Select the best statement

Generally speaking, to estimate process KPIs, one run of a simulation model will suffice

Simulation models can only be reliably used to assess steady$-$state process performance

To reliably assess process risks, one run each of multiple scenarios will usually suffice

To generate reliable confidence intervals for process KPIs, it is a best practice to run a model multiple times $($with different seeds$)$ for at least $1$ million seconds

Simulation models can be used to assess warmup and cool$-$down phases of a process