A production plan for solar power.
Sunshine can be hard to predict. However, new software will make it possible to predict the electricity production from a photovoltaic (solar cell) system.
Just like any other producers, owners of solar cell systems want to know the scale of production in advance. This is valuable knowledge, because it makes it easier to control both the solar cell systems and the overall electricity system. That is why Toyota contacted a research team at DTU Compute.
“After a year’s collaboration, we’ve developed tools capable of predicting the electricity production from solar cells a few hours ahead with the accuracy relevant to industry,”
says Peder Bacher, Assistant Professor at DTU Compute.
Some people might wonder why the car manufacturer Toyota is interested in solar cells. However, the Japanese group has extended its interests considerably in recent decades to include much more than cars. Today, Toyota offers a number of housing products and various innovative sustainable energy solutions.
From rooftops to hybrid cars
The nuclear accident at the Fukushima power plant in 2011 has also increased the interest in sustainable energy in Japan. Despite currently accounting for less than one per cent of the country’s electricity production, the solar cell market is growing, especially in line with the investment increase in solar cells for private homes.
Toyota is interested in the solar cell market for private homes, but to a large extent also the possible future application of solar cells for cars. On a world-wide scale, in 2014, Toyota sold more than eight million hybrid cars, which are driven by an electric motor with a chargeable battery and a gas-fuelled or petrol-fuelled engine. In future, solar cells mounted to the car will possibly be able to contribute to charging the battery.
The ability to predict the electricity production from solar cells is interesting for both private homes and large-scale solar cell systems, and in relation to future applications in cars. The research team at DTU Compute headed by Professor Henrik Madsen and Associate Professor Jens Starke has also taken an interest in the topic for some years.
“We’re pleased to have the opportunity to further develop the tools for this purpose through our collaboration with Toyota,” says Peder Bacher.
Software provides immediate answers
The ability to predict the electricity production from solar cells is interesting for both private homes and large-scale solar cell systems, and in relation to future applications—e.g. for charging car batteries. Therefore, Toyota is working with DTU on developing ‘prediction software’.
The team works together with researchers from different departments at DTU and with meteorologists at the Danish Meteorological Institute (DMI) to optimize renewable energy systems, primarily wind power.
Combining solar cells and wind power is generally a good idea, as solar cells produce maximum output during summer, while wind power production peaks in autumn and winter. Being able to predict the scale of production from various energy sources is, nevertheless, crucial in order for systems with varying output to succeed.
“Our colleagues at DTU Wind Energy and DMI develop complex meteorological models. We develop models that translate their weather forecasts and combine them with other data in a way that provides an estimate of the future production from a given solar cell system,” explains Peder Bacher.
“In this connection, we consider the simplicity of the algorithms a huge benefit. It can take up to four hours to run the calculations in the overall meteorological model, but obtaining estimates of solar cell production is only a matter of seconds. This increases the usability of the solution significantly.”
Enter solar cells
The need to be able to predict the output from solar cells will be reinforced by the growth in the total share of electricity production based on solar power. Today, solar cells account for 1.5 per cent of the electricity produced in Denmark. During summer, it can reach 3.7 per cent—which it did in June 2014.
“Because the share is still so modest, the system operator can balance the electricity system with the current backup reserves. But as the installed capacity increases, the ability to predict solar cell output will soon prove crucial. It’s a good idea to be prepared for the situation before it occurs,” argues Peder Bacher.
“In some Danish areas, for instance Bornholm, solar cells already account for a large share of the electricity production, and the numbers are even higher in some areas abroad. So already today, our work is of both local and international relevance.”
Article by Morten Andersen, DTU Online News