sustainability

Case Report
Spatial Model for Energy Consumption of
LEED-Certified Buildings
Jonghoon Kim 1 , Soo-Young Moon 2 and Daehee Jang 3, *

1 Department of Construction Management, University of North Florida, Jacksonville, FL 32224,

USA; [email protected]
2 Living Environmental Research Center, Korea Institute of Civil Engineering and Building Technology (KICT),
Seoul 411-712, Republic of Korea; [email protected]
3 Green Building Research Center, Korea Institute of Civil Engineering and Building Technology (KICT),
Seoul 411-712, Republic of Korea
* Correspondence: [email protected]

Abstract: In this research endeavor, we undertook a comprehensive examination of the factors

influencing the energy consumption of LEED-certified buildings, employing both a general linear
regression model and a spatial delayed regression model. Gaining a profound understanding of
energy utilization patterns within LEED-certified structures can significantly contribute to advancing
eco-friendly construction practices. Our investigation draws upon data from a 2010 study conducted
at the University of Wisconsin—Madison (UW), encompassing various independent variables, such
as temperature, that exhibit some degree of correlation with energy consumption in LEED-certified
buildings. The principal objective of this study is twofold: firstly, to ascertain the significance of
specific exogenous variables, notably temperature, and secondly, to explore the impact of spatial
factors, such as function and location, on energy usage. Our research framework encompasses
meticulous data collection and rigorous analysis, culminating in the presentation and discussion of
our findings. Notably, our study unveils intriguing insights. Contrary to conventional assumptions,
we discovered that the energy consumption of LEED-certified buildings does not exhibit a robust
linear association with average annual temperature, the count of power plants within a 50 mile radius,
or the LEED rating itself. However, our spatial regression models unveil a compelling narrative:
the geographic distribution and functional diversity of distinct LEED buildings wield discernible
Citation: Kim, J.; Moon, S.-Y.; Jang, D.
influence over energy consumption patterns. The implications of our research resonate profoundly in
Spatial Model for Energy the realm of LEED-certified building design and construction. Architects, builders, and stakeholders
Consumption of LEED-Certified should consider the nuanced interplay of spatial variables and geographical positioning in the pursuit
Buildings. Sustainability 2023, 15, of optimal energy efficiency. Moreover, our findings stimulate further inquiry in this field, paving
16097. https://doi.org/10.3390/ the way for future investigations aimed at refining sustainable building practices and enhancing our
su152216097 understanding of energy consumption within LEED-certified structures.
Academic Editor: Herie Park
Keywords: spatial model; LEED buildings certification; energy consumption; ordinary linear regression;
Received: 15 August 2023 LEED-certified buildings
Revised: 20 October 2023
Accepted: 9 November 2023
Published: 20 November 2023

1. Introduction
In the United States, buildings account for 20–40% of total energy consumption [1,2].
Copyright: © 2023 by the authors. Building operations and maintenance consume nearly 80% of such a large amount of
Licensee MDPI, Basel, Switzerland. energy during their life cycle [3,4]. The bulk of energy used in the U.S. is generated by
This article is an open access article non-renewable sources (e.g., coal, oil, and natural gas). The negative effects of these energy
distributed under the terms and types are partially to blame for global warming, air pollution and energy shortages. To
conditions of the Creative Commons reduce building energy consumption and its associated problems, sustainable development,
Attribution (CC BY) license (https:// especially sustainability in building construction, is a concern of more and more official
creativecommons.org/licenses/by/ and unofficial organizations e.g., government and environmental protection organizations.
4.0/).

Sustainability 2023, 15, 16097. https://doi.org/10.3390/su152216097 https://www.mdpi.com/journal/sustainability

Sustainability 2023, 15, 16097 2 of 15

A recent study in the realm of LEED certification and its relation to energy consumption
in sustainable buildings is “Energy Performance of LEED-Certified Buildings: A Review of
Case Studies and Related Energy Simulation Studies”. The authors of this study delved into
an extensive analysis, reviewing 50 case studies and 25 energy simulation studies pertaining
to LEED-certified buildings, with the primary aim of assessing their energy performance.
The findings of this research indicate that, on the whole, LEED-certified buildings
tend to exhibit superior energy performance when compared with their non-certified
counterparts. Nevertheless, there remains room for enhancement in this domain. The
study also identified various factors that exert influence over the energy performance of
LEED-certified buildings, including considerations such as building type, location, and
design features [5]. The authors of the study put forth a recommendation for future research
endeavors to concentrate on the development of more precise energy simulation models
and the evaluation of the long-term energy performance of LEED-certified structures.
The conditions governing LEED certification maintain a degree of uniformity: LEED
buildings do not consistently outperform non-LEED buildings and, in certain instances, may
even underperform. A notable concern arises from the implementation of recommended
actions, particularly those selected by users, which sometimes yield negligible effects. This
predicament can lead to the installation of superfluous facilities solely to accrue certification
points. An illustrative example is the provision of extensive bicycle parking facilities or an
abundance of windows in the vicinity of military installations where security considerations
take precedence. In government buildings, there is often a need to streamline or reduce
facilities to optimize space utilization and simplify movement routes [6,7]. Furthermore,
a deficiency exists in the number of standards that are outcome-oriented and provide
substantial returns on investment beyond environmental management. This situation
underscores a potential disconnect between LEED certification and actual energy savings.
In response, the author seeks to uncover and efficiently address these discrepancies in a
more visible manner by identifying problem areas and implementing solutions that bridge
the gap between certification criteria and genuine energy efficiency.
In 1998, the United States Green Building Council (USGBC) introduced its building
evaluation system, known as Leadership in Energy and Environmental Design (LEED).
LEED swiftly gained traction in the building evaluation market and has since been widely
embraced, including by various agencies such as the U.S. government. Subsequently,
in 2007, Executive Order (EO) 13423 was issued by the U.S. Army, advocating for the
adoption of the “Federal Leadership in High Performance and Sustainable Buildings”
initiative [8]. EO13423 places a strong emphasis on reducing the life-cycle costs associated
with the environmental and energy attributes of federally owned building facilities by
implementing the overarching principles outlined in the Energy Policy Act (EPA) of 2005 [9].
Enhancing energy efficiency and mitigating greenhouse gas emissions stand as paramount
priorities within this policy framework. In response to Executive Order 13423 (EO13423),
the Naval Facilities Engineering Command (NAVFAC) made revisions to the policy in
2008. These changes mandated LEED Silver certification for all new military construction
projects and significant renovation endeavors undertaken within the United States Navy
and Marine Corps infrastructure.
In 2010 a team from the University of Wisconsin—Madison (UW) compared the energy
consumption of US Navy LEED-certified buildings and a military non-LEED comparable
building, to investigate EO13423’s mandate to meet a thirty percent energy consumption
reduction [10]. Additionally, the study compared the LEED-certified buildings against
the national average from the 2003 Commercial Building Energy Consumption Survey
(CBECS). The research findings suggest that achieving the mandated thirty percent savings
in electricity and water consumption, as stipulated by EO13423, cannot be solely ensured
through LEED certification. Furthermore, the data show that energy savings were not
closely related to the number of points received in the “Energy and Atmosphere” category
of the LEED certification process. On the aspect of cost, it was found that, although some of
the buildings had satisfactory results, half would not be considered economically feasible
Sustainability 2023, 15, 16097 3 of 15

when looking at either the total cost of construction or the costs associated with the LEED
scoring category “Energy and Atmosphere” [11,12].
Most previous research has ignored the spatial distribution information of buildings
in its various energy consumption analyses [13–16]. Using a spatial regression model, it
has been found that the distribution of LEED-certified buildings show a clear shift from an
original concentration in major coastal cities to a more even distribution across the US [17].
Additionally, LEED-accredited professionals, accredited by the USGBC to oversee the cer-
tification process, have shown a similar shift. Another study found that the adoption of
energy-efficient residential heating, ventilating, and air conditioning (HVAC) systems in a
neighborhood has contagious effects and will spill over into adjacent neighborhoods’ HVAC
adoptions [18]. Spatial distribution-related factors include not only information related to
location, but also refers to the economic situation (e.g., average monthly income) of resi-
dents, weather history records (e.g., temperature), the dominant population (e.g., ethnicity),
average education level, local LEED organizations, and available construction materials.
Energy consumption, in a large sense, is quite connected to the local social network.
This research is an extension of a 2010 U W study [10] and focuses on the influence
of both endogenous variables (e.g., temperature) and exogenous variables (e.g., distance)
on the energy consumption of LEED-certified military buildings. Specifically, this study is
expected to answer the following questions: (1) is there some linear relationship between the
energy consumption of U.S Navy buildings and the aforementioned endogenous variables?
(2) Do the energy consumptions of LEED-certified buildings have any connection with each
other in terms of spatial distribution?

2. Previous Studies
A comprehensive literature review was undertaken to understand further the relation-
ships between energy consumption and factors such as temperature, spatial relationships,
and building function in LEED-certified buildings.
The energy consumption of US Navy buildings with LEED-certified and non-LEED-
certified counterparts. According to EO13423, all government departments must reduce
energy consumption by 30% by 2015 [19]. Additionally, the Department of the Navy has
required that all new buildings constructed for the US Navy and US Marine Corps obtain a
LEED Silver certification issued by the USGBC. To find any possible relationship between
LEED-certified buildings and the reduction of energy consumption, the authors studied
11 LEED-certified buildings of the US Navy and their non-LEED-certified counterparts in
the US Navy. The authors found that 9 of 11 LEED buildings did not achieve a 30% savings
in electricity consumption. In addition, this research concluded that the majority of the USN
LEED-certified buildings showed more electricity consumption than the national average.
The modeled and actual energy performance of 21 LEED-certified buildings were
explored between December 2001 and August 2005. The authors collected utility billing
data from 2003 to 2005 and compared the billed energy consumption with the modeled
energy use. The authors found that 18 out of the total 21 buildings have an energy saving
of 26% compared with the baseline buildings [20]. The authors were also aware of the
limitations of their study and took caution against extending the conclusion too broadly.
The limitations of the study include: (1) sample amount; (2) as-built and as-designed
discrepancies; (3) changing of occupancy patterns and densities.
Another study re-analyzed data supplied by the New Buildings Institute and the
USGBC on measured energy use data from 100 LEED-certified commercial and institutional
buildings [12]. The authors found that, on average, LEED-certified buildings used 18–39%
less energy per floor area than comparable non-LEED buildings. However, 28–35% of LEED
buildings used more energy than comparable non-LEED buildings. Further, the measured
energy performance of LEED buildings had little correlation with the certification level of
the building, or the number of energy credits achieved by the building at design time.
One hundred twenty-one LEED-certified buildings were studied through 2006. The
authors used three different comparison baselines to quantify the energy performance of
Sustainability 2023, 15, 16097 4 of 15

those LEED-certified buildings [14]. First, an energy use intensity (EUI) comparison of
LEED and national building stock was completed. It was found that the median measured
EUI was 69 kBtu/sf, 24% below the CBECS national average for all of the commercial
building stock. Additionally, statistics show that the median performance of gold and plat-
inum buildings is very close to achieving the interim goals of Architecture 2030. Secondly,
the authors calculated the average Energy Star rating of these LEED buildings and found
that the rating number was 68 (meaning a better energy performance than 68% of similar
buildings) compared with a median rating of 50 for the complete national building stock.
Additionally, nearly half of LEED buildings had Energy Star ratings of at least 75, meeting
the qualification level for an EPA-certified Energy Star building. Thirdly, the authors mod-
eled the energy consumption of those LEED buildings. Results show that nearly half of
them had an energy performance worse than the simulated energy consumption.
The geography was studied based on the distribution of LEED-certified buildings and
professionals [17]. The author found that the number of LEED-certified buildings across
the United States is not distributed evenly across space. Additionally, the author rejected a
previous conclusion that stated that the increases in income, educational attainment, and
percentage of service-sector employment all correlate with more green buildings, criticizing
suggestions that these are only for the rich.
The literature reviewed above focuses on a comparison of the energy consumption of
LEED-certified buildings with that of similar non-LEED-certified buildings. As highlighted
in the literature, LEED-certified buildings generally exhibit energy-saving benefits; however,
a substantial variation in energy consumption among different buildings is evident. This
prompts a natural question: what factors contribute to this variation, and which elements
influence the energy consumption of both LEED and non-LEED-certified buildings?
In this research endeavor, the authors delve into the influence of spatial factors on
the energy consumption of LEED-certified buildings. Specifically, they employ both an
ordinary linear regression model and a spatial lag model to assess the significance of
selected exogenous variables and spatial factors. By scrutinizing the significance values
through techniques like t-tests and spatial tests such as the robust Lagrange multiplier, we
aim to offer valuable insights for future research based on the outcomes of this study.
The author identifies the current trends in LEED certification and determines the
suitability and efficiency of certification points. The points of LEED certification have
been gradually adding and detailing items from V2.2 to the current V4.1. Although this
certification is quite consistent with the ideology of saving energy consumption in LEED-
certified buildings, there is a uniformity in its items and a narrowness in its judgment. Most
of the entries in this table are based on quantitative criteria without analysis [21,22].
After LEED was introduced, the number of certified buildings increased by more than
1000% in the 10 years from 2000, and by 110% in the following 10 years. However, recent
trends show that the registration rate is slowing to less than 5%. This can be seen as a
decrease in overall investment due to the economic downturn since 2019 and following the
pandemic; however, the registration rate is decreasing due to LEED’s registration system,
excessive costs, low efficiency compared with value, registration procedures and time, and
the appropriateness of registered academic subjects [22].

3. Methodology
The research methodology comprises three key components: (1) data collection and
characterization; (2) data analysis, encompassing ordinary linear regression (OLR), spatial
linear regression, and specifications for the LEED energy case study; and (3) presentation
of results and subsequent discussion. Figure 1 provides an overview of the methodology
employed in this project.
Sustainability 2023, 15, 16097 5 of 15

Figure 1. Methodology for spatially related energy consumption analysis.

To elaborate further, the initial step involves the utilization of an OLR model to
examine the linear relationships between selected exogenous variables (e.g., temperature)
and the dependent variable. Following this, we employ a spatial lag regression model to
simulate the influence of both endogenous and exogenous variables. The inputs for this
model include: (1) the energy consumption of LEED-certified buildings; (2) the amount of
power plants within a 50 mile radius of each building; (3) the LEED rating level (e.g., Gold,
Silver, Platinum); and (4) the temperature records for each year.

4. Data Collection and Characteristics

For this project, the authors utilized energy consumption data of LEED-certified build-
ings used in the previous study that was provided for further study. After obtaining these
data, the authors collected the temperature data of each month for those areas where the
studied buildings were located, shown in Figure 2, and also information regarding the
number of power plants in a 50 mile radius from each building, as shown in Figure 3. Fur-
thermore, the authors searched the historical temperature records database and prepared
monthly temperature data for the study areas, as shown in Figure 4. Among the study areas,
the peak temperatures, which happened in July, did not differ much, but the minimum tem-
peratures varied greatly. In this project, monthly temperature data were used to calculate
the yearly average temperature. This transformation has the unfortunate potential to mask
the temperature differences from region to region. Table 1 lists the information related to
buildings in this project, which includes location, function, LEED certification level, total
electricity consumption in one year, and size. The column for the LEED certification level is
not suitable for quantitative analysis because it refers to a qualitative metric. Therefore, in
order to quantify the LEED certification ratings, each certification level was given a scaled
value, shown in Table 2. For instance, we use 3 to represent the buildings that were certified
LEED Gold. Specifically, this number will quantify the influence of LEED ratings in the
regression models.
Sustainability 2023, 15, 16097 6 of 15

Table 1. Location, functionality, and energy consumption of LEED-certified buildings.

Electricity
LEED
Locations Function Usage Size (ft2)
Certification Level
(MWH)
Naval Station Great Lakes, IL Training and instruction Gold 533.44 58,000
Naval Air Station Oceana, VA Childcare Silver 489.99 29,000
Naval Station, Norfolk, VA Maintenance of aircraft Certified 513.7 40,376
Naval Station, Norfolk, VA Maintenance of aircraft Certified 838.2 28,379
Naval Weapons Station, Yorktown, VA Housing Certified 770.8 48,700
Naval Amphibious Base Little Creek Administration Silver 737.9 37,800
(NABLC), VA
Naval Amphibious Base Little Creek Police and security Silver 559.7 25,000
(NABLC), VA
Naval Vase Ventura County, Port Public works office and Gold 144.9 16,443
Hueneme, C A maintenance shops
Naval Vase Ventura County, Port Research and engineering Silver 1288.5 192,028
Hueneme, C A (includes offices and Labs)
Naval Corps Air Station (MCAS) Golf course clubhouse Gold 487.4 13,437
Miramar, C A

Table 2. LEED ratings and their corresponding numbers in a spatial regression model.

LEED Rating Non-Certified Certified Silver Gold Platinum

Scaling number 0 1 2 3 4

Figure 2. Locations of LEED-certified buildings in a 50-mile radius (blue circles)).

Sustainability 2023, 15, 16097 7 of 15

Figure 3. Locations of power plants in a 50-mile radius from each LEED-certified building.

Figure 4. Temperature history records for the study areas.

5. Data Analysis
The LEED energy case study was verified precisely using general linear regression and
spatial linear regression analyses for various reasons. In the case of general linear regression,
this statistical technique permits the examination of the relationship between two continu-
ous (quantitative) variables. One variable, denoted as x, serves as the independent variable,
while other variables denoted as y are the dependent variables. Given that the LEED energy
case study involves a continuous dependent variable (energy consumption), the use of
general linear regression is deemed suitable.
Sustainability 2023, 15, 16097 8 of 15

• Spatial linear regression, on the other hand, factors in the spatial location of observa-
tions. In the context of an LEED energy case study, a building’s energy consumption
may be influenced by its geographical location due to factors such as local climate
conditions and building codes. Spatial linear regression assists in capturing these
spatial effects.
• The process of model specification entails the determination of which independent
variables to include or exclude from the regression equation. In the context of an LEED
energy case study, there exists a multitude of potential predictors for a building’s
energy consumption, such as building size, building age, and heating system type.
Model specification techniques aid in identifying the most significant variables.
• Collectively, these methods establish a robust framework for the examination of energy
consumption in LEED-certified buildings. They enable the incorporation of multiple
predictors, account for spatial influences, and offer flexibility in model formulation.

5.1. Spatial Regression Model

The spatial regression model evolves from the familiar specification of a linear regres-
sion model [23–25].
y = Xβ + ε

(1)
where y is an n × 1 vector of observations on the dependent variable, x is an n × k matrix
of observations on the explanatory variables, with an associated k × 1 regression coefficient
vector β, and ε is a vector of random error terms. In ordinary linear regression (OLR), the
items in ε are independent from each other.
Spatial regression [23] deals with the specification, estimation and diagnostic checking
of regression models that incorporate spatial effects. Spatial dependence is introduced into
the ordinary linear regression (OLR) model in two major ways, one of which is referred
to as spatial lag dependence and the other as spatial error dependence. While the former
pertains to the spatial correlation in the dependent variable, the latter refers to the error
term. In this project, a spatial lag model was used and takes on the form:

y = ρWy + Xβ + ε

(2)

where W is an n × n spatial weights matrix that formalizes the network structure of LEED-
certified buildings and ρ is the spatial autoregressive parameter.
After some matrix algebra, this follows as:

y = (I − ρ W)−1Xβ + (I − ρ W)−1ε

(3)

The best-known test statistic againsteTspatial

We/ S0 autocorrelation is the application of
I =
Moran’s statistic for spatial autocorrelation to (4)
regression residuals. This statistic corrects
e T e/N
the well-known Moran’s I for the fact that the random variable under consideration is a
regression
where residual.
e is an N × 1 vector of OLS residuals y − Xβ, ˆ and 0 = ∑ ∑ wij is a normalizing factor.
S Moran’s I for the regression residual is then: i j

5.2. Model Specification for LEED Energy Case Study

As denoted in Equation (2), the establishment of a weight matrix is necessary for
a spatial regression model of this nature. Previous conducted research has examined
the spatial distribution of H VAC system adoption in new or renovated buildings and
contended that the incorporation of energy-efficient HVAC systems exhibited a pronounced
spatial distribution pattern [26,27]. However, as observed in our project, the adoption of
Sustainability 2023, 15, 16097 9 of 15

H VAC systems represents just one of several factors characterized by spatial distribution.
Building construction materials and architectural styles also share this spatial attribute.
Furthermore, the year in which buildings were constructed exerts an influence on their
energy performance, as the availability of new construction technologies varies across
different time periods.
Nonetheless, quantifying the energy consumption stemming from external loads
(e.g., computers) presents a challenging task. In this regard, it is important to acknowledge
that a weight matrix can only serve as an approximation of the true situation, as the exact
quantification of such factors may remain elusive. For simplicity, the effects of distance
(L) between buildings in the same state, and the function of those buildings e.g., military
training (F), are considered in the weights matrix. The underlying reasons for the authors’
consideration of these two parameters in the weights matrix are:
• Evaluating the influence of buildings on each other is naturally linked to the distance
between them. A shorter distance implies greater similarities between two buildings.
For example, within the same neighborhood, individuals with similar income and
education levels tend to reside together. These factors significantly impact their
preferences for buildings with similar functions. Additionally, local policies play a
direct role in determining the types of buildings adopted within a neighborhood.
• In the context of the U.S. Navy buildings examined in this research, distance also
holds relevance as these structures are typically situated within close proximity to
one another. Furthermore, in terms of energy consumption, people’s habits can exert
an effect. For instance, individuals sharing similar consumption patterns and cultural
traditions tend to exhibit a comparable propensity for energy conservation. All of
these factors can be effectively simplified and evaluated using distance as a scale
of measurement.
• The function of a building plays a pivotal role in the assessment of its energy con-
sumption. Beyond meeting fundamental requirements (e.g., suitability for habitation),
military buildings must also fulfill specific and unique mandates. For instance, a build-
ing dedicated to weapon manufacturing primarily consumes energy in the production
of weapons. Consequently, variations in functional attributes are expected to result in
differing energy consumption patterns.
However, the authors recognized that directly quantifying the impact of such func-
tional properties for each building can be challenging. To circumvent this challenge, the
concept of weights was employed. These weights establish connections between build-
ings with similar functions. For example, two military training facilities with comparable
energy consumption patterns are assigned a weight of 0.5 each when analyzing their
energy consumption.
Shown below is the way in which the weights matrix that was used in this project
was generated.
Example: assuming buildings A, B, C, D are in the same state (e.g., VA). A and B are
used for military training, while C and D are used for administration. The distance between
them are d A B = 50 miles, d A C = 30 miles, d A D = 60 miles, d B C = 70 miles, d B D = 60 miles,
and d C D = 100 miles.
We assume the weight for distance is generated by Equation (4).
(
0.5 d ≤ 50
wd_Building1,2 = miles dis tan ce_50 (5)
Building1,2 d > 50 miles
0.5 ×
Then, the generated weight for building A and B will be:

(6)
w AB = w F_ AB + w D_ AB
Sustainability 2023, 15, 16097 10 of 15

As specified above, we have the weights matrix as:

 
1 0.5 0.42

W= 0 0 0.36 0.42
(7)
 0 0.75
 Y M 0
S

6. Results and Discussion

In this project, the authors used the following two analyses:
(1) OLR for testing the significance of exogenous variables (explanatory variables);
(2) Spatial linear regression for testing the significance of endogenous variables (the
influence of spatial factors which inserted in the weights matrix).

1. Ordinary Least Regression (OLR)

Figure 5 presents the results of the ordinary linear regression (OLR) analysis. Notably,
the adjusted R-square value was determined to be relatively low, standing at 0.2924. This
implies that the selected independent variables, such as temperature, can only account for
29.24% of the variation in the dependent variable, which is energy consumption. Moreover,
the conducted t-tests reveal that there is no statistically significant linear relationship
between the energy consumption of the selected LEED-certified buildings and the chosen
independent variables.
Several factors may contribute to this lack of statistical significance. Firstly, the tem-
perature data used in this analysis are yearly averaged, which may not accurately represent
the real-world conditions. Temperatures fluctuate on a month-to-month basis, and by using
yearly averaged data, nuanced differences that occur between high- and low-temperature
months may be overlooked due to the averaging process. Secondly, the utilizing the scaled
LEED rating to indicate potential building energy performance has limitations. This ap-
proach overlooks the possibility that some LEED-certified buildings may possess subpar
energy performance despite having relatively high LEED ratings (e.g., Gold).

2. Spatial Linear Regression

Figure 6 shows the spatial linear regression results. The linear regression results are the
same as shown in Figure 5. The difference is that a spatial weights matrix was introduced
in the regression model. This spatial weights matrix correlates the energy consumption of
one LEED-certified building with other buildings. The probability of the robust Lagrange
multiplier is close to 0.05 and shows that there is high potential that a spatial correlation
exists in the studied LEED-certified buildings. This result corresponds with the authors’
assumptions for the following reasons: (1) the energy consumption of buildings with
similar functionality will has some correlations, although this point has not been proven
in literature; (2) the geography distribution of buildings will also influence the energy
consumption as people’s behaviors are similar in a neighborhood. In addition, education
level and income have a strong geographic distribution patterns as verified by previous
research in this area.
Figure 6 illustrates the results of the spatial linear regression analysis. It is important
to note that the linear regression outcomes are consistent with those presented in Figure 5.
The primary distinction lies in the introduction of a spatial weights matrix within the
regression model. This spatial weights matrix establishes correlations between the energy
consumption of one LEED-certified building and its neighboring buildings.
The robust Lagrange multiplier’s probability is approximately 0.05, suggesting a strong
likelihood of the presence of spatial correlation within the LEED-certified buildings under
study. This outcome aligns with the authors’ initial assumptions for the following reasons:
• Energy consumption in buildings with similar functionality is expected to exhibit
correlations, even if this specific point has not been conclusively proven in exist-
ing literature.
Sustainability 2023, 15, 16097 11 of 15

• The geographical distribution of buildings is likely to influence energy consumption

as people’s behaviors tend to be similar within a neighborhood. Furthermore, factors
like education level and income exhibit clear geographical distribution patterns, as
validated by previous research in this field.

Figure 5. Ordinary linear regression results.

Figure 7 visualizes collected data and the fitting results using both OLR and OLR
considering spatial effects. This figure partly explains the way in which the spatial weights
can adjust the fitted polyline locally.
Sustainability 2023, 15, 16097 12 of 15

Figure 6. Spatial linear regression results.

Sustainability 2023, 15, 16097 13 of 15

Figure 7. Monthly energy consumption data vs OLR and spatial fitted results.

7. Conclusions
The primary objective of this research was to investigate the multifaceted factors
influencing energy consumption within LEED-certified military buildings. The study
systematically examined both endogenous factors, like temperature, and exogenous factors,
such as proximity to power plants, to discern their impact on energy usage. While the study
did not reveal a strong linear relationship between energy consumption and variables,
such as annual average temperature, the number of power plants within a 50 mile radius,
and the LEED rating, a spatial regression model unveiled noteworthy insights into the
geographical distribution and functional diversity of various LEED buildings, highlighting
their role in shaping energy consumption patterns. This substantiates existing assumptions
in the literature that energy usage varies significantly among buildings located in different
geographical regions and serving distinct purposes.
The utilization of spatial regression in this study stands out as a potent analytical tool,
opening up the potential to unearth intricate correlations between energy consumption
and a diverse array of contributing factors. This research thus contributes significantly to
our understanding of how energy is consumed within LEED-certified military buildings
and sheds light on the interplay between architectural features, geographical context, and
energy efficiency.
The significance of future US LEED-certified buildings for energy managers/utility
directors is underscored by several key points to consider:
• Energy Efficiency: LEED-certified buildings are intentionally designed to surpass
conventional buildings in terms of energy efficiency. This results in reduced energy
consumption for heating, cooling, lighting, and other operational aspects. Lower utility
costs and a diminished carbon footprint are anticipated as a result of these buildings.
• Performance Monitoring: Vigilant and continuous monitoring of the performance of
LEED-certified buildings is imperative to ensure they consistently meet their energy
efficiency objectives. This entails regular assessments of energy consumption and
comparisons with similar non-LEED-certified facilities. Energy managers/utility
directors should adopt a proactive approach to track and analyze these data.
• Climate Strategy: The release of Climate Action 2030 by the U.S. Navy and Marine
Corps sets forth a commitment to achieving net-zero greenhouse gas emissions by
Sustainability 2023, 15, 16097 14 of 15

2050. This strategy will inherently influence the design, construction, and operation of
forthcoming LEED-certified buildings.
• Training and Education: Remaining updated on the latest best practices for managing
LEED-certified buildings is essential for energy managers/utility directors. This may
entail participation in training sessions or workshops, reading industry publications,
or engaging in networking opportunities with fellow professionals in the field.
• Stakeholder Engagement: Energy managers/utility directors assume a pivotal role
in engaging with various stakeholders, including building occupants, maintenance
personnel, and senior leadership. They are instrumental in educating these groups
about the advantages of LEED certification and illustrating how each stakeholder can
contribute to a building’s energy efficiency goals.
Although LEED-certified buildings come with a multitude of advantages, it is impera-
tive to emphasize that they demand meticulous and diligent management in order to fully
realize their potential for enhanced energy efficiency and sustainability. Proper oversight
and strategic decision-making are essential to ensure that these buildings continue to up-
hold their commitment to environmentally responsible and energy-efficient operations
over time.

Author Contributions: J.K.: Writing—methodology, validation, and original draft preparation; S.-
Y.M.: resource and data curation; D.J.: Writing—review and editing. All authors have read and
agreed to the published version of the manuscript.
Funding: This work was supported by grant [20230080-001], “A Research of Architectural and Urban
Technology for a Safe, Comfortable and Sustainable Built Environment” from the Ministry of Science
and ICT in South Korea.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Nejat, P.; Ferwati, M.S.; Calautit, J.; Ghahramani, A.; Sheikhshahrokhdehkordi, M. Passive cooling and natural ventilation by the
windcatcher (Badgir): An experimental and simulation study of indoor air quality, thermal comfort and passive cooling power.
J. Build. Eng. 2021, 41, 102436. [CrossRef]
2. Fonseca, J.A.; Nevat, I.; Peters, G.W. Quantifying the uncertain effects of climate change on building energy consumption across
the United States. Appl. Energy 2020, 277, 115556. [CrossRef]
3. Sadliwala, M.S.; Gogate, N.G. Life Cycle Costing Methodology for Sustainable Construction. In IOP Conference Series: Earth and
Environmental Science; IOP Publishing: Bristol, UK, 2022; Volume 1084, p. 012023.
4. García Ahumada, F.L.; Gonzá lez Gaya, C.; Sebastiá n Pé rez, M. The Management of Maintainability throughout the Life Cycle of
Buildings and its Contribution to sustainability. Management 2022, 2, 004.
5. Ur Rehman, H.S.; Raza, M.A.; Masood, R.; Khan, M.A.; Alamgir, S.; Javed, M.A.; Roy, C.; Lim, J.B. A multi-facet BIM based
approach for Green Building design of new multi-family residential building using LEED system. Int. J. Constr. Manag. 2023, 23,
2024–2038. [CrossRef]
6. Dinmohammadi, F.; Han, Y.; Shafiee, M. Predicting Energy Consumption in Residential Buildings Using Advanced Machine
Learning Algorithms. Energies 2023, 16, 3748. [CrossRef]
7. Cole, R.J.; Kernan, P.C. Life-cycle energy use in office buildings. Build. Environ. 1996, 31, 307–317. [CrossRef]
8. O’Neil, S.G. Superfund: Evaluating the impact of executive order 12898. Environ. Health Perspect. 2007, 115, 1087–1093. [CrossRef]
9. None, N. Annual Report on Federal Government Energy Management and Conservation Programs, Fiscal Year 2011 (No. DOE/EE-1371);
EERE Publication and Product Library: Washington, D C, USA, 2014.
10. Menassa, C.; Mangasarian, S.; El Asmar, M.; Asce, S.M.; Kirar, C. Energy Consumption Evaluation of U.S. Navy LEED-Certified
Buildings. J. Perform. Constr. Facil. 2012, 26, 46–53. [CrossRef]
11. Carl, V. Cost Analysis of LEED Certified United States Navy Buildings. Ph.D. Thesis, Naval Postgraduate School, Monterey, C A ,
USA, 2011.
Sustainability 2023, 15, 16097 15 of 15

12. Newsham, G.R.; Mancini, S.; Birt, B.J. Do LEED-certified buildings save energy? Yes, but. . .. Energy Build. 2009, 41, 897–905.
[CrossRef]
13. Pé rez-Lombard, L.; Ortiz, J.; Pout, C. A review on buildings energy consumption information. Energy Build. 2008, 40, 394–398.
[CrossRef]
14. Salmon, W.; Turner, C.; Analyst, S.; Owens, B. Energy Performance of LEED for New Construction Buildings. New Build. Inst.
2008, 4, 1–42.
15. Sartori, I.; Hestnes, A .G . Energy use in the life cycle of conventional and low-energy buildings: A review article. Energy Build.
2007, 39, 249–257. [CrossRef]
16. Scofield, J.H. A Re-examination of the NBI LEED Building Energy Consumption Study. In Proceedings of the International
Energy Program Evaluation Conference, Portland, OR, USA, 12–14 August 2009; pp. 764–777.
17. Cidell, J. Building Green: The Emerging Geography of LEED-Certified Buildings and Professionals. Prof. Geogr. 2009, 61, 200–215.
[CrossRef]
18. Noonan, D.S.; Hsieh, L.-H.C.; Matisoff, D. Spatial Effects in Energy-Efficient Residential H VA C Technology Adoption. Environ.
Behav. 2011, 45, 476–503. [CrossRef]
19. Jain, R.K.; Smith, K.M.; Culligan, P.J.; Taylor, J.E. Forecasting energy consumption of multi-family residential buildings using
support vector regression: Investigating the impact of temporal and spatial monitoring granularity on performance accuracy.
Appl. Energy 2014, 123, 168–178. [CrossRef]
20. Diamond, R. Evaluating the Energy Performance of the First Generation of LEED-Certified Commercial Buildings. Build. Environ.
2011, 105, 172–184.
21. Scofield, J.H.; Brodnitz, S.; Cornell, J.; Liang, T.; Scofield, T. Energy and greenhouse gas savings for LEED-Certified US office
buildings. Energies 2021, 14, 749. [CrossRef]
22. U.S. Green Building Council. What Is LEED? 2020. Available online: https://www.usgbc.org/help/what-leed (accessed on 6
August 2023).
23. Anselin, L. Under the hood Issues in the specification and interpretation of spatial regression models. Agric. Econ. 2002, 27,
247–267. [CrossRef]
24. Anselin, L. Spatial Externalities, Spatial Multipliers, and Spatial Econometrics. Int. Reg. Sci. Rev. 2003, 26, 153–166. [CrossRef]
25. Anselin, L. Spatial Regression. In The SAGE Handbook of Spatial Analysis; Sage Publications Ltd.: London, UK, 2006.
26. Blair, C.; Noonan, B.P.; Brown, J.L.; Raselimanana, A.P.; Vences, M.; Yoder, A.D. Multilocus phylogenetic and geospatial
analyses illuminate diversification patterns and the biogeographic history of Malagasy endemic plated lizards (Gerrhosauridae:
Zonosaurinae). J. Evol. Biol. 2015, 28, 481–492. [CrossRef]
27. Truong, L.H.M.; Chow, K.H.K.; Luevisadpaibul, R.; Thirunavukkarasu, G.S.; Seyedmahmoudian, M.; Horan, B.; Mekhilef, S.;
Stojcevski, A. Accurate Prediction of Hourly Energy Consumption in a Residential Building Based on the Occupancy Rate Using
Machine Learning Approaches. Appl. Sci. 2021, 11, 2229. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.