Fix issue with fuel mix in IMAGE scenarios

Introduce battery and hydrogen storage
Add vanadium LCIs

conda/meta.yaml

@@ -20,15 +20,15 @@ requirements:
   run:
     - numpy
     - pandas
-    - bw2io >=0.8.7, != 0.8.8
+    - bw2io >=0.8.10
     - bw2data
     - wurst
     - xarray
     - prettytable
     - pycountry
     - cryptography
     - pyYaml
-    - sparse
+    - sparse>=0.14.0
     - schema
     - datapackage
     - requests

docs/extract.rst

@@ -615,7 +615,7 @@ The datasets introduced are listed in the table below.
  ================================================================== ===========
   Hydrogen distribution                                              location
  ================================================================== ===========
-  Hydrogen refuelling station                                        GLO
+  hydrogen refuelling station                                        GLO
   high pressure hydrogen storage tank                                GLO
   distribution pipeline for hydrogen, dedicated hydrogen pipeline    RER
   transmission pipeline for hydrogen, dedicated hydrogen pipeline    RER
@@ -635,6 +635,19 @@ The datasets introduced are listed in the table below.
 .. _LCI_H2_distr: https://github.com/polca/premise/blob/master/premise/data/additional_inventories/lci-hydrogen-distribution.xlsx
 .. _Cerniauskas: https://doi.org/10.1016/j.ijhydene.2020.02.121
 
+
+Hydrogen turbine
+****************
+
+A dataset for a hydrogen turbine is also imported, to model the production of electricity
+from hydrogen, with an efficiency of 51%. The efficiency of the H2-fed gas turbine is based
+on the parameters of Ozawa_ et al. (2019), consulted here: LCI_H2_turbine_.
+
+.. _Ozawa: https://doi.org/10.1016/j.ijhydene.2019.02.230
+.. _LCI_H2_turbine: https://github.com/polca/premise/blob/master/premise/data/additional_inventories/lci-hydrogen-turbine.xlsx
+
+
+
 Biofuels
 --------
 
@@ -944,6 +957,41 @@ These inventories can be found here: LCI_lithium_.
 .. _LCI_cobalt: https://github.com/polca/premise/blob/master/premise/data/additional_inventories/lci-cobalt.xlsx
 .. _LCI_lithium: https://github.com/polca/premise/blob/master/premise/data/additional_inventories/lci-lithium.xlsx
 
+Vanadium Redox Flow Batteries
+-----------------------------
+
+*premise* imports inventories for the production of a vanadium redox flow battery, used
+for grid-balancing, from the work of Weber_ et al. 2021.
+It is available under the following dataset:
+
+* vanadium-redox flow battery system assembly, 8.3 megawatt hour
+
+The dataset providing electricity is the following:
+
+* electricity supply, high voltage, from vanadium-redox flow battery system
+
+The power capacity for this application is 1MW and the net storage capacity 6 MWh.
+The net capacity considers the internal inefficiencies of the batteries and the
+min Sate-of-Charge, requiring a certain oversizing of the batteries.
+For providing net 6 MWh, a nominal capacity of 8.3 MWh is required for the
+VRFB with the assumed operation parameters. The assumed lifetime of the stack
+is 10 years. The lifetime of the system is 20 years or 8176
+cycle-life (49,000 MWh).
+
+.. _Weber: https://doi.org/10.1021/acs.est.8b02073
+
+These inventories can be found here: LCI_vanadium_redox_flow_batteries_.
+
+.. _LCI_vanadium_redox_flow_batteries: https://github.com/polca/premise/blob/master/premise/data/additional_inventories/lci-vanadium-redox-flow-battery.xlsx
+
+This publication also provides LCIs for Vanadium mining and refining from iron ore.
+The end product is vanadium pentoxide, which is available under the following dataset:
+
+* vanadium pentoxide production
+
+These inventories can be found here: LCI_vanadium_.
+
+.. _LCI_vanadium: https://github.com/polca/premise/blob/master/premise/data/additional_inventories/lci-vanadium.xlsx
 
 Road vehicles
 -------------

docs/transform.rst

@@ -416,6 +416,28 @@ IAM region with their respective current production volumes (also provided by
 ecoinvent). This is not ideal as it supposes that future country-specific
 production volumes will remain the same in respect to one another.
 
+Storage
+-------
+
+If the IAM scenario requires the use of storage, *premise* adds a storage
+dataset to the high voltage market. *premise* can add two types of storage:
+
+* storage via a large-scale flow battery (electricity supply, high voltage, from vanadium-redox flow battery system)
+* storage via the conversion of electricity to hydrogen and subsequent use in a gas turbine (electricity production, from hydrogen-fired one gigawatt gas turbine)
+
+The electricity storage via battery incurs a 33% loss. It is operated by a 8.3 MWh vanadium redox-based flow battery,
+with a lifetime of 20 years or 8176 cycle-lifes (i.e., 49,000 MWh).
+
+The storage of electricity via hydrogen is done in two steps: first, the electricity is converted to hydrogen
+via a 1MW PEM electrolyser, with an efficiency of 62%. The hydrogen is then stored in a geological cavity
+and used in a gas turbine, with an efficiency of 51%. Accounting for leakages and losses, the
+overall efficiency of the process is about 37% (i.e., 2.7 kWh necessary to deliver 1 kWh to the grid).
+
+The efficiency of the H2-fed gas turbine is based on the parameters of Ozawa_ et al. (2019).
+
+.. _Ozawa: https://doi.org/10.1016/j.ijhydene.2019.02.230
+
+
 Medium voltage regional markets
 _______________________________
 

premise/activity_maps.py

@@ -3,6 +3,7 @@
 mapping between ``premise`` and ``ecoinvent`` terminology.
 """
 
+import sys
 from collections import defaultdict
 from pathlib import Path
 from typing import List, Union
@@ -23,11 +24,12 @@
 )
 
 
-def get_mapping(filepath: Path, var: str) -> dict:
+def get_mapping(filepath: Path, var: str, model: str = None) -> dict:
     """
     Loa a YAML file and return a dictionary given a variable.
     :param filepath: YAML file path
     :param var: variable to return the dictionary for.
+    :param model: if provided, only return the dictionary for this model.
     :return: a dictionary
     """
 
@@ -37,7 +39,11 @@ def get_mapping(filepath: Path, var: str) -> dict:
     mapping = {}
     for key, val in techs.items():
         if var in val:
-            mapping[key] = val[var]
+            if model is None:
+                mapping[key] = val[var]
+            else:
+                if model in val.get("iam_aliases", {}):
+                    mapping[key] = val[var]
 
     return mapping
 
@@ -114,12 +120,15 @@ class InventorySet:
     These functions return the result of applying :func:`act_fltr` to the filter dictionaries.
     """
 
-    def __init__(self, database: List[dict], version: str = None) -> None:
+    def __init__(
+        self, database: List[dict], version: str = None, model: str = None
+    ) -> None:
         self.database = database
         self.version = version
+        self.model = model
 
         self.powerplant_filters = get_mapping(
-            filepath=POWERPLANT_TECHS, var="ecoinvent_aliases"
+            filepath=POWERPLANT_TECHS, var="ecoinvent_aliases", model=self.model
         )
 
         self.powerplant_fuels_filters = get_mapping(
@@ -253,4 +262,15 @@ def generate_sets_from_filters(self, filtr: dict, database=None) -> dict:
             tech: act_fltr(database, fltr.get("fltr"), fltr.get("mask"))
             for tech, fltr in filtr.items()
         }
-        return {tech: {act["name"] for act in actlst} for tech, actlst in techs.items()}
+
+        mapping = {
+            tech: {act["name"] for act in actlst} for tech, actlst in techs.items()
+        }
+
+        # check if all keys have values
+        # if not, print warning
+        for key, val in mapping.items():
+            if not val:
+                print(f"Warning: No activities found for {key} -- revise mapping.")
+
+        return mapping