Charge More Electric Trucks Without Grid Upgrades

A single megawatt, shifted a few hours later, can be the difference between a fleet electrifying this year—or sitting in an interconnection queue until 2027.

That’s the uncomfortable truth many utilities and fleet operators are running into as 2025 closes: demand is rising, distribution capacity isn’t evenly available, and traditional “build-first” grid planning often treats a new depot as if it will hit its maximum load at the worst possible hour. The result is predictable: denied service requests, expensive feeder upgrades, and delayed emissions reductions.

There’s a better way to approach this. Flexible service connections—paired with AI-driven load shifting and feeder monitoring—let utilities connect truck depots faster by contracting for when not to charge, rather than forcing everyone to pay for upgrades sized to rare peak hours. Northern California’s experience shows the scale: across roughly 2,000 feeders, managed flexibility could support 25,000–70,000 additional heavy-duty trucks per year without traditional grid upgrades.

Why truck depot interconnection gets stuck (and why it’s fixable)

The core problem is peak coincidence. Distribution planning commonly checks two extremes at once: the feeder’s highest-load hour of the year and the site’s theoretical maximum demand. If those line up on paper, the request can be delayed until upgrades are built—even if that coincidence would only happen for a handful of hours and the fleet isn’t actually planning to charge then.

This matters because heavy-duty depot charging is lumpy: multiple high-power chargers turning on together can create a big step change in load. From a utility risk perspective, conservative rules feel safe.

But from a grid modernization perspective, that conservatism often leaves real capacity stranded.

The myth: “If the grid is constrained, upgrades are the only answer”

Upgrades are sometimes necessary. Still, a large share of “constraints” are timing constraints, not energy constraints. Many feeders have headroom most of the day and most of the year, then get tight during a narrow peak window—often summer evenings (commonly 4–9 p.m.) when solar drops and residential load rises.

If a fleet can avoid those hours (or reduce charging during them), the feeder can host new load immediately.

Flexible service connections: what they are in plain terms

A flexible service connection is a utility-approved interconnection agreement that provides faster access to power in exchange for temporary, targeted load curtailment during feeder peak constraints.

Instead of “you can’t connect until we upgrade,” the utility says: “you can connect now, but on a limited set of peak days/hours, you’ll reduce charging load to stay under the feeder’s safe operating limit.”

For fleets, the value is straightforward:

• Faster energization for depots (months sooner can mean hitting procurement and compliance timelines)
• Lower make-ready and upgrade costs
• More predictable deployment across multiple sites

For utilities, it’s equally practical:

• More customers on existing infrastructure (rate base growth without immediate capex)
• Deferred upgrades that can be timed and targeted better
• A concrete step toward distribution automation and real-time operations

Flexible service connections also fit neatly into the “AI for Energy & Utilities: Grid Modernization” narrative: they depend on the same capabilities utilities are building anyway—forecasting, monitoring, optimization, and automated control.

What the Northern California numbers say (and why they’re credible)

Analysis of feeder constraints and representative truck charging profiles in PG&E territory shows how small operational shifts unlock disproportionate capacity.

One key finding: if a new depot shifts about 1 MW away from peak hours by several hours—only on the limited days required—most constrained feeders could support roughly 10 additional heavy-duty trucks compared to business-as-usual planning.

Scale that across ~2,000 feeders and the technical potential becomes large:

• 25,000–70,000 additional heavy-duty trucks per year (depending on duty cycle) using flexibility instead of upgrades
• With broader flexibility windows, the “extra trucks per feeder” rises sharply

Flexibility windows change the math fast

The practical insight here is simple: the more time you have to move charging, the more trucks you can serve.

When the flexibility window expands from 2 hours to 6 hours, modeled outcomes for the median feeder increase dramatically:

• Regional heavy-duty trucks: +3 to +22 additional trucks per feeder
• Urban heavy-duty trucks: +9 to +64 additional trucks per feeder

Across the full territory, that translates into ranges like:

• 8,000 to 41,000 extra regional trucks
• 22,000 to 112,000 extra urban trucks

Even if real-world constraints and reserve margins reduce those figures, the direction holds: timing flexibility is a capacity resource.

Can fleets actually shift charging without breaking operations?

Yes—most fleets already have the operational “slack” needed. The common fear is that any curtailment requirement will strand trucks uncharged and late to dispatch. In practice, depot dwell time is often long enough that charging can be re-timed.

Average behavior patterns for many heavy-duty operations show:

• Trucks with daily mileage under ~300 miles often return to depot and remain parked roughly 4 p.m. to 5 a.m.
• Charging needs may require less than half of that parked time at around 75 kW (and many fleets deploy higher power)

So even if a utility asks a depot to avoid (say) 4–9 p.m. on a subset of summer days, there’s typically plenty of night-time runway to complete charging.

Managed charging is the real enabler (and it’s software-first)

Here’s what works in the field: don’t treat charging as “plug in and pray.” Treat it as a scheduling problem.

A basic managed charging system can:

• Prioritize vehicles by next departure time and required state of charge
• Throttle or stagger chargers during constraint hours
• Automatically resume charging once constraints clear

This is where AI-driven load shifting becomes concrete. You don’t need a sci-fi control room. You need algorithms that do three things well:

1. Forecast depot demand (what energy each truck needs by when)
2. Optimize charging schedules under constraints (feeder limits, demand charges, TOU rates)
3. Execute and verify (telemetry, charger control, exception handling)

Fleets that already shift for time-of-use rates are halfway there. Many can also reduce electricity costs by up to ~30% through smarter charging strategies—meaning flexible connections often stack speed-to-power benefits on top of bill savings.

What utilities need to operationalize flexible connections at scale

Flexible connections don’t work on trust alone; they work on measurement, forecasting, and enforceable signals. The best programs combine contractual clarity with modern grid monitoring.

The minimum viable “Flex” program

If you’re a utility leader thinking about scaling beyond pilots, focus on four building blocks:

1. Feeder visibility

   • Identify bottleneck feeders and the hours they bind
   • Use AMI, SCADA, line sensors, and state estimation where available

2. Constraint forecasting

   • Day-ahead (or seasonal) predictions of when feeder capacity will be tight
   • Weather-driven load forecasting plus DER and solar shape awareness

3. Customer-facing dispatch signals

   • Clear, machine-readable curtailment windows (day-ahead is strong; intra-day is stronger)
   • A backstop method for emergencies

4. Measurement & verification (M&V)

   • Prove compliance with submetering, charger telemetry, or interval data
   • Make settlement and penalties predictable (and fair)

A flexible service connection is only as good as the grid data behind it.

Start where it’s easiest to win

Not every feeder will benefit equally. Some will be constrained many months; others only a few summer evenings. Prioritization should be explicit:

• Target feeders where constraints occur rarely (high “capacity unlocked per hour curtailed”)
• Prefer customers with long depot dwell time and many chargers (more flexibility)
• Build early case studies to prove cycle time improvements

PG&E’s program has already shown real-world alignment with modeled results—one example often referenced in industry conversations is a fleet adding about 20 additional electric trucks under a flexible approach.

Where AI fits: from “managed charging” to grid orchestration

Flexible service connections are a policy and program design move. AI is the scaling engine.

At small scale, a utility can email curtailment notices and a fleet manager can manually respond. At hundreds of depots, that breaks.

AI and analytics help in three practical ways:

1) Better distribution capacity models

Feeder hosting capacity is not static. It changes with temperature, local DER output, equipment limits, and load behavior. AI models can improve:

• Dynamic hosting capacity estimates
• Detection of abnormal loading patterns
• Faster screening for new depot applications

2) Automated peak avoidance that still meets fleet KPIs

Optimization engines can balance competing objectives:

• Ensure minimum state of charge by dispatch
• Minimize demand charges and peak kW
• Comply with flexible connection constraints
• Prefer charging during high renewable output (when feasible)

3) Utility-to-fleet coordination that feels boring (in a good way)

The goal isn’t fancy dashboards. It’s dependable operations:

• APIs for dispatch signals
• Exception workflows (late returns, failed chargers, unexpected routes)
• Audit trails for regulators and internal governance

When this works, it’s invisible. Trucks leave on time. Feeders stay within limits. Everyone stops arguing about who caused the peak.

Practical playbook: what fleets and utilities should do next

If you’re a fleet operator:

1. Quantify dwell time and required energy by route class (urban vs regional)
2. Install charger-level telemetry (or verify your vendor exposes it)
3. Adopt managed charging with priority rules (depart time first, then cost)
4. Ask your utility about flexible service options and the specific curtailment mechanics
5. Plan a resilience backstop: spare charging capacity, limited on-site storage, or operational buffers for high-stress days

If you’re a utility:

1. Create a standard flexible connection tariff (pilot terms that are easy to replicate)
2. Publish feeder constraint windows (even coarse seasonal guidance helps developers site projects)
3. Invest in feeder monitoring on bottlenecks before you invest in upgrades everywhere
4. Integrate flexible loads into distribution planning so planners can count verified flexibility as a resource
5. Design for enforcement that doesn’t feel punitive: clear thresholds, transparent data, predictable settlement

The bigger picture for grid modernization in 2026

Flexible service connections should become normal in utility toolkits—right alongside demand response, time-varying rates, and hosting capacity maps. The reason is simple: the distribution grid can’t be rebuilt fast enough to match electrification timelines if every new load triggers worst-case upgrades.

For the “AI for Energy & Utilities: Grid Modernization” series, this is a clean example of what modernization actually looks like: not futuristic promises, but measurable capacity gains from better forecasting, better agreements, and better control software.

If your organization is planning electric truck infrastructure for 2026–2028, here’s the question worth sitting with: Are you designing your interconnection strategy around rare peak hours—or around the flexibility you can prove and automate?