Merge pull request #356 from adrienmellot/feature-ev-controlled-charg…

…ing-constraints

Feature EV controlled charging custom constraints

CHANGELOG.md

@@ -10,7 +10,7 @@
 
 * **ADD** fully-electrified heat demand (#284).
 
-* **ADD** fully-electrified road transportation (#270, #271, #358). A parameter allows to define the share of uncontrolled (timeseries) vs controlled charging (optimised) by the solver (#338).
+* **ADD** fully-electrified road transportation (#270, #271, #358). A parameter allows to define the share of uncontrolled (timeseries) vs controlled charging (optimised) by the solver (#338). Data for controlled charging constraints is readily available (#356), but corresponding constraints are not yet implemented (#385).
 
 * **ADD** nuclear power plant technology with capacity limits. Capacity limits can be equal to today or be bound by a minimum and maximum capacity to represent an available range in future. In either case, capacities are allocated at a subnational resolution based on linear scaling from current capacity geolocations, using the JRC power plant database (#78).
 

Snakefile

@@ -34,7 +34,7 @@ wildcard_constraints:
 
 ruleorder: area_to_capacity_limits > hydro_capacities > biofuels > nuclear_regional_capacity > dummy_tech_locations_template
 ruleorder: bio_techs_and_locations_template > techs_and_locations_template
-ruleorder: create_controlled_road_transport_annual_demand > dummy_tech_locations_template
+ruleorder: create_controlled_road_transport_annual_demand_and_installed_capacities > dummy_tech_locations_template
 
 ALL_CF_TECHNOLOGIES = [
     "wind-onshore", "wind-offshore", "open-field-pv",
@@ -178,6 +178,10 @@ rule model_template:
             "build/models/{resolution}/timeseries/demand/uncontrolled-road-transport-historic-electrification.csv",
             "build/models/{resolution}/timeseries/demand/electrified-heat-demand.csv",
             "build/models/{resolution}/timeseries/demand/heat-demand-historic-electrification.csv",
+            "build/models/{resolution}/timeseries/demand/demand-shape-min-ev.csv",
+            "build/models/{resolution}/timeseries/demand/demand-shape-max-ev.csv",
+            "build/models/{resolution}/timeseries/demand/demand-shape-equals-ev.csv",
+            "build/models/{resolution}/timeseries/demand/plugin-profiles-ev.csv",
         ),
         optional_input_files = lambda wildcards: expand(
             f"build/models/{wildcards.resolution}/{{input_file}}",

config/default.yaml

@@ -19,7 +19,7 @@ data-sources:
     swiss-end-use: https://www.bfe.admin.ch/bfe/en/home/versorgung/statistik-und-geodaten/energiestatistiken/energieverbrauch-nach-verwendungszweck.exturl.html/aHR0cHM6Ly9wdWJkYi5iZmUuYWRtaW4uY2gvZGUvcHVibGljYX/Rpb24vZG93bmxvYWQvOTg1NA==.html
     swiss-energy-balance: https://www.bfe.admin.ch/bfe/en/home/versorgung/statistik-und-geodaten/energiestatistiken/gesamtenergiestatistik.exturl.html/aHR0cHM6Ly9wdWJkYi5iZmUuYWRtaW4uY2gvZGUvcHVibGljYX/Rpb24vZG93bmxvYWQvNzUxOQ==.html
     swiss-industry-energy-balance: https://www.bfe.admin.ch/bfe/en/home/versorgung/statistik-und-geodaten/energiestatistiken/teilstatistiken.exturl.html/aHR0cHM6Ly9wdWJkYi5iZmUuYWRtaW4uY2gvZGUvcHVibGljYX/Rpb24vZG93bmxvYWQvODc4OA==.html
-    controlled-ev-profiles: https://zenodo.org/record/6579421/files/ramp-ev-consumption-profiles.csv.gz?download=1
+    controlled-ev-profiles: https://zenodo.org/record/6579421/files/ramp-ev-{dataset}.csv.gz?download=1
     uncontrolled-ev-profiles: https://sandbox.zenodo.org/records/45530/files/uncontrolled-charging-profiles.csv.gz?download=1 # TODO: convert into Zenodo repository
     gridded-temperature-data: https://zenodo.org/records/6557643/files/temperature.nc?download=1
     gridded-10m-windspeed-data: https://zenodo.org/records/6557643/files/wind10m.nc?download=1
@@ -55,6 +55,7 @@ scaling-factors: # values are tuned for models with a few hours resolution and o
     power: 0.00001 # from MW(h) to 100 GW(h)
     area: 0.0001 # from km2 to 10,000 km2
     monetary: 0.000000001 # from EUR to 1 billion EUR
+    transport: 0.01  # from Mio km to 100 Mio km
 capacity-factors:
     min: 0.001
     max: 10 # 0.001 -> 10 leads to a numerical range of 1e5 (hourly resolution)
@@ -167,7 +168,17 @@ parameters:
             coaches-and-buses: Motor coaches, buses and trolley buses
             passenger-cars: Passenger cars
             motorcycles: Powered 2-wheelers
-        uncontrolled-ev-charging-share: 0.5
+        ev-battery-sizes:
+            heavy-duty-vehicles: 0.2  # average from [EUCAR_2019]
+            light-duty-vehicles: 0.1  # own assumption based on passenger cars from [EUCAR_2019]
+            motorcycles: 0.01  # own assumption
+            coaches-and-buses: 0.2  # own assumption based on HDVs from [EUCAR_2019]
+            passenger-cars: 0.08  # average from [EUCAR_2019]
+        uncontrolled-ev-charging-share: 1
+        monthly-demand-bound-fraction:
+            min: 0.9
+            max: 1.1
+            equals: 1
     entsoe-tyndp:
         scenario: National Trends
         grid: Reference

config/schema.yaml

@@ -106,7 +106,7 @@ properties:
             controlled-ev-profiles:
                 type: string
                 pattern: ^(https?|http?):\/\/.+
-                description: Web address of electric vehicle data.
+                description: Web address of electric vehicle data for controlled charging.
             uncontrolled-ev-profiles:
                 type: string
                 pattern: ^(https?|http?):\/\/.+
@@ -198,6 +198,9 @@ properties:
             monetary:
                 type: number
                 description: "Applied to all quantities of monetary cost (base: EUR)."
+            transport:
+                type: number
+                description: "Applied to all quantities of transport (base: Mio km)."
     parameters:
         type: object
         description: Parameter values of the models.
@@ -368,9 +371,43 @@ properties:
                             motorcycles:
                                 type: string
                                 description: JRC-IDEES name of motorcycles.
+                    ev-battery-sizes:
+                        type: object
+                        description: EV battery size per vehicle type in MWh.
+                        additionalProperties: false
+                        properties:
+                            light-duty-vehicles:
+                                type: number
+                                description: Light duty vehicles / trucks.
+                            heavy-duty-vehicles:
+                                type: number
+                                description: Heavy duty vehicles / trucks.
+                            coaches-and-buses:
+                                type: number
+                                description: Coaches and buses.
+                            passenger-cars:
+                                type: number
+                                description: Passenger cars.
+                            motorcycles:
+                                type: number
+                                description: Motorcylces.
                     uncontrolled-ev-charging-share:
                         type: number
                         description: Share of uncontrolled charging.
+                    monthly-demand-bound-fraction:
+                        type: object
+                        description: Charging flexibility parameter constraining the amount of monthly demand to be met.
+                        additionalProperties: false
+                        properties:
+                            min:
+                                type: number
+                                description: Minimum monthly demand range.
+                            max:
+                                type: number
+                                description: Maximum monthly demand range.
+                            equals:
+                                type: number
+                                description: Equal monthly demand range.
             entsoe-tyndp:
                 type: object
                 description: Parameters to define scenario choice for data accessed from the ENTSO-E ten-year network development plan 2020. For more information, see https://2020.entsos-tyndp-scenarios.eu/

rules/transport.smk

@@ -6,7 +6,10 @@ rule download_transport_timeseries:
     params:
         url = config["data-sources"]["controlled-ev-profiles"]
     conda: "../envs/shell.yaml"
-    output: protected("data/automatic/ramp-ev-consumption-profiles.csv.gz")
+    output:
+        protected("data/automatic/ramp-ev-{dataset}.csv.gz")
+    wildcard_constraints:
+        dataset = "consumption-profiles|plugin-profiles"
     localrule: True
     shell: "curl -sSLo {output} {params.url}"
 
@@ -38,16 +41,20 @@ rule annual_transport_demand:
         road_distance_historically_electrified = "build/data/transport/annual-road-transport-distance-demand-historic-electrification.csv",
     script: "../scripts/transport/annual_transport_demand.py"
 
-rule create_controlled_road_transport_annual_demand:
-    message: "Create annual demand for controlled charging at {wildcards.resolution} resolution"
+rule create_controlled_road_transport_annual_demand_and_installed_capacities:
+    message: "Create annual demand for controlled charging and corresponding charging potentials at {wildcards.resolution} resolution"
     input:
         annual_controlled_demand = "build/data/transport/annual-road-transport-distance-demand-controlled.csv",
+        ev_vehicle_number = "build/data/jrc-idees/transport/processed-road-vehicles.csv",
+        jrc_road_distance = "build/data/jrc-idees/transport/processed-road-distance.csv",
         locations = "build/data/{resolution}/units.csv",
         populations = "build/data/{resolution}/population.csv",
     params:
         first_year = config["scope"]["temporal"]["first-year"],
         final_year = config["scope"]["temporal"]["final-year"],
         power_scaling_factor = config["scaling-factors"]["power"],
+        transport_scaling_factor = config["scaling-factors"]["transport"],
+        battery_sizes = config["parameters"]["transport"]["ev-battery-sizes"],
         conversion_factors = config["parameters"]["transport"]["road-transport-conversion-factors"],
         countries = config["scope"]["spatial"]["countries"],
         country_neighbour_dict = config["data-pre-processing"]["fill-missing-values"]["ramp"],
@@ -56,6 +63,23 @@ rule create_controlled_road_transport_annual_demand:
         main = "build/data/{resolution}/demand/electrified-transport.csv",
     script: "../scripts/transport/road_transport_controlled_charging.py"
 
+rule create_controlled_ev_charging_parameters:
+    message: "Create timeseries parameters {wildcards.dataset_name} for controlled EV charging at {wildcards.resolution} resolution"
+    input:
+        ev_profiles = lambda wildcards: "data/automatic/ramp-ev-consumption-profiles.csv.gz" if "demand" in wildcards.dataset_name else f"data/automatic/ramp-ev-{wildcards.dataset_name}.csv.gz",
+        locations = "build/data/{resolution}/units.csv",
+        populations = "build/data/{resolution}/population.csv",
+    params:
+        demand_range = config["parameters"]["transport"]["monthly-demand-bound-fraction"],
+        first_year = config["scope"]["temporal"]["first-year"],
+        final_year = config["scope"]["temporal"]["final-year"],
+        country_neighbour_dict = config["data-pre-processing"]["fill-missing-values"]["ramp"],
+        countries = config["scope"]["spatial"]["countries"],
+    wildcard_constraints:
+        dataset_name = "demand-shape-equals|demand-shape-max|demand-shape-min|plugin-profiles"
+    conda: "../envs/default.yaml"
+    output: "build/models/{resolution}/timeseries/demand/{dataset_name}-ev.csv"
+    script: "../scripts/transport/road_transport_controlled_constraints.py"
 
 rule create_uncontrolled_road_transport_timeseries:
     message: "Create timeseries for road transport demand  (uncontrolled charging)"

scripts/transport/road_transport_controlled_charging.py

@@ -45,7 +45,7 @@ def convert_annual_distance_to_electricity_demand(
     final_year: int,
     conversion_factors: dict[str, float],
     country_codes: list[str],
-):
+) -> pd.DataFrame:
     """
     Convert annual distance driven demand to electricity demand for
     controlled charging accounting for conversion factors.
@@ -65,6 +65,60 @@ def convert_annual_distance_to_electricity_demand(
     return -df_energy_demand
 
 
+def extract_national_ev_charging_potentials(
+    path_to_ev_numbers: str,
+    transport_scaling_factor: float,
+    first_year: int,
+    final_year: int,
+    conversion_factors: dict[str, float],
+    battery_sizes: dict[str, float],
+    country_codes: list[str],
+) -> pd.DataFrame:
+    # Extract number of EVs per vehicle type
+    df_ev_numbers = (
+        pd.read_csv(path_to_ev_numbers, index_col=[0, 1, 2, 3, 4])
+        .squeeze()
+        .droplevel(["vehicle_subtype", "section"])
+    )
+    assert not df_ev_numbers.isnull().values.any()
+    # Compute max. distance travelled per full battery for one EV [in Mio km / vehicle]
+    battery_size = pd.DataFrame.from_dict(
+        {
+            vehicle: battery_sizes[vehicle] / conversion_factors[vehicle]
+            for vehicle in battery_sizes
+        },
+        orient="index",
+        columns=["value"],
+    )
+
+    # Compute available chargeable distance per vehicle type [in transport scaling unit km]
+    df_ev_chargeable_distance = (
+        df_ev_numbers.align(battery_size, level="vehicle_type")[1]
+        .squeeze()
+        .mul(df_ev_numbers)
+        .groupby(level=["country_code", "year"])
+        .sum()
+        .loc[country_codes]
+        .unstack("year")
+        .mul(transport_scaling_factor)
+    )
+
+    if final_year > 2015:
+        # ASSUME 2015 data is used for all years after 2015
+        df_ev_chargeable_distance = df_ev_chargeable_distance.assign(**{
+            str(year): df_ev_chargeable_distance[2015]
+            for year in range(2016, final_year + 1)
+        })
+
+    df_ev_chargeable_distance.columns = df_ev_chargeable_distance.columns.astype(int)
+
+    return df_ev_chargeable_distance[range(first_year, final_year + 1)].T
+
+
+def reshape_and_add_suffix(df, suffix):
+    return df.T.add_suffix(suffix)
+
+
 if __name__ == "__main__":
     resolution = snakemake.wildcards.resolution
 
@@ -83,8 +137,12 @@ def convert_annual_distance_to_electricity_demand(
         .to_dict()
     )
     populations = pd.read_csv(snakemake.input.populations, index_col=0)
+    battery_sizes = snakemake.params.battery_sizes
+    transport_scaling_factor = snakemake.params.transport_scaling_factor
+    path_to_ev_numbers = snakemake.input.ev_vehicle_number
 
-    df = convert_annual_distance_to_electricity_demand(
+    #  Convert annual distance driven demand to electricity demand for controlled charging
+    df_demand = convert_annual_distance_to_electricity_demand(
         path_to_controlled_annual_demand,
         power_scaling_factor,
         first_year,
@@ -93,14 +151,39 @@ def convert_annual_distance_to_electricity_demand(
         country_codes,
     )
 
-    if resolution == "continental":
-        df = scale_to_continental_resolution(df)
-    elif resolution == "national":
-        df = scale_to_national_resolution(df)
-    elif resolution in ["regional", "ehighways"]:
-        df = scale_to_regional_resolution(
-            df, region_country_mapping=region_country_mapping, populations=populations
-        )
-    else:
-        raise ValueError("Input resolution is not recognised")
-    df.T.to_csv(path_to_output, index_label=["id"])
+    # Extract national EV charging potentials
+    df_charging_potentials = extract_national_ev_charging_potentials(
+        path_to_ev_numbers,
+        transport_scaling_factor,
+        first_year,
+        final_year,
+        conversion_factors,
+        battery_sizes,
+        country_codes,
+    )
+
+    # Add prefix for yaml template
+    parameters_evs = {
+        "_demand": df_demand,
+        "_charging": df_charging_potentials,
+    }
+
+    # Rescale to desired resolution and add suffix
+    dfs = []
+    for key, df in parameters_evs.items():
+        if resolution == "continental":
+            df = scale_to_continental_resolution(df)
+        elif resolution == "national":
+            df = scale_to_national_resolution(df)
+        elif resolution in ["regional", "ehighways"]:
+            df = scale_to_regional_resolution(
+                df,
+                region_country_mapping=region_country_mapping,
+                populations=populations,
+            )
+        else:
+            raise ValueError("Input resolution is not recognised")
+        dfs.append(reshape_and_add_suffix(df, key))
+
+    # Export to csv
+    pd.concat(dfs, axis=1).to_csv(path_to_output, index_label=["id"])