors and have an attached camera. This would allow us to use cellular networks (3g or 4g) to transmit the data to be processed. The data needs to be transmitted with no more than a 20 second delay to a server.
This server will analyze the received data and predict future areas of cloud cover and the energy production of the PV array(s). The phones would also need to be modified to include a small PV array to power themselves and possibly a way of diffracting direct sunlight from the camera lens to prolong its lifespan. Our goal would be to have a minimum of 1 - 3 of these sensors deployed and functioning at the end of the process to prove the feasibility of this concept.