Water consumption in public schools: a case study

The objective of this study is to define a consumption indicator (CI) for water that can be used as a reference for developing water conservation plans at public schools. The methodology followed consists of a historical water consumption data survey of all schools in Recife, Brazil, a registration survey of school buildings, the calculation of consumption indicators for the period 2012–2015, and finally, the calculation of reference indicators. The results obtained indicated a reference range for the CI of 13.0± 2.0 L/student/day for water considering the set of all school typologies, with a 95% confidence level. The analysis of the schools in groups, classified into four typologies, led to the following reference ranges: 11.5± 2.0 L/student/day for regular schools, 13.5± 3.5 L/student/day for extended period schools, 22.0± 6.0 L/student/day for full period schools, and 18.5± 6.5 L/student/ day for technical schools. Through the use of a consumption indicator, schools with a limited supply of potable or above average water can be identified and specific actions can be developed to achieve a sustainable use of water in the school environment. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/). doi: 10.2166/washdev.2019.074 s://iwaponline.com/washdev/article-pdf/9/1/119/613285/washdev0090119.pdf Luiz Gustavo Costa Ferreira Nunes (corresponding author) Anna Elis Paz Soares Willames de Albuquerque Soares Simone Rosa da Silva Universidade de Pernambuco Escola Politecnica de Pernambuco, R. Benfica, 455 Madalena, Recife – PE, 50720-001, Brazil E-mail: l.gustavo.nunes@hotmail.com This article has been made Open Access thanks to the generous support of a global network of libraries as part of the Knowledge Unlatched Select initiative.


INTRODUCTION
Population growth, increased urbanization, and industrialization have led to an increase in the demand for drinking water, while also contributing to the degradation of natural resources and compromising the availability and quality of freshwater sources, especially when occurring without planning or organization. Water must therefore be con- Water conservation planning is normally created so that the population can use water in a rationed manner, either through voluntary or non-voluntary measures, many times as a result of an imminent water crisis (Mini et al. ; conserved conditions of their sanitation systems, leading to a high degree of water loss. These factors derive from many causes, including the lack of consumer awareness regarding environmental conservation, as well as the absence or inefficiency of a maintenance system techniques are only apprehensively exploited by the population, such as the use of rainwater, the reuse of gray water, and even the use of water-conserving equipment. Soares et al. () analyzed water consumption at two public state schools in Recife and found that, although the per capita consumption indicator (CI) was relatively low (less than 5 L/student/day), water losses due to leaks were high and the schools did not use water-economizing equipment. These results have aroused interest in understanding the pattern of water consumption in public schools across the region.
This study presents a tool to characterize the water consumption pattern at public schools through the quantification of the Consumption Index. With it, one can identify schools with a limited supply of drinking water or those that are above average, for example. Both cases require attention in the search for the sustainable use of water in the school environment.

WATER CONSUMPTION INDICATORS
In order to meet the guidelines of a Water Conservation Program (WCP), priority must be given to knowing the characteristics that influence water consumption in buildings (typology, construction process, leaks, population, and climate conditions), and the definition of a consumption profile for a consumer agentthe CI (Makki et al. ).
The CI is a tool to characterize and monitor water consumption in buildings, and is very useful for setting targets and implementing water conservation measures.
The consumption of water in buildings is normally estimated through a CI expressed as a volume of water divided by the number of consumer agents, which is the most representative variable of water consumption in the system. The CI is therefore obtained according to the typology of the building, such as L/person/day in a residential building or an office building, L/bed/day at hospitals, and L/student/day at schools. These values can be used as references when assessing the impact of the reduction of water consumption on the implementation of a WCP In the literature, CI values for projects can be found, which are indicators adopted to develop cold water building installations that assure the proper functioning of the sanitary equipment. These CI values can be found in textbooks or government regulations. However, the actual water consumption of a building should be measured and the actual CI may be calculated by relating the volume of water consumed within a certain period of time to the number of consumer agents. A value that represents the real consumption of a building can be obtained, reflecting its particular characteristics as well as the consumption patterns of its population. Through the identification of the actual CI, it is possible to create benchmarks that characterize an ideal consumption situation. Generally, to achieve the benchmark, water conservation plans must be laid out that take into account, among other measures, the use of water-conserving equipment and user consumption habits. Table 1 lists the water consumption indicators for schools from around the world. UNICEF (), in its standards for water, sanitation, and hygiene (WASH) in schools, considers that sufficient water must be available at all times for drinking and personal hygiene, and, when applicable, for food preparation, cleaning, and laundry, considering an indicator of 5 L/person/day for all school children and staff at day schools, and 15-20 L/person/day for boarding schools. In addition, water for flushing toilets must be considered as follows: flushing toilets 10-20 L/person/day for conventional flush toilets; 1.5-3 L/person/day for pour-flush toilets; and 1-2 L/person/day for bidets. These quantities are for day schools and should be doubled for boarding schools. The UNICEF indicators do not consider the use of water conservation equipment or technology; therefore, they are not a benchmark for the rational use of water in schools. They are a CI for school building projects in developing countries.
The city of Sydney, Australia, has already set benchmarks for primary schools based on the use of water conservation techniques. They indicate a range of 3-9 L/student/day to be normal consumption. Indicators below this reference range are considered to be very low consumption, and values above need to be investigated (Sydney Water ).

METHODOLOGY
The methodology employed in this study is subdivided according to the following topics: (a) survey of historical data; (b) calculation of drinking water consumption The students are able to study during one or two shifts.
Each building has very specific working hours, which can involve from two to three shifts per day.
A preliminary analysis was necessary regarding the were deficient for some years at some schools.

Calculation of drinking water consumption statistics
The historical water consumption data for the period from 2012 to 2015, obtained from the local concessionaire, were organized into an Excel spreadsheet for analysis.
Each building received a code from 01 to 180. As was done in other studies carried out across many cities, including Recife, the atypical months corresponding to the periods of vacations were disregarded in this calculation (January, June, July, and December).
The consumption of drinking water is considered here only from the water supplied by the local water company.
In the case of the existence of any alternative supply source such as a well or water trucks, this additional availability was not taken into account. However, the service of water trucks is very uncommon within the city of Recife due to the associated high costs. Regarding supply from wells, no monitoring is generally carried out for the values captured, making it impracticable to obtain data related to water consumption from wells.
For each year analyzed (2012, 2013, 2014, and 2015), the annual measurements of monthly water consumption were calculated for each school. Therefore, each building analyzed can represent up to four samples.

Calculation of water consumption indicator
The CI in L/student/day was calculated according to Equation (1), which related the water consumption in the building to the consumer agents for a given time period.
where Cm is the mean consumption in m³/month, calculated for each year; NA is the number of consumer agents; and Dm is the number of weekdays per month.
For the number of consumer agents (NA), the number of students was used, and the period in which these students were present at school was analyzed. The teachers and remaining employees were not considered since the CI is a reference of consumption as function of a specific agent.
In schools, the reference normally used is the number of students. For the number of weekdays, an average of 22 per month was considered.
As previously mentioned, the students can attend the school at different hours. This study balanced all of the students by considering them to all have studied during a single shift, doubling them in cases where they remained in the school for two shifts, according to the criterion adopted by Nunes ().
Therefore, using this criterion, the number of high school students at a full-period EREM is multiplied by a factor of 2, since they remain at school for two shifts; while those at extended EREMs are multiplied by factor of 1.6, because they study for two shifts during 3 weekdays.

Calculation of the reference indicators
The search for a reference CI for the state public schools in Recife took into consideration all of the CIs calculated for all of the schools during the period studied and was organized in such a way that each indicator could be treated as an independent value. Subsequently, the mean CI was calculated as well as the respective standard deviation in order to calculate the reference range of the samples. The limits of the reference range were defined as the mean CI plus or minus a confidence interval of 95%.
The data obtained are plotted in Figure 1 relating the school code (01 to 180) and the respective CI(s) in L/student/day. A general reference range was thereby obtained as a set of all of the samples.
In addition to the general analysis of the CI dataset, a second analysis was carried out that took school typology into consideration. In order to consider the particularities of the distinct school types, the samples were divided into four subgroups: full-time EREM, extended EREM, technical school, and regular school. Reference ranges for the CIs for each of these school types were established, using the data from each subgroup separately.

RESULTS AND DISCUSSION
The water CI for the total set of schools varied from a low of 1.0 L/student/day at School 159 to a high of 103.5 L/student/day at School 06. The reference range for the water CI calculated for the set of all school typologies in Recife was 13.0 ± 2.0 L/student/day, with 95% confidence, using data from 140 schools.
Values above 200 L/student/day and below 1 L/student/day were not considered in the study because they are likely to not represent an atypical situation and should be thoroughly investigated on a case-by-case basis. These consumptions may have occurred because of water meter measurement errors, the presence of alternative water supply sources, or even the occurrence of large leaks. in this study and discovered that more than 54% of them had no hydraulic equipment with water-conserving technology.
Out of the sample universe, 54 were found to be within the calculated reference range. Among the 140 buildings studied, none of them presented a CI within the reference range for all of the years studied. Among the particularities, School 100 is located in a rented building, where all the sanitary installations for student use are in a precarious condition, jeopardizing water supply. In contrast, a family home exists on the school building grounds that is supplied by the same distribution branch as the school, meaning that all consumption in the residence is counted as part of the school's consumption. This information was obtained in the field in 2014, illustrating the 5% error expected in this study.
Other schools, such as School 68, presented an indicator within the reference range during only one of the four years. This unit is part of the technical school subgroup, which has a daily routine similar to an EREM, where students study full time. Another difference is that the institution is provided with private support and consequently receives preventive maintenance, effectively correcting any leaks discovered. Simultaneously, analyses were carried out for each school subgroup, shown in Table 2.
The analysis of the group of regular schools led to a reference range for the CI of 11.0 ± 2.0 L/student/day. Out of the total 313 CI values analyzed, 181 samples were below the lower limit and 76 samples were above the higher limit. For this typology, none of the schools had all four years within the reference range.
By comparing the reference range obtained for the total set of schools, we can observe that the consumption habits at regular schools make sense with this lower indicator, since at regular schools, students are normally not provided with lunch. In addition, the student body remains at school for only one shift, making the use of showers unnecessary.
Some schools do not even have this type of equipment.
The group of schools from the extended EREM and full period EREM types showed results very similar to those of the regular schools. The reference range of the CI for extended EREM schools was 13.5 ± 3.5 L/student/day, with 14 CIs above the higher limit and 26 below the lower limit out of the total of 53. The reference range for the full period EREM schools was 22.0 ± 7.0 L/student/day, with a universe composed of 59 samples, where 15 fell above the higher limit and 30 below the lower limit.

CONCLUSIONS
The focus of this study was the quantitative assessment of drinking water consumption at the different types of state public schools in Recife, restricting the analysis to historical data from the past few years. Qualitative aspects of water use were not considered. The criterion used to analyze schools (according to the typology) allowed for the consideration of a few important factors for water consumption, such as amount of functioning shifts at the school, period of time students spend in the school, and the preparation of meals.
This study obtained reference values for water consumption at public state schools located in the city of Recife, categorized by the type of school building. It is an important contribution to assessing the general scenario of the city as well as to represent a parameter for the surrounding municipalities or those with similar physical, cultural, social, and economic characteristics.
The reference ranges should not be considered as benchmarks, but as a representation of the actual current scenario in the public schools. These data should be used to prioritize water conservation actions at schools to achieve rational and sustainable water use in these buildings and to ensure that the minimum health requirements are being considered.
For the schools that consumed water above the reference range limit, actions to manage water demand should be considered, such as replacement of conventional equipment by water economizing equipment, and the use of alternative sources such as rainwater and gray water reuse.
The consumption of water below the reference range limit at schools should also be investigated. One of the possible factors for low water consumption values is restriction of the school's water supply. In these situations, the study of possible solutions to increase the water supply from other alternative supply sources is recommended, such as capture of rainwater and/or well water.
This research should continue and further studies should assess other indicators, especially indicators regarding leaks and losses, as well as assessment of consumer perception regarding water consumption in schools, since consumer behavior reflects directly on consumption. It is also important to develop extension activities along with the school community, seeking to create awareness for this target public (students, teachers, and employees) regarding the importance of conscious behavior with respect to water conservation, consumption, and basic hygiene.
In addition to educational campaigns, other actions to manage water demand in buildings need to be developed, such as preventive maintenance of hydro-sanitary equipment and systems, development of studies for the use of new alternative water supply sources such as the harvesting of rainwater, and development of feasibility studies for the replacement of conventional hydro-sanitary equipment with water-economizing models. It would then be possible to elaborate and implement propositions for water conservation planning at public schools.