BMS + EV Charging: What You Can Integrate (and What’s Not Worth It)

A New York property manager called us last month with a familiar problem. They’d dropped twelve Level 2 chargers into a Manhattan office building’s garage, the chargers worked, drivers were happy, and the first electric bill came back about $4,200 higher than projected. The chargers had set a new demand peak on a single Tuesday afternoon. That single peak then governed the demand charge for the rest of the month.

This is the situation BMS integration with EV charging is supposed to prevent — and the reason building owners keep asking us which integrations actually do something useful and which are just things a vendor put in their pitch deck.

Short answer: integrate the parts that touch your electricity bill or your compliance reports. Skip the rest.

BMS integration with EV charging: what’s worth integrating

Load management at the panel. Not glamorous. Probably the single most useful thing a BMS can do with chargers. Most commercial panels in older NYC buildings were never sized for four to twenty Level 2 chargers running at the same time. Adding even four 32-amp chargers naively can push a site past its service capacity. A load-managed integration — chargers reading panel headroom in real time and stepping down their output — keeps you off the service-upgrade carousel. ConEd service upgrades in dense parts of the five boroughs aren’t just expensive. They can take eighteen months.

Demand-charge coordination. If you operate in ConEd territory, demand delivery charges can run 30 to 70 percent of a commercial electric bill, and during summer they often dwarf the kWh charges entirely. Your BMS already knows when HVAC compressors are about to kick on, when elevators cycle, when the chiller plant ramps. Feeding that data into the charger network — usually through an OCPP-to-Modbus gateway like Intesis or HMS — lets the chargers throttle right before the building hits a new peak. We’ve seen this kind of integration pay back in under two billing cycles on properties enrolled in Rider Q.

Sub-metering and reporting. Local Law 88 already requires sub-meters in many qualifying tenant spaces. Local Law 97 requires you to report total emissions, and EV load counts toward that number. If your chargers feed kWh data into the BMS through OCPP MeterValues, you have a clean audit trail for both — without standing up a separate spreadsheet workflow at the property manager’s desk. NYC Accelerator will also want to see this kind of data during a good-faith compliance review.

Alarm forwarding. Charger offline. Ground fault on station 6. Door tamper on a curbside unit. Push those alerts to the facility dashboard. Maintenance is already staring at that screen all day. The alternative is another vendor portal — every charger network ships with one — and in practice those get logged into maybe twice a year, usually after something already failed.

Coordination with on-site storage or solar. If the building has BESS — and post-LL97, more of them do — the BMS can route stored energy to chargers during the four-hour Rider Q super-peak window. ConEd’s SmartCharge Commercial program also pays performance incentives for peak avoidance, and those incentives are easier to capture when the BMS is the one making the throttle decisions automatically rather than someone clicking around in a charger vendor’s web app.

BMS integration with EV charging: What isn’t worth your money

Granular per-session control from the BMS itself. OCPP charging station management systems already handle authorization, billing, RFID, reservations, payment processing. Asking a BACnet or Modbus BMS to replicate that work is reinventing a wheel that’s been standardized since OCPP 1.6. Use the gateway’s “BMS as Central System” mode only if you have no CSMS at all, which is rare in commercial deployments. For most NYC sites, run a real CSMS and let the BMS see metered data, status, and smart-charging setpoints — that’s enough.

Tying chargers tightly to HVAC sequencing. Academic papers love this. In practice, on a Manhattan office tower, the savings from coordinating chiller loads with charger loads are small relative to the engineering cost. HVAC and EV charging operate on different time scales — an air handler responds in seconds, a charge session lasts hours — and the orchestration overhead rarely pays back outside a campus environment with central plant economics.

DCFC plus BMS smart charging. Public DC fast-charging load is essentially inelastic. A driver who pulls into a 150 kW station expects 150 kW. Trying to manage that load through a building system to flatten demand is a losing fight — and the major EV charging trade groups have argued exactly that point to the New York Public Service Commission. If your site leans heavily on DCFC, the real answer is on-site storage or a higher service tier, not BMS finesse.

Full bidirectional BACnet for every charger object. Tempting on paper. Expensive in commissioning. Most operators who go this route end up using maybe ten percent of the available data points and still call the charger vendor when something breaks. Monitoring-only Modbus or BACnet, plus OCPP for control, covers ninety-five percent of real cases.

A note on the law side

Local Law 55 of 2024 is going to push more EV-ready parking spaces into NYC garages — 20 percent installed and 40 percent capable by January 2035 for licensed lots, with the DOB rules expected by January 2027. Buildings hitting LL97 caps have to count EV charging load against their emissions budget, even though the city’s grid is greening fast enough that the long-term math should improve. Owners planning capital projects right now have leverage. Once the 2027 rules drop and everyone is competing for the same ConEd make-ready windows and the same EV-qualified electricians, costs climb.

How to scope it

Start with a one-line sanity check: what is this integration actually trying to prevent or capture? If the answer is a demand charge, a service upgrade, an LL97 reporting headache, or a maintenance blind spot — green light. If the answer is “the vendor said it was a good idea” — keep walking.

The good integrations earn their commissioning cost back inside a year. The bad ones sit in the BMS as orphan objects that nobody touches until the next system upgrade, when somebody finally asks why they’re there.