GFCI vs AFCI Difference: What Each Device Trips On

One device is watching for a person in the current path and the other is listening for a fire, and the panel treats them very differently.

Two small devices in a New York panel look almost the same, cost more than a
plain breaker, and are constantly confused with each other. One of them is
watching where the current goes. The other is listening to how the current
behaves. They catch different failures, they fail to catch different things,
and swapping one for the other leaves a real gap.

GFCI vs AFCI Difference: Two Faults, Two Devices, One Panel

An ordinary circuit breaker has exactly one job: open the circuit when the
current stays above its rating long enough to overheat the wire. That covers
overloads and dead shorts. It does not cover a current leaking out of the
circuit through a person, and it does not cover a loose connection quietly
arcing behind a wall at a current the breaker considers perfectly normal. Those
two blind spots are what the two devices exist to close.

What a ground-fault circuit interrupter measures

On a healthy circuit, every amp that leaves on the hot conductor comes back
on the neutral. A GFCI puts both conductors through a sensing coil and compares
them. If they no longer match, the missing current is going somewhere else:
through damp masonry, through a punctured cable jacket, through a failing motor
winding, or through a person standing on a wet floor. The device does not care
where. It only cares that the two numbers stopped agreeing, and it opens the
circuit in a fraction of a second.

The important part is the size of the imbalance it reacts to. People
protection devices are built to trip on a leakage current in the single-digit
milliamp range, which is far below anything a breaker would notice and far
below the current that runs a light bulb. That threshold was chosen around what
a human body can survive without losing muscular control, not around what
damages equipment. It is why a GFCI can trip while everything on the circuit
still appears to be working.

What it does not do is protect the wire. A ground-fault device that also
carries a breaker rating is doing two separate jobs in one housing.

What an arc-fault circuit interrupter measures

Arcing is current jumping a gap instead of crossing a connection. It happens
at a backed-out screw terminal, at a conductor nicked by a drywall screw, at a
cracked cord where the copper has broken but the ends still touch. The
dangerous case is the series arc: because the arc itself adds resistance, the
circuit draws less current than normal, not more. A breaker
sized for the wire will never see it. The connection can run hot enough to char
a box for a long time while the panel reports nothing wrong.

An AFCI is a small signal processor. It watches the current waveform for the
erratic, high-frequency signature that arcing produces and tries to separate it
from the arcing that is supposed to happen — a switch closing, a brush motor
running, a drill starting. Its target is fire, not shock. That is the whole
difference in one sentence: a ground-fault device is protecting the person, an
arc-fault device is protecting the building.

GFCI vs GFI: the same device with two names

There is nothing to choose here. GFI is the older informal shorthand and
GFCI is the full name of the same class of device; electricians, inspectors and
catalogues use both. If a quote lists a GFI receptacle and the panel schedule
says GFCI, they are describing one thing. The abbreviation that actually
changes the product is AFCI, and the combination units usually sold as
dual-function devices, which contain both sensing circuits.

Side by side

Where each one turns up in electrification work

Most New York homeowners meet these devices while adding load, not while
fixing something. A car charger, an induction range, a heat pump or a dryer all
mean a new circuit, and a new circuit is where the question gets asked.

Car charging is the case that generates the most confusion. Charging
equipment generally contains its own ground-fault sensing, set for the charging
application rather than for a wet bathroom. Whether an additional device is
also called for on that circuit, and of which type, depends on the rules in
force where the permit is pulled — and two ground-fault sensors in series on
one run are a well known source of trips that have nothing to do with a fault.
The wiring side of that job is covered separately in
what a dedicated circuit for an EV charger involves.

Kitchen and laundry circuits raise the same question from the other
direction: appliance circuits have long been a focus for both kinds of
protection, which is part of why
the electrical work an induction stove needs
is rarely just a matter of pulling a bigger cable, and why
circuit sharing devices on a dryer circuit
have to be evaluated as part of the circuit rather than as an accessory to it.

The code question nobody can answer for you in an article

Which circuits require which protection is written into the electrical
standard, and the edition of that standard that governs your job is a local
question with two separate answers in this state. Most of New York works under
the state Uniform Code, which adopts an edition of the electrical standard on
its own schedule; New York City runs a separate Electrical Code with its own
amendment cycle. The two paths do not move together, so no article can tell you
that a particular installation in a particular town is compliant. What it can
tell you is who to ask — the building department that will issue the permit,
before the work is designed rather than after it is installed. The inspection
itself is a short, mechanical event when the paperwork was right from the
start, which is the pattern described in
what happens at an electrical inspection.

Breaker, receptacle, or inside the equipment

The same protection can sit in three places, and the choice changes daily
life more than it changes safety. In the panel, a protective breaker covers
everything on the circuit, and a trip means a trip to the panel. At the head of
the circuit, a protective receptacle covers itself and everything downstream of
it, and a trip means pressing a button in the room. Inside the equipment,
protection travels with the appliance and does not depend on the wiring at
all.

If the panel itself is being replaced, the circuits landing in the new one are
evaluated against the rules that apply to the new work, not the rules that
applied when the house was wired — the usual reason a straightforward swap
turns into
a larger panel upgrade conversation,
and one of the things worth settling before you plan around
what a smart panel can and cannot shed
during an outage.

Trips that are not faults, and trips that are

Both device types have a reputation for tripping when nothing is wrong. Some
of that reputation is earned by the installation rather than the device.

• Long runs and shared neutrals. Every cable has a small
  amount of natural leakage to ground. Enough of it on one protected circuit and
  the device is asked to distinguish a real fault from the sum of many harmless
  ones.
• Moisture. An outdoor device that trips during rain and
  resets when things dry out is not malfunctioning. It is reporting a real
  leakage path — usually a box, a gasket or a cord end letting water in.
• Motors and older electronics. Some equipment produces
  electrical noise an arc-fault device has to work to classify. A trip that
  happens only when one appliance starts is a clue pointing at that
  appliance.
• Two sensors in series. Protective equipment behind another
  protective device can trip together in ways neither would alone.

The useful diagnostic is simple. Unplug everything on the circuit and reset.
If it holds, the problem travels with something you plug in. If it will not
reset with nothing connected, the problem is in the wiring and belongs to an
electrician.

The sensing electronics in both device types can also fail while the device
continues to pass current, so a circuit can look protected and not be. That is
what the test button is for: pressing it makes the device demonstrate that it
can still open the circuit. It costs nothing and takes a few seconds.

What to ask before the work starts

• Which new circuits are getting ground-fault protection, which are getting
  arc-fault protection, and which are getting both?
• Is the protection at the breaker or at the first receptacle, and where will
  I go to reset it?
• Does the equipment already contain its own ground-fault sensing, and does
  that change what the circuit needs?
• Which building department issues the permit for this job, and which edition
  of the electrical standard applies there?
• Does the panel physically accept the breaker type this design requires, or
  does the panel have to change first?

Both devices are cheap relative to the work of running the circuit, and both
are far easier to specify at design time than to retrofit after the walls are
closed. Knowing which one is looking for a person and which one is looking for
a fire is enough to have that conversation properly.

Frequently Asked Questions

Is a GFI the same thing as a GFCI?

Yes. GFI is older informal shorthand for the same ground-fault device that the standards call a GFCI. If a quote says GFI and the panel schedule says GFCI, they refer to one product. The abbreviation that describes a genuinely different device is AFCI, the arc-fault interrupter.

Can one device provide both kinds of protection?

Yes. Dual-function devices contain both sensing circuits in one breaker or one receptacle, and are common where a circuit needs both. They cost more than either single-function device and take the same panel space as a standard breaker, so the practical question is usually whether the panel accepts that breaker family at all.

Will a GFCI protect me from an overloaded circuit?

No. Ground-fault sensing compares current out against current back and has no opinion about how much current is flowing. Overload and short-circuit protection is the ordinary breaker’s job. A GFCI breaker does both because it is two devices in one housing; a GFCI receptacle does not.

My car charger trips its breaker. Does that mean the charger is faulty?

Not necessarily. Charging equipment usually contains its own ground-fault sensing, and a second ground-fault device upstream can react to the same event, so repeated trips on that circuit are as often a design question as an equipment fault. It is worth establishing what protection is on that circuit and what is inside the unit before replacing anything.

Do I need arc-fault protection when I add a circuit to an older New York home?

That depends on the circuit and on the edition of the electrical standard in force where the permit is issued. Most of the state works under the Uniform Code and New York City runs a separate Electrical Code on its own cycle, so the requirement is a local one. Ask the building department that will issue the permit before the circuit is designed.

How do I find out what protection a circuit already has?

Look in two places. In the panel, protective breakers carry a test button that ordinary breakers do not have. In the rooms, a receptacle with test and reset buttons is protecting itself and usually everything downstream of it, so an outlet with no buttons may still be protected by one you have not found yet.

Do these devices wear out?

The sensing electronics can fail while the device still passes current, so a circuit can appear protected when it is not. Pressing the test button monthly makes the device prove it can still open the circuit. One that will not trip on its own test button, or will not reset afterwards, needs replacing.