The grid goes down across town. But the hospital, the water plant, the grocery store and the fire station stay lit. Why them? Some have generators. Some have batteries. Some have a true microgrid. Those are different things. Here’s how to tell them apart, and which one your building actually needs.

First, kW vs. kWh

Two units come up in every conversation about batteries and microgrids:

kW (kilowatts). How fast power can flow. Think of it as the size of the engine.
kWh (kilowatt-hours). How much energy is stored. Think of it as the fuel tank.

Batteries are sized in both, like “100 kW / 200 kWh.” Microgrids are usually described by kW.

What a battery system is

A battery energy storage system (BESS) is batteries plus an inverter, which usually holds the control “brains.” It also has a battery management system that watches cell voltage and temperature for safety. Larger systems often come in an enclosure. A BESS is a building block. It is not a microgrid on its own.

What a battery does on a normal day

Most batteries earn their keep without any outage at all:

Peak shaving. Short demand spikes, like when compressors or refrigeration kick on, can set your demand charges. The battery covers those spikes so the utility doesn’t see them. This is why your 15-minute interval data matters.
Time-of-use shifting. Charge when power is cheap. Use stored power when it’s expensive.
Backup, if designed in. This one is optional.

A battery does not automatically mean backup. Many batteries only do peak shaving and time-of-use shifting. Backup has to be designed in: which circuits, and for how long.

A simple battery backup usually covers a few critical loads, like a server room, for a short outage. The inverter often acts as the controller. And the switchover may not be seamless. Some systems take seconds.

What makes it a microgrid

In plain English, a microgrid is a small local power system that keeps things running when the grid goes down. It’s a mini version of the grid.

The U.S. Department of Energy defines it as a group of interconnected loads and distributed energy resources within clearly defined electrical boundaries that acts as a single controllable entity. It can connect to and disconnect from the grid, running grid-connected or on its own (“island mode”). That gives you four tests:

A clear boundary.
More than one source and load inside it.
One controller running it as a single system.
The ability to island and reconnect.

California’s definition, in Public Utilities Code section 8370, adds two things. The microgrid must be “appropriately sized to meet customer needs.” And it can maintain electrical supply to connected critical infrastructure.

The boundary is set by the point of common coupling (PCC). That’s the switch where the microgrid disconnects from and reconnects to the utility. Everything on the customer side of it is inside the boundary.

Inside, the sources can include solar, batteries, generators (diesel, propane or natural gas), combined heat and power (cogeneration) and fuel cells. Wind and micro-hydro are rare. The loads are everything that uses power.

The controller is the brain

The controller turns the equipment into one system. It:

Watches the grid’s voltage and frequency so it can island seamlessly. Hospitals and emergency sites can’t tolerate even a second of interruption.
Forms the grid when the utility is down (“grid forming”), and follows the grid when it’s up.
Tracks battery charge and health, and generator fuel.
Sheds non-critical loads to stretch a long outage, like a multi-day public safety power shutoff.

Reconnecting is the hard part. The microgrid must match the grid’s voltage and frequency before it closes the switch. That’s why microgrids are engineered with the utility.

Why your solar goes dark in an outage

Normal solar inverters have anti-islanding protection. In an outage, they shut off so they don’t push power onto lines crews are repairing. A microgrid isolates itself at the PCC first, so it can keep running safely.

Side by side

	Battery (savings only)	Battery backup	Microgrid
Main job	Cut demand charges and energy costs	Savings, plus backup for a few loads	Keep the site running through an outage, plus savings
What’s backed up	Nothing	A few critical circuits	Critical loads, with others shed as needed
Outage length	None	Short	Long, even multi-day, with firm power
Switchover	None	May take seconds	Can be seamless
Controller	Inverter	Usually the inverter	Dedicated microgrid controller
Utility involvement	Utility paperwork	Utility paperwork	Engineered with the utility
Typical timeline	Relatively quick	Relatively quick	Much longer; upgrades can take a year or two

Savings help pay for the backup

A microgrid runs in one of three modes. Grid-connected is a normal day, when it does peak shaving and time-of-use savings. Islanded is during an outage. Off-grid is for sites that are never connected, like remote locations.

That first mode matters for your budget. The same batteries and solar that keep the lights on in an outage cut demand charges and energy costs the rest of the year.

Do you need a generator?

The definition doesn’t require one. But in practice, you want some “firm” power source that doesn’t depend on the sun for long outages. Without it, you have to size solar and storage for worst-case winter, which can cost far more. In California, a generator generally needs an air permit from the local air district. That can delay a project.

Who owns it, and who maintains it

Microgrids come in a few types: a single-owner campus, a multi-customer “community” microgrid that crosses property lines, or an off-grid system. Ownership can sit with the customer, a community energy authority or a third-party developer. A developer may own the system and sell you the power under a power purchase agreement (PPA). In that case, the PPA owner maintains it.

One real example is the Redwood Coast Airport Microgrid in Humboldt County. It’s California’s first 100% renewable, front-of-the-meter, multi-customer microgrid. It pairs about 2.2 MW of solar with a 2 MW battery. The Redwood Coast Energy Authority owns the generation, and PG&E owns the microgrid circuit. It keeps the airport, a U.S. Coast Guard air station and neighboring facilities powered during outages.

Whoever owns it maintains it. Projects are typically quoted with long warranties (often 10-year extended) and an operations and maintenance (O&M) contract. Some manufacturers require certified service providers.

Questions to answer first

What’s the goal? Lower bills, backup or both?
How long must you ride through an outage? Hours, days, or a week or more?
Must the switchover be seamless? Or can your loads handle a few seconds of interruption?
Which loads are truly critical? And which can be shed?
What’s the budget, and how will you pay? Cash, financing or a PPA?
What will your load look like in 3-5 years? Think EV chargers, heat pumps and new buildings.
Who will own and maintain it?

Where to start

There’s no rule-of-thumb savings number. Every site is different. So the work starts with an energy audit and a look at your interval data. Then do efficiency first, like lighting and HVAC, so the battery or microgrid is sized to the lower load. That’s also how utility On-Bill Financing is structured: efficiency first. Learn more in our article on EV charging and On-Bill Financing.

EverWatt’s sweet spot is battery systems and microgrids under about 1 MW for a single customer or campus. We start with the audit and efficiency, define your goals with you, design with the utility and our installation partners, and arrange O&M so the system keeps performing.

Talk to an engineer about what your building needs, or read more about our battery storage and microgrid services.

Sources

This article is a plain-language overview, not engineering or legal advice.

Battery Backup or Microgrid? What Your Building Actually Needs