**Utilization of MAR for Wastewater Reuse in Arid Regions**

## **Managed Aquifer Recharge (MAR) Economics for Wastewater Reuse in Low Population Wadi Communities, Kingdom of Saudi Arabia**

### **Thomas M. Missimer, Robert G. Maliva, Noreddine Ghaffour, TorOve Leiknes and Gary L. Amy**

**Abstract:** Depletion of water supplies for potable and irrigation use is a major problem in the rural wadi valleys of Saudi Arabia and other areas of the Middle East and North Africa. An economic analysis of supplying these villages with either desalinated seawater or treated wastewater conveyed via a managed aquifer recharge (MAR) system was conducted. In many cases, there are no local sources of water supply of any quality in the wadi valleys. The cost per cubic meter for supplying desalinated water is \$2–5/m3 plus conveyance cost, and treated wastewater via an MAR system is \$0–0.50/m3 plus conveyance cost. The wastewater reuse, indirect for potable use and direct use for irrigation, can have a zero treatment cost because it is discharged to waste in many locations. In fact, the economic loss caused by the wastewater discharge to the marine environment can be greater than the overall amortized cost to construct an MAR system, including conveyance pipelines and the operational costs of reuse in the rural environment. The MAR and associated reuse system can solve the rural water supply problem in the wadi valleys and reduce the economic losses caused by marine pollution, particularly coral reef destruction.

Reprinted from *Water*. Cite as: Missimer, T.M.; Maliva, R.G.; Ghaffour, N.; Leiknes, T.; Amy, G.L. Managed Aquifer Recharge (MAR) Economics for Wastewater Reuse in Low Population Wadi Communities, Kingdom of Saudi Arabia. *Water* **2014**, *6*, 2322-2338.

### **1. Introduction**

Hundreds of small villages and farms exist in wadi (ephemeral streams) valleys throughout the Kingdom of Saudi Arabia (KSA). For centuries, these agrarian communities relied upon shallow groundwater resources to supply potable and irrigation water demands [1]. Anthropogenic impacts, including over-pumping and contamination, have combined to deplete or render unusable the groundwater in shallow alluvial aquifers underlying the wadis [1–3]. Because of the low population density, generally small contribution of crop production to the national economy, and the arid nature of the climate, there are quite limited options available to supply the necessary water to maintain these populations. Nevertheless, rural communities are considered an important part of the cultural heritage of the Kingdom, and finding a solution to their water challenges is a priority. While the KSA is a wealthy country and has technically feasible options to replace the depleted water supplies for these rural communities, such options are even more limited in other, less prosperous countries in the Middle East–North Africa (MENA) region [4].

Four potential methods of providing a comprehensive and reliable water-supply solution are currently being assessed in the KSA. These options include: (1) the construction of seawater reverse osmosis (SWRO) desalination plants and conveyance of this water from the coastline to the end users via pipelines; (2) desalination of brackish groundwater by reverse osmosis (RO), where brackish-water aquifers are available; (3) construction of wadi dams to trap seasonal stormwater discharges and conveyance of the water to the users via pipelines (treated or untreated); and (4) conveyance of treated domestic wastewater to the users via pipeline with subsequent storage and treatment in the underlying aquifer system using aquifer recharge and recovery (ARR) systems. ARR is a form of managed aquifer recharge (MAR) that takes advantage of natural contaminant attenuation processes to improve water quality. ARR systems have an element of treatment along with the conventional storage functions of aquifer storage and recovery (ASR) systems. Use of cistern water capture and other water harvesting methods have been considered, but are insufficiently robust to meet water supply requirements, especially under future global climate change scenarios.

A more detailed analysis of the region shows that the western part of KSA bounding the Red Sea does not contain significant brackish water aquifers that could produce sustainable quantities of water to become a reliable source of water supply. Also, the construction of wadi dams and development of water supplies is based on storm events that have a very uneven frequency and with global climate change, could become more intense and less frequent [5]. Therefore, only two of the four options (seawater desalination and wastewater reuse with ARR) are technically viable as far as potentially providing sustainable water supplies. The feasibility of the two technically viable options depends upon both costs and social acceptance. It should be noted that if all of the supply options were to be found unfeasible, the population living in the small communities and farms would be forced to leave their lands and move into densely populated urban areas, therefore exacerbating existing water supply and social issues in the region.

It is the purpose of this paper to assess the relative economics of two potential sustainable water supply options for these small communities and farms; use of seawater desalination *versus* use of treated domestic wastewater with ARR storage and treatment for both irrigation and indirect use. This assessment is conducted using unit costs for many of the variables, because there is considerable variation in the transport distances from the sea or sources of treated domestic wastewater to water users and corresponding spatial variation in water demands.

### **2. Background and Methods**

### *2.1. Description of the Rural Wadi Communities and Farms*

Rural communities and small farms are quite common in the wadi valleys of the KSA as well as in many other areas of the MENA region (Figure 1). For centuries, these small communities and farms have been dependent on shallow groundwater for supplies. In the past, there was sufficient recharge to the underlying alluvial aquifer system to maintain the sustainability of the water supply. Aquifer water levels fluctuated seasonally between 1 and 3 m below surface in the early part of the 1900s, depending upon rainfall accumulation and the occurrences of periodic drought conditions [6]. However, in the modern era, population growth and expansion of agricultural activities has caused depletion of the groundwater resources of wadi alluvial aquifers with water levels commonly dropping 20 to 30 m below surface in many areas and causing complete aquifer dewatering in some wadi systems [2] (Figure 2).

**Figure 1.** Large-diameter abandoned well at the center and a dead date palm plantation in western Wadi Qidayd, Saudi Arabia. Aquifer depletion has caused the large-scale failure of small farms and abandonment of some villages as shown by the dead date palms.

The wadi channels in which the villages and farms lie are moderate to low sloped features that contain alluvial sediments and are periodically flooded to variable degrees. A large number of large diameter wells are used to supply groundwater where it is still available. Many of the wells have been abandoned because of resource depletion or contamination with saline water and/or nitrates. Entire conventional treatment facilities have been abandoned (Figure 3). In areas where groundwater depletion has occurred, the only method of obtaining potable and irrigation water is to purchase it from suppliers and have it hauled by tank truck to fill onsite storage tanks. At Wadi Qidayd, the cost for treated water is \$1.60–1.87/m3 and for untreated water \$0.27–0.40/m3 . The source of the truck-transported water is often local wells, the use of which contributes to further aquifer depletion. Use of the wadi aquifers for water supply at current rates is not sustainable. The lack of effective rainfall and associated recharge in the lower part of the Wadi Qidayd basin for the past several years has caused the shutdown of several local water suppliers due to dry wells.

**Figure 2.** Two-meter diameter well showing the water level at about 20 m below surface in Wadi Qidayd, western Saudi Arabia.

**Figure 3.** Abandoned municipal well that served a village water treatment facility.

### *2.2. Estimation of Wastewater Treatment Costs*

Wastewater treatment costs have been estimated from the literature for primarily conventional treatment technologies that will provide a relatively high degree of purity to allow indirect potable use. The technologies evaluated are discharge lagoons/oxidation ponds with natural infiltration (LAG), conventional trickling filters, conventional activated sludge (CAS) with nutrient removal (*i.e.*, secondary treatment) as nitrates can adversely impact drinking water quality, assessing CAS as conventional aeration tanks or oxidation ditches (CAS-OxD), advanced treatment using an integrated membrane bioreactor system (MBR), and conventional activated sludge followed by tertiary filtration (CAS-TF). The final polishing of the treated domestic wastewater is assumed to be aquifer treatment, whereby the treated wastewater is placed into the alluvial aquifer using wells and the extraction for potable use is from wells located down-gradient. It is known that some refractory trace organic compounds will not be removed from the wastewater and further treatment may be required at extraction points closer to larger population centers. Although there are some public concerns regarding possible impacts of these compounds on human health, there are currently no drinking water standards established for them [7]. The available evidence suggests that exposure to trace concentrations of pharmaceuticals (at concentrations found in treated wastewater and water) is unlikely to cause health effects [8,9].

Capital and operating costs for the various wastewater treatment technologies are comparatively developed, which are then systematically compared to seawater desalination costs.

### *2.3. Estimation Methods for Desalination Costs*

Compilation and analysis of desalination costs have been published recently [10] based on past and recently collected cost data. The costs estimated for seawater desalination are focused on SWRO because it is the least costly of the large-scale desalination methods currently being used in the KSA and can be designed and constructed at a variety of capacities. Thermal desalination systems are quite difficult to design, construct, and operate at small capacities, especially in consideration of the rather small water use requirements in some wadi systems. Low-capacity, renewable-energy driven systems, such as solar stills cannot be used because of the lack of any local supply of water, saline or fresh.

SWRO costs are developed for a range of capacities. There is an economy of scale that generally causes larger-capacity SWRO systems to operate at lower costs compared to small capacities. However, many of the wadi communities are widely separated from large population centers and would require the development of comparatively low capacity SWRO systems. It would be less expensive to construct and operate a small scale SWRO plant to serve a number of small communities, than to pipe treated water a great distance from the very large-capacity desalination plants located near major population centers [11].

### *2.4. Estimation of Conveyance Costs*

The cost of conveyance is based on design and construction of the pipelines using a standard diameter high-density polyethylene pipe (HDPE). The pipe would be buried primarily within the wadi channels at the proper location and depth to avoid damage during flow events. The burial depth is estimated to average 1 m below grade. The two pipe diameters considered are OD 1100 mm and OD 630 mm. The strength grade of the HDPE pipe is 16 BAR PE 100. These large-diameter pipeline sizes are used with the assumption that there will be off-takers of water along the trunk lines and reduced diameter pipelines would serve the most distal farms and villages.

It is assumed that a pumping station will be required for each 40 km of pipeline. The average elevation change is estimated to be about 70 m and head losses due to pipe friction limit the overall head loss to no greater than 120 m. Electrical requirements are estimated based on the friction head loss for the two pipe diameters plus the elevation head required.

Costs of the pipeline design and construction and the pumping stations are estimated based on conversations held with contractors in KSA (Jeddah city) and consulting engineers in the United States that have Middle East region experience. Engineering design and construction observation costs for the pipeline and pumping stations are estimated as about 15% of the total construction cost based on KSA practices.

### *2.5. Estimation Methods for Treated Water by ARR/MAR Systems*

ARR/MAR costs are estimated based on the construction of both large (2 m) and smaller (0.5 m) wells using local drilling contractors. The injection of the water is designed to use the line pressure from the pipeline and is essentially either a gravity feed system or a low pressure injection system. The down-gradient recovery well pump costs and electrical use are based on an average lift of 25 m using electric turbine pumps. A few different pumping rates were used based on the well types and desired capacities.

It should be noted that there are hundreds (or thousands) of abandoned or seldom-used, large-diameter wells located throughout the wadis of western KSA. Wherever possible, existing wells would be used. In some cases, the wells would have to be rehabilitated or repaired. Also, these abandoned wells were commonly located adjacent to the villages and farms where the water would be used.

### *2.6. Water Treatment*

Water recovered from the ARR/MAR treatment systems for potable use is of generally high quality. It is assumed that the only post-recovery treatment would be disinfection using chlorine.

Water from the system used for agricultural purposes would not be treated. Some farms may choose to install storage tanks to allow higher irrigation rates during the nighttime. This would ameliorate any supply and demand imbalances and would keep the pipeline costs lower (smaller diameter pipes).

Most houses and public buildings in the wadi communities are equipped with storage tanks, which would allow a lower amount of system common storage to be used for ARR/MAR recovered and treated water. Therefore, only relatively small capacity storage tanks are used for the village supply systems. Distribution piping system costs are not included in these estimates because of the large differences in demand for population centers ranging from a few families to perhaps 3000 people. Currently, many villages are not equipped with distribution systems, particularly the smaller population centers where water is trucked to the users. This practice may remain after system installation, but the quality of water would be truly potable.

### **3. Results and Discussion**

### *3.1. Seawater Desalination Treatment Costs*

The KSA is the world's largest user of desalinated seawater, accounting for about 18% of the total global capacity [12]. The most used desalination technology in the KSA is multi-stage flash (MSF) distillation which is very energy intensive accounting for up to 70% of the desalination costs, and these plants are expensive to maintain [13]. In recent years, the KSA has begun to use large scale SWRO, but most commonly in hybrid facilities containing electric generation, MSF, and SWRO [14]. In addition, many standalone SWRO and brackish water reverse osmosis (BWRO) plants, with capacities ranging between 50 m3 /d and 17,500 m3 /d, have been installed by the private sector [12]. On the other hand, multi-effect distillation (MED) is being used to replace the MSF process in other sites in the KSA, mainly with enhanced performance using thermal vapor compression (TVC) in a hybrid MED-TVC configuration (Table 1). It is difficult to ascertain the true cost of seawater desalination in the KSA because all utilities are subsidized by the government, and, commonly, freshwater is provided at vey low cost to the consumer [10]. Also, there are virtually no legal restrictions on water use and legal guidelines on reclaimed water use in the KSA.


**Table 1.** Water cost of different thermal desalination projects in Kingdom of Saudi Arabia (KSA), including subsidies (Global Water Intelligence/Water Desalination Report (GWI/WDR), 2009–2014).

Some water cost estimates can be made for the various seawater desalination facilities based upon energy consumption for the different technologies being used. These costs are truly scale dependent based on the capacity of the treatment facilities (e.g., large capacity facilities generally produce water at a generally lower unit cost compared to small facilities) [15] (Figure 4). Within the KSA, thermal desalination costs, including subsidies, range between \$0.83 and \$1.5/m3 depending on the technology used, age of the facility, and the plant capacity. Hybrid water desalination facilities costs depend strongly on the hybrid configuration used along with the capacity. The total water cost produced by a MSF-SWRO hybrid system is 5%–10% less compared to MSF standalone plants [16]. On the other hand, SWRO facilities (standalone) produce freshwater at a cost ranging from \$0.5 to \$1.5/m3 , depending on several parameters, such as feed water salinity, requested product quality (includes post-treatment), and electrical energy, land and labors costs. Water cost of some thermal desalination plants in the KSA is presented in Table 1.

Another important factor affecting desalination water cost for rural communities is the distance of the end users from the coast or from large population centers that are served by high-capacity desalination facilities. In wadi communities located within a 50–100 km radius of a major desalination facility, the actual freshwater production cost would be the same as for the city residents, but the additional cost would be for transmission. In remote communities, the design, construction, and operation of standalone SWRO or BWRO facilities with a moderate to low capacity would likely be required and their costs might be very competitive with long water transfer from coastal desalination plants [11]. However, the feed water sources to supply these local plants within the wadi areas are not sustainable and therefore, would not be a reliable water supply. Overall desalination costs could be quite high from these facilities, located near the user (if possible) or at the shoreline, with a probable range from \$1.25–5/m3 .

Installation of small-scale innovative desalination processes at the shoreline, powered with solar energy, may produce water at a lower cost and would require less maintenance and chemical use [17]. Low-energy processes, such as membrane distillation, could be the best solution for communities living in rural areas. Also, each village would be responsible for providing water storage and a distribution system where the population density permits doing so. Existing systems depend upon trucking water from a well (if available) and conveying the water to storage tanks located at each home or cluster of homes. This system also has a cost.

### *3.2. Domestic Wastewater Treatment Costs*

Wastewater treatment plants naturally vary in capacity as a function of the community being serviced as well as the underlying infrastructure approach used (e.g., individual households *versus* decentralized/satellite designs and large centralized technologies). There is also an economy of scale that generally causes larger-capacity treatment plants to operate at lower costs compared to small capacity plants. For smaller communities (e.g., villages and farms) the operating costs for a treatment system will therefore fall in the higher range of cost estimates. In addition, the type of treatment technology selected will impact both capital and operating costs. A complicating factor in such an assessment is the difference in both regional and local parameters, such as land cost (*i.e.*, impact of plant footprint), and the unit cost for energy.

In rural communities there is a tendency to choose low-technology solutions which typically require a lot of space (on the assumption that land costs are minimal) and which will require a minimum of skilled-labor maintenance. However, for a water reclamation and reuse strategy, there is a move to more advanced treatment systems to ensure reliable, safe, and high water quality as defined by the end use. In this cost estimate, representative treatment technologies, which can be defined as low to high technology solutions, have been included in the comparison. These include ponds and lagoons, trickling filters, variations of conventional activated sludge (CAS) (e.g., secondary treatment/oxidation ditches), tertiary treatment of secondary effluent (e.g., membrane filtration/advanced oxidation), and membrane bioreactor technology (MBR) as an alternative advanced tertiary treatment system. A direct comparison of capital costs for these technologies is

**Figure 4.** Investment cost/m3 for seawater reverse osmosis (SWRO) and brackish water reverse osmosis (BWRO) systems. Note the reduction as the capacity increases.

not straight forward, although studies in the literature can be found showing that high technology options are cost competitive to low technology alternatives [18–21].

In most studies assessing operating costs for various treatment technologies, energy is highlighted as a key parameter for defining the operating costs, typically in the range of 40%–60% of total costs [18,19]. The specific energy consumption for wastewater treatment is reported in the range of 0.4–1.0 kWh/m3 of treated water [19,21,22]. Breaking this down to commonly used technologies in terms of sophistication of the treatment plant gives the ranges or 0.08–0.28 kWh/m3 for lagoons, 0.19–0.41 kWh/m3 for trickling filter plants, 0.33–0.61 kWh/m3 for conventional activated sludge, and 0.48–1.03 kWh/m3 for oxidation ditches and tertiary treatment. Membrane bioreactors are perceived as being energy intensive, however recent case studies comparing average energy requirements for tertiary treatment based on conventional activated sludge compared to MBR have shown that a relatively large MBR plant consumes 0.9 kWh/m3 compared to a range of 0.5–1.8 kWh/m3 for the tertiary conventional activated sludge options [18,20,23–28]. On the assumption that energy costs on average are 50% of the total operating costs, energy can be estimated at 0.01–0.210/kWh/m3 , an estimate for a lower and upper range of operating costs for various wastewater treatment technologies can be compared. The results are shown in Figure 5.

For the low technology options (e.g., lagoons, trickling filters) the treatment costs will range between \$0.05–0.20/m3 depending on the criteria chosen, however, it is debatable whether the water quality achieved is well suited for reuse. Conventional activated sludge is a more appropriate technology with respect to treated water quality and design options resulting in costs ranging between \$0.10–0.50/m3 . It is interesting to note that conventional activated sludge designed as an oxidation ditch can be relatively higher in O&M costs, as exemplified by the example of case studies in the KSA [26,27]. Treatment of wastewater to a high quality suitable for reuse can be achieved by conventional activated sludge followed by advanced tertiary treatment, estimated at a cost ranging from \$0.10–0.70/m3 based on a series of assumptions. For this level of treatment, MBR technology is shown to be more efficient with estimated costs of less than \$0.40/m3 [23,24]. In recent research conducted on advanced treatment of wastewater by MBR technology, it was shown to be very competitive an as alternative to desalination options [23,24]. With respect to rural populations having to rely on desalination as a reliable water source, it is apparent for the simple estimations shown above that advanced treatment of wastewater for both non-potable and indirectpotable reuse is a viable and sustainable option.

### *3.3. Conveyance Cost*

The cost to convey water from the treatment plant to the end user is quite significant, especially in the isolated rural environment. Conveyance cost can be broken down into capital and operating costs. Capital costs include the pipeline engineering and construction, the cost of the pumping stations, and some undefined costs of conveyance, such as construction of pipelines crossing roads and municipal infrastructure at large facilities. The operating costs include the electrical costs and are mostly for electricity to run the pumps. Any additional costs associated with the operation of ASR wells are considered to be minor within the overall assessment.

Wadi valley pipeline engineering design and construction are relatively simple, but require consideration of periodic flooding within the wadi channels and potential erosion of the main channel area. The soils are predominantly sands and gravels and are easy to excavate. The preferred pipeline material can be HDPE. The strength of the pipe should be 16 BAR PE 100 to prevent any damage due to movement during earthquakes and by trucks or other farm equipment. The cost for materials and installation of HDPE pipe in the wadi valleys of Saudi Arabia is given in Table 2. The hydraulic gradient from the shoreline to the heads of the wadi valleys is not very steep and the overall elevation rise is likely not more than 70 m over a distance of about 40 km. Pumping station costs were obtained for a variety of facilities ranging in capacity from 5000 to 40,000 m3 /d. The cost for such facilities in western Saudi Arabia is roughly \$500,000/5000 m3 /day of capacity (Table 2). A preliminary assessment shows that a single pumping station can be used to transmit this range of capacities between 40 and 60 km, assuming that the overall head loss is no greater than 120 m.

Only two relatively large diameter pipe sizes are listed in the table. Since wadi systems contain a series of local farms and village occurring along a linear geometry or with a series of branches, these pipeline diameters would be used as trunk lines and could be reduced in diameter from proximal to distal users. For cost estimation purposes, the larger diameters should be used because the cost of construction will likely be nearly the same for the next lower set of pipe diameters.

The electrical use for operation of the pumping stations to convey the water from the source to the use area is dependent on the required capacity (Table 2). The kilowatt-hours of electricity per day are also given in Table 2. The subsidies used in Saudi Arabia make the determination of real electric costs quite difficult to estimate, but the real cost likely ranges from \$0.05–0.15/kw-h. An estimated cost range to convey the water 40 km is \$0.45–1.50/m3 .

Since the key aspect of this research is the comparison of costs between use of desalinated water and reuse of highly treated domestic wastewater indirectly via an MAR system for potable supply and directly for irrigation use, the cost of conveyance of the water will be the same for either option. It can be calculated from the data given in the tables. If the water is conveyed from great distance, the cost of desalinated water delivery will be roughly doubled. The multiplier will be even greater for conveyance of highly treated wastewater because of its lower treatment cost.


**Table 2.** Estimated cost for construction of high-density polyethylene pipe (HDPE) pipelines in wadi systems.

Notes: 1 The assumed total dynamic head is estimated to be 122 m; 2 Real cost of electric power in the KSA is estimated to range from \$0.05–0.15/kw-h.

### *3.4. Cultural and Religious Issues Involving Wastewater Reuse*

A major challenge for indirect potable reuse projects is obtaining public acceptance. Public perception issues associated with reuse of reclaimed water were reviewed by Maliva and Missimer [2]. In general, public acceptance of the reuse of reclaimed water increases with increasing "distance" or isolation from the treated wastewater. There is generally a high level of acceptance for projects with no human exposure and a much lesser support for projects with direct human contact.

The passage of water through a natural environment, such as an aquifer, also reduces its "taint" of being wastewater. Public acceptance also depends upon the recognition by the effected population of the severity of the water shortage and confidence in the agency or organization that will implement the project. Reuse of reclaimed water and even indirect potable reuse are not contrary to Islamic Law. The Council of Leading Islamic Scholars in Saudi Arabia issued a fatwa in 1978, stating that reclaimed water can be used for ablution and drinking if it is sufficiently and appropriately treated to ensure good health, but recommended avoiding use of treated wastewater for drinking purposes to avoid health problems and also in consideration of the negative public sentiment about this water. If drinking is to be avoided, it is to be merely for reasons of public health and safety, not due to any ramifications of Islamic Law [29].

Wastewater is already being recharged to some wadi alluvial aquifers downstream of wastewater treatment plants and through on-site disposal systems, so the introduction of the more controlled upgraded wastewater treatment/ARR could, in some instances would, result in improved water quality. Nevertheless, obtaining local public support will be a critical feasibility issue, which will need to start with a public education campaign. A lack of knowledge on issues such as wastewater quality, health risks, and for farmers, impacts on soils and crops often leads to a negative perception of wastewater reuse.

### *3.5. Cost for ARR (MAR) Construction and Operation*

In most cases, the number of abandoned, large-diameter wells would be sufficient to meet the need for existing small villages and farms, at least for the upgradient injection well or wells for each site. At locations where an additional well is required to recover the injected water, the construction cost for a well ranges from \$5000 to 20,000 depending on the depth and diameter of the well. The recovery pump would be a diesel-powered vertical turbine pump with a head lift maximum of 50 m. Typical pumps used in the wadi systems cost about \$7500. The cost of fuel to power the pumps is subsidized and is about \$0.25/L. Therefore, the operational cost of a small ARR system for a village is <\$0.05/m3 . The treatment cost and conveyance of the source water is greater than this cost.

### *3.6. Indirect Reuse and Irrigation Use Using MAR Treatment of Domestic Wastewater for Wadi Communities in the KSA: Special Circumstances*

The economic analyses developed in this research suggest that the use of treated domestic wastewater combined with ARR polishing for indirect potable use is the most economical solution to meet the rural water supply requirements, but it is still costly. However, there are extenuating circumstances that greatly affect the economics of water reuse which include the current practice of disposal of the treated or untreated wastewater and its adverse environmental effects on the marine environment and some inland aquifer water quality.

Only about 10% of the wastewater generated in the KSA is reused in a beneficial manner. Partial treatment and discharge to tidal water or into channels transmitting into the desert with no users are not economically beneficial. Therefore, a real cost comparison between use of desalinated water and wastewater should consider that there is zero cost for treatment of the wastewater if it is being discharged to waste. In fact, environmental damage caused by inappropriate wastewater disposal practices produces a negative economic impact, which must be considered in this analysis.

Wastewater discharges to tide adversely affect the fringing reef of the Red Sea as occurs in all coral reef ecosystems [30–32], which in turn, adversely affects fisheries and the potential recreational aspects of the reef ecosystem. Coral reef ecosystems provide a diverse variety of goods and services to humanity [33,34]. Goods and services of all natural systems of the Earth affect the human economy and well-being [35]. Anthropogenic impacts on coral reefs have a direct economic

impact on the recreational value of reefs that can be measured [36]. Economic assessments by Cesar [37] and Berg *et al.* [38] found that losses to coral reef tourism caused by the destruction of 1 km2 of reef ranged between \$27,900 and \$100,800 USD and \$5500 and \$368,000 USD, respectively. A loss of \$40 million USD over a 10-year period was estimated by Hodgson and Dixon [39] for tourism and fisheries declines in a coastal area of the Philippines. While the Red Sea of KSA does not have a well-developed ecotourism industry, it is greatly dependent on the fisheries, which may generate an event larger overall economic impact.

There is a negative cost impact on the disposal of each 1 m3 of wastewater discharged to tidal water in the vicinity of a coral reef system. This cost depends on the concentration of nutrients within the wastewater, the degree of treatment for removal of solids and organic carbon, the proximity of the discharge to the reef, and the nearshore current patterns. A crude estimate of this cost range is \$0.05–0.20 USD/m3 for the economic losses associated with marine pollution. The range of loss associated with discharge to wadi aquifers and contamination of groundwater cannot really be estimated for areas where there is no significant water use.

### *3.7. Long-Term Sustainability of Seawater Desalination to Meet Rural Water Demands: Subsidies*

In any economic analysis, the issue of sustainability must be raised within the context of the water supply options being assessed. Based on the economic return of the relatively small population and the farms within the wadi valleys, the cost of supplying desalinated seawater to these areas would have to be subsidized by the government to bring economic viability to the residents and farmers. This issue raises questions concerning the long-term viability of a fully subsidized water supply within the context of the Saudi Arabian economy. However, there may be some mitigating economic issues with regard to food security which cannot be evaluated within the context of this research.

Electricity, fuel, and utilities are all nearly fully subsidized in KSA. The root of economic prosperity in the KSA is the income received from the international sale of petroleum [40]. In 2009, 25% of the petroleum produced in the KSA was consumed domestically and with population growth, this percentage will likely continue to rise [41]. This means that as domestic petroleum consumption rises, the petroleum available for export sale declines, and overall revenue income will decline with time. Also, the rate of domestic energy consumption in the KSA is greater than the United States. Declining revenue raises the question whether significant water use that provides little or no economic return can be maintained.

All other subsidies, including water supply and wastewater treatment are also subsidized to a nearly full degree. However, water and wastewater tariffs are being assessed to a limited degree in an attempt to recover some costs of providing utility service to the public and industry. There has been considerable push-back by the general population and industry that have grown comfortable with free utility services. Ramady [40] suggests that continuation of subsidies is a great challenge that is part of greater economic reform, which will be required in the future. Krane [41] has suggested that most economists believe that continued maintenance of utility subsidies threatens the stability of the Saudi Arabian economy. Therefore, the long-term economic sustainability of providing desalinated water to small villages and farms for drinking and irrigation water is debatable and questionable. This suggests that choosing the low cost water supply alternative, despite religious and cultural questions, may be the only viable long-term water supply option.

### **4. Conclusions**

There are limited options to supply water to the rural villages and farms located in western Saudi Arabia as well as other such communities located in similar global arid lands areas. A comparison of actual treatment costs between providing desalinated seawater for potable and irrigation uses to use of highly treated domestic wastewater with MAR polishing show a difference of nearly 300%. The overall cost of SWRO treatment with conveyance of the water over a distance of 40 km ranges from \$1.70–6.50/m3 . Treatment cost of domestic wastewater ranges from \$0.10–0.80/m3 . Conveyance cost for a distance of 40 km ranges from \$0.45–1.50/m3 . The use of local ARR systems using existing wells and a new well with a new pump is about \$0.05/m3 . Therefore, the water reuse system including treatment, conveyance and the ARR final treatment and operation ranges from \$0.6–2.35/m3 . If it is assumed that the treatment cost of the wastewater is zero, because it is currently not used or is discharged to waste, then the cost range declines to \$0.5–1.55/m3 .

The costs developed herein are rather specific to the western Saudi Arabia region, but can be estimated for any region based on the cost per kw-h for power consumption and correction for local electric rates. Construction costs vary greatly worldwide, but when these costs are amortized over a period of 20 years or greater, the impact on the cost per cubic meter to the consumer is minimal. This is particularly evident in regions where long conveyance of any source water is required.

Use of MAR for storage and polishing treatment of highly treated domestic wastewater is a significant method to minimize cost to supply safe drinking and irrigation water to rural areas in arid lands. Such systems need to be explored for use in areas where wastewater is being discharged with no economic benefit and alternative sources of water are extremely expensive.

### **Acknowledgments**

Funding for this research was provided by the Water Desalination and Reuse Center at the King Abdullah University of Science and Technology and from discretionary faculty funding from the same university. We thank Mohammed Saud, Vice President, Moya Bushnak Water and Environmental Services Company, Jeddah, Saudi Arabia for providing construction cost estimates. We also thank Thomas Burke, Chief Engineer of the Southwest Florida Water Management District for information on pumping station capital and operating costs.

### **Authors Contributions**

Thomas Missimer was the lead author and contributed the text including the introduction, the text on MAR/ARR, conveyance cost, and the special circumstances that impact wastewater reuse costs. Robert Maliva contributed the text on public acceptance of treated wastewater reuse and some of the MAR/ARR text. Noreddine Ghaffour contributed the economics of seawater desalination. TorOve Leiknes contributed the economics of wastewater treatment as applied to arid lands and reuse. Gary Amy provided some of the wastewater reuse text and edited the overall paper based on his wastewater reuse experience.

### **Conflicts of Interest**

The authors declare no conflict of interest.

### **References**


## **Impact Assessment and Multicriteria Decision Analysis of Alternative Managed Aquifer Recharge Strategies Based on Treated Wastewater in Northern Gaza**

### **Mohammad Azizur Rahman, Bernd Rusteberg, Mohammad Salah Uddin, Muath Abu Saada, Ayman Rabi and Martin Sauter**

**Abstract:** For better planning of a managed aquifer recharge (MAR) project, the most promising strategies should analyze the environmental impact, socio-economic efficiency, and their contribution to the existing or future water resource conditions in the region. The challenge of such studies is to combine and quantify a wide range of criteria from the environment and society. This necessity leads to an integrated concept and analysis. This paper outlines an integrated approach considering environmental, health, social and economic aspects to support in the decision-making process to implement a managed aquifer recharge project as a potential response to water resource problems. In order to demonstrate the approach in detail, this paper analysed several water resources management strategies based on MAR implementation, by using treated wastewater in the Northern Gaza Strip and the potential impacts of the strategies on groundwater resources, agriculture, environment, health, economy and society. Based on the Palestinian water policy (Year 2005–2025) on wastewater reuse, three MAR strategies were developed in close cooperation with the local decision makers. The strategies were compared with a base line strategy referred to as the so-called "Do Nothing Approach". The results of the study show that MAR project implementation with treated wastewater at a maximum rate, considered together with sustainable development of groundwater, is the best and most robust strategy amongst those analyzed. The analysis shows the defined MAR strategies contribute to water resources development and environmental protection or improvement including an existing eutrophic lake. The integrated approach used in this paper may be applicable not only to MAR project implementation but also to other water resources and environmental development projects.

Reprinted from *Water*. Cite as: Rahman, M.A.; Rusteberg, B.; Uddin, M.S.; Saada, M.A.; Rabi, A.; Sauter, M. Impact Assessment and Multicriteria Decision Analysis of Alternative Managed Aquifer Recharge Strategies Based on Treated Wastewater in Northern Gaza. *Water* **2014**, *6*, 3807-3827.

### **1. Introduction**

Nowadays, managed aquifer recharge (MAR) is considered as an integral part of integrated water resources management (IWRM). Like the IWRM concept, the interaction of MAR with other sectors of the water resources system, society, and natural processes is inherently strong [1]. Several researchers e.g., [2,3] mentioned that like other IWRM projects, the most promising MAR strategy should study the environmental impact, socio-economic efficiency, and their contribution to the existing or future water resources problem in the region [3]. Proper investigation and planning of MAR projects is important for successful application and can lead to significant risk reduction (e.g., environmental, health) and overall project cost reduction by potentially reducing uncertainties during project implementation. Again, proper planning requires impartiality and transparency in the evaluation of MAR options, considering explicit assessment of feasibility and cost-effectiveness [4]. Up until now, very few research studies have performed an extensive integrated study that consider the potential impacts on the environment, health, economy and society due to MAR project implementation and which select the best project option after intensive impact assessment [5].

The Gaza Strip, located on the eastern coast of the Mediterranean Sea, is a region facing severe water resources problems [6]. Due to the hot and dry climate, little surface water is available. Water supply relies mostly on groundwater resources located in the Northern Coastal Aquifer of Gaza [7]. The Beit Lahia Wastewater Treatment Plant (BLWWTP), located at Northern Gaza Strip, has been dysfunctional for some time now and is creating severe environmental, socio-economic and agricultural impacts for the public health and the environment [8,9]. A detailed description of the water resources problem at the North Gaza strip is given in Section 4.1. A three-phase 20-year project involving the construction of a new WWTP, called the North Gaza Wastewater Treatment Plant (NGWWTP), is planned to be located further to the south near the Israeli border (see Figure 1) [10]. The new wastewater treatment plant will involve MAR of effluents [11]. The Palestinian Water Authority (PWA), along with international support, decided to use practical, already established MAR technologies such as infiltration ponds with Soil-Aquifer Treatment (SAT) to replenish the coastal aquifer in order to meet the continually rising demand of water for domestic, industrial, and agricultural use in this water-parched region [12–14]. Decision support is required to identify the best MAR project option to implement in the study area.

**Figure 1.** Study area map showing the wastewater treatment plants. Data source [9]. Inset picture from Google Earth.

In order to support the decision makers to plan the MAR project, this paper focuses on the impact assessment for several strategies for the implementation and operation of MAR in the Northern Gaza Strip. The strategies were quantitatively analyzed based on their potential impacts on agriculture, environment, health, society, and the economy. Finally, all strategies were compared to each other and ranked according to their ability to promote water resources development at the Northern Gaza Strip. In addition, this paper also describes the optimal MAR strategy of the candidates considered to sustain water resources and groundwater-dependent environment of Northern Gaza.

### **2. Study Area**

With an area of 365 km2 and a population of roughly 1.6 million [9], the Gaza Strip is located on the southwestern part of historical Palestine at the Mediterranean Coast on the edge of the Sinai Peninsula. Precipitation varies between 200 and 400 mm/year, with an average of ca. 300 mm/year [6,15], and temperatures are generally high, ranging between 29 and 9 °C throughout most of the year [16], while 97% of water used in Northern Gaza comes from the Northern Coastal Aquifer [7]. In this study, a part of North Gaza was selected for analysis and comparison of MAR strategies (Figure 1), which is referred to in this paper as the "study area". The study area was delineated based on the boundary selection process using a groundwater flow and transport model. This model simulates the spreading of infiltration water at the new infiltration ponds, which commenced at the beginning of 2008 and will continue until 2040.

### *Geology and Hydrogeology of the Study Area*

According to [17], the Gaza strip is underlain by a series of geological formations from the Mesozoic to the Quaternary. The two main formations are called Tertiary formation and Quaternary formation. The Tertiary formation, a 1200 m thick layer, is composed mainly of Saqiya formation and it consists of clay, marl and shale [14,18,19]. The 160 m thick Quaternary deposits covers the Pliocene Saqiya formation. The overlying Pleistocene deposits "Lower Quaternary" consists of (1) Marine Kurkur Formation (10–100 m thick on the coast); (2) Continental Kurkur Formation (maximum thickness is about 100 m with often-calcareous cement, and Quaternary Deposits. The sand loess and gravel beds formation is considered the main formation of the Gaza strip [17]. A general geological cross section of the coastal plain can be found in a number of sources [17,20–22] and therefore is not included in this paper.

The North Gaza aquifer is a part of the Coastal aquifer that extends north to south from Haifa to the Sinai Coast. The highly permeable shallow vadose zone is mostly sand and gravel [23]. Larger and more consistent clay layers at the coast and extending 2–5 km inland, divide the Coastal Aquifer into several confined permeable layers [23]. The hydraulic connection between groundwater in the different subaquifers and the sea is not well investigated [17]. Beyond this distance, to the east, the Kurkar Group comprises the unconfined aquifer [18,23]. The average thickness of the aquifer at the coast is 150–200 m [23], whereas at the eastern border with Israel, the average thickness varies between 40 and 50 m [18]. The low-permeability Saqiya Formation of tertiary age constitutes the base of the aquifer. The 1 km thick Saqiya Formation is composed of clay, shale and marl [18]. The transmissivity of the Gaza aquifer ranges between 700 and 5000 m2 /d, corresponding hydraulic conductivity ranges between 20 and 80 m/d. Specific yield and Specific storativity values are 0.1–0.3 and 1 × 10<sup>í</sup><sup>4</sup> per meter [19,24].

Rainfall is the main recharge component for the shallow aquifer unit in the study area. Aish *et al.*, (2009) [20] estimated that the average annual recharge of the Gaza strip is 108 mm/year (39–40 Mm3 /year). Around 1016 agricultural wells pump ca. 50 Mm3 /year and 45 urban supply wells abstract approximately 42 Mm3 /year. Irrigation return flow is considered as 30 Mm3 /year [18]. In the Gaza strip, the groundwater abstraction from the drinking water wells constitute more than 50% of the net withdrawal [25]. In the northern part of Gaza, groundwater levels range from about 2 m above MSL at the eastern border with Israel to mean sea level along the shore [18]. A steep groundwater level gradient is seen at the southern part of the Gaza strip. The coastal aquifer possesses 5000 Mm3 storage of groundwater of variable quality of which 30% is fresh [26,27]. In North Gaza, the GWL in the centre of the area is lower than the other parts of the area. So, in this part of the coastal aquifer, the main groundwater flow direction is towards the centre of North Gaza [28]. Besides the water quantity shortage, groundwater quality related problems, especially chloride and nitrate contamination, have been mentioned by several researchers e.g., [18,19,29]. The existing monitoring network in the Gaza strip observes groundwater level, and measures nitrate and chloride concentrations. The network is not suitable for monitoring sea water intrusion [18].

### **3. Methodology**

An integrated approach was formulated in order to select the best strategy for MAR implementation. The approach is integrated in the sense that the study considered the impacts of possible MAR strategies on several sectors such as environment, health, economy and society. The sequential steps to select the best rank MAR strategy, a structured and sequential work flow was prepared, as shown in Figure 2. In general, the entire process involves three main steps to identify the best ranked MAR strategy: (a) water resources system analysis and strategy development (b) strategy ranking: criteria selection, impact assessment and criteria quantification, and (c) Multicriteria decision analysis (MCDA).

**Figure 2.** Overall methodology of the study.

The main objective of water resources system analysis (step-1) is to identify the main water resources drivers and pressures, and the potential responses to solve the impacts. Causal chain analysis using the Driver (D), Pressure (P), State (S), Impact (I) and Response (R), in short DPSIR, methodology [30,31] can be used at this step. Based on the pertaining water resources problem and the potential responses, water resources strategies are developed (step-2). The strategies should comply with the national water policy. In the third step of strategy ranking, relevant environmental, health, social and economic characteristics are selected. Each characteristic is defined as a criterion. The next step involves the decomposition of the ultimate goal into a hierarchy of several levels, following the principle of Analytical Hierarchy Process (AHP). The bottom level is the most specific criteria and the middle levels are more general criteria and can be called the "main criteria". The criteria in the lowest level are related to the main criteria in the middle levels. All levels combined is the goal of the study—the best strategy for MAR implementation, and is positioned at the top of the hierarchy. The next step in the strategy ranking procedure is assigning values of relative importance for each criterion at all levels, which is done by assigning a weight to each criterion. The criteria under each main criterion are compared amongst themselves and a weight is assigned to each one (step-4). The main criteria are also weighted in this way. The next step (step-5) is to quantify the relevant criteria, which is the main focus of the present study. A number of techniques, such as groundwater modelling, GIS and field surveys are available to quantify scores for the criteria. The quantification procedure depends on the type of criterion. After quantifying all criteria, an evaluation matrix is prepared at this step which is one of the principle components for ranking of alternatives. The final step (step-6), strategy comparison and ranking analysis, encompasses two multi-criteria analysis techniques: Weighted Linear Combination (WLC) and PROMETHEE II (Preference Ranking Organisation MeTHod for Enrichment Evaluations) method.

The role of AHP, mentioned earlier, was to construct the hierarchy and to estimate the relative weight by pairwise comparison, after getting the preference information from the researchers, decision makers and stakeholders. Additionally, the role of WLC and PROMETHEE is to rank the alternatives.

### **4. Water Resources Problem Analysis and Strategy Development**

### *4.1. Water Resources Problem Analysis (Step-1)*

With the aim to analyse the existing water resources problems of the study area, causal chain analysis using the DPSIR method was used. The DPSIR concept has been developed for describing interactions between society and the environment [31,32], starting from the assumption that there is an interaction between the two. The water resources problems of North Gaza were analyzed, decomposed, and structured in this method in order to find the potential response of the problem. In brief, the water resources system of North Gaza is affected by two main drivers: population and urbanization. These drivers cause certain pressures on groundwater exploitation, wastewater status, land-use change, salinization, *etc.* The causal chain analysis of surface water is negligible as there are no surface water resources in the area. The DPSIR analysis has identified four potential responses to the current water resources problem. Each response can be considered and studied independently as well as in combination. In this paper, we considered MAR as a potential response due to the following reasons: (1) the poverty level in Gaza is high and many cannot afford the costs of advanced water treatment or desalination (considered as innovative technology) [33]; (2) Treated wastewater reuse will complement the existing water resources and will improve the water supply for agriculture; (3) Use of reclaimed water for agriculture would make fresh groundwater available for domestic and industrial use. In this study, MAR is seen not only as a contribution for a solution to the water supply and groundwater quality issue, but also as a solution to the problematic effluent lake, located at Beit Lahia, as the use of the new infiltration pond would help to rehabilitate the old infiltration lake.

### *4.2. Water Resources Strategy Development (Step-2)*

Based on the water resources problem analysis and considering the water resources management plans for the years 2005–2025 [2,5,9,10], the following four MAR strategies were established in this study (Table 1).

The water management strategies based on MAR presented in Table 1 consider three phases in terms of wastewater resources development at the case study area. Strategy No.1 (Sc-1) represents the strategies if nothing has been changed with respect to the existing water resources structure and no further planning is being considered. Strategy No.2 (Sc-2) is linked to the first phase. This phase considers the diversion of the water from the BLWWTP to the newly constructed infiltration basin, which is located close to the foreseen position of the new North Gaza Wastewater Treatment Plant (NGWWTP) at the Israeli border. The diversion of water will be accomplished via a pressurized pipeline and the effluents will then infiltrate into the aquifer. Strategy No.3 (Sc-3) considers the strategies if the diverted water will be treated in the NGWWTP and then infiltrated into the aquifer. The effluent quality is higher than that of the water used for infiltration in Sc-2. In Phase 3, the NGWWTP is designed to increase the treatment capacity of around 24 Mm3 per year in 2025. It indicates in Sc-3, the effluent water quality is better than that in Sc-2. Strategy No.4 (Sc-4) considers infiltration of this extra volume of treated water to the aquifer. Sc-2, Sc-3, and Sc-4 are considered as MAR management strategies.


**Table 1.** MAR management strategies towards the development of water resources at the Northern Gaza Strip.

Note: \* in natural recharge.

### **5. Criteria Selection and Quantification Procedure**

### *5.1. Criteria Selection (Step-3)*

A wide range of indicators are considered for the selection of criteria. The criteria were derived from the identified sectors of impact and emphasis was given to the availability of information to quantify the criteria. A total of 19 most representative decision criteria were selected in close cooperation with Palestinian researchers and authorities as well as further relevant stakeholders and were discussed with other international experts in related fields. Among the 19 criteria, six criteria represent environment considerations. They consider groundwater level, chloride and nitrate concentration averaged year 2005–2040 and also in year 2040 alone. Four health criteria consider chloride and nitrate concentration at the domestic wells average 2005–2040 and also in year 2040 alone. Seven social criteria consider people's acceptance, convenience, satisfaction with the water quality and quantity, employment and willingness to pay. Affordability to pay and net cost-benefit analysis were considered as economic criteria.

Figure 3 shows the four-level hierarchical structure of the categories and criteria. AHP was used at this step. The AHP, proposed by [34], is a multicriteria analysis technique that enables the explicit ranking of tangible and intangible factors against each other for the purpose of decision-making or conflict resolution. It combines qualitative and quantitative approaches [35].

**Figure 3.** Criteria selection and hierarchy. Italic numbers indicate the number of criteria associated to each item at the fourth level.

Nineteen criteria were grouped into four main criteria groups such as environmental, health, social and economic. At the third level of the hierarchy, social, health, and economic criteria were grouped as "socio-economic" criteria. Socio-economic criteria and environmental criteria group combines the ranking of the strategies.

### *5.2. Criteria Weighting (Step-4)*

The relevant importance of each criterion was defined in close cooperation with local scientists, decision makers and stakeholders. A participatory process was undertaken among the local stakeholders and experts. The participatory process includes scientific meetings, questionnaire surveys and workshops. Judgments of international experts were considered along with the weights from local experts and stakeholders. The pairwise comparison method, originally proposed by [34], was used to transfer the linguistic importance to numeric value and relative weights were estimated. The net cost and groundwater quantity were considered to be the most important criteria. All categories at level 2 and level 3 were considered as being equally important for MAR planning and management.

### *5.3. Criteria Quantification (Step-5)*

The selected criteria were quantified using several state-of-art analysis techniques such as groundwater flow and transport models, field surveys, economic models, *etc*.

### 5.3.1. Quantification of Environmental Criteria (Criteria 1 to 6)

The selected environmental criteria refer to the groundwater quality and quantity status. These criteria were quantified by using groundwater-modelling techniques. A groundwater flow and transport model, developed in this case study using VISUAL MODFLOW (version 4.3, SWS, Vancouver, BC, Canada, 2009) and its integrated modules, was used to quantify the six environmental criteria in this study. The detailed description of the flow model set up and model parameters together with calibration plot can be found in [28]. The transport parameters such as longitudinal and vertical transverse dispersivity were initially assigned values of 4 m and 1 m, respectively (according to [36]). Bulk density of water was considered as 1000 kg/m3 . For Sc-2 and Sc-3, the infiltration starts in 2008 with 9.7 Mm3 of treated water and with an increase of infiltration by 0.08 Mm3 per year until 2012 and afterwards the infiltration volume remains 13 Mm3 until 2040. For Sc-4, the infiltration starts in 2008 with 9.7 Mm3 of treated water and with an increase of infiltration by 0.08 Mm3 per year until 2040. During the analysis and quantification of all the strategies, the current water withdrawal for agriculture was assumed to be constant. Domestic water demand was assumed to increase (based on population growth), according to the estimated demand increase. The model was run until year 2040. Simulation results flow and transport modelling from years 2005–2040 were used to estimate Criteria 01, Criteria 03 and Criteria 05. Simulation results from flow and transport modelling at the end of year 2040 were used to quantify Criteria 02, Criteria 04 and Criteria 06.

### 5.3.2. Quantification of Health Criteria (Criteria 7–10)

The four health-related criteria refer to the water quality status at the domestic water supply wells. Average chloride and nitrate concentration were considered at the places where the domestic wells are situated (Criteria 07 and Criteria 08). Criteria 09 and Criteria 10 were quantified by considering the average concentration of chloride and nitrate in the waters of the study area aquifer. The developed groundwater flow and transport model was also used to quantify the health criteria for the analysis. The water quality in the domestic wells depends on the quality of infiltrated water, quality of native groundwater and the seasons (winter and summer). These three aspects were considered in the model.

### *5.4. Model Simulation for the Health Criteria Quantification for the Strategies*

### 5.4.1. Chloride

Chloride was modelled as a conservative parameter and hence, no sorption or kinetic reaction was considered. The initial concentration, ranges between 40 and 2200 mg/L, of chloride was taken from the trend analysis in [37], considering the data from the years 1984–1998 [37,38]. The chloride concentration of the infiltrated water was considered to be the same as that in the wastewater lake at BLWWTP. The chloride concentration used in the model and during the entire modelling period was 559–857 mg/L for years 2004–2007 and 250 mg/L for years 2008–2040 in all strategies except Sc-1 [9]. For Sc-1, the base condition was maintained. The base condition considers the chloride concentration used in the simulation model from year 2000 to year 2003. The effect of chloride concentration changes as the volume of infiltrated water changes in different scenarios.

### 5.4.2. Nitrate

For nitrate simulation, equilibrium controlled linear isotherm was considered and no kinetic reaction was considered. Similar to chloride, the initial concentration, ranges between 5 and 370 mg/L, of nitrate was taken from the trend analysis from [37] and considered is the data from 1984 to 1998. The nitrate quality of the infiltrated water was calculated based on the quality of the infiltrated water, the infiltration process, and seasonal climatic conditions (after [37,38]) and where dilution and denitrification have been assumed to be the main processes for nitrate reduction in the model simulation. For Sc-1, a base condition was maintained throughout the entire simulation period. A base condition maintains the nitrate source, considering the same land use utilized in the simulation model 2000–2003. The nitrate concentration for Sc-2 used in the model and during the entire modelling period was 20–107 mg/L for years 2004–2007 and 19–43 mg/L for years 2008–2040. The nitrate concentration for Sc-3 and Sc-4 used in the model and during the entire modelling period was 20–107 mg/L, 19–43 mg/L and 7.5–17 mg/L for the period of 2004–2007, 2008–2011, and 2012–2040, respectively.

### 5.4.3. Quantification of Social Criteria (Criteria 11 and 17)

A questionnaire survey was performed by the Palestinian Hydrology Group to get the social aspect of the MAR strategies [33]. The questionnaire was prepared in such a way that it includes criteria that would measure the anticipated level of convenience, perceptions on willingness to use the recharged water for different purposes and the fees that the user would be willing to pay for the supply and the expected level of satisfaction from the quantity and quality of water supplied from each option. A total of 76 questionnaires were filled out by the locals in the area [33]. The number of questionnaire was decided based on statistical analysis and population residing at the study area.

### 5.4.4. Quantification of Economic Criteria (Criteria 18 and 19)

In the present study, two economic criteria were considered. Affordability to pay (criteria 18) was quantified using the surveyed data. Criterion 19 considers the net present cost and benefit of the four strategies implementation. For net present cost and benefit estimation, the following factors were considered (after [37]):


The cost estimation was done using an economic model based on a spreadsheet.

### *5.5. MCDA Analysis and Ranking of Options*

After quantification of all the criteria, the normalized matrix was prepared for multicriteria analysis. The normalization was done using the following formulae:

$$NV = \frac{Max - Value}{Max - Min} \tag{1}$$

Here, NV denotes normalized value, Max and Min indicate the maximum and minimum value among the values to be normalized, respectively. We use Equation 1 to normalize all criteria values.

### 5.5.1. Criteria Aggregation Methods: Weighted Linear Combination (WLC)

WLC combines the criteria and provides the ranking. WLC is the most simple and commonly used aggregation method in decision analysis [35].

$$\mathbf{S(x\_{\\_i})} = \sum \mathbf{w}\_{\\_j} \cdot \mathbf{s}\_{\\_j}(\mathbf{x}\_{\\_i}) \tag{2}$$

where, wj is a normalised weight; and Ȉ wj = 1; and sj(xi) is the normalised criteria function.

After receiving the criteria weights and preparing the evaluation matrix, the role of WLC is to perform weighted summation for each group of criteria at all levels of the hierarchy until the strategy ranking achieves.

### 5.5.2. PROMETHEE

PROMETHEE, developed by [39], is a nonparametric outranking method for a finite set of alternatives. The method was later extended by [40,41]. PROMTHEE I gives partial ranking and PROMETHEE II provides a complete ranking of the strategies by using the net flow [42]. The details of the procedure can be found in many sources such as [39,43–45].

### **6. Results Analysis**

### *6.1. Environmental Criteria*

The simulations show (see Figure 4a) that the maximum average GWL rise in the study area is 6 m by the year 2028 with respect to "Do nothing" (Sc-1). At the end of 2040, the GWLs are estimated to be í2.61 m, 0.81 m, and 3.57 m above sea level (ASL) for Sc-1, Sc-2 & Sc-3, and Sc-4, respectively. 3%–5% of the infiltrated water may flow to Israel each year under the simulation condition of Sc-2 and Sc-3, whereas this outflow was estimated to be 7%–15% per year for Sc-4. The inflow to the study area from the Israeli side will be reduced by 20%, for both Sc-2, Sc-3 and by 30% for Sc-4. Due to the infiltration of treated wastewater, the groundwater level below the infiltration basin would increase and would cause the fresh water flow to be reduced from the Israeli side.

Figure 4b shows the average chloride concentration in the study area for the four strategies. The model results show the average chloride concentrations at the end of 2040 are 522 mg/L, 426 mg/L, and 400 mg/L for Sc-1, Sc-2 & Sc-3, and Sc-4, respectively. Figure 4c shows the average nitrate (expressed as NO3-N) concentration in the study area for the four strategies. The average nitrate concentrations at the end of 2040 are 82.27 mg/L, 67 mg/L, 59 mg/L, and 44 mg/L for Sc-1, Sc-2, Sc-3, and Sc-4, respectively. Implementation of Sc-4 will therefore provide storage in the aquifer with a maximum value of 23 Mm3 per year after the full implementation of north Gaza wastewater treatment plant (NGWWTP), Phase 3 (year 2025).

**Figure 4.** (**a**) Average groundwater level; (**b**) average chloride concentration and (**c**) average nitrate concentration in the study area during year 2005 to year 2040 for the four MAR strategies.

### *6.2. Health Criteria*

A total of 10 domestic wells are located within the study area. Figure 5a shows the average chloride content of the 10 domestic wells for the four strategies until the year 2040. The average chloride concentrations at the end of 2040 are 555 mg/L, 528 mg/L and 407 mg/L for Sc-1, Sc-2 & 3, and Sc-4, respectively. In the case of Sc-1, the average chloride concentration in all domestic wells increases until the year 2040. In the case of Sc-2 & 3 and Sc-4, the average chloride concentration increases until the year 2035 and 2030, respectively, and then the chloride concentration decreases. Figure 5b shows the average nitrate content of the 10 domestic wells for the four strategies until the year 2040. Minimum nitrate concentration was observed in case of Sc-4. The average nitrate concentrations at the end of 2040 are 90 mg/L, 72 mg/L, 68 mg/L, and 49 mg/L for Sc-1, Sc-2, Sc-3, and Sc-4, respectively.

**Figure 5.** (**a**) Average chloride concentration; (**b**) Average nitrate concentration in the ten domestic wells for the entire simulation period (year 2005 to year 2040).

### *6.3. Social Criteria*

The survey results indicate that 86% of the respondents agreed to reuse wastewater for agricultural purposes whereas 67% and 42% agreed to reuse wastewater for industrial and domestic purposes, respectively. Results also show that respondents are willing to pay very little for the infiltrated water regardless of use and claim to be able to afford very small fees. The inhabitants are willing to pay a maximum \$0.37/m3 to use wastewater for irrigation (Figure 6). The survey results indicate that the distribution of acceptance and satisfaction of the public is similar throughout the various MAR strategies. In terms of satisfaction with the water quality, perceptions range from being satisfied to fairly satisfied with Sc-3 and Sc-4 having the greatest level of satisfaction.

**Figure 6.** Willingness to pay of the respondents for the MAR strategies for different usage.

### *6.4. Economic Criteria*

In the study area most of the people depend on agriculture, and many youths and women participate in agricultural activities. The agricultural activities in the study area depend on the groundwater irrigation. Hence, it is important to carefully review the water price (tariffs) for project feasibility. The survey results indicate that the respondents cannot afford to expend more money in order to use the benefit gained due to implementation of Sc-2, Sc-3 and Sc-4.

High investment cost is an important factor that makes a big difference between MAR strategies (Sc-2, Sc-3 and Sc-4) and the "Do nothing approach" (Sc-1). From the net benefit (cost-benefit) estimation (Figure 7), the implementation of a MAR strategy would be beneficial after year 2022 in case of Sc-4 and after year 2024 in case of Sc-2 and 3 (Figure 7). Sc-4 returns the most benefit due to its extended amount of infiltration volume even after year 2012. The net present values of the strategies (years 2005–2040) are \$10.2 M for Sc-2 and Sc-3 and \$28.4 M for Sc-4 whereas for Sc-1 the value is í\$32. 0 M (negative sign indicates net cost). That is, there is a \$60.4 M PV net benefit of switching from strategy Sc-1 to Sc-4 or a \$42.2 M PV net benefit of switching from Strategy Sc-1 to either Sc-2 or Sc-3.

**Figure 7.** Net benefit analysis for the four MAR strategies.

### **7. Strategy Comparison and Ranking**

Figure 8 shows the performance of the four strategies according to the main criteria group (level-2). It is clear from the figure that Sc-4 performs the best in environmental, health and social criteria and Sc-1 performs the worst in these cases. Sc-2 performs better that Sc-3 according to the social and economic criteria but performs worse than Sc-3 for environmental and health criteria. People's affordability, convenience, and acceptance of wastewater seem important for the ranking. The final ranking was achieved after combining the main criteria groups (level-4) and the ranking is Sc-4 > Sc-3 > Sc-2 > Sc-1.

It was found that Sc-4 performs best for all the quantified detailed criteria with the following exceptions; average chloride concentrations in domestic wells over the study period, satisfaction with domestic water quality, willingness to pay and affordability to pay. These deviations are due to temporarily increased salinity of domestic wells in specific locations due to changed flow directions and variable salinity in the aquifer. This also influences criteria for satisfaction with domestic water quality for users of those domestic wells, and willingness to pay. Sc-4 also has the highest capital costs of all options (affecting affordability to pay), although the net benefits are greatest. For these specific criteria, only the "Do Nothing Case" (Sc-1) performs the best, although for other criteria it performs very poorly compared with other options, especially Sc-4.

PROMETHEE I partial ranking also confirms that Sc-4 performs better than the other strategies. No out-ranking relation does exist between Sc-2 and Sc-1; and Sc-1 and Sc-3. PROMETHEE II ranking is similar to that observed using WLC method.

**Figure 8.** Ranking of the strategies according to main criteria group (level 2) using AHP-WLC combination.

### **8. Discussion**

### *8.1. Criteria Quantification*

### 8.1.1. Environmental Criteria

The Sc-1 ("Do Nothing Approach") indicates continuous groundwater level mining over time, whereas Sc-4 indicates higher groundwater development than the other three strategies. Similarly, among the four strategies, Sc-4 shows better conditions in terms of inflow from the sea to North Gaza. Infiltration of excess treated wastewater even after 2012 might help Sc-4 to get more environmental benefit. In general, the problem of water flow from the sea will remain under control by the infiltration of all MAR strategies. It is clear from the results that Sc-1 ("Do Nothing Approach") will lead to deterioration of groundwater quality (*i.e.*, chloride and nitrate increase) with time. However, for other strategies, the groundwater quality will improve with time. The long-term effect of groundwater flow might also control the groundwater quality in the study area as the distribution of chloride and nitrate in North Gaza and the nearby Israel border is complex.

From the groundwater model simulation, we delineated a zone of ca. 200 m from the edge of the infiltration basins receiving the infiltrated water with a residence time of ca. six months. Regarding pathogenic bacteria, residence time of more than 6 months is recommended [46]. In the study area, no domestic wells exist within these 200 m.

### 8.1.2. Health Criteria

The impact of managed aquifer recharge projects on domestic wells is very sensitive to the population living in the area. The simulation result for Sc-4 shows a significant chloride concentration decrease in the study area at the end of the year 2040 in comparison to Sc-1, Sc-2 and Sc-3. By analysing the chloride concentrations in all domestic wells and comparing them with the "Do Nothing" strategy, observations show that the impact on chloride concentrations in all wells will be almost the same. Due to the groundwater flow direction of the infiltrated treated effluent, this would also impact the domestic wells. The increasing trend in the domestic well chloride concentration is due to the higher chloride concentration in the infiltrated water than the native groundwater and groundwater flow direction. In general, the nearby aquifer of the wells and the aquifer beneath the infiltration basin display higher chloride concentration. The infiltrated water would displace this water towards the domestic wells and the chloride concentration rises at the wells. The infiltrated water replaces the worse quality water and chloride concentrations at the wells are expected to decrease with time.

The nitrate concentration at the locations where the domestic wells are located is comparatively higher than the nitrate concentration below the infiltration pond and the nitrate concentration in the infiltrated water. The nitrate concentration in all domestic wells will be slightly improved.

### 8.1.3. Social Criteria

In general, the inhabitants are willing to pay more if fully treated wastewater is reused. Respondents do not agree to use the infiltrated water for domestic purposes but they have higher acceptance to use this water for agricultural or industrial purposes. The reuse of treated wastewater for irrigated agriculture would save higher quality groundwater water for drinking water supply and subsequently may solve some environmental problems. The health and religious aspects could be a major concern of people of Gaza to reuse wastewater [13]. The study found that the education level, standard of living and the environment might be key issues in order to convince the people of Gaza to reuse wastewater in agriculture.

### 8.1.4. Economic Criteria

Implementation of Sc-4 would lead to the maximum benefit. Reuse of wastewater would offer the release of corresponding fresh water resources and will help to expand the overall irrigated area by providing more water to irrigate lands. Hence, the livelihood of the residents may improve. Besides the above-mentioned benefits, more indirect benefits may be gained from improving groundwater quality. These are increased safety and the benefits generated from freeing the land that the current effluent lagoon occupies as well as the other subjective benefits related to seawater intrusion. Finally, the MAR project would create many other supported jobs e.g., related to MAR operation and agricultural activities *etc*.

### *8.2. Strategy Comparison and Ranking*

According to the analysis using WLC and PROMETHEE, the same rankings of options were achieved. The comparison of water management options showed that increasing investments in wastewater collection, treatment, and later MAR would result in an improved water management strategy performance with regards to the considered environmental, social, and health criteria. Obvious drawbacks are the investments for infrastructure and their impact on economic feasibility. This should be discussed in greater depth and should be based on comprehensive cost-benefit analysis (CBA) and cost effectiveness that should refer to cost minimization and the related environmental and health benefits, which are fundamental to guarantee the sustainable development of the Gaza Strip.

### **9. Conclusions and Recommendations**

The present study clearly shows the importance of environmental, health, social and economic impact assessment of MAR strategies performing a case study in North Gaza. The integrated approach of combining field campaign, methodological analysis and mathematical modelling has been proven to be effective for a multicriteria decision analysis. In order to increase water supply and to combat water scarcity, water pollution, and health problems at the Northern Gaza Strip, appropriate water resources planning and management measures are urgently required. Reuse of the treated effluent by MAR would strengthen agricultural development and result in increased groundwater availability for domestic and industrial use. The reuse of treated effluent has already been adapted in the national Water Policy for the Gaza Strip [47]. The present study shows that the so-called "Do Nothing Approach" is no real option for Northern Gaza, contributing to further groundwater level decline and groundwater quality deterioration, and increasing health risks for the population of Gaza. The performance analysis of the developed water resources planning and management strategies clearly shows that managed aquifer recharge by infiltration ponds with proper treatment of the effluents is a viable response to the increasing water resources problems of the region. In order to maximize project benefit, optimal pond operation based on practical experiences and regular cleaning of the pond is required to avoid clogging of the pond bed. Application of several MCDA analysis methods probes the robustness of the ranking analysis.

Ten domestic wells will be affected over time due to displacement of relatively low quality groundwater towards the abstraction wells. However, with time, the low quality water will be replaced by the nearby infiltrated water. Special care for water recovery should therefore be planned to protect the existing domestic wells. Another option could be to use the affected domestic wells for agricultural use and use the nearby unaffected wells for domestic water supply. Nevertheless, regular water quality monitoring of abstracted water and efficient recovery wells should be considered. Tremendous effort is required to increase public awareness for wastewater reuse. Adequate water pricing should be made considering the level of income and economic feasibility of the MAR project.

Additional investments should be undertaken for better maintenance and to further extend the wastewater collection network as well as the capacity of the NGWWTP at the Israeli border, accompanying the rapidly increasing wastewater production. Furthermore, managed aquifer recharge contributes to the control of seawater intrusion and groundwater salinity.

Due to the unavailability of scientific data, a variable-density groundwater flow model was not considered in this case study. As the objective of the study is not to quantify salinity intrusion, rather compare different management scenarios, the fresh water flow model is sufficient. In order to investigate the effect of MAR strategies on saline groundwater intrusion into the coastal aquifer, a variable-density groundwater flow model is recommended.

The approach and techniques used in this study can be applicable not only to MAR project implementation but also to other water resource development projects.

### **Acknowledgments**

The authors gratefully acknowledge the contribution of European Commission (EC) to fund the study under the FP6 program, with the project entitled "Groundwater Artificial recharge Based on Alternative sources of wateR: aDvanced INtegrated technologies and management" (GABARDINE), contract No. 518118. Sincere thanks to the Palestinian Hydrology Group for their helpful support and providing the field data. The authors would also like to express their gratitude to the reviewers and the external editor whose collective input substantially improved this paper. We acknowledge support by Deutsche Forschungsgemeinschaft and Open Access Publishing Fund of Leibniz Universität Hannover.

### **Author Contributions**

The text of this article was written by Mohammad Azizur Rahman, Bernd Rusteberg with contributions from Muath Abu Saada, Ayman Rabi and Martin Sauter. Mohammad Azizur Rahman conducted background research on integrated approach for MAR, developed the model and analysed the results. Bernd Rusteberg was involved in method development and coordinated the study. Mohammad Salah Uddin contributed to the model development and result analysis. Muath Abu Saadah provided support to the model development and field data collection. Ayman Rabi involved in coordination of social surveys and contributed to the economic analysis. Martin Sauter provided content review and helped to shape the presentation of our results.

### **Conflicts of Interest**

The authors declare no conflict of interest.

### **References**

