Table 1 shows the synthesis of thermal and morphological characteristics. Study cases were monitored and evaluated under real conditions of use, and the performance of each case was described in previous works [23-26]. Table 2 reviews the natural gas consumption for heating during 1996-2009, together with the average temperature conditions measured during 2003-2010. It is important to note that, from a climatic point of view, the annual temperature variation, recorded during this period, was 3%, while the inter-annual variation found during winter months only was higher (10%). However, in order to simplify the analysis, pre-set temperature values in winter were used as a constant. Therefore, it allows for taking values resulting from different registration periods. In order to understand the occupant’s interactions with their built environment, a brief description of occupancy behavior, use of heating appliances and perceived comfort based on unstructured interviews are displayed in Table 3.

The envelope’s thermal transmittance of single-family dwellings is greater than the one recommended in a study [19]. The volumetric coefficient of loss, G, exceeds the recommendations in a study, [26] suggesting a value between 1,587 and 2.6 according to volume and heating degree-day (Table 1). The exposure factor is higher in H2 (73%) by having its west-oriented wall exposed to the outside. The compactness index varies between 57 and 73%. H5 shows little compactness (Ic = 63%) and high exposure factor and Envelope/floor area factor FAEP, 1 and 3.7, respectively.

The monitoring results of H1 were between 13-20 July 2010. This period was cold with outside temperatures that reached -5ºC. The average indoor temperature (19.9°C) was 16.3°C above the outside average (3.6°C). Between 11-15 August 2010, in H2, the average indoor temperature was around 21.8°C, 14.1°C above the average external (7.7°C). The average indoor temperature of H3 (July 13-19, 2010) was 16.0°C, 12.4 °C above the average exterior (3.6°C). H4, house facing south, with bedrooms towards the north, was occupied by two retired people. The average indoor temperature was 17.9°C, 10.5°C above the outside (7.4°C) between July 31 and August 6, 2010. H5 dwelling was evaluated during June 2003. The average indoor temperature (20.5°C) was 11.1°C above the outside (9.4°C). According to the psychometric chart of climatic conditions, the average inside temperature in dwellings is within the comfort zone with auxiliary heating energy consumption. H2 reached the maximum level of comfort, H1 and H5 reached the normal level, and H2 and H4 the minimum, according to a study [27].

As seen in Table 2, H1 is the house that shows the lowest coefficient of variation between years in the annual consumption of NG. The values registered during July and August are slightly higher, as expected for the climate harshness of those months. H2, H3 and H4 show coefficients of variation greater than those found in H1. Both H2 and H4 show greater variation in natural gas consumption during July and August within the historical period under consideration. It is possible that the increased exposure of H2, the south orientation of the living room and kitchen-dining room in H4, the habits of owners of H4 who, within the historic period of analysis, retired and stayed more hours at home, and the owner of H2, a more sedentary retired woman, have conditioned consumption.

Energy consumption agrees with the psychometric chart that shows that auxiliary heat is required to enter the comfort zone (Fig. 2). Under less adverse weather conditions, H2 consumed 100% more energy than H1 in order to reach nearly 2°C more. On the other hand, the house has greater exposure to cold winds. It should be noted that H2 had only one user in 90m². An earlier work [26] also showed that the daily average electricity consumption in this dwelling was even lower than the city’s average consumption. This rational use of energy was not associated with the level of income, but to the users’, who expressed to live comfortably under the conditions measured (Table 3). If compared with H1 (same period measured), it can be observed that for almost 4ºC difference in average inside temperature, a decrease in consumption was 10%. According to the Institute for Energy Diversification and Savings of Spain, an increase or decrease of indoor temperature in 2ºC can lead to/result in an energy saving of 10%. Finally, H5 has a perimeter of 39.2 m, a covered area of 48.9m² and a volume of 127.2m³, evidencing little compactness (Ic = 63%), high Exposure factor (Fe =1) and Envelope/floor area factor (FAEP = 3.7). Its Form Factor (FF) is 1.18 and its volumetric coefficient of loss is 4.1W/m³°C, widely exceeding the recommendations of the standard [27]; which, in this case, suggests a value of 2.6W/m^{3 °}C. As seen in Table 2, the house shows one of the highest annual levels of natural gas heating / m² with an average daily heating energy of around 10m³ and 0.20m³/m² the usable area (10.4 m³/inhabitant), doubling that measured in H2 (0.10 m³/m² and 4.5m³/inhabitant). As shown above, H5 and H2 have the normal and minimum comfort level according to a study [28], respectively.

Filippín et al. [25], described and analyzed the monitoring of apartments in multifamily blocks. On each floor, there are 8 units (with 1, 2 or 3 bedrooms) facing NE and SW. Dimensional-morphological and thermo-physical and energetic indicators are shown in Table 2. The buildings have an independent structure of reinforced concrete and hollow ceramic brick walls. The walls and roof have no thermal insulation, and thermal transmittance does not meet the requirements [19]. The FAEP value increases toward the upper floor. Thus, the Global Loss Coefficient ‘G’ partly defined by the thermal resistance of roofs varies between 2.46 and 2.76W/m³ °C and shows higher values than those considered acceptable [27]. On the basis of the analysis of indoor temperature and natural gas consumption during the monitoring period of July-August, the apartment 12 (facing north with its single bedroom facing northwest and with its balcony closed with transparent glass carpentry) showed an average temperature of 21.2ºC (Table 2). In this apartment, the halogen heater was switched on 6h a day (between 9 p.m. and 2 a.m.). Natural gas consumption was 0.13 kWh/day/m². Apartment 15, located at the intermediate level block in the SE corner, consumed 0.74 kWh/day/m² (with an average temperature of 22.2°C). Apartment 23, located in the upper level (above apartment 15) with higher energy loss, consumed 0.84 kWh/day/m² and had an average temperature of 23.7°C. Apartment 18, also on the top floor, and facing north, in the central area with its unique bedroom facing northeast, consumed 0.80 kWh/day/m² with an average temperature of 22.3ºC. Finally, apartment 23, with the input of transferred heat from the lower flat but with greater thermal loss through its envelope, showed the largest consumption per m². The balcony’s carpentry enclosure of apartment 12 enabled a significant reduction in energy consumption, improving comfort conditions during winter. Apartments reached the maximum level of comfort [28]. Following the psychometric chart of climatic conditions, the average inside temperature is within the comfort zone with auxiliary heating energy consumption.

The thermal-energy behavior of a multifamily tower building is studied [26]. The building has two blocks of 11 floors each; each ground has five apartments: A, B, C, D and E. Part of the façade of apartments A and B in the block faces north; part of the façade of apartments C, D and E in the back-block faces south, towards the block’s center (Fig. 4). The technology of T is conventional and commonly used in the country: independent reinforced concrete with a vertical and horizontal envelope without thermal insulation. Aluminum windows with PVC roller shutters, most of them with single glass, as part of the building. Walls and roofs have no thermal insulation, and thermal transmittance does not meet the requirements [19]. Table 3 shows the inside temperature evolution during 15-31 July 2009 of four apartments located on the 11^th floor with mechanical heating and different orientations. In this period, the outdoor minimum temperature reached -5°C, with an external average of 6.3°C. Thermal monitoring during July allowed seeing/determining the influence of orientation and solar resource, in addition, to assess whether inhabitants lived in comfort conditions. Four monitored apartments reached welfare conditions with an average temperature of 20.5, 20.6, 23.9 and 22.4°C, respectively. Following the psychometric chart of climatic conditions, the average interior temperature is within the comfort zone with auxiliary power consumption, reaching the maximum level of comfort as per requirement [28].

It is shown that apartments D and E, facing South and SW, respectively, achieved the highest average weighted temperature: 23.9 and 22.4°C with daily natural gas consumption in heating per square meter of usable area of 0.8 and 2.2 kWh/m²/day. The least energy consumption was found in apartments B and C, with 0.6 kWh/m²/day and 0.7 kWh/m²/day, respectively, with an average weighted temperature of 20.5°C in apartment B and 12.8°C in C. Surely, the sun rays from the north produces a different thermal sensation, allowing inhabitants to live comfortably with lower temperature, and consequently, with lower energy consumption.

To understand the interactions of inhabitants with the built environment, the present analysis was carried out according to inhabitant and availability of useful area per dwelling. These factors deserve their specific analysis since they can mislead the interpretation of quantitative results. In La Pampa, according to the national census, 2.97 inhabitants live per household. Only two of the households register a population of 4 inhabitants (typical family in Argentina). Case studies averaged only 1.85 inhabitants. In addition, in the case studies, most exceeded the minimum surface area suggested by the standards by 30% or more. Of the 13 cases monitored (Table 2), two exceeded by 150% the area of 38 m²/person that we have considered as an area used commonly and B18 is overcrowded with only 17m²/person.

H1 and TB have four inhabitants, with an availability of 35.5 and 32.2 m²/inhabitants, respectively. Dimensional indicators are similar (Table 1). Main differences involve the square meters of envelopes, hence, the values of FAEP are markedly different (H1= 1.53; TB = 0.64). H1 and TB reached the maximum level of comfort as per requirement [28]. According to the psychometric chart of climatic conditions, the average interior temperature is within the comfort zone with auxiliary power consumption. The ΔT indoor-outdoor is +2°C in H1. It should be noted that the energy savings according to average historical natural gas consumption achieved by TB was 15%.

H3, H4, H5, B18 and TE have 2 inhabitants per dwelling. H5 has a G-value much higher than that of the rest of the dwellings (4.10 compared to 2.70 and 3.05). The morphological indicator presents an FAEP 3 times more than the rest. The value of this index, which shows the high exposure of H5, is evidenced here. It is interesting to note that TE, whose area doubles that of H5 and its FAEP is half compared to B18, having the same orientation, shows a similar average of historical heating natural gas consumption (120-122kWh/m²/year). Even with the greater area to heating, the FAEP value seems to be a determining factor in consumption. H4 showed 15% more energy for heating than TE, as evidenced in Table 2. It can be said that facing a bad orientation (south, cold winds), the tower department is more efficient with a considerable decrease of exterior envelope. Main/Major differences lie in the square meters of the envelope, thus, the values of FAEP are largely different.

H2, B15, TC and TD have 1 inhabitant per dwelling. In this case, a comparison of the difference in terms of m2 per inhabitant should be underlined. On the one hand, H2 and TC far exceeded the local estimate of 3 m²/person (H2 = 90 m² and TC =100 m²). Both have the same unfavorable orientation (SW and SE, respectively), equal to G-value and ΔT (Tables 1 and 2). However, the energy heating consumption is particularly dissimilar, where H2 consumes almost four times more than TC in kWh/m²/year. Although the comfort requirements differ since the TC user is in activity and, therefore, does not stay as long at home, compared to H2 who is a retired woman and stays all day in the house (Table 3). On the other hand, we compared B15 and B23 which shared surfaces, orientation and type of construction as well as inhabitants’ lifestyle. The difference in G-value and energy consumption can be attributed to the fact that B23 is in the upper floor with its roof facing the outdoor and a T+1.5C higher than B15.

Regarding the average of energy heating consumption, multifamily buildings (96.3 and 95.4 kWh/m²/year) consume /use 52% less energy than the average single-family dwellings (186.4 kWh/m²/year), under the maximum, normal or minimum level of comfort [29].

Santa Rosa finds itself at a crucial moment in its unplanned historical growth, showing an imbalance in center-periphery density. The strategic plan for the development of Santa Rosa [35] establishes:

• To increase the density in the consolidated central area and to minimize the area of idle land.
• To optimize the consumption of energy according to its different uses.
• To contribute to the complexity of medium-density areas (between 20 and 50 homes per hectare) and to promote typologies of collective housing in 3 or 4 floors, surrounded by open spaces that facilitate nature and recreational activities.
• To regroup homes into a compact design of single volumes, simple and compact, involves ecological and economic strengths (eight units in a multi-family building with three floors reduce land occupation by 66%, compared to an isolated house which reduces the number of materials, infrastructure, etc.).
• To discourage the growth of very-low-density areas (from 4 to 10 houses per hectare).

Based on the comparative synthesis and integration of contextual and social drivers through statistical analysis, an architectural proposal that incorporates the use of resources and energy efficiency strategies is theorized. The proposal must satisfy the total housing deficit: 4300 families, the lowest value in Argentina [36]. The goal is to give an answer to the mitigation of the dwelling deficit and the energy poverty [22]. The authors define the equation Fuel Poverty = energy consumption*Price/Income. According to this equation, the alternative to reduce energy consumption and the energy poverty of the most vulnerable sectors with lowest income is to intervene in urban planning and architectural design. The proposal recommends H and B located in different parts of the city where there is good connectivity and public transport infrastructure and services. These typologies have been selected on the basis that none requires elevators and high maintenance costs [37].

The design, which is energy-efficient, healthy and environmentally sustainable, is associated with the morphological and dimensional indicators from the results of the analysis described. Thus, and in the next phase of this research, design measures are being proposed to reduce the housing deficit and energy poverty, preserving resources.

These examples (Figs. 9 and 10) derive from a logical conclusion following the synthesis analysis as shown in Fig. (5), and seeks to meet the first objective of this article to handing in an architectural proposal considering the use of resources (land, morphology, materials and energy), the lifestyle of the inhabitants (qualitative variables) and the critical analysis of Argentina’s historical problem of housing deficit. Fig. (9) and 10) show different possible sites of the new neighbourhoods with pre-existing infrastructure and service. Proximity and connectivity to main roads, public and private transport have been considered. Fig. (10) shows free land within the urban grid. This is only part of the city; the northern zone, one of the most consolidated one. The strategic plan imposes sanctions for idle urban land. It is important to consider that part of the void land comprises urban barriers like unused railway land and borders of the ring road. The eastern border of the city also represents a great/favourable opportunity for land use with infrastructure and services.

Analysing the housing requirements for the exercise, some aspects must be defined. On the one hand, and according to a study [38] 2.9 inhabitant/dwelling relation [8] was considered to build 60m²/family, reaching a total construction of 258000 m² without galleries or pergolas. On the other hand, complexes of single-family dwelling built by the provincial state in 2005 showed 40 dwelling/ha., and complexes of multifamily dwelling built with loans from a national bank during 1960-1970 showed an 80-96 dwelling/ha rate. Hence, for single-family dwellings (typology H), land could be between party walls of 8 x 20 m depth arranged in city blocks of 170 m and 50 m depth (including the corresponding street surface and sunlight access). In order to this 0,85ha. each city block comprises 40 single-family dwellings aligned by two that help to decrease vertical envelope (Figs. 11 and 12). Plus, the land has convenient N-S orientation, both fundamental bioclimatic strategies. In this case, 90,3 ha would be required for all families.

We propose 2 types of B, including the corresponding street surface (Fig. 12): one of 110 x110 m and another of 50 x170 m with 1.26 ha each block. In this case, 38ha would be required for the 4,300 families divided into 30 blocks with 144 families each. The building has ground floor, 1^st and 2^nd floor without elevator and 8 apartments of 60 m²/floor. By comparing H with B, H uses 58% more of land than B. This is compensated with inhabitant’s access to fertile soil for mini-orchard cultivation and fruit trees, greater privacy, less noise, outdoor areas and the possibility of expanding areas on the ground floor and in height to adapt the design as per the family requirements [39]. Each family in H owns 200m² of land, whereas those in B only own 88.4 m² (North to top).

Considering the recommendations [30] and Sulaiman, Mazzocco and Filippín [40], the height of the vertical envelope has a significant impact on the cost of construction and energy use for acclimatization. An interior height of 2.50 m (2.4 m is the minimum for social houses) is thus proposed here. In agreement with the same authors, the FAEP value is 1.46 and 0.8 for H and B, respectively (smaller enclosure area per m² of useful area). It is important to note that the real envelope does not comply with the hygrothermal conditions of current national regulations.

The strategy of energy conservation is technologically feasible and it would allow achieving thermal transmittance values in accordance with the relevant IRAM recommendation. In winter [19], the study recommends a U-value for walls of 0.30 and 0.80W/m² °C, levels A and B, respectively, according to an outside design temperature of -6°C. Thus, a U-value of 0.47W/m² ºC can be obtained with the addition of a 0.05m thick layer of extruded polystyrene to the exterior protected with coat base, a cementitious material with additive polymer. The standard also recommends a U-value for roofs 0.26 and 0.67W/m² ºC, levels A and B, respectively. Thus, a U-value of 0.38W/m² ºC can be obtained with the addition of a 0.075 m thick layer of extruded polystyrene to the exterior facing roof. Double hermetic glazes and external shutters reduce the U-value around 1.05W/m² ºC. In agreement with the thermal improvement of the entire enclosure, energy loss decreases in both types. Table 6 summarizes all the factors considered in this proposal. The convenient orientation of all H dwellings would promote good energetic behaviour despite having greater external envelope than B. The B dwelling with S orientation, half of the total, has the possibility of accessing natural light with technology available in the market like [41]. In relation to energy saving and, according to the equation [2] provided in the present work and to the FAEP of each typology, heating energy is around 138.4 and 94.4 kWh/ m²/year for H and B, respectively. Consequently, energy saving is 32%, a promising value for a city that imports energy and depends on natural gas in winter (67% of the total annual consumption of the fluid is mainly required for the heating of homes). Regarding the access of natural light (availability of natural daylight) for the illumination of the interior space, both types can use different design components. The southern sector of B could allow entry of natural light from the roof through tubular daylighting (technology available in Argentina). Therefore, the availability of the solar resource in order to heat and illuminate the spaces, in addition to an adequate use by inhabitants of the strategies and design components, would allow reducing energy consumption.

Fig. (5). Study cases in Santa Rosa, La Pampa.

Fig. (6). A methodological diagram that synthesizes the analysis approach.

Fig. (7). Cluster analysis with 11 variables.

Fig. (8). Cluster analysis with 4 variables.

Fig. (9). Possible locations of new neighborhoods with pre-existing infrastructure and service in peri-urban.

Fig. (10). Part of the available urban land with preexisting infrastructure and service.

Fig. (11). Configuration of H in blocks of 50 m x170 m with corresponding street.

Fig. (12). Configurations of B in blocks of 110 x110 m and 50 m x170 m with corresponding street.