Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

Please note that all times are shown in the time zone of the conference. The current conference time is: 21st May 2022, 18:05:00 CEST

 
 
Session Overview
Session
Session T3.7 (Online Track): Ensuring high quality building simulations
Time:
Thursday, 02/Sept/2021:
13:30 - 15:00

Session Chair: Michael Kummert, Polytechnique Montréal
Location: Virtual Meeting Room 1

External Resource: Click here to join the Zoom Meeting
Show help for 'Increase or decrease the abstract text size'
Presentations
13:30 - 13:37

Development of energy use benchmark data with respect to locations, housing types, gross area and construction year from household energy standing survey in Korea

Na Hyeon Lee, Dong Hyun Seo

Chungbuk National University(CBNU), Korea, Republic of (South Korea)

Aim and Approach

(max 200 words)

The aim is to present residential energy use features that can be used in various fields by analyzing monthly consumption of electricity and fuels on the basis of principal variables such as location, housing type, gross area and construction year.

[Approach]

- Architectural properties and residential energy monthly consumption data of sampled houses are collected from the Year 2016 microdata of the Household Energy Standing Sample Survey (HESS). Quality control process is carried out to remove the outlier consumption data.

- Energy consumption is analyzed and energy use features are presented through systematic analysis criteria based on the locations, housing types, gross area, and construction year, which are considered to be principal factors of energy consumption.

- A literature review on residential energy use features is analyzed. And then, the authors identify the current research practices and determines the reliability of HESS energy consumption data which is generated in this research.

Scientific Innovation and Relevance

(max 200 words)

- According to statistics from the EIAS of Ministry of Land, Infrastructure, and Transport in 2017, residential buildings account for about 44% of total energy consumption in Korea. Therefore the government has been enforced with a strong building efficiency policy over the past decade. Despite the efforts of the government, there is a lack of data to identify the energy consumption landscape of residential buildings in Korea.

- This study contributes to presenting detailed residential energy use features through a new analysis system that has not been used in Korea by selecting location, housing type, gross area, and construction year as the main variables. Detailed residential energy use features data to be presented as a result of this study are expected to be more useful to researchers and policymakers as well as to the general public.

Preliminary Results and Conclusions

(max 200 words)

Monthly fuel and electricity consumption are analyzed based on the four major variables such as location, housing type, gross area, and construction year. By doing so detailed energy use benchmark data are derived for residential buildings in Korea.

Main References

(max 200 words)

- Lee, N., Kim, H., Seo, D. (2019). Analysis of Residential energy Use Features with Respect to Location, Housing Type, Gross Area and Construction Year from Household Energy Standing Survey, Journal of KIAEBS, 13(6), 536-549.

- Raffio, G., Isambert, O., Mertz G., Schreier C., Kissock K. (2007). Targeting residential energy assistance, Proceedings of Energy Sustainability Conference, 489-95.

- Filippini, M., Hunt, L. (2012). US Residential Energy Demand and Energy Efficiency: A Stochastic Demand frontier Approach, Energy Economics 34, 1484-1491.



13:37 - 13:44

Integration of probabilistic methods and parametric tools for performance-based building design decision making

Fatemeh Shahsavari, Wei Yan

Texas A&M University, United States of America

Aim and Approach

(max 200 words)

Building performance simulation (BPS) tools are useful in the field of building design optimization but mostly fail to deal with uncertainties. The commonly-used BPS tools, such as EnergyPlus and TRNSYS, act solely upon deterministic sets of input data, disregarding the associated uncertainties (Suna et al., 2014). Solving optimization problems in the field of performance-based building design with deterministic approaches may oversimplify reality and lead to overestimation or underestimation of building performance (Yao, 2014).

Tian et al. (2018) identified the key sources of uncertainties in building energy simulation related to climate change, occupant behavior, and degradation of the HVAC system and building envelope materials over time. Uncertainty analysis methods such as Monte Carlo support robust optimization to find the optimum design alternative subjected to existing uncertainties.

The aim of this research is to deploy the capabilities of design parametric tools to integrate uncertainty analysis methods such as Monte Carlo in performance-based design decision making. The probabilistic methods versus deterministic methods allow considering different attitudes towards risk in design projects. Also, the probabilistic methods allow the calculation of expected value, expected utility, and confidence intervals to help designers make informed design decisions based on the project needs and existing uncertainties.

Scientific Innovation and Relevance

(max 200 words)

The probabilistic methods provide designers with additional information including the mean value and variance of the simulation result for different design alternatives. Also, other factors including expected value and expected utility inform designers about the benefit-cost of each design alternative under different possible conditions. This extra information allows designers to make informed decisions with regard to the requirements of the design project. For instance, occupant thermal comfort may be a higher priority in design compared to construction cost in a health care facility design project, while the order may be different in a storage facility design project.

Preliminary Results and Conclusions

(max 200 words)

Considering uncertainties in design input parameters with associated probability density functions and using Monte Carlo sampling methods allow performing building simulations for a range of inputs and getting the probability distribution of simulation output. We have tested a probabilistic method with a test case of three design alternatives for a modular classroom design and concluded that the ranking order of design alternatives obtained from the probabilistic methods may be different from what deterministic method suggests. Our previous work shows major discrepancies between deterministic and probabilistic results in building energy simulations.

This research will expand the previous experiment with more complicated design scenarios and studying the effect of different probability distribution types of design input parameters on the results. The probabilistic results may provide designers with better insight into design consequences under different possible conditions.

Main References

(max 200 words)

Suna, K., Yana, D., Hongb, D., & Guoa, S. (2014). Stochastic modeling of overtime occupancy and its application in building energy simulation and calibration. Building and Environment, 79(2010), 1–12.

Tian, W., Heo, Y., de Wilde, P., Li, Z., Yan, D., Park, C. S., Feng, X., & Augenbroe, G. (2018). A review of uncertainty analysis in building energy assessment. Renewable and Sustainable Energy Reviews, 93(January 2017), 285–301. https://doi.org/10.1016/j.rser.2018.05.029

Yao, J. (2014). Determining the energy performance of manually controlled solar shades: A stochastic model based co-simulation analysis. Applied Energy, 127, 64–80. https://doi.org/10.1016/j.apenergy.2014.04.046



13:44 - 13:51

Energy prediction impact of the Space Level Occupancy schedule for a Primary School

Yingli Lou1, Yunyang Ye2, Wangda Zuo1,3, Jian Zhang2

1University of Colorado, Boudler, United States of America; 2Pacific Northwest National Laboratory, United States of America; 3National Renewable Energy Laboratory, United States of America

Aim and Approach

(max 200 words)

This research is aimed to explore whether it’s necessary to develop heterogeneous occupancy schedules for building energy models. Both homogeneous and heterogeneous occupancy schedules can describe occupant status in the building. On one hand, the homogeneous occupancy schedule focuses on the total number of occupants in the building, and occupancy schedule is the same for all rooms [1,2]. On the other hand, the heterogeneous occupancy schedule describes occupant status in each room of the building so that each room has its unique occupancy schedule, which may improve the prediction accuracy [2,3]. Collecting data for the homogeneous schedule is easier than the heterogeneous schedule because it only requires an occupancy sensor in building entrance, while the heterogeneous schedule needs multiple sensors and it is labor intensive to identifying those data. Therefore, it is important to quantify the potential accuracy difference between these two schedules so that practitioners can select the schedule according to their needs.

This research uses primary school building as case study and develops homogeneous and heterogeneous occupancy schedules for it. Then, energy saving of primary school model [4] from ASHRAE Standard 90.1-2004 to 90.1-2016 using both schedules are calculated. Energy saving difference between these two schedules is compared.

Scientific Innovation and Relevance

(max 200 words)

In the conventional view, occupancy has minimal impact on building energy prediction for buildings without occupancy-based control (OBC). Therefore, most researches employ a homogeneous schedule to represent building occupancy [5]. However, occupancy does impact the required heating and cooling loads of a building [6,7], which will then impact the building energy performance. Thus, this research evaluates the energy impact of occupancy schedules for buildings without OBC. This research develops homogeneous and heterogeneous occupancy schedules and compared their energy predictions.

Building energy models are critical for various energy related applications, such as building energy performance evaluation, energy efficiency measures selection, and large-scale building energy prediction. Determining the impact of heterogeneous occupancy schedule on building energy modeling will provide a guideline for practitioners in building energy modeling to seek a balance between modeling accuracy and labor time.

Preliminary Results and Conclusions

(max 200 words)

In this case study, Energy savings of primary school from ASHRAE Standard 90.1-2004 to ASHRAE Standard 90.1-2016 using homogeneous and heterogeneous occupancy schedules are shown in Table 1.

Table 1. Energy saving of primary school from ASHRAE Standard 90.1-2004 to ASHRAE Standard 90.1-2016

Climate zones Homogeneous occupancy schedule (%) Heterogeneous occupancy schedule (%) Difference (%)

1A 44.29 45.55 -1.26

2A 43.67 44.53 -0.86

2B 36.57 37.66 -1.09

3A 41.34 41.95 -0.61

3B 36.85 37.60 -0.75

3C 42.81 44.33 -1.51

4A 32.97 32.69 0.27

4B 39.76 39.86 -0.09

4C 43.50 42.36 1.14

5A 47.06 47.42 -0.35

5B 40.75 39.96 0.79

5C 43.38 43.82 -0.44

6A 47.36 47.35 0.01

6B 45.33 45.46 -0.13

7A 47.68 47.56 0.12

8A 44.16 43.31 0.86

Occupancy has impact on building energy prediction for buildings without OBC. And the impact of occupancy on energy saving evaluation is more significant in hot areas (climate zones 1, 2, and 3).

Main References

(max 200 words)

[1]. Oldewurtel, F., Sturzenegger, D., & Morari, M. (2013). Importance of occupancy information for building climate control. Applied energy, 101, 521-532.

[2]. Chen, Y., Hong, T., & Luo, X. (2018, February). An agent-based stochastic Occupancy Simulator. In Building Simulation (Vol. 11, No. 1, pp. 37-49). Tsinghua University Press.

[3]. Agarwal, Y., Balaji, B., Gupta, R., Lyles, J., Wei, M., & Weng, T. (2010, November). Occupancy-driven energy management for smart building automation. In Proceedings of the 2nd ACM workshop on embedded sensing systems for energy-efficiency in building (pp. 1-6).

[4]. DOE. (2020). Commercial prototype building models. https://www.energycodes.gov/development/commercial/prototype_models

[5]. Ye, Y., Lou, Y., Zuo, W., Franconi, E., & Wang, G. (2020). How Do Electricity Pricing Programs Impact the Selection of Energy Efficiency Measures?-A Case Study with US Medium Office Buildings. Energy and Buildings, 110267.

[6]. Kwok, S. S., & Lee, E. W. (2011). A study of the importance of occupancy to building cooling load in prediction by intelligent approach. Energy Conversion and Management, 52(7), 2555-2564.

[7]. Kwok, S. S., Yuen, R. K., & Lee, E. W. (2011). An intelligent approach to assessing the effect of building occupancy on building cooling load prediction. Building and Environment, 46(8), 1681-1690.



13:51 - 13:58

Impact of ELA calibration methods on building energy model fidelity

Mohanned M Althobaiti1,2, Godfried L Augenbroe1

1Georgia Institute of Technology, Atlanta, USA; 2King Saud University, Riyadh, KSA

Aim and Approach

(max 200 words)

As building performance is increasingly improved and building energy consumption decreases, a greater percentage of the total energy loss of a building occurs through envelope leakage. This leakage is characterized by the effective leakage area or ELA, which is a proxy parameter to what is essentially a complex flow phenomenon through cracks driven by pressure differences. Current building performance simulation (BPS) uses software modules that approximately calculate envelope infiltration, but the literature shows that their calibration and validation is still unsatisfactory. In fact, calibration and validation of BPS models is still an important subject of study in our quest to improve the fidelity of simulation-based predictions in various applications. The high level of interaction and subsumption between parameters can result in a model that approximates the measurements well (and thus meets the ASHRAE auditing threshold) but whose “best estimates” of parameters are unreliable. This can be a problem in performance contracting when limits have been agreed on certain parameters such as ELA and U-value. It can also be problematic in the use of the model for certain performance assessments. This paper introduces underlying issues discusses results of direct and indirect calibration with different model fidelities.

Scientific Innovation and Relevance

(max 200 words)

The study focuses on the calibration of building energy models of existing buildings. It does so by conducting calibration for different experiments, i.e., with different data resolutions, and for different model fidelities. The calibration is anchored around ELA and its impact on “best estimates” of other parameters is verified. The study develops a new framework to address calibration and validation for different combinations of data and model fidelity, where each combination leads to probability distributions of the calibration parameter set. For each combination the ultimate aim is to determine the fitness of the resulting building energy model for given application studies such as building energy benchmarking, fault detection, unmet hour verification, etc.

The results are meaningful for better understanding façade infiltration and better understanding of the limits of calibrated models. The paper focuses exclusively on existing buildings, but its findings may lead to large scale data sets of calibrated ELA values in existing buildings, that may find their way into better ELA quantification in energy models of new designs.

Preliminary Results and Conclusions

(max 200 words)

A representative case study is conducted to demonstrate the approach. To explore the role of model fidelity, two tools are chosen, i.e., a reduced-order energy calculator and EnergyPlus, which represents the high-fidelity tool. The results demonstrate the effectiveness of the proposed calibration process in approaching the true values of the calibration parameters with focus on ELA.

Calibration techniques have encountered great improvement in recent years and are well supported by ASHRAE and numerous theoretical and field studies. The criteria laid down in the ASHRAE guideline are useful, but more work is needed to understand the special nature of calibration of building energy models. Such criteria stipulate ranges of admissible error in the total estimated energy consumption of a building but do not address underlying aspects such as uncertainty embedded in the models or initial guesses about the value of building a model. The paper discusses the criteria that need to be met in order to guarantee that a calibration not only meets the minimum required CVRMSE of the predicted outcomes, but also provides a close enough approximation of the best estimate of ELA obtainable with an experimental (tracer gas) set-up.

Main References

(max 200 words)

Augenbroe, G., Heo, Y., & Choudhary, R. (2011). Risk Analysis of Energy-Efficiency Projects Based on Bayesian Calibration of Building Energy Models. 12th Conference of International Building Performance Simulation Association, (pp. 2579-2586). Sydney.

Campolongo, F., Cariboni, J., & Saltelli, A. (2007). An Effective Screening Design For Sensitivity Analysis of Large Models. Environmental Modelling & Software, 22(10), 1509-1518.

Chong, A. (2018). Datasets, Bayesian Calibration of Building Energy Models for Large Datasets. Carnegie Mellon University. Thesis.

Costola, D., Blocken, B., Ohba, M., & Hensen, J. (2010). Uncertainty in airflow rate calculations due to the use of surface-averaged pressure coefficients. Energy and Buildings 42 881–888.

de Wilde, P., van der Vordan, M., Augenbroe, G., & Kaan, H. (2001). The Need for Computational Support In Energy-Efficient Design Projects In The Netherlands. Seventh International IBPSA Conference, (pp. 513-519). Rio de Janeiro.

de Wit, M. S. (2001). Uncertainty in Predection of Thermal Comfort in Buildings. Delft: Delft University of Technology.

Judkoff, R., & Neymark, J. (2006). Model validation and testing: the methodological foundation of ASHRAE Standard 140.

Reddy, T. A. (2006). Literature review on calibration of building energy simulation programs: Uses, problems, procedures, uncertainty, and tools. ASHRAE Transactions, 112(1), 226–240.



13:58 - 14:05

Development of Gray-box modeling technique for HVAC&R systems with automatic calibration process

Hyemi Kim1, Sumiyoshi Daisuke2, Youngjin Choi3

1Ph.D. Student, Faculty of Human-Environment Studies, Kyushu University, Japan; 2Associate Prof., Faculty of Human-Environment Studies, Kyushu University, Japan; 3Assistant Prof., Department of Architectural Engineering, Kyonggi University, Repubpic of Korea

Aim and Approach

(max 200 words)

Key challenges faced with the simulation tools in the building performance simulation(BPS) sectors include: (1) to create a detailed physical-based energy model using measured data, (2) to create a compatible model that data is collected in real time (3) to lack the technician or energy manager with the building simulation knowledge [1]. This study develops an automatic simulation tool of Heating, Ventilation, Air Conditioning, and Refrigeration (HVAC&R) systems based on physics characteristics using the measured data. However, the existing simulation modeling process is costly and time-consuming to collect data in buildings. There are some problems about the limitations of the items that correspond to the physical formula [2]. Moreover, the techniques of systematic calibration are required to reconcile the discrepancies between the predicted and measured data [3].The purpose of this paper is to develop the gray-box modeling method in HVAC&R systems entailed with the automated calibration process, and to validate the reproducing performance of the simplicity systems. At first, we create each equipment model, and implement the modeling of the whole primary systems. Each device model selects a physical-based parameter and those are automatically tuned by machine learning. The reproducibility of the simplified HVAC&R system model using TRNSYS is verified.

Scientific Innovation and Relevance

(max 200 words)

This study systematically founds an automatic modeling procedure of HVAC&R systems in central air conditioning systems. The developed modeling method has two typical characteristics that introduce Grey-box and automatic calibration methods. The existing research [1-4] of two methods have modeled a whole building energy simulation, not about HVAC systems. In addition, the most of research have developed an optimized algorithm or implemented FDD (Fault Detection and Diagnostics) about the specific system or equipment. Those are insignificant to grasp the status of operating system using a systematic modeling method that can cope with numerous systems. This study has the innovations: (1) to make a machine learning model of the physics-based on each equipment, (2) to systematize and categorize the modeling procedure so that the results of (1) can be extended to the system scope to respond various systems and equipment, (3) to improve the accuracy of the physical parameters by automatizing the stage of calibrating it one by one and making the whole steps accurate. The developed modeling method can identify each device’s operation in HVAC&R systems by the physics principle. Afterward, this method will be able to be adapt during the developing optimized algorithm and FDD procedure of each equipment.

Preliminary Results and Conclusions

(max 200 words)

This paper develops the modeling method of air-cooled systems in the initial stages of the whole study. The equipment adopted by the system are heat-source devices (e.g. turbo chiller, heat pump), heat storage tank, inverter pump, heat exchanger, etc. The development process of the modeling method of each device verifies the accuracy by using the various measured data. The modeling method in this study calculates the parameters based on the physics considering the minimized input data of each equipment, and it categorizes characteristics that are different from each manufacturer. Next, the selected parameter is tuned using the machine learning and calibrates by each hour. Each modeling output value is substituted into the input value of the other equipment, and modeling of the simplified air-cooled system is performed with TRNSYS. As a result, while the modeling of the whole system is performed, detailed operation situation will be grasped using the basic input variables. The HVAC&R system modeling method developed in this study uses fewer input variables and considers measured data items compared to the existing simulation method. Although the new data is unstable out of the range of the training data, and the output of the high accuracy is expected.

Main References

(max 200 words)

The main references in this study are as follows. Hong et al. and Coakley et al. are the researches about the calibration for reducing discrepancies during the steps of building modeling and simulating. Harish et al. and Afroz et al. are papers about the classified modeling method with White-box, Black-box, and Grey-box. More than 100 existing studies are organized and analyzed in each research. Based on these researches, this study specified the modeling target, method and process, and calibration method through comparison with the relevant studies.

[Reference]

1) Tianzhen Hong et al., Building Simulation: Ten Challenges, Building Simulation, March 2018

2) V.S.K.V. Harish et al., A review on modeling and simulation of building energy systems, Renewable and Sustainable Energy Reviews, 56 (2016), pp.1272–1292

3) Daniel Coakley et al., A review of methods to match building energy simulation models to measured data, Renewable and Sustainable Energy Reviews,37 (2014), pp.123–141

4) Zakia Afroz et al., Modeling techniques used in building HVAC control systems: A review, Renewable and Sustainable Energy Reviews, 83 (2018), pp.64–84



14:05 - 14:12

Development of a metamodel to assess building thermal performance for naturally ventilated residential buildings

Rodolfo Kirch Veiga, Letícia Gabriela Eli, Amanda Fraga Krelling, Marcelo Salles Olinger, Ana Paula Melo, Roberto Lamberts

Federal University of Santa Catarina, Brazil

Aim and Approach

(max 200 words)

Building energy simulation programs allow the assessment of the impact of different air conditioning systems into the total building final energy consumption, however, it requires high computational cost and effort. On the other hand, metamodels are quick and easy-to-use tools that support experts in decision-making to optimize the thermal performance of buildings. The main objective of this study is to develop a metamodel to predict the thermal performance of multifamily residential buildings. The thermal performance will be evaluated according to a percentage of hours where the operating temperature of the room is within a range of interest (PHFT). To develop the metamodel it will be necessary to define a parameterizable model. The development of the parameterizable model will consist of a sequence of simplifications made on the original base model. The simplifications that will be carried out have as main objective to reduce the amount of descriptions necessary to simulate the room of interest. Simplifications will be assessed through statistical analysis and heat balance in some critical cases. The computer simulation program EnergyPlus was adopted to obtain the PHFT for each case. The study emphasized the residential buildings located in São Paulo, Brazil.

Scientific Innovation and Relevance

(max 200 words)

The development of accurate models for the prediction of the thermal performance of buildings has been intensified regarding the user thermal conditions needs and, consequently, the high energy demand of buildings, directly related to the expressive energy consumption of artificial air conditioning system. Metamodels are quick and easy-to-use tools, capable of identifying low-cost solutions that lead to energy efficiency. These tools can assist not only in the early design stages of new building projects and retrofit projects, but also facilitate the development of thermal performance approaches. Therefore, the development of a metamodel that accurately describes the thermal performance of residential building is a relevant challenge, since it can be applied to an extensive group of buildings. In this way, a tool that can be widely applied becomes viable, indirectly generating an increase in the energy efficiency of the residential building sector.

Preliminary Results and Conclusions

(max 200 words)

The preliminary results showed that the proposed simplifications can compromise the heat balance, and, consequently, the percentage of hours where the operating temperature of the room is within a range of interest (PHFT). It was observed that it is a consequence related to the consideration of adiabatic surfaces. Therefore, in order to continue the process of simplifications on the original base model and to achieve a parameterizable model, it was necessary to adopt new measures that consider the heat exchanges on the room surfaces of interest. The thermal performance of each simplification will be compared with the base model based on the difference in PHFT indicated by root-mean-square error (RMSE) and mean absolute error (MAE) of the descriptive statistics of the reductions in comparison with the base model and illustrative graphics. The simplification process will be interrupted when a parameterizable model is found whose thermal performance is similar to the original base model. In this study, the Keras library - PHYTON language, was adopted as a function for artificial neural networks. The global sensitivity analysis of Sobol will be applied in order to identify the influence of the inputs data on the output data.

Main References

(max 200 words)

AÏT-SAHALIA, Yacine; XIU, Dacheng. Principal component analysis of high-frequency data. Journal of the American Statistical Association, v. 114, n. 525, p. 287-303, 2019.

AMASYALI, Kadir; EL-GOHARY, Nora M. A review of data-driven building energy consumption prediction studies. Renewable and Sustainable Energy Reviews, v. 81, p. 1192-1205, 2018.

DEB, Chirag et al. A review on time series forecasting techniques for building energy consumption. Renewable and Sustainable Energy Reviews, v. 74, p. 902-924, 2017.

HAASE, M.; AMATO, A. An investigation of the potential for natural ventilation and building orientation to achieve thermal comfort in warm and humid climates. Solar energy, v. 83, n. 3, p. 389-399, 2009.

MOHD NAWI, Nazri; ATOMIA, Walid Hasen; REHMAN, Mohammad Zubair. The effect of data pre-processing on optimized training of artificial neural networks. 2013.

VEHTARI, Aki; GELMAN, Andrew; GABRY, Jonah. Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC. Statistics and Computing, v. 27, n. 5, p. 1413-1432, 2017.

WANG, Zeyu; SRINIVASAN, Ravi S. A review of artificial intelligence based building energy use prediction: Contrasting the capabilities of single and ensemble prediction models. Renewable and Sustainable Energy Reviews, v. 75, p. 796-808, 2017.



14:12 - 14:19

A knowledge-based framework for building model performance verification

Yan Chen, Jeremy Lerond, Xuechen Lei, Michael Rosenberg, Draguna Vrabie

Pacific Northwest National Laboratory, United States of America

Aim and Approach

(max 200 words)

Building simulation is widely used by researchers, regulators, and engineers to quantify building energy performance. One application is building energy codes benchmarking, that is, creating code compliant building energy models and simulating their performance. To obtain confident benchmarking results, it is critical to verify that code requirements are properly implemented and simulated. Currently, this is done in a manual and ad-hoc manner. Such a verification process is tedious, error-prone and time-consuming for several reasons: 1) the number of models to be checked is very large, 2) modelers check models differently, which might cause difficulties for quality control, and 3) manual checks can be highly repetitive if not automated. To solve these challenges, we propose a data-driven performance verification framework, which conducts automated output-based verification of building energy code requirements (especially control requirements which have been verified via time-series output). This framework has an extensible knowledge base that contains programmatic interpretation and implementation of code requirements. While this framework was developed for building energy code benchmarking it can be extended and used for other situations such as code compliance performance modeling evaluation, and other control performance verification.

Scientific Innovation and Relevance

(max 200 words)

In the past, whole building energy modeling quality assurance (QA) and quality control (QC) have mainly been done using (or a combination of) input, output summary, or comparative output-based verifications. Most of these approaches tend to minimize the model review time while still being able to identify potential modeling errors within building energy models. The main downside to these approaches is that 1) they lack standardization, and 2) they do not typically evaluate the performance of the model at a granular level (e.g. simulation timestep) and hence might not be able to identify potential system control issues which in turns could have an impact on the simulation results. The proposed framework provides a transparent, reproduceable and automated way of checking the implementation of building system control strategies in a model based on simulation timestep data. The knowledge base developed as part of the framework uses a wide range of methods to verify simulation outputs: from simple rule-based verifications to more sophisticated approaches that leverage applied machine learning techniques.

Preliminary Results and Conclusions

(max 200 words)

We present two example uses of this framework where the implementation of ASHRAE Standard 90.1 requirements are checked: daylighting dimming control and integrated economizer control. Detailed discussion includes how the framework automatically instruments an EnergyPlus model, collects necessary data from various input and output files, and uses different types of algorithms to check the implementation of the code requirements. The examples also illustrate how the framework can provide verification results with quantitative and visual outputs to help users better understand how the requirements are implemented in the model.

Main References

(max 200 words)

BEMLibrary. n.d. Review Checklists and References. https://www.bemlibrary.com/index.php/practicioners/resources/review-checklists-references/.

Butzbaugh J.B., C.A. Antonopoulo. n.d. Introducing the Automated Fault Detection & Diagnostics Technology Challenge. PNNL-SA-153582.

Butzbaugh J.B., C.A. Antonopoulos, A. Tidwell. n.d. Automatic Fault Detection & Diagnostics Market Analysis. PNNL-30077.

Chris Balbach, David Bosworth. 2016. "Automated methods for improving energy model qualkity assurance and quality control." Building Performance Modeling Conference. Salt

Lake City, UT: ASHRAE and IBPSA-USA. 385-392.

COMNET. 2012. COMNET, Standardizing energy modeling for buildings. https://comnet.org/.

Kara Vega, Shelley Beaulieu, Maria Karpman. 2014. "Lessons Learned: Performing QC on Energy Models." ACEEE Summer Study on Energy Efficiency in Buildings. 339-350.

NREL. 2016. BuildingSync XML Schema. https://buildingsync.net/index.html .



14:19 - 14:26

Solar radiation nowcasting through advanced CNN model integrated with ResNet structure

Lei Chen1, Yangluxi Li1, Hu Du1, Yukun Lai2

1Welsh school of Architecture, Cardiff University, United Kingdom; 2School of Computer Science and Informatics, Cardiff University, United Kingdom

Aim and Approach

(max 200 words)

Although a range of solar radiation forecasting methods have been developed for predicting photovoltaic generation, only a few of them pay attention to the solar radiation forecasting for building energy demand. From the perspective of building performance modelling, solar radiation forecasting needs to meet several critical requirements including high spatial resolution (1m-2km) and high temporal resolution (5-60mins), the accurate value of Direct Normal Irradiance (DNI) and Diffuse Horizontal Irradiance (DHI), which differs the requirement for predicting photovoltaic generation. As the geometric sum of DNI and DHI, accurate prediction of Global Horizontal Irradiance (GHI) with high spatial-temporal resolution tends to be the prerequisite for accurate prediction of DNI and DHI.

This research aims to construct a hybrid nowcasting model to predict GHI in high spatial-temporal resolution.

In this article, the authors adopt an advanced Convolutional Neural Network (CNN) model with Residual Neural Network (ResNet) structure to identify the cloud image information and predict the GHI at 10 minutes intervals merely using cloud images captured by a ground-based sky camera. On this basis, several ResNet structures are compared to achieve the optimal nowcasting model for GHI.

Scientific Innovation and Relevance

(max 200 words)

Solar irradiance has a key role in predicting renewable thermal and electrical generation, informing demand response and ensuring high-quality real-time building simulations. In the past several decades, a series of solar irradiance forecasting methods have been developed and they could be classified as Numerical Weather Prediction Methods (NWP), Statistical Methods, Top-Down Methods, Bottom-Up Methods and Hybrid Methods [1]. Each method has its unique feature in term of spatial and temporal resolutions [2], for example, the spatial horizon of Bottom-Up Methods ranging from 1 meter to 2 kilometres and its temporal horizon ranging from 5 to 30 minutes. In recent years, researchers have explored various approaches [3], models [4] and data acquisition tools [5] for achieving high spatial-temporal resolution solar irradiance forecasting.

The key innovations of this study include:

1.Classifying solar radiation forecasting methods and reviewing recent representative articles.

2.Proposing a hybrid method using CNN-ResNet model to nowcast solar irradiance.

3.Only using ground-based cloud images as input to predict solar irradiation.

4.Comparing several ResNet structures to achieve the optimal nowcasting model for GHI.

Preliminary Results and Conclusions

(max 200 words)

This study developed a hybrid nowcasting method for Global Horizontal Irradiance (GHI) through using an advanced CNN model with ResNet structure to identify the ground cloud images information collected by an open database – NREL dataset. Several ResNet structures are compared to achieve optimal nowcasting model for GHI. The results show that the ResNet structure could efficiently capture the significant information of cloud images and the ResNet-152 has the optimal performance on solar irradiance nowcasting. Final result also indicates that a low prediction error and high relative coefficient values are got. The accomplishment of this model apparently decreases the difficulty of solar irradiance forecast because only need to use sky image. Nevertheless, taking consideration of this model generalization capability, future relative investigations had better to be performed to improve precision under other sky condition.

Main References

(max 200 words)

[1] Chen, L., Du, H. and Li, Y., 2019, September. Scoping Low-Cost Measures to Nowcast Sub-Hourly Solar Radiations for Buildings. In IOP Conference Series: Earth and Environmental Science (Vol. 329, No. 1, p. 012041).

[2] Diagne, M., David, M., Lauret, P., Boland, J. and Schmutz, N., 2013. Review of solar irradiance forecasting methods and a proposition for small-scale insular grids. Renewable and Sustainable Energy Reviews, 27, pp.65-76.

[3] Alonso, J., Batlles, F.J. and Portillo, C., 2015. Solar irradiance forecasting at one-minute intervals for different sky conditions using sky camera images. Energy Conversion and Management, 105, pp.1166-1177.

[4] Chu, Y., Li, M., Pedro, H.T. and Coimbra, C.F., 2015. Real-time prediction intervals for intra-hour DNI forecasts. Renewable Energy, 83, pp.234-244.

[5] Du, H., Jones, P., Segarra, E. L. & Bandera, C. F., 2018. Development Of A Rest Api For Obtaining Site-Specific Historical And Near-Future Weather Data In Epw Forma. Building Simulation and Optimization 2018, 2018 Emmanuel College, University Of Cambridge.

[6] Perez, R., Ineichen, P., Maxwell, E., Seals, R., & Zelenka, A., 1991. Dynamic models for hourly global-to-direct irradiance conversion. In Solar world congress: Proceedings of the biennial congress of the international solar energy society (Vol. 1, No. part II, pp. 951-956).



14:26 - 14:33

Intelligent building envelopes energy evaluation: an integration of double-skin facades with earth-tubes

Payman Sadeghi

Drury University, United States of America

Aim and Approach

(max 200 words)

Ecological Architecture inclusively addresses all the matters that should and could be considered including but not limited to economic, social, or environmental dimensions. The integration of an ecological worldview, thus, should be recognized as a fundamental goal in reaching high-performance solutions that are formed as environmental systems rather than as standalone objects. Building envelope, if conceived as a responsive skin, can serve as a critical component of an ecological building. It is strictly linked to the environmental, social, and economical aspects of buildings and their context. Therefore, it should act in a creative, flexible, dynamic, and adaptive manner. A Double-Skin Facade (DSF) is product of this viewpoint: a multilayered formation to provide daylight, indoor air quality, natural ventilation, thermal and visual comfort, acoustics isolation, and aesthetic. In this domain, energy is surely a crucial concern too. This paper's focus is to examine ways to optimize the energy demand_ specifically Load Intensity (LI)_ of buildings with a DSF as an integrated, intelligent, and high performance building envelope system.

Scientific Innovation and Relevance

(max 200 words)

From environmental and social standpoints, Double-Skin Facades (DSFs) are presumed as high-performance systems that integrate ecological solutions. Yet, DSFs’ economic performance is debated mainly due to the overheating issue in hot seasons. That is, a DSF’s goal to minimize buildings’ energy use in hot climatic conditions is controversial, presenting it occasionally as not the most economically viable solution. In this study, it was hypothesized that the energy demand of a building with a naturally ventilated DSF, inspired by Persian wind-catchers, would be improved via linking it into the earth via earth-tubes, inspired by the concept of Roman hypocausts. The study speculates that this combination could resolve the DSF’s overheating issue leading to a lower building load intensity, which would then increase the economic performance of a DSF system. To conduct the inquiry, a simulation method was applied that employed both TRNSYS and CANTAM softwares to model a nonresidential building with integrated DSF and earth-tube systems in three main climatic conditions of the United States. Combination of alternative design options, then, were assessed to compare their load intensity and identification of the optimized proposal.

Preliminary Results and Conclusions

(max 200 words)

Consequently, the energy impact of alternative design options such as climate, DSF depth, shading, ventilation strategy, and earth-tubes integration was examined. The results showed that the assimilation of earth-tubes into a DSF system always increases the system’s energy performance with no overheating issue provided that optimum DSF depth is envisioned. However, the integrated system demonstrated higher positive influence on buildings’ energy performance in cold climates compared to moderate and hot climates where the impact is lower. Also, the energy demand in the integrated system is primarily linked to the building’s ventilation strategy not the DSF’s. In contrast to certain studies on DSFs suggesting that the integration of a DSF system will always increase cooling loads, a DSF system can considerably lower cooling loads on conditions that appropriate shading devices, construction types, DSF depths, and building ventilation strategies are implemented. After all, the study outcomes suggest additional emphasis on the role and the need for systematic communication with buildings science. To conclude, the research process has elaborately showcased how an early theorized design solution can still perform different from what had been known or read in the literature, highlighting the importance of evaluation.

Main References

(max 200 words)

Bartok, J. (2012). Geothermal Heat for Greenhouses, Retrieved from: http://www.extension.org/pages/27790/geothermal-heat-for-greenhouses#.UwO-SvldVqU, Date Accessed: February 22, 2014.

Choi, W., Joe1, J., Kwak., & Huh, J. (2012). Operation and Control Strategies for Multi-Story Double Skin Facades during the Heating Season. In Energy and Buildings. 49. 454–465.

Gratia, E., & De Herde, A. (2004). Natural cooling strategies efficiency in an office building with a double-skin façade. Energy and buildings, 36(11), 1139-1152.

Kalyanova, O., Heiselberg, P., Felsmann, C., Poirazis, H., Strachan, P., & Wijsman, A. (2009, July). An empirical validation of building simulation software for modelling of double-skin facade (DSF). In Proceedings of the 11th IBPSA Conference, July (pp. 27-30).

Poizaris, H. (2004). Double Skin Facades for Office Buildings.

Wong, P. C., Prasad, D., & Behnia, M. (2008). A New Type of Double-Skin Façade Configuration for the Hot and Humid Climate. Energy and Buildings, 40(10), 1941-1945.

Xu, L., & Ojima, T. (2007). Field Experiments on Natural Energy Utilization in a Residential House with a Double Skin Façade System. Building and Environment, 42(5), 2014-2023.

Zhou, J., & Chen, Y. (2010). A Review on Applying Ventilated Double-Skin Facade to Buildings in Hot-Summer and Cold-Winter Zone in China. Renewable and Sustainable Energy Reviews, 14(4), 1321-1328.



14:33 - 14:40

Whole building validation for simulation programs including synthetic users and heating systems: experiment and results

Matthias Kersken1, Paul Strachan2

1Fraunhofer Institute for Building Physics IBP, Germany; 2Energy Systems Research Unit, Department of Mechanical and Aerospace Engineering, University of Strathclyde, Glasgow, UK

Aim and Approach

(max 200 words)

Dynamic Building Energy Simulation (BES) tools are increasingly used in the design of buildings, building services and controls. Especially when considering controls of complex buildings, the tools in use need to reliably respond the occupants’ various inputs, represent the buildings’ service systems and most importantly, capture the interactions between both correctly. Within IEA EBC Annex 58 a side-by-side full scale validation experiment on Fraunhofer IBP’s Twin Houses was undertaken and made public, investigating BES tools predictive capabilities, focusing particularly on climate and the building envelope [1]. Within the current IEA EBC Annex 71 a new experiment was conducted focusing on occupants’ influence and heating systems with controls. To guarantee a known but realistic occupancy, synthetic users with probabilistic elements were included [2]. During the experimental design, a sensitivity analysis was conducted to identify the most critical influences on the validation [3]. After these items were identified, either the experiment was changed or suitable instrumentation was installed. Using the detailed experimental specification and provided boundary conditions, 13 modelling teams using 8 different programs submitted “blind validation” results for the main part of the experiment. Updated submissions to correct modelling errors were made after the full experimental dataset was released.

Scientific Innovation and Relevance

(max 200 words)

Because it is relatively cheap and fast the method of inter model comparison is widespread for the validation of building energy performance simulation software. Well known examples for inter model comparison are BESTEST and CEN 13791. Also, several in situ measurement datasets are available, but most of them are created using very simple test boxes [6]. In addition, often the datasets underlying published empirical validation studies are not available, so the validation can not be repeated with updated versions of the tools validated previously or other tools. This contribution presents a whole building empirical validation, including a publicly available dataset, based on detailed measurements and extensive documentation, including occupancy and different heating systems in a side-by-side configuration, allowing for direct comparison of different heating systems.

Preliminary Results and Conclusions

(max 200 words)

The experiences from Annex 58 clearly show that the publicly available dataset itself is a very valuable outcome of such an experiment. Together the Annex 58 datasets [7] and [8] were downloaded about 300 times since their publication. The current Annex 71 dataset (including documentation) is also published [9] already. Both datasets are suitable for validation, training and educational purposes. This paper briefly summarises the detailed experimental specification and datasets recorded over a 4-month winter period for both electric and underfloor heating systems in two two-storey houses with realistic time–varying occupancy-related heat gains. The experiment consisted of several phases of increasing complexity. The paper will then show selected comparisons of the modelling predictions with the measured data, emphasising areas of good agreement, but also areas where agreement was poorer, indicating deficiencies in the modelling.

Main References

(max 200 words)

[1] P. Strachan, K. Svehla, I. Heusler und M. Kersken, „Whole model empirical validation on a full-scale building,“ Journal of Building Performance Simulation, Bd. 9, Nr. 4, pp. 331-350, 09 2015.

[2] G. Flett und N. Kelly, „An occupant-differentiated, higher-order Markov Chain method for prediction of domestic occupancy,“ Energy and Buildings, Nr. 230, pp. 219-230, 2016.

[3] E. Mantesi, K. Mourkos, C. Hopfe, R. McLeod, P. Vatougiou, M. Kersken und P. Strachan, „Deploying Building Simulation to Enhance the Experimental Design of a Full-scale Empirical Validation Project,“ in Proceedings BuildingSimulation2019, Rome, Italy, 2019.

[4] P. Strachan, „Model validation using the PASSYS Test cells,“ Building and Environment, Bd. 28, Nr. 2, 1993.

[5] „Annex 58 - BES-Model Validation DATA: doi: 10.15129/8a86bbbb-7be8-4a87-be76-0372985ea228,“ [Online]. Available: https://pure.strath.ac.uk/portal/en/datasets/twin-houses-empirical-dataset-experiment-1%288a86bbbb-7be8-4a87-be76-0372985ea228%29.html.

[6] „Annex 58 - BES-Model Validation DATAset 2: doi: 10.15129/94559779-e781-4318-8842-80a2b1201668,“ [Online]. Available: https://pure.strath.ac.uk/portal/en/datasets/twin-houses-empirical-validation-dataset-experiment-2(94559779-e781-4318-8842-80a2b1201668).html.

[7] M. Kersken und P. Strachan, Twin House Experiment IEA EBC Annex 71 Validation of Building Energy Simulation Tools - Specifications and dataset - (DOI: 10.24406/fordatis/76) http://fordatis.fraunhofer.de/handle/fordatis/161, 2020.



14:40 - 14:47

A methodology for generating a synthetic local urban climate weather profile for building energy simulations in hot arid areas

Mohamed elnabawi Mahgoub1, Neveen Hamza2

1Applied Science University, Bahrain; 2Newcastle University, UK

Aim and Approach

(max 200 words)

The paper examines the thermal performance of different shading designs for an urban alley in the hot arid context. Computational fluid dynamics (CFD) programs can provide detailed distributions and calculation of the air velocity and temperature distribution if all of the heat factors are set as dynamic boundary conditions and heat generators. However, this method is computationally very expensive and time demanding to provide predictions for a season or year. Alternatively, an integration of building energy modeling (BEM) and Computational fluid dynamics (CFD) can eliminate many of these assumptions, as the two models exchange the appropriate boundary conditions. Therefore, Designbuilder as BEM handles the external surface temperature for the main building surrounding the semi- enclosed areas, while Fluent as CFD simulates the street air flow and air temperature. Therefore, the paper main aim is to evaluate the operation of an alternative approach through the coupling between the Computational fluid dynamics (CFD) with the building energy modeling (BEM) in order to propose guidelines to mitigate the UHI effect and improve the urban microclimate. The numerical results were validated and compared to the experimental ones.

Scientific Innovation and Relevance

(max 200 words)

• The coupling of building energy simulation and CFD simulations has been used to improve energy modeling performance (Zhai and Chen 2005; Tian and Zuo 2013).). As most of the available studies investigate the effect of outdoor environment on the building performance and indoor environment. However, the current paper examines the building performance on the outdoor microclimate and urban thermal comfort by calculating an outdoor comfort index

• Most studies assumed the reliability and accuracy of the building energy modelling (BEM) and Computational fluid dynamics (CFD) programs and did not conduct or offer a well validation result to support their findings (Zhai et al. 2002 and Tian et al, 2018). Few of the studies provided the validation of either program against experimental data (Du et al. 2015; Tian, Sevilla, Zuo, et al. 2017; Zuo, Wetter, et al. 2016) and small-scale experiments (Novoselac 2004; Wang 2007). Nevertheless, it needs future research efforts to develop small-scale experiments with high-quality measurements to validate the co-simulation (Tian et al, 2018). The current study provides a detailed validation based on the mean air temperature distribution, surface temperature and mean radiant temperature

Preliminary Results and Conclusions

(max 200 words)

• The study has validated a coupled simulation assessment framework against data obtained from the field measurement. The validations verify that the process of can provide reasonable and reliable predictions for mean air temperature distribution for the urban alley and surface temperature for the surrounding buildings. the simulation outcomes showed a consistent trend with the observed ones and that the temperature differences are within 0.8°C to 1.2°C.

• Measurements and modelling of the outdoor thermal conditions has to be seen as a tool for urban planning and mitigation of conditions to increase utilization of outdoor urban spaces, especially in a context of globally increasing air temperatures and UHI in summer.

• The coupled simulation produces more accurate and detailed results than the separate simulations because; first, CFD receives more precise and real-time thermal boundary conditions and can predict the dynamic outdoor environment conditions that are important for the assessment of outdoor air quality and thermal comfort, while BES obtains more accurate convection heat from enclosures and can provide more accurate estimation of dynamic thermal behaviours of building.

Main References

(max 200 words)

• Abdulrahman, A and Sharples, S (2014) 'Analysing and optimizing the performance of hybrid housing design in the Middle East'. In: 2nd IBPSA-England Conference, UCL.

• Allegrini, J., Carmeliet, J. (2018) Simulations of local heat islands in Zürich with coupled CFD and building energy models. Urban climate, 2018; 24: 340-359.

• Barbason, M., and Reiter, S., (2014) 'Coupling building energy simulation and computational fluid dynamics: Application to a two-storey house in a temperate climate',

• Blocken, B., Stathopoulos, T., Carmeliet, J. and Hensen, J.L.M. (2011) 'Application of computational fluid dynamics in building performance simulation for the outdoor environment: an overview', Journal of Building Performance Simulation, 4(2), pp. 157-184.

• Chao, Y., and Ng, E., (2012) 'Building porosity for better urban ventilation in high-density cities A computational parametric study', Building and Environment 50;176e189

• W. Tian. X. Han, W. Zuo, M. Sohn 2018. "Building Energy Simulation Coupled with CFD for Indoor Environment: A Critical Review and Recent Applications." Energy and Buildings, 165, pp.184-199.

• Yi, Y.K., Feng, N. Dynamic integration between building energy simulation (BES) and computational fluid dynamics (CFD) simulation for building exterior surface. Build. Simul. 6, 297–308 (2013).



 
Contact and Legal Notice · Contact Address:
Privacy Statement · Conference: Building Simulation 2021
Conference Software - ConfTool Pro 2.6.143+TC
© 2001–2022 by Dr. H. Weinreich, Hamburg, Germany