Calculate expected decommission year for existing assets

> I guess the idea is that for an asset type with average lifetime $<\tau>$, the probability that the *actual* lifetime, $\tau$ might be roughly distributed:
> ```math
> p(\tau) = \frac{1}{\sqrt{2\pi\sigma}}\exp\left\{-\frac{(\tau-<\tau>)^2}{2\sigma^2}\right\}
> ```
> (supposing that it is normally distributed as a simple go-to example).
>
> When MUSE gets given a bunch of assets at the base year $t_0$, each with comissioning year $t_\mathrm{c}$, it could then calculate the expected decomissioning year $t_\mathrm{d}$, for each asset from that information and the probability distribution:
> ```math
> t_\mathrm{d} = t_\mathrm{c} + \frac{ \int_{t_0-t_\mathrm{c}}^\infty \tau p(\tau) }{ \int_{t_0-t_\mathrm{c}}^\infty p(\tau) }
> ```
> To get this working we can just set $\sigma=0.2\tau$, or similar, as a working default value and then make it an optional user setting later once it's working reliably.
>
> The integral could obviously be tabulated and stored so that it is just look up function $t_\mathrm{d} = f(t_\mathrm{c},\tau)$ that gets called for each asset. These days though, and given the relatively small number of assets (<100) it'd probably be very quick even if calculated per asset.
>
> What do you think is the best way to proceed?

_Originally posted by @martinstringer in [#916](https://github.com/EnergySystemsModellingLab/MUSE2/issues/916#issuecomment-3410181296)_