Using oemof

Oemof is a framework and even though it is in an early stage it already provides useful tools to model energy systems. To model an energy system you have to write your own application in which you combine the oemof libraries for you specific task. The example section shows how an oemof application may look like.

oemof-network

The oemof-network library is part of the oemof installation. By now it can be used to define energy systems as a network with components and buses. Every component should be connected to one or more buses. After definition, a component has to explicitely be added to its energy system. Allowed components are sources, sinks and transformer.

The code of the example above:

from oemof.network import *
from oemof.energy_system import *

# create the energy system
es = EnergySystem()

# create bus 1
bus_1 = Bus(label="bus_1")

# create bus 2
bus_2 = Bus(label="bus_2")

# add bus 1 and bus 2 to energy system
es.add(bus_1, bus_2)

# create and add sink 1 to energy system
es.add(Sink(label='sink_1', inputs={bus_1: []}))

# create and add sink 2 to energy system
es.add(Sink(label='sink_2', inputs={bus_2: []}))

# create and add source to energy system
es.add(Source(label='source', outputs={bus_1: []}))

# create and add transformer to energy system
es.add(Transformer(label='transformer', inputs={bus_1: []}, outputs={bus_2: []}))

The network class is aimed to be very generic and might have some network analyse tools in the future. By now the network library is mainly used as the base for the solph library.

oemof-solph

The oemof-solph library is part of the oemof installation. Solph is designed to create and solve linear or mixed-integer linear optimization problems. It is based on optimization modelling language pyomo.

To use solph at least one linear solver has to be installed on your system. See the pyomo installation guide to learn which solvers are supported. Solph is tested with the open source solver cbc and the gurobi solver (free for academic use). The open glpk solver recently showed some odd behaviour.

The formulation of the energy system is based on the oemof-network library but contains additional components such as storages. Furthermore the network class are enhanced with additional parameters such as efficiencies, bounds, cost and more. See the API documentation for more details. Try the examples to learn how to build a linear energy system.

oemof-outputlib

The oemof-outputlib library is part of the oemof installation. It collects the results of an optimisation in a dictionary holding scalar variables and pandas DataFrame for time dependend output. This makes it easy to process or plot the results using the capabilities of the pandas library.

The following code collects the results in a pandas DataFrame and selects the data for a specific component, in this case ‘heat’.

results = outputlib.processing.results(om)
heat = outputlib.views.node(results, 'heat')

To visualize results, either use pandas own visualization functionality, matplotlib or the plot library of your choice. Some existing plot methods can be found in a separate repository oemof_visio which can be helpful when looking for a quick way to create a plot.

feedinlib

The feedinlib library is not part of the oemof installation and has to be installed separately using pypi. It serves as an interface between Open Data weather data and libraries to calculate feedin timeseries for fluctuating renewable energy sources.

It is currently under revision (see here for further information). To begin with it will provide an interface to the pvlib and windpowerlib and functions to download MERRA2 weather data and open_FRED weather data. See documentation of the feedinlib for a full description of the library.

demandlib

The demandlib library is not part of the oemof installation and has to be installed separately using pypi. At the current state the demandlib can be used to create load profiles for elctricity and heat knowing the annual demand. See the documentation of the demandlib for examples and a full description of the library.