Network modelling subgroup meeting

Date: 11 Jan 2019, 17-18pm (CET)

Attending: Remi Patureau, Enora Garreau, Felix Bünning, Michael Mans, Bram van der Heijde (notes)

Minutes

DESTEST Paper (Building Simulations Conference)

• Contribution to DESTEST paper written by Felix and Michael (correct?) sent to Dirk for review before Christmas break

Comparison of different pipe models.

• Contributions from Alessandro, Enora, Felix and Michael
• There is a difference in the insulation parameter (Dimosim is correct, other models have too high insulation values).
  • This will be changed after paper submission due to lack of time right now.
• Solver aspects: 3 different simulations with cvode, dassl and euler (1s).
  • Big deviations after summer non-heating period
  • Differences etween all solvers, but dassl deviations are the largest
  • Due to error in table handling (timetable is problem): integrator block behind table to fix time variables?
  • Euler gives best result, but takes very long to simulate (2h versus few minutes). This is not fast enough for the envisioned purpose.
  • How to solve?
    • Make issue in MSL repository? Issue is known, but core developers disagree on how it should be solved
    • Ask Michael and/or Filip?
    • Is it a problem inherent to pipe or do all pipes have this error?
    • We should try to simulate a smaller case that is easier for debugging
    • Michael will open new issue inside IBPSA project first
• Next steps?
  • Cooling demand
  • Pipe network structures
  • Different flow temperatures, varying flow temperature, different control strategies
  • But first make sure everything is running properly
• Aim of comparison exercise: comparing/benchmarking different

TO DO

• Michael: make new issue in IBPSA Project 1 repo to discuss solver differences
• Bram: set up doodle for another coordination meeting before end of January?