DC fast charging (DCFC) station design and installation guide

DC fast charging, and why the charger is the easy part

DC fast charging delivers high-power direct current straight to the vehicle battery, which is why a depleted car can take a usable charge in minutes instead of the hours a Level 2 unit needs. The equipment people picture, the stall and the cable, is the small part of the job. The service that feeds it is the project.

Here is the part that surprises people who have only done Level 2 work. A single DC fast dispenser can pull anywhere from 50 kW to 350 kW or more, so a site with a handful of them carries a load measured in megawatts, not the 40 or 80 A branch circuit a wall charger lives on. That load usually means a new transformer, switchgear rated for the available fault current, a large feeder run underground, and a conversation with the utility that starts long before anyone orders a charger.

So the schedule, the cost, and the risk all live upstream of the dispenser. Load management to cap the site peak, on-site battery storage to dodge demand charges, trenching and conduit across a parking field, and the make-ready built ahead of the chargers are the real scope. Price the charger as the last and easiest item on a long list and the estimate comes out right. Treat the service as an afterthought and the utility timeline runs the project for you.

This guide covers the DC fast side. For the Level 2 wall and pedestal install, the GFCI and disconnect detail, and the commissioning charge, see the EVSE installation and commissioning guide. For the conductor and overcurrent math on a single charger circuit, see the EV feeder sizing walkthrough. The point here is the system: the service, the utility, and the equipment architecture that make a fast-charging site work.

What is the difference between Level 2 and DC fast charging?

The difference is where the AC becomes DC and how much power moves. A battery stores DC, so something has to convert the grid's AC. On Level 1 and Level 2, that converter is the onboard charger inside the car, and it is small, commonly 7 kW to 19 kW, which is why an AC charge takes hours. DC fast charging moves the converter out of the car and into the station, so the equipment hands the battery DC directly and skips the onboard bottleneck.

That single change is why the power levels jump. The onboard charger no longer limits the rate. The station's power electronics and the battery's acceptance do. A DC fast site is therefore a different electrical animal. Level 2 is a continuous load on a branch circuit. DC fast is a large, variable, electronically controlled load the site has to feed, manage, and pay demand charges on.

On the install side the trades barely overlap. A Level 2 unit is a mounting detail and a branch circuit. A DC fast station is service-entrance work: medium or low-voltage service, switchgear, large feeders, grounding for the fault current, and utility coordination. If your experience is wall chargers, the jump to DC fast is a jump in scale, not a bigger version of the same job. The EVSE guide covers the Level 2 side in depth.

The power levels: 50 kW, 150 kW, and 350 kW and up

DC fast charging is sold by power level, and the levels cluster around a few points. 50 kW was the early standard and still shows up on legacy and low-cost units. 150 kW is the current workhorse and the floor for federally funded corridor sites. 350 kW and above is the high-power tier, and megawatt charging for heavy trucks is arriving on top of that. The published number is the cabinet's rating, not a promise to every car.

The rate any vehicle actually takes is the lower of three limits: what the station can deliver, what the vehicle's battery and onboard system will accept, and what the battery's state of charge and temperature allow at that moment. A car rated for 250 kW plugged into a 350 kW dispenser still charges at 250 kW at best, and tapers well below that as the battery fills. This is why a stall's nameplate kW and a driver's experience often disagree.

Design for the future, not just today's cars. The vehicles charging on the site in five years will accept more than today's fleet, and high-power equipment paired with load management lets a site deliver its full rating to one car or share it across several. Size the service and the make-ready for where the equipment is going. Confirm the deliverable power and the taper behavior against the manufacturer's specification, because the cabinet rating and the per-port output under sharing are not the same number.

Which connector do you install: CCS, NACS, or CHAdeMO?

In North America the connector decision has consolidated around two: CCS1, the Combined Charging System, and NACS, now standardized as SAE J3400. NACS, the connector Tesla opened up, has become dominant as the automakers adopt it on new vehicles, and most public DC fast deployments now carry both so any car on the lot can charge. CHAdeMO is legacy. It survives on older Japanese models and tops out around 100 kW to 150 kW on most hardware, and new sites rarely add it.

What that means on the ground is a mixed-connector site. Federal corridor funding has required at least one CCS connector per port, while many states now also call for NACS to match the vehicles on the road, so the practical answer is to provide both and confirm the required mix against the program and the manufacturer before you order cabinets. Adapters fill gaps but are not a substitute for the connector the site is required to provide.

The connector also drives the cable. High-power connectors above roughly 200 A run liquid-cooled cables so the conductor stays manageable in a tech's hand instead of growing into something nobody can lift to the port. Get the connector mix and the cable rating from the equipment manufacturer and the funding program, not from a generic spec, because this is the part of the design the market keeps moving.

Why is the electrical service the real project, not the charger?

Because the service is where the money, the time, and the risk concentrate. A DC fast site's connected load runs from hundreds of kilowatts to several megawatts, and a load that size rarely fits the existing service. The common path is a new service: a transformer sized for the peak, switchgear rated for the available fault current, a large feeder, and a metering and protection arrangement the utility will accept.

The charger, by comparison, arrives on a pallet, sets on a pad, and lands its feeder. The hard, slow, expensive work is everything that has to exist before that pallet shows up. Industry timelines for getting high-power chargers energized routinely run 12 to 24 months when transformer or feeder upgrades are involved, and almost none of that clock is the charger itself. It is utility engineering, equipment lead time, and trenching.

This is the framing that keeps an estimate honest. Price the service, the utility work, the civil, and the make-ready as the project, and treat the dispensers as the trim that goes on at the end. Verify the actual service requirement and the available capacity with the utility early, because the difference between a site with spare capacity and one that needs a substation upgrade is the difference between a six-month job and a two-year one.

Medium-voltage vs low-voltage service

The first service decision is whether the site takes power at low voltage or medium voltage. A smaller site, a few 150 kW ports, can often run on a low-voltage service, commonly 480Y/277 V, fed from a utility or customer transformer the way any commercial building is. The chargers' rectifiers take that AC and make the DC.

A large site, many high-power ports or a truck depot in the megawatt range, usually justifies a medium-voltage service, where the utility brings primary voltage to a customer-owned transformer and switchgear pad on site. Medium voltage lets the feeders carry the same power at lower current, which keeps the conductors and the trenching sane, but it adds a transformer, primary protection, and the engineering and clearances that come with primary equipment.

The crossover depends on the total load, the utility's available service at the location, and the cost of conductor and trenching at low voltage against the cost of a medium-voltage transformer and switchgear. There is no universal threshold. The utility's service rules and the available primary or secondary capacity at the site drive the decision, so confirm both before the one-line is set.

How big does the service have to be?

Size the service to the demand load, not the simple sum of every dispenser's nameplate. The connected load is the total if every port ran flat out at once. The demand load is what the site actually draws given diversity: not every stall is occupied, not every car is at its peak rate, and tapering means a port at 80 percent state of charge is pulling a fraction of its rating.

On a public site with no load management, the conservative design feeds the full connected load, because nothing stops simultaneous full-power charging. That is the most expensive service. Add load management and the service can be sized to a capped site peak instead, which is usually the move that keeps the project off a service upgrade. The trade-off is charge speed when the site is busy.

The per-charger conductor and overcurrent math, the 125 percent continuous-load rule, and the voltage drop over the long parking-field runs are covered in the EV feeder sizing walkthrough. At the service level the question is the aggregate: connected load, the diversity you can defend, and whether load management or storage lets you serve it with a smaller service. Confirm the demand basis with the utility and the equipment manufacturer, since both have a say in what counts.

Demand charges: the bill that sinks the economics

A demand charge is a utility charge based on the highest power draw, in kW, during a billing period, separate from the energy charge for total kWh used. It exists because the utility has to build capacity for a customer's peak even if that peak lasts a few minutes. For DC fast charging it is the line item that quietly decides whether a site makes money.

The problem is utilization. A new public site might see a few sessions a day, but the moment two 350 kW cars charge at once, the site sets a 700 kW peak the demand charge bills for the whole month, even though the site sat near idle the rest of the time. Demand charges commonly make up a large share of a commercial electric bill, and in some tariffs a single brief peak sets the charge for the entire period.

This is why a site can be busy with cars and still lose money. The fix is not more sessions alone. It is controlling the peak. Two tools do that: load management to cap how high the site can draw, and on-site battery storage to supply the spikes so the grid draw stays flat. Confirm the actual tariff and the demand-charge structure with the utility, because the rate design varies widely and it changes which fix pays off.

Load management and power sharing

Load management caps the site's total draw by sharing the available power across the dispensers instead of letting each pull its full rating at once. A site with four 150 kW ports has 600 kW of connected load, but a load-management system can hold the site to, say, 300 kW and split it among whoever is plugged in. Two cars get 150 kW each, four cars get 75 kW each, and the service never sees more than the cap.

This does two things that matter to the budget. It lets the service and the transformer be sized to the cap instead of the full connected load, which can be the difference between using the existing service and paying for a utility upgrade. And it holds the demand-charge peak down, directly cutting the operating bill. The NEC recognizes this: where an automatic load-management system controls the load, the maximum load on the service and feeder is the value the system enforces, not the connected total.

Dynamic load management goes further, reading the building or site load in real time and giving the chargers whatever headroom is left under a fixed service limit. On a site shared with a store or a depot, that is what lets the chargers use spare capacity at night and back off when the building load climbs. Size and configure the system to the manufacturer's specification, and make sure the enforced cap is documented, because the inspector and the utility will both ask what holds the load down.

On-site battery storage to shave the peak

Battery energy storage sits between the grid and the chargers and supplies the short, high spikes so the grid sees a flatter draw. When a 350 kW session hits, the battery covers the surge; between sessions it recharges slowly at low power. The grid connection and the demand charge get sized to the average, not the peak, which is the whole point on a low-utilization public site.

This is also the tool that breaks a grid bottleneck. Where the utility cannot deliver the full peak, or where the upgrade would take two years, a battery-buffered site can present a small grid load and still deliver high power to cars from the battery. Projects that would wait a year or more for a transformer have been brought online in months by buffering the connection with storage. Storage pairs naturally with on-site solar where the site has room, offsetting energy cost on top of the demand savings.

The sizing is real engineering, not a bolt-on. Published analysis has put the storage needed at roughly 120 kWh per 150 kW port to hold full power through the first hour, so a multi-port site carries a serious battery. Confirm the capacity, the round-trip behavior, and the interconnection treatment of the storage with the manufacturer and the utility, and design the battery's own fire separation and listing into the site plan from the start.

Power cabinet and dispenser, or all-in-one

DC fast equipment comes in two architectures. An all-in-one unit puts the rectifier, the controls, and the connector in a single enclosure at the stall, which is simpler and cheaper for a one or two-port site. A split architecture separates a central power cabinet, which holds the rectifiers and the heavy power electronics, from the dispensers at the stalls, which carry just the connector, the cable, and the user interface.

Split is what high-power, multi-stall sites use, and the reason is power sharing. One power cabinet feeds several dispensers over a DC bus and allocates its capacity dynamically, so the cabinet's rectifiers follow the cars instead of sitting idle behind a single stall. A common arrangement runs one cabinet to as many as four dispensers. It costs more and takes more room than all-in-one, but it is what makes load sharing and high per-port power work.

Either way the equipment makes heat, and the rectifiers are liquid or forced-air cooled with filters and fans that become a maintenance item. The cabinet footprint, the working clearances, the cooling airflow, and the cable routing from cabinet to dispenser all belong on the site plan early. Take the dimensions, the clearances, and the cabinet-to-dispenser distance limits from the manufacturer, because they vary by product and they constrain the civil layout.

The dispenser, the cable, and the stall

The dispenser is what the driver touches: the connector, the cable, the screen, and the payment terminal. On high-power stalls the cable is liquid-cooled, because a cable that carries 350 kW without cooling would be too thick and stiff to handle. The cooling keeps the conductor smaller and the connector light enough to plug in one-handed, and it is one more fluid loop that needs maintenance.

Stall layout is where the civil and the human factors meet. Cable reach is limited, so the dispenser has to sit where the connector reaches the car's port, and vehicle ports are in different places, so the stall and the cable management have to handle a pickup, a sedan, and a car parked slightly off. Pull-through stalls matter for vehicles towing or for trucks.

Accessibility is not optional. Public charging falls under accessibility requirements for reach range, clear floor space, and an accessible route, and the dispenser interface, the connector height, and the approach have to meet them. The exact dimensions come from the adopted accessibility standard and the AHJ, so confirm the reach ranges and the accessible stall count for the jurisdiction rather than assuming the equipment ships compliant.

The site: trenching, conduit, pads, and protection

The civil scope is most of the field labor on a DC fast site. Large feeders run underground from the service to the power cabinets and from the cabinets to the dispensers, which means trenching across a parking field, conduit and duct banks sized for the conductors and for spares, and concrete pads engineered for the cabinet and dispenser weights. On a medium-voltage site, add the transformer and switchgear pads and their clearances.

Every dispenser in a drive aisle needs protection from vehicles. Bollards rated to take a hit, wheel stops, and curb protection keep a car from driving into a live cabinet, and the equipment manufacturer often specifies a minimum bollard arrangement the install has to follow. Some sites add a canopy for weather and lighting, which then carries its own structural and electrical scope.

Plan the conduit and the trench for the future, not just the first phase. Pulling spare conduit and sizing the duct bank for added ports while the trench is open is cheap. Reopening a finished parking lot is not. This is the make-ready logic applied to the civil: the hole in the ground is the expensive part, so build it once for where the site is going.

Make-ready: the infrastructure built ahead of the chargers

Make-ready is the electrical infrastructure installed before, and sized beyond, the chargers a site starts with: the service, the transformer capacity, the panel or switchgear, the conduit, and the pads, built so adding chargers later is a connection rather than a reconstruction. It is the planning move that separates a site that scales from one that gets torn up every expansion.

Utilities and programs push make-ready because it is cheaper for everyone to build capacity once. Many utilities run make-ready programs that cover part of the infrastructure cost up to the charger, and federal and state funding has favored sites built with headroom. Where such a program applies, its rules shape what gets installed and what gets reimbursed, so confirm the program scope before the design is final.

The judgment call is how far ahead to build. Oversize the service and the duct bank and you spend money on capacity that sits unused. Undersize them and the second phase pays to redo the first. The defensible answer ties the make-ready to a real expansion plan and the utility's available capacity, not to a guess. Build the trench, the conduit, and the service for the planned port count, and land the chargers as the demand arrives.

How long does utility coordination take?

Plan for the utility to be the longest item on the schedule. A DC fast site of any size needs a service application, a load study, utility engineering, possibly a transformer or feeder upgrade, and an interconnection or service agreement, and that sequence routinely runs 12 to 24 months when upgrades are required. The charger lead time disappears next to it.

Start the utility conversation before the design is locked, because the utility's answer changes the design. The available service capacity at the location, the voltage offered, the metering and protection the utility requires, and whether a transformer upgrade falls on the customer or the utility all flow from that early coordination. A site that looks simple on paper can need a substation-level upgrade nobody priced because nobody asked the utility first.

Where a make-ready or interconnection program exists, it can both fund and pace the work, so factor its timeline in. The blunt version: the utility, not the contractor, controls when a DC fast site energizes. Build the schedule around the service application date, keep the utility engaged through the milestones, and treat the energization date as something you confirm with them rather than promise to a client.

Grounding, overcurrent, and surge protection

DC fast equipment is fed like any other large load on the AC side, so the feeder needs overcurrent protection sized for the continuous load, an equipment grounding conductor sized to the conductors, and a disconnect the code requires. The grounding electrode and bonding for a new service or a customer transformer follow the service rules, and on a medium-voltage site the primary grounding is its own design.

The DC side and the high-power electronics add protection the equipment handles internally. The listed charger provides the ground-fault and DC-side protection its standard requires, which is part of why the equipment listing matters and why field-modifying a charger is off the table. Surge protection belongs on the design because the power electronics are sensitive and the equipment sits outdoors exposed to switching and lightning transients.

Treat the listed equipment's internal protection and the site's AC-side protection as two separate responsibilities. The manufacturer owns what is inside the cabinet. The installer owns the feeder, the grounding, the disconnect, and the surge protection feeding it. Size the AC-side protection and grounding to NEC Article 625 and the feeder rules, and confirm the surge and DC-side requirements against the equipment listing and the manufacturer's instructions.

NEC Article 625 and the code that governs EVSE

NEC Article 625 covers electric vehicle supply equipment, and it controls the charger side of a DC fast site. The rule that drives sizing is that EV charging loads are continuous, so the service and feeder must be rated for at least 125 percent of the load, commonly cited at 625.42. Where an automatic load-management system is used, that same section lets the maximum load be the value the system enforces rather than the full connected total.

The disconnect rules matter on DC fast gear because the equipment is large. A disconnecting means that opens all ungrounded conductors is required, and for equipment over 60 A or over 150 V to ground, a lockable disconnect within sight of the equipment is the common requirement, which a DC fast cabinet will trigger. The equipment must be listed for the purpose, and a listed assembly is what carries the internal DC and ground-fault protection.

Article 625 is the charger side, not the whole job. The service, the feeder, the transformer, and the utility interconnection fall under the service and feeder articles and the utility's own rules, which sit outside 625. Section numbers and the continuous-load and disconnect details shift between code cycles, so confirm them against the adopted edition and any local amendments, and treat the AHJ and the equipment listing as the final word.

Networking, payment, and the back office

A public DC fast charger is a networked payment device as much as an electrical one. It talks to a back-office network over a cellular or wired connection, authorizes and bills the session, reports its status, and takes firmware updates. The common protocol between charger and network is OCPP, with corridor funding pushing OCPP 1.6J or higher and newer deployments moving to OCPP 2.0.1.

The payment and connectivity side is where a surprising share of failures live. Industry data has tied over half of charging failures to the station failing to reach its network for authorization rather than to anything electrical, which is why the cellular signal, the antenna placement, and the network configuration are part of the install, not an afterthought the operator deals with later.

Vehicle-to-charger communication rides on its own standard for the high-power handshake and plug-and-charge authentication, commonly ISO 15118. Get the network, the payment integration, and the communication standard from the operator and the manufacturer, and test them at commissioning, because a charger that passes its electrical tests but cannot authorize a payment is, to a driver, a broken charger.

Uptime: the reputation problem public charging lives with

Uptime is the metric public DC fast charging is judged on, and the industry has a credibility gap on it. Networks report 98 to 99 percent uptime while independent testing has found first-time charge success well below that, in the 70s to mid-80s percent range in recent years. Federal corridor funding sets a hard floor: each port must average better than 97 percent uptime on a rolling 12-month basis.

The gap comes from how uptime is counted against what a driver experiences. A station can be reported up while a payment terminal is frozen, a connector is damaged, or the unit derates and charges at a crawl. Real reliability is the power electronics, the cooling, the connectors and cables, the payment system, and the network all working at once, which is a higher bar than the box being energized.

This is an operations problem, not just an install problem, but the install sets it up. Build in monitoring, design the cooling and the cable management for the duty the site will see, and put a maintenance plan and a service-level agreement in place before the site opens. The uptime requirement and the way it is measured come from the funding program and the operator's agreement, so confirm both, because they define what the site is contractually on the hook for.

Fleet and depot charging vs public charging

Fleet and depot charging and public charging look like the same equipment solving two different problems. A depot charges known vehicles on a predictable schedule, usually overnight, returning to the same stalls, which means the load is plannable and managed. Public charging serves strangers who arrive at unpredictable peaks, want the fastest possible charge, and have to pay at the unit.

Those differences change the design. A depot leans on heavy load management and often lower per-port power spread across many vehicles, because there are hours to charge and the goal is to fill the fleet by morning at the lowest demand peak. It may skip the public payment and the corridor-grade uptime requirements. A public site optimizes for peak power, payment, accessibility, and uptime, and it lives or dies on the demand-charge math because its utilization is low and spiky.

Storage and load management pay off in both but for different reasons. A depot uses them to flatten an overnight load and ride a time-of-use rate. A public site uses them to survive the demand charge on a low-utilization spike. Match the architecture to the use case, because a public-grade build on a depot wastes money and a depot-grade build on a public corridor fails the uptime and payment requirements.

Commissioning the site

Commissioning a DC fast site ends with cars charging at the rated power and a payment going through, not with an energized cabinet and a green light. The power test confirms the equipment delivers its rating, alone and under power sharing, and that the load-management cap actually holds when several ports are loaded at once. A site that charges one car fine can still fail when the second and third plug in.

Walk the full chain. Verify each connector and cable on a real vehicle or a test load, confirm the network connection and a live payment transaction, check the ground-fault and safety functions the equipment provides, and confirm the disconnect and the labeling. On a medium-voltage or utility-interconnected site, the utility often witnesses the energization and the protection settings, so coordinate that step rather than discovering it on the day.

Document what you commissioned, not just that you did. The measured power per port, the load-management behavior under simultaneous load, the network and payment test, and the safety checks are the record the operator and the next technician rely on. Run the commissioning to the manufacturer's procedure and the program's acceptance criteria, since both define what counts as a site that is actually open for business.

Maintenance and the lifecycle

DC fast equipment has more moving and wearing parts than a wall charger, and the maintenance plan is what keeps the uptime number real. The connectors and cables take physical abuse and are a top failure point, and the liquid-cooled cables add a coolant loop that needs checking. The cabinet cooling, the fans, and the air filters foul over time and let the rectifiers derate or shut down if they are ignored.

Firmware is part of maintenance now. Chargers get firmware updates for payment, protocol, and reliability fixes, and a unit running old firmware can fail in ways that look like hardware faults. Remote monitoring catches a derating or an offline unit before a driver finds it, which is the difference between a scheduled fix and a one-star review with a truck roll behind it.

Build the maintenance into the project, not after it. A site with a service-level agreement, spare connectors and cables on the shelf, a filter and coolant schedule, and monitoring that actually gets watched will hold its uptime. One without those will drift down to the reputation the industry is fighting. Set the intervals to the manufacturer's schedule and the operator's agreement, and keep the records where the next tech can find them.

What to document

A DC fast site is too big and too long-lived to carry in anyone's head. The service rating, the utility agreement, the equipment, the load-management cap, and the commissioning results are the record that lets the next person service, expand, or troubleshoot the site without reverse-engineering it. A field tool such as FieldOS keeps that record with the site so it survives staff turnover and the years between the install and the next upgrade.

Capture the service and the utility side first, because that is the part nobody can reconstruct: the service size and voltage, the transformer and switchgear ratings, the utility agreement and interconnection terms, and the load-management cap the service was sized to. Then the equipment, the connector mix, the commissioning measurements, and the maintenance plan.

Common mistakes

• Underestimating the electrical service and starting the utility conversation too late, so the interconnection timeline runs the project.
• Skipping load management, so the service is oversized and the demand charge sinks the operating economics.
• Specifying the wrong connector mix for the vehicles and the program, and finding out after the cabinets are ordered.
• Building make-ready with no headroom, so the second phase tears up the first.
• Treating uptime and maintenance as the operator's problem, so cables, cooling, and payment fail and the site earns a bad reputation.
• Ignoring the NEC 625 continuous-load and disconnect rules, or assuming Article 625 covers the service and utility side too.
• Pricing the charger as the project and the service as an accessory, which gets the cost and the schedule backward.

Field checklist

Standards and references

The NEC, NFPA 70, governs the installation, with Article 625 covering the EV supply equipment, the continuous-load sizing commonly at 625.42, and the disconnect rules. The service, feeder, transformer, and grounding sit under the service and feeder articles, not 625. Section numbers move between code cycles, so confirm them against the adopted edition and local amendments.

The connectors follow SAE standards: SAE J1772 for the AC and CCS coupler family and SAE J3400 for NACS, which the market has consolidated around in North America, with CHAdeMO as a legacy standard on older vehicles. Vehicle-to-charger communication and plug-and-charge ride on ISO 15118, and the charger-to-network link commonly uses OCPP.

Where a site is built under the federal NEVI program or a state equivalent, that program sets minimums beyond the NEC: at least four 150 kW ports able to run simultaneously, the required connectors, and the 97 percent per-port uptime floor on a rolling 12-month basis. Confirm the program's current status before counting on it, since federal NEVI funding was paused and placed under review in 2025 even as the technical build standards stayed the common reference. The utility's service and interconnection rules and tariff control the service and the demand charges. Cite the standard that governs the point, and let the AHJ, the utility, the program, and the equipment listing be the final word.

Units and terms

DC fast charging carries its own vocabulary, and the same idea shows up under different names across a manufacturer sheet, a utility tariff, and a funding program.

DC fast charging / Level 3

High-power DC delivered straight to the battery; the station, not the car, converts AC to DC

CCS / NACS (J3400)

The two dominant North American DC connectors; CCS is the Combined Charging System, NACS is SAE J3400

CHAdeMO

Legacy DC connector on older Japanese EVs, commonly limited to about 100 to 150 kW

Connected vs demand load

The sum of all nameplate ratings against what the site actually draws with diversity and tapering

Demand charge

A utility charge based on peak kW draw in a billing period, separate from the kWh energy charge

Load management

Sharing capped power across dispensers so the service and the demand peak stay below the connected total

Make-ready

The service, conduit, and pads built ahead of and beyond the first chargers so expansion is a connection

Power cabinet vs dispenser

In a split architecture, the central rectifier cabinet against the stall unit with the connector and cable

NEC Article 625

The NEC article governing EV supply equipment, including the continuous-load and disconnect rules

NEVI uptime

The federal corridor requirement that each port average over 97 percent uptime on a rolling 12-month basis

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is the difference between Level 1, Level 2, and DC fast charging?
• What size wire for a 48A charger?
• What is an intrusion alarm system?
• How does a septic system work?
• What is a sewage lift station?