Fire Alarm Voltage Drop Calculation: The Real Numbers (and Why Most Submittals Show Zero)

Here's something I noticed after going through a lot of completed submittal packages: the voltage drop section almost always shows all wire lengths as 0 feet.

Volts at end-of-line = source voltage. Drop = 0%. Every time.

And the packages got approved. So what's actually going on here?

The honest answer is that most AHJs accept a structurally present but numerically zeroed VD section on panel swaps and straightforward retrofit jobs, because they're reviewing whether you know what you're doing, not running your math through a simulator. The calculation structure is what signals competence.

But that's not the whole story. There are jobs where a real voltage drop calculation is required, and where zeroing the wire lengths will get you a correction letter. And there's a real calculation to do. So this article covers both: the actual math, and the honest reality of how it gets used in the field.

What the calculation is actually doing

A notification appliance circuit (NAC) is a loop of wire connecting your panel to horn/strobes, bells, or speakers. The panel outputs a nominal 24V DC (sometimes 12V, but most commercial panels are 24V). The appliances at the far end of the wire run need a minimum operating voltage, typically 16V DC for most System Sensor and Wheelock appliances, per their cut sheets.

The wire has resistance. Resistance times current equals voltage drop. If the drop is large enough, the appliances at the end of the circuit get less than their minimum operating voltage and may not activate reliably or at rated output.

The calculation:

Wire resistance per foot: from AWG tables
  (18 AWG: ~0.0064 Ω/ft one-way, or 0.0128 Ω/ft round-trip)
  (14 AWG: ~0.0025 Ω/ft one-way, or 0.0050 Ω/ft round-trip)
  (12 AWG: ~0.0016 Ω/ft one-way, or 0.0032 Ω/ft round-trip)

Total circuit resistance: round-trip resistance/ft × wire run length in feet

Voltage drop: Total circuit resistance × alarm current in amps

Voltage at end-of-line (EOL): Source voltage − Voltage drop

If the voltage at EOL is above the minimum operating voltage for your appliances, you pass.

A note on "round-trip": current has to flow out to the appliances and back, so the wire length the current travels is twice the physical run length. That's why I use a round-trip resistance per foot (double the one-way value) and multiply by the one-way run length. Same result either way, just don't double-count.

A worked example

Let's say you have one NAC circuit with these characteristics:

• Source voltage: 24V DC (from the panel's NAC output spec)
• Wire: 14 AWG, 200-foot run (one way)
• Appliances: 8 × System Sensor P2RL, alarm current 0.111 A each, minimum operating voltage 16V DC
• Total alarm current on circuit: 8 × 0.111 = 0.888 A

Resistance calculation:

Round-trip resistance = 0.0050 Ω/ft × 200 ft run = 1.00 Ω
   (0.0050 already accounts for out-and-back per foot of run)

Voltage drop = 1.00 Ω × 0.888 A = 0.89V

Voltage at EOL = 24V − 0.89V = 23.11V

23.11V is well above 16V. That circuit is fine, with a ton of headroom.

Now push it harder. Same wire, same devices, but a 1,200-foot run:

Round-trip resistance = 0.0050 Ω/ft × 1,200 ft run = 6.00 Ω

Voltage drop = 6.00 Ω × 0.888 A = 5.33V

Voltage at EOL = 24V − 5.33V = 18.67V

18.67V still clears the 16V minimum, but the margin is getting thin. Push it to a 2,000-foot run and you cross the line:

Round-trip resistance = 0.0050 Ω/ft × 2,000 ft run = 10.00 Ω

Voltage drop = 10.00 Ω × 0.888 A = 8.88V

Voltage at EOL = 24V − 8.88V = 15.12V

15.12V is below the 16V minimum. That circuit fails the VD check. You'd need to either shorten the run, increase wire gauge, reduce the number of appliances per circuit, or add a NAC power extender.

Why most submittals show 0 ft

On a standard panel swap in a commercial building, which is probably the most common job type for small shops, the wire is already in the walls. You're replacing a panel that failed or aged out, reusing the existing field wiring and devices.

In that situation, do you actually measure 800 feet of existing conduit runs to fill in the VD calc? Most contractors don't. The existing system presumably worked before the panel died, the wire gauge is documented on the old drawings (or you confirm it's at least 14 AWG), and the runs are probably not extreme. The AHJ knows this. They're reviewing the package to confirm you're using the right calculation method and that the system is fundamentally sound, not because they're going to dispatch an inspector with a tape measure to verify your wire length.

So the convention that developed is: fill out the VD section with 0 ft wire lengths (or omit it on jurisdictions that don't require it), the math produces 0% drop, and the AHJ accepts the structural compliance.

This is not great engineering, strictly speaking. But it is the actual practice, based on what I've seen across a lot of real approved packages.

When you need real numbers

There are situations where you actually have to do the point-to-point calculation with real wire lengths:

New installs with PE-sealed drawings. If your AHJ requires an engineer of record (Collier County in Florida, for example, is stricter than most), the PE will run the numbers. The permitted drawing set will include real circuit lengths from the riser diagram. The VD calc is tied to those drawings.

Long NAC runs. If you have reason to believe a circuit might be marginal, like a large campus, a long single-ended run, or a low-voltage panel (12V) with a lot of appliances, do the real calc. 12V systems have half the headroom before you hit the minimum operating voltage floor.

AHJs that ask for it explicitly. Some fire marshals will specifically ask for "wire lengths on the voltage drop calc." If you get a plan review comment asking for this, you fill it in. For those jurisdictions, leaving 0 ft in is a correction letter.

Speaker circuits (EVAC systems). Distributed audio systems have line loss calculations that are more rigorous than a simple DC drop calc, and audio system engineers typically treat them more seriously. This article is focused on DC NAC circuits for horn/strobes. PA/EVAC is its own category.

The 16V floor and where it comes from

Most System Sensor horn/strobes (L-Series: P2RL, SRL, P2RK, SGRL, and so on) have a minimum operating voltage of 16V DC, per their UL listing and cut sheets. That's the number you're protecting. If your EOL voltage stays above 16V, you're in range. Some older Wheelock appliances and some speaker drivers have different floors, so always check the cut sheet for the specific device.

The panel's NAC output voltage is nominally 24V under full-load alarm (check the panel spec sheet for "NAC output voltage at rated load," which can be slightly below 24V depending on the panel). Some installers use 22V or 23V as the source voltage to be conservative. Others use 24V. I've seen both accepted by AHJs.

The actual wire resistance numbers (AWG table)

These are approximate values for copper conductor at roughly 75°C. The NEC tables (Chapter 9, Table 8) are the authoritative source. Resistance increases with temperature, so a conservative approach is to use the 75°C values rather than the 20°C values.

What the Potter, Fire-Lite, and Silent Knight forms do with this

The panel maker's VD form is a spreadsheet that has columns for circuit ID, wire gauge, wire length, number of appliances, appliance current, calculated resistance, voltage drop, and EOL voltage. On the panels I work with most (Potter PFC-4064, Fire-Lite MS-series, Silent Knight 5208/6808), the VD section is integrated into the same 2 to 3 page document as the battery calc.

When wire lengths are 0, the form spits out EOL voltage equal to source voltage and drop equal to 0%. It's structurally complete. On the jobs where I've had to do real VD, mostly new installs with engineered drawings, I fill in the actual circuit lengths from the riser and let the formula run.

If you want the battery side of that same form, I broke it down step by step in the NFPA 72 battery calculation walkthrough.

What this looks like in a submitted package

A typical approved Sarasota County panel-swap package in my experience has a battery calc page and a VD page. The VD page shows all wire runs as 0 ft and an EOL voltage equal to 24V. The AHJ accepts it.

A typical approved Collier County new-install package (a more demanding jurisdiction) has real circuit lengths on the VD section tied to the riser diagram, reviewed by a PE. The numbers are real.

Both get approved. The difference is the jurisdiction and the job type.

The tool I built to handle this

Full disclosure: I'm the person behind FireDeck, which handles the battery calc and VD calc automatically. For the 0-ft default case (which is most panel swaps), it fills in the VD section structurally, so the form is present, the math produces the right structure, and you get an AHJ-ready document without keying anything in for the VD portion. For jobs where you need real wire lengths, that input is on the roadmap.

First month free. I'm not trying to hard-sell it. The goal of this article is to explain the calculation so you know what you're doing regardless of what tool you use.

If you have questions about a specific job situation, reach out.