Methane and carbon dioxide emissions from
Shan-Chu-Ku landfill site in northern Taiwan
Ullas Hegde
1, Tsan-Chang Chang, Shang-Shyng Yang
*Department of Biochemical Science and Technology, National Taiwan University, Taipei 10617, Taiwan, ROC Received 19 February 2003; received in revised form 10 March 2003; accepted 13 March 2003
Abstract
To investigate the methane and carbon dioxide emissions from landfill, samples were taken of material up to 5 years old from Shan-Chu-Ku landfill located in the northern part of Taiwan. Atmospheric concentrations of carbon dioxide, methane and nitrous oxide ranged from 310 to 530, 2.64 to 20.16 and 0.358 to 1.516 ppmv with the measurement of gas-type open-path Fourier transform infra-red (FTIR) spectroscopy during February 1998 to March 2000, respectively. Average methane emission rate was 13.17, 65.27 and 0.99 mg m2h1measured by the gas chromatography chamber method in 1–2, 2–3 and 5 year-old landfill, respectively. Similarly, average carbon dioxide emission rate was 93.70, 314.60 and 48.46 mg m2h1, respectively. About 2–3 year-old landfill had the highest methane and carbon dioxide emission rates among the tested areas, while 5 year-old landfill was the least. Methane emission rate at night in most tested locations was higher than that in the daytime. Total amount of methane and carbon dioxide emission from this landfill was around 171 and 828 ton in 1999, respectively.
2003 Elsevier Ltd. All rights reserved.
Keywords: Fourier transform infra-red spectroscopy; Greenhouse gas; Landfill; Methane and carbon dioxide emission rate
1. Introduction
Anaerobic decomposition of landfilled solid waste generates significant amounts of greenhouse gas com-prising 60% methane and 40% carbon dioxide (v/v), together with numerous trace gases. Landfills have been implicated as the largest anthropogenic source of at-mospheric methane in the world (Bogner et al., 1997) and as a significant contributor to global warming in greenhouse gas scenarios. On a global scale, approxi-mately 653 Tg yr1 of waste is landfilled and annual
global methane emission from landfills ranged from 9 to 70 Tg (Kreileman and Bouwman, 1994).
Compared to wetlands and paddy fields, landfills function as a closed system because of controlled burial of biodegradable organic materials. The management, composition, moisture content and pH of the refuse influence methane and carbon dioxide production in landfills. Methanogenesis occurs when the organic waste pH is between 6.8 and 7.4 and is stimulated with in-creasing moisture content (Gurijala and Sulflita, 1993; Jang and Yang, 2001). On-field measurements of green-house gas concentrations and emissions from landfill are surprisingly few, despite the fundamental physical dif-ferences between aquatic and terrestrial ecosystems. The Shan-Chu-Ku landfill is the main landfill in northern Taiwan. The methane and carbon dioxide emission pattern of the Shan-Chu-Ku landfill is different from other landfills because of differences in refuse composi-tion, landfill management, environmental conditions and methane oxidation activity by microorganisms www.elsevier.com/locate/chemosphere
*
Corresponding author. Tel.: 2-2362-1519; fax: +886-2-2367-9827.
E-mail address:ssyang@ccms.ntu.edu.tw(S.-S. Yang).
1
Permanent address: Post Graduate School for Biological Studies, Ahmednagar College, Ahmednagar 414 001, Mahara-stra State, India.
0045-6535/03/$ - see front matter 2003 Elsevier Ltd. All rights reserved.
present in the soils covering the landfill (Liao et al., 1998; Kuo et al., 2000). In this study, we have measured the methane and carbon dioxide emissions from mu-nicipal solid waste (MSW) material of various ages, 1–2, 2–3 and 5 year-old, of the Shan-Chu-Ku landfill and the seasonal variation of greenhouse gas concentration in the 2–3 year-old material has also been studied during February 1998–March 2000.
2. Materials and methods 2.1. Site description
The Shan-Chu-Ku landfill (25020N, 120320E) is situated near Taipei City in northern Taiwan. It covers 65 ha of area, out of which 30 ha are currently in use for landfilling of MSW. It was started in June 1994 and is predicted to be in use for at least 10 years. At present it receives daily approximately 1400 ton of MSW (it re-ceived 1100 ton of MSW at the initial period). The compositions of MSW are listed in Table 1. Sampling locations of material deposited 1–2, 2–3 and 5 years ago were chosen in this study. The properties of cover soils at different burial periods are also shown in Table 2. The campus of National Taiwan University (NTU) (25010N, 120310E) is also located in Taipei City. The grass area in the open field near the library is used for comparison of the gas chromatography (GC) and
Fourier transform infra-red (FTIR) spectroscopy as methods of measuring atmospheric methane and carbon dioxide concentrations.
2.2. Gas-type open-path FTIR spectroscopy measurement FTIR spectroscopy method was used to investigate the concentrations of atmospheric greenhouse gas as follows: with a height 197 cm above ground and a scan number 100 in 7 min, absorbance at wave numbers be-tween 2239 and 2393 cm1 for carbon dioxide, 2999 cm1for methane, and between 2214 and 2236 cm1for nitrous oxide measurement. The distance between the light source and the reflecting mirror was 25 m. The details of the experimental set-up adopted for FTIR spectroscopy method was described in our previous paper (Chang et al., 2000). As the interferometer char-acteristics dominated by a high spectral resolution up to 0.06 cm1 and a configuration suitable for field mea-surements, the MB-104 (BOMEM, Hartmann & Braun, Canada) is used for radiation emission as well as ab-sorption measurement. Observing the spectrum on a PC screen and adjusting the mirror achieve signal optimi-zation. The atmospheric concentrations of greenhouse gas from long-path measurement are determined by the differential absorption, a least square fit of measure-ment, and simulated by modeling the air transmittances for various absorbed concentrations.
Table 1
Properties of MSW produced in Taiwan Year 1992 1995 1997 1999 1999a Food waste (%) 25.73 17.94 24.90 21.83 40.29 ± 6.36 Paper (%) 24.86 32.17 29.13 35.83 33.11 ± 4.33 Plastic (%) 19.14 18.27 19.57 19.85 15.10 ± 2.22 Wood (%) 5.06 5.82 4.86 4.89 3.47 ± 2.24
Textile & litter (%) 3.97 6.21 5.80 5.20 1.23 ± 1.26
Leather (%) 1.73 0.88 1.13 0.60 0.95 ± 0.43
Glass (%) 7.69 5.59 4.95 4.99 2.89 ± 2.11
Metal (%) 7.07 6.05 5.33 3.80 2.88 ± 3.27
Porcelain (%) 0.83 1.64 1.26 0.51 0.06 ± 0.06
Others (%) 3.93 5.45 3.07 2.50 0.02 ± 0.01
aMSW produced in Taipei City, means ± S.D.
ðn ¼ 3Þ.
Table 2
Properties of cover soils in the Shan-Chu-Ku landfill Disposal
period (year)
Texture Soil depth
(cm) Moisture content (%) Soil temperature (C) Organic carbon (%) Total nitrogen (%) Vegetation
1–2 Loam, sandy loam 130 23.09 ± 4.02 21.5 ± 3.2 6.73 ± 0.76 0.29 ± 0.05 Grass, fern
2–3 Loam, clay loam 140 22.72 ± 3.98 22.4 ± 4.1 8.40 ± 0.62 0.31 ± 0.07 Grass, fern
5 Loam 180 22.16 ± 2.25 21.0 ± 2.8 6.30 ± 0.46 0.26 ± 0.05 Grass, fern,
small shrubs
Means ± S.D. (n¼ 15).
2.3. Carbon dioxide and methane emission rate
Gas samples were collected using a homemade closed acrylic chamber (length 40 cm, width 40 cm and height 65 cm, about 96 l) that was equipped with an electronic fan, a thermometer and a sampling hole. For measuring carbon dioxide and methane emission rates, four acrylic chambers were installed on the surface of landfill soils in each measurement. Grass and small plants are cut from the cover soil to avoid plant respiration and photosyn-thesis. The chambers were placed on the soil surface with 5 cm inserted into the soil 10 min prior to each sampling for equilibration to reduce the disturbance to the sampling site. 35 ml gas samples were collected from the headspace of the chamber and transferred to a 12.6 ml serum bottle that had been sealed by a butyl rubber stopper and flushed with oxygen-free nitrogen gas (Chang and Yang, 1997). Gas samples were collected from the chamber at 30-min intervals for 1 h, and sampled again 3 h later for comparison. The methane and carbon dioxide concentration and the accumulated time from 0 to 2 h in four chambers had linear rela-tionship and correlation coefficient r2¼ 0:99 and 0.98, respectively. The rates of carbon dioxide and methane emissions were calculated by fitting linear regression to the difference in the carbon dioxide and methane con-centrations and adjusting for the chamber volume and area covered following the equation (Rolston, 1986): f¼ ðV =AÞðDC=DtÞ
where f is the carbon dioxide or methane emission rate (mg m2h1); V , the volume of chamber above soil (m3); A, the cross-section of chamber (m2); DC, the concen-tration difference between time zero and time t (mg m3) and Dt, the time duration between two sampling periods (h). The total carbon dioxide or methane emission from landfills was the summation of carbon dioxide and methane emissions in all year (Yang and Chang, 1998). Atmospheric carbon dioxide and methane were also analyzed by GC method. Gas sample was collected in the center of line determination by FTIR spectroscopy method at 197 cm height as the instrument measure-ment.
Methane concentration was analyzed by a Shimadzu 14A gas chromatograph (Shimadzu Co., Japan) with a glass column (0.26 mm· 2.0 m) and a flame ionization detector, which was packed with Porapak Q (80/100 mesh). The column temperature was set at 100C, the injection and the detector temperatures were set at 130 C. Methane concentration was calculated with a stan-dard curve from 0.1 to 1000 mg kg1(v/v) (Chang and Yang, 1997; Yang and Chang, 1997). For carbon diox-ide analysis, a thermal conductivity detector was used. The column temperature was set at 150C. The injection and the detector temperatures were set at 200C.
Car-bon dioxide concentration was calculated with a stan-dard curve from 0.1 to 1000 mg kg1 (v/v) (Chang and Yang, 2003).
2.4. Analytical methods
Air and soil temperatures were determined with a thermometer. Wind speed was measured with a weath-ercock (Weather link 4.0, USA). Light intensity was measured with a Toshiba SPI-5 photometer. Soil pH was determined directly in the tested soils or on 1:1 (w/v) soil to water suspension using a pH meter. Moisture content was measured by drying a sample at 105C for 24 h to constant weight. Total organic carbon of the soil or leachate was estimated using a TOC-5000A total organic carbon analyzer (Code HI 8424C, Shimadzu, Japan) after proper dilution of the soil or leachate (Yang and Chang, 1997). The redox potential (Eh) of the soil or leachate was measured with a Hanna No. 081-854 po-tentiometer (Yang and Chang, 1997). Total nitrogen was determined with a modified Kjedahl method (Yang et al., 1991). Experiments were carried out in quadruplicate, and flux data were subjected to analysis of the coefficient of variance and DuncanÕs multiple range tests (p¼ 0:05) using the Statistical Analysis System (SAS Institute, 1988).
3. Results
Landfill is the major method of garbage disposal in Taiwan. Landfill disposal was 88.99% in 1989 and de-creased to 71.42% in 1999 for the increasing of inciner-ation treatment and material recovery policy. The predominant part of the landfill waste soon after dis-posal becomes anaerobic, and a consortium of bacteria start degrading the solid organic material. Eventually these materials are converted into carbon dioxide, methane and water. Thus, atmospheric concentrations of greenhouse gas, emission rates of carbon dioxide and methane, and other environmental conditions were measured.
3.1. Greenhouse gas concentration on the campus of National Taiwan University
As a standard, atmospheric concentrations of greenhouse gas on the campus of the NTU were mea-sured by FTIR spectroscopy method and GC method on December 12, 1997, May 16, 1998 and August 20, 1999 between 11 a.m. and 1 p.m. (Table 3). Gas-type open-path FTIR spectroscopy method had a good correlation with open field GC chamber method. The conversion factors of the concentration with FTIR spectroscopy method to the concentration with open field GC method ranged from 0.986 to 1.008, from 0.994 to 1.011, and
Table 3
Atmospheric concentration of greenhouse gas at different landfill periods of the Shan-Chu-Ku measured by FTIR spectroscopy and GC
Sampling date Disposal
period (year)
Detection method
Carbon dioxide (ppmv) Methane (ppmv) Nitrous oxide (ppbv)
Average Minimum Maximum Average Minimum Maximum Average Minimum Maximum
May 14–15, 1998 1–2 FTIR 347 293 367 7.61 3.56 14.61 480 330 760 GC chamber 560 383 907 28.77 8.12 55.64 – – – GC open field 341 322 419 7.50 4.03 16.99 – – – February 25–27, 1998 2–3 FTIR 310 295 345 12.07 4.93 17.27 688 650 841 GC chamber 679 395 1010 35.79 13.90 139.08 – – – GC open field 302 293 422 11.36 5.02 22.11 – – – May 15–16, 1998 5 FTIR 353 325 376 3.40 1.88 7.06 371 320 675 GC chamber 375 353 486 12.82 3.23 53.41 – – – GC open field 348 325 415 3.12 2.20 10.95 – – –
December 12, 1997 Campus FTIR 363 350 372 1.77 1.73 1.79 321 316 333
GC chamber 362 348 374 1.76 1.72 1.78 320 316 330
GC open field 361 355 372 1.75 1.72 1.77 324 317 334
May 16, 1998 Campus FTIR 362 351 370 1.76 1.74 1.79 321 315 333
GC chamber 363 349 373 1.77 1.73 1.80 320 317 330
GC open field 359 351 372 1.76 1.75 1.80 325 317 334
August 20, 1999 Campus FTIR 360 349 370 1.77 1.74 1.78 319 315 334
GC chamber 356 331 381 1.74 1.72 1.76 328 305 351
GC open field 358 342 372 1.76 1.74 1.79 323 315 333
Each value was derived from more than 15 independent measurements on the same day. FTIR spectroscopy and GC open field measurement were at 197 cm height above ground.
1278 U. Hegde et al. / Chemosph ere 52 (2003) 1275–1285
from 0.991 to 1.003 in carbon dioxide, methane and nitrous oxide, respectively. Air temperatures were be-tween 30 and 38 C. Light intensities ranged from 2.50· 105 to 1.45· 106 lx. Diurnal variation of air temperature was similar to that of light intensity. The effect of air temperature and light intensity on atmo-spheric concentrations of greenhouse gas was not sig-nificant. Therefore, greenhouse gas measurement with FTIR spectroscopy method can be applied in open field. 3.2. 1–2 year-old landfill
The diurnal variations of atmospheric concentrations of greenhouse gas, carbon dioxide and methane emission rates and environmental conditions in 1–2 year-old landfill during May 14 and 15, 1998 are presented in
Fig. 1, Tables 3 and 4. The light intensity was high at noon. Atmospheric concentration of carbon dioxide decreased between 3 p.m. and 6 p.m. on May 14, as it rained. Atmospheric concentrations of methane and nitrous oxide were high at night and low in the daytime due to the burning of exit gas in the daytime. The variation of atmospheric concentrations of carbon di-oxide and methane measured by GC method (both chamber and open field) was larger than that by FTIR spectroscopy method because of the point measurement in gas chromatographic method while the line determi-nation and the average of 100 times of open field in FTIR spectroscopic method. GC open field method had lower atmospheric concentrations of greenhouse gas than that of chamber method for the mixing effect of the wind and the diffusion of greenhouse gas from landfill to
CO 2 e m ission rate (m g m -2 h -1 ) 0 100 200 300 CH 4 emissi o n ra te (m g m -2 h -1) 0 10 20 30 40 At m o sp h e ri c C O2 (ppmv ) 0 250 500 750 1000 At m o sp h e ri c C H4 (p p m v ) 0 20 40 60 12 15 18 24 3 6 9 12 May 14 May 15 (a) (b) (c) (d) 21 Sampling time 0.0 0.2 0.4 0.6 0.8 (e) A tmosp h e ric N2 O (p p m v )
Fig. 1. Diurnal variation of atmospheric greenhouse gas and carbon dioxide and methane emission rates in 1–2 year-old landfill of the
Shan-Chu-Ku during May 14–15, 1998. (a) Atmospheric carbon dioxide with FTIR spectroscopy method (
) and with GC method(d). (b) Carbon dioxide emission rate (j). (c) Atmospheric methane with FTIR spectroscopy method (
M
) and with GC method (N
).the atmosphere. The fluctuation of atmospheric con-centrations of greenhouse gas in landfill was larger than that on the campus of the NTU due to the high emission of greenhouse gas from landfill and the low emission from campus. In addition, carbon dioxide and methane emission rates were high at night and low in the daytime for the burning of exit gas of landfill in the daytime.
3.3. 2–3 year-old landfill
The diurnal variations of atmospheric concentrations of greenhouse gas, carbon dioxide and methane emission rates and environmental conditions during February 25–27, 1998 are illustrated in Fig. 2, Tables 3 and 4. Atmospheric concentration of carbon dioxide in the daytime was slightly lower than that at night for the photosynthesis of grass and plants in the landfill. At-mospheric concentrations of methane and nitrous oxide were high at noon and low at night due to the low temperature (10.3C) in the winter season. The highest air temperature was at noon on February 26 and the lowest was at night on February 25. The highest light intensity observed was at noon on February 26. Both carbon dioxide and methane emission rates had high values at noon for high temperature (23.9C). Negative fluxes were sometimes observed in the tested periods when the methane oxidation rate is higher than that of the methane emission rate. Both atmospheric concen-trations of greenhouse gas and carbon dioxide and methane emission rates were the highest in 2–3 year-old landfill area among the tested locations for the active degradation of organic wastes and the high total organic carbon content.
Atmospheric concentrations of greenhouse gas and emission rates of carbon dioxide and methane in this landfill were also measured in January, March, July, September and October 1999, February and March 2000 (Table 5). Atmospheric concentration of carbon dioxide was the highest in March 1999 and the lowest in Feb-ruary 1999. Atmospheric concentration of methane was the highest in January 1999 and the lowest in October 1999. Atmospheric concentration of nitrous oxide was the highest in March 2000 and the lowest in March 1999. Atmospheric concentrations of carbon dioxide, methane and nitrous oxide of tested samples in this landfill had 62.5%, 100% and 100%, respectively higher than those on the campus of NTU. Carbon dioxide emission rate had the highest value in October 1999 and the lowest in February 1999, while the methane emission rate was the highest in February 1999 and the lowest in February 2000. The fluctuation patterns were slight differences between atmospheric concentrations of greenhouse gas and emission rates of carbon dioxide and methane for the former with FTIR spectroscopy measurement and the latter with GC determination.
Table 4 Environmental conditions, carbon dioxide and methane emission rates at different landfill periods of the Shan-Chu-Ku measured by GC Sampling date Disposal period (year)
Temperature ( C) Light intensity (lx) Wind speed (m s 1) Carbon dioxide emission rate (mg m 2h 1) Methane emission rate (mg m 2h 1) Minimum Maximum Minimum Maximum Minimum Maximum Average Minimum Maximum Average Minimum Maximum May 14– 15, 1998 1–2 22.5 32.5 0 9.0 · 10 5 0.0 5.4 93.70 <0 343.15 13.17 <0 36.01 Feb. 25– 27, 1998 2–3 10.3 23.9 0 2.6 · 10 4 0.6 2.4 99.21 <0 337.13 157.56 <0 759.82 May 15– 16, 1998 5 21.0 32.5 0 1.2 · 10 5 0.2 2.1 48.46 <0 200.57 0.99 <0 5.55 Each value was de rived from more than 16 indep endent me asuremen ts using fou r acryl ic ch ambers installed at a dista nce aro und 6 m fr om each oth er on the sam e day.
3.4. 5 year-old landfill
Diurnal variations of atmospheric concentrations of greenhouse gas, carbon dioxide and methane emission rates and environmental conditions in the 5 year-old landfill are presented in Fig. 3, Tables 3 and 4. The highest air temperature and light intensity was at noon on May 16 and the lowest was at night on May 15. Atmospheric concentrations of methane and nitrous oxide were also high at night and low in the daytime because of the burning methane in the daytime and high temperature in May. Atmospheric concentration and emission rate of methane were low in this area for longer burial period. Methane emission rate was only 0.63%
and 7.51% of the 2–3 and 1–2 year-old landfill areas, respectively.
4. Discussion
Compared to other terrestrial ecosystems, landfills are characterized by high rates of methane production and large methane gradients from the deeper production zone to the soil–atmosphere interface. This is in contrast to methane emission from paddy fields, where the emission rate was high from 11 a.m. to 2 p.m. and was low in the early morning (Yang and Chang, 1998, 1999, 2001b). However, no such correlation has been observed in
CO 2 emissio n rate (m g m -2 h -1) 0 100 200 300 CH 4 emission rate (m g m -2 h -1 ) 0 200 400 600 800 Atmospheric C O2 (ppmv) Feb. 26 Atmospheric C H4 (ppmv) 0 50 100 150 200 13 19 1 7 13 19 1 7 13 Feb. 25 Feb. 27 (a) (b) (c) (d) Sampling time 0 300 600 900 1200 0.4 0.5 0.6 0.7 0.8 A tmospheric N2 O (ppmv) (e)
Fig. 2. Diurnal variation of atmospheric greenhouse gas and carbon dioxide and methane emission rates in 2–3 year-old landfill of the
Shan-Chu-Ku during February 25–27, 1998. (a) Atmospheric carbon dioxide with FTIR spectroscopy method (
) and with GCmethod (d). (b) Carbon dioxide emission rate (j). (c) Atmospheric methane with FTIR spectroscopy method (
M
) and with GCTable 5
Average atmospheric concentrations of greenhouse gas and carbon dioxide and methane emission rates in 2–3 year-old landfill of the Shan-Chu-Ku during January 1999–March 2000
Sampling date Carbon dioxide
(ppmv)
Methane (ppmv)
Nitrous oxide (ppbv)
Carbon dioxide emission
rate (mg m2h1) Methane emission rate (mg m2h1) January 1999 417.97 ± 11.00 20.16 ± 1.37 283.26 ± 8.00 248.04 ± 44.50 90.84 ± 29.00 February 1999 310.03 ± 9.72 12.07 ± 3.12 268.78 ± 6.39 99.21 ± 42.43 157.60 ± 25.10 March 1999 530.07 ± 4.54 10.13 ± 1.10 258.18 ± 7.41 309.30 ± 48.67 78.18 ± 27.33 July 1999 400.86 ± 5.82 9.73 ± 2.87 359.76 ± 19.17 296.16 ± 42.67 32.93 ± 12.00 September 1999 376.73 ± 7.90 3.52 ± 0.30 464.14 ± 10.06 314.60 ± 42.83 18.61 ± 5.06 October 1999 397.08 ± 5.00 2.64 ± 0.32 1122.60 ± 21.42 624.60 ± 63.70 13.47 ± 6.76 February 2000 354.87 ± 11.17 8.91 ± 4.26 1460.19 ± 26.00 534.10 ± 26.90 8.91 ± 2.69 March 2000 323.80 ± 11.00 4.46 ± 0.71 1516.17 ± 17.40 125.32 ± 17.65 13.91 ± 1.29 August 1999/NTU 360.31 ± 5.00 1.76 ± 0.01 319.24 ± 4.50 – –
Each value was derived from more than 15 independent measurements on the same day with FTIR spectroscopy in atmospheric concentrations of greenhouse gas and with GC in emission rates of carbon dioxide and methane using four acrylic chambers installed at a distance around 6 m from each other.
CO 2 em issi on rate (m g m -2 h -1) 0 50 100 150 200 CH 4 emi ssion rat e (m g m -2 h -1) 0.0 1.5 3.0 4.5 A tmo s phe ri c C O2 (ppm v ) 0 150 300 450 600 A tmosp h e ri c C H4 (p p m v ) 0 15 30 45 12 15 18 24 3 6 9 12 May 15 May 16 (a) (b) (c) (d) 21 Sampling time 0.0 0.3 0.6 0.9 A tmo s phe ri c N2 O (p p m v ) (e)
Fig. 3. Diurnal variation of atmospheric greenhouse gas and carbon dioxide and methane emission rates in 5 year-old landfill of the
Shan-Chu-Ku during May 15–16, 1998. (a) Atmospheric carbon dioxide with FTIR spectroscopy method (
) and with GC method(d). (b) Carbon dioxide emission rate (j). (c) Atmospheric methane with FTIR spectroscopy method (
M
) and with GC method (N
).(d) Methane emission rate (). (e) Atmospheric nitrous oxide (O) with FTIR spectroscopy method.
landfills during the tested periods. Methane emission rate was high at night and low in the daytime due to the burning of methane in the daytime. Correlation between greenhouse gas emission and air temperature was not obvious in landfill. The fluctuation of soil temperature in the landfill at deeper zone was low, and the effect of air temperature on greenhouse gas emission was not signif-icant when temperature was above 15 C. However, it showed a positive correlation between methane emission and soil temperature (15–37C) in paddy soil (Yang and Chang, 1998). Correlation among air temperature (T ), light intensity (L) and atmospheric greenhouse gas con-centrations in landfill was CH4¼ 18:9287 0:5416T þ 0:6229L with a p-value 0.0001 (r2¼ 0:57) and CO
2¼ 339:6611þ 2:2423T þ 28:2638L with a p-value 2.505 (r2¼ 0:71). Light might affect the oxygen distribution and penetration in the soil–air interface, and enhance the methane oxidation (Khalil, 1995). King (1990) also in-dicated that illumination could possibly induce the growth and photosynthesis of algae and the oxidation of methane, but reduce the emission of methane. In con-trast, the correlation between methane emission and light intensity in paddy fields was not obvious at low light intensity (between 1000 and 2000 lx) (Yang and Chang, 1998). Therefore, the methane emission pattern in land-fills was different from paddy fields. Methane emission from soils depends mainly on soil temperature, moisture content, water table level, organic matter content, redox potential and wind induced ebullition (Bartlett et al., 1992; Yang et al., 1994, 2003; Yang and Chang, 1997, 1998, 2001a,b). However, it is difficult to correlate the exact relationships among greenhouse gas emission rates, temperature, moisture content and composition in landfill due to the heterogeneous nature of the refuse used (Jang and Yang, 2001).
The GC chamber technique gives a point measure-ment, while gas-type open-path FTIR spectroscopy method gives a line determination. The chamber method can be used to measure greenhouse gas emission rates from small areas of the tested sample, typically less than 1 m2 in each measurement (Yang and Chang, 1999, 2001a). Advantages of this technique include simplicity, an appropriate scale for concurrent measurement of controlling variables (e.g. temperature, pH, soil Eh, moisture content, organic matter content), ability to determine the heterogeneity of surface emissions, and ready comparison with other results, since this technique has been widely employed in non-landfill settings for a variety of greenhouse gases (Harriss et al., 1982; Rolston, 1986; Chang and Yang, 1997; Yang and Chang, 1999). However, the variation of GC measurement was high in a long distance heterogeneous field. In contrast, the variation was low using FTIR spectroscopy method due to its ability to compensate for long-distance and more scan number measurements (Chang et al., 2000). Grass-land soil on the campus of NTU is more homogeneous
than that in landfill and the atmospheric concentrations of greenhouse gas in the grassland of the campus is more uniform than that in landfill. The ratios of atmospheric concentrations of greenhouse gas with GC chamber method to FTIR spectroscopy method in grassland of campus was between 0.989 and 1.003, 0.983 and 1.006, and 0.997 and 1.028 for carbon dioxide, methane and nitrous oxide, respectively. The differences between the ratios of two measuring methods were not significant. However, the ratios of atmospheric concentrations of greenhouse gas in the Shan-Chu-Ku landfill were very significant difference. The ratios of atmospheric concen-tration of methane were between 2.28 and 4.79, 2.82 and 8.05, and 1.72 and 7.57 in 1–2, 2–3, and 5 year-old landfills, respectively. The differences of the ratios be-tween campus and landfill might be due to (1) The active decomposition of organic wastes to carbon dioxide, methane and water was found in the landfill. The meth-ane concentration was high in 2–3 year-old landfill. (2) Surface emission measurement with the closed chamber method had higher values than that with open-path field measurement because of the diffusion of methane from the landfill to the atmosphere and the mixing effect of the wind in the open field method. (3) The concentration measurement with FTIR spectroscopy method is the average of 100 scans along the line, while the concen-tration determination with GC chamber method is only the value of a single sampling point. In addition, the ratios of the highest concentration to the lowest one of atmospheric greenhouse gas with GC measurement were larger than that with FTIR spectroscopy method in all tested sample. FTIR spectroscopy method had a narrow range of the ratios for 100 scans and line detection. Methane emission rate increased at night in most tested locations compared with the same day because the methane was burning in the daytime but not at night.
The properties and the thickness of cover soil were also very important in methane emission from landfill. Although the correlation between methane emission (X ) and populations of methane oxidizing microbes (Y ) was low. The regression equation was Y ¼ 2:503X 5 106 and r2¼ 0:078 (Liao et al., 1998). However, the popu-lations of methane oxidizing microbes increased with the period of burial. Methane oxidation rate (X ) had a lin-ear correlation with the populations of methane oxidiz-ing microbes (Y ). The regression equation was Y¼ 8:40 104X3:0106 and r2¼ 0:735 (Liao et al., 1998). Total soil organic carbon of cover soil was 5.97–7.49%, 7.78– 9.02% and 5.84–6.67% in 1–2, 2–3 and 5 year-old land-fills, respectively. Methane emission increased with the increasing of total soil organic carbon and total soil nitrogen content, and decreased with the increasing of thickness of cover soil. Kuo et al. (2000) reported that methane emission rate was 0.01, 0.054 and 18.21 gm2h1with 1.5–2.9, 0.2 and 0.05 m thickness of cover soil in Taoyuan landfill site.
The air temperature of 2–3 year-old landfill site during February 25–27, 1998 was lower than those of 1– 2 year-old and 5 year-old landfill sites during May 14– 16, 1998. Carbon dioxide and methane emissions from 2–3 year-old landfill site were not complete comparable with the other two sites for time difference of two-and-half months earlier than the other sites. Carbon dioxide and methane emissions were still significantly high at the 2–3 year-old landfill site despite the cooler temperature due to the high total organic carbon and total nitrogen contents and high methanogenesis activities. About 2–3 year-old landfill site showed the highest methane emis-sion rate might be due to methanogenesis being fully established. Methane emission rate in the 5 year-old landfill site was very low to negative value. These results suggest that cover soils in this area have high capacity for methane oxidation as Whalen et al. (1990) and Young and Liu (1998) had described. Methane oxida-tion activity by methanotrophs is higher than that of methane production by methanogens. Atmospheric concentration of carbon dioxide in 5 year-old landfill site was low because of the presence of many trees and small shrubs on this area. Active photosynthesis during daytime consumes carbon dioxide production in the landfill.
Methane emission from the landfill of MSW was affected by methane correction factor, fraction of de-gradable organic carbon, biogas formation factor, car-bon conversion rate (ratio of carcar-bon in biogas to methane) and methane oxidation coefficient. Liao et al. (1998) had estimated the methane emission from the Taichung landfill site by 5 different methods. Methane emission from Taichung landfill site with 1.46· 105 ton of MSW in 1995 was 1.9, 1.7, 1.9, 3.3 and 1.4· 104 ton by heat value, carbon content, chemical composition, physical composition and IPCC methods, respectively. Annual methane emission from Taichung landfill with local measurement was 1.3· 104ton that was slight less than IPCC method due to the high methane oxidation activity in cover soil and low dry matter content of MSW. In this study, annual methane emission from the Shan-Chu-Ku landfill was 0.115, 0.572 and 0.009 kg m2 in 1–2, 2–3 and 5 year-old landfill sites, respectively. Methane emission from Shan-Chu-Ku landfill was less than that in Taichung landfill (it was 2.65 kg m2at the top of landfill, and 12.10 kg m2 in the side edge of landfill) because of the high thickness of cover soil in Shan-Chu-Ku landfill.
Gas-type open-path FTIR spectroscopy method has the advantage in long-distance, precision and rapid measurement. In addition, it can detect different com-pounds in the meantime (Sch€aafer et al., 1994) and we have successfully utilized this technique along with GC method to monitor atmospheric concentrations of green-house gas in the landfill.
5. Conclusions
Both gas-type open-path FTIR spectroscopy method and GC chamber method have been successfully used to measure the atmospheric concentrations of greenhouse gas and carbon dioxide and methane emission rates from landfill. Average methane emission rate was 13.17, 65.27 and 0.99 mg m2h1 in 1–2, 2–3 and 5 year-old landfills, respectively. Similarly, average carbon dioxide emission rate was 93.70, 314.60 and 48.46 mg m2h1, respectively. Annual methane and carbon dioxide emis-sion from each hectare of landfill is estimated around 5.717 and 27.61 ton, respectively. About 30 ha is cur-rently in use in the Shan-Chu-Ku landfill, therefore around 171 and 828 ton of methane and carbon dioxide are released to the atmosphere in 1999, respectively. Taiwan produced 8 565 820 ton of MSW in 1999, and 71.42% of the MSW was disposed by landfill. Total carbon dioxide and methane emissions from the landfill were around 10 050 and 2075 ton in Taiwan, re-spectively. Four landfill sites (Shan-Chu-Ku and Fu-Der-Ken in the north, Chi-Shin-Pu in the south and Taichung in the central Taiwan) have set up a gas re-covery system and the maximal electricity generation capacity is estimated 32.7 MW. It reduces 5% methane emission from landfill. In the future 11 potential landfill sites are suitable for methane recovery in Taiwan and the maximal methane reduction expected is 13.4%.
Acknowledgements
The authors thank Dr. E.H. Chang, Mr. Y.C. Leo and I.C. Chen for their helpful assistances in FTIR spectroscopy measurement, Professor C.T. Liao for statistical analysis, and the National Science Council of the Republic of China for their financial support (NSC87-2621-P002-021, NSC88-2811-Z-002-0001 and NSC89-EPA-Z002-003).
References
Bartlett, K.B., Crill, P.M., Sass, R.L., Harriss, R.C., Dise, N.B., 1992. Methane emissions from tundra environments in the Yukon-Kuskowim Delta, Alaska. J. Geophys. Res. 97, 16645–16660.
Bogner, J., Meadows, M., Czepiel, P., 1997. Fluxes of methane between landfills and atmosphere; natural and engineered controls. Soil Use Mgmt. 13, 268–277.
Chang, E.H., Yang, S.S., 2003. Some characteristics of micro-algae isolated in Taiwan for biofixation of carbon dioxide. Bot. Bull. Acad. Sinca 44, 43–52.
Chang, H.L., Yang, S.S., 1997. Measurement of methane emission from soil. J. Chin. Agric. Chem. Soc. 35, 475–484.
Chang, T.C., Leo, Y.C., Yang, S.S., 2000. Determination of greenhouse gases by open-path gas-type FTIR spectro-scopy. Food Sci. Agric. Chem. 2, 7–14.
Gurijala, K.R., Sulflita, J.M., 1993. Environmental factors influencing methanogenesis from refuse in landfill samples. Environ. Sci. Technol. 27, 1176–1181.
Harriss, R.C., Sebachar, D.I., Day, F.D., 1982. Methane flux in the great dismal swamp. Nature 297, 673–674.
Jang, H.D., Yang, S.S., 2001. Greenhouse gases production of municipal solid wastes in column bioreactors. J. Biomass Energy Soc. China 20, 101–112.
King, G.M., 1990. Dynamics and controls of methane oxida-tion in a Danish wetland sediment. FEMS Microbiol. Ecol. 74, 309–324.
Khalil, M.A.K., 1995. Greenhouse gases in the EarthÕs atmo-sphere. Encycl. Environ. Biol. 2, 251–265.
Kreileman, G.J.J., Bouwman, A.F., 1994. Computing land use emissions of greenhouse gases. Water Air Soil Poll. 76, 231– 258.
Kuo, K.T., Chang, C.M., Wang, K.S., 2000. Methane flux of landfill and atmospheric methane concentration. In: Yang, S.S. (Ed.), Flux and Mitigation of Greenhouse Gases II. Global Change Research Center, Department of Agricul-tural Chemistry and Agriculture Exhibition Hall, National Taiwan University, Taipei, Taiwan, pp. 192–204.
Liao, W.P., Horng, C.C., Lee, C.M., 1998. Estimation and monitoring of greenhouse gas emission from Taichung landfill. Final Report of Environmental Protection Admin-istration (EPA87-FA44-03-47). Taipei, Taiwan, p. 320. Rolston, D.E., 1986. Gas flux. In: Klute, A. (Ed.), Methods of
Soil Analysis, second ed. In: American Monograph, vol. 9. American Society Agronomy and Soil Science Society, Wisconsin, pp. 1103–1119.
SAS Institute, 1988. SAS/STAT userÕs Guide, Release 6.03. SAS Institute, Cary, NC.
Sch€aafer, K., Haus, R., Heland, J., 1994. Inspection of non-CO2
greenhouse gases from emission sources and in ambient air by Fourier-transform spectrometry: measurements with
FTIS-MAPS. In: Proceedings of the International
Sympo-sium on non-CO2 Greenhouse Gases. Kluwer Academic
Publishers, Massstricht, pp. 191–196.
Whalen, S.C., Reeburgh, W.S., Sandback, K.A., 1990. Rapid methane oxidation in a landfill cover soil. Appl. Environ. Microbiol. 56, 3405–3411.
Yang, S.S., Chang, E.H., 1997. Effect of fertilizer application on methane production in paddy soils of Taiwan. Biol. Fertil. Soils 25, 245–251.
Yang, S.S., Chang, H.L., 1998. Effect of environmental conditions on methane production and emission from paddy soil. Agric. Ecosyst. Environ. 69, 69–80.
Yang, S.S., Chang, H.L., 1999. Diurnal variation of methane emission from paddy fields at different growth stages of rice cultivation in Taiwan. Agric. Ecosyst. Environ. 76, 75–84. Yang, S.S., Chang, H.L., 2001a. Effect of green manure
amendment and flooding on methane emission from paddy fields. Chem.: Global Change Sci. 3, 41–49.
Yang, S.S., Chang, H.L., 2001b. Methane emission from paddy fields in Taiwan. Biol. Fertil. Soils 33, 157–165.
Yang, S.S., Chang, H.L., Wei, C.B., Lin, H.C., 1991. Reduce waste production in the Kjeldahl methods. J. Biomass Energy Soc. China 10, 147–155.
Yang, S.S., Lin, C.C., Chang, E.H., Chung, R.S., Huang, S.N., 1994. Effect of fertilizer, soil type, growth season on methane production and emission in the paddy soils of Taiwan. J. Biomass Energy Soc. China 13, 68–87. Yang, S.S., Liu, C.M., Lai, C.M., Liu, Y.L., 2003. Estimation
of methane and nitrous oxide emission from paddy fields and uplands during 1990–2000 in Taiwan. Chem.: Global Change Sci. 5.
Young, C.C., Liu, C.W., 1998. Uptake and emission by the addition of greenhouse effect gases in orchard and forest soils in central and southern Taiwan. In: Lu, S.C., Liu, C.M., Yang, S.S. (Eds.), Change of Atmospheric Environ-ments in Taiwan Area III. Global Change Research Center and Department of Agricultural Chemistry, National Tai-wan University, Taipei, TaiTai-wan, pp. 54–69.
