T4.5 Resilience (TNO)

[clizbe]

• Resource adequacy & resiliency (in connection with multi-instance modelling T4.6)
• Is restarting model from previous results already implemented?
  • Check if can improve speed of warm start
  • Related issue #728
• Work with WP2 to define worst-case scenario for dispatchable technologies for solar & wind
• Create tutorials for these resilience analyses:
  • Optimize planning decisions with the uncertainty included (different futures are included in scenario tree branches with probabilities)
  • Running a lot of scenarios that cover different situations and utilize robust decision making practices to find resilient and cost-effective planning solutions
  • Use multi-year time series to capture interannual weather variability and critical events like low VRE with high demand, droughts, or frozen pipelines

[clizbe]

From @DillonJ: Just to note here that as part of a national research project, ER will be implementing a reliability tool based on the Capacity outage probability table approach (COPT). This essentially would take a SpineOpt input DB and output DB and calculate the loss of load probability (LOLP) over the full results horizon. This could be useful... also with some assumptions, it could also be run over an input DB only to calculate LOLP for a large number of weather years very efficiently - but some thought would be needed to see how to model storage in such an approach. One idea would be to use one SpineOpt result run and calculate from that a capacity constribution metric (perhaps using the ELCC approach which in turn uses the COPT approach) from each storage and then use this to perform a multi-year analysis.