How AI Powers Today’s Advanced Microgrids
Microgrids are on the rise, fueled by drivers such as rising demand for clean energy to enable the net zero transition. Along with the growing requirement for reliable & secure global power supply, the microgrid market is projected to reach USD 47.4 billion by 2025 from USD 28.6 billion in 2020, at a CAGR of 10.6% between 2020 and 2025. With the rethinking of critical infrastructure during the 2020 pandemic, the role of microgrids for commercial & industrial sites has become even more focused. Microgrids are self-contained electrical networks connecting distributed energy resources (DERs) and loads, acting as a single controllable entity, and can operate on either grid-connected or islanded/isolated modes. As such, they are a perfect example of digitization enabling the green energy transition. To understand the latest technological advancements of autonomous and dynamic control of energy production and consumption, we met with one of Schneider Electric’s microgrid experts.
François Borghese: Microgrid Marketing Director
Since 2011, François has lead Schneider Electric’s microgrid marketing actions and contributes to its strategic positioning. He has played a key role in the creation of several offers: an energy Management System called EcoStruxure Microgrid Advisor and a Microgrid controller called EcoStruxure Microgrid Operation. Together, these solutions address all types of microgrids globally, from fully connected, to islandable to pure islanded microgrid.
With the increased interest in microgrids, what use cases have you seen evolving for Schneider Electric’s clients in the past two years?
Today, the changing energy landscape is moving the needle on many microgrid benefits, including:
- Strengthening and greening a sites’ local infrastructure
- Connecting onsite renewable power generation and storage
- Meeting zero-emission goals
We see clients gearing up their infrastructure for EVs with microgrids, as existing networks often have capacity limitations to integrate EV charging. With the pandemic and other natural disasters we have experienced in the past years, the need to secure the supply of critical loads has grown. During a blackout, for example, our clients need to prevent production losses, low workforce efficiency and equipment damage, and microgrids can help them to stay open in a time when their customers need them the most. Last but not least, we have clients gaining considerable energy cost savings and accessing new revenue streams with their enhanced energy flexibility.
Through our new optimization algorithms with EcoStruxure Microgrid Advisor, we enable not just a single, one-size-fits-all solution, but many very unique use cases to meet individual organizational needs. You can think of this software as the brain of our advanced microgrid solutions.
Tell us more about how the brain of the microgrid functions, and why is this an example of AI powering the future energy landscape?
EcoStruxure Microgrid Advisor consists of a cloud-based platform running Model Predictive Control algorithms to define the microgrid's optimization strategy. These algorithms combine, for example, price signals, historic data, weather forecast and operating schedule to better forecast energy supply and demand variations within the microgrid’s boundaries. A local edge controller interfaces with the DER on site for monitoring and control, thus making the DERs interact in an optimized way. Today’s advanced microgrids have the power to run real-time optimization, enabling use cases like frequency regulation or demand response that usually need an optimization action faster than 1 sec.
In a nutshell: AI helps to better and faster forecast energy supply and demand variations across a microgrid. With AI, a microgrid can successfully manage a complex energy structure, including new variables such as renewable power generation or rapidly changing energy prices.
Do you have some practical examples of how users can take advantage of this AI-enabled, autonomous and dynamic control?
Our latest update to EcoStruxure Microgrid Advisor, released this November, was about some very practical extensions for the evolving use cases of microgrids. On dynamic tariff management, for example, we see electricity markets with increasing price volatility and even the phenomenon of negative prices happening in more markets, more frequently. Negative pricing can occur when demand for electricity drops to an extent that suppliers are prepared to pay others to take it away – something that we see happening in markets with an increased share of renewables like Germany, the UK, or Australia. Because of this, we include the use case of negative power prices in our optimization algorithms – together with an advanced interface and faster reaction to energy spot markets.
AI-powered forecast algorithms are critical for such cases, as they enable flexibility for the most critical demand. For example, if you have a peak at $1 and in one hour another peak at $1.50, it’s better to have enough energy available for the second peak. This is possible thanks to the forecasting algorithms.
For one of our clients in Australia, the South Australian Produce Market, tariff rates can range from as low as $0.07/kWh* to as high as $1.88/kWh* on a normal day along with negative pricing on certain days in Australia. Our EcoStruxure Microgrid Advisor plays on that flexibility to optimize their operations based on the price signals in 3 ways:
- Discharging its battery loaded with solar energy when the prices are on a high
- Charges battery in the event of negative prices and
- Alternating to self-consumption when the spot market is flat
Schneider Electric Solution for Price Volatility
The customer had installed 1,600 solar panels and a large lithium-ion battery and will cut power bills by around $500k a year. The whole use case is even more complex when you consider that it also increased reliability in case of a power outage and reduces annual greenhouse gas emissions by 32%.
What else is new from a technology perspective, and what capabilities will microgrids need for the future energy landscape?
A big trend is Sector Coupling or Integrated Energy Concepts. We see many clients moving away from a focus on classical electricity generation and loads. A big push for our microgrids actually comes with the EV trend. These microgrids often have to bring intelligence into a site’s or city campus EV charging pattern: when and how do I best charge my electrical fleet, or could I even discharge my cars at certain grid events, using them as a connected battery or virtual power plant? We now have a plug-in load management system (LMS) for EV and battery cycle in our solution and see this as a future area for even more extension and experimenting, for example with buildings as active energy storages.
Today’s complex microgrids require more data to deliver automated, accurate forecasting of short-term production and consumption. It is critical to have an intelligent artificial intelligence system to automatically learn the local context and then issue a forecast. This is also about the ability to absorb and evaluate thousands of touchpoints across the grid. An automated model, with artificial intelligence and machine learning such as our Model Predictive Control, will ultimately be able to scale with fine precision.
What is your advice for clients interested in these advanced microgrids? How can a company find out if this fits their facilities?
Not too long ago, microgrids and AI were considered a thing of the future. But pushed by the pandemic, it’s becoming clear that they are a practical, strategic direction. Many organizations are ready to place their trust in AI and are now considering intelligent microgrid systems to both improve resilience during recovery and to protect their business from future disruptions.
As discussed, microgrids offer advantages that can address a range of facility goals. We have successfully designed, built and maintained over 300 advanced microgrid projects. But to make the most of a microgrid investment, you must understand the bottom-line impacts of how the system is designed, powered and financed. So, if you’ve been wondering if a microgrid is the right solution for your facility, a microgrid feasibility study can help you understand your options.