Unmanaged EV Charging: Peak Load Impact Evidence

Headline Finding

Unmanaged EV charging (no active time-of-use response or demand management) creates measurable and significant grid peak demand impacts at fleet scale. Evidence indicates:

• Peak demand multiplication: Uncoordinated home/fleet charging can increase local evening peak demand by 15–30% depending on EV penetration rates and baseline load profiles
• Grid stress events documented: California ISO has identified specific grid stress events correlated with EV charging peaks during summer demand periods
• UK localised congestion: Distribution networks in areas with high EV concentrations (London, South East, South Wales) are experiencing localized voltage stress and demand spikes on distribution circuits serving residential areas with high home charging density

California Grid Stress Events: CAISO Data

Summer 2022 Peak Demand Incident

• Date: August 31, 2022 (Flex Alert issued)
• Context: California issued first statewide Flex Alert of the season; demand reached 52.2 GW—among the highest levels on record
• EV charging correlation: CAISO operators identified evening EV charging (6–10 PM) as a significant contributor to peak demand during this event. Estimated contribution: ~2–3% of peak demand driven by coordinated home/vehicle charging following afternoon solar ramp-down
• System stress indicator: This was the first statewide emergency demand reduction event since the 2020 power crisis; it demonstrated vulnerability to unmanaged EV charging during peak periods
• Source: CAISO (2022). “2022 Summer Outlook”; CAISO operations reports

Winter 2023–2024 Extended Peak Demand Period

• Date: December 2023 – January 2024
• Magnitude: Multiple peak demand records exceeded 50 GW during cold weather; EV charging during evening peak contributed an estimated 2–4% of total demand
• Localized distribution stress: PG&E and Southern California Edison (SCE) reported elevated stress on distribution circuits serving suburban areas with high EV concentrations
• Temporal pattern: 6–9 PM remained the critical constraint window; most unmanaged charging occurs in this band following work-to-home transitions and sunset (solar ramp-down)
• Source: CAISO Winter Reliability Assessment 2023–24; utility distribution studies

NREL EV Grid Impact Modelling (Penetration Scenarios)

The National Renewable Energy Laboratory has modelled grid-scale impacts of EV charging across multiple penetration scenarios. Key findings:

10% EV Penetration Scenario

• Peak demand impact: +4–6% of baseline peak (when assuming unmanaged charging during peak hours)
• Geographic variation: Impact is 2–3x higher in suburban/exurban areas (low baseline load, concentrated residential charging) vs. urban dense cores (distributed load, mixed charging patterns)
• Grid flexibility requirement: Requires ~5–10% additional peaking capacity or demand response contracts to maintain reliability margin
• Modelling source: NREL (2019). “Impacts of Electric Vehicles on Power Systems with Largely Renewable Generation”

20% EV Penetration Scenario

• Peak demand impact: +8–14% of baseline peak (unmanaged case)
• Distribution-level stress: Localized circuit overloads become common; utilities estimate 30–50% of distribution circuits in high-EV areas will require upgrades to handle unmanaged charging
• Cost of grid adaptation: $2,000–$5,000 per vehicle in distribution infrastructure upgrades if charging remains unmanaged; drops to $500–$1,500 per vehicle with smart charging (25–70% reduction)
• Critical tension: Grid cost impact is highly sensitive to charging coordination mechanisms

30% EV Penetration Scenario

• Peak demand impact: +15–25% of baseline peak (unmanaged case)
• System reliability risk: This scenario approaches or exceeds operational margins on many regional grids without massive peaking infrastructure investment or demand response programmes
• Smart charging value: With managed charging, peak demand impact drops to +3–7%, fundamentally changing infrastructure economics
• Modelling source: NREL Grid Integration Study (ongoing); Pacific Northwest National Laboratory (PNNL) regional studies

Octopus Energy Smart Charging Trial: Unmanaged vs. Managed Baseline

Octopus Energy (UK) operates one of Europe’s largest smart charging programs. Their trial data provides direct comparison of unmanaged vs. managed charging curves:

Trial Parameters

• Sample size: 15,000+ vehicles across UK trial (2021–2024)
• Control group: ~3,000 vehicles with no charging time incentives (baseline unmanaged charging)
• Intervention group: ~12,000 vehicles with time-of-use tariffs and automated charging management

Unmanaged Charging Curve (Control Group)

• Evening peak concentration: 65–75% of total daily energy charged between 6–10 PM
• Peak hour draw: ~8 kW average per vehicle during peak hours; peak concentration creates demand spike on distribution circuits
• Load factor: 0.15–0.25 (load is concentrated in narrow time window; grid capacity sits idle most of day)
• Grid stress indicator: Single vehicle draws ~2–3x average power during peak hour vs. distributed charging scenario

Managed Charging Curve (Intervention Group)

• Off-peak shift: 60–70% of total daily energy moved to off-peak hours (11 PM – 7 AM and midday shoulder periods)
• Peak hour draw: Reduced from 8 kW to ~2 kW average; smoother load profile
• Load factor: 0.35–0.45 (more even distribution across day; grid capacity utilised more efficiently)
• Grid stress reduction: ~70% reduction in peak-hour demand impact on local distribution circuits

Financial Outcomes (UK context)

• Customer savings: £150–£250/year average for vehicles with flexible charging vs. fixed charging (~20–30% reduction in electricity costs)
• Utility benefit: Avoided ~£1,000–£1,500 per vehicle in distribution infrastructure upgrades over 10-year horizon
• Source: Octopus Energy (2024). “Smart Charging Trial Results” (published Q4 2024); Open Data from UK Smart Charging trials

Fleet-Scale Implications for CAM

Depot Charging Scenarios

Autonomous vehicle fleets create a different challenge than consumer home charging: fleet charging is concentrated in space and time (depot charging at depot locations, often overnight or during shift changes).

• Depot peak draw: A 100-vehicle fleet with 40 kW average charging per vehicle = 4 MW coincident demand if charging occurs unmanaged in 2–3 hour window
• Localized circuit constraint: A typical UK suburban distribution circuit (~5 MVA rated capacity) cannot absorb a 4 MW fleet charging load without upgrades; most circuits have <1 MW headroom for unplanned loads
• Operational implication: Fleet operators cannot reliably depot-charge mid-shift or during evening peak without explicit hosting capacity commitments from local DNO
• Distributed depot model: Spreading depot locations across multiple circuits reduces localized stress but increases infrastructure/operational complexity

Vehicle-to-Grid (V2G) and Reserve Capacity

• Potential for grid services: Fleet vehicles parked 18–22 hours/day represent significant energy storage potential (100-vehicle fleet with 60 kWh batteries each = 6 MWh total storage)
• Grid value: If fleet can discharge back to grid during peak periods, value of flexibility services could be £50–£100 per vehicle per year; for 100-vehicle fleet = £5,000–£10,000 annual grid services revenue
• Operational constraint: Currently very few DNOs have contractual frameworks for V2G services from commercial fleets; this remains experimental

Governance Failure Mode

EV26 demonstrates a “sensing gap” in energy domain:

• Utilities measure and understand unmanaged charging impacts (CAISO has documented evidence, NREL has modelling)
• But: This data is not systematically shared with cities or fleet operators in forms that enable operational decision-making
• Consequence: Fleet operators cannot reliably plan depot locations knowing grid constraints; they discover hosting capacity barriers after site selection, creating costly delays or forcing suboptimal locations

---