INTRODUCTION All: Please adapt intro as you see fit. The Intergovernmental Panel on Climate Change (IPCC) stated in their last report that anthropogenic greenhouse gases are responsible for the current climate change. As such urban areas are responsible for more than 70% of the emissions with over half of the world population living in urban areas. It is hence crucial to develop more sustainable urban areas that will significantly reduce the carbon footprint of cities while at the same time taking into account the rising temperatures and the vulnerability of the urban spaces. Multiple tools have been developed in the recent decades for a broad range of applications to address sustainability in urban areas. Meteorological models (for example WRF , MESO-NH ) have been improved to include parameterizations such as BEP-BEM or TEB or UCM , that would better show the influence of urban areas on meteorological variables. These development were an important leap for the representation of the urban heat island phenomena . However, these models do not have a sufficiently high horizontal or vertical resolution to describe properly buildings in urban areas. Other models such as CitySim or Envi-Met provide an enhanced description of buildings. They are often use to analyse the energy consumption of buildings. However, although Envi-Met gives a simulation of the flow around buildings, it cannot be run over a full year to have a full evaluation of the energy consumption of these buildings. CitySim on the contrary can be run for the 8760 hourly time-steps, but it does not take into account the flow and hence lack a proper description of the micro-climate so particular to urban areas. In the recent years, some attemps have been made to couple meteorlogical models with building energy models. CFD models have been used to provide a better description of the flow around buildings . These models require nonetheless significant computational resources and are thus not practical for the evaluation of urban planning scenarios which thus becomes a tedious task. Mauree et al., has thus developed an urban canopy model, CIM to provide high-resolution vertical profiles to building energy models. In a previous study they validated the coupling of CIM with CitySim and demonstrated the advantage of the coupling int he simulation of building energy use in an urban district . There is however still a need to evaluate the performance of the coupling for producing future climatic scenarios adapted to urban areas and to determine the relevance of using the multi-scale coupling to provide useful information to urban planners. The paper is structured as follows. In Section [mandm], a description of the different tools used is given. We explain how the climatic data was generated, how it was used as input for CitySim to calculate the surface temperatures and then used as boundary conditions for CIM. We then show in the Section [res] how the wind speed and air temperature differ in an urban context and how relevant using local climatic data is in the evaluation of energy consumption. We consequently demonstrate the resiliency of the buit areas with a refurbishment scenario for the future climate. Finally in Section [disc], we discuss the implications of the simulations and the results on the energy system sizing as well as on the urban design. MATERIAL & METHODS In this section a brief description of the different methodologies used to create the dataset for the energy simulations tools as well as the building energy model is given. Fig [schema] illustrates the process for the simulation of the energy consumption at the district scale. [Schematic illustration of the simulations] Preparation of climate files for future scenarios To be completed by Vahid. CitySim CitySim is an urban energy modelling tool , able to quantify the energy demand from the building to the city scale. The thermal model of buildings is based on an analogy with the electrical circuit, or more precisely on a simplified resistor-capacitor network . The radiation model, previously validated with Radiance, is based on the Simplified Radiosity Algorithm (SRA). With the SRA, the radiant external environment is represented by two hemispheres, discretized into several solid angles . CitySim provides the energy needs of buildings, as well as the electricity demand and the energy produced by renewable energy sources. Results obtained by the software were previously validated with the BESTEST, showing a sound correlation between them . CitySim works dynamically, providing the results in hourly values, and by including the interactions within the built environment. Among other, the inter-reflections between buildings’ surfaces as well as the mutual shading are calculated. In order to perform the calculations, hourly weather data are required, such as those generated by the software Meteonorm , or by on-site monitoring. Recent development of the model considers the inclusion of the microclimatic conditions, by calculating the evapotranspiration from the ground and the impact of greening on the outdoor human comfort . CIM CIM is an urban canopy model that can be used in an offline mode to provide high resolution data for building energy simulation tools . It has already been coupled with CitySim to take into account the particularities of urban areas, to improve building energy simulations. CIM is a column module where the Navier-Stokes equations are reduced in one-dimension. Flow is resolved for the two components in the horizontal direction and also the the air temperature along the vertical axis. {dt}={dz}\left({\mu }_t{dz}\right)+f^s_u {dt}={dz}\left({\kappa }_t{dz}\right)+f^s_{\theta }, where u is the mean horizontal velocity (m s−1), θ is the potential temperature (K), μt and κt are the momentum and heat viscosity coefficients (calculated using a 1.5 turbulence closure) and fus and fθs are the source terms representing the fluxes that will impact the flow. Additionally, CIM resolves its own equation for the turbulent kinetic energy providing an enhance description of turbulent flow over complex terrain while not significantly using computational resources. More details can be found in Mauree et al., . Study case The campus of EPFL is chosen as the study case. Covering an area of 55 ha, the campus is a comparable to an urban area with over 10,000 students and 5000 staff members. The energy simulations are done for the existing EPFL campus, in Lausanne, Switzerland (see Fig [epfl_campus]) to quantify the impact of the changing climate on the energy consumption of the built stock and on the importance of accounting for the urban climate. Hence two set of scenarios are run: (i) using the climatic data and (ii) using the Climate-CIM-CitySim data. [MAP OF THE EPFL CAMPUS, SWITZERLAND (EXTRACTED FROM PLAN.EPFL.CH) This image is taken from Open Street Map whose copyright notices can be found here: (CC-BY-SA-2.0).] We present here the setup of the model, as well as the physical characteristics of the campus. The energy model of the campus was set up in a previous study : hourly heating and cooling demand of the site were provided by the software CitySim, and validated with on-site monitoring. The geometrical information of the campus was obtained from Carneiro , and the physical data of the buildings were defined according to the phase of construction. Renovation scenarios As stated above, the objective of this paper is to understand the impact of climate change and of the local climatic conditions on the energy demand of buildings, as well as its sensitivity as a function of the envelope characteristics. Simulations of a hypothetical refurbishment of the university campus according to the high energy efficiency standard Minergie-P were performed . Minergie is a well-established standard, commonly applied to the Swiss construction market; a further improvement of this standard is the so called Minergie-P, which implies a lower energy demand. In order to apply the standard, all buildings are well insulated with 35cm of EPS and triple glazing with infrared coating. The novelty in the proposed approach is the fact that the Minergie standard is applied to an entire campus, not only to one building and the simulations are performed for the end of the century. RESULTS Table [tabscen] summarizes the simulations that have been run for the different climatic scenarios. The results from these runs are described hereafter. Years Metenorm CIM Renovation --------------------- ---------- ----- ------------ 1990-2010 X X X 2039 X X 2069 X X 2999 X X X : Set of scenarios Analysis of the future climate in an urban context A first simulation is done with the typically used dataset obtained from Meteonorm. The wind speed and air temperature are averaged climatic values (from 1990-2010) for the location of Ecublens. Figure [meteonorm_cim] shows values for each time step through the year for the data from Meteonorm and the one produced from CIM. 0.75 [Changes between Meteornorm and CIM dataset] 0.75 [Changes between Meteornorm and CIM dataset] --------------------- ---------- ------ ---------- -------- Metenorm CIM Metenorm CIM Mean 1.94 0.37 10.28 9.92 St. Dev. 1.94 0.48 7.74 9.97 Min. 0.00 0.02 -9.50 -14.30 Max. 16.2 4.74 30.00 40.60 --------------------- ---------- ------ ---------- -------- : Statistical analysis from the Meteonorm and CIM dataset Table [ustats] summarizes the statistical analysis conducted for these two datasets. It is clear that there is a notable difference between these two scenarios. For example the mean wind speed is decreased from 1.94 (m s−1) to 0.37 (m s−1) while for the air temperature there is a decrease in from 10.3 to 9.9 (°C). It should be highlighted that although there is an annual decrease in the air temperature, there is a significant increase (>10 °C) in the maximum temperature when considering the local environmental climatic conditions. 0.75 [Change in air temperature for 2039, 2069 and 2099] 0.75 [Change in air temperature for 2039, 2069 and 2099] 0.75 [Change in wind speed (m s−1) for 2039, 2069 and 2099] 0.75 [Change in wind speed (m s−1) for 2039, 2069 and 2099] Figures [T-future] and [U-future] shows the monthly averaged temperatures and wind speed respectively for the three climatic scenarios for the future. The air temperature increase is clear both with and without CIM, with a slightly higher rise during the summer periods. The same trends as with the Meteornorm data, can be seen for the future climate. With the CIM’s simulation there is on average a decrease in the mean annual temperature but for the maximum temperature there is a notable rise (0.6°C in 2039, 0.7°C in 2069 and 0.2°C in 2099). There are no clear trends for the wind speed when looking at the change in the future for the monthly mean values. It can nonetheless be noted that the wind speed in the 2039 scenarios appears to be higher during the winter time as compared to the other two cases. A The impact of considering the micro-climate on the the energy demand as was demonstrated by Mauree et al., , will be explained further in particular with respect to the future climate in the next sections. Energy consumption for the EPFL campus To be completed by Silvia Renovation scenario To be completed by Silvia DISCUSSIONS AND CONCLUSIONS Impact of considering the urban climate To be completed by Das Energy system design To be completed by Dasun Improved urban design To be completed by Emanuele Future transition pathways Will be completed by Das DISCLOSURE/CONFLICT-OF-INTEREST STATEMENT The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. ACKNOWLEDGEMENT The authors would like to thank the CTI for the funding of the SCCER Future Energy Efficient Buildings and Districts – FEEB& D (CTI.2014.0119). SUPPLEMENTAL DATA

Dasaraden Mauree

and 2 more

Abstract Introduction (ATD) Energy requirements in urban areas are rising very fast and this trend is expected to continue with the increase in the urban population (70% by 2050). New buildings are built according to more and more stringent norms which can make them either net-zero or energy positive buildings. However, the core of the existing buildings is still old and renovation strategies and scenarios will not be able to make neighbourhoods or buildings in urban areas fully autonomous. Building optimization has been amply discussed in recent literature considering various aspects of the building especially considering the energy flow [1]. Lately, this has been extended considering building energy systems along with the various aspects of the building including location of windows, thickness of the walls, orientation etc. [2], [3]. However, this is a challenging due to its computationally demanding nature. When moving from building scale to neighbourhood scale, the complexity of the problem increases significantly [4].   A number of recent studies have focused on combining energy system sizing problem with building simulation in both urban and neighbourhood scale [5], [6]. However, most of these studies do not consider the impact adjacent buildings on the thermal and electricity demand (due to lighting). Effect of shadowing and boundary layer is not considered in most of the instances. The proper representation of buildings and their effects such as drag force effects, generation of turbulence etc. are crucial in the evaluation of building energy demand [7] as they can impact the convective heat transfer coefficient [8]. Hence, it is important to represent micro-climate accurately in the building simulation process which will be subsequently used for energy system sizing. More importantly, appropriate representation of micro-climate can be used along with building simulation and energy system sizing to optimally locate buildings with less thermal demand and more opportunity to integrate renewable energy technologies. This can lead to sustainable neighbourhoods with lower carbon impact.                                      In order to fill this gap, in this study we focus on combining building simulation model [9], [10], energy system sizing tool and an urban meteorological model [11] in order to assess the impact of building positioning on energy system designing in neighbourhood scale. We analyse the influence of using local climatic data and architectural aspects on energy demand of the building and subsequently to the energy system sizing for energy sustainable buildings. In the next section we describe the models used in the current study. We then show the results obtained and discuss them. In the final section we conclude on the most important findings in this study and give a few perspectives to further develop the simulation platform.   Overview of the computational platform and case studies (ATD) Computational model for urban micro climate and building simulation (Das & Silvia)The coupling of micro-scale model (such as CitySim) with meteorological models is essential to represent the impact of buildings on the climatic variables and to provide enhanced building energy simulation. Phenomenas such as the Urban Heat Island \cite{OKE_1982} are not represented in TMY or Meteonorm dataset since they are usually collected outside of the city. These data then need to be transformed to account for the specifities of the urban climate and to provide useful data to building energy models. This is why it is proposed here to use the CIM-CitySim coupled models and to extend it further.CIM - CitySim CIM is a 1D meteorological model \cite{Mauree_2017} that can work offline as a stand-alone module while using as input data from the climatic dataset (such as Meteonorm  \cite{Remund} ) or can be coupled with a 3D meteorological model (such as WRF ). CIM calculate high resolution vertical profiles of the variables considering the local environment (for example considering the presence of buildings and their density). The meteorological model is used to generate profiles wind speed, direction and air temperature profiles. CIM uses a diffusion equation derived from the Navier-Stokes equations but reduced to one direction only. It has subsequently been coupled with the CitySim building simulation software (see Figure 1) in order to determine the energy demand of a district \cite{Mauree_2017a}\cite{scartezzini2015}. Equations are taken from \citet{Mauree2017} to account for the obstacles density and height in the canopy. The coupling of CitySim and CIM provides enhanced boundary conditions for both models. On the one hand, CIM gets a calculation of the surface temperature from CitySim, which it then uses to simulate the flow in the column module and recalculate high resolution vertical profile of meteorological variables, such as the air temperature and the wind speed. On the other hand, CitySim is provided with local meteorological data. The modification of the variables influences two main processes that are computed by CitySim: the convective heat transfer coefficient (\(h_c\) \((W m^{-2}K^{-1})\)) given by:\(h_c=2.8+3U \) where \(U \) is the wind speed.  \(h_c\) is then used in the calculation of the flux \(Q_{ch}\):\(Q_{ch}= h_c (T_s-T_{air})\)where \(T_s \) is the surface temperature in K and \(T_{air}\) is the air temperature in K.the longwave exchanges. Indeed the longwave is calculated as a function of difference between the environmental temperature (\(T_{env} \)) and the surface temperature. The longwave flux \(Q_{lw}\) can be computed using : \(Q_{lw}=h_r (T_{env} - T_s)\)where \(h_r \) \((W m^{-2}K^{-1})\) is \(h_r= \epsilon_s \sigma \frac{1}{2}(T_{env}-T_s)\)where \(\epsilon\) is the emissivity of the surface and \(\sigma\) is the Stefan-Boltzmann constant.Figure \ref{211596} shows the flowchart used in the coupling between CIM and CitySim.