The predictability of lumped BES models, a case study

Transcription

1 The predictability of lumped BES models, a case study van Schijndel, A.W.M. Published in: Expert Meeting of the International Energy Agency Annex 58, 2-4 April 2012, Bilbao, Spain Published: 01/01/2012 Document Version Accepted manuscript including changes made at the peer-review stage Please check the document version of this publication: A submitted manuscript is the author's version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. The final author version and the galley proof are versions of the publication after peer review. The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publication General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Users may download and print one copy of any publication from the public portal for the purpose of private study or research. You may not further distribute the material or use it for any profit-making activity or commercial gain You may freely distribute the URL identifying the publication in the public portal? Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Download date: 13. Jul. 2018

2 The predictability of lumped indoor climate models, a case study A.W.M. van Schijndel Eindhoven University of Technology, Department of the Built Environment, Unit Building Physics and Systems, Eindhoven, Netherlands ABSTRACT: The indoor climate is the key element that combines crucial issues like energy and thermal comfort. Lumped models are often used to simulate the indoor climate. The objective of the paper is to investigate the predictability of these type of models. The research approach was to study a well measured and documented case and try to extract some general rules from it. Preliminary results show that: (1) 90% of the time the plant energy measurements are in between the simulated bandwidth when all parameters are varied according to most uncertain case scenario. (2) Similarly, 30% of the time the relative humidity measurements are within the simulated bandwidth. (3) In this specific case, simulation results seem most sensitive for changes in insulation material layer thickness, ventilation rate and the internal surface heat transfer resistance. Most important and contrary to the expectations, it seems that even in the most uncertain scenarios, the measurements do not fit within the simulated bandwidths. 1. INTRODUCTION People spend most of their lives indoors, especially at home. On average they spend about 16 hr/day during the week and 17 hr/day during the weekend inside their homes (WHO 2003). Today, primary energy use in the built environment accounts for about 35% of total EU energy consumption (EU 2008). In housing about twothirds of energy used is required for heat including space heating. The indoor climate is the key element that combines the just mentioned crucial issues. Furthermore, lumped models are often used to simulate the indoor climate. The objective of the paper is to investigate the predictability of these models. When should we measure more, and when should we model better? The research approach was to study a well measured and documented case and try to extract some general rules from it. More specific the methodology consisted of the following four steps: First, simulation results of the energy consumption and relative humidity were compared to real measurements adapted from the IEA Annex 41 (Two coupled test rooms located at the outdoor testing site of the Fraunhofer-Institute Holzkirchen, Germany). Second, to assess the overall sensitivity of simulation results due to uncertainties in input parameters, a number of sensitivity scenarios were created. Third, Monte Carlo Analysis (MCA) was selected as technique for obtaining the total sensitivities in the predictions. Fourth, the predictions produced by sets of input values were repeated many times. Since all the inputs were perturbed simultaneously the method fully accounted for any interactions between the inputs and, in particularly, any synergistic effects. The Sections of the paper is organized similar to above mentioned research approach. 2. MODELING AND SIMULATION OF THE REFERENCE INDOOR CLIMATE The whole building model originates from the thermal indoor climate model ELAN which was already published in 1987 (de Wit et al. 1988). Separately a model for simulating the indoor air humidity (AHUM) was developed. In 1992 the two models were combined (WAVO) and programmed in the MATLAB environment (van Schijndel & de Wit 1999). Since that time, the model has constantly been improved using newest techniques provided by recent MATLAB versions. Currently, the hourly-based model, named HAMBase, is capable of simulating the indoor temperature, the indoor air humidity and energy use for heating and cooling of a multi-zone building. The physics of this model is extensively described by de Wit (2006). Furthermore, it is part of the so-called Heat, Air and Moisture Laboratory (HAMLab 2010), described by van Schijndel (2007, 2009). The reference indoor climate, based on a test building in the framework of IEA Annex 41 (2008), is used in this Section. Measured data are obtained from two test rooms which are located at the outdoor testing

3 site of the Fraunhofer-Institute of building physics in Holzkirchen. Geometry and construction of both rooms are identical. The geometry of one of the rooms is shown in Figure 1. Figure 1. One of the Fraunhofer test rooms The front façade is oriented south. The back wall, the walls separating the test rooms and the roof can be considered adiabatic. One room (the reference room) is plastered with a common used gypsum plaster and in the other room (the test room) the walls and ceiling are covered with aluminum foil. The concrete floor is covered with linoleum and has a small influence on the air humidity. In the test room hygroscopic material can be fixed on top of the aluminum foil at the walls and ceiling. The rooms were heated by electric heating and controlled on 20ºC air temperature. The air change rate was determined with a blower door test at 0.65±0.05 h-1. The moisture production is 2.4kg/day: 25gr/hour from 6 to 8 AM and 400 gr/hour from 16:00 till 22:00. The measurements were carried out during a winter season. The measured data of the hourly mean indoor relative humidity and the hourly mean total energy demand (sum of heating and energy for evaporation) were obtained from IEA Annex 41 (2008). In Figure 2 the measured and simulated results of the reference room are presented. The results of the test room are similar and not shown in this paper. Figure 2. The measured and simulated RH in the reference room

4 The hygroscopic plaster has an important influence on the amplitude of the variation as can be seen from the results of the test room in Figure 2.2. The results of simulation agree well with the measurements (mean error less than 4%). In the reference room the amplitude of the simulation is 10% smaller. This was also concluded in the exact (analytical) test. The HAMBase model is damping 10% more. Also the linearization of the sorption curve and the constant vapour permeability introduce some errors. A comparison of the simulated heat supply and the measured one is shown in Figure 3 Figure 3. The measured and simulated heating power [W] in the reference room Differences can be expected by the infiltration, the one-dimensional approach of the heat flow through the construction, the variation of the indoor temperature and the model approximations. Nevertheless the mean difference between simulation and experiment equals 10W and is less than 2% of the measured mean heating power. 3. METHODOLOGY OF IMPLEMENTING SENSITIVITY 3.1 Scenarios To assess the overall sensitivity of simulation results to uncertainties in input parameters, five sensitivity scenarios were created. Lomas et al. (1992) identify three sensitivity analysis techniques to predict the total sensitivity in both hourly and daily average predictions, due to uncertainties in input parameters: differential sensitivity analysis, Monte Carlo analysis (MCA), and stochastic analysis. It was found that MCA is the preferred technique for obtaining the total sensitivities in the predictions. In general, 10 parameters were grouped into six scenarios (see Table 1, Appendix) to assess the sensitivity of the HAMBase simulation results. 3.2 Selection of input parameters To determine the sensitivity of the results produced by HAMBase, first an overview was created of the input parameters which are required for specifying a HAMBase-model in the MatLab environment. Secondly, the input parameters which were expected to have an impact on simulation results were selected and listed in Table 2. It should be noted that there is considered to be no solar radiation entering the room. This implicates that all parameters which are affected exclusively by solar radiation may be excluded from this sensitivity study. Two input parameters were expected to influence the plant energy usage due to better or worse thermal insulation of the rooms: material layer thickness of the insulation material of the external wall and the U-value

5 of the glass. The insulation layer might (locally) be thinner or thicker than expected, and the U-value might be changed by the woolen blanket, which was used the keep the sun out of the rooms. The airflow inside the rooms is considered to be affected by both the ventilation and convection factor of the HVAC-system. Moreover, different values for ventilation were found and so the sensitivity for this parameter should certainly be part of this sensitivity study. Also it is unrealistic to expect a heating system to be fully convective, because part of the heat transfer is probably caused by radiation. The boundary conditions are expected to be affected by the internal surface heat transfer resistance (Ri). This parameter directly impacts the energy required for heating the test and reference room. Please note that HAMBase uses one Ri-value for calculating the U-value of the construction (air-to-air) and uses another Ri-value for the actual heat transfer calculations. Also, the internal vapour resistance (Zi) needs to be considered. This parameter affects the relative humidity of both rooms, but is not specified in the basic model. Besides these two boundary conditions both the internal vapour diffusion resistance factor (μ) and the internal water vapour diffusion equivalent air layer thickness (sd) should be part of this sensitivity study. Similar to the values for Ri and Zi, these two parameters are assigned a standard value in HAMBase, unless they are explicitly assigned a different value. Finally, the moisture production should be taken into account. In practice, it is quite difficult to produce exactly the amount of moisture as specified in the Annex 41 profile, especially at the start and end of a production period. The total amount of moisture (2.4 kg/day) is of relatively easily managed, and will not be addressed in this sensitivity study. 3.3 Minimal and maximal values The selected parameters were varied between a minimum and maximum value both separately and in a scenario using Latin Hypercube sampling. The following values were used (See Table 3, Appendix) The fluctuations in the moisture load were varied in a scenario. The standard moisture production was varied between 50 to 100 per cent in the first and last hour of a period using Latin Hypercube sampling. 4. SIMULATION RESULTS The basis for all HAMBase simulations in this paper was the existing HAMBase model, which was used to produce the original results. This file was converted into a MatLab-function and one of the eight selected parameters was replaced with a variable. For the simulation of the five scenarios, two parameters were replaced by variables. Figure 4. Flowchart for the simulation of sensitivities. As can be seen from the flowchart in Figure 4, for the simulation of one parameter the original HAMBase model was converted into a MatLab-function, in which the parameter values were replaced by a variable. In the input-file the minimal and maximal value were specified. When the input-file is executed, it calls the calculation-file two times: once for the minimal value and once for the maximal value. The results are saved in two files and which are used for plotting the sensitivity graphs. The simulation process for a scenario is roughly the same as for the separate parameters. There are two important differences: (1) not one but two parameters are replaced in the original model; (2) the input values are randomized between the minimal and maximal value using Latin Hypercube sampling. This last point requires 200 computations in stead of 2. From these

6 files six matrices are created: three for storing the relative humidity results, and three for storing the plant energy results. From these files the maximal and minimal sensitivity graphs are plotted. 4.1 Bandwidth for separate parameters To assess the sensitivity of the HAMBase results for (small) variations in the eight selected parameters, the resulting bandwidth is presented both in graphs and tables. The results in tabular form are presented in the Appendix. A complete overview of all sensitivity graphs can be found in (Schriek & van Schijndel 2007). Here, only the results for ventilation are presented as an example Testroom with aluminium HAMbase (max) HAMbase (min) Measurements 700 Plant power [W] Time [days] Figure 5. Sensitivity graph of plant energy in the test room for ventilation; average simulation bandwidth 47 W; measurements within bandwidth: 65 % Figure 5 compares the simulation results for ventilation with the real measurements. It can be seen that the real measurements of the plant energy occasionally lie in between the minimal and maximal simulation results (i.e. bandwidth) Testroom with aluminium HAMbase (max) HAMbase (min) Measurement Relative Humidity [%] Time [days] Figure 6. Sensitivity graph of relative humidity in the test room for ventilation; average simulation bandwidth 4 %; measurements within bandwidth: 19 %

7 The average simulation bandwidth, as well as the amount of time that measurements are in between the two simulated results, are calculated. These are the values which are reported under the sensitivity graph. Figure 6 is similar to Figure 5, but compares the relative humidity (RH) with the real measurements. Again, the two numbers under the graph represent the average bandwidth of the simulation and the amount of time the real measurements are between the simulated values. For the remaining zones and periods the results are analogue to these two graphs we refer to (Schriek & van Schijndel 2007). Table 4 of the appendix presents a general overview of the simulation results. 4.2 Bandwidth for scenarios The resulting bandwidth for (small) variations in the six selected scenarios is presented both in graphs and tables Testroom with aluminium HAMbase (max) HAMbase (min) Measurements 700 Plant power [W] Time [days] Figure 7. Sensitivity graph of plant energy in the test room (all parameters combined); average simulation bandwidth 106 W; measurements within bandwidth: 94 % Testroom with aluminium HAMbase (max) HAMbase (min) Measurement Relative Humidity [%] Time [days] Figure 8. Sensitivity graph of relative humidity in the test room (all parameters combined); average simulation bandwidth 6 %; measurements within bandwidth: 25 % Here, only the results for a combination of all parameters (scenario 6) are presented as an example. Again, Table 5 at the appendix presents a general overview of these simulation results.

8 5. ANNEX 58 CONTEXT DISCUSSION Preliminary results show that: (1) about 90% of the time the plant energy measurements are in between the simulated bandwidth when all parameters are varied according to most uncertain case scenario. (2) Similarly, about 30% of the time the relative humidity measurements are within the simulated bandwidth. (3) In this specific case, simulation results seem most sensitive for changes in insulation material layer thickness, ventilation rate and the internal surface heat transfer resistance. Most important and contrary to the expectations, it seems that even in the most uncertain scenarios, the measurements do not fit within the simulated bandwidths. It should also be noted that the total simulation bandwidth will further increase when uncertainties in the measurements of the Fraunhofer-Institute for plant energy and relative humidity are also taken into account. This would further improve the simulation accuracy. It is clear that the models including uncertainty are unable to predict bandwidths that contain the expected 100% of the time of the measurements. The discussion focuses now on the question: When should we measure more, and when should we model better? Looking at Figure 3, it is expected that if (realistic) uncertainty is included into the models, 100% of the time of the measured heating energy use should be within the simulated bandwidth. However, Figure 7 shows that in the most uncertain case, by combining all scenarios, 94% of the time of the measured heating energy use is within the simulated bandwidth. Similar to the previous Section, but now it is even worse. Figure 8 shows that in the most uncertain case, just 25% of the time of the measured relative humidity is within the simulated bandwidth. There is a discrepancy between the energy use and relative humidity percentages of time within the bandwidths (respectively 94% and 25%). A major difference between these two parameters is that the energy use is a true global quantity (dependent of energy flows all over the room) and the relative humidity is a local quantity consisting of one (or a couple more) sensor(s). In this specific case, the lumped model (HAMBase) is significant better capable in predicting the uncertainty in the global quantity (energy usage) than in the local quantity (relative humidity). This may lead to a more general conclusion that lumped parameter models are required for the most accurate simulation of global quantities and distributed parameter models for local quantities. REFERENCES EU Energy and transport in figures, statistical pocket book 2007/2008, ISBN HAMLab Lomas, K.J., Eppel, H Sensitivity analysis techniques for building thermal simulation programs. Energy and Buildings 19. Schijndel, A.W.M. van & Wit, M.H. de, A building physics toolbox in MatLab, 7TH Symposium on Building Physics in the Nordic Countries Goteborg, pp81-88 Schijndel, A.W.M. van Integrated heat air and moisture modeling and simulation. PhD thesis, Technische Universiteit Eindhoven, 200 pages. Schijndel, A.W.M. van Integrated modeling of dynamic heat, air and moisture processes in buildings and systems using SimuLink and COMSOL. Building Simulation: An International Journal, 2(2), Schriek, M.H.J., Schijndel, A.W.M. van The effect of uncertainties in the input parameters for step 3. IEA Annex 41 Report A41-T1-NL-07-3, 7th working meeting, Florianopolis, Brazil. WHO Housing and Health: Identifying Priorities, Meeting Report, October 2003, Bonn, Germany. Bonn, Germany: World Health Organization. Wit M.H. de,h.h. Driessen ELAN A Computer Model for Building Energy Design. Building and Environment 23: Wit, M.H. de, HAMBase, Heat, Air and Moisture Model for Building and Systems Evaluation, Bouwstenen 100, Eindhoven University of Technology Woloszyn, M. & Rode, C IEA Annex 41 report on Modelling principles and common exercises, 234 pages

9 APPENDIX: TABLES 1-5 Table 1. Overview of the sensitivity scenarios Scenario: Parameters: (1) Thermal insulation - Material layer thickness (d n ) - U-value of glass (U glass ) (2) Airflow - Ventilation rate (n) - Convection factor of the heating system (CF) (3) Boundary conditions - Int. surface heat transfer resistance (R i ) - Int. vapour resistance (Z i ) (4) Surface humidity - Int. vapour diffusion resistance factor (μ) - Int. water vapour diffusion equivalent air layer thickness (s d ) (5) Moisture load fluctuations - Moisture production at the start and end of period 1 - Moisture production at the start and end of period 2 (6) All parameters combined - All 10 parameters listed above Table 2. Overview the HAMBase input parameters Symbol: Property: Unity: Value: d n Material layer thickness (insulation external wall) m 0.07 U glas U-value glass (without sun blinds) W/m 2 K 1.10 U glasw U-value glass (with sun blinds) W/m 2 K 1.10 V min /v ma Ventilation (test room) 1/h 0.65 x V min /v ma Ventilation (ref. room) 1/h 0.65 x CF h Convection factor of the heating system - 1 R i Internal surface heat transfer resistance m 2 K/W 0.13 Z i Internal vapour resistance m 2 spa/kg (varies) μ Internal vapour diffusion resistance factor kg/kg (varies) s d Internal water vapour diffusion equivalent air layer thickness m (varies) G int Moisture gains kg/s Table 3. Maximum and minimum of the selected parameters Parameter Value Material layer thickness (dn): m U-value of glass (Uglass): W/m 2 K Ventilation rate (n): h -1 Convection of the heating system (CF) multiply factor:: [-] Int. surface heat transfer resistance (Ri): 1/10 1/6 m 2 K/W Int. vapour resistance (Zi) multiply factor: [-] Int. vapour diffusion resistance factor (μ) multiply factor: Int. water vapour diffusion air layer thickness (sd) multiply factor::

10 Table 4. Simulation results for separate parameters Parameter: Zone: Average simulation bandwidth: Measurements within bandwidth: d n = m U glass = W/m 2 K U glass = /m 2 K n = h -1 CF = R i = 1/10-1/6 m 2 K/W Z i = factor μ = factor s d = factor Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Test 42 W 59% Ref. 42 W 59% Test 39 W 4% Ref. 39 W 13% (No change) Test 15 W 23 % Ref. 15 W 23 % Test 14 W 1 % Ref. 14 W 6 % (No change) Test 47 W 65 % Ref. 47 W 65 % Test 45 W 4 % Ref. 45 W 15 % Test 4 % 19 % Ref. 4 % 10 % Test 4 % 31 % Ref. 4 % 15 % Test 8 W 9 % Ref. 8 W 9 % Test 9 W 1 % Ref. 9 W 4 % (No change) Test 47 W 70 % Ref. 47 W 70 % Test 39 W 3 % Ref. 39 W 12 % (No change) (No change) Test 1 % 5 % Ref. 0 0 Test 0 0 Ref. 0 0 (No change) Test 0 1 % Ref. 0 0 Test 1 % 5 % Ref. 0 0 (No change) Test 0 1 % Ref. 0 0 Test 1 % 9 % Ref. 0 0

11 Table 5 Simulation results for scenarios Scenario: (1) Thermal insulation: d n = m U glass = W/m 2 K (2) Airflow: n = h -1 CF = (3) Boundary conditions: R i = 1/10-1/6 m 2 K/W Z i = factor (4) Surface humidity: μ = factor s d = factor (5) Moisture load fluctuations: (moisture production varied between % in the first and last hour of production periods 1 and 2) (6) All parameters combined Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: Plant energy: RH: : Zone Average simulation bandwidth: Measurements within bandwidth: Test 47 W 65 % Ref. 47 W 65 % Test 43 W 4 % Ref. 44 W 15 % (No change) Test 52 W 64 % Ref. 52 W 64 % Test 50 W 4 % Ref. 50 W 17 % Test 4 % 18 % Ref. 4 % 10 % Test 4 % 31 % Ref. 4 % 15 % Test 48 W 72 % Ref. 48 W 72 % Test 41 W 3 % Ref. 42 W 12 % Test 1 % 6 % Ref. 0 % 0 % Test 0 % 2 % Ref. 0 % 0 % (No change) Test 1 % 2 % Ref. 0 % 0 % Test 2 % 14 % Ref. 0 % 0 % (No change) Test 2 % 3 % Ref. 3 % 16 % Test 1 % 10 % Ref. 3 % 13 % Test 106 W 94 % Ref. 106 W 94 % Test 96 W 6 % Ref. 97 W 29 % Test 6 % 25 % Ref. 6 % 33 % Test 5 % 39 % Ref. 6 % 24 %

12 Parameter: Min/max values: Scenario: dn (m) (1) Thermal insulation Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 16

13 Parameter: Min/max values: Scenario: dn (m) (1) Thermal insulation RH (Test room): to RH (Reference room): to Average simulation bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % 17

14 Parameter: Min/max values: Scenario: Uglass (W/m 2 K) (1) Thermal insulation Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): Period: to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 18

15 Parameter: Min/max values: Scenario: Uglass (W/m 2 K) (1) Thermal insulation RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Average simulation bandwidth: % RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % 19

16 Parameter: Min/max values: Scenario: n (1/h) (2) Airflow Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 20

17 Parameter: Min/max values: Scenario: n (1/h) (2) Airflow RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % 21

18 Parameter: Min/max values: Scenario: CF (-) (2) Airflow Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 22

19 Parameter: Min/max values: Scenario: CF (-) (2) Airflow RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Reference room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth: % 23

20 Parameter: Min/max values: Scenario: Ri (m 2 K/W) (3) Boundary conditions Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 24

21 Parameter: Min/max values: Scenario: Ri (m 2 K/W) (3) Boundary conditions RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth % Measurements within bandwidth: % Average simulation bandwidth: % 25

22 Parameter: Min/max values: Scenario: Zi (m 2 spa/kg) (±25%) (3) Boundary conditions Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: 0 W Plant energy (Test room): to Average simulation bandwidth: 0 W Plant energy (Reference room): to Average simulation bandwidth: 0 W Average simulation bandwidth: 0 W 26

23 Parameter: Min/max values: Scenario: Zi (m 2 spa/kg) (±25%) (3) Boundary conditions RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth 0 % Average simulation bandwidth: % 27

24 Parameter: Min/max values: Scenario: μ (kg/kg) (±10%) (3) Boundary conditions Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: 0 W Plant energy (Test room): to Average simulation bandwidth: 0 W Plant energy (Reference room): to Average simulation bandwidth: 0 W Average simulation bandwidth: 0 W 28

25 Parameter: Min/max values: Scenario: μ (kg/kg) (±10%) (3) Boundary conditions RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth % Measurements within bandwidth: % Average simulation bandwidth: % 29

26 Parameter: Min/max values: Scenario: sd (m) (±10%) (3) Boundary conditions Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: 0 W Plant energy (Test room): to Average simulation bandwidth: 0 W Plant energy (Reference room): to Average simulation bandwidth: 0 W Average simulation bandwidth: 0 W 30

27 Parameter: Min/max values: Scenario: sd (m) (±10%) (3) Boundary conditions RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth % Measurements within bandwidth: % Average simulation bandwidth: % Measurements within bandwidth: % 31

28 Appendix B: Scenarios 32

29 Parameter(s): Min/max values: Scenario: Number of samples: dn (m) (1) Thermal insulation n=100 (Latin HyperCubic) Uglass (W/m 2 K) Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 33

30 Parameter(s): Min/max values: Scenario: Number of samples: dn (m) (1) Thermal insulation n=100 (Latin HyperCubic) Uglass (W/m 2 K) RH (Test room): to RH (Reference room): to Average simulation bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % 34

31 Parameter(s): Min/max values: Scenario: Number of samples: n (1/h) (2) Airflow n=100 (Latin HyperCubic) CF (-) Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 35

32 Parameter(s): Min/max values: Scenario: Number of samples: n (1/h) (2) Airflow n=100 (Latin HyperCubic) CF (-) RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % Measurements within bandwidth: % 36

33 Parameter(s): Min/max values: Scenario: Number of samples: Ri (m 2 K/W) (3) Boundary conditions n=100 (Latin HyperCubic) Zi (m 2 spa/kg) (±10%) Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 37

34 Parameter(s): Min/max values: Scenario: Number of samples: Ri (m 2 K/W) (3) Boundary conditions n=100 (Latin HyperCubic) Zi (m 2 spa/kg) (±10%) RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % 38

35 Parameter(s): Min/max values: Scenario: Number of samples: u (kg/kg) (4) Surface humidity n=100 (Latin HyperCubic) sd (m) Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: 0 W Plant energy (Test room): to Average simulation bandwidth: 0 W Plant energy (Reference room): to Average simulation bandwidth: 0 W Average simulation bandwidth: 0 W 39

36 Parameter(s): Min/max values: Scenario: Number of samples: u (kg/kg) (4) Surface humidity n=100 (Latin HyperCubic) sd (m) RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % Measurements within bandwidth: % 40

37 Parameter(s): Min/max values: Scenario: Number of samples: Moisture production starts and ends (50% 100%) (5) Moisture load fluctuations n=100 (Latin HyperCubic) between 50 to 100 per cent Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: 0 W Plant energy (Test room): to Average simulation bandwidth: 0 W Plant energy (Reference room): to Average simulation bandwidth: 0 W Average simulation bandwidth: 0 W 41

38 Parameter(s): Min/max values: Scenario: Number of samples: Moisture production starts and ends (50% 100%) (5) Moisture load fluctuations n=100 (Latin HyperCubic) between 50 to 100 per cent RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % Measurements within bandwidth: % 42

39 Parameter(s): Min/max values: Scenario: Number of samples: All scenarios combined - 1, 2, 3, 4, 5 n=100 (Latin HyperCubic) Plant energy (Test room): to Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Test room): to Average simulation bandwidth: W Measurements within bandwidth: % Plant energy (Reference room): to Average simulation bandwidth: W Measurements within bandwidth: % Average simulation bandwidth: W Measurements within bandwidth: % 43

40 Parameter(s): Min/max values: Scenario: Number of samples: All scenarios combined - 1, 2, 3, 4, 5 n=100 (Latin HyperCubic) RH (Test room): to RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Test room): to Average simulation bandwidth: % Measurements within bandwidth: % RH (Reference room): to Average simulation bandwidth: % Measurements within bandwidth: % Average simulation bandwidth: % Measurements within bandwidth: % 44