ARTICLE IN PRESS
Building and Environment 42 (2007) 1654–1666 www.elsevier.com/locate/buildenv
Rainwater tank capacity and potential for potable water savings by using rainwater in the residential sector of southeastern Brazil Enedir Ghisi, Diego Lapolli Bressan, Maurı´ cio Martini Laboratory of Energy Efficiency in Buildings, Department of Civil Engineering, Federal University of Santa Catarina, Floriano´polis-SC, 88040-900, Brazil Received 25 August 2005; received in revised form 17 January 2006; accepted 17 February 2006
Abstract Rainwater has been used in many countries as a way of minimising water availability problems. In Brazil, it has been reported that the potential for potable water savings by using rainwater may range from 48% to 100% depending on the geographic region. In southeastern Brazil, water availability is about 4500 m3 per capita per year, but it is predicted to be lower than 1000 m3 per capita per year from about 2100 onwards. The main objective of this article is to evaluate the potential for potable water savings by using rainwater in 195 cities located in southeastern Brazil. Rainwater tank sizes are also assessed for some cities in order to evaluate the ideal tank capacity as a function of potable water demand and rainwater demand. Results indicate that average potential for potable water savings range from 12% to 79% per year for the cities analysed. Ideal rainwater tank capacities for dwellings with low potable water demand range from about 2000 to 20,000 litres depending on rainwater demand. For dwellings with high potable water demand, ideal rainwater tank capacities range from about 3000 to 7000 litres. The main conclusion drawn from the research is that the average potential for potable water savings in southeastern Brazil is 41%. It was also concluded that rainwater tank capacity has to be determined for each location and dwelling as it depends strongly on potable water demand and rainwater demand. r 2006 Elsevier Ltd. All rights reserved. Keywords: Potable water savings; Rainwater tank capacities; Rainwater usage in southeastern Brazil
1. Introduction Water availability has become a matter of concern in many countries as their populations increase at a fast rate. In order to ease such a problem, rainwater usage has been acknowledged to promote significant potable water savings in buildings. Reports on this can be found in Herrmann and Schmida [1], Coombes et al. [2], Fewkes [3], Marinoski et al. [4], Appan [5], Handia et al. [6], Li and Gong [7], amongst others. As for Brazil, it has already been reported that the potential for potable water savings by using rainwater varies greatly depending on the geographic region, i.e., it varies from 48% in the southeast region to 100% in the north region as shown in Table 1 [8]. In the south region, a potential for potable water savings of 82% was obtained. However, by performing a detailed analysis in the state of Corresponding author. Tel.: +55 48 3331 5185; fax: +55 48 3331 5191.
E-mail address:
[email protected] (E. Ghisi). 0360-1323/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2006.02.007
Santa Catarina, which is located in the south region, a potential of 69% was obtained [9]. This indicates that detailed analyses may provide more accurate results. Therefore, this article focuses on a detailed analysis of the potential for potable water savings by using rainwater in southeastern Brazil. As for rainwater tank sizing, its accurate determination depends on daily rainfall data and also on potable water demand, rainwater demand, and roof area. There are some reports on rainwater tank sizing that have been published; some examples can be found in Refs. [10–12]. As this subject has not been fully studied in Brazil, this article also assesses rainwater tank sizing for dwellings located in southeastern Brazil. 2. Objective The main objective of this article is to evaluate the potential for potable water savings by using rainwater in the residential sector of 195 cities located in the southeast
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
1655
Table 1 Average rainfall, potable water demand and potential for potable water savings by using rainwater per geographic region of Brazil Region
Average rainfall (mm/year)
Potable water demand (l/capita per day)
Potential for potable water savings by using rainwater (%)
North Northeast Southeast South Central-West
2182 1146 1362 1615 1540
88 97 158 117 120
100 61 48 82 74
Source: based on Ghisi [8].
Fig. 1. Map of Brazil with location of the southeast region and map of (a) Minas Gerais, (b) Espı´ rito Santo, (c) Rio de Janeiro, and (d) Sa˜o Paulo, with location of the 195 cities (grey colour) included in the analysis.
3. Southeast region of Brazil In order to achieve the objectives specified above, rainfall data, potable water demand, population and number of dwellings supplied with potable water in each city included in the analysis were obtained. It was intended to consider all of the cities in southeastern Brazil, but there were no rainfall data available for all of them. The geographic region is composed of four states, i.e., Minas Gerais, Espı´ rito Santo, Rio de Janeiro, and Sa˜o Paulo. Fig. 1 shows a map of Brazil indicating the location of the geographic regions and also the location of the 195 cities considered in the analysis over the southeast region. 3.1. Water availability in southeastearn Brazil Ghisi [8] evaluated the water availability in the southeast region up to the year 2100 and concluded that it will be lower than 1000 m3 per capita per year from 2094 onwards. Fig. 2 shows the results for the four states located in the
100000 Predicted water availability (m3 per capita/year)
region of Brazil. Rainwater tank sizing as well as correlation between the potential for potable water savings and either potable water demand or rainfall are also investigated. Such correlations are deemed appropriate as they may allow for a simple way of estimating the potential for potable water savings over the southeast region.
10000
1000
100 1900
Minas Gerais
Espírito Santo
Rio de Janeiro
São Paulo
1950
2000 Year
2050
2100
Fig. 2. Predicted water availability for the four states over the period 1900–2100.
southeast region analysed separately over the period 1900–2100. Water resources were obtained from ANA [13] and population, from IBGE [14] for the period 1900–2000. The water availability predictions from 2000–2100 were estimated considering the average growth rate of population over the period 1991–2000. It can be noticed that the states of Sa˜o Paulo and Rio de Janeiro have a water availability of about 2000 m3 per capita per year at the moment, which is considered very low by the United Nations Environment Programme (UNEP) [15]. Predictions indicate that water availability will be lower
ARTICLE IN PRESS 1656
E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
than 1000 m3 per capita per year from about 2050 in Sa˜o Paulo, from 2060 in Rio de Janeiro, and from 2100 in Espı´ rito Santo. Such a water availability is considered catastrophically low by UNEP [15].
4.2.3. Number of dwellings supplied with potable water The number of dwellings supplied with potable water was estimated by using Eq. (2). ND ¼
4. Methodology The methodology applied to estimate the potential for potable water savings in southeastern Brazil is similar, but not identical, to the one applied by Ghisi et al. [9] when focusing the state of Santa Catarina, southern Brazil. As for the methodology used to determine the ideal rainwater tank capacity, it is based on computer simulation that takes into account daily rainfall data. 4.1. Rainfall data Daily rainfall data were obtained from the Department of Water and Electricity of Sa˜o Paulo [16] for 163 cities in the state of Sa˜o Paulo. The available data ranged from 1931 to 2000, but did not cover the same period for all the cities and in some cases they were not complete. When there were missing data, they were assumed to be zero. Such an assumption implies that the potential for potable water savings may be higher than the figures obtained for some cities. As for rainfall for the states of Minas Gerais, Espı´ rito Santo and Rio de Janeiro, only monthly data were obtained; and these were available for a few cities for the period 1961–1990 [17]. Twenty cities were considered in Minas Gerais, four in Espı´ rito Santo and eight in Rio de Janeiro. 4.2. Potential for potable water savings The potential for potable water savings by using rainwater in each of the 195 cities was calculated as follows. 4.2.1. Population supplied with potable water and potable water demand The number of people supplied with potable water and the potable water demand in each city were obtained for the year 2002 from the National Database on Sanitation, available online [18]. Arithmetic averages were performed to determine monthly figures. 4.2.2. Number of people per dwelling The number of people living in each city and the number of dwellings were obtained from IBGE [14] for the year 2000. Thus the specific number of people per dwelling was estimated by using Eq. (1). PC , (1) NDC where PD is the number of people per dwelling, PC is the number of people living in the city, and NDC is the number of dwellings in the city.
PD ¼
NP , PD
(2)
where ND is the number of dwellings supplied with potable water, NP is the number of people supplied with potable water per month (as given by the National Database on Sanitation [18]), and PD is the number of people per dwelling. 4.2.4. Total roof area From all dwellings located in southeastern Brazil, 86.7% on average are houses and 13.3% are flats located in multistorey residential buildings [8]. In order to obtain an average roof area in each city, the same procedure as suggested in Ghisi [8] was used. Therefore, an area of 85.00 m2 was assumed for houses and 3.75 m2 per person for flats. A weighted average roof area per dwelling was then determined by using Eq. (3). RA ¼ H 85:00 þ F PD 3:75,
(3)
where RA is the weighted average roof area per dwelling in each city (m2), H is the percentage of houses in each city (non-dimensional), F is the percentage of flats in each city (non-dimensional), and PD is the number of people per dwelling in each city. The total roof area in each city was then obtained considering only the population supplied with potable water. It was determined by using Eq. (4). TRA ¼ RA ND;
(4)
where TRA is the total roof area in each city (m2), RA is the weighted average roof area per dwelling in each city (m2), and ND is the number of dwellings supplied with potable water. 4.2.5. Volume of rainwater The monthly volume of rainwater that could be harvested in each city was determined considering monthly rainfall data, the total roof area, and a runoff coefficient of 0.80. Such a runoff coefficient indicates a loss of 20% of the rainwater that is discarded for roof cleaning and evaporation. Thus, the volume of rainwater that could be harvested in each city was determined by using Eq. (5). VR ¼
R TRA Rc , 1000
(5)
where VR is the monthly volume of rainwater that could be harvested in each city (m3/month), R is the monthly rainfall in each city (mm/month), TRA is the total roof area in each city (m2), Rc is the runoff coefficient (non-dimensional), and 1000 is the conversion factor from litres to m3.
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
4.2.6. Potential for potable water savings The monthly potential for potable water savings was determined for each of the 195 cities by using Eq. (6). VR , (6) PWD where PPWS is the potential for potable water savings in each city (%), VR is the monthly volume of rainwater that could be harvested in each city (m3/month), and PWD is the monthly potable water demand in each city (m3/month). PPWS ¼ 100
4.3. Rainwater tank sizing
1657
where WDD is the daily potable water demand per dwelling in each city (litres/day per dwelling), DWD is the daily potable water demand per capita in each city (l/capita per day), and PD is the number of people per dwelling (people/dwelling). 4.3.5. Roof area The average roof area per dwelling in each city was estimated by using Eq. (3), shown in Section 4.2.4. 4.3.6. Volume of rainwater harvested The daily volume of rainwater that could be collected in each dwelling was determined by using Eq. (9).
In order to investigate rainwater tank sizes, an analysis was performed over nine cities located in the state of Sa˜o Paulo. The selected cities enclose samples with low, medium and high potable water demand. Cities in the other states were not considered as there were no daily rainfall data available.
DVR ¼ DR TRA Rc ,
4.3.1. Computer programme The rainwater tank sizing analysis was performed by using a computer programme called Neptune [19]. Input data are daily rainfall, daily potable water demand per capita, number of people per dwelling, roof area per dwelling, rainwater demand per dwelling, rainwater tank capacity and runoff coefficient. Such a programme determines the potential for potable water savings according to specific rainwater tank sizes. Therefore, the rainwater tank capacity is chosen by the user according to the potential for potable water savings determined by the computer programme. Procedure for obtaining input data are described in Sections 4.3.2–4.3.5; and a brief description of the procedure used by Neptune is presented in Sections 4.3.6–4.3.9.
4.3.7. Volume of rainwater in the tank To estimate the volume of rainwater available in the tank, the daily rainwater demand in each dwelling was subtracted from the daily volume of rainwater that could be harvested in each dwelling as shown in Eq. (10).
4.3.2. Daily potable water demand per capita Daily potable water demand per capita was estimated by using Eq. (7). AWD , (7) 365 NP where DWD is the daily potable water demand per capita in each city (litres/capita per day), AWD is the annual potable water demand in each city (litres/year), and NP is the number of people supplied with potable water in each city. DWD ¼
4.3.3. Number of people per dwelling The number of people per dwelling was determined as shown in Section 4.2.2. 4.3.4. Daily potable water demand per dwelling The daily potable water demand per dwelling was estimated by using Eq. (8). WDD ¼ DWD PD;
(8)
(9)
where DVR is the daily volume of rainwater that could be harvested in each dwelling (litres/day per dwelling), DR is the daily rainfall in each city (mm/day ¼ litres/m2 per day), TRA is the total roof area in each dwelling (m2), and Rc is the runoff coefficient (non-dimensional).
VRT ¼ VRT pd þ ðDVR P WDDÞ pRainwater tank capacity;
ð10Þ
where VRT is the daily volume of rainwater in the tank (litres/day), VRTpd is the volume of rainwater in the tank from the previous day (litre), DVR is the daily volume of rainwater that could be harvested in each dwelling (litres/ day per dwelling), WDD is the daily potable water demand per dwelling in each city (litres/day per dwelling), and P represents an incremental percentage ranging from 0 to 100% at increments of 10%, so that the product P WDD represents the rainwater demand in the dwelling. The rainwater demand is taken from 0% to 100% of the potable water demand as there is no ultimate data on this in Brazil and also in order to evaluate such an influence on the rainwater tank capacities. In the context of this article, rainwater demand means the water end-uses in which there is no need for potable water, such as toilet flushing, car and clothes washing, garden watering, etc. In the UK, for example, the rainwater demand in dwellings is about 46% of the potable water demand [20], 40% in the USA [21], 51% in Switzerland [21], and 55% in Colombia [21]. 4.3.8. Potential for potable water savings The potential for potable water savings for the nine cities was estimated by using Eq. (11). Pd i¼1 RC PWS ¼ 100 , (11) WDD d where PWS is the potential for potable water savings over the period being analysed (%), RC is the rainwater that is
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
4.3.9. Rainwater tank capacity The procedure described above was performed for incremental rainwater tank capacities ranging from 1000 to 30,000 litres at increments of 1000 litres. The ideal rainwater tank capacity was assumed as the one in which the potential for potable water savings increased less than 0.5% by increasing the rainwater tank capacity in 1000 litres. 4.3.10. Correlations Correlations between rainwater tank capacities and either potable water demand or rainwater demand were also investigated for the nine cities in the state of Sa˜o Paulo. As for the influence of the roof area on the rainwater tank capacities, this was investigated by ranging such an area from 50% to +50% for three cities. 5. Results 5.1. Rainfall data
5.00 Number of people per dwelling
actually consumed in the dwelling, d is the number of days in the period being analysed, and WDD is the daily potable water demand per dwelling in each city (litres/day per dwelling).
4.00 3.00 2.00 1.00
Minas Gerais
Rio de Janeiro
Espírito Santo
São Paulo
0.00 1
21
41
61
81 City
101
121
141
161
Fig. 4. Number of people per dwelling over 195 cities in Minas Gerais, Rio de Janeiro, Espı´ rito Santo, and Sa˜o Paulo.
350 Potable water demand (litres per capita per day)
1658
300 250 200 150 100
Minas Gerais
Rio de Janeiro
50
Espírito Santo
São Paulo
0
Amongst the 195 cities, rainfall ranged from 470 mm per year in the city of Sa˜o Luis do Piraitinga to 3395 mm per year in the city of Santos, both located in the state of Sa˜o Paulo. Fig. 3 shows the annual rainfall in ascending order for the cities analysed in each of the four states. 5.2. Number of people per dwelling
Rainfall (mm per year)
3500
2500
Minas Gerais
Rio de Janeiro
Espírito Santo
São Paulo
1500 1000 500 0 21
41
61
81 City
41
61
81 City
101
121
141
161
Fig. 5. Potable water demand over 195 cities in Minas Gerais, Rio de Janeiro, Espı´ rito Santo, and Sa˜o Paulo.
101
121
Fig. 5 shows the potable water demand obtained for the cities analysed over the four states. Amongst all cities, it ranged from about 89 to 307 litres per capita per day. The average potable water demand was 150 litres per capita per day in Minas Gerais, 217 litres per capita per day in Rio de Janeiro, 166 litres per capita per day in Espı´ rito Santo, and 162 litres per capita per day in Sa˜o Paulo. 5.4. Average roof area
2000
1
21
5.3. Potable water demand
The number of people per dwelling ranged between 2.94 and 4.55 over the 195 cities, with an average of 3.65 people per dwelling. The result is similar to the figure for all of the cities located in the southeast region, which is 3.52 people per dwelling for the year 2000 [14]. Fig. 4 shows the results for the four states.
3000
1
141
161
Fig. 3. Annual rainfall over 195 cities in Minas Gerais, Rio de Janeiro, Espı´ rito Santo, and Sa˜o Paulo.
In order to determine an adequate roof area per dwelling, the percentage of houses and flats in multi-storey residential buildings was obtained for all of the cities. Such a distinction was deemed appropriate as the specific roof area per person is lower in multi-storey buildings. Fig. 6 shows, as an example, the results for the 163 cities located in the state of Sa˜o Paulo, where the average roof area ranged from 38 to 85 m2. Such a low figure as 38 m2 was obtained for the city of Santos (first bar in Fig. 6) due to the high percentage of flats in multi-storey residential buildings (64%) in that city.
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
5.5. Volume of rainwater
savings for the cities of Belo Horizonte, Rio de Janeiro, Vito´ria and Sa˜o Paulo, where a potential ranging from 4% in June–August to 83% in December was observed. Fig. 8 shows the results for the two cities with the minimum and maximum potential for potable water savings observed along the year. The average potential for potable water savings observed in Salto was 12% and in Saleso´polis, 92%. Both cities are located in the state of Sa˜o Paulo. Fig. 9 presents the maximum, average and minimum potential for potable water savings observed for all of the 195 cities. On average, such a potential ranges from 12% in August to 79% in January. The overall average observed in
Houses
Flats
Average roof area
0 1
21
41
61
81 101 City
121
141
40 20
Dec
Nov
Oct
Sep
Aug
Jul
0 Jun
20
60
May
40
80
Apr
60
100
Mar
80
Potential for potable water savings by using rainwater (%)
100 90 80 70 60 50 40 30 20 10 0
Average roof area (m2)
Percentage of houses and flats (%)
By comparing the monthly volume of rainwater that could be harvested with the monthly potable water demand in each city, the potential for potable water savings could be estimated. Fig. 7 shows the potential for potable water
Feb
5.6. Potential for potable water savings
Jan
The monthly volume of rainwater that could be harvested in each one of the 195 cities was calculated through the procedure described in the methodology. Table 2 presents an example for four cities, which are the capital cities of each state. Number of people per dwelling, potable water demand, average roof area per dwelling and total roof area for the four cities are also shown in Table 2.
100
1659
Month Belo Horizonte (MG) Vitória (ES)
161
Rio de Janeiro (RJ) São Paulo (SP)
Fig. 7. Potential for potable water savings by using rainwater over the four capital cities.
Fig. 6. Percentage of houses and flats in multi-storey residential buildings and average roof area over the 163 cities in the state of Sa˜o Paulo.
Table 2 Results for the cities of Belo Horizonte, Rio de Janeiro, Vito´ria and Sa˜o Paulo Belo Horizonte (MG)
Rio de Janeiro (RJ)
Vito´ria (ES)
Month
Rainfall (mm/month)
Rainfall (mm/month)
Rainfall (mm/month)
January February March April May June July August September October November December Number of people per dwelling Potable water demand (litres/capita per day) Average roof area per dwelling (m2) Total roof area (m2)
296 188 164 61 28 14 16 14 41 123 228 319
Volume of rainwater (m3)
9,644,327 6,132,269 5,321,794 1,992,011 904,868 458,944 511,022 445,924 1,318,243 4,006,806 7,408,198 10,396,214 3.56
114 105 103 137 86 80 56 51 87 88 96 169
Volume of rainwater (m3)
8,424,642 7,774,889 7,627,218 10,145,012 6,320,328 5,936,383 4,164,328 3,728,698 6,431,081 6,512,300 7,058,684 12,478,217 3.25
143 82 111 89 81 65 78 55 78 127 171 195
Sa˜o Paulo (SP) Volume of rainwater (m3)
522,305 300,964 405,790 326,166 294,755 237,776 286,354 200,886 285,989 462,404 622,748 712,964 3.42
Rainfall (mm/month)
216 226 170 79 58 42 36 38 76 125 122 174
Volume of rainwater (m3)
34,111,076 35,572,403 26,790,551 12,443,583 9,185,446 6,637,806 5,642,576 6,043,767 11,967,871 19,757,965 19,246,193 27,366,161 3.54
178
288
248
176
65
56
53
67
42,172,778
92,294,504
4,724,113
.198,936,026
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
1660
100 Potential for potable water savings by using rainwater (%)
Potential for potable water savings by using rainwater (%)
100
80 60
40 20
90 70 60 50 40 30 20 10 0 50
Salto (SP)
Fig. 8. City with maximum and city with minimum potential for potable water savings by using rainwater.
100 80 60 40
Potential for potable water savings by using rainwater (%)
Fig. 10. Correlation between the potential for potable water savings and potable water demand over 195 cities in the four states of the southeast region.
Month Salesópolis (SP)
100 90 80 70 60 50 40 30 20 10 0 75
20
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
0
Month Maximum
Average
100 150 200 250 300 Potable water demand (litres/capita per day)
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
0
Potential for potable water savings by using rainwater (%)
y = -0.1519x + 66.31 R2 = 0.3092
80
Minimum
Jan Jul
100
125 150 175 200 225 250 275 Potable water demand (litres/capita per day) Feb Aug
Mar Sep
Apr Oct
May Nov
300
Jun Dec
Fig. 11. Monthly correlation between the potential for potable water savings and potable water demand over 195 cities in the four states of the southeast region.
Fig. 9. Maximum, average and minimum potential for potable water savings by using rainwater over the 195 cities.
the southeast region was 41%, which is lower than the figure obtained for the same region (48%) as presented in Ghisi [8]. By analysing the four states separately, an average potential for potable water savings of 42% was obtained for the states of Minas Gerais and Sa˜o Paulo, 35% in the state of Espı´ rito Santo and 29% in the state of Rio de Janeiro. These figures are significantly lower than the figure that Ghisi et al. [9] obtained for the state of Santa Catarina (69%) as rainfall is lower and potable water demand is higher in the states of the southeast region. 5.7. Correlations Fig. 10 shows the correlation between the average potential for potable water savings and potable water demand over all the cities analysed in the southeast region. It can be observed that there is not a good correlation between the average potential for potable water savings
and average potable water demand. But there is a tendency for the potential for potable water savings by using rainwater to be higher in those cities where the potable water demand is lower. By analysing the correlations on a monthly basis, interesting results were obtained and these are shown in Fig. 11. Higher potential for potable water savings by using rainwater occurs from October to March, when it is warmer and there is more rainfall in the southeast region. The potential for potable water savings along the year varies more significantly for cities with low potable water demand. For cities whose potable water demand is as low as 75 litres per capita per day, the potential for potable water savings ranges from 16% to 100%, while for cities with potable water demand of 300 litres per capita per day, it ranges from 6% to 44% along the year. The equations and coefficients of determination (R2) for the 12 best-fit straight lines of Fig. 11 are shown in Table 3.
ARTICLE IN PRESS Table 3 Equations and R2 for the best-fit straight lines of Fig. 11 Month
Equation
R2
January February March April May June July August September October November December
y ¼ 0:2565x þ 121:24 y ¼ 0:2326x þ 107:38 y ¼ 0:2124x þ 92:27 y ¼ 0:0797x þ 42:41 y ¼ 0:0870x þ 36:89 y ¼ 0:0652x þ 27:72 y ¼ 0:0471x þ 20:70 y ¼ 0:0427x þ 18:98 y ¼ 0:1040x þ 41:96 y ¼ 0:1839x þ 75:42 y ¼ 0:2239x þ 90:71 y ¼ 0:2773x þ 117:84
0.3349 0.2387 0.2203 0.0703 0.0967 0.0758 0.0560 0.0510 0.1285 0.2928 0.2524 0.3326
Note: y represents the potential for potable water savings by using rainwater (%) and x, the potable water demand (litres/capita per day).
y = 23.863Ln(x)- 129.78
90
R2 =0.2852
80 70 60 50 40 30 20 10 0 0
500
1000 1500 2000 2500 Rainfall (mm per year)
3000
3500
1661
100 80 60 40 20 0 75
100 125 150 175 200 225 Potable water demand (litres/capita per day) Southeast region
250
State of SC
Fig. 13. Correlation between the potential for potable water savings and potable water demand—comparison between the southeast region and the state of Santa Catarina.
Potential for potable water savings by using rainwater (%)
Potential for potable water savings by using rainwater (%)
100
Potential for potable water savings by using rainwater (%)
E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
100 80 60 40 20 0 1000
1250
1500 1750 Rainfall (mm per year)
Southeast region
2000
State of SC
Fig. 12. Correlation between the potential for potable water savings and rainfall over 195 cities in the four states of the southeast region.
Fig. 14. Correlation between the potential for potable water savings and rainfall—comparison between the southeast region and the state of Santa Catarina.
The correlation between the potential for potable water savings and rainfall was also investigated (Fig. 12). It was observed that such a correlation is weaker than that shown in Fig. 10. From the correlations shown in this section, it can be implied that the potential for potable water savings of cities located in the southeast region cannot be estimated by using the average potable water demand or the average rainfall as the coefficients of determination are low.
1200 mm per year (Fig. 14), the potential for potable water savings by using rainwater in Santa Catarina is likely to be higher than in the southeast region.
5.7.1. Comparison with results for the state of Santa Catarina By comparing the results obtained for the southeast region with the results obtained for the state of Santa Catarina [8], it can be observed that the potential for potable water savings in Santa Catarina is more dependable on the variation of potable water demand and on rainfall. For potable water demand lower than 180 litres per capita per day (Fig. 13) and rainfall higher than
5.8. Rainwater tank capacities Rainwater tank capacities were determined for nine cities located in the state of Sa˜o Paulo. Table 4 shows the average potable water demand, weighted average roof area, average number of people per dwelling and rainfall in the nine cities. Fig. 15 shows, as an example, the potential for potable water savings as a function of the rainwater tank capacity and the rainwater demand (10–100% of the potable water demand) in the cities of Itaquaquecetuba, Jales and Santos. These data were obtained by using the computer programme Neptune. It can be observed that the potential for potable water savings increases as the rainwater demand increases from 10% to 100%; such a potential decreases as the potable water demand increases (from Figs. 15a–c); and
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
1662
Table 4 Potable water demand, weighted average roof area, average number of people per dwelling and rainfall in the nine cities in the state of Sa˜o Paulo City
Potable water demand
Itaquaquecetuba Alumı´ nio Paulistaˆnia Jales Espı´ rito Santo do Pinhal Sa˜o Paulo Indaiatuba Jundiaı´ Santos
Potential for potable water savings (%)
0
Potential for potable water savings (%)
Low Low Low Medium Medium
82 85 83 84 83
3.97 3.79 3.54 3.32 3.61
1340 1541 1255 1166 1524
176 238 244 255
Medium High High High
66 81 74 38
3.54 3.66 3.50 3.19
1363 1243 1140 3395
5000 10000 15000 20000 25000 30000 Rainwater tank capacity (litres)
Potential for potable water savings (%)
10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
5000 10000 15000 20000 25000 30000 Rainwater tank capacity (litres)
100 90 80 70 60 50 40 30 20 10 0
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0
(c)
94 106 114 161 166
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0
Rainfall (mm/year)
Classification
100 90 80 70 60 50 40 30 20 10 0
(b)
Number of people per dwelling
(litres/capita per day)
100 90 80 70 60 50 40 30 20 10 0
(a)
Roof area (m2)
5000 10000 15000 20000 25000 30000 Rainwater tank capacity (litres)
Fig. 15. Potential for potable water savings as a function of rainwater tank capacity and rainwater demand ranging from 10% to 100%. (a) City of Itaquaquecetuba. (b) City of Jales. (c) City of Santos.
for a specific rainwater demand, the potential for potable water savings increases less steeply as the rainwater tank capacity increases.
As explained in the methodology, the ideal rainwater tank capacity was chosen as the one in which the potential for potable water savings increased 0.50% or less when increasing the rainwater tank capacity by 1000 litres. For example, in the city of Itaquaquecetuba (Fig. 15a), for a rainwater demand of 10% of the potable water demand, the potential for potable water savings is 9.32% for a rainwater tank of 1000 litres and 9.78% for a rainwater tank of 2000 litres; as the difference between these two figures is 0.46%, the chosen rainwater tank capacity was 2000 litres. Following such a procedure, 10 rainwater tank capacities (one for each rainwater demand) were determined for each city. Figs. 16–18 show the results for the nine cities. It can be observed that for very low potable water demand, such as in Itaquaquecetuba, Alumı´ nio and Paulistaˆnia (Fig. 16), rainwater tank capacities increase according to rainwater demand of up to 60–70%; for rainwater demand higher than 70% of the potable water demand, tank capacities decrease. A similar trend was also observed for the other cities (Figs. 17–18), but rainwater tanks are smaller than those for cities with low potable water demand. For cities with medium potable water demand (Fig. 17), the maximum rainwater tank capacity ranges from 7000 to 12,000 litres; for cities with high potable water demand (Fig. 18), it ranges from 6000 to 8000 litres. For cities with either medium or high potable water demand, the rainwater tank capacity does not vary significantly as a function of the rainwater demand; for cities with high potable water demand, for example, the rainwater tank capacity is about the same size for rainwater demand above 20% of potable water demand. 5.8.1. The influence of the potable water demand From the results shown in Figs. 16–18, another correlation was deemed appropriate. Correlation between rainwater tank capacities and potable water demand for the nine cities, for rainwater demand ranging from 10% to 100% at increments of 10%, is shown in Fig. 19. An interesting trend is observed. For rainwater demand lower
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
30 Rainwater tank capacity (1000 litres)
Rainwater tank capacity (1000 litres)
30 25 20 15 10 5
25 20 15 10 5 0
0 0 (a)
20 40 60 80 Rainwater demand (% of potable water demand)
0
100
Rainwater tank capacity (1000 litres)
Rainwater tank capacity (1000 litres)
100
30
25 20 15 10 5 0 0
(b)
20 40 60 80 Rainwater demand (% of potable water demand)
25 20 15 10 5 0
100
0
20 40 60 80 Rainwater demand (% of potable water demand)
(b)
30
100
30 Rainwater tank capacity (1000 litres)
Rainwater tank capacity (1000 litres)
20 40 60 80 Rainwater demand (% of potable water demand)
(a)
30
25 20 15 10 5 0 0
(c)
1663
20 40 60 80 Rainwater demand (% of potable water demand)
25 20 15 10 5 0 0
100 (c)
20 40 60 80 Rainwater demand (% of potable water demand)
100
Fig. 16. Correlation between rainwater tank capacity and rainwater demand in cities with low potable water demand. (a) City of Itaquaquecetuba. (b) City of Alumı´ nio. (c) City of Paulistaˆnia.
Fig. 17. Correlation between rainwater tank capacity and rainwater demand in cities with medium potable water demand. (a) City of Jales. (b) City of Espı´ rito Santo do Pinhal. (c) City of Sa˜o Paulo.
than 30% of potable water demand, there is a slight increase in the rainwater tank capacity as the potable water demand increases from 75 to 275 litres per capita per day. On the other hand, for rainwater demand higher than 30%, there is a decrease in the rainwater tank capacity as the potable water demand increases; and the higher the rainwater demand, the steeper the decrease in the rainwater tank capacity, being maximum for 60–70% of rainwater demand. Table 5 shows the equations and coefficients of determination (R2) for the 10 trend lines that can be seen
in Fig. 19. It can be noticed that there is scarcely any correlation between rainwater tank capacity and potable water demand when the rainwater demand is lower than 30% of the potable water demand, implying in a weak coefficient of determination. 5.8.2. The influence of the roof area In order to evaluate the influence of the roof area on the rainwater tank capacities, an analysis was performed for the cities of Alumı´ nio (low potable water demand), Espı´ rito Santo do Pinhal (medium potable water demand)
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
30
Table 5 Equations and R2 for the best-fit straight lines of Fig. 19
25 20 15 10 5 0 0
20 40 60 80 Rainwater demand (% of potable water demand)
Rainwater tank capacity (1000 litres)
(a)
100
30
Rainwater demand (% of potable water demand)
Equation
R2
10 20 30 40 50 60 70 80 90 100
y ¼ 0:0026x þ 2:45 y ¼ 0:0042x þ 4:17 y ¼ 0:0056x þ 7:96 y ¼ 0:0318x þ 14:22 y ¼ 0:0660x þ 21:84 y ¼ 0:0964x þ 28:60 y ¼ 0:0986x þ 28:65 y ¼ 0:0699x þ 22:07 y ¼ 0:0554x þ 18:81 y ¼ 0:0466x þ 16:53
0.2210 0.1819 0.1576 0.6644 0.8221 0.8001 0.8217 0.8072 0.8038 0.8509
Note: y represents the rainwater tank capacity (m3) and x, the potable water demand (litres/capita per day).
25 20 15
10%
20%
30%
40%
50%
10
60%
70%
80%
90%
100%
5
25
0 0
20 40 60 80 Rainwater demand (% of potable water demand)
(b)
100
Rainwater tank capacity (1000 litres)
30 25 20 15 10
20
15
10
5
0 75
5 0 0
(c)
Rainwater tank capacity (1000 litres)
Rainwater tank capacity (1000 litres)
1664
20 40 60 80 Rainwater demand (% of potable water demand)
100
125 175 225 Potable water demand (litres/capita per day)
275
Fig. 19. Correlation between rainwater tank capacity and potable water demand for rainwater demand ranging from 10% to 100% of potable water demand.
Fig. 18. Correlation between rainwater tank capacity and rainwater demand in cities with high potable water demand. (a) City of Indaiatuba. (b) City of Jundiaı´ . (c) City of Santos.
6. Conclusions
and Santos (high potable water demand). Results are shown in Fig. 20. In general, it can be observed that larger rainwater tanks are needed for larger roof areas. For the city of Alumı´ nio (Fig. 20a), however, this is not always the case, as the rainwater tank size may decrease by increasing the roof area. This happens because dwellings in such a city present low potable water demand. When the rainwater demand is lower than 30%, rainwater tanks for roof areas of 85 m2 and 127 m2 are smaller than those for roof area of 42 m2. For rainwater demand ranging from 30% to 70%, rainwater tank sizes for roof area of 42 and 127 m2 are smaller than tanks for roof area of 85 m2.
The potential for potable water savings by using rainwater in the states of Minas Gerais, Espı´ rito Santo, Rio de Janeiro and Sa˜o Paulo, located in southeastern Brazil, has been assessed for a sample of 195 cities. Rainwater tank capacities as a function of potable water demand, roof area, and rainwater demand were evaluated for nine cities located in the state of Sa˜o Paulo. Results of the research performed over 195 cities located in southeastern Brazil indicate that potable water demand in the residential sector ranges from about 90 to 300 litres per capita per day and that there is an average rainfall ranging from about 500 to 3400 mm per year. The average potential for potable water savings is 41%, ranging from
ARTICLE IN PRESS E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
Rainwater tank capacity (1000 litres)
30 25
42sqm 85sqm
20
127sqm
15 10 5 0 0
(a)
20 40 60 80 Rainwater demand (% of potable water demand)
100
Rainwater tank capacity (1000 litres)
30 42sqm
25
83sqm 125sqm
20 15
Acknowledgements 5
0 (b)
20 40 60 80 Rainwater demand (% of potable water demand)
100
30 Rainwater tank capacity (1000 litres)
[9]. However, the same trend was observed, i.e., the lower the potable water demand and the higher the rainfall, the higher the potential for potable water savings by using rainwater. As for rainwater tank capacities for the nine cities located in the state of Sa˜o Paulo, it was observed that they depend not only on roof area, rainfall and potable water demand, but also on rainwater demand. Cities with higher potable water demand will need rainwater tanks with lower capacities and less dependence on the rainwater demand. It was observed that maximum rainwater tank capacities are needed for rainwater demand of 60–70% of potable water demand. As for the roof area, it was noted that by increasing it, the rainwater tank capacities in general need to be larger. However, as observed for the city of Alumı´ nio, which has a low potable water demand compared to the other cities, the increase in the roof area does not always result in a larger rainwater tank capacity. Therefore, it is recommended that rainwater tank capacity be determined for each dwelling and location, as it depends on potable water demand, rainwater demand, roof area and daily rainfall.
10
0
19sqm
25
57sqm
15 10 5
0
20 40 60 80 Rainwater demand (% of potable water demand)
Dr. E. Ghisi would like to thank CAPES—Fundac- a˜o Coordenac- a˜o de Aperfeic- oamento de Pessoal de Nı´ vel Superior, an agency of the Brazilian Government for post-graduate education, for the scholarship that allowed him to supervise this research. References
38sqm 20
0 (c)
1665
100
Fig. 20. Rainwater tank capacities as a function of roof area and rainwater demand. (a) City of Alumı´ nio. (b) City of Espı´ rito Santo do Pinhal. (c) City of Santos.
12% to 79% on average. Such a potential is very significant and indicates that rainwater could probably be used for both potable and non-potable purposes. In case of using rainwater for potable purposes, it must be emphasized that it should go through proper treatment in order to avoid the spread of diseases. Correlations between the potential for potable water savings and either potable water demand or rainfall were weaker than those verified for the state of Santa Catarina
[1] Herrmann T, Schmida U. Rainwater utilisation in Germany: efficiency, dimensioning, hydraulic and environmental aspects. Urban Water 1999;1(4):307–16. [2] Coombes PJ, Argue JR, Kuczera G. Figtree Place: a case study in water sensitive urban development (WSUD). Urban Water 1999;1(4):335–43. [3] Fewkes A. The use of rainwater for WC flushing: the field testing of a collection system. Building and Environment 1999;34(6):765–72. [4] Marinoski DL, Ghisi E, Go´mez LA. Aproveitamento de a´gua pluvial e dimensionamento de reservato´rio para fins na˜o pota´veis: estudo de caso em um conjunto residencial localizado em Floriano´polis-SC [Rainwater harvesting and rainwater tank sizing: a case study in a multi-storey residential building located in Floriano´polis, Southern Brazil]. CLACS04—I Confereˆncia Latino-Americana de Construc- a˜o Sustenta´vel e ENTAC04—101 Encontro Nacional de Tecnologia do Ambiente Construı´ do, Sa˜o Paulo-SP, CD Rom, 2004. (in Portuguese). [5] Appan A. A dual-mode system for harnessing roofwater for nonpotable uses. Urban Water 2000;1(4):317–21. [6] Handia L, Tembo JM, Mwiindwa C. Potential of rainwater harvesting in urban Zambia. Physics and Chemistry of the Earth 2003;28(20–27):893–6. [7] Li X- Y, Gong J- D. Compacted microcatchments with local earth materials for rainwater harvesting in the semiarid region of China. Journal of Hydrology 2002;57(1-4):134–44. [8] Ghisi E. Potential for potable water savings by using rainwater in the residential sector of Brazil. Building and Environment, in Press, Corrected Proof, Available online 12 July 2005.
ARTICLE IN PRESS 1666
E. Ghisi et al. / Building and Environment 42 (2007) 1654–1666
[9] Ghisi E, Montibeller A, Schmidt RW. Potential for potable water savings by using rainwater: an analysis over 62 cities in southern Brazil. Building and Environment 2006;41(2):204–10. [10] Lee KT, Lee CD, Yang MS, Yu CC. Probabilistic design of storage capacity for rainwater cistern systems. Journal of Agricultural Engineering Research 2000;77(3):343–8. [11] Villarreal EL, Dixon A. Analysis of a rainwater collection system for domestic water supply in Ringdansen, Norrko¨ping, Sweden. Building and Environment 2005;40(9):1174–84. [12] Chilton JC, Maidment GG, Marriott D, Francis A, Tobias G. Case study of a rainwater recovery system in a commercial building with a large roof. Urban Water 1999;1(4):345–54. [13] ANA Ageˆncia Nacional da A´gua [Water Information Agency of Brazil]. Water Resources Management in Brazil. Information available at http://hidroweb.ana.gov.br/HidroWeb/doc/WRMB/ index.htm. Accessed in April 2005. [14] IBGE Instituto Brasileiro de Geografia e Estatı´ stica [Brazilian Institute for Geography and Statistics]. Obtained from: /http:// www.ibge.gov.br/S. Accessed in April 2005. [15] UNEP United Nations Environment Programme. Global Environment Outlook 3: past, present and future perspectives. Earthscan, UK, 2002. [16] Department of Water and Electricity of Sa˜o Paulo. Obtained from: /www.daee.sp.gov.brS. Accessed in July 2004.
[17] BRASIL. Normais Climatolo´gicas (1961–1990). Ministe´rio da Agricultura e Reforma Agra´ria. Secretaria Nacional de Irrigac- a˜o. Departamento Nacional de Meteorologia. Brası´ lia. 1992. [Climatic data for 1961–1990, Meteorology Information Agency of Brazil] (in Portuguese). [18] SNIS Sistema Nacional de Informac- o˜es sobre Saneamento. Diagno´stico dos servic- os de a´gua e esgotos–2001 [National Database on Sanitation. Diagnosis on water and sanitation services in the year 2001]. Brası´ lia: Secretaria Especial de Desenvolvimento Urbano da Presideˆncia da Repu´blica—SEDU/PR: Instituto de Pesquisa Econoˆmica Aplicada—IPEA. Available at: /http://www.snis.gov.br/ diag_2001.htmS. Accessed in July 2004. [19] Ghisi E, Tre´s ACR, Kotani M. Netuno—aproveitamento de a´guas pluviais no setor residencial. 2004 [Neptune—a computer programme to evaluate potable water savings and rainwater tank capacity in the residential sector] (in Portuguese). [20] Environment Agency. Harvesting rainwater for domestic uses: an information guide. Available at www.environment-agency.gov.uk, Bristol, UK, 2003. [21] De Oreo WB, Mayer PW. Residential end uses of water. AWWA Research Foundation 1999. Data also available from http:// www.sabesp.com.br/pura/noticias_dados/dados_distribuicao_agua1. htm
