Electrical Load Calculation Worksheet: Build and Check It

The load calculation worksheet is the most consequential single page in a set of electrical drawings, and the page most owners never look at. Everything downstream depends on it: transformer size, switchgear lead time, electrical room area, whether the building can accept chargers without a service upgrade. Wrong high, and the owner pays for capacity nobody uses. Wrong low, and the fix arrives during construction, after the utility has set a transformer.

Budlong’s guide to electrical load calculation for commercial buildings walks the six step method end to end. This is its practical companion: how the worksheet is built line by line, and then the half rarely written down, which is how a reviewer reads someone else’s and finds what is wrong with it. Budlong provides electrical engineering services across California, and the wider range of that work sits on our expertise overview.

1. What Is an Electrical Load Calculation Worksheet?

The worksheet, also called the load schedule or service load summary, has four columns that matter and a fifth that is usually missing. The four are load category, connected load, demand factor and demand load. The fifth is the basis: whether the connected load came from a fixture schedule, an equipment schedule, a nameplate, a code minimum or an allowance.

Its legal purpose is narrow: service and feeder conductors must be sized to carry the load calculated under NEC Article 220. Its commercial purpose is wider, because the same number drives the service application, transformer procurement, switchgear duty and the electrical room area the architect has to find.

Three numbers that are not the same

Connected load is the arithmetic sum of everything that could draw power. It is the largest of the three and almost never the right basis for sizing anything.

Calculated demand load is what remains after the code’s demand factors are applied. It is a legal construct rather than a physical prediction, and deliberately conservative.

Measured maximum demand is what an existing building actually drew, recorded as the highest fifteen minute average. For an occupied building it usually sits well below the calculated demand, sometimes by half. NEC 220.87 permits it to be used, which makes it one of the code’s most valuable tools for renovation work.

2. How the Worksheet Actually Works

Every load calculation follows the same five movements. The complexity is in the detail of each category, not the shape of the process.

Inventory the connected load by category

Loads are grouped because the code assigns demand factors by group, not by device: general lighting, exterior lighting, general purpose receptacles, dedicated equipment, kitchen equipment, motors, heating, cooling, elevators, charging equipment, and any process load.

Separate continuous from non-continuous

A continuous load runs at maximum current for three hours or more. Conductors and overcurrent devices are sized at not less than 125 percent of the continuous load plus 100 percent of the non-continuous load. This is where a worksheet quietly gains or loses fifteen percent.

Apply the demand factor for each category

This step needs the most judgement, because each category has its own rule and the rules are not interchangeable. Lighting demand factors apply only to certain occupancies. Non-dwelling receptacles have their own table. Kitchen equipment has a table keyed to appliance count, elevators one keyed to car count. Nothing else gets a reduction at all.

Apply the special rules

Two rules sit outside the category structure. The largest motor rule adds 25 percent of the largest motor’s full load current. The noncoincident rule in NEC 220.60 counts only the largest of a group of loads that will never run together.

Convert and select

For three phase, amperes equal volt-amperes divided by line voltage multiplied by 1.732. A 239 kVA demand at 480 volts is roughly 288 amperes, which points at a 400 ampere service and a 300 kVA transformer. The gap to that standard size is spare capacity, whether or not anyone named it.

3. The NEC Article 220 Framework

Article 220 of the National Electrical Code, published by the National Fire Protection Association, is organised into parts, and knowing which part you are in prevents most of the confusion. Part I sets general rules including how floor area is determined. Part II covers branch circuit loads, where per outlet and per square foot unit loads live. Part III is the standard method for feeder and service calculations, which is what a service worksheet is built on. Part IV holds the optional methods.

The 2023 renumbering catches people out

The 2023 NEC reorganised Article 220. The non-dwelling lighting provisions moved from 220.12 to 220.42, with the unit load table becoming Table 220.42(A). General lighting demand factors became Table 220.45. Receptacle provisions moved from 220.44 to 220.47. A worksheet still citing 220.12 and Table 220.44 is not automatically wrong, but it signals a template untouched since the last code cycle, and stale templates carry other stale assumptions.

Standard method and optional methods

The standard method is the category by category calculation. It is the default for commercial and institutional work and what a plan checker expects unless told otherwise. It is conservative by design, and on a genuinely diverse building it produces a number well above anything that building will draw.

Part IV offers alternatives where enough statistical evidence existed to justify a simpler and lower calculation: optional calculations for dwelling units and multifamily buildings, one for schools based on volt-amperes per square foot with its own demand factor table, and one for new restaurants keyed to whether the building is all electric. Separately, 220.87 lets an existing building’s measured maximum demand stand in for a calculation. None is free. Each carries conditions, and using one means accepting the whole method rather than the favourable half of it.

4. Which Projects Need One, and When

Any project that establishes or modifies a service or feeder needs a load calculation. The harder question is when it has to be right.

New construction

A preliminary calculation is needed at the end of schematic design, because the utility service application and the electrical room area both depend on it and both carry long lead times. It is refined through design development and finalised in the construction documents package as a permit deliverable.

Tenant improvements and fit-outs

A fit-out needs two calculations. The first establishes the demand of the new work. The second confirms the existing service and feeder to the suite can carry it, which is where NEC 220.87 and recorded demand data earn their keep. Skipping the second is how a landlord discovers, on the fourth fit-out in a building designed for six, that the service is full.

Additions, change of occupancy and electrification

An addition loads existing distribution as well as adding new. A change of occupancy resets the analysis entirely, because lighting unit load, receptacle count, kitchen equipment and ventilation change together. Converting warehouse to laboratory use is the extreme case: modest architecture, effectively new electrical construction. Replacing gas equipment with heat pumps or installing chargers does the same to an existing service, and is now the most common reason a building needs a fresh worksheet. The impact of electrification on MEP engineering shows up first, and most expensively, in service capacity.

5. What the Worksheet Asks of Each Load Category

Each row has its own rule. Here is what governs the main categories.

General lighting

NEC 220.42 requires a unit load of not less than the Table 220.42(A) value for the occupancy, applied to the floor area. Those values were substantially revised in the 2020 code cycle using lighting power density data from ASHRAE and the model energy codes, and now sit far below the legacy figures that had stood since the 1970s. Where a building complies with an adopted energy code, 220.42(B) permits the energy code values instead, but only if a power monitoring system reports the general lighting load continuously, alarms alert the owner when it exceeds the code value, the lighting demand factors are not applied, and 125 percent is applied for continuous operation. A lighting load below the table minimum without those four conditions is not compliant.

Receptacle loads

General purpose receptacle outlets in non-dwelling occupancies are counted at 180 volt-amperes per single or multiple receptacle on a yoke. Above a threshold a demand factor applies: the first portion at full value, the remainder at a reduced percentage under Table 220.47. It is one of the few genuine reductions in a commercial calculation and one of the most frequently misapplied, because engineers extend it to dedicated equipment circuits.

Continuous and non-continuous loads

Interior and exterior lighting, sign circuits, IT and server equipment, refrigeration and vehicle charging are normally continuous. Convenience receptacles, most kitchen appliances and intermittently cycling motors are not. The 125 percent factor applies at the branch circuit, the feeder and the service, and it is the item most commonly omitted from a spreadsheet based calculation.

Motors and the largest motor rule

Motor loads come from the full load current tables in Article 430 rather than from nameplate current, and 25 percent of the largest motor’s full load current is added to the total. Where that motor is a chiller compressor the adder is not trivial, and where several motors are close in size, confirm it was taken on the genuinely largest rather than whichever appeared first in the schedule.

Kitchen equipment

Commercial kitchen equipment takes a demand factor from Table 220.56 based on the number of units of thermostatically controlled or intermittently used equipment, the reduction increasing as units are added and reaching 65 percent at six or more. Two limits apply: the result must never fall below the sum of the two largest kitchen equipment loads, and the table does not cover general purpose motor loads, ventilation fans or space conditioning, which are calculated separately.

HVAC and the noncoincident load rule

NEC 220.60 permits, where it is unlikely that two or more loads will be in use simultaneously, that only the largest be counted. Electric space heating and mechanical cooling are the textbook pair, and in a genuinely single mode building the smaller is dropped.

The rule is heavily abused. A heat pump with electric resistance backup can run both. A variable air volume system with reheat runs heating and cooling simultaneously by design. A north zone calling for heat while a south zone calls for cooling is not noncoincident in any meaningful sense. Before accepting a dropped heating load, ask for the sequence of operation. The discipline that produces a credible HVAC load calculation is what makes the electrical claim defensible.

Elevators

Elevator feeders may take the demand factors of Table 620.14, which start at 100 percent for a single car and fall as cars are added, reaching about 72 percent where ten or more share a feeder. The table applies only to feeders serving elevators, not to the individual branch circuits, which are sized under Article 430. Consolidating every car onto one feeder to capture the reduction also consolidates the failure mode, which is why many buildings split elevator loads across distribution boards.

Electric vehicle charging and energy management

Charging loads are continuous, and NEC 220.57 requires each piece of supply equipment to be calculated at 7,200 volt-amperes or its nameplate rating, whichever is larger. The escape valve is NEC 625.42: where an energy management system complying with Article 750 controls the equipment, the maximum load on the service and feeder is the maximum that system permits, and equipment with a restricted access adjustable ampere setting may be taken at the adjusted rating.

This is the difference between a service upgrade and no service upgrade on a great many California retrofits, and it is the line most often taken on faith. Chargers shown at an aggregate limited value must be backed by a specified, listed and installed system with the limit marked on the equipment. A note reading “assume load management” is not compliance.

A worked worksheet

Below is an illustrative service load summary for a 24,000 square foot two storey building with a small commercial kitchen, one hydraulic elevator, electric space heating and six Level 2 charging points at 480Y/277 volts three phase. The figures are representative rather than taken from a project.

Three things are worth noticing. The demand load is 73 percent of the connected load, a normal ratio for a mixed use building of this type. The calculated demand works out at roughly 10 volt-amperes per square foot, high for an office but reasonable with a commercial kitchen, electric heat and six chargers. And the spare capacity is stated as a percentage on the face of the document rather than left as an unremarked gap.

6. Title 24 and the California Layer

California does not use the NEC unmodified. The California Electrical Code is Part 3 of Title 24, adopted under the California Building Standards Commission, and it is the NEC with state amendments. The 2025 edition is based on the 2023 NEC and applies to permit applications submitted on or after 1 January 2026.

Title 24 Part 6 and power distribution

The California Energy Code, Part 6, is administered by the California Energy Commission and shapes the distribution system without appearing anywhere in Article 220. Section 130.5 imposes four things. Each service or feeder needs permanently installed metering, with capability increasing as service size increases. The distribution system must be arranged so load types can be monitored separately, which affects how many panelboards you need and how loads are grouped. Combined feeder and branch circuit voltage drop to the farthest connected load must not exceed 5 percent. And specified space types need controlled 120 volt receptacles, which changes how receptacle circuits are laid out.

EV charging and battery storage

CALGreen, Part 11 of Title 24, sets electric vehicle infrastructure requirements for nonresidential parking that scale with the number of spaces, reaching 20 percent of total parking as EV capable at the largest tiers, with a defined proportion built out as charging stations and office and retail carrying the higher requirement. Every one of those spaces is a load on the worksheet.

Part 6 separately requires battery energy storage alongside solar photovoltaics for many nonresidential occupancies, and the California Energy Commission publishes the sizing formulas with each code cycle. Storage is not a load in the conventional sense, but it changes the service arrangement, the point of interconnection and the protection scheme, and it belongs on the one line diagram beside the worksheet. Coordinating it with photovoltaic design at concept stage is far cheaper than retrofitting the interconnection later.

7. Specialist Cases

Four building types break the standard worksheet in predictable ways.

Healthcare

Hospitals and outpatient facilities carry an essential electrical system under Article 517, divided into life safety, critical and equipment branches, each with its own transfer scheme, so the calculation is done twice: once for the normal service and once for the alternate source. Imaging equipment adds momentary demands a calculated average does not describe, which is why manufacturers publish both a nameplate and a momentary rating. See our healthcare facilities work.

Laboratories

Laboratory buildings combine high receptacle density, high process load and high exhaust fan power, and the standard receptacle demand factor understates what a bench full of instruments draws. Load densities in research space are frequently double those of office space in the same building, and the flexibility lab tenants expect means the spare capacity policy has to be explicit. See our laboratory and industrial capability.

Data centres and mission critical

In a data centre the IT load is continuous, dense and known, so calculated demand approaches connected load rather than sitting well below it. The worksheet becomes a capacity plan rather than a code exercise, with redundancy configuration, uninterruptible power supply efficiency and the distinction between design and commissioned capacity all sitting on top of the Article 220 baseline. Our articles on power distribution system design for data centres and why data centre power systems fail cover the ground Article 220 does not, and our mission critical practice works to those standards.

Aviation and education

Airport terminals combine very large baggage handling motor loads, dense passenger charging demand and mandated standby systems, which makes both the largest motor rule and the noncoincident rule consequential. Schools qualify for the Part IV optional calculation, which can materially reduce the calculated demand. Our aviation and education teams handle both.

8. Choosing a Calculation Method

Where more than one method is available, the choice should be deliberate and recorded. The table below sets out how it usually resolves.

For an existing building, 220.87 deserves particular care. Thirty days recorded in a mild spring, in a half occupied building, will not represent the peak. The code requires the recording to reflect the maximum demand of the feeder or service, taken while occupied, with heating or cooling equipment and seasonal loads included. Data from the wrong month is worse than no data, because it produces a defensible looking number that is wrong.

9. How to Check Someone Else’s Worksheet

Reviewing a load calculation is a different skill from producing one. The reviewer does not have the equipment schedules in their head and does not need them. What they need is a sequence of tests that surfaces an error without reproducing the work.

The five sanity checks

Volt-amperes per square foot. Divide the calculated demand by conditioned floor area and compare it to what that occupancy normally produces. An office shell landing at 12 volt-amperes per square foot is either wrong or carrying something unusual, and either way it needs explaining. A laboratory landing at 4 is understated. Benchmark data on measured commercial electricity intensity comes from the U.S. Energy Information Administration, and research from Lawrence Berkeley National Laboratory covers plug and process loads specifically, where measured intensity most often diverges from assumption.

Transformer and service against the demand. Convert the demand to amperes yourself and check it against the transformer kVA and the main breaker rating. Thirty seconds, and it catches transposed digits, kVA quoted where kW was meant, and single phase arithmetic on a three phase service.

Panel schedule totals against the service calculation. Add the connected loads on every panel schedule and compare to the worksheet. They should reconcile. When they do not, either a panel was added after the service calculation was finalised or a load exists on the worksheet that no panel serves. The second is more dangerous because it hides in plain sight, and MEP BIM coordination workflows increasingly catch the mismatch automatically.

Demand factor by category. Read down the demand factor column and check each factor belongs to the category on its row. The receptacle reduction applied to dedicated equipment, the kitchen table applied to ventilation fans, and a lighting demand factor used in an occupancy that does not receive one are all caught by a single column read.

A named spare capacity allowance. Look for an explicit line. If future capacity is not on the worksheet as a number with a percentage beside it, whatever headroom exists is an accident of standard equipment sizes and will not survive the first fit-out.

The errors that recur

Spare capacity is an owner decision

The code does not require spare capacity, so nobody will produce it unless the owner asks. A single tenant owner occupied building can run on a modest allowance. A multi tenant shell with unknown fit-outs cannot, and 20 to 25 percent at the service and at each distribution board is a common target. Make the decision once, write it down, and defend it when value engineering arrives, because switchgear is always among the first places a contractor looks for savings. Our note on estimating MEP engineering costs covers where that trade off usually lands.

Key Takeaways

• A worksheet with no basis column cannot be reviewed. Insist on knowing where every connected load number came from.
• Connected load, calculated demand load and measured maximum demand are three different quantities and are not interchangeable.
• Article 220 was renumbered in the 2023 NEC. A worksheet still citing 220.12 and Table 220.44 is running on an old template.
• The 125 percent continuous load factor is the item most commonly omitted from a spreadsheet based calculation.
• The noncoincident rule in 220.60 only applies where loads genuinely never coincide. Ask for the sequence of operation.
• Kitchen equipment demand factors have a floor: never below the sum of the two largest appliance loads.
• Reduced EV charging loads require a specified, listed and installed energy management system under Article 750, not an assumption.
• Spare capacity that is not written on the worksheet as a percentage is not spare capacity. It is a gap waiting to be consumed.

10. Where Budlong Works

Load calculations are governed by a statewide code but reviewed locally, and local practice differs. Utility service application procedures vary between territories, plan check expectations differ between jurisdictions, and reach codes in several California cities add electrification and charging requirements above the state baseline. Budlong maintains teams across California so local plan check and utility experience sits behind every service submission. More on the practice is on our company story page.

Sectors where the worksheet carries the most risk

Healthcare carries a second calculation for the essential electrical system. Education qualifies for an optional method that changes the answer materially. Commercial buildings absorb the charging and battery storage requirements first. Mission critical facilities push calculated demand close to connected load.

11. Related Reading

12. Frequently Asked Questions