Renewable Energy 35 (2010) 15511558

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Assessing the solar potential of low-density urban environments in Andean cities

with desert climates: The case of the city of Mendoza, in Argentina. 2nd. Part
M. Arboit a, *, A. Mesa a, A. Diblasi b, J.C. Fernandez Llano a, C. de Rosa a
a
Instituto de Ciencias Humanas, Sociales y Ambientales. (INCIHUSA CONICET), RD Unit: Laboratorio de Ambiente Humano y Vivienda, Adrian Ruiz Leal s/n. Parque General San
Martn. (5500) Mendoza, Argentina
b
Facultad de Ciencias Economicas, UNCuyo. Area Ciencias Exactas,(CRICYT CONICET), Adrian Ruiz Leal s/n. Parque General San Martn. (5500) Mendoza, Argentina

a r t i c l e i n f o a b s t r a c t

Article history: Energy use in the built environment is globally recognized as a key issue for sustainable urban devel-
Received 11 June 2009 opment. In temperate-cold arid regions with a generous solar resource, such as those of western
Accepted 10 November 2009 Argentina, adequate design and technology can substantially reduce the energy demand for space and
Available online 16 December 2009
water heating in urban buildings. The solar potential of low-density residential urban areas in the city of
Mendozas Metropolitan Area (MMA), has been studied earlier in this research [1]. Several indicators of
Keywords:
the solar potential were elaborated. They provide necessary information when planning and designing
Sustainable urban development
new urban structures or refurbishing existing ones. However, a more direct indicator, relating the
Building morphology
Solar access available solar radiation during a heating season to the space volume to be heated, the Volumetric
Energy efciency Insolation Factor (VIF), seems to be of most practical use as far as contributing a helpful evaluation
indicator, to the above mentioned design processes. The present study follows the methodological steps
used in the former research, evaluating comparatively the results of a Graphic-Computational Model and
a Multiple Linear Regression Statistical Model. As in the earlier study, the good t of both models results
clearly point at the reliability of the statistical procedure and its valuable contribution of a simplied
calculation tool as its by-product.
2009 Elsevier Ltd. All rights reserved.

1. Introduction unsustainable situation of energy use in Mendozas urban building

Many studies have evaluated the interrelationship between
urban form and energy use [28]. Energy consumption in urban
environments can roughly be divided in two main sectors: trans-
portation and buildings. The energy consumed by the building
sector is mainly dependent on the conditions of urban climate and
the micro-scale of the citys inner structure, mainly the congura-
tion of neighbourhoods, urban spaces and morphology (design)
and materials (technology).
While constant advances in energy-efcient building tech-
nology have provided signicant and well known contributions to
energy conservation, progress in urban morphology has been more
complex, limited and conictive [911]. Physically, economically
and legally feasible alternatives are required to minimize energy
waste and maximize the potential use of renewable energies,
namely, solar radiation in urban buildings [1214].
The features of the reference urban environment in MMA
provided the justication for the research, given the present
* Corresponding author. Tel.: 54 26 1 524 4054; fax: 54 261 428 7370.
E-mail address: marboit@lab.cricyt.edu.ar (M. Arboit).
sector. They have extensively been dealt-with in the previous paper
[1] and will be only summarily addressed in this presentation. It is
however necessary to emphasize the importance of the Volumetric
Insolation Factor on north facing walls (VIFnw) (space heating) and
on horizontal roofs or horizontal projection of tilted roofs (VIFhr)
(domestic water heating), as the most expressive and useful indi-
cator relating the energy demand for space and water heating and
the actual insolation of the building components performing as
potential solar collectors.
2. Reference situation and methodology: synthetic overview
2.1. Site and climate
MMA, an urban conglomerate of nearly one million is settled on
the mid-latitude (32.86) arid region of central-western Argentina.
Presently low-density urban environments account for almost 90%
of the urbanized land area.
It features a mesothermal arid climate (1384 heating DD, base
18  C; cooling DD: 163, base 23  C) with intense solar radiation
throughout the year (18.06 MJ/m2 day) (Figs. 1 and 2).

0960-1481/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2009.11.027
1552 M. Arboit et al. / Renewable Energy 35 (2010) 15511558
Fig. 1. Mendozas metropolitan area aerial view.
2.2. Methodology
Since the methodological development of this research has been
informed in the previous publication [1], it was considered neces-
sary to present here only a short synthesis of it, in order to ease the
comprehension of recent studies results, presented for the rst
time in this paper with greater degree of detail, particularly on the
statistical procedures used. The already published stages have
been:
Denition of a set of urban and building morphological vari-
ables, for the analysis of their incidence on the access to the solar
resource in urban buildings. The urban city-block is taken as
analysis unit.
The variables considered in the analysis can be grouped into two
main types: urban and building. The urban variables included: city-
block shape, city-block orientation, vial channels width and urban
forest features; these including: magnitude, solar permeability of
trees and completeness of tree stocks per unit. The building vari-
ables included: morphology (homogeneily/heterogeneily), form
factor (FF), soil occupation factor (SOF), and total occupation factor
(TOF) [15].
The last 3 (three) variables were calculated as the sum of the
total built-up areas on a city-block as if dealing with a single
building (Table 1).
The data collection procedure included: ofcial cadastral les,
satellite images, on-site photographic survey of units, photo digi-
talization and integration of 3D models, with and without the
inclusion of the urban forest. The determination of the
Fig. 2. Typical downtown forested street.
representativity degree of the analysis units was performed
through the use of a statistical procedure and their spatial distri-
bution through a random routine. A set of 32 analysis units (city-
blocks) was nally determined.
The actual insolation of building volumes (north facing faades
and roofs) were calculated through the use a graphic-computa-
tional model for the two situations relating the inclusion of trees.
Detailed data of seasonal crowns permeability for tree species was
considered (Figs. 3 and 4).
Determination of the collecting area (CA) for direct gain: The
potentially available CA (m2) for space heating is the effectively
insolated net glass area of north facing faades (/ 15 ). The glazed
surfaces that receive shadows cast by solid elements (neighbouring
buildings) and by permeable masses (trees) are deducted and the
reduction percentages due to sash mullions, security bars and side
frames (25%) are taken into account as well. (Fig. 5)
Energy calculations for solar space heating were performed for
the present situation, without changes and a target theoretical
situation after solar refurbishing of all sample units. The Load
Collector Ratio (LCR) method of Los Alamos National Laboratory
(LANL) was used.
3. Solar indicators
A numerical indicator relating urban morphology and heated
space volume of analysis units was developed:
3.1. Volumetric insolation factor (VIF)
Expresses the relationship between the total net energy
impinging on unmasked north facing vertical surfaces (/ 15 ),
during a heating season, and the total space volume to be heated.
The indicator is probably the one that best describes the space
heating solar potential through passive solar systems, particularly
Direct Gain, in Mj/m2 year. Its mathematical expression is:
P08 P30 P14:30
m:4 d:1  SMA PMA$1  P $R
h:9:30 TCA
VIF
Volume to be heated
(1)
Where:
TCA: Total potentially collecting areas of north facing faades
(m2)
SMA: Solid masked area (buildings): potentially collecting
faade shaded by neighbouring buildings (m2).
PMA: Permeable masked area (trees): potentially collecting
faade shaded by urban trees, the typical permeability values of
each species are applied to determine the actual collecting area
(m2).
P: Permeability factor: solar permeability percentage of each
species (%).
R (m-d-h): Daily mean solar radiation on north faades for each
month the heating season (Wh/m2). number of heating months (n),
d: number of days per month (5), h: number of hours per day
(Fig. 6).
These values include the diffuse radiation (considered isotropic)
as a function of the sky view factor and the reected radiation from
the immediate environment as a function of urban morphology and
its corresponding albedo. Recent literature on related studies for
the development of this stage is currently being consulted [1619].
4. Results
VIF values are presented on Table 2 for the 32 selected analysis
units. Existing urban morphology congurations that allowed
 M. Arboit et al. / Renewable Energy 35 (2010) 15511558 1553

Table 1
Values of urban and building variables of the sample set.

Scenario Explanatory variables

Urban Building

N8 City blocks Canyon width (m) Trees Morphology FF % SOF TOF

Shape Orientation degrees Magnitude Permeability % Completeness %

1 4 5 20.00 2 51.63 57.35 1 0.49 0.74 0.97
2 3 6 20.00 2 51.63 55.40 1 0.64 0.56 0.78
3 4 6 20.00 2 51.63 65.46 1 0.60 0.72 0.98
4 3 87 20.00 2 42.79 69.22 1 0.61 0.69 0.91
5 3 86 20.00 2 51.63 70.43 1 0.59 0.74 1.00
6 2 6 20.00 2 35.04 77.26 1 0.68 0.60 0.66
7 1 5 20.00 2 42.79 70.52 1 0.63 0.68 0.75
8 1 1 20.00 1 51.03 80.65 1 0.89 0.57 0.61
9 2 1 20.00 1 51.03 79.43 1 0.64 0.58 0.61
10 2 74 20.00 2 42.79 76.34 1 0.65 0.62 0.62
11 2 74 20.00 2 42.79 67.51 1 0.88 0.52 0.52
12 2 74 16.00 2 42.79 68.74 1 0.57 0.67 0.67
13 4 23 13.00 2 42.79 63.89 0 0.62 0.55 0.57
14 4 23 13.00 2 42.79 73.82 0 0.63 0.55 0.56
15 4 23 13.00 2 42.79 85.11 0 0.64 0.51 0.52
16 4 23 13.00 2 42.79 75.93 0 0.59 0.53 0.54
17 4 23 13.00 2 42.79 83.32 0 0.53 0.53 0.80
18 4 23 13.00 2 42.79 67.01 0 0.54 0.52 0.79
19 5 65 15.00 4 42.79 53.65 0 0.68 0.52 0.55
20 5 65 15.00 4 42.79 57.30 0 0.56 0.61 0.61
21 5 65 15.00 4 42.79 54.28 0 0.62 0.57 0.59
22 4 65 15.00 3 38.55 47.13 0 0.62 0.58 0.58
23 4 65 15.00 3 38.55 67.32 0 0.64 0.55 0.56
24 4 65 15.00 3 38.55 56.35 0 0.70 0.52 0.53
25 5 24 15.00 2 42.79 65.85 0 0.63 0.60 0.60
26 4 24 15.00 2 42.79 70.75 0 0.61 0.57 0.57
27 5 24 15.00 2 42.79 67.22 0 0.61 0.58 0.58
28 4 24 15.00 2 42.79 56.93 0 0.64 0.54 0.54
29 3 67 18.00 2 42.79 78.50 0 0.58 0.63 1.05
30 3 67 18.00 2 42.79 67.96 0 0.56 0.65 1.08
31 3 67 18.00 2 42.79 74.04 0 0.56 0.66 1.08
32 3 67 18.00 2 42.79 67.73 0 0.53 0.69 1.16

obtaining maximum, minimum and close to mean values of solar
collection are also identied.
VIF values for north facing walls (VIFnw) vary substantially from
a mean value of 75 Mj/m3 for a heating season, reaching down to
a minimum of 30 Mj/m3 and up to a maximum of 116 Mj/m3; in the
present conguration of low-density urban city-blocks, a very high
percentage of the total radiation impinging on north facing walls is
masked by neighbouring buildings (solid) and urban trees
(permeable) (Fig. 25). The VIFnw values represent a mean 23% of the
total energy demand for space heating during a heating season
(324 Mj/m3) (Fig. 7).
Scenario 29 presents the lowest value of VIFnw (30 Mj/m3)
among the urban city blocks analyzed; several reasons support this
result:
 It presents an unfavourable orientation (67 east from north)
which masks the insolation of the longest longitudinal wall,
which, combined with the homogeneous building morphology,
substantially reduce the access to solar energy.
 The high building density expressed by high values of soil
occupation (SOF) and of total occupation (TOF), at 0.63 and
1.05 m2/m3.

Fig. 3. Volumetric integration of the urban environment. Fig. 4. Insertion of urban trees.
1554 M. Arboit et al. / Renewable Energy 35 (2010) 15511558

Fig. 5. Graphic presentation of area calculations of openings and elevations of an analyzed building group.

 The magnitude of urban trees (2nd) and the high plenitude
levels (79%) in the city-block add one more reason for the low
insolation.
Scenario 8 presents the highest insolation values on north facing
walls, as a consequence of:
related to the built volume (Scenario 1), the heterogeneous
building morphology represents a disadvantage due to the
irregular roofs slopes that partially block the impinging radiation
on the surfaces, presents the lowest Form Factor of the sample
group (0.49) and a substantial increase of building density: SOF
0.74 and TOF 0.97.

 The convenient orientation of the city-block (1 west from

north).
 The highest value of Form Factor (0.89), which implies rela-
tively large potentially collecting faades.
 Soil Occupation Factors (SOF) values below the mean
(0.57 m2/m3) and low Total Occupation Factors (TOF) values
(0.61 m2/m3).
 The slightly higher winter permeability of tree crowns (51%)
increases the available radiation on north facing walls.
The range of values of the Volumetric Insolation Factor on
horizontal roofs (VIFhr) varies from a maximum of 427 MJ/m3 to
a minimum of 346 MJ/m3 (Fig. 8).
In the rst case, Scenario 8 presents a low Form Factor (0.89
m2/m3) and low SOF and TOF values of 0.57 and 0.61, respec-
tively. In the case of the least insolation availability on roofs,
5. Statistical analysis of results
Given the quantity and complexity of the intervening variables,
their interrelationships and different incidence levels, it was
decided to resort to the use of Multiple Linear Regression, Statistical
Analysis (MLRSA), with the aim to determining the incidence level
of morphologic variables (explanatory variables) on the consequent
energy indicator (response variables) and to be able to discard the
variables of minimum impact. Simultaneously, a simple equation is
obtained that allows calculating the indicators value, avoiding the
cumbersome task of running 60 times the graphic-computational
model, for each analyzed city-block. A Multiple Linear Regression
Statistical Model (MLRSM) for the Volumetric Insolation Factor
(VIF) indicator was elaborated.

Fig. 6. Descriptive schemes of TCA, SMA, PMA and AIA.

 M. Arboit et al. / Renewable Energy 35 (2010) 15511558 1555

The statistical analysis contribution is twofold: i. It provides

37
235
32
a hierarchical listing according to the explanatory variables
(energy indicators) incidence level, discarding all the non-signi-

41
245
31
cant variables; ii. It contributes a simple calculation tool from the
data obtained from the sample-set and the results of the two

38
245
30
models, graphical and statistical, by means of a simple equation.
With the aim of generalizing the results, departing from

30
253
29
a limited number of cases, the MLRSM allowed correlating the solar
potential values of the existing building stock, resulting from the

83
416
28
simulation performed with the graphic-computational model, to
a set of urban and building explanatory variables. Those correla-
85
418
27

tions allowed identifying the most efcient technological alterna-

tives for the analyzed situations.
97
412
26

The steps to construct the statistical model were the following: i.

Verifying the models assumptions. ii. Applying the Box & Cox
100
416
25

transformates. iii. Selecting the most signicant set of variables in

order to explain the response variable through a step-back proce-
43
421
24

dure. iv. Analyzing each variables signicance level (p-level), v.

44
394

Testing the models good tness, correlation coefcients to the

23

square that symbolize the variance explained by the model and vi.
63
395

Obtaining the residual standard error [2022]. The obtained values

22

are presented below:

49
401

Urban and building variables analysis: (Table 3)

21

VARIABLE: VIFnw
39
381

After variables selection: (Table 4)

20

VARIABLE VIFhr (Tables 5 and 6)

46
404
19

6. Conclusions
100
245
18
Volumetric insolation factor (VIF) comparative values of the calculated indicator for north facing vertical surfaces and roofs.

From the conventional analysis of results, it can be stated, that

96
226
17

the solar potential of urban environments is essentially dependent

on their spatial congurations. If the mean values of potentially
80
385
16

available north facing walls and horizontal roofs are added (421 MJ/
m3) and the resulting data are compared to the energy demand for
80
388
15

space heating, (324 MJ/m3), it could be stated that high percentages

of conventional energy savings can be obtained by implementing
88
384
14

morphologic and technologic strategies aimed at optimizing the

use of the regions solar resource.
91
383
13

The obtained results demonstrate that in low-density urban

areas, the energy availability on roofs (70%) is considerably higher
69
321
12

than that on north facing walls (30%). For this reason, it is essential
to maximize the solar energy use for domestic water heating, space
100
394
11

heating and provide for eventual PV generation (Table 7).

It is possible to obtain an equation for each indicator that can be
81
377
10

used as a practical tool that allows inferring indicators values

without resorting to the complex graphic-computational proce-
75
285
9

dure to determine the actual insolation. The equations corre-

116
427

sponding to both indicators are presented below: (Table 8)

8

The results obtained through both methods, graphic-computa-

86
346

tional and statistic, are close for the 32 cases of the sample (Fig. 5),
7

thus conrming the validity of using the statistical model to assess

86
330

the solar potential of low-density urban environments in cities of

6

the region (Fig. 9).

71
257

An Excel format sheet work calculation tool was developed from

5

the data obtained of the sample set and the graphic and statistic
91
272

models results. It allows inferring the solar potential indicator

4

values of low-density urban environments through a simple

110
383

procedure.
3

The statistical analysis of an important number of case studies

106
269

allowed the linear correlation of calculated values of the selected

2

indicators to the simulation results of the graphic-computational

79
193

model, identifying the most efcient morphological alternatives for

1

the analyzed reference situation (present without changes). The

VIFnw
Table 2

VIFhr

tool allows obtaining results in an easy way, when compared to the

3D modelling and the graphic-computational model and to expand
1556 M. Arboit et al. / Renewable Energy 35 (2010) 15511558

Fig. 7. Calculated values of Volumetric Insolation Factors on north facing walls (VIFnw).

Fig. 8. Calculated values of Volumetric Insolation Factors on roofs (VIFhr).

Table 5
Table 3
Analyzed variables by MLRSM for VIFhr.
Analyzed variables by MLRSM for VIFnw.
Coefcients Standard Error p-value
Coefcients Standard error p-value
VIFhr (Intercepcion) 114.5414 0.439 0.665
VIFnw (Intercepcion) 66.9313 1.477 0.15448
Shape of city-blocks (sha.) 6.6961 0.786 0.441
Shape of city-blocks (sha.) 4.0424 2.410 0.02520*
Orientation of city-blocks (orie.) 6.3749 0.517 0.611
Orientation of city-blocks (orie.) 3.8528 3.718 0.00127**
Street width (st. wi.) 3.4321 0.940 0.358
Street width (st. wi.) 2.1490 2.808 0.01053*
Trees Magnitude (magn.) 9.6944 0.302 0.766
Trees magnitude (magn.) 6.0683 2.063 0.05170
Trees Completion (comp.) 0.4842 0.310 0.760
Trees completion (comp.) 0.3056 1.024 0.31770
Trees permeability (perm.) 1.1301 0.199 0.845
Trees permeability (perm.) 0.7159 1.247 0.22623
Building Morphology (B. Morp.) 18.1269 5.894 9.15e06***
Building morphology (B. Morp.) 11.0255 5.153 4.18e05***
Form Factor (FF) 80.5018 5.303 3.44e05***
Form factor (FF) 47.5472 3.119 0.00519**
Soil Occupation Factors (SOF) 102.4719 6.881 1.10e06***
Soil occupation factors (SOF) 63.3348 0.241 0.81152
Total Occupation Factor (TOF) 39.8857 9.873 3.92e09***
Total occupation factor (TOF) 23.9571 1.985 0.06034
***Highly signicant.
*Signicant, **Very signicant, ***Highly signicant.

the number of study cases to the total number of low-density units,
allowing determining the available energy for the mentioned uses.
The tool derived from MLRSM attains a mean condence level of
99%.
Table 4
Selected variables by MLRSM for VIFnw.
A later verication and validation of the model outside the main
sample-set was performed, obtaining an explained variance of
87.90%.
The results obtained yielded the following information:
i. Correlation between the indicators values and the maximum and
minimum found in the sample-set units. ii. Comparative analysis of

Coefcients Standard error p-value

Table 6
VIFnw (Intercepcion) 37.415 0.917 0.368127 Selected variables by MLRSM for VIFhr.
Shape of city-blocks (sha.) 3.403 2.486 0.020288*
Orientation of city-blocks (orie.) 3.672 3.804 0.000864*** Coefcients Standard Error p-value
Street width (st. wi.) 2.014 2.818 0.009519** VIFhr (Intercepcion) 62.879 1.472 0.153
Trees magnitude (magn.) 5.357 1.613 0.119747 Building Morphology (B. Morp.) 8.847 10.338 1.06e10***
Building morphology (B. Morp.) 9.322 5.430 1.40e05*** Form Factor (FF) 50.263 9.519 5.86e10***
Form factor (FF) 41.018 3.439 0.002139** Soil Occupation Factors (SOF) 84.295 8.582 4.62e09***
Total occupation factor (TOF) 21.513 1.662 0.109601 Total Occupation Factor (TOF) 23.480 15.906 6.48e15***
*Signicant, **Very signicant, ***Highly signicant. ***Highly signicant.
 M. Arboit et al. / Renewable Energy 35 (2010) 15511558
Table 7
Statistical model results.
Residual standard error DF degrees of freedom Multiple R-squared
VIFnw 10.89 24 0.849
VIFhr 16.07 26 0.9593
Table 8
Equations resulting from the statistical model.
exp34:319512 8:45980 sha  13:968244 orie:  5:675619 st:wi:  8:642648 magn: 50:619810 B:Morp: 141:079372 FF 35:746419 TOF
VIFnw
1 exp34:319512 8:45980 sha  13:968244 orie:  5:675619 st:wi:  8:642648 magn: 50:619810 B:Morp: 141:079372 FF 35:746419 TOF
exp92:57562  91:46402 B:Morp: 478:43853 FF 723:43454 SOF  373:47541 TOF
VIFhr
1 exp192:57562  91:46402 B:Morp: 478:43853 FF 723:43454 SOF  373:47541 TOF
the results of substituting the morphological variables statistically
determined, grouping the most and the least favourable in each
case. iii. City-blocks presenting extreme values on all signicant
variables for the statistical model. iv. The impact of varying all
values of explanatory variables is being studied.
Finally, the calculation tool derived from the statistical analysis,
allows simulating situations where the best and worst conditions of
the studied indicators are encountered, which are obtained by
grouping the most and the least variables values of the most
signicant variables ordered by MLRSM. In this case, the maximum
and minimum resulting values are: 207.52 MJ/m3 and 5.13 MJ/m3
for VIF north facing walls and 623.86 MJ/m3 and 2.46 MJ/m3 for VIF
on horizontal roofs. The great amplitude of the ranges puts into
evidence the important possibilities for improvement, which could
be obtained through the adequate design of urban and building
morphologies in the region.
Considering again the methodological aspect of this study, it is
evident that the good t of the results obtained through the
Graphic-Computational Model and the MLRSM is indicative of the
statistical methods dependability for complex situations analysis,
such as those in the present research, where the consideration of
a signicant number of variables (explanatory variables) must be
integrated, determining the relative weight of each variable on
a determined indicator (response variables). The possible extension
Fig. 9. Compared values of VIFnw from the graphic-computational and statistical
model for the 32 sample-set units.
1557
Adjusted R-squared F-statistic DF degrees of freedom p-value
0.805 19.28 724 2.028e08
0.9531 153.3 426 <2.2e16
of the equations derived from the analysis to structurally similar
situations but quantitatively diverse, establishes a valuable
instrumental contribution to establishing urban and energy plan-
ning criteria, basic to the elaboration of norms aimed at the
consecution of a greater energetic and environmental sustainability
share in urban planning towards the future.
Acknowledgements
Most of the work presented in this paper was conducted within
the research project: PICT No.13-12734/02, Calidad de Vida Urbana
y Desarrollo Urbano Ambientalmente Sustentable en Ciudades
Andinas de Clima Arido (Urban Life Quality and Sustainable Urban
Development in Andean Cities of Arid Climate), which is funded by
the Agencia Nacional de Promocion Cientca y Tecnologica (ANP-
CyT), (National Agency for Scientic and Technological Promotion)
and is coordinated by Carlos de Rosa at the Laboratorio de Ambiente
Humano y Vivienda (LAHV), (Laboratory for Human Environment
and Housing), an R D unit of the Instituto de Ciencias Humanas,
Sociales y Ambientales (INCIHUSA) (Institute for Human, Social and
Environmental Sciences), belonging to the Consejo Nacional de
Investigaciones Cientcas y Tecnicas (CONICET), (National Council
for Scientic and Technological Research) of Argentina.
References
[1] Arboit M, Diblasi A, Fernandez Llano JC, de Rosa C. Assessing the solar
potential of low density urban environments in andean cities with desert
climates the case of the city of Mendoza, in Argentina. Renewable energy.
Elsevier Editorial System. Renewable Energy 2008;33(8). 09601481.
[2] Meadows DH, Randers J, Meadows DL. Limits to growth. The 30-year update.
USA: Chelsea Green Publishing Company; 2002.
[3] Dresner S. The principles of sustainability. London: Earthscan Publications
Ltd; 2002.
[4] Daly H. Beyond growth. Boston: Beacon Press; 1996.
[5] Jenks M, Burton E, Williams K. Compact cities and sustainability: an intro-
duction to the compact city. London: E &FN Spon; 1996.
[6] Owens S. Energy planning and urban form. London: Pion Ltd; 1986.
[7] Blowers A, editor. Planning for a sustainable environment. A report by the
town and country planning association. London: Earthscan Publications
Limited; 1993.
[8] Rogers R. Cities for a small planet. USA: Westview Press; 1997.
[9] Watson D, Labs K. Climatic design. New York: McGraw-Hill Book Company; 1983.
[10] Givoni B. Climate considerations in building and urban design. USA: John
Wiley & Sons, Inc; 1998.
[11] Balcomb JD, Dennis Barley, Robert McFarland, Joseph Jr Perry, William Wray,
Scott Noll. Passive solar design handbook, vol. 3. New York: American Solar
Energy Society, Inc; 1983.
[12] Brager G, de Dear R. Thermal adaptation in the built environment: a literatura
review. Energy and Buildings 1998;27:8396.
[13] Brager G and de Dear R. Climate, comfort and natural ventilation: a new
adaptive comfort standard for ASHRAE standard 55. In: Proceedings of the
moving thermal comfort standards into the 21st century, Windsor, UK; 2001.
pp. 0118.
[14] Nicol F, Humphreys MA. Adaptive thermal comfort and sustainable thermal
standards for buildings. Energy and Buildings 2002;34(6):56372.
[15] Codigo Urbano y de Edicacion. Mendoza: Municipalidad de Mendoza; 2000.
1558 M. Arboit et al. / Renewable Energy 35 (2010) 15511558

[16] Kristl Z, Krainer A. Energy evaluation of the urban structure and dimen-
sioning of building site using the ISO-shadow method. Solar Energy
2001;70(1):2334.
[17] Capeluto IG, Shaviv E. On the use of solar volume for determining the urban
fabric. Solar Energy 2001;70(3):27580.
[18] Ratti C, Raydam D, Steemers K. Building form and environmental perfor-
mance: archetypes, analysis and an arid climate. Energy and Buildings
2003;35:4959.
[19] Robinson D. Urban morphology and indicators of radiation availability. Solar
Energy 2006.
[20] Faraday JA. Linear models with R. London, New York: Chapaman & HLL/
CRC; 2005.
[21] Venables WN, et al. Modern applied statistics with S-Plus. Berlin, New York
Heidelberg, London: Springer-Verlag; 2002.
[22] Box GRP, Cox DR. An analysis of transformations. Journal of the Royal Statis-
tical Society, Series B 1964;26:21152.