Inspiration - Usage and billing data is a great resource that allows us to more efficiently allocate electrons in a microgrid.

What it does - Use case: shared solar system for individual dwellings in a townhouse/apartment/community. Each dwelling has a smart meter. In the business as usual case, power is allocated based on individual demand so as to minimise the amount of electricity required from the grid. This is the case since in most jurisdictions the feed in tariff is a lot less than the retail price of energy.

This isn't the most efficient way to allocate energy however. If we incorporate another objective into the algorithm, namely, that we allocate more solar energy to those whose instantaneous electricity price is highest at any given time, we are minimising the collective bill. Since in both cases (BAU and our case), the savings are shared between all occupants, adding this extra rule into the algorithm means that the occupants save more money than BAU.

How I built it. We have explored the problem space by first obtaining recorded power output of a real PV system and several examples of real world usage (half hour interval data, measured over 1 day in each case). Next we took this data and structured it such that it is appropriate for a decision making algorithm to process.

We also did some experimentation with decision making using the processed data.

Finally, the basic rules by which the algorithm should allocate power amongst stakeholders has been established. These are mentioned above.

Challenges I ran into

-idea complexity initially under estimated -business case took longer than expected to refine

Accomplishments that I'm proud of

-We came up with a viable (early/tentative) business proposal

-We evaluated our idea using real world data

What I learned

What's next for AirPV

Share this project: