Why are global parameters not updated in nested product systems when calculating in python?

Question

I have created a product systems that contains other product systems as providers for the main process. They contain processes that depend on a global parameter. When I run them in openLCA and change the global parameter and run again, the results change. If I do the same from a python script (inside openLCA), the results stay the same. For everything else (e.g. the same use case where I uses processes as providers and not product systems), the python script works fine.

The script looks something like this:

for prod_system in prod_systems:

  for param_val in param1_vals:

    system = dao.getForName(prod_system)[0]

    solver = JuliaSolver()

    m_cache = MatrixCache.createLazy(db)

    calculator = SystemCalculator(m_cache, solver)

    setup = CalculationSetup(CalculationType.UPSTREAM_ANALYSIS, system)

 

    ### Parameter für Berechnung festlegen

    p1 = model.ParameterRedef()

    p1.value = param_val

    p1.name = 'param1'

    setup.parameterRedefs.add(p1)

 

    ### Die Wirkungsabschätzungsmethode auswählen

    method_dao = ImpactMethodDao(db)

    impactMethod = method_dao.getForName('My_Method')[0]

    setup.impactMethod = Descriptors.toDescriptor(impactMethod)

 

    result = calculator.calculateFull(setup)

Answer 1

And I found a solution that makes sense. If I update the parameter in the database, it works fine. So ParameterRedef only seems to update the foreground system. Here ist the working code:

for prod_system in prod_systems:

  for param_val in param1_vals:

    system = dao.getForName(prod_system)[0]

    solver = JuliaSolver()

    m_cache = MatrixCache.createLazy(db)

    calculator = SystemCalculator(m_cache, solver)

    setup = CalculationSetup(CalculationType.UPSTREAM_ANALYSIS, system)

 

    ### Parameter für Berechnung festlegen

    param_dao = ParameterDao(db)

    param = param_dao.getForName('param1')[0]

    param.value = param_val

    param_dao.update(param)

 

    ### Die Wirkungsabschätzungsmethode auswählen

    method_dao = ImpactMethodDao(db)

    impactMethod = method_dao.getForName('My_Method')[0]

    setup.impactMethod = Descriptors.toDescriptor(impactMethod)

 

    result = calculator.calculateFull(setup)