Fleet operators managing a single depot in a single utility territory have a complex enough optimization problem: one tariff structure, one demand charge account, one EVSE system, one set of vehicles. Add a second depot — potentially in a different city, a different utility territory, with different EVSE hardware and a different operational profile — and the complexity doesn't merely double. The interdependencies multiply, the data infrastructure becomes more demanding, and the management overhead grows in ways that are hard to anticipate before you're in the middle of it.
The upside is real: larger fleet operations with multiple depots have more total optimization surface area. The aggregate demand charge exposure across three or four depots is larger in absolute terms, and a well-coordinated multi-site system can capture savings that a single-site approach can't access — including load balancing between sites for vehicles that have flexibility in where they return, and consolidated demand response program participation that single-site fleets may not qualify for.
This article covers the specific challenges and opportunities of multi-site fleet depot load management, and what a scheduling platform needs to handle to do it well.
The Utility Zone Problem
Each depot sits in a utility service territory with its own tariff structure. For a Pacific Northwest fleet operator with depots across the region, this commonly means a mix of PGE service (Portland metro), Pacific Power service (eastern Oregon, southern Oregon), Clark PUD service (Vancouver, WA), and potentially other PUDs or IOUs depending on the specific locations.
As covered in our article on Pacific Northwest utility rate structures, these territories have materially different commercial rate structures. PGE's Schedule 32 is a two-period TOU (on-peak / off-peak) with a specific demand ratchet. Pacific Power's Schedule 47 has a three-period TOU structure. Clark PUD has lower base rates with different demand charge mechanics.
A multi-site scheduling platform must maintain separate tariff objects per depot — not a single "utility rate" parameter applied uniformly. The optimization logic running at each depot must reference that depot's tariff, with the correct peak window hours, correct demand charge parameters, and correct ratchet tracking for the specific billing account at that facility. Any cross-site coordination logic (discussed below) must be aware that a scheduling decision that's cost-optimal at a PGE depot may have different implications at a Pacific Power depot.
Demand Account Isolation
Each depot has its own utility account, and each account's demand charge is measured independently. This has an important implication: demand reduction at one site does not affect the demand charge at another site. The demand charge problem must be solved at each site individually.
However, demand account isolation also means that the demand ratchet at one site doesn't contaminate other sites. A mismanaged charging event that triggers a high demand peak at Depot A in January doesn't affect the demand baseline at Depot B. For multi-site operators managing a rolling transition where some depots are deploying EVSE before others, this isolation is genuinely helpful — the fleet can learn from the first depot's experience and improve demand management before it affects other sites.
The operational discipline implication is that multi-site operations need per-site demand charge tracking, not just aggregate energy cost tracking. A dashboard that shows total fleet charging spend across all sites is useful but obscures whether individual sites are accumulating demand charge exposure from poor scheduling. Each site needs its own demand curve monitoring, its own ratchet tracking, and its own alert thresholds.
Cross-Site Vehicle Routing Flexibility
Some fleet operations — regional delivery networks, multi-terminal transit operations, utility maintenance fleets — have vehicles that can return to different depots on different days based on route assignment. When this operational flexibility exists, it creates a charging optimization opportunity that single-site systems can't capture.
Consider a scenario: Depot A (Portland, PGE territory) is approaching its monthly demand peak for the billing period. Several vehicles that would normally return to Depot A tomorrow have routes that could terminate at Depot B (Gresham, still PGE territory) without significant operational inconvenience. If the scheduling system can communicate with the route planning system and flag that Depot A has elevated demand charge risk for the next few days, routing decisions can be adjusted to distribute the charging load.
This kind of cross-site coordination requires integration between the charging management system and the dispatch/routing system — an integration point that doesn't exist in most fleet software architectures today. Where it does exist, the value can be meaningful: shifting 15–20% of a high-demand depot's charging volume to a lower-demand site during a critical billing period can materially change the monthly demand charge outcome.
We're not saying this integration is straightforward or available off-the-shelf. It isn't, for most fleet operators today. What we're saying is that it represents the kind of optimization opportunity that becomes relevant as multi-site fleet management matures, and that scheduling systems should be architected with this integration point in mind even if it isn't immediately implemented.
Consolidated Demand Response Participation
Utility demand response programs — programs where commercial customers receive payments or bill credits for reducing load during utility-dispatched curtailment events — often have minimum load reduction thresholds for enrollment. A single small depot with 50 kW of interruptible charging load may fall below the threshold for program enrollment. A multi-depot operator whose aggregate interruptible load across three sites totals 250 kW may qualify for programs that each individual site could not access.
In Oregon, PGE's demand response programs for commercial customers have included both direct load control and customer-managed load reduction programs. The specific available programs, capacity requirements, and payment structures vary by program and are subject to change through PUC proceedings. Fleet operators interested in demand response program participation should work with their utility key account manager and review current program terms.
What the multi-site aggregation enables is meeting participation thresholds that require dispatch across multiple sites simultaneously. When a utility dispatches a demand response event, the scheduling system must curtail charging across all enrolled depots within the required response window — typically 10–30 minutes from dispatch notification. This requires centralized coordination: the scheduling system needs to know the total enrolled curtailable load across sites, calculate a per-site curtailment allocation that satisfies the program commitment while respecting each site's operational constraints (morning dispatch readiness), and execute the OCPP commands across all sites in parallel.
EVSE Hardware Heterogeneity
Multi-site fleets acquired their EVSE over time, often from different vendors at different sites. The result is a common pattern: one depot has ChargePoint hardware running OCPP 1.6, another has BTC Power units also on OCPP 1.6 but with different smart charging profile support, and the newest depot has ABB Terra hardware running OCPP 2.0.1. Each charger network uses a different vendor CSMS (Charge Station Management System) with different API access models for third-party integration.
This is the integration complexity that makes multi-site fleet management genuinely hard. The scheduling optimization layer needs to send coherent charging commands across all these systems. The standard path is through each CSMS's API — which means maintaining separate integrations for each vendor's management system, each with different authentication flows, different data models, and different rate limiting policies.
The alternative is direct OCPP connectivity: if each charger is directly accessible via OCPP WebSocket, the scheduling platform can connect directly without going through vendor CSMS APIs. This requires network access to the charger endpoints, which may require coordination with IT and site network administrators, but reduces the dependency on vendor-specific API contracts that can change without notice.
Reporting and Visibility Across Sites
Multi-site operators need consolidated reporting that makes cross-site comparison meaningful without obscuring site-specific issues. The useful multi-site reporting dimensions:
- Per-site demand charge exposure: Monthly demand peak recorded at each site vs. the prior 11-month high (ratchet tracking)
- Off-peak charging percentage per site: What fraction of kWh delivered at each site fell in the off-peak window — a quick indicator of scheduling quality
- Demand response performance per site: For enrolled programs, actual load reduction delivered vs. committed reduction at each curtailment event
- Cost per kWh delivered (fully burdened) per site: Total utility charges (energy + demand + fees) divided by total kWh delivered — allows cross-site comparison that accounts for different tariff structures
The cost-per-kWh comparison across sites is particularly useful for identifying which sites have the most optimization upside. A site with a high fully-burdened cost per kWh is either on an unfavorable tariff structure, running poor demand management, or both. That metric flags where to focus improvement effort across a portfolio of depots.
Multi-site fleet charging is where the operational discipline of scheduling management meets the complexity of operating in multiple regulatory and tariff environments simultaneously. The tools are available; the challenge is integrating them coherently across a diverse infrastructure landscape.