JosePauloSavioli / Lavoisier

Format converter for LCI datasets

Python library for conversions between Life Cycle Assessment (LCA) inventory formats. Currently, it does both ways of conversion between EcoSpold 2 and ILCD 1 formats. Other versions and other dataset types are to be included due time as Lavoisier is a work in progress.

Lavoisier's objective is to make possible a cohesive conversion with lower loss of information between LCA inventory formats. This objective is in line with the efforts in the LCA community for a higher interoperability within data formats, in a way having a dataset of one inventory format does not limit its usability with other inventory formats.

Right now, Lavoisier is incorpored in the Brazilian Life Cycle Inventory database (namely SICV-Brasil) to make possible the conversion of Ecospold 2 datasets from Brazilian reseach to ILCD 1 format on the upload process.

The field conversion is based on Greendelta's converter https://github.com/GreenDelta/olca-conversion-service, but Lavoisier can overcome major shortcomings from OpenLCA converter such as loss of Pedigree matrix uncertainty and coefficient information, some input flows not converted and loss of all parameters and mathematical equations.

The project is still updating frequently and feedback is welcomed. The main branch has the most stable version, while the development branch has the latest updates of the library. Tests are made mainly on Linux (Ubuntu).

Installation

Lavoisier will require the following packages:

• pathlib to help with file and diretory paths
• pint to help with unit conversion
• xmltodict to help with XML parsing
• openturns to help with uncertainty conversion
• pycryptodome to help with UUID conversion

To install the dependencies beforehand, use the commands below

• Install dependent libraries

pip install pathlib pint xmltodict openturns pycryptodome

• Install Lavoisier

python3 pip install Lavoisier

If you want to install it via github clone, just execute the setup.py file. In the case of uninstalling or trying to install it by pip again, it is recomended to run as administrator or sudo, since problems related to .pyc files could raise. Follow the steps bellow:

• Clone github repository

git clone https://github.com/JosePauloSavioli/IBICT-converter

• Go to repository and install the library

cd path/to/cloned/library
python3 setup.py install

Conversion

Lavoisier works by creating a Converter class using the method get_converter(input_: tuple, output: tuple, path: str, save_path: str, mode: str). The input_ and output tuples are the format and elementary flow mapping to be used in conversion. Currently, available format options are "ILCD1" and "EcoSpold2" and available elementary flow mappings are "EF3.0" for ILCD and "ecoinvent3.7" for EcoSpold 2. The mode is an entry to specify if the conversion is to single file(s) of the output format (to file) or to a single database (to database - available only for EcoSpold2 to ILCD1 conversion).

The Converter class has the information about the two formats and is changeable within that conversion. It has a main attribute that can be changed before conversion, the elementary flow mapping file (attr: elem_flow_mapping) and other attributes that are set depending on the conversion being carried out. To see these attribute options one can print the converter.

For example, to convert a file and a directory of files from EcoSpold 2 to ILCD 1:

# To convert a file
converter = get_converter(("EcoSpold2", "ecoinvent3.7"), ("ILCD1", "EF3.0"),
                           "path_to_file", "path_to_save_directory", "to_file")

# To convert a directory with files
converter = get_converter(("EcoSpold2", "ecoinvent3.7"), ("ILCD1", "EF3.0"),
                           "path_to_directory", "path_to_save_directory", "to_file")

# To see the available options
print(converter)

The Converter class has one method called convert which is used to carry out the conversion at any given time.

converter.convert()

Support

This project was developed in the Center for Life Cycle Sustainability Assessment (GYRO) of the Federal University of Technology - Paraná (UTFPR) with the support of the Brazilian Institute of Information in Science and Technology (IBICT). It began with the support of REAL (Resource Efficiency through Application of Life Cycle Thinking) of the UN Environment and The Life Cycle Initiative funded by the European Commission.

-> GYRO website: http://gyro.ct.utfpr.edu.br

Code developed by José Paulo Pereira das Dores Savioli. Logo created by Larissa Ugaya Mazza.