Using AI in battery energy storage systems keeps the power on 24/7

It’s no surprise that more industrial and commercial businesses are embracing green practices in a big way. With almost a quarter (24.2%) of global energy use attributed to industry, its rapid decarbonization is a critical component of a net-zero future and remains the subject of new sustainable standards and government regulations across the world.

Adding further pressure are increasingly eco-conscious consumers, demanding the companies they spend with go the extra mile to be as environmentally friendly as possible. This is seen in a recent analysis of the stock market, which revealed a direct link between pro-sustainability activity and positive stock prices impact.

More than ever, though, going green isn’t just about ticking the environmental, social, and governance (ESG) boxes, but an issue of energy security. For years, traditional fossil-based systems of energy production and consumption — including oil and gas — have become increasingly expensive. Add to that the current energy crisis, and businesses now face historic energy price highs not seen since the early ’70s as well as widespread supply issues. For energy-intensive industrial and commercial premises, where continuous power supply is often mission critical, this places an even greater onus on sustainability to mitigate the risks of escalating fuel prices and market volatility.

The result is a profound shift in the energy landscape, as more companies move away from the entrenched centrally run energy model and transition to self-generation for a more sustainable and secure future.

Decarbonization, decentralization, and digitalization

As with most aspects of the highly complex energy category, this transition is not necessarily a simple one.

To understand why, it’s important to first consider what are widely established as the key drivers of this change: decarbonization, decentralization, and digitalization. While they each bring their own set of benefits, they also bring challenges too.

In terms of decarbonization, global industry continues to make progress toward reducing emissions and, in turn, energy costs by ramping up the pace and scale of renewable investments. But, while this shows progress, the reality is that the inherent variability of wind and solar poses some limitations.

Solar, for example, will only generate electricity in line with how much sunshine there is and will not match the same profile of the electricity that a site is using. Used in silo, companies are left with having to top off with electricity from the grid or waste any excess generated.

Adding further complexity is the opportunity for decentralization. The decentralized nature of renewable generation holds the potential for power users to not only produce much of the electricity they need locally but also transition to an independent energy system, such as a microgrid, for the ultimate in self-sufficiency. One of the major benefits of a microgrid is that it can act as part of the wider grid while also being able to disconnect from it and operate independently, for example, in the event of a blackout. Of course, this presents a huge advantage for mission critical applications, where even a moment’s downtime can entail huge operational and financial implications.

But this also brings challenges. Although a decentralized approach makes for a more resilient and secure system, it must be carefully synced to ensure stability and alignment between generation, demand, and the wider central network.

Achieving this and meeting decarbonization goals requires digitalization. This will lead to a shift toward advanced energy management software that allows real-time automated communication and operation of energy systems. Such software will allow businesses to optimize the generation, supply, and storage of renewable generation according to their requirements, the market, and other external factors.

In the future, it is predicted that companies could even go beyond self-sufficiency and leverage a lucrative new revenue stream by reselling excess generation, not just back to utilities but even direct to consumers or other businesses.

But for now, the focus is on finding the most suitable framework for delivering this new layer of next-generation intelligence to the evolving energy system.

Battery energy storage systems

The answer to this and many of the other key challenges facing this energy transition lies in battery energy storage systems (BESSs).

Behind-the-meter BESS solutions already form a central part of decarbonization strategies, enabling businesses to store excess energy and redeploy it as needed for seamless renewable integration. When partnered with an energy management system (EMS), monitoring, and diagnostics, BESSs allow operators to optimize power production by leveraging peak shaving, load-lifting, and maximizing self-consumption.

Another big advantage is that these systems can provide critical backup power, preventing potential revenue losses due to production delays and downtime. But, there’s more.

Smart gets smarter

Beyond tackling decarbonization, applying AI takes BESSs to a completely new level of smart operation.

As many operatives know, energy storage operations can be complex. They typically involve constant monitoring of everything, from the BESS status and solar and wind outputs to weather conditions and seasonality. Add to that the need to make decisions about when to charge and discharge the BESS in real time, and the result can be challenging for human operators.

Introducing AI can now achieve all of this for a much more effective and efficient energy storage operation.

This unique innovation takes a four-pronged approach: data acquisition, prediction, simulation, and optimization. Using advanced machine learning, the system is able to constantly handle, analyze, and exploit data. This data insight is partnered with wider weather, seasonality, and market intelligence to forecast future supply and demand expectations. As a final step, a simulation quantifies how closely the predictions resemble the real physical measures to provide further validation.

The result is radical new potential for energy and asset optimization. Predictive analytics will allow commercial and industrial operators to save and distribute self-generated resources more effectively and better prepare for upcoming demand. It can also ensure “business as usual” in the ability to identify and address issues before they escalate and anticipate similar failures or performance constraints.

Greater intelligence is incorporated throughout the system, which allows operators to understand everything from the resting state of charge to the depth of discharge and how these factors can degrade the battery over time. This intelligence makes it easier to predict wear and tear. It also increases overall life span of the equipment and, ultimately, the return on the investment for the end user.

There is no doubt the energy transition is on, as decarbonization, decentralization, and digitalization continue to redefine everything we thought we already knew about how to produce and consume energy. While this brings new complexity for industrial and commercial operators, it also provides an opportunity to reimagine environmental strategy and take advantage of innovation. With benefits that include significant energy reductions, asset optimization and mission critical reli-ability, the transition to AI-enabled BESS is an inevitable and intelligent one.

Carlos Nieto

Carlos Nieto is global product line manager for energy storage at ABB.