Environmental metabolism of educational services. Case study of nursery schools in the city of Barcelona

The environmental analysis of public nursery schools is of great interest since they are crucial in the early education of children and are expected to show high environmental standards. This paper aims to analyse the environmental profile (energy, water and transport flows) of this sector. A sample of 12 public nursery schools belonging to the Scholar Agenda 21 (SA21) of the city of Barcelona were selected given their data quality (eight centres applied to all analysis) to determine their energy and water consumption, as well as the greenhouse gas emissions resulting from energy consumption and transport use. For each centre, energy and water consumption were obtained from bills and surveys were conducted to get data regarding the transport associated with the centre. Results show that, on average, a child consumes 966 kWh of energy (electricity and gas) and 12.9 m3 of potable water every year. Nursery schools with more energy-efficient devices hold lower energy consumption, a trend which could not be found in the case of water and water-efficient devices. Regarding transport, car usage was the flow with highest impact, since it accounts for 69 % of CO2eq emissions, although only 19 % of the children commute by car.


Sustainability and cities
European cities hold around 70 % of the energy (LCSCCI 2011) and 17 % of water consumption (Ecologic 2007). Among the economic activities that take place in urban areas, the service sector represents 60-70 % of the gross domestic product in developed countries (Carpintero 2003) and accounts for 11.3 % of the total energy consumption in the EU-27 (IEA 2008). Therefore, assessing the environmental performance of urban services is crucial.
In this sense, the Agenda 21 (A21), which was created in the Rio Earth Summit (1992) (Our Common Future 1987), consists of a comprehensive action plan that can be applied and adapted globally, nationally and locally (United Nations 2013). Since 1992, it has been embraced by more than 5000 cities all over Europe (ESDN 2014). In the framework of the A21, the Scholar Agenda 21 (SA21) was created to involve educational centres in the environmental improvement of the city. Apart from helping schools to increase their level of sustainability, SA21 seeks to include environmental education in the centres and involve children in different environmental issues. This should not only reflect on the children's behaviour but also on household and centre management, given that teachers and parents play a key role in the development of this initiative. Therefore, an assessment of this service would be useful for making decisions about its resource and energy consumption along with the environmental awareness of the users. Especifically, the focus of this article is the environmental assessment of nursery schools.

Nursery schools and the environment
Previous studies provided environmental information on service buildings such as hotels (Deng 2003;Priyadarsini et al. 2009;Xin et al. 2012;Wang 2012), office buildings (Edwards et al. 2012;Kong et al. 2012), schools (Santamouris et al. 1994;Pons and Wadel 2011), museums (Farreny et al. 2012), commercial malls (Zhisheng et al. 2012), retail parks (Farreny et al. 2008) and hospitals (Moghimi et al. 2014). In the case of hotels, for instance, an average annual electricity consumption of 361 kWh/m 2 was found in Singapore (Priyadarsini et al. 2009), 143.6-280.1 kWh/m 2 in Taiwan (Wang 2012) and 542 kWh/m 2 in Hong Kong (Deng 2003). Moreover, Xin et al. (2012) showed that the building area was highly correlated with the energy consumption in four-and five-star hotels in Hainan (China), while building age and occupancy rate presented weak correlations with energy consumption.
Educational facilities have a high potential to encourage and change the habits of children and their families with respect to the environment. Environmental projects and participation programmes linked to the environment increase the students' interest in these issues (Uittoa et al. 2010). Furthermore, environmental education is important at early life stages because children are already able to understand their surroundings and the effects of environmental changes (Palmer and Suggate 2004). In spite of this potential, environmental education in nursery schools is very limited due to the lack of adequate formation of the teachers (Flogairis 2005). In this context, the environmental assessment of nursery schools will provide useful information about the most relevant environmental impacts and will guide teachers and parents on how to deal with them at school and at home. This approach would close a social and environmental loop that links the environment, families and education.
Regarding previous environmental studies on educational centres, McNichol et al. (2011) analysed the ecological footprint of an early childhood centre in Australia (for children from 2.5 to 5.5 years). The study concludes that the largest contributors to the environmental impacts are food, transport choices and energy use. Another study by Pons and Wadel (2011) conducts an environmental analysis of 200 prefabricated preschool and primary centres in Catalonia. The results show that the phases contributing most to the environmental impacts are the manufacture and use of the building, with similar contributions to the total impact. However, the elements that affected the operational phase were not further analysed. In another case, the energy efficiency of a nursery school in Greece was addressed considering the installation of an experimental green roof, which improved the insulation and reduced the conditioning requirements (Santamouris et al. 2007). A specific assessment of a sample of nursery schools-which in Spain host children aged 0-3 years old-would be of interest since they have some specific environmental impacts due to the early age of children. For instance, thermal requirements are stricter than in other educational services in order to preserve the children's health (Mitchell 2007). As a matter of fact, Oliver-Solà et al. (2013) reported that, in Spain, nursery schools are ranked eighth in a group of 23 different municipal facilities in terms of energy consumption per square metre.
This article focuses on the operational phase of nursery schools. This phase consists of the energy, water and material flows in the daily life of the centres as well as the transport related to the service. Some previous studies (Pons and Wadel 2011;Santamouris et al. 2007) only considered the energy requirements for acclimatisation.
In general, there is a lack of data regarding the metabolism of nursery schools.

Objectives
This article aims to evaluate the environmental profile of the SA21 nursery schools, which is expected to be better than that of conventional nursery schools, and to identify which centres in the sample have the greatest impact. The specific goals of the study consist of the following: to elaborate an environmental profile of a sample of nursery schools belonging to the SA21 in the city of Barcelona through the inputs of energy, water and transport and the output of equivalent CO 2 of the operational phase; to identify the main variables affecting the results, e.g. building area, opening hours and school age; and to evaluate good environmental practices in the centres. Both the scale of the system (i.e. the nursery school) and the expanded system, which includes the neighbourhood, were considered when assessing the transport.

Materials and methods
This section describes the case study selection and the process and methods used for data collection and treatment, as well as the assumptions considered in the analysis.

Sample selection
In the present paper, nursery schools located in the city of Barcelona (Catalonia, Spain) were analysed as a first approach to determining the consumption patterns in Mediterranean regions. To do so, 12 public nursery schools that take care of children aged 0-3 years old were considered (Table 1) so as to represent 5 % of the centres included in the Barcelona SA21 programme.
In general, centres belonging to the SA21 tend to show greater interest in the environment. Originally, 25 nursery schools were willing to participate in the analysis, but only data from 12 of the centres met the quality required to conduct the environmental assessment (8 out of 12 provided information for all the flows). Given that the sample represents 35 % of the nursery schools involved in this initiative (SA21 2015), they might illustrate the consumption patterns of this type of facility. As a result, the findings of this study may be representative for centres belonging to the SA21 in Barcelona, as their inclusion in the programme might have represented more sustainable behaviour given these local environmental initiatives and policies. Moreover, Barcelona is a compact city with a welldeveloped public transport network, which a priori enables sustainable commuting.

Scope of the analysis
The operational phase of nursery schools was addressed in this paper. Fig. 1 shows the inputs and outputs to the system. Energy, water and transport were the only inputs considered, as data on the material goods (e.g. diapers, toys and office material) were not available. Similarly, waste generation could not be estimated as an output, whilst CO 2 emissions deriving from the energy use and transport were accounted for. However, food consumption and waste generation were addressed in terms of best practices in BApplication of efficient devices^. Moreover, the use of eco-efficient devices was also taken into account in order to find a relationship between the consumption patterns and their application in nursery schools. All these elements constitute the operational phase of this system; when dealing with the maintenance activities, cleaning tasks were accounted for in terms of water consumption.

Fieldwork activities
To obtain the necessary data, fieldwork was required in order to have an accurate idea of each nursery school. This procedure was needed given that in some cases, no record of the energy and water consumption was provided by the city council-which is responsible for managing these facilities-and further information was required. Moreover, it was necessary to get acquainted with the habits of the managers, teachers and families in order to assess their level of environmental awareness. This issue consisted of two different aspects. On the one hand, onsite observations helped to determine environmental actions developed inside the boundaries of the nursery school, including the occurrence and use of different types of devices, such as low-energy lighting, aerated and timed faucets, toilets with dual-flush capacity and drip irrigation, as well as ecological food consumption and the separation of waste fractions. A comparison was made between the degree of energy and water consumption and the implementation of efficient devices in the nursery schools under analysis.
On the other hand, surveys were conducted during the visits to gather information regarding the commuting habits of the families. A set of questions was made when parents picked up their children or it was handed in to be completed at home. Families were asked about the means of transport they used to commute from home to the nursery school and whether they lived in the same neighbourhood where the centre was located or not. As a result, an average distance from homes to the school of 20 km (round trip) was considered.

Data from registers
In the case of energy and water, data could be mostly retrieved from registers for the 2008-2009 period that were provided by the city council (since the centres do not hold these data). Energy was represented in terms of electricity and gas consumption, and it was assumed that the largest consumption took place in the built areas, i.e. the interior of the nursery school, including the kitchen, toilets, laundry, office and rooms. The values per unit of area were thus calculated using the built area. In contrast, it was considered that water was also consumed for watering gardens and playgrounds; as a result, the whole area of the nursery school was taken into account when analysing the water-area relationship. It must be pointed out that, in case of having a single educational complex for different school levels, the area considered only covered the nursery school.
The energy and water consumption could not be found for two of the case studies, as the registers were not available. In those cases, the nursery school was part of an educational complex together with primary and secondary schools and data could not be directly obtained. Also for transport, there were no data for two of the centres because the surveys could not be implemented.

Data treatment
After data collection, an environmental analysis was conducted to estimate the carbon footprint. For each type of energy consumption and transport, the equivalent carbon dioxide (CO 2 eq) emissions were calculated to determine the impacts of nursery schools on the environment. Different conversion factors were used to translate each flow to CO 2 eq emissions. For electricity consumption, the Spanish electricity mix for 2011 was used with an emission factor of 366 g of CO 2 eq per kWh of electricity, according to the results obtained applying the CML 2 baseline 2000 method V2.05 (Guinée et al. 2001) and the ecoinvent 2.2 database (Swiss Centre for Life Cycle Inventories 2009). Similarly, 262 g of CO 2 eq per kWh of natural gas were considered. In the case of transport, an emission factor was considered for each mean of transport, which was converted after using the appropriate CO 2 eq value, depending on the source of energy used in each means of transport (Catalan Office for Climate Change 2012; BEA 2011; Biomass energy data book 2011). The emissions for 2011 were considered in the electricity, gas and transport in order to be able to compare the carbon footprint on the same annual basis.
In addition, statistical analyses included descriptive statistics and correlations, which were developed with the aid of PASW Statistics 17 software, from the Statistical Package for the Social Science (SPSS)-developed by EBM (Armonk, NY, USA). The Pearson's correlation index (r) was analysed to determine the strongest and weakest correlations between variables, considering that a p value <0.05 indicates a significant relationship.

Results and discussion
In the following sections, the flows and elements considered (i.e. energy, water and transport) are analysed, with a focus on possible causal relationships between each flow and the features of the nursery school (e.g. area, number of children, building age and opening hours). Table 2 shows a compilation of the data related to the energy and water consumption for each of the nursery schools included in the sample.

Energy and water flows
In this section, analyses were carried out to study water and energy consumption in ten nursery schools using two different reference units: consumption per unit of area (m 2 ) and consumption per child, as observed in Figs. 2 and 3. The assumptions made are presented in BMaterials and methods^.
First, it can be observed that the energy and water flows show similar trends in both the consumption per child and per unit of area. In terms of energy consumption, all the cases are homogeneously distributed between 50 and 200 kWh/m 2 , whereas in terms of water consumption 8 out of the 10 values are within the narrow range of 0.4-0.8 m 3 of water/m 2 (Figs. 2 and 3). Thus, nursery schools present much dispersed results for energy than for water consumption.
There are variations in the behaviours of the centres. The maximum values are fourfold to sixfold greater than the minimum values for both energy and water flows. Considering that they belong to the same geographical area and they carry out the same activities as educative facilities, because they all belong to the SA21, these are important variations. They could be related to different levels of environmentally friendly behaviour in the nursery schools (BApplication of efficient devices^), so differences in the management and the structure of the centres are likely to play an important role.
Drawing comparisons to existing data, the study of Oliver-Solà et al. (2013) evaluated the metabolism from  (2011) show that heating a school requires between 20 and 30 kWh/ m 2 (1000-1700 kWh/m 2 in a time span of 50 years), although the energy source (gas/electricity) remains unknown. In our case, nursery schools used gas boilers and the average gas consumption was 72 kWh/m 2 because of the need for stricter climatic conditions in these facilities. Further explanations to these results were searched for in features of the nursery schools. The relationship between energy and water flows and the area, number of children, number of yearly opening hours and building age was analysed by means of regression statistics (Table 3). By so doing, it was seen that the area is the main factor influencing the performance of the facilities. In all cases, the area refers to the built area except for the correlations with water consumption (BData from registers^). First of all, a significant correlation (p<0.05) was obtained in the water-area relationship with a Pearson's correlation index (r) of 0.680. This result must be related to the need for watering the garden, given that this area is accounted for in the water requirements.
Other findings show that more children are hosted in bigger centres (r=0.740), what is actually an intuitive result. An interesting relationship occurs between the area and the building age. In this case, there is an inverse effect (r=−0.629), which means that older nursery schools are smaller. Here, the historical background of each centre and the facilities that accommodate them play an important role. For instance, centre B was founded in the 1970s in a small room inside a market and centre K is located in an old rehabilitated house from the early 1980s. In contrast, centre J was given a large area in a multi-functional public building of big dimensions seven years ago. This indirectly affects the water consumption, given that older facilities could have lower water requirements. Regarding the remaining variables, no significant correlations were found. The opening hours, for instance, could hardly explain the variations in energy and water consumption given that most of the nursery schools open during the same period of time (Table 1).

Transport and CO 2 emissions
Regarding the transport, the percentages of the mean of transport used by children (and their families) to commute from home to school were determined, as well as the transport-related CO 2 emissions. Table 4 shows the  percentages for the use of each transportation type, CO 2 emissions derived from cars and public transport, and their relative contribution to the transport-related emissions of nursery schools. According to the surveys (BFieldwork activities^), a daily average distance of 20 km (round trip) covered by car or public transport was considered. More than half of the children (61 %) commute on foot or by bicycle, and 20 % use public transport. This means that most children do not produce any CO 2 emissions related to motorised transport. Thus, nursery schools have sustainable mobility habits in general terms. In addition, it can be observed that the average percentage of children using cars is less than one fifth (19 %), which is similar to the average percentage using public transport (20 %). However, most of the emissions derive from the former, being more than 50 % of the emissions in almost all the cases, while the latter produces a smaller part (31 %). Commuting by private car is, thus, the most important factor to take into account to reduce transport-related emissions, even though it represents a small percentage of commuters.
However, great differences can be observed in the percentages of each means of transport. Sustainable transport patterns (i.e. a higher percentage in the use of public transport, bicycle and on-foot) can be related to factors that shape the neighbourhood where the nursery school is located, such as topography, average economic income, population density or the presence of adequate infrastructure (e.g. train station and bus stop). In addition, personal decisions play a key role in these results. The school-home distance and the socio-economic position of the families can be associated with the parents that decide to commute by car or prioritize the education of their children in centres that are not so close from their home. Therefore, both the structure of the neighbourhood and personal decisions should be further analysed in future studies.
Nevertheless, CO 2 not only results from use of fuel when commuting but also from the energy use in the facilities. Therefore, the emissions of both systems were compared in each nursery school, as illustrated in Table 5. Globally, transport emissions account on average for approximately 159.7 kg CO 2 eq per child, representing 40 % of of the total emissions. This result must be taken into account in the management of these educational facilities in order to achieve a reduction in the environmental impacts by actions taking place outside the boundaries of the centre. This goal could be achieved by implementing more environmental education programmes, but nevertheless, the location of the centre and the place of origin of the children are also fundamental.
In addition, one can see that the energy use in the building is an important vector, since it accounts for more than 70 % of the emissions in several nursery schools. Indeed, previous articles on service buildings focused on energy consumption, which was considered one of the main contributors to the environmental impacts of the system Santamouris et al. 1994;Zhisheng et al. 2012). Therefore, energy saving must be another important objective of the centres and it can be achieved by means of different alternatives such as reducing the energy demand or applying efficient devices. The latter were analysed in order to determine their effect on the saving task, together with the application of other environmentally friendly actions (BApplication of efficient devices^).

The environmental profile of a child
To establish the environmental profile of a child, it was determined that on average, each of them requires 966.3 kWh of energy and 12.9 m 3 of water every year to fulfil his or her educational needs, which means, in terms of daily consumption, 5.03 kWh and 67.3 L of water. As observed in BEnergy and water flows^, data related to energy consumption could be compared with existing analyses while there is a lack of studies on the metabolism of nursery schools for other flows. Moreover, the centres present an average carbon footprint of 420 kg of CO 2 per child. Oliver-Solà et al. (2013) also focused on the school system and found that, on average, a student consumed 172.6 kWh annually, which is approximately one fifth of the calculated value for nursery schools. This highlights the importance of the nursery school system, which was found to be one of the greatest energy consumers in the education sector. For water, a comparison could be drawn in relation to the daily consumption of an individual throughout a 24h period. An average inhabitant of Barcelona consumes between 130 and 200 L of water every day (CWA 2012). The lower consumption per child (67.3 L/day) could be explained considering age and habits. In general, children under 2 years old are not able to use toilets because they wear diapers, and this fact reduces the number of children consuming water. Furthermore, children do not spend the whole day at the nursery school, but around a third of it.
Application of efficient devices Efficiency also contributes to the level of consumption of the facilities, and it should be a subject of study. Thus, the infrastructure of the nursery schools was analysed to identify possible relations between the level of the facilities' efficiency and the noted environmental flows. Table 6 shows the degree to which nursery schools adopted measures to save energy and water and manage their waste. This degree has been symbolised using crosses: B+^, B++^and B+++^represent low, medium and high application of a specific device, respectively. Details of the ranks considered for each measure are found in the legend. The measures considered were: the application of devices for low-energy lighting, time/ aerated faucets and toilets with dual-flush capacity; the presence or absence of solar panels and drip irrigation; and the number of waste fractions being separated. Additionally, the ratio of ecological food consumption was also assessed. According to the food regulations of nursery schools , at least 5 % of the food consumed at school must have an ecological background. It was observed that low-energy lighting, solar panels and toilets with dual-flush capacity present a low degree of implementation. Drip irrigation and separated waste fractions have been broadly adopted (+++) by over half of the centres. Timed or aerated faucets and ecological food have different degrees of implementation among nursery schools. Although, as mentioned above, 5 % of the food must be ecological at every nursery school, four centres did not reach this minimum. In the so-called green schools framework, this fact must be taken into account, as it is compulsory but not accomplished by 26 % of the centres.
According to the total punctuation marks in the water and energy categories (i.e. adding all the + symbols) and the water and energy consumption per unit of area in each nursery school, it could be determined whether applying efficient devices had an impact on the total consumption. In the case of energy, this relationship was positive because nursery schools with a higher use of efficient devices (e.g. E, F, H, O) generally presented a lower level of energy consumption (88 kWh/m 2 ) than the others (131.3 kWh/m 2 ).
This relationship is not so clear in the case of water. An approximation could not be developed because there were some difficulties in the determination of open areas. The garden/playground uses were heterogeneous with combinations of vegetable gardens, areas with regional plant species and paved playgrounds. Thus, finding a relation held great difficulty, and the use of these types of water-efficient measures could not be directly associated with lower water consumption. In addition, these results could be in line with previous findings of this study. Given the complexity in finding correlations between the energy and water consumption and the building features (i.e. size, opening hours, etc.), it is also controversial to find accurate relationships with the use of efficient devices. It is believed that future investigations should focus on these assumptions, given that it is of great importance to manage and analyse the real efficiency of these facilities. Still, this paper provides new information on the performance of nursery schools. The novelty of this approach lies on the interrelation with local environmental policies, the effect on educational services as public infrastructures and a positive social and environmental behaviour adopted by teachers, families and children.

Conclusions
The environmental analysis of the case study nursery schools, considering the inflows of energy, water and transport, shows that this system has an important demand for resources. For instance, around 123 kWh/m 2 of energy are annually consumed on average, which is significantly higher compared to educational facilities in general (87 kWh/m 2 ) or to sports facilities (90 kWh/m 2 ) (Oliver-Solà et al. 2013). Thus, it is confirmed that these facilities have a high impact on the environment and its reduction is desirable.
Attempts to reduce the energy consumption through the improvement of a centre management should be adapted to its specific conditions, provided that there is considerable variability among the annual consumption of the nursery schools. The highest value is four times greater than the lowest (from 4.92 to 36.2 kWh), and the centres are homogeneously distributed along this range. Previous articles on service buildings also show variability regarding the energy consumption (Farreny et al. 2012;Xin et al. 2012). Larger nursery schools should be a priority when reducing water consumption in the centres, given that it is highly correlated with the building area (r=0.68). However, there is a narrow range of water consumption in the centres (between 0.4-0.8 m 3 of water/m 2 ), which might indicate that these values are already low and that there is small room for improvement.
Regarding the transport, the reduction of the car use must be a priority. This mean of transport represents an average of 69 % of the transport-related CO 2 eq emissions, in spite of being used by only 19 % of children. This vector presents a high potential for environmental savings, which could be achieved through the promotion and improvement of sustainable means of transport. However, this might exceed the competence of the nursery school management and require the cooperation of both families and local authorities.
The application of energy-saving devices should be promoted since nurseries with a higher use of these devices show lower energy consumption (88 kWh/m 2 ) than the rest (131.3 kWh/m 2 ). However, the presence of these devices might be a consequence of a higher environmental awareness and the savings might be at some point a result of a more conscientious management.
Future studies should focus on the development of effective measures that reduce the environmental impacts of nursery schools and services in general. These measures should especially address the energy and transport vectors, as they have shown the greater impacts as well as the highest potential for improvement. Furthermore, measurement methods should be developed in order to check the effect of these measures on the resource consumption of the centres. In the case of energy and water, the centres are unable to know its consumption, given that the expenses are paid by the city council. Thus, reporting these consumptions to the employees might help with the monitoring and the awareness.