Two homes can have similar solar panels and batteries yet behave very differently when the utility fails. One goes dark. Another keeps a few essential circuits running. At a remote property with no utility connection, the same event is irrelevant because the site already operates on its own electrical network.
The difference is not simply battery size. On-grid, off-grid, and hybrid describe different relationships with the utility grid. To understand them, ask three questions: Is the grid present during normal operation? What establishes voltage and frequency when it is absent? Which energy sources can keep the loads supplied?
On-grid storage works alongside the utility
An on-grid, or grid-connected, system operates in parallel with the public electricity network. The grid provides an established AC voltage and frequency for grid-following equipment to match. It also covers the difference between local supply and demand: a site can import electricity when solar and battery output are insufficient and may export surplus energy where local rules allow.
An on-grid battery may store surplus solar, shift consumption to another time, or respond to an operating schedule. “On-grid” therefore does not mean “solar without a battery.” It means that the storage system is designed to interact with an available utility network during normal operation.
That connection does not automatically provide backup. The U.S. Department of Energy explains that traditional grid-following inverters use the grid as their waveform reference and disconnect when disturbances move sufficiently outside their permitted range. UL’s application guide for inverter systems likewise distinguishes utility-interactive equipment from stand-alone and multimode equipment, and describes anti-islanding protection that prevents continued export after a utility outage.
As a result, a charged battery may be physically present while the building still loses power. Continuing to supply local loads requires a system designed to separate from the failed grid and operate as an intentional island. Battery capacity answers how much energy is available; it does not prove that the installation can use that energy during an outage.
Off-grid operation requires a local electrical reference
An off-grid system has no public grid to lean on. Where the site supplies AC loads, equipment within the system must establish the local voltage and frequency while generation, storage, and demand remain balanced from moment to moment. In an inverter-based system, that role is normally performed by grid-forming or stand-alone-capable equipment rather than a conventional grid-following inverter.
The battery and inverter together help manage both short and long imbalances. They can respond when a motor starts or a cloud reduces PV output, then carry loads after sunset. They are still only part of the supply system. Solar, wind, hydro, or a fuel-powered generator may contribute energy, and the control system has to coordinate charging, discharging, generation, reserve levels, and—where the design permits it—non-essential load shedding.
This is why off-grid design is more demanding than adding a large battery. The design must cover the highest simultaneous demand among the loads it is intended to serve and carry those loads through expected gaps in renewable generation, using stored energy and any dispatchable local source. Seasonal generation, weather variability, load growth, equipment losses, and reserve margins all affect the result.
An external generator does not make a system “on-grid.” It remains off-grid as long as it is not connected to the public utility. The IEA Photovoltaic Power Systems Programme documents off-grid systems that combine PV, batteries, and generator sets, while SMA’s stand-alone system documentation provides a concrete example of a battery inverter forming the local grid and coordinating batteries, generators, and loads. In this context, a generator is another local energy source, often retained for periods when renewable generation and stored energy are insufficient.
“Hybrid” describes capability, not a guaranteed outcome
For this comparison, hybrid means a grid-connected storage system intended to support more than one operating mode. With the utility available, it can import, export, charge, and discharge according to site objectives and local rules. When an outage is detected, a backup-capable configuration can isolate protected circuits and establish a local supply. The DOE’s resilience guidance describes solar-plus-storage systems switching into an islanded mode after loss of grid power and reconnecting when service returns.
The word “hybrid,” however, is used loosely. The IEA PVPS report, for example, also uses “PV hybrid system” for off-grid combinations of PV, batteries, and diesel generation. In practice, the label leaves several questions unanswered: which circuits remain powered, how quickly transfer occurs, and whether the inverter can start and support an island without the utility.
Those functions must be verified at system level. A working backup installation needs appropriate power-conversion capability, isolation or transfer equipment, compatible batteries and controls, and a defined backup-load boundary. Some systems support only a dedicated essential-load output; others can be designed for a larger part of the building. Available battery energy affects runtime, while inverter power and the backed-up circuit design determine which loads can run together.
Seen this way, the three categories describe operating arrangements rather than product grades. A grid-connected system may be enough where utility service is reliable and outage supply is unnecessary. A site with no utility needs a self-contained power network. A site expected to use the grid normally and keep selected loads running during outages needs a verified multimode configuration. Its islanding, transfer, load-support, and control capabilities are more informative than the category printed on the inverter.
