DHNx Addons

This package contains a collection of functions useful for workflows with DHNx, LPagg, and GIS-data in general in the context of municipal heat planning.

Some functions of the script are specific to Germany.

This is not a stable release and breaking changes will occur often and without warning.

Example workflow

This package provides a default workflow that perfoms the following:

• Take a polygon defining an area as input
• Download OpenStreetMap building and street data
• Assign a status "heated" depending on the type of each building
• Assign a random distribution of construction years
• Assign a random refurbishment status depending on the building type and construction year based on typical distributions from the literature
• Assign a specific heat demand based on construction year, refurbishment status and building type from the literature
• Estimate domestic hot water demand based on the building type
• Calculate the heated reference area based on the building ground area from the OpenStreetMap-Data and an estimation of the number of floors
• Apply climate correction factor based on the TRY-region
• Based on the gathered heat demand for each building, create load profiles for each building with LPagg
• As weather data, the old DWD TRY (2011) is used for the appropriate region
• (It is recommended to download and use the current DWD TRY (2017) data for your location from https://kunden.dwd.de/obt/)
• Choose a random building as a producer for a district heating grid
• Optimize the installation of a district heating grid along the streets with DHNx, choosing paths and required diameters for the pipes
• Simulate the heating grid to determine pressure loss, flow rate and temperature distribution within the network

Installation

TLDR

If you already have Anaconda installed, the easiest way to get everything set up is to create a dedicated python environment from a yaml. Save the file environment_user.yaml to a folder where you run the following command. Change the environment name my-env-name to something useful like work or dhnx:

conda env create --name=my-env-name --file=environment_user.yaml

(A simple conda install dhnx_addons will install all dependencies available on conda, but cannot install the required pip dependencies.)

To update an existing environment, you can use:

conda env update --name=my-env-name --file=environment_user.yaml

If you want to instead clone this repository to develop it further, create a development environment with all the dependencies:

conda env create --name=my-env-name --file=environment_dev.yaml

Detailed Information

• Download and install Anaconda. On windows, if winget is installed, the fastest way is winget install anaconda3
• On windows, if the Terminal is used with PowerShell, do not forget to run conda init powershell (which might require administrator rights)
• Download and install git to clone this repository, e.g. with winget install Git.Git
• It is recommended to install the dependencies into a dedicated python environment. This repository contains an example environment file that can be used like this: conda env create --file environment_dev.yaml
• Activate the new environment with conda activate work
• If you want to use your new environment in Spyder, you will likely need to install spyder-kernels. But Spyder will inform about the required version if necessary
• Some dependencies (namely dhnx and oemof) can only be obtained via pip, not conda
• Installation instructions for dhnx can sometimes change, since it is often worked on. The published pip version is not up-to-date, but environment_dev.yaml always contains the current installation target, e.g. the dev branch or a special feature branch. Perform a dry-run first to see if any packages other than dhnx and oemof would be installed. Try to install those with conda, if possible. (Mixing pip and conda installations in generally not recommended, because conda does not keep track of the pip installations properly.) To install the dev branch of dhnx, use pip install https://github.com/oemof/DHNx/archive/dev.tar.gz --dry-run
• If everything seems fine, perform the actual installation with pip install https://github.com/oemof/DHNx/archive/dev.tar.gz
• Development on dhnx is ongoing. Newer versions might require updates to the current workflow. These can be tested by installing directly from the repository at https://github.com/oemof/DHNx
• dhnx requires a solver to perform its optimization, e.g. the free cbc or gurobi (which is faster)
• The solver cbc (https://github.com/coin-or/Cbc/releases/latest) is installed automatically to ~\coin-or-cbc by the example workflow, if no solver is detected. Its location is added to the system path only during runtime, so it might not be available in other scripts
• If the user is eligible, an academic license for gurobi can be obtained at https://www.gurobi.com/downloads/end-user-license-agreement-academic/
• To test the example OpenStreetMap workflow, run python dhnx_addons.py