Feasibility of rainwater harvesting from residential rooftops in Jordan

Rainwater harvesting is adopted to face water scarcity in arid regions. Many studies were developed in Jordan to estimate the potential of water harvesting for several uses. However, the precise estimation of water saving and cost benefits for the potential of rainwater harvesting from the roofs of residential urban areas is insufficient. Therefore, the objective of this study was to investigate the feasibility of rainwater harvesting from residential rooftops in all the eighty-nine Jordanian districts. The forecast number of buildings was calculated from 2016 to 2025 using building growth rate, where the number of houses and villas was adopted only. The long-term rainfall average from 1937 to 2017 was used. Two scenarios were used to assess the potential of rainwater harvesting; numerical (as scenario 1) and tabulated by plumbing code (as scenario 2) for the years from 2019 to 2025. Also, the growth rate of one cubic meter of water cost was calculated to find the money saving potential for the water companies in Jordan. The results indicated that the water harvesting potential was different between two scenarios in the districts which have annual rainfall more than 100 mm was efficient in scenario 1 compared to scenario 2 with the projected financial return which was increased from $5.4 million at 2019 to reach $33.4 million at 2025, while in the districts which have annual rainfall less than 100 mm was more efficient in scenario 2 compared to scenario 1 with the projected financial return which was increased from $2.4 million at 2019 to reach $14.6 million at 2025.


Introduction
Water shortage is currently a universal problem facing more than eighty countries of the world, where about two billion people lack access to clean water and this percentage is increasing by time (Alois 2007). By results of many research projects, several adaptation plans in different countries were suggested one of which is implementing rainfall water harvesting. Others are collection of rainfall from the rooftops of buildings and storing it in the artificial reservoirs for using it as a backup supply for daily domestic requirements and minimizing the pressure on the natural water resources.
The potential of water harvesting is different from a country to another depending on amount of precipitations and amount of water supply in each country. Sahin and Manioğlu (2019), Summerville and Sultana (2019), Baby et al. (2019) and Abu-Zreig et al. (2013) found that for countries with low rainfall, only 3.4-8.5% of the domestic annual water requirements were covered by water harvesting, for medium rainfall reached to 14.7-17.7% (Ghisi and Ferreira 2007) and 43% in countries with heavy rainfall (Hofman-Caris et al. 2019). Most of Jordan is desert land that has 75% of the total area of Jordan with annual rainfall reaches to be less than 120 mm. The highest annual rainfall around 300-500 mm is found in the northwestern part of Jordan. In Jordan, the rainwater harvesting is one of the additional water resources that is adopted in Jordan's water demand management policy (MWI 2015), and the Jordan uniform plumbing code (JNBC 2013). In Jordan, 87% from the total water consumption in the municipal sector went to the residential uses (MWI 2012).
Several studies were developed in Jordan for estimating the potential of water harvesting for several purposes (Saidan et al. 2015;Hadadin et al. 2014;Abu-Zreig et al. 2013;Awawdeh et al. 2012). The Abu-Zreig et al. (2013) and Saidan et al. (2015) depend on the rooftops areas of the apartments to build their studies, and the Saidan et al. the GIS to estimate the catchment areas of rooftops. According to the previous approaches, the number of apartments does not mean the number of buildings, where the buildings have many apartments, and they do not have enough space to build rainfall storage tanks and if storage tank was built, it cannot serve high number occupancies in those apartments. In addition to the studies that used the GIS to estimate the catchment areas of rooftops, the roads and open spaces were captured also as polygons that have areas more than the actual rooftops areas. However, the lack of precise estimation of water saving and cost benefits for the potential rainwater harvesting from the roofs of residential buildings in populated urban areas justifies this research and adopts the rooftops of the houses and villas only. The objective of this study was to investigate the feasibility of rainwater harvesting from residential rooftops in Jordan at governorates districts level.

Methodology
In this study, the estimation of potential water saving was built in the eighty-nine districts, and the estimation of potential money saving was built in the twelve governorates in whole Jordan. Jordan is located in the middle east at the crossroads of Asia, Africa and Europe as shown in Fig. 1.
Jordan is relatively small in area, semiarid and arid, almost landlocked country with an area of 89,342 km 2 and its population around ten million (DOS 2016). Jordan has twelve governorates (Amman is the capital city, Balqa, Zarqa, Irbid, Mafraq, Ajloun, Jerash, Madaba, Karak, Tafila, Maan-which is the largest province by area-and Aqaba). Amman has 42% of Jordan's total population. The country is characterized by diverse range of habitats, ecosystems and biota due to its varied landscapes and environments.

Forecasting the number of buildings
This study adopted houses and villas only, because they have enough space to build rainfall storage tank comparing with the apartments. The number of houses and villas with their areas was obtained from the department of statistic (DOS). The percentage of the houses and the villas from all buildings in all governorates was calculated and found 13%. These data were divided into the level of district per governorate, where there is a fluctuation in rainfall on the level of district and governorate. The areas of buildings in each district are important to calculate the volume of water that can be harvested. The growth rate of buildings was calculated based on census (2004) and census (2015), and then the forecasting number of buildings was calculated from 2016 to 2025.

Amount of long-term average of rainfall per district
Based on the data of the ministry of water and irrigation (MWI 2018a), there are 256 rainfall stations were distributed in the whole of Jordan. By using GIS ArcMap version 10.3, the clip method was used to match the districts with their rainfall stations. There are eight districts out of 89 districts that do not have rainfall stations. For those districts, the surrounding stations were used to interpolate the average rainfall volume as a function of global mean similar to spatial kriging analysis. Thus, the average annual rainfall (mm) in each of those eight districts was the average of the two rainfall depths of the Isohyetal lines, multiplied with their area to get on the volume between any two Isohyetal lines. All the volumes were summed and divided by the total area of each district calculated by using the following equation (Nielsen and Wendroth 2003): (1) where global mean is the mean average rainfall along the whole study area, average rainfall depth is the rainfall depth estimated from the isohyets generated from two weather stations in mm, and area is the area between two isohyets, and ∑ Areas of district is a total area of the whole district.

Investigation scenarios
In this study, two scenarios were used to estimate the potential of rainwater harvesting. In the first scenario, the rainfall volume for each district was calculated based on the following (Gould 1993) equation: where V is the annual potential of rainfall that can be collected by the roofs in cubic meter, I is the average annual rainfall in mm, A is the total roofs of the same area of houses and villas, C is the runoff coefficient that 0.85 was accepted for asphalt and concrete, and 1000 is a conversion factor. In the second scenario, the water harvesting table in the Jordan uniform plumbing code (JUPC) was used. The areas of rooftops that found in the table are different about the areas of rooftops that accredited by DOS. So, the storage tank size for areas that accredited by DOS was interpolated with the storage tank size that was obtained from the JUPC table and multiplied with the total number of buildings that have the same area. The potential of water harvesting by the two scenarios was calculated for years from 2019 to 2025.

Cost-benefit assessment
The water portion for each governorate at year 2017 with its cost was obtained from the Jordan water authority (WAJ) (2) V = SUM (C.I.A∕1000) (MWI 2018b). So the total volume of water for all governorates by WAJ reached to be 227.6 MCM. Besides, the cost of one cubic meter of water at the same year was $2.06 with growth rate reaching to 2% to get on estimation for cost of one cubic meter in years from 2019 to 2025. Based on the results of the two scenarios, the potential water saving for each governorate was calculated and multiplied with the estimation cost for each governorate. The potential of money saving for the water companies in Jordan was calculated for years from 2019 to 2025.

Number of forecasted buildings
The total number of buildings in Jordan is 1776135 (DOS 2016). Most of those buildings found in the Amman with percent reached to be 42%. The growth rates of buildings were calculated and vary between the governorates. And the highest growth rate for all types of buildings is found in Amman with 14%, but the percentage of the houses and villas was the lowest there and reached to be 5%; because most of the commercial and apartment buildings are found in the capital. The growth rate and percentages of houses and villas with average size of rooftop (m 2 ) in each governorate are shown in Table 1. In addition to the comparison between Amman and Irbid, the number of the houses and villas in Irbid in year 2016 was larger than Amman, but with the growth rate of buildings, the number of the houses and villas in Amman will be larger than Irbid in year 2025 as shown in Table 2 and forecasts number of houses and villas in each governorate from 2016 to 2025.

Distribution of rainfall in Jordan
Jordan has three topographic regions where the highlands are located on the northwestern part that have abundant rainfall with the highest annual rainfall reaching to 519 mm in Ajloun, and the Jordan rift valley is located on the west and it contains the Jordan River and the Dead Sea (the lowest point in the world) and the eastern and northeastern areas of the country are arid desert with the lowest annual rainfall reaching to 26 mm in Aqaba. So, Fig. 2 shows the average annual rainfall distribution in the governorates of Jordan.

Potential water harvesting volume from rooftops
The potential water harvesting volume has been depended on two main factors, namely number of houses and villas with their areas (m 2 ) and the average annual rainfall governorate. As comparison between two governorates, Amman has the highest growth rate for the number of houses and villas, while Ajloun has the highest number for annual average rainfall as explained in two previous sections. Moreover, the results show that the number of houses and villas has more effective potential of water harvesting than the average annual rainfall, and this is obvious in Amman, which the potential of water harvesting in 2019 showed as shown in Table 3. This result refer to that the large volume of rainfall found in the governorates have percentage of agricultural areas more than residential areas. In addition, the volume of harvested water in scenario 1 more than scenario 2, because the storage tank size that was calculated in scenario 1 had a maximum size, while the storage tank size that was determined by (JUPC) in scenario 2 had an optimum size as shown in Tables 3 and 4. The third result showed that scenario 1 had more volume of harvested water than scenario 2 in the districts which have annual rainfall more than 100 mm. However, scenario 2 estimated higher water harvesting potential at districts receiving average annual rainfall less than 100 mm. Accordingly, this result is clear in Aqaba governorate where all districts have annual rainfall less than 100 mm. The results of the two scenarios are less than the results of Abu-Zreig et al. (2013) where the apartments were included in its study.
Doing specific enumeration to collect data about urban water harvesting is important for the government to estimate actual water volume that collected and used in the  municipality sector. In particular, there are many buildings that have water harvesting tanks, but there is no information about number of tanks, their sizes and times of filling tanks in the one winter season.

Cost feasibility analysis
The total volume of water for municipality sector reached to 227.6 MCM at the year 2017 in all governorates of Jordan (MWI 2018b). Its operational cost per one cubic meter of the same year was calculated to be $2.06 with growth rate reaching to 2% (MWI 2018b). The portion of water has been differentiated between the governorates, and their growth rate has been differentiated also. Hence, the largest water portion found in Amman with growth rate reached to 2.9%, because Amman has 42% from the total population in Jordan (DOS 2016) and most of commercial activities found there. However, Mafraq governorate has the highest growth rate in water portion that reached to 5.5%, and this referred to that most of Syrian refugees have settled in Mafraq governorate. To get on the percentage of water saving in each governorate, the estimated volume of water harvesting was divided by the value of estimated water portion in each governorate from 2019 to 2025. Thus, in scenario 1, the percentage of water saving was 1% in all governorates at 2019, and increased to reach 4.6% at 2025, with a financial return of $5.4 million at 2019, and a projected return of $33.4 million at 2025. On the other hand, in scenario 2, the percentage of water saving was 0.4% at 2019, and increased to be 2% at 2025, with a financial return of $2.4 million at 2019, and projected to be $14.6 million at 2025. Supporting the government to the water harvesting approach from residential rooftops by mandatory process will reflex many benefits on it, as providing extra source for drinking purposes without having any capital or operational loads compared to dam building. On the other hand, the government will save its budgets to other its services.

Conclusion
The rainfall water harvesting is one of the several adaptation plans that can minimize the domestic annual water requirements. In Jordan, the rainwater harvesting is one of the additional water sources adopted in Jordan's water demand management policy, and the Jordan uniform plumbing code. Several studies were developed in Jordan for estimating the potential of water harvesting for several purposes. The rooftop areas of the apartments were used in those studies, but the rooftops areas of the apartments were excluded in this study for several considerations; the number of apartments does not mean the number of buildings because the buildings that have many apartments, and the buildings that have many apartments which do not have enough space to build rainfall storage tanks; then if storage tank was built, it cannot serve high number occupancy in those apartments. The number of houses and villas had more effective potential of water harvesting than the average annual rainfall that the regions that have high annual rainfall occupy agricultural areas more than residential areas. Scenario 1 had more volume of harvested water than scenario 2 in the districts which have annual rainfall more than (100 mm), while scenario 2 estimated higher water harvesting potential at districts receiving average annual rainfall less than (100 mm).
Jordan is suffering recently from huge population growth, which causes additional stress on water demand. This study concluded that adaptation of water harvesting from residential rooftops has positive effects on the communities which encounter real pressure on water resources, and increase in percentages of water saving. These savings are reflected positively on the financial status of water supplier companies.