AC Coupled vs DC Coupled: Where the Battery Meets the Array

One architecture converts once and one converts three times, and the efficiency argument that follows applies to a smaller share of your energy than it sounds like.

The Question Is Where the Conversions Happen

Solar panels produce direct current. Batteries store and deliver direct
current. A house runs on alternating current. Somewhere between those three
facts, conversions have to happen, and AC coupling and DC coupling are two
different answers about where.

In a DC coupled system, the array and the battery meet on the direct current
side, through a charge controller inside one piece of equipment, and a single
inverter converts to alternating current once. In an AC coupled system, the
array has its own inverter producing alternating current, and the battery has
its own inverter too, so power going from panel to battery is converted from DC
to AC and back to DC again.

What each architecture costs you

The efficiency argument, honestly stated

DC coupling is more efficient at the specific task of moving solar energy
into a battery, because it skips two conversion stages. That is real and it is
the strongest technical argument for it.

It is also frequently overstated, for two reasons. The energy that goes
through the battery is only part of what the array produces — a great deal of
solar output goes straight to the house or to the grid, and on that path the
two architectures are identical. And the loss per conversion stage in modern
equipment is modest, so the difference applies to a fraction of the energy at a
small rate.

Where it does start to matter is a system that cycles the battery hard every
day, which usually means one being used to shift energy against a tariff rather
than one waiting for outages. The size of the prize there depends on the
tariff, which is a separate question covered in
how the value stack compares with net metering and in
how storage and net metering interact.

When AC coupling is simply the right answer

The commonest situation in New York is a house that already has solar and
wants storage added. In that case AC coupling is the straightforward path: the
existing array and its inverter stay exactly as they are, and the battery
arrives with its own inverter and connects on the alternating current side.

The alternative is replacing a working inverter to get DC coupling, and that
is rarely worth the efficiency difference. There are other reasons it can make
sense — an inverter near the end of its life, an array being expanded anyway —
but the arithmetic has to be done rather than assumed. The sequencing for the
retrofit case is in
adding a battery to an existing solar system.

Two other cases favour AC coupling regardless of retrofit:

• Microinverters or module-level electronics on the array. If
  each panel already has its own inverter, the DC side does not exist as a place
  to couple anything.
• A desire to keep the two systems independent. Different
  manufacturers, different upgrade cycles, and a fault in one that does not take
  the other down with it.

When DC coupling earns its place

• New installations designed as one system. If the array and
  the battery are being specified together, the DC coupled route gives fewer
  boxes, one manufacturer and one place to call.
• An array deliberately oversized relative to the inverter.
  On the DC side, production above the inverter’s rating can go to the battery
  instead of being clipped away. On the AC side it is lost at the array inverter
  before the battery ever sees it.
• Daily deep cycling. Where most of the solar energy is
  routed through the battery every day, the conversion saving accumulates.

What neither of them decides

Several things get attributed to this choice that have nothing to do with
it, and separating them keeps a quote honest.

• Whether you have backup power at all. That depends on
  whether the system includes an automatic transfer arrangement and a backup
  panel, not on which side the coupling happens. Both architectures can do it and
  both can be sold without it.
• How much of the house runs in an outage. That is set by
  the battery’s continuous power output and by how the loads are separated, which
  is the whole subject of
  whole-house backup against critical loads.
• Whether the array keeps producing during an outage. This
  one is worth checking explicitly, because it is a common and expensive
  surprise: some arrangements let the array keep charging the battery through a
  multi-day outage and others do not, and it is a property of the specific
  equipment rather than of the coupling.
• The battery chemistry. Independent of both — see
  what the cathode chemistry changes.
• Whether the installation is acceptable where you live.
  Interconnection requirements, permits and inspection depend on the utility and
  on where the property sits, because New York State’s Uniform Code and New York
  City’s own requirements are separate paths on separate adoption cycles. That is
  a question for the authority having jurisdiction and the utility rather than
  for a datasheet.

The grounding detail that catches retrofits

A system capable of running as an island has to establish exactly one bond
between the neutral and the grounding system while it is disconnected from the
grid, and exactly one while it is connected. Getting that wrong is a real
hazard rather than a paperwork issue, and it is more easily got wrong in an AC
coupled retrofit because two separate pieces of equipment are involved and each
may have its own opinion.

The principle is set out in
exactly one bond during island operation. What it means
for a quote is that the installer should be able to say, without looking it up,
which device establishes the bond in each state.

How to use this when comparing proposals

Ask what the architecture is and why, and treat “it is more efficient” as an
incomplete answer. The questions that produce a useful conversation:

• What is the round-trip efficiency of the whole system, measured at the
  meter rather than at the battery terminals? The conversion architecture is one
  input to that, and the honest figure is the one that includes everything —
  which is the framing in
  where the conversions happen.
• What fraction of the array’s annual production do you expect to pass
  through the battery? That is what the efficiency difference applies to.
• Can the array charge the battery during a grid outage, and for how long?
• If I want to add storage capacity in three years, what does this
  architecture allow?

A proposal that answers those four is describing a designed system. One that
leads with the coupling type as a selling point is describing a product.

Frequently Asked Questions

What is the difference between AC coupled and DC coupled?

It is where the array and the battery meet. In a DC coupled system they connect on the direct current side through a charge controller, and a single inverter converts to alternating current once. In an AC coupled system the array has its own inverter and the battery has another, so solar energy going into the battery is converted from DC to AC and back to DC again.

Which one is more efficient?

DC coupled, at the specific task of moving solar energy into the battery, because it skips two conversion stages. The difference is real and usually a single-digit percentage, and it applies only to the energy that actually passes through the battery. Solar going straight to the house or the grid takes the same path in both architectures.

I already have solar. Which should I add?

AC coupling, in almost every case. The existing array and inverter stay as they are and the battery arrives with its own inverter on the alternating current side. Getting DC coupling instead usually means replacing a working inverter, which is rarely justified by the efficiency difference unless that inverter is near the end of its life or the array is being expanded anyway.

Does my array have to be off during an outage?

That is a property of the specific equipment rather than of the coupling type, and it is worth asking explicitly because it is a common and expensive surprise. Some arrangements let the array keep charging the battery through a multi-day outage; others do not, which turns a battery into a single night of runtime no matter how much sun there is.

Does the coupling decide whether I get backup power?

No. Backup depends on whether the system includes an automatic transfer arrangement and a backup panel, and both architectures can be built with or without it. How much of the house runs during an outage is set by the battery’s continuous power output and by how the loads are separated, not by which side the battery couples on.

Why does an oversized array favour DC coupling?

Because production above the inverter’s rating can be routed to the battery on the direct current side instead of being clipped. In an AC coupled system that surplus is lost at the array inverter before the battery ever sees it. This matters where an array is deliberately specified larger than its inverter, which is a reasonable design in a cloudy climate.

Is one easier to expand later?

AC coupling is generally more modular, because another battery with its own inverter can be added alongside. A DC coupled system is bounded by what its single inverter supports, so the expansion path has to be checked at purchase rather than assumed. Ask specifically what adding capacity in a few years would involve.

What is the grounding issue with AC coupled retrofits?

A system that can run as an island must establish exactly one bond between the neutral and the grounding system while disconnected from the grid, and exactly one while connected. That is easier to get wrong when two separate pieces of equipment are involved and each may have its own arrangement. An installer should be able to say which device establishes the bond in each state without looking it up.