stefanpauliuk
commented
6 years ago In a standard lifetime model (convolution), this would not work. But we can modify that, e.g., by saying ‘actual outflow’ is ‘delay factor’*baseline outflow. Cf. model equation 7 in ODYM-RECC docu.
As long as we keep the mass balance we can delay scrapping at any time, leading to a lifetime extension. Just need to be careful to not label these cohorts as ’15 years lifetime’ or so, when in fact, their average lifetime got extended. May need to recalculate the actual lifetime after the fact, or adjust (extend the tail) of the lifetime pdf on the fly, according to what the RE strategy scenario sais.
Another strategy is reuse, e.g., giving cars a second (maybe shorter) life.
nheeren
TomerFishman
@TomerFishman realised this yesterday, documenting it here for further discussion.
Lifetime is an attribute of a given age-cohort. If we want to assume the RE strategy of lifetime extension it would be necessary to redefine the lifetime within a model run.
An example: The building cohort "residential, single-family, 1970" has the expected median lifetime of 50 years and a standard deviation of +/-10. That means, in 2015 already a fair share of this cohort will have left the stock. Now, we assume a RE strategy that increases lifetime by +20%, i.e. median life increases to 60 years. How do we account for this?
From how I understand the code right now, this might be tricky to account for in the current model.