Comparative Environmental Life Cycle Analysis of Stone Wool Production Using Traditional and Alternative Materials

The mineral wool sector represents 10 % of the total output tonnage of the glass industry. The thermal, acoustic and fire protection properties of mineral wool make it desirable for use in a wide range of economic sectors especially in the construction industry for the creation of low energy buildings. The traditional stone wool manufacturing process involves melting raw materials, in a coke-fired hot blast cupola furnace, fiberization, polymerization, cooling, product finishing and gas treatment. The use of alternative raw materials as torrefied biomass and sodium silicate, is proposed as an alternative manufacturing process to improve the sustainability of stone wool production, particularly the reduction of gas emissions (CO2 and SO2). The present study adopts a life cycle analysis (LCA) approach to measure the comparative environmental performance of the traditional and alternative stone wool production processes; process data are incorporated into a LCA model using SimaPro 8 software with the Ecoinvent version 3 life cycle inventory database. The CML 2000 and Eco-Indicator99 methods are used to estimate effects on different impact categories. The Minerals and Land use impacts in Eco-Indicator99 and the Eutrophication impact in CML2000 increase between 2 and 4 % for the alternative process instead of the traditional one. Similarly, the ecotoxicity-related impacts increase between 9 and 24 % with the use of the alternative process. However these increases are compensated by concomitant impact decreases in other categories of impact; consequently, the three areas of impact grouped by individual Eco-indicator 99 impacts, show environmental benefits improvements between 6 and 15 % when using the alternative process based on torrefied biomass and silicate instead of the traditional process based on coke and cement use.


Introduction
Fibrous materials may be naturally occurring or synthetically manufactured by thermal or chemical processes. Refractory ceramic fibre, fibre glass and mineral (or stone) wool belong to a class of materials known as synthetic vitreous fibres [1]. Mineral wool is typically used in the construction industry for heat insulation, cold and fire protection, and noise insulation [2]. In 2011, traditional mineral wool prevailed in the world thermal insulating materials market, with a 52 % market share. The technical, environmental and public health aspects of the insulation materials, play an increasing role in the highly competitive building construction market [3], and more environmentally friendly buildings outlines developing opportunities for improved, new and alternative sustainable insulating materials [4,5].
In Europe mineral wool production directly employed over 21,000 people at 62 installations in 2005. The total volume of rock wool production in EU27 countries between the years 2003 and 2011 is highly variable (between 1.95 and 2.5 million tonnes), but the annual production volume showed an average growth rate of 0.91 % as the general trend [6][7][8].
The most common melting technique for the production of traditional stone wool is the coke-fired hot blast cupola furnace. Typically used raw materials are: (1) igneous rocks such, as diabase, gabbro or basalt; (2) briquettes, made from a blend of various minerals, such as olivine or basalt, diabase and/or gabbro, together with recycled waste stone wool with cement as binder; and (3) limestone added to adjust the viscosity of the melt to the requirements of the spinning process.
Molten material, from 1300 to 1500°C, gathers at the bottom of the furnace and flows out of a notch and along a short trough positioned above the spinning machine. Air is blasted from behind the rotating wheels to attenuate the fibres and to direct it onto the collection belt to form a mattress. An aqueous phenolic resin solution is sprayed over the fibres. The mattress passes through an oven, which dries the product and cures the binder. The product is then cooled and cut to size before packaging. Gases emitted during the production process are cleaned in gas treatment systems to minimize the environmental impact. Water use in the process is generally confined to closed circuit systems.
A set of best available techniques (BAT), with the potential for achieving a high level of environmental protection, can be applied to stone wool manufacturing installations; these BAT are focused to avoid, reduce and control dust and gaseous emissions from melting and downstream manufacturing processes. Environmental management systems, process-integrated techniques and end-of-pipe measures, waste minimization and recycling procedures, and techniques for reducing the consumption of raw materials, water and energy, are proposed as BAT [9].
Alternative raw materials derived from mineral wastes can play an important role in manufacture of mineral wool [10]. Several methods have been developed to return fine rock wool production waste and to recycle mineral wool waste to the manufacturing process through briquetting mineral wool waste with a binder material [7,11]. An alternative process patented by VL Ambiental Company [12], proposes the use of briquettes formed by waste rock wool agglomerated with torrefied biomass (e.g., conventional biomass, sewage sludge) for production as alternative fuel briquettes. The binder used is a non-fibrous inorganic material such as sodium silicate, which replaces the cement used in the traditional process.
Using a sulphur free binder and a CO 2 neutral biomass fuel has the advantage of reducing both the emissions of CO 2 and SOx. In addition to the low nitrogen content of biomass, the fuel nitrogen in biomass is converted to NH radicals during combustion providing an in situ thermal DeNOx source and can also result in lower NOx levels [13].
Life cycle analysis (LCA) is a methodological tool that is used to measure the environmental impact of a product, process or system throughout its life cycle. It is based on the collection and analysis of the inputs and outputs of a system to obtain results that show its potential environmental impacts; LCA results can be used to identify strategies for reducing environmental impact and to improve industrial processes to become more environmentally friendly under a cradle to gate approach [14]. The LCA approach has been applied extensively to construction materials and insulation materials, particularly to mineral wool products [15][16][17][18].
The main objective of the present study is to utilize the LCA approach to measure the comparative environmental performance of the different stages of traditional and alternative stone wool production processes. A process flow diagram is built and mass and energy balance are performed. A comparative LCA is applied to both processes determining the inventory analysis and impact assessment.

Raw Materials and Manufacturing Processes
The raw materials used are the main difference between traditional and alternative manufacturing processes, that is, the use of petroleum coke (petcoke) or metallurgical coke (metcoke) in the traditional process and torrefied biomass in the alternative process. In stone wool the main oxides are silicon dioxide and oxides of alkali earth metals (predominantly calcium and magnesium). Silicon dioxide is principally derived from basalt and blast furnace slag. These raw materials are used in both processes studied. Alkali earth metal oxides are derived from the briquetted recycled material. In traditional briquettes, cement is used as a binder, while in the alternative process, cement is replaced by sodium silicate ( Table 1). The composition of traditional raw materials is obtained from the Integrated Environmental Authorisation of the company Rockwool Peninsular 2005 [19]. The amount of biomass needed is greater (215 kg) than petcoke (155 kg) and metcoke (167.8 kg) due to a lower heat capacity of the biomass (5618 kcal/kg) versus petcoke (7792.3 kcal/kg) and metcoke (7200 kcal/kg). The manufacturing steps, equipment and energetic requirements are the same for the three raw materials.
The process for production of stone wool comprises melting, fiberization, polymerization, cooling, product finishing and gas treatment. A detailed process flow diagram (PFD), built in Aspen software, can be found in Fig. 1. This PFD is the same for traditional and alternative process with only differences in the inputs (selected raw materials) and outputs (emissions) variables from the mass balance analysis.
The stone wool production in the blast oven includes coke, which is used for heating and melting rocks, melting the raw materials and additives (Table 1) and forming fibre on rotating wheels under the influence of a powerful airflow. The product is cured in a polymerization chamber at 200°C and after cooling, the stone wool is cut to the desired dimensions and packaged in polyethylene foil. The flue gas treatment system includes cooling and particle separation before burner systems. Off-cuts and other mineral wool scraps are recycled back into the production process, which further reduces the inputs and energy requirements. The significant impact of the coke properties and reactivity on the cupola operation during stone wool production has been reported [20]. With a more reactive coke less heat is lost to the cooling water, and therefore, coke can be saved. The coke reactivity must be determined before each specific application. The reactivity of metallurgical coke is slightly lower than that of petroleum coke [21]. However, petcoke has a higher sulphur content than metcoke ( Table 7, Appendix). On the other hand, sodium silicate is widely used as a raw material in alternative inorganic   thermal insulation material and is fundamental in geopolymer technology [22,23]. The use of torrefied biomass and sodium silicate as alternative materials in the present study has been shown to reduce the carbon and sulphur content of the raw materials and reduce outflow gas emission of carbon dioxide and sulphur dioxide (Table 1).  Fig. 4 Characterization of the Eco-Indicator99 impacts at different stages of the life cycle for the traditional stone wool manufacturing process using metcoke life cycle inventory data base. The CML 2000 and Eco-Indicator99 methods were used to estimate effects on different impact categories. According to the standards, LCA methodology is divided into four steps.

Goal and Scope Definition
The objective of the study is to evaluate and compare the environmental impacts from the production of stone wool using traditional and alternative manufacturing processes.
The functional unit selected for this analysis is 1 tonne of final finished product.

System Boundaries
The system boundaries determine which unit processes should be included in the LCA study. Defining system boundaries is partly based on a subjective choice that is made during the scope phase when the boundaries are initially set. Figure 2 shows the different steps of the life cycle of the traditional and alternative stone wool products. In this figure, the production, use and end of life stages are represented. Only the raw materials used and the production of the stone wool changes between the two processes, because the products obtained in both processes have the same technical and environmental properties. Thus, this work limited the application of life cycle analysis to the extraction of raw materials and industrial production of stone wool. In the studied industrial process, coke (petcoke and metcoke), biomass, raw materials consumptions and the gas emissions, are taken into account.

Inventory Analysis
The life cycle inventory involves the collection of the necessary data using specific methods. These data were then analysed comparatively with studies from the literature and software databases, involving materials, energy and fuels. Each stage in the stone wool manufacturing process is analysed. These production stages are melting of  [19]. Air emission data are the emission limit values authorized to this company for the minority compounds studied; for carbon dioxide and sulphur oxides, the data are the stoichiometric quantities assuming that all of the carbon and sulphur from the raw materials (coke, biomass and cement) completely react. The inventory generated for the life cycle study is shown in Tables 1, 2, 3, 4, 5 and ''Appendix Table 7''. These tables show the input and output materials and the gas emissions for the traditional stone wool manufacturing process using petcoke and metcoke and for the alternative process studied. The last column of the inventory tables refers to the nomenclature used by the software to introduce the inputs and outputs. It can be seen that the inputs are materials (products and wastes) and the outputs are gaseous emissions. All products used as raw materials have a Simapro reference to account for the impacts related to its production. In Table 1, the collected inventory data for melting stage for both processes of manufacturing stone wool are shown. Table 2 refers to the inventory data for the fiberization stage where the inflows are similar for traditional and alternative processes. The inventory data for the polymerization stage is shown in Table 3. This step is entirely controlled by outputs in the form of emissions to the atmosphere, which are identical in both traditional and alternative processes. The inventory data for the cooling stage are summarized in Table 4, showing no differences between the outflows (gas emissions) of the studied processes. Finally, Table 5 gives the inventory data for the product finishing consisting of particulate emissions generated due to the cutting process of the final product.
In the ''Appendix'', the specific composition of the materials that were used to calculate the stoichiometric gaseous emissions, and the composition of the different waste flows used as raw materials are compiled in ''Appendix Tables 7  and 8

Impact Assessment
The purpose of the life cycle impact assessment is to evaluate the quantity and significance of the potential environmental impacts of a defined system throughout its whole life cycle. For this study, LCA is conducted based on the cradle to gate approach, including the raw materials extraction. The study begins with the input of materials to the production system and ends with the product output of the system (Fig. 2). The methods applied for the impact assessment are twofold: (1) the CML 2000 method, developed using the mid-point approach, which is widely used in the construction sector that assesses the effect on ten categories of impact, and (2) the Eco-Indicator 99 method, developed using the end-point approach, which assesses the effect on eleven categories of impact and is

Results and Discussion
For the traditional process that uses metcoke as a raw material, and using the CML2000 method, the results obtained show that all effects for all categories are not good for the environment (Fig. 3). The stages that generates the greatest impacts in most of the impact categories analysed are the melting and fiberization stages. This is because during these steps most of the materials used in the process are introduced and involve a greater extraction of natural resources, which has a greater impact on the Abiotic depletion category. Furthermore, coke and cement are fed  ) and Marine aquatic ecotoxicity (60.1 %) categories, mainly due to the phenolic resin introduced at this stage as an additive. Eutrophication, Terrestrial ecotoxicity and Photochemical oxidation, are impacts that come from more than two process stages. The impacts obtained with Eco-Indicator99 are also not good for the environment for all categories (Fig. 4). Melting and the fiberization stages generates the greatest effects in most of the impact categories analysed. Melting has an effect particularly in the Minerals (98.6 %), Land use (97.7 %), Climate change (92.6 %), Fossil fuels (69.6 %) and Ecotoxicity (63.9 %) categories, while fiberization has the greatest effect on the Radiation (70.9 %), Ozone layer (68.5 %) and Carcinogens (63.0 %) categories. Respiratory organics, Respiratory inorganics and Acidification/Eutrophication categories come from all stages of the production process.
For the damage assessment based on areas of protection from Eco-Indicator99, all of the effects are negative for the environment (Fig. 5). The stage with the greatest negative effects is the melting stage, reaching 77.7 % of the impact in Resources; this process stage consumes most of the raw materials used in manufacturing and is the main stage generating flue gas emissions. Thus, in the Human Health and Ecosystem quality areas of protection, this stage produces the greatest impacts, followed by the fiberization stage.
The results of the traditional process using petcoke do not show many differences (between 0 and 15 %) in any impact category compared to those for the process using metcoke; however, the melting and fiberization process stages interchange as the main responsible stages for Abiotic depletion and Ozone layer depletion in CML2000 and Fossil fuels in Eco-Indocator99. The impacts remain negative in the same studied categories of impact; the melting and the fiberization stages are still responsible the greatest impacts in different categories for the CML2000, Eco-Indicator 99 and Areas of Protection, as discussed in the methodologies (Figs. 12, 13 and 14 of the ''Appendix'').
Assessment of the alternative stone wool manufacturing process with the CML2000 method shows that the fiberization stage dominates the negative effects generated in most of the impact categories analysed (Fig. 6). The effects in impact categories, such as Abiotic depletion and Human toxicity, are mainly generated at this stage, with contributions of 76.0 and 75.1 %, respectively. The melting stage has an important impact contribution to the Global warming category (81.3 %).
The melting stage also generates the greatest effects in four impact categories, Ecotoxicity (72.0 %), Climate change (81.6 %), Land use (97.8 %) and Minerals (98.6 %) (Fig. 7), when analysed by the Eco-indicator99 method. In addition to the melting stage, the fiberization stage is important in the Respiratory organics, Fossil fuels, Radiation and Ozone layer categories, with contributions of 86.5, 78.2, 66.2 and 62.5 %, respectively. The impact category Carcinogens is equally divided between the melting and fiberization stages. However, all of the process stages contribute effects in the Respiratory inorganics and Acidification/Eutrophication impact categories. The finishing stage only contributes to one category of impact (Respiratory inorganics) with a low value of 0.56 %.
The four production stages dominate the effects in the Human health and Ecosystem Quality areas, while only the melting and fiberization production stages have influences in the Resources impact area (Fig. 8).
The comparative LCA results of both studied processes are shown in Figs. 9, 10, 11 and Table 6. Comparing these results shows that the use of metallurgical coke (S: 0.7 %) and petroleum coke (S: 2.8 %) in the traditional stone wool production process and the alternative production process have similar effects on the categories Eutrophication, Human toxicity and Terrestrial ecotoxicity. However, the alternative process decreases the effects in the Photochemical oxidation, Acidification, Global warning and Abiotic depletion impacts by between 24 and 61 % for metcoke and increases Ozone layer depletion, Marine aquatic ecotoxicity and Fresh water aquatic ecotoxicity between 5 and 21 % with the CML2000 method. The effects on Minerals and land use impacts are similar in all processes when analysed using the Eco-Indicator99 method. The alternative process decreases the effects on Acidification/Eutrophication, Respiratory organics, Respiratory inorganics, Climate change and Fossil fuels impacts studied between 14 and 61 % for metcoke, while the effects on Ozone layer, Radiation, Carcinogens and Ecotoxicity impacts increase between 4 and 22 % for metcoke. The alternative process reduces all of the areas of protection by between 6 and 44 % ( Table 6) as studied by Eco-Indicator99. The impact reductions obtained using the alternative process are more significant when compared to using petroleum coke instead of metallurgical coke.

Conclusions
The melting stage, which occurs in a blast cupola furnace and the fiberization stage are the most intensive steps in the stone wool manufacturing process due to resources extraction and consumption and pollutant emissions. Minimization of the environmental impact of the final product can be promoted by a combined strategy of material recycling and selection of raw materials with a low sulphur content. The alternative rock wool manufacturing process, in which torrefied biomass is used in place of coke and sodium silicate is used instead of cement, is able to reduce both the emissions of CO 2 and SO 2 . The impact categories Minerals and Land use in Eco-Indica-tor99 and the Eutrophication impact in CML2000 increase between 2 and 4 % for the alternative process compared to the traditional process. However, with the use of the alternative process, the ecotoxicity-related impacts increase between 9 and 24 %. These increases are compensated for by decreases in other impact categories; in consequence, the three impact areas composed of individual Eco-indicator 99 impacts show environmental benefits between 6 and 15 % when using the alternative process with torrefied biomass, instead of the traditional process based on coke use. The modeling, simulation and optimization of the stone wool manufacturing process, and the use of real emission data for both traditional and alternative processes, make the LCA results more representative of real life cases, and are a fruitful direction for future work.   Characterization of the Impacts for Traditional Stone Wool Manufacturing Process Using Petcoke The following figures are the result of analyzing the traditional process of manufacturing stone wool with CML200, Eco-Indicador99 and damage assessment methods (Figs. 12, 13, 14).