fix how t&d are matched against consumption mix; addresses USEPA#237

electricitylci/__init__.py

@@ -545,6 +545,7 @@ def run_net_trade(generation_mix_dict):
         end user within the named region with a 1 MWh of electricity as
         its product flow.
     """
+    # Run ba_io_trading.
     logging.info("using alt gen method for consumption mix")
     regions_to_keep = list(generation_mix_dict.keys())
     cons_mix_df_dict = get_consumption_mix_df(regions_to_keep=regions_to_keep)
@@ -564,17 +565,18 @@ def run_net_trade(generation_mix_dict):
             cons_mix_dicts[subreg]
         )
 
-    logging.info("get distribution mix")
+    logging.info("get t&d losses")
     dist_mix_df_dict = {}
     for subreg in cons_mix_dicts.keys():
         dist_mix_df_dict[subreg] = get_distribution_mix_df(subregion=subreg)
 
+    # NOTE: fails to find 'New Smyrna Beach' and 'City of Homestead'
     logging.info("write dist mix to dict")
     dist_mix_dicts = {}
     for subreg in dist_mix_df_dict.keys():
         dist_mix_dicts[subreg] = write_distribution_mix_to_dict(
-            dist_mix_df_dict[subreg],
-            cons_mix_dicts[subreg],
+            dist_mix_df_dict,
+            cons_mix_dicts,
             subregion=subreg
         )
 
@@ -588,19 +590,22 @@ def run_net_trade(generation_mix_dict):
 
 
 def run_post_processes():
-    # There is no easy way of editing files archived in a zip and no
-    # easy way of remove files from a zip without just creating a new zip,
-    # so that's what's done here.
-    # 1. FIXED. There should be only one instance of each .json file;
-    #    addressed in the new _save_to_json (olca_jsonld_writer.py)
-    # 2. Remove flows/*.json that are not found in a process exchange.
-    #    See cleanup SQL queries in GitHub issue
-    #    https://github.com/USEPA/ElectricityLCI/issues/216
-    # 3. Remove zero flows from quantitative reference exchanges
-    #    https://github.com/USEPA/ElectricityLCI/issues/217
-    # 4. Add NETL TRACI 2.1 characterization factors (impact category)
-    #    NOT TO BE ADDED TO BASELINES
-    # 5. Create product systems for select processes (@user, consumption mixes)
+    """Run post processes on JSON-LD file.
+
+    There is no easy way of editing files archived in a zip and no
+    easy way of remove files from a zip without just creating a new zip,
+    so that's what's done here.
+
+    1. There should be only one instance of each .json file;
+       addressed in the new _save_to_json (olca_jsonld_writer.py)
+    2. Remove flows/*.json that are not found in a process exchange.
+       See cleanup SQL queries in GitHub issue
+       https://github.com/USEPA/ElectricityLCI/issues/216
+    3. Remove zero flows from quantitative reference exchanges
+       https://github.com/USEPA/ElectricityLCI/issues/217
+    4. DO NOT ADD NETL TRACI 2.1 characterization factors
+    5. Create product systems for select processes (user, consumption mixes)
+    """
     from electricitylci.olca_jsonld_writer import build_product_systems
     from electricitylci.olca_jsonld_writer import clean_json
 
@@ -640,14 +645,15 @@ def write_distribution_dict():
     return distribution_mix_dictionary()
 
 
-def write_distribution_mix_to_dict(dist_mix_df, gen_mix_dict, subregion=None):
+def write_distribution_mix_to_dict(dm_dict, gm_dict, subregion=None):
     import electricitylci.eia_trans_dist_grid_loss as tnd
 
     if subregion is None:
         subregion = config.model_specs.regional_aggregation
 
+    # HOTFIX: send full dicts to method, not region-specific [2024-03-21;TWD]
     dist_mix_dict = tnd.olca_schema_distribution_mix(
-        dist_mix_df, gen_mix_dict, subregion=subregion
+        dm_dict, gm_dict, subregion=subregion
     )
     return dist_mix_dict
 

electricitylci/eia_trans_dist_grid_loss.py

@@ -251,31 +251,61 @@ def generate_regional_grid_loss(year, subregion="all"):
 
 
 def olca_schema_distribution_mix(td_by_region, cons_mix_dict, subregion="BA"):
-    """Create dictionaries for openLCA"""
+    """Create dictionaries for openLCA
+
+    Parameters
+    ----------
+    td_by_region : dict
+    cons_mix_dict : dict
+    subregion : str
+
+    Returns
+    -------
+    dict
+
+    Notes
+    -----
+    This method was incorrectly using td_regions as the primary list of
+    balancing authority and FERC region names. It turns out this dataset
+    is not 100% matching with the regions that make up the processes in
+    cons_mix_dict; therefore, the cons_mix_dict regions should be the
+    region reference. These regions are searched within the correct subregion
+    data frame within td_by_region and, when not found, uses the US average.
+    """
     distribution_mix_dict = {}
-    if subregion == "all":
-        aggregation_column = "Subregion"
-        region = list(pd.unique(td_by_region[aggregation_column]))
-    elif subregion == "NERC":
-        aggregation_column = "NERC"
-        region = list(pd.unique(td_by_region[aggregation_column]))
-    elif subregion == "BA":
-        aggregation_column = "Balancing Authority Name"
-        region = list(pd.unique(td_by_region[aggregation_column]))
+    if subregion == "BA":
+        td_col = "Balancing Authority Name"
+        region = [
+            x.replace(" - BA", "") for x in cons_mix_dict[subregion].keys()]
     elif subregion == "FERC":
-        aggregation_column = "FERC_Region"
-        region = list(pd.unique(td_by_region[aggregation_column]))
+        td_col = "FERC_Region"
+        region = [
+            x.replace(" - FERC", "") for x in cons_mix_dict[subregion].keys()]
     else:
-        aggregation_column = None
+        td_col = None
+        subregion = "US"
         region = ["US"]
+
     for reg in region:
-        if aggregation_column is None:
-            database_reg = td_by_region
+        # Get the T&D losses for this region.
+        database_reg = td_by_region[subregion].copy()
+        if td_col is not None:
+            database_reg = database_reg.loc[database_reg[td_col] == reg, :]
+
+        if database_reg.empty:
+            logging.warning(
+                "Failed to find T&D losses for '%s', using US average." % reg)
+            td_val = td_by_region['US']['t_d_losses'].values[0]
+        elif len(database_reg) == 1:
+            td_val = database_reg['t_d_losses'].values[0]
         else:
-            database_reg = td_by_region.loc[
-                td_by_region[aggregation_column] == reg, :
-            ]
+            logging.warning(
+                "Found too many regions in T&D table for '%s'! "
+                "Using first value." % reg)
+            td_val = database_reg['t_d_losses'].values[0]
+
         exchanges_list = []
+
         # Creating the reference output
         exchange(
             ref_exchange_creator(electricity_flow=electricity_at_user_flow),
@@ -287,15 +317,17 @@ def olca_schema_distribution_mix(td_by_region, cons_mix_dict, subregion="BA"):
         )
         exchanges_list[1]["input"] = True
         exchanges_list[1]["quantitativeReference"] = False
-        exchanges_list[1]["amount"] = 1 + database_reg["t_d_losses"].values[0]
+        exchanges_list[1]["amount"] = 1 + td_val
+
         matching_dict = None
-        for cons_mix in cons_mix_dict:
+        for cons_mix in cons_mix_dict[subregion]:
             if (
-                cons_mix_dict[cons_mix]["name"]
+                cons_mix_dict[subregion][cons_mix]["name"]
                 == f"Electricity; at grid; consumption mix - {reg} - {subregion}"
             ):
-                matching_dict = cons_mix_dict[cons_mix]
+                matching_dict = cons_mix_dict[subregion][cons_mix]
                 break
+
         if matching_dict is None:
             logging.warning(
                 f"Trouble matching dictionary for {reg}. "