You are on page 1of 15

Environ Earth Sci (2013) 68:1375–1389

DOI 10.1007/s12665-012-1836-3

ORIGINAL ARTICLE

Minimization of environmental risk of landfill site using fuzzy


logic, analytical hierarchy process, and weighted linear
combination methodology in a geographic information system
environment
Abdelwaheb Aydi • Moncef Zairi • Hamed Ben Dhia

Received: 20 January 2012 / Accepted: 9 July 2012 / Published online: 25 July 2012
 Springer-Verlag 2012

Abstract This paper presents a GIS-based multi-criteria Abbreviations


site selection for municipal solid waste landfilling in Ari- GIS Geographic information systems
ana Region, Tunisia. Based on the regional characteristics, AHP Analytical hierarchy process
literature related to disposal sites and waste management, MCE Multi-criteria evaluation
local expert, data availability and assessments via ques- CR Consistency ratio
tionnaires, 15 constraints, and 5 factors were built in the WLC Weighted linear combination
hierarchical structure for landfill suitability by multi-cri-
teria evaluation. The factors are divided into environmental
and socio-economic groups. The methodology is used for
preliminary assessment of the 20-year most useful lifetime
Introduction
suitable landfilling sites by combining fuzzy set theory,
weighted linear combination (WLC) and analytic hierarchy
Municipal solid waste (MSW) management is one of the
process (AHP) in a GIS environment. The criteria stan-
major problems facing city planners worldwide. The
dardization is undertaken by application of different fuzzy
problem is especially severe in developing countries where
membership functions. The fuzzy membership functions
urbanization, poor planning, and the lack of adequate
shape and their control points are chosen through assess-
resources contribute to the poor state of solid waste man-
ment of expert opinion. The weightings of each selection
agement practices. The improper management of MSW
criterion are assigned depending on the relative importance
may cause a serious threat to human health and environ-
using the AHP methodology. The WLC approach is
ment including diseases transmission, fire hazards, odor
applied for alternative landfill sites prioritization. The
nuisance, atmospheric and water pollutions, aesthetic nui-
results of this study showed five potential candidate sites,
sance, and economic losses (Abu-Qdais and Abu-Qudais
which are generated when the environmental factors are
2000; Kontos et al. 2003). As sanitary landfilling is an
valued higher than socio-economic factors. These sites are
inevitable part of MSW management system (Tchobanog-
ranked in descending order using the ELECTRE III
lous et al. 1993), appropriate landfill site selection can
method. However, the final decision will require further
reduce the risk of contamination to the surrounding envi-
detailed geotechnical and hydrogeological analyses toward
ronment. However, landfills have become more difficult to
the protection of groundwater as well as surface water.
site because of their increasing cost, community opposition
to landfill siting in their vicinity, and more restrictive
Keywords MSW landfill  Multi-criteria evaluation 
regulations regarding the siting and operation of landfills
Analytical hierarchy process  ELECTRE III
(Mahini and Gholamalifard 2006).
Landfill site selection can generally be divided into two
A. Aydi  M. Zairi (&)  H. B. Dhia main steps: the identification of potential sites through
Laboratory of Water, Energy and Environment, preliminary screening and the evaluation of their suit-
National School of Engineers, PB 1173, 3038 Sfax, Tunisia ability (Chang et al. 2008). Geographic information
e-mail: moncef.zairi@enis.rnu.tn

123
1376 Environ Earth Sci (2013) 68:1375–1389

systems (GIS) approaches are popular for this kind of Siting methodology
preliminary studies due to their ability to manage large
volume of spatial data from a variety of sources (Kontos A substantial multi-disciplinary evaluation process with
et al. 2003). In the last few decades, GIS have been multiple sets of criteria is used to identify the best available
increasingly used as an important spatial decision support locations for a new MSW landfill in Ariana Region. The
system for evaluating suitable landfill locations (Done- final goal of the present work is to meet the regulatory
vska et al. 2011). A number of GIS methods and tech- requirements and minimize economic, environmental,
niques have been proposed to evaluate suitable landfill health, and social costs of landfill through its optimized site
location (Siddiqui et al. 1996; Al-Jarrah and Abu-Qdais selection. Hence, the siting process is divided into five
2006; Chang et al. 2008; Sumathi et al. 2008; Guiqin (a–e) sequential steps (Fig. 1):
et al. 2009; Sharifi et al. 2009; Nas et al. 2010; (a) The thematic maps of geologic, hydrological and
Khamehchiyan et al. 2011; Nazari et al. 2012). Some of hydrogeological characteristics, and demographic data
those techniques take advantage of GIS-based multi-cri- were represented as layers of georeferenced data and
teria evaluation (MCE) (Mahini and Gholamalifard 2006; constituted the GIS project using ArcGis environment.
Sener et al. 2006; Chang et al. 2008; Hasan et al. 2009; Then thematic maps were transformed on raster grid with
Sharifi et al. 2009; Khamehchiyan et al. 2011) and fuzzy cells of 100 9 100 m2 and exported to the IDRISI soft-
set theory (Chang et al. 2008; Hasan et al. 2009; ware. The multi-criteria factors are considered homoge-
Moeinaddini et al. 2010; Donevska et al. 2011; Gorsevski nous in each cell of the grid.
et al. 2012; Zelenović et al. 2012). In MCE, the weighted (b) For constraints, a Boolean criterion was used to
linear combination (WLC) is one of the most popular differentiate areas considered suitable for a waste disposal
methods because of its simplicity (Voogd 1983). Several
WLC-based approaches for landfill siting can be found in Data collection
the literature (Mahini and Gholamalifard 2006; Hasan
et al. 2009; Moeinaddini et al. 2010; Donevska et al.
Conversion to raster maps of
2011; Sahnoun et al. 2011). In the WLC procedure, uniform grid sizes
analytical hierarchy process (AHP) (Saaty 1980) is often
applied to elicit criteria weights and to better represent Selection of decision criteria
relationships between criteria and alternatives (Donevska
et al. 2011; Gorsevski et al. 2012). Another family of
MCE models uses the ‘‘outranking relations’’ to rank a set
Constraints Factors
of alternatives. A prominent role in this group is played
by the ELimination Et Choix Traduisant la REalité
(ELECTRE) method and its derivatives (Xiaoting and Reclassification Standardization of factors
Triantaphyllou 2008). It is a comprehensive evaluation
approach to rank a number of alternatives described by a Computation of weighted
Distance operation
number of criteria (Sahnoun et al. 2011). Today, the most factors:
widely used versions are known as ELECTRE III (Shabou Establish a pair-wise matrix
Calculate factors weighted
et al. 2009). The ELECTRE methods have proved useful Boolean map Calculating Consistency Ratio
decision aids in civil and environmental engineering
(Hobbs and Meier 2000). Applications include the
Multi-criteria evaluation
assessment of complex civil engineering projects, selec- - Suitabilty maps for each set of
tion of highway designs, energy-planning problem, site factors
- Create 3 scenarios: Weighted
selection for the disposal of nuclear waste, and solid linear combination of sets of
waste management (Hokkanen and Salminen 1997; factors

Xiaoting and Triantaphyllou 2008, Shabou et al. 2009;


Sahnoun et al. 2011). The ELECTRE III method was Localization of appropriate site
chosen from its family methods, because of the impreci-
sion and uncertainty of some available data. ELECTRE III method for
The purpose of this paper is to evaluate the suitability of the identification of priority
the study region to optimally site a landfill for MSW list for MSW landfill sites

Ariana using fuzzy set theory, AHP, and WLC in a GIS


environment. Then, ELECTRE method was applied to rank Fig. 1 The flow chart illustrating the methodology used in the MSW
the best alternative MSW landfill sites. landfill siting

123
Environ Earth Sci (2013) 68:1375–1389 1377

site from those considered unsuitable under any conditions. urban and the remaining 9 % being rural (Fig. 2). As per
The Boolean image was affected by the values of 1 for the the census of Tunisia, 2004, the total population of Ariana
first case (suitable site) or 0 in case of an inconvenient site. is about 422,246 inhabitants, living in 117,100 households,
(c) The factors are standardized to a continuous scale of and it is rising at an annual rate of 3.81 % (INS 2004). The
suitability from 0 (least suitable) to 10 (most suitable) in a average population density in this region is about 876.5
GIS environment by fuzzy membership functions, then inhabitants per km2. The daily average MSW generation
weighted and combined using the AHP method. for Ariana was about 0.6 kg per persons per day in the year
The AHP is a widely accepted decision-making method 2004 (Aouadhi and Ben Settela 2006).
based on pairwise comparisons. It is used to determine the The overall climate of Ariana is of Mediterranean type,
relative importance of a criterion in a specified decision- with 460 mm of annual precipitation, an annual tempera-
making problem. In the construction of a pairwise com- ture average of 19 C (max: 27 C in August and min:
parison matrix, the factors are rated one against other by 12 C in January).
assigning a relative dominant value between 1 and 9 to the The municipal authorities and the National Agency for
intersecting cell (Saaty 1977). The consistency ratio (CR) Waste Management are responsible for the management of
indicates the overall consistency of the pairwise compari- the MSW generated in the city. Until 1999, all municipal
son matrix. The CR should have a value of less than 0.1 to wastes in Ariana were buried at unapproved open dumps
indicate a consistency method (Kontos et al. 2005). such as Nahli and Raoued (Fig. 2).
(d) The determined weights are used with their corre-
sponding individual standardized criteria as input for the
WLC aggregation method. A total score is then obtained as
a sum of the products of each criteria and its weight
(Eastman 2001):
X
n
S¼ w i xi
i¼1

where S is the suitability index; w1, w2, …, wn are the


weights of the criteria constrained to sum to 1 and x1, x2,
…, xn are the standardized scores of the criteria i. As the
sum of the weights is constrained to 1, the final combined
estimate is presented on the same scale.
In this study, suitability maps for each set of factors
were combined to create three scenarios for determination
of the best suitable sites.
(e) The ELECTRE III method was applied to rank the
most favored site. This model is a highly developed multi-
criteria analysis method, based on the definition of
concordance and the discordance matrices (Vallée et
Zielniewicz 1994). Three parameters are introduced in this
method which are: indifference, preference, and veto
(Ulubeyli and Kazaz 2009). Based on this information, an
outranking for each alternative was determined, consider-
ing the preference of decision makers.
In this study, IDRISI (ver. 32) software was used to
buffer the constraints, calculate the weights of criteria,
standardize the factors by fuzzy functions, and combine the
constraints and factors by MCE procedure.

Study area

The Ariana, located in the far north-eastern part of Tunisia,


covers a total area of 482 km2, with 91 % of its area being Fig. 2 Location of the study area

123
1378 Environ Earth Sci (2013) 68:1375–1389

The Tunisian national strategic plan (PRONAGDES) criteria that limit analysis to particular geographic regions.
started in 1999 imposed dump sites closure and new san- In contrast, factors define some degree of suitability for all
itary landfill construction. Hence, the Raoued and Nahli geographic regions; they permit an evaluation of alterna-
dumping sites were closed in January 2000. It is worth tives issued from the previous step.
noting that, National Agency for Waste Management has The constraints criteria used in this study include soil
supervised landfill siting and construction projects at a permeability, elevation, slope, coastal zone, soil charac-
national scale. The actual MSW generated in Ariana is teristics, depth to impermeable layer, depth to ground water
transported to the sanitary landfill of Jebel Chakir, located table, water supply, wetlands, distance from rivers, prox-
15 km to the south of Ariana City. The Jebel Chakir imity to irrigational canals, land use, distance from roads,
landfill will be soon closed. The municipalities consider the protected areas, and residential areas. The factors include
site as too far from waste production centers in Ariana land use, olfactory and sonorous impacts, proximity to an
Region, which increase the transportation cost and need area of dense population, distance from waste production
additional investments in the infrastructure and subse- centers, and distance from road.
quently intensify their financial problems.
Constraints: buffer maps

Data collection According to Dikshit et al. (2000), a landfill site must be


situated at a far distance away from biophysical elements
GIS maps at 1:50,000 scale of land use, slope, elevation, such as water, wetlands, habitats, and wells to reduce the
road network, urban area extent, waste production centers, contamination risk. Thus, buffer maps indicating regions
permanent and temporary rivers, water supply, wetland, encompassing appropriate areas from these elements are
ground water wells, irrigational canals, soil types, protected generated to act as areas of constraints where a disposal site
area, and coastal zone limits maps were obtained from the is unsuitable. Literature review, and a set of questionnaires
Regional Commission for Agricultural Development of sent to the town administrators, developers, environmen-
Ariana. talists, economists, and geologists were analyzed and per-
Data on groundwater levels and impermeable layer mitted to identify the most important constraints as: soil
properties (depth and thickness) were obtained through permeability, elevation, slope, coastal zone, soil charac-
field investigations of piezometers and via geostatistical teristic, depth to impermeable layer, depth to ground water
analysis using the Kriging Interpolator 3.2 SA tool of table, water supply, wetland, distance from river, proximity
ArcView. to irrigation canal, land use, distance from road, protected
The population data were obtained from the National area, and residential area.
Institute of Statistics (NIS). The population density
required the building of a geographical database linked to Soil permeability
an alphanumeric database that considered spatial variabil-
ity in the 2004 year population and base maps compiled at According to the characteristics of local geology of the
1:50,000 scale. region, this criterion categorizes the area into three distinct
classes: soils having high permeability rate (sandy and
sandy-loam soils) are considered unsuitable for being used
Criteria description and application as a landfill while soils with medium and relatively low
permeability rate (sandy clay) and very low permeability
Criteria were selected according to the regional charac- (clay and clay loam soils) are considered fairly suitable and
teristics, literature related to disposal sites and waste optimal for landfill siting, respectively (Fig. 3).
management (Al-Jarrah and Abu-Qdais 2006; Mahini and
Gholamalifard; Chang et al. 2008; Sumathi et al. 2008; Elevation
Guiqin et al. 2009; Hasan et al. 2009; Sharifi et al. 2009;
Moeinaddini et al. 2010; Nas et al. 2010), local expert, data Elevation was considered as high areas are inaccessible and
availability, and assessments via questionnaires. transportation difficulties. Thus, areas of more than 200 m
In the present study, 15 influential criteria are consid- elevation were considered unsuitable (Fig. 3).
ered in the computation process and distinguished as con-
straints and factors. These attributes are clustered in five Slope
main groups including geologic criteria, hydrologic and
hydrogeologic criteria, environmental criteria, social cri- The slope is a crucial factor when there are risks of
teria, and economic criteria. Constraints are the Boolean leachate pollution since higher slopes would increase

123
Environ Earth Sci (2013) 68:1375–1389 1379

Soil permeability Elevation Slope

Depth to impermeable layer


Coastal Zone Soil characteristic

Depth to groundwater Water supply Wetlands

Distance from rivers Proximity to irrigation canals Land use

Distance from roads Protected area Residential area

Fig. 3 Boolean images of constrained maps

runoff of pollutants from the landfill (Lin and Kao 1999; study, a 5 % slope was considered as the maximum slope
Sharifi et al. 2009). Steeper areas would not be economi- value. Areas with higher slope values were excluded from
cally appropriate because they would require excessive further consideration, as not appropriate for landfill siting
excavation (Gemitzi et al. 2007; Guiqin et al. 2009). In this (Fig. 3).

123
1380 Environ Earth Sci (2013) 68:1375–1389

Coastal zone Distance from rivers

The landfill disposal site must be far from costal line. According to the Tunisian legislation, the disposal of solid or
Therefore, a buffer of 3 km is applied around the coastal liquid waste in any water body such as sea, lakes, and rivers is
shoreline (Fig. 3). prohibited. However, some studies have suggested different
buffer distances from stream and rivers based on the size of
Soil characteristics the watershed, such as buffer of 100 m (Sumathi et al. 2008),
180 m (Zeiss and Lefsrud 1995), 200 m (Mahini and Gho-
Clay availability is one of the best parameters for landfill lamalifard 2006; Akbari et al. 2008; Shabou et al. 2009),
siting. Leachate migration from the landfill could be a 500 m (Kontos et al. 2003; Gemitzi et al. 2007; Guiqin et al.
potential source of surface and groundwater contamina- 2009; Hasan et al. 2009), 800 m (Siddiqui et al. 1996),
tions. In order to avoid these risks for soil and groundwater, 1,000 m (Chang et al. 2008; Sharifi et al. 2009) and 2–3 km
the selected site must be implemented on an impermeable (Lin and Kao 1999). A buffer distance of 200 m for perma-
clayey layer, which is naturally present with a minimum nent and temporary rivers was considered unsuitable for
thickness of 5 m (Fig. 3). allocating landfills in this study (Fig. 3).
Proximity to irrigation canals
Depth to impermeable layer
The landfill disposal site must be far away from irrigation
A depth of the top clayey layer till 9 m was considered canals—a 200-m buffer zone of protected areas around
(Fig. 3). irrigational canals was assigned (Fig. 3).
Land use
Depth to groundwater
Land use is important for resolving public conflicts over
Considering ground water contamination, depth of water
the acceptance of unwanted facility siting. In this study,
table must be taken into consideration as a highly effective
agricultural area, urban area, rivers, wetlands, and pro-
factor. The groundwater table level should be sufficiently
tected area were considered unsuitable for landfill siting.
low and must be obligatory under the impermeable clayey
Areas with low economic advantages and unused lands are
layer. In this study, the minimal depth of the groundwater
considered suitable for MSW landfill site (Fig. 3).
was 14 m (9 ? 5 m) (Fig. 3).
Distance from roads
Water supply (reservoirs, wells, boreholes, springs)
Landfill location must be close to roads network in order to
The waste disposal area should be away from drinking and facilitate transportation and consequently reduce relative
irrigation water sources (wells); otherwise, it can have costs (Zeiss and Lefsrud 1995; Chang et al. 2008). The
irretrievable human and environmental effects. Therefore, Tunisian law does not specify minimum distances of san-
wells must be protected from the runoff and leaching of the itary landfills from roads. The road network in the region
landfill (Akbari et al. 2008). According to the EU direc- consists of highways, main roads, secondary roads, and
tives, a landfill should not be close to any source of water. tracks. Considering the huge cost of transportation and
However, some researchers have suggested a distance of up environmental concern (visual impact, odor, excessive
to 500 m away from a freshwater body (Kontos et al. 2003; noise, traffic, and scavengers), a 200-m buffer for roads
Akbari et al. 2008; Sharifi et al. 2009, Nas et al. 2010). In was finally fixed for this study (Fig. 3).
this study, a 2-km buffer was placed using the function in
IDRISI software, which will be used to generate the buffer Protected area
around all water supplies (Fig. 3).
Since a landfill site should not be placed on a site close to
Wetlands protected area, a buffer zone of 300 m from Ennahli
National park was used (Fig. 3).
The waste disposal areas should not be in the vicinity of
wetland (lakes, or swamps). For this reason, a 1-km buffer Residential area
was placed using the function in GIS software, which was
used to generate the buffer around all wetland areas (lakes The landfill should be situated at a significant distance
or swamps) (Fig. 3). away from urban residential areas due to public concerns,

123
Environ Earth Sci (2013) 68:1375–1389 1381

such as aesthetics, odor, noise, decrease in property value, roads were considered as components of socio-economic
and health concerns, which may avoid contamination of factors.
freshwater aquifers through leaching (Tagaris et al. 2003, These factors are standardized in a range from 0 to 10,
Zeiss and Lefsrud 1995; Nagar and Mizra 2002). A buffer relatively to their relevance to site conditions.
distance of 2 km was chosen for the study area. In this
case, suitable areas will have to be at least 2 km from the Standardization of factors
residential zone (Fig. 3).
The potential area is obtained through the overlay of The use of fuzzy sets within a multi-criteria, GIS-based
layers relative to the 15 previous constraint criteria environmental management and decision making has been
(Fig. 4). Figure 4 shows the 44 candidate sites subjected to demonstrated to have a good effect (Shabou et al. 2009).
the advanced assessment in the second-stage analysis. The Hence, the factors were standardized to a byte-level range
areas shown in white were excluded for consideration as a of 0–10, providing the maximum differentiation possible
landfill disposal site. (Eastman 2001). Zero is assigned to the least suitable areas
and 10 to the most suitable ones, transforming the different
Factors for landfill site selection measurement units of the factor images, which served as
GIS map layers, into comparable values using fuzzy
The factor is defined as a criterion enhancing or detracting membership functions. Four fuzzy set membership func-
from the relative suitability of an area for landfill siting tions are provided in IDRISI: sigmoidal, J-shaped, linear,
(Eastman et al. 1993). and user-defined (Eastman et al. 1993). The selection of
In this study, factors were categorized into environ- appropriate fuzzy functions and their control points are
mental and socio-economic groups. The environmental based on the local expert knowledge and literature review
factor consisted of land use and olfactory and sonorous (Table 1).
impacts.
The proximity to an area of dense population, the dis- Environmental factors
tance from waste production centers and the distance from
The environmental factor comprises two sub-factors
namely land use and olfactory and sonorous impacts.
Rescaling categorical data of land use requires giving
a rating to each category based on the results of
questionnaire investigations of experts (agronomist,
environmentalists…).
The land use was categorized on the suitability basis on
a numeric scale containing 10 points (Table 1). Urban
areas, protected area, wetlands, and water were excluded
from consideration as they are least suitable with a score of
0. The most suitable land use is naked soil (score of 10)
while the vine and mariachi culture is assigned with the
lowest suitability (Fig. 5).
For the olfactory and sonorous impacts factors, a
monotonically increasing linear fuzzy membership func-
tion was used. The first control point (a = 200 m) indi-
cates the least suitable distance for siting a landfill while
the second control point (b = 3,000 m) and beyond indi-
cates the most suitable distance for siting a landfill (Fig. 5).

Socio-economic factors

The socio-economic factors comprises three sub-factors


namely proximity to an area of dense population, distance
from waste production centers and distance from road.
The proximity to an area of dense population considers
Fig. 4 Potential area for landfill siting obtained from constraint the costs associated with economic distances from the
criteria source of the waste that are adjusted based on the population

123
1382 Environ Earth Sci (2013) 68:1375–1389

Table 1 Fuzzy set memberships and membership functions with Weighting and aggregation of factors
control points used for MSW landfill site selection
Factors Sub-factors Standardisation of factors After standardization, criteria weightings were performed.
Pairwise comparison method was used to assign weights
Control point Fuzzy
function/ and establish importance of environmental and socio-eco-
membership nomic criteria using experience of local expert knowledge
and literature review.
Environmental Land use (no
Table 2 shows the pairwise comparison matrix of rela-
units)
tive importance and weights associated to the environ-
Urban areas 0
mental factors. The land use is valued higher than the
Protected area 0
olfactory and sonorous impacts. The CR is 0 indicating a
Wetlands 0
good consistency of the judgments used for the
Water 0
comparison.
Vine 2
On the other hand, the Table 3 shows the pairwise
Mariachi 2
culture
comparison matrix of relative importance and weights
Cereals 3
associated to the socio-economic factors. The highest
weights assigned to the distance from waste production
Olive trees 3
centers were considered the most important factors. The
Forager culture 4
distance from roads is valued higher than the proximity to
Course 6
an area of dense population. The CR is 0.03 signifying a
Naked soil 10
good consistency of the judgments used for the
Olfactory and 200 and 3,000 m Linear,
sonorous increasing comparison.
impacts Finally, the WLC is used to aggregate suitability maps
Socio- Proximity to an 200 and 1,000 Sigmoidal, created by each factors. The possible suitability maps
economic area of dense inhabitants/km2 decreasing derived by this methodology are discussed in the section
population below.
Distance from 2,000 and J-shaped,
waste 5,000 m decreasing
production
centers Results and discussion
Distance from 200 and 3,000 m J-shaped,
roads decreasing WLC is applied to compute and display two decision
alternatives for landfill suitability associated with the
environmental and socio-economic factors. The two alter-
density. Thus, close proximity to the waste source decreases natives were generated by the overlay of the final constraint
the lengths of hauling and promotes better long-term eco- maps (Fig. 4) and the assigned weight of the standardized
nomic operational costs (Baban and Flannagan 1998). The criteria (Fig. 5) for the environmental and socio-economic
population density map was standardized by sigmoidal factors (Fig. 6).
monotonically decreasing fuzzy function controlled by two
points (c = 200 hab km-2, d = 1,000 hab km-2) where Scenario design
the density population less than 200 hab km-2 are the most
suitable (full membership) and more than 1,000 hab km-2 The WLC was used to create the final landfill suitability
are not suitable (full-non-membership) (Fig. 5). maps using different weights applied to the environment
The distance from waste production was standardized by and socio-economic factors (Hasan et al. 2009; Shabou
J-shaped monotonically decreasing fuzzy function, et al. 2009; Donevska et al. 2011; Sahnoun et al. 2011;
assigning highest suitability to the distances less than Gorsevski et al. 2012). The WLC alternatives are applied
2,000 m and lowest suitability to the distances more than for three scenarios generated by different weights applied
5,000 m. to the factors. For instance, the scenario (1) assigns a same
The distance from roads was standardized by using a weight of 0.5 to the environmental and the socio-economic
J-shaped monotonically decreasing fuzzy function con- factors. The scenario (2) assigns a weight of 0.75 to the
trolled by two points (c = 200 m, d = 3,000 m), where environmental factors and 0.25 to the socio-economic ones.
distances less than 200 m are the most suitable (full For the scenario (3), weights are of 0.25 and 0.75,
membership) and more than 3,000 m are not suitable (full respectively, for environmental and socio-economic factors
non-membership) (Fig. 5). (Table 4).

123
Environ Earth Sci (2013) 68:1375–1389 1383

Fig. 5 Standardized maps for


the environmental and socio-
economic factors

Proximity to dense population


Land use Olfactory and sonorous area
impacts

Distance from waste production Distance from roads


centers

Table 2 Pairwise comparison matrix for assessing the weights of environmental factors
Land use Olfactory and sonorous impacts Eigenvector Weight

Land use 1 5 2.23 2.23/(2.23 ? 0.45) = 0.83


Olfactory and sonorous impacts 1/5 1 0.45 0.45/(2.23 ? 0.45) = 0.17
kmax = 2, CI = 0.00, CR = 0.00 (consistency is acceptable)

Using an equal interval classification method, landfill It is important to note that the most suitable sites are
suitability values of the Ariana City were reclassified into concentrated to the west of the study area. This area con-
three classes: class 1 (0–3), class 2 (3–7), class 3 (7–10). tains lowly permeable soils and poor land use without
This method creates an easy to understand legend and residential inhabitants. Thus, this area can be fairly con-
works best with continuously distributed data (Fig. 7). sidered for landfill construction. The location of these sites,

123
1384 Environ Earth Sci (2013) 68:1375–1389

Table 3 Pairwise comparison matrix for assessing the weights of socio-economic factors
Proximity Distance Distance Eigenvector Weight
to an area of dense from waste from road
population production centers

Proximity to an area of dense population 1 1/5 1/3 0.44 0.11


Distance from waste production centers 5 1 3 2.44 0.63
Distance from road 3 1/3 1 1 0.26
kmax = 3.038, CI = 0.019, RI = 0.58, CR = 0.03  0.1 (consistency is acceptable)

Fig. 6 Multi-criteria
aggregation of factors by WLC

(a) Environmental factors (b) Socio-economic factors

Table 4 Pairwise comparison matrix for assessing the weights of ranged on class 3, was almost the same in the three con-
factors concerning three scenarios sidered scenarios.
Decision factor Environmental Socio- Eigenvector Weight
economic
Sizing procedure
Scenario (1) (same weighting for all sets of factors)
Environmental 1 1 1 0.5
To get suitable landfill surface area, the formula given by
Socio-economic 1 1 1 0.5
Jaramillo (2003) was adopted considering the following
Scenario (2) (environmental [ socio-economic)
assumptions: the average daily MSW production per capita
Environmental 1 3 1.73 0.75
in Ariana in 2004 is 0.6 kg/capita/day and MSW generation
Socio-economic 1/3 1 0.57 0.25
growth is set to 2 % per year (Aouadhi and Ben Settela
Scenario (3) (socio-economic [ environmental)
2006). The MSW collection coverage in urban areas is set to
Environmental 1 1/3 0.57 0.25
95 % and to 90 % in rural area (World Bank 2007). The
Socio-economic 3 1 1.73 0.75
estimated MSW quantity to be landfilled (taking into account

123
Environ Earth Sci (2013) 68:1375–1389 1385

Scenario (1) (environemental= socio- Scenario (2) (environemental >soci-


economic) economic)

Scenario (3) (socio-economic > environemental)

Fig. 7 Possible suitability maps derived by both factors using different weights: scenario (1) 0.5 and 0.5; scenario (2) 0.75 and 0.25; and
scenario (3) 0.25 and 0.75 for the environmental and the socio-economic factors, respectively

5 % recycling and composting of 0.5 %) (Saidi et al. 2008) is compacted volume of MSW and the additional area
4.3 million tons for a 20-year operation period assuming a required for operation facilities, penetration roads, border
3.81 % of the population increase per year. setback areas, control building and sanitary facilities, etc.,
In Tunisia, the waste thickness is almost set to 20 m and is set to 15 %.
the average waste density after compaction is 950 kg/m3. Thus, the required area is set to 28 ha, which will be
The amount of cover material is set to 15 % of the filled in the 20 years lifespan of the landfill (Table 5). It

123
1386 Environ Earth Sci (2013) 68:1375–1389

Table 5 Calculation of required landfill area Table 6 Family and criteria


No. Description Value Family Criteria Direction
3
Q1 Total waste to be filled (10 kg) 4,251,783 Social Cr 1: problems with local communities Min
Q2 Density of compacted MSW (kg/m3) 950 (scores 1–10)
Q3 Landfill volume required (m3)(Q1/Q2) 4,475,561 Cr 2: distance from residential area (km) Max
Q4 Cover rate 10 % Geometric Cr 3: geometric form of the site Max
Q5 Total landfill volume (waste ? cover) (m )3
4,923,117 (scores 1–10)
Q6 Landfill depth (m) 20 Economic Cr 4: availability and cost of cover soil Min
2
($/m3)
Q7 Required landfill net area (m ) (Q6/Q6) 246,156
Cr 5: possibility of extension (ha) Max
Q8 15 % Extra area required for operation facilities 36,923
(m2) (20 % Q7)
Q9 Total landfill area (m2) (Q7 ? Q8) 283,079
Table 7 Synthesis performances matrix based on the environmental
scenario

should be mentioned that we selected the scenario (2) Site Cr 1 Cr 2 Cr 3 Cr 4 Cr 5


which assigned the highest weight to the environmental
1 10 2.53 10 2.65 98.15
factors. Then, we grouped similar and adjoining pixels to
2 4 2.10 6 3.05 25.88
keep sites having an area greater than 28 ha, roughly 28
3 1 4.85 1 2.85 6.79
cells (100 9 100 m2). Thus, five appropriate sites for
4 5 4.05 4 2.35 4.05
MSW landfill were identified.
5 7 6.07 5 3.50 2.26

Outranking of candidate sites using ELECTRE III


method
Table 8 Weights, thresholds, and directions of criteria

The selected sites, according to the environmental sce- Site Cr 1 Cr 2 Cr 3 Cr 4 Cr 5


nario, are classified in a predefined category using the Indifference 2 350 1 0.9 4.05
ELECTRE III method. The evaluation criteria used are Preference 7 950 5 1.07 9.50
the problems with local communities, the distance from
Veto 0 0 0 0 0
residential area, the geometric form of the site, the
Weight 12 18.5 22.5 23.15 23.85
availability and cost of cover soil and the possibility of
Direction Min Max Max Max Max
extension (Table 6).
The five possible sites were then assigned a numerical
value according to each criterion (Table 7).
For the geometric form, the square or rectangular form The outranking of these candidate sites using ELECTRE
of the candidate site (with area of 28 ha) were assigned a III method according to both distillations were divided into
grade of 10 (best case) (length B 1.5 width). However, the ‘‘appropriate’’, ‘‘fairly appropriate’’ and ‘‘inappropriate’’
elongated form was taken as the worst case (grade of 1). An landfill areas (Fig. 8). The site number 1 is displayed as the
intermediate grade was used (1–10) for other form. best for a possible siting of the MSW landfill. The selection
The acceptance level of the surrounding communities of the final MSW site, however, requires further geotech-
must at last be satisfactory. A grade of 10 was given to the nical and hydrogeological analyses toward the protection
best candidate (no problems with local communities) and 1 of groundwater as well as surface water.
to the worst.
The possibility for landfill extension was considered by
assuming an area greater than the required landfill site size Conclusion
(28 ha). It can help the waste management authorities to
expand the capacity of the selected landfill and conse- Landfill site selection is a complex and time-consuming
quently extend their useful lifetime. process, which requires evaluation of several factors where
Once the criteria are estimated, we determined weights many different attributes are taken into account. The
to express the relative importance granted to every crite- present study shows a methodology for MSW landfill
rion. In this context, we solicited the appreciations from selection based on the application of fuzzy set theory,
emanating appreciative experts in the waste management AHP, and WLC, as applied to the Ariana Region in
field (Table 8). Tunisia. Fuzzy set theory enables criteria standardization

123
Environ Earth Sci (2013) 68:1375–1389 1387

Fig. 8 Final landfill suitability


map showing the selected sites
based on the environmental
scenario

through the application of different fuzzy membership in the infrastructure. Initially, landfill site selection criteria
functions. Expert knowledge and literature review are used were determined depending on the regional characteristics,
for the selection of the appropriate fuzzy functions and literature related to disposal sites and waste management,
their control points. GIS is used for analyzing and dis- local expert, data availability, and assessments via
playing geospatial data. AHP methodology is applied for questionnaire.
establishing the relative importance of selection criteria The proposed method considers 15 constraints and 5
and weights generation, providing an objective weight factors. These criteria are clustered in five main groups
assessment. The landfill suitability maps are created by including geology, hydrology and hydrogeology, envi-
WLC. Finally, ELECTRE III method was applied to rank ronment, societal, and economy. The constraints criteria
the proposed candidate sites and identified the best option. used in this study include soil permeability, elevation,
The Ariana Region considered in this study is too far slope, coastal zone, soil characteristics, depth to imper-
from the actual operating Jebel Chakir landfill, inducing meable layer, depth to ground water table, water supply,
high MSW transportation cost and additional investments wetlands, distance from rivers, proximity to irrigational

123
1388 Environ Earth Sci (2013) 68:1375–1389

canals, land use, distance from roads, protected areas, and Gemitzi A, Tsihrintzis VA, Christou O, Petalas C (2007) Use of GIS
residential areas. The factors were categorized into envi- in siting stabilization pond facilities for domestic wastewater
treatment. J Environ Manag 82(2):155–166
ronmental and socio-economic groups. The environmental Gorsevski PV, Donevska KR, Mitrovski C, Frizado JP (2012)
factor consisted of land use and olfactory and sonorous Integrating multi-criteria evaluation techniques with geographic
impacts. The proximity to an area of dense population, information systems for landfill site selection: a case study using
the distance from waste production centers and the dis- ordered weighted average. Waste Manag 32(2):287–296
Guiqin W, Li Q, Guoxue L, Lijun C (2009) Landfill site selection
tance from roads were considered as the components of using spatial information technologies and AHP: a case study in
socio-economic factor. Beijing, China. J Environ Manag 90:2414–2421
The combination of different sets of factors according Hasan MR, Tetsuo K, Islam AS (2009) Landfill demand and
to the environmental scenario indicated five possible allocation for municipal solid waste disposal in Dhaka city—
an assessment in a GIS environment. J Civil Eng 37(2):133–149
disposal sites. This area contained lowly permeable soils Hobbs BF, Meier P (2000) Energy decisions and the environment: a
and poor land use without residential inhabitants. The guide to the use of multi-criteria methods. Kluwer, Boston (MA)
ELECTRE III method was applied to rank the proposed Hokkanen J, Salminen P (1997) Choosing a solid waste management
candidate sites and identified the best option. These sites system using multi-criteria decision analysis. Eur J Oper Res
98:19–36
generally satisfy the minimum requirements of a landfill INS: Institut National de la Statistique (2004) Données générales sur
site. However, the selection of the final MSW site la population: données démographiques et sociales. http://
requires further detailed geotechnical and hydrogeologi- www.ins.nat.tn
cal analyses toward the protection of groundwater as Jaramillo J (2003) Guidelines for the design, construction and
operation of manual sanitary landfills. A solution for the final
well as surface water. disposal of municipal solid wastes in small communities.
Universidad de Antioquia, Colombia
Acknowledgments The authors acknowledge the technical support Khamehchiyan M, Nikoudel MR, Boroumandi M (2011) Identifica-
from the Commissariat Régional de Développement Agricole de tion of hazardous waste landfill site: a case study from Zanjan
l’Ariana for the data reports cited and used in this analysis. province, Iran. Environ Earth Sci 64(7):1763–1776
Kontos TD, Komilis DP, Halvadakis CP (2003) Siting MSW landfills
on Lesvos island with a GIS-based methodology. Waste Manag
Res 21(3):262–278
References Kontos TD, Komilis DP, Halvadakis CP (2005) Siting MSW landfills
with a spatial multiple criteria analysis methodology. Waste
Abu-Qdais M, Abu-Qudais HA (2000) Energy content of municipal Manag (Oxford) 25:818–832
solid waste in Jordan and its potential utilization. Energy Lin HY, Kao JJ (1999) Enhanced spatial model for landfill siting
Convers Manag 41(9):983–991 analysis. J Environ Eng 125(9):845–851
Akbari V, Rajabi MA, Chavoshi SH, Shams R (2008) Landfill site Mahini AS, Gholamalifard M (2006) Siting MSW landfills with a
selection by combining GIS and Fuzzy multi-criteria decisions weighted linear combination methodology in a GIS environment.
analysis, case study: Bandar Abbs, Iran. World Appl Sci J 3(1): Int J Environ Sci Technol 3(4):435–445
39–47 Moeinaddini M, Khorasani N, Danehkar A, Darvishsefat AA, Zienalyan
Al-Jarrah O, Abu-Qdais H (2006) Municipal solid waste landfill siting M (2010) Siting MSW landfill using weighted linear combination
using intelligent system. Waste Manag 26:299–306 and analytical hierarchy process (AHP) methodology in GIS
Aouadhi H, Ben Settela M (2006) Modes et Plans de Gestion des environment (case study: Karaj). Waste Manag 30(5):912–920
Déchets Ménagers dans le Grand Tunis. Projet de Fin d’étude. Nagar BB, Mizra UK (2002) Hydrogeological environmental assess-
Institut Supérieur des Technologies de l’Environnement de ment of sanitary landfill project at Jammu City, India. Green J
l’Urbanisme & du Bâtiment 17(8):223–245
Baban SMJ, Flannagan J (1998) Developing and implementing GIS- Nas B, Cay T, Iscan F, Berktay A (2010) Selection of MSW landfill
assisted constraints criteria for planning landfill sites in the UK. site for Konya, Turkey using GIS and multi-criteria evaluation.
Planning Pract Res 13:139–151 Environ Monit Assess 160(1–4):491–500
Chang NB, Parvathinathan G, Breedenc JB (2008) Combining GIS Nazari A, Salarirad MM, Bazzari Aghajani A (2012) Landfill site
with fuzzy multi-criteria decision-making for landfill siting in a selection by decision-making tools based on fuzzy multi-
fast-growing urban region. J Environ Manag 87:139–153 attribute. Environ Earth Sci 65(6):1631–1642
Dikshit AK, Padmavathi T, Das RK (2000) Locating potential landfill Saaty TL (1977) A scaling method for priorities in hierarchical
sites using geographic information systems. J Environ Syst structures. J Math Psychol 15(3):234–281
28:43–54 Saaty TL (1980) The analytical hierarchy process. McGraw Hill, New
Donevska KR, Gorsevski PV, Jovanovski M, Pesevski I (2011) York
Regional non-hazardous landfill site selection by integrating Sahnoun H, Serbaji MM, Karray B, Medhioub K (2011) GIS and
fuzzy logic, AHP and geographic information systems. Environ multi-criteria analysis to select potential sites of agro-industrial
Earth Sci. doi:10.1007/s12665-011-1485-y complex. Environ Earth Sci. doi:10.1007/s12665-011-1471-4
Eastman RJ (2001) Guide to GIS and image processing, vol 2. Clark Saidi N, Cherif M, Jedidi N, Mahrouk M, Fumio M, Boudabous A,
University, USA Hassen A (2008) Evolution of biochemical parameters during
Eastman JR, Kyem PAK, Toledano J (1993) A procedure for composting of various wastes compost. Am J Environ Sci
multiobjective decision making in GIS under conditions of 4(4):332–341
conflicting objectives. In: Proceedings of European conference Sener B, Süzen ML, Doyuran V (2006) Landfill site selection by
on geographical information systems, EGIS’93. EGIS Founda- using geographic information systems. Environ Geol
tion, Utrecht, pp 438–448 49:376–388

123
Environ Earth Sci (2013) 68:1375–1389 1389

Shabou R, Zairi M, Kallel A, Neji J, Ben Dhia H (2009) GIS and Ulubeyli S, Kazaz A (2009) A multiple criteria decision-making
multi-criteria analysis for OMW disposal site choice. Waste approach to the selection of concrete pumps. J Civ Eng Manag
Resour Manag 162(2):99–108 15(4):369–376
Sharifi M, Hadidi M, Vessali E, Mosstafakhani P, Taheri K, Shahoie Vallée D, Zielniewicz P (1994) ELECTRE III-IV, version 3.x—
S, Khodamoradpour M (2009) Integrating multi-criteria decision Aspects méthodologiques, Université de Paris-Dauphine, Docu-
analysis for a GIS-based hazardous waste landfill sitting in ment du LAMSADE no 85
Kurdistan Province, Western Iran. Waste Manag 29(10):2740– Voogd H (1983) Multi-criteria evaluations for urban and regional
2758 planning. Pion, London
Siddiqui M, Everett JM, Vieux BE (1996) Landfill siting using World Bank (2007) Sustainable municipal solid waste management
Geographic Information Systems: a demonstration. J Environ project to the republic of Tunisia
Eng 122(6):515–523 Xiaoting W, Triantaphyllou E (2008) Ranking irregularities when
Sumathi V, Natesan U, Sarkar C (2008) GIS-based approach for evaluating alternatives by using some ELECTRE methods.
optimized siting of municipal solid waste landfill. Waste Manag Omega 36:45–63
28(11):2146–2160 Zeiss C, Lefsrud L (1995) Analytical framework for facility waste
Tagaris E, Sotiropolou RE, Pilinis C, Halvadakis CP (2003) A siting. J Urban Planning Dev 121(4):115–145, ASCE
methodology to estimate odors around landfill sites: the use of Zelenović VT, Srdjević Z, Bajčetić R, Vojinović MM (2012) GIS and
methane as an odor index and its utility in landfill siting. J Air the analytic hierarchy process for regional landfill site selection
Waste Manag Assoc 53(5):629–634 in transitional countries: a case study from Serbia. Environ
Tchobanoglous G, Theisen H, Vigil SA (1993) Integrated solid waste Manag 49:445–458
management: engineering principles and management issues.
McGraw-Hill, New York

123

You might also like