Grid Focused Model Scenario Combinations
When comparing energy system simulation models it makes sense to establish a common test base on which they can actually be compared.
The Application/scenarios
example module aggregates explanatory details on how to use tessif’s
comparing utilities on a set of grid based bench
mark scenarios to obtain meaningful results.
Further more it illustrates how these utilities can be used in a scientific context.
Introduction
This section includes a basic grid-based energy system model as well as various variations of this model.
These scenarios do not attempt to imitate an existing energy supply system, but are oriented towards a possible future energy supply system for Germany. In particular, the structure of the energy system is reflected in the models.
These scenarios are modeled and optimized with Oemof and PyPSA using Tessif to allow comparison of Oemof and PyPSA.
Reference Energy Systems
The basic energy system model create_grid_kp_es consists
of three different voltage levels (high, medium and low voltage).
These individual networks are interconnected by power transformers
(connector component).
Various consumers (households, commercial consumers, industrial consumers and charging stations for electric vehicles),
renewable energy sources (solar and onshore/offshore wind energy) and conventional power plants
(coal-fired cogeneration plant, combined cycle power plant and biogas CHP) are connected to the voltage levels.
In addition, there is a sector coupling with the heat sector. The electricity can be converted into heat directly
via Power-to-Heat.
The base model is considered lossless and without limitation of the transmission power between the voltage levels.
Energy System Graph
>>> import matplotlib.pyplot as plt >>> import tessif.visualize.nxgrph as nxv >>> grph = es.to_nxgrph() >>> drawing_data = nxv.draw_graph( ... grph, ... node_color={ ... 'Coal Supply': '#666666', ... 'Coal Supply Line': '#666666', ... 'HKW': '#666666', ... 'HKW2': '#666666', ... 'Solar Thermal': '#b30000', ... 'Heat Storage': '#cc0033', ... 'District Heating': 'Red', ... 'District Heating Demand': 'Red', ... 'Power to Heat': '#b30000', ... 'Biogas plant': '#006600', ... 'Biogas': '#006600', ... 'BHKW': '#006600', ... 'Onshore Wind Power': '#99ccff', ... 'Offshore Wind Power': '#00ccff', ... 'Gas Station': '#336666', ... 'Gaspipeline': '#336666', ... 'GuD': '#336666', ... 'Solar Panel': '#ffe34d', ... 'Commercial Demand': '#ffe34d', ... 'Household Demand': '#ffe34d', ... 'Industrial Demand': '#ffe34d', ... 'Car charging Station': '#669999', ... 'Low Voltage Powerline': '#ffcc00', ... 'Medium Voltage Powerline': '#ffcc00', ... 'High Voltage Powerline': '#ffcc00', ... 'High Voltage Transformator': 'yellow', ... 'Low Voltage Transformator': 'yellow', ... }, ... title='Energy System Grid "Kupferplatte" Graph', ... ) >>> # plt.show() # commented out for simpler doctesting
Model variations
There are different model variations to allow both transmission losses, but also grid limitation. Furthermore, the optimization capability of the energy system is improved by adding various components. A basic distinction can be made between models that are capable of limiting the transmission power between voltage levels and those that are not.
To enable network limitation, the voltage levels must be connected by two
transformer components each instead of connectors.
Models without grid limitation
The model create_grid_cs_es is able to represent
transmission losses of the grid transformer and also has an energy storage connected to the high-voltage level as a
further component.
The model create_grid_cp_es is also capable of representing
transmission losses. Instead of the storage unit, an unlimited power source/sink is integrated into the high-voltage
level.
These can be used to compensate for power surpluses and deficits.
Models with grid limitation
In addition to transmission losses, the
model create_grid_ts_es can also take into account a
limitation of the network. This model is also extended by an energy storage system.
However, a separate energy storage system is installed in each voltage level.
The model create_grid_tp_es can also represent the losses
as well as the grid limitation. This model is equipped with an unlimited power source/sink in each voltage level.
Model Results
In the following section, it is shown how to perform the optimization of the basic model using Oemof and PyPSA. Furthermore, it is shown how to evaluate the results with Tessif. As a result, the summed input and output powers of the components are given in GW. For the model variations, the optimization and evaluation can be performed in the same way. Therefore, only the results are shown.
Basic model (Kupferplattenmodell, create_grid_kp_es)
>>> import tessif.frused.configurations as configurations
>>> configurations.spellings_logging_level = 'debug'
>>> import pandas as pd
>>> # Import hardcoded Tessif energy system
>>> import tessif.examples.data.tsf.py_hard as tsf_py
>>> es = tsf_py.create_grid_kp_es()
>>> # Oemof optimization
>>> # Transform Tessif energy system into Oemof energy system
>>> from tessif.transform.es2es.omf import transform
>>> oemof_es = transform(es)
>>> # Optimize Oemof energy system
>>> import tessif.simulate as simulate
>>> optimized_oemof_es = simulate.omf_from_es(oemof_es)
>>> # Import the post processing utilities and conduct the post processing
>>> import tessif.transform.es2mapping.omf as oemof_results
>>> resultier_oemof = oemof_results.LoadResultier(optimized_oemof_es)
>>> resultier_oemof2 = oemof_results.IntegratedGlobalResultier(optimized_oemof_es)
>>> # PyPSA optimizartion
>>> # Transform Tessif energy system into PyPSA energy system
>>> from tessif.transform.es2es import ppsa as tsf2pypsa
>>> pypsa_es = tsf2pypsa.transform(es)
>>> # Optimize PyPSA energy system
>>> import tessif.simulate as simulate
>>> optimized_pypsa_es = simulate.ppsa_from_es(pypsa_es)
>>> # Import the post processing utilities and conduct the post processing
>>> from tessif.transform.es2mapping import ppsa as pypsa_results
>>> resultier_pypsa = pypsa_results.LoadResultier(optimized_pypsa_es)
>>> resultier_pypsa2 = pypsa_results.IntegratedGlobalResultier(optimized_pypsa_es)
>>> # Evaluation and output of results
>>> # Integrated Global Results
>>> IGR = pd.DataFrame([[resultier_oemof2.global_results['emissions (sim)'], resultier_pypsa2.global_results['emissions (sim)']],
... [resultier_oemof2.global_results['costs (sim)'], resultier_pypsa2.global_results['costs (sim)']]],
... columns=['Oemof', 'PyPSA'], index=['Emissions in t CO2', 'Costs in €'])
>>> print(IGR)
Oemof PyPSA
Emissions in t CO2 443159.0 443159.0
Costs in € 202259102.0 202259102.0
>>> # Summed input and output powers
>>> result1 = pd.concat([resultier_oemof.node_load['Low Voltage Powerline'].sum()*-0.001,
... resultier_pypsa.node_load['Low Voltage Powerline'].sum()*-0.001],
... keys=['Oemof', 'PyPSA'], axis=1)
>>> result2 = pd.concat([resultier_oemof.node_load['Medium Voltage Powerline'].sum()*-0.001,
... resultier_pypsa.node_load['Medium Voltage Powerline'].sum()*-0.001],
... keys=['Oemof', 'PyPSA'], axis=1)
>>> result3 = pd.concat([resultier_oemof.node_load['High Voltage Powerline'].sum()*-0.001,
... resultier_pypsa.node_load['High Voltage Powerline'].sum()*-0.001],
... keys=['Oemof', 'PyPSA'], axis=1)
>>> result4 = pd.concat([resultier_oemof.node_load['District Heating'].sum()*-0.001,
... resultier_pypsa.node_load['District Heating'].sum()*-0.001],
... keys=['Oemof', 'PyPSA'], axis=1)
>>> result = pd.concat([result1, result2, result3, result4], keys=['Low Voltage Powerline', 'Medium Voltage Powerline',
... 'High Voltage Powerline', 'District Heating'])
>>> print(result)
Oemof PyPSA
Low Voltage Powerline BHKW 8.774095 8.774096
Low Voltage Transformator 644.028627 644.028627
Solar Panel 593.346832 593.346832
Commercial Demand -582.459558 -582.459558
Household Demand -552.564843 -552.564842
Low Voltage Transformator -111.125154 -111.125154
Medium Voltage Powerline High Voltage Transformator 871.089257 871.089257
Low Voltage Transformator 111.125154 111.125154
Onshore Wind Power 1099.866419 1099.866419
Car charging Station -37.026338 -37.026338
High Voltage Transformator -0.000000 -0.000000
Industrial Demand -1229.008356 -1229.008357
Low Voltage Transformator -644.028627 -644.028627
Power to Heat -172.017506 -172.017506
High Voltage Powerline GuD 121.996114 121.996114
HKW 237.669821 237.669820
HKW2 201.601795 201.601796
High Voltage Transformator -0.000000 -0.000000
Offshore Wind Power 309.821525 309.821525
High Voltage Transformator -871.089257 -871.089257
District Heating BHKW 13.825848 13.825848
HKW 594.174551 594.174550
Power to Heat 172.017506 172.017506
Solar Thermal 85.441944 85.441944
District Heating Demand -865.459850 -865.459849
No differences in the optimization can be seen. It should be noted that Tessif determines the emissions for PyPSA in post processing via the energy flows, since the values of PyPSA cannot be tapped directly. Thus, different emissions may occur internally to the model.
Model without grid limitation and storage unit (create_grid_cs_es)
>>> print(IGR)
Oemof PyPSA
Emissions in t CO2 460354.0 454697.0
Costs in € 200821160.0 199646851.0
>>> print(result)
Oemof PyPSA
Low Voltage Powerline BHKW 2.909165 2.029346
Low Voltage Transformator 649.893559 650.773377
Solar Panel 593.346832 593.346832
Commercial Demand -582.459558 -582.459558
Household Demand -552.564843 -552.564842
Low Voltage Transformator -111.125154 -111.125154
Medium Voltage Powerline High Voltage Transformator 823.327705 819.948388
Low Voltage Transformator 110.013902 111.125154
Onshore Wind Power 1099.866419 1099.866419
Car charging Station -37.026338 -37.026338
High Voltage Transformator -0.000000 -0.000000
Industrial Demand -1229.008356 -1229.008357
Low Voltage Transformator -656.458141 -650.773377
Power to Heat -110.715195 -114.131888
High Voltage Powerline GuD 78.789074 72.550369
HKW 265.887430 265.075304
HKW2 198.048698 193.560970
High Voltage Transformator -0.000000 -0.000000
Offshore Wind Power 309.821525 309.821525
Pumped Storage 59.533591 59.980085
High Voltage Transformator -831.644150 -819.948388
Pumped Storage -80.436169 -81.039866
District Heating BHKW 4.584138 3.197758
HKW 664.718572 662.688259
Power to Heat 110.715195 114.131888
Solar Thermal 85.441944 85.441944
District Heating Demand -865.459850 -865.459849
When comparing Oemof and PyPSA, it is noticeable that the PyPSA model is both cheaper and has lower emissions. This is due to the fact that PyPSA cannot represent transmission losses via connector components. The lower cost and higher emissions compared to the basic model are due to the impact of energy storage.
Model without grid limitation and power source/sink (create_grid_cp_es)
>>> print(IGR)
Oemof PyPSA
Emissions in t CO2 448623.0 443159.0
Costs in € 203465454.0 202259102.0
>>> print(result)
Oemof PyPSA
Low Voltage Powerline BHKW 10.422229 8.774096
Low Voltage Transformator 642.380495 644.028627
Solar Panel 593.346832 593.346832
Commercial Demand -582.459558 -582.459558
Household Demand -552.564843 -552.564842
Low Voltage Transformator -111.125154 -111.125154
Medium Voltage Powerline High Voltage Transformator 872.781594 871.089257
Low Voltage Transformator 110.013902 111.125154
Onshore Wind Power 1099.866419 1099.866419
Car charging Station -37.026338 -37.026338
High Voltage Transformator -0.000000 -0.000000
Industrial Demand -1229.008356 -1229.008357
Low Voltage Transformator -648.869188 -644.028627
Power to Heat -167.758037 -172.017506
High Voltage Powerline GuD 126.908111 121.996114
HKW 238.334786 237.669820
HKW2 206.533148 201.601796
High Voltage Transformator -0.000000 -0.000000
Offshore Wind Power 309.821525 309.821525
Power Source -0.000000 -0.000000
High Voltage Transformator -881.597570 -871.089257
Power Sink -0.000000 -0.000000
District Heating BHKW 16.422906 13.825848
HKW 595.836962 594.174550
Power to Heat 167.758037 172.017506
Solar Thermal 85.441944 85.441944
District Heating Demand -865.459850 -865.459849
Again, the differences in the modeling programs are due to the transmission efficiency of the connector. Since this is 100% for PyPSA, the optimization is identical to the base model.
Model with grid limitation and storage units (create_grid_ts_es)
>>> print(IGR)
Oemof PyPSA
Emissions in t CO2 467832.0 467832.0
Costs in € 197240294.0 197240294.0
>>> print(result)
Oemof PyPSA
Low Voltage Powerline BHKW 3.748359 3.748359
Medium Low Transformator 640.488831 640.104346
Pumped Storage LV 53.042050 51.950018
Solar Panel 593.346832 593.346832
Commercial Demand -582.459558 -582.459558
Household Demand -552.564843 -552.564842
Low Medium Transformator -83.942598 -83.942598
Pumped Storage LV -71.659072 -70.182555
Medium Voltage Powerline High Medium Transformator 764.228788 762.902417
Low Medium Transformator 83.103172 83.103172
Onshore Wind Power 1099.866419 1099.866419
Pumped Storage MV 56.145835 53.481680
Car charging Station -37.026338 -37.026338
Industrial Demand -1229.008356 -1229.008357
Medium High Transformator -0.000000 -0.000000
Medium Low Transformator -646.958414 -646.570046
Power to Heat -14.495461 -14.495461
Pumped Storage MV -75.855645 -72.253488
High Voltage Powerline GuD 11.500603 11.500603
HKW 303.846376 303.846376
HKW2 164.165787 164.165787
Medium High Transformator -0.000000 -0.000000
Offshore Wind Power 309.821525 309.821525
Pumped Storage HV 49.545702 53.350973
High Medium Transformator -771.948270 -770.608502
Pumped Storage HV -66.931723 -72.076762
District Heating BHKW 5.906505 5.906505
HKW 759.615942 759.615940
Power to Heat 14.495461 14.495461
Solar Thermal 85.441944 85.441944
District Heating Demand -865.459850 -865.459849
Identical results for Oemof and PyPSA. Both transmission efficiencies and network limitations can be achieved identically for both programs using Transformer components. Minor differences in the individual components are due to the effects of the energy storage.
Model with grid limitation and power sources/sinks (create_grid_tp_es)
>>> print(IGR)
Oemof PyPSA
Emissions in t CO2 445925.0 445925.0
Costs in € 208059511.0 208059511.0
>>> print(result)
Oemof PyPSA
Low Voltage Powerline BHKW 27.008983 27.008984
Medium Low Transformator 607.595379 607.595379
Power Source LV 18.198362 18.198362
Solar Panel 593.346832 593.346832
Commercial Demand -582.459558 -582.459558
Household Demand -552.564843 -552.564842
Low Medium Transformator -111.125154 -111.125154
Power Sink LV -0.000000 -0.000000
Medium Voltage Powerline High Medium Transformator 837.645113 837.645112
Low Medium Transformator 110.013902 110.013902
Onshore Wind Power 1099.866419 1099.866419
Power Source MV -0.000000 -0.000000
Car charging Station -37.026338 -37.026338
Industrial Demand -1229.008356 -1229.008357
Medium High Transformator -0.000000 -0.000000
Medium Low Transformator -613.732706 -613.732706
Power Sink MV -0.000000 -0.000000
Power to Heat -167.758037 -167.758037
High Voltage Powerline GuD 83.140093 83.140093
HKW 227.880105 227.880103
HKW2 225.264453 225.264453
Medium High Transformator -0.000000 -0.000000
Offshore Wind Power 309.821525 309.821525
Power Source HV -0.000000 -0.000000
High Medium Transformator -846.106174 -846.106174
Power Sink HV -0.000000 -0.000000
District Heating BHKW 42.559610 42.559611
HKW 569.700257 569.700258
Power to Heat 167.758037 167.758037
Solar Thermal 85.441944 85.441944
District Heating Demand -865.459850 -865.459849
Again, the optimization results of PyPSA and Oemof are identical. However, if the transmission of the grid is limited to 30,000 MW, PyPSA can no longer optimize the model. This is because PyPSA cannot make sinks variable and thus the excess power due to the grid limitation cannot be removed from the low voltage level. The following shows the command and the results of the grid limitation to 30,000 MW.
>>> # Import hardcoded Tessif energy system
>>> import tessif.examples.data.tsf.py_hard as tsf_py
>>> esys = tsf_py.create_grid_tp_es(24, 0.99, 30000, None, None)
>>> print(IGR)
Oemof
Emissions in t CO2 429778.0
Costs in € 253837370.0
>>> print(result)
Oemof
Low Voltage Powerline BHKW 97.417180
Medium Low Transformator 360.224135
Power Source LV 198.469179
Solar Panel 593.346832
Commercial Demand -582.459558
Household Demand -552.564843
Low Medium Transformator -110.645128
Power Sink LV -3.787798
Medium Voltage Powerline High Medium Transformator 596.902146
Low Medium Transformator 109.538677
Onshore Wind Power 1099.866419
Power Source MV 6.151233
Car charging Station -37.026338
Industrial Demand -1229.008356
Medium High Transformator -11.353142
Medium Low Transformator -363.862763
Power Sink MV -0.000000
Power to Heat -171.207877
High Voltage Powerline GuD -0.000000
HKW 182.121669
HKW2 99.748658
Medium High Transformator 11.239610
Offshore Wind Power 309.821525
Power Source HV -0.000000
High Medium Transformator -602.931459
Power Sink HV -0.000000
District Heating BHKW 153.505860
HKW 455.304168
Power to Heat 171.207877
Solar Thermal 85.441944
District Heating Demand -865.459850
Conclusion
In summary, some differences exist between the optimizations using Oemof and PyPSA when considering grid-based energy systems. These are mostly due to limitations of the model components of Oemof or PyPSA. Whereas Oemof has no problems modeling grid-based energy systems. However, PyPSA is not able to specify transmission efficiencies for connectors. Furthermore, sinks cannot be made variable, so it is not possible to integrate an unlimited power sink into the model. Apart from the previously mentioned aspects, Oemof and PyPSA behave quite similarly in terms of optimization.
Used Utilities
grid_scenarios is a tessif
module aggregating the research results of a project thesis titled
Developing grid based scenarios for comparing free open source energy
supply system simulation software implemented in Python conducted by
YOUR_NAME.