Workflow design for the analysis of transitional space microclimates application to the analysis of strategies for the improvement of the energy performance and comfort of courtyard buildings in the mediterranean climate

PhD Thesis WORKFLOW DESIGN FOR THE ANALYSIS OF TRANSITIONAL SPACE MICROCLIMATES Application to the analysis of strategies for the improvement of the energy performance and comfort of courtyard buildings in the Mediterranean climate DISEÑO DE UNA METODOLOGÍA DE ANÁLISIS DEL MICROCLIMA DE LOS ESPACIOS DE TRANSICIÓN Aplicación en el análisis de estrategias de mejora energética y confort en arquitecturas con patio del clima mediterráneo Victoria Patricia López-Cabeza TESIS DOCTORAL 2023 Universidad de Sevilla PhD Thesis WORKFLOW DESIGN FOR THE ANALYSIS OF TRANSITIONAL SPACE MICROCLIMATES Application to the analysis of strategies for the improvement of the energy performance and comfort of courtyard buildings in the Mediterranean climate Victoria Patricia López-Cabeza PHD THESIS 2023 Universidad de Sevilla Tesis doctoral DISEÑO DE UNA METODOLOGÍA DE ANÁLISIS DEL MICROCLIMA DE LOS ESPACIOS DE TRANSICIÓN Aplicación en el análisis de estrategias de mejora energética y confort en arquitecturas con patio del clima mediterráneo Victoria Patricia López-Cabeza TESIS DOCTORAL 2023 Universidad de Sevilla 1 PhD Thesis Architecture Workflow Design for the Analysis of Transitional Space Microclimates Application to the analysis of strategies for the improvement of the energy performance of courtyard buildings in the Mediterranean climate Author Victoria Patricia López-Cabeza Supervisors: Prof. Dr. Carmen Galán Marín Prof. Dr. Carlos Rivera Gómez Submitted in fulfilment of the requirements for the Degree of Doctor of Philosophy at the University of Seville Doctorate in Architecture Institute of Architecture and Building Science (IUACC) School of Architecture Universidad de Sevilla Sevilla, 2023 Workflow Design for the Analysis of Transitional Space Microclimates 2 3 Acknowledgments This document is the summary of the work realized through five years of continuous learning and personal development withing the research group TEP 206 of the Department of Building Construction I at the University of Seville. I greatly thank my advisors, Carmen Galán-Marín and Carlos Rivera-Gómez, for their continuous support and guidance in this journey, their ideas, and especially their time in helping me to achieve my goals. I am very fortunate to be part of the team they have built, and to all the help I have had from mathematicians (Prof. Chacon, Prof. Gómez and Prof. Fernández among others) and engineers (Prof. Sánchez and Prof. Ruiz) working with us. I need to thank the Solar Decathlon Aura team for opening the university back to me to collaborate with them to raise the Solar Decathlon Europe initiative at the University of Seville. And many other people in the department who have helped me and trust me in much collaborative work, teaching, and administration. I am grateful to have had the opportunity to have spent my time in the department and outside it with great colleagues with who I have shared work, breaks, meals, nightmares, and successes. Eduardo, David, Javier, thanks for being there. And Alix, Fernanda, Sofía, Jose, I hope you are fine wherever you are now. I am in debt to the people who made my stay at Harvard fruitful and unforgettable. My deepest gratitude to Holly Samuelson for her support and lessons, and to my colleague and friend, Aditi Agarwal, who now I consider my family. And many others, Mirah, Bohan, SungHwan, Elence, Katerina, you made my stay enjoyable. This trip would not have been possible without the support of my family and friends, who were always there if I needed help and counseling and also understood when I was too distant because of work. I am sorry and I feel lucky for their consideration. Workflow Design for the Analysis of Transitional Space Microclimates 10 engineers, and software developers to achieve our goals. This has given me multidisciplinary abilities to apply in my research activity and grow as a professional in the field. To sum up, this predoctoral period is about to conclude with the participation in 3 private and public research projects in regional, national, and international spheres, 15 articles published in indexed scientific journals and 11 communications at international conferences, obtaining an h6 index and a total of 141 citations, according to Scopus. Notes: 1. Work experience: Arquia Intern Scholarship in Emilio Tuñón Arquitectos / Collaborator in Juan Pablo Lopez Recio Arquitecto / Research Collaboration Grant with Narciso Vazquez / Architect in Lucena City Council. 2. Research contract: Research contract from the Young Employment Plan of the University of Seville (PEJUS) in the framework of the National Youth Guarantee System and the Youth Employment Operational Program 2014-2020. 3. Fulbright grant to develop higher studies in USA. Research stay in Harvard University, Graduate School of Design. (64200€) /Real Colegio Complutense Scholarship to fund the learning stay in Graduate School of Design (10000€). 4. Research projects during PhD: ROM Optimization for Architecture and Design Benchmarking Thermal Behavior in Transitional Spaces as Energy Efficiency Resource ROAD BETTER (RTI2018-093521-B-C33), Eco efficient Design of Courtyards in Buildings through Reduced Order Modeling MORE PATIO (MTM2015-64577-C2-2-R) and Direct applications of “Aura Strategy” from Solar Decathlon US Team on the restoration of Andalusian obsolete neighborhoods (US.20-11) Publications and Awards 11 Publications and Awards This PhD thesis is presented in the form of a compendium of publications, according to Article 9 of the regulations governing the doctoral thesis regime of the University of Seville (Agreement 9.1 / CG 19-4-12), within the framework of the national regulation RD 99/2011. It involves a list of several original publications developed by the PhD candidate and her co-authors during the thesis period. The PhD candidate has the explicit authorization of all the co-authors to use their contents as a part of this thesis. A complete copy of the publications is presented in the Appendix. In addition, several contributions have been made to international conferences as a part of the PhD work, listed below. Papers in indexed journals Publication 1 V.P. López-Cabeza, C. Galán-Marín, C. Rivera-Gómez, J. Roa-Fernández, Courtyard microclimate ENVI-met outputs deviation from the experimental data, Building and Environment. 144 (2018) 129–141. https://doi.org/10.1016/j.buildenv.2018.08.013. JCR: Construction and Building Technology (3/132) Q1 SJR: Building and Construction (8/173) Q1 Publication 2 V.P. Lopez-Cabeza, F.J. Carmona-Molero, S. Rubino, C. Rivera-Gómez, E. Fernández-Nieto, C. Galán-Marín, T. Chacón Rebollo, Modelling of surface and inner wall temperatures in the analysis of courtyard thermal performances in Mediterranean climates, Journal of Building Performance Simulation. 14 (2021) 181–202. https://doi.org/10.1080/19401493.2020.1870561. JCR: Construction and Building Technology (32/68) Q2 SJR: Architecture (8/147) Q1 Workflow Design for the Analysis of Transitional Space Microclimates 12 Publication 3 V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gómez, C. Galán-Marín, H.W. Samuelson, Thermal comfort modelling and empirical validation of predicted air temperature in hot-summer Mediterranean courtyards, Journal of Building Performance Simulation. 15 (2022) 39–61. https://doi.org/10.1080/19401493.2021.2001571. JCR: Construction and Building Technology (32/68) Q2 SJR: Architecture (8/147) Q1 Publication 4 V.P. Lopez-Cabeza, S. Alzate-Gaviria, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, Albedo influence on the microclimate and thermal comfort of courtyards under Mediterranean hot summer climate conditions, Sustainable Cities and Society. 81 (2022) 103872. https://doi.org/10.1016/J.SCS.2022.103872. JCR: Construction and Building Technology (2/68) Q1 SJR: Civil and Structural Engineering (12/324) Q1 Publication 5 V.P. Lopez-Cabeza, C. Rivera-Gomez, J. Roa-Fernández, M. Hernandez- Valencia, R. Herrera-Limones, Effect of thermal inertia and natural ventilation on user comfort in courtyards under warm summer conditions, Building and Environment, 109812 (2022). https://doi.org/10.1016/j.buildenv.2022.109812. JCR: Civil Engineering (10/138) Q1 SJR: Building and Construction (11/191) Q1 Publication 6 V.P. Lopez-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, Shade and thermal comfort in courtyards: experimental versus simulation results, Buildings. 12 (2022) 1961. https://doi.org/10.3390/BUILDINGS12111961. JCR: Construction and Building Technology (28/68) Q2 SJR: Architecture (14/147) Q1 Publication 7 V.P. Lopez-Cabeza, A. Agarwal, Sustainable strategies for improving passive survivability of the traditional Pol House in Ahmedabad, India, Building and Environment. 207 (2022) 108503. https://doi.org/10.1016/j.buildenv.2021.108503. JCR: Civil Engineering (10/138) Q1 SJR: Building and Construction (11/191) Q1 Publication 8 V.P. López-Cabeza, J. Lizana, E. Diz-Mellado, C. Rivera-Gómez, C. Galán- Marín, Outdoor Microclimate Influence on Building Performance: Simulation Tools, Challenges, and Opportunities. Book Chapter in: New Technologies in Building and Construction, Springer, 2022: pp. 103-. https://doi.org/https://doi.org/10.1007/978-981-19-1894-0_7. SPI: General (2018) (4/96) Q1 Publications and Awards 13 Participation in International Conferences • V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, A comparative numerical simulation study for outdoor comfort indexes in courtyards. I3M-SESDE 2022: 10th International Workshop on Simulation for Energy, Sustainable Development and Environment. Rome, Italy. • S. Alzate-Gaviria, V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galán-Marín, Albedo impact on the microclimate of courtyards, potential implications on the UHI. A case study in Seville, Spain. BS2021 IBPSA: Building Simulation Conference 2021. Bruges, Belgium. • J. Lizana, V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galán-Marín, Impact of multiple urban microclimates in building performance. A simulation approach to support climate-resilient building design. BS2021 IBPSA: Building Simulation Conference 2021. Bruges, Belgium. • E. Diz-Mellado, A. Redondo, V.P. López-Cabeza, C. Rivera-Gomez, C. Galan-Marin, Thermal comfort assessment in courtyards under heat waves in a Mediterranean climate. CATE21: Comfort At The Extremes 2021. Sultanato de Oman. • V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galán-Marín, UTCI improvement quantification inside a courtyard using a shading device. 15th SDEWES Conference on Sustainable Development of Energy, Water and Environment Systems. COLOGNE, Germany. 2020. • E. Diz-Mellado, V.P. López-Cabeza, C. Rivera-Gomez, J. Roa-Fernandez, C. Galan-Marin, Comparison of albedo and vegetation effect on the courtyards microclimatic variations. 15th SDEWES Conference on Sustainable Development of Energy, Water and Environment Systems. COLOGNE, Germany. 2020. • V.P. López-Cabeza, C. Galán-Marín, C. Rivera-Gomez, H. Samuelson, Comparison and validation of two simulation workflows for courtyard microclimates. 35th PLEA Conference Sustainable Architecture and Urban Design: Planning post-carbon cities. A CORUÑA, Spain. 2020. • E. Diz-Mellado, C. Galan-Marin, C. Rivera-Gomez, V.P. López-Cabeza, Facing climate change overheating in cities through multiple thermoregulatory courtyard potential case studies appraisal. REHABEND: Construction Pathology, Rehabilitation Technology and Heritage Management 2020. Granada, Spain. • C. Rivera-Gomez, C. Galan-Marin, E. Diz-Mellado, V.P. López-Cabeza, Thermodynamic behavior of transitional spaces in architecture. Reviewing the main thermal tempering parameters. International Conference on GREEN CONSTRUCTION. Universidad de Córdoba. 2019. Workflow Design for the Analysis of Transitional Space Microclimates 14 • C. Galan-Marin, V.P. López-Cabeza, C. Rivera-Gomez, On the influence of shade in improving thermal comfort in courtyards. ESDEIC: Economy, Sustainable Development and Energy International Conference. Edinburgh, Great Britain, 2018. https://doi.org/10.3390/proceedings2221390 • V.P. López-Cabeza, C. Galan-Marin, C. Rivera-Gomez, J. Roa-Fernandez, Comparative analysis of thermal performance between courtyards in Mediterranean climate. ESDEIC: Economy, Sustainable Development and Energy International Conference. Edinburgh, Great Britain, 2018. https://doi.org/10.3390/proceedings2221381 Papers and works not directly related to the thesis The PhD candidate has also collaborated on other papers not directly related to the subject of the thesis but developed within the research group framework. For this reason, these publications are not included in the appendix. • C. Rivera-Gómez, E. Diz-Mellado, C. Galán-Marín, V. López-Cabeza, Tempering potentialbased evaluation of the courtyard microclimate as a combined function of aspect ratio and outdoor temperature, Sustainable Cities and Society. 51 (2019) 101740. https://doi.org/10.1016/j.scs.2019.101740. • E. Diz-Mellado, V.P. López-Cabeza, C. Rivera-Gómez, J. Roa-Fernández, C. Galán-Marín, Improving School Transition Spaces Microclimate to Make Them Liveable in Warm Climates, Applied Sciences. 10 (2020) 7648. https://doi.org/10.3390/app10217648. • C. Rivera-Gómez, C. Galán-Marín, V.P. López-Cabeza, E. Diz-Mellado, Sample key features affecting mechanical, acoustic and thermal properties of a natural-stabilised earthen material, Construction and Building Materials. (2020). https://doi.org/10.1016/j.conbuildmat.2020.121569. • E. Diz-Mellado, V.P. López-Cabeza, C. Rivera-Gómez, C. Galán-Marín, J. Rojas-Fernández, M. Nikolopoulou, Extending the adaptive thermal comfort models for courtyards, Building and Environment. 203 (2021) 108094. https://doi.org/10.1016/j.buildenv.2021.108094. • J. Lizana, V.P. L, R. Renaldi, E. Diz-mellado, C. Rivera-Gomez, C. Galan-Marin, Integrating courtyard microclimate in building performance simulation to mitigate extreme urban heat impacts, Sustainable Cities and Society. (2021) 103590. https://doi.org/10.1016/j.scs.2021.103590. • C. Galán-Marín, C. Rivera-Gómez, J. Lizana, J. Roa-Fernández, E. Diz-Mellado, V.P. Lopez- Cabeza, Enhancing Urban Microclimates Towards Climate-Resilient Cities: The Potential of Courtyards, Springer International Publishing, 2021. https://doi.org/10.1007/978-3-030-78566-6. Publications and Awards 15 • E. Diz-Mellado, V.P. López-Cabeza, J. Roa-Fernández, C. Rivera-Gómez, C. Galán-Marín, Energy-saving and thermal comfort potential of vernacular urban block porosity shading., Sustainable Cities and Society. (2022) 104325. https://doi.org/10.1016/j.scs.2022.104325. Awards & Honors The work in this research has been awarded by different national and international institution. The communications that have received recognition are: • Award for Best Scientific Publication in Architecture and Construction Sciences 2022, granted by the Institute of Architecture and Building Science (IUACC) of the University of Seville, with a prize of 200€ (Seville, 2019). V.P. Lopez-Cabeza, S. Alzate-Gaviria, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, Albedo influence on the microclimate and thermal comfort of courtyards under Mediterranean hot summer climate conditions, Sustainable Cities and Society. 81 (2022) 103872. • Best paper award presented to the 10th International Workshop on Simulation for Energy, Sustainable Development and Environment (SESDE 2022, Rome). V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, A comparative numerical simulation study for outdoor comfort indexes in courtyards. I3M-SESDE 2022: 10th International Workshop on Simulation for Energy, Sustainable Development and Environment. Rome, Italy. Workflow Design for the Analysis of Transitional Space Microclimates 16 17 List of abbreviations AR Aspect Ratio BES Building Energy Simulation CFD Computational Fluid Dynamics CS Case Study (CV)RMSE Coefficient of Variation of the Root Mean Square Error DB T Dry-Bulb Temperature (ºC) EU European Union GH Grasshopper software HULC Herramienta Unificada LiIDER-CALENER (Unified Tool LIDER-CALENER) IPCC Intergovernmental Panel on Climate Change LBT Ladybug Tools MAPE Mean Absolute Percentage Error (%) MRT Mean Radiant Temperature (ºC) NMBE Normalized Mean Bias Error OTC Outdoor Thermal Comfort PDE Partial Differential Equations PET Psychological Equivalent Temperature (ºC) R2 Coefficient of Determination Workflow Design for the Analysis of Transitional Space Microclimates 18 RH Relative Humidity (%) RMSE Root Mean Square Error RMSEs Systematic Root Mean Square Error RMSEu Unsystematic Root Mean Square Error SET Standard Equivalent Temperature (ºC) SVF Sky View Factor TD Thermal Delta (Outdoor Temperature-Courtyard Temperature) (ºC) TG Thermal Gap (Outdoor Temperature-Courtyard Temperature) (ºC) UHI Urban Heat Island UTCI Universal Thermal Climate Index (ºC) WBGT Wet Bulb Globe Temperature (ºC) Chapter 1: Research Framework 19 Chapter 1: Research Framework 1.1 Introduction 1.2 State of Art 1.3 Objectives and Scope 1.4 Methodology 1.5 Thesis Structure Summary This thesis develops a new workflow for the simulation of the thermodynamic effects in outdoor microclimates and applies it to the analysis of building with courtyards to improve user comfort and building energy performance in the Mediterranean climate. In this first chapter, the background and objectives of the thesis are presented, the scope of the study is delimited, the state of the art is analyzed, the methodology is described, and the whole structure of the document is explained. Related Publications This chapter incorporates analysis, data, and conclusions from the peer-reviewed publications: “ Outdoor Microclimate Influence on Building Performance: Simulation Tools, Challenges, and Opportunities ”, in New Technologies in Building and Construction (Publication 8); and “ Thermal comfort modelling and empirical validation of predicted air temperature in hotsummer Mediterranean courtyards ¨ in Journal of Building Performance Simulation (Publication 3). Workflow Design for the Analysis of Transitional Space Microclimates 26 buildings. Some researchers proposed modifications to this model because they suggest that it does not consider cultural and climatic diversity in the world, proposing new boundaries for the comfort zone [64]. Outdoor thermal comfort (OTC) is affected by many parameters such as air temperature relative humidity, wind speed, direct and diffuse radiations from the sun, infrared radiation, human activity, and clothing level (see Figure 1.4-3 ). Among the different indexes developed, the most commonly used are the Physiological Equivalent Temperature (PET), the Standard Equivalent Temperature (SET*), and the Universal Thermal Comfort Index (UTCI) [65]. PET [66] is defined as the air temperature, without wind speed and solar radiation (indoor conditions) at which the heat balance of the human body is maintained with the same core and skin temperature as under the conditions to be quantified in the outdoor thermal environment. However, some researchers found it not suitable for all hot and humid climate zones [67] and proposed some adaptations. SET is defined by ASHRAE 55 [62] as the temperature of the hypothetical environment at 50% of relative humidity, air speed lesser than 0.1 m/s and air temperature equivalent to the mean radiant temperature, in which the heat loss from the skin of an imaginary respondent with 1.0 met of activity and 0.6 clo of clothing level is equal to a respondent in the actual environment wearing actual clothing and doing the actual activity. This index is recommended for moderate to hot climates [68]. Finally, UTCI was developed by the International Society of Biometeorology looking for a model valid for a wide range of outdoor conditions, including extreme environments [69,70]. It is commonly understood as the “feels like” temperature [71]. Figure 1.2-1. Parameters affecting outdoor thermal comfort. Adapted from [72] . Different software can be used to simulate outdoor comfort (refer to Table 1.2-1 ). Some of the most used ones are Rayman, ENVI-met, or the Ladybug Tools. Each of them can compute different parameters needed for the thermal comfort calculation, so the inputs required are different, and their suitability depends on the study characteristics and available data. Rayman [73] is a microscale model than can calculate radiation fluxes in the urban environment. This means that it can compute local Mean Radiant Temperature (MRT) based on solar radiation Chapter 1: Research Framework 27 inputs and a geometry model. However, it cannot simulate the thermodynamic processes, so the air temperature and wind velocity inputs to calculate thermal comfort need to be supplied by the user. ENVI-met uses some of the outputs of a CFD simulation to calculate thermal comfort. Different indexes can be computed using its BioMet application, which is specialized in urban thermal comfort. Ladybug tools also compute many thermal indexes using outputs from the different plugins that they include. Both tools, ENVI-met and LBT, require a meteorological weather file and a geometry and materials characterization to compute the different thermal comfort indexes. Among the different thermal comfort indexes previously described, UTCI is preferred to be used in the analysis of outdoor comfort in the courtyards in the Mediterranean climate in this research. The reasons are two: it is considered the most appropriate to the analysis of extreme weather conditions [74] and it is universal, meaning that the comfort range can be applied to different location and climates, making the results comparable [75]. For indoor conditions, adaptive thermal comfort is preferred. However, some other indexes have been used, and this is justified in each section. Figure 1.2-2. Andalusian courtyard as an outdoor comfortable space to live in. Cordoba. (Own photo). Workflow Design for the Analysis of Transitional Space Microclimates 28 1.2.3 Opportunities and needs After analyzing the state of the art, the detected research gap and resulting goal proposed in this thesis go through the identification of some opportunities and needs that are going to be addressed: 1- Classification and analysis of existing outdoor simulation tools and approaches applicable to the prediction of the thermal performance of outdoor spaces and evaluate their suitability to be used in the design process and simulation of courtyard buildings. 2- Definition of the desired characteristics of a simulation procedure of the microclimate of courtyards to be incorporated in the design stage of a project, considering that needs at this stage are substantially different from needs at later stages of the process when the design is already fixed. 3- Development of a simulation workflow for the early design process of courtyard spaces and their improvement from a professional perspective, that fulfills the previously defined characteristics. This workflow needs to combine BES and CFD tools to achieve an acceptable accuracy to link building and courtyard microclimate physics. 4- Development of experimental campaigns to record data in courtyard spaces in the hot and warm seasons, to be used in the calibration and validation of the simulations to optimize their performance. Monitored case studies need to be representative of the existing built environment, comprise different kinds of courtyards, and include different improving performance strategies data. 5- Validation and calibration of simulation methodologies applied to the simulation of outdoor microclimates using monitored data in order to achieve accurate data from the simulations. To this end, studies contrasting monitoring and simulation data applied to the study of the microclimate of courtyards are needed. 6- Numerical quantification of the benefits of different strategies to improve the thermal performance of courtyards. The different strategies are known by experience, but their effectiveness needs to be measured to inform the designer to decide the optimal one. 7- Development of simulation studies focused on the analysis of the thermal comfort of users in outdoor spaces. The focus on energy demand has been traditionally preferred. However, users comfort is gaining attention as the main performance indicator of the suitability of outdoor spaces. 8- Development of a methodology that incorporates the performance of the courtyard on the design of the building, able to predict the effects of the courtyard on the indoors. 9- Provide a tool for designers which can be used alongside existing tools or be easy to implement in the design process, to incorporate the benefit of courtyards in building design. 10- Design recommendations for buildings with courtyards, and the optimization of existing courtyards, to improve the benefits of this passive strategy for the general user. 11- Raise awareness about the benefits of courtyard spaces and the need for their accurate performance simulation. Chapter 1: Research Framework 29 1.3 Objectives and scope Taking into consideration the problems detected in the field of outdoor microclimate simulation affecting the sustainable design of buildings, previously described in Section 1.1, and focusing on our Mediterranean climate and local architecture, the main and specific objectives of this work are outlined below. 1.3.1 General objective The general objective of this work is the design of a new simulation methodology applied to the microclimate of courtyards to understand their thermodynamic performance considering the different factors involved. This methodology will be applied to the analysis of different strategies to improve the performance of the courtyard as a passive element to reduce the energy consumption of buildings and improve user comfort. 1.3.2 Specific objectives To achieve the previous general objective, four specific objectives are defined: O1. Identification of the existing approaches to the simulation of outdoor microclimates and their applicability to the specific case of inner courtyards of buildings through a detailed analysis of literature, considering different approaches and scales, in response to needs 1 and 2. This includes a characterization of the existing simulation tools from a designer's perspective. O2. Development of a new simulation methodology for the analysis of the different factors that affect the microclimate in courtyards of the Mediterranean climate and their implications on the outdoor comfort and energy efficiency of buildings. This process includes the monitoring of existing buildings as case studies from which to collect data to be compared to the simulation results in order to validate and calibrate the methodology, in response to needs 3, 4, and 5. O3. Application of the designed methodology to the evaluation of the implementation of different strategies to improve the performance of courtyards and their potential to increase user comfort in existing buildings, in response to needs 6 and 7. This includes the analysis of the factors that can be altered such as geometry, vegetation, shading devices, material properties such as albedo or thermal inertia, and ventilation. O4. Application of the designed methodology to the comprehensive evaluation of thermal performance of a building, focusing on user comfort, with especial emphasis on the user experience when using the proposed tools, in response to needs 8 and 9. O5. Definition of recommendations for the intervention in courtyards of existing buildings to improve energy efficiency and comfort and result dissemination to society in form of Workflow Design for the Analysis of Transitional Space Microclimates 30 research communications and informative interventions to the public, in response to needs 10 and 11. 1.3.3 Scope Courtyards are a traditional element present in many building typologies around the world ( Figure 1.3-1 ). In addition, the scientific interest in courtyards in the recent years has increased exponentially. Before the beginning of this PhD work (year 2018), the number of articles in English found in Scopus using the keywords courtyard , microclimate , simulation, and thermal comfort was only 100. At the time of writing these lines (end of 2022), the number of scientific papers has risen to 326 (226% increase over this period). This manifests what a cutting-edge topic is the courtyards. Figure 1.3-1. Distribution of courtyard buildings around the world. Adapted from [76] . This work is focused on the specific strategy of courtyards in the Mediterranean climate, and the recommendations for courtyard improvement consider existing buildings and low-cost interventions. However, the simulation procedures that are developed intend to be suitable to any other outdoor microclimate and weather conditions, and the applicability of the results is wider, aiming for a greater impact of the results on the practitioner world. The same methodologies proposed here could be implemented in other climates, building typologies, or urban microclimate analyses affecting buildings. Chapter 1: Research Framework 31 1.4 Thesis Methodology The methodology followed in this thesis combines a critical analysis of the state of the art, the experimental analysis of case studies, and the development of simulation outputs to achieve the goals exposed in the objectives section. In this sense, the methodology is divided into four parts and ten tasks as follows: A. Analysis of the state of the art M1. Literature review in the field of outdoor microclimate simulation and building simulation for the identification of promising simulation tools. M2. Application of existing simulation tools for their evaluation as an option to simulate courtyard microclimates. B. Research methodology development M3. Selection of a representative number of case studies of building with courtyards in the Mediterranean climate. M4. Planning field monitoring campaigns to collect experimental data from the case studies selected. M5. Design of a new and optimized simulation methodology for the specific analysis of the microclimate of courtyards. M6. Validation of the novel simulation methodology through the comparison of monitored and simulated data, based on statistical parameters for error analysis. C. Research methodology application M7. Application of the simulation methodology to the analysis of the influence of the different factors on user comfort. M8. Integration of the simulation methodology to the comprehensive analysis of building thermal performance. M9. Summary of recommendation and implication of the findings from previous steps. D. Dissemination plan M10. Publications and participation in informative events. This thesis is developed during a four-year (plus five months due to the COVID pandemic) predoctoral contract (FPU17/05036), supported by the Spanish Ministry of Education, Culture and Sport, at the University of Seville. This period includes nine months of temporary research stays at the Harvard Graduate School of Design, supported by a Fulbright grant for Study Abroad 2018/2019 and a Real Colegio Complutense in Harvard Scholarship. Workflow Design for the Analysis of Transitional Space Microclimates 32 The methodology workflow is diagramed in Figure 1.4-1 , relating each methodology phase (M1-10) with the specific objectives (O1-5) and the needs defined in the introduction section of this thesis (N1-11). Figure 1.4-1. Methodology workflow followed in this thesis. Materials, experimental procedures, numerical simulation methods, and validation indicators are next defined. 1.4.1 Context and Case Studies The Mediterranean geographical context of this thesis has a climate defined as Csa according to the Köppen classification [77], specifically, hot dry summers with temperatures reaching above 40ºC, mild winters, and little mean precipitation throughout the year. This weather is shared by the two cities where most of the selected case studies are located: Sevilla (37°17′01″N 5°55′20″W, elevation 42 m a.s.l) and Cordoba (4°46'21.9"W, 37°53'29.58N, elevation 106 m a.s.l). Existing case studies are selected for experimental campaigns needed to test and validate the procedures developed in this work. Selection criteria respond to two main ideas: representativity of the selected case studies of the built environment of the studied area, and accessibility and flexibility for the analysis of the different parameters that are being analyzed: geometry, albedo, shading, thermal inertia, and vegetation. Chapter 1: Research Framework 33 The geometry of the courtyards is defined by their Aspect Ratio (AR), that is the relation between their height and width (see eq.1), and the Sky View Factor (SVF) which is the ratio of the sky that can be directly seen from a specific point to the whole hemispheric sky which would be seen without any elements blocking the view (see eq.2 and Figure 1.4-2 ). 𝐴𝐴𝐴𝐴 =𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻ℎ𝑡𝑡 𝑊𝑊𝐻𝐻𝑊𝑊𝑡𝑡ℎ (1) 𝑆𝑆𝑆𝑆𝑆𝑆 =𝑉𝑉𝐻𝐻𝑉𝑉𝐻𝐻𝑉𝑉𝑉𝑉𝐻𝐻 𝑆𝑆𝑆𝑆𝑆𝑆 𝑃𝑃𝑃𝑃𝑃𝑃𝑡𝑡𝐻𝐻𝑃𝑃𝑃𝑃 𝑇𝑇𝑃𝑃𝑡𝑡𝑇𝑇𝑉𝑉 𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝑉𝑉𝐻𝐻ℎ𝐻𝐻𝑃𝑃𝐻𝐻 (2) 1.4.2 Experimental procedure The field monitoring campaigns were carried out during the hottest months of the year, in a minimum period of one week up to one month in each courtyard. The variables recorded are air temperature, relative humidity, and wind speed and direction outside the courtyard, using PCE-FWS 20 weather stations. Inside the courtyard, the variables recorded are air temperature and relative humidity in all cases using TESTO 174H sensors, and some of them include mean radiant temperature using a black globe thermometer QUESTemp° 34 model. Table 1.4-1 shows the technical information of the measurement instrument used. Figure 1.4-2. The sky view factor is ratio of the sky that can be directly seen from a specific point to the whole hemispheric sky. Courtyard in the Royal Tobacco Factory of Seville. Source: Universidad de Sevilla [78] . Workflow Design for the Analysis of Transitional Space Microclimates 34 Table 1.4-1. Technical data of the measurement instruments. Instrument Variable Accuracy Resolution TESTO 174H Dry bulb Temperature ±0.5 °C 0.1 °C Relative Humidity ±0.1% 2% PCE-FWS 20 Dry bulb Temperature ±1 °C 0.1°C Relative Humidity ±5% 1% Wind speed ±1 m/s - QUESTemp° 34 Globe Temperature ±0.5 °C 0.1°C The setting up follows a methodology previously determined by the research group TEP 206 and already published [79], i.e., the weather station is located on the roof of the building, separated from the floor to avoid overheating, and the sensors in the courtyard are located at different heights from the floor, close to the shaded wall of the courtyard to avoid direct solar radiation. They are protected by open shields and separated from the wall by at least 10 cm. Figure 1.4-3 shows some examples of the instrument installed in a case study. a) TESTO 174H b) PCE-FWS 20 c) QUESTemp° 34 Figure 1.4-3. Images of the measuring instrument installed in the building. 1.4.3 Computational methods The numerical simulations in this work are performed using two existing software programs selected from the literature review according to the needs of the analysis: they are CFD tools, allow the analysis of outdoor microclimates, include thermal comfort analysis and are available to users. These tools are ENVI-met and the Ladybug Tools. They have been incorporated into a novel workflow simulation methodology that is explained in depth in Chapter 2. ENVI-met is one of the most widely used Computational Fluid Dynamics (CFD) software for urban microclimate simulations on a global scale. In addition to considering variables such as solar radiation, wind, soil, and vegetation, it uses the BioMet tool to calculate comfort indexes Chapter 1: Research Framework 35 and allows the definition of materials by thermal characteristics, such as albedo and emissivity, which makes it appropriate for the analysis of the influence of different parameters on the thermodynamic performance of the courtyard and users’ comfort. ENVI-met uses the Finite Difference numerical method to solve the multitude of partial differential equations (PDE) and other aspects of the model. The Ladybug Tools are a set of open-sourced plugins for Grasshopper that allows the connection between different validated simulation software such as EnergyPlus (energy simulation), OpenFOAM (CFD simulation), or Daysim (daylighting analysis) to the graphical user interface of the modeling tool Rhinoceros [42]. It is open-source, and a great community of researchers and developers are continuously updating their components. Its flexibility and interoperability make it suitable for the development of new workflows and procedures, also allowing parametric analysis. For the validation of the results from the simulations the statistical parameters used are the Coefficient of Determination (R2), which indicates how close simulated values are to the regression line of the measured values; the Root Mean Square Error (RMSE), the Systematic Root Mean Square Error (RMSEs), and the Unsystematic Root Mean Square Error (RMSEu), which measure the variability of the errors between measured and simulated values; and the Mean Absolute Percentage Error (MAPE). An explanation of these metrics can be found in [80]. The desired values for these indexes should tend to R2 → 1, RMSE → 0, RMSEs → 0, RMSEu → RMSE, and MAPE → 0 [81]. There are no guidelines for the validation of outdoor simulations, for that reason, the comparison with previous studies and validated tools will be the criteria for validation. Another model validation criterion that is used is the ASHRAE Guideline 14-2014 [82]. Although this guideline is for energy simulations in conservation projects, the recommendations of accuracy that it provides are to be considered for good reliability: If hourly calibration data are used (which is our case), the Normalized Mean Bias Error (NMBE) should be lower than 10% and the Coefficient of Variation of the Root Mean Square Error (CV-RMSE) should be lower than 30%. Additionally, it is recommended to obtain a value of R2 higher than 0.75. Workflow Design for the Analysis of Transitional Space Microclimates 42 will be validated in this work using typical courtyard dimensions and characteristics of the Mediterranean climate. The second software selected to be analyzed is Ladybug Tools. This set of plugins for Grasshopper is gaining attention recently for its advantages, such as the union of many software into one single interface, being developed for designers, and being open-source [103]. Some researchers [53] used the Ladybug Tools to analyze the impact of courtyard design variables on thermal comfort using parametric analysis with 8600 alternatives. A similar workflow was used by Natanian et al. [40] to optimize the urban design under three performance indicators: energy load, daylight, and thermal comfort. They found that the courtyard building typology achieved the optimal combination with the tested criteria. However, they noted the necessity of including numerical models to increase accuracy in the predicted microclimate conditions and recommended further validation of the tools. There are some CFD modules for the Ladybug Tools used by previous research [104] to simulate wind microclimatic conditions in the urban environment, nevertheless, they did not account for temperature variations, a variable that is crucial in the performance of courtyard spaces. The developers of the Ladybug Tools published a study analyzing a workflow they designed for the simulation of urban microclimates and they stated in the “ Limitations and future work ” section that “…the surface temperatures from the EnergyPlus simulation were not used to inform the 36 CFD simulations… Needless to say, future research should still include a validation of this method against fully-integrated engines as well as empirically measured climate conditions of the urban environment. ” [51], which is the gap to be overcome with this thesis. These tools will be used in this chapter and validated using experimental data. A detailed description of the simulation inputs and workflow are shown in Section 2.2 and the results comparison and validation in Section 2.3. A discussion of the results finalizes this chapter. Chapter 2: Analysis of simulation procedures for courtyards 43 2.2 Workflow methodology 2.2.1 ENVI-met simulation workflow This software was tested using three case studies, CS1, CS2 and CS3 in Seville, described as follows: • CS1: A multi-family building located in an area of high compactness but low-height edifications and few green areas nearby in Dos Hermanas (Seville). The building typology is characterized by deep courtyards to provide daylight to the indoor bedrooms. The walls, displaying some windows, have a coating material of cement mortar white colored. The ground surface is covered with ceramic tiles. It lacks any kind of shading, although its depth protects most of the wall surface against direct solar radiation ( Figure 2.2-1 a ). • CS2: A two-story building in Dos Hermanas (Seville) used as a primary school organized around the inner courtyard employing a closed gallery allowing access to the classrooms. The coating material of the walls is cement mortar white colored and it has large windows. The ground of the courtyard is covered with flowerbeds and tiles. Two palm trees in the center of the courtyard protect it from direct solar radiation during the middle hours of the day ( Figure 2.2-1 b ). • CS3: A multi-family building located in a densely populated area displaying a space of vegetation around it in Dos Hermanas (Seville). The building presents three inner courtyards divided by the main stairwells which provide light to the bedrooms. The cladding of the courtyard walls is cream color stone, red-painted cement mortar, and small glazed windows. The ground is covered with grey tiles ( Figure 2.2-1 c ). The geometry characteristics of the courtyards are shown in Table 2.2-1 . a) CS1 b) CS2 c) CS3 Figure 2.2-1. Images of the buildings selected as case studies. These cases were monitored in June 2017, between the fourth and the ninth days. Air temperature, relative humidity, wind speed, and direction were monitored with the weather station located on the roof of one of the case studies. The air temperature was monitored inside the courtyards at 1.5 m above the floor. Then, the eighth of June was selected to be compared with simulated data. Workflow Design for the Analysis of Transitional Space Microclimates 44 Table 2.2-1. Case Studies' main characteristics. Adapted from Publication 1. Case Study Width (m) Length (m) Height (m) AR (Axis 1, Axis 2) 3D Sky View Factor Albedo (Wall, Floor) CS 1 3.1 7.4 12.6 4.1, 1.7 0.03 0.8, 0.4 CS 2 7.0 11.0 8.9 1.3, 0.8 0.12 0.8, 0.4 CS 3 7.5 13.2 10.7 1.4, 0.8 0.12 0.4, 0.5 The model geometry of each case study is shown in Figure 2.2-2 , although more information can be found in Publication 1, as well as the main input variables for the material properties and vegetation. The model was designed following the recommendations of previous research about grid size, context domain, and model size to balance accuracy and computer power requirements [105,106], following an iterative process in which the resolution and the time frame were modified to obtain the best possible results. In this case, the model included the building of the courtyard and some surroundings, although previous research indicates that the surroundings of the building with the courtyard are not needed, the outdoor environment is simulated to compare with experimental data. CS1 CS2 CS3 Figure 2.2-2. ENVI-met model of the courtyards. Top: perspective. Down: Top view. Extracted from Publication 1. The hourly temperature and relative humidity monitored outside the building by the weather station were inputs for the Envi-met simulation. The temperature of the soil at different depths was also needed and obtained from previous studies [87,107], and the specific humidity at 2500 m height was calculated using the method described by Aysan Forouzandeh [35]. The ENVI-met software solves the Reynolds-averaged non-hydrostatic Navier-Stokes equations for each time step and grid in the model. The turbulence in the air is predicted using the 2- equation Turbulence Kinetic Energy (TKE) model [108]. The orthogonal Arakawa C-grid Chapter 2: Analysis of simulation procedures for courtyards 45 numerical discretization scheme [109] is used and the Finite Difference numerical method solves the multitude of partial differential equations (PDE) and other aspects of the model. The simulation requires an initialization time to provide accurate data, which is why 40 hours were simulated but only the last 24 were used as results, as suggested by other researchers [35,87]. 2.2.2 Ladybug Tools simulation workflow The specifications previously described of the Ladybug Tools allow for the combination of different software and the design of scripts appropriated for the specific objectives of the desired simulations. In this thesis, a novel script has been designed for the specific simulation of microclimates where temperature variations are important, such as courtyards. The wind flows inside the courtyard may not be the main driver for the microclimate but the convention flows generated by surface temperature differences, which affect the performance of the space. For that reason, the Ladybug Tools’ plugin Butterfly is used, which links to the CFD software OpenFOAM that has a heat transfer solver allowing CFD calculations using surface temperature as boundary conditions. Given that it is a new contribution to the research community, it needed validation, which was done by contrasting simulation and monitoring results. Four case studies, CS4, CS5, CS6 and CS7 were used. The main geometric characteristics of the courtyards are shown in Table 2.2-2 and the buildings are described as follows: • CS4: A residential building in the city center of Cordoba, in an area of high compactness close to the historic quarter of the city. It is a two-story traditional courtyard house with a nearly square courtyard surrounded by an arcade on three of its four sides. The building is currently unoccupied. The walls, which include windows on the three arcaded sides, are coated with white lime, and the ground is covered with light concrete ( Figure 2.2-3 a ). • CS5: A two-story house with a small courtyard in the dense area of the city center of Cordoba. The courtyard’s walls include large windows on the lower level and small windows on the upper level. The coating is cement mortar painted in white ( Figure 2.2-3 b ). • CS6: A multifamily residential building located in a dense area of Seville. It is a six-story building and the courtyard studied is used mainly for ventilation and illumination of interior areas; thus, it has a high AR. The wall’s coating is cement mortar painted white ( Figure 2.2-3 c ). • CS7: The inner courtyard of the CITIUS building of the University of Seville. The building is in an urban area with a high density of mid-rise constructions. The walls’ coating is white cement mortar. There are small, glazed openings distributed sporadically in the walls. There is no shading element or vegetation in the courtyard ( Figure 2.2-3 d ). Workflow Design for the Analysis of Transitional Space Microclimates 46 Table 2.2-2. Case Studies main geometry characteristics. Adapted from Publications 2 and 3. Case Study Width (m) Length (m) Height (m) AR (Axis 1, Axis 2) 3D Sky View Factor Albedo (Wall, Floor) CS 4 7.8 8.4 6.8 0.9, 0.8 0.17 0.7, 0.4 CS 5 4.5 4.5 7.0 1.5, 1.5 0.07 0.7, 0.4 CS 6 3.5 3.5 16.0 4.6, 4.6 0.01 0.7, 0.4 CS 7 6.6 7.3 14.0 2.1, 1.9 0.04 0.8, 0.4 a) CS4 b) CS5 c) CS6 d) CS7 Figure 2.2-3. Image of the buildings selected as case studies. The field monitoring campaign was carried out during the summer (June to September) of 2017, in the same procedure that was described before. One day for each case was selected to compare simulation and experimental results. The monitored data outside the buildings were used as inputs to modify an .epw file which is the weather file used by the software. The geometry of the building was modeled in Rhinoceros in a simplified way, following the general recommendations of the software. The script of the workflow is represented in Figure 2.2-4 . The three plugins used are Ladybug (which manages the weather data inputs for simulations), Honeybee (which links to EnergyPlus Software to perform BES), and Butterfly (which links to OpenFOAM to perform the CFD outdoor simulation). The workflow functions as follows: after modeling the building and incorporating all the materials and internal loads, an energy simulation is performed by Honeybee, using the weather data imported by Ladybug. Then, the surfaces’ outdoor temperature is incorporated in the Butterfly simulation as boundary conditions to perform the CFD simulation, also using weather data from Ladybug (outdoor temperature and wind speed). This workflow provides the temperature inside the courtyard used for validation, but it can also provide other data such as mean radiant temperature, wind speed, or solar radiation inside the courtyard, data that is needed if comfort calculations are considered. More information about the configuration of the simulations can be found in Publication 3. Chapter 2: Analysis of simulation procedures for courtyards 47 Figure 2.2-4. Flowchart of how the inputs and outputs relate to the simulation tools and validation. One of the most critical aspects of the designed workflow is to ensure that the CFD simulation is correct. To guarantee that, the model was defined for each case following the general recommendations for best practices in CFD simulation of urban environments [110,111]. A wind tunnel was modeled around the building: the inlet, top, and lateral boundaries were placed 5H away from the building model, being H the height of the building, and the outlet 15H away, as shown in Figure 2.2-5 . Figure 2.2-5. Computational domain in Ladybug Tools models. Extracted from Publication 3. Workflow Design for the Analysis of Transitional Space Microclimates 48 The Butterfly plugin only incorporates two CFD solvers with limited functionalities included in its components. For that reason, the automatic definition of the atmospheric boundary layer in the inlet is not possible, and it was done manually. The inlet was divided into different surfaces to which a wind speed velocity was assigned, in order to define the logarithmic profile for the wind (Figure 2.2-5 .). The discretization of the different surfaces and velocities was done every 0.2 m/s speed increase of the wind profile. The logarithmic equation used was: 𝑢𝑢=𝑢𝑢∗ 𝑘𝑘 𝑙𝑙𝑙𝑙�𝑧𝑧−𝑑𝑑+𝑧𝑧0 𝑧𝑧0� 𝑣𝑣=𝑤𝑤= 0 where 𝑢𝑢 = Ground-normal streamwise flow speed profile [m/s] 𝑣𝑣 = Spanwise flow speed [m/s] 𝑤𝑤 = Ground-normal flow speed [m/s] 𝑢𝑢∗ = Friction velocity [m/s] 𝑘𝑘 = von Kármán constant [-] 𝑧𝑧 = Ground-normal coordinate component [m] 𝑑𝑑 = Ground-normal displacement height [m] 𝑧𝑧 0 = Aerodynamic roughness length [m] The friction velocity 𝑢𝑢∗ is defined as follows: 𝑢𝑢∗=𝑢𝑢𝑃𝑃𝐻𝐻𝑟𝑟𝑘𝑘 𝑙𝑙𝑙𝑙�𝑧𝑧𝑃𝑃𝐻𝐻𝑟𝑟 +𝑧𝑧0 𝑧𝑧0� where 𝑢𝑢 𝑃𝑃𝐻𝐻𝑟𝑟 = Reference mean streamwise wind speed at z ref [m/s] 𝑧𝑧 𝑃𝑃𝐻𝐻𝑟𝑟 = Reference height being used in 𝑢𝑢∗ estimations [m] Figure 2.2-6 shows an example of the wind speeds defined at different heights for a wind profile generated for a wind speed of 2 m/s at a reference height of 10 meters. The blockMesh and snappyHexMesh utilities were used to define the mesh of the model with a refined area around the building and the courtyard area of three or four levels depending on the model. The final meshes were 2.5x106, 2.3x106, and 2.0x106 cells in CS4, CS5, and CS6 respectively. Figure 2.2-7 shows images of the mesh in each case. For more information about the construction of the domain and the boundary conditions see Publication 3. Figure 2.2-6. Wind profile in Ladybug Tools models for a wind speed of 2 m/s at 10 m height. Extracted from Publication 3. Chapter 2: Analysis of simulation procedures for courtyards 49 CS4 CS5 CS6 Figure 2.2-7. Ladybug Tools models’ grid. Extracted from Publication 3. These meshes were selected through a grid sensitivity analysis performed with different levels of refinement: coarse, medium, and fine, with a 1.5 linear factor between the cells. Figure 2.2-8 shows the results of temperature at different heights in a vertical line in the center of the courtyard in the three levels of refinement in each case. The average percentage error between the coarse and the medium was 1.8%, 1.2%, and 2.4% in CS4, CS5, and CS6 respectively, and between the fine and the medium was 1.4%, 3.9%, and 1.8% in CS4, CS5, and CS6 respectively. These results allowed us to conclude that the grid dependency was acceptable in the three cases using the medium refinement, optimizing this way the computational time. CS4 CS5 CS6 Figure 2.2-8. Results of temperature in a vertical line of the model for the grid sensitivity analysis. Extracted from Publication 3. Workflow Design for the Analysis of Transitional Space Microclimates 50 The heat transfer solver incorporated in Butterfly is the buoyantBoussinesqSimpleFoam, a steady-state solver for buoyant, turbulent flows of incompressible fluids that uses the Boussinesq approximation [46], and the SIMPLE (Semi-Implicit-Method-Of-Pressure-Linked- Equations) coupling velocity-pressure scheme [112]. The turbulence model is the RNG kepsilon. OpenFOAM uses an upwind finite volume method, meaning the divergence terms are discretized by the bounded Gauss linear upwind method, cell limiters for the gradient approximations are considered, and linear interpolation. Initial parameters for the simulation, boundary conditions for the domain limits, inlet and outlet, and the building geometry are detailed in Publication 3. Chapter 2: Analysis of simulation procedures for courtyards 51 2.3 Simulation Results and Validation 2.3.1 Monitoring Results This section analyses the outcomes of the field monitoring campaigns performed in the six courtyards described in the previous section. All of them were used to validate the simulation methodologies. Results obtained from the monitoring campaigns confirmed the previously reported tempering effect of these spaces, even if this effect was not equally beneficial in every case. Figure 2.3-1 shows the results of the three courtyards monitored for the validation of ENVI-met software. The maximum temperature on the 8th of June in 2017 reached 38ºC and minimum temperatures did not fall under 21ºC. Wind speed at the top of the building was low, with a mean value of 3 Km/h, the typical wind speed in this area. The temperature inside the courtyards at 1.5 m above the floor showed a significant reduction in the maximum temperature compared with the outdoors. The best performance case was CS1, the courtyard with the highest AR, given that the depth of the courtyard prevents solar radiation from reaching the lowest parts of the courtyard. In contrast, it showed an overheating effect during the night, given that the lower the SVF, the more difficult to release the heat accumulated during the day. The other two courtyards had a similar AR, however, CS2 performed better than CS3. This can be explained by their characteristics: CS2 had vegetation providing shade, while CS3 had no vegetation or shading elements. Furthermore, their surfaces had a lower albedo, storing more heat from solar radiation than surfaces in CS2. During the night, these two cases showed no overheating effect, as their lower AR allowed heat to be released. Figure 2.3-1. Monitoring data of Air Temperature and Wind Speed in CS1, CS2, and CS3 on June 8th. Adapted from Publication 1. Workflow Design for the Analysis of Transitional Space Microclimates 58 CS4. Monitoring and Simulation Results. August 20th CS5. Monitoring and Simulation Results. August 19th CS6. Monitoring and Simulation Results. June 16th Figure 2.3-7. Temperature comparison of monitoring, ENVI-met, and Ladybug Tools workflows results. Extracted from Publication 3. Figure 2.3-8. Ladybug Tools simulation results and monitoring. a) Hourly data (August 2nd, 2018). b). East-west section of the courtyard at different times. Extracted from Publication 2. Chapter 2: Analysis of simulation procedures for courtyards 59 The statistical analysis of the simulated air temperature results at 1.5 m was performed using the Coefficient of Determination (R2), the Root Mean Square Error (RMSE), Systematic Root Mean Square Error (RMSEs), Unsystematic Root Mean Square Error (RMSEu) and Mean Absolute Percentage Error (MAPE) (see Chapter 1, Methodology, for further explanation about these parameters). These values per each case study and simulation workflow are shown in Table 2.3-2 . In ENVI-met simulations, R2 ranged between 0.84 and 0.96, which was a good value, but MAPE ranged between 6.85 and 14.38, and RMSE between 2.61 and 5.22, which were slightly high, although in accordance with the validations done before. In Ladybug Tools simulations, R2 was lower than in ENVI-met (0.73 to 0.94), however, this coefficient is considered insufficient and unreliable by itself [81]. For that reason, this work considers these values less significant than the other parameters. Results for MAPE (3.81 to 7.55%) and RMSE (1.37 to 2.59ºC) showed better accuracy in Ladybug Tools than ENVI-met workflow. Table 2.3-2. Quantitative evaluation of the simulation’s performance for the courtyard temperature output. Adapted from Publications 2 and 3. AR R2 MAPE (%) RMSE (ºC) RMSEu (ºC) RMSEs (ºC) ENVI-met simulation CS4 0.9 0.96 6.85 2.61 3.96 1.42 CS5 1.5 0.92 8.74 3.46 5.54 2.20 CS6 4.6 0.84 14.38 5.22 6.76 1.66 CS7 2.1 0.89 10.08 4.19 6.21 2.24 Ladybug Tools simulation CS4 0.9 0.94 3.81 1.37 2.41 1.12 CS5 1.5 0.73 5.07 2.00 2.41 0.96 CS6 4.6 0.78 7.55 2.29 3.09 1.03 CS7 2.1 0.89 7.18 2.59 4.39 1.93 Legend: The best result of each parameter between the two simulation workflows. The worst result of each parameter between the two simulation workflows. Workflow Design for the Analysis of Transitional Space Microclimates 60 2.4 Summary and next steps This chapter shows the validation process of two different workflows for the simulation of the outdoor microclimate applied to the specific case of courtyards. The first one using an existing tool already known for urban outdoor microclimate simulation, and the other by means of a novel workflow using the Ladybug Tools including the use of the Butterfly plugin for CFD calculations with temperature variations. Although the results from both methodologies are considered valid, the Ladybug Tools workflows were found more accurate than the ENVI-met ones. In order to explain these differences, given that both methods use the same RANS turbulence model with the Boussinesq approximation, other aspects of the CFD model need to be analyzed, such as the model boundary conditions or the grid resolution: • The mesh generated in the Ladybug Tools model is finer than in ENVI-met, which could lead to more accurate results. This is because, in ENVI-met, the grid size is limited to a minimum of 0.5 m, which is usually enough to simulate large urban spaces but could be insufficient for smaller spaces such as courtyards. • The boundary conditions in ENVI-met are obtained through a simplified method to simulate building temperatures that affect the surface temperatures in the courtyard. In LBT, the surface temperatures in the courtyard are obtained using a BES for the whole building in EnergyPlus, which may produce more accurate results. In the two methodologies, one interesting aspect is that the higher the AR, the lower the accuracy of the simulation results. This correlation is relevant considering that, regarding experimental results, the deeper the courtyard, the higher the tempering effect that produces. The reason could be that in deeper courtyards, the microclimate is more influenced by the inner conditions of the building than in shallower courtyards, which are more influenced by outdoor conditions. This makes it essential for achieving better results, good modeling of material properties, and building operations. Results shown in this chapter contribute to meeting the needs 1, 2, 3, 4, and 5 identified in Chapter 1, fulfilling the specific objectives 1 and 2 of this thesis. Further steps in this thesis include the use of these workflows for the analysis of different strategies to improve the performance of courtyards (Chapter 3) and their influence on the comfort of users both, in the courtyard and inside the building (Chapter 4). Chapter 3: Simulation analysis to improve the thermal performance of courtyards 61 Chapter 3: Simulation analysis to improve the thermal performance of courtyards 3.1 Introduction 3.2 Geometry and orientation 3.3 Albedo and vegetation 3.4 Shading device 3.5 Ventilation and thermal inertia 3.6 Summary and next steps Summary This chapter shows the results of using the simulation methodologies analyzed in the previous chapter on the analysis of the effectiveness of different strategies to improve the thermal performance of courtyards: the geometry, albedo, vegetation, shading devices, ventilation, and thermal inertia. The performance is measured using different indicators, focusing on two: air temperature and thermal comfort, as a measure of energy performance and user’s suitability. Related Publications This chapter incorporates analysis, data, and conclusions from the peer-reviewed publications: “ Albedo influence on the microclimate and thermal comfort of courtyards under Mediterranean hot summer climate conditions ” in Sustainable Cities and Society, (Publication 4); “ Effect of thermal inertia and natural ventilation on user comfort in courtyards under warm summer conditions ” in Building and Environment (Publication 5), and “ Shade and thermal comfort in courtyards: experimental versus simulation results” in Buildings (Publication 6). Workflow Design for the Analysis of Transitional Space Microclimates 62 3.1 Introduction Courtyards are present in the architecture of warm climates of many cultures around the world and throughout history [113]. Traditionally, it has been known to function as a thermal regulator, in addition to cultural and social implications inherent in the architecture with courtyards. The thermal performance of the courtyard is driven by the thermodynamic effect that occurs inside them due to differences in surface temperature produced by the shading of the surfaces Figure 3.1-1 . The different temperatures generate convective flows in the air which involve heat transfer between the surfaces and the air. Heat is also stored in the thermal mass of the walls and released during the night as infrared radiation. Another part of the heat is transferred through conduction and ventilation between the courtyard and the outdoor or indoor. Figure 3.1-1. Thermodynamic effects inside a courtyard. Adapted from [114]. These thermodynamic effects determine the performance of the courtyard and can be regulated to improve it. Different design factors can be employed to improve the thermal and microclimatic performance of the courtyard [115]: • Related to heat gain: Solar radiation is the main source of heat inside the courtyard. For that reason, blocking it helps to regulate the amount of heat that reaches the courtyard. Their geometry, that is, the relation between the height and width (Aspect Ratio) or the proportion of sky that is visible from within the courtyard (Sky view factor), together with Chapter 3: Simulation analysis to improve the thermal performance of courtyards 63 their orientation and location, are the main factors that determine the solar radiation that reaches the courtyard [116]. However, other elements can block solar radiation: shading elements such as awnings [117,118], galleries around the courtyard [119], or even vegetation [13]. They can help to reduce solar radiation that reaches the surfaces of the courtyard. Once the radiation reaches the surfaces, there are also strategies to regulate the heat that is stored in the material, which can help to improve the performance. These two variables are albedo and thermal inertial [115]. High albedo means that the proportion of solar radiation absorbed by the surface is low and a higher proportion is reflected, an effect that reduces the temperature reached by the surfaces in comparison to another one with a lower albedo. Thermal inertial helps to regulate not only the amount of heat stored by a surface but also the speed to release that heat. Albedo and thermal inertia are properties that depend on the materials, and intelligent management of them can optimize solar gains and releases inside the courtyard during the daily cycle. • Related to humidity: humidity is a factor that influences thermal comfort. Furthermore, the evaporative cooling of water and vegetation transpiration also influence the temperature inside the courtyard. The addition of vegetation, the incorporation of water nebulizers, or water elements inside the courtyard can regulate this effect [120]. The benefit of these strategies is limited to specific climatic zones, given that if the climate already has high levels of humidity, evaporative cooling is not possible. Figure 3.1-2. Example of courtyards with the presence of vegetation and water elements. (Own photos). Workflow Design for the Analysis of Transitional Space Microclimates 64 • Related to wind flow: The natural wind flow patterns generated inside the courtyard are related to other characteristics such as geometry, material properties, and elements inside the courtyard, in addition to weather conditions such as wind speed and direction [121]. Deeper courtyards are less affected by outside conditions than shallow courtyards. In contrast, they are more affected by the indoor functioning of the building. Openings can generate flow patterns between the building’s indoor and the courtyard, which may influence comfort in both spaces. Higher wind speeds could be desired concerning thermal comfort, and convective or wind patterns could not be enough. In these cases, forced wind flows could be an option to be analyzed. All these factors that intervene in the performance of the courtyard are interrelated, making it impossible to predict the performance of a courtyard without computer simulation, which is required to foresee these complex interactions. Even though, previous chapters have shown that there is no single tool that meets all the requirements to simulate these spaces. The two methodologies analyzed in the previous chapter are promising. This chapter uses them to predict the thermal performance of courtyards where the described design strategies have been implemented, to numerically quantify their effect. The design strategies analyzed are geometry, albedo, vegetation, shading elements, ventilation, and thermal inertia. An important aspect is that all the simulations are validated using experimental data. Chapter 3: Simulation analysis to improve the thermal performance of courtyards 65 3.2 Geometry and orientation 3.2.1 Temperature results. The previous section has described the influence and decisive role of the geometry of the courtyard on its thermal performance. The geometry here is understood as the combination of the AR and the SVF, concepts described in Section 1.4. The geometry of the courtyard is restricted by urban norms or building program necessities most of the time, and it generally cannot be modified in a retrofitting intervention. However, knowing its influence on courtyard performance parameters is useful to provide design recommendations and understand the limitations of other interventions due to predetermined configurations. The previous chapter showed results of air temperature inside the courtyard with different geometries. Table 3.2-1 summarizes the data of those courtyards which also share other characteristics, in order to be able to compare the results and associate the differences only with geometry: they have similar albedo, no vegetation, no shading devices, and were simulated using the same workflow, LBT. In this table, the thermal range is understood as the difference between the maximum and minim temperatures on the analyzed day. This information is a summary of the simulations performed in the publications in Publications 2 and 3. Table 3.2-1. Summary of simulation results for courtyards with different geometries. Case Study AR (Axis 1, Axis 2) SVF Outdoor thermal range (ºC) Courtyard thermal range (ºC) Maximum Thermal Delta (ºC) Minimum Thermal Delta (ºC) CS 4 0.9, 0.8 0.17 42.1 - 26.1 36.5 - 28.4 6.0 -2.0 CS 5 1.5, 1.5 0.07 44.1 - 25.1 37.0 – 28.0 7.7 -2.9 CS 6 4.6, 4.6 0.01 42.8 – 22.9 33.7 - 24.4 9.1 -2.5 CS 7 2.1, 1.9 0.04 45.8 – 26.0 36.8 – 29.1 9.0 -3.7 This exemplifies how the geometry correlates with the thermal delta that the courtyard can achieve: the lower the SVF, the higher the maximum thermal delta in the courtyard. This does not pretend to be an exhaustive analysis of the geometry, which would have been done with parametric analysis, but a validation of the tool as a way of reproducing the benefits of geometry. 3.2.2 Thermal comfort results. Another parameter that was tested for this simulation workflow for courtyards using these case studies was thermal comfort. Using LBT, thermal comfort can be incorporated easily into the workflow, as shown in Figure 3.2-1 : the butterfly simulation already provides air temperature Workflow Design for the Analysis of Transitional Space Microclimates 66 and wind speed. Then, using the mean radiant temperature from the BES simulation and the relative humidity from the weather file, the UTCI can be calculated. Figure 3.2-1. Ladybug Tools workflow to obtain comfort results. The results of UTCI obtained with this workflow are represented in Figure 3.2-2 , at different hours of the day (8:00, 12:00, 16:00, and 20:00 hours). These are the same courtyards used to validate the LBT methodology in the previous chapter. CS4 showed the highest values of UTCI provoked by the sun hitting the walls and floor of the courtyard, which was the one with the lowest AR, thus the self-shading of the walls was limited. Solar radiation increased the surface temperatures in the courtyard, especially during the early afternoon. The arcade on three of the walls of the courtyard provided a positive effect with an area without solar radiation around the courtyard, although the elevation of the sun made it reach these areas in the late afternoon. Despite UTCI values not falling into the comfort range (it was an extremely hot day), the difference in the values between shaded and non-shaded areas made the difference between slight and intense hot stress, up to 5ºC. CS5 also received the sun radiation in the lowest parts of the courtyard, although it was deeper thus the time suffering direct solar radiation was shorter, and only in the central hours of the day (noon). For the rest of the day, the UTCI values were homogenous in the courtyard with peak values of 40ºC when the sun reached the floor. CS6 is the deepest courtyard, and this was manifested by the absence of solar radiation in the lowest part of the courtyard, where no sudden changes in the UTCI appeared at any time. The UTCI was relatively constant during the day and provided the lowest values of all the courtyards analyzed here. Finally, note that the UTCI values in the courtyard were generally higher than the air temperature simulated, especially during the afternoon, when the sun was high and solar radiation hit the walls and the floor of the smaller AR courtyards. This was mainly due to MRT results, given that RH and wind speed were low. Chapter 3: Simulation analysis to improve the thermal performance of courtyards 67 This simulation was also performed using ENVI-met and contrasting the results between the two software. This analysis is included in Publication 3. CS4 CS5 CS6 8:00 h 12:00 h 16:00 h 20:00 h Figure 3.2-2. UTCI results at 1.5 m above the ground for CS4, CS5, and CS6 from Ladybug Tools workflows at different hours of the day. Adapted from Publication 3. Workflow Design for the Analysis of Transitional Space Microclimates 74 3.4 Shading device 3.4.1 Case study and methodology. CS5 courtyard had the possibility of being analyzed with a shading device installed. For that reason, it was selected to compare the performance of a building with and without a shading device installed and evaluate the accuracy of the simulation workflow which was also used to quantify the thermal comfort. The courtyard was monitored during the summer of 2017. Two days with similar highest temperatures, around 44ºC, were selected to be simulated: August 4th (with the shading device installed) and August 18th (without the shading device installed). These days the building was unoccupied due to summertime vacations. The shading device was a black polyethylene fabric of 60–70% UV filter, and around 70 g/m2 density, tied to the top of the railing on the flat roof of the building, as shown in Figure 3.4-1 . This allowed the ventilation of the courtyard in the space between the walls and the canvas. This system was a very cheap and easy-to-install element that provided comfort in the courtyard. Figure 3.4-1. Left: Image of CS5’s courtyard with the shading device installed. Right top: Section of the building. Right down: Plan of the building. Extracted from Publication 6. The courtyard was simulated using ENVI-met and the Ladybug tools novel workflow designed in this thesis. The LBT results are shown in this chapter, which produced the more accurate results. Chapter 3: Simulation analysis to improve the thermal performance of courtyards 75 3.4.2 Monitoring results. Figure 3.4-2 shows the monitored results of air temperature on the selected days and the thermal difference between the courtyard and the outdoors (thermal delta). Outdoor temperatures showed a similar trend on both days, close to 45ºC of maximum temperature and 25ºC minimum. Inside the courtyard, the maximum temperature reached was different between the two days. Without the shading installed, the maximum temperature inside the courtyard reached 37ºC. However, with the installation of the canvas, the temperature inside the courtyard did not rise above 31ºC. In terms of thermal delta, the tempering effect of the courtyard was improved from 7.4ºC maximum without the shading device to 13.7ºC with the shading installed. During the night, a slight overheating effect is shown on both days, with and without the shading device installed. Figure 3.4-2. Monitored Outdoor and Courtyard Dry Bulb Temperature on the selected days. Extracted from Publication 6. 3.4.3 Air temperature simulation results. The simulation results of air temperature inside the courtyard are shown in Figure 3.4-3 . During the no-shaded day, the maximum temperature inside the courtyard reached 36ºC, and the maximum thermal delta was 8.2ºC, results that were closed to monitored data. On the shaded day, in contrast, the maximum thermal delta achieved was only 9.5ºC, and not 13.7ºC as monitored. This means that the simulation was not able to reproduce accurately the improved performance of the courtyard with the shading, but still, it was able to increase the thermal delta, not only the maximum delta but also during most of the warm hours. Another effect that the simulation was able to reproduce is the overheating effect during the night, although in this case, it was higher in the simulations than in the monitored data. Workflow Design for the Analysis of Transitional Space Microclimates 76 The statistical parameters R2, RMSE, CV(RMSE), and NMBE were calculated to validate the results of the simulation. They are shown in Table 3.4-1 . The simulation without the shading device installed showed higher accuracy than the shaded simulation. The reasons can be two: it is needed more precise modeling of the device (transparency parameters for example), and the BES simulation may need better calibration in this case, because the indoor parameters can influence more in the shaded option than in the no-shaded. Table 3.4-1. Quantitative evaluation of the simulation’s results for the courtyard temperature output. R2 RMSE (ºC) CV(RMSE) NMBE No-Shaded 0.73 1.90 6% -3.2% Shaded 0.64 4.26 16% -12.7% Figure 3.4-3. Ladybug simulated courtyard air temperature in the selected days. Extracted from Publication 6. 3.4.4 Comfort simulation results. The simulated air temperature and wind speed inside the courtyard were used to calculate thermal comfort using the UTCI index, using the methodology already shown in Figure 3.2-1 . The UTCI was obtained at 1.5 m above the floor of the courtyard, at four different hours on both days, as shown in Figure 3.4-4 . The results ranged from 28.3ºC to 34.5ºC when the courtyard was shaded and from 25.9ºC to 34.1ºC when it was shaded. The effect of the shading device was more evident in the morning when the solar radiation is partially blocked by it. In this case, the mean value of the UTCI inside the courtyard is a few degrees lower with the shading installed. In the afternoon, the shading benefit is reduced to only a few tenths of a degree. Chapter 3: Simulation analysis to improve the thermal performance of courtyards 77 Not shaded Shaded 8:00 h 12:00 h 16:00 h 20:00 h Figure 3.4-4. UTCI values at 1.5 m height in the courtyard at different hours from the Ladybug Simulation Workflow. Left: without shading. Right: Shaded. Extracted from Publication 6. Workflow Design for the Analysis of Transitional Space Microclimates 78 3.4.5 Discussion and recommendations. Monitoring results showed the huge benefit of installing a shading device over the top of a courtyard in terms of air temperature reduction at peak times. The thermal delta achieved up to 13.7ºC with the device installed, while without the shading the thermal delta was only 7.1ºC. The simulation was able to reproduce a similar thermal delta without the shading device installed, but the simulated thermal delta with the shading device was lower, up to only 9.3ºC. This comparison is represented in Figure 3.4-5 , showing different hours. Figure 3.4-5. Simulated and Monitored Thermal Delta (TD) at different hours for the different configurations analyzed. Adapted from Publication 6. The lower accuracy of the simulation with the shading device does not discard the use of the methodology to analyze the shading effect of devices on the courtyard, but only emphasizes the necessity of a better calibration of the model to achieve accurate data. Especially with the shading installed, the performance of the courtyard is more sensible to other variables that can affect its performance such as material properties or building functioning. For that reason, having a precise model of the building is also needed to simulate the microclimate of the courtyard and more precisely to achieve accurate data for the thermal comfort results. Chapter 3: Simulation analysis to improve the thermal performance of courtyards 79 3.5 Ventilation and thermal inertia 3.5.1 Case study and methodology. Ventilation and thermal inertia analysis was performed using the proposal presented by the University of Seville [26] for the Solar Decathlon Europe competition in 2019, in Szentendre, Hungary. It consisted of different modules plugged into a building according to their retrofitting needs. To represent this in the competition site, a square plan structure served as a connection between the different modules located at different heights. The space in the middle of the structure was configured as a courtyard which functions as an air exchanger with the rooms that surround it. The plan of the courtyard is a square of 4.9 m and it is 5.2 m in height, with an AR of 1.1 and an SVF of 0.09. The lightweight construction of the building was different from traditional materials. The connection structure façade was a double skin of semipermeable textile material. The modules were constructed with sandwich panels with insulation. This case study is selected to analyze how different materials perform in the courtyard. The monitoring campaign took place during the Solar Decathlon competition in Hungary in July 2019 to calibrate and validate the simulations. The methodology followed is displayed in Figure 3.5-1 : the monitoring campaign was used to validate the simulation model with Ladybug Tools and construct the weather data file. Then, different configurations for the construction solutions of the courtyard walls were simulated, in two different operational configurations, ventilated and non-ventilated courtyard. The building is modeled in passive mode. The ventilation of the courtyard was achieved considering all the openings of the building open or closed in the case of the non-ventilated courtyard. These openings allowed the direct connection of the courtyard with the outdoors of the building. However, in all cases, the ventilation was natural and depends solely on the outdoor wind speed. Figure 3.5-1. Methodology workflow for ventilation and thermal inertia analysis. Extracted from Publication 5. The three types of walls used in the simulations are described as follows. The characteristics of the materials are displayed in Table 3.5-1 : • CS9-BC: The courtyard wall had no inertia, just a PVC canvas covering the structure of the gallery, providing shade and wind protection to the building. This was the built state. Workflow Design for the Analysis of Transitional Space Microclimates 80 • CS9-1: The construction system in this model was a conventional brick wall with thermal insulation in the indoor layer. It had a higher heat capacity. • CS9-2: This construction was a traditional earth wall. It had a very high thermal capacity and is representative of the traditional architecture of many cultures. The performance indicators used to evaluate the simulations were air temperature, mean radiant temperature, and thermal comfort inside the courtyard, measured at 2 m height above the floor of the courtyard. The thermal comfort index used was the UTCI in this case. Table 3.5-1. Thermal properties of each simulation variation. Extracted from Publication 5. Type Description Thickness (mm) Density (kg/m3) Thermal conductivity (W/mk) Specific Heat Capacity (J/kgK) Thermal inertia (J/m2K√s) U-value (W/m2K) Brick wall (CS9-1) Brick wall 100 1920 0.89 790 1161.9 0.53 Air gap - - - - - Rockwool 45 40 0.032 800 32 Gypsum panel 15 700 0.25 1000 418.3 Earth wall (CS9-2) Rammed earth 250 1540 1.25 1260 1557.4 0.49 Air gap - - - - - Rockwool 45 40 0.032 800 32 Gypsum panel 15 700 0.25 1000 418.3 Gallery membrane (Base case) PVC canvas 2 150 0.17 840 146.4 3.68 3.5.2 Simulation results. Results of air temperature, MRT, and thermal comfort are shown in Figure 3.5-2 for the simulated day. In terms of air temperature, the CS9-BC, with no thermal inertia in the walls, achieved the highest temperatures inside the courtyard during the day, reaching up to 34.8ºC at 16 hours (higher temperature than outdoors), and the lowest during the night. The other two cases, both with thermal inertia in the walls, performed similarly to each other, producing lower temperatures than the BC during the day (up to 1ªC less), and higher temperatures during the night (up to 0.5ºC higher). This means that only with thermal inertia the courtyard performed the way it usually does (tempering peak temperatures and slightly overheating during the night), although this courtyard did not produce a very good performance compared to others. MRT was highly affected by thermal inertia. The pattern was similar to air temperature, but up to 2.5ºC degrees difference. This was expected to influence thermal comfort. The UTCI is the parameter that clearly showed the benefits of the courtyard. Outside the building, the UTCI reached 42ºC, while inside the courtyard the maximum UTCI was 35ºC for the BC. Between the three configurations, the lower the thermal inertial, the higher the UTCI during the day, and the opposite during the night. The UTCI was the parameter that showed a larger difference between the CS9-1 and CS9-2 during the daytime, the highest inertia showing the best results. Chapter 3: Simulation analysis to improve the thermal performance of courtyards 81 a) Air temperature b) MRT c) UTCI Figure 3.5-2. Closed courtyard simulation results at 1.5 m above the ground. Adapted from Publication 5. Figure 3.5-3 shows the simulation results of air temperature and UTCI when the courtyard was opened to the outdoors to allow natural ventilation. It is noted that in this case, differences in air temperature between cases were negligible, only a few tenths of a degree. This temperature was close to the outdoor temperature, which is expected given that the outdoor air is coming into the courtyard continually. In terms of thermal comfort, the UTCI differences were slightly lower than in the closed courtyard simulation, with a similar tendency, and differences between CS9-1 and CS9-2, in this case, were reduced. a) Air temperature b) UTCI Figure 3.5-3. Closed courtyard simulation results at 1.5 m above the ground. Adapted from Publication 5. Simulation results showed that in general, the thermal inertia benefits the courtyard performance during the day, and slightly harms it during the night. Furthermore, opening the Workflow Design for the Analysis of Transitional Space Microclimates 82 courtyard reduces its tempering effect, although, in terms of thermal comfort, the results are not so straightforward. They depend on many variables, not only air temperature or mean radiant temperature, but also relative humidity and wind speed. Wind speed is highly affected by the opening state of the courtyard. Figure 3.5-4 shows the differences in wind speed between the opened courtyard and closed courtyard at 16:00 hours. While in the closed courtyard the wind speed was very low, in the opened courtyard it reached up to 2 m/s. The wind speed that a courtyard can achieve through natural ventilation depends on many factors such as openings distribution or outdoor wind speed. Furthermore, although higher wind speed can benefit thermal comfort, at high temperatures it can be detrimental [122]. In conclusion, predicting the impact of opening a courtyard on thermal comfort is difficult for each specific case. Closed courtyard Ventilated courtyard Figure 3.5-4. Wind flow pattern in a cross-section of the courtyard in the wind direction at 16.00 hours. Extracted from Publication 5. 3.5.3 General recommendations. General recommendations for courtyard spaces in this specific climate are that higher thermal inertia and natural ventilation are strategies that improve the thermal performance of the courtyard, provided that the natural ventilation is high enough and activated when necessary. To know when is beneficial, Figure 3.5-5 has been constructed to inform decisions about that. The graphs show UTCI inside any courtyard according to outdoor temperature and the thermal Chapter 3: Simulation analysis to improve the thermal performance of courtyards 83 gap that the courtyard achieves when it is closed (dashed green lines), and the UTCI when the courtyard is ventilated at different wind speeds (yellow line), assuming, in this case, no thermal gap, as seen in the previous analysis. To use the figure, for a specific wind speed that can be reached inside the courtyard, one graph is selected. Then, at a given outdoor temperature and a thermal gap that the courtyard can achieve, the comparison of the dotted line of the specific thermal gap and the continuous line provides two UTCI values, the lower one indicating the option (closed or opened) that perform the best in terms of thermal comfort. These abacuses could be used together with temperature and wind speed sensors inside and outside the courtyard to select the best courtyard configuration at any time. If this is not possible, the general recommendation would be to keep the courtyard closed during the day and open it during the night. This would take maximum advantage of the thermal inertia and performance of the courtyard. Wind speed= 2 m/s Wind speed= 4 m/s Wind speed= 6 m/s Wind speed= 8 m/s Figure 3.5-5. Courtyard UTCI for different wind speeds, ventilated and non-ventilated variations. Extracted from Publication 5. Workflow Design for the Analysis of Transitional Space Microclimates 90 4.2.3 Indoor adaptive comfort results. The mean adaptive comfort in the house per simulated case is represented in Figure 4.2-4 . Adaptive comfort was very compromised in Cases A, B, and C of the June week (hot and dry period), and only Case C achieved an acceptable mean adaptive comfort. In the August week (warm and humid), the mean adaptive comfort was acceptable in all cases, reaching 100% of the time in Case D. The implementation of the strategies progressively increased comfort in the two seasons. Figure 4.2-5 shows a breakdown of the mean adaptive comfort time per space in the house. In general, adaptive comfort was higher in the lower level of the house, and in the rooms next to the courtyard. In June, the shading strategy is enough to achieve a high adaptive comfort in the lower levels of the house, but only air movement achieved it for all the spaces. Figure 4.2-4. Mean comfort time per strategy in June (Hot-Dry) and August (Warm-Humid) weeks. Extracted from Publication 7. Chapter 4: Analysis of indoor comfort and energy consumption of courtyard buildings 91 June week August week CASE A - BASE CASE CASE B - INCREASED INSULATION CASE C - SHADING DEVICES CASE D - IMPROVED AIR MOVEMENT Figure 4.2-5. Mean adaptive comfort map per strategy. Extracted from Publication 7. Workflow Design for the Analysis of Transitional Space Microclimates 92 4.2.4 Wet Bulb Globe Temperature results. Wet Bulb Globe Temperature results are needed to better understand the thermal comfort in the warm and humid season, given that the adaptive comfort model does not include an input for the humidity, which can be very uncomfortable at high levels. Figure 4.2-6 shows hourly WBGT results for the best and worst performance rooms (the passage and the family space) in Case A and Case D simulations. In June, when the outdoor temperature reached 42ºC and RH was between 35 to 80%, the indoor WBGT reached 33ºC in the Family space in Case A, the highest value in the house. After the application of all the strategies, this value fell to 31ºC. This was a reduction of 2ºC in the WBGT and two categories of stress. The passage space (on the ground floor, connecting the rooms with the courtyard) was always under 30ºC, and the application of the strategies produced a lower improvement. In August, outdoor temperatures were lower than in August, up to 35ºC, but RH was much higher, between 60 to 98%. With these conditions, the family space reached 31ºC of WBGT, and only after the application of the strategies this value could be reduced to under 30ºC. In the other space, where WBGT did not go over 28ºC, the improvement was reduced. June week August week Figure 4.2-6. Wet Bulb Globe Temperature results for Case A and Case D in the best and the worst performance rooms of the house. Extracted from Publication 7. WBGT is divided into five categories of heat stress: no heat risk (category 0), moderate heat risk (categories 1 and 2), high risk (categories 3 and 4), and extreme heat risk (category 5). Figure 4.2-7 shows the percentage of time under each category for the previously analyzed spaces. The air movement strategy increased the percentage of time in the lower categories in all cases. This was especially relevant in the family space, the one with the highest percentage of time in the highest categories. It is also interesting to notice that considering all rooms, there is always a space that is comfortable in Case D. Chapter 4: Analysis of indoor comfort and energy consumption of courtyard buildings 93 Figure 4.2-7. Percentage of time at each category of WBGT in the June and August weeks. Extracted from Publication 7. 4.2.5 Outdoor comfort results. This section analyses outdoor comfort in the courtyard, using the UTCI index, for cases A and D, that is before and after the installation of the shading device, which is the strategy that affected outdoor thermal comfort the most. The UTCI index was calculated for the hottest day of each week. Figure 4.2-8 shows the UTCI results at 2 pm in a vertical section of the courtyard. It can be seen how the uncovered courtyard (Case A) reached up to 45 ºC except for the covered areas under the gallery, which ranged from 41ºC to 43ºC. The installation of the shading canvas had a positive effect on the courtyard, reducing the UTCI 5ºC in the courtyard. UTCI values in August were lower than in June, although shading also had a very positive effect this season. Figure 4.2-9 shows the hourly air temperature and UTCI values inside the courtyard, and the outdoor air temperature and humidity. The tempering effect of this courtyard is limited in Case A, but the installation of the shading device increased the thermal difference with the outdoor in the hottest hours around 1 ºC. UTCI values were higher than the air temperature, as UTCI also includes MRT, RH, and wind speed in their calculations, and MRT in this case is high, with direct solar radiation hitting the courtyard. However, in the UTCI values, the effect of the shading is much more important, reaching a difference of around 4ºC. Note the difference between air temperature and UTCI in Case D in each season, being much higher in August, an effect of the high relative humidity of that season. Workflow Design for the Analysis of Transitional Space Microclimates 94 June 20th Aug 27th Case A (Non-shaded Courtyard) Case D (Shaded courtyard) Figure 4.2-8. UTCI results in the courtyard for the two selected days at 2 p.m. Extracted from Publication 7. Figure 4.2-9. UTCI and Dry bulb temperature results in the courtyard for the two selected days. Extracted from Publication 7. Chapter 4: Analysis of indoor comfort and energy consumption of courtyard buildings 95 The analysis of the wind speed inside the courtyard showed that in this narrow space, the wind speed was very low, especially at ground level. This penalized the possible positive effect that the wind speed could produce. Wind speed inside the courtyard could also be increased by forcing air movement with fans. This effect is analyzed in Figure 4.2-10 , using 1 m/s and 2 m/s airspeeds. It showed that the forced wind speed was not very effective in the hottest hours of the day in June, but it could reduce UTCI up to 2ºC in the warmest hours in August. During the night, it was effective in both seasons. This agrees with what has been described in previous literature that noted that wind speed is only beneficial for the UTCI at a certain range of temperatures and humidity values. Figure 4.2-10. UTCI results in the courtyard for the two selected days with increased wind speed. Extracted from Publication 7. 4.2.6 Discussion. This section analyzes the improvement of indoor and outdoor thermal comfort of a case study with a courtyard through different strategies. The analysis is performed with the novel workflow designed using the Ladybug Tools. The thermal benefits of the courtyard are manifested by the relatively higher thermal comfort produced in the rooms close to it. In addition, the performance of the courtyard and the indoor rooms has been improved by installing a shading device and enhanced air movement strategies. This workflow has one limitation, which is the lack of information going back from the CFD analysis to the energy analysis (the stream is onedirectional). This is due to the lack of a Honeybee component to modify the boundary conditions of the walls. For that reason, the benefits of the courtyard are underestimated given that is not possible to include the tempering effect of the courtyard in the energy simulation, and only shading effects are considered. Workflow Design for the Analysis of Transitional Space Microclimates 96 4.3 Coupling the proposed workflow with an energy modeling tool. 4.3.1 Introduction and objectives. In Section 4.1 the problem of energy tools not being able to simulate outdoor microclimates was introduced. This section analyses the possibilities of coupling the proposed LBT method with an existing energy simulation tool: HULC. It is the Spanish computer software required to comply with existing energy certification requirements. The HULC tool runs with an hourly time step in a transitory state and includes the additional capability (CA) to modify the boundary conditions of one or more of the elements of the building envelope. However, this software does not have a plugin to calculate specific microclimates, it is only able to incorporate externally calculated or monitored data. This means that this additional capability is in practice not very useful, given that outdoor microclimates require CFD simulation tools not always known by users. This problem is solved by introducing the microclimate effects of courtyards predicted with the proposed workflow with LBT. Another main difference between BES and CFD tools is the speed of simulation. While BES can simulate long periods in a few minutes, CFD requires hours to simulate just one or two days. Although the methodology designed in this thesis is faster than other CFD tools, there is still a gap in the simulation time required by HULC and the designed workflow. In order to speed up the simulation time of the CFD, it is proposed to simulate just two hours each day, (the maximum and the minimum outdoor temperatures outdoors) and interpolate the rest of the hours. In this way, it would be easier to calculate longer periods with the coupled workflow, up to a month, in a reasonable time. This section explains the coupling process of LBT and HULC in the Grasshopper interface and presents the results of its application in one case study in Cordoba. An objective evaluation of the potential of the use of the tool by practitioners is presented. 4.3.2 Coupling description. The coupling of the two tools was designed with the idea of reducing as much as possible the need for changing from one software interface to another, being able to do most of the simulation directly from the Grasshopper interface. To achieve that, several blocks of code in Python components in Grasshopper had to be written. The most complex part was the communication between the two software in relation to the geometry of the case study, which had to be modeled in Rhinoceros and exported to HULC. The workflow structure is shown in Figure 4.3-1 . Chapter 4: Analysis of indoor comfort and energy consumption of courtyard buildings 97 Figure 4.3-1. Coupling workflow for CFD and BES tools using LBT and HULC. Despite the number of steps shown, most of them are controlled in the Grasshopper interface, thus the process is quite easygoing. The only steps that happen outside that interface are those marked as “HULC” in the process of the figure. The process goes as follows: • The geometry of the case study is modeled using the Rhinoceros CAD software and imported to Grasshopper (GH) ( Figure 4.3-2 ). It is also possible to directly model using GH, especially for parametric analysis. Solids or polysurfaces are supported, representing the spaces in a simplified way, following the rules of LBT analysis. Figure 4.3-2. Rhinoceros and Grasshopper interface. Import of geometry into Grasshopper. • The first step before the simulation of the microclimate of the courtyard is the analysis of the weather file. This is done to find maximum and minimum outdoor temperatures and when they happen. Workflow Design for the Analysis of Transitional Space Microclimates 98 • The CFD simulation of the microclimate of the courtyard using the novel workflow in LBT designed in this thesis is run for the maximum and minimum outdoor temperatures found in the previous step. • A Python code is written to interpolate the results of the temperature in the courtyard at the maximum and minimum temperatures between the different hours of the day. The BES tools (in this case HULC) need averaged hourly data. • The HULC CA files require one data per hour and surface of the building. A code in Python extracts that information from the CFD simulations (which have multiple points per surface) and stores the information in the appropriate format to copy it later in the CA files. • The most complex part of the workflow is the export of the geometry in Rhinoceros to HULC. Each software has its language to define the model, thus, a Python code was created to read the geometry in Rhinoceros and translate it into HULC language. At this moment, the code only exports geometry. The material properties of the constructions still need to be defined in the HULC interface. However, this is something that could be added in the future. • In HULC ( Figure 4.3-3 ), the building construction, openings, systems, and operating characteristics should be defined, in addition to the location of the case study. Figure 4.3-3. Completed model in HULC. • In HULC, CA files are generated. Each of the surfaces of the building that is going to have their boundary conditions changed is selected and an Initial Conditions File (ICF) is generated. This file contains information about the surface, including one column for outdoor temperature that is the one to be altered. Chapter 4: Analysis of indoor comfort and energy consumption of courtyard buildings 99 • Once generated, the ICF files need to be modified with the temperature data simulated with the CFD simulation and already stored in GH. This is automatically done through a Python piece of code that creates the modified files (MCF) ( Figure 4.3-4 ). Figure 4.3-4. Modified Condition Files created using Python code in Grasshopper interface. • The MCF files are imported into HULC, and the energy simulation is run. • All the information generated by HULC can be imported into the GH interface using Python code to be represented and analyzed using the LBT. 4.3.3 Discussion. The process is still a proof of concept that is being refined and tested in some case studies, and the results will be presented in future publications. The main contribution of this thesis is the connection between the tools, leaving the application to future work. This workflow allows the practical use of the additional capacities tool in HULC for the introduction of the courtyard performance in the energy simulation. Otherwise, this capability of the software is rather limited. This is the reason why this thesis first focused on the connection of this tool which is provided by the Spanish government to perform the energy certification of buildings. In the future, the linkage of the novel LBT methodology developed in this thesis with other BES tools will be addressed, especially with the Honeybee plugin in LBT, which allows the whole process in the same interface, which is also already widely used by the scientific community. Another aspect that needs to be addressed is finding a solution for the different simulation times required by each tool. The interpolation used in this section is one possible solution, but it affects accuracy. Even faster CFD simulation or the use of other methods are some possibilities that need to be analyzed. Workflow Design for the Analysis of Transitional Space Microclimates 106 Regarding professional use. The last analysis is made considering the professional usability of these tools. In this sense, three criteria are considered: accuracy, computational resources, and ease of use and access. • Accuracy: It has already been said that LBT performed more accurately than ENVI-met. However, this does not discard ENVI-met because it is the only one currently that can simulate the evapotranspiration effect of vegetation. This is not included in LBT yet, although it is a tool continuously under development and being open source, this issue may be overcome soon. • Computational resources: Simulation time in CFD tools is usually long, and it depends on many factors such as computational resources and model configurations. Both methodologies required similar simulation time to provide data for a whole day. However, LBT has one advantage: it is a steady-state solver which does not need initialization time to get accurate results, so it can calculate specific hours (for example peak hours), not needing to perform the whole day, and thus reducing the simulation time. This makes it more suitable for the early design stage of projects. • Ease of use and access: Considering that the LBT links the simulation software to the model design software, this tool is highly recommended for the early design stage, showing the results directly on the model. The fact that it is open-access makes it also available to everyone, in contrast to ENVI-met, which is a commercial tool needing a paid license. On the other hand, CFD modeling requires more knowledge in LBT than ENVI- met, to understand all the options provided and the errors that may occur. Further considerations. The conclusions and results derived from this work must be understood within the limits of the analysis carried out. First, the weather conditions in this study are limited to hot summer conditions, both dry and humid, and with low wind speeds. To generalize the results to the whole year or to other climates, further research is needed. Second, there are a limited number of case studies. Although chosen to be representative of the most common building typologies [79], different urban configurations or building uses can produce different results. The same methodology presented in this thesis can be used in any climate or building typology and is the main contribution of this work. Another important aspect is that, although the methodology has been applied to the study of courtyard buildings, which are the focus of this PhD, it has the potential to be used to analyze outdoor spaces, which means that the same methodology can be used to analyze urban microclimates. Considering the emergency situation in this field, this option is also a valuable contribution that needs further consideration in the future. Chapter 5: Discussion and conclusion. Recommendations 107 5.2 Novel research contributions The main contributions of this thesis in the field of building microclimate analysis and specifically, the simulation of courtyard spaces are summarized in this section. Chapter 1 identified the existing simulation tools and approaches applicable to the analysis of outdoor spaces, in response to need 1, through a critical analysis of the state of the art. BES, CFD, hybrid, and ML methods were discussed. The most promising tools were identified. In Chapter 2, the desired characteristics of a simulation procedure for the microclimate of courtyards in the design stage of a project were defined in response to need 2. This led to the selection of the ENVI-met and the LBT to be adapted and tested for the simulation of courtyards. A novel simulation workflow using LBT was designed in response to need 3. The main novelty of this workflow is the link between a BES tool and a CFD tool in LBT capable of simulating air temperature variations in a courtyard. To calibrate and validate the simulations, experimental campaigns were developed during the hot dry period in different case studies, and monitored data were compared with simulation data, using statistical variables, in response to needs 4 and 5. It was found that the proposed simulation workflow has some advantages for the early design in comparison to ENVI-met, and its accuracy was higher. However, the ENVI-met tool is still needed in some cases when vegetation or water is present, given that LBT still does not include the simulation of the evapotranspiration effect. In Chapter 3, the proposed methodologies are applied to the numerical quantification of the benefits achieved by different strategies to improve the thermodynamic performance of existing courtyards, in response to need 6. The effects of geometry, albedo and vegetation, shading elements, and thermal inertia coupled with ventilation were studied. Furthermore, the influence of these strategies on the user's thermal comfort was evaluated, in response to need 7. General recommendations for the design of improvement strategies in existing courtyards were given, although the main contribution of this thesis is the methodology applied, which can be adapted to different case studies and climates. Chapter 4 proposed two methods to apply the novel workflow to the analysis of indoor thermal comfort and energy demand of buildings with courtyards, in response to needs 8 and 9. In this way, it is possible to numerically analyze the use of courtyards as a passive strategy to improve building performance. All required tools are open-access and available to designers, contributing to the dissemination of the findings of this work and the improvement of the field of energy certification and sustainable building design in Spain. Finally, Chapter 5 summarized the recommendations provided for the design of courtyard spaces to optimize their performance. The publication of this work aims to raise awareness about the need to achieve better microclimate simulations, in response to needs 10 and 11. Workflow Design for the Analysis of Transitional Space Microclimates 108 5.3 Main research conclusions This thesis proposes a novel simulation methodology applied to the microclimate of courtyards to understand their thermodynamic performance considering the different factors involved. This methodology was applied to the analysis of different strategies to improve the performance of the courtyard as a passive element to reduce the energy consumption of buildings and increase user comfort. The main conclusions of this work concerning the specific objectives initially proposed are summarized below: In relation to O1: “Identification of the existing approaches to the simulation of outdoor microclimates and their applicability to the specific case of inner courtyards”, many different approaches were analyzed, from very established ones (BES and CFD) to new ones (hybrid or ML methods) with a very promising future. Two tools were selected to be applied in this thesis: ENVI-met as an existing and already widely used tool for urban microclimates that was not validated enough for its application on courtyards, and LBT, a new hybrid tool that needed a new script design to implement CFD simulation with heat transfer solvers in courtyards. In relation to O2: “Development of a new simulation methodology for the analysis of the varied factors that affect the microclimate in courtyards of the Mediterranean climate and their implications on the outdoor comfort and energy efficiency of buildings”, the methodology using LBT proposed was validated and meets the necessities of the early design stage of projects with some advantages over other existing tools like EVI-met, such as accuracy, flexibility in the simulation and speed. However, ENVI-met was still necessary if vegetation or water ponds exist in the simulation study, given that it is the only software capable of incorporating their effects on the simulations. This software was also linked to LBT to manage all the tools in one single interface. The validation analysis of both tools provided an RMSE between 1.52ºC and 3.40ºC for ENVI-met and between 1.37ºC and 2.29ºC for LBT, both considered validated for the analysis of courtyards. In relation to O3: “Application of the designed methodology to the evaluation of the implementation of different strategies to improve the performance of courtyards and their potential to increase user comfort in existing buildings”, the effectiveness of many strategies to improve courtyard performance was numerically demonstrated. The inclusion of shading devices had the largest influence on air temperature (monitored up to 13ºC of thermal delta). Albedo, in contrast, had less effect on the air temperature (around 0.2ºC TD) but had an important influence on user comfort (lowering up to 5ºC in PET index if optimized), mainly for their implication in surface temperature and thus mean radiant temperature. Thermal inertia and night ventilation were found to improve the courtyard's performance up to 1.6ºC and 1.4ºC UTCI. The proposed methodology using LBT and ENVI-met can be used to analyze all these strategies in the courtyard but also the outdoor space. Chapter 5: Discussion and conclusion. Recommendations 109 In relation to O4: “Application of the designed methodology to the comprehensive evaluation of the thermal performance of a building, focusing on user comfort, with special emphasis on the user experience when using the proposed tools”, a methodology to increase indoor and outdoor comfort in a house with a courtyard was applied using the LBT methodology. It was found that using some low-carbon strategies such as shadings and air movement using fans, indoor adaptive comfort increased from 39 % of the time to up to 90- 100%. In the courtyard, the use of a shading device reduced air temperature by 1ºC and UTCI up to 5ºC. The different characteristics of BES and CFD simulation approaches make them difficult to couple for a long simulation. To analyze the influence of the courtyard over a longer period than a few hours, the coupling of LBT and HULC included an interpolation code to extract hourly data from two steady-state simulations using the LBT novel workflow and use the results as boundary conditions for the surfaces of the courtyard in HULC. This way, up to a month could be analyzed in a reasonable simulation time. In relation to O5: “Definition of recommendations for the intervention in courtyards of existing buildings to improve energy efficiency and comfort and result dissemination” recommendations for the improvement of the microclimate of courtyards and user comfort were given, and a simulation methodology is available for designers. The results of this thesis have been published in several high-impact journals, fulfilling the dissemination objective. The tools designed have been uploaded to a public repository to allow their use by anyone interested. This thesis aims to contribute to the field of simulation of buildings and cities considering strategies to mitigate and adapt to rising temperatures in the expected future of climate change, providing the tools that architects, designers, and urban planners need to approach the challenge, and recommendations to understand urban thermodynamics. Workflow Design for the Analysis of Transitional Space Microclimates 110 5.4 Future work Despite these research contributions, there is room for further research and improvement in the field of microclimate simulations for the analysis of their impact on building performance. Future research in the field should focus on some challenges that have not yet been overcome: • The analysis of the evapotranspiration effect of vegetation and water is limited to very few software tools such as ENVI-met. Therefore, most of the energy simulation or CFD tools do not consider the possible benefits of vegetation or water, even though experimental research has found that they can be effective passive strategies against urban dry heat. For that reason, the addition of this capability to other tools, specifically in the workflow designed in this thesis is an issue that needs to be addressed in order to improve the analysis of urban heat mitigation strategies. • The simulation methodologies analyzed in this work can be further optimized. Although the results have been numerically validated using statistical parameters, the accuracy in some situations can be improved (shading devices, deep courtyards, etc). Furthermore, computational time could be reduced. For the specific case of CFD methods, there is much research that can be applied to the development of new CFD algorithms. As an example, reduced-order modeling (ROM) is a strategy that provides reductions of several orders of magnitude in the computational cost of numerical simulation of parametric design processes and problems [131]. This is an option to reduce computing time that is under research in the scope of the project RTI2018-093521-B-C33 of our research group. • The methodology applied in this work to courtyards can be used in the analysis of other urban spaces affecting buildings. To do that, it needs further studies to validate it in other applications, different weather, and building typologies. • Despite the tools provided here, their accessibility to the professional sector still needs development. Some of them require advanced knowledge or require a large expenditure to be acquired. This is a problem that needs to be addressed to transfer research knowledge into society. References 111 References [1] IPCC, Annex I: Glossary, in: J.B.. Matthews (Ed.), Global Warming of 1.5°C, Cambridge University Press, Cambridge, 2018: pp. 541–562. https://doi.org/10.1017/9781009157940.008. [2] IPCC, Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, 2021. [3] M. Santamouris, N. Papanikolaou, I. Livada, I. Koronakis, C. Georgakis, A. Argiriou, D.N. Assimakopoulos, On the impact of urban climate on the energy consumption of building, Solar Energy. 70 (2001) 201–216. https://doi.org/10.1016/S0038- 092X(00)00095-5. [4] In focus: Energy efficiency in buildings | European Commission, Brussels, 2020. https://ec.europa.eu/info/news/focus-energy-efficiency-buildings-2020-feb-17_en (accessed June 30, 2020). [5] European Parliament, Directive (EU) 2018/844 on the energy performance of buildings, 2018. [6] U. Nations, Sustainable Development Goals: Sustainable Development Knowledge Platform, Web Page. (n.d.). https://sustainabledevelopment.un.org/?menu=1300 (accessed July 9, 2019). [7] T. Huang, J. Li, Y. Xie, J. Niu, C.M. Mak, Simultaneous environmental parameter monitoring and human subject survey regarding outdoor thermal comfort and its modelling, Building and Environment. 125 (2017) 502–514. https://doi.org/10.1016/j.buildenv.2017.09.015. [8] N.A. Al-Hemiddi, K.A. Megren Al-Saud, The effect of a ventilated interior courtyard on the thermal performance of a house in a hot–arid region, Renewable Energy. 24 (2001) 581–595. https://doi.org/10.1016/S0960-1481(01)00045-3. [9] M.A. Cantón, C. Ganem, G. Barea, J.F. Llano, Courtyards as a passive strategy in semi dry areas. Assessment of summer energy and thermal conditions in a refurbished school Workflow Design for the Analysis of Transitional Space Microclimates 112 building, Renewable Energy. 69 (2014) 437–446. https://doi.org/10.1016/j.renene.2014.03.065. [10] T. Kubota, M.A. Zakaria, S. Abe, D.H.C. Toe, Thermal functions of internal courtyards in traditional Chinese shophouses in the hot-humid climate of Malaysia, Building and Environment. 112 (2017) 115–131. https://doi.org/10.1016/j.buildenv.2016.11.005. [11] M. Taleghani, M. Tenpierik, A. van den Dobbelsteen, D.J. Sailor, Heat in courtyards: A validated and calibrated parametric study of heat mitigation strategies for urban courtyards in the Netherlands, Solar Energy. 103 (2014) 108–124. https://doi.org/10.1016/j.solener.2014.01.033. [12] H.A. Abdulkareem, Thermal Comfort through the Microclimates of the Courtyard. A Critical Review of the Middle-eastern Courtyard House as a Climatic Response, Procedia - Social and Behavioral Sciences. 216 (2016) 662–674. https://doi.org/10.1016/j.sbspro.2015.12.054. [13] A. Ghaffarianhoseini, U. Berardi, A. Ghaffarianhoseini, Thermal performance characteristics of unshaded courtyards in hot and humid climates, Building and Environment. 87 (2015) 154–168. https://doi.org/10.1016/j.buildenv.2015.02.001. [14] J.M. Rojas, C. Galán-Marín, E.D. Fernández-Nieto, Parametric study of thermodynamics in the mediterranean courtyard as a tool for the design of eco-efficient buildings, Energies. 5 (2012) 2381–2403. https://doi.org/10.3390/en5072381. [15] D. Lai, C. Zhou, J. Huang, Y. Jiang, Outdoor space quality : A field study in an urban residential community in central China, Energy & Buildings. 68 (2014) 713–720. https://doi.org/10.1016/j.enbuild.2013.02.051. [16] Y. Li, Z. O’Neill, L. Zhang, J. Chen, P. Im, J. DeGraw, Grey-box modeling and application for building energy simulations - A critical review, Renewable and Sustainable Energy Reviews. 146 (2021) 111174. https://doi.org/10.1016/j.rser.2021.111174. [17] D. Coakley, P. Raftery, M. Keane, A review of methods to match building energy simulation models to measured data, Renewable and Sustainable Energy Reviews. 37 (2014) 123–141. https://doi.org/10.1016/j.rser.2014.05.007. [18] M.-A. Hamdaoui, M.-H. Benzaama, Y. El Mendili, D. Chateigner, A review on physical and data-driven modeling of buildings hygrothermal behavior: Models, approaches and simulation tools, Energy and Buildings. 251 (2021) 111343. https://doi.org/10.1016/j.enbuild.2021.111343. [19] J.H. Choi, Investigation of the correlation of building energy use intensity estimated by six building performance simulation tools, Energy and Buildings. 147 (2017) 14–26. https://doi.org/10.1016/j.enbuild.2017.04.078. [20] O.S. Asfour, A comparison between the daylighting and energy performance of courtyard and atrium buildings considering the hot climate of Saudi Arabia, Journal of Building Engineering. 30 (2020) 101299. https://doi.org/10.1016/J.JOBE.2020.101299. [21] F. Soflaei, M. Shokouhian, H. Abraveshdar, A. Alipour, The impact of courtyard design variants on shading performance in hotarid climates of Iran, Energy and Buildings. (2017). https://doi.org/10.1016/j.enbuild.2017.03.027. [22] A. M’Saouri El Bat, Z. Romani, E. Bozonnet, A. Draoui, Thermal impact of street canyon microclimate on building energy needs using TRNSYS: A case study of the city of Tangier in Morocco, Case Studies in Thermal Engineering. 24 (2021) 100834. https://doi.org/10.1016/j.csite.2020.100834. [23] E. Walter, J.H. Kämpf, A verification of CitySim results using the BESTEST and References 113 monitored consumption values, Second Building Simulation Applications Conference. (2015) 215–222. https://infoscience.epfl.ch/record/214754/files/p%28139%29v2.pdf%0Ahttps://infosci ence.epfl.ch/record/214754 (accessed June 17, 2020). [24] Y. Toparlar, B. Blocken, B. Maiheu, G.J.F. van Heijst, A review on the CFD analysis of urban microclimate, Renewable and Sustainable Energy Reviews. 80 (2017) 1613– 1640. https://doi.org/10.1016/j.rser.2017.05.248. [25] P. Kastner, T. Dogan, A cylindrical meshing methodology for annual urban computational fluid dynamics simulations, Journal of Building Performance Simulation. 13 (2020) 59–68. https://doi.org/10.1080/19401493.2019.1692906. [26] B. Blocken, J. Carmeliet, T. Stathopoulos, CFD evaluation of wind speed conditions in passages between parallel buildings-effect of wall-function roughness modifications for the atmospheric boundary layer flow, Journal of Wind Engineering and Industrial Aerodynamics. 95 (2007) 941–962. https://doi.org/10.1016/j.jweia.2007.01.013. [27] J. Rojas-Fernández, C. Galán-Marín, C. Rivera-Gómez, E.D. Fernández-Nieto, Exploring the interplay between CAD and FreeFem++ as an energy decision-making tool for architectural design, Energies. 11 (2018) 1–19. https://doi.org/10.3390/en11102665. [28] M.A. Padilla-Marcos, J. Feijó-Muñoz, A. Meiss, Eficiencia isoterma de los modelos de ventilación exterior en patios de edificios residenciales. Estudio de casos, Informes de La Construcción. 67 (2015) e121. https://doi.org/10.3989/ic.15.029. [29] DesignBuilder, DesignBuilder Software Ltd - CFD, (2021). https://designbuilder.co.uk/cfd (accessed October 22, 2021). [30] E. Naboni, M. Meloni, C. Mackey, J. Kaempf, The Simulation of Mean Radiant Temperature in Outdoor Conditions: A review of Software Tools Capabilities, Building Simulation Conference Proceedings. (2019). https://doi.org/10.26868/25222708.2019.210301. [31] R. Willis, Validation of CFD Predictions of Urban Wind - Developing a Methodology, Victoria University of Wellington, 2018. http://researcharchive.vuw.ac.nz/handle/10063/7086 (accessed October 22, 2021). [32] D. Mauree, E. Naboni, S. Coccolo, A.T.D. Perera, V.M. Nik, J.-L. Scartezzini, A review of assessment methods for the urban environment and its energy sustainability to guarantee climate adaptation of future cities, Renewable and Sustainable Energy Reviews. 112 (2019) 733–746. https://doi.org/10.1016/J.RSER.2019.06.005. [33] C.K.C. Lam, H. Lee, S.-R. Yang, S. Park, A review on the significance and perspective of the numerical simulations of outdoor thermal environment, Sustainable Cities and Society. 71 (2021) 102971. https://doi.org/10.1016/j.scs.2021.102971. [34] Complete Guide - ENVI-met, (n.d.). https://www.envi-met.com/learningsupport/unfolded-the-complete-guide/ (accessed July 21, 2021). [35] A. Forouzandeh, Numerical modeling validation for the microclimate thermal condition of semi-closed courtyard spaces between buildings, Sustainable Cities and Society. 36 (2018) 327–345. https://doi.org/10.1016/j.scs.2017.07.025. [36] S. Huttner, Further development and application of the 3D microclimate simulation ENVI-met, 2012. http://ubm.opus.hbz-nrw.de/volltexte/2012/3112/. [37] E. Erell, T. Williamson, Simulating air temperature in an urban street canyon in all weather conditions using measured data at a reference meteorological station, International Journal of Climatology. 26 (2006) 1671–1694. Workflow Design for the Analysis of Transitional Space Microclimates 114 https://doi.org/10.1002/joc.1328. [38] D. Mauree, S. Coccolo, J. Kampf, J.-L. Scartezzini, Multi-Scale Modelling to Assess Human Comfort in Urban Canyons, in: G. Habert, A. Schlueter (Eds.), Expanding Boundaries, vdf Hochschulverlag AG an der ETH Zürich, Zurich, 2016. https://doi.org/10.3218/3774-6_15. [39] X. Yang, L. Zhao, M. Bruse, Q. Meng, An integrated simulation method for building energy performance assessment in urban environments, Energy and Buildings. 54 (2012) 243–251. https://doi.org/10.1016/j.enbuild.2012.07.042. [40] J. Natanian, T. Auer, Beyond nearly zero energy urban design: A holistic microclimatic energy and environmental quality evaluation workflow, Sustainable Cities and Society. 56 (2020). https://doi.org/10.1016/j.scs.2020.102094. [41] K. Perini, A. Chokhachian, S. Dong, T. Auer, Modeling and simulating urban outdoor comfort: Coupling ENVI-Met and TRNSYS by grasshopper, Energy and Buildings. 152 (2017) 373–384. https://doi.org/10.1016/j.enbuild.2017.07.061. [42] Ladybug Tools | Home Page, (n.d.). https://www.ladybug.tools/ (accessed September 4, 2020). [43] D.B. Crawley, L.K. Lawrie, F.C. Winkelmann, W.F. Buhl, Y.J. Huang, C.O. Pedersen, R.K. Strand, R.J. Liesen, D.E. Fisher, M.J. Witte, J. Glazer, EnergyPlus: Creating a newgeneration building energy simulation program, Energy and Buildings. 33 (2001) 319– 331. https://doi.org/10.1016/S0378-7788(00)00114-6. [44] TRNSYS : Transient System Simulation Tool, (n.d.). http://www.trnsys.com/index.html (accessed July 2, 2020). [45] ENVI-met, (n.d.). https://www.envi-met.com/ (accessed October 28, 2019). [46] OpenFOAM User Guide, (n.d.). https://www.openfoam.com/documentation/user-guide/ (accessed January 22, 2020). [47] FreeFEM, (n.d.). https://freefem.org/ (accessed December 8, 2022). [48] Autodesk, Overview | CFD | Autodesk Knowledge Network, (2021). https://knowledge.autodesk.com/support/cfd (accessed October 22, 2021). [49] B. Bueno, L. Norford, J. Hidalgo, G. Pigeon, The urban weather generator, Journal of Building Performance Simulation. 6 (2013) 269–281. https://doi.org/10.1080/19401493.2012.718797. [50] F. Lindberg, B. Holmer, S. Thorsson, SOLWEIG 1.0 – Modelling spatial variations of 3D radiant fluxes and mean radiant temperature in complex urban settings, International Journal of Biometeorology. 52 (2008) 697–713. https://doi.org/10.1007/s00484- 008-0162-7. [51] C. Mackey, T. Galanos, L. Norford, M.S. Roudsari, Wind , Sun , Surface Temperature , and Heat Island : Critical Variables for High-Resolution Outdoor Thermal Comfort, in: Proceedings of the 15th IBPSA Conference San Francisco, CA, USA, San Francisco, CA, 2017: pp. 985–993. https://doi.org/10.26868/25222708.2017.260. [52] I. Elwy, Y. Ibrahim, M. Fahmy, M. Mahdy, Outdoor microclimatic validation for hybrid simulation workflow in hot arid climates against ENVI-met and field measurements, in: Energy Procedia, 2018. https://doi.org/10.1016/j.egypro.2018.10.009. [53] F. Soflaei, M. Shokouhian, A. Tabadkani, H. Moslehi, U. Berardi, A simulation-based model for courtyard housing design based on adaptive thermal comfort, Journal of Building Engineering. 31 (2020) 101335. References 115 https://doi.org/10.1016/j.jobe.2020.101335. [54] G. Evola, V. Costanzo, C. Magrì, G. Margani, L. Marletta, E. Naboni, A novel comprehensive workflow for modelling outdoor thermal comfort and energy demand in urban canyons: Results and critical issues, Energy and Buildings. 216 (2020) 109946. https://doi.org/10.1016/j.enbuild.2020.109946. [55] C. Reinhart, T. Dogan, A. Jakubiec, T. Rakha, A. Sang, Umi – An Urban Simulation Environment For Building Energy Use, Daylighting And Walkability, in: 13th Conference of International Building Performance Simulation Association, Chambery, 2013: pp. 476–483. https://doi.org/10.26868/25222708.2013.1404. [56] K. Fabbri, A. Di Nunzio, J. Gaspari, E. Antonini, A. Boeri, Outdoor Comfort: The ENVIBUG tool to Evaluate PMV Values Output Comfort Point by Point, Energy Procedia. 111 (2017) 510–519. https://doi.org/10.1016/j.egypro.2017.03.213. [57] F. Binarti, M.D. Koerniawan, S. Triyadi, S.S. Utami, A. Matzarakis, A review of outdoor thermal comfort indices and neutral ranges for hot-humid regions, Urban Climate. 31 (2020) 100531. https://doi.org/10.1016/j.uclim.2019.100531. [58] P.O. Fanger, Thermal Comfort: analysis and applications in environmental engineering, (1970). [59] S. Salimi, E.E. Guillén, H. Samuelson, Exceedance Degree-Hours: A new method for assessing long-term thermal conditions, Indoor Air. (2021). https://doi.org/10.1111/INA.12855. [60] R. Maiti, PMV model is insufficient to capture subjective thermal response from Indians, International Journal of Industrial Ergonomics. 44 (2014) 349–361. https://doi.org/10.1016/J.ERGON.2014.01.005. [61] R.J. de Dear, G.S. Brager, Richard J. de Dear, Gail Schiller Brager, Developing an adaptive model of thermal comfort and preference, 1998. [62] ASHRAE-55, Thermal environmental conditions for human occupancy, ANSI/ASHRAE Standard - 55. 7 (2017) 6. [63] EN 16798-1:2019 - Energy performance of buildings - Ventilation for buildings - Part 1: Indoor, (n.d.). https://standards.iteh.ai/catalog/standards/cen/b4f68755-2204- 4796-854a-56643dfcfe89/en-16798-1-2019 (accessed September 17, 2021). [64] S. Kumar, J. Mathur, S. Mathur, M.K. Singh, V. Loftness, An adaptive approach to define thermal comfort zones on psychrometric chart for naturally ventilated buildings in composite climate of India, (2016). https://doi.org/10.1016/j.buildenv.2016.09.023. [65] P. Kumar, A. Sharma, Study on importance, procedure, and scope of outdoor thermal comfort –A review, Sustainable Cities and Society. 61 (2020) 102297. https://doi.org/10.1016/j.scs.2020.102297. [66] P. Höppe, The physiological equivalent temperature - A universal index for the biometeorological assessment of the thermal environment, International Journal of Biometeorology. 43 (1999) 71–75. https://doi.org/10.1007/s004840050118. [67] T.-P. Lin, S.-R. Yang, Y.-C. Chen, A. Matzarakis, The potential of a modified physiologically equivalent temperature (mPET) based on local thermal comfort perception in hot and humid regions, Theoretical and Applied Climatology 2018 135:3. 135 (2018) 873–876. https://doi.org/10.1007/S00704-018-2419-3. [68] A.P. Gagge, A.P. Fobelets, L.G. Berglund, A standard predictive Index of human reponse to thermal enviroment. - OceanRep, Transactions / American Society of Heating, Workflow Design for the Analysis of Transitional Space Microclimates 122 Table 3.4-1. Quantitative evaluation of the simulation’s results for the courtyard temperature output. ...................................................................................................................... 76 Table 3.5-1. Thermal properties of each simulation variation. Extracted from Publication 5. ................................................................................................................................ 80 Chapter 4: Table 4.2-1. Strategies' constructive and operational description. Extracted from Publication 7. ................................................................................................................................ 89 List of Figures 123 List of Figures Chapter 1: Figure 1.1-1. Thermodynamic effects in courtyards. CFD simulation. a) Stratification. b) Convection. c) Flow patterns from the effect of wind. Adapted from [14]. ......................... 21 Figure 1.2-1. Parameters affecting outdoor thermal comfort. Adapted from [72]. .............. 26 Figure 1.2-2. Andalusian courtyard as an outdoor comfortable space to live in. Cordoba. (Own photo). ...................................................................................................................... 27 Figure 1.3-1. Distribution of courtyard buildings around the world. Adapted from [76]. ...... 30 Figure 1.4-1. Methodology workflow followed in this thesis. ........................................... 32 Figure 1.4-2. The sky view factor is ratio of the sky that can be directly seen from a specific point to the whole hemispheric sky. Courtyard in the Royal Tobacco Factory of Seville. Source: Universidad de Sevilla [78]. ........................................................................................ 33 Figure 1.4-3. Images of the measuring instrument installed in the building. ...................... 34 Chapter 2: Figure 2.2-1. Images of the buildings selected as case studies. ....................................... 43 Figure 2.2-2. ENVI-met model of the courtyards. Top: perspective. Down: Top view. Extracted from Publication 1. .................................................................................................... 44 Figure 2.2-3. Image of the buildings selected as case studies. ......................................... 46 Figure 2.2-4. Flowchart of how the inputs and outputs relate to the simulation tools and validation. ................................................................................................................. 47 Figure 2.2-5. Computational domain in Ladybug Tools models. Extracted from Publication 3. ................................................................................................................................ 47 Workflow Design for the Analysis of Transitional Space Microclimates 124 Figure 2.2-6. Wind profile in Ladybug Tools models for a wind speed of 2 m/s at 10 m height. Extracted from Publication 3. ...................................................................................... 48 Figure 2.2-7. Ladybug Tools models’ grid. Extracted from Publication 3. .......................... 49 Figure 2.2-8. Results of temperature in a vertical line of the model for the grid sensitivity analysis. Extracted from Publication 3. ......................................................................... 49 Figure 2.3-1. Monitoring data of Air Temperature and Wind Speed in CS1, CS2, and CS3 on June 8th. Adapted from Publication 1. ......................................................................... 51 Figure 2.3-2. Monitored air temperature and Thermal gap in CS4, CS5, and CS6. Adapted from Publication 3. .................................................................................................... 52 Figure 2.3-3. Air temperature at 1.5 m above the ground for each courtyard at 16.00h on June 8th. Extracted from Publication 1. ............................................................................... 53 Figure 2.3-4. Monitored and simulated air temperature results from ENVI-met for CS1, CS2, and CS3 on the 8th of June 2017. Adapted from Publication 1....................................... 54 Figure 2.3-5. Simulated and monitored temperature gap in CS1, CS2, and CS3 on the 8th of June 2017. Adapted from Publication 1. ...................................................................... 55 Figure 2.3-6. Air Temperature results at 1.5 above the ground (plan and section) from Envimet and Ladybug Tools workflows at 16:00 hours. Extracted from Publication 3. ............. 57 Figure 2.3-7. Temperature comparison of monitoring, ENVI-met, and Ladybug Tools workflows results. Extracted from Publication 3. ........................................................................... 58 Figure 2.3-8. Ladybug Tools simulation results and monitoring. a) Hourly data (August 2nd, 2018). b). East-west section of the courtyard at different times. Extracted from Publication 2. ................................................................................................................................ 58 Chapter 3: Figure 3.1-1. Thermodynamic effects inside a courtyard. Adapted from [114]. ................. 62 Figure 3.1-2. Example of courtyards with the presence of vegetation and water elements. (Own photos). .................................................................................................................... 63 Figure 3.2-1. Ladybug Tools workflow to obtain comfort results. ..................................... 66 Figure 3.2-2. UTCI results at 1.5 m above the ground for CS4, CS5, and CS6 from Ladybug Tools workflows at different hours of the day. Adapted from Publication 3. ....................... 67 Figure 3.3-1. Albedo analysis methodology workflow using ENVI-met. Extracted from Publication 4. ............................................................................................................ 68 Figure 3.3-2 Simulated scenarios of the courtyard. The colors represent the distribution of materials on the surfaces. Extracted from Publication 4. ................................................. 69 Figure 3.3-3. Air and surface temperature simulation results for the different model configurations. Extracted from Publication 4. ................................................................. 70 Figure 3.3-4. MRT and PET simulation results for the different model configurations. Extracted from Publication 4. .................................................................................................... 71 Figure 3.3-5. Simulation results for the Best-Case configuration. Extracted from Publication 4. ................................................................................................................................ 72 List of Figures 125 Figure 3.4-1. Left: Image of CS5’s courtyard with the shading device installed. Right top: Section of the building. Right down: Plan of the building. Extracted from Publication 6. ..... 74 Figure 3.4-2. Monitored Outdoor and Courtyard Dry Bulb Temperature on the selected days. Extracted from Publication 6. ...................................................................................... 75 Figure 3.4-3. Ladybug simulated courtyard air temperature in the selected days. Extracted from Publication 6. ............................................................................................................ 76 Figure 3.4-4. UTCI values at 1.5 m height in the courtyard at different hours from the Ladybug Simulation Workflow. Left: without shading. Right: Shaded. Extracted from Publication 6. . 77 Figure 3.4-5. Simulated and Monitored Thermal Delta (TD) at different hours for the different configurations analyzed. Adapted from Publication 6. ..................................................... 78 Figure 3.5-1. Methodology workflow for ventilation and thermal inertia analysis. Extracted from Publication 5. ............................................................................................................ 79 Figure 3.5-2. Closed courtyard simulation results at 1.5 m above the ground. Adapted from Publication 5. ............................................................................................................ 81 Figure 3.5-3. Closed courtyard simulation results at 1.5 m above the ground. Adapted from Publication 5. ............................................................................................................ 81 Figure 3.5-4. Wind flow pattern in a cross-section of the courtyard in the wind direction at 16.00 hours. Extracted from Publication 5. .................................................................. 82 Figure 3.5-5. Courtyard UTCI for different wind speeds, ventilated and non-ventilated variations. Extracted from Publication 5. ....................................................................... 83 Chapter 4: Figure 4.2-1. Schematic description of the CS10 geometry. Extracted from Publication 7. . 88 Figure 4.2-2. Workflow of inputs and outputs streaming between LBT plugins. Extracted from Publication 7. ............................................................................................................ 88 Figure 4.2-3. Model for the simulations. Extracted from Publication 7. ............................. 89 Figure 4.2-4. Mean comfort time per strategy in June (Hot-Dry) and August (Warm-Humid) weeks. Extracted from Publication 7. ............................................................................ 90 Figure 4.2-5. Mean adaptive comfort map per strategy. Extracted from Publication 7. ....... 91 Figure 4.2-6. Wet Bulb Globe Temperature results for Case A and Case D in the best and the worst performance rooms of the house. Extracted from Publication 7. .............................. 92 Figure 4.2-7. Percentage of time at each category of WBGT in the June and August weeks. Extracted from Publication 7. ...................................................................................... 93 Figure 4.2-8. UTCI results in the courtyard for the two selected days at 2 p.m. Extracted from Publication 7. ............................................................................................................ 94 Figure 4.2-9. UTCI and Dry bulb temperature results in the courtyard for the two selected days. Extracted from Publication 7. ...................................................................................... 94 Figure 4.2-10. UTCI results in the courtyard for the two selected days with increased wind speed. Extracted from Publication 7. ............................................................................ 95 Figure 4.3-1. Coupling workflow for CFD and BES tools using LBT and HULC. ................. 97 Workflow Design for the Analysis of Transitional Space Microclimates 126 Figure 4.3-2. Rhinoceros and Grasshopper interface. Import of geometry into Grasshopper. 97 Figure 4.3-3. Completed model in HULC. ..................................................................... 98 Figure 4.3-4. Modified Condition Files created using Python code in Grasshopper interface.99 Appendix Complete copy of the publications 1 V.P. López-Cabeza, C. Galán-Marín, C. Rivera-Gómez, J. Roa-Fernández, Courtyard microclimate ENVI-met outputs deviation from the experimental data, Building and Environment. 144 (2018) 129–141. https://doi.org/10.1016/j.buildenv.2018.08.013. 2 V.P. Lopez-Cabeza, F.J. Carmona-Molero, S. Rubino, C. Rivera-Gómez, E. Fernández-Nieto, C. Galán-Marín, T. Chacón-Rebollo, Modelling of surface and inner wall temperatures in the analysis of courtyard thermal performances in Mediterranean climates, Journal of Building Performance Simulation. 14 (2021) 181–202. https://doi.org/10.1080/19401493.2020.1870561. 3 V.P. López-Cabeza, E. Diz-Mellado, C. Rivera-Gómez, C. Galán-Marín, H.W. Samuelson, Thermal comfort modelling and empirical validation of predicted air temperature in hot-summer Mediterranean courtyards, Journal of Building Performance Simulation. 15 (2022) 39–61. https://doi.org/10.1080/19401493.2021.2001571. 4 V.P. Lopez-Cabeza, S. Alzate-Gaviria, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, Albedo influence on the microclimate and thermal comfort of courtyards under Mediterranean hot summer climate conditions, Sustainable Cities and Society. 81 (2022) 103872. https://doi.org/10.1016/J.SCS.2022.103872. 5 V.P. Lopez-Cabeza, C. Rivera-Gomez, J. Roa-Fernández, M. Hernandez-Valencia, R. Herrera- Limones, Effect of thermal inertia and natural ventilation on user comfort in courtyards under warm summer conditions, Building and Environment. (2022) 109812. https://doi.org/10.1016/j.buildenv.2022.109812. 6 V.P. Lopez-Cabeza, E. Diz-Mellado, C. Rivera-Gomez, C. Galan-Marin, Shade and thermal comfort in courtyards: experimental versus simulation results, Buildings. 12 (2022) 1961. https://doi.org/10.3390/BUILDINGS12111961. 7 V.P. Lopez-Cabeza, A. Agarwal, Sustainable strategies for improving passive survivability of the traditional Pol House in Ahmedabad, India, Building and Environment. 207 (2022) 108503. https://doi.org/10.1016/j.buildenv.2021.108503. 8 V.P. López-Cabeza, J. Lizana, E. Diz-Mellado, C. Rivera-Gómez, C. Galán-Marín, Outdoor Microclimate Influence on Building Performance: Simulation Tools, Challenges, and Opportunities, Book Chapter in New Technologies in Building and Construction, Springer, 2022: pp. 103-. https://link.springer.com/10.1007/978-981-19-1894-0_7. Publication 1 Authors: Victoria Patricia López-Cabeza, Carmen Galán-Marín, Carlos Rivera-Gómez, Jorge Roa- Fernández. Title: Courtyard microclimate ENVI-met outputs deviation from the experimental data. Journal, year, volume, pages: Building and Environment, 2018, 144, 129–141. doi: https://doi.org/10.1016/j.buildenv.2018.08.013 Journal metrics: JCR Impact Factor (2018): 4.82 Construction and Building Technology (3/132) Q1 / T1 / D1 SJR Impact Factor (2018): 1.879 Building and Construction (8/173) Q1 / T1 / D1 Contents lists available at ScienceDirect Building and Environment journal homepage: www.elsevier.com/locate/buildenv Courtyard microclimate ENVI-met outputs deviation from the experimental data V.P. López-Cabeza, C. Galán-Marín ∗ , C. Rivera-Gómez, J. Roa-Fernández Departamento de Construcciones Arquitectónicas 1, Escuela Técnica Superior de Arquitectura, Universidad de Sevilla, Avda. Reina Mercedes, 2, 41012, Seville, Spain ARTICLE INFO Keywords: Inner courtyards Patio Simulation ENVI-met Mediterranean climate ABSTRACT It is necessary to determine if the tools used in the process of building performance calculation accurately reflect the real conditions, with the objective of introducing simulation in the design process of buildings. The aim of this study is the assessment of the effectiveness of the software ENVI-met v4 on the prediction of the thermodynamic performance of courtyards by means of comparing between field data obtained from simultaneous monitoring of three courtyards and the results obtained from the software simulations. The results of the study show a significant difference between monitored and simulated data for air temperature inside the courtyards. And the difference between outdoor temperature and courtyard temperature is too important not to be considered in building efficiency calculations. For that reason, ENVI-met it is not an accurate software to be used in the process of design of this kind of architectural transitional spaces that can be very important in order to design nZEB in some climates. 1. Introduction The sustainable development of our society is one of the main challenges we face nowadays. Considering a future of climate change, in which the reduction of the energy consumption will be needed, the research into ways of saving energy is essential. Within the European framework for 2030, sustainable cities and communities is one of the 17 goals of sustainable development. The Directive 2010/31/UE, EPBD, of the European Parliament, relative to the energy performance of buildings, declares that Member States must assure A fulfilment of the Zero Energy Consumption (nZEB) requirements by every new building by 31 December 2020. There is a necessity of taking advantage of the opportunities to reduce the energy consumption of buildings both in the previous design and in the refurbishing of the existing ones. Geometry and surface properties of buildings generate microclimates, formerly analysed by Ogyay [1], Oke [2] and Givoni [3], which can be the key in the energy development of buildings. In this sense, special attention must be paid to the elements of the building that contribute to passive conditioning, such as courtyards. The courtyard has been traditionally used in hot climates as an element providing light, ventilation and helps conditioning of the building. Previous research manifests the relationship between geometry, energy performance and the context of the architecture [4,5]. As an element of passive conditioning, the courtyard is especially relevant in the present situation of climate change, accelerated by the high energy consumption we are facing nowadays [6]. It has been proven to be a significant element with a cooling effect on buildings [7–10]. The thermodynamic effects explaining the performance of the courtyard are affected by a variety of factors such as geometry, the presence of vegetation or water, or wind among others [11,12]. All of them have an influence on the temperature distribution inside the courtyard. This distribution is explained by the physical phenomena known as stratification, convection and flow patterns [13]. Courtyard could be a key element in the design of nearly zero energy consumption buildings at some latitudes. They are an interesting element not only from the point of view of thermal comfort in the space of the courtyard, but also because of its direct effect on the reduction of the energy consumed by the building. Being able to predict these effects and decide which design techniques can make the courtyard more efficient is an objective of the present research, as they are key elements that can be employed to achieve the goals of sustainable development of cities. Current tools and regulations do not allow an efficient use of the environmental benefits of courtyards, or even penalize it. Design tools are not able to simulate the microclimate created by courtyards, and even consider them as detrimental to the building due to exposure to outdoor climatic conditions. There is a general belief that a compact building is more efficient than a https://doi.org/10.1016/j.buildenv.2018.08.013 Received 8 June 2018; Received in revised form 24 July 2018; Accepted 8 August 2018 ∗ Corresponding author. E-mail address: [email protected] (C. Galán-Marín). %XLOGLQJDQG(QYLURQPHQW² $YDLODEOHRQOLQH$XJXVW (OVHYLHU/WG$OOULJKWVUHVHUYHG porous one. Conversely, some studies have monitored a reduction of the temperature in courtyards up to 8 °C in hot climates [13,14], which means an important benefit regarding energy consumption and comfort, given that a reduction of 1 °C in air conditioning has been reported to entail an energy consumption increase of 8%, according to IDAE (Spanish Institute of Diversification and Energy Consumption Reduction). Hence the possibility of courtyard modelling for the prediction of their performance is a fundamental objective for the improvement of the efficient design of buildings. Therefore, the analysis of the available software is needed to find an adequate tool that takes into account all these variables in the process of design. This analysis and calculation is a complex issue which requires a powerful simulation tool providing accurate results and flexibility: powerful enough to calculate the Navier-Stokes equations that define the behaviour of the fluid mechanics coupled with those equations defining thermodynamic behaviours; with sufficient flexibility, so that the model can be modified according to the adjustments needed when comparing them to the physical reality and the possibility to include all the parameters affecting the performance of the courtyard [15]. Thus, the use of powerful calculation software based on CFD (Computational Fluid Dynamics) to finite elements is required. Among the different tools capable of simulating outdoor environment based on CFD we can find ANSYS Fluent, IES. VE, Design Builder and ENVI-met as the most used ones. Since ANSYS Fluent is a software mainly designed to simulate fluid models and turbulence, it requires a not proportional amount of time and effort to be applied to buildings. Design Builder is a specific software for architecture that has been reported not to accurately simulate the temperature in the courtyards [16]. Since we are interested in the performance of outdoor spaces, the most suitable software seems to be ENVI-met, which we describe as follows. The aim of this paper is the assessment of the effectiveness of the software ENVI-met v4 in the prediction of the thermodynamic performance of courtyards by means of comparing field data obtained from simultaneous monitoring of three courtyards and the results obtained from the software simulations. 2. Materials and methodology 2.1. Software ENVI-met is designed to analyse microclimates through the fundamental laws of fluids and thermodynamics, being able to simulate interactions between buildings, soil, vegetation and air. For this reason, it is widely used in the study of urban microclimates. It is designed for 3d modelling with a typical horizontal resolution from 0.5 to 5 m and a time frame of 24–48 h with a time step of 1–5 s. This resolution allows the analysis of small-scale interactions between individual buildings, surfaces and plants [17]. In contrast, it presents some limitations when it comes to defining the initial boundary conditions for the calculation. The most significant one is the constant wind speed and direction during the simulation, and the same occurs with the cloudiness rate, unlike the air temperature and humidity which can be modified throughout the day. This software has been accepted for its use in the simulation of urban microclimates in several previous studies. Aysan Forouzandeh [18] validates it by contrasting data from the simulation with field data obtained from a courtyard in Hanover, Germany. However, errors are detected in the mean radiant temperature for sun-exposed areas. The same inaccuracy is detected by Acero and Herranz-Pascual [19], and it affects the software accuracy to predict thermal comfort parameters. On the other side, Gusson et al. [20] ratify the use of ENVI-met in urban zones of the subtropical climate of Sao Paulo. Salata et al. [21] suggest a procedure to fit ENVI-met simulation outputs to experimental data. Several studies, once they validate the use of the software with a case study, use ENVI-met to predict the effect of different factors on the urban microclimate, such as geometry and orientation [10,12,22], urban morphology [23], albedo of the surfaces [12], the presence of vegetation [12,24–26], and its implications in the thermal comfort of users [19,27]. It has been used even to analyse the role of vegetation in archaeological sites [28]. Table 1 shows a selection of papers that use ENVI-met along with some statistical parameters used to validate the simulations. These parameters are the coefficient of determination (R2) and the Root Mean Square Error (RMSE) which, in order for a model to be considered reliable, must tend to the following values: R2 →1, RMSE →0. It can be seen that the results are quite variable, ranging from a R2 of 0.52–0.96, and a RMSE from 0.26 to 4.83. Furthermore, not many of the studies that analyse ENVI-met are focused mainly on the microclimate generated in courtyards, especially in hot dry climates, and the courtyards are not the typical dimensions of Spanish building courtyards. For that reason, the main objective of this research is the demonstration of the software as a tool capable of predicting the thermodynamic performance of this kind of courtyards, where temperatures 8 °C below the outdoor temperature can be reached during the hottest periods of the year. In this way, ENVI-met could be a useful tool for the architectural design of these traditional spaces of the Mediterranean climate. This research uses the software ENVI-met v.4.3.1 to perform the microclimate simulations. Table 1 Results of experimental validations carried out in other studies. Author City R 2 RMSE Area Scale (m) Taleghani et al. [10] Netherlands –0,26 Courtyards 10 × 50 Ghaffarianhoseini [12] Kuala Lumpur 0,96 –Courtyard 24 × 24 Salata et al. [21] Rome 0,88 1,89 Courtyard 27 × 30 Nasrollahi et al. [22] Shiraz 0,87 –Courtyard 12 ×12 h = 8 Forouzandeh [18] Hanover –3,05 Courtyard 14 × 6 h = 7 Middel et al. [23] Phoenix –1,74 Urban neighbourhood 115 × 120 Song and Park [26] Changwon 0,52 4,83 Urban open spaces 200 × 200 Hedquist and Brazel [29] Phoenix 0,89 2,9 Urban neighbourhood 200 × 200 Acero et al. [19] Bilbao 0,92 –Urban squares and park 300 × 200 Lee, Mayer and Chen [25] Freiburg 0,85 0,66 Urban neighbourhood 150 × 150 Duarte et al. [30] Sao Paolo –1,61 Urban 300 × 300 Ketterer and Matzarakis [31] Stuttgart 0,88 0,28 Urban 180 × 180 Wang and Zacharias J [32] Beijing 0,81 –Urban district 300 × 600 Wang et al. [33] Toronto 0,6–0,83 –Urban 300 × 300 Jänicke et al. [34] Berlin 0,89 1.35 Urban façade 80 × 110 Qaid and Ossen [35]. Putrajaya 0,69 1,82 Urban Boulevard 460 × 380 Taleghani et al. [36] Netherlands 0,8 1,2 Urban neighbourhood 90 × 90 Yang et al. [37] Guangzhou 0,94 1,01 Urban park 255 × 255 Gusson and Duarte [20]. Sao Paulo –1,9 Urban – V.P. López-Cabeza et al. %XLOGLQJDQG(QYLURQPHQW²  The process used in the study consists of two steps. First, the simultaneous monitoring of the selected courtyards for the gathering of field data. Second, the simulation of the courtyards with ENVI-met v.4. Results of both steps are contrasted in order to find the conclusions of the research about the effectiveness of the software for its use in the building design process. 2.2. Courtyard description The first step of monitoring has been designed to obtain data from courtyards of different geometries in identical climate conditions. Three courtyards in Seville (Spain 37°17′01″N 5°55′20″W, elevation 42 m a.s.l.) have been monitored during the same period of time. The location of the buildings can be seen in Fig. 1. Seville is located in an area of Csa category of Köppen climate classification [38]. It consists of hot and dry summers with maximum mean temperatures of 36 °C in July and warm winters with mean temperatures of 10.9 °C in January. Mean annual precipitation is 539 mm [39]. The studied courtyards belong to buildings located in the urban area, and they have been selected according to the common dimensions of courtyards in Spain [4]. In terms of Spanish regulations, the area is classified as B4, which implies summers of the highest climate severity in Spain [40]. The characteristics of each courtyard (Fig. 2) are described and summarized in Table 2. The geometry of each courtyard is described by means of its Aspect Ratio (AR), i.e., the correlation between the height and width of the courtyard as described by Hall [41]: Fig. 1. Location of the cases studies. Fig. 2. Images of the studied courtyards. Table 2 Geometric characteristics of the courtyards. Courtyard Dimensions (m) Height (m) AR (H/W) Sky View Factor Albedo Case 1. Residential. 7.4 × 3.1 12.6 4.1 0.03 0.8 Case 2. School. 7.0 × 11.0 8.9 1.3 0.12 0.8 Case 3. Residential 7.5 × 13.2 10.7 1.4 0.12 0.4 V.P. López-Cabeza et al. %XLOGLQJDQG(QYLURQPHQW²  R2 above 0.83, present a RMSE of 3.35 °C in the worst case. The reliability of the results from the simulation is checked through the values of RMSEu and RMSEs. The value of RMSEu is close to RMSE and RMSEs tends to 0, which means that the results of the simulation are reliable. From the results, it is concluded that ENVI-met simulations show acceptable accuracy regarding outdoor temperatures, in the urban area surrounding the buildings, as it has been previously reported by other studies [19,23,25,30,31,36,37]. However, the simulated air temperature in the courtyards do not present such a fair agreement. Therefore, ENVI-met is able to simulate a decrease in the temperature of the courtyards of more than two degrees, thus demonstrating their tempering effect. Such prediction is however not as favourable as the real monitored performance to vindicate the role of the courtyard as a passive cooling system. 4. Discussion The results that have been obtained through analysis of the Fig. 11. Monitored and simulated temperature in Case 1. Residential. Fig. 12. Monitored and simulated temperature in Case 2. School. V.P. López-Cabeza et al. %XLOGLQJDQG(QYLURQPHQW²  parameters are in accordance with those published in other research about the software. In this sense, even the value of R2 = 0.84 obtained for the courtyard of Case 1, showing the largest difference between monitored and simulated values, is inside the range of results considered as valid by other studies. (See Table 1). Regarding the value of RMSE = 3.35, slightly higher than the mean value obtained in other studies, it is also considered as valid in previous studies (up to 4.83 obtained by Song and Park [26]). Nevertheless, in this work, whose objective is to validate the use of the software ENVI-met for the simulation of the performance of courtyards in hot climates in the design process of a building, it is Fig. 13. Monitored and simulated temperature in Case 3. Residential. Fig. 14. Simulated and monitored temperature gap. Table 8 Quantitative evaluation of the ENVI-met performance on the evaluation of R2, RMSE, RMSEs and RMSEu. R 2 RMSE (°C) RMSEu (°C) RMSEs (°C) Case 1 Outdoor 0.99 0.77 1.33 0.62 Courtyard 0.84 3.35 4.50 1.33 Case 2 Outdoor 0.99 0.73 1.46 0.83 Courtyard 0.88 2.92 4.25 1.60 Case 3 Outdoor 0.99 0.82 1.41 0.66 Courtyard 0.93 1.52 2.10 0.88 V.P. López-Cabeza et al. %XLOGLQJDQG(QYLURQPHQW²  considered that errors obtained in air temperature in the courtyards are too significant when taking into account its proportion with respect to the total amount of reduction of the monitored temperature of the courtyard. For that reason, it is not a satisfying value in the process of design of this kind of architectural spaces. The limitations of the software in the simulation of courtyards may be due to different factors. First, an insufficient resolution of the software for the small scale simulation of Mediterranean courtyards. Despite being a software design for the simulation of microclimates, the majority of the validations have been carried out for broader scenarios. This explanation is supported by the fact that the best performance in our simulation has taken place in the courtyard displaying the largest area (Case 3). Furthermore, deeper courtyards seem to be worse simulated, since the results of Case 1 with an SVF of 0.03 are not so good as those of Case 2 and 3 with an SVF of 0.12. Differences between Case 2 and 3 are because of the addition of other factors such as vegetation. Nevertheless, considering the case studies analysed, a correlation between Sky View Factors and the results in ENVI-met is difficult to stablish. It is also expected that with a longer time frame or a more defined model the results would be more similar to those obtained by experimental monitoring. This entails, however, an important increase in the calculation time and the needed computational power, which limits the usability of the software. Hence, although the results might be better, the practical use of the software in the design process would also be impossible for the majority of professionals. 5. Conclusion From a previous analysis of the existing studies that validate the use of ENVI-met in the simulation of microclimates, it has been noticed a lack of case-studies about courtyards. The effectiveness of courtyards as elements of passive conditioning in hot and dry climates has been previously reported in several studies. Since our objective is the assessment of the possibility of using ENVI-met in the design process of a building for the prediction of positive performance of courtyards, a case-study validation of this software using typical courtyard size in our latitudes justifies the present research. It is concluded that, although the software is able to simulate a reduction in the temperature of the courtyards, it is not as high as in the real cases. The larger the size and higher SVF of the courtyard, the better the results of the performed simulation. For that reason, although the software can be employed to understand the general tendency of a courtyard, the data it provides is not accurate enough to be used as input parameters in the process of energy consumption calculation of the building. This software is aimed at larger urban scale simulations. On the contrary, designers would require a more helpful tool for building courtyard scale simulation. Future studies must assess the possibility of testing ENVI-met on the simulation of other parameters such as mean radiant temperature in order to validate the accuracy of the software in the prediction of thermal comfort. The analysis of more cases studies is also necessary to stablish a solid correlation between the geometry of the courtyard, daily temperature and daily temperature gap. It may also be interesting the comparison of ENVI-met with other software as Rayman or Honeybee which are also widely used in this field. Acknowledgments The authors wish to acknowledge the IUACC “Instituto Universitario de Arquitectura y Ciencias de la Construcción”for the necessary support to develop this research. They thank the two anonymous reviewers whose comments/suggestions helped to improve and clarify this manuscript. This work has been supported by the National Government of Spain Research Project MTM2015-64577-C2-2-R. References [1] M.V. Olgyay, Design with Climate, Princeton University Press, Princeton, NJ, USA, 1963. [2] T.R. Oke, Boundary Layer Climates, second ed., Routledge, London, 1992. [3] B. Givoni, Climate Considerations in Building and Urban Design, Van Nostrand Reinhold, New York, 1998. [4] J. Rojas-Fernández, C. Galán-Marín, J. Roa-Fernández, C. Rivera-Gómez, Correlations between GIS-based urban building densification analysis and climate guidelines for Mediterranean courtyards, Sustain. Times 9 (2017), https://doi.org/ 10.3390/su9122255. [5] M. Taleghani, M. Tenpierik, A. van den Dobbelsteen, Environmental impact of courtyards - a review and comparison of residential courtyard buildings design in different climates, J. Green Build. 7 (2012) 113–136, https://doi.org/10.3992/jgb. 7.2.113. [6] T.F. Stocker, Q. Dahe, G.-K. Plattner, L.V. Alexander, S.K. Allen, N.L. Bindoff, F.- M. Bréon, J.A. Church, U. Cubash, S. Emori, P. Forster, P. Friedlingstein, L.D. Talley, D.G. Vaughan, S.-P. Xie, Technical summary, clim. Chang. 2013 phys. Sci. Basis. Contrib. Work. Gr. I to fifth assess, Rep. Intergov. Panel Clim. Chang. (2013) 33–115, https://doi.org/10.1017/CBO9781107415324.005. [7] N.A. Al-Hemiddi, K.A.M. Al-Saud, The effect of a ventilated interior courtyard on the thermal performance of a house in a hot–arid region, Renew. Energy 24 (2001) 581–595 www.elsevier.nl/locate/renene. [8] M.A. Cantón, C. Ganem, G. Barea, J.F. Llano, Courtyards as a passive strategy in semi dry areas. Assessment of summer energy and thermal conditions in a refurbished school building, Renew. Energy 69 (2014) 437–446, https://doi.org/10. 1016/j.renene.2014.03.065. [9] T. Kubota, M.A. Zakaria, S. Abe, D.H.C. Toe, Thermal functions of internal courtyards in traditional Chinese shophouses in the hot-humid climate of Malaysia, Build. Environ. 112 (2017) 115–131, https://doi.org/10.1016/j.buildenv.2016.11. 005. [10] M. Taleghani, M. Tenpierik, A. van den Dobbelsteen, D.J. Sailor, Heat in courtyards: a validated and calibrated parametric study of heat mitigation strategies for urban courtyards in The Netherlands, Sol. Energy 103 (2014) 108–124, https://doi.org/ 10.1016/j.solener.2014.01.033. [11] H.A. Abdulkareem, Thermal comfort through the microclimates of the courtyard. A critical review of the middle-eastern courtyard house as a climatic response, Procedia - Soc. Behav. Sci. 216 (2016) 662–674, https://doi.org/10.1016/j.sbspro. 2015.12.054. [12] A. Ghaffarianhoseini, U. Berardi, A. Ghaffarianhoseini, Thermal performance characteristics of unshaded courtyards in hot and humid climates, Build. Environ. 87 (2015) 154–168, https://doi.org/10.1016/j.buildenv.2015.02.001. [13] J.M. Rojas, C. Galán-Marín, E.D. Fernández-Nieto, Parametric study of thermodynamics in the mediterranean courtyard as a tool for the design of eco-efficient buildings, Energies 5 (2012) 2381–2403, https://doi.org/10.3390/en5072381. [14] E.Á. Rodríguez Jara, F.J. Sánchez de la Flor, S. Álvarez Domínguez, J.M. Salmerón Lissén, A. Rincón Casado, Characterizing the air temperature drop in Mediterranean courtyards from monitoring campaigns, Sustain. Times 9 (2017), https://doi.org/ 10.3390/su9081401. [15] J.M. Rojas-Fernández, Thermodinamics of Mediterranean Courtyards: Quantification and Applications in Eco-efficient Architectural Design, University of Seville, 2012. [16] O. Al-Hafith, S. B K, S. Bradbury, P. de Wilde, Simulation of courtyard spaces in a desert climate, Energy Procedia, 9th International Conference on Applied Energy, Cardiff, UK, 2017, https://doi.org/10.1016/j.egypro.2017.12.401. [17] Introduction –ENVI_MET, (n.d.). http://www.envi-met.com/introduction/(accessed March 20, 2018). [18] A. Forouzandeh, Numerical modeling validation for the microclimate thermal condition of semi-closed courtyard spaces between buildings, Sustain. Cities Soc. 36 (2018) 327–345, https://doi.org/10.1016/j.scs.2017.07.025. [19] J.A. Acero, K. Herranz-Pascual, A comparison of thermal comfort conditions in four urban spaces by means of measurements and modelling techniques, Build. Environ. 93 (2015) 245–257, https://doi.org/10.1016/j.buildenv.2015.06.028. [20] C.S. Gusson, D.H.S. Duarte, Effects of built density and urban morphology on urban microclimate - calibration of the model ENVI-met V4 for the subtropical Sao paulo, Brazil, Procedia Eng 169 (2016) 2–10, https://doi.org/10.1016/j.proeng.2016.10. 001. [21] F. Salata, I. Golasi, R. de Lieto Vollaro, A. de Lieto Vollaro, Urban microclimate and outdoor thermal comfort. A proper procedure to fit ENVI-met simulation outputs to experimental data, Sustain. Cities Soc. 26 (2016) 318–343, https://doi.org/10. 1016/j.scs.2016.07.005. [22] N. Nasrollahi, M. Hatami, S.R. Khastar, M. Taleghani, Numerical evaluation of thermal comfort in traditional courtyards to develop new microclimate design in a hot and dry climate, Sustain. Cities Soc. 35 (2017) 449–467, https://doi.org/10. 1016/j.scs.2017.08.017. [23] A. Middel, K. Häb, A.J. Brazel, C.A. Martin, S. Guhathakurta, Impact of urban form and design on mid-afternoon microclimate in Phoenix Local Climate Zones, Landsc. Urban Plann. 122 (2014) 16–28, https://doi.org/10.1016/j.landurbplan.2013.11. 004. [24] F. Salata, I. Golasi, A.D.L. Vollaro, R.D.L. Vollaro, How high albedo and traditional buildings' materials and vegetation affect the quality of urban microclimate. A case study, Energy Build. 99 (2015) 32–49, https://doi.org/10.1016/j.enbuild.2015.04. 010. [25] H. Lee, H. Mayer, L. Chen, Contribution of trees and grasslands to the mitigation of human heat stress in a residential district of Freiburg, Southwest Germany, Landsc. V.P. López-Cabeza et al. %XLOGLQJDQG(QYLURQPHQW²  Urban Plann. 148 (2016) 37–50, https://doi.org/10.1016/j.landurbplan.2015.12. 004. [26] B. Song, K. Park, Contribution of greening and high-albedo coatings to improvements in the thermal environment in complex urban areas, Adv. Meteorol. 2015 (2015) 12–21, https://doi.org/10.1155/2015/792172. [27] K. Fabbri, A. Di Nunzio, J. Gaspari, E. Antonini, A. Boeri, Outdoor comfort: the ENVI-BUG tool to evaluate PMV values output comfort point by point, Energy Procedia 111 (2017) 510–519, https://doi.org/10.1016/j.egypro.2017.03.213. [28] K. Fabbri, G. Canuti, A. Ugolini, A methodology to evaluate outdoor microclimate of the archaeological site and vegetation role: a case study of the Roman Villa in Russi (Italy), Sustain. Cities Soc. 35 (2017) 107–133, https://doi.org/10.1016/j.scs.2017. 07.020. [29] B.C. Hedquist, A.J. Brazel, Seasonal variability of temperatures and outdoor human comfort inPhoenix, Arizona, U.S.A, Build. Environ. 72 (2014) 377–388, https://doi. org/10.1016/j.buildenv.2013.11.018. [30] D.H.S. Duarte, P. Shinzato, C. Dos, S. Gusson, C.A. Alves, The impact of vegetation on urban microclimate to counterbalance built density in a subtropical changing climate, Urban Clim 14 (2015) 224–239, https://doi.org/10.1016/j.uclim.2015.09. 006. [31] C. Ketterer, A. Matzarakis, Comparison of different methods for the assessment of the urban heat island in Stuttgart, Germany, Int. J. Biometeorol. 59 (2015) 1299–1309, https://doi.org/10.1007/s00484-014-0940-3. [32] Y. Wang, J. Zacharias, Landscape modification for ambient environmental improvement in central business districts - A case from Beijing, Urban For. Urban Green. 14 (2015) 8–18, https://doi.org/10.1016/j.ufug.2014.11.005. [33] Y. Wang, U. Berardi, H. Akbari, Comparing the effects of urban heat island mitigation strategies for Toronto, Canada, Energy Build. 114 (2016) 2–19, https://doi. org/10.1016/j.enbuild.2015.06.046. [34] B. Jänicke, F. Meier, M.-T. Hoelscher, D. Scherer, Evaluating the Effects of Façade Greening on Human Bioclimate in a Complex Urban Environment, Adv. Meteorol. 2015 (2015) 1–15, https://doi.org/10.1155/2015/747259. [35] A. Qaid, D.R. Ossen, Effect of asymmetrical street aspect ratios on microclimates in hot, humid regions, Int. J. Biometeorol. 59 (2015) 657–677, https://doi.org/10. 1007/s00484-014-0878-5. [36] M. Taleghani, L. Kleerekoper, M. Tenpierik, A. Van Den Dobbelsteen, Outdoor thermal comfort within five different urban forms in the Netherlands, Build. Environ. 83 (2015), https://doi.org/10.1016/j.buildenv.2014.03.014. [37] X. Yang, L. Zhao, M. Bruse, Q. Meng, Evaluation of a microclimate model for predicting the thermal behavior of different ground surfaces, Build. Environ. 60 (2013) 93–104, https://doi.org/10.1016/j.buildenv.2012.11.008. [38] M. Kottek, J. Grieser, C. Beck, B. Rudolf, F. Rubel, World map of the Köppen-Geiger climate classification updated, Meteorol. Z. 15 (2006) 259–263, https://doi.org/10. 1127/0941-2948/2006/0130. [39] Resúmenes climatológicos - España - Anuales - Agencia Estatal de Meteorología - AEMET. Gobierno de España, (n.d.). http://www.aemet.es/es/serviciosclimaticos/ vigilancia_clima/resumenes?w=0&datos=2 (accessed March 21, 2018). [40] CTE. Código técnico de la Edificación. Documento Básico HE Ahorro de energía, (n. d.). https://www.codigotecnico.org/images/stories/pdf/ahorroEnergia/DBHE.pdf (accessed April 16, 2018). [41] D.J. Hall, S. Walker, A.M. Spanton, Dispersion from courtyards and other enclosed spaces, Atmos. Environ. 33 (1999) 1187–1203 https://ac.els-cdn.com/ S1352231098002842/1-s2.0-S1352231098002842-main.pdf?_tid=7e7c7a95- 8bda-4803-b12e-8861f38c9061&acdnat=1526296658_ 0b6eaa7b484bfd71dc581e6b4d9531e2. [42] E.R. Marciotto, A.P. Oliveira, S.R. Hanna, Modeling Study of the Aspect Ratio Influence on Urban Canopy Energy Fluxes with a Modified wall-canyon Energy Budget Scheme, (2010), https://doi.org/10.1016/j.buildenv.2010.05.012. [43] X. Yang, Y. Li, The impact of building density and building height heterogeneity on average urban albedo and street surface temperature, Build. Environ. 90 (2015) 146–156, https://doi.org/10.1016/j.buildenv.2015.03.037. [44] G.T. Johnson, I.D. Watson, The determination of View-Factors in Urban Canyons, J. Appl. Meteorol. Climatol. 23 (1984) 329–335, https://doi.org/10.1175/1520- 0450(1984)023<0329:TDOVFI>2.0.CO;2. [45] A. Al-Sudani, H. Hussein, S. Sharples, Sky View Factor Calculation A computationalgeometrical approach, Sp. Syntax Ontol (2017) 673–682 2-ECAADe 35. 2 http:// papers.cumincad.org/data/works/att/ecaade2017_240.pdf. [46] Agencia Estatal de Meteorología - AEMET. Gobierno de España, (n.d.). http://www. aemet.es/es/portada (accessed April 17, 2018). [47] J. Busby, M. Lewis, H. Reeves, R. Lawley, Initial geological considerations before installing ground source heat pump systems, Q. J. Eng. Geol. Hydrogeol. 42 (2009) 295–306, https://doi.org/10.1144/1470-9236/08-092. [48] C.J. Willmott, Some Comments on the Evaluation of Model Performance, (n.d.). https://journals.ametsoc.org/doi/pdf/10.1175/1520-0477%281982%29063% 3C1309%3ASCOTEO%3E2.0.CO%3B2. V.P. López-Cabeza et al. %XLOGLQJDQG(QYLURQPHQW²  Publication 2 Authors: Victoria Patricia Lopez-Cabeza, Francisco Javier Carmona-Molero, Samuele Rubino, Carlos Rivera-Gómez, Enrique Fernández-Nieto, Carmen Galán-Marín, Tomás Chacón Rebollo. Title: Modelling of surface and inner wall temperatures in the analysis of courtyard thermal performances in Mediterranean climates. Journal, year, volume, pages: Journal of Building Performance Simulation, 2021, 14, 181–202. doi: https://doi.org/10.1080/19401493.2020.1870561 Journal metrics: JCR Impact Factor (2021): 2.867 Construction and Building Technology (32/68) Q2 / T2 / D5 SJR Impact Factor (2021): 0.814 Architecture (8/147) Q1 / T1 / D1 JOURNAL OF BUILDING PERFORMANCE SIMULATION 2021, VOL. 14, NO. 2, 181–202 https://doi.org/10.1080/19401493.2020.1870561 Modelling of surface and inner wall temperatures in the analysis of courtyard thermal performances in Mediterranean climates V.P. López-Cabeza a, F.J. Carmona-Molerob,S.Rubino b, C. Rivera-Gómeza, E.D. Fernández-Nietoc, C. Galán-Marín aand T. Chacón-Rebollob aDepartamento de Construcciones Arquitectónicas 1, Instituto Universitario de Arquitectura y Ciencias de la Construcción, Escuela Técnica Superior de Arquitectura, Universidad de Sevilla, Avda. Reina Mercedes, 2, 41012, Seville, Spain; bDepartamento de Ecuaciones Diferenciales y Análisis Numérico, Facultad de Matemáticas and Instituto de Matemáticas (IMUS), Universidad de Sevilla, C. Tarfia s/n, 41012, Seville, Spain; cDepartamento de Matemática Aplicada I, Escuela Técnica Superior de Arquitectura, and Instituto de Matemáticas (IMUS), Universidad de Sevilla, Avda. Reina Mercedes, 2, 41012, Seville, Spain ABSTRACT Courtyards are an effective passive strategy for improving the energy performance of buildings. However, there is a lack of accurate simulation tools for their thermal performance due to their complex thermodynamics. This paper’s contribution is the coupling of a CFD model with a system of differential equations at the walls, governing surface and inner wall temperatures, providing an accurate computation of courtyard thermal performance. On this basis, boundary conditions for the standard Boussinesq equations governing temperatures, wind velocity, and pressure within the courtyard are established. Modelling results are compared with monitored data in two days of different months in a courtyard in Seville (Spain). Simultaneously, simulations carried out by existing software were performed. Resulting data showed more accuracy than existing tools, with a Root Mean Square Error of 1.19 (August) and 1.59 (October), while the corresponding values with ENVI-met were 3.31-3.4, and with Ladybug Tools 2.59-4.49. ARTICLE HISTORY Received 8 June 2020 Accepted 23 December 2020 KEYWORDS Courtyards; ENVI-met; numerical model; model evaluation; microclimate simulation Table of symbols: θTemperature inside the courtyard θin Inner temperature of the wall θwSurface temperature of the wall θ(x◦δ,t)Air temperature near the wall ρwWall density CwSpecific Heat of the wal VwWall volume Uext Transmittance coefficient of the air AAreaofthewall Uin Transmittance coefficient of the wall Emissivity σStefan-Boltzmann constant αwAbsorption coefficient I Solar radiation Q Heat transfer θsSky temperature μ Sky function e Wall thickness Domain TFinaltime L Characteristic flow length CONTACT C. Galán-Marín [email protected] Departamento de Construcciones Arquitectónicas 1, Instituto Universitario de Arquitectura y Ciencias de la Construcción, Escuela Técnica Superior de Arquitectura, Universidad de Sevilla, Avda. Reina Mercedes, 2, 41012, Seville, Spain βCharacteristic maximum temperature difference uAir velocity field p Pressure νViscosity RH Relative Humidity lαThermal expansion coefficient θref Reference temperature αThermal diffusion g Gravity constant ρ0Reference density iBoundaries of the system usWind velocity at the upper part of the courtyard Pr Prandtl number Ra Rayleigh number AR Aspect Ratio of the courtyard (hmax/W) hmax Maximum height of the courtyard W Width of the courtyard R2Coefficient of determination RMSE Root Mean Square Error © 2021 International Building Performance Simulation Association (IBPSA) 182 V. LÓPEZ-CABEZA ET AL. RMSEs Systematic Root Mean Square Error RMSEu Unsystematic Root Mean Square Error MAPE Mean Absolute Percentage Error 1. Introduction Climate change is one of the main challenges our society is facing nowadays. Projected global mean surface temperature will rise over all possible scenarios and heat waves will occur more often and last longer (IPCC 2014). Moreover, this problem will be intensified in urban environments, where heat island effect may increase urban microclimate up to 10°C (Santamouris and Yun 2020), reducing inhabitants’ comfort and increasing the energy consumption of buildings to meet cooling necessities (Santamouris et al. 2001). Aiming at mitigating global warming effects and impacts, the decarbonization and implementation of climate-resilient design solutions in buildings are critical. They are responsible for almost 40% of energy- and process-related emissions, taking climate action in buildings is among the most cost-effective strategies (IEA/UN 2019). Energy consumption reduction in buildings is possible through advanced design, as it is possible to achieve zero energy consumption buildings or even positive. Furthermore, through the design of climate-resilient urban environments, it is possible to reduce the heat island effect (Srebric, Heidarinejad, and Liu 2015). From the analysis of traditional architecture that is adapted to specific climate conditions, it is possible to learn climate-resilient strategies for new constructions and increase user comfort (Chandel, Sharma, and Marwah 2016). The design of transitional spaces such as courtyards is one of the best strategies that can reduce energy consumption as well as temper the outdoor climatic conditions creating a microclimate in the courtyard that also helps to reduce the heat island effect in the urban scale (Kubota etal.2017;Rojas,Galán-Marín,andFernández-Nieto2012; Rojas-Fernández et al. 2017). However, a deep understanding of the thermodynamic effects that occur in the courtyards, i.e. stratification, convection, and flow patterns, is needed to decide the optimum design (Rojas, Galán-Marín,andFernández-Nieto2012).Plentyoffactors take part in the performance of a courtyard, not only its geometry but also shading elements, vegetation, water elements and the albedo of the surfaces among others (Abdulkareem2016; Ghaffarianhoseini,Berardi, andGhaffarianhoseini 2015). A simulation procedure able to incorporate all these factors used during the design process would be helpful to make decisions about the project, based on accurate data. Furthermore, if the simulated microclimate is incorporated into an energy modelling tool, it would accurately predict the energy consumption of the building. Although there are many simulation tools to predict the energy performance of buildings, the complex thermodynamic effects that happen in courtyards require simulation tools with high computational resources and thousands of simulation hours to calculate and simulate a representative period, based on CFD (Computational Fluid Dynamics) and finite elements (Blocken et al. 2011). In the analysis performed by Rojas et al. (Rojas-Fernández et al. 2018), the main available simulation tools are analysed concerning their capacity to simulate transitional spaces from the point of view of a professional of the building sector. They found that there is no available software that meets all requirements for this reliable microclimate simulation. Even though some tools have been developed in recent years concerning the urban microclimate simulation, the efforts are focused on the integrated systemic approach, which includes urban modelling of microclimatic effects, transient heat flows, plants and equipment and occupant behaviour (Mauree et al. 2019). However, this approach requires a huge computational power and may lack accuracy in favour of other aspects such as managing a huge amount of data. This is the case of one of the most widely used dynamic tools for the simulation of urban microclimates: ENVI- met (Tsoka, Tsikaloudaki, and Theodosiou 2018). ENVI- met is a three-dimensional non-hydrostatic model for the simulation of surface-plant-air interactions (“ENVI-Met” 2019). The software integrates an atmospheric model, vegetation model, soil model, and the built environment and building systems. However, it has some limitations: first, it needs a great amount of time to provide results given the integration of such quantity of data. Second, it is developed for urban microclimates where wind causes turbulent situations. But in the case of courtyards, generally protected from the wind, the thermodynamic effects are governed by convection patterns. Thus, existing CFD models usually don’t provide accurate results. The Ladybug Tools are an emerging methodology that combines different simulation engines into the same graphical interface of Grasshopper for Rhinoceros (Sadeghipour Roudsari, Pak, and Smith 2013). The Ladybug Tools allows the connection of the building energy simulation by EnergyPlus and the microclimate CFD calculation by OpenFOAM. It is open source, being continuously enhanced by the research community. It has some limitations too. First, there is only one OpenFOAM heat transfer solver linked to the Ladybug interface; thus, it may not be the best one for outdoor microclimate analysis. Second, although the different simulation engines are linked through Grasshopper, it holds the potential to JOURNAL OF BUILDING PERFORMANCE SIMULATION 183 Figure 1. Location of the building. increase error as data is passed from one engine to the other. In this context, it is necessary the development of new modelspecificallycreatedfortheperformancesimulation ofcourtyards, aimingtogain accuracyand computational speed. It should take into account buoyancy, solar radiation, wind and thermal exchange with the surrounding building. Once developed, it should be easy to implement in a ‘hybrid’ simulation workflow (Elwy et al. 2018; Mackey et al. 2017) to link with other simulation engines if needed. This paper develops and validates a novel numerical procedure to evaluate the courtyard thermal performance as a climate-resilient strategy of buildings. The proposed model is developed in FreeFEM software and validated in a case study located in the city of Seville, Spain. Moreover, the model accuracy is compared with other existing commercial tools, such as the previously described ENVI-met and the Ladybug tools. Both are also contrasted with the monitored data. The paper is structured as follows. First, the case study for data collection and model validation is characterized. Second, the methodology is presented, which is divided into three sections: section 3.1 shows the novel coupled model for courtyard; section 3.2 details the data collection procedure for model validation; and section 3.3 provides details about the comparison with existing simulation tools. Third, the results of novel procedure and existing simulation tools are presented and discussed, highlighting the accuracy of each of the simulation tools described. Finally, conclusions are reported, and future research steps are described. 2. Definition and characterization of the case study for model validation The selected case study to test the proposed numerical model is the courtyard of the CITIUS building of the University of Seville. This city (Spain 37°1701N 5°5520W, elevation 42 m a.s.l.) is Csa in the Köppen classification (Kottek et al. 2006), characterized by hot and dry summers with mean temperatures of 36°C in July and mild winters with mean temperatures of 10.9°C in January. The mean annual precipitation is 539 mm (“Resúmenes Climatológicos - España - Anuales - AgenciaEstatalde Meteorología- AEMET. Gobiernode España” 2018). The building is located in an urban area of a high density of mid-rise constructions (Figure 1). The courtyard, graphically defined in Figure 2and Figure 3, has an approximately squared floor plan of 7.3 ×6.6 and 14.0 m height. To define the geometry of a courtyard, the concept Aspect Ratio is used, which is defined as the relation between the width and the height as follows: AR =hmax W, where hmax is the maximum height and W is the width of the courtyard. The aspect ratio of the case study courtyard is 2.02. The walls’ coating is white cement mortar. The thermal transmittance of the wall is 0.474 W/m2K. There are small glazed openings distributed sporadically in the walls. There is no shading element or vegetation in the courtyard. The interior composition of all walls is illustrated in Figure 4, needed to predict the temperature of the surfaces. They are cavity walls; the outdoor layer consists of 11.5 cm brick coated by 1 cm white mortar. The air cavity separating the outdoor layer from the inside functions as an insulator and it is followed by more layers of insulation material. Consequently, the rest of the layers of the wall are not relevant for temperature purposes because the inner temperature that affects the surface temperature is independent of the indoor. Therefore, a reduction 184 V. LÓPEZ-CABEZA ET AL. Figure 2. Images of the building. a. View. b. Courtyard. Figure 3. Courtyard drawings. Figure 4. Construction section of the wall. Units in metres. JOURNAL OF BUILDING PERFORMANCE SIMULATION 185 Table 1. Wall properties. Wall thickness (m) Brick thermal Conductivity (W/(mK)) Emissivity 0.10 0.67 0.93 Specific heat (Ws/(KgK)) Density (kg/m3) Absorption 1000 1140 0.73 assumed in our simplified courtyard is that the effective thickness of the wall is only the 11.5 cm brick (highlighted area in Figure 4). In Table 1it can be found some data relative to walls, like the effective thick mentioned. 3. Methodology Themethodology isdivided intofour sections.Section 3.1 shows the novel coupled model for courtyard simulation. Section 3.2 details the data collection procedure for model validation. Section 3.3 explains the indicator used for validation. And section 3.4. provides details about the comparison with other existing simulation tools. The methodological workflow and the relationship between the inputs and outputs is illustrated in Figure 5. Three stages can be seen in the graph. At the input data collection stage, the monitoring campaign is performed to record outdoor temperature, relative humidity and wind speed and direction to be used as inputs for the simulations. At the simulation stage, temperatures inside the courtyard are simulated using the proposed model describedinthe next sectionand two otherexistingtools: ENVI-met and the Ladybug tools. Finally, the results of the three different methodologies for simulating the temperature inside the courtyard are compared to monitoring temperature using statistical parameters in order to evaluate the performance of the proposed model with the reference of two of the most used tools to date. 3.1. Novel coupled model for the for the temperature simulation of courtyards The accurate simulation of the thermal behaviour of the courtyard needs a CFD model considering buoyancy, solar radiation, and external wind velocity. The combination of these three effects generates a genuine 3D behaviour of the airflow within the courtyard, so lower dimensional models are not appropriate. No turbulence modelling is needed, as we are interested in studying thermally stable configurations, in which courtyards are thermally efficient. This is the most common case in practice, in particular for the test cases considered. Summer climate in the Seville area is quite dry, with high temperatures, so no moisture is present during summertime. For this reason, we do not include the moisture as state variable in our model. The main modelling challenge is the accurate computation of the heat exchange with the surrounding walls, of special importance to evaluate the energy needs of the building. This requires a precise mathematical model for the thermal boundary layer within the courtyard and the thermal behaviour of the courtyard walls. With this purpose, we propose a coupled 2D problem at courtyard walls describing wall surface temperature and inner wall temperature. This model is coupled with the standard 3D Boussinesq model for the inner courtyard aerothermal flow. The length of the mesh corresponding to the numerical simulation is 0.2 m. 3.1.1. 2d surface temperature model To compute the temperature at the walls of the courtyard we consider a system of 2D differential equations. This problem involves three different temperatures: the wall surface temperature (θw), the wall inner temperature (θin) and the courtyard temperature (θ). Actually, the relevant value of θin this model is the temperature at the top of the thermal boundary layer in the air flow within the courtyard. The problem is then solved using the followings equations, that hold on the walls of the courtyard: ⎧ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎩ Cw ∂θw ∂t(x,t) =UextA(θ(x◦δ,t)−θw(x,t)) +UinA(θin(x,t)−θw(x,t)) +εwσA(θ4(x◦δ,t)−θ4 w(x,t)), +αwAI(x,t)+Q(x,t), x∈,t∈[0, T], Cw ∂θin ∂t(x,t) =Uint(θin(x,t)−θw(x,t)), x∈,t∈[0, T], (1) where Cw=ρwcwVwwith ρwthe wall density, cwthe specific heat of the wall, and Vwthe wall volume. Also, δ denotes the thickness of the wall thermal boundary layer within the air flow in the courtyard. Note that θ(x◦δ,t) is the courtyard temperature at the top of the thermal boundary layer. In system (1), Uext is the transmittance coefficient of the air (25 W/m2K), Ais the area of the considered wall, Uin is the transmittance coefficient of the wall(ratiobetween theheatconductivityandtheefficient thickness), εis the emissivity, σis the Stefan–Boltzmann constant (5.57e-8 W/(m2K4)), αwis the absorption coefficient, Iis the solar radiation and Q is a net heating radiation transfer simplification. In particular, it is only taken into account the sky temperature transmitted as a long wave being Q=μεwσA(θs(x,t)−θw(x,t)),(2) %XLOGLQJ DQG (QYLURQPHQW     @'647"" <>=@ <*=+$" + ""        6C7     5" 5" ",+%5 5@+' @    > "    @   86G7+ "**)6.7    + $     8 #%::6107%&14C.G6117 +" I+ "    6197+ $  >    <> =" #%%&" 61-7+' +$A "+ +61;72) 3$<2)3$=  "5A+ 2)3$" ,< "=61:7+#5 "6147+           +"2)3$ "" "61C7+,   +  2)3$ "    ", 61G1.7+ ( "  6907+$ I" " +5  A          %A$<%$=>$B $ 5<B$ =%A$ <%$!=6917+%$@ ""<=" ", A +" 6997+$> 69-7+$B$ " )" 5, "5 69;7+ “,” A+%$!@ :0/# 0+1F$=>#$", "1+0<=0+4 <=A " 6107+$5 69:7+ $5" + "   +   "                 +$  + $ 2)3$        +'B$  +$" +$ ",, +$ +2" +   " +$""'+1+$5  """"  +$      + $   ", 5<""= +'< = "B$  5"<$= +' "5" +  ( $* <J¨ 3@E) ++ =6947 ;0–:0◦ "+$ H "+" ""+$ "",69C7+$ ? "" 5+$"" +$ '+9 ""'+-+ ""," 5 69G7+ <= < '+;=+$"" """ '+:+$5      + $   +$A ",@"+$ ,"" + $" " ""I /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     I<-+C:×9+C:= +$ ,"<'+4=+  0$  $$)$    5        69.7+$ @ "",6-07+$  >",I"+  *)<=<=+ /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     .0/< @#%::901-=-◦  – 9% 6-17+$ ) %< )%=% @ @+ $<=" <=+$ ?"+B  @$1E  3))%)435"6 "2)3$$" ),J86-97%A+ <1=+ WBGT =0.67 T wb +0.33 T db – 0.048 log v (T db – T wb ) <1= "$ " "$ + 2)3$ ",61.7+$ 6147+", @@B6--7 <$9=+ $2)3$’+ ","@1+9<= +$ , +>"       ",    <"=+ * 7 5B$  56-;7 “,” A+$  2?H2<F=#– *)$<◦ =+  J69G7+  + /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     B$ 5   + "  <>#$= " "+$A B$ "%A+<9=6-:7E UTCI (T a , MRT, V a , pa) =T a +Offset (T a , MRT, V a , pa) <9= " $       <◦ = >#$    <◦ =  "<F=" <=( + $B$  F "$-+  )%(46 < %="8 <%%=<(%=+%%            6-47+% 8,<>?F,= +@"  8+ ' 6-C7+'  <'+915$G=+ $E 2E$," +$      ,,  "   "  + '     +2  "" 5"+" " 5@ ""+&" + #E$A ""+      "    ,  "5 +  >H3@I+  + > "$<◦ = $<◦ = ? 9:+C0 -1+C0  9:+:- -1+:-  2)3$+  $ <◦ = * #,  <94 &A+ &,  94–9C+C &A+B " + >  9C+C–9.+; ( 5 :0+  9.+;–-1+1 ( 5 ;0+   -1+1–-9+9 ( 5 -0+  >-9+9 ( 5 90+ %5 /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     2"E" +$ """ ""+ *  $      $  3 "" +$"$>K5@6-G7" "901G+ "3 %<= # < =  ('(> < '* =  ,+'+C" +$ # "@+$       '*  " )I "" "@"I" +$, + , '*" @+&",I"", , '*< =+$, "+$  I"+ *  $  @5<$;=+ $A+ $ "+ @,0+0004 - FA ",  , +"E  +$    + $       "1:-9◦  "".0/<" "=,@0+1C + * % $'+G+%@   "  8      "" +@5"  B$ +   >    & >   −9C−1-◦  −1-0◦  0–.◦  .–94◦  94–9G◦  9G–-9◦  -9–-G◦   2,I"""$+ /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     "$: +$"""<22#=90/ ,+ +"       + '   @ +$I "" I" +$ +"" "",+ " +""  E '* ,+' A<=1F+$ 5 + ** #8(2 $ ",@%" @<+$$=+$  +$",1C9-? ",941 +  %+    >#  %AE92F 9 HE:2F 9 H(E0+0:F 9 @ 0+0004 - F 9 L > 0+0009<=+0+001<= - F+ 9 & E1:◦ <<-9◦ H""E.0/H*@E0+1C  '5<"I=  &<&= *  >+  + &+ * 2 #    ) )" )# "" & )  " # "" &    " # ""+ "+ + & *  " # ""+ "+ + &<1F= /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     *  $ "$4+&B  "6-.7+ *9 :$ $B$ 5 “(  ”  6;07 "    '    6;17+ $ , " 5+' ")I " '*"('(>+$ "@ $ +$"@" "< ="<  *=+$"5  '* <" -9◦ $;=+ $" "@ '*('(>"I" +'"@ +"" " '* 6;9;-7E :1: +@ )@+$ "");# ,(!<#@ +$@" 1+-9×10 4 + $C+ $)A' I" )A5>%<> (,%A=+$ #&3,"+     #+' +$   2)3$ +        + * 7$ $" '+."+"  + $ * $,<= *<,F 9 = $<2F,= @<?F,J= B<2F 9 J= )2  90 1G40 0+C9 G;0 -+04 )," 9-0 1.90 0+G. C.0  90 1;;0 0+;G .C0 2  90 1G40 0+C9 G;0 1+44 ), 110 1.90 0+G. C.0  -0 190; 0+9: 100G ), 9-0 1.90 0+G. C.0  90 1;;0 0+;G .C0 )#  :0 1C00 0+G G:0 9+9: 2 90 1900 0+1- 400  -0 190; 0+9: 100G  ;0 9:00 0+C9 ..1 #  :0 1C00 0+G G:0 0+:9 2 90 1900 0+1- 400 ' 9: 1;00 0+9: G;0  :0 4: 0+0; 9090 ' 9: 1;00 0+9: G;0 ' I 9:0 19G0 0+4. G;0 1+;; 3 3 :00 1--0 1 19:: 9+0 2"  90/22# – – – – B=:+G 3 =0+G$=0+G  ('(>+    '%2"@+ # '%2"@+ 2 -JF"10+ ) (% $ #&3, $, 0+1 9 F 9 $ 0+01 9 F - #$ --◦  ),> 9:0×100×90 #@ -  ! >       <?=  > <=",+ /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     I #1I'#- I+') < = <?=+(-./+$  +" ;4/C) +$       4;/ 1G  + "     .:/       + <=",  >+ /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     .1/)+$   "    100/"+ '+10'+11"," +$" I I+?", + I"I @ +5  + $","" I C0/<" =+$ 100/" + * 3))%) $ + "" +$" "+ '+19"2)3$ """< I=<@I=+ $ @      "  "+@  2)3$+ ",<?= ;9◦ +* -0◦ @+$ "-:G0/+2 2)3$--◦  +$: 2)3$+ 2)3$-1◦ " -+$"2)3$+ "-0◦ +$ ,2)3$1◦ +$ "+ ",<=" ?-:◦ 5+" ".G/40/+$ 2)3$" -1◦ +2  -0 ◦ +$"" 2)3$9G◦ + @ + $ 2)3$<0–:='+1-" "    ,+ $      "+$ "" ;:   + $       -"8+$ ",+ '+1;   +      " + $ <,,09= <-:  5  =+           -9/1;/+ ** :$ $ I+$   *"+ '+1:"B$ 9+00++       +"? ;:◦ 5 +$ ""B$ :◦ +( B$  ? "+" *  B$ + "B$  <'+14=B$    "               <>#$=     >",+ /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     +$ ",>#$ AB$ +$"B$             9C I  ’ + $       ,1◦  B$ -◦ + $"I""'+1C ""  + 3    "    " +" "" +'+1G"B$  """ 1F9F+ "" B$ 9◦ 9 F""+$ @ +2B$ " "-:◦ 6-:7+ * )%( 2  F + $        2)3$)"+$E?",+*"E ",+  2)3$?",+ /012.)2%, %XLOGLQJ DQG (QYLURQPHQW     6-G7 ++<++=+EFF++F< '149091=+ 6-.7 B++*%%.+:+0*+% #<9091=+ 6;07 (   3R3 *<++=+EFF+FFF  +<?-9090=+ 6;17 B$ B$ 5<++=+EFF"""++F< ?-9090=+ 6;97 ?+',++P)+ 3 '* '"B%E ($C-9Q >>>900C+ 6;-7 ?+', +3+?+JP+>>++2 ?+2&+2# '*"E  B2% )900;+ /012.)2%, Publication 8 Authors: Victoria Patricia López-Cabeza, Jesus Lizana, Eduardo Diz-Mellado, Carlos Rivera-Gómez, Carmen Galán-Marín. Title: Outdoor Microclimate Influence on Building Performance: Simulation Tools, Challenges, and Opportunities. Book: New Technologies in Building and Construction, pp. 103-121. Series: Lecture Notes in Civil Engineering, Springer Nature, 2022. doi: https://doi.org/10.1007/978-981-19-1894-0_7 Book series metrics: ISBN: 978-981-19-1894-0 ISSN/E-ISSN: 2366-2557/2366-2565 SPI General (2018) (4/96) Q1 IE-CSIC Alto Outdoor Microclimate Influence on Building Performance: Simulation Tools, Challenges, and Opportunities Victoria Patricia López-Cabeza, Jesus Lizana, Eduardo Diz-Mellado, Carlos Rivera-Gómez, and Carmen Galán-Marín Abstract This chapter reviews the different approaches that currently exist to evaluateoutdoormicroclimates andtheirinfluenceonbuilding performance.Considering specific outdoor microclimates in building design flow can enable additional passive cooling strategies to mitigate climate risks in buildings and cities, improving their resilience capacity under extreme heat events. The available methods are defined and compared through different case studies of buildings with an inner courtyard, a traditional microclimate for passive cooling in hot climates. The results show the advantages and disadvantages of the different approaches and highlight the high interest in hybrid simulations coupling building energy simulation (BES) and computational fluid dynamics (CFD) tools for early design stages. Keywords Courtyard microclimate ·Heat mitigation ·Urban heat island ·Urban simulation ·Climate-resilient design ·Climate responsive design 1 Introduction Buildings have an important role in minimizing climate change, given that they are responsible for approximately 40% of energy consumption in the EU [1] and 35% in the world and increasing [2]. Furthermore, they not only affect climate at a global scale but also at a local scale through the urban heat island effect (the higher urban temperatures in the city in comparison to rural areas), which is related to urban compactness and energy performance of buildings [3]. In the last decade, regulations that aim to reduce the energy consumption of buildings have been developed, especially in Europe. The European Directive 2018/844 of theEuropean Parliament is one example [4], conducting net zero energy consumption requirements for buildings. V. P. López-Cabeza (B)·E. Diz-Mellado ·C. Rivera-Gómez ·C. Galán-Marín Departamento de Construcciones Arquitectónicas 1, Instituto Universitario de Arquitectura y Ciencias de la Construcción, Universidad de Sevilla, Avda. Reina Mercedes, 2, 41012 Seville, Spain e-mail: [email protected] J. Lizana Department of Engineering Science, University of Oxford, Parks Road, Oxford OX1 3PJ, UK © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 D. Bienvenido-Huertas and J. Moyano-Campos (eds.), New Technologies in Building and Construction, Lecture Notes in Civil Engineering 258, https://doi.org/10.1007/978-981-19-1894-0_7 103 104 V. P. López-Cabeza et al. Researchers have studied the cost-effectiveness of different energy-saving measuresformanyyears through building simulationtoolsand computer technology. However, there are still many challenges to be considered in ensuring high quality and useful building simulations: the importance of weather data selection, the need to include passive performance and thermal comfort indices in design processes, and the consideration of the potential benefits of specific microclimates on building performance. This last point is critical since these microenvironments have been proven to provide thermal benefits to buildings, especially in hot and warm climates. Enclosed courtyards are spaces surrounded by buildings and open to the sky. Their geometry, materials, water, vegetation, or shading elements produce thermodynamic effects that help to temper the extreme outdoor conditions. However, the courtyard performance is not considered in the existing building energy simulation (BES) tools and procedures [5]. This is due to two facts: It needs multi-outdoor climate conditions, which is not an option in most BES tools, and courtyard performance is hard to compute, given the multiple factors that have a role in it. The research aims to identify and compare existing and new procedures to account for the benefits of these specific microclimates, discussing advantages and drawbacks, and highlighting future needs toward a more accurate climate-resilient design process. In the next section, a review of existing tools and approaches is shown, and in the rest of the chapter, some case studies using different approximations are analyzed, providing some challenges, opportunities, and recommendations. 2 Existing Tools and Approaches Review Currently, different approximations to estimate the energy performance of buildings areavailable.Theycanbedividedintotwogroups:whiteboxmodels,whicharebased on the laws of physics that govern a system (e.g., conservation of mass, energy, and momentum) [6,7], and black box or data-driven models which are based on statistical methodsusingmachinelearning[8].Forthesimulationofoutdoormicroclimate,both models have been used. White box models Physics-driven models were the first to be developed, and most of the existing tools and methodologies rely on this approach. BES and computational fluid dynamics (CFD) are two of the main numerical approaches in this area: •Building Energy Simulation Tools (BES) BES estimates the physical parameters in one single node per zone, representing a uniform profile of parameters in the region. This allows for long-term unsteady analysis although it simplifies too much, and the differences in the area represented byasinglenodecannotbeconsidered.Forthatreason,outdoorsimulationcanbedone only in a simplified way. Some examples of software of this kind are EnergyPlus, TRNSYS, or DesignBuilder, among many others. These tools have been used to Outdoor Microclimate Influence on Building Performance … 105 predict the influence of semi-outdoor spaces on energy consumption and indoor comfort. However, these models underestimate the potential benefits of these spaces, given that BES tools cannot simulate the microclimatic effects that occur outside. The temperature in the transitional space is assumed to be the same as outside, which is inaccurate most of the time. A study analyzing the impact of urban geometry on the energy consumption of buildings using TRNSYS concluded that if the microclimate isnotincluded; thissoftwarecanlead toinaccuracies [9].Inthe lastyears, sometools, such as CitySim or UMI, have been specifically developed to account for the urban form on the energy consumption of districts. However, they still rely on microclimate simplifications. •Computational Fluid Dynamic Tools (CFD) In contrast to BES, these tools account for the whole volume of the space being considered, which is meshed in a few thousand to a few million control volumes, and the conservation equations are then applied to each of them. The basic that uses the CFD approach is the resolution of the Navier–Stokes’s equations which can be solved using finite element or finite volume methods for each control volume. This means that this approach requires considerable computing power to analyze unsteady states for a longer period than one day. However, they achieve high accuracy in the results. COMSOL Multiphysics, ANSYS FLUENT, OpenFoam, or FreeFEM++ are some of the many choices of CFD tools available. They provide a variety of numerical solutions, which are not limited to building simulation. For the specific case of urban microclimates, the most widely used CFD software is ENVI-met. It can analyze small-scale interactions between soil, water, air, vegetation, and buildings at different scales. This software has renounced some of the broader capabilities of non-specific CFD tools to simplify the simulation analysis process. For example, the meshing options in ENVI-met are greatly limited. •Hybrid workflows There are currently many approaches that combine BES and CFD approaches to take advantage of the benefits of each one. In the hybrid workflows, one software outputs become inputs for another software. The aims of combining tools are diverse. First, to achieve higher accuracy in the simulation predictions. Second, to improve the workability of these tools to make them appropriate for the different design phases. Third, to add some other capabilities to the simulation from one tool, for example, the calculation of comfort indexes. For this reason, the hybrid workflows generally combine one BES tool with CFD software or comfort-calculation-specific tools. For example, EnergyPlus or Trnsys are combined with ENVI-met in many studies [10– 12], or with Ladybug Tools [13–16], which are a set of plugins for Grasshopper that link the simulations engines, such as EnergyPlus or TRNSYS, with climate analysis, CFD procedures, and some comfort components. 106 V. P. López-Cabeza et al. Black box models This approach, also called data-driven models, is based on statistical predictions using machine learning techniques. No physical model is required, but only large sets of input data to train a selected machine learning model to obtain the predicted outputs. Nowadays, these methods are recognized for their exceptional performance describing the overall behavior of a system based on its input–output relationship without needing any physical knowledge. Machine learning methods can be classified into supervised or unsupervised. In supervised models, the learning is based on input–output pairs to guide the process. Examples used in building simulation are the artificial neural network (ANN) or the support vector machine (SVM). In unsupervised learning methods, there is no outcome, and the goal is to describe the associations and patterns among a set of inputs. Unsupervised methods are more practical and promising in discovering novel knowledge given limited prior knowledge than supervised methods. For this reason, it is preferred in mining building operational data, when many inputs are not known. Despite a large amount of research into data-driven models, recently, its application by professionals of the building sector is more limited, and no specific simulation tool is based on this modeling approach. Still, they are a promising tool with many advantages that will be later analyzed. The following section shows some case studies using each of the previously described methods for the simulation of courtyard buildings. 3 Courtyard Simulation Studies In this section, different examples of courtyard performance simulation are analyzed. The case studies selected share the same general methodology, which is represented in Fig. 1. First, a monitoring campaign is carried out to gather weather data used as Fig. 1 General workflow methodology followed in all the case studies Outdoor Microclimate Influence on Building Performance … 107 inputs for the simulations and courtyard data to validate the simulations. Then, the simulations are performed, each case using one of the previously described tools or approaches. Then, statistical parameters are calculated in order to measure the accuracy of the simulations and validate the results. The section is structured according to the classification previously used. BES Many researchers have used building energy simulation to analyze the performance of courtyard buildings [17,18]. However, few studies have considered the courtyard as a different microclimate to outdoor conditions. This means that the benefits of the courtyard in these simulations are limited to the self-shading that the courtyard provides to the building. Recent studies have acknowledged that the courtyard has a singular microclimate, totally different to the outdoor climate, and they have found a way to include multi-nodal outdoor conditions in the simulations. Lizana et al. [5] quantified the effect of an inner courtyard microclimate on building performance using TRNSYS. The case study was analyzed with two outdoor weather conditions, an inner courtyard, and a local urban climate, as the most realistic case. Then, the building performance was compared to three single outdoor weather conditions associated with the urban climate, weather data from a rural station, and a typical year weather file. The selected case study was a dwelling unit on the sixth floor of a multifamily building in Seville, with one façade to an inner courtyard and the other to the outdoor. The outdoor and inner courtyard environments were monitored using weather stations and data loggers to get information for the microclimatic conditions. The building was numerically modeled as a multi-zone in TRNSYS. Information about geometry, external shadings, constructive elements, internal gains, infiltration, natural ventilation, thermal bridges, and internal heat capacity was required by the software. Surfaces linked to the courtyard microclimate were numerically modeled as an equivalent resistance layer with a boundary condition linked to the courtyard temperature, where courtyard solar radiative and convective gains were previously obtained and introduced per zone. An iterative calibration process was performed using the monitored air temperature data in four rooms and the standard statistical indices for model validation recommended by ASHRAE Guidelines [19] (Table 1). The results were analyzed using two building performance indicators, illustrated in Fig. 2: the percentage of indoor discomfort hours in free-running conditions (Fig. 2a) and the cooling energy demand using an idealized cooling system (Fig. 2b). Theresultsshowedthatthesimulationconsideringtwooutdoormicroclimates(the local weather file and the courtyard) mitigates severe hot hours by 88% compared Table 1 Statistical indices for model validation following ASHRAE Guidelines NMBE (%) CV-RMSE (%) R2 Bedroom 1 −2.0 (<±10%) 3.9 (<30%) 0.77 (>0.75) Bedroom 2 −0.5 (<±10%) 2.0 (<30%) 0.93 (>0.75) Bedroom 3 −1.5 (<±10%) 3.0 (<30%) 0.84 (>0.75) Corridor −1.2 (<±10%) 3.5 (<30%) 0.76 (>0.75) 108 V. P. López-Cabeza et al. Fig. 2 Discomfort hours and cooling demand results per simulated scenario in [5] to the simulation with only the local climate. The discomfort hours were reduced by 26% during the measured period. Additionally, it can be noted in Fig. 2how the comfort period in the courtyard microclimate (S1) is similar to the simulation where the rural weather file is used. These results demonstrate that the courtyard benefits the building and can mitigate extreme urban heat events. In terms of cooling energy balance, the courtyard was able to reduce the impact of urban overheating by 15%. This is very influenced by cooling-related occupant behavior patterns, as shown in the bars chart (Fig. 2b), with the possibility of a reduction up to 29% in the case of a low energy consumption pattern. Another study aimed to quantify the benefits of the courtyard in terms of cooling demand reduction in a building in Seville (Spain) [20]. This time, the simulation was performed with the Spanish BES software HULC and using a university office building with a courtyard as a case study. In this study, only two scenarios were analyzed: a reference case, in which it was assumed that the air temperature throughout the whole building envelope was the outside monitored temperature (Fig. 3a) and the proposed methodology, varying the boundary conditions in the courtyard according to the monitoring campaign (Fig. 3b), thus, considering the effect of the courtyard on air temperature. The results showed a reduction of the cooling demand in the simulated case considering the monitoring results in the courtyard that reached 10% in adjacent roomslocatedonthegroundfloorofthecourtyard.Forthewholebuilding,thecooling demand obtained was 7% in the spaces adjacent to the courtyard. The variation in cooling demand reduction according to the level of the room is because of the characteristic stratification effect of the courtyards (Table 2). The results from these studies emphasize the importance of selecting an accurate weather file in the simulation. Local climate weather file increases discomfort hours andcooling demandincomparison tothe scenariowiththe ruralweather file,whichis not considering the urban heat island effect. Moreover, the unsuitability of the typical year weather file to evaluate the climate resilience of buildings is demonstrated. The analysis of climate resilience building design requires the use of weather files that includes the extreme events that will happen more often in the future. It is also demonstrated how it is needed to include multi-nodal outdoor conditions in order Outdoor Microclimate Influence on Building Performance … 109 Fig. 3 Section of the building showing the boundary conditions considered in the reference case aand the proposed methodology, bfor thermal simulation of a building including a courtyard Table 2 Accumulated cooling demand reduction in the case study by Sanchez de la Flor et al. [20] Room Case study (kWh/m2) Reference case (kWh/m2) Absolute difference (kWh/m2) Percentage difference (%) Third floor room 1 23.67 25.74 2.07 9 Fourth floor room 1 18.38 20.17 1.79 10 Fifth floor room 1 27.62 29.27 1.65 6 Third floor room 2 15.47 15.96 0.49 3 Fourth floor room 2 16.09 16.65 0.56 4 Fifth floor room 2 17.20 17.75 0.55 3 to account for the benefits of passive strategies associated with urban microclimate strategies, such as the use of courtyards. However, this is not straightforward for two reasons. First, the introduction of different boundary conditions for the building envelopes in the simulation has been possible following different assumptions, which are even not possible in other tools. And second, in these studies, the courtyard 116 V. P. López-Cabeza et al. Fig. 8 Example of air temperature simulation results from the data-driven model by Diz-Mellado et al. [27] (Fig. 9a) corresponds to the relative error of the predicted thermal gap and is always under 8%, and the graphic on the right (Fig. 9b) corresponds to the relative error of the predicted temperature inside the courtyard, providing results under 0.1% in the two cases. The values confirm that the strategy used is quite accurate compared to other, more computationally expensive CFD model simulations. In addition, the results show more accuracy than other existing commercial tools. The root mean square error (RMSE) of this investigation is in a similar range to the values obtained with the CFD model in FreeFEM++ described before (1 °C) and lower than in other commercial tools that are around 3 °C. The new methodology proposed for the ML Fig. 9 Relative error calculated for the validation of the data-driven methodology by Diz-Mellado et al. [27] Outdoor Microclimate Influence on Building Performance … 117 method in this research is helpful for developing design tools capable of modeling the complex microclimate of semi-outdoor spaces such as courtyards without the complexity of defining a physical model. Other advantages of using ML techniques are related to the identification of fundamental variables, simplifying the calculation process, and better accuracy by including a more significant amount of training datasets. 4 Discussion of Challenges and Opportunities The previous section analyzed different approaches to simulate the thermal performance of buildings with courtyards. From that analysis, some general conclusions about the strength and weaknesses of each methodology are discussed here: •BES models are the tool commonly used to simulate the energy performance of buildings. The nodal approach allows for quick computing time when simulating dynamic evolutions throughout the whole year. However, BES models cannot be used to analyze outdoor microclimate influence on building performance if that dataarenotmanuallyincludedasboundaryconditions.ThismeansthatBESneeds to be coupled with other simulation approaches that can predict microclimates in order to measure their influence on the building performance. For that reason, the use of hybrid workflows is needed when analyzing passive strategies to enhance the resilience of buildings facing climate change. •CFD allows performing a detailed analysis of fluxes inside and outside a building on a small scale, being one option to obtain the microclimate data to be coupled in the BES simulations. However, this methodology has some weaknesses. First, these tools are difficult to implement without previous knowledge of fluid dynamics. Some of them are easier but rely on simplifications, leading to a lack of accuracy. And they require a high computation time. We have seen that it is possible to optimize the simulation model for specific simulations, but this has not been implemented yet in a tool accessible to the public, so currently, programing knowledge is also required to do this. •Data-driven models are a promising alternative to CFD models for the prediction of courtyard performance based on ML. They have another advantage, which is the possibility of being applied even without detailed information of the model. However, their accuracy depends on the quality of the training datasets, and large and representative data are required. This is a novel approach that is currently limited to researchers with programing knowledge of ML. This analysis highlighted how the best choice depends on the objectives of the analysis and the available data. For example, when speed and easily accessible feedback is needed for the early design, hybrid options using currently available tools are probably the best option. However, research is still needed to improve the connectivity among tools. On the other hand, data-driven models could be the best option if an analysis is to be done about an existing building from which there is not much 118 V. P. López-Cabeza et al. constructive information. Being this said, future research on this field should still focus on some essential challenges that are not yet overcome: •The integration of evaporative effects on the simulations is limited to very few software applications currently available for professionals, like ENVI-met. This means that the evapotranspiration effect of vegetation and the evaporative cooling of water are not being considered in many other tools, but experimental research shows that they are effective passive strategies against urban heat. For that reason, it is important to work on this issue to improve our analysis of urban mitigation strategies. •The simulations can be further optimized, aiming to reduce computational time and increase accuracy in all methods. For the data-driven models, improving the datasets to train the algorithm or developing other ML techniques applied to urban microclimates is ways of future research. For the CFD and hybrid methods, much research exists that can be applied to the development of new CFD algorithms. As an example, reduced-order modeling (ROM) is a strategy that provides reductions ofseveralordersofmagnitudeinthecomputationalcostofnumericalsimulationof parametric design processes and problems. This is an option to reduce computing time that is currently under research [28]. •Finally, although many options and tools have been presented here, better accessibility to the professional sector is still required. Some methodologies are still not incorporated into existing commercial or open-source tools accessible for architects, designers, or city planners. Others require advanced knowledge that makes them difficult to apply practically. This is a problem that needs to be addressed in order to transfer research knowledge into society. 5 Conclusion Thischapterpresentsanoverviewofthedifferentmethodstosimulateoutdoormicroclimates and their influence on building performance. They were tested, discussed, and compared through various case studies based on a very common microclimate, the traditional inner courtyard. Although the examples presented in this chapter focused on courtyards, these methodologies apply to a large variety of small-scale or medium-scale urban outdoor spaces capable of generating microclimates. The results demonstrate the lack of capabilities of current building energy simulation (BES) tools to consider the benefits of these specific microclimates on building performance. Furthermore, despite different computational fluid dynamics (CFD) software that can predict the performance of these urban environments, the high computational effort and the lack of accuracy limit its potential application for preliminary design stages. The analysis in different case studies also shows how multi-nodal outdoor climate simulations considering these microclimates can present high benefits in building performance. These specific microclimates can solve the problem of the urban overheatingeffect,forexample,mitigating peakdaytime temperatures. Theconsideration Outdoor Microclimate Influence on Building Performance … 119 of these benefits in design workflow can enable additional passive cooling strategies in the building design, and BES tools should be improved to account for them. In this sense, the potential of hybrid simulations coupling a variety of software (BESandCFD)hasbeendescribedandtested,presentinganovelpromisingapproach to combine the benefits of different methods in early design stages. However, it is still essential to enable better accessibility to the professional sector to promote a better climate change adaptation of the built environment. The optimal integration of these microenvironments in urban planning can enable new actions to mitigate climate risks in cities and buildings. From a practical view, the most important issue is to make it simpler to incor-porate the microclimate data in decision-making processes. Only then it will be possible to broader analyze urban resilience strategies under climate change projections. Acknowledgements This work was supported by the grant RTI2018-093521-B-C33 funded by MCIN/AEI/ 10.13039/501100011033 and by “ERDF A way of making Europe” and the Spanish Ministry of Education, Culture, and Sport via a pre-doctoral contract granted to V.P. L-C. (FPU17/05036) and E. D-M (FPU18/04783). The research was also supported by the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 101023241. References 1. Eurostat. Database—Eurostat. https://ec.europa.eu/eurostat/en/web/main/data/database. Accessed 14 December 2021 2. Global Alliance for Buildings and Construction. 2020 Global Status Report For Buildings And Construction Towards a Zero-Emissions, Efficient and Resilient Buildings and Construction Sector, 2020. www.globalabc.org. Accessed 14 December 2021 3. Santamouris M, Papanikolaou N, Livada I et al (2001) On the impact of urban climate on the energy consumption of building. Sol Energ 70(3):201–216. https://doi.org/10.1016/S0038-092 X(00)00095-5 4. Directive (EU) 2018/844 of the European Parliament and of the Council of 30 May 2018 amending Directive 2010/31/EU on the energy performance of buildings and Directive 2012/27/EU on energy efficiency (Text with EEA relevance). Official Journal of the European Union, 2018 5. Lizana J, Lopez-Cabeza VP, Renaldi R, Diz-Mellado E, Rivera-Gomez C, Gal C (2021) Integrating courtyard microclimate in building performance simulation to mitigate extreme urban heat impacts. Sustain Cities Soc 2021:103590. https://doi.org/10.1016/j.scs.2021.103590 6. Li Y, Neill ZO, Zhang L, Chen J, Im P, Degraw J (2021) Grey-box modeling and application for building energy simulations—a critical review. 146 7. CoakleyD,RafteryP,KeaneM(2014)Areviewofmethodstomatchbuildingenergysimulation models to measured data. Renew Sustain Energ Rev 37:123–141. https://doi.org/10.1016/j.rser. 2014.05.007 8. Hamdaoui M-A, Benzaama M-H, El Mendili Y, Chateigner D (2021) A review on physical and data-driven modeling of buildings hygrothermal behavior: models, approaches and simulation tools. Energ Build 251:111343. https://doi.org/10.1016/j.enbuild.2021.111343 9. M’Saouri El Bat A, Romani Z, Bozonnet E, Draoui A (2021) Thermal impact of street canyon microclimate on building energy needs using TRNSYS: a case study of the city of Tangier in Morocco. Case Stud Therm Eng 24:100834. https://doi.org/10.1016/j.csite.2020.100834 120 V. P. López-Cabeza et al. 10. Yang X, Zhao L, Bruse M, Meng Q (2012) An integrated simulationmethod for building energy performance assessment in urban environments. Energ Build 54:243–251. https://doi.org/10. 1016/j.enbuild.2012.07.042 11. Natanian J, Auer T (2020) Beyond nearly zero energy urban design: a holistic microclimatic energy and environmental quality evaluation workflow. Sustain Cities Soc 56. https://doi.org/ 10.1016/j.scs.2020.102094 12. Perini K, Chokhachian A, Dong S, Auer T (2017) Modeling and simulating urban outdoor comfort: coupling ENVI-Met and TRNSYS by grasshopper. Energ Build 152:373–384. https:// doi.org/10.1016/j.enbuild.2017.07.061 13. Mackey C, Galanos T, Norford L, Roudsari MS (2017) Wind, sun, surface temperature, and heat island: critical variables for high-resolution outdoor thermal comfort. In: Proceedings of the 15th IBPSA conference San Francisco, CA, USA, pp 985–993. https://doi.org/10.26868/ 25222708.2017.260 14. SoflaeiF, ShokouhianM,TabadkaniA, MoslehiH, BerardiU (2020)Asimulation-based model forcourtyardhousingdesignbasedonadaptivethermalcomfort.JBuildEng31:101335.https:// doi.org/10.1016/j.jobe.2020.101335 15. EvolaG,Costanzo V,Magrì C,MarganiG,Marletta L,NaboniE(2020) Anovelcomprehensive workflow for modelling outdoor thermal comfort and energy demand in urban canyons: results and critical issues. Energ Build 216:109946. https://doi.org/10.1016/j.enbuild.2020.109946 16. Elwy I, Ibrahim Y, Fahmy M, Mahdy M (2018) Outdoor microclimatic validation for hybrid simulation workflow in hot arid climates against ENVI-met and field measurements. Energ Procedia. https://doi.org/10.1016/j.egypro.2018.10.009 17. Asfour OS (2020) A comparison between the daylighting and energy performance of courtyard and atrium buildings considering the hot climate of Saudi Arabia. J Build Eng 30:101299. https://doi.org/10.1016/J.JOBE.2020.101299 18. Soflaei F, Shokouhian M, Abraveshdar H, Alipour A (2017) The impact of courtyard design variants on shading performance in hotarid climates of Iran. Energ Build. https://doi.org/10. 1016/j.enbuild.2017.03.027 19. ASHRAE (2014) ASHRAE Guideline 14-2014. Measurement of energy, demand, and water savings. https://energywatch-inc.com/ashrae-guideline-14/ 20. Sánchez de la Flor FJ, Ruiz-Pardo Á, Diz-Mellado E, Rivera-Gómez C, Galán-Marín C (2021) Assessing the impact of courtyards in cooling energy demand in buildings. J Cleaner Prod 320:128742. https://doi.org/10.1016/j.jclepro.2021.128742 21. Rojas JM, Galán-Marín C, Fernández-Nieto ED (2012) Parametric study of thermodynamics in the mediterranean courtyard as a tool for the design of eco-efficient buildings. Energies 5(7):2381–2403. https://doi.org/10.3390/en5072381 22. Toparlar Y, Blocken B, Maiheu B, van Heijst GJF (2017) A review on the CFD analysis of urban microclimate. Renew Sustain Energ Rev 80:1613–1640. https://doi.org/10.1016/j.rser. 2017.05.248 23. López-CabezaVP,Galán-MarínC,Rivera-GómezC,Roa-FernándezJ(2018)Courtyardmicroclimate ENVI-met outputs deviation from the experimental data. Build Environ 144:129–141. https://doi.org/10.1016/j.buildenv.2018.08.013 24. Huttner S (2012) Further development and application of the 3D microclimate simulation ENVI-met. Johannes Gutenberg-Universitat in Mainz, Mainz. http://ubm.opus.hbz-nrw.de/vol ltexte/2012/3112/ 25. Lopez-Cabeza VP, Carmona-Molero FJ, Rubino S et al (2021) Modelling of surface and inner wall temperatures in the analysis of courtyard thermal performances in Mediterranean climates. J Build Perform Simul 14(2):181–202. https://doi.org/10.1080/19401493.2020.1870561 26. López-Cabeza VP, Diz-Mellado E, Rivera-Gómez C, Galán-Marín C, Samuelson HW (2022) Thermal comfort modelling and empirical validation of predicted air temperature in hotsummer Mediterranean courtyards. J Build Perform Simul 15(1):39–61. https://doi.org/10. 1080/19401493.2021.2001571 Outdoor Microclimate Influence on Building Performance … 121 27. Diz-Mellado E, Rubino S, Fernández-García S, Gómez-Mármol M, Rivera-Gómez C, Galán- Marín C (2021) Applied machine learning algorithms for courtyards thermal patterns accurate prediction. Mathematics 9(10). https://doi.org/10.3390/math9101142 28. Azaïez M, Chacón Rebollo T, Gómez Mármol M et al (2021) Data-driven reduced order modeling based on tensor decompositions and its application to air-wall heat transfer in buildings. SeMA 78:213–232. https://doi.org/10.1007/s40324-021-00252-3 This thesis develops a new simulation methodology applied to the microclimate of courtyards to understand their thermodynamic performance considering the different factors involved. Using the new methodology, different strategies to improve the performance of the courtyard as a passive element to reduce the energy consumption of buildings and improve user comfort are analyzed. This work is structured into five chapters. Chapter 1 presents the state of the art and the objectives, methodology, and structure of the thesis. Chapter 2 summarizes the findings of this research about existing tools applied to the simulation of the microclimate of courtyards and the development and validation of the novel workflow to their simulation, one of the main contributions of the thesis. Chapter 3 shows the results of using the methodology previously described in the analysis of the effectiveness of different strategies to improve the thermal performance of courtyards. Geometry, material properties such as albedo or thermal inertial or the inclusion of shading devices are simulated. Chapter 4 presents the linkage of the methodology of simulation of the courtyard to the analysis of the indoor comfort of buildings and energy demand using existing building energy simulation tools. Finally, Chapter 5 discusses the main results and conclusions, the contributions of this work, and some recommendations for professionals in the sector of building design. The results of this thesis have been published in several high-impact journals. Victoria Patricia López-Cabeza [email protected] Doctorado en Arquitectura Instituto Universitario de Arquitectura y Ciencias de la Construcción (IUACC) Universidad de Sevilla Sevilla, 2023