Prospective environmental and economic life cycle assessment of vehicles made blazing fast.

A fully parameterized Python model developed by the Technology Assessment group of the Paul Scherrer Institut to perform life cycle assessments (LCA) of vehicles.

See the documentation for more detail, validation, etc.

• Background
  • What is Life Cycle Assessment
  • Why carculator
• Install
• Usage
  • As a Python library
  • As a web app
• Support
• Maintainers
• Contributing
• License

Life Cycle Assessment (LCA) is a systematic way of accounting for environmental impacts along the relevant phases of the life of a product or service. Typically, the LCA of a passenger vehicle includes the raw material extraction, the manufacture of the vehicle, its distribution, use and maintenance, as well as its disposal. The compiled inventories of material and energy required along the life cycle of the vehicle is characterized against some impact categories (e.g., climate change).

In the research field of mobility, LCA is widely used to investigate the superiority of a technology over another one.

carculator allows to:

• produce life cycle assessment (LCA) results that include conventional midpoint impact assessment indicators as well cost indicators
• carculator uses time- and energy scenario-differentiated background inventories for the future, based on outputs of Integrated Asessment Model REMIND.
• calculate hot pollutant and noise emissions based on a specified driving cycle
• produce error propagation analyzes (i.e., Monte Carlo) while preserving relations between inputs and outputs
• control all the parameters sensitive to the foreground model (i.e., the vehicles) but also to the background model (i.e., supply of fuel, battery chemistry, etc.)
• and easily export the vehicle models as inventories to be further imported in the Brightway2 LCA framework or the SimaPro LCA software.

carculator integrates well with the Brightway LCA framework.

carculator was built based on work described in Uncertain environmental footprint of current and future battery electric vehicles by Cox, et al (2018).

carculator is at an early stage of development and is subject to continuous change and improvement. Three ways of installing carculator are suggested.

We recommend the installation on Python 3.7 or above.

conda install -c romainsacchi carculator

pip install carculator

Calculate the fuel efficiency (or Tank to wheel energy requirement) in km/L of petrol-equivalent of current SUVs for the driving cycle WLTC 3.4 over 800 Monte Carlo iterations:

    from carculator import *
    import matplotlib.pyplot as plt
    cip = CarInputParameters()
    cip.stochastic(800)
    dcts, array = fill_xarray_from_input_parameters(cip)
    cm = CarModel(array, cycle='WLTC 3.4')
    cm.set_all()
    TtW_energy = 1 / (cm.array.sel(size='SUV', year=2017, parameter='TtW energy') / 42000) # assuming 42 MJ/L petrol

    l_powertrains = TtW_energy.powertrain
    [plt.hist(e, bins=50, alpha=.8, label=e.powertrain.values) for e in TtW_energy]
    plt.xlabel('km/L petrol-equivalent')
    plt.ylabel('number of iterations')
    plt.legend()

Compare the carbon footprint of electric vehicles with that of rechargeable hybrid vehicles for different size categories today and in the future over 500 Monte Carlo iterations:

    from carculator import *
    cip = CarInputParameters()
    cip.stochastic(500)
    dcts, array = fill_xarray_from_input_parameters(cip)
    cm = CarModel(array, cycle='WLTC')
    cm.set_all()
    scope = {
        'powertrain':['BEV', 'PHEV'],
    }
    ic = InventoryCalculation(cm.array, scope=scope)

    results = ic.calculate_impacts()
    data_MC = results.sel(impact_category='climate change').sum(axis=3).to_dataframe('climate change')
    plt.style.use('seaborn')
    data_MC.unstack(level=[0,1,2]).boxplot(showfliers=False, figsize=(20,5))
    plt.xticks(rotation=70)
    plt.ylabel('kg CO2-eq./vkm')

For more examples, see examples.

carculator has a graphical user interface for fast comparisons of vehicles.

Do not hesitate to contact the development team at [email protected].

• Romain Sacchi
• Chris Mutel

See contributing.

BSD-3-Clause. Copyright 2020 Paul Scherrer Institut.