A homeowner in Sacramento wants a heat pump. Her gas furnace is 19 years old, her utility bills hit $340 in January, and the 30 percent federal tax credit makes the switch feel overdue. She calls an HVAC contractor, who sends an electrician to assess the panel. He opens the box in her garage, counts breakers, sees 100-amp service, and delivers the verdict in under two minutes: she needs a panel upgrade before the heat pump can go in, $1,600 for the panel, another $2,200 if the utility has to pull new service from the transformer. She asks how he arrived at those numbers. He taps the panel door. "Standard practice."
He never ran a load calculation.
That 600,000 figure comes from PG&E's February 2026 announcement of its PanelBoost program with SPAN, and it represents a staggering infrastructure bottleneck: one in five homes in their Northern California service territory lacks the electrical capacity to add an EV charger, a heat pump, or both without modification. Nationally, half of all U.S. homes still run on 100-amp service, a capacity standard designed for houses that heated with gas, cooled with window units, and charged nothing more demanding than a vacuum cleaner.
But "needs an upgrade" and "needs a $3,800 panel swap" are not the same sentence, and the gap between them is where AI-powered assessment tools and smart load management devices are quietly rewriting the economics of home electrification.
One Photo, 269,000 Data Points
Qmerit, which has installed over 269,000 home EV chargers across North America, launched Panel Insights in late 2023: a computer vision system trained on its installation database that evaluates an electrical panel's capacity from a single smartphone photograph. Point your camera at the open panel door. Within seconds, the AI identifies available circuit spaces, counts tandem breakers, reads the main breaker rating, and flags the panel's manufacturer and model. That assessment feeds into LoadCRE, Qmerit's Load Capacity Recommendation Engine, which calculates whether the existing panel can absorb an EV charger, a heat pump water heater, or an induction cooktop without tripping the main breaker during peak demand.
Developed in collaboration with Schneider Electric's AI Hub team, the system sidesteps the crudest failure in residential electrical work: the rule-of-thumb assessment. Most electricians dispatched for pre-installation evaluations do not perform NEC-compliant load calculations. They eyeball the panel, see 100 amps, know the heat pump draws 30 to 50 amps at startup, and conclude the service is insufficient. ChargeRight, founded by IBEW Licensed Master Electrician Jason Walls, offers free assessments using the NEC Section 220.82 "Optional Method" calculation, and Walls estimates that 70 percent of homeowners told they need a panel upgrade do not, in fact, need one when the math is done properly.
That 70 percent claim deserves scrutiny, and I will return to it. But the underlying logic is sound: NEC 220.82 permits demand-factor adjustments that recognize a home rarely runs every load simultaneously, where a dryer, an oven, and a central air conditioner firing at once is the exception that happens perhaps three or four days per year, not the assumption that should drive permanent infrastructure decisions costing thousands of dollars.
What the NEC 2026 Actually Changed
For decades, load calculations have been the domain of a formula that assumes the worst. You list every appliance, assign each its nameplate wattage, apply modest demand factors for lighting and small appliances, add them up, and divide by 240 volts to get your required amperage. A conservative method and a necessary one, designed for an era when the inspector filling out the permit had no way to know what the homeowner would plug in next year.
NEC 2026, published by the NFPA in late 2025 with adoption rolling through state and local jurisdictions now, introduced a structural shift in Chapter 1. For the first time, the code formally integrates Power Control Systems (PCS) and Energy Management Systems (EMS) into load calculation methodology. According to Schneider Electric's technical overview, design professionals can now reference PCS settings when sizing service entrance equipment, recognizing that a software-managed home that dynamically throttles loads in real time does not require the same headroom as a dumb panel where everything runs unchecked.
Put concretely: if your panel has a certified EMS that will automatically reduce your EV charger from 48 amps to 16 amps whenever the dryer kicks on, the code no longer demands that you size your service as though both are drawing full power at once. It is a recognition, belated but consequential, that the electrical panel of 2026 is not the electrical panel of 1996, and the sizing rules should reflect what the hardware actually does.
SPAN's Bet: Bypass the Upgrade Entirely
SPAN, the smart panel company founded by former Tesla engineer Arch Rao, has built its entire product strategy around this gap. Its PowerUp technology, a certified EMS, continuously monitors the current on incoming service conductors and automatically manages individual circuits to keep total draw below the panel's rated capacity. When your heat pump compressor cycles on while the EV charger is running at 48 amps, PowerUp can temporarily throttle the charger to 16 amps for the 20-minute compressor cycle, then ramp it back up. Brief. Automatic. Invisible to anyone not staring at the SPAN app. Partners include Mitsubishi Electric Trane HVAC, SMA for solar inverters, and FranklinWH for battery storage, enabling coordinated load orchestration across the highest-draw systems in the home.
"Adding a marginal appliance that triggers an expensive electrical service upgrade is neither cost-effective nor scalable if every single home has to undergo this open-heart surgery," Rao told Solar Builder Magazine. SPAN's installed cost runs approximately $4,500, which replaces the existing panel entirely rather than supplementing it, meaning the homeowner gets both the smart load management and the panel hardware in one installation.
Then there is SPAN Edge, announced in the PG&E partnership. Unlike the full SPAN panel that replaces your breaker box, Edge is an at-the-meter device. It sits between your utility meter and your existing panel, monitoring incoming current and managing loads without requiring a panel replacement at all. PG&E is deploying Edge through PanelBoost as a lower-cost on-ramp for the hundreds of thousands of homes that need capacity headroom but do not need, or want, a $4,500 smart panel.
Running the Numbers Yourself
I modeled a common electrification scenario for a 1,800-square-foot California home built in 1995: 100-amp service, gas furnace, gas water heater, electric dryer, central air conditioning, no EV. The homeowner wants to add a 48-amp Level 2 EV charger and replace the gas furnace with a heat pump.
| Path | Cost | Timeline |
|---|---|---|
| Traditional: panel upgrade to 200A | $1,600 panel + $2,200 service upgrade + permit fees = $4,100–$5,500 | 3–6 months (utility coordination) |
| SPAN Panel (PowerUp) | $4,500 (replaces panel, no service upgrade) | 1–2 weeks |
| SPAN Edge (at meter) | Pricing TBD (positioned below $4,500) | Days, via utility program |
| NEC 220.82 recalculation only | $150–$300 (electrician's load calc) | Same day |
Option four is the one nobody talks about. A proper NEC 220.82 Optional Method calculation on this house, using demand factors for general lighting (first 10 kVA at 100%, remainder at 40%), applying the heating-versus-cooling non-coincident load rule where only the larger of the two counts, and factoring in that the gas water heater draws no electrical load, often shows existing 100-amp service can absorb a 40-amp EV circuit without modification. No new panel. No utility truck. No three-month wait for a service drop. Just an electrician who knows which section of the code to apply and a permit office willing to accept the calculation.
When it works, it saves the homeowner $3,800 to $5,200 and months of scheduling. When it does not work, the SPAN route still avoids the utility coordination timeline that kills more electrification projects than cost alone.
Why This Is Not Happening at Scale
Contractor education. Full stop.
A SEPA report published in January 2026, authored in collaboration with Schneider Electric and SPAN, identified the core problem: "Scaling smart panels requires aligned incentives, consistent program design, and broad customer and contractor education." Most licensed electricians learned load calculations on the NEC 220.82 Standard Method, which produces higher capacity requirements by design because it applies fewer demand factors. Many do not routinely use the Optional Method because their training never emphasized it, because the Standard Method is simpler to defend to inspectors, and because recommending a panel upgrade guarantees a larger job with higher margins.
I am not suggesting corruption. Recommending an upgrade is the safe call; nobody gets sued because they oversized a panel. But the cumulative effect is hundreds of thousands of homeowners who hear "$4,000 before we can even start on the heat pump" and walk away from electrification entirely, not because the grid cannot handle it, but because nobody ran the calculation that proves it can.
Strongest Case Against
Smart panels are proprietary technology with a finite corporate lifespan. SPAN was founded in 2018, has raised approximately $125 million, and is not yet publicly reporting profitability. If SPAN fails, your $4,500 intelligent panel becomes a dumb breaker box with a touchscreen that no longer receives updates. Traditional panel upgrades use commodity hardware from Eaton, Square D, and Siemens with 40-year operational track records and replacement parts available at any electrical supply house in America. A 200-amp Square D Homeline will outlast every software company currently pitching smart energy management.
Load management also introduces real constraints. PowerUp's circuit throttling means your EV charger will sometimes deliver 16 amps instead of 48. On most nights that difference is invisible because the car charges for eight hours regardless. On the night you arrive home at 11 PM with a 15 percent battery and need to leave at 6 AM, the throttled charge rate delivers roughly 60 miles of range instead of 180, and you either accept the reduced range or manually override the load management from your phone, which may trip the main breaker if multiple loads are running. Edge cases are not the same as no cases, and the homeowner who discovers this constraint at midnight, in winter, is not going to appreciate the demand-factor math that made it possible.
What It Means If You Are Electrifying This Year
Before accepting any panel upgrade quote, ask your electrician to perform an NEC 220.82 Optional Method load calculation on your home. Not a rule-of-thumb assessment, not a glance at the breaker count, but the actual calculation with demand factors applied. If they cannot or will not, find an electrician who can, and Qmerit's Panel Insights AI provides a free initial assessment from a smartphone photo that gives you a baseline before the electrician arrives.
If the calculation shows your existing service is genuinely insufficient, compare the full panel upgrade path (cost plus utility coordination timeline) against a SPAN panel or SPAN Edge installation. In PG&E territory, the PanelBoost program may subsidize the Edge device. Inquire directly with PG&E; the program is active but details on customer eligibility and cost-sharing are still emerging.
If you are building new construction, specify at least 200-amp service and wire a dedicated 50-amp circuit to the garage whether or not you own an EV today. A $200 circuit during construction avoids a $1,500 retrofit later. Consider pre-wiring for a heat pump water heater as well, since the incremental cost during framing is negligible compared to cutting into finished walls.
Limitations
ChargeRight's "70 percent" figure is a practitioner's estimate, not an independently verified study. No peer-reviewed data set confirms what percentage of 100-amp homes can absorb electrification loads without modification under the NEC Optional Method. SPAN's PowerUp technology is certified as an EMS but has not published long-term reliability data across diverse climates, particularly in regions with sustained sub-zero temperatures where heat pump compressor loads peak and EV charging efficiency drops simultaneously. NEC 2026 has been published but not adopted by most local jurisdictions; permitting offices typically lag the national code by one to three years, meaning a homeowner's inspector may not accept PCS-adjusted load calculations even if the math is valid. PG&E's PanelBoost is a pilot-stage program with no published enrollment numbers, subsidy amounts, or scaling timeline. SPAN Edge pricing has not been publicly disclosed.