Electrical Load Calculation Basics (NEC Article 220 Overview)
Before the permit application goes in, before the panel schedule gets drawn up, an electrician doing a load calculation sits down with a set of plans and works through a question that has nothing to do with any single circuit: how much total power does this building actually need its service sized for? The answer is never simply "add up every device that could ever be plugged in." It's a smaller number than that, arrived at through a specific, tiered method the National Electrical Code lays out in Article 220, and understanding why the number shrinks is the whole point of a load calculation.
Two figures do the real work in that process, and confusing them is one of the more common mistakes on a first-pass calculation. Connected load is the straightforward sum: every lighting circuit, every dedicated appliance circuit, every piece of equipment on the drawings, added up in volt-amps as if all of it could run at once. Demand load is what a service actually gets sized against, the connected load after a set of demand factors have been applied to account for the fact that a building's lighting, receptacles and equipment don't all draw their full rated load simultaneously. A service sized to the connected load would be safe but badly oversized; a service sized to the demand load reflects how the building is actually used.
In practice, connected load is built from two different kinds of numbers. Part of it is a general load calculated from floor area, lighting and general-use receptacles, sized off a unit load in volt-amps per square foot that depends on the occupancy. The rest is a set of specifically calculated loads for named equipment that doesn't scale with floor area at all: dedicated circuits for fixed appliances, motors, HVAC equipment, kitchen equipment and anything else that gets its own circuit on the panel schedule. A load calculation adds both pieces together before any demand factor gets applied.
The "1,500 VA circuits" input used in the calculator and worked example below isn't an arbitrary round number: 1,500 VA is a figure commonly associated with small-appliance and laundry branch circuits in dwelling-style load calculations, used here as a stand-in unit for any dedicated circuit whose VA rating isn't already folded into the general area-based load. A commercial worked example substitutes its own dedicated-equipment VA ratings into that same slot; the mechanic doesn't change, only the number being fed into it.
Why the total gets smaller, not bigger
The reasoning behind a demand factor is simple even though the tables that encode it aren't: lighting circuits get switched off, most receptacles sit unused most of the day, and the odds that every dedicated appliance circuit in a building draws its full nameplate load at the exact same instant drop fast as a building gets bigger. NEC Article 220 turns that real-world behavior into a set of demand factors, fixed percentages applied to different portions of the connected load, so a calculated service size tracks how a building is actually operated rather than a worst-case scenario that will almost never occur.
The calculator below runs the same math this page walks through by hand. A floor area and a general unit load in VA per square foot produce the general connected VA; a count of dedicated 1,500 VA-equivalent circuits adds the appliance VA; and the total gets run through the tiered demand factor to produce a demand load in kW. Adjust any of the three inputs to see how the demand figure moves. It is offered here as a supporting tool for working through the same mechanic interactively, not as a replacement for it.
- General VA
- 7,500
- Circuit VA
- 6,000
- Connected VA
- 13,500
- Demand load
- 6.68 kW
The specific demand-factor percentages, and the general unit load that gets multiplied by floor area in the first place, are what the rest of this page unpacks.
The tiered demand-factor structure
The demand-factor tiers commonly used to teach this mechanic split the connected general load into three bands, each carrying a different multiplier: the first 3,000 VA counts in full, the next 117,000 VA (the range from 3,001 VA up to 120,000 VA) is derated to 35%, and anything above 120,000 VA is derated further, to 25%. The effect compounds as a building gets larger: a small space's demand load sits close to its connected load, while a large building's demand load can end up a fraction of what's actually connected, because so much of its VA total falls into the 35% and 25% bands.
| Portion of connected VA | Demand factor | What it means |
|---|---|---|
| First 3,000 VA | 100% | Every volt-amp in this band counts toward demand in full. |
| Next 117,000 VA (3,001 to 120,000 VA) | 35% | Only just over a third of this band is added to the demand load. |
| Remainder above 120,000 VA | 25% | Only a quarter of anything beyond that is added. |
This is the tiered structure most commonly used to teach the demand-factor mechanic, not a claim that every occupancy and every code edition applies these exact percentages unmodified.
This general tiered mechanic isn't the only demand factor at work in a real calculation, either. Named load categories (kitchen equipment in a commercial building, laundry circuits in a dwelling, certain motor and HVAC loads) commonly carry their own separate demand factors in Article 220, distinct from the general lighting and receptacle tiers described above, and those factors have shifted between code cycles as usage patterns changed. That layering is exactly why a real load calculation works from the current adopted code edition for a given jurisdiction rather than a single memorized rule of thumb.
The VA-per-square-foot figure that produces the general load in the first place is not one number that applies everywhere, either. NEC Table 220.12 sets a different general unit load depending on occupancy: a warehouse, an office, a school and a retail space are not assigned the same figure, so the 3 VA/sq ft used in the worked example below is illustrative of the mechanic only, not a value to carry over to a different occupancy without checking the applicable table. Treat any unit-load figure on this page as a stand-in for the number a real load calculation would pull from the code for that specific building type.
Worked example: a small retail space
Here's the same mechanic worked by hand, for a 2,500 sq ft small commercial retail space using an illustrative general unit load of 3 VA/sq ft, plus four dedicated 1,500 VA-equivalent circuits for fixed equipment.
As with the calculator above, this is general reference arithmetic for a single illustrative occupancy and unit load. A real load calculation applies the specific NEC Table 220.12 unit load for the actual occupancy and the full set of applicable Article 220 provisions, and it's performed and stamped by a licensed electrician before it goes into a permit application.
The 6.675 kW demand load isn't the end of the calculation; it's the number that everything downstream gets sized against. Converting it to amps at the service voltage gives the current the service conductors and main breaker have to carry, and that current is what actually drives conductor size, breaker rating and equipment ratings on the drawings. Skipping straight from a connected load to a breaker size, without running the demand factor first, is a common way to land on an oversized and unnecessarily expensive service.
Step by step
2,500 sq ft retail space, 3 VA/sq ft general unit load, four dedicated 1,500 VA circuits, 240 V single-phase, power factor 1.0.
- 1
Calculate the general connected VA
2,500 × 3 = 7,500 VA
- 2
Add dedicated circuit VA and total the connected load
4 × 1,500 = 6,000 VA → 7,500 + 6,000 = 13,500 VA
- 3
Apply the tiered demand factor
3,000 + (13,500 − 3,000) × 0.35 = 6,675 VA (6.675 kW)
- 4
Convert the demand load to amps at 240 V
6,675 / 240 = 27.8125 A
General reference arithmetic only. A real service or feeder is sized from the full applicable code provisions for the actual occupancy, confirmed by a licensed electrician.
Connected load vs. demand load at a few sizes
NEC Article 220 also offers more than one route to a demand load for some occupancies, a standard method that walks through the general and named loads and applies tiered demand factors the way this page does, and, for dwellings, an optional method that reaches a comparable result through a different, simplified set of steps. This page works through the standard-method mechanic because it's the one that generalizes most directly across occupancies; a real calculation picks whichever method the applicable code section permits for that specific building.
The same tiered demand factor applied to a range of connected-VA totals, showing how the gap between connected and demand load widens as the total grows.
| Connected VA | Demand VA | Demand load (kW) |
|---|---|---|
| 5,000 VA | 3,700 VA | 3.70 kW |
| 13,500 VA | 6,675 VA | 6.67 kW |
| 50,000 VA | 19,450 VA | 19.45 kW |
| 150,000 VA | 51,450 VA | 51.45 kW |
Illustrative figures from the tiered demand-factor mechanic above, not measured or code-certified values for any real building.
The pattern across every row is the same one the worked example showed: connected load answers "what's installed," demand load answers "what does the service actually have to carry," and the distance between those two answers is exactly what the tiered demand factor is measuring. Getting comfortable with that distinction is most of what makes an NEC Article 220 load calculation legible before pulling in the specific tables a real project needs.
Load figures in practice
Questions
Load calculation FAQ
Questions that come up once connected load and demand load enter the conversation.
Why is a demand load almost always lower than the connected load?
Because a demand load accounts for diversity: the reality that lighting, receptacles and equipment across a building don’t all draw their full rated load at the same instant, while a connected load simply adds up everything that could draw power. The tiered demand factors compress that difference into a single number: the first 3,000 VA of general load counts in full, but everything above that is derated, so the gap between connected and demand load widens as a building’s connected total grows.
Do dwellings and commercial buildings use the same demand-factor tiers?
Not in general. NEC Article 220 sets out separate general unit loads by occupancy in Table 220.12 and separate demand-factor provisions in Table 220.42 and elsewhere in the article, rather than one universal set of numbers for every building type. The 100%/35%/25% tiered structure this page uses to teach the mechanic is one commonly cited example, not a figure that applies unmodified to every occupancy. Confirm the specific table and demand factors that apply to a given building type against the applicable code edition.
Why does a service get sized to the demand load instead of the connected load?
Sizing to the full connected load would mean paying for service equipment, conductors and transformer capacity that will almost never be used, since diversity means a building’s real peak draw sits well below the sum of everything installed. Demand load is the code’s way of estimating that realistic peak, so the service, feeder or panel gets sized to what the building is actually expected to draw rather than to a theoretical maximum that isn’t a practical design target.
What happens if a building’s actual usage exceeds what the demand calculation assumed?
A demand calculation is an estimate built on standardized assumptions about how a building’s loads behave, not a guarantee about how the building will actually be used once it’s occupied: added equipment, a change in occupancy, or genuinely simultaneous heavy use can push real draw above what was calculated. That’s why an electrical system carries margin at multiple points, and why a significant change in a building’s equipment or use is a reason to have a licensed electrician re-run the load calculation rather than assume the original numbers still hold.
What’s the difference between a feeder calculation and a service calculation?
A service calculation sizes the conductors and equipment bringing power in from the utility to the whole building, based on the building’s total demand load, while a feeder calculation sizes a subsidiary circuit, from a main panel out to a sub-panel or a separate structure, based on just the demand load of what that feeder serves. The same connected-load-to-demand-load mechanic underlies both; they’re just applied at a different point in the distribution system and against a different scope of loads.
Why do the demand-factor tiers apply to portions of the load rather than the whole total?
Applying a single percentage to the entire connected load would either overstate a small building’s real demand or understate a large one’s, so the tiered structure derates larger portions of the load more aggressively as the total grows. That’s also why doubling a building’s connected load doesn’t double its demand load: a bigger share of the increase lands in the 35% or 25% tiers, so the demand load grows more slowly than the connected load does.
Can this kind of calculation replace an electrician’s load study for a permit application?
No. This page and its calculator walk through the demand-factor mechanic as general reference so the concept makes sense before a real project starts, but a permit-ready load calculation has to use the specific occupancy’s unit loads, the full applicable code provisions, and a licensed electrician’s judgment about the actual building. Treat any number produced here as an illustration of the method, not a figure to submit on an actual application.