Energy and Environmental Aptitude (EEA) to Assess So- lar Energy Exploitation in Cities

In Argentina, the number of residential buildings (homes) has grown 20% in the last decade (2000-2010), thus increasing the level of non-renewable energy consumption of the residential area per inhabitant (0.25 TOE per capita per year in 2000, 0.3 TOE per capita per year in 2010) and lowering the construction quality. This situation has a significant impact on the GHG emissions to the environment. W ithin this context, this paper aims to assess the energy and environmental aptitude (EEA) of residential building groups in characteristic urban areas (urban mosaics), to estimate the potentiality of solar energy and energy efficiency explo itation. To begin with, the following urban area variab les were calculated from aerial and satellite image classification techniques: i. Horizontal heat d issipation area; ii. Construction system of horizontal heat dissipation area – Roofs; iii. Vert ical heat dissipation perimeter – Walls; iv. Bu ilt-up area height. Then, obtained results were exposed on two characteristic urban areas, differing in the compactness of its fabric, what allowed to evaluate that the urban area of greater compactness presented lower energy loss and similar EEA in relation to the urban area of lower compactness. This conclusion contributes to the discussion about the diffuse city and the compact city, since it reinforces the need to develop urban conglomerates with greater building compactness.


Introduction
Urban planning allo ws city council authorit ies to have a mid-and long-term vision since it influences the future development of cities. This is why it plays a key role in the creation of co mmunit ies that lead this develop ment to a mo re sustainable position between culture and nature.
There are two key factors to consider for a sustainable urban planning: occupation and socio-territorial models and patterns, and urban energy services production (demand-supply), such as natural gas or electrical energy. They determine the economic and p roductive activities as well as mobility, production and building construction; consequently, they influence the volu me of generated GHG emissions.
As regards the urban occupation model, it is remarkable the role played by buildings as urban environ ment and not as individual technical objects. The reduction in energy consumption and current emissions will not be achieved unless the built-up area is considered as a whole; in particu lar, the existing built-up area has to be taken into account. However, the biggest efforts in the last years have been centered in the design of new efficient build ings while less attention has been paid to the improvement of existing buildings and the infrastructure supporting them [1].
As far as energy services are concerned, it is understood that the incorporation of solar origin renewable services (electric or thermal generation) implies a reduction in GHG emissions. In this respect, Rueda[2] stands out that it is fundamental to incorporate and sustain urban organization with renewable energies as basic resource to reduce impact over the at mosphere, understanding the city as an open system where energy planning is necessary [3]; he proposes to use them as an essential and applicable strategy to promote sustainable development [4].
In Argentina, the growth of the residential build ings is still sustained (20% in the 2000-2010 decade according to data fro m the INDEC -Nat ional Institute of Statistics and Censuses, according to its acronym in Spanish) [5]. But this growth has been developed without incorporating an improvement in the construction quality or other energy sources that allo w energy savings given the current levels of consumption or interior co mfort improvements. This is shown in the national energy balance[6] that p resents -in the national energy matrix, non-renewable energy consumption levels in the residential sector (main ly natural gas, electrical energy and bottled gas) in the year 2000 of 0.25 TOE per capita per year, with an increase of 0.3 TOE per cap ita per year in 2010. It is then essential to improve this situation to avoid a greater impact on the environment and the depletion of the resources available. Now, how can we know the po-tentiality to explo it solar energy and improve energy efficiency in build ings?
Given this situation, it is necessary to develop tools, such as energy models, for the decision-makers to help in the follo wing aspects: (i) to manage quality imp rovements in the built-up environment; (ii) to determine where to invest resources to achieve greater benefits [7]; (iii) to accompany improvement processes of build ing energy and thermal quality with the mod ification of the current leg islation. At present, most tools tend to consider the building as a sole entity and not from the urban scale perspective [8].
Therefore, to further these aspects, it is necessary to study the relation between buildings and their energy and environmental aptitude (EEA) to use energy in a more efficient way or replace it with alternative sources. It will also be useful for the application in building massive recycling strategies, which have proved to be very successful in d ifferent countries, getting at present up to 80% of energy savings with d ifferent technical-econo mic viab ility [10] [1]. At methodological level, several research analyzing the urban block (urban fabric min imal unit that circu mscribes buildings and that is limited by streets, also called four-sided square in chessboard layout) and the building envelope back up this development [11] [12][13] [14]. The EEA is analyzed with a methodology applied to urban mosaics previously developed [15].

Urban Mosaics
Urban mosaics (UM ) are characteristic areas wh ich are representative of bigger ones. They have been studied by different disciplines, such as art, landscape ecology and sociology [16] [17][18] [19]; in this case, they strictly refer to the urban space, what poses the need to go into the morphological aspects for their classification using different variables and indicators ( Figure 1). UMs main co mponents are build ing units associated to recognizable building typologies, located in parcels. Many o f them will make up an urban block and they, in turn, associate among themselves through the public spaces of the streets, pavements and voids, at a part icular d istance. The disposition, characterizat ion, shape and occupation of their components over the territory shape one kind of urban mosaic which is defined as a characteristic pattern [20].
Its analysis and processing can be done in manual, semi-automat ic or auto matic way.
In the manual procedure, there is qualified observation of aerial o r satellite images and the survey of the areas is done in situ; then, the data collected is drawn in two and three dimensions (with CA D programs) to calculate size and shape of build ings. Finally, through direct observation, surveyed building units are assimilated to the study area historical records about recognizable building typologies. This procedure is really t ime-consuming.
{0>Ante esta situación, se observaron los avances desarrollados por la ecología del paisaje y por la teledetección.<}100{>Given this situation, the advances made by landscape ecology and remote sensing were considered.<0} {0>La primera, incorporó el análisis de la fotografía satelital y aérea, y el uso de los sistemas de información geográfica (SIG) que co mp lementan bases gráficas y numéricas de distintas áreas e xtensas [21].<}94{>The former incorporates the analysis of satellite and aerial photography, and the use of geographic informat ion systems (GIS) that complement graphic and numerical databases of extended areas [21].<0} {0>Por su parte, la teledetección en ámb itos urbanos sobre imágenes, permite numerosas aplicaciones que van desde estimación de la población, la cuantificación de camb ios, la estimación de la densidad de edificación, la detección de asentamientos urbanos y rurales, en forma continuada y actualizada, lo que le confiere una indudable utilidad de cara a la gestión y planificación del desarrollo urbanístico [22].<}92{>The latter, in urban environ ments, has a lot of applications that range fro m population estimation to change quantification, building density estimation, and urban and rural detection in a continuous and updated way. This is why remote sensing is so useful for urban development planning and management.<0} According to these antecedents, it was observed that, through the combination of resources and techniques, it is possible to compare and devise indexes for studying UMs in an automatic way, simp lify ing the in situ survey work.
Consequently, this paper aims to expose a methodology that complements the concept of UM previously developed. The proposed methodology determines the EEA inferred and calculated fro m the detection of the following variab les: horizontal heat dissipation area, construction system o f the horizontal heat dissipation area, vertical heat dissipation perimeter and built-up area height on satellite and aerial urban images. Such methodology is developed and applied on two characteristic sectors to evaluate their EEA.

Methodology
Build ings are modelled in their exterior envelope by applying segmentation techniques on urban images (details are omitted because volumetry influence exceeds the importance of details) [11]. In remote sensing, the image segmentation process is defined as the search for ho mogeneous regions in an image and the classification of these regions [19]. It allows to extract d ifferent types of characteristics fro m the outstanding objects. For this methodology, we are specifically interested in the following characteristics -geo metry (shape and size), localizat ion (height, width, area, perimeter, shape factor, etc.), intensity and brightness of the reg ion and neighbourhood. The different techniques are applied through the functions developed in the Image Processing Toolbox 7, MATLAB software type. For the analysis, d igitalized analogical aerial images in scale 1:20,000 [22] were used, as well as free access satellite images[23]. The following variab les used to calculate the EEA of an urban sector were analized, and they are su mmarized in Figure 2: i. Delimitation and measuring of the built-up area and the empty area of the urban mosaic to get the horizontal heat dissipation area (HHDA).
ii. Ho rizontal heat dissipation area classification to get the constructive system of the horizontal heat dissipation area -Roofs (tiles, reinforced concrete slab, metal sheet) iii. Delimitation and measuring of the built-up area outline to get the vertical heat dissipation perimeter (VDP).
iv. Delimitation of buildings shadows to know the built-up area height (BA H).
The informat ion collected is represented as vector file (dwg, d xf, etc.), wh ich is used in computer-aided drawing programs (CAD) or geographic information systems (GIS). In this way, a database with the urban-morphological characteristics of a city sector or UM is achieved.
Fro m the operation of the exposed variables, the EEA of an urban sector can be calculated, and different types of urban fabric co mpositions can be evaluated or co mpared according to the best energy performance or energy potential. For this purpose, ad hoc models are used in compliance with e xisting standards, regulations and estimat ions.
Up to now, seven EEA indicators were used, the first three related to energy loss and the remain ing four, to solar gain potentials: • Bu ildable surface explo itation (BSE).
• Energy loss through roofs per built m 2 in winter period ( E LR ).
• Solar gain through windows per built m 2 in winter period (SGW).
• Use of roof surface fo r solar water heating (URSWH).

Horizontal Heat Dissipation Area (HHDA)
This variable is analyzed in four stages: i. thr es hol din g segmentation on greyscale images; ii. morphological gr adients on the aforementioned resulting images; iii. labelling o f each region on the p revious image; iv. HHDA q u antifi cati on. For stage i., the thresholding segmentation tec hni qu e was applied (separating the objects of interest fro m the rest on the basis of pixel value) on d igitalized analogical aeri al images in greyscale in scale 1:20,000 [24]. The image w as segmented according to the histogram, g iven that the vari ous objects of the image present different grey levels. Fro m the histogram, a threshold was chosen since it is the point who s e intensity separates, in this case, the pixe ls belonging to the buildings fro m the background (vegetation and land), with a threshold value 200. In the resulting image, the reg ions detected are very irregular, they are open o isolated little detections that can be deemed wrong since they do not belo ng to any building.
For stage ii., the starting point is the resulting image where the regions are closed and the object structure is simplified by the application of morphological gradients (dilation and erosion). For this, an image sweep is performed with a 9x9 square structuring element in order to quantify the way it is confined, getting as a result a new image with the simplification of the objects making up the urban area in regular shapes. In this way, detections are mo re defined, even though small objects persist that have to be eliminated since they do not correspond to buildings.
For stage iii., a labelling of each region was perfo rmed on the previous image, and those regions with a size lo wer than 20% of average size were eliminated. The result is a very well-defined image, with regular objects, very close to reality.
For stage iv., binary objects are surveyed (buildings-background) defining the HDA (white colour areas). The quantity of pixels making it up is quantified and it is turned into a metric system scale. Figure 3 shows the images resulting fro m i, ii, iii and iv.

Stagei
Stage ii Stage iii Stageiv In this case, the thresholding segmentation technique is also used, with the only difference that colour satellite images are used here to know the textures of the objects' horizontal area. Th resholds are chosen in the points dividing the dif-ferent colours of each roof constructive system. Broadly, there are three usual kinds of constructive systems: concrete slab (white colour, brighter), metal sheet (grey) and tiles (mostly red). To these three classes there correspond two thresholds, one to divide the whites in the image fro m the rest, and the other to divide the red parts from the rest.
In Figure 4, the original satellite image is observed, together with the thresholding segmentation on the satellite image separating colour red fro m the rest. A threshold that classified build ings with tile roofs (red) was incorporated.

Vert ical Dissipation Perimeter (VDP)
The borders of an object in a greyscale image can be defined as the transitions between two reg ions with sign ificantly different levels of grey. To delimit borders and measure buildings' perimeter, the technique used is border detection fro m images resulting fro m the HHDA.
Bo rder detect ion was previously perfo rmed on the t wo types of HHDA result ing images used, greyscale aerial and colour satellite. It was observed that the results fro m the aerial image present great er regu larity in the bo rders than in the satellite image; so the fo rme r was used. Results are observed in Figu re 5. The information about the height was obtained fro m the shadows shown by the shadows of the buildings. In this case, the thresholding segmentation technique and the morphological g radients techniques were used on aerial images for shadow detection. An in itial thresholding with a threshold value 67 made it possible to get a binary image with every type of shadow conveyed (fro m build ings, vegetation, common walls, or any other badly classified dark object).
The ne xt step was to eliminate those shadows of no use for this analysis by means of morphological operations. It was then necessary to have the follo wing additional info rmation : the binary images resulting fro m HHDA delimitat ion, and the sun position that could be determined by the operator. Then all those shadows conveyed by built-up areas were selected. For the case of this particular image (figure 6, third image), they were all those shadows (blue colour) which were belo w the built-up area (red colour). All the zones that did not correspond to this pattern were eliminated, that is vertical shadows corresponding to fences, common walls and the like.
Then, the open irregular shadows were closed with a 5x5 rectangular structuring element with a dilation operation. Last, an image sweep is performed to eliminate shadows that did not correspond to any construction detected.
In Figure 6, a succession of images comb ining the HHDA with its respective shadow conveyed is presented (useless shadows were eliminated). The construction height is deduced relating the solar height angle and the length of the shadows conveyed by the buildings through trigonometric functions.

EEA Indicator Calculation
Energy loss indicators are calculated in the follo wing w a y: • Bu ildable surface explo itation (BSE): it indicates the proportion of land occupied with construction. BSE = built-up area (m 2 ) / buildable area (m 2 ) • Exposed envelope of built-up volume (EEBV): it indicates that to greater value, greater thermal loss due to the fact that there is mo re exposed envelope to the exterior in relation to the built-up volume. EEBV = exposed envelope (m 2 ) / built-up volume ( m 3  SGW (kW m 2 in winter) = energy gained through windows (kWh) / built-up area (m 2 ) • Use of roof surface for solar water heating (%) (URSW H): considering only 50% of the roofs available, it indicates in a relative way the surface used to cover the hot water demand by the inhabitants of the UM. URSWH = necessary surface (m2) to produce SWH per inhabitant * total number of UM inhabitants / sunny roof surface available (m2) * 100

Case Study
The methodology was applied on two urban mosaics in L a Plata, Buenos Aires, Argentina.
This city is located in the Northeast of Buenos Aires province, 60 km away fro m the City of Buenos Aires, -34° 55' latitude (South) and -57° 17' longitude (West). Its total surface is 821 km 2 . Its height above sea level ranges between 0 and 15 meters, and it is geomorphologically characterized by the plain (Pampeana plain); it has temperate-humid climate [24]. The city of La Plata was funded in 1882 as the capital of the province and built according to the layout of the urban engineer Pedro Benoit. It was materialized as a reflect ion of the hygienist urbanism o f the end of the XIX century as regards street amplitude and wooded avenues, which ensured comfort, ventilation and cleanliness [25].
Two sectors of the urban area were selected. They are both similar regard ing: land use (mainly residential), regular urban layout (10 m frontage parcels) and orientation (NE-SW). The differences lay in the co mpactness of the urban fabric (the degree of co mpactness indicates the predominance o f built -up volu mes over empty spaces), which is represented by the urban consolidation (see Figure 7). Fro m this classificat ion, it was calculated that average consolidation areas (20-40 homes per hectares) represent 17.30% (1332 hectares) of the urban area extension of La Plata with 154,091 inhabitants (116 inhabitants per hectare). Low consolidation areas (less than 20 homes per hectares) represent 80.49% of the total (6,196 hectares), with a population of 377,107 inhabitants (65 inhabitants per hectare). High consolidation areas (more than 40 homes per hectare) representing around 2.2% of territorial extension are not considered. This first diagnosis shows the disperse fabric characteristic of most areas in the city.

Energy and Environmental Aptitude Calcul ati on
In Table 1, the result of the variables calculated and of those resulting fro m the quantificat ion of the urban sector volumetry (HHAD, CSHHDA, VDP, volu me, facade surface) is observed. In Figure 8, volu metry is observed.
Fro m results in Table 1, Tab le 2 shows the calculus of EEA indicators for both areas.

Discussion
Fro m the EEA indicators applied to the two sectors, we can observe: • The average consolidation mosaic (UM1) has a higher BSE indicator (0.41) than the lo w consolidation mosaic (UM2) (0.12). This means that the buildable surface is used more efficiently in the former.
• The average consolidation mosaic has a lower EEBV indicator (0.45) than the low consolidation one (0.69), which translates into fewer losses per envelope.
• The average consolidation mosaic (UM1) has fewer losses through roofs (1.90 kW m 2 winter) in co mparison with the other mosaic (UM2) (2.67 kW m 2 winter). This shows that the former has the best thermal quality of all the roof constructive systems.
• Both mosaics present similar values in the SAF and SAR indicators (between 85% and 97%), wh ich means that the solar obstruction degree due to self-portrayed shadow is not significant in either of them.
• The average consolidation mosaic (UM2) has a lo wer SGW indicator (4.73 kW/m 2 winter) in co mparison with the low consolidation mosaic (UM1) (7.63 kW/ m 2 winter). This evidences that the surfaces of potential sun collecting facades are proportionally bigger in the lo w consolidation mosaic.
• Both mosaics present low URSW H indicators (30% and 40% in UM 1 and UM2, respectively). This shows that roofs, as potential surfaces to incorporate solar water heaters, give the possibility to be also used for other applications, such as photovoltaic power generation.
To sum up and to relate thermal loss indicators with thermal gain ones, we can conclude: • In co mparison with the low consolidation mosaic (UM 2), urban areas represented by the average consolidation mosaic (UM1) p resent better exploitation of the buildable area, s maller exposed envelope area, fewer energy losses through roofs, similar percentages of solar blocking in facades and roofs, significant ly lower solar gain through windows, and larger surface of availab le roofs to incorporate power generation solar systems.
• Consequently, these urban areas represented by the UM2 have a better EEA for the explo itation of renewab le energies.
• Considering population and territorial extension of the areas studied, it is possible to conclude that 17% of the total territory of La Plata (home to 27% o f the population) p res e nt b etter E E A in dic ato rs. Lik e w is e, a pp ro ximately 80% of the territorial extension has a lower EEA , even though there is better potential for energy gain explo itation due to the lower building co mpactness.

Instrumentation
The results exposed show that semiauto matic object interpretation procedures are suitable for UM's requirements.
These procedures have simplified urban survey, minimized field study and reduced operator's time. They significantly contribute to the improvement in the efficiency of the interpretation of urban areas, co mbin ing measuring speed with the operator's interpretation ability.
The applicat ion of this UM p rocessing technique allowed to know the characteristics of both UMs and, with such information, perfo rm the spatial modelling and the shape and quantitative synthesis of its variables to calculate EEA indicators.
As regards the images used for this work in particular, some considerations about the pros and cons of aerial and satellite photography. The advantage of aerial photography is its good spatial resolution and, consequently, good border definit ion. It also allows to know the date and time of the photo taking and thus estimate sun position. Within its disadvantages, we may say aerial photography has fallen into disuse due to its high costs -so the information surveyed becomes easily out of date, and the grey scale does not allow to positively spot the constructive system of the horizontal surfaces (roofs).
On the other hand, satellite photography does allow to discriminate the constructive system of horizontal surfaces thanks to the existence of different color bands (RGB). However, as the images are made up of several takings (different dates and times), sun position cannot be determined. Moreover, free-access satellite images do not have appropriate spatial resolution and border definit ion, what leads to imprecise detections.

Energy Pl anni ng and Management
By means of this methodology, areas can be evaluated to improve their energy and environmental conditions and to know those residential gatherings more appropriate to promote more sustainable occupation models.
Likewise, it allows to propose measures to improve energy management in the city according to the pot enti alities of each sector.

Energy and Environmental Aptitude (EEA)
As regards indicators devising, we can conclude that it has allo wed us to establish the differences between both sectors and to elaborate conclusions for their imp rovement.
Thanks to them, mo re sectors can be studied with the same methodology, broadening the knowledge o f such a vast and complex area as the residential sector of the city is.
Moreover, more specific indicators can be incorporated: energy loss through all the elements of the envelope (windows, floors, walls, and so on); gains fro m other energy generation sources, such as thermal storage walls, solar hot air collector, or photovoltaic systems, among others. They will be studied for future application.