Researchers at the Australian National University (ANU) are looking for volunteers across the Australian Capital Territory to take part in a new study that analyzes how user-installed solar panels interact with the power grid.
Every homeowner who chooses to participate will get data logger installed in their meter boxes to see how much solar energy is being produced. This will allow engineers to predict the amount of energy these panels pump into the grid and adjust the supply of conventional power stations accordingly. While it’s relatively easy to measure the total output of these panels over long periods of time, estimating the amount of energy they can generate at different times of day is another matter altogether, says Dr. Chrisfried Webers from ANU’s College of Engineering and Computer Science.
The University called on Data61’s Machine Learning Group – formerly National ICT Australia (NICTA) – who are working to develop methods of forecasting power output from rooftop solar energy systems.
“What we need is to be able to predict how much energy will be produced over five minutes to 60 minutes,” Webers said. “That’s necessary information for the energy market operator — they need information on what’s coming from hour to hour.”
Around 14 % of households in the ACT have operational photovoltaic solar panel installations, which means that one in ten houses can cover their energy requirements at least in part.
The facts show solar power is great for the users and the environment both, but engineers take this number with a pinch of salt – should the panels stop producing energy, these households will start drawing on the main grid. If this happens and the power stations are not prepared, it could cause rolling blackouts, especially as more panels are installed.
“It’s important for the local utility providers because they have a spinning reserve running and if they can anticipate an energy drop, they can ramp that up when they need to. ” Dr. Webers said.
Dr. Webers hopes that the data gathered from this volunteer project will allow his team to develop software that can accurately predict solar output from each suburb using low-cost data-logging devices installed on individual homes.
“On a clear sky day it’s quite easy to predict the amount of energy produced by solar panels, but the real challenge is when we have clouds because a cloud will lead to a drop in the energy production. We want to be able to counteract that drop in energy production by predicting the energy drop [by aligning it with the weather forecast].”
Dr. Webers said data from 60 houses across the ACT would be required to make the study viable. Should you choose to sign up, a data logger will be installed in your home’s meter box, allowing researchers to track how much energy the panels generate and when they’re the most productive during the day.
By Alexandru Micu, ZMEScience