Use of MRF residue as alternative fuel in cement production

Waste Management xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Waste Management journal homepage: www.elsevier.com/locate/wasman

Use of MRF residue as alternative fuel in cement production John R. Fyffe, Alex C. Breckel, Aaron K. Townsend, Michael E. Webber ⇑ Department of Mechanical Engineering, The University of Texas at Austin, 1 University Station C2200, Austin, TX 78705, United States

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: Solid recovery fuel Refuse derived fuel Waste-to-energy Waste management Recycling Energy recovery

a b s t r a c t Single-stream recycling has helped divert millions of metric tons of waste from landfills in the U.S., where recycling rates for municipal solid waste are currently over 30%. However, material recovery facilities (MRFs) that sort the municipal recycled streams do not recover 100% of the incoming material. Consequently, they landfill between 5% and 15% of total processed material as residue. This residue is primarily composed of high-energy-content non-recycled plastics and fiber. One possible end-of-life solution for these energy-dense materials is to process the residue into Solid Recovered Fuel (SRF) that can be used as an alternative energy resource capable of replacing or supplementing fuel resources such as coal, natural gas, petroleum coke, or biomass in many industrial and power production processes. This report addresses the energetic and environmental benefits and trade-offs of converting non-recycled post-consumer plastics and fiber derived from MRF residue streams into SRF for use in a cement kiln. An experimental test burn of 118 Mg of SRF in the precalciner portion of the cement kiln was conducted. The SRF was a blend of 60% MRF residue and 40% post-industrial waste products producing an estimated 60% plastic and 40% fibrous material mixture. The SRF was fed into the kiln at 0.9 Mg/h for 24 h and then 1.8 Mg/h for the following 48 h. The emissions data recorded in the experimental test burn were used to perform the life-cycle analysis portion of this study. The analysis included the following steps: transportation, landfill, processing and fuel combustion at the cement kiln. The energy use and emissions at each step is tracked for the two cases: (1) The Reference Case, where MRF residue is disposed of in a landfill and the cement kiln uses coal as its fuel source, and (2) The SRF Case, in which MRF residue is processed into SRF and used to offset some portion of coal use at the cement kiln. The experimental test burn and accompanying analysis indicate that using MRF residue to produce SRF for use in cement kilns is likely an advantageous alternative to disposal of the residue in landfills. The use of SRF can offset fossil fuel use, reduce CO2 emissions, and divert energy-dense materials away from landfills. For this test-case, the use of SRF offset between 7700 and 8700 Mg of coal use, reduced CO2 emissions by at least 1.4%, and diverted over 7950 Mg of energy-dense materials away from landfills. In addition, emissions were reduced by at least 19% for SO2, while NOX emissions increased by between 16% and 24%. Changes in emissions of particulate matter, mercury, hydrogen chloride, and total-hydrocarbons were all less than plus or minus 2.2%, however these emissions were not measured at the cement kiln. Co-location of MRFs, SRF production facilities, and landfills can increase the benefits of SRF use even further by reducing transportation requirements. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Managing our waste has been an issue for thousands of years affecting sanitation and the use of land, water and natural resources. Society’s municipal solid waste (MSW) management solutions have evolved significantly over time due to policy, public opinion, economics, and technological capability (Rada, 2014). In light of the inherent complexities of handling our waste, the EPA

⇑ Corresponding author.

has developed a waste management hierarchy to describe its regulatory policy preference for solid waste management in an effort to minimize environmental harm (U.S. EPA, 2011a). Under this framework, hierarchical preference is given to waste management options that conserve resources, recycle material and recover energy as alternatives to simply landfilling the waste. Preferentially, as much waste as possible is ‘‘moved up’’ the hierarchy to more useful and benign end-of-life pathways. Single-stream residential recycling, where all recyclables are commingled and collected from a single bin and sorted at a centrally located materials recovery facility (MRF), has become

http://dx.doi.org/10.1016/j.wasman.2015.05.038 0956-053X/Ó 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

2

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

increasingly common waste management solution in the United States. The move to single-stream has increased recycling rates but has also led to increased contamination rates compared to content-separated recycling where consumers separated items before collection. As a result, MRF operators often dispose of a larger portion of incoming materials than comparable mixed stream facilities (Mueller, 2013). This discarded material, called residue, is currently landfilled, despite the fact that it is primarily composed of plastics and fibrous1 materials with an energy density comparable to coal. According to the EPA’s waste management hierarchy, recovering the energy of this residue is preferential to simply landfilling (U.S. EPA, 2011a). The focus of this research is to determine the energetic and environmental tradeoffs of converting this energy-dense material into solid recovery fuel (SRF), which can be used in the form of fuel pellets to replace fossil fuels at a cement kiln. Cement kilns offer a flexible means for testing SRF because of their high energy demand and fuel flexibility as well as the benefit of incorporating the ash from fuels in the product, requiring no ash disposal processes (Sarc, 2014). The tradeoffs are non-obvious due to the need to transport typically low density waste materials to and from SRF production facilities and the electricity requirements to produce the SRF. Additionally, the heterogeneity of the MRF residue makes it difficult to predict emissions without accurate characterization. The methodology was developed specifically to highlight process and logistical barriers to SRF adoption, determine the amount of fossil fuels displaced when using SRF, amount of material displaced from the landfill, and the net change in emissions of seven primary air emissions (NOX, SO2, CO2, PM, THC, Hg, and HCl). This analysis was conducted in two parts; first through an experimental test burn of 118 Mg of SRF in a cement kiln, and second through a systems analysis considering several possible scenarios while using emissions production data from the experimental section. 2. SRF experimental analysis The experimental approach was to obtain MRF residue and ship it to an SRF production facility to be processed into SRF. Lab tests were performed to characterize the feedstock residue and SRF, and determine if the fuel met cement kiln specifications. At the cement plant, the SRF was introduced to the precalciner tower of the cement kiln (see Fig. 1) immediately after baseline data was taken. During the test burn NOX and SO2 emissions were continuously monitored. The amount and type of other fuels used and internal measurements indicating combustion characteristics were also tracked during the experiment. 2.1. SRF production and testing To reduce the variability of the fuel and to protect the processing equipment at the SRF facility, 68 Mg of MRF residue from a commercial MRF was blended with 50 Mg of industrial waste of known composition to produce the SRF, yielding a mixture of approximately 60% plastics and 40% paper scrap. The plastic component of the industrial waste is commonly used at the SRF facility and is predominantly polyethylene and polypropylene, similar to typical MRF residue composition (Thorington, 2011). The paper scrap consisted of primarily cardboard and shredded fiber based on observation. While mixing well-defined industrial waste with the MRF residue makes it more difficult to analyze SRF from MRF residue directly, it was determined that the uncertainties introduced would not outweigh the benefit of being able to produce 1

Fiber refers to paper, paperboard, cardboard, etc.

the SRF at the commercial SRF facility (which would not otherwise have been possible). Additionally, mixing of industrial wastes with MRF residue is likely a stepping stone toward fully MRF residue SRF use. Elemental analyses of multiple SRF samples were conducted to determine the amount of heavy metals in each sample. Three bulk samples (around 4 L each) of the fuel were analyzed as well as four small samples (around a half quart each) from the first bulk sample. The tests showed: (1) results varied drastically sample-to-sample, (2) levels higher than allowable of cadmium and antimony appeared in one of the small samples each, and (3) all three bulk SRF samples met the cement kiln’s fuel specifications for energy density, moisture content, and composition. The composition requirements are a combination of environmental (Hg, Pb, etc.) and process (S, Na, K, etc.) limits. The elemental results are discussed in greater detail in a previous report (Fyffe et al., 2012). Separate samples of the SRF delivered to the cement kiln were sent to an outside laboratory for further testing. Standard American Society for Testing and Materials (ASTM) tests for solid fuels were used to characterize the SRF. CEN TC 343 is the European Committee for Standardization’s set of methods, guidelines, and specifications. Standardized tests ensure consistency when comparing SRF to coal and other fuels (Table 1). While our SRF results are not expected to be representative of SRF in general, the test results align with previous studies of residue-derived fuel as seen in Table 1 (Thorington, 2011; U.S. EPA, 2011b; Gera and Gautam, 1993; Buah et al., 2007; Velis et al., 2012). A comparison of the SRF characteristics with sub-bituminous coal, petroleum coke, wood, and tire chips, which are common cement kiln fuels, is presented in Table 2. The SRF tested has an energy content on par with several traditional secondary cement kiln fuels, but pollutants are a mixed story. 2.2. Experimental test burn The 118 Mg of SRF were produced and delivered to a relatively modern cement kiln with a precalciner tower (discussion of production methods can be found in Fyffe et al., 2012). The cement kiln operators introduced the SRF to the precalciner tower via an existing conveyor designed to deliver solid fuels such as wood or tire chips. Solid fuels are typically used in the precalciner tower instead of the kiln because the kiln requires a more precisely controlled temperature profile and therefore a more precise fuel injection system. The SRF was fed and combusted steadily at 0.9 Mg/h for 24 h and at 1.8 Mg/h for 48 h. The experiment was conducted between 11/13/2011 and 11/17/2011. The average energy content of SRF was calculated by performing an energy balance on the remaining amount of other fuels being consumed. This calculation was done under the assumption that the cement kiln was operating at a constant energy requirement, which is reasonable because the energy intensity of the cement manufacturing process is not a function of the fuel type. The plant was burning coal and natural gas at the main burner of the kiln with natural gas and some liquid fuels in the precalciner tower. While the liquid fuel was not disclosed for competitive reasons, it was given that a liter of the liquid fuel contained the energy equivalent of 43 m3 of natural gas. The precalciner is a special combustion chamber for cement kilns that serves to pre-heat and decarbonate raw materials before entering the kiln. The plant was operating at roughly 80% capacity during the test, which provided additional flexibility for various process related systems that might not otherwise have been available. During the test burn the SRF averaged a contribution of 24.5 GJ/Mg during the 0.9 Mg/h test and 29.7 GJ/Mg during the 1.8 Mg/h testing period, which agrees well with the laboratory SRF test results of 28.3 GJ/Mg. The variability is attributed to the

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

3

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

Fig. 1. The precalciner tower is used to preheat and decarbonate the incoming raw materials. They increase the overall energy efficiency by lowering the waste heat and also allow for the use of smaller kilns. Roughly 55% of the total energy is consumed in the precalciner tower (Mueller, 2013).

Table 1 Summary of SRF sample analysis results compared to previously published data on residue derived fuels (Gera and Gautam, 1993; Buah et al., 2007; Velis et al., 2012). Parameter

Test method

Experimental

Literature

Calorific value Ash content Sulfur content Moisture Chlorine content

ASTM ASTM ASTM ASTM ASTM

28.28 5.02 0.07 4.27 0.05

11.55–26.06 1.55–22.00 0.20–0.74 4–21

D5865 D7582 D4239 D7582 D7359

GJ/Mg wt % wt % wt % wt %

Table 2 Comparison of cement manufacturer fuel specifications to several alternative kiln fuels (ICF International, 2008; SWANA, 2002; Serre and Lee, 2009; Conesa et al., 2008; Wang et al., 2004; Tillman, 2008).

Fuel specifications Tested SRF Sub-bit. coal Pet. coke Wood chips Tire chips

LHV (Btu/lb)

LHV (GJ/Mg)

Ash (wt%)

Sulfur (wt%)

Chlorine (wt%)

8000 12,150 11,300 14,600 7300 14,000

18.6 28.28 26.3 34.0 17.0 32.6

N/A 5.02 4.04 0.48 2.82 1.61

2.00 0.07 0.25 4.00 0.07 2.25

0.05 0.05 0.04 0.00 0.09 0.50

heterogeneous nature of the SRF, the imprecise SRF feeding process, variable moisture content, and the effect of changing fuels and production levels by the cement kiln. The actual variability in the SRF itself is therefore expected to be lower than the net variability seen in the test burn results. The SO2 and NOX emissions averages are shown in Table 3 and are compared to the cement kiln’s reference data taken over the 2 days prior to the experiment and the permitted levels of emissions (TCEQ, 2009). The expected reduction of SO2 emissions because of the low sulfur SRF was confirmed by the test burn where the SO2 emissions were reduced by 47% during the entire testing period. The reason for the increase in NOX emissions is not certain, but is hypothesized to be from the increased generation of thermal NOX due to local high-temperature zones near where the SRF was inserted into the precalciner tower. If the

Table 3 Table of averaged NOX and SO2 emissions from the cement kiln before (Reference Case) and during SRF use (SRF Case) compared to the cement kiln’s permitted levels. Compared to the Reference Case, SO2 decreases while NOX increases but all emissions remain below permitted levels.

Reference Case 0.9 Mg/h SRF Case 1.8 Mg/h SRF Case Permitted

SO2 (g/Mg clinker)

NOX (g/Mg clinker)

8.5 4.5 4.5 40.0

505 630 980 1300

increase in NOX emissions is from thermal NOX generation, they could be alleviated with SRF feed systems designed to reduce high-temperature zones that can lead to NOX formation (TCEQ, 2009). 3. Systems analysis An analytical framework was constructed to compare two primary cases: (1) the Reference Case and (2) the SRF Case. The Reference Case assumes the only fuel consumed at the cement kiln is coal and that all MRF residue is landfilled. In the SRF Case MRF residue is converted to solid recovery fuel and used to offset some coal use at the cement kiln. The type of coal used at the cement kiln is accounted for in the emissions calculations in Section 3.5.5. In both cases, coal is assumed to be the only non-SRF fuel used at the cement kiln. The two cases compare energy consumption, emissions, and landfill avoidance to provide an understanding of the various trade-offs. Fig. 2 shows the overall pathways for the two cases, tracking the various material, energy, and emission flows through the control boundaries indicated for this framework. 3.1. System boundaries and flows The system boundaries in Fig. 2 were selected to include the energy, emissions and material flows of interest in this research. The system boundary was drawn to exclude any processes upstream of the MRF residue being delivered to the SRF production

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

4

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

Fig. 2. Two scenarios are considered: (1) the Reference Case, with material recovery facility (MRF) residue being disposed of in a landfill and the cement kiln using fuel such as coal or natural gas (top panel) and (2) the Solid Recovered Fuel Case (SRF Case), where MRF residue is processed by the Solid Recovered Fuel Facility (SRFF) into SRF that is then used to offset fuel use at the cement kiln and to avoid landfilling the residue (bottom panel).

facility under the assumption that the MRF does not operate any differently in the proposed alternative pathway. The transportation of the MRF residue, SRF, and non-SRF kiln fuel (for solid fuels only) were included in the system boundaries. The transportation of these components add to the emissions and energy consumption of the different scenarios depending on the location of the MRF, SRF facility (SRFF), cement kiln, and fuel supply. Including these internal transportation steps in the energy and emissions costs elucidates the benefits of co-location. The SRF facility in this analysis uses electricity to process the MRF residue into SRF. The electricity used at the SRF facility must be included in order to capture the effect of that energy consumption to produce SRF from the raw MRF residue. The source of electricity varies spatially and temporally, which affects the amount of fuel consumed and emissions produced. The spatial variation in electricity related emissions and fuel consumption is captured in the analysis by considering a few representative electric grids that emulate broad regions of the country with different fuel consumption and emission rates. The temporal effect on electricity generation is not included in this study. This assumption is not expected to add significant errors because the SRF facility is assumed to operate throughout the year, averaging out temporal variation in fuel consumption or emissions. Energy flows into the system are tracked by type to tally fuel consumed at each step. The final analysis tracks the total amount of non-SRF fuel consumed in each case for all scenarios. Any renewable energy accounted for in the analysis, primarily hydro and wind used to generate electricity, was converted from electricity produced in kWh to GJ by a direct conversion factor. The SRF is not included in fuel flows in either case. The energy contained within the residue could be added to the primary energy accounting, but would be equal in both cases and is left off for clarity and to highlight the energy consumption that varies between the two cases studied.

pathway with varying degrees of SRF adoption and infrastructure build-out. These scenarios were designed to roughly mimic the temporal advancement as the industry matures. The variables of interest in each scenario are: (1) the distances traveled for each transportation step, (2) the electric grid used to power the SRF facility, (3) the landfill recovery efficiency,2 and (4) the use of an internal-combustion engine to provide energy from captured landfill gas (LFG-to-energy). The scenarios are developed in a general way to highlight stark differences in possible implementations of SRF use. For all scenarios the analysis was conducted with the assumption that coal was replaced by SRF at a cement kiln at a rate of 0.9 Mg/h over a one year period. 3.2.1. Early Scenario The Early Scenario is used to approximate the experimental SRF test burn method and an early stage SRF infrastructure. The inputs for the Early Scenario as well as the other two scenarios, Near-Term Scenario and Future Scenario, are shown in Table 4. In the Early Scenario the distances traveled are very high indicating an immature infrastructure for SRF use and accurately approximates the experimental conditions of the study, for which the cement kiln was in central Texas, the MRF residue was from New England, and the SRFF was located in Arkansas. The distance between the SRF facility and the MRF to a landfill approximates the distance from a central metropolitan area to a remotely located landfill, rounded to a reasonable distance that is consistent with the experiment. The distance from the MRF to the SRF facility and between the SRF facility and the cement kiln are similar to the distances in the experimental test. Finally, the distance for the cement kiln fuel was chosen to approximate the use of Colorado coal at a central Texas cement kiln, again, similar to that of the experimental test. The transportation method was set as diesel truck for all shipping except for the fuel being sent to the cement kiln, which was designated as diesel rail. These transportation methods are the same in the other two scenarios as well. The

3.2. Analysis Scenarios Three different scenarios were developed to study the energetic and environmental trade-offs associated with the proposed SRF

2 Landfill recovery efficiency refers to the percentage of methane generated that is captured and either used for LFG-to-energy or simply flared. The remaining methane is assumed to have escaped from the landfill.

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx Table 4 The inputs for each scenario consist of transportation distances and methods, grid characteristics, landfill gas recovery efficiencies, and landfill gas-to-energy use. The Early Scenario is representative of today’s situation, where the MRF and SRFF are separated by 2400+ kilometers, which we expect to decrease significantly in future commercial operations where the MRF and SRFF would be nearby or co-located to reduce superfluous costs. Distance/parameter

MRF to landfill MRF to SRFF SRFF to landfill SRFF to kiln Kiln fuel to kiln Recovery efficiency LFG-to-energy U.S. electric grid

Scenario

Units

Early

Near term

Future

81 2414 81 885 2092 55 No MRO

81 322 81 81 322 75 No U.S.

0 0 0 81 322 90 Yes WECC

km km km km km %

Midwest Reliability Organization (MRO) grid, located in the mid-west US, was chosen because of its high coal and low renewable components, a worst case scenario for electricity related emissions and fossil fuel consumption. The landfill recovery efficiency was chosen on the low end of the typical range for landfills in the U.S. (typically between 55% and 90%, dependent upon landfill construction and waste composition) (Townsend and Webber, 2012; Cruz and Barlaz, 2010). Finally, landfill gas-to-energy (or LFG-to-energy3) was not utilized in this scenario. 3.2.2. Near-Term Scenario The Near-Term Scenario is developed with moderate considerations for energy efficiency, SRF infrastructure, and environmental impact in mind. In the Near-Term Scenario the distances traveled are reduced for the MRF to SRF facility and the SRFF to the cement kiln, representing a distance from one major metropolitan area to another. The distance to the landfill was kept the same as in the Early Scenario. The fuel transportation distance to the cement kiln was reduced, representing an in-state fuel supply. U.S. average grid data were used as a moderate example of an electricity production fuel-mix. The landfill recovery efficiency was increased to a typically used average value of 75% (Townsend and Webber, 2012; Cruz and Barlaz, 2010). No landfill gas-to-energy was considered. 3.2.3. Future Scenario The Future scenario was chosen to be as energetically and environmentally advantageous as possible within the realm of the current assumptions. The MRF, SRF facility, and landfill are assumed to be co-located and therefore there is no transportation required between any of these facilities. The distance between the co-located facilities and the cement kiln was maintained as it is expected the co-located facilities and a cement kiln are both in rural environments but not necessarily also co-located. The fuel transportation distance to the cement kiln was kept the same as in the Near-Term Scenario. The Western Electricity Coordinating Council (WECC, west coast of the US) grid data were used due to WECC’s low coal and high renewable fractions. The landfill recovery efficiency was set at the highest level and landfill gas-to-energy was used in this scenario. 3.2.4. Additional scenario assumptions A few parameters were kept constant throughout the scenarios but are important assumptions made in the analysis. The mass conversion rate of MRF residue to SRF was assumed to be 90% and the electricity required to produce 0.9 Mg of SRF was fixed at 3 LFG-to-energy refers to the process of using captured landfill methane to produce electricity on site.

5

110 kWh/Mg, based on the data from the SRF production facility used in this experiment (Fyffe, 2011; Fyffe et al., 2012). The transportation methods and their associated energy intensities and emission rates were held constant, though the efficiency and emissions of transportation are expected to improve in the future. The residue composition was assumed to be unchanged in all scenarios. 3.3. Cement kiln energy balance The energy requirement at the cement kiln plays a vital role in the overall energy balance of this analysis. The cement kiln energy requirement is the amount of energy the cement kiln requires to produce 1.0 Mg of clinker (the primary prerequisite for cement produced by the cement kiln). A value of 3.9 GJ/Mg of clinker was used for this analysis (Thorington, 2011). Depending on the cement kiln equipment or operational limitations and SRF specifications, there might be a maximum fractional amount of SRF that can be used to offset fossil fuels. Limiting factors could be process related such as not being able to control temperature precisely enough with solely SRF, emissions limits related, or equipment related due to a lack of infrastructure for using solely solid fuels. The cement kiln energy requirement is constant in both the Reference Case and the SRF Case because there is no expected change to the operation of the cement kiln due to a fuel switch (Thorington, 2011; Brooks, 1991). 3.4. Primary energy consumption The total primary energy consumed in the Reference Case is the sum of the transportation energies and the non-SRF energy used by the cement kiln over a one year study period. The two transportation steps are the shipping of fossil fuels to the cement kiln and shipping the MRF residue to the landfill. The total primary energy consumed in the SRF Case is the sum of the transportation energies, the total non-SRF energy used by the cement kiln, and the electricity related energy use to produce the SRF. The transportation steps are the shipping of fossil fuels to the cement kiln, shipping of MRF residue to the SRFF, shipping of SRF to the cement kiln, and shipping of SRF waste to the landfill. The amount of electricity used to produce SRF is a function of the SRFF process equipment, MRF residue composition, and the characteristics desired for the SRF product. The grid-specific method in this study can compare the energy use from the following categories: U.S. average, the WECC, and the MRO. These grids were chosen to represent a range of possible location-dependent electricity implications. The MRO grid represents a coal-heavy grid while the WECC grid has a significant fraction of hydro and little coal-based generation. The U.S. average is used as a moderate example. The resource mix for each of the grids studied is from eGRID data maintained by the EIA and EPA for the year 2007, the most recent full data set available at the time, shown below (U.S. EPA, 2011c). 3.5. Emissions The total amount of emissions avoided by using the SRF method is the difference between the total emissions in the SRF Case and the Reference Case. This comparison is done on a pollutant by pollutant basis for each pollutant generating step. The pollutants tracked in this analysis are carbon dioxide (CO2), sulfur dioxide (SO2), oxides of nitrogen (NOX), particulate matter (PM), total hydrocarbons (THC), mercury (Hg), and hydrogen chloride (HCl). The landfilling of waste and the combustion of fuels both produce harmful gaseous emissions. Landfill emissions occur in both the Reference Case and the SRF Case, though to differing

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

6

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

degrees due to partial landfill avoidance in the SRF Case. The landfill emissions are analyzed for each scenario using the same emission generation rate and landfill gas collection efficiency for consistency. Combustion emissions occur during fuel transportation, electricity generation, and cement production. The emissions from the electricity generation in the SRF Case are determined using average emission rates for the selected electricity grids of interest. The details for how these emission rates were calculated can be found in Fyffe et al. (2012). Cement kiln nitrogen oxide (NOX) and sulfur dioxide (SO2) emissions data were taken from on-site emissions monitoring systems before (Reference Case) and during (SRF Case) the SRF trial burn. Due to a lack of on-site measured data for the remaining pollutants, EPA average values of carbon dioxide (CO2), particulate matter (PM), total hydrocarbons (THC), mercury (Hg), and hydrogen chloride (HCl) production rates for cement kilns were used in conjunction with the measured NOX and SO2 data (U.S. EPA, 2011d). Therefore, further emissions monitoring at kilns using SRF is suggested for these pollutants. 3.5.1. Quantifying emissions The emission rates for each process must be determined in order to compare the total emissions in each case analyzed. Emission rates of SO2 and NOX from the cement kiln were experimentally measured. The remaining emission rates from the transportation steps, landfill, cement kiln, and electricity generation were found from public sources of emission rates and energy intensities. 3.5.2. Electricity emission rates Electricity is consumed in the SRF Facility in order to process the MRF residue into SRF. The source of the electricity is dependent on the location of the SRF Facility. The emissions from the electricity production vary on a diurnal and seasonal basis but, assuming the SRF facility continually produces SRF, yearly grid averages are useful representations. Electricity generated in Texas is produced by burning a different mixture of fuels than electricity generated in California or Pennsylvania, for example. Even if the fuel mixture was the same at two different locations, different climates and different power plant efficiencies could lead to different fuel consumptions and emissions for every kilowatt-hour generated. For this study a range of emission rates are developed for the various grids chosen. Emission rates for carbon dioxide (CO2), nitrogen oxides (NOX), sulfur oxides (SO2), particulate matter (PM), total hydrocarbons (THC), mercury (Hg), and hydrogen chloride (HCl) are used to maintain consistency with emission rates from the cement kiln and transportation methods. Table 5 shows the emission rates calculated for the U.S. Average, WECC, MRO, and ERCOT. The values in Table 5 are projected emissions rates for 2011 using historical data from the EPA (U.S. EPA, 2007a, 2007b, 2011c, 2011d). 3.5.3. Landfill emission rates In both cases, some material is disposed of in a landfill, leading to emissions from the decomposition of the waste. The primary air emission from landfills is methane (CH4) and is the only one Table 5 Emission rates (per MWh) for electricity generated in different regions of the United States (U.S. EPA, 2011c, 2011d, 2007a, 2007b).

U.S. Average WECC MRO ERCOT

CO2 (kg)

NOX (g)

SO2 (g)

PM (g)

THC (g)

Hg (lg)

HCl (g)

540 386 712 467

635 499 907 181

3040 408 2090 771

141 68 132 77

10.9 10.4 15.4 9.1

9.0 4.1 15.9 13.6

23.6 1.1 10.9 5.9

considered in this analysis. The methane produced in the landfill could either escape directly into the atmosphere or be captured and converted. The captured methane is either flared and converted directly to CO2 with no beneficial product or used in a combustion application to produce additional electricity. The method of using the methane to produce electricity is typically referred to as LFG-to-energy.4 The flaring of the methane is done for safety reasons and because methane has a global warming potential (GWP) of 25, meaning methane is about 25 times as potent a GHG as CO2 over a 100-year period (Forster and Ramaswamy, 2007). The amount of methane emitted from a landfill in this analysis is a function of the amount of waste and the landfill gas capture efficiency. Landfill emissions are a function of composition of the waste as well, but this level of detail is outside the scope of this paper as the MRF residue composition not used in the SRF was not analyzed nor is it expected to be constant over time. 3.5.4. Transportation emission rates The energy intensities of some major transportation methods were taken from the NREL Life-Cycle Inventory (LCI) Transportation Data and are shown in Table 6. The emission rates for the transportation methods were also gathered from the NREL LCI and are shown below in the emission rates section (Section 3.5.4) (NREL, 2012). Of the 7 pollutants tracked in this study, only Hg and HCl were not accounted for by NREL. We expect these to be minimal, and are neglected in this portion of the analysis. 3.5.5. Cement kiln emission rates The cement kiln emission rates (Table 7) were derived from two primary sources: (1) experimental results from the SRF test burn and (2) EPA emission rate data for cement kilns similar to the one used in the SRF test burn. Because emissions of PM, THC, Hg, and HCl were not measured during the test burn, these emissions rates are assumed to be equal in both the Reference Case and the SRF Case. Further studies regarding the emission of these pollutants from the combustion of SRF at cement kilns is therefore recommended. The CO2 emission rate for the cement kiln in Table 7 was adjusted to account for the fuel switch between coal and SRF between the two cases considered. In order to do this, EPA data for the CO2 emission rate for coal, refuse derived fuel, and similar cement kilns was used. The Reference Case CO2 emission rate was chosen to be equal to the average cement kiln CO2 emission rate. The SRF Case CO2 emission rate was calculated by subtracting the CO2 emissions from the replaced coal and adding the CO2 emissions from the amount of SRF used. The CO2 emission rate for SRF from the EPA was 47.3 kg/GJ, based on data from a single refuse derived fuel combustion facility (U.S. EPA, 2011b). The CO2 emissions rate of 87.3 kg/GJ was used for coal. While the SRF value is not representative of all SRF, it is a figure supported by the U.S. EPA and fits with the understanding that plastic waste typically has lower carbon intensities than coal. The NOX and SO2 emissions are derived from the experimental data taken during the SRF test burn. Due to the limited duration and scope of the experimental analysis, all emissions results discussed here should be considered preliminary and require further testing to confirm. During the experimental test burn, liquid fuels were used at certain times adding complexity in determining the effect of SRF on the emissions of the cement kiln. For this reason, the SO2 emission rate was calculated using data during times with no liquid fuel use. The data showed an SO2 emission rate of 8.5 g SO2 per Mg of clinker produced (gSO2/Mgclinker) when no SRF was 4

LFG = landfill gas.

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

7

Table 6 Energy intensity values for various transportation methods from NREL Life-Cycle Inventory (NREL, 2012). Transportation method

Energy intensity (kJ/Mg-km)

Emission rate (kg/GJ) CO2

NOX

SO2

PM

THC

Train (diesel) Truck (diesel)

251 1052

75.5 76.0

1.98 0.52

.017 .017

.052 .009

.073 .025

Table 7 The cement kiln emission rates were calculated using the EPA and experimental data (U.S. EPA, 2011b). The experimental data was scaled from the 3 day experiment to a full year of operation combusting SRF at a rate of 0.91 Mg per hour. Pollutant

Reference Case

SRF Case

Units

CO2 NOX SO2 THC HCl PM Hg

816.5 458.1 7.8 54.4 22.2 21.8 99.8

807.4 571.5 4.0 54.4 22.2 21.8 99.8

Gg/year Mg/year Mg/year Mg/year Mg/year Mg/year kg/year

being used. This number decreased to 4.5 gSO2/Mgclinker when the SRF was being used at 0.9 and 1.8 Mg/h feed-rates. While these numbers are significantly lower than the EPA average for kilns of this type, the specific type of kiln fuel, raw materials, and efficiency of the pollution control systems can dramatically affect the SO2 emission rate. Therefore, the 8.5 gSO2/Mgclinker rate was used for the Reference Case and the 4.5 gSO2/Mgclinker value was used for the SRF Case. The NOX emission rate was derived in a similar way to the SO2 emission rate. Again, the times when liquid fuels were being used is neglected due to added complexity in determining the effect of SRF use. The NOX emissions average was 505 gNOX/Mgclinker during the time period before SRF was used, 630 gNOX/Mgclinker during the 0.9 Mg/h SRF test, and 980 gNOX/Mgclinker during the 1.8 Mg/h SRF test. It is possible that the method of injection, which had not been optimized, caused higher temperatures within localized zones of the precalciner tower, leading to higher thermal NOX formation rates. The amount of nitrogen in the fuel is unknown, so the possibility of the increase due to non-thermal NOX formation cannot be determined. 3.6. Analytical results The results from the analysis of the three scenarios described above are given here. The energetic trade-offs between the Reference Case and the SRF Case are compared for each of the scenarios considered. Additionally, the environmental trade-offs are analyzed, focusing on the seven species of air pollutants and the total mass of material landfilled. 3.6.1. Energy results The SRF Case reduced primary energy consumption across all scenarios examined versus the Reference Case. The total energy use, not including SRF energy, was reduced by 5.5% in the Early Scenario, 6.2% in the Near-Term Scenario, and 6.3% in the Future Scenario over a projected one year period, as seen in Fig. 3. Including the large transportation distances in the Early Scenario, the use of SRF still reduced the total energy consumption by 5.5%, or by 198 TJ over the course of a year. In the Near-Term Scenario and Future Scenario the reduction totaled 6.2% and 6.3%, or 219 and 222 TJ, respectively. The additional energy consumption from electricity is only a fraction of the total energy used in any of

Total Annual Non-SRF Energy Consumed (TJ)

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

3650 3600

Electricity

Reference Case

3550 3500

Transportaon Fuel

Non-SRF Kiln Fuel

Reference Case Reference Case

5.5%

6.2%

6.3%

3450 SRF Case

3400

3350

SRF Case

SRF Case

3300 3250 3200 3150

Early Scenario

Future Scenario

Near-Term Scenario

Fig. 3. The use of SRF offsets at least 5.5% of the total energy required in the Reference Case for all of the scenarios considered. Operating at one ton SRF per hour for a one year period.

the scenarios and is barely visible in Fig. 3. Table 8 shows the equivalent amount of coal, natural gas, or oil use that could be avoided using SRF under these assumptions. These results were for a 0.9 MgSRF/h feedrate (about 5% of a kilns total energy requirement) at one cement kiln. Therefore, these numbers could be scaled linearly for higher feedrates and across many cement kilns. 3.6.2. Emissions results Fig. 4 shows the emissions reductions (% basis) in the SRF Case compared to the Reference Case for PM, THC, Hg, and HCl. In both the Near-Term Scenario and Early Scenario the emissions changes are all below 1%. In the Future Scenario the PM emissions were reduced by 1.6% while the THC emissions were reduced by 2.2% in the SRF Case. However, the reduction in PM and THC primarily comes from the small decrease in rail transportation of fuel to the cement kiln, which in the Future Scenario was chosen to be the longest distance. Therefore, the transportation effect has only a minimal effect on the emissions in Fig. 4. As mentioned previously, the cement kiln emissions rate for PM and THC was kept constant for both the Reference Case and the SRF Case. While keeping these emissions rates constant was done because insufficient data were gathered, further investigation into the effect of using SRF to replace fossil fuels is recommended in regard to the emissions of PM, THC, CO2, HCl, and Hg at the cement kiln. The SO2 emissions reduction ranged from as low as 19% in the Near-Term Scenario to 44% in the Future Scenario, as seen in Fig. 5. The SO2 emission rate for the cement kiln is the driving factor in total SO2 emission reductions in the SRF Case. The reduction in SO2 emissions at the cement kiln in the SRF Case are somewhat offset by the increased SO2 emissions from transportation and electricity production. In the Near-Term Scenario the electricity production contributes roughly 20% of the total SO2 emissions in the SRF Case, lowering the SO2 emissions benefit to only 19%, while the transportation for both the Reference Case and the SRF

Table 8 The amount of fossil fuel avoided by using SRF under the given assumptions at a rate of 0.9 Mg of SRF per hour (or about 5% replacement) over a one year period. Fuel

Early Scenario

Near-Term Scenario

Future Scenario

Units

Coal Natural gas Oil

7780 5077 32,364

8615 5615 35,797

8705 5692 36,287

Mg Thousand m3 Barrels

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

8

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

PM

Hg

THC

Electricity

HCl

0.6%

0.04%

7 0.08%

0.1%

0.01%

0.00% 0.00%

0.01%

-1% -1% -2% -1.6%

6

20%

-2.2%

24%

Near-Term Scenario

4

3 2

Early Scenario

Future Scenario

Fig. 4. The percent reduction of emissions in the SRF Case is minimal for PM, THC, Hg, and HCl. Because the cement kiln emission rate of these pollutants was the same in both Cases, further investigation into the effect of SRF on these emissions is recommended.

SO2 Emisssions (Mg)

7

Electricity

LF Captured

Transportaon

Kiln

Reference Case

21% SRF Case

Reference Case

Reference Case

19% SRF Case

44%

6 5

SRF Case

4 3

2 1 0

16% SRF Case Reference Case

Reference Case

5

0

Early Scenario

Near-Term Scenario

Future Scenario

Fig. 5. The SO2 emissions are lower in all 3 scenarios, with reductions between 19% and 44%. The reduction comes from the reduced cement kiln emissions which are large enough to offset any increase due to electricity consumption or increased transportation requirements.

Case in the Near-Term Scenario contributes only a small fraction of the total SO2 emissions. The large SO2 emissions increase from the electricity in the Near-Term Scenario is due to the high SO2 emission rate of the U.S. average electricity production. The MRO electric grid has a slightly lower SO2 emission rate than the U.S. average, leading to a smaller component of SO2 emissions in the Early Scenario coming from electricity production. The Early Scenario SO2 emissions are much larger due to large transportation distances contributing a significant fraction of the total SO2 emissions for both the Reference and SRF Cases. The reduction of SO2 from the cement kiln is significant enough to result in a total reduction of 21% SO2 in the Early Scenario for the SRF Case compared to the Reference Case. The NOX emissions increased by 16% in the Future Scenario, 24% in the Near-Term Scenario, and 20% in the Early Scenario (Fig. 6). The NOX emissions for each scenario increased in the SRF Case compared to the Reference Case due to the previously mentioned result of increased NOX emissions at the cement kiln when using SRF (Table 3). In the Future Scenario some of the increases were offset by reductions in NOX emissions from the combustion of landfill gas because less landfill gas was created in the SRF Case.

Near-Term Scenario

Future Scenario

Fig. 6. Emissions of NOX increased slightly due to the higher emission rate of NOX at the cement kiln during the SRF test burn. Total NOX emissions remained below permitted levels.

This effect is not seen in the Near-Term Scenario or the Early Scenario because the landfill gas was flared in these cases, simply converting the captured landfill gas directly to CO2. The effect of transportation difference is minimal in all scenarios compared to the difference in NOX production at the cement kiln. However, about 11,800 Mg more NOX is emitted in the Early Scenario compared to the Future Scenario in both the Reference Case and the SRF Case, showing that co-location of facilities and the proximity of fuel to the cement kiln lead to large reductions in life-cycle emissions. The CO2 emissions in the SRF Case were lower in every scenario compared to the Reference Case, as seen in Fig. 7. There are CO2 reductions from the cement kiln, which is the largest overall effect, but a significant fraction of the reduction also comes from the reduced landfill mass. The effect of the landfill reduction is more pronounced, as expected, in the Near-Term Scenario and even more so in the Early Scenario. These more pronounced reductions in CO2 are directly related to the landfill gas recovery efficiencies of the landfill, revealing that the use of SRF is even more beneficial where landfills are poorly designed or maintained. The transportation component of the CO2 emissions is the greatest in the Early Scenario where the SRF Case has nearly 50% more CO2 from transportation than the Reference Case. Finally, the electricity

Electricity 835 830

CO2 Emissions (Gg)

Early Scenario

8

SRF Case

Reference Case

-3%

9

Kiln

1

-2%

10

Transportaon

SRF Case

0.4%

1%

0%

LF Captured

8

0.7%

NOX Emissions (hundred Mg)

Percent Change from Reference to SRF Case

1%

LF Captured

LF Leaked

Transportaon

Kiln

Reference Case

825 1.6%

Reference Case Reference Case

820 SRF Case

815

1.5% SRF Case

810

1.4% SRF Case

805 800 795

Early Scenario

Near-Term Scenario

Future Scenario

Fig. 7. The CO2 emissions were reduced by at least 1.4% or 12,000 tons over the course of the year using SRF compared to the Reference Case.

Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038

J.R. Fyffe et al. / Waste Management xxx (2015) xxx–xxx

production in the SRF Case shows only limited contribution to the total CO2 emissions, showing that the amount of energy needed to produced SRF is minimal compared to the amount it displaces. In addition to the CO2 emissions shown in Fig. 7, it is important to remember the policy discussion on carbon related to the use of non-recycled materials. Depending on which policies are adopted in the future, it is possible that some or all of the CO2 from the SRF will be considered biogenic in nature. While the total CO2 emissions will not be changed, the accounting of CO2 emissions as biogenic or not has important implications in terms of renewable energy credits or any costs that are attached to CO2 emissions. Other possible policies treat the avoided emissions from the landfill as credits for the SRF producer or user or discount emissions from energy recovery from waste streams, which would likely apply to SRF. While these policies are still in discussion, they are important to keep in mind when thinking about the CO2 emissions in the context of using SRF as an alternative fuel (Lariviere, 2007). 4. Conclusion The experiment and analysis presented here elucidates the potential energetic and environmental trade-offs of using residue-derived SRF as an alternative fuel in a cement kiln. These are some of the key findings of the report: The use of SRF is expected to significantly reduce CO2 emissions at both the cement kiln and on a life-cycle basis. CO2 reductions could be amplified if policies are in place that consider landfill emissions avoidance or claim non-recycled materials as partially or wholly renewable due to the biogenic nature of some carbon contained in SRF as well as the second-use aspect of SRF. The experimental test burn of the SRF indicated a decrease in SO2 but a slight increase in NOX compared to baseline measurements at the kiln. In addition to air emission benefits, the use of SRF as an alternative fuel reduces the amount of material being landfilled by the amount of SRF that is created. SRF creates a second use for non-recycled material that would otherwise be buried in landfills, saving land space and the emissions associated with the breakdown of biodegradable materials underground. Therefore, the use of SRF adds auxiliary benefits that are amplified when material is diverted from poorly designed or maintained landfills and extends the life of well-designed landfills. Finally, the use of SRF was demonstrated to be beneficial on an energy consumption basis, requiring overall less energy (not including energy contained in SRF that would otherwise be buried) than if no SRF was used. The use of SRF was shown to reduce the equivalent of between 7710 and 9070 Mg of coal use at feed rates used during the test burn. The primary reduction in energy comes from the 1-to-1 energy replacement of fossil fuels with SRF at the cement kiln. Energy used to process and transport the SRF were found to be significantly less than the energy contained in the SRF in all cases. Co-location of facilities would increase all of the benefits mentioned, with the co-location of an SRF production facility and a materials recovery facility being the most obvious candidates. This analysis suggests that using SRF to replace fossil fuels at cement kilns is beneficial on an energetic and environmental basis, with the possible exception of increased NOX emissions. Further study into the other pollutant emissions associated with combustion of SRF at cement kilns is suggested as well. Co-location, likely due to the similarity in operations, of MRFs, SRFs, and landfills would increase the benefits of using SRF even further.

9

Foundation’s graduate research fellowship program, and the Cockrell School of Engineering’s Department of Mechanical Engineering at the University of Texas at Austin. The sponsors have not officially endorsed this publication, and the views expressed herein may not reflect the views of the sponsors. The authors would like to acknowledge the contributions from the following for aiding in the synthesis of this research: Craig Cookson from American Chemistry Council, Mike Goins and his team from Total Petrochemicals, Greg Mayes and Derek Thorington from TXI, Steve Stone, Randy Wolf, and Kerry Getter from Balcones Resources, Herb Northrup from Blue River Resources, Tim Horchar and Jonathan Sloan from Canusa Hershman, and the team from Colgate Paper Stock Company, Inc. References Brooks, C.L., 1991, An Analysis of Refuse Derived Fuel as an Environmentally Acceptable Fuel Alternative for the Cement Industry. M.S. Thesis, University of North Texas. Buah, W., Cunliffe, A., Williams, P., 2007. Characterization of products from the pyrolysis of municipal solid waste. Process Saf. Environ. 85, 450–457. Conesa, J., Galves, A., Mateos, F., Martin-Gullon, I., Font, R., 2008. Organic and inorganic pollutants from cement kilns stack feeding alternative fuels. J. Hazard. Mater. 158, 585–592. Cruz, F.D., Barlaz, M., 2010. Estimation of waste component-specific landfill decay rates using laboratory-scale decomposition data. Environ. Sci. Technol. 44, 4722–4728. Forster, P., Ramaswamy, V., 2007. Changes in atmospheric constituents and in radiative forcing. Tech. Rep. Fyffe, J.R., 2011, Balcones Fuel Technology Site Visit. Fyffe, J.R., Breckel, A.C., Townsend, A.K., Webber, M.E., 2012, Residue-Derived Solid Recovered Fuel for Use in Cement Kilns. Tech. Rep., The University of Texas at Austin. Gera, D., Gautam, M., 1993. Emissions from RDF/coal blended fuel combustion. Chemosphere 27 (12), 2353–2363. ICF International, 2008, U.S. EPA Cement Sector Trends in Beneficial Use of Alternative Fuels and Raw Materials. Tech. Rep. Lariviere, M., 2007, Methodology for Allocating Municipal Solid Waste to Biogenic and Non-Biogenic Energy. Tech. Rep., U.S. EIA Office of Coal, Nuclear, Electric, and Alternate Fuels. Mueller, W., 2013. The effectiveness of recycling policy options: waste diversion or just diversions? Waste Manage 33, 508–518. National Renewable Energy Lab (NREL), 2012, U.S. Life Cycle Inventory Database. www.nrel.gov/lci. Rada, E.C., 2014. Energy from municipal solid waste. WIT Trans. Ecol. Environ 190, 945–958. Sarc, R., Lorber, K.E., Pomberger, R., Rogetzer, M., Sipple, E.M., 2014. Design, quality and quality assurance of solid recovered fuels for the substitution of fossil feedstock in the cement industry. Waste Manage. Res. 32, 565–585. Serre, D.S., Lee, C.W., 2009, Evaluation of the Impact of Chlorine on Mercury Oxidation in a Pilot-Scale Coal Combustor – The Effect of Coal Blending. Tech. Rep. U.S. EPA. Solid Waste Association of North America (SWANA), 2002, Successful Approaches to Recycling Urban Wood Waste. Tech. Rep. U.S. Department of Agriculture – Forest Service. Texas Commission on Environmental Quality (TCEQ), 2009, Cement Plant Emissions Permit 5933 and PSD-TX-63M3. Thorington, D., 2011, Personal Interview at TXI Hunter Cement Plant. Tillman, D.A., 2008, Chlorine in Solid Fuels Fired in Pulverized Coal Boilers – Sources, Forms, Reactions, and Consequences: A Literature Review, Tech. Rep. Foster Wheeler. Townsend, A.K., Webber, M.E., 2012. An integrated analytical framework for quantifying the LCOE of waste-to-energy facilities for a range of greenhouse gas emissions policy and technical factors. Waste Manage. 32, 1366–1377. U.S. EPA, 2007a, eGRID2007 Version 1.1 Year 2005 Summary Tables. U.S. EPA, 2007b, eGRID2006 Version 2.1 (April 2007) Year 2004 Summary Tables. U.S. EPA, 2011a. Waste Management Hierarchy. . U.S. EPA, 2011b. Emissions Factors and AP 42, Compilation of Air Pollutant Emission Factors. . U.S. EPA, 2011c, eGRID2010 Version 1.1 Year 2007 Summary Tables. U.S. EPA, 2011d, Toxic Release Inventory Database. . Velis, C., Wagland, S., Longhurst, P., Robson, B., Sinfield, K., Wise, S., Pollard, S., 2012. Solid recovered fuel: influence of waste stream composition and processing on chlorine content and fuel quality. Environ. Sci. Technol. 46, 1923–1931.

Acknowledgments The research presented in this report has been funded in part by the American Chemistry Council (ACC), National Science Please cite this article in press as: Fyffe, J.R., et al. Use of MRF residue as alternative fuel in cement production. Waste Management (2015), http:// dx.doi.org/10.1016/j.wasman.2015.05.038