EV Charging Infrastructure Sizing: kW vs kWh

Last updated: November 2026

Fifty chargers rated at 11.5 kW represent 575 kW of connected charging capacity. That number alone cannot tell an engineer whether the existing electrical service is adequate. The answer also depends on the equipment’s input rating, other building loads, phase allocation, approved load controls and the utility’s connection requirements.

This guide separates peak power from energy delivery, explains how to use demand profiles, and works through five illustrative sites: a multifamily garage, retail parking lot, fleet depot, workplace and fast-charging corridor. The examples are planning calculations, not installation designs or accounts of completed projects.

SurgePV publishes this post and sells solar design software.

The Peak kW vs Energy kWh Distinction

Power in kilowatts (kW) describes the rate of electricity transfer. Energy in kilowatt-hours (kWh) describes the amount transferred over time. Electrical capacity and charging throughput need separate calculations, even when they use the same load profile.

Peak kW helps determine:

• The highest operating load at the site or a particular feeder
• Apparent power and current after accounting for power factor and voltage
• The operating conditions the transformer and switchgear must support
• Potential demand charges, subject to the utility’s measurement interval
• The control limit needed to prevent charging from exceeding an approved capacity

Energy kWh helps determine:

• How many vehicles can receive their required energy before departure
• The energy component of the electricity bill
• The potential contribution from solar generation
• The battery energy needed to shift charging between periods
• Whether the proposed service can meet the charging schedule at all

For a simple planning calculation, average power equals energy divided by time. The table holds required input energy at 200 kWh and changes only the delivery window. It assumes constant power and excludes other loads; it is not a service-rating table.

Basis: calculated as 200 kWh divided by the assumed window. If 200 kWh means energy reaching batteries rather than input energy, include charging losses before calculating grid demand.

A useful design question is whether drivers can tolerate a longer charging window. Extending that window lowers the minimum average power needed, but it does not automatically lower the legally calculated service load. The physical charging equipment and any approved controls must make the proposed operating limit enforceable.

A short peak can matter even when daily consumption is modest. Conversely, a site can stay below its electrical limit yet fail to deliver enough energy before vehicles leave. Keep both the maximum-power check and the energy-by-departure check in the design record.

Service Capacity Calculation

Start with equipment input ratings and a complete site load calculation. Use estimated concurrent use to study operation, but obtain approval before relying on a reduced load for service, feeder or circuit sizing.

A screening equation is:

Connected charging input kW = sum of the maximum input kW of all charging equipment.

For identical units this becomes ports multiplied by input kW per port, provided each port has an independent power allocation. Shared cabinets can have a total input limit different from the sum of their advertised dispenser outputs.

Step by Step

1. Record each charger’s input voltage, phase arrangement, maximum input current or kVA, output rating and power-sharing arrangement.
2. Identify existing non-charging loads and the relevant metering boundary. Do not treat a building’s average load as its maximum coincident load.
3. Map chargers to panels and phases. A balanced site total can conceal an overloaded branch or phase.
4. Document any proposed charging limit, how it is enforced and what happens if a meter, controller or network connection fails.
5. Have the electrical designer apply the adopted code and equipment instructions. Check the standard for continuous-load treatment and permitted load-management methods.
6. Calculate operating current using the appropriate voltage, phase arrangement and power factor. Keep operating current distinct from final conductor or protective-device sizing.
7. Obtain the utility’s transformer, metering and service requirements before selecting final equipment.

The DOE’s charging infrastructure development checklist recommends involving the utility and a qualified electrical contractor during planning. It also separates equipment procurement, permits and installation responsibilities. A charging vendor’s capacity estimate is one input to that process.

Worked Number: 20 Port Office Garage

Assume twenty chargers each draw 11.5 kW at maximum input. Their combined input is 230 kW. A proposed control limit of 138 kW would be 60% of that value. That is a hypothetical control setting, not an empirically justified demand factor or an approved service calculation.

If the building simultaneously draws 25 kW, total operating demand at that control limit is 163 kW. Assuming a balanced 415 V three-phase supply and overall power factor of 0.95:

Operating current = 163,000 / (√3 × 415 × 0.95) = about 239 A.

Apparent power is 163 / 0.95 = about 172 kVA. Neither result specifies the final service rating. The design still needs code treatment, equipment ratings, phase checks, fault-level assessment and utility acceptance. If the control system is not acceptable for limiting the calculated load, the electrical design must use another approved basis.

Before selecting the 138 kW setting, verify that it can deliver the expected daily energy. If the charging window were eight hours, it could supply at most 1,104 kWh of input energy at a constant limit. Late arrivals, vehicle acceptance limits and outages reduce what can actually be delivered.

Demand Factor and Concurrent Use by Site Type

Demand profiles depend on vehicle behavior and site operation. There is no universal concurrency percentage that makes a workplace, garage or retail project electrically compliant.

Define terms before comparing datasets. An occupied port can remain plugged in after charging ends. An actively charging port can draw far below its rating. Energy utilization, time utilization and peak simultaneous demand measure different things, so one cannot be substituted for another without a model.

Evidence to Collect Instead of a Generic Factor

This is a planning checklist based on the DOE development process, not measured network utilization. Use site telemetry or a documented demand model to supply the numerical assumptions.

Why the Range Matters

For an illustrative 100-port installation with 7 kW maximum input per port, a 50% operating scenario draws 350 kW and a 70% scenario draws 490 kW. That 140 kW difference is useful for sensitivity analysis. It is not a permitted reduction in the electrical load calculation. Demand charges may also change, depending on when and how the utility measures demand.

If a lower site limit is proposed, define it in the control specification and verify it with the authority having jurisdiction, or AHJ, and the utility. The design should identify protected circuits, phase limits, response behavior, fallback operation and access to settings. A commercial promise to keep usage low is not equivalent to an engineered control.

When no operating data exist, model more than one plausible charging schedule. Include concentrated arrivals and high-energy days, not only an average weekday. Keep the assumptions separate from any code-approved calculation so a later reviewer can see which conclusions depend on driver behavior.

Level 1, Level 2, and DCFC Power Envelopes

The charger class describes how electricity reaches the vehicle, but the selected equipment and vehicle determine the actual charging rate. Use the manufacturer’s input specification for electrical design and its output specification for the charging service offered.

Level 1 AC Charging

The DOE’s charging equipment overview describes US Level 1 charging as using a 120 V AC connection. It can suit long parking periods and modest energy needs. Check the proposed circuit and cordset instructions before treating an existing outlet as charging infrastructure.

For illustration, 120 V at 12 A and unity power factor is 1.44 kW of input. Over ten hours, constant operation would supply 14.4 kWh before charging losses. The energy reaching the battery will be lower, and the vehicle may draw less than the available power. This arithmetic is an example, not a permitted circuit-loading instruction.

Level 2 AC Charging

The same DOE overview describes US Level 2 charging on 208 V or 240 V supplies. Its broad equipment range is 2.9 to 19.2 kW. The vehicle’s onboard charger can limit the rate below the charging station’s advertised capacity.

Be careful with voltage assumptions. A single-phase unit drawing 48 A at unity power factor corresponds to 11.52 kW at 240 V, but 9.984 kW at 208 V. Connecting multiple single-phase chargers to a three-phase distribution system does not make each charger a three-phase load.

European AC equipment uses different supply arrangements and terminology. Check the selected unit’s voltage, phase support, installation instructions and applicable standard. Avoid transferring a US Level 2 rating directly to a European mode classification or assuming every vehicle can accept all available phases.

DC Fast Charging

DC fast charging converts AC input to DC before supplying the vehicle. DOE identifies vehicle characteristics and battery state of charge as influences on charging power. The advertised output rating therefore differs from both average session power and maximum AC input.

Ask whether a dispenser has dedicated conversion capacity or shares a cabinet with other dispensers. Document what each vehicle receives when every port requests maximum power. A pair of connectors on one port also does not necessarily represent two simultaneous charging positions.

Use model-specific charging curves to estimate session duration and queues. For electrical capacity, include conversion losses, auxiliary loads and the maximum permitted simultaneous output. A lower average session power is not proof that a higher simultaneous demand cannot occur.

For illustration, delivering 40 kWh over half an hour corresponds to 80 kW average output. That session could include an initial period above 80 kW and a later taper below it. An energy average alone cannot select the upstream breaker or transformer.

Diversity Factor by Site Type

Time-series modeling can show how charging loads overlap, but a statistical forecast and an enforceable maximum are different design inputs. State which one each result represents.

Workplace Charging Curves

A workplace scenario might concentrate arrivals in the morning and departures late in the afternoon. Model charging from each vehicle’s arrival, requested energy and accepted power, then sum the profiles. Add the building load using matching timestamps. The maximum of the combined profile may occur at a different time from either individual maximum.

Compare immediate charging with a scheduled alternative. In the scheduled case, verify each vehicle’s departure energy rather than reporting only a smaller peak. A result that reduces peak demand by leaving vehicles undercharged is not an equivalent service.

Use sub-hourly information where it affects queues, controller response or billing demand. Hourly energy totals can hide brief coincidence. Keep the time resolution and missing-data treatment visible when presenting a modeled peak.

Multifamily Overnight Pattern

An overnight scenario can provide scheduling flexibility, but residents do not all arrive or leave at the same time. Some vehicles may arrive late with a large energy requirement. Others may need an early departure. An equal allocation to every connected vehicle may therefore produce a different outcome from a departure-priority schedule.

Shared parking introduces another issue: a charged vehicle may occupy the space for the rest of the night. Model access and turnover separately from electrical utilization. A site can have spare electrical power while a driver cannot reach a free charging port.

DCFC Corridor Pattern

Public fast-charging arrivals can be clustered around travel patterns, events and holidays. Do not infer simultaneous peak demand directly from a network’s annual utilization percentage. A local station also has a different vehicle mix and traffic pattern from the network average.

As a purely mathematical illustration, suppose each of four ports independently has a 0.40 probability of being active, and an active port has a 0.35 probability of being in a defined high-power state. Each port’s high-power probability is 0.14, and all four being in that state has probability 0.14⁴, or about 0.0384%. The expected number in that state is 4 × 0.14 = 0.56.

Those assumed probabilities are not operating evidence. Independence may fail during clustered arrivals, and ports outside the defined high-power state can still draw substantial power. This calculation therefore cannot justify a service rating, a transformer size or compliance with a simultaneous-power obligation.

Transformer and Service Entry Sizing

Transformer selection requires an agreed load basis, equipment input characteristics and utility requirements. A generic percentage margin cannot replace that review.

Transformer Sizing Formula

For an operating load, apparent power is approximately real power divided by power factor:

kVA = kW / power factor.

For an illustrative 200 kW load at 0.95 power factor, apparent power is 210.5 kVA. If the utility separately requests a planning allowance, record that allowance and its purpose. Do not apply a growth factor, continuous-load factor and thermal allowance interchangeably or assume they always compound.

Obtain input power-factor information for the actual charging equipment across relevant operating conditions. Include auxiliary loads, harmonic characteristics, voltage tolerance and any intermediate transformers. The DC output power factor is not the quantity used to calculate the AC supply current.

Available Transformer Ratings

Ask the utility which transformer types and ratings are available for the connection. Voltage, impedance, ownership, installation location and metering requirements can affect the choice. A generic list of standard sizes across countries is insufficient for a purchase decision.

Document future expansion separately. Installing spare conduit may be worthwhile even if the first phase uses a smaller power allocation. Adding chargers later still requires a review of the service, control limit and utility agreement; physical parking space does not imply electrical headroom.

Service Entry Conductor Sizing

For a balanced three-phase operating load:

Current in A = input kW × 1,000 / (√3 × line-to-line voltage × power factor).

For a single-phase operating load:

Current in A = input kW × 1,000 / (voltage × power factor).

These equations calculate current, not final conductor sizes. Check the standard for applicable load factors, conductor ampacity, temperature correction, grouping, terminal ratings, protection and other installation conditions. Check voltage drop and fault performance as part of the detailed design.

As a calculation check, 71.9 kW at 240 V single phase and unity power factor is about 300 A. It is not 173 A; that lower value would incorrectly introduce a three-phase factor. Always write the supply arrangement next to the formula so an otherwise plausible result does not conceal the wrong assumption.

Get the utility’s connection schedule and equipment availability in writing. Lead times vary by location and specification; use a project quote rather than an unsupported national median. Assign an owner to service-design approval before the charging equipment order is released.

kWh Throughput Modeling

Daily charging energy equals the number of sessions multiplied by average energy per session. State where that energy is measured, since charger input, charger output and battery-stored energy are different quantities.

For a mixed site, calculate each user group separately. Fleet vehicles with known routes may have different requirements from visiting customers. Keep weekday, weekend and seasonal assumptions distinct where they affect the business case. A single daily average can conceal a capacity problem on the busiest day.

Illustrative Session Arithmetic

A retail charger supplying 7.2 kW for a 45-minute visit delivers at most 5.4 kWh if output stays constant. At an assumed 75% average-power fraction, delivery is 4.05 kWh. An 8 kWh session cannot be obtained from those same inputs; either the visit must last longer or the average power must rise.

Write the session calculation before multiplying by annual visits. This catches an error that otherwise propagates into revenue, energy cost, solar offset and battery sizing. The operating forecast should also distinguish successful charging sessions from vehicles that arrive but cannot charge.

Energy Cost Math

For an illustrative workplace, assume input energy of 1,200 kWh per workday and 22 workdays in a billing month:

• Monthly input energy: 1,200 × 22 = 26,400 kWh
• Assumed flat energy rate: $0.12/kWh
• Energy charge: 26,400 × $0.12 = $3,168
• Assumed billing demand: 280 kW
• Assumed demand rate: $18/kW-month
• Demand charge: 280 × $18 = $5,040
• Combined modeled charge: $8,208 before fixed fees, taxes and other tariff terms

These are hypothetical tariff inputs, not a utility quote. Multiplying by twelve is valid only if all months have the same assumed energy and demand. For a real annual model, use each billing period’s workdays, rate windows, peaks and any demand ratchet.

Demand charges can be material, but their share of the bill is not a universal percentage. A controller may reduce a billing peak while shifting energy into a more expensive time period. Compare the full tariff result and the charging-service outcome, not just the maximum kW.

NEC, IEC, and EU AFIR Compliance Anchors

Electrical installation rules, communication protocols and infrastructure-program requirements address different questions. The design must identify which requirements apply to the location, equipment and funding agreement.

The Open Charge Alliance describes OCPP as communication between charging stations and management systems. Its versions support different functions, including smart charging. A protocol claim alone does not demonstrate that a site controller meets the installation’s electrical requirements.

For a US installation, check the adopted NEC edition and local amendments with the electrical designer and AHJ. For an IEC-based installation, check the standard and local implementation rather than importing a clause or percentage from another jurisdiction. The approved control arrangement, equipment listing and installation instructions matter as much as the software’s scheduling features.

Record the compliance basis in the drawing package: jurisdiction, adopted edition, relevant equipment documentation, utility conditions and any public-program obligations. If the design depends on a control limit, document who can change it and how that change is prevented from invalidating the approved design.

During commissioning, verify the response to loss of communication, meter failure, controller restart and conflicting schedules. Define the acceptable fallback behavior before procurement. Retain the settings and verification record with the installation documents so a maintenance change does not silently remove the assumed capacity limit.

Worked Example 1: 10 Port Multifamily Garage

This illustrative garage compares an energy requirement with a proposed controlled input limit. It does not specify a final service, conductor or transformer rating.

Assumed inputs

• Ten ports with 11.5 kW maximum AC input each
• Total connected charging input of 115 kW
• Required charging input energy of 200 kWh per night
• Eight-hour charging window
• Proposed charging input limit of 69 kW
• Other coincident building input of 30 kW
• For the current calculation only, 240 V single phase and unity power factor

Without a controller-enforced reduction, the combined maximum input in this simplified model is 115 + 30 = 145 kW. Its operating current is 145,000 / 240 = about 604 A. With an effective 69 kW charging limit, combined operating input becomes 99 kW and current becomes 412.5 A.

These figures precede the required installation design checks. They do not show that a particular existing service is adequate. The 30 kW building load is an assumption that must be replaced with an approved existing-load assessment. The control system also needs acceptance for the intended load-limiting role.

The energy requirement provides a separate check: 200 / 8 = 25 kW average charging input. At a constant 69 kW limit, the theoretical delivery time is about 2.90 hours. Real operation can take longer because vehicles arrive at different times, accept different rates and may not be connected throughout the window.

Before recommending the lower operating limit, inspect the distribution path from service to each charger. A shared garage feeder, phase imbalance or individual circuit can constrain the design even when the building total is acceptable. Define how charging priority works when all residents connect and how an urgent departure is handled.

Compare costs only after obtaining like-for-like quotes for the electrical upgrade and the accepted control solution. Include commissioning, communications, ongoing support and any replacement obligations. The arithmetic does not support a guaranteed dollar saving.

Worked Example 2: 20 Port Retail Site

This illustrative retail site shows why session energy must match parking duration. All usage, control and tariff values are assumed for the example.

Assumed inputs

• Twenty ports, each delivering up to 7.2 kW
• Three sessions per port per day
• Each session lasts 45 minutes
• Average session output is 75% of the port rating
• An assumed overall input-to-output efficiency of 90% for the energy estimate

Session output energy is 7.2 × 0.75 × 0.75 = 4.05 kWh. Across 60 sessions, daily output is 243 kWh. At the assumed efficiency, charging input is 243 / 0.90 = 270 kWh per day, or 8,100 kWh over an assumed 30-day month.

If the hypothetical energy rate is $0.10/kWh, the energy charge is $810. If the billing demand is separately assumed to be 60 kW and the demand rate is $14/kW-month, the demand charge is $840. The modeled combined amount is $1,650 before other charges. Neither the 60 kW demand nor the efficiency is a measured result.

A 60 kW charging-input limit could constrain customer service when many vehicles arrive together. Twenty simultaneous sessions following the assumed average-output behavior would request 108 kW of output, already above that input limit. A schedule that achieves the daily energy total may still disappoint lunchtime customers.

The electrical designer needs the actual equipment input ratings and control architecture. A 144 kW sum of output ratings is not the site’s AC input nameplate. Include conversion losses and any separate auxiliaries according to the product documentation.

For solar planning, compare the charging profile with an hourly or finer PV production estimate. Do not use an assumed daily solar yield to claim a fixed self-consumption percentage. Solar energy available after customers leave does not help those sessions unless another load, storage system or export arrangement can use it.

Worked Example 3: 50 Port Fleet Depot

This illustrative fleet depot uses the departure schedule to check whether a lower charging limit can deliver the required energy. The example is not an account of a municipal installation.

Assumed inputs

• Fifty vehicles and fifty charging ports
• Maximum charging input of 19.2 kW per port
• Twelve-hour overnight availability for every vehicle
• Each vehicle requires 60 kWh of charging input energy
• Proposed total charging-input limit of 310 kW
• Other depot demand of 40 kW at the same time

Total connected charging input is 50 × 19.2 = 960 kW. Required nightly input energy is 50 × 60 = 3,000 kWh. Spread perfectly across twelve hours, that requires 250 kW average charging input. A 310 kW cap offers a theoretical maximum of 3,720 kWh across the window.

Now examine a phased schedule with groups of 17, 17 and 16 vehicles. Running the first group at every port’s maximum would request 326.4 kW, so the controller must limit that group. If it allocates the 310 kW cap equally, each vehicle gets about 18.24 kW and takes about 3.29 hours to receive 60 kWh. The second group is identical; the third takes about 3.13 hours at 19.2 kW each. The simplified schedule totals about 9.71 hours.

That leaves time within the assumed window, but only under the stated ideal conditions. Test late returns, early dispatch, a failed port, higher route energy and vehicle power limits before selecting the operating cap. If 60 kWh means battery-stored energy instead of input energy, revise the energy balance to include losses.

With 40 kW of other demand, the proposed controlled operating total is 350 kW. At an assumed balanced 480 V three-phase supply and 0.95 power factor, operating current is about 443 A and apparent power is about 368 kVA. Final electrical ratings require the accepted load calculation and utility design.

A useful operating specification names the dispatch-critical vehicles and explains how priorities change after a route delay. It also identifies who receives an alert if the requested departure energy cannot be delivered. Capacity control should protect both the electrical installation and the fleet’s operational commitments.

Worked Example 4: 100 Port Workplace Site

This illustrative workplace separates the charging budget from the solar contribution. None of the assumed production or tariff values are a forecast for a named location.

Assumed inputs

• One hundred ports at 7.7 kW maximum input per port
• Ninety charging sessions per workday
• Twelve kWh of input energy per session
• Eight-hour charging window
• Two hundred forty workdays per year
• Hypothetical energy price of $0.15/kWh

Connected charging input is 770 kW. Daily energy is 90 × 12 = 1,080 kWh, requiring 135 kW average over the window. Annual input energy is 259,200 kWh and the flat-rate energy cost is $38,880. These calculations exclude demand charges, fixed fees and time-of-use differences.

Suppose a separate operating scenario has 500.5 kW of charging input and 80 kW of building demand. Combined input is 580.5 kW. At 480 V balanced three phase and assumed 0.95 power factor, current is about 735 A. This is an operating scenario, not proof of a permitted service rating or a prediction of billing demand.

Solar Integration

At a particular moment, if the same site has 200 kW of AC solar generation behind the same meter, grid import falls from 580.5 to 380.5 kW, ignoring other losses and flows. When solar output falls to zero, import returns to 580.5 kW unless charging or another load changes. A service design must address that low-solar condition.

For energy savings, align the solar and charging time series and apply the tariff to the resulting meter flows. A modeled annual PV total cannot establish the highest import interval or the share of generation consumed by vehicles. Building loads may consume solar first, depending on the metering and allocation method.

SurgePV supports solar yield and financial modeling. Use those solar assumptions alongside a dedicated EV charging and control assessment; do not assume the platform performs EV arrival simulation, controller programming or automatic electrical-code approval. Retain the input files and metering diagram used in the combined assessment.

Worked Example 5: 4 Port DCFC Corridor Site

This illustrative fast-charging site begins with simultaneous output capability rather than average session utilization. Four ports rated at 350 kW each can represent 1,400 kW of output if the conversion equipment supports all four at once.

Assumed input calculation

• Four ports at 350 kW simultaneous DC output
• Assumed conversion efficiency of 95%
• Forty kW of site auxiliaries outside that efficiency boundary
• Assumed aggregate AC power factor of 0.99
• Balanced 480 V three-phase supply

Maximum operating AC input in this simplified case is 1,400 / 0.95 + 40 = about 1,514 kW. Apparent power is about 1,529 kVA and operating current is about 1,839 A. Use the manufacturer’s actual maximum input specification for detailed design; these assumed factors are not equipment specifications.

If the site instead uses shared conversion capacity, document the cabinet limit and the power available to each vehicle during simultaneous demand. Check whether that allocation meets customer promises and any funding-program requirements. Do not describe a statistical 95th-percentile demand as a hard maximum.

Battery Storage Option

A battery can supply part of the charging demand, but both power and usable energy matter. A hypothetical 500 kW battery inverter supplying a constant 300 kW shortfall for half an hour must deliver 150 kWh at its AC output. The battery needs additional energy for losses, reserves and degradation, and it must recharge before the next event.

A nominal 1 MWh battery label alone cannot demonstrate this service. Review usable capacity, state-of-charge limits, inverter rating, temperature behavior and the controller’s response when storage is unavailable. If grid capacity depends on the battery, explain how charging is limited during a depleted or failed-battery condition.

Model consecutive busy periods and the recharge requirement, not just one favorable event. A battery that successfully covers the first arrival cluster may be unable to cover the next. Economic payback requires the actual tariff, installed cost, maintenance and replacement assumptions; there is no supported universal payback period for this example.

How Software Supports EV Charging Sizing

Software can help compare charging schedules and energy flows when its functions match the design question. Ask which calculations are supported, what inputs they require and how the results can be reviewed.

Useful capabilities to evaluate in a dedicated EV planning tool include arrival and departure modeling, vehicle power limits, session energy, queue behavior, control constraints and utility tariff calculations. These are procurement criteria, not claims that every solar design platform includes them.

For a solar-plus-charging project, SurgePV supports the PV design work: roof modeling, array layout, shading, energy yield, financial modeling and proposals. EV controls, electrical protection and compliance need their own engineering workflow. Bring the supported solar outputs into that process with clear timestamps, units and assumptions.

Use the DOE EVI-X toolbox when evaluating charging infrastructure needs and electrical demands under defined EV adoption scenarios. Choose the specific model and geographic scope that fit the question. An adoption scenario does not replace installation design for an individual site.

A useful model handoff includes its version, weather basis, charging inputs, time resolution, tariff and operating constraints. Reconcile totals before comparing scenarios. If two tools report different peaks, first check whether one is reporting charger output and the other is reporting total grid input.

Common Sizing Errors

The most important checks concern the load boundary, control assumptions and energy balance. Review these before treating any model output as a procurement specification.

Error 1: Treating Estimated Concurrency as Approved Capacity

A forecast of low simultaneous use is useful for operations and economics. It is not automatic permission to reduce the electrical load calculation. Use the adopted rules and an accepted control arrangement where required. Keep the forecast and the design limit in separate fields.

Error 2: Using the Wrong Multipliers or Supply Formula

Changing the order of two multipliers does not change their product. The important questions are which factors apply and what quantity each factor modifies. Do not apply a blanket continuous-load allowance to an energy total, or use a three-phase current equation for a single-phase supply.

Error 3: Sizing from Session-Average DC Output

Session-average output describes energy delivery over time. The upstream electrical equipment must address the maximum load on the approved design basis. Include conversion and auxiliary requirements, and check the actual power-sharing architecture.

Error 4: Assuming Network Connectivity Guarantees Control

A networked charger may support scheduling without providing the load-limiting behavior required for the installation. Specify failure behavior, protected limits, supported equipment combinations and commissioning evidence. Verify the installation as a system, not just a list of protocol-compatible products.

Error 5: Leaving the Utility Decision Until Procurement

Service capacity, metering, protection and equipment availability can affect the layout and schedule. Obtain the utility’s requirements while options remain open. Keep the connection approval and the charging-equipment order aligned to the same design revision.

Solar PV Integration

Solar can reduce grid energy when its generation overlaps charging or other site loads. Its nameplate capacity cannot simply be subtracted from a required charging-service rating.

A daytime workplace may have better direct overlap than an overnight depot, but the result depends on location, weather, parking schedules and meter configuration. Model the actual operating hours. A hotel can include daytime visitors as well as overnight guests, and a fleet may return for daytime charging between routes.

For each interval, track building demand, charging demand, PV generation, storage charging or discharging and grid import or export. Keep the energy balance consistent across the same meter boundary. If the PV array and charging ports use separate meters, the commercial allocation can differ from the physical site total.

A battery can shift solar energy, but doing so consumes capacity that might otherwise support a demand peak or backup reserve. State the operating priority and model conflicts between those uses. Do not count the same stored energy twice in the savings calculation.

For a proposed import limit, examine low-solar conditions, a full or depleted battery and unavailable equipment. A controller may need to reduce EV charging to protect the limit. Document the resulting effect on customer service and departure energy rather than treating curtailment as free capacity.

Use the solar yield and financial workflow to assess PV assumptions, then reconcile those outputs with the EV engineer’s load model. Retain site-specific shading and roof constraints so the proposed solar capacity is physically credible. The combined analysis should show energy savings separately from any approved capacity benefit.

EU AFIR and NEVI Sizing Constraints

Public infrastructure obligations can require simultaneous charging performance that an average-use model would not provide. Read the requirements for the specific jurisdiction and funding agreement before setting a power-sharing limit.

EU AFIR

The European Commission’s AFIR overview states that the regulation has applied since 13 April 2024. It includes national fleet-based targets and distance-based targets for light- and heavy-duty vehicles on the TEN-T networks, alongside interoperability, information and payment requirements.

Determine which vehicle category, road network, deadline and permitted exception apply to the project. A national target is not a universal transformer-sizing instruction. Verify the applicable charging-pool and point requirements in the regulation and national implementation before finalizing the electrical brief.

NEVI and Covered US Corridor Stations

The opened 23 CFR 680.106 requires covered corridor-serving stations to have at least four network-connected DC fast-charging ports capable of charging four vehicles simultaneously. Each must meet a vehicle’s request up to 150 kW simultaneously; power sharing must preserve that capability.

Four ports therefore require at least 600 kW of simultaneous DC output capability under that requirement. Applying a 0.55 concurrency assumption to reduce the commitment to 330 kW would not meet it. The upstream input must also account for equipment losses and auxiliaries. The regulation does not specify one universally compliant transformer size.

Review the applicable award conditions and other program requirements with the responsible agency. A design can meet an average-energy forecast and still fail its simultaneous-power obligation.

UK Public Charge Point Regulations

The UK government’s public charge point guidance describes 99% reliability averaged across an operator’s rapid public network of 50 kW and above over the calendar year. This is broader than a requirement applying only to strategic-road locations, and it is not a universal site power minimum.

Include monitoring, maintenance and reporting responsibilities in the operating plan. Electrical sizing supports reliable operation, but reliability also depends on equipment service, payment systems and the treatment of outages under the applicable rules.

Conclusion

EV charging infrastructure sizing needs two connected assessments: an energy-delivery plan and an approved electrical capacity design. Start with input ratings, realistic sessions, departure requirements and the existing site load. Then evaluate a control arrangement and utility connection that can support the intended service.

The five worked examples show how to check the arithmetic without confusing a planning assumption with a required equipment rating. Replace every assumed efficiency, power factor, tariff and charging schedule with project evidence before using the result commercially.

For a solar component, book a SurgePV walkthrough to review the supported layout, shading and yield workflow. Bring those PV assumptions to the charging engineer so both sides of the project use the same site and metering plan.

Frequently Asked Questions

How do you calculate EV charging infrastructure size?

Calculate both required charging energy and the maximum electrical load on an approved design basis. Use equipment input ratings, other site loads and any accepted control limits. Have the electrical designer and utility determine the final service and equipment ratings.

What is the difference between kW and kWh in EV charging?

Kilowatts measure power, while kilowatt-hours measure energy delivered over time. Delivering 200 kWh over eight hours requires 25 kW average before additional losses. Peak demand and required electrical ratings can be higher than that average.

Can a demand factor reduce an EV charging service rating?

An assumed utilization or concurrency factor alone cannot justify a smaller electrical service. Any reduction must follow the adopted rules and accepted control arrangement. A forecast of typical usage and an enforceable maximum are different inputs.

How much energy can a 7.2 kW charger deliver in 45 minutes?

At constant full output, a 7.2 kW charger delivers 5.4 kWh in 45 minutes. Lower vehicle acceptance or interrupted charging reduces delivery. Battery-stored energy also depends on losses.

Does OCPP support prove electrical load-management compliance?

OCPP support alone does not prove that a site meets electrical load-management requirements. It describes communication capabilities between equipment and management systems. Verify the complete control implementation, failure behavior and installation acceptance.

Can solar reduce the service capacity needed for EV charging?

Solar can reduce grid import while it is generating, but it cannot automatically be deducted from the required service capacity. The design must address low-solar periods and unavailable equipment. Any capacity reduction needs an accepted operating and control basis.

Can covered NEVI corridor sites use only average charging demand?

Covered corridor sites must preserve the required simultaneous charging capability. The cited rule requires at least four ports that can each supply a vehicle’s request up to 150 kW simultaneously. Average utilization cannot replace that obligation.