Life cycle assessment of heated apron pavement system operations

Transportation Research Part D 48 (2016) 316–331

Contents lists available at ScienceDirect

Transportation Research Part D journal homepage: www.elsevier.com/locate/trd

Life cycle assessment of heated apron pavement system operations Weibin Shen a, Halil Ceylan b, Kasthurirangan Gopalakrishnan c,⇑, Sunghwan Kim d, Peter C. Taylor e, Chris R. Rehmann f a

24 Town Engineering Building, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011, USA 406 Town Engineering Building, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011, USA c 354 Town Engineering Building, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011, USA d 24 Town Engineering Building, Institute for Transportation, Iowa State University, Ames, IA 50011, USA e National Concrete Pavement Technology Center, Iowa State University, Ames, IA 50011-3232, USA f 482B Town Engineering Building, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011, USA b

a r t i c l e

i n f o

Article history: Available online 27 August 2016 Keywords: Greenhouse gas Energy Pavements Snow and ice Heating Sustainability

a b s t r a c t Heated pavement systems (HPS) offer an attractive alternative to the cumbersome process of removing ice and snow from airport pavements using traditional snow removal systems. Although snow and ice removing efficiency and economic benefits of HPS have been assessed by previous studies, their environmental impact is not well known. Airport facilities offering public or private services need to evaluate the energy consumption and global warming potential of different types of snow and ice removal systems. Energy usage and emissions from the operations of hydronic heated pavement system using geothermal energy (HHPS-G), hydronic HPS using natural gas furnace (HHPS-NG), electrically heated pavement system (EHPS), and traditional snow and ice removal system (TSRS) are estimated and compared in this study using a hybrid life cycle assessment (LCA). Based on the system models assessed in this study, HPS application in the apron area seems to be a viable option from an energy or environmental perspective to achieve ice/snow free pavement surfaces without using mechanical or chemical methods. TSRS methods typically require more energy and they produce more greenhouse gas (GHG) emissions compared to HPS during the operation phase, under the conditions and assumptions considered in this study. Also, HPS operations require less energy and have less GHG emissions during a snow event with a smaller snowfall rate and a larger snow duration. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction The global climate change is likely to lead to lower average temperatures in the mid-west and northeast parts of the United States as well as more frequent winter precipitation events. Therefore, the use of life-cycle assessment (LCA) methodology to identify the appropriate technology to help airports adapt to climate change while also mitigating future emissions is warranted. A primary objective of the U.S. Federal Aviation Administration’s (FAA’s) Next Generation Air Transportation System (NextGen), currently being implemented in stages, is to reduce delays and interruptions to flight operations. ⇑ Corresponding author. E-mail addresses: [email protected] (W. Shen), [email protected] (H. Ceylan), [email protected] (K. Gopalakrishnan), [email protected] (S. Kim), [email protected] (P.C. Taylor), [email protected] (C.R. Rehmann). http://dx.doi.org/10.1016/j.trd.2016.08.006 1361-9209/Ó 2016 Elsevier Ltd. All rights reserved.

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

317

Nomenclature Ar C COP c1 cp,ice cp,water d DT E

es H hc hf hfg hm M MU n P

qdry_air qwater

Q Qt qh qe qi qm qo qs SG s Tf TMR t ta tf ts Wa Wf

ratio of snow-free area to total area, dimensionless specific heat of concrete pavement, kJ/kg °C or Btu/lb °F coefficient of performance of geothermal heat pump, decimal number conversion factor, 1000 mm/m or 12 in./ft specific heat of ice, kJ/kg °C or Btu/lb °F specific heat of water, kJ/kg °C or Btu/lb °F Stephan-Boltzmann constant, Btu/h ft2 °R4 temperature difference, °C or °F Energy consumption of geothermal heat pump, Btu/h or kJ/h emittance of wet slab, dimensionless total head, ft convection heat transfer coefficient for turbulent flow, Btu/h ft2 °F heat of fusion for water, kJ/kg or Btu/lb heat of evaporation at the film temperature, kJ/kg or Btu/lb mass transfer coefficient of concrete slab, m/h or ft/h mass of concrete pavement, kg or lb flow rate increase multiplier, 0.085 for 40% by volume glycol mixture pump efficiency, decimal number power required for circulating pump, hp density of dry air, kg/m3 or lb/ft3 density of water equivalent of snow, kg/m3 or lb/ft3 flow rate, gallon/min total heat rate required for pavement idling and snow melting, Btu/h or kJ/h heat transfer rate by convection and radiation, Btu/h ft2 or kJ/h m2 heat rate of evaporation, Btu/h ft2 or kJ/h m2 heating rate required for concrete pavement idling, Btu/h or kJ/h heat rate of fusion, Btu/h ft2 or kJ/h m2 heat rate required for melting snow using heated pavement system, Btu/h ft2 or kJ/h m2 sensible heat rate transferred to the snow, Btu/h ft2 or kJ/h m2 specific gravity of heated solution, 1 of water and 1.034 of 40% propylene glycol rate of snowfall, mm or inches of water equivalent per hour liquid film temperature, °R mean radiant temperature of surroundings, °R snow period, h ambient temperature coincident with snow fall, °C or °F liquid film temperature, °C or °F melting temperature, °C or °F humidity ratio of ambient air, lbvapor/lbair humidity ratio of saturated air at film surface temperature, lbvapor/lbair

Traditional snow and ice removal methods could cause airline delays, high operation costs and airside incidents involving airport crew during snow and ice removal activities. In order to prevent these problems, heated pavement systems (HPS) are being studied as an alternative strategy, in the context of FAA’s NextGen and efficient adaptation to climate change, to traditional snow and ice removal systems (TSRS) applied in apron areas (Ceylan, 2015). Shen et al. (2015) recently reported that hydronic HPS, the most common type of HPS, could have environmental benefits when used to remove snow and ice from aprons. The primary goal of this study is to provide a more comprehensive understanding of different snow and ice removal system (SRS) operations not only from an energy consumption perspective but also from an environmental impact aspect and to help the airport authorities make a more informed decision. To accomplish this goal, this study aims to identify the inventories or steps that burden each SRS operation the most so that energy usage and environmental impacts can be reduced. The energy consumption and contributions to global warming of four snow and ice removal systems as applied to the airport apron are evaluated and compared for different snowfall conditions. These systems are hydronic HPS using geothermal energy (HHPS-G), hydronic HPS using natural gas furnace (HHPS-NG), electrically heated pavement system (EHPS) using electricity, and TSRS. Heating energy sources are the primary differences among HPS types evaluated. As one of the first life cycle assessment (LCA) studies on different types of HPS, this paper focuses on the impacts of HPS operation phase and related life cycle stages in comparison to TSRS. For simplicity, system boundaries of four different SRS include only sectors defined as processes of snow and ice removal operation. SRS can be generally classified into four sub-system processes: power generation, material production, snow and ice removal application, and waste treatment.

318

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

Therefore, the operation system boundary in this study includes these four sectors. A ‘‘well to gate” assessment for the power generation facility was applied to understand the greenhouse gas (GHG) emission from the power production phase. An economic input-output life cycle assessment (EIO-LCA) on-line model (GDI-CMU, 2008) was applied in material production stage, and its system boundary was defined in the 2002 US Benchmark version of the EIO LCA model (Weber et al., 2010). Life cycle inventories are significantly related to the system boundary (SAIC, 2006). Because this study is to evaluate the energy consumption and global warming potential of different airport apron SRS operations, inventories that contribute efforts (e.g., increasing thermal conductivity or preventing heat lost) to snow and ice removal and their upstream stage life cycle (e.g., raw material extraction) are assessed. Life cycle inventories of the snow removal systems were collected from previous research reports, government official documents or company manual scripts and are defined in following sections. 2. Analysis methods Four different SRS, viz., HHPS-G, HHPS-NG, EHPS and TSRS, are individually analyzed. Each system is designed for a short to medium range airliner apron area of 1765 m2 (19,000 ft2) (Airliners, 2015). The systems are analyzed at
6.7 °C (20 °F) air temperature, 16 km/h (10 mile per hour) wind speed, and under 12.7 mm/h, 25.4 mm/h, and 50.8 mm/h (0.5 in./h, 1 in./h, and 2 in./h) snowfall rate conditions. This study uses a time-based functional unit and allocates energy consumption of pavement idling depending on the snow durations, assumed to be 1 h, 4 h, 8 h, and 12 h. As a typical airport concrete pavement is designed to serve for 20 years, HPS are assumed to be designed for a 20-year life (WisDOT RD&T Program, 2004). Note that pavement LCA is still a relatively new topic and a number of challenges and research gaps have been identified for the practical use of LCA for pavements (UCPRC, 2010; FHWA, 2014; Gopalakrishnan et al., 2014). It is also acknowledged that there are currently fewer HPS in operation around the globe and among them, most have focused on HHPS-G. Most recently, a HHPS-G was constructed to hear an aircraft parking ramp at the Greater Binghamton Airport (BGM). This geothermal system uses heat from the ground to melt ice/snow on an 24-m  24-m (80-ft  80-ft) boarding area and passenger walkway from the terminal. Thus, cost and LCA-related data needed for conducting a full LCA is just not available for these technologies in an airport environment. Consequently, this study has focused on just the operation phase of the LCA. 2.1. Modeling equations The equations utilized for modeling in this study are summarized and presented in the following sub sections. The detailed definitions along with unit for equations are provided in nomenclature section at the end of paper. 2.1.1. Pavement idling energy consumption The heating rate qi required for concrete pavement idling, a heating procedure that raises the HPS surface temperature to 0 °C (32 °F), is calculated with

qi ¼

C  DT  M t

ð1Þ

where C = specific heat of concrete pavement, DT = temperature difference, M = mass of concrete pavement, and t = snow period. 2.1.2. Snow melting energy consumption After the concrete slab surface is heated to 0 °C, the HPS uses heat to melt snow. The heat rate qo required for melting snow using HPS is the sum of the sensible heat rate qs transferred to the snow, the heat rate of fusion qm, the heat rate of evaporation qe, and the heat transfer rate by convection and radiation qh (ASHRAE Handbook, 2003):

qo ¼ qs þ qm þ Ar ðqe þ qh Þ

ð2Þ

where Ar = ratio of snow-free area to total area. The sensible heat rate qs to bring the snow to 0 °C is

qs ¼ s  qwater  ½cp;ice ðt s
t a Þ þ cp;water ðt f
ts Þ=c1

ð3Þ

where s = rate of snowfall; qwater = density of water equivalent of snow; cp,ice = specific heat of snow; cp,water = specific heat of water; ts = melting temperature; tf = liquid film temperature; ta = ambient temperature; c1 = conversion factor. The heat rate of fusion qm to melt the snow is

qm ¼ s  hf  qwater =c1

ð4Þ

where hf = heat of fusion for water. The heat rate of evaporation qe is

qe ¼ qdry

air

 hm ðW f
W a Þhfg

ð5Þ

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

319

where qdry_air = density of dry air, hm = mass transfer coefficient of concrete slab, Wf = humidity ratio of saturated air at film surface temperature, Wa = humidity ratio of ambient air, and hfg = heat of evaporation at the film temperature. The heat transfer rate by convection and radiation (qh) to melt the snow is:

  qh ¼ hc ðtf
ta Þ þ d  es T 4f
T 4MR

ð6Þ

where hc = convection heat transfer coefficient for turbulent flow, d = Stephan-Boltzmann constant, es = emittance of wet slab, Tf = liquid film temperature, and TMR = mean radiant temperature of surroundings. 2.1.3. Geothermal heat pump operating energy demand Energy consumption E of geothermal heat pump is calculated with (Mix, 2006):

E¼

Qt COP

ð7Þ

where Qt = total heat rate required for pavement idling and snow melting and COP = coefficient of performance. 2.1.4. Hydronic system flow rate The flow rate Q (in gallon/min) is computed with (Viega, 2015)

Q¼

q0  ð1 þ MUÞ 500  DT

ð8Þ

where MU = flow rate increase multiplier. 2.1.5. Circulating pump operating energy demand To calculate the energy demand for circulating pump, the power P (in horsepower) required was computed with (Viega, 2015):

P¼

Q  H  SG 3960  n

ð9Þ

where Q = flow rate; H = total head; SG = specific gravity of heated solution; n = pump efficiency. 2.2. HPS operation: energy requirements A HPS applies hydronic heat or radiant heat through conductive media to melt snow (FAA, 2011). To evaluate how much energy is needed to operate each HPS, the energy requirement for pavement idling and snow melting should first be analyzed. Because of the insulation installed in top 102 mm (or 4 in.) of concrete of each HPS, zero back and edge losses were assumed. The Portland Cement Concrete (PCC) as normal concrete is assumed to be composed of 12% cement, 82% aggregates and 6% water by total volume. Typically, it is assumed to have a density of 2402 kg/m3 (150 lb/ft3) (WS DOT, 2015) and a specific heat of 0.837 kJ/kg/°C (0.2 Btu/lb °F) (Lamond and Pielert, 2006). As the HPS goes into operation, it is expected to supply heat to the slab surface for melting the snow. At 0 °C, snow is able to melt with extra energy input. So, the strategy of using a HPS idling operation in this study is to maintain a pavement surface temperature of 0 °C. Based on Eq. (1), a warming of the 1765 m2 normal concrete slab surface from
6.7 °C to 0 °C requires 2405 MJ/snow period. This study uses a time-based functional unit and allocates energy consumption of pavement idling depending on the snow durations, assumed to be 1 h, 4 h, 8 h, and 12 h. Based on the snow periods evaluated in this study, energy consumption for normal concrete pavement idling are 2405 MJ/h, 601 MJ/h, 301 MJ/h, and 200 MJ/h. By using Eqs. (2)–(6), the energy requirements for snow melting under different assumed snow rates, viz., 0.5 in./h, 1 in./h, and 2 in./h, can be calculated as 1521 kJ/h m2, 1965 kJ/h m2, and 2850 kJ/h m2, respectively. 2.3. HPS operation: GHG emissions factors for different fossil fuel applications Three different types of fossil fuel power plants are considered for electricity generation: coal (bituminous), natural gas and distillate oil. Based on the information provided by U.S. Energy Information Administration (EIA) (EIA, 2015a), among these three types of power plant, 58% utilize coal as the energy source, 40% use natural gas, and only 2% use distillate oil to generate electricity. The phases of coal-fired power plant life cycle include coal mining, coal preparation/cleaning, all necessary transportation of coal to the power plant, and grid electricity production. A natural gas-fired power plant life cycle includes natural gas extraction, natural gas pretreatment and transportation, and grid electricity production (NETL, 2000). Shen et al. (2015) estimated the whole life cycle electricity GHG emission factors from the coal fired power plant and the natural gas-fired power plant in consideration of each life phase in which GHG emission factors are reported in literature. The estimated whole life cycle natural gas (production and combustion) GHG emission factor is 0.185 kgCO2eq/kWh whereas the estimated whole life cycle electricity (production and combustion) GHG emissions factor is 0.42 kgCO2eq/kWh when the electricity is generated from natural gas-fired power plant. Detailed estimation procedures are presented in Shen et al. (2015).

320

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

Since the fuel-fired (distillate oil-fired) power plant GHG emissions factor is highly site-specific, a reasonable value of 0.778 kgCO2eq/kWh was assumed based on a previous study (Gagnon et al., 2002) and US EIA database (EIA, 2015b). The GHG emission factor from natural gas production (i.e., extraction, pretreatment, and transportation) is about 0.004 kgCO2eq/kWh (NETL, 2000) and the GHG emission factor from natural gas combustion is 0.181 kgCO2eq/kWh (EIA, 2013). Thus, the whole life cycle GHG emission factor for natural gas production and combustion can be estimated as 0.185 kgCO2eq/kWh. The GHG emission factor from diesel oil production is about 0.19 kgCO2eq/kWh (Shen et al., 2014) and the GHG emission factor from diesel oil combustion is 0.27 kgCO2eq/kWh (AFDC, 2014). Thus, the whole life cycle GHG emission factor for diesel oil can be estimated as 0.46 kgCO2eq/kWh. To sum up, GHG emission factors of different fossil fuel applications utilized in this study are shown in Table 1. 3. Operation of hydronic heated pavement system using geothermal energy 3.1. HHPS-G operation system boundary HHPS-G utilizes geothermal energy as a heating source to warm up antifreeze solution circulating under the pavement in order to keep the concrete slab surface without snow. Based on the methodology, the HHPS-G operation life cycle can be divided into 4 sub-life cycles: power- generation life cycle, snow-removal operation life cycle, material-production life cycle, and antifreeze-wastewater treatment life cycle. The HHPS-G operation flow chart and system boundary of the HHPS-G operation life cycle is shown in Fig. 1. 3.2. HHPS-G operation model The energy used for operating a HHPS-G includes the energy used for geothermal heat pump and circulating pump operation, antifreeze solution production, insulation production, and solution waste treatment. A HHPS-G uses a direct exchange ground source heat pump (GSHP) to extract geothermal energy from the ground to heat solution flowing through embedded pipes in the pavement to increase the pavement surface temperature above 0 °C and melt the ice/snow. Energy required for pavement idling and snow melting is calculated by applying Eqs. (1)–(6). Based on geothermal heat pumps key product criteria, the coefficient of performance (COP) of a direct ground exchange heat pump can be as high as 3.6 (Energy Star, 2015). To understand the behavior of HHPS-G applied in different geothermal conditions, the COP of geothermal heat pump is assumed to be 2.0 as minimum and 3.6 as maximum in this study. Energy consumption of heat pump can be calculated using Eq. (7). System design is based on the energy requirement for snow melting (Viega, 2015). The heaviest snowfall in this study is 50.4 mm/h; therefore, systems at least need to be feasibly operational under 50.4 mm/h snow rate conditions. Based on the operational energy requirement for snow melting under 50.4 mm/h snow rate, cross-linked polyethylene (PEX) pipe spacing in normal concrete is assumed to be 229 mm in order to support enough energy. Tubing length has a multiplier of 4.92 m/m2, therefore 8689 m of PEX tube is required to be installed under a 1765 m2 apron area. Piping circuit length is designed to be 122 m and a total of 71 circuits are required. HHPS-G circulates 40% by volume of propylene glycol solution in 19 mm (¾ in.) PEX pipes. Based on the design for 50.8 mm snow accumulation, the circulating solution flow rate can be calculated using Eq. (8) to obtain a flow rate of 26 L/min (6.9 gallons per minute, gpm). Thus, the total flow rate is 1866 L/min (493 gpm) and the total pressure drop is about 38 m (125 ft) of head. A glandless circulating pump with 50–70% efficiency (Wilo, 2009) can be used, and so 60% efficiency circulating pumps are applied in the systems in this situation. The energy circulating pump demand is calculated as 19.4 kW (26 hp) using Eq. (9). Because 40% by volume of propylene glycol solution has a very similar density to water, the unit volume of solution in 19 mm PEX pipe is about 0.22 L/m (0.018 Gal/ft), and a total of 1941 L (513 gallon) of solution is required for HHPS-G

Table 1 GHG emission factors of electricity, natural gas and diesel oil. Fossil fuel application emission factors Electricity emission factor

Value (kgCO2eq/kWh) From coal power plant From natural gas power plant From distillate oil power plant

Natural gas emission factora Diesel oil emission factorb a b

Natural gas upstream and combustion stages are included. Diesel oil upstream and combustion stages are included.

0.960 0.420 0.778 0.185 0.460

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

321

Fig. 1. Schematic view of HHPS-G: (a) HHPS-G operation flow chart and (b) system boundary of HHPS-G operation.

operation. By using the IO-LCA software, GREET (ANL, 2013), the energy to produce 1 kg propylene glycol is 27.57 kWh which will release 6.46 kgCO2eq. The propylene glycol solution needs to be checked and replaced every year (Raypak, Inc., 2015), and the waste solution can be discharged and treated in a municipal wastewater treatment plant. Propylene glycol has a Chemical Oxygen Demand (COD) content of about 1680 g/kg (USC, 2008) and produces 0.15 kgCOD/h of waste antifreeze solution for the 1765 m2 apron area. In general, aerobic wastewater treatment energy requirement is 1 kWh/kgCOD (Geest and Kiechle, 2010). Since such treatment requires electrical power, there is no direct GHG released from the wastewater plant itself, so the GHG emission is actually from the power generation phase. Calculations (Shen, 2015) show that this is about 104 kgCO2eq/h for propylene glycol wastewater treatment. A Polyiso insulation layer is installed on the bottom and edge of top 102 mm (4 in.) of the concrete slab to prevent heat loss. The back and edge heat loss of the HPS is assumed to be 0% (Viega, 2015). Based on the description of life cycle inventory assessment, a 38 mm (1.5 in.) thick layer of Polyiso insulation with a 9.8 of thermal resistance RIP (US unit, using Inch-Pound measures) was assumed to be used in the HPS whose life time is about the same as that of normal concrete pavement life time. The life cycle of insulation layer manufacturing has been studied, and its GHG emission factor is 0.39 kgCO2eq/ft2, with an energy consumption factor of 8.66 MJ/ft2 (PIMA, 2015).

322

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

4. Operation of hydronic heated pavement system using natural gas 4.1. HHPS-NG operation system boundary The HHPS-NG operation flow chart and system boundary are shown in Fig. 2. The only difference assumed between the HHPS-G and the HHPS-NG is that the HHPS-NG utilizes a fossil fuel heater (i.e., natural gas furnace in this study) as a heating source to warm up the antifreeze solution. 4.2. HHPS-NG operation model A natural-gas furnace with a 90% efficiency, considered to have higher efficiency than a traditional gas boiler but reflect current improvements on the efficiency of the heating technique, is applied in the HHPS-NG. A heat exchanger is required in

Fig. 2. Schematic view of HHPS-NG operation: (a) HHPS-NG operation flow chart and (b) system boundary of HHPS-NG operation.

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

323

the HHPS-NG because propylene glycol is used as antifreeze to prevent heat transfer medium from freezing, and propylene glycol solution cannot be directly heated by the furnace. Therefore, the HHPS can be divided into two subsystems, a waterheating system and a pavement-heating system. The water-heating system uses a natural gas furnace to heat up water and circulates heated water through a 70% efficiency heat exchanger using a circulating pump. Forty percent by volume of propylene glycol solution extracts heat from the water heating system through the heat exchanger and is circulated under the normal concrete slab surface by the circulating pump to heat the pavement surface. As demonstrated by the model and boundary of each HHPS operation (see Fig. 1 for HHPS-G and Fig. 2 for HHPS-NG), the difference between the HHPS-G and the HHPS-NG is that a natural gas furnace unit along with a heat exchanger are required in the HHPS-NG instead of geothermal heat pump unit in the HHPS-G. However, the system design of HHPS-NG is generally similar to HHPS-G, according to the ‘‘Viega Heated Pavement Design Manual Script” (Viega, 2015). Because the only difference between the HHPS-NG and the HHPS-G is the heating source, the piping design, circulating pump selection, insulation layer design, propylene glycol solution usage, and solution waste treatment for the HHPS-G will be the same as those of HHPS-NG. 5. Operation of electrically heated pavement system 5.1. EHPS operation system boundary An electrically-heated pavement system utilizes electric mats or cables to transform electricity into radiant heat for pavement heating. The EHPS operation system boundary is similar to the other heated pavement system boundaries (i.e., HHPS-G and HHPS-NG). The only difference for the EHPS boundary is that it does not include the wastewater treatment stage. An EHPS operation flow chart and system boundary are shown in Fig. 3. 5.2. EHPS operation model An EHPS utilizes an electrically-heated cable as a heating source to warm concrete pavement directly instead of using a heated solution. Studies (Ceylan, 2015; Gopalakrishnan et al., 2015) have been done in which conductivity of concrete was increased to improve EHPS efficiency, and one approach is to mix conductive material inside the concrete pavement. Carbon fiber is one such conductive material that has been reported to effectively increase the conductivity of concrete EHPS. A 1.2 m-long, 0.9 m-wide, 102 mm thick (or 4-ft-long, 3-ft-wide, and 4 in.) thick EHPS slab was tested in an on-going study (Ceylan, 2015) at Iowa State University (ISU) to evaluate its energy efficiency. The electrical input was 950 W, and edge and bottom insulation layers were installed to prevent heat lost through those surfaces. About 0.8% carbon fiber (by volume of total concrete mix) was mixed in the concrete to increase its conductivity (Ceylan, 2015). The result was that it took 20 min to warm the 1.1 m2 (12 ft2) slab from
6.7 °C to 0 °C, so energy consumption for conductive concrete pavement idling was 0.07 MJ/m2. Based on experimental test results, a 1765 m2 apron can utilize 102 mm thick concrete pavements containing 0.8% of carbon fiber (by volume of total concrete mix) as effective EHPS. The total concrete volume required for this is 221 m2. In using 0.8% of carbon fiber (by volume of total concrete mix), 1.77 m3 of the total volume of active carbon is required in concrete paving construction for EHPS. Carbon fiber has a density of 1550 kg/m3 (Clearwater Composites, LLC, 2015), and the total mass of carbon fiber required for a 1765 m2 apron is about 2736 kg. Since concrete pavement service life was assumed to be 20 years, the lifetime of carbon fiber is also assumed to be 20 years. Allocating carbon fiber usage on an hourly basis, 16 g/h of carbon is required. Based on a previous study, carbon fiber production life cycle has an energy consumption factor of 704 MJ/kg and 31 kgCO2eq/kg of GHG emission factor (Das, 2011). Because EHPS utilizes electricity as the only energy input for heating, and the insulation layer is installed in the system to prevent heat loss, all electrical power is assumed to transform into radiant heat for snow and ice melting. 6. Operation of traditional snow and ice removal system 6.1. TSRS operation system boundary Traditional snow and ice removal systems use mechanical equipment like snow plows or snow brooms to remove snow first and then apply de-icing chemicals on the pavement to prevent snow formation. The chemically polluted water is subsequently treated in municipal wastewater treatment plant. Therefore, life cycles of de-icing chemical production, power generation, snow and ice removal operations and wastewater treatment should be included. Fig. 4 shows the TSRS flow chart and system boundary that differs from that of HPS. 6.2. TSRS operation model Potassium acetate, sodium acetate, and propylene glycol are the chemicals commonly used in airport pavement de-icing (EPA, 2012). A 50% by weight potassium acetate solution, a 60% by weight propylene glycol solution, and a sodium acetate solid de-icer are assessed in this study. These three chemicals are applied at levels of 75 g/m2 (MeltSnow, 2015), 65 g/m2

324

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

Fig. 3. Schematic view of EHPS operation: (a) EHPS operation flow chart; (b) system boundary of EHPS operation.

(FAA, 1992), and 50 g/m2 (Cryotech NAACÒ, 2008), respectively. Because the amount of chemical for a de-icing application is based on air temperature, and the air temperature is constant under different snow rate conditions, de-icer usage is the same for all snow rates. Among these three chemicals, EPA (2012) has been reported that 67% of airports in U.S. use potassium acetate, 11% use propylene glycol and 22% use sodium acetate. De-icing chemicals are sprayed on the pavement once every hour during snow. Software such as GREET (ANL, 2013) and on-line software Economy Input-Output Life Cycle Assessment (EIO-LCA) (GDICMU, 2008) were used to calculate the energy consumption and GHG emission of de-icer production. To produce 1 kg of potassium acetate requires 18 kWh of energy releasing 3.82 kgCO2eq, 1 kg of propylene glycol requires 28 kWh of energy releasing 6.46 kgCO2eq, and 1 kg of sodium acetate requires 12 kWh of energy releasing 2.73 kgCO2eq. A multifunctional vehicle, 1104D-E44TA, with a diesel engine requiring 97 kW and a transmission power demand of 68 kW (RPM Tech, 2015), is used for spraying the de-icing chemical on the 1765 m2 apron pavement to prevent ice adhesion. An hour after the application of the chemicals, the multifunctional vehicle is converted into a snowplow to remove snow from the apron area. Considering that the apron area is relatively small, the total operating time of the vehicle, including chemical spraying and snow plowing, is assumed to be 10 min.

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

325

Fig. 4. Schematic view of TSRS: (a) TSRS operation flow chart and (b) system boundary of TSRS operation.

Because a de-icing chemical is used in the TSRS, a wastewater treatment process is required. In general, most of the airport pavement runoff is treated in municipal wastewater treatment plant using an aerobic treatment. Based on a previous study, the chemical oxygen demand (COD) of the different chemicals are 1050 g/kg for potassium acetate, 1680 g/kg for propylene glycol, and 1010 g/kg for sodium acetate (USC, 2008). About 1 kWh of electricity demand per kg COD is assumed for such aerobic treatment (Geest and Kiechle, 2010). Since such treatment requires electrical power, there is no direct GHG released from the wastewater plant itself, so the GHG emission is actually from the electrical power generation phase. Calculations (Shen, 2015) show that 129 kgCO2eq/h will be released for potassium acetate waste water treatment, 104 kgCO2eq/h for propylene glycol wastewater treatment, and 56 kgCO2eq/h sodium for aerobic wastewater treatment. 7. Results and discussion Four case studies involving TSRS operations and three alternative HPS have been analyzed to evaluate the sustainability of such systems. As the analyses for different SRS operations demonstrate, energy consumption conditions and environmental

326

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

impact are influenced by several different factors such as snow rates, snow period, and system efficiency, and these factors vary among the four types of system operations. Energy consumption and GHG emissions have also been compared to estimate the most sustainable system for removing snow. 7.1. Comparison of energy consumptions Snow period and snow rate have significant effects on energy consumption (Fig. 5). As snow period increases, energy consumption of all SRS operations increases. Among SRS, TSRS will require more energy for snow and ice removal operation than HPS operations when snow duration becomes longer. As snow rate increases, energy consumption of HPS operations increases but that of TSRS does not. Considering that energy consumption of HPS operations will increase when snowfall rate increases, energy consumption of HHPS-NG operation might be higher than that of TSRS when the snow rate exceeds 50.8 mm/h (see Fig. 5(c)). However, energy consumption of the other HPS operations will be smaller under various snow rates. Although institutively HPS operation life cycle is expected to consume relatively more energy than TSRS operation life cycle considering that HPS requires energy to heat the pavement surface to remove snow, it was found that more energy is consumed in TSRS operation life cycle as Fig. 5 demonstrates. Also, among the 3 kinds of different HPS operations, the HHPS-NG operation has a higher energy consumption compared to energy consumption of HHPS-G and EHPS operations. To understand the inventories causing such differences, the energy consumption contributions of different inventories in different systems have been analyzed and are summarized in Table 2. As seen in Table 2, heating source productions for all HPS systems and deicer productions for TSRS are contributing the most and are the most sensitive inventory factors to energy consumption results than other inventory factors with uncertain values. Most of the energy used in TSRS operation is related to de-icing chemical production and wastewater treatment. Large amounts of de-icing chemicals are required for a 1765 m2 apron area, and the energy demand for de-icer manufacturing and the waste solution (from used de-icer) treatments is relatively high. This results in a higher energy consumption for the TSRS operation life cycle than for the heated pavement systems that do not require de-icing materials. Therefore, if an airport is conducting an ‘‘Airport Sustainability Planning” program and wishes to reduce their energy consumption during snow and ice removal, using less de-icer is an effective way to reduce much of the energy demand. More than 99% of the total energy consumed in HPS operation is used for heating, as shown in Table 2. Due to differences in system models and equipment used for HPS, energy consumption may vary. Taking HHPS-NG as an example, the system utilizes a 90% efficient natural gas furnace, a 60% efficient circulating pump, and a 70% efficient heat exchanger. Compared to the other two system models, the HHPS-NG exhibits more heat loss during the heating process, so the HHPS-NG operation requires relatively more energy among the three different kinds of HPS. HHPS-G efficiency is highly dependent on COP which is related to the geothermal condition of the area. Since analysis for HHPS-G operation assumes that geothermal energy is sufficient for heating support, HHPS-G with a high COP still has the least energy demand among the three types of HPS. 7.2. GHG emissions comparison Because GHG emissions are dictated by energy consumption, they increase with an increase in snow period (Fig. 6). The increase in snow rates results in more GHG emissions from the HPS operations requiring more energy but has little effect on TSRS operation. Considering GHG emissions changes under different snow periods and rate conditions, three types of HPS operations will have less GHG emissions than TSRS applied in apron snow and ice removal under 12.7 mm/h snow rate conditions when the snow period is greater than 9 h. GHG emissions also depend on type of energy source. Since different energy sources have different emission factors, a system operation consuming more energy does not necessarily release more GHG than others. For example, HHPS-NG requires about 1.6 times more energy for snow and ice removal operation than EHPS; however, HHPS-NG releases only half the GHG. Also, although HHPS-G requires much less energy than the other snow and ice removal systems, HHPS-G with a COP of 2 can possibly release more GHG than the amount of GHG released from HHPS-NG, because natural gas combustion has a much lower GHG emission factor than electrical power generation, as Table 1 shows. Although it would not increase system efficiency, switching the energy source to natural gas could dramatically reduce GHG emissions. Most of the GHG emissions result from heating energy production in HPS and de-icer production and waste de-icer solution treatment in TSRS (Table 3). These results indicates that these inventory factors are the most sensitive ones to GHG emissions results than other inventory factors with uncertain values. Since GHG emissions are significantly positively correlated to energy consumption, the more energy used, the more GHG will be released, as shown in Tables 2 and 3. Using a similar strategy to reduce energy consumption of snow and ice removal operation, using less de-icer can be a significant way to significantly reduce GHG emissions in TSRS, and using a HPS instead of de-icing chemical application has the potential for reducing GHG emissions. In summary, analysis of energy consumption and GHG emissions from different snow and ice removal system operations show that, under lower snow rate and longer snowfall conditions, operations of HPS produce less energy consumption and GHG emissions than a TSRS operation.

327

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

Fig. 5. Energy consumption comparison of different snow and ice removal system operations against different snow periods: (a) under 12.7 mm/h (0.5 in./h) snow rate, (b) under 25.4 mm/h (1 in./h) snow rate, and (c) under 50.8 mm/h (2 in./h) snow rate.

Table 2 Operation energy contributions of different inventories in different snow and ice removal systems.

a–c d e

Energy consumption (%)

HHPS-G (COP = 3.6 as max)a

HHPS-G (COP = 2.0 as min)b

HHPS-NGc

EHPSd

TSRSe

Geothermal heat pump + circulating pump Natural gas furnace + circulating pump Electrically heating Deicer production + wastewater treatment Other

99.04 – – – 0.96

99.45 – – – 0.55

– 99.82 – – 0.18

– – 99.68 – 0.32

– – – 98.80 1.20

Other includes insulation layer production stage, antifreeze production stage, and antifreeze waste treatment stage. Other includes insulation layer production and carbon fiber production stage. Other includes diesel oil for mechanical equipment operation.

328

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

GHG emissions (kgCO2eq)

Snow rate: 12.7mm/h (0.5in/h)

12,000 10,000 8,000 6,000 4,000 2,000 0 1h

4h

12 h Snow period (h)

8h

HHPS-G COP=3.6 (Max.) HHPS-G COP=2.0 (Min.) HHPS-NG EHPS TSRS

(a)

GHG emissions (kgCO2eq)

Snow rate: 25.4 mm/h (1.0in/h)

12,000 10,000 8,000 6,000

HHPS-G COP=3.6 (Max.)

4,000

HHPS-G COP=2.0 (Min.)

2,000

HHPS-NG EHPS

0 1h

4h

8h

12 h

Snow period (h)

TSRS

(b)

GHG emissions (kgCO2eq)

Snow rate: 50.8 mm/h (2.0in/h)

12,000 10,000 8,000 6,000 HHPS-G COP=3.6 (Max.)

4,000

HHPS-G COP=2.0 (Min.)

2,000

HHPS-NG

0

EHPS

1h

4h

8h

12 h

Snow period (h)

TSRS

(c) Fig. 6. GHG emissions comparison of different snow and ice removal system operations against different snow periods: (a) under 12.7 mm/h (0.5 in./h) snow rate, (b) under 25.4 mm/h (1 in./h) snow rate, and (c) under 50.8 mm/h (2 in./h) snow rate.

Table 3 Operation GHG emissions of different inventories in different snow and ice removal systems.

a–c d e

GHG Emission (%)

HHPS-G (COP = 3.6 as max)a

HHPS-G (COP = 2.0 as min)b

HHPS-NGc

EHPSd

TSRSe

Geothermal heat pump + circulating pump Natural gas furnace + circulating pump Electrically heating Deicer production + wastewater treatment Other

99.67 – – – 0.33

99.81 – – – 0.19

– 99.77 – – 0.23

– – 99.93 – 0.07

– – – 97.78 2.22

Other includes insulation layer production stage, antifreeze production stage, and antifreeze waste treatment stage. Other includes insulation layer production and carbon fiber production stage. Other includes diesel oil for mechanical equipment operation.

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

329

8. Conclusions and recommendations Ineffective snow and ice removal activities can result in airline delays, employee injuries, and potential environmental risks from overuse of de-icers or anti-icers (FAA, 2010). As an industry with facilities that must pay attention to environmental impact and sustainability of its products or systems under conditions of increased environmental awareness, airports are seeking more sustainable systems able to replace conventional snow and ice removal system (FAA, 2010). This study was carried out with the specific goal of applying a hybrid LCA approach for evaluating energy consumption and GHG emissions from the operations of HHPS-G, HHPS-NG, and HHPS-E. The findings and future recommendations of the study are summarized below. 8.1. Findings  Heated pavement system apron application seems to be a viable option from an energy or environmental perspective for achieving pavement surfaces free of ice/snow without using mechanical or chemical methods. It is acknowledged that a thorough and detailed LCA that considers other environmental indicators (not just GWP) needs to be conducted to be able to conclusively justify the use of HPS for airport apron application.  De-icing chemical production requires a high energy demand and produces GHG emissions over a TSRS operation life cycle. Using heated pavement systems instead of de-icers thus enables effective snow and ice removal with reduced energy consumption and GHG emissions.  Energy demand and GHG emissions from operation of heated pavement systems are significantly determined by snowfall rate and snow period.  Compared to TSRS, heated pavement system operations have a greater advantage during a snow event with a small snow rate and a long snow period. It is acknowledged that an in-depth analysis is required to accurately quantify the impacts and benefits. This will be part of the future research.  Energy production (i.e., electrical power generation) and energy consumption phases (i.e., natural gas combustion) for heating require the most energy and contribute the most GHG emissions in a heated pavement system operation life cycle.  HHPS-G using geothermal heat pumps with a higher COP results in less energy consumption and fewer GHG emissions than other types of snow and ice removal systems under the same snow rate conditions. From an environmental impact perspective, hydronic heated pavement system using natural gas furnace, with high heating efficiency and a relatively low emission factor, has the potential to be used as a sustainable snow and ice removal system for places with lower potential for geothermal energy.  Analysis of energy consumption and GHG emissions from different snow and ice removal system operations show that, under lower snow rate and longer snowfall conditions, operations of HPS produce less energy consumption and GHG emissions than a TSRS operation. Although this study only focused on the operation phase of both heated pavement systems and traditional snow and ice removal systems, it provides a decision maker or airport manager a more informed view of operating HPS in removing snow in terms of energy saving and global warming potential aspects. However, it should be stressed that the theoretical models in this study used to calculate energy consumption and GHG emissions from different types of apron snow and ice removal systems are still under development, so the study’s results should be regarded as only a qualitative view, and more comprehensive assessments that include broader system boundaries are required for future study. 8.2. Recommendations  Based on the assumptions for system boundaries defined in this study, most of the energy is used for heating and causes relatively higher GHG emissions. Thus, heating source efficiency and COP are critical in HPS if they are to be more energyefficient and environmental-friendly.  Future studies may focus on different weather conditions, different de-icing chemical usage strategies, and other potential factors that might influence energy consumption and GHG emissions from different types of SRS.  The entire life cycle of a HPS, including construction and maintenance stages and a more comprehensive life cycle of traditional SRS, could be assessed to provide more information.  Previous studies have suggested that the use of de-icer chemicals on airport pavement surfaces tends to cause and/or accelerate distress and lead to more frequent repairs (Shi et al., 2009). Studying the full life cycles of SRS may reveal an increase in the energy spent during the pavement maintenance phase, so it will be interesting to study the life cycles of both SRS from this perspective.

Acknowledgements This paper was prepared from a study conducted at Iowa State University under the Federal Aviation Administration (FAA) Air Transportation Center of Excellence Cooperative Agreement 12-C-GA-ISU for the Partnership to Enhance General

330

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

Aviation Safety, Accessibility and Sustainability (PEGASAS). The authors would like to thank the current project Technical Monitor (Interim), Mr. Jeffrey Gagnon, and the former project Technical Monitor, Mr. Donald Barbagallo and Dr. Charles A. Ishee for their invaluable guidance on this study. Although the FAA has sponsored this project, it neither endorses nor rejects the findings of this research. The presentation of this information is in the interest of invoking comments by the technical community on the results and conclusions of the research. References Airliners, 2015. Aircraft Technical Data & Specifications Available from (28 February, 2015). Alternative Fuels Data Center (AFDC), 2014. Fuel Properties Comparison. U.S. Department of Energy, Washington, DC. Available from (28 February, 2015). Argonne National Laboratory (ANL), 2013. The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (GREET): Version 1, Chicago, IL. ASHRAE Handbook, 2003. HVAC Applications. Snow Melting and Freeze Protection. ASHRAE, Atlanta, GA, pp. 50.1–50.20 (Chapter 50). Ceylan, H., 2015. FAA PEGASAS COE project 1: heated airport pavements. In: Presented at FAA PEGASAS COE 3nd Annual Meeting. Purdue University, West Lafayette, IN (May 27–28, 2015). Clearwater Composites, LLC, 2015. Properties of Carbon Fiber, Duluth, MN Available from (18 February, 2015). Cryotech NAACÒ, 2008. Solid Commercial Deicer, Fort Madison, IA Available from (16 February, 2015). Das, S., 2011. Life cycle assessment of carbon fiber-reinforced polymer composites. Int. J. Life Cycle Assess. 16, 268–282. Energy Information Administration (EIA), 2013. Carbon Dioxide Emissions Coefficients, Washington, DC Available from (20 February, 2015). Energy Information Administration (EIA), 2015a. What Is U.S. Electricity Generation by Energy Source? EIA FAQ, Washington, DC. Available from (20 February, 2015). Energy Information Administration (EIA), 2015b. How Much Carbon Dioxide Is Produced Per Kilowatthour When Generating Electricity With Fossil Fuels? EIA FAQ, Washington, DC. Available from (24 February, 2015). Energy Star, 2015. Geothermal Heat Pumps Key Product Criteria, Washington, DC Available from (25 February, 2015). Environmental Protection Agency (EPA), 2012. Technical Development Document for the Final Effluent Limitations Guidelines and New Source Performance Standards for the Airport, EPA-821-R-12-005, Washington, DC. Available from: EPA Portal (4 March, 2015). Federal Aviation Administration (FAA), 1992. Airport Snow and Ice Control Equipment Publication FAA 150/5220-20A. U.S. Department of Transportation, Federal Aviation Administration, Washington, DC. Available from FAA Portal: Advisory Circular (12 March, 2015). Federal Aviation Administration (FAA), 2010. Memorandum. U.S. Department of Transportation, Washington, DC. Available from (21 March, 2015). Federal Aviation Administration (FAA), 2011. Airside Use of Heated Pavement Systems Publication FAA 150/5370-17. U.S. Department of Transportation, Federal Aviation Administration, Washington, DC. Available from FAA Portal: Advisory Circular (16 February, 2015). Federal Highway Administration (FHWA), 2014. Life Cycle Assessment of Pavements TechBrief FHWA-HIF-15-001. U.S. Department of Transportation, Federal Highway Administration, Washington, DC. Available from (6 May, 2015). Geest, V.J., Kiechle, C., 2010. Anaerobic Wastewater Treatment, No. 1. Brewing and Beverage Industry International, Retrieved from (17 February, 2015). Gagnon, L., Belanger, C., Uchiyama, Y., 2002. Life cycle assessment of electricity generation options: the status of research in year 2001. Energy Policy 30, 1267–1278. Gopalakrishnan, K., Steyn, W., Harvey, J. (Eds.), 2014. Climate Change, Energy, Sustainability and Pavements. Green Energy and Technology Book Series. Springer-Verlag, Berlin, Germany. Gopalakrishnan, K., Ceylan, H., Kim, S., Yang, S., Abdullah, H., 2015. Electrically conductive mortar characterization for self-heating airfield concrete pavement mix design. Int. J. Pavement Res. Technol. 8 (5), 315–324. Green Design Institute – Carnegie Mellon University (GDI-CMU), 2008. Economic Input-Output Life Cycle Assessment (EIO-LCA) Available from (28 March, 2015). Lamond, J.F., Pielert, J.H., 2006. Significance of Tests and Properties of Concrete and Concrete-Making Materials. STP169D, ASTM International, West Conshohocken, PA. MeltSnow, 2015. Acetates. Franklin, MA Available from (28 March, 2015). Mix, J., 2006. HVAC Efficiency Definition. Carrier Corporation, Syracuse, NY. Available from (28 March, 2015). National Energy Technology Laboratory (NETL), 2000. Life Cycle Analysis: Natural Gas Combined Cycle (NGCC) Power Plant. Publication DOE/NETL-403/ 110509, Pittsburgh, PA. Available from (15 October, 2015). Polyisocyanurate Insulation Manufactures Association (PIMA), 2015. Polyiso wall insulation boards Available from (25 March, 2015). Raypak, Inc., 2015. Anti-Freeze in Hydronic Systems, Oxnard, CA Available from (25 March, 2015). RPM Tech, 2015. The CameleonÒ: A Multifunctional Vehicle, Canada Available from (21 March, 2015). Scientific Applications International Corporation (SAIC), 2006. Life Cycle Assessment: Principles and Practice EPA/600/R-06/060. U.S. Environmental Protection Agency, Cincinnati, OH. Shen, W., 2015. Life Cycle Assessment of Heated Airfield Pavement System for Snow Removal MS Thesis. Iowa State University, Ames, IA. Shen, W., Gopalakrishnan, K., Kim, S., Ceylan, H., 2014. Greenhouse gas emission analysis for heated pavement system. In: Proceedings of the 2014 FAA Worldwide Airport Technology Transfer Conference, P10065, Galloway, NJ. Available from: FAA Portal: Conference-and-Workshop (24 February, 2015). Shen, W., Gopalakrishnan, K., Kim, S., Ceylan, H., 2015. Assessment of the greenhouse gas emissions from geothermal heated airport pavement system. Int. J. Pavement Res. Technol. 8 (4), 233–242. Shi, X., Akin, M., Pan, T., Fay, L., Liu, Y., Yang, Z., 2009. Deicer impacts on pavement materials: introduction and recent developments. Open Civ. Eng. J. 3, 16– 27. Available from (2 June, 2015). University of California Pavement Research Center (UCPRC), 2010. Pavement Life Cycle Assessment Workshop: Discussion Summary and Guidelines UCPRCTM-2010-03. University of California Pavement Research Center, Davis, CA. Available from: (6 May, 2015). University of South Carolina (USC), 2008. Formulations for Aircraft and Airfield Deicing and Anti-Icing: Aquatic Toxicity and Biochemical Oxygen Demand, Airport Cooperation Research Program (ACRP) Web-Only Document 3. Transportation Research Board. Available from (13 March, 2015).

W. Shen et al. / Transportation Research Part D 48 (2016) 316–331

331

Viega, 2015. S-no-IceÒ Snow Melting System Installation Manual, IM-SNO-01/05, Wichita, KS Available from (28 February, 2015). Washington State Department of Transportation (WS DOT), 2015. Metrics International System of Units, Olympia, WA Available from (28 April, 2015). Weber, C., Matthews, D., Venkatesh, A., Costello, C., Matthews, H.S., 2010. The 2002 US Benchmark Version of the Economic Input-Output Life Cycle Assessment (EIO-LCA) Model. Green Design Institute, Carnegie Mellon University. Available from (17 March, 2015). Wilo, 2009. Circulating Pumps, Dortmund, Germany Available from (25 February, 2015). WisDOT RD&T Program, 2004. Pavement Service Life-Rev.2. CTC & Associates LLC and Wisconsin Department of Transportation, Madison, WI. Available from (26 February, 2015).