Decision-making for sustainable location of a cement plant in the state of Florida

Abstract

Site selection is a typical strategic decision for many industries that deals with specifying the most appropriate location for a facility. In the context of sustainable development, site selection decisions need to be improved by adopting environmental, economic and social requirements. This study aims to frame sustainable location decisions by conducting a case study of siting a new cement plant in the state of Florida. As a part of the study, a wide range of technical and sustainability criteria has developed. These criteria can help decision-makers in the cement industry evaluate the selection of a location with the requirements of sustainable development. The sustainability characteristics of alternative sites in the state of Florida are evaluated based on the criteria to identify the most appropriate ones. The analytical hierarchy processand geographic information system techniques are utilized to weight the criteria and evaluate the characteristics of site candidates.

Keywords::

• site selection
• facility location
• cement plant
• sustainability
• AHP
• GIS
• state of Florida

Introduction

Cement is the main component of concrete, which is the most consumed material after water. As a result of this high level of consumption, the cement industry is one of the major industries affecting the built environment. In 2012, the USA produced and imported 79.1 million metric tons of Portland and Blended cement (van Oss and Kraft 2013). Although, due to the economic crisis, cement consumption has been significantly reduced (PCA 2013; van Oss and Kraft 2012, 2013), this number will increase in the coming years due to the economic recovery of the USA. It is noteworthy that, in 2012, about 6.3 million tons of cement were imported into the USA. According to Holcim (US) Inc. (2003), 5.8 million tons of imported cement can result in $1.5 billion lost income for cement manufacturers of the USA. Therefore, filling this gap is crucial for job creation and the economic recovery of the USA.

However, the cement industry is one of the most controversial industries due to its significant environmental impacts. As Klee and Coles (2004, 114) stated, “[t]he industry is faced with increasingly strong legislative and stakeholder pressure, specifically regarding how to meet environmental and corporate social responsibilities.”

The major environmental issues of the cement industry are energy consumption and air emissions (Marlowe and Mansfield 2002). A typical cement manufacturing plant generates nearly 0.85 kg of CO₂ for every kilogram of cement (WBSCD/CSI 2009). In 2008, the cement industry accounted for approximately 7% of global CO₂ emissions related to primary energy consumption. The total CO₂ emissions related to primary energy consumption, in 2008, was about 29Gt, a quarter of which, were attributed to industrial activities and fuel transformation. Within industrial activities and fuel transformation, 27% of the emissions were due to cement production (IEA and UNIDO 2011).

Therefore, the environmental impacts of the cement industry are enormous and, as a result, the cement manufacturing industry is under increasing scrutiny.

The location decision of a cement plant is very challenging and can be totally affected by local communities and other involved stakeholders. Although some of the impacts of the cement industry are not location-specific, such as CO₂ emissions, they are inevitable inputs for siting decisions. Based on the public-hearing report of the US Fish and Wildlife Service (2008), some of the concerns of local communities regarding the location of cement plants included: future air quality, potential long-term impacts on the local public health, endangered species, wildlife, aquifers, archaeological and historical resources, natural conservation, tourism activity, infrastructure, growth of the community, economic impacts and employment opportunities of plants.

Therefore, to achieve the global objectives of sustainable development, the cement industry should fundamentally re-assess its business operations and strategic plans, and its long-term survival depends upon its compliance to the principles of sustainable development. Site selection decisions are among the strategic decisions that need to be made in the context of sustainable development.

Most of the facility location theories examine the optimum site alternative that minimizes the associated cost or maximizes serviceability of one or more facilities. However, in some facility location studies, social and environmental factors, in addition to conventional criteria, were taken into account. Most of these studies are associated with undesirable facilities. The studies of Gros (1975, 281–292), Muntzing (1976, 3–11), Barda et al. (1990, 332–346), Queiruga et al. (2008, 181–190), Sumathi et al. (2008, 2146–2160), and Achillas et al. (2010, 870–879) are some examples. Nevertheless, the existing literature of facility location theories does not effectively address all the requirements of sustainable development and the need for framing the concept of sustainability in location decisions was evident.

Although various siting guidelines and instructions have been proposed in theories and practices, most of them focused on environmental issues, and only a few of them addressed the other aspects of sustainability. The objective of this research is to fill in this gap in the literature and define and frame the sustainable facility location problem by considering all aspects of sustainability. For this purpose, a case study is designed to analyse the sustainability characteristics of multiple site location alternatives and identify the most appropriate ones for locating a cement plant in the state of Florida. The geographic information system (GIS) technique is utilized to evaluate the spatial characteristics of site candidates. Examples of these characteristics include technical requirements of sites, such as access to raw materials and market, and environmental requirements in terms of distance from certain areas, or population centres. GIS analysis helps in effectively screening out inappropriate site options.

Background

In 1999, Cement Sustainability Initiatives (CSI) was organized as one of the key management initiatives for the cement industry. CSI was established with the cooperation of 10 major cement companies and under the supervision of the World Business Council for Sustainable Development (WBCSD). The main target of this initiative is to evaluate the operations and policies of cement plants across the world in pursuit of the sustainability goals. In 2002, CSI published its first sustainability report titled Towards a Sustainable Cement Industry (Cement Sustainability Initiative 2002). The purpose of this report was to establish guidelines for cement companies across the world for conducting their businesses in the sustainability realm (Klee 2004, 9–12; Humphreys and Mahasenan 2002). The report includes 13 sub-studies, one of which is more relevant to the current research: Sub-Study 11: Management of Land Use, Landscape, and Biodiversity. By illustrating several case studies, sub-study 11 recommended the land management practices, which contribute to sustainable development in each phase of the plant lifecycle. Two of the basic recommended land management actions for improving the sustainable development during locating a cement plant are:

• Using techniques, such as GIS, to manage ecosystem processes for siting a cement plant. According to the report, “GIS techniques are also very valuable in monitoring land use patterns including habitation, housing, agriculture patterns etc. around existing and prospective sites.”(Misra 2002, vi).

• Minimizing the area affected by plant and quarrying operations.

One of the other valuable resources developed by CSI is the Environmental and Social Impact Assessment (ESIA) guideline. The purpose of the guideline is to address positive and negative impacts of a cement plant on the environment and the local community during the different phases of development, operation and closure. ESIA's high-level considerations in the site assessment phase of locating a cement plant include stakeholder mapping, land use, social structure and population, public health, biodiversity and ecosystem, cultural heritage, and landscape (WBCSDCSI 2005).

The focus of the following section of the literature review is on the facility location studies, which have addressed the concept of sustainability.

In their paper, Dudukovic et al. (2005, 1085–1088) used the term sustainable industrial siting. The main focus of their study was developing a site suitability map for placing industrial facilities in a case study. The study did not differentiate between desirable and undesirable industrial facilities. In that study, one of the considered evaluation criteria was proximity to high population density areas, which is in contrast with the targets of undesirable facility location decisions. In addition, some of the key technical and sustainability criteria, such as the closeness to raw material, market, and more importantly criteria pertaining to biodiversity were overlooked.

One of the other relevant studies is the work of Tsoutso et al. (2007), in which they proposed a sustainable siting procedure for small hydroelectric plants (SHPs) in Greek. According to them, some of the influencing sections on sustainable sitting of hydroelectric plants (HPs) are planning procedure, cases of obstructing the procedure, legal framework, bodies of implementation (public or private) and measures (social, environmental and technical).

In their study, Farahani et al. (2010, 1689–1709) recommended sustainable facility location for future research: “{t}he most important thought that we took into consideration, is how to measure these attributes [related to social and environmental objective functions]. Therefore, we can think of a term like sustainable facility location.” (Farahani, SteadieSeifi, and Asgari 2010, 1689–1709). In another article, Boloori Arabani and Farahani (2012, 408–420) suggested that the concept of sustainable facility location be considered in future research studies. As they mentioned, “[r]egarding this fact, a trade-off must be set between handling changing parameters of the problem and maintaining the sustainability of the location decision.”

In their study, Cattafi et al. (2011) found a sustainable site for locating a biomass power plant. By taking into account a set of environmental criteria, such as proximity to water wells, unstable terrain and archaeological sites, they found a feasible area in the GIS environment. By using linear integer programming and considering mostly economic factors, such as demand and profit, they then tried to find the optimum location for the plant.

Terouhid et al. (2012, 18) have presented an extensive review of literature on the issue of facility location considering sustainability aspects and proposed a framework for location decision-making as well as essential aspects for adapting sustainability in facility location decisions.

There are numerous studies in the literature that have addressed various aspects of sustainability, such as social, economic or environmental factors, without any premeditation to implement the concept of sustainability with a holistic approach. Some of the relevant studies are the ones that focus on undesirable and semi-desirable facilities. The studies of Gros (1975, 281–292), Muntzing (1976, 3–11), Barda et al. (1990, 332–346), Queiruga et al. (2008, 181–190), Sumathi et al. (2008, 2146–2160), and Achillas et al. (2010, 870–879) are some examples.

Nevertheless, the existing literature of facility location theories does not effectively address all the requirements of sustainable development. The sustainable facility location concept is a holistic approach encompassing all sustainability dimensions including environmental, economic, and social in addition to the technical dimension. Based on the literature review, the need for framing the concept of sustainability in location decisions is evident. In this research, the author tried to fill in this gap in the literature and define and frame the concept of sustainable facility location by considering all aspects of sustainability.

Since in the current research, AHP, as an MCDM method, was selected for weighting the decision criteria of the case study, the background of the technique and its application in site selection studies are briefly described below.

AHP, known as an effective and flexible method for ranking qualitative and quantitative data, has been used in many site selection studies. In their paper, Yang and Lee (1997, 241–254) presented an AHP decision model for selection of a facility location. As they proposed, the main steps of the model include: 1. identify relevant facility location factors, 2. develop priority weights of factors, 3. collect data and rank alternative locations, 4. analyse comparative results, 5. identify preferred sites and 6. final recommendations.

Wang et al. (2009, 2414–2421) utilized AHP technique to weight the economic and environmental criteria for a landfill site selection. For analysing the spatial data, they used GIS technique. In their research, Moeinaddini et al. (2010, 912–920) used AHP to select solid waste disposal site in Iran. Sener et al. (2010, 2037–2046) combined AHP technique with GIS to evaluate the economic and environmental criteria for landfill site selection. For weighting the criteria with AHP technique, they decomposed the criteria into three levels 1. Main Criteria (environmental and economic), 2. Criteria (for example criteria under the main criteria of economic include height, slope, and distance from road) and 3. Sub-Criteria (for example the sub-criteria of slope include 0–10, 10–20, >20). By using GIS and AHP, Jack (2012) identified the potential sites for a limited impacted design storm water management project.

Environmental, Social, and Economic Impact Assessment (ESCIA) of a cement plant

Environmental, social, and economic impact assessment is an inevitable input for siting decisions. Therefore, it is suggested that a primary impact analysis of the targeted facility be conducted regarding its whole lifecycle. This assessment helps companies identify the major negative and positive impacts of the facility and customize the siting decision process and the evaluation criteria accordingly. As a result, companies would be able to utilize a prevention strategy rather than mitigation or compensation approaches for reducing the negative impacts of their facilities.

Production of cement has positive and negative, direct and indirect, social, environmental and economic impacts at the national and local levels. For the purpose of the case study, the major expected environmental, socialand economic impacts of a new cement plant during its operational phase are outlined later.

The main environmental issues of cement plants are greenhouse gas emissions (e.g. CO₂), air emissions (e.g. NO_x, SO₂ and dust/ particulate), and local nuisance (noise/vibration, dust and visual impact) (Marlowe and Mansfield 2002; WBCSD/CSI 2005). Toxic emissions of cement plants are very harmful to humans and wildlife (Cho and Giannini-Spohn 2007). The primary greenhouse gas emission is significant amounts of CO₂ generated in cement production. The main sources of CO₂ emission in cement plants are the manufacturing process – converting limestone to calcium oxide (accounted for 50% of total CO₂ emission of a cement plant) – fossil fuel combustion for manufacturing operations (40%), transportation (5%) and fossil fuel combustion for on-site electricity generation (5%) (Humphreys and Mahasenan 2002). The main sources of dust emissions are kilns, raw mills, clinker coolers and cement mills; and the principal sources of noise are blasting or drilling operation of quarrying and the cement grinding equipment.

Cement plants have negative traffic impacts due to hauling of raw materials and final products. These impacts can create soil contamination, noise, vibration and dust causing potential health and safety risks.

The main environmental concerns of quarrying activities of a cement plant are associated with dust pollution, noise, vibration and impacts on land use and biodiversity. Cement plants impact land use and biodiversity in the following manner: 1. displacement of current activities, flora and fauna on the land where quarry and cement plant are to be located; 2. land degradation and disturbance of biodiversity on the land and 3. creation of a land use that is incompatible with adjacent land uses, thereby affecting neighbouring landscapes (Misra 2002).

Cement plants are often located in karst areas and may create threats to these relatively sensitive areas. “Karsts are specialized geological formations related to erosion and dissolution of carbonate minerals contained in limestone and other sedimentary” (WBCSD/CSI 2005, 44). One of the key attributes of karst areas is that they have rich and productive soil for agriculture uses, and they are respected as valuable lands. Some of the karst landscapes are valuable sources of information from past environmental and historical conditions (WBCSD/CSI 2005). Furthermore, according to the study of Misra (2002), one-quarter of all people worldwide get their water supplies from karst areas.

In terms of social and economic impacts, the cement industry is very labour intensive (Marlowe and Mansfield 2002); therefore, a new cement plant can provide significant employment opportunities for the local community and it can enhance local workers' skills. A cement plant can also be a source of income and welfare for companies, local communities, and governments. In addition, locating a cement plant in a region makes the most consumed construction material readily available for local industries. Nevertheless, a cement plant may have negative social impacts; for instance, it might induce the migration of strangers to the local community, and subsequently, create adverse impacts on the cohesion and social networks of the community and its values. These plants may also increase the local traffic load, which has adverse impacts on the amenities of people.

Figure 1 summarizes a cement plant's impacts. It should be noted that the outlined impacts are basic and they are not prescriptive for all cement projects. The detailed impacts are directly dependent on various factors, such as business plans and national or local environmental conditions. Therefore, it is suggested that after selecting the site alternative, a secondary ESCIA in association with the site be conducted. Knowing the physical characteristics of the site and the local condition, companies can better evaluate the impacts of the plant on the surrounding areas and address the stakeholders' concerns. This level of ESCIA will also help companies to manage and minimize their adverse impacts in a more effective way during the whole lifecycle of the plant.

Figure 1 Environmental, social and economic impacts of a cement plant.

Framing the problem of siting a cement plant in the state of Florida

The state of Florida has six existing cement plants with eight active kilns located in the counties of Alachua, Hernando, Miami-Dade, Sumter and Suwannee. In 2012, total Portland and Blended cement production in Florida was about 3,846,862 metric tons, which accounted for 5.3% of the total cement production of the USA. In 2012, the total Portland and Blended cement consumption in Florida was about 3,883,252 metric tons, which accounted for 5% of total Portland and Blended cement consumption of the USA (Van Oss and Kraft 2013).

The purpose of the current study is to hypothetically locate a new cement plant in the state of Florida with consideration of sustainability criteria. The following assumptions are considered for the case study:

• In this study, a cement plant is considered as a semi-desirable facility. Semi-desirable facilities are desirable facilities for people in terms of their services, but these facilities have negative environmental impacts and cause inconveniences to the neighbouring areas (Colebrook and Sicilia 2007, 1491–1514).

• In the problem, only limestone is considered as the primary raw material for cement production.

• Limestone, as a county's natural resource, can be transported to a cement plant located in another county.

• In the demand analysis, the population of Florida is considered as the only factor that determines the demand. It is also assumed that exporting cement to the other states or countries will have the same condition for all site alternatives.

• All the existing limestone reserves of Florida are potentially cement-quality limestone. In other words, the reserves meet the required quality for production of cement.

• The considered limestone reserves in different areas of Florida provide enough raw materials for a cement plant to operate according to its business plan.

• The limestone reserves of the site alternatives are similar in terms of quality.

Sustainability requirements include various factors, such as socio-economic ones, which need to be addressed within a large area, e.g. a state, rather than a small one, e.g. a few parcels of a region. Considering a small area at a local scale for sustainable siting of a facility would make the decision process inefficient. Therefore, to conduct the study, siting analysis was performed at the macro-level to identify candidate counties of the state of Florida and at the micro-level to select appropriate parcels within the candidate counties. In addition, since considering sustainability requirements makes location decisions more complicated, for the sake of simplicity, both macro- and micro-level analyses were conducted in two phases of screening and suitability. The purpose of the screening phase is to identify feasible location options based on compulsory and cautionary criteria. The screening phase not only facilitate the decision-making process by reducing the numbers of candidates which do not meet the compulsory and precautionary criteria, but also help considering sustainability and environmental criteria earlier in decision-making processes. The suitability criteria measure the appropriateness of the feasible options, resulted from the screening phase. In the next sections, the decision criteria and the siting analyses are briefly described.

Evaluation criteria

Decision criteria are the foundation of any assessment, and measuring the sustainability of locations is not possible without identifying the evaluation criteria. To characterize sustainable location of a cement plant, a list of criteria developed. In the process of developing the list of the criteria, first, a primary list including sustainability and conventional decision criteria was developed based on the literature, exiting guidelines, and the reviewed ESCIA. In extracting the criteria from the literature, those that were relevant to the current research and were in line with the assumptions of this research were considered. At the next step, the evaluation criteria were reviewed by the researchers to eliminate the superfluous ones and to classify them. Expert judgments were also solicited in this phase. Four categories of dimension, theme, sub-theme and index in a hierarchical order were identified for the classification of the criteria.

Dimensions represent general objectives of the problem of sustainable semi-desirable facility location including social, economic, environmental and technical principles. Themes, sub-themes and indices are grouped under dimensions. Because the sustainability dimensions including environmental, social and economic, have limitations that cannot cover the operational and technical issues of a facility, such as the market status, the fourth technical dimension is also considered in the framework of the criteria.

Themes and sub-themes are defined to translate the goals of dimensions into measurable indices. Themes and sub-themes have more specific meanings and are more concrete compared to dimensions. The main purpose of indices is to provide meaningful measures and information for evaluation and final decision-making (Worrall et al. 2009, 1426–1434). A good indicator should be specific, measurable, sensitive to change, achievable in terms of available data, and should have analytical and scientific soundness and policy relevance (Niemeijer and de Groot 2008, 14–25).

The evaluation criteria were also categorized and customized into the following four major classes:

• Mandatory and discretionary screening criteria for identifying feasible counties

• Suitability criteria for prioritizing the feasible counties and identifying candidate counties

• Mandatory and discretionary screening criteria for identifying potential parcels in the candidate counties

• Suitability criteria for selecting appropriate parcels among the potential ones

In the following, the primary steps taken in siting analysis are described

Macro-level analysis

In this phase, both the screening and suitability evaluations were performed. First, by applying the screening criteria, the counties in which cement production was feasible were identified. The suitability of these feasible counties then was evaluated, and they were ranked based on the suitability criteria. In the following sections, the evaluations are briefly explained.

Screening analysis

Due to the heavy weight of raw materials and high transportation costs, the proximity of raw materials to a cement plant is a crucial factor in its economic feasibility. Proximity to raw materials reduces fossil fuel consumption and provides regular and consistent material supply as well as savings of storage spaces and costs. Therefore, in this case study, the availability of limestone was considered as the screening criterion for identifying the feasible areas and counties in the state of Florida. The areas within 5 km buffer distance of limestone were identified as the feasible areas, and the counties in which the feasible areas are located were considered as the feasible counties (see Figure 2). The buffer distance of 5 km, for the proximity of plants to limestone, was selected based on the reviewed literature. As the result of macro-level screening analysis, out of 67 counties of Florida, 40 counties were identified as the feasible counties.

Figure 2 Feasible areas (areas located in 5 km buffer of limestone).

Suitability analysis

In this phase, the suitability of the feasible counties was evaluated to identify those with higher priorities. The evaluation was done based on the macro-level suitability criteria including demand, number of competitors, operational cost, availability of infrastructure, environmental status, poverty, public health status, social acceptability, regional economic performance and employment status (See Figure 3). Some of the suitability indices that need more clarification are explained in the following.

Figure 3 Hierarchy of criteria evaluated by AHP method.

Regarding the environmental status of Florida, a project called Critical Land and Waters Identification Project (CLIP) was developed through a collaboration of the Florida Natural Areas Inventory (FNAI), the University of Florida Center for Landscape Conservation Planning, and the Florida Fish and Wildlife Conservation Commission. The CLIP project includes a spatial database of natural resources of Florida and can significantly facilitate spatial planning decisions at the state level. In the project, an aggregate priority was identified based upon three resources of biodiversity, landscapes and surface water.

Since the Aggregate CLIP project is very comprehensive and covers key natural resources of Florida, it was selected for evaluating the environmental status of the feasible counties. For this purpose, the following CLIP index (Equation (1)) was defined:

(1)

In the CLIP project, priority 1 is the highest priority. As a result of consulting with a CLIP project member, the areas with CLIP priorities 1 and 2 were considered as critical environmental areas. In the context of sustainability, the feasible counties with lower CLIP index values were preferable.

In the suitability analysis, for the purpose of measuring the availability of infrastructure, the distances between the centres of the feasible areas in each feasible county and the nearest infrastructure were measured in the ArcGIS environment..

In this study, poverty was measured by the percent of residents below the 100% poverty line, and income inequality was measured by the Gini Coefficient, which is the most common index for measuring income inequality. The Gini Coefficient is a dimensionless index and can be defined as “{a} summary measure of the extent to which the actual distribution of income, consumption expenditure, or a related variable, differs from a hypothetical distribution in which each person receives an identical share.” (UN DSD 2001, 62) Considering the economic objectives of sustainable development, the feasible counties with higher Gini Coefficient and higher percent of families under poverty line were preferred for siting a new plant due to the expected economic growth. A new industrial plant is more desirable for undeveloped communities because it creates new jobs and economic opportunities.

The County Health Rankings and Roadmaps organization has developed a health ranking system for nearly all US counties. The program can significantly facilitate health improvement decision-making processes (County Health Rankings and Roadmaps Organization 2012). These rankings were used for evaluating public health status and healthcare factor of the feasible counties.

Social acceptability is one of the considered criteria under social dimension. Social acceptability is about the willingness of the targeted community to accept a facility in their region, and it has a significant impact on the location decision of a new industrial plant (Garrone and Groppi 2012). In this research, the “percentage of high school graduates” was selected to measure the social acceptability of the feasible counties. Higher education enables communities to collect information about potential negative impacts of the plant, and consequently, they will probably have higher propensity to reject the plant. Although the goal of the current study is to locate a cement plant in a sustainable manner, some of the potential negative impacts of the plant on the targeted community are unavoidable. Therefore, social acceptability is a key criterion in locating an industrial plant. After selecting a site, companies should devise strategies to relocate the community or compensate them for any potential negative impacts of the plant.

To evaluate the list of feasible counties, the relative importance of the decision criteria had to be identified. In this research, the AHP method was selected for weighting the suitability criteria of the first level analysis (the county selection process) to derive the relative importance of the criteria based on the opinions of a panel of experts and by pair-wise comparisons of the criteria in a hierarchical process. In addition, creating a balance among the weights of conventional and sustainability criteria was one of the objectives of this research; therefore, soliciting expert judgments from different academic and industrial sections was crucial to make balance among judgments. In this regard, AHP method significantly facilitated the process.

However, it should be noted that the AHP method was used only for weighting the dimensions and the sub-themes. Using AHP to determine the weights of all dimensions, sub-themes, and indices was impractical due to high number of responses required from the participants. Therefore, for weighting the indices, a consensus was achieved among a small group of experts.

To perform the AHP analysis, a questionnaire was developed and distributed among experts who had knowledge and expertise in the areas of sustainability and the cement industry. The experts were selected from both industry and academia. The questionnaire included four comparison tables related to the main dimensions and the sub-themes of each dimension. Figure 3 shows the hierarchy of the criteria evaluated by the AHP method.

To compare the criteria, the scale of 1–9 was adopted if the criteria had direct relationship and the scale of 1/2 to 1/9 was used if the criteria had an inverse relationship. Table 1 presents the rating scale used in the AHP method.

Table 1 The Rating Scale table (developed based on the rating scale proposed by Saaty (1994)).

Comparative importance	Description	Explanation
1	Equally important	Two criteria contribute equally to the objective
3	Moderately more important	Experience and judgment slightly favour one criterion over another
5	Strongly more important	Experience and judgment strongly favour one criterion over another
7	Very strongly more important	An activity is strongly favoured and its dominance demonstrated in practice
9	Extremely more important	The evidence favouring one criterion over another is of the highest possible order of affirmation
2,4,6,8	Intermediate values between the two adjacent judgments	When compromise is needed

Twenty six responds were received. The average total year of experience of the participants was 28.3 years. Furthermore, 38.5% of the participants were involved in site selection processes. The results of the pair-wise comparisons of criteria were filled in matrices using excel spreadsheets to calculate the eigenvectors (individual preferences).

One of the key steps in the AHP method is checking the consistency of the results through the consistency ratio. If the consistency ratio exceeds 0.10, the comparison matrix should be re-assessed. The consistency ratios of individual judgments for each comparison matrix were checked and those with C.R. greater than 0.1 were excluded from the calculation of the preferences (weight factors).

By calculating the normalized principal eigenvectors of each comparison matrix, the individual preferences were computed. Table 2 shows the opinion of one of the experts in comparison of the dimensions along with the eigenvector of the matrix. According to Forman and Peniwati (1998), when more than one respondent are involved, only geometric mean should be used for aggregating individual judgments; however, either arithmetic or geometric mean can be used for the aggregation of preferences. In this research, arithmetic mean was used to aggregate individual preferences.

Table 2 Opinion of Expert 1 in comparison of the dimensions.

	Technical	Environmental	Social	Economic	Eigenvector (preference)
Technical	1.00	1.00	3.00	5.00	0.42
Environmental	1.00	1.00	2.00	3.00	0.33
Social	0.33	0.50	1.00	1.00	0.14
Economic	0.20	0.33	1.00	1.00	0.11
Landa max	4.053
C.I.	0.018
C.R.	0.012

Tables 3456 show the weights of all dimensions, sub-themes and indices.

Table 3 Weight values (W.V.) of technical indices.

Dimension	Sub-theme	W.V.	Index	W.V.
Technical 0.35	Demand	0.33	Population in 300 km buffer distance	1.00
	No. of competitors	0.13	Max capacity of existing kilns	1.00
	Operational cost	0.26	Industrial power rate	1/3
			Gas rate	1/3
			Median hourly wage of production workers	1/3
	Availability of infrastructure	0.28	Distance to the nearest major road (km)	1/2
			Distance to the nearest railroad (km)	1/6
			Distance to the nearest sea port (km)	1/6
			Distance to the nearest navigable waterway (km)	1/6

Table 4 Weight values (W.V.) of environmental indices.

Dimension	Sub-theme	W.V.	Index	W.V.
Environmental 0.30	Environmental status	1.00	% of area with CLIP priority 1,2	1.00

Table 5 Weight values (W.V.) of social indices.

Dimension	Sub-theme	W.V.	Index	W.V.
Social 0.21	Poverty & income inequality	0.29	Gini Coefficient	1/2
			Percent of families below 100% poverty	1/2
	Public health status	0.21	The rank of health outcomes	1.00
	Health factors	0.23	The rank of health factors	1.00
	Social acceptability	0.27	High school graduation (%)	1.00

Table 6 Weight values (W.V.) of economic indices.

Dimension	Sub-theme	W.V.	Index	W.V.
Economic 0.14	Regional economic performance	0.52	Money income (per_capita)	1.00
	Employment status	0.48	Unemployment rate %	1.00

Comparison of the feasible counties

After identifying the weights of the criteria and collecting all the data in association with each of them, the feasible counties were evaluated and ranked. The spatial data were analysed in the ArcGIS environment. Since the spatial and non-spatial data had different measuring units, they were first normalized. The normalization process converts different scales of criteria's values into common measurable units to be able to compare the alternatives, based on the criteria.

Then the normalized data were added up for calculating the final scores of the feasible counties. For this purpose, SAW (simple additive weighting) method (Equation 2) was used.

(2)

where i = (1,…,m) is the number of the alternatives, j = (1,…,n) is the number of the criteria, W_j is the weights of the suitability criteria, x′_ij is the normalized jth criterion's value for ith alternative.

Based on the scores of counties, eight priority groups were identified. For the micro-level analysis, the counties with the highest priority, priority one, were selected as the candidate counties including: 1. Polk, 2. Putnam, 3. Pinellas, 4. Collier, 5. Dixie, 6. Glades, 7. Pasco, 8. Hillsborough and 9. Hendry.

It is noteworthy that among the candidate counties, Putnam, Dixie, Glades and Hendry were designated as Rural Areas of Critical Economic Concern (RACEC). Under Rural Economic Development Initiative (REDI) program, the Department of Economic Opportunity (DEO) has designated some of the rural areas of Florida as Rural Areas of Critical Economic Concern (RACEC). These areas have been adversely impacted by extraordinary economic events, natural disasters, or they have lack of natural resources (Florida Department of Economic Opportunity 2013).

Micro-level analysis

After performing the macro-level analysis, the micro-level evaluation was conducted on the candidate counties to select the appropriate parcels for siting a cement plant with sustainability and technical considerations. In the following sections, the evaluation phases of the micro-level analysis (screening and suitability) are briefly described.

Screening analysis

As the first step of the micro-level analysis, the screening micro criteria were evaluated to identify the candidate parcels. The screening micro criteria include: 7 km buffer distance from Florida urbanized area and conservation lands, FFBOT and FF Acquired projects, CLIP priority 1 & 2, land use and land size.

Two of the considered screening environmental criteria were a 7 km buffer distance from Florida urbanized areas with more than 50,000 population, and a 7 km buffer distance from Florida conservation lands. For evaluating the ambient air quality of a cement plant, CSI proposed 7–10 km as the impacted region (Misra 2002). As a result, 7 km was assumed for the buffer distance.

The other screening environmental criteria were Florida Forever Board of Trustees (FFBOT) and Florida Forever Acquisitions (FF Acquired) projects. FFBOT projects include the lands approved by the State's Acquisition and Restoration Council and administered by the Florida Department of Environmental Protection, Division of State Lands, for the State Board of Trustees. These lands are notable for their distinguishable natural, historical, and archaeological resources. FF Acquired projects include the lands purchased through the Florida Forever Program beginning in 2001. FFBOT and FF Acquired can be considered as mandatory environmental criteria for this case study.

Consideration of urban planning designations is a crucial prerequisite for any site selection decision. Accurate spatial planning can preserve natural resources and protect humans from potential hazards and health risks. It may also reduce the risks of unknown spatial factors for cement companies. Therefore, in this study, land use was considered as another screening criterion. For this purpose, industrial (heavy and light manufacturing), mineral, not-identified land uses and land uses for future development were considered for the evaluation. However, this assumption resulted in no potential parcels in all candidate counties. Therefore, agricultural land parcels were also considered in the screening phase due to the fact that, under special conditions, the land use can be changed. Nevertheless, the codes of ordinances of all the candidate parcels were checked in the suitability phase for the purpose of locating a cement plant.

Plant sizing is a major component of the business plan of a cement factory. It is proposed that plant sizing be implemented before any site evaluation. Since developing a business plan for a new cement plant was out of the scope of this research, the required land size was based on the minimum land size of the existing cement plants in Florida, 120 acres. Although the acquisition of several adjacent small lands can be another option for locating a cement plant, this was not considered in this study due to its complexity. In the screening phase, lands with CLIP priorities 1 and 2 were also excluded from feasible areas.

As the results of the screening phase in the micro-level analysis, 11 candidate parcels were identified for the suitability analysis. The candidate parcels were located in Collier, Glades and Hendry counties. Figures 4 and 5 show the location of the candidate parcels in relation with limestone and the candidate areas (Figure 6).

Figure 4 Collier potential and candidate parcels.

Figure 5 Hendry candidate parcels.

Figure 6 Glades candidate parcels.

Suitability analysis

At this phase, the suitability of 11 candidate parcels resulting from the screening phase was evaluated based on all micro-level suitability criteria to identify the most appropriate parcels. Figure 7 shows the hierarchy of the evaluated criteria.

Figure 7 Hierarchy of micro-level suitability criteria.

According to the micro-level analysis, only the planning ordinance of Glades County has flexibility and allows for the placement of a heavy manufacturing plant in the candidate parcels with “Transition” future land use. In Glades County, there is a policy that “{l}and parcels within areas designated ‘Transition’ on the Future Land Use Map may be converted from agriculture to commercial or industrial use” (Glades County Government 2010). Since the candidate parcels of Glades County are adjacent, their attributes are approximately the same.

Nevertheless, decision-making about the final parcel is a multi-disciplinary process, which should be made by a panel of experts, and in reality, several other key issues need to be considered before making the final decision. The quality and quantity of the available limestone in each parcel is a key technical factor for siting a cement plant. The composition of limestone not only impacts the final products, but also affects the process design, and subsequently, waste, water and air pollution. The quantity of the limestone greatly impacts the economic feasibility of the plant. Therefore, any final decision regarding the location of a cement plant directly depends on the composition and amount of the available limestone in that area. In this case study, due to the limited resources, it was not feasible to check this technical factor.

If it is assumed that embracing sustainability aspects might create possibilities to change the planning designations of Collier and Hendry parcels, there is still not one single dominant solution with the best attributes. Therefore, the candidate parcels were ranked based on the suitability criteria. The performance of the candidate parcels in terms of different suitability criteria was evaluated. In cases where the attributes of all candidate parcels were almost the same in terms of a criterion, that criterion was excluded from the ranking process. Therefore, the following criteria were not considered in the ranking phase:

• Traffic volume: The total volume of traffic on a highway segment for one year, divided by the number of days in the year

• Major hazardous lines: It includes major gas transmission pipelines, major hazardous liquid pipelines, LNG plants and power lines

• CLIP priority: It is explained in previous sections

• Fault line: The intersection line between the earth's surface and the surface of a fault

In addition, DVI and proximity to water wells criteria, for which companies might be able to develop a remedy action plan, were not considered in the ranking phase. DVI is a standardized grading system for the evaluation of ground water pollution potential based on hydrologic factors, including depth of water, net recharge, aquifer media, soil media, topography, impact of vados zone media, and hydraulic conductivity of the aquifer. The combination of the weighted factors produces DVI. The areas with higher DVI have more potential to affect ground water quality.

In this phase, the micro-level suitability criteria were weighted equally and in identifying the suitability scores, the parcels' attribute values were normalized. According to the results listed in Table 7, the parcels of Glades County and parcels 4 and 5 of Hendry County have the highest suitability scores, respectively. All these parcels are located in the limestone areas. In general, it is proposed that once the most suitable site is selected based on the macro and micro level evaluations, the site should be physically analysed in more detail. The availability of the land for the acquisition process might be a challenge for companies. Although the main target of the current research is to minimize potential challenges by addressing them during the decision-making process, companies might still face them during the land acquisition phase. Responding to the concerns of all parties along with satisfying all decision criteria is hardly possible in reality; therefore, an appropriate solution that balances different requirements needs to be pursued.

Table 7 Suitability scores of the candidate parcels for cement plant purpose (excluding quarrying operation).

ID	Urban planning flexibility	Proximity to limestone (km)	Proximity to main road	Min 500 m distance to national highways	Proximity to navigable waterway	Proximity to railroad	DOR land value/acre
Glades_3	1.00	1.00	0.18	1.00	1.00	0.42	1.00
Glades_2	1.00	1.00	0.19	1.00	1.00	0.46	0.76
Glades_1	1.00	1.00	0.22	1.00	1.00	0.49	0.73
Hendry_4	0.00	1.00	0.99	0.00	0.95	0.38	0.79
Hendry_5	0.00	1.00	0.99	0.00	0.92	0.37	0.75
Hendry_3	0.00	0.21	1.00	1.00	0.85	0.02	0.42
Hendry_6	0.00	0.50	1.00	1.00	0.87	0.19	0.81
Hendry_1	0.00	0.42	0.87	1.00	0.88	0.00	0.19
Collier _1	0.00	0.09	0.11	1.00	0.00	1.00	0.05
Hendry_2	0.00	0.30	0.50	1.00	0.88	0.00	0.31
Collier _2	0.00	0.00	0.00	1.00	0.00	1.00	0.00
ID	Low risk flood zone	Proximity to medical services	Affected population	Equity	Final score
Glades_3	0.00	0.84	0.62	0.52	7.58
Glades_2	0.00	0.90	0.07	1.00	7.36
Glades_1	0.00	0.94	0.00	0.94	7.32
Hendry_4	1.00	0.51	0.95	0.05	6.62
Hendry_5	1.00	0.43	0.84	0.03	6.34
Hendry_3	1.00	0.00	1.00	0.05	5.56
Hendry_6	0.00	0.13	0.99	0.01	5.51
Hendry_1	0.00	0.14	0.94	0.02	4.46
Collier _1	0.00	0.95	0.85	0.00	4.06
Hendry_2	0.00	0.07	0.88	0.03	3.97
Collier _2	0.00	1.00	0.81	0.14	3.95

After checking the availability of the land for acquisition, its physical characteristics should be visually inspected. In the real world, it is impractical to take this step during the analysing phase for all candidate parcels in different locations. This phase includes, but is not limited to, the soil evaluation, hydrological characteristics, topographic maps, and existing vegetation of the site. If a plant is dependent on on-site raw material, the accessibility and quality of the raw material should be examined in detail. Site evaluation will also help in determining any potential obstacles and constraints.

Aerial photographs (such as the photographs available on Google Earth), topographic maps, ground-level photographs, and websites of property appraisers can be good sources for physical assessment of the land. Historical information and photographs of the site can also be considered as helpful evaluation resources. They can assist in determining previous land uses, the archaeological and historical value of the site, historical environmental contamination in the site, and any other ambiguous physical constraints. Although this information might be unofficial or unwritten, it may include crucial elements that need to be addressed before placing the facility (LaGro 2011).

Addressing stakeholders' concerns is another essential step in a site selection process. Stakeholders are agencies or people who can potentially be affected by or might affect an industry. Therefore, building trusting, open, and inclusive relationships with stakeholders and addressing their concerns are unavoidable parts of siting decision processes (WBCSD/CSI 2005). A public hearing is one of the effective tools for addressing stakeholders' concerns in siting or expansion of semi-desirable facilities; therefore, it can be considered as a contributory step of the sustainable facility location. Although this step is intuitive, it is not a common practice among all practitioners.

In this study, considering precautionary sustainability criteria in the micro level analysis of the case study resulted in 11 candidate parcels, and only 5 of them were located on limestone. This means that in order to site a cement plant in one of the other candidate parcels, additional property for the purpose of quarrying is required. Since only a few candidate parcels are available, companies may face challenges in acquiring one of them.

One of the methods that may address this issue to increase the number of candidate parcels is Sensitivity analysis to investigate the impacts of the decision criteria on the results. Relaxing some criteria with significant impacts may result in acquiring more candidate parcels. In the next section, the results of the sensitivity analysis conducted in this study are reported.

Sensitivity analysis of the parcel selection process

Four of the precautionary sustainability criteria, which resulted in limiting the number of candidate parcels, were 7 km buffer distance from conservation lands, 7 km buffer distance from urbanized areas, land size, and land use. To have more candidate parcels, some compromises may be made regarding these criteria. For this purpose, it was decided to conduct a sensitivity analysis to investigate the impacts of these criteria on the number of candidate parcels. Figure 8 demonstrates the relationship between the number of candidate parcels and the buffer distances. In this analysis, all the other screening criteria were kept the same. As shown, reducing the sustainability buffer distances from 7 to 3.5 km increases the number of candidate parcels from 11 to 70.

Figure 8 The result of sensitivity analysis on buffer distance from conservation lands and urbanized areas.

Figure 9 shows the relationship among the number of candidate parcels, land size and land use. In this analysis, the criteria of 7 km buffer distance from urbanized areas, FFBOT and FF Acquired projects, and CLIP priority 1 and 2 were kept the same. The number of the candidate parcels without considering land use and land size criteria is 754. By considering land use, the number will decrease to 139, whereas by considering land size the number will decrease to 18. Therefore, the impact of land size is greater than that of land use.

Figure 9 The result of sensitivity analysis on land size and land use.

Conclusion

The Cement Industry, as one of the controversial industries in terms of its negative environmental impacts, has been under scrutiny in the last two decades. To pursue sustainable development, this industry should reassess its operation and strategic plans and consider sustainability requirements in depth. The main motivation of this research was to minimize the environmental footprint of industrial facilities through strategic decisions. Site selection is one of the strategic decisions with long term impacts on companies, environment and local communities.

Based on the conducted literature review in this research, the need for framing the concept of sustainability in location decisions was evident. Although various siting guidelines and instructions have been proposed in theories and practices, most of them focused on environmental issues, and only a few of them addressed the other aspects of sustainability. By conducting a case study, this research has filled in this gap and contributed to both theories and practices of facility location and sustainable development. Having a comprehensive approach combined with sustainability considerations in site selection ensures that the adverse impacts of facilities on their surrounding environment are minimized. It can also help decision-makers to make better location decisions to minimize construction, operational, and maintenance costs of various facilities.

In this study, the sustainable location problem of a cement plant in the state of Florida was examined. For this purpose, an Environmental, Social, and Economic Impact Analysis (ESCIA) associated with cement plants was first conducted. By developing and categorizing the decision criteria, the macro- and micro-level analyses were conducted and each level of analysis was performed in two phases of screening and suitability. Making location decisions based on the proposed criteria helps in addressing the challenging requirements of sustainability. However, it should be noticed that the proposed criteria needs to be regarded as a guideline and it needs to be customized for any projects based on the companies' objective, type of facility, the applied technology, plant design and local circumstances.

To address sustainability requirements more effectively, companies are suggested to conduct a comprehensive siting evaluation process at both macro- and micro-levels. The purpose of the macro-level analysis is to identify the candidate counties and micro-level analysis is conducted to select a suitable parcel in the candidate counties. With this classification, companies ensure that all technical and sustainability factors are taken into account in their decisions. For instance, if companies, without any evaluation, choose the county or community in which they want to place the facility, some of the socioeconomic sustainability requirements, such as health factors and economic status, may not be satisfied.

Creating a balance among the importance weights of sustainability and technical criteria is a crucial step in a siting process. Developing a strategic plan demonstrating how different approaches would be balanced might facilitate this process. In this study, AHP method was used for weighting the macro-level suitability criteria. For creating a balance among the weights of conventional and sustainability criteria, the opinions of experts from different sectors of industry and academia were solicited. The experts were selected based on their knowledge and expertise in the areas of sustainability and the cement industry.

During sustainable siting process, decision-makers may face various difficulties. Considering tight sustainability criteria may significantly limit the number of potential candidate sites. For instance, the consideration of the 7 km buffer distance from the urbanized areas and conservation lands were two of the sustainability screening criteria that significantly limited the number of potential site candidates. Therefore, meeting all sustainability and technical criteria in site selection decisions can be a challenging process that requires reaching a trade-off among the criteria. In some cases, sustainability or technical constraints may need to be relaxed to some extent to increase the number of potential site candidates. For this purpose, in the case study, a sensitivity analysis was performed, and it was showed that by reducing the buffer distance from the conservation lands and urbanized areas, the number of potential parcels significantly increases.

An accurate spatial planning by local planning departments is a prerequisite for sustainable facility location decisions. Although sustainability principles were considered in the case study for identifying the candidate parcels, the zoning ordinances of most of the candidate parcels did not allow for the placement of a heavy manufacturing plant. Therefore, addressing these conflicts requires taking a more flexible and holistic approach in spatial planning to facilitate sustainable location decisions.

Overall, this research demonstrated that integrating sustainability requirements into location decisions is not only possible but is also necessary. This study provides interested cement companies and researchers with an insight into how to holistically include the concept of sustainability in siting decisions.

In this study, it was assumed that the location decision is a state-wide decision; however, in reality it might be a nationwide or an international decision. Therefore, new research studies can be framed for these types of problems to outline the characteristics of national or international sustainable facility location problems. Framing the concept of sustainable facility location for other types of facilities can also be considered as future research areas. Although the focus of this research was on cement plants, the applied methodology can be utilized for other facilities with some modification. However, for some of the undesirable facilities, the list of the criteria should thoroughly be reviewed to assess the need for the inclusion of more stringent criteria. In addition, in the case of locating undesirable facilities, the political (governmental) aspect might need to be considered as an additional dimension to the problem.

Disclosure statement

No potential conflict of interest was reported by the authors.