Waste Management & Research2014, Vol. 32(8) 772 –781© The Author(s) 2014Reprints and permissions: sagepub.co.uk/journalsPermissions.navDOI: 10.1177/0734242X14543813wmr.sagepub.com

Introduction

The management of municipal solid waste (MSW) is gaining more attention because landfilling space near cities is becoming scarce and because it is recognized that valuable energy and materials can be recovered from waste streams. On average, about 2.5 kg of MSW per day per capita is generated in the US (Van Haaren et al., 2010) and about 1.4 kg per day per person is generated in Europe (Eurostat, 2013) and Japan. In 2011 it was estimated that 2 billion tonnes of waste were generated world-wide and this number is expected to grow to 2.9 billion tonnes by 2022 (Lawrence and Adamson, 2012). Apart from recycling, this growth rate can only be managed by two processes: landfill-ing or thermal conversion systems. Thermal processes are envi-ronmentally preferable to landfilling because they result in a large reduction in mass and volume of the waste (Malkow, 2004; Scholz et al., 1994) and because they concentrate inorganic con-taminants, making them easier to sequester and treat (Brunner and Rechberger, 2004). Thermal conversion systems include combustion, gasification or pyrolysis as a thermal pre-treatment technology. At this time, combustion of as-received MSW on a moving grate is the dominant thermal treatment process com-posing 84% of the total (Themelis et al., 2013). Recently, com-bustion systems have been shown to achieve up to 32% efficiency comprised of about 28% from electricity generation (Murer

et al., 2011). While increasing the efficiency of the plant opera-tion, the potential to increase the energy to the public is very promising. For example, in Vienna both electricity and heat are provided to the surrounding community. However, there is a potential to provide an additional 30–50 GWh year−1 for cooling purposes in the summer months (Ragoßnig et al., 2008). These and other systems with different efficiencies can be quantita-tively evaluated using a rigorous calculation model. The models vary by country but are similar in that they define a specific plant boundary and consider the energy (electricity and heat) produced from the process. For a good review on the topic see Gohlke (2009).

Gasification is an alternate thermal treatment method with the potential to be more efficient and versatile than complete

Technical assessment of the CLEERGAS moving grate-based process for energy generation from municipal solid waste

Marcella R Lusardi1, McKenzie Kohn2, Nickolas J Themelis1 and Marco J Castaldi2

AbstractA technical analysis has been completed for a commercial-scale two-stage gasification-combustion system. The CLEERGAS (Covanta Low Emissions Energy Recovery GASification) process consists of partial combustion and gasification of as-received municipal solid waste (MSW) on a moving grate producing syngas followed by full combustion of the generated syngas in an adjoining chamber and boiler. This process has been in operation since 2009 on a modified 330-tonne day−1 waste-to-energy (WTE) line in Tulsa, Oklahoma. Material balances determined that the syngas composition is 12.8% H2 and 11.4% CO, the heating value of the gas in the gasifier section is 4098 kJ Nm−3, and an aggregate molecular formula for the waste is C6H14.5O5. The analysis of gas measurements sampled from the Tulsa unit showed that the gasification-combustion mode fully processed the MSW at an excess air input of only 20% as compared to the 80–100% typically found in conventional WTE moving grate plants. Other important attributes of the CLEERGAS gasification-combustion process are that it has operated on a commercial scale for a period of over two years with 93% availability and utilizes a moving grate technology that is currently used in hundreds of WTE plants around the world.

KeywordsGasification, large-scale, municipal solid waste, syngas, moving grate, NOx reduction

1Earth Engineering Center, Columbia University, USA2 Earth Engineering Center, City College of New York, USA

Corresponding author:Marco J Castaldi, Chemical Engineering Department, Earth Engineering Center, The City College of New York, City University of New York, 140th Street, Convent Avenue, Steinman Hall, New York, NY 10031, USA. Email: mcastaldi@ccny.cuny.edu

543813WMR0010.1177/0734242X14543813Waste Management & ResearchLusardi et al.research-article2014

Original Article

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

Lusardi et al. 773

combustion. Typical gasification systems currently process hundreds of tonnes per day, whereas moving grate combustion systems process thousands of tonnes per day. Gasification is the partial combustion of a fuel, such as MSW, in sub-stoichi-ometric oxygen to produce H2 and CO, or syngas. Gasifiers can be used for heat and power production and may be heated directly by partial combustion of the MSW or indirectly by electricity or fuel combustion external to the system. The fuel (MSW) in gasifiers may be dried, sorted or shredded before gasification depending on the configuration. The energy in the syngas may be recovered by gas-cleaning followed by com-bustion in an external gas engine or turbine (power produc-tion), or combustion in the boiler followed by flue gas-cleaning producing heat or electricity with a steam cycle (Arena, 2012). There also is the possibility offered by some direct-melting system gasification plants of a complete and immediate reuti-lization of solid residues (Arena and Di Gregorio, 2013; Jung et al., 2005; Rocca et al., 2012; Parkpain et al., 2000). The pressure in the gasifier is typically atmospheric, but can be higher, and the temperatures range from 550°C to 900°C for air gasification. In addition to H2 and CO, the syngas product may contain CO2, H2O and CH4. For a comprehensive cover-age of MSW thermal conversion systems, the reader is referred to Combustion and Incineration Processes (Niessen, 2002) and Waste to Energy Conversion Technology (Klinghoffer and Castaldi, 2012).

There are many benefits to MSW gasification compared to combustion. The gasification reaction atmosphere reduces NOx (Malkow, 2004), thus reducing the burden on the down-stream air pollution control (APC) systems. Also, the resulting syngas can be burned more efficiently than solids in a gas tur-bine or boiler (Arena, 2012). The low excess air requirement should result in more compact and less costly waste-to-energy (WTE) plants. However, fouling and slagging of refractory materials and the boiler due to ash melting increases. There are many endothermic and exothermic reactions involved in the conversion of solid waste to syngas that occur simultane-ously (Arena, 2012). Therefore, it is valuable to have meas-urement of large-scale systems that operate on realistic fuels to further the understanding and development of the technology.

A pilot-scale reverse acting grate apparatus was developed and tested in 1994 to understand the influence of various oper-ating parameters on the gasification post-combustion process. Special emphasis was given to the decomposition of organic trace compounds and the measurement of polychlorinated dibenzodioxin/furan PCDD/F compounds utilizing an online resonance-enhanced multiphoton ionization mass spectrometer. The temperature, oxygen concentration and residence time was considered simultaneously. It was determined that a high tem-perature and mixing rate in the sub-stoichiometric region fol-lowed by near plug flow conditions downstream resulted in good organic and nitric oxide decomposition (Beckmann and Zimmermann, 1998).

Companies currently operating commercial-scale MSW gasi-fiers include AlterNRG, Ebara, EnEco, Energos, Hitachi Zosen, JFE, Thermoselect, Kobelco, Mitsui, Nippon Steel Engineering, Plasco Energy Group and Takuma (Arena, 2012). Most of the gasification systems in operation are atmospheric pressure and require fuel pre-treatment, such as separation or shredding (Malkow, 2004). Combustion of the syngas in an adjoining chamber is the most common method of energy recovery. However, a rare example of a system in which clean syngas is extracted from the system is the Thermoselect process. In this process as-received waste is compacted using a hydraulic press and pushed into an indirectly heated degassing channel at 600°C. The degassed waste then enters a high-temperature pyrolysis (i.e. oxygen-free) zone at 800°C, followed by high-temperature (1200–2000°C) gasification with oxygen of the pyrolysis prod-ucts. The syngas product is then cooled and cleaned and may then be used for external applications (Consonni and Vigano, 2012; Malkow, 2004). The production of clean syngas is an advantage to this system; however, there are energy losses involved with rapidly cooling the hot syngas and pre-heating the MSW in the degassing channel (Themelis, 2007).

More often the syngas is combusted internally to produce heat or electricity from a steam cycle. For example, the Ebara gasifier is an atmospheric pressure, low-temperature fluidized bed reac-tor (Arena, 2012). Shredded MSW is gasified in suspension with oxygen resulting in a fluidized bed reactor, and the produced syn-gas is combusted in a higher temperature cyclone furnace (Consonni and Vigano, 2012; Suzuki, 2007). The JFE direct-melting process utilizes a high-temperature system where oxygen gases flow from top to bottom (i.e. down-draft) through a packed bed of refuse-derived fuel (RDF).

Another example of a gasifier that does not require pre-treat-ment of MSW is the Direct Melting System of Nippon Steel Engineering. This is an atmospheric pressure, oxygen-enriched fixed bed down-draft gasifier. MSW is directly deposited into the gasification furnace along with solid carbon (coke) and limestone that act as a reducing agent and viscosity regulator, respectively (Tanigaki, 2012). The syngas is transferred to a downstream cham-ber for complete combustion and energy recovery. The Energos process also does not require pre-treatment of MSW or fuel addi-tives. This system is an atmospheric pressure, high-temperature (900°C) air gasifier with a moving grate reactor design. The syngas is produced in a gasification stage and then combusted in a second stage, called the gas-phase oxidation phase (Alamo et al., 2012). The process produces syngas with approximately half the oxygen required for complete combustion and with a gross calorific value of 4.34 MJ Nm−3 (Alamo et al., 2012; Consonni and Vigano, 2012).

The Covanta Low Emissions Energy Recovery Gasification (CLEERGAS®) process has been operated for over two years at a capacity of 330 tonnes day−1 of MSW by modifying one of the three units of the Tulsa, Oklahoma, WTE facility. This is the largest of the gasifier systems currently in operation and produces 40 metric tonnes (88,500 pounds) per hour of high-pressure (40 bar (670 psia); 346°C (656 °F)) process steam. The CLEERGAS process in

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

774 Waste Management & Research 32(8)

Tulsa consists of partial combustion (estimated at about 45%) and gasification of as-received MSW on a moving grate in the bottom half of the combustion chamber followed by full combustion of the syngas in the upper part of the chamber by injecting additional air. The design utilizes a conventional combustion grate and feed sys-tem yet stages the air to achieve gasification conditions in a targeted section of the boiler. The average NOx emissions vary between 60 and 90 mg Nm−3, with values as low as 45 mg Nm−3, and CO emis-sions are approximately 18 mg Nm−3. This work will present an analysis of the CLEERGAS process, including material balances, determination of the composition and heating value of the syngas and estimation of the chemical and thermal efficiency of the gasifi-cation stage and of the overall process determined by steam output, overall air input, boiler temperature and enthalpy of syngas.

Experimental methods

In order to test the CLEERGAS process at an industrial scale, Covanta modified one of the three units of their Tulsa, Oklahoma, WTE plant. Boilers 1 and 2 of this plant are conventional com-bustion chambers using under-fire and over-fire air injection. Boiler 3 was retrofitted by Covanta R&D to operate in a two-stage mode: the lower stage, extending to a few metres above the grate, is operated at a sub-stoichiometric air-to-fuel ratio that results in the drying of the MSW and the generation of syngas. This syngas is then combusted fully by the injection of additional air in the upper part of the chamber (Figure 1(a)).

During the observed operation period, the MSW supplied by trucks was stored in a conventional below-ground waste storage pit and supplied to the bins of the three boilers via overhead crane deliv-ery. Observations made from the crane-control deck confirmed that

the supply of feed to the CLEERGAS® unit (Boiler 3) was identical to that provided to the conventional units (Boilers 1 and 2). There was no pre-sorting, special waste selection, processing or special attention of any kind to the material fed into the gasification boiler. Also, a tour of the entire facility confirmed that the operations of the Tulsa facility are typical of conventional WTE operations.

A water-cooled probe and an Agilent/Inficon micro gas chro-matograph (GC) were used to analyse gas samples obtained from two locations in Boiler 3. Figure 1(b) shows the two ports where flue gas was extracted using a water-cooled probe and transferred to the on-site GC used for chemical species identification. Port B was located at the bottom of the boiler, directly above the gasifi-cation zone, while port A was near the top of the boiler, above the zone where secondary air was injected to combust the syngas.

Because the air-to-fuel ratio is a critical parameter in combus-tion and gasification systems, the equivalence ratio (φ) is used throughout this work to compare conditions throughout the boiler. The equivalence ratio is defined as the ratio of actual oxy-gen in the process to the oxygen required stoichiometrically for full combustion. Therefore, an equivalence ratio of less than one indicates that the process is deficient in air, promoting gasifica-tion, while a ratio greater than one indicates that the process is operating in excess air, promoting complete combustion.

ResultsAnalysis of gas sampling at port A (gasification stage)

Sampling port B was just above the waste bed and at the level of the lower “bullnose”, a protrusion into the furnace boiler that deflects the gas flow, as shown in Figure 1(b). The section of the

Figure1. A cross-section illustration of Boiler 3: (a) air-to-fuel variation from port B to port A; (b) location of the gas sampling ports B and A.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

Lusardi et al. 775

furnace between the grate and this port is considered to be the gasification zone and includes both the primary air flow under the bed and the secondary air injected above the bed. The total airflow is controlled to be below the stoichiometric amount required for complete combustion, giving an equivalence ratio of approximately 0.6.

Figure 2 shows two datasets of gas concentration profiles obtained through port B, as a function of distance from the fur-nace wall. The concentrations labelled “#1” indicate a previous dataset (Storm Technologies, Inc. North Carolina, USA), while concentrations labelled #2 were measured by the authors. Figure 2 shows that the maximum gas concentrations were at the centre-line of the furnace, 2.5 metres (8 feet) from the wall. The gas contained 12.8% H2, 11.4% CO and 3.7% CH4, that is, a H2/CO ratio of 1.1. Further from the centreline, closer to the sides of the boiler, the syngas and CH4 concentrations decrease, indicating a lesser extent of gasification. This is likely due to air bypass at the walls of the boiler, resulting in complete combustion closer to the sides. This air bypass is commonly observed where the boiler waterwall section interfaces with the grate section below. Since the grate is designed to move (to mix the waste) and is cooled by air upon operation the relative position of the boiler waterwall and the side-end of the grate can change, leading to some air bypass through that gap.

The later dataset obtained included H2 and CO concentra-tions as well as CO2 and O2 concentrations in the boiler. The goal during the testing was to obtain samples across the area of the boiler, but due to access limitations, only 0.625 metre (2 feet) and 1.25 metre (4 feet) penetrations were obtained. These data points are also shown in Figure 2 at 3.65 and 4.27 metres (12 and 14 feet) into the boiler. The H2 and CO con-centrations decreased closer to the boiler wall, consistent with the earlier dataset. As expected, O2 concentration was

higher closer to the boiler wall, indicating higher O2 content that would favour complete combustion over gasification at the edges of the boiler. Finally, CO2 concentration remains steady at 7.04 ± 0.08% at the 3.65 and 4.27 metre positions into the boiler.

Figure 3 shows the product gases as a function of time at the 4.27 metre (14 foot) position from the boiler wall. This test set consisted of inserting the probe approximately 0.610 metres (2 feet) into the side wall of port B and taking samples every 3 min-utes. The total test set took approximately 25 minutes. This time differential sequence provides insight into the fluctuations occur-ring within the gasification zone. These fluctuations are due to mixing, heterogeneity of the waste and air flow modifications. Figure 4 shows the concentration profiles of CO2, CO and H2 from this dataset as a function of O2 concentration.

Figures 3 and 4 show the trade-off between O2, CO2, H2 and CO. At low O2 values (0–11 minutes in Figure 3), CO2, CO and H2 values are higher and vice versa for the 14–23 minute time range. The timeframe reported (23 minutes) is representative of the type of fluctuation observed at this location near the wall. Nitrogen, methane and hydrocarbons were measured directly using the GC. The CO2 variation is directly matched by O2 variation as expected in this part of the boiler. This is shown more clearly in Figure 4; in general, as O2 concentration increases, CO2, CO and H2 concen-trations decrease, which is consistent with gasification conditions when air is utilized. Throughout the test set, the H2/CO ratio is approximately 0.59, which is very close to a typically desired ratio of 0.5 for many chemical synthesis processes, such as Fisher–Tropsch systems producing liquid fuels. This data shows that reaction is occurring and the presence of H2 and CO show that the waste is being partially gasified. The low concentration of H2 and CO relative to CO2 is due to the proximity of the boiler wall. The probe had a maximum penetration distance into the boiler from

H2

CO

CH4

CO2

O2

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

0.5 1 1.5 2 2.5 3 3.5 4 4.5Distance into Boiler (m)

Con

cent

rati

on (

Mol

e %

)

H2, #1

CO, #1

CH4, #1

CO2, #2

O2, #2

CO, #2

H2, #2

H2

CH4

CO2

O2

H2

Figure 2. Concentration profile of H2, CO, CH4, O2 and CO2 in the boiler (#1 indicate Storm dataset; #2 measured by the authors).

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

776 Waste Management & Research 32(8)

the sidewall of approximately 1.22 metres (4 feet) of the overall (16 feet); therefore, a larger than expected amount of O2 was observed. This O2 is likely from air bypass near the wall of the boiler or air intrusion from the outside and is consistent with the higher oxygen and lower syngas concentrations for these meas-urements. This also highlights the importance of airtight construc-tion when erecting large-scale gasification systems. The variation in H2 and CO concentrations at relatively constant O2 concentra-tions is due to mixing at the boiler wall.

Figure 5 shows a similar test set with an extension attached to the probe to 1.22 metre (4 feet) penetration into the boiler from

the same port B, corresponding to the 3.65 metre (12 foot) posi-tion in Figure 2. Again a series of samples were taken at a fixed probe position over a time period of about 23 minutes (approxi-mately 3 minutes per sample). O2, CO2, CO, H2 and N2 concen-trations are shown as a function of time. Figure 6 shows the concentration profiles of CO2, CO and H2 from this dataset as a function of O2 concentration. Here again, the CO concentration was slightly higher than the H2, resulting in a H2/CO ratio of 0.83. Also, similar to the 4.27 metre (14 foot) position, as O2 concentration increased, on average, H2, CO and CO2 concentra-tions decreased, indicating reduced partial oxidation of the MSW.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0 5 10 15 20 25

Time (minute)

Gas

Mol

ar C

once

ntra

tion

(%

)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

N2 C

oncentration (mole %

)

CO2 CO O2 N2

N2

CO2

O2

H2

CO

CO2 H2 O2 N2

Figure 3. Lower level boiler species concentrations at the 4.27 metre (14 foot) location. N2 is on the secondary axis.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

O2 molar concentration (%)

Gas

Mol

ar C

once

ntra

tion

(%

)

CO2 CO H2CO2

H2

CO

H2CO2

Figure 4. H2, CO and CO2 in the lower boiler as a function of O2 concentration at the 4.27 metre (14 foot) location.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

Lusardi et al. 777

Table 1 summarizes the results in Figures 3–6, showing aver-age values and standard deviations for O2, N2, CO, H2, CH4 and CO2 at two and four feet from the boiler wall, corresponding to 3.65 and 4.27 metre (12 and 14 foot) positions in Figure 2. It should be noted the GC was capable of measuring down to parts

per million mole fraction reliably, thus the many decimal places for methane.

In addition to the CO2, CO and H2 produced in the lower level of the boiler, smaller amounts of hydrocarbons were also meas-ured. Because gasification occurs in a lean-oxygen atmosphere

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

O2 molar concentration (%)

Gas

Mol

ar C

once

ntra

tion

(%

)

CO2 COCO2

H2

CO

CO2 H2

Figure 6. H2, CO and CO2 in the lower boiler as a function of O2 concentration at the 3.65 metre (12 foot) location.

Table 1. Flue gases from port B (gasification zone) at 3.65 and 4.27 metre (12 and 14 foot) positions in the boiler.

Location O2 (%) N2 (%) CO (%) H2 (%) CH4 (%) CO2 (%)

Port B, 3.65 m 2.81±1.11 68.98±2.61 2.68±1.04 2.36±1.16 0.00012±0.00034 7.12±0.43Port B, 4.27 m 4.01±1.51 70.91±2.53 1.56±1.36 1.15±1.21 0.00022±0.00035 6.97±0.56

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0 5 10 15 20

Time (minute)

Gas

Mol

ar C

once

ntra

tion

(%

)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

N2 C

oncentration (mole %

)

CO2 CO O2 N2

N2

CO2

H2

CO

O2

CO2 H2 O2 N2

Figure 5. Lower level boiler species concentrations at the 3.65 metre (12 foot) location. N2 is on the secondary axis.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

778 Waste Management & Research 32(8)

(equivalence ratio less than one), higher order hydrocarbons (HOHCs) that would be combusted in an oxygen-rich combus-tion atmosphere remain either unreacted in the gas phase or are aerosolized and can create issues downstream if not reacted or removed. Table 2 shows the range of HOHCs that were present in the gasification regime of Boiler 3, grouped from C2 to C9. The largest concentration of HOHC was of nonanes, between 0.6% and 2.0%. Two, three, seven and eight carbon species were the next most prominent HOHC, of the order of magnitude of 1000 ppm. Finally, four, five and six carbon species were at the lowest concentration (about 100 ppm). The presence of these HOHCs again confirms that gasification is occurring in Boiler 3 and points to the vigilance that must be maintained during operation of such as system. The equivalence ratio and temperature associ-ated with gasification create an environment conducive to the release of HOHCs from the MSW feed. Some of these species remain in the gas phase, while others form particulates and lodge onto the boiler wall, which are responsible for the smooth tar (highly recalcitrant carbonaceous material) layers on the inner wall of the boiler in the port B region. Neither the tar nor these species are present in the upper level of the boiler, as expected, as

the oxygen-rich combustion environment is capable of oxidizing the HOHCs into CO2 and H2O.

Upper boiler gas analysis

Gas analysis from the upper section of the boiler, where complete combustion was targeted, resulted in the values shown in Figures 7 and 8 and Table 3. Figure 7 shows an analysis of the temporal data at 1.8 metre (6 foot) penetration horizontally into the boiler and commensurate changes in CO2. The concentrations are quite constant due to the good mixing conditions of the secondary air injection. The data showed the expected trade-off between CO2 and O2: as CO2 is produced, O2 is consumed, as shown in Figure 8. The major species present in the upper section of the boiler are inert N2, unreacted O2, CO2 and H2O. CO is present at the ppm level. This confirms that complete combustion of syngas occurs in the upper part of the gasification-combustion chamber.

Table 3 summarizes the results for the gas species in the upper boiler, showing average values and standard deviations for O2, N2, CO, H2, CH4 and CO2 at 2.4, 3.05, 3.65 and 4.27 metre (8, 10, 12, 14 foot) positions in the boiler. The N2, O2 and CO2 values do not show much variation, indicating that the upper complete oxi-dation zone of the boiler is more homogeneous. Furthermore, the HOHCs observed in the lower section of the boiler were not observed in the upper section, indicating that they are completely combusted in the upper section.

Material balances

Material balances were conducted based on the gas measurements discussed above. This was done because the scale of the unit (330 tonnes day−1) does not enable a representative sample to be taken. The waste is transferred from collection trucks to a holding pit where it is then crane-fed to the top of the gasification boiler

Table 2. The higher order hydrocarbon species present in the lower level of boiler 3.

Species Concentration (± 50%)

C2 1620 ppmC3 1730 ppmC4 15 ppmC5 110 ppmC6 50 ppmC7 2850 ppmC8 990 ppmC9 1.3%

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0 5 10 15 20

Time (minute)

Gas

Mol

ar C

once

ntra

tion

(%

)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

N2 C

oncentration (mole %

)

CO2 CO O2 N2

O2

H2CO

N2

CO2

CO2 H2 O2 N2

Figure 7. Upper level boiler species concentrations at the 1.8 metre (6 foot) location. N2 is on the secondary axis.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

Lusardi et al. 779

untreated. The amounts of MSW and air supplied to the plant were provided by Covanta. The molecular formula of the MSW was determined using the overall combustion reaction with added air and assumed moisture content of the waste at 20%:

C H O O CO H O gx y z 2 2 2+ → + ( )β γ δ

along with the following simultaneous equations:

y 2= δ

z 2 2+ = +β γ δ

x = γ

Additional constraints in this analysis were that the composite chemical should match the higher heating value for typical MSW and the C/H, C/O and H/O molar ratios should match as closely as possible to measured values. In this case the ratios from the calculations were C/H = 0.41, C/O = 0.31 and H/O = 0.76. The calculation resulted in a molecular formula of C6H14.5O5, which is close to the formula C6H10O4 developed by Themelis et al. (2002). In that study, the chemical thermodynamics of the MSW combustion reaction were modelled by representing the compos-ite combustible MSW by an established hydrocarbon compound. It is interesting to note that on the basis of the above formula and

the assumption that the Tulsa MSW contains 30% carbon, which is typical of US MSW received at WTE plants (Bahor et al., 2009), the heating value of this MSW is calculated to be 11 MJ kg−1, comparable to the 10.5 MJ kg−1 provided by Covanta, which is typical of US MSW according to previous studies (Themelis et al., 2013).

For the gas mixtures present in the lower and upper portion of the boiler, shown in Tables 1 and 3, respectively, H2O con-centration, adiabatic temperature, equivalence ratio, heating value and atomic molar ratios were calculated. These values are provided in Table 4. It should be noted the location of port B only allowed for a short probe to be inserted into the boiler; therefore, the locations reported for port A enable direct com-parisons with the port B location and connection to Table 3. The H2O concentration was calculated because water vapour in the sample gas was condensed prior to introduction into the GC, thus dry samples were taken. The temperature shown is the esti-mated adiabatic temperature inside the chamber. The equiva-lence ratio is defined as the O2 concentration injected into the boiler in comparison to the stoichiometric concentration required for complete combustion. Values less than one indicate low O2 atmospheres that favour gasification, while higher val-ues favour complete combustion.

The equivalence ratio results are highly significant because they demonstrated that full combustion was achieved in the pilot unit with only 20% excess air versus the conventional operation

Table 3. Flue gases from port A (oxidation zone), at 2.4, 3.05, 3.65 and 4.27 metre (8, 10, 12 and 14 foot) distances in the boiler.

Location O2 (%) N2 (%) CO (%) H2 (%) CH4 (%) CO2 (%)

Port A, 2.4 m 7.48±1.20 71.30±0.59 0 0 0 5.49±0.68Port A, 3.05 m 3.98±0.91 72.24±0.47 0 0 0 7.22±0.35Port A, 3.65 m 5.63±0.55 71.95±0.19 0 0 0 6.54±0.21Port A, 4.27 m 6.97±0.41 71.34±0.20 0 0 0 5.87±0.15

4.5

5.0

5.5

6.0

6.5

7.0

6.0 6.5 7.0 7.5 8.0 8.5

O2 Concentration (mole %)

CO

2 C

once

ntra

tion

(m

ole

%)

Figure 8. CO2 in the upper boiler as a function of O2 concentration at the 1.8 metre (6 foot) location.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

780 Waste Management & Research 32(8)

where excess air is of the order of 80–100%. Port A, which is at

the top of the boiler, near the outlet of the first pass, shows φ

values ranging between 1.13 and 1.24 or between 13% and 24%

excess air. The heating value of the syngas was determined to be

approximately 616 kJ per normal cubic metre for the average

measurements taken with the 1.22 metre (4 foot) penetration

probe at port A. The data obtained by Storm Technologies (shown

in Figure 8) results in a syngas heating value of 4098 kJ per nor-

mal cubic metre near the centreline of the gasification chamber.

These are typical heating values of syngas produced via air

gasification.

Based on the calculated formula of MSW and the feed rate of

MSW into Boiler 3, the theoretical maximum values of CO and

H2 that could be produced can be calculated using C and H bal-

ances. For every tonne of MSW processed, 1.00 tonne of CO and

0.88 tonnes of H2 can be produced theoretically. Assuming that

the mass flow rate in the lower level of the chamber is equal to

that in the upper level, this corresponds to maximum theoretical

concentrations of 13.0% and 10.1% for CO and H2, respectively.

This compares well to the data taken by Storm at the boiler cen-

treline shown in Figure 2 where the CO and H2 were measured

at 11.4% and 12.8%, respectively. The actual concentrations of

CO and H2 measured by the Earth Engineering Center (EEC)

were 2.3% and 1.0%, respectively, 1.22 metres (4 feet) into the

boiler for comparison. At the edge of the boiler, primary air

bypasses the MSW/grate, creating an oxygen-rich atmosphere in

an already low syngas environment. As a result, this air fully

oxidizes a large amount of the syngas in the region and yields

lower CO and H2 percentages than theoretically expected. This

is shown by the comparatively large amounts of O2 (2 mol%)

and CO2 (7 mol%) present in the “syngas” stream. A correspond-

ing equilibrium calculation showed that a temperature of 950 K

or greater must be maintained in the gasification stage of the

pilot unit in order to obtain significant gasification products.

Utilizing the determined MSW formula with an inlet tempera-

ture of 300 K, a chemical equilibrium calculation at a defined

temperature and pressure was performed. The pressure was held

at one bar and the temperature was changed until the calculated

gaseous concentrations were close to those measured. A lower

temperature would result in higher concentrations of the oxida-

tion products of CO2 and H2O and lower amounts of CO and H2.

Emissions impactThe operation of Boiler 3 in a gasification mode was shown to have a positive effect on emissions, resulting in a lower air pollu-tion control (APC) system load downstream. In particular, the NOx production was significantly lower due to the reducing envi-ronment just above the waste bed. The NOx emissions (@7% O2) during the testing from Boiler 3 were reported by Covanta to be 60–90 mg Nm−3, which is nearly 40% below normal operation with normal CO emission approximately 18 mg Nm−3. Importantly, the particulate matter (PM) or dust measurements were approximately 10% of the total ash generated, which is sig-nificantly less than normal excess air operation. This is consistent with the overall airflow reduction of 38% reported by Covanta during gasification. The under-grate air is controlled to be sub-stoichiometric in relation to the amount of waste on the grate, resulting in complete oxygen utilization and leaving partially oxidized products (CO and H2), lighter hydrocarbons (CH4, C2H4, etc.) and lower concentration of nitrogen oxides. These emission results are achieved with an estimated overall boiler temperature increase of 400 K, leading to an estimated overall thermal efficiency improvement of approximately 5%. In con-trast, a conventional WTE boiler maintains an oxidizing environ-ment through and above the waste bed, which does not allow in-situ NOx reduction.

The main conditions that contribute to NOx formation in WTE boilers are intimate mixing between the gases emanating from the bed and the air flow introduced to the furnace off-gases in the waste and the chemical composition in the gas phase immedi-ately above the bed. While intimate mixing of the air within the fuel bed is desired from a combustion perspective, it is undesir-able from a NOx formation perspective. Since most of the NOx is formed from fuel-bound nitrogen, it is important to maintain a reducing environment immediately above the waste bed. Goh et al. (2003) demonstrated this effect via modelling and sug-gested that the waste bed should have a char portion to promote NOx reduction. Also, EEC researchers Frank and Castaldi (2011) have shown via computational fluid dynamics (CFD) coupled to detailed chemical kinetic modelling that it is imperative to pro-vide sufficient time for the fuel-bound nitrogen to form N2, which then passes through the secondary air injection stage as an inert gas. Therefore, the avoidance of NOx formation immediately

Table 4. H2O calculation, adiabatic temperature calculation, equivalence ratio, syngas heating value, C/H, C/O and H/O ratios for the boiler locations analysed.

Location H2O (%) calc

T ad (K)

φ Syngas HV kJ m−3

C/H C/O H/O

Port B, 0.610 m 15.40 1540 1.08a 616.29 0.26 0.22 0.85Port B, 1.22 m 16.04 1540 1.09a 616.27 0.27 0.25 0.95Port A, 0.610 m 15.82 1375 1.22 N/A 0.19 0.14 0.76Port A, 1.22 m 15.87 1447 1.16 N/A 0.21 0.16 0.79Port A, 1.83 m 16.56 1485 1.13 N/A 0.22 0.19 0.85Port A, 2.44 m 15.73 1353 1.24 N/A 0.17 0.13 0.76

aφ is slightly above one due to the proximity of the sample near the boiler wall.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from

Lusardi et al. 781

above the waste bed from fuel-bound nitrogen is one of the most critical steps that can be taken to reduce overall NOx formation in WTE boilers.

Conclusions

The CLEERGAS® process has been tested extensively on an industrial scale (330 tonnes day−1). The process consists of two stages: the lower stage, extending to just above the grate, is oper-ated at a sub-stoichiometric air-to-fuel ratio that results in the gen-eration of syngas. The syngas produced in the lower part of the chamber is then combusted fully by injecting additional air in the upper part of the chamber. Gasification in the lower part of the unit and subsequent combustion of the syngas in the upper part resulted in full combustion of as-received MSW at an excess air input of only 20%, versus conventional WTE grate combustion units where 80–100% excess air is typically used. Another important attribute of the CLEERGAS® process is that gasification takes place on a moving grate, a technology that has been tested, proven and used in over 600 existing WTE plants around the world.

In the prototype of the process, now under design by Covanta Energy, the amount of air infiltration in the gasification chamber will be reduced and the conditions for combustion of the syngas in the separate gas combustion unit will be improved. Therefore, it is expected that the amount of excess air required in the proto-type will be substantially lower than in conventional grate com-bustion WTE plants. Lower excess air will result in higher thermal efficiency of the process and, most importantly, in lower capital and operating costs per tonne of MSW processed. The two-stage process of gasification followed by syngas combustion also enables better control of NOx generation by properly design-ing the air injection to the syngas combustion chamber.

AcknowledgementsThe cooperation of Covanta Energy and the Tulsa, Oklahoma, WTE plant in providing data and hosting the researchers during their col-lection of operating data and measuring gas compositions in the gasification-combustion chamber are gratefully acknowledged.

Declaration of conflicting interests None declared.

FundingThis work was supported by the EEC of Columbia University and City College of New York.

ReferencesAlamo G, Hart A, Grimshaw A, et al. (2012) Characterization of syngas pro-

duced from MSW gasification at commercial-scale ENERGOS plants. Waste Management 32: 1835–1842.

Arena U (2012) Process and technological aspects of municipal solid waste gasification. A review. Waste Management 32: 625–639.

Arena U and Di Gregorio F (2013) Element partitioning in combustion- and gasification-based waste-to-energy units. Waste Management 33(5): 1142–1150.

Bahor B, Weitz K and Szurgot A (2009) Analysis of GHG emissions from MSW combustion. In: proceedings of the global waste management symposium.

Beckmann M and Zimmermann R (1998) Gasification-post-combustion of waste materials; influencing parameters and on-une monitoring of organic substances. In: AIChE-PTF topical meeting “advanced technolo-gies for particle processing”, 15–20 November, Miami Beach (USA), pp.661–671.

Brunner PH and Rechberger H (2004) Practical Handbook of Material Flow Analysis. Advanced Methods in Resource and Waste Management. Boca Raton, FL: CRC/Lewis, p.318

Consonni S and Vigano F (2012) Waste gasification vs. conventional waste-to-energy: A comparative evaluation of two commercial technologies. Waste Management 32: 653–666.

Eurostat (ed.) (2013) env_wasgen; v2.10.4.6–20131018–5035-PROD_EUROBASE, available at: http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Waste_indicators_on_generation_and_landfilling_-_monitoring_sustainable_development#.

Frank A and Castaldi MJ (2011) Numerical modeling of pollution formation in waste-to-energy systems using computational fluid dynamics. In: 19th North American waste to energy conference, NAWTEC 19–5450.

Goh YR, Zakaria R, Yang YB, et al. (2003) Reduction of NOx during incin-eration of municipal solid waste by a fundamental combustion technique. Journal of the Institute of Energy 76(508): 72–79.

Gohlke O (2009) Efficiency of energy recovery from municipal solid waste and the resultant effect on the greenhouse gas balance. Waste Management & Research 27(9): 894–906.

Jung CH, Matsuto T and Tanaka N (2005) Behavior of metals in ash melt-ing and gasification-melting of municipal solid waste (MSW). Waste Management 25(3): 301–310.

Klinghoffer NB and Castaldi MJ (2012) Waste to energy conversion technol-ogy. Energy 29: 256.

Lawrence M and Adamson K (2012) Waste-to-energy technology markets. Renewable power and heat generation from municipal solid waste: Market outlook, technology assessments, and capacity and revenue fore-casts, Pike Research.

Malkow T (2004) Novel and innovative pyrolysis and gasification tech-nologies for energy efficient and environmentally sound MSW disposal. Waste Management 24: 53–79.

Murer MM, Hartmut Spliethoff H, de Waal CMW, et al. (2011) High effi-cient waste-to-energy in Amsterdam: Getting ready for the next steps Waste Management & Research 29(10): S20–S29.

Niessen W (2002) Combustion and Incineration Processes. 3d Edition. Volume 25. New York: Marcel Dekker Inc.

Parkpain P, Sreesai S and Delaune RD (2000) Bioavailability of heavy met-als in sewage sludge amended Thai soils. Water, Air and Soil Pollution 122: 163–182.

Ragoßnig AM, Wartha C and Kirchner A (2008) Energy efficiency in waste-to-energy and its relevance with regard to climate control. Waste Management & Research 26(1): 70–77.

Rocca S, van Zomeren A, Costa G, et al. (2012) Characterisation of major component leaching and buffering capacity of RDF incineration and gasification bottom ash in relation to reuse or disposal scenarios. Waste Management 32(4): 759–768.

Scholz R, Beckmann M and Schulenburg F (1994). Experimental research on gasification of coarse waste on a Stoker system and separate after burning as well as optimization with the aid of a process model. In: proceedings of the 1994 incineration conference, Houston, Texas.

Suzuki S (2007) The Ebara advanced fluidization process for energy recovery and ash vitrification. In: 15th NAWTEC, May, Miami, FL.

Tanigaki N (2012) Co-gasification of municipal solid waste and mate-rial recovery in a large-scale gasification and melting system. Waste Management 32: 667–675.

Themelis N (2007) Thermal treatment review. Waste Management World 8(4): 37–45.

Themelis NJ, Diaz-Barriga ME, Estevez P, et al. (2013) Guidebook for the application of waste to energy technologies in Latin America and the Caribbean, InterAmerican Development Bank, New York, USA, WTE Guidebook for Latin America. Available at: www.wtert.org.

Themelis NJ, Kim YH and Brady MH (2002) Energy recovery from New York City municipal solid wastes. Waste Management & Research 20(3): 223–233.

van Haaren R, Themelis NJ and Goldstein N (2010) 17th nationwide sur-vey of MSW management in the U.S.: The state of garbage in America. Biocycle, October, 16–23.

at COLUMBIA UNIV on September 2, 2015wmr.sagepub.comDownloaded from