The concept of microgrids is appealing. Intelligent independent energy that is oftentimes associated with renewable resources making a difference to local people. There are of course drawbacks, usually the cost, but with the help of government grants and the newest breakthroughs in AI technology microgrids are becoming more accessible. So what is a microgrid and how does AI make them better?
What Makes a Microgrid?
A microgrid can be defined by three things. The first is that it needs to be local to its clients, often being near or even in the same building as those using its energy. This is done to lessen the loss of energy transferred from it being generated to where it needs to go.
The second aspect is its ability to be independent. This isn’t to say a microgrid can’t be connected to a central power station, but if that station goes down it shouldn’t affect a microgrid. This is done to lessen the amount of things that can go wrong and speed up how long it may take to find issues that occur.
The last part is intelligence. What is considered intelligence is actually the microgrid controller. This is the system that controls generators, batteries, where to store energy, and even which buildings get power at certain times. This last aspect often involves AI’s the most.
Specific Consumption Needs
When looking at the importance of AI in microgrids it's best to look at a study done by Jan-Peter Doomernik and Age van der Mei. Their 2018 study outlined that artificial intelligence can be beneficial for microgrid planning, specifically how AI dispersed energy to certain areas.
Their work reported that “authors assess that artificial intelligence would be able to plan with a higher detail: load per connection and potentially even load per application/device.” The study goes on to say that, “the planning could be done once a day and even once an hour if enough computational capacity is available.” This shows that AI controlled microgrids can solve energy consumption problems unique to local areas. They can get the right amount of energy to the right locations based off of previous demands. It lowers the stress on the microgrid only and helps tackle hyper specific issues a normal power grid couldn’t do while maintaining peak energy levels.
Business Taking It Seriously
To see the impact AI’s have on a microgrid's ability to operate it’s best to look at when technology company Sustainable Power Systems was helped by AI company BluWave-ai to update their microgrid. When asked why AI was being implemented it was explained by BluWave-ai’s CEO Devashish Paul that, “using AI for forecasting allows the system to find unique, optimal solutions for a particular situation.”
He also said that “for example, forecasting the load of a community at a certain location, the AI can take in various inputs including weather, situational data, and historical data to forecast better than a fixed rules-based system.” He’s basically repeating what was learned by Jan-Peter Doomernik and Age van der Mei’s study that AI is making extremely powerful strides to improve microgrids, and a business as large as Sustainable Power Systems is getting AI involved. It really is the best tool for the job.
Helping To Create More Efficient Microgrids
Jan-Peter Doomernik and Age van der Mei have also published work on how AI can better help create microgrids as well. In a 2019 study was conducted AI and human microgrids. This study followed two different music and culture festivals with about 7000 guests each. One had humans set up the grid the other had an AI do it. In The second study showed that AI’s did have a positive impact on the design of microgrids.
The study mentioned, “the AI output results in somewhat higher power, compared to the human designs, mainly due to the redundancy factor applied, as well achieving as a saving on the grid length.” The report did mention that “an estimated 30% cost reduction is possible compared to the human grid plan and 40% cable length reduction compared to the realized grid plan, highlighting there is a small but significant difference between a grid plan and a realized grid.” This is great news, it shows that AI can be used to help get microgrids to their most efficient.
It was pointed out that, “at the current stage of artificial intelligence development, the best approach seems to assist human planners not relegate them to the role of ‘checking’ such as in certain legal and medical tasks.” This is also good news, it means that humans won’t be replaced, just greatly aided when it comes to how and where microgrids are built.
So with studies and businesses showing how AI has improved microgrids, what are some real world applications? One example is Metron. The program has been created by a number of different groups like Accenture, Alstom, Engie, and Schneider Electric to assist in the microgrid REID. Renewable Energy Integration Demonstrator or REID has been designed to speed up the distributed energy systems in South East Asia, and because of AI like Meron, it is having success.
Metron uses “multiple data sources from industrial systems and others such as the weather, while interfaced to energy markets to deliver energy efficiency and savings in real-time.” This AI is designed to use all the available data to make the most money for selling the energy generated as transparent as possible. It also has the extra task of tracking different types of energy gathers like solar, wind, and diesel power. Across every section, there was profit because of Metron. It has shown that AI’s can make microgrids more profitable.
Taking It Seriously
Microgrids are providing accessible energy to local areas that may not get the attention they need. AI is assisting these processes making them more profitable, and efficient than ever before. AI is allowing these grids to operate at the best of their ability making not just corporations but nations take microgrids more seriously.