Accepted Manuscript

Title: A system-level approach for designing multi-family

Sustainable and energy-efficient housing communities

Authors: Li Cheng, Yeonjin Bae, W. Travis Horton

PII: S2210-6707(17)31403-8
DOI: https://doi.org/10.1016/j.scs.2018.09.017
Reference: SCS 1255

To appear in:

Received date: 6-10-2017

Revised date: 7-9-2018
Accepted date: 14-9-2018

Please cite this article as: Cheng L, Bae Y, Horton WT, A system-level approach
for designing multi-family Sustainable and energy-efficient housing communities,
Sustainable Cities and Society (2018), https://doi.org/10.1016/j.scs.2018.09.017

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
A system-level approach for designing multi-family
Sustainable and energy-efficient housing communities
Li Cheng a, Yeonjin Bae b, W. Travis Horton c,*
a School of Mechanical Engineering, Purdue University, 585 Purdue Mall, West Lafayette, IN 47907, USA
b School of Civil Engineering, Purdue University, 550 W Stadium Ave, West Lafayette, IN 47907, USA
c School of Civil Engineering, Purdue University, 550 W Stadium Ave, West Lafayette, IN 47907, USA

PT
*
Corresponding author. E-mail address: [email protected]

Authors’ E-mail addresses:

RI
Li Cheng: [email protected] Yeonjin Bae: [email protected] W. Travis Horton: [email protected]

SC
U
Highlights


N
A system-level approach for designing multi-family sustainable housing is developed.
A
 Energy variables of building envelope, window, HVAC and DHW are investigated.
M

 The methodology here is user-friendly and flexible to be implemented.

D

 The implementation of the design will be carried out by a developer in Indiana, U.S.
TE

 The study is enriched by reliable and up-to-date industrial information in U.S.

EP
CC
A

i
Abstract

This study introduces a system-level methodology for designing multi-family sustainable housing communities. This

methodology is flexible and user-friendly, and can be implemented by both researchers and design practitioners. It

can be applied in assessment of various energy efficiency measures and energy delivery options that can be considered

PT
during the early stages of design and development. The overall goal of such a system-level process is to provide a

quick and efficient approach for identifying various design trade-offs early in the design process, which leads to an

RI
acceptable final design.

SC
The authors have utilized a sustainable housing community that is under development in Bloomington, IN, U.S. as the

case study to illustrate this methodology in the design process. Ten building variables that were deemed to have a

U
significant influence on energy consumption and/or construction cost were considered for potential inclusion in the

N
final design. During the process, a baseline building model was carefully defined according to applicable energy
A
standards, available reference building models in the U.S., and community and developer preferences. All energy
M

saving measures are considered relative to this baseline building model. The results of a sensitivity analysis show that

advanced energy delivery options (such as geothermal heat pumps, energy recovery ventilators and heat pump water
D

heaters) can significantly reduce the Energy Use Intensity (EUI) of the building as compared with adjustments to the
TE

envelope. In the final scenario, more than 45% of the annual baseline energy consumption is reduced from the initial

design.
EP

Keywords:
Sustainable housing development; system-level approach; case study of true application
CC
A
1 Introduction
In the United States, residential and commercial buildings consume 40% of the total primary energy and 73% of the

electricity [1]. There exist various opportunities to reduce primary energy use and CO 2 emissions towards energy

efficiency. In residential building sector people are becoming more and more interested in sustainable communities

which are planned, built, or modified to promote sustainable living, environmentally and economically. Therefore, as

PT
one step towards sustainability, energy-efficient housing communities are receiving extensive focus by building

designers, researchers and scientists with emerging methodologies and case studies.

RI
1.1 Previous studies

SC
Over the last few years, several studies have been carried out to investigate different energy efficiency measures,

energy delivery options, and building configurations. Liu et al. [2] studied the energy savings of energy recovery

U
ventilators on residential apartments with different enthalpy efficiency, fan power, and air change rates in different

cities in China. Kjellsson et al. [3] evaluated a hybrid system of solar collectors and ground source heat pumps as a
N
solution to space heating and domestic hot water in Swedish residential buildings. Georges et al [4] investigated the
A
performance of various heating systems, including electric boilers, stoves, and heat pump systems. Not only did they
M

compare individual systems, but also considered the combined performance of the heating systems together with the

building envelope. Furthermore, Aste et al. [5] simulated several heat pump options such as air source, water source,
D

ground source, and ground water heat pumps in Italy. However, Hachem et al. [6] considered electricity production
TE

by building integrated solar photovoltaic (PV) systems as a solution for sustainable buildings instead of Heating

Ventilation and Air Conditioning (HVAC) options. They studied the impacts of the number of floors and configuration
EP

on the ratio of electricity generation to total consumption in residential buildings.

Some research focused on advanced design of building envelopes, windows, and whole building architectural systems.
CC

R. Pacheco et al. [7] provided an overview of building design criteria that can reduce the energy demand for residential

buildings including building envelope system, glazing, shading, orientation, shape and passive heating and cooling
A

mechanisms. Granadeiro et al. [8] presented a new building efficiency indicator, the ERED (Envelope-Related Energy

Demand), indicating envelope-related surface areas, material U-values, Solar Heat Gain Coefficient (SHGC) of

windows, etc. Dodoo et al. [9] analyzed three multi-family buildings with cross laminated, beam-and-column and

modular prefabricated timber building systems, separately. Echenagucia et al. [10] investigated the influence of

1
various window choices on building performance in several locations around Europe. They suggested a preference for

choosing larger south-facing windows together with lower Window-to-Wall Ratios (WWR) on north, west, and east

faces. B. Andersson et al. [11] investigated heating and cooling loads for a prototype residential building at different

orientations in U.S. They provided sensitivity of results to different window design and thermal mass. Additionally,

J. Morrissey et al. [12] assessed the implications of orientation on residential building, modeled and experimented

PT
with two different building energy efficiency standards. Takano et al. [13] considered the energy performance of

current typical Finnish residential buildings with four different building types (detached house, row house, town house

RI
and apartment block). They claimed that life cycle energy efficiency of a building increased as the number of floors

and floor area increased, while the structural framing material had a relatively minor effect. From the studies above,

SC
there are five major categories of energy efficiency measures or alternative technologies [14] that can be applied to

sustainable housing developments. These include: 1) advanced design integration of the building as a whole, 2)

U
enhanced envelope design including windows and shading, 3) improved HVAC systems, 4) efficient domestic hot

N
water (DHW) systems, and 5) inclusion of energy efficient appliances and lighting.
A
In early stages of community development, designers and architects could do better to evaluate impacts of the various
M

building energy and architectural features mentioned above with help of building modeling by computer. S.

Bambardekar et al. [15] investigated ways to better inform architects and designers of building energy modeling and
D

simulation. For the whole design process, S. Macmillan et al. [16] developed a framework which included 12 activities
TE

in five phases to help decision making.

EP

Various design targets are evaluated including economic, environmental or social impacts of building during the

assessment of building technologies and scenarios. Building researchers typically look at life cycle cost, net present
CC

value, primary energy consumption, CO2 emissions, and thermal comfort as the performance measures of residential

buildings. Meanwhile, various methodologies and analysis approaches are utilized to evaluate design targets.

Sadeghifam et al. [17] investigated the energy demand of tropical residential buildings by statistical analysis to
A

conclude that the most significant envelope factor for a building in Malaysia was the ceiling design. Yu et al. [18]

used a genetic algorithm, with back-propagation neural network, to perform a multi-objective optimization study.

Zaca et al. [19] implemented design scenario analysis to evaluate primary energy consumption and life cycle cost of

new buildings around the Mediterranean Fesanghary et al. [20] conducted a multi-objective optimization based on a

2
harmony search algorithm to minimize the life cycle cost and life cycle CO 2 emissions of a single family residential

building in Baton Rouge, LA, U.S.

Four methodologies are mentioned above: 1) sensitivity analysis, 2) scenario analysis, 3) statistical analysis, and 4)

optimization (single objective and multi-objective), during which, sensitivity analysis has drawn the attention of many

building scientists since they are simple to perform and don’t require mathematical optimization techniques to

PT
efficiently search the large design space of construction options. This approach can be used to identify how individual

key building energy variables can affect building performance. As Tian [21] has summarized, a general sensitivity

RI
analysis process for building performance involves the following steps: 1) input variations, 2) create building energy

SC
models, 3) run building energy models, 4) collect simulation results, 5) run sensitivity analysis, and 6) presentation of

sensitivity results. Lam et al. [22] introduced sensitivity coefficients describing the response of building energy outputs

U
to inputs. Following case studies were performed. An energy-efficient municipal building in Portugal was developed

N
by Tavares and Martins [23] by applying sensitivity analysis regarding energy variables such as wall structure and

materials, window frames, and HVAC systems. Wang et al. [24] ranked retrofit options for representative Swedish
A
residential buildings. Hemsath et al. [25] evaluated building design choice with local and global sensitivity. In this
M

case, global sensitivity is a measure of how sensitive the impacts of particular design elements on building

performance when coupled with all the other elements.

D
TE

After the literature review above, following issues are identified:

 Many researchers only consider different energy efficiency measures or building energy variables
EP

individually, and don’t always consider the interactive effects of these adjustments which result in

underestimates or overestimates synergies.

CC

 Energy performance of multi-family buildings is not investigated thoroughly to include interactive effects

among these dwellings through demising walls, corridors and stairways, which may alter the building
A

performance to a large extent.

 Optimization, mathematical programing, and statistical analysis with specific search algorithms tend to

require considerable computational efforts and significant mathematical knowledge. Cost function for every

project may differ due to various construction crew experience and proficiency for a given technology, which

3
 may lead optimization results into different directions. Consequently, the current methodologies are not

efficient and flexible enough to be implemented at a large scale.

 Although research data in the literature mostly come from real data collected from reference buildings or

specific building types, almost no implementation in practice has been carried out. Intensive communication

and cooperation between building researchers (academia) and commercial developers (industries and market)

PT
is generally not found in the literature.

RI
1.2 Aims and objective
This study aims at introducing a systematic approach for designing multi-family sustainable and energy-efficient

SC
housing community along with a case study for the application. The methodology that is presented should be user-

friendly, modular and flexible in its implementation. In the case study, the authors assess the benefits and trade-offs

U
of various energy efficiency measures and energy delivery options in the early stages of engineering design and

development. Cooperating with a local building developer, the authors have designed a multi-family residential
N
building in Bloomington, IN as a case study in the Midwest, U.S. The methodology, together with reliable and up-to-
A
date industrial information can be used to support future developers in the U.S. in designing sustainable residential
M

buildings in the early stages of development.

D

2 Methodology
A system-level approach to designing a multi-family sustainable housing community is described by the following
TE

steps from section 2.1 to 2.4 and evaluated in section 2.5.

EP

2.1 Establishing a baseline building model

The first step is to establish a baseline building model. As is shown in Figure 1, designers can identify a schematic
CC

building plan and combine that information with standardized building codes and community preferences to establish

a baseline, which incorporates the primary variables that contribute strongly to annual energy consumption. The
A

baseline energy model involves detailed building geometry information that comes from schematic building plans

provided by the developer and architect. The whole building is then partitioned into thermal zones into which are

applied the appropriate thermal gains, occupancy characteristics, thermal mass, envelope choices and energy supply

options.

4
Designers can evaluate the energy performance of the baseline after its establishment. If designers aim at conducting

a life cycle cost analysis, then the analysis of the baseline should include both upfront cost (material cost, labor cost,

overhead and profit) and long-term cost (annual utility cost and replacement cost over the life cycle of the building).

If designers are interested in other life cycle environmental impacts then a product-level relevant conversion and

calculation should be included. In the case study of this paper, EnergyPlus 8.4 [26] was utilized to simulate the

PT
comprehensive building energy model and evaluate the total energy consumption and the Energy Use Intensity (EUI)

of the baseline building design.

RI
SC
2.2 Identifying building energy variables and technologies
The next step is to identify all the major building energy variables that will be parametrically studied for their impact

U
on energy consumption. The schematic is shown in Figure 2. Decisions that identify which variables will or will not

N
be considered in a parametric study are often made based on a developer’s familiarity, experience, and comfort with
A
the various available technology options. One benefit of the proposed methodology is that it takes these factors into
M
account during both the brainstorming stage as well as the expert analysis (introduced in section 2.4) so that the final

design choices reflect a developer’s true capabilities. Specifically, in the case study of this paper, alternative lighting
D

technologies were investigated, so they were defined as a significant energy variable; while a decision may be made,

for example, relative to infiltration, to have it remain fixed in the baseline model and all subsequent permutations.
TE

At this point, designers and the developer begin brainstorming to identify an extensive list of potential alternative
EP

technologies for each energy variable. In the current case study, all of the energy efficiency measures and energy

delivery options that are considered come from real industrial and market-available choices in the Midwest, U.S.
CC

2.3 One-at-a-time building technology sensitivity analysis

A

After identifying all of the appropriate technologies, the impact of each technology is simulated individually as a

permutation in the energy variable of the baseline model to determine the sensitivity of building energy consumption

to each technology. The reduction in Energy Use Intensity (EUI) of the building is compared to the baseline model to

indicate each technology’s impact. After simulating the building performance by changing each variable individually,

5
all of the sensitivity results are compared together to identify the best alternative technology for each energy variable.

At this point, all of the best alternative technologies will be combined to suggest an energy-efficient design scenario.

2.4 Expert analysis, 2nd stage sensitivity analysis (optional), and design

PT
finalization
The final step in the process is for designers and developer to conduct an expert analysis to determine a pre-final

RI
design based on the most energy-efficient scenario that comes from the one-at-a-time parametric study, coupled with

the experiences and preferences of developer. When the energy performance of the pre-final design is accepted, the

SC
building design will be finalized. The corresponding flowchart is shown in Figure 4.

Combined with the expert analysis, a 2nd stage one-at-a-time sensitivity analysis is optional to be carried out. With the

U
involvement of the 2nd stage analysis, the step of generating a pre-final design will be more systematic. In the 2nd stage
N
of analysis, the most energy-efficient model will be regarded as the updated baseline and impacts of the previously
A
defined technologies will be updated. More affordable technologies will be permuted into corresponding energy
M

variables to generate the pre-final design with the help of expert analysis.

This process leads to several potential building designs based on a combination of energy modeling results and expert
D

analysis to ultimately arrive at a solution that is sufficiently affordable from the standpoint of the developer, while
TE

also further reducing energy consumption relative to a code-minimum building. This process continues through

various iterations until a final design is achieved.

EP

2.5 Evaluation of the methodology

CC

Effective methodologies involving sensitivity analyses have been proposed in several previous studies. One of the

merits of the proposed methodology compared to those contained in previous literature studies is the presentation of
A

an overall framework that can provide a more methodical justification for each decision made, especially in the

detailed design process. The baseline building model, which plays a significant role in the sensitivity analysis, is

developed and the significant variables that will be used for building energy simulation permutations are identified

during the baseline analysis. By employing the results obtained from a “one-at-a-time” sensitivity analysis, designers

and developers can see the trade-offs between various energy efficiency measures and can incorporate subjective

6
opinions such as designers’ preference and experience in the expert analysis. This process continues iteratively until

the final goal of the design is achieved.

The other merit is that the overall methodology is versatile and flexible in different applications. This approach can

be combined with any objective function such as life-cycle analysis, provided that the relevant embodied energy and

other environmental impacts or costs at the construction and demolition phase are given. The same methodology can

PT
be applied to other countries, other climate zones, other standard building codes and other building types i.e. a single-

family housing unit and/or commercial buildings by different building designers and developers to explore the various

RI
benefits. The parameters that are not presented in this study but can be identified during the baseline analysis such as

SC
building geometry and orientation related parameters, can also be considered in the sensitivity analysis. Designers can

still follow this approach (baseline analysisvariable identificationsensitivity analysisexpert analysis) to

U
accelerate the design process effectively, to acquire a near-optimal design.

N
Although in this approach, designers evaluate the sensitivity of building technology one-at-a-time, the coupling effects
A
of several technologies can be interpreted in expert analysis during the evaluation of several overall building designs.
M

When coupled together, the energy performance of an overall design seems to be slightly more moderate than the

predicted performance by summation of all involved individual technologies. However, the one-at-a-time sensitivity
D

analysis can still give the design guidance in the ballpark. This system-level process is not a strict optimization process
TE

because it cannot perfectly and exactly locate the optimum design. However, this process can begin to approach the

optimum design with less efforts than a full optimization process during the conceptual design phase. Additionally,
EP

performing a true optimization process with a combination of continuous and discrete variables can be time-consuming

and challenging for many practitioners in a conceptual design phase, while the method proposed in this paper only
CC

requires sufficient knowledge and expertise to perform annual energy simulations using software. In addition, the final

design that is chosen will be more acceptable to the developer and community since their input is used during the

expert analysis to eliminate impractical design choices.

A

3 A case study of sustainable community development in

Bloomington, IN
In this section, a case study is presented to illustrate the system-level design approach that has been described above.

The methodology is clearly employed in which a baseline model is developed, energy variables are selected, a

7
parametric analysis is performed to determine the sensitivity of all the energy variables, and potential design scenarios

of a sustainable community are analyzed.

3.1 Overview of the building

3.1.1 Building plan
The authors teamed up with a developer who plan to build a multi-family sustainable housing community in

PT
Bloomington, IN. The architect’s conceptual plan from a south-facing viewpoint is shown in Figure 5.

RI
The building has four floors, and each floor has one corridor that provides passage to all of the living spaces. There is

one stair tower on the west end and one on the east end of the building connecting each corridor from the 1 st floor to

SC
the 4th floor. The 2nd and 3rd floors have residential units on both the north and south sides of the corridor, while the

1st floor only has units on south side and 4th floor only has units on north side of the corridor. This occurs because the

U
building will be constructed on a gentle slope and it needs a slanted south-facing roof above the 4th floor to collect
N
solar energy. There is storefront glazing on each floor and curtain wall glazing for a community center on the 1st floor.
A
The floor plan for the 1st floor is shown in Figure 6, in which there is one community center, one business center, four
M

public toilets and one fitness center.

D

3.1.2 Building geometry and zone configuration

TE

Building geometry and zone configuration is necessary for energy modeling purposes. Building geometry is simplified

from the building plan and different thermal zones are specified based on the floor plan. These simplifications and
EP

specifications, which are based on reasonable assumptions, will lead to quicker convergence for the building

simulations. Therefore, the authors make the following assumptions:

CC

 Each residential unit is treated as a single zone that is controlled by a thermostat located in the living room.


A

The detailed internal structure (walls, partitions, etc.) of each zone is ignored but the internal thermal mass

is accounted for.

 The building’s mechanical rooms are merged in with the corridors. Balconies are merged in with the

residential units. The fitness center, business center, and public toilets are merged into the community center.

8
 These design simplifications will not change the overall energy performance of the building significantly,

but the doors to the balconies will have a much larger effect on the zone heat balance, so they will be reserved

in the simplified building geometry.

 Each corridor is split into 3 rectangular zones which speeds up convergence in the simulation as compared

to solving a single concave-shaped corridor.

PT
 For each zone, all windows facing in a common direction are merged into one single window with the same

RI
surface area as the sum of the originals.

SC
Figure 7 shows the simplified building geometry which can be compared with the conceptual building plan in Figure

5; while Figure 8 shows the zone configuration for the 1st floor which can be compared with the corresponding floor

U
plan in Figure 6.
N
A
The zone configurations for this particular building geometry are summarized in Table 1. There is a total of
13 1-bedroom units, 17 2-bedroom units, and 6 3-bedroom units. The total conditioned square footage of the
building is 56,266 ft2 (5227m2).
M

3.1.3 Internal gains, ventilation and infiltration

D

Internal heat gains in the building include occupancy, electric equipment and appliances, lighting etc. The occupancy

schedule in each zone is defined relative to the peak occupancy rate shown in Table 2, and a daily percentage schedule.
TE

The daily schedules for occupancy, lighting, and electric equipment are shown in
EP
CC
A

9
Figure 9. All of these schedules are taken from the multifamily prototypical reference building models from the Pacific

Northwest National Laboratory (PNNL) [28].

The peak occupancy of community center in the 1st floor is defined to be 7 since the zone is composed of a business

center, fitness and 4 toilets. The peak occupancy rates of 1.5, 2.5, or 3.5 in a 1-bedroom, 2-bedroom, or 3-bedroom

PT
unit zone, respectively, simply represents an average occupancy of each living unit type in this building, since there

may be slightly different numbers of children and adults in each unit.

RI
The peak electric equipment power density in the residential zones and in the community center is defined to be 0.62

SC
W/ft2 (6.67 W/m2), which is also taken from the PNNL reference models [21]. This electric equipment power density

includes daily usage of a cooking range, plug loads, and a clothes washer and dryer. There is also an elevator in the

U
building, which has its motor located on the 1st floor with a peak power consumption rate of 1,314Watts.

N
According to the requirements of ASHRAE (American Society of Heating, Refrigerating and Air-conditioning
A
Engineers) Standard 62.1 2007 [29], the outdoor air ventilation rate and exhaust air requirements for each zone are
M

configured as follows:

 5 cfm (2.36L/s) per occupant and an additional 0.06 cfm per ft2 (0.30L/s-m2) of outdoor air is introduced to
D

residential zones and the community center zone as mechanical ventilation. This ventilation air is brought in
TE

through each air handler whenever the air handler is operating. Therefore, this ventilation air also cycles on

and off due to the thermostat control.

EP

 50 cfm (23.6L/s) continuous exhaust air ventilation in each residential zone accounts for kitchen exhaust air.
CC

Blower door test results that this particular developer can typically achieve on a project like this is an average outdoor

air infiltration rate of 3.5 ACH (Air Changes per Hour) under a differential pressure of 50Pa. This can be converted
A

to an Effective Leakage Area (ELA) of 1,845 in2 (1.19m2) for the whole building and results in 0.22 ACH with a

typical pressure differential of 4Pa. The operation of doors and windows is also taken into consideration so the average

total infiltration rate for the whole building is set to 0.45 ACH. Tavares and Martins [23] used an infiltration rate of

0.5 ACH in their research to design a municipal building.

10
3.2 Baseline model development and major energy variables
3.2.1 Baseline model configuration
The baseline model is configured according to ASHRAE Standard 90.1 2007 [30] together with preferences of the

community and developer. In this case, for example, the developer prefers to use an insulation value of R18 in the

exterior walls of the baseline model instead of the R13 wall that is required by ASHRAE Standard 90.1 2007. The

detailed input information to the baseline model is summarized in Table 3.

PT
RI
Lighting power density (LPD) in each zone is shown in Table 3. The LPD of compact fluorescent (CFL) and LED

lights are modified from the LPD of incandescent lighting while keeping the lumens constant. The daily lighting

SC
schedule is shown in

U
N
A
M
D
TE
EP

Figure 9.

The baseline exterior walls and roof are constructed in layers according to the assemblies shown in Table 4.
CC

In the baseline model an air source heat pump is chosen as the primary cooling and heating equipment. The indoor air
A

handler is equipped with a supplementary 5kW electric heater and a 0.5 inH2O (124.5 Pa) pressure rise fan with 565

cfm for 1-bedroom units and 800 cfm for 2,3-bedroom units. The authors in this study have processed heating and

cooling performance data from the heat pump and generated corresponding performance curves. The performance

curves have been inserted into the heat pump portion of the building model to capture detailed HVAC annual

11
performance characteristics. The heating performance curves are shown in Figure 10. The horizontal axis represents

outdoor air temperature in Fahrenheit and the vertical axis represents normalized heating performance of the heat

pump. Equations 1, 2 and 3 define the normalized heating performance characteristics (normalized capacity, power

input, and COP, respectively).

PT
𝑎𝑐𝑡𝑢𝑎𝑙 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦
𝑛𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 = (Equation 1)
𝑟𝑎𝑡𝑖𝑛𝑔 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦

RI
𝑎𝑐𝑡𝑢𝑎𝑙 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛𝑝𝑢𝑡
𝑛𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛𝑝𝑢𝑡 = (Equation 2)

SC
𝑟𝑎𝑡𝑖𝑛𝑔 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛𝑝𝑢𝑡

𝑛𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦
𝐶𝑂𝑃 = × 𝐶𝑂𝑃𝑟𝑎𝑡𝑖𝑛𝑔 (Equation 3)
𝑛𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛𝑝𝑢𝑡

U
𝑟𝑎𝑡𝑖𝑛𝑔 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦
𝐶𝑂𝑃𝑟𝑎𝑡𝑖𝑛𝑔 =
N
𝑟𝑎𝑡𝑖𝑛𝑔 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛𝑝𝑢𝑡
(Equation 4)
A
When the actual COP of the heat pump is smaller than 1, it doesn’t make sense to allow the heat pump to provide heat
M

to the zone. Based on the curve, an outdoor air temperature of 13F is used as the cutoff below which the heat pump

is locked out and the supplemental heat source is used.

D
TE

3.2.2 Baseline model results

In addition to the baseline building energy model, the authors have also put together an ASHRAE standard 90.1 2007

[30] building energy model as a further comparison to the baseline model. The ASHRAE standard model is developed
EP

based on the same building geometry, but with major differences as shown in Table 5.
CC

In addition, RECS [31] data from the Midwest region has been collected and used as a means of roughly validating

the baseline model just to ensure that no major errors exist in the baseline energy model. A comparison of results
A

between the baseline model and RECS annual energy consumption data is shown in Figure 11. Here the category of

Others includes the energy consumption of lighting, fans, appliances, plug loads, etc.

The baseline model can be validated in the following aspects:

12
  Total energy consumption and end use energy consumption of the baseline model is consistent with those of

ASHRAE Standard 90.1 2007 model, although the baseline model is slightly more energy-efficient, which

would be anticipated.

 The baseline model consumes less heating energy, but more energy in the Cooling and Others category than

the RECS data. This is also consistent with expectations since the RECS survey is based on older existing

PT
residential buildings in the Midwest region. Most of the current buildings are not as air tight and well

RI
insulated as a newly constructed code-minimum building, which will result in greater heating loads and less

cooling loads. Also, the baseline model is located at Bloomington, IN where the weather is generally warmer

SC
as compared with most of the locations in Midwest, so the small discrepancies in heating and cooling energy

consumption are considered acceptable. In addition, the energy consumption in the category of Others in the

U
baseline model primarily come from electric equipment consumption, which is determined by the electric

N
equipment density and schedule provided by the PNNL reference building model [28], which was not altered

by the authors for the current study.

A
M

Based on these observations a decision was made that the baseline energy model is sufficiently accurate to be used as

the primary point of comparison in this study.

D

3.2.3 Significant energy variables

TE

The significant energy variables are those building design variables that will most strongly influence building energy

consumption when they are changed. For purposes of this study, building envelope related variables, lighting, HVAC,
EP

and DHW (Domestic Hot Water) related variables are those that are considered in the parametric study. The 10 major

energy variables are listed in Table 6.

CC
A

3.3 Single energy variable study

Starting from the baseline model, the authors and developer brainstormed alternative technologies for each energy

variable with the goal of finding the best combination of all these alternatives that also maximizes energy-efficiency.

The parametric study approach that was adopted is to change the value of each energy variable one-at-a-time while

keeping all other variables fixed at the baseline configuration. This approach allows researchers to gain an

13
understanding of the energy consumption sensitivity of each energy variable. Finally, the most efficient variable from

each category is selected to represent the building with the highest energy efficient scenario. The single variable

parametric study approach was selected because full optimization with specific search algorithms tends to require

considerable computational effort and significant mathematical knowledge. Some life cycle analysis methods may

require information, or are based on assumptions that are not always available at the conceptual design phase of a

PT
building.

However, this one-at-a-time parametric study approach requires less search effort when using discrete energy

RI
variables. Additionally, it is a more user-friendly and flexible to approach to identify a reasonable and acceptable

SC
design.

U
3.3.1 Potential alternative technologies
N
Table 7 shows a list of the available alternative technologies for each energy variable that are also considered
A
acceptable by the developer.
M

In the first two choices for the exterior wall, staggered stud layers are chosen to replace the original 2x6 framing layer
D

in the baseline model. Additional exterior wall constructions that were considered include, SIPs (Structural Insulation

Panels) with different R-values. SIPs exterior walls are also assumed to reduce ELA due to their enhancement in wall-
TE

to-wall joint tightness and ceiling-to-wall joint tightness. The ELA of wall-to-wall and ceiling-to-wall joints for

typical construction and tight construction are assumed to be 0.07 in2/ft (1.48cm2/m) and 0.02 in2/ft (0.42cm2/m),
EP

respectively [32]. According to Sherman and Grimsrud’s model [33], the infiltration rate is calculated on an hourly

basis using Typical Meteorological Year (TMY) data. Thus, a decrease in infiltration due to tightness of SIPs
CC

construction is 0.02 ACH.

A

The HVAC system energy variables include a higher efficiency air-source heat pump, or the use of a geothermal heat

pump instead. Since geothermal heat pumps use the ground as their heat source they do not experience a sharp drop

in performance as the outdoor air temperature drops similar to air-source heat pumps. Also, the use of an energy

recovery ventilator is considered as an alternative technology to harvest energy from the exhaust air. Finally, heat

14
pump water heaters are considered as an alternative to standard electric water heaters for generating domestic hot

water.

3.3.2 Single energy variable study: part I

After simulating each energy variable individually among the alternatives based on the baseline model, the EUI results

were collected and are shown in Table 8, with exception of the HVAC and DHW related variables.

PT
The results of percentage change in EUI are better shown in the tornado chart of Figure 12. The name of each

RI
horizontal bar is consistent with the alternative technology listed in Table 7. The higher the alternative technology is

SC
on the tornado chart, the greater EUI reduction it gives relative to the baseline model. The bars in Figure 12 are also

colored to imply which energy variable category the specific alternative technology belongs with. For example, the

U
blue bars represent a family of exterior wall alternative technologies. In addition, the bars with angled lines represent

the selections for the pre-final design scenario for the building after expert analysis was completed by the developer.
N
A
M

3.3.3 Single energy variable study: part II

Similar to the study above, researchers also varied the HVAC and DHW energy variables one-at-a-time and the
D

sensitivity of each is shown in

TE
EP
CC
A

15
Table 9. The related description of these alternative technologies is given in Table 7.

In

PT
RI
SC
U
N
A
M
D
TE
EP
CC
A

16
Table 9, when applying geothermal heat pump technology both residential spaces and public spaces are upgraded

together since it makes little sense to use a combination of air source and ground source heat pumps while the boreholes

for ground heat exchanger exist. The results showing percentage change in EUI for these variables are shown in Figure

13.

PT
Figure 13 clearly shows that the geothermal heat pump provides the highest energy efficiency among all of the

variables studied in this paper. Additionally, the use of heat pump water heaters shows a significant positive impact

RI
on energy consumption.

SC
3.4 Scenario comparison: proceeding to final design
3.4.1 Expert analysis
After the single variable parametric study is complete the next step is to select the most energy-efficient alternative

U
technologies within each energy variable category as a potential design scenario. Based on this scenario, researchers
N
and the developer can conduct an expert analysis and adjust some energy variables to propose a pre-final design
A
scenario. The criteria of the expert analysis include community preferences, proficiency of the developer and
M

construction team in implementing the selected technologies, cost considerations, maintenance concerns, etc. If a

proposed technology is not deemed practical for any reason then the designers can simply choose from the following
D

technologies according to their rank in the energy variable parametric study. After the expert analysis of the current
TE

study, a pre-final design scenario was selected and is compared with the most energy-efficient scenario in Table 10.
EP

The alternative technologies for the pre-final design scenario that were changed from the most energy-efficient

scenario, have been colored in blue in Table 10. For this project, SIPs construction was not chosen after the expert
CC

analysis due to a lack of familiarity from the construction crew in working with SIPs panels. Also, R20 vertical

insulation is much more expensive than R15 insulation and the developer felt that it doesn’t show a high enough
A

energy consumption reduction to justify the additional cost. Storefront glazing with shading is chosen to accommodate

the placement of photovoltaic arrays on the shading structures for electricity production.

3.4.2 Energy results for the two scenarios

Both of the preceding scenarios were simulated and the EUI results are summarized in Table 11. Both scenarios save

over 45% of the annual energy consumption as compared to the baseline model; and it is noted that both of the

17
scenarios do not differ significantly from one another in EUI, which means that the decision to move away from SIPs

construction to accommodate the preferences of the construction crew does not have a large impact on overall building

energy efficiency.

The coupling effects of several building technologies can also be interpreted from energy results of these two design

PT
scenarios as is shown in Table 11. For both pre-final scenario and most energy-efficient scenario, the summation of

EUI change of all technologies from baseline in one-at-a-time study exaggerates the energy efficiency of the involved

RI
technologies compared to the results when they are coupled together in energy modeling. However, the prediction of

a simple summation is still in the ballpark by giving tolerable differences.

SC
U
The final results from this analysis, then were used to finalize the design scenario for moving forward. If for some

reason the final EUI was not acceptable then researchers could begin the analysis again using the most energy-efficient
N
scenario as the baseline building and carry out the parametric study and expert analysis again.
A
A detailed comparison of building energy end uses for the baseline building is shown in Figure 14. The figure shows
M

the baseline energy, best-case scenario, and the final design scenario that was selected through expert analysis. Both
D

of the design scenarios show heating energy savings of more than 70%, along with over 55% cooling energy savings,

and more than a 70% reduction in domestic hot water energy consumption over the baseline model.
TE
EP

3.5 Comparison of computational efforts

One of the strength of the systematic approach is being capable to finalize building design through a quick and efficient
CC

manner. This efficiency results from not only the systematic and modular structure of the methodology but also the

incorporation of community’s and developers’ preference through the steps of baseline establishment, energy variable
A

identification, alternative technology selection and expert analysis.

To illustrate this efficiency, computational efforts for two design methods applied to this case study are evaluated and

compared in Table 12 by counting the number of whole building simulation throughout the design process. Two

different design methods are compared as follow.

18
First method evaluates all possible designs coming from the design space which is composed of 10 design variables

and corresponding 22 alternative technologies. The system-level approach suggested by this paper, requires building

simulation of building baseline model, energy-efficient building model, final building model and models which have

one technology permuted from the baseline. Time duration of the simulations is estimated in Table 12 under the

PT
assumption of 5 min per simulation.

RI
The computational effort of a full optimization, which is not shown in this paper, is difficult to estimate because a

total required number of simulation is highly dependent on the cost function and the optimization methodology;

SC
however, at least a few hundreds of simulations are required for the full optimization.

U
The efforts of the design process can only be partly reflected by evaluation of computational efforts. For example, a

N
full optimization requires relevant knowledge such as mathematical and computer programming background.

However, the system-level approach introduced in this paper show an advantage in computational efforts over the
A
method of evaluating the whole design space. Compared to optimization, it is more flexible and simple to setup the
M

solution procedure before computation.

D

4 Conclusion and future steps

TE

4.1 Conclusion and contributions

This study has introduced a standardized and efficient system-level approach for designing a sustainable housing

community along with a case study for true application during which, various energy efficiency measures and energy
EP

delivery options are assessed in the early stages of the design and development. The methodology is flexible, modular
CC

and user-friendly. It can be implemented through the collaboration of designers, researchers and developers.

This methodology can approach the optimum design with less efforts than a full optimization process. Also, it is
A

friendly since it requires only sufficient knowledge and expertise to perform building modeling and energy simulation.

This approach is effective due to its integration of developers’ and community’s preference to eliminate impractical

design choices.

19
In addition, designers can proceed with or without embodied energy or environmental footprint when they are

expert analysis). The

design process will be accelerated in either way and a near-optimum design will be acquired.

For the case study, the authors have cooperated with a developer to determine an energy-efficient final design of a

sustainable housing community in Bloomington, IN, U.S. Ten building energy variables were considered during the

PT
evaluation process as the team worked to identify the final design.

RI
A baseline building energy model was carefully defined according to ASHRAE standards, reference building models

in the U.S., and community’s and developer’s preference. The baseline model was shown to give results that are

SC
consistent with the RECS data in the Midwest and a typical ASHRAE standardized building model, which provides

confidence in the definition of all the model parameters. The baseline model includes a detailed description of the

U
building geometry, thermal zoning, typical envelope, HVAC configuration, ventilation, infiltration, internal gains and

their schedules. N
A
After the definition of the baseline building model, the results of a parametric study show that advanced exterior wall
M

construction options are more energy-efficient than those typically used in multi-family buildings. In addition, for this

particular building, advanced energy delivery options such as geothermal heat pumps, energy recovery ventilators and
D

heat pump water heaters can reduce building EUI more significantly than adjustments on the envelope. From a
TE

comparison of the final design scenario with the baseline model, the authors found annual energy consumption savings

of more than 45%.

EP

As a case study, except that it illustrates the procedure of implementing the system-level approach, this work also sets
CC

a good example for modeling a multi-family residential building which can be referenced for future work. Specifically,

future designers and researchers can benefit from the reasonable configuration of internal thermal gains, assumptions

for simplification of building geometry and thermal zones, approach for validating the baseline. Also, practical
A

information is provided by this case study due to its intensive integration with industry in Midwest, U.S.

4.2 Future steps

In the step of expert analysis (see section 2.4), authors have mentioned that a 2nd stage one-at-a-time sensitivity analysis

can be performed using the best-case scenario as the new baseline building if desired. Therefore, it would be beneficial

20
to see how this 2nd stage analysis would affect the expert analysis and final design in a detailed case study. Also, these

new results will better reflect the synergy that will be encountered by changing multiple construction variables at the

same time.

In future, designers and researchers can also implement this approach with life cycle analysis in the design process.

Additionally, this methodology can be applied to design of single family housing and commercial buildings to explore

PT
the benefits.

RI
5 Acknowledgments
The authors would like to thank the Indiana Housing and Community Development Authority for authorizing the

SC
development of the community. Additionally, the authors would like to thank Pedcor companies for their intensive

communication and cooperation with the authors during this process. In the end, the authors would like to thank Center

U
for High Performance Building, Purdue University for their one-year-long sponsorship on this study.

N
A
M
D
TE
EP
CC
A

21
References
[1] U.S. Department of Energy, Buildings energy databook, Energy Effic. Renew. Energy Dep. (2012) 286.

[2] J. Liu, W. Li, J. Liu, B. Wang, Efficiency of energy recovery ventilator with various weathers and its energy
saving performance in a residential apartment, Energy Build. 42 (2010) 43–49.
doi:10.1016/j.enbuild.2009.07.009.

[3] E. Kjellsson, G. Hellström, B. Perers, Optimization of systems with the combination of ground-source heat
pump and solar collectors in dwellings, Energy. 35 (2010) 2667–2673. doi:10.1016/j.energy.2009.04.011.

PT
[4] L. Georges, C. Massart, G. Van Moeseke, A. De Herde, Environmental and economic performance of
heating systems for energy-efficient dwellings: Case of passive and low-energy single-family houses,

RI
Energy Policy. 40 (2012) 452–464. doi:10.1016/j.enpol.2011.10.037.

[5] N. Aste, R.S. Adhikari, M. Manfren, Cost optimal analysis of heat pump technology adoption in residential
reference buildings, Renew. Energy. 60 (2013) 615–624. doi:10.1016/j.renene.2013.06.013.

SC
[6] C. Hachem, A. Athienitis, P. Fazio, Energy performance enhancement in multistory residential buildings,
Appl. Energy. 116 (2014) 9–19. doi:10.1016/j.apenergy.2013.11.018.

U
[7] R. Pacheco, J. Ordóñez, G. Martínez, Energy efficient design of building: A review, Renew. Sustain. Energy
Rev. 16 (2012) 3559–3573. doi:10.1016/j.rser.2012.03.045.

[8]
N
V. Granadeiro, J.R. Correia, V.M.S. Leal, J.P. Duarte, Envelope-related energy demand: A design indicator
A
of energy performance for residential buildings in early design stages, Energy Build. 61 (2013) 215–223.
doi:10.1016/j.enbuild.2013.02.018.
M

[9] A. Dodoo, L. Gustavsson, R. Sathre, Lifecycle primary energy analysis of low-energy timber building
systems for multi-storey residential buildings, Energy Build. 81 (2014) 84–97.
doi:10.1016/j.enbuild.2014.06.003.
D

[10] T. Méndez Echenagucia, A. Capozzoli, Y. Cascone, M. Sassone, The early design stage of a building
envelope: Multi-objective search through heating, cooling and lighting energy performance analysis, Appl.
TE

Energy. 154 (2015) 577–591. doi:10.1016/j.apenergy.2015.04.090.

[11] B. Andersson, The impact of building orientation on residential heating and cooling, 8 (1985) 205–224.
EP

[12] J. Morrissey, T. Moore, R.E. Horne, Affordable passive solar design in a temperate climate: An experiment
in residential building orientation, Renew. Energy. 36 (2011) 568–577. doi:10.1016/j.renene.2010.08.013.

[13] A. Takano, S.K. Pal, M. Kuittinen, K. Alanne, Life cycle energy balance of residential buildings: A case
CC

study on hypothetical building models in Finland, Energy Build. 105 (2015) 154–164.
doi:10.1016/j.enbuild.2015.07.060.

[14] L. De Boeck, S. Verbeke, A. Audenaert, L. De Mesmaeker, Improving the energy performance of residential
A

buildings: A literature review, Renew. Sustain. Energy Rev. 52 (2015) 960–975.

doi:10.1016/j.rser.2015.07.037.

[15] S. Bambardekar, U. Poerschke, The Architect As Performer of Energy Simulation in the Performance Based
Design, Elev. Int. IBPSA Conf. (2009) 1306–1313.

[16] S. Macmillan, J. Steele, P. Kirby, R. Spence, S. Austin, A. Mitchell, I. Frame, A. Coday, M. Hoxley,
Engineering , Construction and Architectural Management Mapping the design process during the
conceptual phase of building projects Article information : To cite this document : About Emerald

22
 www.emeraldinsight.com Mapping the design process during the co, (2002).

[17] A.N. Sadeghifam, S.M. Zahraee, M.M. Meynagh, I. Kiani, Combined use of design of experiment and
dynamic building simulation in assessment of energy efficiency in tropical residential buildings, Energy
Build. 86 (2015) 525–533. doi:10.1016/j.enbuild.2014.10.052.

[18] W. Yu, B. Li, H. Jia, M. Zhang, D. Wang, Application of multi-objective genetic algorithm to optimize
energy efficiency and thermal comfort in building design, Energy Build. 88 (2015) 135–143.
doi:10.1016/j.enbuild.2014.11.063.

PT
[19] I. Zacà, D. D’Agostino, P.M. Congedo, C. Baglivo, Assessment of cost-optimality and technical solutions in
high performance multi-residential buildings in the Mediterranean area, Energy Build. 102 (2015) 250–265.
doi:10.1016/j.enbuild.2015.04.038.

RI
[20] M. Fesanghary, S. Asadi, Z.W. Geem, Design of low-emission and energy-efficient residential buildings
using a multi-objective optimization algorithm, Build. Environ. 49 (2012) 245–250.
doi:10.1016/j.buildenv.2011.09.030.

SC
[21] W. Tian, A review of sensitivity analysis methods in building energy analysis, Renew. Sustain. Energy Rev.
20 (2013) 411–419. doi:10.1016/j.rser.2012.12.014.

U
[22] J.C. Lam, S.C.M. Hui, Sensitivity analysis of energy performance of office buildings, Build. Environ. 31
(1996) 27–39. doi:10.1016/0360-1323(95)00031-3.

[23] N
P.F. de A.F. Tavares, A.M. de O.G. Martins, Energy efficient building design using sensitivity analysis-A
case study, Energy Build. 39 (2007) 23–31. doi:10.1016/j.enbuild.2006.04.017.
A
[24] Q. Wang, S. Holmberg, A methodology to assess energy-demand savings and cost effectiveness of
M

retrofitting in existing Swedish residential buildings, Sustain. Cities Soc. 14 (2014) 254–266.
doi:10.1016/i.scs.?.014.10.007.

[25] T.L. Hemsath, K.A. Bandhosseini, Sensitivity analysis evaluating basic building geometry †TM s effect on
D

energy use, Renew. Energy. 76 (2015) 526–538. doi:10.1016/j.renene.2014.11.044.

TE

[26] EnergyPlus 8.4. Retrieved from https://energyplus.net. Last accessed on 10/05/2017

[27] S. Attia, E. Gratia, A. De Herde, J.L.M. Hensen, Simulation-based decision support tool for early stages of
zero-energy building design, 49 (2012) 2–15. doi:10.1016/j.enbuild.2012.01.028.
EP

[28] Residential Prototype Building Models. Retrieved from

https://www.energycodes.gov/development/residential/iecc_models. Last accessed on 10/05/2017
CC

[29] D.a Stanke, R.L. Hedrick, L.E. Alevantis, M.G. Apte, L.G. Bellenger, D.C. Bixby, D.R. Conover, L. a
Damiano, R. a Danks, F.M. Gallo, J.R. Girman, T.P. Houston, E.P. Howard, R.L. Howard, C.R. Magee, C. a
Marbery, C.O. Muller, D.E. Knebel, S.D. Kennedy, M.F. Beda, J.F. Hogan, J. a Kohler, H.M. Newman,
ASHRAE STANDARD Ventilation for Acceptable Indoor Air Quality, Health Care (Don. Mills). 2007
A

(2007) 1-4. doi:http://dx.doi.org/10.1108/17506200710779521.

[30] J.M. Calm, S.S. Hanson, C.E. Marriott, K. Amrane, R.T. Higa, J.F. Hogan, P. a Baselici, D.L. Beaty, E. a
Conrad, M.D. Lane, C.C. Cottrell, R. Crane, R. Lord, J.J. Deringer, K. Luther, K.I. Emerson, R. Majette, A.
Fraser, M.F. Mcbride, J. a Garrigus, J. Montgomery, D. Weitz, R. Wilson, M.W. Woodford, ASHRAE
STANDARD Energy Standard for Buildings Except Low-Rise Residential Buildings, Society. 8400 (2007)
404–636. doi:http://dx.doi.org/10.1108/17506200710779521.

[31] Residential Energy Consumption Survey (RECS), 2009 Survey Data. Retrieved from

23
 http://www.eia.gov/consumption/residential/data/2009/. Last accessed on 10/05/2017

[32] S. Emmerich, A. Persily, Multizone modeling of three residential indoor air quality control options, (1996).

[33] Sherman, M. H. "Air infiltration measurement techniques." First International Energy Agency Symposium
of the Air Infiltration Centre, entitled "Instrumentation and Measuring Techniques", Windsor, England,
October 6-8, 1980. 2013

PT
RI
SC
U
N
A
M
D
TE
EP
CC
A

24
Figure captions

PT
RI
SC
Figure 1. Step 1: Establishing a baseline building model

U
N
A
M
D

Figure 2. Step 2: Identifying building energy variables and technologies

TE
EP
CC

Figure 3. Step 3: One-at-a-time building technology sensitivity analysis

A

25
 PT
Figure 4: Step 4: Expert analysis and design finalization

RI
SC
U
N
A
Figure 5. Conceptual plan for a sustainable community
M
D
TE
EP
CC
A

Figure 6. Floor plan of the 1st floor

26
 PT
Figure 7. Building geometry

RI
SC
U
N
A
M
D

Figure 8. Zone configuration for the 1st floor

TE
EP
CC
A

Figure 9. Schedule of internal gains

27
 PT
RI
SC
U
N
Figure 10. Heat pump heating performance curves
A
M
D
TE
EP
CC

Figure 11. Comparison of energy model results

A

28
 PT
RI
SC
U
N
A
Figure 12. Sensitivity of single energy variables, except HVAC and DHW
M
D
TE
EP
CC

Figure 13. Sensitivity of single energy variable, HVAC and DHW

A

29
 Energy end uses [kWh]
2,00,000

1,50,000

1,00,000

PT
50,000

0
Heating Cooling Lighting Electric Fans and Domestic

RI
equipment pumps hot water

Final design scanario The most energy-efficient scenario Baseline

SC
Figure 14. Comparison of energy end uses

U
N
A
M
D
TE
EP
CC
A

30
Table

Table 1. Zone configuration

1-bedroom 2-bedroom 3-bedroom Corridor Stairway Community

Floor
unit zone unit zone unit zone zone zone center zone
1st floor 0 3 1 3 2 1
nd
2 floor 5 5 2 3 2 0

PT
3rd floor 5 6 2 3 2 0
th
4 floor 3 3 1 3 2 0

RI
Table 2. Peak occupancy of each zone

SC
1-bedroom 2-bedroom 3-bedroom Community
Zone type
unit zone unit zone unit zone center zone

Peak occupancy rate 1.5 occupants 2.5 occupants 3.5 occupants 7 occupants

U
Categories
N
Table 3. Baseline model configuration

Design details Annotation

A
Exterior wall 2
R18 (3.17Km /W) 2x6@12” on center batt framing
M

Slab R10 (1.76Km2/W) vertical insulation 2” rigid insulation to frost line

U=0.3 SHGC=0.47 VT (Visible Low emissivity, double panes,
Window Transmittance) =0.58, argon gap
D

size of 3’x5’
Storefront glazing U=0.52 SHGC=0.47 VT=0.58
TE

R40 (7.04Km2/W) roof with blown-in 2x4 wood trusses@24” on center

Roof
insulation in resilient channel
LPD (Lighting Power Density) =0.1W/ft2 LED (Light-Emitting Diode)
EP

Stairway lighting
(1.08W/m2)
Community center lighting LPD=0.36W/ft2 (3.88W/m2) LED
2 2
CC

Corridor lighting LPD=0.16W/ft (1.72W/m ) LED

Residential unit lighting LPD=0.44W/ft2 (4.74W/m2) CFL (Compact Fluorescent Lamp)
Heat pump for 1-bedroom COP=3.6(cool), 3.8(heat) 5kW supplementary heater in air
unit 18MBH capacity handler
A

Heat pump for 2,3-bedroom COP=3.5(cool), 3.8(heat) 5kW supplementary heater in air
unit 24MBH capacity handler

Multi-split system for public COP=2.6(cool), 3.0(heat) Up to 9 indoor units

zones 56MBH capacity

31
 R10 (1.76Km2/W) water tank
4500W, 40gal (151.4L) for 1-bedroom unit,
Domestic hot water insulation, supply 120F (49C) hot
50gal for 2,3-bedroom unit
water
Annotation: R1=0.17611m2K/W, U1=5.678W/m2K
Table 4. Exterior wall and roof assembly

Exterior wall Roof

Inside 5/8” gypsum board 5/8” gypsum board

PT
2x4 wood trusses @ 24” on
2x6 framing @ 12” on center with
center with 11” blown-in
5.5” fiberglass batt insulation
insulation

RI
1/2” OSB (Oriented Strand Board) 1/2” OSB
Outside 1.3” fiber cement siding Metal roof

SC
Table 5. Difference in configuration

Category ASHRAE standard model Baseline model

U
Slab R5 (0.88Km2/W) R10 (1.76Km2/W)
Exterior wall
Roof
N
R13 (2.29Km2/W)
R38 (6.69Km2/W)
R18 (3.17Km2/W)
R40 (7.04Km2/W)
A
Window U0.35 SHGC0.4 U0.3 SHGC0.47
M

LPD=0.44 W/ft2
Lighting for units LPD=0.7 W/ft2 (7.54W/m2)
(4.74W/m2)
LPD=0.3 or 0.16 W/ft2
Lighting for public space LPD=0.5 W/ft2 (5.38W/m2)
D

(3.23W/m2 or 1.72W/m2)
Annotation: R1=0.17611Km2/W, U1=5.678W/m2K
TE

Table 6. Major energy variables

EP

Envelope related energy Lighting and glazing HVAC and DHW

variables
Exterior wall Residential lighting Residential HVAC
CC

Roof Window Public space HVAC

Slab Storefront glazing Energy recovery ventilator
Domestic hot water
A

Table 7. Alternative technologies for each energy variable

Exterior ①R19 wall, 2x4 staggered studs ②R25 wall, 2x4 staggered studs ③R24 wall, SIPs
wall ④R28 wall, SIPs ⑤R40 wall, SIPs ⑥R50 wall, SIPs
①R60 roof, 2x4 wood trusses, 17” blown-in insulation
Roof
②R59 roof, 2x4 wood trusses, 11” blown-in insulation with 4” rigid insulation

32
 Slab ①R15 slab, vertical insulation ②R20 slab, vertical insulation
① Window, double-pane, U=0.3 SHGC=0.47 VT=0.58, size of 3’x6.7’
Window ② Window, double-pane, U=0.2 SHGC=0.19 VT=0.37, size of 3’x6.7’
③ Window, triple-pane, U=0.22 SHGC=0.53 VT=0.68, size of 3’x5’
Storefront
①Storefront U0.29 ②Storefront U0.52 with 2’ overhang shading
glazing
Residential
①LED lighting, 0.36 W/ft2 lighting power density

PT
lighting
Residential ① Residential ASHP, cooling COP4.2, heating COP4.0
HVAC ② Residential GSHP, cooling COP5.5, heating COP3.8

RI
Energy
recovery ① ERV 0.79 sensible effectiveness, 0.40 latent effectiveness
ventilator

SC
Public ① Public space multi split system, cooling COP3.5, heating COP3.8
space
HVAC ② Public space GSHP, cooling COP6.1, heating COP5.9

U
Domestic ① Heat pump water heater, 1800W heat pump capacity, 4500W backup electric power, 3.25
hot water energy factor, 50 gal (189.3L)
Annotation: N
1. For ①R19 wall, 2x4 staggered studs, studs are on 2x6 sill plate with blown-in insulation. However, for ②R25
A
wall, 2x4 staggered studs, studs are on 2x8 sill plate with blown-in insulation.
2. Due to the benefit of SIPs wall construction, zone infiltration is reduced from 0.45ACH to 0.43 ACH.
3. R1=0.17611Km2/W, U1=5.678W/m2K
M

Table 8. The sensitivity of single energy variable, except HVAC and DHW
D

Single Single Percentage of

Percentage of
energy EUI (kBtu/ft2) energy EUI (kBtu/ft2) change in EUI
change in EUI (%)
TE

variable variable (%)

Exterior Exterior
30.27 -0.23 29.95 -1.29
wall① wall②
EP

Exterior Exterior
29.60 -2.44 29.48 -2.83
wall③ wall④
Exterior Exterior
29.24 -3.63 29.11 -4.05
CC

wall⑤ wall⑥
Roof① 30.13 -0.69 Roof② 30.19 -0.49
Slab① 30.25 -0.30 Slab② 30.23 -0.36
A

Window① 29.79 -1.81 Window② 29.78 -1.85

Window③ 29.77 -1.88
Residential
30.13 -0.69
lighting①
Storefront Storefront
30.17 -0.56 30.34 0.00
glazing① glazing②
Annotation: EUI: 1kBtu/ft2=3.155kWh/m2

33
 A
CC
EP
TE
D
M

34
A
N
U
SC
RI
PT
 Table 9. The sensitivity of single energy variable: HVAC and DHW

Percentage Percentage
EUI EUI
Single energy variable of change in Single energy variable of change in
(kBtu/ft2) (kBtu/ft2)
EUI (%) EUI (%)
Residential air source
29.88 -1.52
heat pump
Geothermal heat pump 22.25 -26.66
Public space multi-split
29.75 -1.94
system

PT
Energy recovery
28.66 -5.54 Heat pump water heater 27.85 -8.21
ventilator
Annotation: EUI: 1kBtu/ft2=3.155kWh/m2

RI
Table 10. Configuration of two scenarios of building design

SC
The most energy-efficient
Energy variable Pre-final design scenario
scenario
Exterior wall R25 wall, 2x4 staggered studs R50 wall, SIPs

U
R60 roof, 2x4 wood trusses, 17” R60 roof, 2x4 wood trusses, 17”
Roof

Slab
N
blown-in insulation
R15 slab, vertical insulation
blown-in insulation
R20 slab, vertical insulation
A
Window, double-pane, U=0.2
Window, triple-pane, U=0.22
Window SHGC=0.19 VT=0.37, size of
SHGC=0.53 VT=0.68, size of 3’x5’
M

3’x6.7’
Storefront glazing Storefront U=0.52 with 2’ shading Storefront U=0.29
LED lighting, 0.36 W/ft2 lighting LED lighting, 0.36 W/ft2 lighting
D

Residential lighting
power density power density
Residential HVAC Geothermal heat pump Geothermal heat pump
TE

Public space HVAC Geothermal heat pump Geothermal heat pump

ERV 0.79 sensible effectiveness, ERV 0.79 sensible effectiveness,
Energy Recovery Ventilator
EP

0.40 latent effectiveness 0.40 latent effectiveness

Domestic hot water Heat pump water heater Heat pump water heater
Annotation: R1=0.17611Km2/W, U1=5.678W/m2K, 1W/ft2=10.76W/m2
CC

Table 11. EUI comparison

Pre-final design The most energy-

A

Baseline
scenario efficient scenario
EUI (kBtu/ft2) 30.34 16.65 16.34
Percentage of EUI change from
0 -45.12 -46.14
baseline (%)
Sum of EUI change of all
technologies from baseline in 0 -45.23 -48.64
one-at-a-time study (%)

35
Annotation: EUI: 1kBtu/ft2=3.155kWh/m2

PT
RI
SC
U
N
A
M
D
TE
EP
CC
A

36
 Table 12. Comparison of computational efforts in this case study

Evaluation of whole design space System-level approach

Number of building simulation 54,432 24
Estimated time duration of
4,536 2
simulation [hours]

PT
RI
SC
U
N
A
M
D
TE
EP
CC
A

37