Signature of Author:
Department of Architecture
May 8, 2015
Certified by:
Leslie K. Norford
Professor of Building Technology
Thesis Supervisor
Accepted by:
Takehiko Nagakura
Associate Professor of Design and Computation
Chair of the Department Committee on Graduate Students
2
Urban Weather Generator (UWG) is the urban design simulation tool that provides
climate-specific advice for cityscape geometry and land use to assist the development of
energy-efficient cities that are also thermally comfortable. The software enables urban
designers to parametrically test built densities for masterplanning and urban planners to
advocate zoning regulations such as building height and land use as well as policies for
traffic intensity with energy and thermal implications of these interventions. UWG is the
first tool publicly available that incorporates microclimatic considerations in urban design
and energy simulations.
The project succeeds the work of Bueno et al. (2014) to develop a useful and accessible
urban design tool to model urban heat island effect (UHI) from measurements at an
operational weather station based on neighborhood-scale energy balances. The sensitivity
analyses for Boston, MA, USA, and Punggol, Singapore identify as key parameters the
building morphologies such as site coverage ratio and façade-to-site ratio; building surface
albedo and emissivity; and sensible anthropogenic heat in the urban canyon. The
consistency of results for these cities reduced required user inputs to the model by 46%
without decreasing the simulation accuracy.
The developed software is available as a stand-alone tool as well as a new plug-in for
the Rhinoceros-based urban modeling interface (umi) to integrate the microclimate
analysis in the formal design process. The graphical user interface is written in
programming language C# in the Microsoft .NET platform and is available free of charge at
http://urbanmicroclimate.scripts.mit.edu/uwg.php.
The newly proposed workflow for energy- and thermal comfort-driven urban design
and planning is demonstrated through a case study of the new 130 thousand square meter
development on the MIT East Campus in Cambridge, MA, USA. An IPCC-based climate
change prediction is considered along with UHI to evaluate the proposed massing models
at each design phase to ensure thermally comfortable urban development along the way.
Thesis Supervisor: Leslie Keith Norford, PhD
Title: Professor of Building Technology
3
Leslie Keith Norford, PhD
George Macomber Professor of Construction Management
Professor of Building Technology
MacVicar Faculty Fellow
Thesis Advisor
Christoph Reinhart, PhD
Associate Professor of Building Technology
Thesis Reader
4
This research is funded by the National Research Foundation (NRF) through the SingaporeMIT Alliance for Research and Technology Center for Environmental Sensing and Modeling.
This project would not have been possible without the help from many people who
willingly contributed their time and knowledge into improving the tool. Since the beginning
of my Master’s degree, my goal has been to create a useful tool to promote the integration
of engineering and design. I am happy to report that with their help I have been able to
create and distribute the tool that is already being used by many practitioners and
researchers.
I would like to first thank my advisor Dr. Les Norford for his guidance and insights
throughout the last two years. I appreciate your pedagogy to always think about the big
picture and your encouragement to pursue my interests. A big thank you also goes to Dr.
Christoph Reinhart for helping me shape the scope of this project in relation to other
building science and urban design tools. I also would like to thank Bruno Bueno for your
patience in helping me understand your algorithm for UWG. Lingfu Zhang and Bokil LopezPineda made possible many functionalities of the user interface, and Cody Rose was
instrumental in implementing the Rhino integrated version of the software.
Additionally, the tool improvements are based on constructive feedbacks from many
people including Hope Stege and Celina Guzman from the MIT Center for Advanced
Urbanism as well as my colleagues at the Sustainable Design Lab.
Thank you to everyone at the Sustainable Design Lab and the Building Technology Lab for
your feedbacks and friendships. A special thanks to Julia Sokol, Manos Saratsis, and Denise
Rivas for working with me on the project “Sustainable Urban Development: Modeling
Future Growth Scenarios” that motivated the MIT East Campus case study in this work.
Ariel Noyman helped me understand the urban design process behind the Kendall - East
Campus development. Manos also helped me understand the general urban design process
and helped develop my graphic design skills.
The MIT East Campus case study is my tribute to MIT for providing me with so many longlasting friendships and learning opportunities during my six years for my undergraduate
and graduate studies. It is the people who make this place the best. I hope this work will
contribute towards shaping a sustainable future for MIT and beyond, by always putting
5
people and their comfort first. Many inspirations for the case study design developments
come from my experiences at climate engineering firm Transsolar and architectural design
studio Höweler + Yoon Architecture. Meejin, Eric, and Erik, thank you for your guidance.
Finally I would like to thank my parents, my brother, and friends for their unwavering love
and support. Jason, thank you for always being there for me. I’m so glad I shared my
proudest days with you and I had your humor to ride through the toughest moments of
graduate school. You and Hunter bring so much joy to my life.
6
Chapter 1
Introduction: Purpose and Scope of Study ................................................................................. 11
1.1
1.2.
1.3.
Context................................................................................................................................................ 11
1.1.1.
1.1.2.
Rapid urbanization ....................................................................................................................................................11
Urban heat island effect (UHI) .............................................................................................................................12
Need for Microclimatic Design Tool.......................................................................................... 14
1.2.1.
1.2.2.
Current urban design and planning processes .............................................................................................14
Proposing microclimatic urban design ............................................................................................................15
1.3.1.
1.3.2.
Methodology ................................................................................................................................................................16
Scope of study and structure of this thesis .....................................................................................................16
Research Goal: Urban Energy and Thermal Comfort Software Development ........... 16
Chapter 2
Literature Review: State-of-the-Art Simulation Tools for Modeling Urban Heat Flows
and Integrated Design ....................................................................................................................... 18
2.1
Weather Morphing Models for Capturing UHI ...................................................................... 18
2.1.1
2.1.2
2.1.3
2.1.4
2.1.5
2.2
2.3
Mesoscale atmospheric modeling and Computational Fluid Dynamics approaches ...................18
Crawley’s temperature alteration algorithm .................................................................................................19
Urban canopy models as urban-climate prediction tools ........................................................................19
Urban Weather Generator .....................................................................................................................................19
[Lack of] weather morphing tools in the urban design context ............................................................20
Integrated Design Simulation Tools ......................................................................................... 21
2.2.1
2.2.2
2.2.3
Current trend in integrated tools ........................................................................................................................21
Rhinoceros-based tool: umi ..................................................................................................................................22
Grasshopper-based tool ..........................................................................................................................................22
UWG as integral component of family of design tools ........................................................ 23
Chapter 3
Platform and Engine for the Simulation: Urban Weather Generator, umi, and
EnergyPlus ............................................................................................................................................. 24
3.1
3.2
3.3
3.4
UHI Modeling Engine: Urban Weather Generator .............................................................. 24
3.1.1
3.1.2
3.1.3
3.1.4
3.1.5
Model description ......................................................................................................................................................24
Required inputs ..........................................................................................................................................................26
Model validation in Singapore .............................................................................................................................27
Applicability as a design tool ................................................................................................................................29
Limitations ....................................................................................................................................................................30
City-Scale Energy Simulation Platform: umi .......................................................................... 30
Energy Simulation: EnergyPlus .................................................................................................. 33
System Integration and New Workflow: Single Workflow in Rhinoceros .................. 33
7
Chapter 4
Towards a Usable Design Tool: Identification of Key Design Parameters ...................... 35
4.1.
4.2.
4.3
4.4
Goal of the Sensitivity Analysis................................................................................................... 35
Methodology ...................................................................................................................................... 35
4.2.1.
4.2.2
4.2.3
Parameter selections and extraction processes for UWG ........................................................................35
IDF File creation for EnergyPlus .........................................................................................................................41
Evaluation metrics .....................................................................................................................................................41
4.3.1
4.3.2
4.3.3
Boston simulation results ......................................................................................................................................43
Singapore simulation results ................................................................................................................................43
Analysis of the combined results ........................................................................................................................43
Key Parameters for UHI ................................................................................................................ 43
Implications of the Sensitivity Analysis for Creating the Design Tool ......................... 50
Chapter 5
Design Tool Development: Stand-alone and Rhino Integrated Design Tool .................. 51
5.1 Overview of the User – Centric Design and Design Process .................................................... 51
5.2
User Interface Design .................................................................................................................... 52
5.3
5.2.1
5.2.2
5.2.3
5.2.4
5.2.5
Software development environment and C# programming language ...............................................52
First prototype: testing of user interaction and designing proper interface ..................................52
Current user interface design ...............................................................................................................................57
Integrated Design Tool: Rhino Version ............................................................................................................68
Results viewer .............................................................................................................................................................70
5.3.1
5.3.2
5.3.3
Morphed diurnal dry bulb temperature ..........................................................................................................70
Outdoor thermal comfort metric: Universal Thermal Climate Index .................................................71
Heating and cooling energy ...................................................................................................................................75
Evaluation Metrics ......................................................................................................................... 70
5.4 Proposed New Workflow Using the Tool ....................................................................................... 75
Chapter 6
Case Study: Demonstration of the Tool ....................................................................................... 76
6.1. Design Brief: MIT East Campus development, Cambridge, MA............................................. 76
6.1.1
6.1.2
6.1.3
Context and MIT 2030 initiative .........................................................................................................................76
Zoning and total build capacity............................................................................................................................79
City of Cambridge Sustainable Design Initiatives ........................................................................................80
6.2.1
6.2.2
6.2.3
Concepts .........................................................................................................................................................................80
Simulation workflow using the UWG and Energy components of umi ..............................................91
MIT plan simulation results ..................................................................................................................................95
6.2. MIT’s Urban Design Plan .................................................................................................................... 80
6.3
6.4
Integrated Design using the Urban Weather Generator .................................................. 97
6.3.1
6.3.2
6.3.3
6.3.4
6.3.5
Urban design concept ...............................................................................................................................................97
Basic climate analysis to determine design strategy .................................................................................97
Formal design and simulation workflow ..................................................................................................... 101
Simulation process, results, and recommended urban design ........................................................... 102
Application: Holistic evaluation of urban heating with UHI and climate change ....................... 117
Comparison of the current and proposed urban design workflows .......................... 121
Chapter 7
Summary of Contributions and Discussion ............................................................................. 122
7.1
7.2
Development of Microclimatic Design Tool ........................................................................ 122
Notes on Limitations and Future Work ................................................................................ 123
References
8
Appendix
A.1
B.1
B.2
C.1
D.1
D.2
Glossary of Terms .......................................................................................................................... 133
Sensitivity Analysis: Singapore Parameters ........................................................................ 134
Singapore IDF Parameters ......................................................................................................... 135
Design Tool: Sample Input Xml File ........................................................................................ 135
Ordinance Number 1355 Section 13.81 (2013) ................................................................. 140
Updated MIT East Campus design released on April 14, 2015 ..................................... 140
9
10
In 2007, more than fifty-percent of the human population was living in cities that are
trending towards increasing population densities. In China, 51.3% of the population lived
in major urban centers in 2011 and other developing countries are projected to follow suit
(Figure 1). Furthermore, over 250 million rural Chinese inhabitants are planned to move to
urban areas over the next fifteen years [Johnson, The New York Times, 2013]. This means
that in China alone we need to design and build for an influx equivalent to the current
urban population in the US within the next two decades.
The overall density is projected to increase in all major areas except Europe by 2050,
and the increases are substantial for Asia and Africa [United Nations Department of
Economic and Social Affairs, 2004] (Figure 2). With buildings responsible for 40% of
carbon emission, we need a tool to strategically design cities to support this rapid urban
growth.
Figure 1
Over 50% of Chinese population now live in urban areas. Southeast Asia and India will be
following the trend in the next 30 years. (Figure from The Economist Online (2012))
11
Figure 2
Density for major areas (1950 - 2300) (Figure 44 from the United Nations Department of
Economic and Social Affairs, 2004).
As cities develop, open spaces are filled with tall buildings and form canyons with dense
and extruded blocks of structures (urban canyon) and natural terrains are replaced with
artificial building materials. These modifications in urban morphology and surface
materials lead to warmer temperatures in cities than in rural areas at night, a phenomenon
known as urban heat island effect (UHI). UHI tends to be most intense at the center of the
city and has a diurnal pattern, reaching minima in the later afternoon and maxima during
the night (Figure 3). The thermal modification can be as high as 12 °C [Oke, 1987]. This
behavior has been observed in numerous field studies around the world, including Nanjing,
China [Huang et al., 2008], London, U.K. [Kolokotroni et al., 2012] and for a variety of
climate regions [Crawley, 2008].
The diurnal cycle of the air temperature is delayed in cities because urban surfaces tend
to have higher volumetric heat capacity. These materials efficiently absorb shortwave
radiation and decrease the convective heat removal due to reduced mean wind velocity
12
caused by surface roughness. Furthermore, decreased vegetation reduces evaporative
cooling and tall urban canyons block the nighttime longwave exchange to the relatively
cold sky (i.e. reduced sky view factor). As a result, more heat is absorbed and retained in
the urban than rural terrain. Added to this is the heat gain due to human activities such as
from transportation and heat sources in buildings [Sailor, 2011]. The heat island intensity
is proportional to the logarithm of population size [Oke, 1987] and is accelerated with the
current trend in urban population growth.
In general, UHI increases overall electricity demand and peak demand, which generally
occurs on hot summer weekday afternoons when offices and homes are running cooling
systems, lights, and appliances.
The UHI has contrasting energy implications for different climates. Warmer summers
imply increased use of air conditioning and thus cooling energy consumption, while
Figure 3
Typical diurnal pattern of the (a) urban and rural air temperatures, (b) cooling and warming
rates, and (c) the resulting heat island intensity (Figure 8.14 from Oke, 1987)
13
warmer winters may reduce heating needs. UHI creates a vicious cycle in tropical cities
because air conditioners release heat into the urban canyon and increase the outdoor
temperatures, thus further increasing the cooling demand. On the other hand, cold cities
with moderate summers such as Boston may be helped by their dense urban landscapes to
reduce their heating energy consumption. Crawley (2008) showed that the energy use is
reduced by 10% or more in cold climates, increased by 20% for cooling needs in tropical
climates, and reduced by 25% in heating and increased by 15% in cooling in temperate
mid-latitude climate as a result of the UHI.
In order to meet the increased demand for air conditioning in the summer, electricity
companies rely on electricity generation by power plants leading to higher emissions of
sulfur dioxide, carbon monoxide, nitrous oxides, and suspended particulates, as well as
carbon dioxide, a greenhouse gas known to contribute to global warming and climate
change [Gorsevski et al., 1998].
Summer heat islands often accelerate the formation of harmful smog, as ozone
precursors such as nitrous oxides (NOx) and volatile organic compounds (VOCs) combine
photochemically to produce ground level ozone [SOS, 1995].
Warmer days and nights along with higher air pollution levels can contribute to general
discomfort, respiratory difficulties, heat cramps and exhaustion, non-fatal heat stroke, and
heat-related mortality. Heat islands can also exacerbate the impact of heat waves, which
are periods of abnormally hot, and often humid, weather. Sensitive populations, such as
children, older adults, and those with existing health conditions, are at particular risk from
these events.
A study in Shanghai [Tan et al., 2010] (subtropical climate) shows that in the 1998 heat
wave, the excess mortality rate in the urban area is higher than that in the exurban districts
(27.3/100,000 compared to 7/100,000). A comparison between excess deaths and the
spatial coverage of the heat wave shows that the extent of high temperatures played an
important role in the number of excess deaths for that year as well as for another heat
wave in 2003.
The last section discussed how UHI can influence the livability of the urban
environment including public health as well as heating and cooling energy consumption of
buildings. Unfortunately, neither urban planners and designers nor energy consultants
currently have tools or methods at their disposal to incorporate this effect into their design
of a new or renovated neighborhood.
The current urban design and planning processes are described below according to
Besserud and Hussey (2011). The masterplanning process is rather linear and begins with
14
the design of urban morphology and land use by a community of urban planners, urban
designers, and architects.
Urban planner
The urban planner sets regulations for zoning and building codes
and top-level land use requirements.
Urban
designer
The urban designer then takes these inputs to masterplan the
neighborhood in the following order:
1. layout of the street grid, which is the primary organizational
element that helps to establish the flows of vital infrastructural
systems such as energy, water, data, and waste systems
2. location for public spaces such as parks and plazas
3. subdivision and organization of space into districts and
neighborhoods including proposals for city centers. The
process of defining neighborhoods include target land use and
type distribution within parcels
4. layout of the public transit to create a robust and convenient
transit
5. development densities defined by floor area ratio and
architectural massing. The massing designs are often based on
building regulations such as setbacks and maximum heights
without considerations for changes in microclimatic conditions
6. adjustments and refinements to previous steps as well as
additional agendas such as sustainability
Architect +
Engineer
Architects work on individual building schemes with engineers to
establish each specific building massing and system
Approval
The design proposals are presented to different stakeholders
including real estate developers and governmental regulatory
groups for zoning
As described above, the current practice of building massing design is shaped by
building codes without regards to energy performance. Though slowly changing, building
massing designs are often fully developed with little involvement from engineering
consultants even though building orientation and built density affect outdoor thermal
comfort and building energy use via daylight availability and urban heating.
Besserud and Hussey (2011) call for a need of simulation tools to model massing of
buildings to create comfortable experience with respect to microclimate conditions, wind,
daylighting, and energy consumption. The intensity of UHI is a function of how buildings
are clustered together in a city, and that is why we propose an intervention in the urban
design process when the urban canyon forms take shape.
15
The vision for this research is then to create a software that can help urban designers
and engineers to evaluate their massing designs for thermal comfort and energy efficiency
early in their workflow. The architects and engineers can then refine the individual
building designs in combination with high performance building system for building
envelopes, cool roof, and lighting control systems.
The goal of this research is to create software for urban designers to help masterplan
cities that are energy efficient and thermally comfortable to support sustainable rapid
urban growth. We envision a future where cities are designed with microclimatic
conditions. With this tool urban planners can advocate zoning regulations for building
height and land use as well as policies for traffic intensity and cool and green roofs with
energy and thermal implications. Urban designers can now articulate their design with
microclimatic conditions and parametrically test built densities and vegetation for master
planning. Thermal comfort and energy efficiency are considered together to mitigate the
UHI in order to improve public health as well as to ensure that energy efficiency measures
are sustained. Once developed, this tool will be the first publicly available software to help
designers and engineers model UHI. It will be available as a stand-alone tool as well as a
plug-in for Rhinoceros, a CAD-based modeling environment used by urban designers and
architects to facilitate the integration of microclimatic considerations with the massing
design.
The UHI will be modeled with Urban Weather Generator (UWG) [Bueno et al., 2014],
which estimates the hourly urban canopy air temperature and humidity based on weather
data from a rural weather station. In an effort to create a usable tool, a sensitivity analysis
will be performed to reduce the number of inputs and allow urban designers to focus their
design iterations on the key parameters that affect the UHI. The graphical user interface
(GUI) is developed and simplified based on the result of this sensitivity analysis.
The design tool is developed with inputs and feedbacks from urban designers and
planning practitioners as well as energy consultants to tailor to the needs of the target
users towards a usable tool for a fully integrated climate based design in architecture.
This thesis aims to create a design tool to evaluate and identify design strategies for an
energy-efficient and thermally comfortable urban development. The focus is on building
massing as they are of the greatest concern to the design community as well as to the UHI.
In the following Chapter 2, existing simulation tools are reviewed and discussed for
their appropriateness as design tools. Integrated design tools are introduced to position
this work as an integral component of the family of design tools that are currently being
developed. UWG and umi, which are selected simulation engines and platform for this
software are explained in detail in Chapter 3.
16
We then describe the methodology and results of the sensitivity analysis towards the
creation of a usable tool (Chapter 4). Chapter 5 explains the development process of the
user interface and the usability considerations. The newly proposed workflow of the
microclimatic urban design using this tool is demonstrated through a case study in
Cambridge, MA, USA (Chapter 6). Finally, this thesis concludes with a summary of
contribution and discussion in Chapter 7.
The developed wrapper for UWG can be downloaded at and is documented in
http://urbanmicroclimate.scripts.mit.edu/uwg.php.
17
This chapter provides an overview of currently available urban heat models and integrated
design tools to locate this work in the context of urban design simulation tool for modeling
urban heat island effect.
There have been significant efforts to incorporate the UHI effect in simulation models,
including complex mesoscale atmospheric modeling and computational fluid dynamics
(CFD), simple and analytical method, as well as physics-based urban canopy models.
Mesoscale models are considered the state-of-the-art in atmospheric weather
prediction. Due to the chaotic nature of turbulence, it is inherently difficult to resolve
numerical solutions of atmospheric flow. The latest generation is coupled with urban
canopy models to improve the representation of the thermal effects of urban areas on the
atmosphere. Oxizidis et al. (2008) proposed generating urban weather files by coupling
EnergyPlus [Crawley et al., 2001] with micro-scale Computational Fluid Dynamics (CFD)
and mesoscale atmospheric simulations. Santiago and Martilli (2010) developed a scheme
for a micro-scale CFD that is able to obtain accurate information about the canopy-scale
UHI distribution at a particular location.
The ability to visualize results and identify improvements for specific buildings is very
powerful as a design tool. However, due to the high computational cost of CFD simulations
their scope needs to be spatially and temporally limited, making it undesirable for annual
energy calculations and for analyses of neighborhoods that are larger than a few blocks.
Furthermore, the accuracy of CFD simulations and mesoscale atmospheric modeling
strongly depend on the boundary conditions, for which detailed empirical data is not
readily available.
18
Crawley (2008) introduced a simple and empirical method to create weather files to
represent IPCC’s climate change scenarios in 2100 and UHI effects today. The study
performed energy simulations on prototypical small office buildings with typical, good, and
low-energy practices for 25 locations (20 climate regions) for a typical year represented by
a typical meteorological year (TMY) as well as years with low and high energy use in the
same period. The dry bulb temperature, dew point temperature, and relative humidity
were modified through simple algorithms under IPCC climate change scenarios, and the dry
bulb temperature and humidity ratio were modified for the UHI effect based on diurnal
patterns of UHI [Oke, 1987].
Urban canopy models (UCMs) have been developed to represent urbanized surfaces in
atmospheric numerical simulations and are being used as urban-climate prediction tools.
The Town Energy Balance (TEB) scheme [Masson, 2000] is a well-established example of a
physically-based UCM. The TEB scheme is based on a two-dimensional approximation of an
urban canyon formed by three generic surfaces: a wall, a road, and a roof. The urban
canyon air is represented as a well-mixed thermal zone that exchanges heat with the
generic wall, the generic road, and the atmosphere above the urban canopy layer. The
generic roof exchanges heat directly with the atmosphere. Further developments of the
TEB scheme divide the urban canyon in multiple horizontal layers [Hamdi and Masson,
2008].
In order to improve the representation of buildings in urban canopy models, Kikegawa
et al. (2003) and Salamanca et al. (2010) proposed new building energy models (BEM)
integrated in the urban canopy model. These BEMs are able to capture the main heat
transfer processes that occur inside buildings as well as calculate building energy demand
and waste heat emissions from AC systems.
Bueno et al. (2012c) developed a BEM-TEB model that includes specific models for
passive systems, such as window shadowing devices and natural ventilation. This model is
able to represent the energy effects of buildings and building systems on the urban climate
and to estimate the building energy consumption at city scale (~10 km) with a resolution
of a neighborhood (~100 m).
Bueno (2012a) developed Urban Weather Generator (UWG) using his BEM-TEB scheme
and energy conservation applied to a control volume in the urban canopy layer and the
urban boundary layer (Figure 4). The UWG calculates the hourly values of urban air
temperature and humidity based on rural weather data measured outside a city. The model
considers radiation, precipitation, air velocity, and humidity measured at the weather
station as well as heat fluxes from building walls, windows, and roads, waste heat from
HVAC equipment and other anthropogenic heat sources in the city. It has been tested for
Toulouse, France; Basel, Switzerland [Bueno, 2012a, 2012b]; and Singapore [Bueno, 2014].
The recent evaluation against field data from a network of weather stations in Singapore
demonstrated a range of land uses, morphological parameters and building usages that the
UWG is able to simulate. The performance of UWG is comparable to a more
19
computationally expensive mesoscale atmospheric model. The model shows satisfactory
performance for all weather conditions and for different reference sites. In the Singapore
case study, Bueno et al. (2014) showed that a similar urban air temperature profile can be
achieved when using Changi Airport and Seletar Farmway as reference sites, although the
average maximum temperature difference between the urban sites and the reference site is
significantly different. For these reasons UWG is selected as urban simulation engine for
this project and the model is described in detail in Chapter 3.
Sullivan created the basic executable file for running the UWG algorithm through the
command line interface. Required input files are an EnergyPlus weather file and an xml file
that describes the characteristics of the urban site. This thesis continues Sullivan’s effort to
create a usable UHI modeling tool for urban designers and engineers.
Figure 4
The schematic drawing of the UWG model
No known publicly available tools incorporate UHI in the urban design process. Current
city-scale simulation tools are focused on visualization such as ESRI’s CityEngine (2013)
and land use and transportation such as UrbanSim [Waddell, 2002, 2011]. They do not
include microclimatic effects such as UHI effect and local wind conditions that affect energy
consumption.
SUNtool [Robinson, 2011] is an urban modeling platform that includes integrated
custom modules for modeling microclimatic effects, transient heat flow, plants and
equipment as well as occupant presence and behavior. Its atmospheric flow modeling is
based on mass, momentum, and energy conservation equations.
The purpose of this work is to propose a new tool and workflow of the urban design
process with UHI considerations and its implications for the thermal comfort and energy.
20
The relatively fast algorithm of UWG developed by Bueno allows users to parametrically
test and iteratively improve their massing designs.
There has been much effort in creating simulation tools that interface with architectural
and urban design models in the recent years. Most common are structural and energy
simulation tools. We will focus on the latter for the purpose of this work.
A popular energy simulation engine is the U.S. Department of Energy’s EnergyPlus
[Crawley, 2001] (see Section 3.3 for further details). Many Graphical User Interfaces (GUI)
have been developed for EnergyPlus including DesignBuilder (2014) (Figure 5). It allows
users to import geometry from their 3D modeling software of choice as long as the
geometry is convex and has planar surfaces with collinear vertices. These limitations
present obstacles for an integrated architectural design with energy considerations,
Figure 5
DesignBuilder’s GUI consistently uses tab-based structure for setting different simulation
settings including building schedules [Image credit: http://www.designbuilder.co.uk/]
]
21
especially when the user needs to modify or even simplify their building geometries to
comply with the energy simulation platform. In a recent survey of energy modelers and
architects by Samuelson et al. (2012), 23 out of 62 participants (37%) answered that the
results of energy simulations “rarely” or “occasionally” had impact on design decisions
even in AEC (Architecture, Engineering, and Construction) firms which employ in-house
energy models. This is a direct result of this delayed use of tools within the design process,
and therefore it is crucial that we create a tool within the designers’ current design
platform to encourage early integrations of energy and thermal comfort concepts with
massing design.
Urban Modeling Interface (umi) [version 02.0039; Reinhart et al., 2013], Archsim
[Dogan, 2015; Dogan & Reinhart, 2013], as well as Ladybug [version 0.0.59; Roudsari,
2015] and Honeybee [version 0.0.56; Roudsari, 2015] are a few of the many simulation
tools developed for Rhinoceros (Rhino) [version 5; 2015], which is a three-dimensional
CAD modeling software. Rhino is widely utilized by leading architecture and urban design
schools and practices globally. All of these tools use U.S. Department of Energy’s
EnergyPlus as their energy simulation engine and are described below.
Energy simulation tools have also been developed for other CAD modeling platforms.
For example, SketchUp (2015)’s plug-in OpenStudio [version 1.7.0; 2015] is a whole
building energy modeling platform using EnergyPlus and daylighting simulation tool
Radiance [version 4.2; 2015].
Urban Modeling Interface (umi) is developed in an effort to streamlines the workflow
from formal design conceptualization through energy simulation within single design
platform. It is a plug-in tool for Rhino version 5 for simulating urban-scale operational
energy, walkability, and daylighting. umi’s custom toolbar guides the necessary user inputs
along the way, requiring minimum training to start using the tool. The tool enables
designers to evaluate the environmental performance aspects of urban design within Rhino
and iteratively improve their designs based on these metrics.
Grasshopper [version 0.9.0076; Davidson, 2015] is an algorithmic modeling plug-in tool
for Rhinoceros. Its application include geometry generation as well as structural and
energy simulations. The strength is in its parametric nature, where users can change inputs
quickly to generate series of forms and results. The tool can reference geometry in Rhino or
be created through algorithmic definition in Grasshopper.
Archsim allows users to create complex multi-zone energy models, simulate these
models using EnergyPlus, and visualize results without switching between tools. Ladybug
and Honeybee perform energy and daylighting simulation using Grasshopper. Ladybug
imports standard EnergyPlus Weather files (.EPW) into Grasshopper and provides 3D
graphics to support the decision-making process during the initial stages of design.
Honeybee runs EnergyPlus [Crawley, 2001], Radiance [version 4.2; 2015], Daysim
[Reinhart, 2012] and OpenStudio [version 1.7.0; 2015] inside the Grasshopper
environment for building energy and daylighting simulation. It can also calculate the
22
thermal comfort metric Universal Thermal Climate Index. Users can then connect the
results back to Ladybug to visualize the simulation results.
Both of these tools are in constant development. The simulation inputs such as model
geometry, materiality, constructions and zone usage profiles are fully parametric and can
be coupled with optimization algorithms within Grasshopper. Grasshopper has become
increasingly popular in architectural design to generate design proposals and thus provides
a low cost of entry for users to start using these environmental performance design tools.
Its procedural nature, however, requires initial training as users need to know which
components can be used with one another.
A number of simulation tools are currently developed for Rhino, Grasshopper, and
SketchUp with the same goal to integrate environmental performance simulations within
the urban design modeling process. This chapter surveys only a fraction of them and
particularly those focused on energy and thermal comfort analyses as these metrics are the
main focus of this thesis work. There are other tools that simulate other performance
metrics such as daylighting (i.e. DIVA (2015) for Rhino), structural considerations (i.e.
Millipede [Michalatos, 2015] also for Rhino) among many others.
The wrapper for UWG documented in this thesis serves as the integral tool of
aforementioned family of design tools for two reasons. Firstly it enables designers for the
first time to predict urban heating/cooling based on their designs within their workflow in
Rhino. Secondly when used in conjunction with energy simulation tools such as the ones
mentioned in this chapter, urban energy consumption predictions are improved compared
to our current practice of using weather file from rural weather stations which do not
reflect the microclimatic conditions of the urban sites. For this reason the stand-alone GUI
for UWG is created along with the Rhino-based version for people who are interested in
UWG for improving energy simulations. The users of 3D modeling tools other than Rhino
can also use the stand-alone version.
This chapter introduced various models for UHI including mesoscale atmospheric
modeling involving CFD, Crawley’s empirical algorithm, and urban canopy models. UWG
uses the BEM-TEB model and is selected as the most appropriate UHI modeling approach
for the design tool because of its computational efficiency and its consideration for
interactions between buildings and urban climate. This new software is created for Rhino
for its popularity and more specifically as a new plugin for umi to take advantage of and to
complement umi’s ability to simulate different aspects of environmental performance
including energy. Both UWG’s model and umi’s user interface are explained in detail in the
following chapter.
Our overall goal is to promote integrated design. Together with umi’s energy, mobility,
and life cycle analyses, UWG can contribute towards designs of sustainable cities.
23
We explain in detail the simulation platforms and engines used in this tool development.
Each of UWG, umi, and EnergyPlus is described for its appropriateness for the new urban
design tool. Their interactions are then described in the last section of this chapter.
As introduced in Chapter 2, Urban Weather Generator (UWG) [Bueno et al., 2012a,
2014] estimates the hourly urban canopy air temperature and humidity using weather data
from a rural weather station. The model is based on energy conservation principles and is a
bottom-up building stock model. The model has been tested for Toulouse, Basel, and
Singapore (described in Section 3.1.3) and shows that it can satisfactorily estimate urban
temperatures in different climates, weather conditions, and urban configurations to obtain
an estimation of UHI. The performance of UWG is comparable to a more computationally
expensive mesoscale atmospheric model.
UWG has four sub-models: (1) rural station model (RSM), (2) urban diffusion model
(VDM), (3) urban boundary layer model (UBLM), and (4) urban canopy and building energy
model (UC-BEM), as shown in Figure 6.
The RSM reads in the hourly values of meteorological data from the rural reference site.
The sub-model is based on an energy balance at the soil surface and calculates the rural
sensible heat flux used by the VDM and UBLM. A transient heat diffusion equation is solved
via finite differences, representing the storage and heat released from the ground.
The VDM takes the sensible heat flux from the RSM as well as air temperatures and
velocities measured at the weather station to calculate the vertical profile of air
temperature above the weather station based on heat diffusion. Its output is then provided
to the UBLM.
24
The UBLM [Bueno et al., 2012a] calculates air temperatures above the urban canopy
layer based on an energy balance for a selected control volume (Figure 4). It numerically
solves the advection integral [Bueno et al., 2014] by using the vertical profile of air
temperature provided by the VDM and the rural and urban sensible heat fluxes from RSM
and UC-BEM, respectively.
If the model considers multiple neighborhoods, the temperature calculated for an
upwind section of the urban boundary layer is used as the reference temperature for
downwind sections.
The UC-BEM calculates urban canyon air temperature and humidity using radiation and
precipitation data, air velocity and humidity measurements at the reference site as well as
air temperatures from the UBLM. The model is based on the Town Energy Balance (TEB)
scheme [Masson, 2000] and is further improved to integrate building energy model [Bueno
et al., 2012c] as described in Section 2.1.4.
For multiple neighborhood simulations, the building characteristics are defined by the
distribution of the building types and are simulated in parallel in the UC-BEM. All building
Figure 6
The four sub-models of UWG and their interactions
25
types in the neighborhood are affected by the same urban climate and their effect on the
urban climate is weighted by the distribution of building types.
UWG takes urban and rural reference site descriptions to morph a weather file from the
closest reference site. Specifically, we need (1) an EnergyPlus weather (epw) file and (2) an
Extensible Markup Language (xml) file describing the urban and rural site characteristics.
The output is an urban weather file in the epw format that incorporates the urban heat
island effect and that is compatible with many building performance simulation programs.
The first input, the epw file, can be obtained from the U.S. Department of Energy
(DOE)’s website [U.S. Department of Energy, Weather Data, 2013]. It has weather data for
more than 2100 locations, including 1042 locations in the U.S.. UWG can also take custommade weather files as inputs as long as they are in the epw file format. Theoretically, the
reference weather station should capture the climate conditions upwind for the city for all
observed wind directions. However, often times it is difficult to find a suitable reference
site. Bueno et al. (2014) showed that the choice of the reference weather reference site is
not crucial for the model.
The weather morphing is based on descriptions of the meteorology, urban morphology
and geometry and reference site parameters, schematically shown in Figure 7. They are
passed to UWG as an XML file. The urban morphology information includes building
characteristics related to energy balance that are required by most energy simulations,
such as building construction materials and properties as well as building use schedules.
The urban geometry parameters define the shape of the urban canyon and can be
computed from a user-defined Rhino massing model or extracted from a geographic
information system (GIS) data for existing cities. Finally, the meteorological parameters
describe the urban boundary layer, derived through experimentation and observations.
The sensitivity analysis in Chapter 4 determines which of these parameters are the most
important to the UHI.
26
Figure 7
Overview and examples of UWG inputs that describe the urban and reference site environments
UWG has been tested for Basel, Switzerland; Toulouse, France; and recently in several
neighborhoods in Singapore [Bueno et al., 2014]. The findings for Singapore are
summarized below. Singapore represents a heterogeneous morphology with abundant
vegetation in contrast to the previously tested European cities. The climate is tropical and
is thus cooling dominated, which is vastly different from the mild climates of the European
cities. This case study also serves as a validation for the new UWG model that incorporates
multiple neighborhood descriptions.
The performance of UWG is tested by comparing the urban air temperatures calculated
by UWG with measurements from a network of weather stations in Singapore (Figure 8),
representing a range of land uses, morphological parameters and building usages. The
average diurnal cycles of canopy-layer air temperatures calculated by UWG and observed
at urban sites are compared in Figure 9. The error of the model stayed within the range of
air temperature variability observed in different locations within the same urban area for
all neighborhoods. An error (RMSE) of 1K is consistent with the previous case studies and
is comparable to that of the significantly more complex and computationally expensive
simulation platforms (1.7K) when compared with the same set of temperature
measurements.
Bueno also shows the model’s robustness for different reference sites by comparing
simulation results from Seletar Farmway and Changi Airport (Table 1). Seletar Farmway is
used as the temperature reference site for this study as it is upwind of all neighborhoods;
however its data is not public. The temperature data for this site as well as those for each
neighborhood were specifically obtained for this study using a network of ONSET HOBOTM
sensors.
Another concern for the reference site is the proximity to large bodies of water. Many
airports are located near the water, as in the case for Changi Airport. The rural site should
not be affected by its site-specific microclimatic conditions produced by orography or the
presence of large bodies of water. Bueno showed in the case of Singapore that the choice of
reference site has little influence on the computed urban temperatures, although it affects
the computation of temperature differences between the rural and urban sites.
27
Figure 8
Neighborhoods modeled in Singapore validation. Changi Airport (right side of image) is the
reference site for all weather data except for air temperatures from Seletar Farmway
Figure 9
Average diurnal cycle of canopy layer air temperatures calculated by UWG and observed at
urban sites. The results correspond to the average of seven urban sites shown in Figure 8. The error bar
represents the standard deviation of simulated and measured air temperatures for the different sites. Finally,
the air temperatures measured at Seletar Farmway (reference site) are plotted to visualize the UHI [Bueno et
al., 2014]
28
By showing that similar urban temperature profiles can be produced by using weather
data from Changi Airport (publicly available and located near the coastline) and Seletar
Farmway (privately collected and inland), we have increased the number of cities for which
we can confidently model the UHI using UWG. This robustness is pertinent to the design
tool as data may not be available for a site that captures climate conditions upwind of the
city.
The UWG model can thus be applied to different climates and urban configurations to
capture the UHI based on these results.
Table 1 UWG is able to produce similar urban temperature profiles for the two different reference sites. The
UHI is estimated by ΔTmax, the average maximum temperature difference observed at urban and reference
sites. The last two columns indicate the modeled and observed average standard deviation among the seven
urban weather stations.
The model shows satisfactory performance for all tested weather conditions and
reference sites [Bueno et al., 2014], which validates UWG’s suitability as an urban
simulation engine. Its output is a morphed weather file (epw) that is compatible with many
building performance simulation programs including those introduced in Chapter 2.
UWG is robust and shows similar results for different weather stations, as discussed in
the prior section 3.1.3. This allows users to use only publicly available data for their
simulations. The recent Singapore case study also showed that the model performs well for
all weather conditions including rainy, dry/cloudy, and dry/clear. UWG’s results are
comparable to a more computationally expensive mesoscale atmospheric model. It is much
faster and thus is appropriate to be used for iterative design process, but the trade-off of its
simplifications and assumptions of the model is that it prevents the model from capturing
very site-specific microclimate effects. Yet the model is still robust enough to produce
plausible values across urban morphology and vegetation parameters based on model
validation in three different sites.
Street (2013) compared Crawley (2008)’s temperature alteration algorithm introduced
in 2.1.2 against UWG. Crawley’s method is recommended for applications that lack urban
site data and for order of magnitude estimations to provide extremes of the predicted
energy use intensity. On the other hand, UWG is appropriate for applications that either
require feedback with the urban design or if there is access to extensive data on the urban
morphology. For the purpose of this study, UWG is more suitable to be used for an iterative
design process.
29
As introduced in the last section, UWG is not able to capture very site-specific
microclimate effects beyond spatially-averaged results due to its simplification of the
model for computational efficiency. This implies that the urban design tool is not able to
point to which building should be the focus of alteration in order to improve the thermal
comfort of the neighborhood. Although a numerical solution of the momentum equation via
computational fluid dynamics (CFD) would be able to provide the spatial mapping
described here, it is not the focus of this study as each simulation would take significantly
longer time. Instead this study focuses on design improvement via iterative design, by
allowing users to explore different design options in a short time frame to quickly evaluate
modeling hypothesis.
UWG still requires further investigation into how it captures the effects of vegetation as
well as advection from rural to urban boundary layers. The sensitivity analysis for Boston
(Chapter 4) does not capture the effects of urban vegetation. Earlier studies by Kurn et al.
(1994) showed that the vegetation may lower urban temperatures by 1K, based on a
network of twenty-three weather stations in Southern California.
The Boston case study also revealed that advection can play a relevant role in the
energy balance of the UBL for cities with high wind velocities. This is contrary to Bueno et
al. (2014)’s conclusion that the low influence of the advective heat flux on the UBL energy
balance infers small sensitivity to city size. Yet this sensitivity analysis shows that city size
(characteristic length) is not a key parameter for Boston or Singapore.
Users should also note that UWG model does not compute local wind speeds, though the
calculated urban air temperatures and humidity use wind direction and magnitude
information from the reference site epw file. An approximation of the urban wind velocity
can be made (see Section 5.3.2) to calculate outdoor thermal comfort.
Finally, UWG requires over fifty user inputs to run a single simulation. These inputs
include boundary layer information, which is not easily available to designers nor
engineers. The sensitivity analysis presented in Chapter 4 is designed to reduce such inputs
to facilitate the iterative design process.
The urban modeling interface (umi) is a Rhino-based design environment for architects
and urban planners interested in modeling the environmental performance of
neighborhoods and cities with respect to operational and embodied energy use, walkability
and daylighting potential. Umi was developed with the goal to introduce urban designers
and architects to building performance simulations within a familiar modeling
environment and to thus allow them to combine urban environmental performance
assessments with computational design approaches such as parametric modeling and
optimization [Reinhart, 2013]. Umi is being actively developed by the Sustainable Design
Lab at the Massachusetts Institute of Technology. The first public version was released in
2013, followed by version 2.0 in November 2014, which can be downloaded from
http://urbanmodellinginterface.ning.com/. The version used in this study is 2.0039 (2015).
Umi was selected for our design platform to work with its existing energy simulation
component. The tool walks users through the energy simulation process as shown in Figure
30
10. The user sets the weather file and creates massing model of the city. Each building is
assigned customizable templates (Figure 11) for construction materials, glazing systems,
internal loads, and building use schedules. The massing is approximated into four small
shoeboxes facing each direction to reduce the simulation time [Dogan & Reinhart, 2013].
When the simulation is run a comma separated values (csv) file is saved in the umi
directory. A results viewer is currently being developed to facilitate comparison and
analysis of simulation results.
Figure 10 Typical workflow for umi– (1) Set location (for weather file), (2) assign building templates and
(3) run simulation. In step (1) users are able to add and modify templates.
31
Figure 11
32
Building construction and schedule templates for umi
The U.S. Department of Energy’s whole building energy simulation program EnergyPlus
models heating, cooling, lighting, ventilation, and other energy flows through multi-zone
airflow, thermal comfort, and natural ventilation analyses. EnergyPlus calculates energy
usage of a building based on climate and building property information related to energy
balance. The simulation requires an epw and input data (IDF) files.
EnergyPlus was developed by Crawley et al. (2001) and was first released in April 2001.
The engine has been tested thoroughly and continues to be extended; the most recent
version is Version 8.3 (as of 1 April, 2015). EnergyPlus is available free of charge and is
open source, leading to development of many third-party interfaces including
DesignBuilder, Archsim, and Honeybee as mentioned in Chapter 2. umi also uses
EnergyPlus (version 8.1.0) as its energy simulation engine.
For the purpose of this study EnergyPlus is used to evaluate the energy implications of
UWG’s parameters in the sensitivity analysis as well as in the umi-integrated version of the
software.
One goal of this work is to create a design software in which formal and urban
microclimate design can be done within the same platform. As previously discussed, Rhino
is selected as it is used widely in architectural practices and schools.
The envisioned workflow is schematically shown in Figure 12. Once urban designers
define the massing of the neighborhood as well as neighborhood characteristics they can
use the umi-plugin to run UWG and EnergyPlus. Some urban morphology parameters are
automatically extracted from their massing model. In this way, our wrapper for the UWG
assists users to evaluate their design for urban microclimate and energy consumption so
that users can iteratively improve their massing model for sustainability and thermal
comfort.
33
Figure 12 Vision for integrated tool in Rhino. UWG and EnergyPlus run within the Rhinoceros environment
so designers can iteratively improve their designs for thermal comfort and energy efficiency during the
formal design process
34
As UWG requires over 50 parameters, sensitivity analyses are performed to identify the
most important parameters and reduce the number of user inputs. Fewer required
parameters facilitate an iterative design process, allowing urban designers to focus on
improving their designs around the key parameters that affect the UHI.
The goal of the sensitivity analysis is two-fold: (1) test significance of parameters that
are of high interest to urban designers and planners, such as massing and land use as well
as (2) ensure that the inputs that are difficult to obtain (i.e. meteorological parameters) are
not significant to the UHI nor site-specific and therefore can be assigned default values in
the user interface. Each parameter is evaluated against the thermal and energy metrics to
align with the goal of the tool to achieve thermal comfort with energy efficiency.
An earlier study for Toulouse and Basel (mild climates) [Bueno et al., 2012a] has shown
that site coverage ratio, façade-to-site ratio, and vegetation are the most sensitive
parameters. Additional studies for Punggol, Singapore (tropical, residential district) and
Boston Financial District, Boston, MA (cold, commercial and densest district in Boston) are
conducted to determine the most effective design strategies for each climate. If a parameter
is determined to be significant across all climates, then we can conclude that such
parameter is a key contributor to the UHI. Similarly, if it has a minor impact across all
climates, then we can assign a default value. The Boston parametric study is demonstrated
in this chapter and the setup for the Singapore study is provided in Appendix B.
The base value for each parameter under study is extracted from geographic
information system (GIS) or satellite image. The high and low ranges are obtained based on
empirical data from other cities or previous research (Table 2). Their simulation results are
compared against the base case to determine the parameter’s effect on UHI.
35
Table 2 UWG Input Parameters for Boston
Notes:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
36
Bueno et al. (2014)
Comparable values assigned to Basel, Switzerland [Bruno et al., 2012a]
Annual average of vehicular anthropogenic heat is 45.2% of total anthropogenic heat [Sailor, 2006; Stewart & Oke, 2012].
Low and high values obtained by ranging the average value by ±25%
Same as urban albedo
Maximum of LCZ 1-6 from Stewart and Oke (2012)
Minimum and maximum of LCZ A-Z from Stewart and Oke (2012)
Small Office from US Department of Energy, Commercial Reference Buildings (n.d.)
Midrise Apartment from US Department of Energy, Commercial Reference Buildings (n.d.)
Minimum and maximum of LCZ 1-6 from Stewart and Oke (2012)
Façade area implied from horizontal building density as well as floor area and façade area from respective reference building
Initial temperature for all surfaces: wall, roof, mass, rural road, urban road
Average of eight neighborhoods, mix of commercial and residential
Residential and commercial, respectively
Assumed maximum of 10.0% of the anthropogenic sensible heat
Table 3
URBAN
LCZ Key
Example
Local climate zone for urban classification [Stewart & Oke, 2012], used as reference values for obtaining base values for Table 2
Average Values
Aspect Ratio H/W
Mean Building/Tree Height
Building Surface Fraction
Pervious Surface Fraction
Albedo
Anthropogenic Heat Flux
SUBURBAN
LCZ Key
Example
Average Values
Aspect Ratio H/W
Mean Building/Tree Height
Building Surface Fraction
Pervious Surface Fraction
Albedo
Anthropogenic Heat Flux
RURAL
LCZ Key
Example
Average Values
Aspect Ratio H/W
Mean Tree Height
Building Surface Fraction
Pervious Surface Fraction
Albedo
Anthropogenic Heat Flux
1 Compact High Rise
Core downtown
Min
2
25
0.4
0
0.1
50
Max
Ave
3
0.6
0.1
0.2
300
2 Compact Mid Rise
Core (old city, inner city)
2.5
25
0.5
0.05
0.15
175
7 Lightweight Low-rise
Periphery of developing city
Min
1
2
0.6
0
0.15
0
Max
Ave
2
4
0.9
0.3
0.35
35
1.5
3
0.75
0.15
0.25
17.5
A Dense Trees
Natural forest or park
Min
1
3
0
0.9
0.1
Max
Ave
30
0.1
1
0.2
0
Min
0.75
10
0.4
0
0.1
0
Max
2
25
0.7
0.2
0.2
75
3 Compact Low Rise
Core (old city, inner city)
Ave
Min
Max
1.375
0.75
17.5
3
0.55
0.4
0.1
0
0.15
0.1
37.5
0
8 Large Low-Rise
Light industrial, commercial
Min
0.1
3
0.3
0
0.15
0
Max
Ave
0.3
10
0.5
0.2
0.25
50
0.2
6.5
0.4
0.1
0.2
25
B Scattered Trees
Lightly wooded parks
1
16.5
0.05
0.95
0.15
0
Min
0.25
3
0
0.9
0.15
Max
Ave
0.75
15
0.1
1
0.25
0
1.5
10
0.7
0.3
0.2
75
5 Open MId Rise
Periphery
6 Open Low Rise
Suburbs
Ave
Min
Max
Ave
Min
Max
Ave
Min
Max
Ave
1.125
0.75
1.25
1
0.3
0.75
0.525
0.3
0.75
0.525
6.5
25
25
10
25
17.5
3
10
6.5
0.55
0.2
0.4
0.3
0.2
0.4
0.3
0.2
0.4
0.3
0.15
0.3
0.4
0.35
0.3
0.5
0.4
0.3
0.6
0.45
0.15
0.12
0.25
0.185
0.12
0.25
0.185
0.12
0.25
0.185
37.5
0
50
25
0
25
12.5
0
25
12.5
9 Sparsely Built
10 Heavy Industry
Extended Metropolitan regions Industrial
Min
0.1
3
0.1
0.6
0.12
0
Max
0.25
10
0.2
0.8
0.25
10
Ave
Min
Max
0.175
0.2
6.5
5
0.15
0.2
0.7
0.4
0.185
0.12
5
300
C Bush, Shrub
Mediterranean Scrubland
0.5
9
0.05
0.95
0.2
0
4 Open High Rise
Periphery
Min
0.25
0
0
0.9
0.15
Max
1
2
0.1
1
0.3
0
Ave
0.5
15
0.3
0.5
0.2
0.35
10
0.25
0.45
0.16
300
D Low Plants
E Bare Rock or Paved
Urban recreation (grassy park) Desert or car parks
F Bare soil or sand
Natural desert, barren land
Ave
Min
Max
Ave
Min
Max
Ave
Min
0.625
0
0.1
0.05
0
0.1
0.05
1
0
1
0.5
0
0.25
0.125
0.05
0
0.1
0.05
0
0.1
0.05
0.95
0.9
1
0.95
0
0.1
0.05
0.225
0.15
0.25
0.2
0.15
0.3
0.225
0
0
0
0
0
0
0
0
0
0.2
Max
Ave
0.1
0.25
0.1
0.9
0.35
0
0.05
0.125
0.05
0.45
0.275
0
G Water
Lakes, rivers, reservoirs
Min
0
0
0
0
0.02
Max
Ave
0.1
0
0.1
0.9
0.1
0
0.05
0
0.05
0.45
0.06
0
[page intentionally left black]
As described in Section 3.1.2, UWG’s meteorological parameters are empirically derived
and designers rarely have access to this information. The existing data from Toulouse and
Basel [Bueno et al., 2012a] are used as the starting point and are varied ±25% for the high
and low values, unless otherwise noted in Table 2 to achieve a more realistic range of
values.
ESRI’s GIS data [ArcGIS/ArcMAP 10.2 for Windows, 2014] is used for collecting urban
morphology data. ArcGIS contains the 3D CAD model of the city as well as the twodimensional map (Figure 13), and the compilations of building information, such as
building height, footprint, and base perimeter. These are extracted from the CAD model
using Grasshopper, an algorithmic modeling plugin for Rhinoceros (Figure 14).
Parameters that are determined after the urban design process (i.e. architectural design
process) such as building construction materials and building use are not studied in the
sensitivity analysis and thus are treated as constants. However, building insulation levels
and building use schedule can affect the energy consumption. Thus, the methods to obtain
these values are detailed in Section 4.2.2.
The sensible anthropogenic heat is estimated as the vehicular contribution of
anthropogenic heat flux [Sailor, 2011] for compact high rise neighborhoods based on
Stewart’s Urban Classification in Table 3 [Stewart & Oke, 2012].
Urban tree coverage is estimated from satellite images, as it is not provided in the GIS
dataset (Figure 15). The area covered with trees are outlined and then computed as a ratio
of the total site area in Rhino. For most cities it can be estimated as the park footprint.
Rural vegetation parameters and obstacle heights are estimated from satellite images,
as they are not provided in the GIS dataset (Figure 16). The vegetated areas are computed
similarly as the urban tree coverage using Rhino.
Figure 13
Boston Financial District’s GIS data is used to extract the urban geometry parameters. Rhino
model (left) and site selection in blue (right)
39
Figure 14
Grasshopper definition to extract the urban geometry parameters. This Grasshopper
definition is available for download at http://urbanmicroclimate.scripts.mit.edu/uwg_parameters.php
Figure 15
Figure 16
40
Boston Financial District satellite image obtained using Google Maps
Boston reference site: Logan Airport obtained using Google Maps
The energy implications of UWG input values are measured using EnergyPlus
simulations following weather morphing using UWG. Care must be taken to ensure
consistent parameter use in UWG and EnergyPlus simulations. The building construction
and building use values are obtained from the U.S. Department of Energy’s Commercial
Reference Buildings (n.d.) templates summarized in Table 4. Hourly building use schedules
are provided for a typical week for infiltration and internal heat gains, including gain from
occupants, lights, and equipment. Weighted average values for all zones are used to
extrapolate each parameter, where daytime is defined between 7am and 6pm based on
Boston’s monthly sunlit hours.
The base case urban simulation result is generated using the actual values measured for
Boston Financial District. Each parameter is changed one at a time and its simulation result
is evaluated for its impact on temperature and energy use against the base case. Four
metrics are used, which are described in detail below: (a) temperature change compared to
urban base case, (b) percent change in annual heating and cooling energy consumption, (c)
percent change in winter heating energy consumption, and (d) percent change in summer
cooling energy consumption. A parameter is determined to be significant if it fails one or
more metric for either low or high end of the sensitivity range.
The hourly impact of changing a parameter is evaluated through this metric. A sample
distribution is shown in Figure 17, with the vertical dotted line representing the cut-off for
sensitivity. A parameter is considered to be significant if more than 0.5% of the 8760 hours
in a year deviate more than 0.5K from the original temperature profile.
Using the morphed weather file for the urban site, EnergyPlus energy simulation results
are compared. Hourly simulation results for “Heating: Gas” and “Cooling: Electricity” are
aggregated for heating and cooling energy consumptions, respectively. A parameter is
considered significant if the total difference in heating and cooling energy consumption is
greater than 2.0%.
Figure 17 Temperature metric: temperature change compared to urban base case. A sample for site
coverage ratio, high value.
41
Table 4
Building construction and use values based on DOE’s Commercial Reference Buildings (n.d.) for
Boston small office
Notes:
1
Weighted average based on hourly schedule. Cooling for April through September, and heating from October to March. Daytime and
nighttime determined based on Climate Consultant's sun path chart
2
Assumed that of the apartments, as it is 89.5% of the total volume (apartments, corridor, and office).
3
Weighted average of each of hourly schedule and energy use for all programs including apartments, corridor, and office. Weighted
by floor area
4
Includes internal gain from occupants, lights, and equipment. Assumed 100W per occupant. Weighted average of each of hourly
schedule and energy use for all programs including apartments, corridor, and office. Weighted by floor area. Day and night time
split based on Boston settings
5
Does not take into account the attic
6
Heat gain from people [EnergyPlus Engineering Reference, 2013]
7
Weighted average by the hourly internal load from people, lighting, and equipment
42
The change in seasonal energy consumption is analyzed through this metric.
EnergyPlus is used here as well to compare heating energy consumption for November
through January.
Similarly to the above, change in air conditioning use between the months of June
through August are simulated to observe any change in seasonal energy use during the
summer months.
Simulation results for Boston are summarized in Table 5, and detailed analysis for key
parameters is shown in Figure 18. Boston Logan Airport is used as the reference site.
Similarly to the earlier studies in Toulouse and Basel, geometric parameters such as site
coverage ratio and façade-to-site ratio are important to the urban heat island effect, as well
as anthropogenic heat and roof materials.
The sample results for meteorological parameters are shown in Figure 18 and Figure
19. While most parameters did not significantly change the temperature profile or energy
consumption of the cities, the parameter that defines the boundary condition for the
vertical diffusion model called “reference height at which vertical profile of potential
temperature is assumed uniform” seems to affect the UHI. For cities with high wind
velocities such as Boston, advection can play a relevant role in the energy balance of the
urban boundary layer. This represents a limitation of the UWG’s model as explained earlier
in Section 3.1.5.
The sensitivity analyses results for Punggol, a residential district in Singapore, are
shown in Table 6. Changi Airport is used as the reference site. The cooling loads for the
northeast monsoon period (June through August) as well as that for the southwest
monsoon period (November through January) are calculated. As mentioned before, the
parameter selection process is documented in the Appendix.
The summary of the key parameters for Boston and Singapore are shown in Table 7 and
are namely site coverage ratio, façade-to-site ratio, and sensible anthropogenic heat.
One can observe a higher sensitivity to site coverage ratio for Boston, which has
buildings closer together and in turn has narrower canyon width. This configuration traps
heat in the urban canyon and thus has detrimental effect to the UHI.
The façade-to-site ratio is sensitive for Boston and the European cities. This parameter
was not significant to Punggol, perhaps due to the fact that the variations were too small
for the low and high ranges.
43
Finally, sensible anthropogenic heat plays a significant role for UHI. This is as predicted,
as automobile fumes and waste heat are directly put in the urban canyon to warm the city.
Table 7 shows results for Boston Financial District using two different reference sites:
Logan Airport and Bedford Hanscom Field. As an aside, we can observe that the same
sensitivity metrics were triggered for Boston regardless of their reference site.
Figure 18
Boston sensitive parameters
Figure 19
The meteorological parameters are not important for Boston
44
Table 5
Boston sensitivity analysis results
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
Urban Road K
Low
1.00
0.75
(25.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Urban Road VHC
High
1.25
25.0%
99.9%
(0.1%)
(0.4%)
0.6%
Minor
Minor
Minor
Minor
Latent Fraction of Grass
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
Low
0.60
0.45
(25.0%)
99.5%
0.1%
(0.0%)
0.8%
Minor
Minor
Minor
Minor
High
0.75
25.0%
99.4%
0.1%
0.0%
0.7%
Minor
Minor
Minor
Minor
Rural Vegeration Fraction
Average Value
H/L
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
Low
0.50
0.00
(100.0%)
99.9%
(0.0%)
(0.3%)
0.5%
Minor
Minor
Minor
Minor
High
1.00
100.0%
99.6%
(0.1%)
(0.5%)
0.8%
Minor
Minor
Minor
Minor
Temp Meas Height at Ref
Height
Average Value
H/L
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
Low
2.0
0
(25.0%)
99.9%
(0.0%)
(0.3%)
0.5%
Minor
Minor
Minor
Minor
High
10
25.0%
99.6%
(0.1%)
(0.5%)
0.8%
Minor
Minor
Minor
Minor
1,600,000
Low
High
1,200,000
2,000,000
(25.0%)
25.0%
99.9%
99.9%
(0.0%)
(0.0%)
(0.4%)
(0.4%)
0.7%
0.7%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Albedo of Vegetation
Low
0.25
0.10
(60.0%)
99.4%
0.1%
(0.0%)
1.0%
Minor
Minor
Minor
Minor
High
0.30
20.0%
99.4%
0.1%
(0.0%)
0.8%
Minor
Minor
Minor
Minor
Wall Vegetation Coverage
Low
0.001
0.00
(100.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
10.0
0
(0.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Rural Road K
0.17
1.00
Low
0.08
(51.5%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Roof Vegetation Coverage
0.01
0.10
9900.0%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
0.00
(100.0%)
99.9%
20
1.0%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
0.25
51.5%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
200.00
Low
High
150.00
250.00
(25.0%)
25.0%
99.4%
99.4%
0.1%
0.1%
(0.0%)
(0.0%)
0.8%
0.8%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Low
High
High
Heat flux threshold for
daytime conditions
High
Air Velocity Measurement
Height
Low
Urban Road Albedo
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
High
0.10
900.0%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Heat flux threshold for
nighttime conditions
Low
50.0
37.5
(0.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
High
62.5
0.2%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Low
0.75
(25.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Rural Road VHC
High
1.25
25.0%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Minimum Wind Velocity
Low
0.10
0.00
(99.9%)
99.3%
0.1%
(0.0%)
0.8%
Minor
Minor
Minor
Minor
High
1.00
900.0%
99.4%
0.1%
(0.0%)
0.8%
Minor
Minor
Minor
Minor
Initial Temperature
Low
20
15
(25.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
High
25
25.0%
99.9%
(0.1%)
(0.5%)
0.7%
Minor
Minor
Minor
Minor
Begin month for veg
participation
Low
1.0
3
(500.0%)
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
High
6
500.0%
99.9%
(0.1%)
(0.5%)
0.7%
Minor
Minor
Minor
Minor
1,600,000
Low
High
1,200,000
2,000,000
(25.0%)
25.0%
99.9%
99.9%
(0.1%)
(0.0%)
(0.5%)
(0.4%)
0.7%
0.6%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Maximim Film Water
Depth
0.005
Low
High
0.004
0.006
(25.0%)
25.0%
99.4%
99.4%
0.1%
0.1%
(0.0%)
(0.0%)
0.8%
0.8%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Rural Average Obstacle
Height
Low
0.10
0.08
(25.0%)
100.0%
(0.2%)
(0.6%)
0.8%
Minor
Minor
Minor
Minor
High
0.13
25.0%
99.9%
0.1%
(0.3%)
0.6%
Minor
Minor
Minor
Minor
End month for veg
participation
Low
12.0
6
(0.2%)
100.0%
(0.2%)
(0.6%)
0.8%
Minor
Minor
Minor
Minor
High
9
0.2%
99.9%
0.1%
(0.3%)
0.6%
Minor
Minor
Minor
Minor
Rural Road Albedo
Daytime BL Height
Nighttime BL Height
Reference height
0.17
700
80
150
Low
0.08
(51.5%)
99.9%
(0.0%)
(0.4%)
0.6%
Minor
Minor
Minor
Minor
High
0.25
51.5%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Exchange velocity
coefficient
Low
0.30
0.23
(25.0%)
32.2%
(4.2%)
(6.4%)
5.6%
Significant
Significant
Significant
Significant
High
0.38
25.0%
98.5%
2.5%
3.4%
(1.6%)
Significant
Significant
Significant
Minor
Characteristic Length
Low
1,000
100
(90.0%)
99.9%
0.1%
(0.2%)
0.7%
Minor
Minor
Minor
Minor
High
3,000
200.0%
99.7%
(0.3%)
(0.6%)
0.5%
Minor
Minor
Minor
Minor
Night setpoint start time
Low
19.0
17
(4736.8%)
99.9%
0.1%
(0.2%)
0.7%
Minor
Minor
Minor
Minor
High
21
10526.3%
99.7%
(0.3%)
(0.6%)
0.5%
Minor
Minor
Minor
Minor
Low
500
(28.6%)
99.5%
0.1%
(0.0%)
0.9%
Minor
Minor
Minor
Minor
High
1000
42.9%
99.4%
0.1%
(0.0%)
0.8%
Minor
Minor
Minor
Minor
Floor Height
Low
3.05
2.29
(25.0%)
99.0%
0.3%
(0.3%)
1.8%
Significant
Minor
Minor
Minor
High
3.81
25.0%
100.0%
(0.2%)
(0.5%)
(0.1%)
Minor
Minor
Minor
Minor
Sensible Anthropogenic
Heat
Low
67.8
0.0
(100.0%)
99.5%
1.6%
2.2%
(2.2%)
Minor
Minor
Significant
Significant
High
135.5
100.0%
94.6%
(1.4%)
(2.7%)
3.2%
Significant
Minor
Significant
Significant
Night setpoint end time
Low
5.0
3
(1355.0%)
99.5%
1.6%
2.2%
(2.2%)
Minor
Minor
Significant
Significant
High
8
1355.0%
94.6%
(1.4%)
(2.7%)
3.2%
Significant
Minor
Significant
Significant
Low
50
(37.5%)
99.3%
0.0%
(0.1%)
0.8%
Minor
Minor
Minor
Minor
High
100
25.0%
99.4%
0.1%
(0.0%)
0.8%
Minor
Minor
Minor
Minor
Average Building Height
Low
51.4
38.6
(25.0%)
100.0%
0.2%
0.3%
(0.2%)
Minor
Minor
Minor
Minor
High
64.3
25.0%
99.6%
(0.2%)
(0.8%)
1.2%
Minor
Minor
Minor
Minor
Latent Anthropogenic Heat
Low
6.8
0.0
(100.0%)
99.9%
(0.1%)
(0.4%)
0.6%
Minor
Minor
Minor
Minor
High
13.6
100.0%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Maximum wind velocity
Low
10.0
1
(67.8%)
99.9%
(0.1%)
(0.4%)
0.6%
Minor
Minor
Minor
Minor
High
20
67.8%
99.9%
(0.0%)
(0.4%)
0.7%
Minor
Minor
Minor
Minor
Low
100
(33.3%)
91.4%
0.5%
0.3%
1.0%
Significant
Minor
Minor
Minor
High
200
33.3%
85.2%
(0.1%)
(0.0%)
0.5%
Significant
Minor
Minor
Minor
Site Coverage Ratio
Low
0.5
0.2
(60.8%)
97.3%
4.9%
6.2%
(2.1%)
Significant
Significant
Significant
Significant
High
0.7
37.3%
25.9%
(7.0%)
(10.1%)
5.3%
Significant
Significant
Significant
Significant
Percent Heat Released to
Canyon
Low
0.5
0.0
(100.0%)
100.0%
0.2%
0.6%
(2.1%)
Minor
Minor
Minor
Significant
High
1.0
100.0%
92.6%
(0.2%)
(1.4%)
3.8%
Significant
Minor
Minor
Significant
Max discritization length
for UBL model
Low
500.0
375
(0.1%)
100.0%
0.2%
0.6%
(2.1%)
Minor
Minor
Minor
Significant
High
625
0.1%
92.6%
(0.2%)
(1.4%)
3.8%
Significant
Minor
Minor
Significant
Urban Breeze Scaling
Coefficint
1.2
Low
High
0.9
1.5
(25.0%)
25.0%
99.4%
99.5%
0.1%
0.1%
(0.0%)
(0.0%)
0.8%
0.8%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Façade-to-site Ratio
Low
4.9
3.7
(25.0%)
99.1%
2.0%
2.0%
(0.0%)
Minor
Minor
Minor
Minor
High
6.1
25.0%
99.8%
(1.0%)
(1.4%)
0.5%
Minor
Minor
Minor
Minor
Cooling Capacity
Low
205
75
(63.4%)
99.8%
0.0%
(0.4%)
0.9%
Minor
Minor
Minor
Minor
High
335
63.4%
100.0%
(0.1%)
(0.4%)
0.4%
Minor
Minor
Minor
Minor
Latent Fraction of Trees
Low
0.7
0.5
(25.0%)
99.3%
0.1%
(0.0%)
0.9%
Minor
Minor
Minor
Minor
High
0.9
25.0%
99.4%
0.1%
0.0%
0.6%
Minor
Minor
Minor
Minor
Urban Road Vegetation
Fraction
Low
0.3
0.0
(100.0%)
99.9%
(0.1%)
(0.5%)
0.7%
Minor
Minor
Minor
Minor
High
0.6
100.0%
99.9%
(0.0%)
(0.4%)
0.6%
Minor
Minor
Minor
Minor
Tree Coverage
Low
0.11
0.00
(100.0%)
99.9%
(0.2%)
(0.7%)
1.3%
Minor
Minor
Minor
Minor
High
0.20
81.8%
100.0%
0.1%
(0.1%)
0.1%
Minor
Minor
Minor
Minor
[page intentionally left black]
Table 6
Singapore sensitivity analysis results
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Cooling % Change
NE Monsoon Season Cooling % Change
SW Monsoon Season Cooling % Change
Checks:
Urban Road K
1.00
Low
High
0.75
1.25
(25.0%)
25.0%
100.0%
100.0%
0.0%
(0.0%)
0.0%
(0.0%)
0.0%
(0.0%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Latent Fraction of
Grass
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Cooling % Change
NE Monsoon Season Cooling % Change
SW Monsoon Season Cooling % Change
Checks:
H/L
%Change from Average
∆T below 0.5K
Annual Cooling % Change
NE Monsoon Season Cooling % Change
SW Monsoon Season Cooling % Change
Checks:
H/L
%Change from Average
∆T below 0.5K
Annual Cooling % Change
NE Monsoon Season Cooling % Change
SW Monsoon Season Cooling % Change
Checks:
Albedo of Vegetation
0.25
Low
High
0.10
0.30
(60.0%)
20.0%
100.0%
100.0%
(0.1%)
(0.2%)
(0.0%)
(0.2%)
(0.1%)
(0.2%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
0.73
Low
High
0.00
1.00
(100.0%)
37.0%
100.0%
95.8%
(0.6%)
1.1%
(0.6%)
1.0%
(0.5%)
0.9%
Minor Significant
Minor
Minor
Minor
Minor
Minor
Minor
0.001
Low
High
0.00
0.10
(100.0%)
9900.0%
100.0%
100.0%
------Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Temp Meas Height at
Ref Height
Average Value
1,600,000
Low
High
1,200,000 2,000,000
(25.0%)
25.0%
100.0%
100.0%
0.0%
(0.0%)
0.0%
(0.0%)
0.0%
(0.0%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
0.60
Low
High
0.45
0.75
(25.0%)
25.0%
100.0%
100.0%
(0.3%)
(0.1%)
(0.2%)
(0.1%)
(0.2%)
(0.2%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Rural Vegeration
Fraction
Average Value
Urban Road VHC
Low
2
High
0
10
(100.0%)
400.0%
100.0%
95.8%
(0.6%)
1.1%
(0.6%)
1.0%
(0.5%)
0.9%
Minor Significant
Minor
Minor
Minor
Minor
Minor
Minor
Wall Vegetation
Coverage
Air Velocity
Measurement Height
Low
0.001
0
(100.0%)
100.0%
---Minor
Minor
Minor
Minor
High
10
NA
100.0%
---Minor
Minor
Minor
Minor
Urban Road Albedo
0.10
Low
High
0.08
0.25
(20.0%)
150.0%
100.0%
100.0%
0.0%
(0.0%)
0.0%
(0.0%)
(0.0%)
0.0%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Heat flux threshold for
daytime conditions
Rural Road K
1.00
Low
High
0.75
1.25
(25.0%)
25.0%
100.0%
100.0%
(0.0%)
0.0%
(0.0%)
0.0%
(0.0%)
0.0%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minimum Wind
Velocity
200
0.1
Low
High
Low
High
150
250
0.0001
1.0
(25.0%)
25.0%
(99.9%)
900.0%
100.0%
100.0%
93.1%
95.8%
(0.2%)
(0.2%)
0.1%
(0.6%)
(0.2%)
(0.1%)
0.1%
(0.6%)
(0.2%)
(0.2%)
(0.0%)
(0.4%)
Minor
Minor Significant Significant
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Roof Vegetation
Coverage
0.00
Low
High
0.00
0.10
(100.0%)
9900.0%
100.0%
100.0%
0.0%
(0.0%)
(0.0%)
(0.0%)
0.0%
(0.0%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Heat flux threshold for
nighttime conditions
50.0
Low
High
37.5
62.5
(25.0%)
25.0%
100.0%
100.0%
0.0%
(0.0%)
(0.0%)
(0.0%)
0.0%
(0.0%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Initial Temperature
Low
27
20
(25.0%)
100.0%
(0.0%)
(0.1%)
0.0%
Minor
Minor
Minor
Minor
High
34
25.0%
99.8%
0.0%
0.1%
(0.0%)
Minor
Minor
Minor
Minor
Begin month for veg
participation
Low
1
3
200.0%
100.0%
(0.0%)
(0.1%)
0.0%
Minor
Minor
Minor
Minor
High
6
500.0%
99.8%
0.0%
0.1%
(0.0%)
Minor
Minor
Minor
Minor
Rural Road VHC
1,600,000
Low
High
1,200,000 2,000,000
(25.0%)
25.0%
100.0%
100.0%
0.2%
(0.1%)
0.2%
(0.1%)
0.1%
(0.1%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Maximim Film Water
Depth
0.005
Low
High
0.004
0.006
(25.0%)
25.0%
100.0%
100.0%
(0.2%)
(0.2%)
(0.1%)
(0.1%)
(0.2%)
(0.2%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Rural Average
Obstacle Height
0.10
Low
High
0.08
0.13
(25.0%)
25.0%
100.0%
100.0%
0.2%
(0.1%)
0.2%
(0.1%)
0.1%
(0.1%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
End month for veg
participation
Low
12
6
(50.0%)
100.0%
0.2%
0.2%
0.1%
Minor
Minor
Minor
Minor
High
9
(25.0%)
100.0%
(0.1%)
(0.1%)
(0.1%)
Minor
Minor
Minor
Minor
Rural Road Albedo
0.10
Low
High
0.08
0.25
(20.0%)
150.0%
100.0%
100.0%
(0.1%)
0.1%
(0.1%)
0.1%
(0.1%)
0.1%
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Exchange velocity
coefficient
0.30
Low
High
0.23
0.38
(25.0%)
25.0%
100.0%
99.9%
0.8%
(0.7%)
0.8%
(0.7%)
0.7%
(0.7%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Characteristic Length
1,000
Low
High
100
3,000
(90.0%)
200.0%
100.0%
99.9%
0.0%
(0.1%)
0.1%
(0.1%)
(0.0%)
(0.0%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Night setpoint start
time
Low
19
17
(10.5%)
100.0%
0.0%
0.1%
(0.0%)
Minor
Minor
Minor
Minor
High
21
10.5%
99.9%
(0.1%)
(0.1%)
(0.0%)
Minor
Minor
Minor
Minor
Daytime BL Height
Nighttime BL Height
700
80.0
Low
High
Low
High
525
875
60
100
(25.0%)
25.0%
(25.0%)
25.0%
100.0%
100.0%
96.1%
100.0%
(0.4%)
(0.0%)
(0.0%)
(0.3%)
(0.3%)
(0.0%)
0.0%
(0.2%)
(0.3%)
(0.1%)
(0.0%)
(0.2%)
Minor
Minor Significant
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Floor Height
3.00
Low
High
2.25
3.75
(25.0%)
25.0%
100.0%
100.0%
0.2%
(0.2%)
0.3%
(0.2%)
0.2%
(0.2%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Sensible
Anthropogenic Heat
Low
4.0
2.0
(50.0%)
50.0%
(2.6%)
(2.8%)
(2.5%)
Significant
Significant
Significant
Significant
High
6.0
50.0%
99.3%
2.3%
2.5%
2.2%
Significant
Significant
Significant
Significant
Night setpoint end
time
Low
5
3
(40.0%)
50.0%
(2.6%)
(2.8%)
(2.5%)
Significant
Significant
Significant
Significant
High
8
60.0%
99.3%
2.3%
2.5%
2.2%
Significant
Significant
Significant
Significant
Average Building
Height
Reference height
150.0
Low
High
113
188
(25.0%)
25.0%
99.9%
90.6%
(1.2%)
0.9%
(0.9%)
0.7%
(1.0%)
0.7%
Minor Significant
Minor
Minor
Minor
Minor
Minor
Minor
Site Coverage Ratio
26.0
0.4
Low
High
Low
High
19.5
32.5
0.2
0.7
(25.0%)
25.0%
(47.2%)
84.7%
99.9%
100.0%
99.8%
21.7%
(0.4%)
0.4%
(2.7%)
4.3%
(0.4%)
0.4%
(2.7%)
4.4%
(0.5%)
0.4%
(2.7%)
4.3%
Minor
Minor
Minor Significant
Minor
Minor Significant Significant
Minor
Minor Significant Significant
Minor
Minor Significant Significant
Latent Anthropogenic Percent Heat Released
to Canyon
Heat
Low
0.0
0.0
NA
100.0%
(0.0%)
(0.0%)
(0.0%)
Minor
Minor
Minor
Minor
High
0.6
NA
100.0%
0.0%
0.1%
0.0%
Minor
Minor
Minor
Minor
0.4
Low
High
0.20
1.0
(50.0%)
150.0%
99.9%
34.6%
(4.2%)
4.8%
(3.9%)
4.5%
(4.3%)
5.0%
Minor Significant
Significant Significant
Significant Significant
Significant Significant
Maximum wind
velocity
Max discritization
length for UBL model
10
500
Low
High
375
625
(25.0%)
25.0%
99.9%
34.6%
(4.2%)
4.8%
(3.9%)
4.5%
(4.3%)
5.0%
Minor Significant
Significant Significant
Significant Significant
Significant Significant
Low
1
(90.0%)
100.0%
(0.0%)
(0.0%)
(0.0%)
Minor
Minor
Minor
Minor
High
20
100.0%
100.0%
0.0%
0.1%
0.0%
Minor
Minor
Minor
Minor
Urban Breeze Scaling
Coefficint
1.2
Low
High
0.9
1.5
(25.0%)
25.0%
100.0%
100.0%
(0.2%)
(0.1%)
(0.2%)
(0.1%)
(0.2%)
(0.1%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Façade-to-Site Raio
Low
1.6
1.2
(25.0%)
100.0%
(0.0%)
(0.1%)
0.0%
Minor
Minor
Minor
Minor
High
225
75
(66.7%)
100.0%
(0.1%)
(0.1%)
(0.2%)
Minor
Minor
Minor
Minor
Urban Road Vegetation
Fraction
Low
0.2
High
1.9
25.0%
100.0%
(0.1%)
(0.1%)
(0.1%)
Minor
Minor
Minor
Minor
0.0
(100.0%)
100.0%
0.1%
0.1%
0.1%
Minor
Minor
Minor
Minor
High
335
48.9%
100.0%
(0.2%)
(0.2%)
(0.1%)
Minor
Minor
Minor
Minor
0.19
Low
High
0.00
0.29
(100.0%)
50.0%
100.0%
99.9%
0.4%
(0.4%)
0.4%
(0.4%)
0.4%
(0.4%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Cooling Capacity
Low
Latent Fraction of
Trees
0.7
Low
High
0.5
0.9
(25.0%)
25.0%
100.0%
100.0%
0.0%
(0.4%)
0.0%
(0.3%)
(0.0%)
(0.3%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
0.6
200.0%
100.0%
(0.1%)
(0.1%)
(0.1%)
Minor
Minor
Minor
Minor
Tree Coverage
[page intentionally left black]
Table 7 Combined sensitivity analysis results for Boston, MA using Logan and Bedford reference sites and for
Punggol, Singapore
SITE COVERAGE RATIO
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
FAÇADE-TO-SITE RATIO
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
SENSIBLE ANTHROPOGENIC HEAT
Average Value
Value
%Change from Average
∆T below 0.5K
Annual Heating/Cooling % Change
Winter Heating % Change
Summer Cooling % Change
Checks:
Boston-Logan
0.51
Low
High
0.20
0.70
(60.8%)
37.3%
97.3%
25.9%
4.9%
(7.0%)
6.2%
(10.1%)
(2.1%)
5.3%
Significant Significant
Significant Significant
Significant Significant
Significant Significant
Boston-Bedford
Punggol
0.51
0.38
Low
High
Low
High
0.20
0.70
0.20
0.70
(60.8%)
37.3%
(47.2%)
84.7%
90.9%
16.2%
99.8%
21.7%
5.6%
(8.0%)
(2.7%)
4.3%
7.8%
(11.5%)
(2.7%)
4.4%
(3.1%)
5.0%
(2.7%)
4.3%
Significant Significant
Minor Significant
Significant Significant Significant Significant
Significant Significant Significant Significant
Significant Significant Significant Significant
Boston-Logan
Boston-Bedford
4.87
4.87
Low
High
Low
High
3.65
6.09
3.65
6.09
(25.0%)
25.0%
(25.0%)
25.0%
99.1%
99.8%
97.2%
97.9%
2.0%
(1.0%)
1.7%
(1.1%)
2.0%
(1.4%)
1.9%
(1.2%)
(0.0%)
0.5%
(0.1%)
(0.2%)
Significant
Minor Significant Significant
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Punggol
1.55
Low
High
1.16
1.94
(25.0%)
25.0%
100.0%
100.0%
(0.0%)
(0.1%)
(0.1%)
(0.1%)
0.0%
(0.1%)
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Minor
Boston-Logan
Boston-Bedford
Punggol
67.75
67.75
4.00
Low
High
Low
High
Low
High
-135.50
-135.50
2.00
6.00
(100.0%)
100.0%
(100.0%)
100.0%
(50.0%)
50.0%
99.5%
94.6%
91.4%
82.0%
50.0%
99.3%
1.6%
(1.4%)
1.5%
(1.4%)
(2.6%)
2.3%
2.2%
(2.7%)
2.8%
(2.7%)
(2.8%)
2.5%
(2.2%)
3.2%
(3.2%)
3.1%
(2.5%)
2.2%
Minor Significant Significant Significant Significant Significant
Minor
Minor
Minor
Minor Significant Significant
Significant Significant Significant Significant Significant Significant
Significant Significant Significant Significant Significant Significant
Of the 42 parameters tested in the sensitivity analysis, only three parameters were
determined to be significant. From user interface perspective, this means that the required
user inputs can be greatly reduced and we can assign default values for most of these
parameters including the empirical and city-specific values such as meteorological
parameters. The metrics of the sensitivity analysis shows that the urban temperature
estimate will be within 0.5% accurate even when UWG is ran with default values for these
parameters. These less significant parameters will be moved to the advanced setting, so the
advanced users can choose to run in-depth analyses to refine their assumptions and
designs.
50
The wrapper for Urban Weather Generator was developed with feedback from urban
design and planning practitioners as well as energy consultants to create a useful and
usable tool towards a fully integrated climate-based design in architecture. The current
version of the user interface design for both the stand-alone and umi-integrated versions
are shown with design considerations for the tools’ usability.
This chapter explains the design process of the wrapper for UWG. As this is the first
design tool for UHI modeling, the initial interface design is based on the existing and widely
used interfaces (i.e. DesignBuilder and umi, introduced in Section 2.2 and explained in the
following section) so that users can easily familiarize themselves with the environment.
The design is improved with feedback from potential users for new functionalities and
interface design.
The user base includes urban designers and planners as well as energy consultants.
Based on the understanding that these users have different workflows, two interfaces were
created based on the following assumptions:
1. Stand-alone version: The intended users of the stand-alone version are energy
consultants, urban planners, and other people who are interested in evaluating the
effects of urban morphologies or building use on urban thermal comfort. They do not
generate urban forms using 3D modeling tools themselves but may work with urban
designers who can provide them with these inputs. The frequent users are experts of
energy simulation tools such as those based on EnergyPlus introduced in Section 2.2.
2. Rhino-integrated version: The second version of the tool will be developed for urban
designers and architects who use Rhinoceros to perform massing and formal studies.
With this integrated tool in umi, users can incorporate UHI as a driver for decision
making in their formal designs. To develop the user interaction and workflow, we will
leverage the fact that some users are already familiar with existing energy simulation
tools for Rhino such as umi, Archsim, and Hobeybee.
51
Users of 3D modeling tool other than Rhinoceros can also use the stand-alone version of
UWG’s wrapper to evaluate their designs.
The C# programming language is developed by Microsoft using Visual Studio and the
.NET framework and is designed to be strongly-typed and object-oriented. C# is selected
for UWG’s development as it is one of the languages supported by RhinoCommon
[RhinoCommon SDK, 2014]. This allows us to access commands unique to Rhino and thus
develop a plug-in for Rhino. umi is also developed in C#. The Visual Studio .NET platform is
a software framework also developed by Microsoft for Windows operating system and
includes a large class library to facilitate programming as well as Windows Presentation
Foundation (WPF) graphical subsystem for designing the user interface used here.
Currently UWG’s Matlab algorithm is packaged into an executable file, which runs in the
Windows environment. The program can be used from command line prompts or be
programmatically called using C#. This chapter explains how the wrapper was created for
this executable file succeeding the earlier works by Sullivan (Section 2.1.4). The standalone UWG UI is an executable file for Windows operating system. The Rhino-version will
be distributed as an RHP file (plugin file primarily associated with Rhino) and loaded using
Rhino's PlugInManager command. From here on, the wrapper for UWG will simply be
referred to as UWG. The output of UWG simulation is a morphed epw file and that of the
template editor is an xml file.
The first version of the user interface (UI) for UWG is developed to test the organization
of the xml data as well as to provide the most basic functionality to help create the input
xml files for UWG. A few images of the initial UI design is shown in Figure 20 and Figure 21.
There are no existing UHI modeling tools available, so the interface will be modeled
after existing environmental performance simulation tools, and in particular energy
simulation tools such as DesignBuilder and umi shown in Figure 5 and Figure 10. A tabbased organization similar to these tools is used for UWG to help guide through simulation
steps and makes clear the hierarchy of information. Each subsequent tabs from top to
bottom represents different tasks for creating the xml file followed by the tab for running
the simulation. This structure helps to organize the large number of parameter inputs as
follows and is diagrammatically represented to construct the data structure for
programming and the resulting xml input file (Figure 22):
1. Building construction: assignments of wall, roof, building mass, glazing, and
urban and rural road material properties and thicknesses
2. Building: parameters defining internal loads from occupants, equipment, and
lighting.
52
3.
4.
5.
Urban area: geometric parameters and anthropogenic heat defining the urban
space
Reference site: information concerning the measurement of the reference weather
data at the rural site
Run UWG: file selections for running the simulation
53
54
Figure 20 The first GUI prototype. The interface is tab-based and guides users through the xml file creation
process (top). The advanced settings are provided with default values to speed up the setup process (middle).
The users then save the xml file in the desired location (bottom).
Figure 21
In the final “Run UWG” tab, a user selects the xml and epw files before they start the simulation.
55
Figure 22 The data structure for the GUI and xml file. Each layer of the building construction except for
glazing takes the same hierarchical structure as Walls > Layer 1 and similarly for building schedules
The focus of this version is to test the data organization using the most basic UI and so
there is no conceptual separation between creating the xml file (tabs 1 – 4) and running the
simulation (tab 5). This is improved in the next iteration. For the same reason this version
does not yet have the editing capabilities for existing xml files. The tab-based design
pattern used for UWG is similar to wizards that are often used in software installation for
improving the learnability of a complex interaction, by structuring it as a step-by-step
process. If the user makes a mistake, he/she can easily go back to the correct location to fix
it.
Using the results from the sensitivity analyses from Chapter 4, the parameters that do
not contribute to the UHI are moved to the Advanced Setting to facilitate the simulation
setup process. The advanced users who are familiar with urban heat flow and
thermodynamics are able to change these parameter values from the Advanced Setting
expander or directly in the xml files to fine-tune their assumptions.
56
The program was tested with seven urban design practitioners and novice users who
have not used design simulation tools before as well as five energy consultants who have
previously used other environmental performance simulation tools. The testers were asked
to interact with the tool independently with the author in the same room. The observations
and their feedback are documented below.
All users were able to create the xml files successfully without intervention. They
navigated the tab-based organization when asked to change certain input parameters. The
energy consultants were especially adept because the interface structure is similar to other
simulation tools. This confirms the success of the basic data structure in Figure 22.
The problem arose in the “Run Simulation” tab when some users skipped input fields in
the xml file creation process (tabs 1 – 4). This phenomenon could be from fatigue or
because users intentionally skipped them. They explained that they meant to go back to fill
in the input field after they look up the appropriate value for that field. Also, some people
prematurely pressed the “Run Simulation” button. As there are no error checking
implemented in this version, the UWG simulation freezes when there are missing inputs.
This poses a challenge to the usability of the tool and confirms the necessity for including
safety measures to prevent system crash as well as to provide feedback to the users to go
back and correct their xml input file when necessary.
Many energy consultants suggested including template editors for construction
assemblies and building use schedules so that they can easily recreate the same building
settings in the future as well as easily parameterize particular inputs such as insulation
material properties and thicknesses. The default template would also help novice users to
understand the expected value ranges for technical inputs. Therefore, this functionality has
been included in the next iteration of the tool development.
Based on the feedback from the first prototype, the structure of the tool is maintained
and new features are added. In addition to the template editor, the results viewer is added
to the UI to help analyze the results of the UWG simulation. As there are now three tasks
that can be done with the tool, namely (1) make xml; (2) run simulation; and (3) view
simulation results, these items are organized into the high- level hierarchy instead of a
linear structure used in the first version. Other helper functions (i.e. Open, Save) will be
placed in the dropdown (second-level hierarchy).
Figure 23 demonstrates the new structure of the tool with three separate tasks and
their make-up. The template editor is used to (1) make xml, so the editor can be called from
that section. The simulation can only be run from (2) run simulation tab to prevent
accidental button presses. (3) View simulation results tab asks users to provide the
morphed epw file (i.e. output of the tool) to visualize and plot the results. The interaction of
inputs and outputs for each tab are also shown in Figure 23. The xml file from the current
or previous sessions can be used for (2) run simulation. The simulation allows for up to 6
parallelized runs. Similarly, (3) results viewer can be used with morphed epw file from the
57
current or previous sessions to compare results from previous simulations. This diagram is
uses as the basis to design the improved interface.
Following the success of the data organization (Figure 22), the template editor is
similarly structured for the building construction and schedules (Figure 24). In fact, the
building construction follows exactly the same structure. The user interface previously
required users to provide a numerical input value for daytime and nighttime internal loads
(in the building tab) that altogether represent equipment, lighting, and occupancy loads. On
the other hand, the new UI includes the template editor to calculate these required UWG
inputs from a bottom-up approach using hourly building use schedules and maximum
values for each internal load types. This is the approach used by other energy simulations
and should be familiar to energy consultants. Using the template inputs the daytime and
nighttime internal heat gains as well as infiltration and ventilation rates are calculated as
the weighted average of each internal heat gain contributions (i.e. occupancy, lighting, and
equipment) for 6am - 5pm and 6pm – 5am, respectively.
The developed tool and the detailed documentation is available from our research
website http://urbanmicroclimate.scripts.mit.edu/. Some snapshots of the current version
of the tool (as of April 29, 2015) are shown in the next section (collectively referred to as
Figure 25) to help illustrate the usability concepts of the new features in the context of user
interface and interaction designs. The usability is defined as how well users can use the
system’s functionality and is evaluated based on learnability, efficiency, and safety. Each
aspect is explained in the following section.
Figure 23 UWG is organized by the users’ goals. The interaction between different components of the UWG
UI is shown, where the output from each functions or already existing files can be used to run another
successive feature of the GUI.
58
Figure 24 The data structure for the template library (right). The interaction between the template library
and xml file editor is exhibited on the left
59
The tab-based navigation is used. A user can select templates created via the template editor (see below)
to assign building construction materials for each building element as well as for building internal loads.
If “New template” is selected from the dropdown menu the template editor is invoked.
60
The “Urban Area” tab asks for inputs specific to the site modeled. The parameters under the advanced
setting have small influence on the UHI and are given default values. Advanced users can change the
parameter values by expanding this section (increase learnability and efficiency).
If the user tries to exit the template construction without saving, the program asks them if they want to
save it before creating a new file or running a simulation. This prevents accidental loss of files and helps
to teach users to save (increase learnability).
61
Template editor
The template editors help to speed up the simulation set up time. The selections in the assembly tab is
displayed in the main interface so the users do not need to go back to the template editor each time to
remember details of their selection (increase efficiency).
62
The glazing inputs are also similar to those required by existing energy simulations.
The building internal gain schedules and values can be set in the “Building” tab similarly to the
construction tab. The user can choose to either type the values in the textbox input or drag the bars for
setting hourly building use schedule (increase efficiency).
63
The user uses the dropdown menu to select schedules. They can take advantage of the autofill feature
(increase efficiency).
The dialogue box appears when a user tries to overwrite existing file to reduce accidental data loss
(increase safety).
64
Run Simulation
“Run simulation” tab helps user select the input files and set the location to save the output files. Up to six
simulations can be run simultaneously (increase efficiency).
65
The “Start UWG” button is deactivated until all of the simulation inputs are entered (increase safety).
A popup opens up to indicate the progress of the simulation once it is invoked (increase learnability).
Figure 25 Select snapshots of the UI. Individual captions indicate how each design considerations improve
the usability of the tool
66
The learnability aspect of UI design concerns how easy it is to learn to use the tool. It
increases with consistency of layouts, affordance (actionable properties), and wording as
well as clarity of navigation and feedbacks.
Our interface uses a grid structure to give a sense of consistency in layout such that
users can expect to see the parameter name and corresponding input fields in the same
general location for each section of the GUI. The clarity and consistency of the affordance is
trivial yet important in our form-based tool so that the UI can communicate to the user how
it should be used. Specifically, users need to know that input values are expected for all text
fields, buttons triggers an event (i.e. run simulation, open file selection dialogue), and
expanders appear optional yet still able to be expanded to change values for advanced
users.
The wording and parameter names are consistent with existing energy simulation tools.
The parameters specific to UWG such as site coverage ratio and façade-to-site ratio are
renamed from horizontal building density and vertical to horizontal built ratio based on
existing zoning terms [New York Department of City Planning, 2011] that are familiar to
urban designers. In case a user needs to refer to a definition or example values for any
parameter, the glossary of terms, sample values, and references are available on the
research website and can be directed from the Help button on the UI. The appendix of
terms is also provided in Appendix A.1 in this document.
In addition, a tab-based navigation is used in this iteration following the success from
the first prototype. The highlights allow users to know which tab they are on and how far
they are from completing all entries. For advanced users, the parameter names and the xml
tag names are matched so they can easily identify the parameter and modify the xml files
directly. Furthermore, the xml file structure matches that of the tab names to facilitate
navigation.
Finally, the feedbacks are major improvements for the second version of the UI. The
users get visual feedback when the simulation is starting and the progress of the simulation
so they can estimate the completion time. In addition, the template editor asks if they want
to overwrite a file to avoid accidental overwriting. If an error occurs, the dialogue box
indicates how the user can correct the issue to prevent the same error in the future. The
error messages are translated to natural language, from command language in
programming (i.e. Matlab and Visual Studios) to provide a constructive feedback.
The efficiency of the UI is defined as how fast the tool can be used once a user learns the
tool. Since we are aiming for repeated use of the tool, efficiency is an important aspect.
As such, template editors for building construction and internal loads are added to this
iteration. Templates can be reused and edited to quickly test different insulation
thicknesses, building assembly, building use schedule, etc. Additionally, the installation
package of the UWG includes a default template file for common building construction
materials and internal loads to help get users started quickly.
The advanced settings are efficiency features of the tool. The default values are assigned
to parameters that have relatively small influences to the UHI. Advanced users can tweak
these values but it is not necessary for running the simulation. As a result of the sensitivity
67
analysis in Chapter 4, 46% of the required inputs are removed from the main interface into
advanced setting.
Finally, the “Run simulation” tab has the ability to run up to six variations of the xml
input file concurrently as a time-saving measure.
The UI should cause few errors, and if any, be recoverable. In UWG’s UI, the safety
measures are implemented for reducing accidental loss of data and to prevent the
simulation from crashing when an invalid input file is provided.
For example, when the user tries to exit from the current xml file without saving (i.e.
selects “Create a new xml file” or hits “Run simulation” tab prematurely), a dialogue box
pops up to ask if they want to save the current file. This ensures that users do not lose their
newly created xml file. Similarly, in the template editor the program confirms with users
that they want to overwrite an existing template.
Good feedback and safety mechanisms are necessary for the Matlab simulation because
it is a time consuming process (8 minutes on Intel® Core™ i3-3225 CPU @3.30GHz on
Windows 7 operating system). Safeguards are implemented along each step of the
simulation. First, the simulation button is deactivated until all required file inputs and
names are provided. Second, once the “Run UWG” button is pressed, the program checks if
all input values are present in the xml file before calling the Matlab compiled executable
file. Third, after the simulation starts, it can still be aborted if the user realizes that he/she
made an error with a simple click of a “Cancel” button. As mentioned in the Learnability
section, the error messages provide constructive feedback on how to fix the problem to
help prevent the same errors in the future.
The Help button on the UI takes the user to the detailed documentation provided on the
research website. It is created with the same idea of user-friendliness and usability as the
UI development. The website includes a searchable glossary that provides the definition
and recommended values for each parameter. It also gives recommendation on which
parameters to focus their design efforts based on the sensitivity analysis.
The Rhino version of the GUI has a similar UI as the stand-alone version and is created
as a modal that is invoked by typing “UmiRunUWG” in Rhino’s command line while the user
is in an umi project (Figure 27). It takes advantage of Rhino and umi’s functions to
automatically calculate geometric parameters and other model information. Seven
geometric parameters are calculated from the user’s model and passed to UWG through
umi. These parameters are average building height, site coverage ratio, façade-to-site ratio,
characteristic length, average window-to-wall ratio, average U-value, and gross floor area
as described in Figure 27. It eliminates the current process of using Grasshopper (Figure
14) to extract such parameters and thus further improves the efficiency of the UI. We have
also removed these parameter inputs from the “Urban Area” tab in the GUI as it is no longer
necessary to provide manual inputs for them.
As mentioned, the interaction of the Rhino version is similar to that of the stand-alone
tool, including the template editors. This version will be available in the next release of umi
and will be available at http://urbanmodellinginterface.ning.com/. Using the morphed
68
weather file from the UWG, users can then run energy simulation within umi. The new
workflow is demonstrated in the next chapter.
Figure 26 The interaction of UWG with umi. Urban area geometric parameters are extracted directly from
the user’s Rhino model
Figure 27 The UWG GUI is invoked inside umi by calling “UmiRunUWG”. The interaction is similar to that of
the stand-alone version
69
The results viewer is currently under development. The goal is to help visualize and
compare results from the UWG simulation. The mockup is shown below in Figure 28. The
user can compare up to six simulation runs based on the UHI (dry bulb temperature
change) as well as thermal comfort metrics explained in the following section 5.3. The same
design language as the rest of the tool will be used for the UI design.
Figure 28
Mockup of the results viewer for the stand-alone version
The main focus of this thesis work is to create a new architectural design simulation
tool for thermal comfort and energy performance. The metrics therefore will be based on
these concepts and include Universal Thermal Climate Index and dry bulb air temperature
for measuring thermal comfort. Heating and cooling energy will be evaluated for the umiversion of this tool. These metrics will be used for results viewer functionality of the UI.
The diurnal dry-bulb temperatures are used to analyze how UHI shifts the diurnal
temperatures as shown in Figure 3. Monthly diurnal averages are observed to evaluate the
effects for heating and cooling seasons. The heating effect of each model will be evaluated
70
based on the temperature difference between the urban and the rural reference sites,
comparing the average dry bulb temperatures for each hour in the particular month.
The Universal Thermal Climate Index (UTCI) [Bröde et al., 2010] is an outdoor thermal
comfort metric where the UTCI Equivalent Temperature of an actual thermal condition is
the air temperature of the reference condition causing the same dynamic physiological
response. In 2005, the European Union funded within the European Cooperation in Science
and Technology (COST) Action 730 the development of UTCI, which aims at the assessment
of the outdoor thermal conditions in the major fields of human biometeorology. This is
based on the physiological response of the human body and considers (1) the behavioral
adaptation of clothing insulation observed for the general urban population in relation to
the actual environmental temperature, (2) the distribution of the clothing over different
body parts providing local insulation values for the different model segments, and (3) the
reduction of thermal and evaporative clothing resistances caused by wind and the
movement of the person walking at 4 km/h on a leveled surface.
UTCI is based on the concept of an equivalent temperature (Figure 29). A reference
environment is defined: 50% relative humidity (but not exceeding water vapor pressure of
20hPa) with still air and radiant temperature equal to air temperature. Equal physiological
conditions are based on the equivalence of the dynamic physiological response predicted
by the model for the actual and the reference environment. As this dynamic response is
multidimensional (body core temperature, sweat rate, and skin wettedness at different
exposure times), a single dimensional strain index is calculated by principal component
analysis. The UTCI equivalent temperature for a given combination of wind, humidity,
mean radiative temperature (MRT), and air temperature is then defined as the air
temperature of the reference environment which produces the same strain index value. As
the thermoregulation model is computationally expensive, polynomial regression
equations are used to predict the UTCI equivalent temperature values. The code is written
in Fortran [UTCI_a002, 2009] and is available from http://www.utci.org.
The metric has been validated for reference conditions, and sensitivity analysis has
been done for each input parameter. The responses of UTCI to changes in air temperature
and relative humidly are in agreement with the psychrometric charts obtained for
physiological data from human experiments [Kampmann & Bröde, 2009]. In the cold, UTCI
indicates a more pronounced effect for wind speeds above 3 m/s compared to the wind
chill temperature because UTCI considers the dynamic response of the whole body. Finally,
the radiation effect on UTCI is within the magnitudes obtained for Ergonomics Standards
concerned with the thermal environment Predicted Heat Strain [ISO 7933, 2004] index and
Wet Bulb Globe Temperature [ISO 7243, 1989].
The associated assessment scale for UTCI (˚C) is derived from the simulated
physiological responses and comprises ten thermal stress categories ranging from extreme
cold stress to extreme heat stress is:
T ≥ +46
+38 ≤ T< +46
+32 ≤ T< +38
extreme heat stress
very strong heat stress
strong heat stress
71
+26 ≤ T< +32
+9 ≤ T< +26
0≤ T< +9
-13 ≤ T< 0
-27 ≤ T< -13
-40 ≤ T< -27
T < -40
moderate heat stress
no thermal stress
slight cold stress
moderate cold stress
strong cold stress
very strong cold stress
extreme cold stress
(1)
Figure 29 Concept for UTCI, defined as air temperature of the reference condition yielding the same
dynamic physiological response. Image credit: Figure 1 from Bröde et al. (2010)
As mentioned, UTCI takes air temperature, relative humidity, air velocity, and mean
radiative temperature. In assessing the UTCI of the urban site, the UWG outputs of the
morphed air temperature and relative humidity can be directly used. UWG does not
calculate urban wind velocity or output MRT, but the following methods can be used to
estimate these parameters.
The average wind speed UH at height H above the average height of local obstacles can
be calculated from the wind velocity measured at the reference site [ASHRAE Handbook of
Fundamentals, 2013]:
𝛿𝑚𝑒𝑡 𝑎𝑚𝑒𝑡 𝐻 𝑎
𝑈𝐻 = 𝑈𝑚𝑒𝑡 (
)
( )
𝐻𝑚𝑒𝑡
𝛿
The boundary thickness δ and exponents a for the local building terrain and for the
meteorological station can be determined using the correlation found in the reference
above. For example, for city centers in which at least 50% of buildings are higher than 25m
over a distance of at least 0.8km or 10 times the height of the structure upwind (whichever
is greater) ,we can use a = 0.33 and δ = 460m. For meteorological stations with
72
anemometer placed 10m above the ground level, amet = 0.14 and δmet = 270m. Then we have
urban wind velocity UH ≃ 0.25Umet.
The MRT is calculated from radiant temperatures for the sky, wall, and road calculated
from UWG [Bueno et al., 2012a]. The MRT is the uniform temperature of an imaginary
enclosure in which radiant heat transfer from the human body equals the radiant heat
transfer in the actual non-uniform enclosure. It can be calculated from the measured
temperature of surrounding walls and surfaces and their positions with respect to the
person. For surfaces with high emittance ε, we can assume a black body and MRT can be
calculated as the following [ASHRAE Handbook of Fundamentals, 2013]:
𝑇𝑀𝑅𝑇 = 𝑇14 𝐹𝑝−1 + 𝑇24 𝐹𝑝−2 + ··· + 𝑇𝑁4 𝐹𝑝−𝑁
where
TMRT = mean radiant temperature, °C
TN = surface temperature of surface N, °C
Fp-N = angle factor between a person and surface N
In our case we are interested in the MRT of a person standing in the urban canyon,
surrounded by two walls and road, and the sky. The equation above then becomes
4
4
4
𝑇𝑀𝑅𝑇 = 𝑇𝑤𝑎𝑙𝑙
𝐹𝑝−𝑤𝑎𝑙𝑙 + 𝑇𝑠𝑘𝑦
𝐹𝑝−𝑠𝑘𝑦 + 𝑇𝑟𝑜𝑎𝑑
𝐹𝑝−𝑟𝑜𝑎𝑑
The view factors are calculated for (1) a person standing on the side walk (0.76m away
from the building, or middle of the sidewalk with width of 1.525m (60in) [U.S. Department
of Transportation Federal Highway Administration, 2014] and (2) in the middle of the
canyon (i.e. campus planning and parks). The view factor for rectangular surfaces depends
on the position and orientation of the person [Fänger, 1982] as shown in Figure 30.
Observing that Fänger’s graphs rise with exponential law, Cannistraro et al. (1992)
developed an algorithm to fit his angle factor graphs for horizontal and vertical rectangular
surfaces:
𝐹𝑝−𝑁 = 𝐹𝑠𝑎𝑡,𝑚𝑎𝑥 (1 − 𝑒 −
𝑎/𝑐
𝜏 ) (1
− 𝑒
−
𝑏/𝑐
𝛾 )
where
𝑎
𝜏 =𝐴+𝐵𝑐
𝑏
𝑎
𝛾 =𝐶+𝐷𝑐+𝐸𝑐
and a/c and b/c are determined according to Fänger’s conventions for the vertical and
horizontal surfaces per Figure 30. As with Fänger’s method, Cannistraro’s algorithm uses
the additive property of the angle factors and is calculated separately for the wall, sky, and
road for below and above the center of gravity for a standing person at 1.1m.
For our case with the rectangle on the side wall, forward or behind, above or below the
center of the standing person (STK2): Fsat,max = 0.119, A = 1.50894, B = 0.13674, C =
0.66134, D = 0.07363, and E = .0.06443. Variables b and c are canyon heights or width
depending on the surface for which the view factor is calculated.
73
The view factors are calculated for UWG simulation for each hour. The UI visualizes
UTCIs for both the sidewalk and the center of the urban canyon through a dropdown
selection. The calculated values conform to those calculated using Fänger’s method. The
code used in UWG is re-written in C# from Fortran and can be viewed and downloaded
from https://github.com/aiko-nakano/Urban-Weather-Generator/tree/master/UTCI_calc.
Figure 30 Mean value of angle factor between seated person and horizontal or vertical rectangle when
person is rotated around vertical axis [Fänger, 1982]
The UTCI values are calculated for each hour of the simulation and then put into the
appropriate bins to categorize the thermal comfort for the whole year. The UHI and climate
change are on the order of 2˚C – 5.4˚C and thus the standard bin sizes for UTCI in (1) are
too large. Thus, the bin size is doubled to observe better the change in thermal comfort.
There are 18 bins with the following temperature ranges. Note that the bin sizes are not all
equal, as with the conventional UTCI index.
T ≥ +46
+42 ≤ T< +46
+38 ≤ T< +42
+35 ≤ T< +38
+32 ≤ T< +35
+29 ≤ T< +32
+26 ≤ T< +29
+17.5 ≤ T< +26
+9 ≤ T< +17.5
74
extreme heat stress
very strong heat stress
very strong heat stress
strong heat stress
strong heat stress
moderate heat stress
moderate heat stress
no thermal stress
no thermal stress
+4.5 ≤ T< 9
+0 ≤ T< 4.5
-6.5 ≤ T< 0
-13 ≤ T< -6.5
-20 ≤ T< -13
-27 ≤ T< -20
-33.5 ≤ T< -27
-40 ≤ T< -33.5
T < -40
slight cold stress
slight cold stress
moderate cold stress
moderate cold stress
strong cold stress
strong cold stress
very strong cold stress
very strong cold stress
extreme cold stress
To evaluate the energy implications of the UHI, heating and cooling energy
consumptions will be used as the metrics for the umi-version of this tool. The underlying
simulation engine is EnergyPlus.
As discussed in this chapter, the stand-alone and Rhino versions of the GUI are
developed for UWG with usability principles from the user interface design. The user
interface evolved as the result of iterative design and user testing, and the metrics were
chosen to measure the thermal comfort and energy implications of the UHI. The goal of this
project is to create an integrated design tool such that the urban designers can consider the
UHI effect as they evaluate their massing design. For this reason the Rhino version of the
tool is created to encourage early integration in the massing design process to help develop
urban designs that are thermally comfortable and also energy-efficient. The new workflow
is demonstrated through a case study in Cambridge, MA in the next chapter.
75
The goal of this case study is to demonstrate the new workflow for urban design using
UWG. We use MIT’s East Campus development as a quintessential example of current
urban design process where the urban design concepts and massing designs are developed
independently of the environmental performance. We propose alternative massing design
strategies to MIT’s plan by using thermal comfort and energy efficiencies as drivers for the
design selection. The newly proposed scheme is then evaluated with the effects of climate
change along with UHI to ensure thermal comfort at every phase of the development.
MIT’s campus is situated across the Charles River from Boston (Figure 31) and is
bordered by research centers and high technology companies in Kendall Square (Figure
32) The Cambridge Research Park/Kendall Square PUD (1999) masterplan [Cambridge
Community Development Department, 2013] helped expand the biotech emphasis, creating
a successful public space with a plaza and canoe access to Charles River, high-rise housing
complex, and ground floor retail.
The MIT East campus urban design study is part of the MIT 2030 initiative [MIT 2030
East Campus Urban Design Study, n.d.] that aims to improve the MIT campus and Kendall
Square to meet future academic and research needs. The vision is to create a gateway to
Kendall Square, enhancing the connection and foster innovation between MIT and
commercial partners nearby:
“The goal of the East Campus Urban Design Study is to help MIT create a long-range
development framework that shapes future academic, residential, and commercial
uses for its properties in the Kendall Square area. The initiative also seeks to create a
vibrant gateway into MIT’s east campus
— MIT 2030
76
The City of Cambridge’s approval for rezoning in 2013 (PUD-5 ordinance) gives MIT the
ability to propose over a million square feet of new development—a combination of
housing, retail, lab, commercial, and open space. MIT also continues to have the right to
pursue 800,000 square feet of new academic development.
The urban design study was performed by three design groups: Elkis Manfredi, MIT
faculty design group [MIT Design Committee, 2013] (referred from here on as “MIT faculty
study”), and Mack Scogin Merrill Elam Architects and Michael Van Valkenburgh Landscape
Architects (2014a) (referred from here on as “East Campus Study”).
2009
MIT began working with Elkus Manfredi Architects to create a
conceptual approach. Over the next three years, the Institute
engaged in a community-wide effort to align the interests of a broad
group of stakeholders around a future vision for the area
April 2013
The Cambridge City Council approved MIT’s rezoning petition for the
new development around Kendall Square (Section 6.1.2)
A faculty design group formed by Adèle Santos, the dean of the
School of Architecture and Planning at the time, proposed an
alternative plan
Fall 2013
MIT commissioned an urban design study and assembled a team
lead by Mack Scogin Merrill Elam Architects and Michael Van
Valkenburgh Landscape Architects
July 2014
Requests for Proposals (RFPs) sent to architectural design teams
Figure 31 MIT and surrounding Kendall districts in Cambridge, MA, U.S.A. (Image credit: Cambridge
Community Development Department (2013))
77
Figure 32 MIT is bordered by innovative companies in the Tech and Biotech industries on Main Street (solid orange line). The urban design concept
is to provide a better gateway between the MIT campus and these companies in Kendall Square and extend it to the MIT campus on both east and
west directions (hashed orange lines) explained in detail in Section 6.2 (Image credit: MIT Design Committee (2013)).
78
The new Section 13.8 PUD-5 ordinance, excerpted in Appendix D [MIT Office of the
Executive Vice President and the Treasurer, 2013], is intended to allow mixed-use
development with increased development densities and heights on the site of East Campus.
It should be noted that the City of Cambridge is proposing a rezoning of the surrounding
Kendall Square area to allow for further commercial, residential and institutional growth.
MIT’s Campus Planning and Investment Management Company is part of this advisory
board.
MIT has 426,000m2 building capacity in East Campus (Table 8). Figure 33 shows the
new height restriction per the PUD-5 ordinance. The maximum floor area ratio (FAR), the
ratio of a building's gross floor area to the size of the land upon which it is built, is 3.9 and
the minimum publicly beneficial open space for PUD-5 is 15% (PUD-5 13.87.1) [Ordinance
Number 1355, 2013].
Table 8 Allowable program and building capacity in PUD-5 from City of Cambridge ordinance Number 1355
157,000m2
211,370m2
30,660m2
27,560m2
426,590m2
Figure 33
New gross floor area by use on area of study
maximum 82,700m2 commercial
maximum 74,300m2 academic
Total existing buildings on area of study
New gross floor area by use on One Broadway site
minimum 22,300 m2 residential
8,360m2 commercial
Existing building on One Broadway Site
Total building capacity on PUD-5
Building height zoning for East Campus (East Campus Study page 15 of Final Report 1 of 3)
79
Cambridge has adopted the optional Massachusetts Stretch Energy Code [Summary of
the Massachusetts Stretch Energy Code, 2011], requiring new buildings to reduce energy
by at least 20% over the state energy code. Furthermore, the city has developed zoning
amendments requiring LEED for projects, rules to encourage passive energy strategies and
green roofs, and guidelines for solar and wind energy installations. The Eastern Cambridge
Kendall Square Open Space Planning Study is developing a plan for improving and
expanding public open spaces in the neighborhood adjacent to the East Campus area
[Cambridge Community Development Department, 2013].
In addition, the City of Cambridge is considering passing the Net Zero Carbon
Ordinance, which proposes to amend the Zoning Ordinance of the City of Cambridge to
ensure that all new construction in Cambridge achieves net zero annual greenhouse gas
emissions. Net zero is defined as an annual balance of zero greenhouse gas emissions from
building operations citywide, achieved through improved energy efficiency and carbon-free
energy production [Executive Summary Prepared for the Cambridge Getting to Net Zero
Task Force, 2015].
If this petition is approved, MIT will need to realize any future campus expansion under
the constraint that its overall greenhouse gas emissions do not rise above the current level.
The case study will consider energy reduction measures to help meet this initiative.
The design proposal is based on the aforementioned East Campus Urban Design Study
by Mack Scogin Merrill Elam Architects and Michael Van Valkenburgh Landscape Architects
(2014a) as well as the East Campus / MIT Gateway Alternative Approaches prepared by the
working group composed of faculty and staff of the School of Architecture and Planning
[MIT Design Committee, 2013].
MIT plans to expand the lots that are closest to the existing campus between the MIT
Media Lab and the Sloan School of Management (Figure 34). Most of the site is currently
used as outdoor parking lots. The open spaces in the East Campus and the off-campus
public space in the north of Main Street are currently experienced as isolated islands rather
than as an integrated system (East Campus study). Thus, the concept developed by the
design team of MIT faculty (Figure 35) is to enhance the connection and continuity of the
surrounding areas.
More specifically, the goal is to create a better arrival experience to Kendall Square and
the MIT campus from the Kendall Square subway station as well as to enhance the
connection to the main campus with the Infinite Corridor which spans 251 meters and runs
through the main buildings of the MIT campus. The idea is to use this new development as
an opportunity to increase the sense of connectivity throughout campus from the east to
the west end. The connection extends beyond the MIT campus to its surrounding
technology hubs in Kendall Square to foster innovation and collaboration between MIT
researchers and industry partners and entrepreneurs. The plan also involves an outdoor
public space that connects to the Charles River to open the campus to the view of Boston.
80
The plan as of February 2015 [Schmidt & Ruiz, 2014] is shown in Figure 36. We will
note here that an updated massing designs based on this scheme was released on April 14,
2015 [C. Barnhart, M. A Schmidt, and I. Ruiz, “Campus Planning / Kendall Update and
Community Meetings”, email, 14 April, 2015] and is included in Appendix D2 for reference.
The previous design proposal from East Campus Study Scheme B [Mack Scogin Merrill
Elam Architects and Michael van Valkenburgh Associates Inc., 2014a] is shown in Figure 37
to indicate the expected 3D model, because the height information is not yet available (as of
May 4, 2015).
The gateway is framed by spaces for the MIT Museum, an Innovation Space, and a
graduate student residence. The new residential building has the capacity to replace all of
the housing in Eastgate (demolished in this new development E55, 201 units), and provide
approximately 270 additional units of graduate housing. [C. Barnhart, M. A Schmidt, and I.
Ruiz, “Campus Planning / Kendall Update and Community Meetings”, email, 14 April, 2015].
The new residential building will be built before repurposing the Eastgate site for a new lab
as shown in the two-step phasing plan in Figure 38. The new space will be filled with ample
greenery and public space along with local attractions and office spaces to vitalize the site
(Figure 39 and Figure 40).
Figure 35 Existing East Campus conditions (Image credit: Mack Scogin Merrill Elam Architects and Michael
van Valkenburgh Associates Inc., 2014b)
Figure 34 Design concept for the new East Campus, continued to the next page (Image credit: MIT Design
Committee, 2013)
81
[page intentionally left blank]
82
83
[page intentionally left blank]
84
Figure 36
Plan for the East Campus Gateway project as of February 2015. Image credit: Schmidt and Ruiz (2014)
85
Figure 37 Approximate three dimensional model and height information from previous proposal. Image
credit: Scheme B by Mack Scogin Merrill Elam Architects and Michael van Valkenburgh Associates Inc. (2014a)
86
Figure 38 Phasing plan from previous proposal. Image credit: Scheme B by Mack Scogin Merrill Elam
Architects and Michael van Valkenburgh Associates Inc. (2014a)
87
Figure 39 Kendall gateway (Carleton Street): Existing conditions (above) and proposed design rendering
(bottom). Image credit: Scheme B by Mack Scogin Merrill Elam Architects and Michael van Valkenburgh
Associates Inc. (2014b)
88
Figure 40
Heyward Street existing conditions (above) and proposed design rendering for a public plaza
(bottom). Image credit: Scheme B by Mack Scogin Merrill Elam Architects and Michael van Valkenburgh
Associates Inc. (2014b)
89
Sustainability concepts are developed by environmental design consultants Atelier Ten
(Figure 41). They suggest the East Campus extends MIT’s existing steam and chilled water
loops to benefit from the highly efficient central plant infrastructure. This allows the
campus to deliver hot and chilled water more efficiently, while eliminating unsightly
cooling towers from building rooftops and greatly reducing greenhouse gas emissions. The
rooftops can then be used as active, occupiable spaces, potentially with green roof
vegetation. Commercial projects should evaluate the potential of on-site cogeneration and
the efficiencies of other shared systems such as waste heat recovery for domestic hot water
generation or heating. Waste heat could also be used to heat sidewalks during winter to
maintain safe, ice-free surfaces for pedestrians and minimize maintenance requirements.
Building system strategies include high-efficiency variable air volume systems, heat
recovery, chilled beams, high-efficiency lighting, advanced controls and high performance
building envelopes. Other strategies include water management and conservation, bicycle
lanes, green space, and recycling. The plans are not yet finalized.
Figure 41 Sustainability concepts include district-level and building system energy efficiencies. Image
credit: East Campus Study Appendix A: Sustainability Report [Mack Scogin Merrill Elam Architects and
Michael van Valkenburgh Associates Inc., 2014a]
90
In a typical urban design project, at this stage the massing design for each building is
further explored and modified by different group of architects. The program layout is
determined to encourage a particular circulation (movement) of occupants through the
building. With this model energy consultants are able to provide the energy use target for
the buildings and recommend building envelope and systems (evaluation and refinement
of strategies suggested by Atelier Ten) to meet that goal using energy simulation.
This case study modifies this urban design process by incorporating the UHI in order to
improve the outdoor thermal comfort. With UWG users are able to predict more accurately
the cooling and heating energy use and test building envelope designs. Before suggesting
alternative designs, we first assess MIT’s design for thermal comfort to serve as the
baseline as well as outline the evaluation process for an existing site using UWG and umi.
The GIS file is obtained for the MIT campus to extract the building massing and street
layouts. The 3D models of the new development is based on Figure 36 and is extruded
based on the previous proposal East Campus Study Scheme B (Figure 37) because building
heights are not yet announced. This scheme most closely resembles the most recent plan in
Figure 36. The resulting 3D model is shown in Figure 42. The existing buildings on site are
colored in light cyan color and the new developments are shown in turquoise.
Figure 42 Three dimensional model of MIT’s plan based on East Campus/ Kendall Square Update [Schmidt
& Ruiz, 2014] and East Campus Study Scheme B [Mack Scogin Merrill Elam Architects and Michael van
Valkenburgh Associates Inc., 2014a]. MIT’s existing buildings are shown in light blue and the planned new
developments are shown in turquoise.
91
The urban geometric characteristics are extracted from the 3D model using
Grasshopper as shown in Figure 43. The building height, site area, and building footprints
are extracted from the model to calculate the required geometric parameters for UWG.
The building construction and intended building schedules are not yet determined by
the design team at this stage of the urban design process, so building templates (Table 9)
are constructed based on metered data from the existing buildings on campus based on
campus energy use from 2008 – 2012 obtained from the Department of Facilities team at
MIT. The annual energy consumption data for 2012 is shown in Table 9 and Table 10. The
new construction templates are modeled after relatively new buildings on campus
(completed after 2010) to be consistent with the construction materials and building use
expected for this new site. As UWG currently only takes one construction template for the
neighborhood, the contributions from each building types are weighed based on built
volume (Figure 44) and consequently the built floor area because we assume a constant
floor height of 3m.
Figure 43 Grasshopper definition used to extract the height and area of each building. In the above image,
the top black number represents the site coverage area and the bottom cyan shows the building height. The
gray area is the site area, defined by the PUD-5 boundary. This Grasshopper definition is available at
http://urbanmicroclimate.scripts.mit.edu/uwg_parameters.php
92
Table 9 Building templates for the new buildings on East Campus are based on existing buildings on campus.
Office, retail, and classroom are based on the Sloan School of Management (Building E62, LEED Gold,
completed in 2010). Lab building is a hybrid of the cancer research center Koch Institute (Building 76, LEED
Gold, completed in 2010) and the Sloan Building. The building schedule is based on the Koch Institute.
Residential template is modeled after a steel framed building and uses Massy Dormitory (W1) for building
use estimates
Building Construction Materials
% Total
Wall Layer 1 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Albedo
Emissivity
Wall Layer 2 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Wall Layer 3 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Wall Layer 4 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Roof Layer 1 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Albedo
Emissivity
Roof Layer 2 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Roof Layer 3 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Roof Layer 4 Material
Thickness [m]
K [W/m-K]
VHC [J/m^3-K]
Office, Retail,
Classroom
Lab
Residential
71% ¹
13% ¹
15% ¹
Cement plaster
Cement plaster
Cement plaster
0.08
0.10
0.03
0.72
0.72
0.72
1.56E+06
1.56E+06
1.56E+06
0.1
0.1
0.1
0.7
0.7
0.7
Spray- glass fiber Spray- glass fiber
Steel
insulation
insulation
0.08
0.08
0.15
0.04
0.04
45.30
1.02E+05
1.02E+05
2.50E+06
Concrete
Concrete Spray- glass fiber
insulation
0.15
0.15
0.08
1.13
1.13
0.04
1.65E+06
1.65E+06
1.02E+05
Gypsum
Gypsum
Expanded
polystyrene
insulation
0.02
0.02
0.01
0.58
0.58
0.03
8.72E+05
8.72E+05
3.51E+04
Built-up roofing
0.01
0.16
1.64E+06
0.05
0.98
Sheathing
0.01
0.06
3.21E+06
cellular
polyurethane
insulation
0.20
0.02
3.82E+04
Gypsum board
0.01
0.58
8.72E+05
Built-up roofing
0.01
0.16
1.64E+06
0.05
0.98
Sheathing
0.02
0.06
3.21E+06
Expanded
polystyrene
insulation
0.01
0.16
5.20E+05
Clay slate
0.20
1.10
1.61E+06
Built-up roofing
0.01
0.16
1.64E+06
0.05
0.98
Sheathing
0.01
0.06
3.21E+06
Expanded
polystyrene
insulation
0.13
0.16
5.20E+05
Steel
0.001
45.30
2.50E+06
Weighted
Average
100%
Layer 1
0.07
0.72
1.56E+06
0.1
0.7
Layer 2
0.09
7.01
4.71E+05
Layer 3
0.14
0.96
1.41E+06
Layer 4
0.02
0.49
7.41E+05
Built-up roofing
0.01
0.16
1.63E+06
0.05
0.98
Sheathing
0.01
0.06
3.21E+06
Insulation
0.17
0.06
1.77E+05
0.04
7.54
1.22E+06
93
Note:
1
Weighted by volume according to the Rhino massing model
Figure 44 This Grasshopper definition calculates the volume for each program types. The yellow panels on
the right show the percentage of the total built area used for weighting of building templates for UWG inputs
Table 10
Template buildings’ annual energy use in 2012, obtained from the MIT Department of Facilities
Program
Labs
Office, retail, and
classroom
Residential
Note:
1. No air conditioning
94
Template building
Koch Institute (76)
Sloan School of Management
(E62)
Massy Dormitory (W1)
Cooling Energy
[kWh/m2]
462.8
Heating
Energy
[kWh/m2]
497.6
52.0
-- ¹
58.6
105.6
The framework for using UWG and umi is as explained in Chapter 3. The interface is
shown in Figure 25 in Chapter 5. UWG simulation is performed through the umi plugin
using the xml file created in step 2 as well as the weather file for Boston (USA_MA_BostonLogan.Intl.AP.725090_TMY3.epw) obtained from the Department of Energy website [U.S.
Department of Energy, Weather Data, 2013].
Using the newly morphed weather file from UWG, EnergyPlus simulation is performed
using umi. The same building templates used for UWG is used in umi as well for
consistency. We assume the coefficient of performance, the ratio of heating or cooling
provided to electrical energy consumed, of 1 for both heating and cooling as we are not
modeling the heating and cooling equipment but rather are interested in the energy
demand on site. Our metric is the sum of ideal air loads for heating and cooling, calculated
separately for each of core and perimeter zones for the four facades, using the shoebox
methodology [Dogan & Reinhart, 2013] explained in Section 3.2. Natural ventilation is not
modeled because the current algorithm for UWG does not account for it.
The simulation results for the MIT design are shown in Figure 45 and Figure 46.
Figure 45 shows the dry bulb air temperatures compared against the current campus
for July and December to represent the summer and winter outdoor conditions,
respectively. We can observe a slight delay in the diurnal cycle of the urban heating as
observed in Figure 3 [Oke, 1987]. The difference between the cyan (MIT design) and gray
(current campus) represents the additional UHI from the new development. The minimum
and maximum UHI intensity over the course of the day is 0.07K and 0.32K for July and
0.34K and 0.63K for December. A larger impact on the thermal comfort is seen for the
winter than for the summer because Cambridge is heating dominated and the addition of
new buildings increase the total heat output from the buildings.
Figure 46 shows the annual thermal comfort for the MIT design based on the expanded
UTCI metric (18 bins) for the sidewalk and the center of urban canyon, as explained in
Section 5.3. The two UTCI histograms are nearly identical and shows that the East Campus
is thermally comfortable for 50% of the year.
This MIT case will serve as the “base” case to evaluate the alternative scenarios
explored in the next section.
95
28
25
22
19
1
5
9
13
17
21
1
5
9
13
17
21
6
3
0
-3
Figure 45 Dry bulb temperature for current campus (gray) and MIT plan (turquoise). The difference
between the two lines represents the change in UHI for the site
25%25%
25%25%
14% 14%
11%
5%
1%
14%13%
11%
6%
3%
1%
1 2 3 4 5 6 7 8 9 101112131415161718
1%
3%
1%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Figure 46 UTCI comfort index for the urban canyon on the sidewalk (left) and middle of the canyon (right).
The turquoise bars represent the hour count of no thermal stress, which is between +9 < T < +26˚C.
96
We propose an alternative to the MIT’s plan for development in the East Campus by
incorporating outdoor thermal comfort as one of the drivers for the urban design process.
The design concept will retain the goal to increase the campus’ connectivity to the
surrounding neighborhoods. Similar to MIT’s case, the East Campus extends the Infinite
Corridor to enhance the sense of community and unify the identity of the campus from
west to east. In the north to south direction, we create an inviting gateway to the MIT
campus from Kendall Square and subway station towards the waterfront through the open
public space on the site.
In our design we will treat the East Campus gateway as a public space that is
surrounded by both academic and commercial buildings to provide space for students to
interact with industry partners. The public space can be used for outdoor social and
academic events. We put an emphasis on the environmental performance, in particular for
the new development to have minimal impact on the energy consumption and be as
thermally comfortable as possible to enhance the outdoor experience.
The alternative scenarios introduced in this section will be framed by the same
programmatic requirements as MIT’s proposal and require the same amount of new floor
area for the commercial, residential, and academic spaces per the City of Cambridge
ordinance Number 1355 (2013) shown in Table 8. The study is focused on testing the
parameters that affect the UHI (i.e. building height, built density and roof materials) based
on the sensitivity analysis performed in Chapter 4.
The first step in bioclimatic design is to understand the overall climate of the site. The
Kendall Square is located in the Middlesex County, Massachusetts, and is categorized as the
Dfd climate zone [Kottek et al., 2006] in the Köppen-Geiger climate classification. It has a
snow climate (D) that is fully humid (f) with warm summer (d).
Climate Consultant [version 6.0; 2015] is used for visualizing the weather data for the
Logan Airport weather file in order to perform a preliminary analysis on Boston’s climate
and understand the basic design strategies that could be effective for Cambridge, MA. Of
particular interest are the average monthly diurnal temperatures (Figure 47),
psychrometric chart (Figure 48) measured against comfortable zones [ANSI/ ASHRAE
Standard 55, 2013; ASHRAE Handbook of Fundamentals, 2013], and wind rose (Figure 49
and Figure 50).
Boston’s weather is below the comfort zone for most days based on dry bulb
temperature (Figure 47) and psychrometric chart that also incorporates relative humidity
(Figure 48). Figure 47 also shows that temperature change can be around 20K over the
course of a day. This indicates that we need to focus on the heating strategies and good
thermal mass.
97
Figure 47 Climate consultant interface: monthly diurnal averages and radiation for Boston Logan. As seen,
Boston is a heating dominated city. The comfort zone for the summer and winter are shown in gray
Figure 48 Climate consultant interface: Psychrometric chart with comfortable hours based on ASHRAE
Handbook of Fundamentals (2013) and ANSI/ ASHRAE Standard 55 (2013). Comfortable hours (blue) for
summer (right) and winter (left)
98
The annual wind rose (Figure 49) shows that the dominant wind direction is northwest
and the average wind speed is below 6 m/s for all directions. Figure 50 separately shows
wind conditions for the summer (July, hottest month) and winter (December, one of the
coldest month). The dominant wind direction is the northwest in the summer with mild
average wind speed of around 4m/s. Based on this we can design the operable glazing to
face this direction to take advantage of natural ventilation (theoretically, though not
incorporated in this case study as stated above). In the winter the site gets westerly wind.
The average wind velocity is relatively mild at around 6 m/s. However, we get moderate
breeze from northeast at around 10m/s for a number of hours. We can protect the outdoor
public space by placing buildings upwind of that direction.
Figure 49 Climate consultant: Annual wind rose showing the frequency of wind speed and temperatures
for all directions
99
Figure 50
100
Wind rose for July (top, summer) and December (bottom, winter)
Based on the design concept the massing design is developed. The basic design concept
is similar to that of MIT’s but the design exploration will incorporate urban heating. As
mentioned, the new designs comply with the programmatic requirements and zoning
height restrictions.
There are two main schemes (MIT and connection schemes) with parametric variations
of the building heights and thermal masses. Note that these cases are developed during the
simulation process and each is guided by simulation results. The strategies are focused
around average building height, site coverage ratio, and average façade-to-site ratio that
are known to affect the UHI based on the sensitivity analysis in Chapter 4. The effect of
green roofs is also explored in the final alternative.
The six alternatives explored here are:
1. MIT scheme with high rise: The buildings are constructed as tall as possible and
this option removes the new lab building R to increase the green space.
2. MIT scheme with low rise: The building heights are minimized and all of available
site area is used for buildings. As a consequence the open lawn space is minimized.
3. MIT scheme with better building envelope: This alternative uses the same
geometry design as MIT’s plan. The wall insulation thickness is doubled from
0.09m to 0.2m.
4. Connection scheme (new scheme): The buildings envelope the open space better
and there are no additional buildup on top of E38 and E39 as planned in MIT’s plan
(this new volume is shifted to other new buildings). Shorter buildings at the
gateway creates a more welcoming arrival experience to the MIT campus and
create a sense of openness and connectivity to Kendall Square. This opening is also
oriented towards the summer breeze direction for natural ventilation (not
modeled).
5. Connection scheme with low rise: The massing is further modulated to decrease
average building height and thus shading on buildings.
6. Connection scheme with better building envelope: The wall insulation level is
doubled from Alternative 5. A half of the new buildings have green roofs and the
vegetation on the sidewalks is increased by 25%.
The morphology characteristics of each alternative is summarized below in Table 11.
The models are shown with simulation results in Figure 51.
Table 11
design
Summary of the building characteristics for each alternative as well as current and planned MIT
Current MIT campus
MIT’s plan
Alternative 1: MIT scheme with high rise
Alternative 2: MIT scheme with low rise
Avg
building
height [m]
30.52
34.26
41.16
29.41
Avg site Avg façadecoverage to-site ratio
ratio
0.36
1.19
0.48
0.40
0.53
2.52
2.33
2.43
101
Alternative 3: MIT scheme with better
insulation
34.26
0.48
2.52
Alternative 4: Connection scheme
Alternative 5: Connection scheme with low
rise
Alternative 6: Connection scheme with
increased insulation and vegetation
35.65
34.61
0.43
0.47
2.31
2.31
34.61
0.47
2.31
As mentioned, the proposed urban design process is an iterative process where the
results of the simulation results drive the next rounds of designs. This section first
introduces the results and the thought process behind each step of the simulation. Finally,
we select the recommended scheme based on the thermal comfort and energy simulation
results.
The simulation steps are identical to that of the MIT design in Section 6.2.2. The results
are shown in Figure 51. The metrics are as described in Section 5.3: average monthly dry
bulb temperature and UTCI for measuring thermal comfort. Similar to the MIT design
simulation, dry bulb temperatures are shown for July and December to demonstrate the
conditions for the summer and winter months, respectively. Results for other months can
be seen from the UWG user interface. Additionally, the UTCI values are only shown for the
sidewalk case as they are nearly identical for the sidewalk and the canyon center as shown
in Figure 46. The UTCI histogram for both cases can also be seen in the UI.
102
Current MIT campus
MIT Eastgate plan
Avg h: 30.52m
Site coverage: 0.36
Facade-to-site ratio: 1.19
Summer [July]
Avg h: 34.26m
Site coverage: 0.48
Facade-to-site ratio: 2.52
28
28
25
25
22
22
19
19
1.00
1
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
-
1
5
9
13
17
21
-1.00
Winter [December]
0
1
5
9
13
17
21
1
5
9
13
17
21
-1
6
6
3
3
0
0
-3
-3
1.00
1
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
-
1
5
9
13
17
21
-1.00
Annual UTCI
< -40˚C
-40˚C to -27˚C
-27˚C to -13˚C
-13˚C to 0˚C
0˚C to 9˚C
9˚C to 26˚C
26˚C to 32˚C
32˚C to 38˚C
38˚C to 46˚C
>46˚C
extremely cold stress
very strong cold stress
strong cold stress
moderate cold stress
slight cold stress
no thermal stress
moderate heat stress
strong heat stress
very strong heat stress
extreme heat stress
-1
25% 24% 25% 12% 6% 2% 0% 0
25% 25% 25% 14% 13% 11% 3% 5% 1% >-40
0 8 +9
1 2 -27
3 4 -13
5 6 7 9 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 1% 0% 14% 14% 3% 1% >-40
0 8 +9
1 2 -27
3 4 -13
5 6 7 9 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 [page intentionally left black]
Alternative 2
Alternative 1
Summer [July]
Alternative 3
Minimize average building height and minimal open space
Avg h: 29.41m
Site coverage: 0.53
Facade-to-site ratio: 2.43
Max height for new buildings, removed lab building R
Avg h: 41.46m
Site coverage: 0.40
Facade-to-site ratio: 2.33
Same geometry as the MIT plan. Insulation thickness is oubled
Avg h: 34.26m
Site coverage: 0.48
Facade-to-site ratio: 2.52
28
28
28
25
25
25
22
22
22
19
19
19
1
1
1
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
0
Winter [December]
1
5
9
13
17
21
0
1
5
9
13
17
21
0
-1
-1
-1
6
6
6
3
3
3
0
0
0
-3
-3
-3
1
1
1
1
5
9
13
17
21
1
5
9
13
17
21
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
0
1
5
9
13
17
21
-1
Annual UTCI
< -40˚C
-40˚C to -27˚C
-27˚C to -13˚C
-13˚C to 0˚C
0˚C to 9˚C
9˚C to 26˚C
26˚C to 32˚C
32˚C to 38˚C
38˚C to 46˚C
>46˚C
extremely cold stress
very strong cold stress
strong cold stress
moderate cold stress
slight cold stress
no thermal stress
moderate heat stress
strong heat stress
very strong heat stress
extreme heat stress
1
5
9
13
17
21
-1
25% 25% 25% 11% 6% 1% 0% 0
-1
25% 25% 25% 14% 13% 11% 3% 5% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 1% 0% 0
25% 25% 25% 14% 14% 11% 3% 6% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 1% 0% 14% 13% 3% 1% >-40
0 8 +9
1 2 27
3 4 -13
5 6 7 9 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 [page intentionally left black]
Alternative 5
Alternative 4
Summer [July]
Alternative 6
Building forms improved to reduce average height
Avg h: 34.61m
Site coverage: 0.47
Facade-to-site ratio: 2.31
Massing design altered to enclose the public spaces better
Avg h: 35.65m
Site coverage: 0.43
Facade-to-site ratio: 2.31
Insulation thickness doubled, increased green space and 50% green roofs
Avg h: 30.52m
Site coverage: 0.36
Facade-to-site ratio: 1.19
28
28
28
25
25
25
22
22
22
19
19
19
1
1
1
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
0
Winter [December]
1
5
9
13
17
21
0
1
5
9
13
17
21
0
-1
-1
-1
6
6
6
3
3
3
0
0
0
-3
-3
-3
1
1
1
1
5
9
13
17
21
1
5
9
13
17
21
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
0
1
5
9
13
17
21
-1
Annual UTCI
< -40˚C
-40˚C to -27˚C
-27˚C to -13˚C
-13˚C to 0˚C
0˚C to 9˚C
9˚C to 26˚C
26˚C to 32˚C
32˚C to 38˚C
38˚C to 46˚C
>46˚C
extremely cold stress
very strong cold stress
strong cold stress
moderate cold stress
slight cold stress
no thermal stress
moderate heat stress
strong heat stress
very strong heat stress
extreme heat stress
1
5
9
13
17
21
-1
25% 25% 25% 11% 6% 1% 0% 0
-1
25% 25% 25% 14% 13% 11% 3% 5% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 1% 0% 0
25% 25% 25% 14% 13% 12% 3% 6% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 1% 0% 14% 13% 3% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 [page intentionally left black]
Year 2020
Current MIT campus
Summer [July]
ΔTair
new Tair
current MIT campus
ΔHeat island intensity
[vs current MIT campus]
31
31
28
28
28
25
25
25
22
22
22
19
19
19
1.00
4
4
2
2
1
5
9
13
17
21
0
-1.00
ΔTair
new Tair
current MIT campus
ΔHeat island intensity + climate
change [vs current MIT campus]
5
9
13
17
21
0
9
9
6
6
6
3
3
3
0
0
0
-3
-3
-3
1.00
4
4
2
3
1
5
9
13
17
21
0
-1.00
Annual UTCI
extremely cold stress
very strong cold stress
strong cold stress
moderate cold stress
slight cold stress
no thermal stress
moderate heat stress
strong heat stress
very strong heat stress
extreme heat stress
1
9
-
< -40˚C
-40˚C to -27˚C
-27˚C to -13˚C
-13˚C to 0˚C
0˚C to 9˚C
9˚C to 26˚C
26˚C to 32˚C
32˚C to 38˚C
38˚C to 46˚C
>46˚C
East Campus development is complete
Avg h: 34.61m
Site coverage: 0.47
Facade-to-site ratio: 2.31
31
-
Winter [December]
Year 2050
One of each of residential (L), lab (N) and commercial (S) built
Avg h: 32.48m
Site coverage: 0.36
Facade-to-site ratio: 1.78
Avg h: 34.26m
Site coverage: 0.48
Facade-to-site ratio: 2.52
25% 24% 25% 12% 6% 2% 0% 1
5
9
13
14% 13% 12% 6% 1% >-40
0 8 +9
1 2 273 4 -13
5 6 7 9 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 1% 0% 21
24% 25% 25% 3% 17
15% 2
1
5
9
13
17
21
1
5
9
13
17
21
23% 25% 25% 13% 9% 4% 4% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 0% 12% 14% 7% 4% 1% >-40
1 2 273 4 -13
5 6 07 8 +99 10 +26
11 12 +32
13 14 +38
15 16 +46
17 [˚C]
18 [page intentionally left black]
Figure 51 [Preceding pages] UWG simulation results for alternatives 1 – 6 compared against the current
campus and MIT’s design. The development stages of the selected case 6 are shown on the final page and are
discussed in detail in Section 6.3.5
The diurnal UHI intensities are compared against the current campus in Figure 52. The
minimum and maximum UHI intensities are -0.1K and 0.4K for the summer and 0.2K and
0.7K for the winter. The negative values indicate that there is urban cooling between the
hours of 9am – 1pm in July for all cases. This is as expected because much of the parking lot
(concrete) is replaced by vegetation. Alternative 2 (MIT scheme with low rise) has the most
cooling effect likely due to the fact that urban canyon height is short and thus heat can
easily escape from the urban canyon. Based on this simulation run, the future schemes will
explore shorter urban canyon heights. This is exactly done for Alternative 5 and 6
(derivatives of the Connection scheme) as a derivation of Alternative 4.
Alternative 6 (Connection scheme with increased vegetation and insulation) achieves
urban cooling via shorter canyon heights as well as through cool roof and increased
vegetation and shading on the streets. We also observe the urban cooling is greater for
cases with higher levels of insulation (Alternatives 3 and 6) because the building
construction are on average improved when new buildings with higher levels of insulation
replace old buildings.
Furthermore, Case 6 (turquoise) has the least amount of urban heating in the summer
from 3pm – 9pm. It has the third smallest increase in the winter. As the initial climate
analysis in Figure 47 indicates, Boston has highest dry bulb temperatures in July. It hovers
around the comfort band, so the effect of urban heating/cooling is more relevant in the
summer months than in the winter. Thus, we select Alternative 6 as our best design for
improving the thermal comfort.
In addition, the energy demand values from umi are shown in Figure 53. They are the
normalized energy demand for heating and cooling loads for the new buildings,
representing a mix of 51.6% lab, 29.8% commercial, and 18.6% residential buildings. The
setup is as explained in Section 6.2.2. The energy consumption estimations for each
program are in-line with those used as templates (Table 10). Observing the trend for the
MIT schemes (MIT design and alternatives 1 - 3) reveals the direction for successful energy
performance design for the East Campus. The comparisons of the MIT design with each of
the variations reveal the following effects of changing the urban design parameters:
1. Alternative 1 (MIT scheme with high rise): This case has lower cooling energy
consumption than the original MIT case. This could be due to the increase in open
lawn space.
2. Alternative 2 (MIT scheme with low rise): Alternative 2 is the extreme case for
minimizing the average building height. It has the lowest energy consumption for
heating possibly because buildings do not shade each other.
3. Alternative 3 (MIT design with increased insulation): As expected, improving
the envelope insulation improves the energy performance. The same strategy will
be tested for the Connection scheme designs.
111
Figure 52 Comparison of UHI intensity for all 6 cases and the MIT design case (black) against the current
campus, The selected case 6 is highlighted in cyan.
112
Figure 53
Heating and cooling energy simulation results from umi for each scenarios
These three observations show that shorter buildings, open space (to mitigate shading),
and thermal mass are effective strategies for improving the energy performance.
Alternatives 4 – 6 will test the same strategies.
We should note here that building energy consumption is very specific to the building
use schedule and building envelope design that is not explored rigorously for the purpose
of this case study, which focuses on the geometric design and the amount of green space.
Building energy consumption in turn affects the UHI, so UWG should be also used during
the architectural design process to get an accurate estimation of the UHI.
The insulation levels are tested here for the demonstrative purpose and to highlight the
necessity for such consideration later in the design process. In addition to the insulation
levels, lighting controls and operable windows and controls for natural ventilation should
be considered in detail in the architectural design phase.
Based on the simulation results above, we recommend Alternative 6 for improved
thermal comfort particularly because urban heating is minimized during prolonged period
on summer afternoons when cooling is most desired.
The suggested plans are shown in Figure 54 with a two-step phasing plan. Rendered
images are shown in Figure 55. For the first phase, the graduate housing development is
113
prioritized. The demolition of existing buildings and the conversion of the parking lot space
to green space happen during this stage too.
Figure 54 Proposed urban development for the East Campus (Alternative 6). The top image shows the
development after the first phase and the bottom shows that after the second phase
114
115
Figure 55 Rendered images of the proposed alternative for MIT’s East Campus development. The northeast
and northwest views are shown for each phase.
116
Urban heating is the local heating effect of the urbanization. Here we discuss UHI in
combination with a global heating effect – climate change – to holistically capture urban
thermal comfort and energy consumption over time. Specifically we will evaluate the
thermal comfort of the suggested Alternative 6 shown in Figure 54. This allows us to
evaluate a thermally comfortable campus throughout the development.
This section also represents how UWG can be used in conjunction with other tools
towards a more holistic urban design process.
The Intergovernmental Panel on Climate Change (IPCC) was jointly established by the
World Meteorological Organization (WMO) and the United Nations Environment Program
(UNEP) to assess the scientific, technical and socio-economic information relevant for
understanding the risk of human-induced climate change. Since its inception the IPCC has
produced a series of comprehensive Assessment Reports on the state of understanding of
causes of climate change, its potential impacts and options for response strategies. The new
scenarios provide input for evaluating climatic and environmental consequences of future
greenhouse gas emissions and for assessing alternative mitigation and adaptation
strategies. They include improved emission baselines and latest information on economic
restructuring throughout the world, examine different rates and trends in technological
change and expand the range of different economic-development pathways, including
narrowing of the income gap between developed and developing countries. The six IPCC
scenarios’ prediction on greenhouse gas emissions and their impacts on urban heating are
shown in Figure 56.
The climate change world weather file generator [CCWorldWeatherGen, 2008] uses
IPCC model summary data of the UK Met Office Hadley Center experiment ensemble
(HadCM3 A2) in order to morph an existing epw file for a medium to high emissions
scenario (A2) [Jentsch et al., 2013]. The A2 model is based on Atmospheric Stabilization
Framework Model (ASF) [Nakicenovic & Swart, 2000]. The tool uses Microsoft Excel Macro
and the output is an epw file that captures the A2 climate change scenarios for years 2020,
2050, and 2080.
The A2-ASF quantification is based on the following assumptions:
1. Relatively slow demographic transition and relatively slow convergence in regional
fertility patterns
2. Relatively slow convergence in inter-regional GDP per capita differences
3. Relatively slow end-use and supply-side energy efficiency improvements (compared
to other storylines)
4. Delayed development of renewable energy
5. No barriers to the use of nuclear energy
117
Notes:
a) Temperatures are assessed best estimates and likely uncertainty ranges from a hierarchy of models of varying complexity as well
as observational constraints.
b) Year 2000 constant composition is derived from Atmosphere-Ocean General Circulation Models (AOGCMs) only.
c) All scenarios above are six SRES marker scenarios. Approximate CO2-eq concentrations corresponding to the computed radiative
forcing due to anthropogenic GHGs and aerosols in 2100 (see p. 823 of the Working Group I TAR) for the SRES B1, AIT, B2, A1B, A2
and A1FI illustrative marker scenarios are about 600, 700, 800, 850, 1250 and 1550ppm, respectively.
d) Temperature changes are expressed as the difference from the period 1980-1999. To express the change relative to the period
1850-1899 add 0.5°C.
Figure 56 Six IPCC scenarios and their impacts on greenhouse gas emission and temperature change
predictions. (Credit: Nakicenovic & Swart (2000) from
https://www.ipcc.ch/publications_and_data/ar4/syr/en/spms3.html)
118
The change in monthly average dry bulb temperature on the MIT East Campus using the
CCWorldWeatherGen tool is shown in Figure 57. The annual average temperature
increases compared to the current MIT campus in 2015 are 0.9K by 2020 and 2.2K by 2050.
The temperature increase are most prominent for the summer and winter months.
35
Annual Average
Temperature
Temperature [C]
25
13.2
11.9
11.0
15
5
-5
1
2
3
4
5
2015
Figure 57
6
2020
7
8
9
10
11
12
2015 2020 2050
2050
Monthly average temperature for East Campus using IPCC-A2 scenario
As mentioned, this section tests the thermal comfort at each phase of the urban
development on the East Campus shown in Figure 54. We assumed Phase 1 is in 2020 and
phase 2 in 2050. The combined effects of the climate change and UHI are shown on the last
page in Figure 51. Compared to the East Campus today, 1˚C increase in urban heating is
predicted in 2020 and 2 – 3˚C increase by 2050. The predicted maximum and minimum
monthly average temperatures on the East Campus site in 2050 are 26.1˚C and 0.5˚C. The
average annual temperature increased from 11.3˚C to 13.5˚C from 2015 to 2050 when UHI
and climate change are considered.
Figure 58 summarizes the urban temperature changes for the average monthly dry bulb
temperatures from 2015 base (Boston Logan Airport reference site) to that on the East
Campus in 2050. It breaks out the contributions from UHI and climate change for Phase 1
and Phase 2 for UHI and climate change. The average contribution of UHI is about a tenth of
that from the climate change. We observe urban cooling in 2020 is possibly from the
increase in open space (i.e. the urban canyon is wider and less heat is trapped) as some
existing buildings are removed. In other words, the climate change is mitigated via a local
change in the site morphology.
In Figure 51, we can also observe an upward shift in the UTCI histogram: the hours
above “no thermal stress” increase from 5% to 12% in 2050 compared to the current
campus in 2015. The hourly count of the thermally comfortable hours (i.e. no heat stress)
decreased by 2%.
119
Figure 58 Changes in urban dry bulb temperatures from 2015 (Boston Logan Airport) through phase 1 in 2020 and phase 2 in 2050. The average
contribution from each phenomenon is shown on the right. Urban heat and cool island intensities are the changes in between the years 2015, 2020, and
2050.
120
This chapter introduced and demonstrated the new urban design workflow through a
case study of the MIT East Campus development.in the context of campus expansion to
promote research and meet student housing demand. This redevelopment is particularly
focused on improving the public space and campus life by creating connections to the
surrounding innovation center and Charles River.
By incorporating the UHI consideration early in the design process, we showed how we
can develop a massing design that can improve the comfort and energy performance of the
outdoor urban site compared to the traditional urban design process. Increase in the urban
heating is inevitable as we increase the building density and that will have a negative
impact on the public health and air quality. However, it can help cool the site in the summer
if designed properly. Furthermore, in Section 6.3.5, we showed how an urban designer can
evaluate the thermal comfort of the urban site at each phase of the urban development.
The recommended case evolved from iteratively testing the massing design using UWG.
The UWG’s interface is designed to promote such a design process. By running parametric
designs in a batch, the user can decide to explore further the successful schemes based on
the previous iterations. For example, in this case study as the flatter geometry is more
successful for mitigating urban heating for the MIT scheme, the same method is tested for
the Connection scheme.
Compared to the rate of the global warming, the gain from urban cooling via local
change (UHI) through massing design is very small and is within 1K. We also observe only a
small change in the UTCI thermal comfort metric on the annual basis between MIT’s
proposed design and our proposed Alternative 6. We’ve also seen that there is a trade-off
between selecting the best case for thermal comfort and for energy.
In the next step of urban design process (detailed architectural designs) other local
changes such as thermal mass, building envelope, shading device, lighting control, green
roofs, and natural ventilation should be considered in detail to reduce the energy
consumption along with other sustainability strategies proposed by Atelier Ten. If internal
energy consumption is reduced the outdoor thermal comfort can also be improved as a
result of reduction in heat output into the urban canyon.
This case study demonstrated the methodology to improve thermal comfort and energy
performance of an urban development. Other aspects of environmental performance such
as daylighting, mobility, and embodied energy should be considered for a complete
evaluation of the performative urban design.
121
This thesis introduces the new workflow for urban design process with thermal
comfort and energy considerations as well as serves as the documentation of the tool
development process. UWG is the first design tool to model UHI. The goal of the tool is to
enable urban designers to design with thermal comfort and energy considerations in mind
as well as allow urban planners to articulate zoning and policies with these considerations.
The stand-alone and Rhino-integrated versions are created for different types of users,
namely energy consultants and urban designers, respectively, to promote early integration
of urban heating considerations in the urban design process.
This new software is developed using Urban Weather Generator developed by Bueno as
the underlying UHI modeling engine. A sensitivity analysis is conducted for Boston, MA and
Singapore, Singapore to reduce the number of required input parameters. The
commonality of results allowed us to decrease the number of inputs by over 46% and thus
increase the speed at which the users can evaluate their designs for thermal comfort and
energy usage. The key parameters are site coverage ratio, façade-to-site ratio, and sensible
anthropogenic heat, which are designed during the master planning phase of the urban
design process.
The user interface is designed with careful considerations for usability based on the
learnability, efficiency, and safety (prevention of error and recoverability) of the tool as
described in Chapter 5. It has been tested with urban designers, energy consultants, and
urban planners, who are the target users of the tool. The results of the sensitivity analysis
allowed us to simplify the user interface.
The new workflow using UWG is demonstrated through a case study of MIT’s East
campus development. Alternative designs are developed for MIT’s plan to improve the
thermal comfort and energy efficiency on campus. With the new East Campus we aim to
improve the quality of academic and social experiences for students by creating outdoor
public spaces that are thermally comfortable as well as to provide spaces for interaction
with industry partners. The goal is a connected space with the rest of campus and with its
neighbors (Kendall Square and Charles River) that bring people together. We showed how
UWG can help develop massing design that can mitigate the UHI. Finally, the UHI is
evaluated with climate change to evaluate the overall thermal comfort of the site at every
phase of development.
122
The current version of the GUI and the underlying algorithm by Bueno present some
limitations that require further development.
The UWG’s algorithm only morphs the dry bulb temperatures and relative humidity. In
the calculation of UTCI in Chapter 5, the urban wind velocity is an estimation of that of the
undisturbed wind approaching a building in the urban site and does not capture the
turbulence inside the street canyon. The impact of vegetation on relative humidity, natural
ventilation, and window shading system are not yet included in the UWG algorithm.
Furthermore, the energy simulation uses “shoebox” representation of zoning [Dogan &
Reinhart, 2013] in the umi version of the tool. The results provide sensible results based on
the author’s experience, and it is in the process of being validated for modeling
neighborhood-scale simulations.
User interface design is an iterative process that can be improved with additions of new
features and further user testing. So far two versions have been released and tested with
potential users. The following can be considered to further improve the usability of the
tool:
1. Library integration with umi: Currently the template editors for umi and UWG use
different data structures. Sharing of the building template libraries would increase
the tool efficiency and promote faster simulation setup.
2. UI improvements for parametric design: The umi version removed the need to
calculate urban site parameters. UWG can be developed for Grasshopper in the
future iteration to work directly with Galapagos and other parametric design tools.
The UWG serves as an integral part of the family of urban design simulation tools that
are developed to encourage early integration of engineering concepts in the design process.
With UWG, urban designers can understand the impact of their designs on the livability of
the space and energy consultants can provide a more accurate predictions of the urban
energy consumption to properly design building systems Together with other
environmental performance simulation tools, the author hopes for sustainable future
urban developments that are thermally comfortable and energy efficient.
123
124
ANSI/ ASHRAE Standard 55-2013 Thermal environmental conditions for human
occupancy. 2013, American Society of Heating, Refrigerating, and Air-Conditioning
Engineers: Atlanta, GA.
ArcGIS/ArcMAP [Computer software]. (2014). ESRI. Retrieved from
http://www.esri.com/software/arcgis/arcgis-for-desktop
ASHRAE Handbook of Fundamentals (2013): 9.11, 24.3. Web. 24 Apr. 2015.
Besserud, K., Hussey, T. (2011). Urban design, urban simulation, and the need for
computational tools. IBM Journal of Research and Development, 55(1.2), 2:1–2:17.
DOI:10.1147/JRD.2010.2097091
Bröde, P., Jendritzky, G., Fiala D., Havenith G. (2010). The Universal Thermal Climate Index
UTCI in Operational Use. Adapting to Change: New Thinking on Comfort Cumberland
Lodge, Windsor, UK, 9-11 April 2010.
Bueno, B., Norford, L., Hidalgo, J., Pigeon, G. (2012a). The urban weather generator. Journal
of Building Performance Simulation 6(4):269–81.
Bueno, B. (2012b). Study and prediction of the energy interactions between buildings and
the urban climate. (Doctoral dissertation). Retrieved from MIT DSpace.
Bueno, B., Pigeon, G., Norford, L. K., Zibouche, K., Marchadier, C. (2012c). Development and
evaluation of a building energy model integrated in the TEB scheme. Geoscientific Model
Development, 5 (2), 433–448.
Bueno, B., Roth, M., Norford, L., Li, R. (2014). Computationally efficient prediction of canopy
level urban air temperature at the neighbourhood scale. Urban Climate, 9 (2014) 35–53.
Cambridge Community Development Department (2013). K2C2 Kendall Square Final
Report 2013[PDF file] http://www.cambridgema.gov/CDD/Projects/Planning/K2C2.aspx
Cannistraro, G., Franzitta G., Giaconia C. (1992). Algorithms for the calculation of the view
factors between human body and rectangular surfaces in parallelepiped environments.
Energy and Buildings, 19 (1992), 51–60.
125
CCWorldWeatherGen tool [Computer software]. (2008). Retrieved from:
http://www.energy.soton.ac.uk/ccworldweathergen/
CityEngine [Computer software]. (2013). ESRI. Retrieved from: http://www.esri.com
City of Cambridge Community Development Department (2015). PUD-KS Rezoning Initial
Proposal. 2 April, 2015 [PDF File]
http://www.cambridgema.gov/~/media/Files/CDD/ZoningDevel/Amendments/2015/PU
DKS/zngamend_pudks_cdddocs_20150402.ashx
Climate Consultant [Computer software]. (2015). Retrieved from: http://www.energydesign-tools.aud.ucla.edu/climate-consultant/request-climate-consultant.php
Crawley, D. B., Lawrie, L. K., Winkelmann F. C., Buhl W.F., Huang, Y. J., Pedersen, C.O., Strand
R. K., Liesen, R. J., Fisher, D.E., Witte, M. J., Glazer, J. (2001). EnergyPlus: creating a newgeneration building energy simulation program.
Crawley, D. B. (2008). Estimating the impacts of climate change and urbanization on
building performance. Journal of Building Performance Simulation, 1(2), 91–115.
Davidson, S. (2015) Grasshopper [Computer software]. Retrieved from
http://www.grasshopper3d.com/
DesignBuilder [Computer software]. (2014) Retrieved from:
http://www.designbuilder.co.uk/component/option,com_docman/task,cat_view/gid,11/It
emid,30/
DIVA [Computer software]. (2015). Retrieved from http://diva4rhino.com/
Dogan, T., Reinhart, C. (2013). Automated conversion of architectural massing models into
thermal ‘shoebox’ models. Proceedings of BS2013: 13th Conference of International
Building Performance Simulation Association, Chambéry, France, August 26-28
Dogan, T. (2015). Archsim [Computer software]. Retrieved from http://archsim.com/
The Economist Online. “Going to Town.” The Economist. The Economist Newspaper, 18
January, 2012. Web. 10 December, 2013.
http://www.economist.com/blogs/graphicdetail/2012/01/daily-chart-6
EnergyPlus [Computer Software]. (2013). Retrieved from
http://apps1.eere.energy.gov/buildings/energyplus/energyplus_about.cfm. EnergyPlus
EnergyPlus Engineering Reference (2013), U.S. Department of Energy [Online] Available:
http://apps1.eere.energy.gov/buildings/energyplus/energyplus_about.cfm. EnergyPlus
126
Executive Summary Prepared for the Cambridge Getting to Net Zero Task Force, (2015). 16
March 16, 2015 [PDF]
http://www.cambridgema.gov/CDD/Projects/Climate/~/media/1AE81F30B646463AABC
DFA488278C938.ashx
Fänger, P.O. (1982). Thermal comfort. Malabar, FL, USA: Robert E. Krieger Publishing
Company; 1982.
Gorsevski, V., Taha, H., Quattrochi, D., and Luvall, J. (1998). Air pollution prevention
through urban heat island mitigation: An update on the Urban Heat Island Pilot Project.
Proceedings of the ACEEE Summer Study, Asilomar, CA. 9: 23-32.
Hamdi, R. and Masson, V. (2008). Inclusion of a drag approach in the town energy balance
(TEB) scheme: Offline 1d evaluation in a street canyon. Journal of Applied Meteorology and
Climatology 47(10), 2627–2644. 16, 24, 42, 96, 110
Huang, L., Li J., Zhao D., Zhu J. (2008). A fieldwork study on the diurnal changes of urban
microclimate in four types of ground cover and urban heat island of Nanjing, China.
Building and Environment, 43(1), 7–17
ISO 7243 (1989), Hot Environments; Estimation of the Heat Stress on Working Man, Based
on the WBGT-Index (Wet Bulb Globe Temperature). International Organization for
Standardization, Geneva.
ISO 7933 (2004), Ergonomics of the Thermal Environment - Analytical Determination and
Interpretation of Heat Stress Using Calculation of the Predicted Heat Strain. International
Organization for Standardization, Geneva.
Jentsch, M.F., James, P.A.B., Bourikas, L. and Bahaj, A.S. (2013) Transforming existing
weather data for worldwide locations to enable energy and building performance
simulation under future climates, Renewable Energy, Volume 55, pp 514-524
Johnson, I. “China’s Great Uprooting: Moving 250 Million into Cities”. The New York Times.
The New York Times Company, 16 June, 2013. Web. 15 December, 2013.
http://www.nytimes.com/2013/06/16/world/asia/chinas-great-uprooting-moving-250million-into-cities.html?pagewanted=all&_r=1&
Kampmann, B., Bröde, P. (2009), Physiological Responses to Temperature and Humidity
Compared with Predictions of PHS and WBGT. In: Castellan, JW; Endrusick, TL (eds.):
Environmental Ergonomics XIII, University of Wollongong, Wollongong, pp 54-58.
Kikegawa, Y., et al., (2003). Development of a numerical simulation system toward
comprehensive assessments of urban warming countermeasures including their impacts
upon the urban buildings’ energy-demands. Applied Energy, 76 (4), 449–466.
127
Kolokotroni, M., Ren, X., Davies, M., Mavrogianni, A. (2012). London’s urban heat island:
Impact on current and future energy consumption in office buildings. Energy and Buildings
47 (2012) 302–311
Kottek, M., Grieser, J., Beck, C., Rudolf, B., and Rubel, F. (2006) World Map of the KöppenGeiger climate classification updated. Meteorologische Zeitschrif, 15, 259-263. DOI:
10.1127/0941-2948/2006/0130.
Kurn, D.M., Bretz, S.E., Huang, B., Akbari, H (1994). The potential for reducing urban air
temperatures and energy consumption through vegetative cooling. Heat Island Project, 1
May, 1994.
Mack Scogin Merrill Elam Architects and Michael van Valkenburgh Associates Inc. (2014a)
Massachusetts Institute of Technology East Campus Urban Design Study: Final Report. 14
March 2014. [PDF file]
http://web.mit.edu/mit2030/documents/MIT_EastCampusStudy_FinalReport_1of3.pdf
http://web.mit.edu/mit2030/documents/MIT_EastCampusStudy_FinalReport_2of3.pdf
http://web.mit.edu/mit2030/documents/MIT_EastCampusStudy_FinalReport_3of3.pdf
http://web.mit.edu/mit2030/documents/MIT_EastCampusStudy_Appdx_A_Sustainability.
pdf
Mack Scogin Merrill Elam Architects and Michael van Valkenburgh Associates Inc. (2014b)
Massachusetts Institute of Technology East Campus Urban Design Study MIT and
Cambridge Community Forum. 12 February 2014. [PDF file]
http://web.mit.edu/mit2030/documents/2014-0212_MIT_Comm_Forum_EC.pdf
Masson, V. (2000). A physically-based scheme for the urban energy budget in atmospheric
models. Boundary-Layer Meteorology 94, 357-397. 16, 18, 23, 26, 27, 41, 107, 117, 123
Michalatos, P. (2014). Millipede [Computer software]. Retrieved from http://sawapan.eu/
MIT 2030 East Campus Urban Design Study. n.d. Web. 21 February, 2014.
http://web.mit.edu/mit2030/projects/eastcampusgateway/index.html
MIT Design Committee (2013). East Campus/ MIT Gateway Alternative Approaches. 19
July, 2013 [PDF file] http://capitalprojects.mit.edu/sites/default/files/imce/MIT_eastcampus_gateway_rpt_07-2013.pdf
MIT Office of the Executive Vice President and the Treasurer (2013). MIT PUD-5 petition to
amend the Cambridge Zoning Ordinance and Cambridge Zoning Map. 13 December 2012
[PDF file]
http://www.cambridgema.gov/~/media/Files/CDD/ZoningDevel/Amendments/2012/mi
tpud5/zngamend_mit_pud5_petition.ashx
Nakicenovic, N., Swart, R. “IPCC Emissions Scenarios” (2000). Web. 30 April 2015.
http://www.ipcc.ch/ipccreports/sres/emission/index.php?idp=98
128
New York Department of City Planning. (2011). Zoning handbook. New York City: New
York City Department of City Planning, Print.
Oke, T.R. (1987). Boundary Layer Climates. London: Methuen and Company, Print.
OpenStudio [Computer software]. (2015). Retrieved from https://www.openstudio.net/
Ordinance Number 1355 (2013). Final Publication Number 3303. First Publication in the
Chronicle on March 21, 2013, City of Cambridge
Oxizidis, S., Dudek, A., Papadopoulos, A. (2008). A computational method to assess the
impact of urban climate on buildings using modeled climatic data. Energy and Buildings, 40
(3), 215–223.
Radiance [Computer software]. (2014). Retrieved from
http://radsite.lbl.gov/radiance/framed.html
Reinhart, C. (2012). Daysim [Computer software]. Retrieved from http://daysim.ning.com/
Reinhart, C., Dogan, T., Jakubiec, A., Rekha, T., and Sang, A. (2013). umi - An Urban
Simulation Environment for Building Energy Use, Daylighting, and Walkability. Proceedings
of BS2013: 13th Conference of International Building Performance Simulation Association,
Chambéry, France, August 26-28, 476–483.
Rhinoceros [Computer software]. (2014). http://www.rhino3d.com/
RhinoCommon Plug-in SDK. (2014). Retrieved 1 May, 2015 from the McNeel Wiki:
http://wiki.mcneel.com/developer/rhinocommon
Robinson, D. (2011). Computer Modelling for Sustainable Urban Design: Physical
Principles, Methods and Applications. London: Earthscan, Print.
Roudsari, M. (2015) Ladybug + Honeybee [Computer software]. Retrieved from:
http://www.grasshopper3d.com/group/ladybug?xg_source=activity
Sailor, D.J., (2011). A review of methods for estimating anthropogenic heat and moisture
emissions in the urban environment. International Journal of Climatology, 31 (2),
189–199.
Salamanca, F., Krpo, A., Martilli, A., Clappier, A. (2010). A new building energy model
coupled with an urban canopy parameterization for urban climate simulations - part i.
formulation, verification, and sensitivity analysis of the model. Theoretical and Applied
Climatology 99, 331-344. 16, 19, 24, 41, 107
129
Samuelson, H., Lantz, A., Reinhart, C. F. (2012) Non-technical barriers to energy model
sharing and reuse. Building and Environment, 54, 71-76
Santiago, J. and Martilli, A. (2010). A dynamic urban canopy parameterization for mesoscale
models based on computational fluid dynamics Reynolds-averaged Navier-Stokes
microscale simulations. Boundary-Layer Meteorology 137, 417–439.
Schmidt, M. and Ruiz, I. (2014). Letter to the community on East Campus/Kendall Square
design firm selection [Press release]. Retrieved from
http://newsoffice.mit.edu/2014/schmidt-ruiz-letter-design-firm-selection-0923
SketchUp [Computer Software]. (2015) Retrieve from http://www.sketchup.com/
SOS (1995). "The State of the Southern Oxidants Study: Policy-Relevant Findings in Ozone
Pollution Prevention Research 1988-1994." Southern Oxidants Study. Raleigh, NC: 94
Stewart, I.D., Oke, T. (2012). Local Climate Zones for Urban Temperature Studies. Bulletin
of American Meteorological Society, 92, 1879–1900
Street, M., Reinhart, C., Norford, L and Ochsendorf, J (2013). Urban heat island in Boston –
Evaluation of urban air-temperature models for predictive building energy use, Building
Simulation 2013, Chambery, France
Summary of the Massachusetts Stretch Energy Code. (2011) [Online] Retrieved from
http://www.mass.gov/eea/docs/doer/green-communities/grant-program/stretch-codesummary-jun20-2011.pdf
Tan J., Zheng Y, Tang X., Guo, C., Li L., Song G., Zhen X., Yuan D., Kalkstein A., Li A., Chen H.
(2010) The urban heat island and its impact on heat waves and human health in Shanghai.
International Journal of Biometeorology (2010) 54:75–84
umi [Computer software]. (2014) Retrieved from:
http://urbanmodellinginterface.ning.com/
United Nations Department of Economic and Social Affairs (2004). World Population to
2300.
http://www.un.org/esa/population/publications/longrange2/WorldPop2300final.pdf
U.S. Department of Energy (n.d.). Commercial Reference Buildings. [Data file] Retrieved
from http://www1.eere.energy.gov/buildings/commercial/ref_buildings.html
U.S. Department of Energy (2013). Weather Data. In EnergyPlus Energy Simulation
Software [Data file] Available:
http://apps1.eere.energy.gov/buildings/energyplus/weatherdata_about.cfm
130
U.S. Department of Transportation Federal Highway Administration. "Chapter 4 - Sidewalk
Design Guidelines and Existing Practices”. Designing Sidewalks and Trails for Access. 10
Feb. 2014. Web. 04 May 2015.
http://www.fhwa.dot.gov/environment/bicycle_pedestrian/publications/sidewalks/chap
4a.cfm
UTCI_a002. [Computer software]. (2009). Retrieved from: http://www.utci.org/
Waddell, P. (2002) UrbanSim: Modeling Urban Development for Land Use, Transportation
and Environmental Planning. Journal of the American Planning Association, Vol. 68 No. 3,
pp. 297-314.
Waddell, P. (2011). Integrated Land Use and Transportation Planning and Modelling:
Addressing Challenges in Research and Practice. Transport Reviews, 31: 2, 209-229.
131
132
Definitions of parameters specific to UWG is included below. Complete set of the definitions
are on the research website with typical values, units, and design tips:
http://urbanmicroclimate.scripts.mit.edu/uwg_parameters.php
Average building height: average
building height of the site area weighted
by building footprint
Facade-to-site ratio: ratio of the vertical
surface area [walls] to the urban plan
∑𝑃ℎ
area. Defined by 𝐴 𝑤𝑡𝑑 where P = building
Average SHGC: average of the solar heat
gain coefficient weighted by facade area
perimeter, hwtd = average building height,
and Asite = total site area. This parameter
is used to calculate canyon height and
thus solar radiation received by building
facade. Formally called vertical to
horizontal urban area ratio
𝑠𝑖𝑡𝑒
Average U-value: average of the glazing
U-value weighted by facade area
Average WWR: average of the windowto-wall ratio weighted by facade area
Characteristic length: the proxy of the
site size. Calculated as the square root of
the site area
Site coverage ratio: describes how close
buildings are built in the city. Defined by
∑𝐴𝑏𝑙𝑑𝑔
where Abldg = building footprint and
𝐴
𝑠𝑖𝑡𝑒
Asite = site area. This is similar to lot
coverage ratio in the New York
Department of City Planning (2011)
Vegetation coverage: approximated as
sum of park area in umi
Figure 59
Urban site geometric parameters (extracted from Figure 7)
133
Note:
Values derived from Bueno et al. (2014)
134
Note:
Values derived from Bueno et al. (2014)
A sample xml input file is shown here. The file includes the parameter inputs required
for UWG. The sample file shown here is the recommended design (Alternative 6) in the MIT
East Campus case study in Chapter 6.
135
5/17/2015
MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof50%25,veg.xml
This XML file does not appear to have any style information associated with it. The document
tree is shown below.
<xml_input>
<construction>
<wall>
<albedo>0.1</albedo>
<emissivity>0.7</emissivity>
<materials>
<names>
<item>layer1</item>
<item>layer2</item>
<item>layer3</item>
<item>layer4</item>
</names>
<thermalConductivity>
<item>0.72</item>
<item>8.23</item>
<item>0.93</item>
<item>0.48</item>
</thermalConductivity>
<volumetricHeatCapacity>
<item>1562135.954</item>
<item>536408.2344</item>
<item>1366946.681</item>
<item>720350.9993</item>
</volumetricHeatCapacity>
<thickness>[0.07,2.0,0.14,0.02]</thickness>
</materials>
<vegetationCoverage>0</vegetationCoverage>
<inclination>0</inclination>
<initialTemperature>20</initialTemperature>
</wall>
<roof>
<albedo>0.07</albedo>
<emissivity>0.98</emissivity>
<materials>
<names>
<item>built‐up_roofing</item>
<item>sheathing</item>
<item>insualtion</item>
</names>
<thermalConductivity>
<item>0.16</item>
<item>0.057</item>
<item>0.07</item>
</thermalConductivity>
<volumetricHeatCapacity>
<item>1634923.651</item>
<item>3212768.95</item>
<item>194877.3112</item>
</volumetricHeatCapacity>
<thickness>[0.01,0.01,0.16]</thickness>
</materials>
<vegetationCoverage>0.5</vegetationCoverage>
<inclination>1</inclination>
file://win.mit.edu/dfs/profiles/a/anakano/WinData/My%20Documents/Downloads/MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof…
1/4
5/17/2015
MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof50%25,veg.xml
<initialTemperature>20</initialTemperature>
</roof>
<mass>
<albedo>0.1</albedo>
<emissivity>0.7</emissivity>
<materials>
<names>
<item>concrete</item>
</names>
<thermalConductivity>
<item>1.13</item>
</thermalConductivity>
<volumetricHeatCapacity>
<item>1839689.04</item>
</volumetricHeatCapacity>
<thickness>[0.10]</thickness>
</materials>
<vegetationCoverage>0</vegetationCoverage>
<inclination>1</inclination>
<initialTemperature>20</initialTemperature>
</mass>
<glazing>
<glazingRatio>0.382</glazingRatio>
<windowUvalue>4.41</windowUvalue>
<windowSHGC>0.55</windowSHGC>
</glazing>
<urbanRoad>
<albedo>0.165</albedo>
<emissivity>0.95</emissivity>
<materials>
<names>
<item>asphalt</item>
</names>
<thermalConductivity>
<item>1</item>
</thermalConductivity>
<volumetricHeatCapacity>
<item>1600000</item>
</volumetricHeatCapacity>
<thickness>1.25</thickness>
</materials>
<vegetationCoverage>0.75</vegetationCoverage>
<inclination>1</inclination>
<initialTemperature>20</initialTemperature>
</urbanRoad>
<rural>
<albedo>0.165</albedo>
<emissivity>0.95</emissivity>
<materials>
<names>
<item>asphalt</item>
</names>
<thermalConductivity>
<item>1</item>
</thermalConductivity>
<volumetricHeatCapacity>
<item>1600000</item>
file://win.mit.edu/dfs/profiles/a/anakano/WinData/My%20Documents/Downloads/MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof…
2/4
5/17/2015
MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof50%25,veg.xml
</volumetricHeatCapacity>
<thickness>1.25</thickness>
</materials>
<vegetationCoverage>0.5</vegetationCoverage>
<inclination>1</inclination>
<initialTemperature>20</initialTemperature>
</rural>
</construction>
<building>
<floorHeight>3</floorHeight>
<dayInternalGains>25.89</dayInternalGains>
<nightInternalGains>10.50</nightInternalGains>
<radiantFraction>0.5</radiantFraction>
<latentFraction>0.09</latentFraction>
<infiltration>0.31</infiltration>
<ventilation>1.32</ventilation>
<coolingSystemType>air</coolingSystemType>
<coolingCOP>3.7</coolingCOP>
<daytimeCoolingSetPoint>25</daytimeCoolingSetPoint>
<nighttimeCoolingSetPoint>26</nighttimeCoolingSetPoint>
<daytimeHeatingSetPoint>20</daytimeHeatingSetPoint>
<nighttimeHeatingSetPoint>17</nighttimeHeatingSetPoint>
<coolingCapacity>205</coolingCapacity>
<heatingEfficiency>0.8</heatingEfficiency>
<nightSetStart>19</nightSetStart>
<nightSetEnd>5</nightSetEnd>
<heatReleasedToCanyon>0</heatReleasedToCanyon>
<initialT>20</initialT>
</building>
<urbanArea>
<averageBuildingHeight>34.607324</averageBuildingHeight>
<horizontalBuildingDensity>0.471381</horizontalBuildingDensity>
<verticalToHorizontalUrbanAreaRatio>2.311519</verticalToHorizontalUrbanAreaRatio>
<treeCoverage>0.225519</treeCoverage>
<nonBldgSensibleHeat>6.8</nonBldgSensibleHeat>
<nonBldgLatentAnthropogenicHeat>0</nonBldgLatentAnthropogenicHeat>
<charLength>1000</charLength>
<treeLatent>0.7</treeLatent>
<grassLatent>0.6</grassLatent>
<vegAlbedo>0.25</vegAlbedo>
<vegStart>1</vegStart>
<vegEnd>12</vegEnd>
<daytimeBLHeight>700</daytimeBLHeight>
<nighttimeBLHeight>80</nighttimeBLHeight>
<refHeight>150</refHeight>
</urbanArea>
<referenceSite>
<latitude>42.3598</latitude>
<longitude>‐71.063694</longitude>
<averageObstacleHeight>0.1</averageObstacleHeight>
</referenceSite>
<parameter>
<tempHeight>2</tempHeight>
<windHeight>10</windHeight>
<circCoeff>1.2</circCoeff>
<dayThreshold>200</dayThreshold>
<nightThreshold>50</nightThreshold>
file://win.mit.edu/dfs/profiles/a/anakano/WinData/My%20Documents/Downloads/MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof…
3/4
5/17/2015
MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof50%25,veg.xml
<windMin>0.1</windMin>
<windMax>10</windMax>
<wgmax>0.005</wgmax>
<exCoeff>0.3</exCoeff>
<simuStartMonth>1</simuStartMonth>
<simuStartDay>1</simuStartDay>
<simuDuration>365</simuDuration>
</parameter>
</xml_input>
file://win.mit.edu/dfs/profiles/a/anakano/WinData/My%20Documents/Downloads/MIT30_2_CASE6_case5mproved_thickerIns,urbanroadvegupby0.25,greenroof…
4/4
Purpose. The PUD-5 District is intended to provide for Kendall Square’s continued
prominence as a world-renowned center of innovation and a vibrant neighborhood
through the creation of a mixed-use district of high quality general and technical office and
laboratory uses with significant retail activity proximate to the MBTA station. The PUD-5
District helps organize placement of commercial and institutional buildings and establishes
an additional mixed-use development containing a significant residential component to
support the burgeoning residential corridor along Third Street and the strong links to
existing neighborhoods and the riverfront. The PUD-5 District allows for continued support
of the academic mission at MIT and encourages connective links, physical and otherwise,
between the Institute and adjacent neighborhoods.
The PUD-5 District responds to the Kendall Square planning process and is intended to
be a smart-growth, transit-oriented district and therefore allows for replacing surface
parking lots with larger scale development in Kendall Square and the major public transit
services located there. The PUD-5 District encourages low parking ratios, shared parking
strategies, the use of public transportation and improved pedestrian and bicycle
environments. The PUD-5 District furthers the City’s goals for sustainable development
through buildings and sites that are planned, designed and constructed in a sustainable
way so as to minimize adverse environmental impacts as they are initially constructed and
as they are occupied and operated over the course of their useful lives.
The PUD-5 District promotes the creation of a strong retail corridor along Main Street
and the enhancement of Broad Canal Way. Combined, this new public crossroads will have
broad appeal as a desirable destination during and beyond the traditional workday by
providing a critical mass of diverse restaurants, shops, entertainment and programming.
The ground floor space will engage pedestrians and provide a variety of indoor and
outdoor gathering spaces, including retail that can address the needs and reflect the
creativity of the local community.
A new update to the plan has been released since the case study is developed and is
included here for reference [C. Barnhart, M. A Schmidt, and I. Ruiz, “Campus Planning /
Kendall Update and Community Meetings”, email, 14 April, 2015]. The fundamental design
strategy and geometries remain the same as Figure 36.
140
141