IESopt: Integrated Energy System Optimization

Note: We are currently moving from our internal version control to GitHub. Missing contents will be added during the upcoming days ...

IESopt (Integrated Energy System Optimization) is a modeling and optimization framework for integrated energy systems.

It is developed and maintained at the Center for Energy at AIT Austrian Institute of Technology GmbH. The framework is designed to support the optimization of energy systems that are characterized by a high degree of integration between different energy carriers and sectors. It focuses on offering a modular and adaptable tool for modelers, that does not compromise on performance, while still being user-friendly. This is enabled by reducing energy system assets to abstract building blocks, that are supported by specialized implementation, and can be combined into complex systems without the need of a detailed understanding of mathematical modeling or proficiency in any coding-language.

MERGING DOCUMENTATIONS

We are currently merging this documentation into the (now) central one, over at working iesopt. Besides migrating everything, this requires some clean-up of (potentially) confindential left-overs (e.g., from projects), so the documentation needs to be checked page-by-page. If you are trying to get started before we manage to fix everything, get in touch with us directly - we'll help you set up everything you need. Leftovers can be seen in #4c2a59c or earlier. Information in the current Julia documentation as well as in this README may be outdated!

Background
Install
Usage
API
Contributing
License

Background

IESopt has been in development at AIT since 2021, and was moved to GitHub in June, 2024, based on a cleaned version with slightly more than 1000 commits. IESopt has been applied in a variety of projects, ranging from small-scale energy system optimization, to large-scale models spanning multiple sectors and regions. Check out the list of references for more information.

Install

Make sure to check out the detailed installation guides in the documentation, both for Python and Julia. Depending on your use-case, and choice of programming language / setup, there are different ways to get started. If you are experienced with Julia, and want to interact with the core model itself, then using IESopt.jl directly will suit you best. If you are looking for a more user-friendly interaction, or are new to energy system modeling and/or coding at all, you might want to check out the Python wrapper iesopt.

If you are not 110% sure where to start... start here: iesopt

Quick setup for Julia

In an open REPL, type ] to enter the package mode, make sure that your environment is activated (e.g., do activate .), then add the package with the following command:

(your-env) pkg> add IESopt

Quick setup for Python

Head over to iesopt and follow the instructions there, or - if you already have a working Python environment - install the package, e.g., via poetry:

poetry add iesopt

Usage

IESopt requires a configured model to run. You can start with the extensive first model tutorial, or checkout IESoptLib which includes many more examples.

Basic usage

The basic workflow to get results from a model, defined by a top-level configuration file config.iesopt.yaml, is as follows:

Parse, generate, and build the model from config.iesopt.yaml.
Optimize the model (the chose solver is specified in config.iesopt.yaml).
Get the results from the model.

Steps 1. and 2. can be combined in a single call, which the convenience function run(...) provides.

Using Julia

using IESopt

model = IESopt.run("config.iesopt.yaml")
results = model.ext[:iesopt].results

Using Python

import iesopt

model = iesopt.run("config.iesopt.yaml")
df_results = model.results.to_pandas()

API

Check out the full API reference in the documentation, and most importantly the YAML reference, that documents how the required model configuration files should be structured.

Basic API

A short overview is given below.

Julia API

"""Builds and returns a model using the IESopt framework."""
IESopt.generate!(filename::String)

"""Optimize the given model, optionally saving model results to disk."""
IESopt.optimize!(model::JuMP.Model; save_results::Bool=true, kwargs...)

"""Build, optimize, and return a model, in a single call."""
IESopt.run(filename::String; verbosity=nothing, kwargs...)

"""Get the component with the name `component_name` from the `model`."""
IESopt.get_component(model::JuMP.Model, component_name::String)

"""Compute the Irreducible Infeasible Set (IIS) of the model."""
IESopt.compute_IIS(model::JuMP.Model; filename::String = "")

ait-energy / IESopt.jl

readme