How to Use AI to Design Commercial Facility Solar and Storage Systems

Businesses have long spent a significant portion of their budgets on energy. The average commercial enterprise typically spends between $2 to $4 per square foot of facility space every year to power its day-to-day operations. In the past, the objective of business leaders was to identify the most reliable energy source of electricity, balancing it against cost. Unfortunately, utility providers were the only option available to them.

While cost sits squarely at the top of the list of factors, two other key realities continue to rub up against it:

1. ESG (environmental, social, and governance) targets continue to put pressure on businesses to green their operations. And consumers now more than ever prefer to do business with companies that are taking tangible steps to achieve these goals.
2. Green technologies have become more affordable, enabling businesses to reduce their carbon footprints. As a result of econcomies of scale and an increase in global competition, there’s no excuse now not to rely more on greener technologies.

But despite these two forces, new technology means new hurdles that businesses have to overcome to effectively adopt and implement a greener operation.

Avoiding Dirtier and Expensive Operations when Attempting to go Green 

The challenge of relying on renewable energy sources, such as wind or solar, is that the energy they produce is intermittent. For this reason, it makes it extremely difficult, if not impossible, to predict how much power each site will produce at any given time. Installing batteries helps reduce this risk, but businesses continue to struggle with constantly changing weather conditions, happening everyday in real time.

Assuming your photovoltaic (PV) panels and batteries are sized correctly to your specifications, running commercial properties with solar and battery storage installed against the above factors is no small task. What typically happens paradoxically, businesses who try to green their operations end up with more waste, dumping, device failures, and energy shortages.

Sometimes, when done incorrectly, going green can be dirtier and more expensive. Even if you and the developer did their due diligence, such as:

• Developers take careful measurements based on the site’s geographic location and historic energy consumption before sizing distributed energy resources (DERs) including solar, wind, storage, and electric vehicle-charging.
• Property managers manually create a predetermined schedule to balance on-site power generation, consumption, and storage.

Unfortunately, both approaches do not produce the best results given how quickly weather can change week to week, day to day, and hour by hour. When humans are left squarely in the driver seat, there will always be slower reaction times and, as a result, imbalances that can cost money.

On the other hand, machines today are built to take on large data sets to create more accurate predictions and forecasts.

How AI Drove the Design of a Commercial Solar and Storage Operation

The recent winter storm in Texas during the 2020-2021 winter seasons created massive electricity spikes. The colder temperatures impacted access to fuel and made other conventional and renewable energy sources unavailable. One office complex in Plano, TX experienced quite the shock when their utility bill was the equivalent to three years worth of energy consumption.

While these facilities had a rooftop PV system, it lacked batteries to store any excess power during times when demand was low. If they had these batteries installed, they could have potentially reduced their need for energy from the grid and avoided the extremely high prices that drove their bill so high.

To future-proof their energy system, they turned to Empower Energies, a leading provider of turnkey solar and storage solutions for commercial, industrial, and municipal stakeholders. To adequately plan a design that would meet the needs of the office complex, Empower Energies leaned on Veritone iDERMS Designer, a capacity sizing solution.

Using patterned artificial intelligence algorithms created by Veritone, they were able to input years of price, load, and solar generation data to inform three key decisions:

1. How to size batteries to generate the most value at minimal cost
2. How to design a battery-coupled PV system to maximize possible revenue streams
3. How to coordinate battery discharging and charging for optimal performance and savings

Executing this project through a process that was highly collaborative between teams, an optimal BESS design was identified as the best option to generate an increase in revenue through energy arbitrage. Arbitrage is the process of buying grid electricity when inexpensive to charge the battery and selling when utility rates go back up. 

The process also revealed even more revenue opportunities through value stacking, which is layering multiple revenue streams and engaging in each when most beneficial. The total profit forecast from participating in ERCOT’s Ancillary Service programs, demand response, and the non-monetary value in providing backup power together surpassed the value through arbitrage alone.

Robert Duva, the Vice President of Engineering at Empower Energies noted that “to avoid the worst effects of climate change, deploying solar and storage isn’t enough – not even at scale. We need better tools to help us navigate the unprecedented environmental challenges on the horizon. And so far, AI has consistently proven to be the most promising technology at our disposal.”

With artificial intelligence informing human decisions on grid development, businesses with unsustainable operations today can adopt greener solutions. With the processing power behind the humans, developers can work more effectively with clients to uncover the most environmentally friendly, cost-effective, and optimal configuration for their specific use case.

Learn more about this project.

Sponsored by: