merging script in retrieve_gas_input_location with build_gas_input_location

Snakefile

@@ -114,12 +114,6 @@ if config["sector"]["gas_network"] or config["sector"]["H2_retrofit"]:
         "IGGIELGN_PipeSegments.geojson",
     ]
 
-
-    rule retrieve_gas_infrastructure_data:
-        output: expand("data/gas_network/scigrid-gas/data/{files}", files=datafiles)
-        script: 'scripts/retrieve_gas_infrastructure_data.py'
-
-
     rule build_gas_network:
         input:
             gas_network="data/gas_network/scigrid-gas/data/IGGIELGN_PipeSegments.geojson"
@@ -134,7 +128,7 @@ if config["sector"]["gas_network"] or config["sector"]["H2_retrofit"]:
 
     rule build_gas_input_locations:
         input:
-            lng="data/gas_network/Europe-Gas-Tracker-August-2022.geojson",
+            lng="https://globalenergymonitor.org/wp-content/uploads/2022/09/Europe-Gas-Tracker-August-2022.xlsx",
             entry="data/gas_network/scigrid-gas/data/IGGIELGN_BorderPoints.geojson",
             production="data/gas_network/scigrid-gas/data/IGGIELGN_Productions.geojson",
             regions_onshore=pypsaeur("resources/regions_onshore_elec_s{simpl}_{clusters}.geojson"),

scripts/build_gas_input_locations.py

@@ -18,12 +18,27 @@ def read_scigrid_gas(fn):
     df.drop(["param", "uncertainty", "method"], axis=1, inplace=True)
     return df
 
+def retrieve_gem_lng_data(lng_url):
+    df = pd.read_excel(lng_url,storage_options={'User-Agent': 'Mozilla/5.0'}, sheet_name = 'LNG terminals - data')
+    df = df.set_index("ComboID")
 
-def build_gas_input_locations(lng_fn, entry_fn, prod_fn, countries):
+    remove_status = ['Cancelled']
+    remove_country = ['Cyprus','Turkey']
+    remove_terminal = ['Puerto de la Luz LNG Terminal','Gran Canaria LNG Terminal']
+
+    df = df.query("Status != 'Cancelled' \
+              & Country != @remove_country \
+              & TerminalName != @remove_terminal \
+              & CapacityInMtpa != '--'")
+    
+    geometry = gpd.points_from_xy(df['Longitude'], df['Latitude'])
+    return gpd.GeoDataFrame(df, geometry=geometry, crs="EPSG:4326")
+
+
+def build_gas_input_locations(lng_url, entry_fn, prod_fn, countries):
     
     # LNG terminals
-    lng = gpd.read_file(lng_fn)
+    lng = retrieve_gem_lng_data(lng_url)
-
     lng.CapacityInMtpa = lng.CapacityInMtpa.astype(float)
 
     # Entry points from outside the model scope

scripts/retrieve_gas_input_locations.py

@@ -1,39 +0,0 @@
-"""
-Retrieve LNG gas infrastructure data from "https://globalenergymonitor.org/wp-content/uploads/2022/09/Europe-Gas-Tracker-August-2022.xlsx"
-Note: This is the script that removes non EU grid countries and terminals that are not closely connected to the europe network
-"""
-
-import logging
-import pandas as pd
-import geopandas as gpd
-
-logger = logging.getLogger(__name__)
-
-
-if __name__ == "__main__":
-    if 'snakemake' not in globals():
-        from helper import mock_snakemake
-        snakemake = mock_snakemake('retrieve_gas_input_locations')
-        rootpath = '..'
-    else:
-        rootpath = '.'
-
-    # LNG terminals
-
-    lng_url="https://globalenergymonitor.org/wp-content/uploads/2022/09/Europe-Gas-Tracker-August-2022.xlsx"
-    df = pd.read_excel(lng_url,storage_options={'User-Agent': 'Mozilla/5.0'}, sheet_name = 'LNG terminals - data')
-    df = df.set_index("ComboID")
-
-    remove_status = ['Cancelled']
-    remove_country = ['Cyprus','Turkey']
-    remove_terminal = ['Puerto de la Luz LNG Terminal','Gran Canaria LNG Terminal']
-
-    df = df.query("Status != 'Cancelled' \
-              & Country != @remove_country \
-              & TerminalName != @remove_terminal \
-              & CapacityInMtpa != '--'")
-    
-    geometry = gpd.points_from_xy(df['Longitude'], df['Latitude'])
-    lng = gpd.GeoDataFrame(df, geometry=geometry, crs="EPSG:4326")
-
-    lng.to_file(snakemake.output[0])