Skip to content

2018 09 05 Task 3 Coordination Meeting

Jump to bottom Edit New page
InaDJ edited this page Sep 6, 2018 · 21 revisions

Contents

Task 3 Coordination Meeting

Date: September 5, 2018, 5pm Brussels time (UTC+2)

Agenda

  • Overview of the activities
  • Research plan
  • Preparation for Paris meeting

Meeting information

Minutes

Present

Present: Dirk Saelens (notes), Konstantin Filonenko, Krzysztof Arendt, Ilaria Vigna, Ina De Jaeger (notes)

Apologies: Bram van der Heijde, Annelies Vandermeulen, Alessandro Maccarini

Overview of activities

  • Building Modeling team
    • Different separate meetings took place
    • Alessandro: looking into the deviations between the IDEAS building model and the Buildings building model. In contact with Michael W and Filip J. Current deviations: probably caused by solar radiation.
    • Paris meeting: looking into the overall model
  • Network Modeling team
    • Different separate meetings took place
    • Konstantin: works with Bram on modeling pipe network by hand. Some issues with temperature drop within the network. Felix has also developed a model for the pipes. Felix compared different pipe models of the different libraries. He was going to upload it on GitHub. Konstantin agreed with Bram to continue the work and to debug in Paris, so they can discuss in person.
  • Action points
    • Ina and Felix to provide some slides with an overview of both subgroups to Dirk. At the Paris meeting, Dirk has to present the progress in the plenary session. He’ll show the overall model that we have developed. He’ll show that we already started the modelling work and got some first results. To this end, he asks for 2 or 3 slides per subgroup with the main results. They, of course, welcome the input of all participants. Konstantin will send a wrap-up of his work to Felix. Felix compiles and sends to Dirk. Alessandro should also send a wrap-up of his work to Ina. Ina compiles and sends to Dirk.

Research plan

  • In Paris we have to finalize the research plan.

  • Some ideas on the research plan were already discussed:

    • Description of the increase in complexity that we want to achieve. The idea is to start simple, but gradually increase the complexity (for both the building models and the network models).
    • We started with a decoupling between modeling the heat demand of the neighborhood and modeling of the network. The research plan should also include a coupling of both.
    • Compare simulation against simulation is the first step. Of course, if deviations are too large, data would be helpful to see which models comes closest to the measured data.

  • Dirk will draft a first version of the research plan. We’ll discuss and finalize it with the WP3 group at the Paris meeting.

Preparation for Paris meeting

  • Call for papers (is actually WP3.1)
  • 3 sessions
    • First session to present our work outside of the project in short + to present the work of the building modelling subgroup + to present the work of the network modelling subgroup
      • Presentation on the work of Krzysztof: Centralized MPC of district heating and cooling networks modeled in Modelica: TERMONET project (Krzysztof Arendt, Konstantin Filonenko, Christian T. Veje)
      • Presentation on the work of Ina
      • Presentation on the work of Bram and Annelies
    • Second session to work within the subgroups
    • Third session to discuss and conclude with the whole group
  • Peter Riederer expressed the interest of CSTB/Efficacity in the DESTEST work and will join the Paris meeting
  • No need to invite other contributions for presentations?
    • Contributions are welcome, but we also would like to spend some time on really working together (debugging etc.)

Network modelling subgroup meeting

Date: 31st July 2018, 10:00 AM (UTC+2)

Agenda

  • Discuss automatic model generation using uesgraphs
  • First comparison of simulation results
  • Discuss research plan
  • Discuss contribution to IBPSA Conference paper on DESTEST
  • Free presentations in Paris
  • ... (Feel free to suggest other agenda points)

Meeting information

Minutes

Present

Present: Felix Bünning, Annelies Vandermeulen, Bram van der Heijde (notes)

Comparison of simulation results

  • Shortly discussed Konstantin's simulations and bug (boiling or freezing medium)
    • Boiling could be caused by addPowerToMedium=True for very slow flow speeds, where heat is dissipated in the fluid.

Discuss contribution to IBPSA Conference paper on DESTEST

  • Annelies explained plans to submit a DESTEST Abstract to IBPSA conference 2019.
  • Dirk is working on abstract and will send it out to all WP3 participants to get feedback and open for contributions to the paper.

Free presentations in Paris

  • Presentation on simulation results and comparison of splitter/IBPSA pipe model

Actions

  • Bram: ask Michael Mans if he is going to implement UESgraphs
    • Felix can compare with his own package
  • Bram: send a mail to check for interest in another breakout meeting because of low presence during this meeting.

Building modelling subgroup meeting

Date: July 31, 2018, 5pm Brussels time (UTC+2)

Agenda

  • Cause of the discrepancies? (Alessandro)
  • Report on proposals for other typologies
  • First draft of research plan
  • Suggestions for free presentations in Paris
  • Next steps

Meeting information

Minutes

Present

Present: Alessandro Maccarini, Ilaria Vigna, Ina De Jaeger (notes), Michael Mans

Discrepancies between IDEAS model (Ina) and Buildings model (Alessandro)

  • Alessandro
    • Talked to Michael and Filip, but did not yet come to a conclusion (holiday period!)
    • Issue on Github: 3 minutes delay between IDEAS and Buildings
    • Results are very similar in free-floating for 1 room with 1 window on the south. As soon as window is on other facades, discrepancies increase (so it must have something to do with the solar gains; maybe reflection on surfaces is different; also when the sun is entering and leaving the building?)

Progress on building typologies

  • Everyone updates the progress on their typology
  • Everyone will put the demand profile in csv format on the github, once ready prepared a typology
  • Everyone also intends to model the other typologies in their library to check the discrepancies
  • Michael:
    • Multiple zones in TEASER are possible. However, the zones are decoupled (no thermal heat exchange, no infiltration air exchange)
    • Typologies in TEASER: office building and institute building
  • Alessandro:
    • Focused on an office building
    • The TEASER office building is a six zone model (50 percent is office, 25 percent is floor, …). We all agree that this is a good template, as it has different uses (office, storage, rest rooms, meeting rooms, …)
    • Looked up the report that is given as a reference for the office building, but it's in German
    • Michael: they indeed modelled this office building, according to the reference of the former BMVBS (now BMVI German Ministry of Transport and Digital Infrastructure). The BMVBS has investigated the non-residential building stock in Germany and has come up with this definition of an office building. In the IBPSA Project 1 (WP2), the IWU are also working on renewing this study (and gather more information on the non residential building stock). So, we can collaborate on that point. Additionally, in TEASER, there is the typology of the institute buildings. To setup this typology, they performed a regression analysis of the 200 institute buildings of the Jülich campus, which resulted in the definition of the areas of the different zones.
    • Ina: nice that it's not just one real building, but it's based on 200 buildings
    • Alessandro: Michael and his group performed a regression analysis, but it's based on data. Where does this data come from? Is the data only recorded for the purpose of building energy modelling or is it setup for another purpose?
    • Michael: I can look up some additional information for the next meeting, but I thought they recorded the data because they wanted to get an idea about the general areas as well as the technical installations inside the buildings. Michael will look up the additional information and update Alessandro Results: Within the scope of the research project, the geometric dependencies between the net ground area and the energy-relevant building envelopes are examined in order to simplify the geometric analysis for the creation of energy certificates in the non-residential building sector. So it was indeed a energy related purpose.
    • Ina: Alessandro, are you planning to model this office building in the Buildings library manually or to implement the automatic export from TEASER to the Buildings library?
    • Alessandro: Yes, it's the plan to model, but is it that straightforward to automate it? Ina thinks it is, but Michael raises the fact that it is easy for single-family dwellings, but it might not be for other types, such as offices (because of the air handling unit that is specific for AixLib). Alessandro and Michael will work on it together when Alessandro visits Aachen in September.
  • Ilaria
    • Focused on an apartment building
    • Approach currently works for the building structure only, not yet for the heating system
    • Apartments are currently not thermally coupled (adiabatic walls between the apartments)
    • Apartment definition will make use of the typologies (construction layers and materials) of Italian TABULA project
    • Work in progress for conference paper in December (BS2019)
    • Michael: apartments, that are not thermally coupled, are a huge drawback on apartment level, but he thinks the influence on building level or district level will be okay. So, if you want to compare the single-apartment heat demand, it will be a big influence (indicated by the user), on urban level: it might be okay
    • Ina: personal interest: I want to know what the influence is
    • Michael: very interesting, in his experience, the number of persons and the set-points are most important. They investigated measured data in apartment blocks and they found that these are the largest drivers (it's actually really hard to do some correlations: between set point, air temperatures, and number of occupants). Michael will forward an article in German on this topic to Ina
  • Ina:
    • Shifted to German materials
    • Modelled different ensemble of real buildings in Genk (Belgium)
    • 2 SD: Gustave Francotte straat 9-11 > OLD + Gierensstraat 49-51 > NEW
    • 2 SD + 1 T: Albert Forgeurstraat 8-10-12 > OLD
    • 2 SD + 2 T: Postwinningstraat 17-23 > NEW
    • 2 SD + 4 T: AugustCollonstraat 14-22 > OLD
    • Discussion of these ensembles actually coincided with the discussion on the work plan for the buildings
    • In short: single-family dwellings don't vary that much in geometry (net leased area is always between 100 and 200 m²), whereas it differs way more for office buildings (from 200 to 2000 m²). Therefore, we concluded that 1 terraced, 1 semi-detached and 1 detached single-family dwelling should suffice

Proposal for work plan?

  • We need to present our research plan in Paris. How do we see it? Dirk will also submit an abstract to BS2019 in Rome, discussing our plans. Contributions to this paper will be discussed more elaborately during the next WP3 Coordination meeting.
  • Proposal for work plan
  • 1: 16 nodes with identical loads
    • Old detached dwelling
  • 2: 16 nodes with different loads
    • Single-family dwellings (1 terraced, 1 semi-detached and 1 detached)
    • Office building
    • Apartment building
    • Need for other typologies?
      • Ina: TEASER institute buildings
      • Michael: for sure, TEASER institute buildings can be contributed as typology. The main difference between institute buildings and other buildings are the laboratory parts (AHU, so different heating and cooling demand). We can use it to get different load profiles, but not sure if it's a good idea. They use the institute buildings, but it's for a research campus, not sure if it's representative for typical districts
      • Ilaria: most typical in cities are these considered 3 typologies (single-family dwelling, office building, apartment building)
      • Michael: agrees
      • Alessandro: these special buildings are not really common. Nevertheless, they might be interesting due to totally different loads, so maybe it could be interesting to consider them in the future
      • Alessandro: supermarkets/commercial buildings can also be part of a typical district
      • Ina: how to model supermarkets? Anyone an idea?
      • Alessandro: maybe look in American guidelines/standard (?) for a definition, probably there will be a commercial building in there…
      • Ina: we'll put it on our dream list!
    • Ilaria: what is the role of the cooling demand and the cooling system? Do we consider it or not?
      • Ina: good point! Cooling demand should be included, since we are looking on the district level, where we have offices with a high cooling demand and housing with a much lower cooling demand. However, including cooling will make the network more complex, so to be discussed with the network group
      • Michael: it would indeed be nice to have, but in a later step: it introduces a bi-directional network, which is complex. It will take some time to introduce such models in the network. However, we can keep it in mind when implementing our models: turn on cooling, if cooling needed: turn on, but for now off!
    • Need for other construction years?
      • Ina: I think it's important to introduce some variations. This is partially through variation in thermal performance of the building envelope (related to the construction year), partially through the variation in geometry and partially through variation in occupancy
      • Michael: he's in favor of using different construction years. We don't need to use really really different geometries for the office buildings, but maybe different percentages for the zoning and different total net leased areas (200 to 200 000 m²). For the SFDs, no variation in geometry (1 terraced, 1 semi-detached and 1 terraced will suffice)
      • Everyone agrees
    • What about the occupant behaviour?
      • Ina: currently we use standard occupants. Shouldn't we shift to stochastic occupants?
      • Michael: influence of occupants is definetely big, so we should use different user profiles! But he's not sure what's the best method to setup these different profiles. We could use some stochastic profiles of Richardson (Residential and UK Based).
      • Ina: I'll put it on the list
  • 3: different layout of the district: x nodes
    • Multiple definitions of typical neighbourhoods?
      • Ina: Urban tissue + Suburban district + Country side district ? All three in old and new ?
      • We all agree on 3 to 4 different districts. Size? Layout? > to be discussed
      • Ina: the size of the district depends on how we model the district. I think we split up the building and the network and decided to communicate via CSVs in the beginning for the sake of simplicity. I think we should aim for coupled simulations between buildings and network, but that has a high computational cost.
      • Michael: what's the advantage of a coupled simulation?
      • Ina: building-network interactions
      • Michael: he does not know if the interactions are that important if you only couple the heat demand directly to substation and network? The largest dynamic influence (that you are able to capture with coupled simulations) will be indicated by the heating system inside the building (so then we also need to model the radiators inside the building and it gets even more complicated, with even less buildings in our district)
      • Michael: influence would indeed be nice, so maybe we can combine this, since we already agreed on modelling 3 to 4 districts > districts with different scales
      • Ina: so different building models for different districts: 1 districts with only the csv on the substation (100 to 200 buildings), 1 district with detailed building models with the building systems coupled to the network (10 to 20 buildings). We think both will be very interesting from a network point of view
      • We all agree

Free presentations

Alessandro and Ina will be present in Paris, Michael and Ilaria have to check their agenda/funding

Actions

  • Everyone proceeds with the definition or modelling of their typology
  • We discuss the work plan on the coordination meeting with the whole group
  • Michael informs Alessandro on the purpose of the data collection of the Jülich campus
  • Michael forwards an article in German on the regression analysis of measured data in apartment blocks to Ina
Add a custom footer

Wiki

Main project site

https://ibpsa.github.io/project1

Clone this wiki locally