## Abstract

The objective of this study is to assess the investment feasibility analysis of rainwater harvesting in a building located in Brazil. Rainwater was used to supply toilets, cleaning and irrigation. The costs of materials, labour and maintenance were obtained to carry out the economic analysis; the indicators used were the net present value, internal rate of return and payback period. The investment feasibility analysis and the potential for potable water savings were obtained by means of computer simulation. The potential for potable water savings ranged from 26.70% to 64.70%. The net present value, internal rate of return and payback period for the best scenario were, respectively, R$132,801.47, 3.73% per month and 32 months. For the worst scenario the net present value was R$9,451.26, the internal rate of return was 0.91% per month and the payback period was 166 months. Thus, rainwater can be used as a sustainable alternative and be financially feasible.

## INTRODUCTION

Climate change and the increasing demand for potable water for agricultural and urban development require water beyond water resources available. Arid and semi-arid regions have been facing problems of water scarcity, both for potable and non-potable purposes (Adham et al. 2016). Lee et al. (2016) state that rainwater harvesting has to be incorporated into the water resource and climate change policy, especially in urban areas where potable water availability is decreasing due to the rapid increase in population and water consumption. In the current scenario, rainwater harvesting is one of the possible alternatives to be considered due to future climate changes and lack of potable water (Kahinda et al. 2010).

This study presents an analysis of rainwater harvesting in a building located in Brazil. Brazil is a country whose population distribution and water resources vary greatly over its territory. According to the Brazilian National Water Agency (ANA 2017), the population density of the south-eastern region is 86.92 inhabitants per square kilometre and has only 6.0% of the available water resources. On the other hand, the northern region has 68.5% of the water resources, but only 4.12 inhabitants per square kilometre. In densely urbanized regions, where roofs account for large areas, rainwater harvesting is an interesting alternative to supply water for non-potable purposes. Domestic rainwater harvesting system is a compatible water supply technique in order to allow for potable water savings in urban areas where there is water scarcity and high demographic concentration (Campisano et al. 2015; Palla et al. 2017).

The water supply to cities requires a notable amount of other resources such as energy and infrastructure. Also, around the world, the availability of water resources is decreasing due to water pollution as a consequence of urbanization and industrialization. The practice of using rainwater must be widespread in order to lessen the pressure on main water supply (Haque et al. 2016). The feasible beneﬁts that result from rainwater harvesting are the reduction of potable water demand, decreasing of runoff into the urban drainage system and a reduced risk of overﬂow from storm events (Zhang et al. 2009).

Zhang et al. (2009) studied the potential for potable water savings by using rainwater and efficient domestic facilities in buildings in various cities in Australia and concluded that a good water demand management was a quite efﬁcient measure to replace potable water. Potable water savings ranging from 29.9% (Perth) to 32.3% (Sydney) were obtained.

From several studies, it can be seen that it is necessary and important to promote the development of rainwater harvesting as an alternative water resource.

## OBJECTIVE

The objective of this paper is to estimate the potential for potable water savings and perform an economic analysis for different scenarios of rainwater demand in a building in the city of Florianópolis, southern Brazil.

## METHOD

### The city

Florianópolis is located between the latitudes 27°22′S and 27°51′S, and between the longitudes 48°20′W and 48°37′W. The climate is subtropical humid, with rainfall well distributed throughout the year, without dry season (PMF 2014). According to the Köppen Geiger classification, the climate is described as humid mesotherm (Andrade 1996). On average, there is rainfall over 140 days per year. Average rainfall is 1,500 mm per year average and air relative humidity is 85% (Andrade 1996). Over the summer season, there is usually lack of water in some areas of Florianópolis, mainly in the northern island. This is due to the great number of tourists.

### The building

The object of study in this work is a public building located in northern Florianópolis. The building is composed of two blocks as shown in Figure 1. The roof of the blocks and part of the parking area is composed of photovoltaic panels, with a floor plan area equal to 526.00 m2. Rainwater will be used to flush toilets, cleaning (external taps), and irrigation (garden taps).

Figure 1

View of the two-block building (WAZ 2014).

Figure 1

View of the two-block building (WAZ 2014).

### Rainwater tank sizing

#### Potable water demand

As the building is new with no occupation when this study was performed, the potable water demand was estimated according to the city legislation (Law #060/2000 of the city of Florianópolis). The total number of users was estimated based on the floor plan area; and the potable water demand per person was obtained according to the activities in the building.

#### Rainwater demand

In a rainwater harvesting system it is necessary to know the rainwater demand. In this work, the rainwater demand was obtained based on studies that estimated the water end-uses for actual buildings. From the end-uses obtained, a range of figures was chosen to perform the simulations in order to account for the unknown rainwater demand in the building.

#### Rainfall

Rainfall data were obtained from a meteorological station located at the latitude 27°36′S and longitude 48°37′W, i.e. near the study area. Simulations were performed using daily rainfall from 2001 to 2014.

#### Roof area

The catchment area considered was the horizontal projection of the impermeable surface, that is, the roof area, as recommended by the NBR 15527 (ABNT 2007). The shape, orientation and slope of the roof in relation to the highest occurrence of winds were verified in order to ease the collection of rainwater.

#### Runoff

The runoff coefficient was obtained from the literature based on the type of roof used in the building.

#### Tank sizes

Two rainwater tanks were sized, i.e. a lower and an upper tank. Rainwater was pumped from the lower to the upper tank. The capacity of the upper tank was estimated to store rainwater to supply the rainwater demand over a day. The capacity and location of such a tank took into account the space available in the building. As for the lower tank, technical and economic aspects were also taken into account. Its capacity was estimated by means of computer simulations using the Netuno computer programme, version 4 (Ghisi & Cordova 2014). The maximum capacity considered in the simulations was defined according to the available area and existing restrictions for placing the tanks.

### Rainwater quality

In order to guarantee the quality of rainwater, a first-flush device will be used to divert the rainwater used to clean the roof. In addition, rainwater stored in the lower tank will receive physical-chemical treatment according to NBR 15527 guidelines (ABNT 2007).

### Economic analysis

The economic feasibility analysis of the rainwater harvesting system was carried out using the Netuno computer programme. The net present value, internal rate of return, and discounted payback period of the investment were estimated.

As input data to the programme, the costs of the following components are required: pipes, motor pump and accessories, as well as labour and maintenance costs. Water, sewage and electricity tariffs were obtained from the utilities, as well as taxes and the frequency tariffs increase.

The minimum acceptable rate of return was taken as the average value of the savings rate, based on the benchmark rate plus 0.5% interest, obtained from Banco do Brasil for the months of July to December 2014 (BB 2014). The monthly inflation rate was taken as the average of the monthly figures over 2014 obtained from the Broad Consumer Price Index (IPCA) and the Brazilian Institute of Geography and Statistics (IBGE 2014).

## RESULTS

### Potable water demand

The building is classified as a building providing a teaching service, which, in accordance with current legislation, the number of users was estimated considering one person for each 7.5 m² of floor plan area, and water consumption equal to 50 L/person·day. The floor plan areas of blocks A and B were 492.78 m² and 277.68 m², and the number of users was estimated as 66 and 37 people, respectively. Thus, daily water consumption in blocks A and B were estimated as 3,300 and 1,850 L, respectively, and the total daily water consumption in the building as 5,150 L. In order to account for variations on the water consumption, the simulations were performed for potable water demands equal to 15, 20, 25 and 50 L/person·day. In similar buildings, Fasola et al. (2011) obtained 25.3 and 28.8 L/person·day, Marinoski & Ghisi (2008) obtained 15.5 L/person per day, and Ywashima (2006) obtained 22.2 L/person·day.

### Rainwater demand

Based on the literature (Ywashima 2006; Marinoski & Ghisi 2008; Fasola et al. 2011), it was observed that rainwater demand ranged from 50% to 80% of the water demand. Therefore, rainwater demands of 50%, 60%, 70% and 80% were used to perform the simulations. Table 1 shows the places and quantity of appliances which will be supplied with rainwater.

Table 1

Quantity of appliances to be supplied with rainwater

ApplianceDescriptionBlock
Total
AB
Toilet Toilet flushing (6 L bowl-and-tank toilet) 10
Outdoor taps Cleaning the rooms and outdoor area, watering plants and grass
Green roof tap Watering plants –
Balcony tap Cleaning the balcony –
ApplianceDescriptionBlock
Total
AB
Toilet Toilet flushing (6 L bowl-and-tank toilet) 10
Outdoor taps Cleaning the rooms and outdoor area, watering plants and grass
Green roof tap Watering plants –
Balcony tap Cleaning the balcony –

### Rainfall

Daily rainfall from 1st September 2001 to 30th June 2014 was used to perform the simulations (Figure 2). The average daily rainfall was 4.9 mm.

Figure 2

Daily rainfall over 1st September 2001 to 30th June 2014.

Figure 2

Daily rainfall over 1st September 2001 to 30th June 2014.

### Roof area

The sloped metal roof of block B was taken in its entirety to collect rainwater; its horizontal projection area was 180.60 m². As for the curved metal roof of block A only the area facing north was considered to collect rainwater; the area facing south was not taken into account because it is small. Therefore, the horizontal projection area of block A was then 354.40 m². The total catchment area considered to collect rainwater was 526.00 m².

### Runoff

The metal roof of both blocks is covered with photovoltaic panels, which are smooth and vitreous. Therefore, a runoff coefficient equal to 0.90 was adopted.

### Upper rainwater tank

The maximum capacity of each upper rainwater tank would be 2,640 L in block A and 1,480 L in block B, considering the water demand equal to 50 L/person·day and rainwater demand equal to 80% of the water demand. However, as there are taps and limitation of space where the tanks will be placed, four upper rainwater tanks were used in block A and two in block B. Figure 3 shows the details for two rainwater tanks in block A. Such configuration was chosen in order to allow for maintenance and cleaning the tanks. The 100 L tank was used to supply the taps; it was placed at a higher level in order to guarantee head pressure in the taps. The 1,000 L tank was used to supply the other appliances. A similar configuration was also used in another place in block A, but with tank capacities of 100 and 1,500 L. Therefore, the total upper rainwater tank capacity in block A was 2,700 L. As for block B, one 100 L and one 1,500 L upper tanks were used. The total upper rainwater tank capacity in block B was 1,600 L.

Figure 3

Upper rainwater tanks in block A. (a) Front view. (b) Cross section.

Figure 3

Upper rainwater tanks in block A. (a) Front view. (b) Cross section.

### Lower rainwater tank

Table 2 shows the input data used to perform the computer simulations to select the ideal lower tank capacity and the corresponding potential for potable water savings. Due to limitation of space, the maximum lower tank capacity was taken as 20,000 L; the simulations were performed for lower tank capacities increments of 1,000 L. The ideal lower tank capacity was taken as the one in which the potential for potable water savings increased 2% or less when increasing the tank capacity in 1,000 L.

Table 2

Input data used to perform the computer simulations to obtain the lower tank capacity

InputFigure
Roof area 526.00 m²
Water demand 50.0 L/person·day
Number of users 103
Runoff coefficient 0.90
First flush 2.0 mm
Daily rainfall 1st Sep 2001 to 30th Jun 2014
Rainwater demand (% of water demand) 50%, 60%, 70%, 80%
Upper tank capacity 4,300 L
Upper tank capacity in which pumping starts 5% of the capacity
Maximum lower tank capacity 20,000 L
Increments for the lower tank capacity 1,000 L
InputFigure
Roof area 526.00 m²
Water demand 50.0 L/person·day
Number of users 103
Runoff coefficient 0.90
First flush 2.0 mm
Daily rainfall 1st Sep 2001 to 30th Jun 2014
Rainwater demand (% of water demand) 50%, 60%, 70%, 80%
Upper tank capacity 4,300 L
Upper tank capacity in which pumping starts 5% of the capacity
Maximum lower tank capacity 20,000 L
Increments for the lower tank capacity 1,000 L

Figure 4 shows the potential for potable water savings obtained from the computer simulations using the input data shown in Table 2. It is possible to observe that the potential for potable water savings increases as increases the lower rainwater tank capacities. Table 3 shows the lower tank capacities chosen for each rainwater demand, as well as the corresponding potential for potable water savings and percentage of days in which the rainwater demand is supplied with rainwater. The chosen lower tank capacity was 10,000 L. Then, two 5,000 L tanks were used (one 5,000 L tank was already included in the design).

Table 3

Lower tank capacities and corresponding results for water demand equal to 50 L/person·day

Rainwater demand (% of water demand)Lower tank capacity (L)Potential for potable water savings (%)Percentage of days in which the rainwater demand is supplied with rainwater:
FullyPartlyNot at all
50 10,000 26.70 46.57 12.87 40.56
60 9,000 26.90 36.18 16.22 47.60
70 9,000 27.30 29.07 18.64 52.29
80 9,000 28.70 26.57 19.69 53.74
Rainwater demand (% of water demand)Lower tank capacity (L)Potential for potable water savings (%)Percentage of days in which the rainwater demand is supplied with rainwater:
FullyPartlyNot at all
50 10,000 26.70 46.57 12.87 40.56
60 9,000 26.90 36.18 16.22 47.60
70 9,000 27.30 29.07 18.64 52.29
80 9,000 28.70 26.57 19.69 53.74
Figure 4

Potential for potable water savings for each rainwater demand.

Figure 4

Potential for potable water savings for each rainwater demand.

### Potential for potable water savings

Table 4 shows the potential for potable water savings for the different water and rainwater demands when the lower rainwater tank capacity is 10,000 L. For water demand ranging from 15 to 25 L/person·day and rainwater ranging from 50% to 80% of water demand, the potential for potable water savings may range from 39.74% to 64.70%. And the percentage of days in which the rainwater demand is fully supplied with rainwater ranges from 56.94% to 92.51%.

Table 4

Potential for potable water savings

Water demand (L/person·day)Rainwater demand (% of water demand)Potential for potable water savings (%)Percentage of days in which the rainwater demand is supplied with rainwater
FullyPartlyNot at all
15 50 46.53 92.51 1.25 6.23
60 53.64 88.47 2.23 9.30
70 59.40 83.28 3.22 13.50
80 64.70 78.64 4.42 16.94
20 50 43.20 84.98 2.86 12.16
60 48.53 78.64 4.42 16.94
70 53.07 72.83 5.81 21.36
80 56.66 67.11 7.20 25.70
25 50 39.74 77.16 4.63 18.21
60 43.85 69.90 6.57 23.53
70 47.00 62.64 8.27 29.09
80 50.02 56.94 9.97 33.09
50 50 26.70 46.57 12.87 40.56
60 27.93 38.59 15.46 45.94
70 28.42 30.77 18.12 51.11
80 29.76 27.42 19.26 53.32
Water demand (L/person·day)Rainwater demand (% of water demand)Potential for potable water savings (%)Percentage of days in which the rainwater demand is supplied with rainwater
FullyPartlyNot at all
15 50 46.53 92.51 1.25 6.23
60 53.64 88.47 2.23 9.30
70 59.40 83.28 3.22 13.50
80 64.70 78.64 4.42 16.94
20 50 43.20 84.98 2.86 12.16
60 48.53 78.64 4.42 16.94
70 53.07 72.83 5.81 21.36
80 56.66 67.11 7.20 25.70
25 50 39.74 77.16 4.63 18.21
60 43.85 69.90 6.57 23.53
70 47.00 62.64 8.27 29.09
80 50.02 56.94 9.97 33.09
50 50 26.70 46.57 12.87 40.56
60 27.93 38.59 15.46 45.94
70 28.42 30.77 18.12 51.11
80 29.76 27.42 19.26 53.32

### Rainwater quality

Considering that rainwater will be used for toilet flushing, garden irrigation and cleaning, the treatment chosen was filtration of solids and the application of chlorine.

### Economic analysis

Tables 5 and 6 show the input data used to perform the economic analysis using the Netuno computer programme. The economic analysis was performed over a period of 20 years due to the lifespan of motor pumps. Energy and water tariffs were taken as per January 2015. The starting time of motor pumps was 216 seconds (multiplied by six according to the number of upper rainwater tanks). Table 7 shows the net present value, internal rate of return, and discounted payback period for each water demand and rainwater demands equal to 50% and 80% of water demand.

Table 5

Input data for the economic analysis

InputFigure
Inflation 0.53% per month
Correction of water and energy tariffs 12 months
Lifespan 20 years
Minimum acceptable rate of return 0.59% per month
Sewage tariff 100% on the water bill
Energy tariff
Tariff 0.38 R$/kWh Tax 25% Pipes R$ 5,844
Motor pumps
Unit cost R$645 Efficiency 40% Water flow 2,150litres/hour Starting time 1,296 seconds Power 1HP Accessories R$ 5,173.00
Labour
Excavation R$1,112 Plumber and assistant R$ 2,465
Maintenance
Tanks R$790/semester Motor pumps R$ 129/semester
Chlorine doser R$13/month Filter R$ 60/semester
InputFigure
Inflation 0.53% per month
Correction of water and energy tariffs 12 months
Lifespan 20 years
Minimum acceptable rate of return 0.59% per month
Sewage tariff 100% on the water bill
Energy tariff
Tariff 0.38 R$/kWh Tax 25% Pipes R$ 5,844
Motor pumps
Unit cost R$645 Efficiency 40% Water flow 2,150litres/hour Starting time 1,296 seconds Power 1HP Accessories R$ 5,173.00
Labour
Excavation R$1,112 Plumber and assistant R$ 2,465
Maintenance
Tanks R$790/semester Motor pumps R$ 129/semester
Chlorine doser R$13/month Filter R$ 60/semester
Table 6

Water tariff for the buildings sector (CASAN 2014)

Water consumption (m3)Tariff
Less than or equal to 10 R$47.32 per month Greater than 10 and less than or equal to 50 R$ 7.85 per m3
Greater than 50 R$9.88 per m3 Water consumption (m3)Tariff Less than or equal to 10 R$ 47.32 per month
Greater than 10 and less than or equal to 50 R$7.85 per m3 Greater than 50 R$ 9.88 per m3
Table 7

Results for the economic analysis

IndexRainwater demand (% of water demand)Water demand (L/person·day)Figure
Net present value (R$) 50 15 9,451.26 20 37,790.04 25 64,368.56 50 112,551.52 80 15 37,831.12 20 65,890.40 25 91,770.16 50 132,801.47 Internal rate of return (% per month) 50 15 0.91 20 1.66 25 2.26 50 3.30 80 15 1.67 20 2.30 25 2.85 50 3.73 Discounted payback (months) 50 15 166 20 86 25 58 50 38 80 15 86 20 57 25 44 50 32 IndexRainwater demand (% of water demand)Water demand (L/person·day)Figure Net present value (R$) 50 15 9,451.26
20 37,790.04
25 64,368.56
50 112,551.52
80 15 37,831.12
20 65,890.40
25 91,770.16
50 132,801.47
Internal rate of return (% per month) 50 15 0.91
20 1.66
25 2.26
50 3.30
80 15 1.67
20 2.30
25 2.85
50 3.73
Discounted payback (months) 50 15 166
20 86
25 58
50 38
80 15 86
20 57
25 44
50 32

The best scenario for the economic analysis, i.e. water demand of 50 L/person·day and rainwater demand of 80% of the water demand, resulted in a net present value of R$132,801.47, internal rate of return equal to 3.73% per month and a payback period of 32 months. On the other hand, the worst scenario, i.e., water demand of 15 L/person·day and rainwater demand of 50% of the water demand, resulted in a net present value of R$ 9,451.26, internal rate of return equal to 0.91% per month and a payback period of 166 months. It is possible to observe that, even for the worst scenario, the internal rate of return (0.91% per month) is higher than the minimum acceptable rate of return (0.59% per month) and the payback period (166 months) is lower than the lifespan (20 years). Therefore, all scenarios can be considered feasible. The economic analysis showed that the greatest the rainwater demand, the more feasible the investment is.

## CONCLUSIONS

This paper assessed the possibility of using rainwater for non-potable purposes in a building located in southern Brazil. Rainwater will be used for flushing 10 toilets and supplying seven taps that will be used for cleaning and irrigation. Based on computer simulations that took into account daily rainfall, and different water and rainwater demands, a lower rainwater tank capacity of 10,000 L was considered adequate. As for the upper tanks, their total capacity equals to 4,300 L. Motor pumps are used to pump rainwater from the lower to the upper tanks.

For all scenarios (water demands from 15‒50 L/person·day and rainwater demands from 50‒80% of the water demand), the potential for potable water savings ranged from 26.70% to 64.70%. Based on the economic analysis all scenarios were considered feasible as the lowest internal rate of return (0.91% per month) is higher than the minimum acceptable rate of return (0.59% per month) and the longest payback period (166 months) is shorter than the lifespan (20 years).

The method shown herein can be easily applied by others to assess the potential for potable water savings by using rainwater and also to perform the investment feasibility analysis. It takes into account the assessment of potable water savings by means of computer simulation on a daily basis using daily rainfall, which leads to more accurate results.

## REFERENCES

REFERENCES
ABNT
2007
NBR 15527: água de Chuva: Aproveitamento de Coberturas em áreas Urbanas Para Fins não Potáveis: Requisitos [Rainwater Harvesting From Roofs in Urban Areas for non-Potable Purposes: Requirements]
.
Brazilian Standard.
Rio de Janeiro
,
Brazil
(in Portuguese)
.
,
A.
,
Riksen
,
M.
,
Ouessar
,
M.
&
Ritsema
,
C.
2016
Identiﬁcation of suitable sites for rainwater harvesting structures in arid and semi-arid regions: a review
.
International Soil and Water Conservation Research
4
,
108
120
.
ANA
2017
Brazilian National Water Agency, http://www2.ana.gov.br/Paginas/default.aspx
,
visited April 2017 (in Portuguese)
.
,
S. F.
1996
Estudo de Estratégias Bioclimáticas no Clima de Florianópolis [Study of Bioclimatic Strategies in Florianópolis]
.
Masters Dissertation
,
School of Civil Engineering, Federal University of Santa Catarina
,
Brazil
(in Portuguese)
.
BB
2014
,
http://www.bb.com.br, visited December 2014 (in Portuguese)
.
Campisano
,
A.
,
Gnecco
,
I.
,
Modica
,
C.
&
Palla
,
A.
2015
Designing domestic rainwater harvesting systems different climatic regimes in Italy
.
Water Science and Technology
67
,
2511
2518
.
CASAN
2014
Local water and sewage utility, http://www.casan.com.br/
,
visited December 2014 (in Portuguese)
.
Fasola
,
G. B.
,
Ghisi
,
E.
,
Marinoski
,
A. K.
&
Borinelli
,
J. B.
2011
Potencial de economia de água em duas escolas em Florianópolis, SC [Potential for potable water savings in two schools in Florianópolis, SC]
.
Ambiente Construído
11
,
65
78
(in Portuguese)
.
Ghisi
,
E.
&
Cordova
,
M. M.
2014
Netuno 4 Computer Programme
.
School of Civil Engineering, Federal University of Santa Catarina
,
Brazil
,
http://www.labeee.ufsc.br/, visited July 2014 (in Portuguese)
.
Haque
,
M. M.
,
Rahman
,
A.
&
Samali
,
B.
2016
Evaluation of climate change impacts on rainwater harvesting
.
Journal of Cleaner Production
137
,
60
69
.
IBGE
2014
Brazilian Institute for Geography and Statistics, http://www.ibge.gov.br
,
visited July 2014 (in Portuguese)
.
Kahinda
,
J. M.
,
Taigbenu
,
A. E.
&
Boroto
,
R. J.
2010
Domestic rainwater harvesting as an adaptation measure to climate change in South Africa
.
Physics and Chemistry of the Earth
35
,
742
751
.
Lee
,
K. E.
,
Mokhtar
,
M.
,
Hanafiah
,
M. M.
,
Halim
,
A. A.
&
,
J.
2016
Rainwater harvesting as an alternative water resource in Malaysia: potential, policies and development
.
Journal of Cleaner Production
126
,
218
222
.
Marinoski
,
A. K.
&
Ghisi
,
E.
2008
Aproveitamento de água pluvial para usos não potáveis em instituição de ensino: estudo de caso em Florianópolis – SC [Rainwater harvesting for non-potable uses in educational institutions: a case study in Florianópolis – SC]
.
Ambiente Construído
8
,
67
84
(in Portuguese)
.
Palla
,
A.
,
Gnecco
,
I.
&
La Barbera
,
P.
2017
The impact of domestic rainwater harvesting systems in stormwater runoff mitigation at the urban block scale
.
Journal of Environmental Management
191
,
297
305
.
PMF
2014
,
http://www.pmf.sc.gov.br, visited August 2014 (in Portuguese)
.
WAZ
2014
, ,
visited August 2014 (in Portuguese)
.
Ywashima
,
L. A.
2006
Avaliação do uso de água em Edifícios Escolares Públicos E Análise de Viabilidade Econômica da Instalação de Tecnologias Economizadoras nos Pontos de Consumo [Evaluation of Water use in Public School Buildings and Economic Feasibility Analysis of Water-Saving Technologies]
.
Masters Dissertation
,
School of Civil Engineering, State University of Campinas
,
Brazil
(in Portuguese)
.
Zhang
,
Y.
,
Chen
,
D.
,
Chen
,
L.
&
Ashbolt
,
S.
2009
Potential for rainwater use in high-rise buildings in Australian cities
.
Journal of Environmental Management
91
,
222
226
.