Welcome to the third installment in our 5-part series on Smart Grid Technology. This article explores how autonomous microgrids are increasingly being used to deliver cleaner, cheaper, and more reliable electricity as the world transitions away from fossil fuels.

Table of Contents

What Is Microgrid Technology?

A microgrid is a collection of energy generation technologies used to provide continuous power to small local communities – from medical complexes to college campuses. Most microgrids use a combination of solar, wind, battery storage systems, diesel generators, electric vehicle (EV) charging, and other distributed energy resources (DER’s) that work independently of and in concert with the central utility grid. In fact, the line between traditional grids vs. microgrids is quite blurred since they both serve similar functions. The 2 primary differences are that:

  • Microgrids are usually privately owned and operated – lying outside the direct control of grid operators
  • The service area is more localized, with the majority of power generated going to stakeholders connected to that specific microgrid

However, because microgrids often take electricity from the network and feed power into it, they are also part of the macro grid. And the ability to balance energy supply and demand across the larger network is an essential component of smart grid design. This is especially important given society’s growing reliance on edge devices built around intermittent renewable energy generation technologies like solar photovoltaics (PV) and wind power.

How Does a Microgrid Work?

There are many different types of microgrids, with each relying on its own unique blend of:

  • Legacy energy technologies like diesel generators and utility power made from fossil fuel
  • Sustainable energy technologies like solar, wind, battery storage systems, and EV-charging

Most microgrids also have control capabilities, allowing them to decouple from the utility network and provide reliable power to local communities during grid malfunctions. In addition, microgrid operators increasingly use this autonomy to help control energy costs. For example, if demand for electricity suddenly spikes during peak demand when utility rates are high, that local community can draw more of its power from closer and cheaper sources within the microgrid itself.

Some microgrid control tools are designed to disconnect from the grid automatically, while others require that someone manually decouple from the network. Either way, it’s best to think of the entire microgrid as a self-contained “edge device” since all electricity flowing in and out passes through a single connection point linking that microgrid to the utility network.

In other words, microgrids are standalone power stations – regardless of how many distributed energy resources are contained within.

Examples of Microgrids in Action

Thomas Edison’s Manhattan Pearl Street Station was the world’s first electricity plant. Built in 1882, it helped lay the foundation for what would eventually become America’s utility grid. However, it is also technically the earliest example of a microgrid given that it was self-contained, served fewer than 100 residents, and relied on a mix of technologies, including:

  • Coal-fired steam engines and on-site battery storage
  • Combined heat and power (CHP) from steam generation

Due to their large campuses and fairly predictable loads, universities were some of the earliest pioneers of modern microgrid technology. Harvard, for example, introduced microgrid plans as far back as 2009, with many other schools quickly following suit.

Other early adopters of microgrid technology are critical service providers that require reliable, 24/7 access to electricity. Military installations like Fort Bragg and data centers like Google’s, Facebook’s, and Apple’s all fall into this category.

Because of the many advantages that microgrids offer, however, the technology is becoming more popular with research parks, medical facilities, and entire municipalities. Even Alcatraz Island has embraced the concept. Now a tourist destination, America’s most famous prison is home to one of the largest microgrids in the country – using a mix of solar PV, storage systems, and diesel generators to power its operations.

What Are the Benefits of Microgrid Technology?

Although microgrids aren’t entirely new, there are several factors driving the technology’s growing popularity.

1. Grid Resilience

Microgrids can disconnect from the utility network and run autonomously to supply reliable power to local residents – even when the rest of the grid goes down. In fact, access to emergency backup power is arguably the main driver behind the larger microgrid trend.

2. Improved Efficiency

On-site power generation for local loads reduces transmission losses – resulting in lower prices, more reliable power, and less overall waste.

3. Carbon Offsets

Increased reliance on solar power and wind energy is necessary in the fight against climate change. But both power technologies are intermittent – making it difficult to anticipate and balance energy supply for real-time needs. The growing use of battery storage helps to mitigate this. But so too does tighter integration among microgrids, the larger utility network, and any DERs connected to either.

For example, a sudden dip in solar production within the microgrid can be offset by drawing more power from the electric grid or from on-site batteries. Similarly, an unexpected surge in wind production can be fed into the utility network or stored in on-site batteries.

4. Costs and Savings

Microgrids cost money to set up, with the investment spread across all the privately-owned distributed energy resources used to generate and store local power. Once created, however, microgrids shield users from network-level power outages, which, according to the US Department of Energy, cost Americans an estimated $150 billion annually (or about $500 for every citizen).

In addition, microgrids give local communities the option of using whatever electricity is cheapest at that precise moment in time. In a typical scenario, microgrid operators might use:

  • Solar energy during peak production to provide continuous power to a college campus. Any excess electricity is used to top up on-site batteries or charge EV’s.
  • Stored solar power during the early evening until the batteries are depleted (although some storage systems might remain fully charged for emergency backup purposes).
  • Utility power for nighttime electricity needs – until the sun rises and solar PV panels can begin generating energy again.

In many jurisdictions, microgrids can also feed electricity into the utility grid in exchange for credits (under net energy metering programs) or actual cash (under feed-in tariffs). In both cases, these incentives help to shorten the payback periods of microgrids while increasing their ROI’s.

Are Microgrids the Future?

The ability to deliver cheaper, cleaner, and more reliable electricity to local communities is a key selling point of microgrid technology. And despite their initial price tag, they’re actually an affordable option since the investment can be shared across all energy-paying customers within the microgrid – similar to how we already finance community solar projects or even America’s legacy electric grid.

Better still, costs will only continue to fall as the technology improves – making the economics of microgrids even harder to ignore. Already, experts estimate that total installed microgrid capacity throughout the US will grow 115% over the next 5 years to reach 4.3GW.

But while microgrids have very clear benefits for stakeholders, they present the same utility-scale challenges outlined in our previous installment about distributed energy resource management systems (DERMS’s). Grid operators and asset managers struggle to balance supply and demand when integrating distributed energy resources that are behind-the-meter and outside the direct control of traditional gatekeepers.

If a microgrid suddenly disconnects from the electric grid to avoid peak charges, for example, the utility provider doesn’t automatically know to scale back its own production to avoid surpluses and dumping. The same is true if solar PV generation spikes in the microgrid – sending a surge of electricity into the network.

This is a simplified version of the problem – one in which the microgrid itself is a distributed energy resource that only has a single metered connection point with the central utility grid. But in truth, most microgrids are actually a collection of many DER’s with varying levels of production, consumption, and storage – all of which change in real time. To complicate things even more, many microgrids are now networked or nested, which adds additional layers of complexity.

The short-term goal of smart grid design involves equipping microgrids and integrated edge devices with tools to send data and receive instructions. Doing so gives grid operators and asset managers greater control in their effort to balance the grid and optimize electricity delivery.

The long-term goal, however, is to remove human decision-making from the process entirely so that local microgrid control and network-wide optimization both happen automatically.

And this is the promise of artificial intelligence (AI).

AI and Microgrids

Grid operators and asset managers already use DERMS’s to receive data from and relay instructions to edge devices across the electricity network. But even under ideal circumstances, this level of coordination is impossible given the many terabytes of data collected when analyzing weather conditions and grid-wide energy production, consumption, and storage levels – all of which evolve in real-time.

But while human actors struggle to manage this much data, AI was purpose-built for this type of real-time analysis.

At Veritone Energy, for example, our AI software can monitor changing grid conditions to determine the optimal allocation of resources in the network – including grid-connected devices like:

  • Residential PV systems
  • Commercial wind farms
  • Municipal EV-charging stations
  • University microgrids

Moreover, our AI technology also uses machine learning to self-correct and make accurate predictions about how these conditions will change seconds, days, and weeks into the future. These forecast-based models automatically drive edge controllers, removing the need for grid operators and asset managers to intervene. Instead, energy is automatically delivered at the right price, place, and time – without human actors making any decisions or pulling any levers.

To see our AI solutions in action, request a free demo from us today.


Microgrids are following a very similar trajectory already witnessed in the rooftop solar revolution. In both cases:

  • Private investment is helping to boost the adoption of renewable technology, which benefits end-user(s) and the planet as a whole
  • This investment creates problems for the grid operators and asset managers normally tasked with balancing energy supply and demand
  • Energy imbalances eventually extend to the entire community in the form of curtailments, blackouts, higher prices, and unnecessary pollution

The growing popularity of distributed energy resource management systems helps grid operators and asset managers more fully integrate renewables into the larger utility network. And artificial intelligence is increasingly demonstrating its capacity to balance resources in real-time and deliver electricity to users more cheaply and reliably.

But even with a well-designed and fully integrated microgrid, disruptions are inevitable – whether they come from natural disasters or from human error. And no discussion about smart grid design would be complete without grid resilience technology – which is the topic of our next installment.

If you have questions about microgrids or want to see our AI software in action, schedule a free consultation with our experts today.

To learn more about the other key technologies at play in smart grid design, visit the chapters below.

You may also enjoy this brief 30-minute podcast that introduces the challenges of smart grids and highlights some of the benefits of AI to improve energy and utility operations.