Demand-Side Flexibility#

PyPSA-GB supports three integrated demand-side flexibility mechanisms that allow demand to shift in time or reduce during peak periods. These are modelled using PyPSA’s native components (stores, links, generators) and can be enabled independently or together.

The methodology is built upon this work:

Franken, L., Hackett, A., Lizana, J., Riepin, I., Jenkinson, R., Lyden, A., Yu, L. and Friedrich, D., 2025. Power system benefits of simultaneous domestic transport and heating demand flexibility in Great Britain’s energy transition. Applied Energy, 377, p.124522. doi:10.1016/j.apenergy.2024.124522

Overview#

Flexibility Type	Mechanism	PyPSA Model	Duration
Heat Pumps	Pre-heat water tanks or building fabric	Store + Link + Load	Hours
Electric Vehicles	Smart charging and Vehicle-to-Grid	Store + Link + Load	Hours
Event Response	Saving Sessions style load reduction	Generator (negative load)	Minutes-hours

All three are controlled via the demand_flexibility section of the configuration and orchestrated through the finalize_demand Snakemake rule.

How It Works#

        flowchart TB
    subgraph Disaggregation["Demand Disaggregation"]
        HP_DIS["Disaggregate Heat Pumps"]
        EV_DIS["Disaggregate EVs"]
    end

    subgraph Profiles["Flexibility Profiles"]
        COP["COP Profiles (Atlite)"]
        EV_AVAIL["EV Availability"]
    end

    subgraph Integration["Flexibility Integration"]
        HP_FLEX["Heat Pump Flexibility"]
        EV_FLEX["EV Flexibility"]
        EVENT["Event Response"]
    end

    BASE["Base Demand"] --> HP_DIS
    BASE --> EV_DIS
    HP_DIS --> HP_FLEX
    COP --> HP_FLEX
    EV_DIS --> EV_FLEX
    EV_AVAIL --> EV_FLEX
    BASE --> EVENT

    HP_FLEX --> FINAL["Finalized Network"]
    EV_FLEX --> FINAL
    EVENT --> FINAL

Heat Pump Flexibility#

Heat pump flexibility exploits thermal storage to shift electrical demand. Two modes are available:

TANK Mode (Hot Water Cylinder)#

Pre-heats a hot water tank (~200L, 50-65 C) during low-price periods and draws from it during peaks.

PyPSA components per bus:

1 Bus (thermal carrier)
1 Store (hot water tank)
1 Link (heat pump: electricity to heat at COP efficiency)
1 Load (hot water demand)

Key parameters:

Storage capacity: determined by tank volume and temperature range
Standing loss: ~1% per hour (well-insulated cylinder)
Efficiency: COP (temperature-dependent, typically 2-5)

COSY Mode (Building Thermal Inertia)#

Pre-heats building fabric during morning (07:00-09:00) and evening (17:00-19:00) windows, then allows temperature to drift down during peak demand.

PyPSA components per bus:

1 Bus (thermal inertia carrier)
1 Store (building thermal mass, state = temperature deviation)
1 Link (charging: pre-heating with efficiency boost during windows)
1 Link (discharging: using stored thermal energy)
1 Load (space heating demand)

Key parameters:

Maximum temperature deviation: 2 C from setpoint
Standing loss: ~5% per hour (building envelope losses)
Pre-heat window boost: 50% extra capacity during morning/evening windows

MIXED Mode#

Splits heat pump demand between TANK and COSY using configurable shares (default 50/50).

Spatial Allocation#

Heat pump demand is distributed across buses using one of four methods:

Method	Description
`proportional`	Allocate by existing base demand
`uniform`	Equal across all buses
`urban_weighted`	Weighted toward high-demand areas
`fes_gsp`	Use FES GSP-level distribution

Electric Vehicle Flexibility#

EV flexibility models smart charging across three tariff types:

GO Tariff (Fixed Window)#

Incentivises charging during a cheap 4-hour night window (00:00-04:00) using marginal cost signals. Charging outside the window is possible but expensive.

PyPSA components per bus:

1 Bus (EV battery carrier)
1 Store (fleet battery)
1 Link (charger with cost incentive)
1 Load (driving demand)

INT Tariff (Intelligent Smart Charging)#

Fully optimisable charging whenever vehicles are plugged in. The optimiser shifts charging to minimise system cost while meeting driving demand and minimum state-of-charge constraints.

PyPSA components per bus:

1 Bus (EV battery carrier)
1 Store (battery with time-varying e_min_pu)
1 Link (charger with availability profile p_max_pu)
1 Load (driving demand)

V2G Tariff (Vehicle-to-Grid)#

Bidirectional charging allows EVs to discharge back to the grid during peak demand. Includes a degradation cost to account for battery wear.

PyPSA components per bus:

1 Bus (EV battery carrier)
1 Store (battery with minimum SOC constraint)
1 Link (charger: grid to battery)
1 Link (V2G discharge: battery to grid, with degradation cost)
1 Load (driving demand)

V2G capacity sources (priority order):

FES Srg_BB005 (V2G MW availability) if use_fes_capacity: true
Calculated from fleet size and charger power

MIXED Mode#

Splits the EV fleet across GO, INT, and V2G tariffs. Shares can be calculated from FES data (mode: "fes") or set manually.

FES mode share calculation:

V2G share = FES V2G capacity / total EV capacity
INT share = FES smart charging capacity - V2G share
GO share = remainder

Event Response (Demand-Side Response)#

Models Saving Sessions style demand response events where households reduce load during specific windows.

Event Schedule#

Events are generated based on the configured mode:

Mode	Description
`regular`	2 events/week year-round
`winter`	5 events/week Oct-Mar only
`both`	2/week year-round + 5/week extra in winter

Events only occur during the configured window (default 07:00-22:00) on weekdays.

Capacity Calculation#

Two methods:

User-defined: Set dsr_capacity_mw directly (e.g., 5000 MW), distributed proportionally to peak demand
Calculated (fallback): base_demand x participation_rate x max_reduction_fraction

PyPSA Representation#

Demand response is modelled as generators (negative loads):

p_nom: maximum load reduction (MW)
p_max_pu: event schedule (0/1 time series)
marginal_cost: incentive payment (dispatched only when cost-effective)

Configuration Reference#

All demand flexibility settings live under demand_flexibility in config/defaults.yaml:

demand_flexibility:
  enabled: true                       # Master switch

  heat_pumps:
    enabled: true
    mode: "MIXED"                     # TANK, COSY, or MIXED
    flex_share: 0.2                   # Fraction using smart flexibility
    mix:
      tank_share: 0.5
      cosy_share: 0.5
    peak_thermal_kw_per_dwelling: 10.0
    link_sizing_margin: 1.5
    cosy:
      morning_window: ["07:00", "09:00"]
      evening_window: ["17:00", "19:00"]
      max_temp_deviation_celsius: 2.0
      thermal_mass_hours: 0.3
      standing_loss_per_hour: 0.05
    tank:
      volume_liters: 200
      temp_range: [50, 65]
      standing_loss_per_hour: 0.01
      heater_power_kw: 3.0

  electric_vehicles:
    enabled: true
    tariff: "MIXED"                   # GO, INT, V2G, or MIXED
    flex_share: 0.2
    battery_capacity_kwh: 60.0
    charge_efficiency: 0.90
    flexibility_participation: 0.10
    urban_weight: 2.0
    energy_per_vehicle_kwh_per_day: 10.0
    charging_hours_per_day: 4.0
    go:
      window: ["00:00", "04:00"]
      window_cost: 0.0
      offpeak_cost: 100.0
    int:
      min_soc: 0.20
      target_departure_soc: 0.80
      charger_power_kw: 7.0
    v2g:
      participation_rate: 0.30
      discharge_efficiency: 0.90
      max_discharge_soc: 0.80
      degradation_cost_per_mwh: 50.0
      use_fes_capacity: true
    mixed:
      mode: "fes"                     # "fes" or "manual"
      go_share: 0.30
      int_share: 0.50
      v2g_share: 0.20

  event_response:
    enabled: true
    mode: "both"                      # "regular", "winter", or "both"
    events_per_week_regular: 2
    events_per_week_winter: 5
    event_window: ["07:00", "22:00"]
    dsr_capacity_mw: null             # Set directly or leave null for calculated
    participation_rate: 0.33
    max_reduction_fraction: 0.10
    marginal_cost: 100.0
    winter_months: [10, 11, 12, 1, 2, 3]

PyPSA Component Summary#

Flexibility Type	Buses	Stores	Links	Loads	Generators
HP TANK	`{bus} heat`	Hot water tank	Heat pump (elec to heat)	Hot water demand	-
HP COSY	`{bus} thermal inertia`	Building thermal mass	Charge link + Discharge link	Space heating	-
EV GO	`{bus} EV battery`	Fleet battery	Charger (with cost)	Driving demand	-
EV INT	`{bus} EV battery`	Battery (with DSM)	Charger (with availability)	Driving demand	-
EV V2G	`{bus} EV battery`	Battery (with min SOC)	Charger + V2G discharge	Driving demand	-
Event Response	-	-	-	-	Demand response generator

Carriers#

The following carriers are automatically added to the network:

Heat-related: heat, hot water, hot water demand, heat pump, thermal inertia, space heating, thermal demand

EV-related: EV battery, EV charger, EV driving, EV driving demand, V2G

Demand response: demand response

Example Scenario#

# In config/scenarios.yaml
HT30_flex:
  description: "2030 Holistic Transition with demand flexibility"
  modelled_year: 2030
  FES_scenario: "Holistic Transition"
  network_model: "ETYS"

  demand_flexibility:
    enabled: true
    heat_pumps:
      enabled: true
      mode: "MIXED"
      flex_share: 0.2
    electric_vehicles:
      enabled: true
      tariff: "MIXED"
      flex_share: 0.2
    event_response:
      enabled: true
      mode: "both"
      dsr_capacity_mw: 5000

Tutorials#

For hands-on examples, see the tutorial notebooks:

Heat Pump Flexibility - Heat pump flexibility (TANK/COSY modes, COP, thermal storage)
Electric Vehicle Flexibility - EV flexibility (GO/INT/V2G tariffs, smart charging)
Demand Side Response - Event response (Saving Sessions, demand reduction)

Demand-Side Flexibility#

Overview#

How It Works#

Flex Share#

Heat Pump Flexibility#

TANK Mode (Hot Water Cylinder)#

COSY Mode (Building Thermal Inertia)#

MIXED Mode#

Spatial Allocation#

Electric Vehicle Flexibility#

GO Tariff (Fixed Window)#

INT Tariff (Intelligent Smart Charging)#

V2G Tariff (Vehicle-to-Grid)#

MIXED Mode#

Event Response (Demand-Side Response)#

Event Schedule#

Capacity Calculation#

PyPSA Representation#

Configuration Reference#

PyPSA Component Summary#

Carriers#

Example Scenario#

Tutorials#