Skip to content

2019 10 07 Task 3 Coordination Meeting

Jump to bottom Edit New page
InaDJ edited this page Jan 22, 2020 · 7 revisions

Contents

Task 3 Coordination Meeting

Date: October 7, 2019, 5pm Brussels time (UTC+2)

Agenda

  • Update on buildings subgroup
    • office building?
    • insulation standards of single-family dwelling?
  • Update on network subgroup
  • Action points?

Meeting information

Minutes

Present:

  • Alessandro Maccarini
  • Katy Hinkelman
  • Enora Garreau
  • Ina De Jaeger
  • David Rulff
  • Gaëlle Faure
  • Nicholas Long
  • Michael Mans (notes)

Excused:

  • Annelies Vandermeulen
  • Dirk Saelens

Agenda:

Buildings / Increasing complexity

  • Stochastic user profiles

    • Buildings, Dimosim, IDEAS already simulated the profiles, Discussion on detail will take place in the a buildings subgroup meeting
    • Initialise a new case report, all duplicate sections from the old report will be linked but not copied & pasted / will be copy & pasted when the first case report is finalized

  • Renovations / Retrofit:

    • First results shown by IDEAS leading to a very high impact of the renovations. Hughe differences there, we will have a further look there.

  • Office:

    • Start with Dirks proposal and 5 zones, maybe refined in later cases or with different modelling approches / 1-zone vs 5-zones / orientation based vs usage based

  • making results available:

    • Michael will ask Michael W. how to proceed with that. Should we host and only link the files to the github page for example?

Networks / next steps

  • Network layouts:

    • 2 more layouts one with 8 buildings and one with 32 buildings

  • First use case:

    • Final result comparison needs to be provided by Michael M.

  • Michael will send a doodle, katy and gaelle will join the network group

Action points

  • Ina
    • Send a doodle for a buildings meeting and next coordination meeting
    • Discuss more in-depth the differences for the 16 occupants with Enora and Alessandro
    • Take a look at the different renovations and their simulation results
    • Send Dirk’s office to Michael
  • Michael
    • Send a doodle for a networks meeting

Doodles for next meetings

Building modelling subgroup meeting

Date: October 28, 2019, 5pm Brussels time (GMT+1)

Agenda

  • Building model comparison
  • Office definition

Present

  • Hicham Johra
  • Vasco Zeferina
  • Alessandro Maccarini
  • Konstantin Filonenko
  • Michael Mans
  • Jes Stershic

Minutes

Building model comparison / Hicham’s proposal:

  • Hicham proposed to use the

    • median (or alternatively the mean average) of the all model’s output as a reference
    • Use the model vs reference graph to graphically and qualitatively assess correctness of the tested model (are all the points on or close to the 45deg angle strait line Y=X)
    • angles between linear regression models for different simulation tools wrt the reference
    • additional indicators to quantify the dynamics of the system, incl. those over the specific time step, fex. RMSE of the amplitude over 24 hours
    • the above can also be applied, when the average is calculated for the output of the specific modeling tool and taken as a reference instead of the total mean (for all the tools) to see how the new case fits to the reference and if there is any constant bias
    • this is taken from Hicham’s experience of checking sensors behavior and calibration correctness in experimental investigations
    • Have a look at Google Document with examples

  • Michael pointed out that the approach is similar to coefficient of determination and simple but effective

  • Konstantin said the approach is good for visualization even if it is not new conceptually

  • Alessandro/Michael: Integrate the visualization in the report, that we could use few indices, 2 or 3 – not all of them, to keep the description simple (also many of the parameter are visible from the graph, like median, mean and angle). Hicham commented that mean and RMS could be the best two choices.

  • Actions on Hicham’s proposal (Michael):

    • more or less copy-pasted the proposal in the existing report (but first discuss it with everyone and see next point)
    • ask everyone give their revisions to the document and give their input (as comments in the document linked above)
    • discuss at the next coordination meeting (include in the agenda?), especially with Dirk and Ina
    • Should we pick average as a reference and RMSE and RMSE of the amplitude over 24h as KPIs?

Office Building Modelling

  • Office building definition:

    • document link was sent before about office building, please see the recent
    • M will integrate additional information about the building, fex that we don’t simulate the whole building, just 3 floors.
    • We can already now start integrating the geometrical data into the table of the Office proposal document. M did not start the simulation yet, but he will start very soon.
    • does somebody want to help out with filling up the document, in the first place, the information in the tables – mostly copy pasting work. M will send around the TEASER file with materials and usage in json format. The geometrical information with all the information about the building (such as floor plan) is already integrated in the document

  • Alessandro: what is the number of thermal zones we need to model? Michael answers based on typical floor layout for the office Dirk sent, which can proposed in some of or all of the following 3 approaches:

    • simple case (northern/southern zone, 2 for offices and 1 internal zone, doubt: 3 thermal zones in total or just in a single floor? This should be clarified with others);
    • intermediate case, still simplified, but with larger amount of zones: probably picke up later on it, now start with the simplest case a.;
    • complex case: detailed with different office zones and in between some stairs, office zones with 2 orientations (this one is too detailed for the urban scale).
    • most people said that they want to start modeling the a.

  • Vasco: why not consider a single zone with aggregated parameters and synthetic boundary conditions for the whole office building? (inspired by Energy+)

    • M: if we take this approach, how do we take into account that different zones are used for different purposes, difference in heating/cooling, also can boundary conditions of different kinds be taken properly into account in different surfaces?
    • K: in the libraries we want to use, there is a nice opportunity to couple zones, which we want to explore -> not very interesting to aggregate everything in A single zone. However, can Vasco (1) simulate one of the simplified models from point 2 (3 zones in total) as a single zone in E+, (2) compare to Modelica results, which other people obtain with different libraries? Vasco said that we can try this.
    • M: Maybe doing single zone model in E+ is not the best idea, since we have to agree on a specific exercise for everyone and models with different number of zones constitute different exercises

  • Michael: how to model the internal zone in the simplified model we have agreed on?

    • after discussion it was agreed to consider the internal zone as an empty space, a corridor
    • differences: another usage of the zone than of the office space, boundary conditions
    • Alessandro said would be nice to include loads later showing how the zone is used

Actions on the office building:

  • Michael will send the proposal for the office, comparison of different simulation tools in this respect and the office building slides
  • People start modeling the simplified 3 zone building as a common exercise (2 office spaces, 1 corridor – empty space for now) – probably the deadline for the exercise will be given, when more people join in the subsequent meetings?

Excluded topics:

  1. Stochastic user behavior has to be discussed at the next building meeting (maybe Ina will make an extra meeting for this for those who is involved in this part)
  2. Renovation for the single-family dwelling- energy demands were a bit high, Ina will look into it and then again discuss in one of the next meetings

Meeting information

Add a custom footer

Wiki

Main project site

https://ibpsa.github.io/project1

Clone this wiki locally