1. Introduction
With population growth and expanding economies, wastewater treatment is becoming increasingly important in many countries. It has been estimated that 90% of wastewater in undeveloped countries is dumped directly into streams, ponds, or seas [
1]. This may cause human health problems, damage to water ecosystems, and ultimately lead to a shortage of clean water [
2]. At present, the construction of wastewater treatment plants (WWTPs) is an important method to solve problems of water pollution, and many treatments are shifting toward resource recovery [
3,
4] to address the water pollution crisis. In developed countries, such as Spain, there is an increasing number of WWTPs oriented towards chemical recovery or sludge thickening [
5]. However, resource recovery is very uncommon in Cambodia [
6]. Infrastructure for wastewater collection and treatment is lacking in Phnom Penh. There are lots of polluted rivers and lakes in the southern and northern regions of Phnom Penh due to the direct discharge of wastewater. Additionally, the discharges may pollute nearby surface and underground water bodies through seepage, which will exacerbate environmental difficulties and the health of the surrounding water resources [
7,
8,
9,
10]. In order to reduce these negative effects, the construction of WWTPs should be improved. Therefore, the reasonable location selection of WWTPs is especially important [
11].
Numerous studies have investigated the location selection of urban infrastructure [
12,
13,
14,
15], hospital sites [
16,
17,
18], emergency shelters [
19,
20,
21], landfill sites [
22,
23,
24] and power plants [
25,
26,
27]. However, the theoretical model for WWTP location selection is different from that of other types of urban infrastructure. Due to the special characteristics of the WWTPs, the location selection should consider not only the demand for WWTPs and the suitability of the construction site, but also forbidden construction zones according to laws and regulations. Since geographic information system (GIS) can manage multiple sources and large amounts of spatial data [
28], the location of WWTPs can be determined using GIS by analyzing various indicators, including topography, prevailing wind direction, hydrology, land use type, and spatial distribution of surface water [
29,
30,
31]. Various location selection models have been investigated thoroughly for WWTPs, such as a model combining artificial neural networks [
32] with the genetic algorithm, a model based on fuzzy logic and multi-criteria decision creation [
33], a multi-objective optimization model for two-fold optimization of price and energy consumption [
34], models combining GIS and the analytic hierarchy process (AHP) with remote sensing (RS) techniques [
35,
36,
37], and models using GIS integrated with the AHP [
38,
39,
40].
However, the majority of current WWTPs location selection models only focus on geographic elements, while ignoring dynamic changes to the water ecosystem caused by WWTPs and local socio-cultural indicators. In the current work, for water ecosystems, time series RS data are used to derive the spatial and temporal distribution of the eutrophication level of water bodies, which reveals patterns of surface water ecological change, and is included in the demand indicators. Additionally, in terms of social-cultural indicators, WWTPs should be located far away from religious sites. Hence, the buffer zones determined from religious point of interest (POI) data should be integrated into the exclusion indicators. Finally, the model of minimized facilities number and the model of maximum coverage are combined to determine the locations of WWTPs. An empirical study is conducted in Phnom Penh, Cambodia, in order to enhance the facility service coverage of WWTPs effectively while respecting local socio-cultural customs fully and considering dynamic changes in water ecology.
2. Methods
The proposed WWTPs location selection method consists of three steps: Firstly, a novel indicator system is established. Due to the varying degrees of influence among indicators, the AHP is used to determine each indicator’s weight. Secondly, a weighted overlay analysis is applied to obtain the demand points map and the suitability map of WWTPs by the established indicator system. Finally, the locations for WWTPs are chosen by minimized facilitates and maximum coverage models. The main steps are depicted in
Figure 1.
2.1. The Indicator System
The indicator system adopted by this study consists of exclusion indicators, suitability indicators and demand indicators. The exclusion indicators are utilized to determine forbidden construction zones based on laws, regulations, experience, etc., which are determined by residence (f1), social-culture (f2), transportation (f3) and water body (f4). The suitability indicators are determined by slope (s1), hydrology (s2), river (s3), transportation (s4) and wind direction (s5), which evaluate suitable locations for the construction of WWTPs. The demand indicators represent the demand for WWTPs, which are determined by population density (q1), industrial zone (q2) and water ecological condition(q3).
2.1.1. Exclusion Indicators
In Phnom Penh, religion plays an important role in society. In consideration of local socio-cultural factors, WWTPs should not be constructed close to religious sites (f
1) in order to respect socio-cultural customs and prevent sludge and exhaust gas from polluting these areas. WWTPs affect public health, so a specified distance should be set to separate WWTPs and residential areas (f
2). Additionally, WWTPs close to roads (f
3) affect the landscape, and those close to water bodies (f
4) contaminate water and fish. In conclusion, the buffer zones for neighborhoods, religious sites, main roads, and water bodies are established as forbidden construction zones, within which WWTPS cannot be constructed. The buffer distance of 300 m [
41] is set for residential areas and religious sites, the buffer distance of 50 m is set for roads and water bodies; and the reclassification level of these regions is set to zero in
Table 1, indicating WWTPs mustn’t be constructed in these areas.
2.1.2. Suitability Indicators
WWTPs should be built on mild-to-moderate slope (s
1) to make use of the natural flow of wastewater. Additionally, taking into account the direction of water flow, WWTPs should be built downstream of rivers (s
2). WWTPs locations should be selected not too far from rivers (s
3) and roads (s
4), in order to reduce the expense of transporting extra sludge and building additional pipeline networks [
42]. Given that WWTPs emit hazardous gases, their locations should be in downwind (s
5) areas in order to promote the diffusion of harmful gases during wastewater processing and reduce contamination of the environment.
The suitability indicators are listed in
Table 2, which are reclassified into five levels, with the higher the level, the more suitable for WWTP construction.
2.1.3. Demand Indicators
Residential domestic wastewater, industrial wastewater, and a small amount of wastewater released from other sources make up the majority of sources of urban wastewater [
43]. In this study, population density is used to assess the demand for WWTPs in residential regions, and the demand for WWTPs is higher in more densely populated areas. The demand from industrial zones for WWTPs should receive more attention since wastewater is discharged at higher rates from industrial zones than from residential domestic areas. The demand from industrial zones for WWTPs is assessed based on the scale of industrial activity, which depends on the category, quantity, layout, construction and local environment conditions [
44]. For the same scale of industrial activity, the demand from industrial zones is conducted by distance analysis, from proximity to distance, with a higher demand level for closer areas.
Due to the impact of human activity and climate change, the water ecosystem is constantly changing. Water-color RS is a trustworthy method for evaluating the degree of eutrophication of water bodies, and it is beneficial in monitoring the dynamics of water ecosystems with a strong correlation to chlorophyll concentration and other water quality parameters. In water-color RS, the Forel–Ule Index (FUI) can reflect the water quality state [
45]. By analyzing the degree of eutrophication of a water body derived from FUI acquired over a number of years, the demand for WWTPs can be obtained from the ecological state of a surface water body. Specifically, the FUI can be used to determine five states [
46]: ① “poor eutrophic” state (0 < FUI < 5,) ② “relatively poor eutrophic” state (5 ≤ FUI < 9); ③ “medium eutrophic” state (9 ≤ FUI < 13); ④ “relatively eutrophic” state (13 ≤ FUI < 17,); ⑤ “eutrophic” state (17 ≤FUI ≤ 21). The higher the eutrophication state, the large the FUI value, and the greater the demand for WWTPs. Additionally, the frequency of occurrence of eutrophication is defined as the frequency of occurrence of the “eutrophic” state (17 ≤ FUI ≤ 21) in the monitoring period by RS data; while eutrophication proportion is calculated according to Equation (1), which reflects the water ecosystem condition.
where E denotes the eutrophication proportion, G is the number of annual quarters with FUI greater than 16 in each grid, T is the total number of quarters, and n is the n-th grid.
In
Table 3, the demand indicators are reclassified into five levels, and the higher the reclassification level, the higher the demand for WWTPs.
2.2. Indicator Layer Processing
Outside the forbidden construction zones, we resample five suitability indicators (slope s
1, flow direction s
2, distance from roads s
3, distance from rivers s
4 and wind direction s
5) into 500 m × 500 m grids, respectively. Then, these are combined into an overall suitability level S according to Equation (2). The higher the value of the indicator S, the more suitable these areas are for constructing WWTPs. The APH is applied to determine the five indicators’ weight.
where v denotes the v-th suitability indicator,
is the weight of the v-th indicator, and s
v is the suitability of the v-th indicator. The grid center points in the suitability map are regarded as candidate points.
We use town geometric center points with high population density, the center points of industrial zones and high eutrophication proportion points to form total demand points, and the density of demand points reflects the level of demand for WWTPs in the region.
2.3. Location Selection Model
The GIS spatial location selection model for WWTPs is introduced in this section. To select WWTP locations reasonably, a model combining the minimized facilities number method and the maximum coverage method is applied. When the effective service radius of the WWTPs covers all demand points, the model of minimized facilities number is used to choose the smallest number of WWTPs needed. In order to enable the effective service radius of the given number of WWTPs to be the largest, the locations of WWTPs are solved using the maximum coverage model. The mathematical model is as follows:
Decision variables x
ij should be binary value variables:
Subject to:
where i represents the i-th demand point, m is the total number of demand points, and the point
j represents the j-th WWTP candidate point, n represents the total number of candidate points, and candidate point
. Equation (3) is the objective function—to select as few points P as possible among candidate points and maximize the total demand covered by WWTPs; r represents the service radius of the WWTP, assuming that the service radius of each WWTP is the same; P is the number of WWTPs to be built. The d
ij is the distance from i(a
i, b
i) to j(a
j, b
j),
. Equation (4) indicates that each WWTP serves at least one demand point. Equation (5) determines whether the demand point is covered by the WWTP.
5. Conclusions
In order to scientifically plan the location of new WWTPs, this study proposes a novel indicator system based on a GIS spatial analysis location selection model. The indicator system selects 12 criteria to compose exclusion indicators, suitability indicators and demand indicators. Considering socio-cultural factors, the buffer zones of religious POI are included in the exclusion indicators. In addition, the spatial distribution changes in the eutrophication state of water bodies are extracted by RS time series data as demand indicators. Finally, we utilize the proposed novel indicator system of WWTP location selection to determine the location of new WWTPs using the minimized facilities number and maximum coverage models. The results show that this multi-factor indicator system enriches the siting model by considering the water ecosystem and socio-cultural indicators, and can effectively delineate the optimal locations for WWTPs. The location selection results can provide a scientific reference for the construction of WWTPs in the field.
Finally, it is highly recommended that further location surveys of selected locations should be carried out in the presence of professional engineers to determine the optimal location of sewerage facilities.