Home battery storage explained: architecture, sizing, economics

Home battery storage has moved from a niche add-on for off-grid enthusiasts to a mainstream component of residential solar systems. The core idea is simple: capture surplus solar generation during the day and release it in the evening, when household consumption peaks and the sun is gone. The engineering behind that simple idea, however, involves choices about coupling architecture, inverter topology, and sizing that materially affect cost and payback.

This article walks through how a home battery system actually works from the DC bus to the wall socket, compares AC-coupled and DC-coupled architectures, and gives a practical sizing method with a worked example. It closes with an honest look at the economics, because storage remains one of the more expensive ways to shave a household electricity bill.

For readers who have not yet sized a system, the battery sizing calculator is a useful companion to the method described here, and background on how PV modules generate the current in the first place is covered in how a solar cell turns light into current.

How a home battery system works

A residential battery system consists of four functional blocks: the battery pack (cells plus a battery management system, or BMS), a power conversion stage, an energy management controller, and the metering/communication layer that decides when to charge, discharge, or hold.

The BMS is the part most buyers never see but that determines the practical lifetime of the system. It balances individual cell voltages, enforces charge and discharge current limits, tracks state of charge, and disconnects the pack on over-temperature, over-voltage, or short-circuit conditions. Lithium iron phosphate (LFP) packs, now the default chemistry for stationary storage, tolerate a wider state-of-charge window than the nickel-based chemistries used in early residential products, which simplifies BMS design and improves usable capacity.

The energy management controller is the layer that decides what to do at any given moment: charge from excess PV, discharge to cover load, or, on systems with grid-services contracts, discharge to export at a scheduled time. This logic runs either in the inverter's firmware or in a separate home energy management system that also handles EV charging and heat pump scheduling.

AC coupling versus DC coupling

The single biggest architectural decision in a retrofit or new-build system is whether the battery sits on the DC side of the solar inverter or is coupled through its own inverter on the AC side.

In a DC-coupled system, the PV array, the battery, and (usually) a single hybrid inverter share one DC bus. Solar power charges the battery directly through a DC-DC converter, without an extra DC-AC-DC conversion step, and only one inverter is needed to serve both PV export and battery discharge. This gives DC coupling a round-trip efficiency advantage, typically a few percentage points, because energy that goes straight from panel to battery avoids one inversion stage entirely.

In an AC-coupled system, the battery has its own inverter/charger and connects to the household AC bus alongside the existing (or new) solar inverter. Surplus solar power is inverted to AC, then rectified back to DC to charge the battery — two extra conversion steps compared with the DC-coupled path. AC coupling is the natural choice for retrofits onto an existing solar installation, since the original inverter is left untouched and the battery is simply added as a new AC-connected device.

Neither architecture is universally better; the right choice depends heavily on whether PV already exists on site and whether backup power during a grid outage is a requirement.

Backup power and islanding

A battery that only shifts self-consumption does not automatically provide power during a grid outage. To do that, the system needs an automatic transfer switch and an inverter capable of forming its own AC waveform (grid-forming mode) once it is disconnected from the utility. This is a safety requirement, not a convenience feature: IEC 62109-2 and the anti-islanding provisions referenced in grid codes require that inverter-based sources disconnect from a de-energised grid to protect utility line workers, and only a properly designed backup circuit is allowed to re-energise a subset of the home's circuits independently.

Most residential backup configurations only cover a "critical loads" sub-panel — fridge, some lighting, internet, well pump — rather than the whole house, because the battery's continuous power rating (commonly 3-6 kW for a home unit) is well below the instantaneous demand of an oven, electric shower, or air-source heat pump compressor running simultaneously with other loads.

Sizing a home battery

Sizing starts from the evening and overnight load profile, not from the PV array size. The two things that matter are usable energy capacity (kWh) and continuous/peak power rating (kW).

A practical method:

1. Identify the evening-to-morning consumption that PV cannot cover directly (from a smart meter or monitoring portal), typically 18:00-07:00.
2. Subtract loads you are willing to shift or exclude from backup (large appliances run during the day instead).
3. Add a buffer of 10-20% for winter months, when PV contribution before evening is lower and heating loads are higher.
4. Divide by the battery's usable-capacity fraction (LFP packs typically expose 90-100% of nameplate capacity to the user; older chemistries closer to 80%).

Worked example

A household consumes 11 kWh overnight (18:00-07:00) with a monitored profile showing 7 kWh of that is essential (fridge, lighting, router, occasional cooking) and 4 kWh is discretionary (dishwasher, laundry) that could run mid-day instead.

• Essential overnight load: 7 kWh
• Winter buffer at 15%: 7 x 1.15 = 8.05 kWh
• Required usable capacity, rounding to available product sizes: 9.6 kWh (a common modular LFP pack size)
• At a nameplate-to-usable ratio of 97% for LFP, nameplate capacity needed: 9.6 / 0.97 = 9.9 kWh, matching the 9.6-10 kWh product tier

For power, if the peak simultaneous essential load (fridge compressor start plus lighting plus router) is 2.2 kW, an inverter rated for at least 3 kW continuous gives adequate headroom for surge currents. The battery sizing calculator automates this arithmetic across seasonal load profiles.

Depth of discharge and usable capacity

Nameplate capacity is measured at a reference discharge rate and temperature; usable capacity in the field is lower once the BMS reserves headroom at the top and bottom of the state-of-charge range to protect cell longevity. LFP systems commonly allow 90-100% depth of discharge (DoD) within warranty terms, while some NMC-based residential products are limited to 80-90% DoD to manage degradation, a topic covered in depth in battery degradation and cycle life.

Economics: what storage actually saves

Storage income comes from three possible sources, and not every market offers all three: increased self-consumption of already-generated solar (avoiding import at retail price), arbitrage against time-of-use tariffs, and grid services payments (frequency response, capacity markets) where regulation allows aggregation of household batteries.

Battery payback is materially longer than PV-only payback in most European markets today, often 8-14 years against product and inverter warranties of 10 years, because installed cost per kWh remains well above the cost of grid electricity saved per cycle. Households with high self-consumption already achieved by PV alone, high export tariffs, or flat electricity tariffs will see the weakest battery-specific payback. The honest payback calculation article discusses how to avoid overstating solar and storage savings, and the same caution applies when a battery is added.

Grid interaction and export rules

In most jurisdictions, a hybrid inverter with battery still needs to comply with the same grid-connection requirements as PV-only inverters: voltage and frequency ride-through, anti-islanding protection, and power quality limits. Rules for exporting stored energy that was originally imported from the grid ("energy arbitrage export") vary by market and sometimes require separate metering or contractual permission; see grid connection and export rules explained for the underlying framework.

Standards and references

• IEC 62109-1/-2 — safety requirements for power converters used in photovoltaic power systems, including battery inverters.
• IEC 62619 — safety requirements for secondary lithium cells and batteries used in industrial and stationary applications, commonly applied to residential storage packs.
• IEC 62477-1 — safety requirements for power electronic converter systems and equipment, relevant to the DC-DC and DC-AC stages in hybrid inverters.
• IEEE 1547 — interconnection and interoperability requirements for distributed energy resources, including battery storage, with the grid (US context).
• IEC 60364-7-712 — low-voltage electrical installation requirements specific to PV power supply systems, referenced for wiring of combined PV-battery installations.

Sources

• NREL — Energy Storage research
• Fraunhofer ISE — Electrical Energy Storage
• IEA PVPS
• PVEducation — Battery based systems

Frequently asked questions

Can I add a battery to an existing solar system without replacing the inverter?

Yes, in most cases. AC-coupled battery systems have their own inverter/charger and connect alongside the existing PV inverter on the household AC bus, so the original inverter is untouched. This is the standard retrofit path. DC coupling would require replacing the existing inverter with a hybrid unit that has a DC battery input, which is usually only worthwhile if the original inverter is near end of life anyway.

Does a battery work during a power cut?

Only if the system includes backup/islanding hardware — an automatic transfer switch and an inverter capable of forming its own waveform when disconnected from the grid. A grid-tied battery without this hardware will shut down during an outage along with the solar inverter, for the same anti-islanding safety reasons that apply to PV-only systems.

How much of my nameplate battery capacity can I actually use?

It depends on chemistry and BMS settings. LFP packs typically expose 90-100% of nameplate capacity as usable, while older or NMC-based residential products may be limited to 80-90% depth of discharge to manage degradation. Always ask for the usable capacity figure, not just nameplate capacity, when comparing products.

Is DC coupling always more efficient than AC coupling?

DC coupling avoids one or two inversion stages when charging the battery directly from solar, giving a round-trip efficiency advantage of roughly a few percentage points in typical conditions. However, the difference is small relative to other factors like correct sizing and tariff structure, and AC coupling is often the only practical option for retrofits.

How many years does it take for a home battery to pay for itself?

In most European markets today, battery-specific payback runs 8-14 years, longer than PV-only payback, because installed cost per kWh of storage remains high relative to the value of electricity shifted per cycle. Households with strong time-of-use tariff spreads or limited export compensation tend to see faster payback than those with flat tariffs and generous export rates.