Content uploaded by Tianzhen Hong
Author content
All content in this area was uploaded by Tianzhen Hong on Jul 05, 2016
Content may be subject to copyright.
CityBES: A Web-based Platform to Support City-Scale
Building Energy Efficiency
Tianzhen Hong, Yixing Chen, Sang Hoon Lee, Mary Ann Piette
Building Technology and Urban Systems Division
Lawrence Berkeley National Laboratory
One Cyclotron Road, Berkeley, CA 94720, USA
thong@lbl.gov, yixingchen@lbl.gov, sanghlee@lbl.gov, mapiette@lbl.gov
ABSTRACT
Buildings in cities consume 30 to 70% of the cities’ total primary
energy. Retrofitting the existing building stock to improve energy
efficiency and reduce energy use is a key strategy for cities to
reduce green-house-gas emissions and mitigate climate change.
Planning and evaluating retrofit strategies for buildings requires a
deep understanding of the physical characteristics, operating
patterns, and energy use of the building stock. This is a challenge
for city managers as data and tools are limited and disparate. This
paper introduces a web-based data and computing platform, City
Building Energy Saver (CityBES), which focuses on energy
modeling and analysis of a city’s building stock to support district
or city-scale efficiency programs. CityBES uses an international
open data standard, CityGML, to represent and exchange 3D city
models. CityBES employs EnergyPlus to simulate building energy
use and savings from energy efficient retrofits. CityBES provides
a suite of features for urban planners, city energy managers,
building owners, utilities, energy consultants and researchers.
CCS Concepts
• Information systems
➝
Database management
systems
➝
Database design and models • Information
systems
➝
Information systems applications
➝
Spatial-temporal
systems • Information systems
➝
Information systems
applications
➝
Decision support systems
➝
Data analytics
• Computing methodologies
➝
Modeling and
simulation
➝
Model development and analysis.
Keywords
Building stock; building simulation; urban computing; CityGML;
EnergyPlus; CityBES; building retrofit; energy savings.
1. INTRODUCTION
Urbanization is one of the great challenges of this century, with
linkages to climate change and the need to develop sustainable use
of energy and other natural resources. Urban energy models aim
to explore opportunities to address these issues by combining the
data generated in cities with new energy simulation tools. Urban
computational tools combine urban sensing, data management,
and data analytics, to evaluate city-scale energy and
environmental systems. Urban computing is an interdisciplinary
field where computer science meets city-related fields, like
transportation, civil engineering, energy supply and demand
analysis, environmental science, economics, ecology, and
sociology in the context of urban spaces [1].
With buildings responsible for about one-third of global energy
consumption and a quarter of CO2 emissions, there is a huge,
untapped opportunity to create and transform cities to more
sustainable environments through improving building energy
efficiency [2]. More efficient buildings can generate economic
benefits, reduce environmental impacts and improve people’s
quality of life. More than two-thirds of people in the U.S. live in
urban areas. These areas face growing challenges to accelerate
building retrofit activities and expand operational efficiency to
reduce energy use and greenhouse gas emissions and to meet
sustainability goals. Urban energy analysis is a complex, multi-
scale, multi-sector challenge. Cities need to be able to evaluate
their current energy use and explore how to compare, rank,
contrast, and estimate strategies to reduce energy use and
environmental impacts. Cities need to evaluate building retrofit
opportunities for their local stock considering energy usage,
vintage, size, type, ownership, and socioeconomic capabilities of
each neighborhood. Advanced shared energy infrastructure, such
as district heating and cooling systems, can provide huge
increases in energy efficiency by combining diverse loads where
the integrated energy performance of a group of buildings can be
less than the simple sum of individual buildings.
Cities need quantitative decision analysis tools that combine
measured data, physics- and data-driven models to support both
new and retrofits building systems. Designing and operating such
systems requires dynamic computer simulation and optimization
to account for the complexities of energy systems, such as
different types of building systems, operating patterns, uncertainty
and variability of weather, user behavior. Recent efforts to
develop these tools have integrated these computation models
with geographical information system (GIS) to obtain input data
for hundreds to thousands of buildings, and to visualize results in
a form that is accessible to urban planners and designers.
Several urban energy simulation tools have been developed [3],
[4], including the Urban Building Energy Models (UBEM) [4],
[5], CitySim [6] and the Urban Modeling Interface (UMI) [7].
UBEM estimates citywide hourly energy demand from energy
simulations of individual buildings in a city, supporting city
policy makers to evaluate strategies on urban building energy
efficiency. 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 creates EnergyPlus models using
simplified zoning and HVAC systems. Rhino is a commercial 3D
computer graphics and computer-aided design (CAD) application
software. CitySim provides decision support for urban planners to
SAMPLE: Permission to make digital or hard copies of all or part of this
work for personal or classroom use is granted without fee provided that
copies are not made or distributed for profit or commercial advantage
and that copies bear this notice and the full citation on the first page. To
copy otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Urban Computing, August 14, 2016, San Francisco, California, USA.
Copyright 2016 ACM 1-58113-000-0/00/0010 …$15.00.
DOI: http://dx.doi.org/10.1145/12345.67890
minimize energy and emissions by simulating the energy demand
of buildings. CitySim uses its own XML schema to represent
building information and a reduced order energy models assuming
simplified zoning and HVAC systems. However, these tools are
isolated, limited to specific applications, and not using open
standards, which are key to share and exchange information
across a wide array of urban modeling tools.
This paper introduces City Building Energy Saver (CityBES),
describing the data structures, workflow automation, integration
with existing urban data models, energy models, and calibration
techniques. As an energy modeling and analysis urban computing
platform, CityBES can provide insights to inform city
stakeholders on the opportunities for new energy technologies and
retrofit policies. For example, CityBES can help identify the
technologies and strategies need to retrofit city buildings to save
30% to 50% of total energy. CityBES can also be used to evaluate
the impact of climate change (long term warming and extreme
heat waves) on building performance and occupant health, and
strategies to mitigate such impact. The tool can be used to
evaluate the impact of urban heat islands on building performance
and strategies to mitigate such effects. One can also evaluate the
feasibility of new advanced district heating and cooling (DHC)
systems that target 50% energy savings through EnergyPlus’ co-
simulation with DHC models in Modelica. CityBES will in future
be able to evaluate the potential of on-site renewable energy
generation (PV or solar thermal) in buildings.
The paper begins with a review of energy modeling approaches
for urban energy analysis, and highlights approaches used in
CityBES. We then describe the software architecture, modeling,
and analysis capabilities of CityBES. Potential use cases,
challenges and the future development of CityBES are discussed
as well.
2. OVERVIEW OF ENERGY MODELING
METHODS
There are three approaches commonly used to model energy use
in buildings. First, physics-based models use first principals to
consider detailed heat and mass balance and heat transfer within
and across systems and components in buildings. Second,
reduced-order models simplify the spatial and temporal dynamics
in buildings and their energy systems. Third, data-driven models
correlate output results with limited available independent
parameters using mathematical regressions, data mining, or
neutral network methods to predict energy use.
2.1 Physics-based Models
Physics-based energy modeling, the highest fidelity and the most
complex option, can provide accurate energy use results of real
buildings. Most of the retrofit toolkits based on physics-based
energy modeling utilize publicly available simulation engines,
such as EnergyPlus [8], ESP-r [9], and DeST [10], and DOE 2.2
(eQuest) [11]. EnergyPlus use heat and mass balance equations to
model detailed dynamics of complex mechanical systems, e.g.
variable refrigerant systems, radiant cooling and heating systems,
and natural ventilation. One challenge with physics-based models
is the difficulty determining many of the input parameters.
Another challenge is the need of measured data to calibrate energy
models. The advantage of using detailed physics-based models
can be the evaluation of an integrated effect. An example of an
integrated effect would occur during a lighting retrofit. Upgrading
the lighting system not only contributes to lighting energy
savings, but also reduces the cooling load, thus reducing the space
cooling energy consumption.
2.2 Reduced-order Models
A reduced-order model uses simple input and output data
providing a quick evaluation of the energy performance of a
building, requiring an appropriate model structure and normative
values of the model parameters. There are a variety of forms of
reduced order models with a resistor-capacitance (RC) model
being a common model form. A well-known reduced-order
model, the normative method, is a first order energy model based
on quasi-steady-state heat balance equations. The normative
method follows calculation standards developed by the European
Committee for Standardization (CEN) and the International
Organization for Standardization (ISO) [12] which defines the
calculation method with a set of normative statements containing
physical building parameters and building systems for different
building types. The normative model based on ISO 13790 [13] is
a well-known reduced-order model. The method calculates the
energy use at different levels of the thermal energy demand,
delivered energy per carrier, primary energy and emissions.
Through simplified and unified modeling assumptions, the
method forms the basis for assessing building energy performance
in a standardized and transparent way. Traditionally used for
energy performance ratings [14], [15], normative calculations can
support retrofit analysis for large-scale energy assessment [16],
[17]. Reduced-order models may not be as accurate as detailed
physics-based models, yet offer advantages for simple energy
analysis because of computational efficiency with fewer inputs
required.
2.3 Data-driven Models
Data-driven models have been used to predict building energy
consumption using simple benchmarking or more complex
regression modeling, to relate building design and operational
parameters with energy consumption. These methods rely on
measured data, such as hourly electric data and energy use
intensity databases for benchmarking. For example, many
building energy baseline characterization models for measurement
and verification fall into this category [18]. Some of the
challenges with empirically data-driven models include: (1) the
requirement of having training data to develop the model, (2) the
model is limited to a specific building and may not be applicable
to other buildings, and (3) there lacks a physical explanation of
certain parameters of building performance. The regression model
derived from statistical methods can be used to solve certain
inverse problems. Differing from the conventional energy
modeling processes, the inverse statistical model derives inputs
from known outputs [19], allowing a building design or
operational parameter to be inferred when energy consumption
data are available. Regression methods applied to existing data
and inverse solving techniques can be used to quickly estimate the
energy consumption of individual buildings with a few parameters
or to be used to derive additional information from city-wide
energy consumption data. However, there is a major gap in this
approach in that the energy model does not capture the dynamics
of the integrated effects of energy conservation measures (ECMs).
3. CityBES
3.1 CityBES Overview
CityBES is a web-based platform to simulate energy performance
of a city’s building stock, from a small group of buildings in an
urban district to all buildings in a city. CityBES builds upon the
LBNL Commercial Building Energy Saver Toolkit [20], which
provides retrofit analysis of individual commercial buildings of
small and medium offices and retails. CityBES will add other
commercial buildings types (e.g., large offices, hotels, hospitals)
as well as residential buildings (single family and multi-family).
In addition, district heating and cooling systems will be added as
retrofit options. CityBES also adds new ECMs for new
commercial and residential building types. To handle simulation
of many buildings simultaneously, CityBES implements a parallel
computing architecture to utilize high-performance computing
(HPC) clusters.
CityBES uses CityGML as the data schema to represent the urban
building stock. It provides 3D visualization as displayed in Figure
1, which shows color coded simulated energy performance of
buildings in New York City.
CityBES is a tool to help city managers and stakeholders evaluate
options to reduce energy use by quantifying and prioritizing
building retrofit solutions at a large scale. The tool is capable of
modeling 10,000 or more buildings and identifying deep energy
savings of 30% to 50%. This concept is intended to exceed the
capabilities of the current practice of evaluating retrofits of single
buildings one at a time and with limited energy savings of 10% to
20%. CityBES is also designed to enable research to explore
opportunities of interactions (such as simultaneous heating and
cooling or opportunities for energy storage) between buildings in
district-scale energy systems.
Figure 1: CityBES: Buildings in Manhattan New York (for illustrative only, using mockup building data)
Figure 2: Software Architecture of CityBES
3.2 Software Architecture
Figure 2 shows the three layers of the software architecture of
CityBES: the Data layer, the Algorithms and Software layer, and
the Use Cases layer. The Data layer includes the weather data, and
the 3D city model represented in CityGML compiled from
building stock, GIS and other database. The Software layer
includes EnergyPlus, OpenStudio [21] and CityBES. The Use
Cases layer provides examples of potential applications, including
energy benchmarking, urban energy planning, energy retrofit
analysis, building operation improvement, as well as performance
visualization.
3.3 Open 3D Data Model for Cities
Urban data models serve as the core layer of CityBES as they
store data from various sources and provide inputs to the
analytics, modeling, and GIS visualization. CityBES uses
CityGML [22], the international open standard of the Open
Geospatial Consortium (OGC), to represent and exchange 3D city
models. CityGML, an XML-based open data model, is an
application schema for the Geography Markup Language (GML),
which provides a standardized geometry model. CityGML
includes modules to represent bridges, buildings, city furniture,
land use, transportation, tunnels, vegetation, water bodies, etc.
Figure 3 shows some examples of CityGML objects. The
CityGML provides virtual 3D city models for advanced analysis
and visualization in a variety of application domains such as urban
planning, indoor/outdoor pedestrian navigation, environmental
simulations, cultural heritage, or facility management [23]. The
CityGML version 1.0 was released in 2008, and an extended
version 2.0 was adopted in March 2012.
Figure 3: Examples of CityGML objects (source [24])
Figure 4: Five levels of details (LODs) to represent buildings
in CityGML
CityGML provides a common definition of the basic entities,
attributes, and relations of a 3D city model, allowing the reuse of
the data in different applications. City officers are often required
to provide city data in different formats for different applications,
which is unnecessary redundant work. To make the city data
reusable, more and more cities are creating their 3D city models
using CityGML by consolidating the data from different sources.
For example, Berlin 3D Portal provides CityGML data in LOD 2
for their entire 550,000 buildings in Berlin [25].
CityGML is mapped to a database as the data structure of
CityBES, which is used for data management. CityGML (Figure
4) has five levels of details (LOD) to represent the city, landscape
and infrastructure for building energy modeling and urban micro-
climate simulations. CityGML allows user-defined objects and
attributes to extend the data model for domain specific data
elements. The Energy Application Domain Extension (ADE),
which is under development, is used in CityBES to represent and
exchange other essential data needed for building energy models,
e.g., constructions and materials, operation schedules, and energy
systems for lighting, plug-loads, and heating, ventilation, and air
conditioning (HVAC).
3.4 Energy Modeling Approach
CityBES enables energy modeling through the OpenStudio
software development kit (SDK) and the EnergyPlus simulation
engine. EnergyPlus is the U.S. Department of Energy (DOE)’s
flagship simulation program for modeling dynamic energy and
environmental performance of buildings. It conducts detailed heat
and mass balance calculations for each room in a building, and
can estimate sub-hourly (from 1 to 60 minutes) energy use of
building systems, including lighting, plug-loads, process-loads
(e.g., elevators), HVAC and service water heating. EnergyPlus
has been widely used by engineers, architects, and researchers to
support the design of new buildings and the retrofit of existing
buildings to increase efficiency and reduce energy use and carbon
emissions. EnergyPlus is also used to support building energy
codes and standards development, as well as support utility
incentive programs and state and federal energy policies.
EnergyPlus has about 800,000 lines of C/C++ code. It is a
console-based program that reads input and writes output to text
files. EnergyPlus is free, open-source, and cross-platform; it runs
on the Windows, Mac OS X, and Linux operating systems.
EnergyPlus simulation of individual buildings has been verified
and validated using test cases from ASHRAE Standard 140 [26].
Datasets from cities’ public building energy use disclosure and
benchmarking ordinance can be used to validate the simulation
results of the baseline buildings at a portfolio level from CityBES.
Another dataset can be used in future is DOE’s Building
Performance Database (BPD) [27], which currently has more than
800K buildings and provides an estimate of energy savings from
the retrofit for a group of buildings
OpenStudio is an SDK for interfacing with EnergyPlus input and
output files as well as managing simulations. The OpenStudio
SDK provides an object-oriented interface to the building energy
model. The OpenStudio SDK can be used to rapidly create full
building energy models based on limited available data; it can also
be used to alter existing building energy models.
3.5 Energy Conservation Measures
CityBES integrates more than 75 ECMs from various sources,
including the Database for Energy Efficiency Resources (DEER)
[28], the Advanced Energy Retrofit Guide for offices and retails
[29]–[31], and RSMeans (rsmeans.com). The ECM database
includes detailed descriptions of the technical specifications,
modeling methods and investment costs for each ECM. In
general, typical and emerging building technologies of the
building envelope, HVAC, indoor lighting, plug-loads, service
water heating, outdoor lighting, and building operation and
maintenance were specified. The measures and modeling of those
building systems are systematically applied to the CityBES
framework through EnergyPlus simulation for the city building
stock retrofit analysis.
3.6 Data
CityBES leverages existing data from several different sources
that are compiled into a central database. The database includes a
CityGML file representing 3D city model that provides the
majority of the data for building energy model. Other essential
data in the database are weather, building characteristics, and
ECMs. CityBES uses typical meteorological year weather data in
EnergyPlus simulations [32], and allows user-defined weather
data measured at local stations. Building characteristic data will
be extracted from various city data sources, including assessors’
records, GIS data, public building energy use disclosure, and
energy benchmarking ordinance. CityBES also has prototype
buildings that meet minimal requirements of ASHRAE 90.1 [33]
and California Title 24 [34] standards at various vintages,
providing default efficiency levels for buildings built at various
vintages and climates. For retrofit analysis, economic data such as
energy costs, investment costs, discount rate and payback years
are used.
CityBES provides a rich dataset as the result of the urban scale
energy simulation. CityBES generates the energy consumption of
the current building stock as a baseline of the city’s building
energy performance. Annual, monthly, and hourly energy usage
data are available to characterize current energy use. Energy end
uses data are available to help identify energy saving potentials
for different building systems. Taking the current energy usage
data as the baseline, CityBES can offer a wide array of analysis
suited for city’s energy efficiency program, including energy
benchmarking, energy savings, greenhouse gas reductions,
operation improvements, and energy costs reductions. Table 1
shows a sample list of ECMs that can be applied during energy
retrofit analysis.
Table 1 Energy Conservation Measures (A Sample List) Used in CityBES
Category Component Name Description
Lighting
Interior Lighting
Equipment
Retrofit
Replace existing lighting with
LED upgrade (6.5 W/m2)
Replace existing lighting to LEDs with 6.5 W/m2. LEDs consume less power and last
longer than fluorescent lamps. A retrofit kit is recommended for converting ballasts.
Replacement may improve lighting quality.
Plug Loads Equipment
Control
Use Plug Load Controller (30%
efficient from Baseline)
Connect plug loads to a smart plug strip with some or all of the following functions:
Occupancy sensing, load sensing, timers, remote control.
Envelope -
Exterior Wall Exterior Wall Apply Wall Insulation (R21) Apply blown-fiberglass insulation (R21) to wall cavity to maintain thermal comfort.
Insulation provides resistance to heat flow, taking less energy to heat/cool the space.
Envelope -
Roof Roof Reroof and Roof with Insulation Demolish existing roof, install insulation (R24.83) and reroof to reduced unwanted
heat gain/loss. This measure is most applicable to older roofs.
Envelope -
Window Window Replace fixed-window to U-
factor (0.25) and SHGC (0.18)
Replace existing window glass and frame with high-performance windows by
changing U-factor and SHGC of window material. U-factor is a measure of thermal
transmittance and SHGC stands for Solar Heat Gain Coefficient, values taken as 1.42
W/(K·m2), SHGC: 0.18.
Service Hot
Water Storage Tank Efficiency Upgrade of the Gas
Storage Water Heater
Replace existing service hot water heater with more efficient gas storage unit, better
insulation, heat traps and more efficient burners to increase efficiency of (0.93).
HVAC -
Cooling Cooling System Packaged Rooftop VAV Unit
Efficiency Upgrade (SEER 14)
Replace RTU with higher-efficiency unit with reheat, SEER 14. Cooling only; include
standard controls, and economizer.
HVAC -
Economizer Ventilation Add Economizer Install economizer for existing HVAC system (includes temperature sensors, damper
motors, motor controls, and dampers).
Envelope -
Infiltration Infiltration Add Air Sealing to Seal Leaks Air sealing can reduce cold drafts and help improve thermal comfort in buildings. Air
sealing is a weatherization strategy, which will change air exchange rate and IAQ.
4. CityBES MAIN FEATURES
CityBES implements a suite of analytics, modeling, simulation
and visualization features to support its use for district and city-
scale building energy efficiency analysis by urban planners, city
energy managers, as well as energy consultants and researchers
for city projects.
4.1 Filtering the building stock
Depending on the use case and analysis, the city building stock
may need to be filtered to a subset of buildings by building type,
year built, total floor area, energy use intensity (EUI), and peak
electricity load per area. Figure 5 shows the design of the building
stock filters, with an example to select sites built between 1950 to
1980, medium and large office buildings, with higher energy use
intensities.
Figure 5: Building stock filters
4.2 Energy Benchmarking
The data needed for common energy use benchmarking include
building characteristics (e.g. type/use, vintage, location and floor
area), and 12 months of energy usage data. Following the data
input, the Energy Star Portfolio Manager [35] and Building
Performance Database (BPD) [27] application program interfaces
(APIs) are provided to perform a benchmarking analysis.
The U.S. Environmental Protection Agency’s Energy Star
program developed an energy performance rating system, using a
scale of 1 to 100, to provide a means for benchmarking the energy
efficiency of an individual building to evaluate its energy
performance. For assessing the Energy Star score, a minimum
score of 75 is required for Energy Star certification [36]. CityBES
obtains Energy Star scores for each selected buildings with
monthly utility data through the Energy Star Portfolio Manager
APIs; visualizes the scores by color-coding the 3D building
shapes; and further filters the building stock by the score. For
example, city managers may be interested to know buildings with
Energy Star score lower than a certain value, say 50.
BPD is the U.S. largest dataset of information about the energy-
related characteristics of commercial and residential buildings. It
combines, cleanses and anonymizes data collected by Federal,
State and local governments, utilities, energy efficiency programs,
building owners and private companies, and makes it available to
the public. CityBES can compare the EUI distribution of the
selected buildings with the peer buildings in BPD, to benchmark
the building energy performance in the district scale.
4.3 Building Energy Use Data Analytics
CityBES provides operational improvement recommendations
from the result of the smart meter interval data analysis using an
algorithm developed by Mathieu et al. [37].
CityBES analyzes weekly 24-hour daily electric load to identify
higher energy consumption during non-operating hours relative to
normal operating hours. This trend may indicate that the HVAC
system and other electrical equipment are operating despite
occupant absence during the nighttime or weekends. CityBES
shows the weekly 24-hour electric load profiles, which provide
electricity use patterns during operating and non-operating hours.
Figure 6 shows an example of the analytic result: average weekly
operating and non-operating hours, and their average electric load
densities. CityBES also provides a sensitivity analysis of the
whole building electricity use as a function of outdoor air
temperature during four periods of the day (early morning,
morning, afternoon, and night), which can infer energy use
patterns, and building tightness and ventilation rates.
Figure 6: The weekly operating and non-operating hours, and
their average power demand and energy use
4.4 Building Stock Energy Retrofit
Energy retrofit analysis is the primary use case provided by
CityBES. First, building stock can be filtered based on criteria of
interest (EUI, vintage, building types). Secondly, a suite of ECMs
(individual and packages) can be selected from the database.
Thirdly economic data such as energy costs and discount rate can
be specified. CityBES runs a series of energy simulations to
evaluate the selected ECMs and provides results such as energy
savings, energy costs reduction, and simple payback. Retrofit
measures cover most building systems and components (lighting,
building envelope, equipment (i.e. plug loads), heating,
ventilation, and air conditioning (HVAC), or service hot water
systems) as well as operation and maintenance strategies. Some
specific examples of ECMs include installing daylighting sensors
for interior lighting control, replacing wall and ceiling, or roof
insulation, upgrading an HVAC rooftop unit with a high-
efficiency unit, adding an economizer, or upgrading to light-
emitting diode (LED) lights. Hong et al. [38] identified a package
of measures resulting in an estimated 20% improvement in the
whole building electricity consumption. Li et al. [39] showed the
three most commonly installed energy-efficient technologies in
high-performance buildings are daylighting, high-efficiency
HVAC systems, and improved building envelope.
For a group of buildings, CityBES will also evaluate district
heating and cooling (DHC) systems as a retrofit option. DHC
systems have potential of reusing thermal energy between
buildings (for example waste heat from data centers can be used
to heat buildings nearby), and reducing the capacity of central
plant equipment (chillers and boilers) by taking advantage of the
diversity of loads from different buildings.
4.5 Automated Model Calibration
There are always concerns regarding potential discrepancies
between the actual and simulated energy use in buildings.
CityBES adopts an automated model calibration [40] using
monthly utility data to fine tune the baseline model before retrofit
analysis to estimate energy savings of ECMs.
The automated model calibration uses logic linking parameter
tuning with bias pattern recognition to overcome some of the
disadvantages associated with traditional calibration processes.
The pattern-based process contains four key steps: (1) running the
original pre-calibrated energy model to obtain monthly simulated
electricity and gas use; (2) establishing a pattern bias, either
universal or seasonal bias, by comparing load shape patterns of
simulated and actual monthly energy use; (3) using programmed
logic to select which parameter to tune first based on bias pattern,
weather and input parameter interactions; and (4) automatically
tuning the calibration parameters and checking the progress using
pattern-fit criteria.
The pattern-based calibration approach is fully automatic without
any need of manual intervention. The approach employs pre-
defined rules, determined by characteristics of bias patterns, to
adjust the model parameters. This method is different from the
traditional optimization-based automatic calibration approach
which searches a parameter space according to a specific
optimization algorithm to minimize the difference between the
simulated and measured energy use of the building [41].
4.6 Visualization
The CityBES main screen (Figure 2) shows a 3D view of the city
building stock. Users can apply filters (in a floating window) to
select a subset of the building stock. When a specific building is
highlighted, a list of characteristics (building name, type, vintage,
and total floor area) and its energy use and potential retrofit
saving data are displayed in a floating window. CityBES can
visualize a suite of performance metrics of buildings by color-
coding the 3D view of the buildings.
Performance metrics that can be visualized include: site or
primary energy use (absolute amount or per floor area),
greenhouse gas emissions, whole building peak electric demand,
Energy Star score, retrofit energy savings (absolute amount and
percentage), weekly operating hours, energy use breakdown into
end uses (lighting, plug-loads, cooling, heating, and process
loads), and code and compliance status.
5. COMPUTING REQUIREMENTS
Modeling an entire building stock in a city using CityBES can
require significant computing resources. For example, a detailed
EnergyPlus model for a typical commercial building can take 10
minutes to run on a 3.6 GHz desktop computer for an annual
simulation at a five-minute time step. This is a computing
requirement of 1013 FLOP. Assuming a large city such as New
York with one million buildings, 20 ECMs (individual and
packages) to explore for each building, each model calibration
effort requiring 25 iterations, inter-building effect increasing the
effort by 10X, and an integration effort of 2X, running all this in
say three hours, we have 1013 X 106 X 20 X 25 X 10 X 2 / 104 =
1019 FLOPS, an exascale problem.
Of course, quite often an urban application concerns only a subset
of city buildings and energy models can be simplified in certain
ways, which can be handled by today’s powerful servers or HPC
clusters.
6. DISCUSSION
CityBES can serve as a data and computing urban platform to
help city policymakers and their consultants to evaluate district
and city-scale energy efficiency issues and opportunities in
buildings. CityBES is targeted for analysis of city building stocks
using CityGML which provides four levels of details to represent
city buildings, and allows energy simulation with different
fidelities of modeling options. The data model using CityGML
can help exchange data between the building energy model and
other urban environmental analysis models. The integration of
city building stock data in CityBES will enable integration with
other tools such as the U.S. Department of Energy’s Building
Energy Asset Score [42] and the Building Performance Database
[27]. In future, CityBES will also support other analysis such as
urban energy planning and design, carbon emissions tracking
system, and local laws and code compliance.
CityBES will be a valuable platform assisting users to answer
important questions about technology deployment and policy such
as:
1) Which types of buildings have the greatest potential for energy
savings and cost-effective retrofits? This is part of a crucial effort
of CityBES to show a city-wide building energy map with actual
energy consumption, and estimate of energy potential savings
from retrofit as well as operational improvements, for every
building in the city.
2) Which energy efficiency technologies can help achieve the
greatest energy savings?
3) Where in the city are there districts with the right mix of load
density and diversity to support district energy systems, or local
energy storage to reduce energy use?
4) How much energy savings can be expected if all buildings in a
city use a specific retrofit, such as single pane to double pane
windows, or fluorescent to LED lights?
5) If all buildings in a city upgrade to meet the current building
energy code, how much energy savings and peak electricity
demand reduction can be achieved?
6) What is the impact of climate change on energy use in the
building stock and in occupant comfort in the next 30 or 50 years?
7) What is the impact of extreme heat waves on building energy
demand and occupants’ health?
8) If solar PV is installed on all available roof spaces of all
buildings in a city, how much electricity can be generated? How
does this meet city’s renewable energy goal? What is the cost of
such a PV deployment plan?
To develop a city-scale platform like CityBES, there are five
technical challenges. The first challenge is to collect data and
create the 3D city models with CityGML. The second challenge is
to get the energy use data of the buildings. The third is to model
all types of retrofit measures using EnergyPlus and the ECM
database. Fourth, the computing power required to run tens of
thousands or more building energy simulations for retrofit
analysis on demand is a challenge. Last but not least, we need to
enhance EnergyPlus to consider inter-building effect in the urban
environment, e.g. radiant heat exchange between exterior surfaces
of buildings, and integration with urban microclimate simulation
to consider the urban heat island effect as well and local climate
conditions. More features will be added in CityBES including
urban energy planning to evaluate technologies and strategies (e.g.
DHC systems and community scale renewable systems) to design
and optimize net-zero energy or carbon neutral communities.
7. CONCLUSIONS AND FUTURE WORK
CityBES will be a publically available web-based data and
computing platform providing a suite of analytic and modeling
features to improve energy efficiency and reduce carbon
emissions in city buildings. CityBES uses the international
standard CityGML to represent the 3D building stock in cities,
and uses EnergyPlus for detailed building energy simulation to
evaluate energy savings potential of a wide array of building
technologies. CityBES can visualize various performance metrics
of buildings by color-coding the 3D view of buildings in cities.
CityBES targets audience of urban planners, energy consultants,
city and utility energy program managers, and urban systems
researchers. Further research is needed to understand the usability
of the tool and evaluate the availability of such data to populate
the analysis. Further research is also needed to obtain feedback
from city stakeholder groups to organize the data outputs to
provide actionable information and salient feedback for urban
energy planning.
8. ACKNOWLEDGMENTS
The development of CityBES is funded by Lawrence Berkeley
National Laboratory through the Laboratory Directed Research
and Development (LDRD) Program. This work was supported by
the Assistant Secretary for Energy Efficiency and Renewable
Energy, the U.S. Department of Energy under Contract No. DE-
AC02-05CH11231.
9. REFERENCES
[1] Y. Zheng, L. Capra, O. Wolfson, and H. Yang, “Urban
Computing,” ACM Transactions on Intelligent Systems
and Technology, vol. 5, no. 3, pp. 1–55, 2014.
[2] International Energy Agency, “Transition to Sustainable
Buildings Strategies and Opportunities to 2050,” 2013.
[3] J. Keirstead, M. Jennings, and A. Sivakumar, “A review
of urban energy system models: Approaches, challenges
and opportunities,” Renewable and Sustainable Energy
Reviews, vol. 16, no. 6, pp. 3847–3866, Aug. 2012.
[4] C. F. Reinhart and C. C. Davila, “Urban Building Energy
Modeling – A Review of a Nascent Field,” Building and
Environment, vol. 2016, 2016.
[5] MIT Sustainable Deisgn Group, “Boston Citywide
Energy Model,” 2016. [Online]. Available:
http://web.mit.edu/sustainabledesignlab/projects/BostonE
nergyModel/. [Accessed: 15-May-2016].
[6] W. Emmanuel and K. Jérôme, “A verification of CitySim
results using the BESTEST and monitored consumption
values,” in Building Simulation Applications BSA 2015,
2015.
[7] C. F. Reinhart, T. Dogan, J. A. Jakubiec, T. Rakha, and
A. Sang, “Umi - an Urban Simulation Environment for
Building Energy Use , Daylighting and Walkability,” in
Proceedings of BS2013: 13th Conference of
International Building Performance Simulation
Association, 2013, pp. 476–483.
[8] DOE, “EnergyPlus,” 2016. [Online]. Available:
https://energyplus.net/. [Accessed: 18-May-2016].
[9] University of Strathclyde, “ESP-r,” 2016. [Online].
Available: http://www.esru.strath.ac.uk/Programs/ESP-
r.htm. [Accessed: 18-May-2016].
[10] D. Yan, J. Xia, W. Tang, F. Song, X. Zhang, and Y.
Jiang, “DeST — An Integrated Building Simulation
Toolkit Part Ⅰ : Fundamentals,” Build Simul, vol. 2008,
no. 1, pp. 95–110, 2008.
[11] J. J. Hirsch, “eQUEST the QUick Energy Simulation
Tool,” 2016. [Online]. Available:
http://www.doe2.com/equest/. [Accessed: 18-May-2016].
[12] CEN, CEN/TR 15615:2008 Explanation of the general
relationship between various CEN standards and the
Energy Performance of Buildings Directive (EPBD).
Umbrella document. 2008.
[13] ISO, “ISO 13790: 2008 Energy performance of buildings
—Calculation of energy use for spaceheating and
cooling.” 2008.
[14] S. H. Lee, F. Zhao, and G. Augenbroe, “The use of
normative calculation beyond building performance
rating systems,” in IBPSA, 2011, pp. 2753–2760.
[15] B. Poel, G. van Cruchten, and C. A. Balaras, “Energy
performance assessment of existing dwellings,” Energy
and Buildings, vol. 39, no. 4, pp. 393–403, Apr. 2007.
[16] W. J. Cole, K. M. Powell, E. T. Hale, and T. F. Edgar,
“Reduced-order residential home modeling for model
predictive control,” Energy and Buildings, vol. 74, no.
2014, pp. 69–77, 2014.
[17] Y. Heo, F. Zhao, S. H. Lee, Y. Sun, J. Kim, G.
Augenbroe, D. Graziano, L. B. Guzowski, and R. T.
Muehleisen, “Scalable Methodology for Energy
Efficiency Retrofit Decision Analysis,” in SimBuild,
2012.
[18] J. Granderson, S. Touzani, C. Custodio, M. D. Sohn, D.
Jump, and S. Fernandes, “Accuracy of automated
measurement and verification (M&V) techniques for
energy savings in commercial buildings,” Applied
Energy, vol. 173, pp. 296–308, 2016.
[19] F. Zhao, S. H. Lee, and G. Augenbroe, “Reconstructing
building stock to replicate energy consumption data,”
Energy and Buildings, vol. 117, no. 2016, pp. 301–312,
2015.
[20] T. Hong, M. A. Piette, Y. Chen, S. H. Lee, S. C. Taylor-
Lange, R. Zhang, K. Sun, and P. Price, “Commercial
Building Energy Saver: An energy retrofit analysis
toolkit,” Applied Energy, vol. 159, pp. 298–309, 2015.
[21] DOE, “OpenStudio,” 2016. [Online]. Available:
https://www.openstudio.net/. [Accessed: 18-May-2016].
[22] Open Geospatial Consortium, “OGC City Geography
Markup Language (CityGML) En-coding Standard,”
2012.
[23] G. Gröger and L. Plümer, “CityGML - Interoperable
semantic 3D city models,” ISPRS Journal of
Photogrammetry and Remote Sensing, vol. 71, pp. 12–
33, 2012.
[24] R. Laurini, “Visual Information Systems Chapter V:
Virtual 3D Cities.” 2015.
[25] Business Location Center, “Berlin 3D - Download
Portal,” 2016. [Online]. Available:
http://www.businesslocationcenter.de/en/downloadportal.
[Accessed: 17-May-2016].
[26] ASHRAE, ANSI/ASHRAE Standard 140-2011 Standard
method of Test for the Evaluation of Building Energy
Analysis Computer Programs. 2012.
[27] DOE, “Building Performance Database,” 2016. [Online].
Available: http://energy.gov/eere/buildings/building-
performance-database. [Accessed: 17-May-2016].
[28] Inc., Itron Inc. and JJ Hirsh & Associates Synergy
Consulting Quantum, “Database for Energy Efficiency
Resources (DEER) Updated Study. Final Report,” 2005.
[29] PNNL and PECI, “Advanced Energy Retrofit Guide,
Retail Buildings. Prepared for the US Department of
Energy PNNL- 20814,” 2011.
[30] PNNL and PECI, “Advanced Energy Retrofit Guide,
Office Buildings, Prepared for the US Department of
Energy, PNNL- 20761,” 2011.
[31] ASHRAE and DOE, “Advanced Energy Design Guide
for Small to Medium Office Buildings, Achieving 50%
Energy Savings Toward a Net Zero Energy Building,”
2014.
[32] DOE, “Weather Data,” 2016. [Online]. Available:
https://energyplus.net/weather. [Accessed: 16-May-
2016].
[33] ASHRAE, ANSI/ASHRAE/IES Standard 90.1-2013 --
Energy Standard for Buildings Except Low-Rise
Residential Buildings. Atlanta, GA: American Society of
Heating, Refrigerating and Air-Conditioning Engineers,
2013.
[34] California Energy Commission, 2016 Building Energy
Efficiency Standards For Residential and Nonresidential
Buildings Title 24, Part 6, And Associated Administrative
Regulations In Part 1. 2015.
[35] EPA, “ENERGY STAR,” http://www.energystar.gov/,
2015. [Online]. Available: http://www.energystar.gov/.
[Accessed: 07-Apr-2015].
[36] F. Fuerst, “Building momentum: an analysis of
investment trends in LEED and Energy-Star-certified
properties,” Journal of Retail and Leisure Property, vol.
8, no. 4, pp. 285–297, 2009.
[37] J. L. Mathieu, P. N. Price, S. Kiliccote, and M. A. Piette,
“Quantifying Changes in Building Electricity Use , With
Application to Demand Response,” IEEE
TRANSACTIONS ON SMART GRID, vol. 2, no. 3, pp.
507–518, 2011.
[38] T. Hong, L. Yang, D. Hill, and W. Feng, “Data and
Analytics to Inform Energy Retrofit of High Performance
Buildings,” Applied Energy, vol. 126, pp. 90–106, 2014.
[39] C. Li, T. Hong, and D. Yan, “An insight into actual
energy use and its drivers in high-performance
buildings,” Applied Energy, vol. 131, pp. 394–410, 2014.
[40] K. Sun, T. Hong, S. C. Taylor-Lange, and M. A. Piette,
“A pattern-based automated approach to building energy
model calibration,” Applied Energy, vol. 165, no. 2016,
pp. 214–224, 2016.
[41] J. Sanyal, J. New, R. E. Edwards, and L. Parker,
“Calibrating building energy models using
supercomputer trained machine learning agents,”
Concurrency Computation Practice and Experience, vol.
26, pp. 2122–2133, 2014.
[42] DOE, “Building Energy Asset Score,” 2016. [Online].
Available: http://energy.gov/eere/buildings/building-
energy-asset-score. [Accessed: 17-May-2016].
