Economic Feasibility of Rainwater Harvesting and Greywater Reuse in a Multifamily Building

Abstract

This study aimed to evaluate the financial feasibility of rainwater harvesting and greywater reuse in a multifamily building located in Florianópolis, Brazil. A building, consisting of two blocks with 60 flats each, was chosen to obtain data about the number of residents, building characteristics, potable water consumption, and rainwater and greywater demands (obtained by means of questionnaires and water measurements). The financial feasibility analyses considered rainwater and greywater systems separately and together. The impact on the urban stormwater drainage system was evaluated through the reduction of stormwater runoff. The energy consumption in the operational phase of each system was estimated through the amount of energy consumed by the motor pumps to supply one cubic meter of water. The potential for potable water savings through the use of rainwater—that supplies water for washing machines—was approximately 6.9%. The potential for potable water savings through the use of greywater—that supplies water to toilets—was approximately 5.7%. Both systems were feasible. The payback period for rainwater harvesting systems ranged from 57 to 76 months. For greywater systems, the payback period ranged from 127 to 159 months. When considering both systems working together, the payback period ranged from 89 to 132 months. The rainwater harvesting system can reduce 11.8% the stormwater volume destined to the urban stormwater drainage system in relation to the current contribution volume. Energy consumption was approximately 0.56 kWh/m³ of treated water for the rainwater harvesting system and 0.89 kWh/m³ of treated water for the greywater system. Rainwater and greywater were considered economically feasible, especially for higher inflation scenarios. Furthermore, such systems are interesting alternatives in terms of impacts considering urban drainage and energy consumption.

1. Introduction

Rainwater and greywater are sources of water that could be used in buildings in order to meet some of the Sustainable Development Goals (SDGs), such as SDG 6 (clean water and sanitation) and 11 (sustainable cities and communities) [1]. Both rainwater and greywater can be considered as a stand-alone water supply or a strategy to supplement the main water supply [2]. Rainwater harvesting can also mitigate floods [3]. Therefore, rainwater and greywater could be regarded as local solutions to supplement the urban water supply that usually takes long distances to reach the buildings [4]. To identify if both or either could be used in buildings, it is very important to know the water end-uses.

The water end-uses can be separated into two main groups: potable uses and non-potable uses. Potable uses are those that involve the user’s direct contact with water, such as cooking, drinking, tooth brushing, and personal consumption in general. Non-potable uses are those that do not require drinking-water quality, such as toilet flushing, washing machines, and washing floors and vehicles. Among the non-potable uses, toilet flushing and washing machines represent some of the highest water demands in the residential sector [2,5,6]. These uses can be supplied by alternative sources of non-potable water, such as rainwater and greywater [7].

Studies on rainwater harvesting systems have shown that rainwater can be used for different purposes in a safe manner. The flexibility of such systems allows them to be installed in different types of buildings and with different components, such as catchment areas, storage tanks, and treatment. Some studies have verified that these systems may be economically viable in multifamily buildings depending on the parameters and characteristics of each building [8,9,10]. Rainwater harvesting is a strategy that can promote sustainability through lower energy consumption and lower greenhouse gas emissions compared to central supply systems [11]. It can decrease surface runoff and peak water flow during rainfall. Furthermore, some extreme climate effects caused by climate change that affect hydrological cycles, such as floods and droughts, can also be mitigated through the use of rainwater in buildings.

Studies on greywater reuse systems have shown that greywater can also be used for different non-potable uses [12,13,14] and that these systems may also be economically viable in different buildings [1,15,16].

In addition to the economic benefits, some researchers have also evaluated the environmental impacts of these systems, such as the potential to reduce flooding through the use of rainwater harvesting systems [17,18,19] and the potential for reducing greenhouse gas emissions and energy consumption through the use of rainwater harvesting and greywater reuse systems [20,21,22]. Environmental impact analyses, considering the entire life cycle of these systems, also show benefits and disadvantages in relation to traditional drinking water supply and sewage treatment systems.

There is a great potential for replacing drinking water with rainwater and greywater in residential buildings that needs to be explored in more depth. These systems can generate great economic, social, and environmental benefits. Therefore, to achieve these objectives, these systems must be designed, evaluated, and studied in more depth. Rainwater harvesting and greywater reuse systems can generate economic and environmental benefits, especially when adopted on a large scale, helping to solve problems of lack of drinking water and poor distribution of water in many regions. Thus, this work seeks to contribute to the development of studies on the feasibility of using rainwater and reusing greywater in multifamily buildings.

The main objective of this work is to assess the economic viability of rainwater harvesting and greywater reuse in a multifamily building. The water end-uses, the reduction of water released into the urban drainage system when rainwater is used, and the energy consumption to operate the motor pumps in rainwater harvesting and greywater reuse systems were also estimated.

2. Methodology

2.1. City and Building

The study area chosen for the development of this work is the city of Florianópolis, located in the state of Santa Catarina, southern Brazil (Figure 1). The city has a population of 516,524 people and an area of 674,844 km², with an average density of 623.68 inhabitants/km² [23]. The region’s climate is humid subtropical (CFa), according to the Köppen–Geiger classification.

This study focuses on the Recanto dos Imigrantes, a multifamily building located at Córrego Grande neighbourhood, in Florianópolis. The building is composed of two blocks, with 60 flats each, and has a total floorplan area of 18,008 m² (Figure 2).

2.2. Building Diagnosis

2.2.1. Number of Residents

The initial phase of the building’s diagnosis consisted of applying questionnaires to residents to obtain the characteristics of the residents. The objective of this initial process was to find the number of residents per flat and the total number of residents in the building.

2.2.2. Water Consumption

The water consumption was obtained through the consumption history available in the water bills for a period of one year. Such a consumption history only presented the sum of consumption of all flats together.

2.2.3. Water End-Uses

The water end-uses were obtained by asking residents, by means of questionnaires, about the uses of water in each flat and by measuring the average water flow rates in the water appliances. In the questionnaire, residents recorded the number of times they used each water appliance and the average time of use. The flow rates of the water appliances were obtained by measuring the time needed to fill containers of known capacities in one of the flats. Three water flow measurements were carried out for each water appliance, using the average as a standard flow. The total water use in each water appliance was estimated by using Equation (1).

$$C_i=N\times D\times V$$

(1)

where C_i is the total use of water in water appliance i; N is the number of uses; D is the average duration of use (s); V is the water flow rate (L/s).

The total water use in the flats was estimated by summing the water use of each water appliance (Equation (2)).

$$C_f=\sum _{i=1}^n{C}_i$$

(2)

where C_f is the total water use in the flat; C_i is the total use of water in water appliance i.

The total water consumption in the flats, calculated through the application of questionnaires, was compared to the consumption history obtained through the water bills in order to verify the validity of the data collected.

Data on water end-uses were used to verify the volume of water consumed for potable and non-potable purposes, allowing to verify the percentage of the current demand that could be met with rainwater or greywater.

2.2.4. Non-Potable Water Demand

The demand for non-potable water in buildings could be supplied by rainwater or greywater. The drinking water used for flushing toilets represents one of the largest non-potable demands in residential buildings, and could be replaced with either rainwater or greywater. The drinking water used in washing machines is also significant in residential buildings, but the quality of water required for this purpose is higher than the one for flushing toilets.

The volume of rainwater that can be supplied through the use of rainwater harvesting systems in multifamily buildings is limited, among other factors, by the small catchment area in relation to the high demand. For greywater reuse systems there is also a limitation on the volume of greywater that can be supplied, among other factors, due to the limited area available for installing wetland systems and the treatment capacity of such systems. In view of this, the rainwater harvesting system analysed herein was considered to provide water for washing machines only; and the greywater reuse system was considered to provide water for toilet flushing only. Therefore, the rainwater demand considered in this work is equal to the volume of water used in washing machines. In addition, scenarios with variations of ±5% in demand were considered, with the aim of reducing imprecision in defining the water end-uses by residents and expanding the scenarios analysed. The greywater demand considered in this study is equal to the volume of water used for flushing toilets.

2.3. Rainwater Harvesting

The rainwater harvesting system proposed herein consists of harvesting rainwater through the roof, then there is the first flush, the conveying of rainwater to a lower tank by gravity, pumping the rainwater to an upper tank, and disinfection by chlorination (Figure 3). From the upper tank, rainwater is conveyed by gravity to the flats to be used in washing machines.

2.3.1. Roof Area

The catchment area is the horizontally designed waterproof area where rainwater is harvested [24]. The rainwater harvesting area is made up of the roof areas of the two blocks, approximately 785 m² each, and the areas of the leisure spaces located between the blocks, with approximately 185 m², and behind the blocks, with approximately 75 m². The total area for rainwater harvesting is 1830 m².

Rainwater collected from roofs suffers losses due to evaporation, absorption of materials, and cleaning before being collected and directed to storage tanks. Therefore, it is necessary to consider such losses when analysing rainwater harvesting systems using a surface runoff coefficient. As the roof is composed of ceramic tiles and concrete slabs, a runoff coefficient of 0.8 was considered [25].

2.3.2. Rainfall Data

Rainfall data for the city of Florianópolis are available on the HidroWeb platform of the National Water Agency [26]. A series of daily rainfall lasting 12 years (January 2002 to December 2013) was used, which was the longest series available without missing data. According to Geraldi and Ghisi [27], the use of short rainfall data series in the evaluation of rainwater harvesting systems is valid if the rainfall characteristics in the region are maintained, such as the number of days with and without rainfall throughout the year.

2.3.3. Rainwater Tank Sizing

The sizing of the rainwater harvesting system was carried out using the Netuno computer programme, version 4 [28]. Input data are daily rainfall data, first flush, catchment area, per capita water demand, number of residents, rainwater demand, and surface runoff coefficient. The programme estimates, among other output data, the potential for potable water savings for certain tank capacities.

In order to discard dirt accumulated on roofs and gutters, in addition to reducing the contamination load of rainwater collected, a first flush of 2 mm was adopted according to the Brazilian standard NBR 15527 [24].

Netuno also allows for finding the rainwater tank capacity considered ideal for the system. The capacity of the upper rainwater tank was considered equal to the daily rainwater demand. We considered that when the upper tank is below 20% of its capacity, the motor pump pumps rainwater from the lower tank, if available, to the upper tank. For finding the ideal tank capacity, it is necessary to input into the programme the minimum and maximum capacities to be simulated and the interval of capacities to be tested by the programme.

The Netuno computer programme, after estimating the potential for potable water savings of each capacity tested, indicates the ideal capacity of the lower rainwater tank. In this study, lower tank capacities between 1000 and 50,000 L were tested, with intervals of 1000 L. According to the variation in the potential for potable water savings for each tank capacity, the ideal capacity can be chosen. The variation adopted to choose the ideal tank capacity was 2%/m³.

The motor pumps power was estimated considering the Brazilian standards. Head loss, water flow, pump efficiency, static head, and manometric head were taken into account.

2.4. Greywater Reuse

The greywater treatment system considered herein is a wetland of the vertical-flow type, which requires little maintenance, occupies small spaces, and has high efficiency in removing organic compounds and suspended solids compared to other types of systems [29]. Effluents from washbasins and showers were considered greywater, as they present lower risks of contamination compared to other sources [30]. As a disinfection process to achieve parameters necessary for use, the effluent from the wetland underwent a chlorination process before being used [31,32]. In this study, greywater was used exclusively for flushing toilets.

The area available for installing the greywater treatment system is the garden. This area was selected due to the possibility of using ornamental plants in the wetland [33], which can be enjoyable to the residents. The total garden area is approximately 375 m². However, a large part of the garden area is spread out in small flowerbeds, which would make it impossible to use the entire green area for implementing a wetland-type treatment system. Therefore, only the largest garden area, measuring approximately 80 m², located at the back of the building, was used to implement the system.

For sizing the greywater system, the same standards and parameters used in another vertical wetland system operating in the city of Florianópolis were adopted. Such a system, developed with the support of the Decentralized Sanitation Study Group (GESAD), was monitored and had its efficiency proven by Monteiro [34]. Therefore, the wetland system proposed herein is composed of Cyperus papyrus as the vegetation layer, a superficial layer of 10 cm of gravel for distributing the effluent, 50 cm of coarse sand as an intermediate layer, and 10 cm of gravel at the bottom for drainage (Figure 4).

The system evaluated by Monteiro [34] has a surface area of 9.24 m² and a design flow of 250 L per day, i.e. approximately 27 L per day for each square meter of surface area. Maintaining this design flow in the wetland system proposed, with 80 m² of surface area, the potential volume of greywater that the system would be able to supply daily was considered as 2160 L (2.2 m³).

The parameters for waterproofing and plant management were selected using a guide for sizing constructed wetlands [36], which was organised by GESAD and based on a consensus between researchers and builders.

A schematic of the greywater system is shown in Figure 5.

The potential for potable water savings through the use of greywater was obtained based on the volume of water that the reuse system can supply in relation to the total water demand of the building.

2.5. Economic Analysis

The economic analysis was carried out considering the construction of a new building with characteristics similar to the building used as a case study. This analysis was performed by considering the cash flow over a period of twenty years. Cash flow considers the initial outlay, operation costs, and economic benefits obtained through the financial savings in the water and sewage bills.

The costs of rainwater harvesting and greywater reuse systems were estimated based on the costs of the main components of such systems, that is, those that represent the highest costs (motor pumps, water tanks, and materials used in the layers of the wetland system). The costs of piping and accessories were considered to be 15% of the cost of the main components.

The costs of the main components were obtained through quotes from three stores located in Florianópolis. For the analysis, the lowest costs found were adopted. The labour costs required to install the systems were considered to be 15% of the cost of the main components.

The operating costs were taken as the total cost of energy consumption for motor pumps and the cost of chlorine disinfection over the twenty-year analysis period. The concentration of chlorine used for disinfection was 1.0 mg/L, within the range recommended by Brazilian standards for the use of alternative sources of non-potable water in buildings, i.e. between 0.5 and 2 mg/L [7].

Adjustments to water, sewage, and energy tariffs were adopted annually and considered equal to the inflation rate. The energy tariff for the residential sector, obtained by consulting the local energy company, is BRL 0.57302/kWh. The water and sewage tariffs, obtained by consulting the local water company, are shown in

Table 1. Water and sewage tariffs for the residential sector.

Water Consumption (m3)	Tariff (BRL/m3)
Flat rate	35.08
1 to 10	2.33
11 to 25	10.84
26 to 50	14.49
51 to 999,999	18.23

.

Three inflation-rate scenarios were considered, with the aim of verifying the impact of the national economic scenario on the viability of the systems. The monthly inflation used in the average scenario was 0.93% per month (11.73% per year)—obtained through the Broad National Consumer Price Index [37]—which represents the average monthly inflation over one year, i.e. April 2021 to May 2022. For the other scenarios, variations of ±5% were adopted, resulting in an average inflation of 0.54% per month (6.73% per year) in the lowest inflation scenario and 1.30% per month (16.73% per year) in the highest inflation scenario.

The minimum attractive rate of return adopted was 1% per month (12.75% per year), a value equal to the interest rate defined by the Central Bank of Brazil for the month of May 2022. Such a rate represents the return on some of the most common investments in Brazil.

The financial benefit was estimated through the difference in water and sewage bills. The bills in the current situation, i.e. without rainwater harvesting and greywater reuse systems, were compared to the bills after the installation of the systems, with lower costs due to lower consumption. Three financial indices were analysed, i.e. net present value (NPV), payback period, and internal rate of return (IRR).

The results of the economic analyses were evaluated and compared to verify whether both systems are economically viable when installed separately or together.

2.6. Urban Drainage

The volume of stormwater contribution to the urban stormwater drainage system is directly related to the types of surfaces that make up the building, as each material has a different surface runoff coefficient. This coefficient represents the relationship between the total volume that flows superficially and the total volume precipitated [24].

The building has areas considered permeable, such as gardens and lawns, and impermeable areas, such as roofs, sports courts, sidewalks, and swimming pools. Much of the rainwater in the building is currently drained and sent to the urban drainage system. Following the installation of the rainwater harvesting system, part of the rainwater that would flow to the drainage system will be stored in water tanks and used by the residents, thus reducing the pressure in the drainage system.

The impact of rainwater harvesting systems on the urban drainage system was obtained through the comparisons of the total volume of rainwater stored and used by residents that would no longer flow to the urban drainage system with the total volume of stormwater that flows to the drainage system in the scenario with no rainwater harvesting (Equation (3)).

$$I_d=\frac{V_p}{V_c}$$

(3)

where I_d is the impact of the rainwater harvesting system on the urban drainage system (%); V_p is the volume of rainwater harvested and used in the building (m³/year); and V_c is the volume of stormwater that flows to the urban drainage system (m³/year).

The volume of stormwater that flows to the urban drainage system was obtained using Equation (4). The runoff coefficients adopted were 0.25 for lawn and garden areas, 0.75 for stone sidewalks, 0.80 for roofs, and 0.95 for concrete areas [25].

$$V_c=\sum _{i=1}^{n}P\times A_i\times C_i$$

(4)

where V_c is the volume of stormwater that flows to the urban drainage system (m³/year); P is the annual rainfall (m/year); A_i is the catchment area i (m²); C_i is the runoff coefficient for area i (non-dimensional); and n is the number of areas with different runoff coefficients.

2.7. Energy Consumption

Rainwater harvesting systems with lower and upper water tanks need motor pumps to pump rainwater stored in the lower tank to the upper tank for subsequent use by gravity. Greywater reuse systems also need motor pumps, but with the purpose of pumping greywater to the wetland system and from there to the upper tank for subsequent use by gravity.

The impact of rainwater harvesting and greywater reuse systems in relation to energy consumption was obtained by analysing the energy consumption required in each system to supply one cubic metre of water in comparison to centralised supply systems of drinking water and sewage treatment. The energy consumption per cubic metre of water supplied was obtained using Equation (5).

$$C=\frac{{E^i}_c}{{V^i}_s}$$

(5)

where C is the energy consumption per cubic metre of water supplied (kWh/m³); Eⁱ_c is the energy consumed by the motor pumps during period i (kWh); and Vⁱ_s is the volume of non-potable water (rainwater or reuse) supplied during period i (m³).

The energy consumed during the operation phase was estimated using Equation (6).

$${E^i}_c=\frac{Pw}{n}\times t$$

(6)

where ${E^i}_c$ is the energy consumed by the motor pumps during period i (kWh); Pw is the power of the motor pump (kW); n is the efficiency of the motor pump (non-dimensional); and t is the total time that the pump was on over period i (hours).

For the analysis of the energy consumption of the greywater reuse system, the energy that is no longer consumed for sewage treatment in traditional centralised treatment plants was considered. The energy balance for this system was estimated using Equation (7).

$$B_r=C_r-C_{cs}$$

(7)

where B_r is the energy balance of the greywater reuse system during its operational phase per cubic metre of greywater (kWh/m³); C_r is the energy consumption per cubic metre of greywater supplied (kWh/m³); and C_cs is the energy consumption per cubic metre of sewage treated in the centralised sewage treatment system (kWh/m³).

3. Results and Discussion

3.1. Building Diagnosis

Based on the questionnaires, the average population was 2.3 people per flat. This average population, if considered constant for all flats (120), results in an estimated total population of 276 people (138 in each block of 60 flats).

The drinking water consumption was obtained by checking the water bills over a period of one year, from July 2021 to June 2022. As shown in Figure 6, the average monthly consumption was 1137 m³ per month, which represents approximately 9.5 m³ per month per flat. December was the month with the highest consumption, 1268 m³ in the building and approximately 10.6 m³ per flat. January presented the lowest consumption, i.e. 968 m³ in the building and approximately 8.1 m³ per flat.

Considering a 30-day month and the estimated population of 276 people, the average per capita consumption is 137 L per person per day. Such a consumption is approximately 9.8% lower than the average per capita consumption in the state of Santa Catarina obtained in 2019, i.e. 152 L per person per day [38].

The average frequencies and durations of use of drinking water in each water appliance in the flats are shown in

Table 2. Frequencies and average duration of drinking water use per person.

Appliance	Frequency (Times/Day)			Duration (Min)
	Minimum	Maximum	Average	Minimum	Maximum	Average
Shower	1.0	2.0	1.1	2.0	20.0	8.5
Toilet flushing	2.3	10.0	4.7	-	-	-
Tooth brushing	2.0	4.0	2.4	0.1	2.0	0.6
Hands washing	1.7	10.0	5.4	0.1	0.5	0.2
Face washing	1.0	2.3	1.3	0.1	0.7	0.3
Cooking	0.3	3.3	1.3	0.1	5.0	1.5
Dishwashing by hand	0.3	2.0	1.0	0.5	15.0	6.4
Dishwasher	0.2	0.3	0.3	-	-	-
Outros	0.1	1.0	0.2	-	-	-
Washing machine	0.5	3.0	1.6	-	-	-
Washing clothes by hand	0.5	5.0	1.2	1.0	10.0	4.1

. In other uses, reported by some residents, the use of water for watering plants, cleaning, personal consumption, and animal consumption was mentioned.

To calculate the water end-uses, water consumption for toilet flushing was considered equal to 6 L per flush, as the flats have bowl-and-tank toilets. The volume of water consumed in each use of the dishwasher and washing machine was considered equal to 25 L and 110 L, respectively. Such volumes were obtained through the analysis of the capacities and types of toilets mentioned by residents in the questionnaire. All residents who had dishwashers, 14.3%, had models that used 25 L per cycle. The average capacity of the washing machines, present in all flats, was 11.2 kg. The average flow rates of the water appliances, necessary to calculate the volume of water consumed for each end-use, were measured in one of the flats and are shown in

Table 3. Water flows measured and considered in the analysis.

Appliance	Water Flow (Ls/s)
	Minimum	Maximum	Average
Shower	0.09	0.12	0.10
Lavatory taps	0.03	0.04	0.04
Kitchen taps	0.07	0.09	0.08
Laundry taps	0.05	0.06	0.05

.

The daily per capita water consumption in the building, estimated through information recorded by residents in questionnaires and the average water flow measured, was approximately 163 L per person per day. Such a consumption is approximately 18.5% higher than the average daily per capita consumption estimated through the water consumption history, which was 137 L per person per day. This difference shows the importance of adopting scenarios with different demands for non-potable water in the analysis in order to reduce the effects of possible inaccuracies and increase the scope and confidence of the results. Some possible explanations for this difference are inaccuracies in recording frequencies and durations of use of water appliances in the questionnaires, variation in the average flow rates of water appliances in each flat and for each user, low response rate to questionnaires (11.7%), and non-occupied flats.

The largest water end-uses in the building were showers (35.8%), washing dishes manually (19.1%), flushing toilets (17.5%), and washing machines (15.1%), representing approximately 87.5% of the total water consumption (

Table 4. Daily per capita water consumption for the building analysed herein.

Appliance	Water Consumption
	(Ls/Person.Day)	(%)
Shower	58.3	35.8
Dishwashing by hand	31.1	19.1
Toilet flushing	28.4	17.5
Washing machine	24.6	15.1
Cooking	9.0	5.5
Tooth brushing	3.7	2.3
Hands washing	2.9	1.8
Washing clothes by hand	2.1	1.3
Dishwasher	1.6	1.0
Face washing	1.0	0.6
Others	0.2	0.1
Total	162.7	100.0

). These results are similar to those found in the pieces of research cited in Table 2. It is possible to observe that non-potable water end-uses, i.e. toilet flushing and washing machines, represent approximately 32.6% of the total drinking water consumption.

In this study, it was considered that the rainwater harvesting system provides water for use in the washing machines. Therefore, the rainwater demand obtained through the application of the questionnaires is approximately 15.1%. Adopting variations of ±5% in the estimated demand to carry out the analyses, scenarios with rainwater demands of 10.1%, 15.1%, and 20.1% were evaluated. The greywater reuse system provides water for flushing toilets. Therefore, the demand for greywater obtained through the application of the questionnaires is approximately 17.5%.

3.2. Rainwater Harvesting

The rainfall data for Florianópolis were obtained from the HidroWeb platform of the National Water Agency [26]. According to the largest series of complete data available, from 1 January 2002 to 31 December 2013, the average annual precipitation in the municipality was 1766 mm/year (Figure 7). The average monthly rainfall was 147 mm/month.

Figure 8 shows the minimum, average, and maximum monthly rainfall in Florianópolis over the period 2002–2013. Blue bars show the average monthly rainfall, and the vertical lines show the minimum and maximum monthly rainfall. For example, over 2002–2013, the average rainfall in January was 228 mm, but the minimum was 87 mm and the maximum was 375 mm.

Rainfall in Florianópolis is well distributed throughout the year, with monthly averages between 100 and 200 mm most of the year. January presents rainfall greater than the average. June and July have the lowest monthly average rainfall, being the only months with average rainfall below 100 mm/month. This balanced distribution of precipitation throughout the year indicates that rainwater harvesting systems may be able to constantly supply water for non-potable purposes throughout the year. However, it is expected that in the months of June and July the rainwater supply will be relatively lower, thus increasing the consumption of drinking water in such periods.

3.2.1. Potential for Potable Water Savings

The input data for the Netuno programme, necessary to carry out the sizing of rainwater tanks and calculate the potential for potable water savings, are shown in

Table 5. Input data used to carry out the simulations in Netuno programme.

Input Data
First flush (mm)	2
Catchment area (m2)	1830
Total water demand (Ls/person.day)	137
Number of people	276
Rainwater demand (% of the total water demand)	10.1; 15.1; 20.1
Runoff coefficient	0.8
Lower tank capacities (Ls)	1000 to 50,000 *
Upper tank capacites (Ls)	2000; 3000; 5000

Note: * At intervals of 1000 L.

.

The capacity of the upper rainwater tank was calculated to meet the daily rainwater demand. Considering the daily drinking water consumption recorded by the water company of approximately 38 m³ per day and the estimated daily rainwater demand, equal to the water demand of washing machines (15.1%), the volume found was 5722 L (2861 L per block). Therefore, a 3000-L tank, which is available for sale, was used as the upper tank in each block. For the scenario with rainwater demand equal to 10.1%, the volume found was 3827 L (1923 L per block), thus 2000-L tanks were used. For the scenario with rainwater demand equal to 20.1%, the volume found was 7616 L (3808 L per block), and thus 5000-L tanks were used.

For determining the lower tank capacity, the potential for potable water savings for each tank capacity was estimated by Netuno considering the three rainwater demands. Results are shown in Figure 9.

Considering the rainwater demand of 15.1%, the variation in the potential for potable water savings between one capacity and another was less than 2%/m³ for the capacity of 12,000 L, therefore such a capacity was adopted for the lower tank in this scenario. For the demand of 10.1%, this variation in potential savings also occurred for the capacity of 12,000 L. And for the demand of 20.1%, a lower tank with 11,000-L capacity was obtained. However, as the nearest tank capacity found in local stores is 12,000 L, such a capacity was also adopted for this scenario.

In the scenario with rainwater demand equal to 15.1%, adopting upper tanks with 6000-L capacity (3000-L for each block) and a lower tank with 12,000-L capacity, the potential for potable water savings found by Netuno was 6.9%. In the scenario with rainwater demand equal to 10.1%, the potential for potable water savings was 5.8%. For the scenario with the highest rainwater demand, 20.1%, the potential savings was 7.2% (

Table 6. Potential for potable water savings for each scenario analysed.

Rainwater Demand (%)	Upper Tank Capacities (L)	Lower Tank Capacities (L)	Potential for Water Savings
			(%)	(m3)
10.1	2000	12,000	5.8	67
15.1	3000	12,000	6.9	78
20.1	5000	12,000	7.2	82

). Such savings are lower than the demand for rainwater for washing machines, i.e. 15.1%. The water savings could be greater through the use of larger tanks, which significantly increase outlay costs. However, even using large tank capacities, the rainwater demand could not be fully met due to the limited catchment area and local rainfall.

By analysing the monthly water demand of the building, which is approximately 1137 m³, the potential for potable water savings found means that the rainwater harvesting system could provide from 67 m³ to 82 m³ of water per month. For flats, with an average monthly water consumption of approximately 9.5 m³, the potential savings found means that the rainwater harvesting system could provide from 0.5 m³ to 0.7 m³ of water per month for each flat.

3.2.2. Economic Analysis

The total cost of installing the rainwater harvesting system in a building with characteristics similar to Recanto dos Imigrantes is made up of the costs of the upper and lower tanks, two motor pumps for each block (main and reserve), automatic chlorinator, labour, and piping and accessories. Based on a survey in local stores in September 2022, the total cost of installing the rainwater harvesting system ranged from BRL 17,661.25 to BRL 21,509.51, depending on the rainwater demand.

The costs and monthly financial benefits obtained through the use of rainwater harvesting systems, resulting from reductions in water and sewage bills, are shown in

Table 7. Costs and monthly benefits for the three scenarios of rainwater harvesting.

Rainwater Demand (%)	Potential for Water Savings (%)	Costs (BRL)	Savings (BRL/Month)
10.1	5.8	17,661.25	314.22
15.1	6.9	19,036.91	368.67
20.1	7.2	21,509.51	386.79

.

The results of the economic analyses for the three different inflation scenarios are shown in

Table 8. Results of the economic analysis for the rainwater harvesting system.

Rainwater Demand (%)	Annual Inflation (%)	Financial Indicator
		NPV (BRL)	Payback (Months)	IRR (% per Month)
10.1	Low (6.73)	22,067.90	76	2.1
	Average (11.73)	41,290.64	68	2.4
	High (16.73)	82,278.81	60	2.8
15.1	Low (6.73)	27,578.58	70	2.2
	Average (11.73)	50,147.71	62	2.6
	High (16.73)	98,271.29	57	3.0
20.1	Low (6.73)	27,420.55	75	2.1
	Average (11.73)	51,104.81	67	2.4
	High (16.73)	101,606.17	60	2.9

. The positive net present values indicate that the system is economically viable for all rainwater demands and inflation scenarios analysed. The internal rates of return were higher than the minimum attractive rate of return in all scenarios, indicating that this system would be a great investment option in economic terms compared to other types of investments.

The payback periods found varied between 57 months (4.75 years) and 76 months (6.33 years). These payback periods are within the range of results found by Bashar et al. [9] in multifamily buildings in Bangladesh, from 3 to 10 years. Maykot and Ghisi [10] found a payback period of 57 months for a rainwater harvesting system in a multifamily building in Florianópolis. Ghisi and Ferreira [1], also in Florianópolis, found shorter payback periods, between 2.4 and 5 years.

In summary, rainwater harvesting systems were economically viable in all tested scenarios. However, the best financial benefits (higher net present values, shorter payback periods and higher internal rates of return) occurred in scenarios of higher inflation. In these scenarios, the financial benefits (reduction in water and sewage tariffs) have their results positively impacted during the analysis period, as it is considered that water and sewage tariffs are adjusted annually according to inflation. Regarding the rainwater demand, the scenario with intermediate demand, equal to 15.1%, presented the best financial benefit in terms of payback period and internal rate of return.

3.3. Greywater Reuse

3.3.1. Potential for Potable Water Savings

The greywater reuse system used as a reference was designed for a daily flow of 250 L on a surface area of 9.24 m² [34]. For the system proposed herein, located in the same city (Florianópolis), with similar characteristics of raw effluents (from wash basins and showers) and treated effluents (toilet flushing), flow (vertical), size, and layer compositions, and type of plant used (Cyperus papyrus), with 80 m² of surface area, the volume of greywater that the system could supply daily was considered as 2160 L (2.2 m³).

Considering the total water demand of approximately 38 m³, the potential for drinking water savings through the use of the greywater reuse system is approximately 5.7% (2.2 m³). Considering the monthly water demand of approximately 1137 m³, the potential volume of drinking water savings is approximately 65 m³. Such potential savings are lower than the demand for greywater for toilet flushing, which is 17.5%. The savings could be greater through the use of greater daily effluent flows in the wetland system proposed, which could affect the efficiency of the system, proven by Monteiro [34] for the parameters and flows proposed in this study. The use of other compact treatment technologies could also be an alternative to increase the supply of treated greywater and the potential for drinking water savings.

3.3.2. Economic Analysis

The initial outlay of installing a greywater reuse system, using a vertical-flow wetland with a surface area of 80 m² as the main treatment and chlorination as final disinfection, in a building with characteristics similar to Recanto dos Imigrantes is BRL 37,433.87.

This cost considered the use of two ½ HP motor pumps (main and reserve) to pump the greywater from the equalizing tank to the wetland system and two 1 HP motor pumps (main and reserve) to pump the greywater from the lower tank to the upper tank on the roof of each block. The capacity adopted for the equalizing tank and upper tanks was 3000 L, as it was the cheapest commercial capacity available above the daily treatment capacity of the system (2.2 m³). For the lower tank, a commercial capacity of 5000 L was adopted, which is the commercial capacity available above twice the system’s proposed daily treatment capacity (4.4 m³).

The monthly financial benefit obtained through the use of the greywater reuse system was BRL 306.80 (savings on water and sewage bills). The results of the economic analyses for the three different inflation scenarios are shown in

Table 9. Results of the economic analysis of the greywater reuse system.

Annual Inflation (%)	Financial Indicator
	NPV (BRL)	Payback (Months)	IRR (% per Month)
Low (6.73)	−83.77	-	1.0
Average (11.73)	17,601.54	159	1.4
High (16.73)	47,620.57	127	1.7

. The net present value was positive for the medium and high inflation scenarios, indicating the economic viability of these systems in these scenarios. However, in the low inflation scenario, which presents lower economic benefits, the net present value was negative, indicating that the reuse system is not economically viable in a lower inflation scenario.

The payback periods found were shorter than the analysis period in the average inflation scenario, 159 months (13.25 years), and the high inflation scenario, 127 months (10.58 years), indicating that greywater is economically viable. In the low inflation scenario, the reuse system did not have a payback period shorter than the analysis period (twenty years). The payback periods found herein were longer than those found in the literature for multifamily buildings; Kotsia et al. [33] found 4.7 years in Greece, Mourad et al. [39] found between 3 and 7 years in Syria, and Ghisi and Ferreira [1] found between 2.1 and 5 years in Florianópolis.

In summary, greywater reuse systems are economically viable in medium and high inflation scenarios. However, in low inflation scenarios, the financial feasibility can be considered low and even negative. Therefore, in low inflation scenarios there may be other investment alternatives that provide better financial benefit for the investor. Like rainwater harvesting systems, the best financial benefit (higher net present values, shorter payback periods, and higher internal rates of return) occurs in scenarios of higher inflation, as water and sewage tariffs are adjusted annually for inflation.

3.4. Rainwater and Greywater Combined

3.4.1. Potential for Potable Water Savings

The potential for water savings in scenarios with rainwater harvesting and greywater reuse systems combined, obtained by summing the potentials of each system individually, is shown in

Table 10. Potential for water savings when rainwater and greywater are combined.

Potential for Water Savings (%)			Water Consumption Savings (m3/Month)
Rainwater	Greywater	Total
5.8	5.7	11.5	132
6.9	5.7	12.6	143
7.2	5.7	12.9	147

.

It is possible to observe that the potential for water savings varies between 11.5% and 12.9%, depending on the rainwater demand. The monthly reduction in drinking water consumption varied between 132 m³ and 147 m³ for the building and between 1.1 m³ and 1.2 m³ for each flat.

3.4.2. Economic Analysis

The initial outlay of installing the rainwater harvesting and greywater reuse systems is approximately BRL 56,470.78. The monthly financial benefits obtained through the use of the two systems are shown in

Table 11. Monthly financial benefits provided by the rainwater harvesting and greywater reuse systems combined.

Potential for Water Savings (%)			Savings in the Water Bill (BRL/Month)
Rainwater	Greywater	Total
5.8	5.7	11.5	621.02
6.9	5.7	12.6	675.47
7.2	5.7	12.9	693.59

for each non-potable water demand analysed.

The results of the economic analysis are shown in

Table 12. Economic analysis of rainwater harvesting and greywater reuse systems combined.

Annual Inflation (% per Year)	Financial Indicator
	NPV (BRL)	Payback (Months)	IRR (% per Month)
Low (6.73)	27,494.81	132	1.4
Average (11.73)	67,749.25	105	1.8
High (16.73)	145,893.84	89	2.2

. It is possible to observe that the payback periods varied between 89 months (7.42 years) and 132 months (11 years). Such payback periods are longer than those found by Ghisi and Ferreira [1], between 3.4 and 8 years. Internal rates of return varied between 1.4% and 2.2% per month, above the minimum attractive rate of return in all inflation scenarios analysed.

In short, the better financial indicators of rainwater harvesting systems offset the financial indicators of greywater reuse systems, which are less attractive, especially in the low inflation scenario, in which the reuse system alone is not economically viable. Therefore, when considering the installation of both systems together, the investment is feasible in all inflation scenarios analysed.

3.5. Urban Drainage

The building is located on a piece of land of approximately 4540 m². The total permeable area, consisting of gardens and lawns, is approximately 375 m² (8.3% of the land surface). The total impermeable area, consisting of roofs, sports courts, sidewalks, swimming pool, and access to the garage, is approximately 4165 m² (91.7% of the land surface).

As for the impermeable area, the roofs represent 1830 m² (40.3% of the surface area), the sidewalks represent 402 m² (8.9% of the surface area), and the concrete surfaces represent approximately 1933 m² (42.6% of the surface area).

The current volume of rainwater that flows to the urban stormwater drainage system, obtained through Equation (4), is approximately 668 m³ per month (8018 m³ per year). The rainwater harvesting system proposed, with a potential for water savings of 6.9%, is capable of supplying approximately 78 m³ of water per month (949 m³ per year) for the washing machines. The impact on the urban stormwater drainage system, represented by the potential reduction in the building’s contribution volume, is approximately 11.7%. In Italy, Palla et al. [17] found an average potential for reducing runoff volumes of 26%.

In the rainwater harvesting system, using a lower tank of 12,000 L and an upper tank of 6000 L (3000 L for each block), approximately 130 m³ of water per month (1556 m³ per year) overflows due to limited storage capacity. Such a volume of rainwater represents approximately 19.4% of the volume of rainwater that flows to the urban drainage system in the scenario without a rainwater harvesting system. Therefore, the adoption of larger tanks, despite generating higher initial outlays and directly impacting economic viability, can be an alternative to obtaining a greater reduction in the volume of stormwater to the urban drainage system.

3.6. Energy Consumption

Energy consumption represents one of the main costs of centralised water supply and sewage systems. According to the National Sanitation Information System, the electricity consumption rate in water supply systems in Brazil in 2020 was approximately 0.73 kWh/m³ [40]. The northern region has the lowest rate, approximately 0.62 kWh/m³. The highest rate of energy consumption in drinking water supply is found in the northeastern region, approximately 0.84 kWh/m³. In the southern region of Brazil, this rate is approximately 0.69 kWh/m³. These figures only take into account the energy consumption in the treatment and distribution processes. Therefore, there is still energy consumption for pumping drinking water from the lower tanks to the upper tanks in the building.

In centralised sewage treatment systems in Brazil, the energy consumption rate in 2020 was approximately 0.27 kWh/m³. The southern and northeastern regions of Brazil have the highest rates, approximately 0.29 kWh/m³. The southeastern and northern regions have the lowest rates, approximately 0.26 kWh/m³ [40].

The rainwater harvesting system proposed herein takes 1-HP motor pump to pump water from the lower tank to the upper tank. The annual volume of rainwater that the system can supply is approximately 949 m³. The annual energy consumption due to motor pumps is approximately 536 kWh. Therefore, the energy consumption of the rainwater harvesting system proposed for the building with characteristics similar to the one studied herein is approximately 0.56 kWh/m³. Such a consumption is lower than the one found by Vieira and Ghisi [22] for low-standard single-family homes in Florianópolis, which was 0.86 kWh/m³.

The greywater reuse system takes a ½ HP motor pump to pump water from the equalizing tank to the wetland system and a 1 HP motor pump to pump water from the lower tank to the upper tank. The annual volume of greywater that the system can supply is approximately 790 m³. The annual energy consumption due to motor pumps is approximately 707 kWh. Therefore, the energy consumption of the greywater reuse system proposed for the building with characteristics similar to the one studied herein is approximately 0.89 kWh/m³. Such a consumption is greater than the consumption found by Vieira and Ghisi [22] for low-standard single-family homes in Florianópolis, which was 0.54 kWh/m³.

It is necessary to emphasise that greywater reuse systems reduce the volume of sewage produced and sent to centralised treatment systems. Considering that the energy consumption of greywater treatment systems in the southern region of Brazil is approximately 0.29 kWh/m³, it is possible to verify that the energy balance in the operational phase of the greywater reuse system would be approximately 0.60 kWh/m³. Therefore, this type of system, by reducing sewage production and providing the treatment and supply of water for non-potable purposes in a decentralised manner, is a good alternative in terms of energy consumption during the operation phase in comparison to centralised systems for water supply and sewage treatment.

4. Conclusions

The objective of this study was to verify the economic viability of rainwater harvesting and greywater reuse in a multifamily building located in Florianópolis in southern Brazil. Furthermore, the impacts in relation to urban drainage and the energy consumption of the systems were estimated. The two systems proved to be economically viable when installed in a building with characteristics similar to the building analysed herein, especially in scenarios with high inflation.

The water end-uses were estimated by applying questionnaires to residents, who responded about the frequencies and durations of use of the water appliances. The greatest end-uses were showering (35.8%), washing dishes manually (19.1%), toilet flushing (17.5%), and using the washing machine (15.1%). Therefore, it is possible to observe that non-potable uses, such as toilet flushing and washing machines, correspond to 32.6% of the water consumption in the building. This represents a great potential for exploring other non-potable sources of supply in multifamily buildings, such as the use of rainwater and the reuse of greywater.

The rainwater harvesting system evaluated herein, which was intended to supply rainwater for washing machines, was economically viable in all rainwater demand and inflation scenarios analysed. The payback period found varied between 57 and 76 months.

The greywater reuse system, aiming to meet part of the water demand for toilet flushing, was economically viable for medium and high inflation scenarios. However, in the scenario with low inflation, such a system was not economically viable, as the financial benefits resulting from the reductions in water and sewage bills did not make up for the high initial outlay. The payback period found for the average inflation scenario was 127 months. For the high inflation scenario, the payback period was 159 months.

Rainwater and greywater considered together were economically viable in all scenarios evaluated. The high viability of rainwater harvesting made up for the lack of feasibility for the greywater system in the low inflation scenario. The payback period for the two systems together varied between 89 and 132 months.

It is necessary to highlight that the new water and sewage tariff policy of the local water company, which does not apply a flat rate anymore for water consumption of up to 10 m³ per month per flat, shows potential to make water-saving strategies economically viable. Such a new tariff policy allows rainwater harvesting and greywater reuse systems to be economically viable and have shorter payback periods in low consumption scenarios. If the previous tariff policy was still applied, no system would be economically viable for the flats with water consumption less than 10 m³ per month.

The rainwater harvesting system had the potential to reduce the volume of stormwater to the urban drainage system by approximately 11.7% in relation to the current volume. This figure could be higher if larger rainwater tank capacities were adopted, which would also increase the potential for drinking water savings, but would increase costs and affect the economic viability. In this way, the large-scale implementation of rainwater harvesting systems in multifamily buildings can contribute to reducing the demand for urban drainage systems.

The energy consumption of the rainwater harvesting system is 0.56 kWh/m³ of rainwater supplied. The energy consumption of the greywater reuse system is 0.89 kWh/m³ of treated greywater supplied. Considering the average Brazilian energy consumption and the reduction in the volume of sewage through greywater reuse systems, both systems proved to be efficient in terms of energy consumption in the operational phase to provide the same volumes of water as in centralised systems. Therefore, the installation of such systems on a larger scale could represent a potential reduction in electricity consumption in the country.

The results of this study demonstrate that rainwater harvesting and greywater reuse systems in multifamily buildings can bring positive benefits to users and society. Therefore, more studies on the aspects of such systems and their impacts when adopted on a large scale are needed. Furthermore, greater incentive to use these systems should be globally debated.