• 222 views | 5 messages Discussion: LEAP
    Topic: Planning Reserve Margin and Reserve Margin -- two violations?Subscribe | Previous | Next
  • Diego García 1/20/2025

    Hi,

    I’ve run a model for the optimisation of the electricity generation in a country. Everything works fine but I see problems in the Planning Reserve Margin and Reserve Margin variables.

    The “Planning Reserve Margin calculated endogenously” is activated. And I’ve introduced the Planning Reserve Margin value with 30.

    When I run a scenario (8760 timeslices, horizon: 2023-2030), it happens that Planning Reserve Margin (PRM) does not appear as “forced” in 30 but like this:

    So, concerning PRM, why am I not getting a flat line in 30%?

    While at the same time, the Reserve Margin variable appears in this manner:

    i.e. Reserve Margin violates the Planning Reserve Margin introduced (30%) between 2029-2030.

    I understand that the violation of Reserve Margin happens because I’ve checked in the Settings/Optimization screen the option for Handling of insufficient capacity additions: “Continue calculations and create a warning diagnostic message in LEAP”. However, the warning that appears says “Optimization unable to build all needed capacity in year 2027”. Why 2027 and not 2029?

    Thanks in advance

    Diego

  • Charlie Heaps 1/24/2025
      Best Response

    Hi Diego - what exact version of LEAP are you using? We made some fixes recently (20204.1.1.10) to address problems with reserve margin calculations. See this: https://leap.sei.org/new.
  • Diego García 1/27/2025
      Best Response

    Hi Charlie,

    I had a successful model built on v.2024.1.1.9. After that, I upgraded to v10 and v12 and, in both versions, I did experience problems. In fact, something has happened with v12 since previous versions of my model have been re-configured or so, and I get infeasible solutions (even alleviating restrictions as much as possible). My NEMO version is the same for all (2.10).

    I'm going to re-install the v2024.1.1.9 in the meantime in order to get results again. I hope that works.

    Cheers
    Diego
  • Charlie Heaps 1/27/2025
      Best Response

    The latest version is 2024.1.1.15. You might want to give that a try.
  • Diego García 1/31/2025
      Best Response

    Hi Charlie,

    I've just cheched version 2024.1.1.15 and it fails.
    My model found an optimal solution with version 2024.1.1.9 and when using updated versions it fails.

    It happens that before (v9) a single run for a unique scenario took 1 hour or so and it reached a solution.
    Note that my point was the violation of Planning Reserve Margin and Reserve Margin variables.

    Now, the run takes around 3 hours and the optimisation fails at the end:

    From OPDebug.txt:

    [...]
    Model status : Infeasible
    Simplex iterations: 674139
    Objective value : 1.6212815748e+05
    HiGHS run time : 10813.47
    ERROR: No LP invertible representation for getDualRay
    2025-31-Jan 11:17:36.715 Solved model. Solver status = INFEASIBLE.
    2025-31-Jan 11:17:36.791 Solver did not find a solution for model. No results will be saved to database.
    2025-31-Jan 11:17:36.806 Dropped temporary tables.
    2025-31-Jan 11:17:36.806 Finished modeling scenario.
    MathOptInterface.INFEASIBLE


    I'm gonna go back to v2024.1.1.9 in the meantime.
    Thanks for your help.

    Best,
    Diego