The Automation Core

Megawatt Truck Charging Explained: Tesla's 1.2 MW Megacharger, 1.5 MW MCS Hubs and the Substation Problem

· CharIN, electric trucks, heavy-duty EV charging, liquid-cooled charging cable, MCS, medium-voltage grid connection, megawatt charging, Megawatt Charging System, SAE J3271, Tesla Megacharger, Tesla Semi,

 


Bottom line: The real obstacle for battery-electric trucks has never been just the battery. It is pushing several hundred kilowatt-hours into a truck during a coffee break. In the first week of October, the first public megawatt-class truck charging sites opened in the US and France. Run the numbers and a truck charging site turns out to be a small substation with plugs. Four takeaways follow. Figures come from operator releases and standards bodies; the conversions are my own illustrative arithmetic, not measurements.

What happened

  • Oct 8, Pilot and Tesla: the first Tesla Megacharger site opened at Pilot’s Ellabell, Georgia travel center with six heavy-duty stalls. Twelve more sites are under construction; Tulare, California was set to open Oct 9, with Fort Stockton, Texas and North Las Vegas next. Each site gets four to eight stalls on Tesla’s V4 cabinet hardware, up to 1.2 MW per post, which Tesla says restores up to 60% of range in 30 minutes.
  • Prefab hardware: since September Tesla has shipped factory-assembled Megacharger units, one V4 cabinet feeding two posts, to cut site work.
  • Context: Tesla formally launched high-volume Semi production in Sparks, Nevada on Sept 24, in a plant designed for 50,000 trucks a year. Published specs: about 1.7 kWh per mile, 325 miles (Standard) and 500 miles (Long Range).
  • Oct 2, Voltix (France): four 1.5 MW charge points opened at Vierzon, where the A20, A71 and A85 meet, each with both CCS and MCS (Megawatt Charging System) connectors and a plan to grow to eight. A 30 to 45 minute charge lines up with the EU rule of a 45-minute break after 4.5 hours of driving.

Scope check: our Sept 22 piece covered the MAN eTGX truck and its MCS inlet. This one is about the infrastructure side: how the cable sheds heat, how much power a site pulls from the grid, and whether the energy math works.

Engineering point 1: more power means more current

CharIN recommends an MCS operating window of 400 to 1,250 V, with continuous current validated to 3,000 A, for a ceiling of about 1,250 × 3,000 = 3.75 MW. Why not just raise the voltage? Pack voltage, creepage and clearance distances, and power semiconductor ratings already sit around the 1,000 V class, which is exactly what 800 V silicon carbide platforms are wrestling with. At an assumed ~1,000 V, 1.2 MW is about 1,200 A. A 350 kW car charger on an 800 V platform is roughly 440 A. Almost all of the extra truck power is current. Note: Tesla has not published the Semi’s pack voltage or charge current, so the 1,000 V figure is my assumption.

Engineering point 2: triple the current, nine times the heat

Heat in conductors and contacts is I²R. Assume 1 mΩ total for cable plus contacts (illustrative only): at 1,200 A that is about 1.4 kW of heat; at 3,000 A it is about 9 kW, a space heater inside the handle. That is why the standard is tiered. Roughly 350 to 500 A can run without active cooling; around 1,000 to 1,500 A needs a liquid-cooled connector; at 3,000 A the vehicle inlet must be cooled too. INL’s interoperability testing caps contact temperature at 100°C, and CharIN recommends cables no longer than 15 m, and as short as practical, because cooled cables are heavy and stiff.

Engineering point 3: the site is a substation

Ellabell is 6 × 1.2 MW = 7.2 MW nameplate. Voltix is 4 × 1.5 = 6 MW today and 12 MW at eight points. That is small-factory load, far beyond a low-voltage service, so each site needs a dedicated medium-voltage connection, transformers and switchgear, then rectification to high-voltage DC. Not every post will run flat out at once, but the utility has to plan for it, and interconnection often takes longer than installing chargers: Pilot originally targeted summer for its first site and opened in October. My inference is that two things become important next. One is on-site battery buffers to shave peaks and demand charges. The other is solid-state transformers (SSTs) that convert medium voltage straight to DC and drop the bulky line-frequency transformer, the same logic behind 800 VDC power in AI data centers.

Engineering point 4: does the energy math close?

Use Tesla’s 1.7 kWh/mi. An EU driver doing 4.5 hours at an average 80 km/h covers about 360 km (224 miles) and uses about 380 kWh. A 45-minute break at an average 1 MW delivers about 750 kWh, so on paper there is plenty of margin. Now flip Tesla’s “60% in 30 minutes.” If the Long Range pack really is 822 kWh, 60% is about 490 kWh, an average close to 1 MW, meaning the curve barely tapers over most of the window. That is a demanding cell-cooling target. One big caveat: Electrek notes that Tesla did not restate pack size after switching to 4680 cells, so treat 822 kWh and that average as an upper-bound estimate.

Open questions

  • Tesla has not published the Semi’s pack voltage, peak charge current or full charging curve, and has not restated pack capacity after the 4680 switch.
  • These releases do not say whether Tesla Megacharger posts are MCS-interoperable.
  • Grid connection capacity, transformer ratings, any battery buffering and build costs are undisclosed for both sites; Voltix explicitly did not share cost.
  • Tesla’s year-end goal of 30+ sites and 200+ megawatt-capable posts is a target, not a delivered number.
  • The 1,000 V, 1 mΩ and 80 km/h figures above are illustrative assumptions; real values vary by truck and site.

Related videos

Quick poll: where do electric trucks get stuck first? Cable cooling, grid capacity at the site, or drivers who simply will not wait 45 minutes? And if a highway rest area needs a 6 MW connection, who should pay for the line: the utility, the charging operator or the fleets? Tell me in the comments.

← Newer articles Older articles →