Natural Gas and Indoor Air Pollution: A Comparison With Coal Gas and Liquefied Petroleum Gas

BIOMEDICAL AND ENVIRONMENTAL SCIENCES 16, 227-236(2003)Natural Gas and Indoor Air Pollution A comparison With coalGas and liquefied Petroleum GasR AYUE ZHANG", BAO-SHENG CHEN, GUANG-QUAN LIU, JU-NINGZHEN-HUA ZHAO, AND LIAN-QING LIN* Institute of Environmental Health and Related Product Safery, Chinese Center for Disease Centroland Prevention, Beiing 100050, China; Beiing Research institute of Environmental Protection,Beijing 100037, China;Beijing Institute of Radiation Protection, Beijing 100013, Chinato compare the combustionpetroleum gas and natural gas in relation to indoor air pollution. Methods Regular pollutantsluding B(a)P were monitored and 1-hydroxy pyrene were tested in urine of the enrolled subjecadon concentrations and their changes in four seasons were also monitored in the city natural gasfrom its source plant and transfer stations to final users. To analyze organic components of coal gas,liquefied petroleum gas and natural gas, a high-flow sampling device specially designed was used tocollect their combustion products, and semi-volatile organic compounds contained in the particleswere detected by gas chromatograph-mass spectrograph(GC/MS). Results Findings in the studyhowed that the regular indoor air pollutants particles and CO were all above the standard in winterhen heating facilities were operated in the city, but they were lowest in kitchens using natural gas;furthermore, although NO and Co were slightly higher in natural gas, B(a)P concentration wawer in this group and l-hydroxy pyrene was lowest in urine of the subjects exposed to natural gasOrganic compounds were more complicated in coal gas and liquefied petroleurhan in naturalgas. The concentration of radon in natural gas accounted for less than 1%e of its effective doseontributing to indoor air pollution in Beijing households. Conclusion Compared to traditionalKey words: Natural gas; Indoor air polhution; Organic compounds; Radon; 1-hydroxy pyrene; EnvironmentalINTRODUCTIONChina is abundant in resources of natural gas (NG), which with its high-energy valuehas become one of the increasingly growing sources of domestic energy in this country. Urto the end of 1993, NG users increased to 240 000 households and its supply volumeeached15×10°m/a; it was expected to rise to8.5×10°m3 a by the year of200The health impact of NG has been widely reported in foreign literature, largely inregard with consequences of the pollutants produced in the process of its exploitation,processing and refining, for instance, acute intoxication, ill health of the population andrising morbidities of respiratory diseases and cancersCorrespondence should be addressed to Yue ZHANG Institute of Environmental Health and Related ProductSafety, Chinese Center for Disease Control and Prevention, Beijing 100050, ChinaBiographical note of the first author Yue ZHANG female, born in 1953, senior technical scientist withlogical and genetic assessment of environmental pollutants0895-3988/2003Copyright 2003 by中国煤化工CNMHGIANG ETALThe health consequences of NG as a domestic energy source might well be focused onits impacts on living environment, or indoor air pollution. Over the recent years, bothnational and international scientific communities have taken great interest in indoor airpollution as a major public health problem. However, no impacts of NG on indoor airpollution have so far been systematically reported in either national or internationalMATERIALS AND METHODSMonitoring of Indoor Air pollutionTwelve families living in apartment houses were selected and each four of them usedNG, coal gas(CG)and liquefied petroleum gas( LPG)respectively. They were monitored for3 successive days in 3 separate groups in summer(between June 24 and July 4, 1992)andwinter(between December 10 and 19, 1992). To avoid interference with cooking smoke inkitchens, samples were collected twice in the morning and in the afternoon by closing alldoors and widows and turning off kitchen ventilators. During the sampling kitchen stoveswere turned on at an angle of 45 and kettles with water were then placed on them for halfan hour to keep the water at an un-boiling stateInhalable particles(IP), CO, SO2, NO2, COz and B(a)P were used as indicatorindoor air pollution, and their respective sampling methods are shown in Table 1TABLE IMethods of Sampling and AnalysisIndicators Sampling DevicesKC 8301 Sampling Deviceonventional mePlastic BagsAnalyzed by EcolyzerSpectophotometry With Para-roEthylene Naphthalene HydrochloridepectrophotometryPortable Infrared Analyzer GXH-3010C(Only in winterB(a)Pand Glass Filter( GF)Collectors PUFP PAHs Were Analyzed After Extraction by InverseAir and Particles Were Collected byPhased-high Pressure Liquid ChromatographyHPLC)and FluoroscopyOrganic Components of Gas Combustion ProductsAir samples in the households exposed to the three respective types of gases werecollected in large quantity by high flow collectors. The samples containing particlesproduced in the process of combustion of the 3 gases, controls and blank filter membraneswere separately placed in 250 mL soxhlet extractors. To improve extraction rates all thesamples were immersed in dichloromethane for 20 h at room temperature before they wereextracted at 70C-80C for 12 h. the extracts were dried to constant weight by anhydrouscalcium chloride after they were processed by K-D concentrator. The contents of organicmpounds on the particles as products of gas combustion were calculated according to theweight of the samples, blank filters and extracts(gravimetric method). The extracts which中国煤化工CNMHGNATURAL GAS AND INDOOR AIR POLLUTIONhad been solved into dichloride methane to reach an appropriate concentration wereanalyzed by GC/MS(Finnigan-MAT 4510)Determination of Radon Concentrations in NG in BeijingNatural gas in Beijing is originated largely from Renqiu Oil Field located in north ChinaTo have a clear picture of radon(Rn) concentrations in different oil wells and their possiblechanges in transportation processes, NG samples from 5 source wells, as well as from itsexits, entries to Beijing and final users were detected respectively. To identify changes of Rnconcentrations over time, NG samples were collected from final users at 7: 00-7: 30 a.m11: 30-12: 00 a.m. and 6: 00-6: 30 p.m. for a total of 8 days(two days per week) beginningfrom March 1992, thereafter from March 1992 to February 1993, gas samples werecollected from final users at 11: 30-12: 00 on one day randomized in the middle of eachmonth, and Rn concentrations were determined by FD-125 indoor Rn analyzerDetermination of I-hydro Pyrene in Subjects'UrineFrom each of the respective 4 households using different types of gases as their cookingfuel, the family members doing most cooking were defined as study subjects arconcentrations of 1-hydro pyrene were determined in their urine samplesPretreatment of urine samples: Urine samples of 10 mL were added into 5 mL 0.5 mol/Lacetate buffer(pH 5)containing 1 000 units of glucuronidase(type H-l Sigma Corp. USA)The mixtures were then kept at 37C for I h and isolated after hydrolyzation by SEPPAK C18 column(Wates Corp). They were concentrated at 60C up to 0.5 mL after elution bymethanol and then tested for l-hydro pyrene by hPlcExpression of 1-hydro pyrene in urine: as the samples tested were not the whole urinewithin 24 h, the following two correcting methods were adopted to eliminate possible errorsderived from urine samples themselves: creatine correction- corrected by creatine values inthe same sample and expressed by umoL/L 1-OH-pyrene/mol creatine; and gravitycorrection-corrected by the Chinese standard gravity of urine(. 020)RESULTSMonitoring of indoor Air PollutantsAs shown in Table 2, IP and B(a)P concentrations in the NG C and LPG were higherin winter than in summer, but not significantly different in summer. IP concentrations wereall around 150 Hg/m, being little higher in the CG group. In winter IP concentrationsexceeded hygienic standards in all the three groups, being highest in the LPG group (31ug/m). As for IP concentrations, there was generally no significant difference in the 3groups, either in winter or in summer. In contrast, great seasonal differences were observedfor B(a)P concentrations on particles, by several times and even dozens of times higher inwinter than in summer, while the lowest was in NG and the highest in LPGAs shown in Table 3, Co concentrations were higher in winter than in summer, andhigher in kitchens than in living rooms. They were much higher than the standard either inher, notably in the former. If the 3 types of gas were compared, COconcentrations in NG and CG were similar while those in LPG were markedly higher than inhe others. SO2 concentrations were quite low, at the normal range either in the 2 seasons orin all the 3 groups, and were only slightly higher than the standard in winter in the Cg grouppossibly due to sulfur contents in coal gas. NO2 concentrations were determined only in中国煤化工CNMHGZHANG ETALwinter and were low in all the 3 groups, indicating that indoor NO2 came merely fromkitchen stoves and its differences in winter and summer were insignificant and irreAlthough the highest NO2 concentrations were found in kitchens using LPG in summer, theywere relatively higher even in kitchens using NG in winter. They were somehow lower inthe CG group, but were higher than the standard in the 3 groups. Since nitrogen is alwayspresent in natural gas, it should be given serious attention in the studies concerning theelationship between indoor air pollution and natural gasTABLE 2IP and B(a)P Indoor Concentrations in NG CG and LPGIP(mg/m)B(a)P(ng/m)Types of Gas HouseholdsSummer0.1650.27200980.2982.160.14±0.0330.242±0058128±0.5213.79±47300890.12l1.15l12.0230.2610.1590.17±00610.196±0.052202土0431620±2.81PG236.1930.1670328x±s012士0.0330.317±0.040339士092755±1140CO, So and NO2 Concentrations in Indoor Air in the 3 Groupsof GasSitesK, Average10.83.8-125009-0.130.020.,200.23-1430.15-3.04LR.Average003L.R. Range14-10.019-12.5K, Range44-10.66.3-32.5007-0.220.072690.10-0.390.070.38L.R. AveraL.R., Range19503.8-150.42LPG K, Range0-56.38.8-4880.10-0.340.100.00.152190.130.85LPG LR,Average10.3Note. K: Kitchens; L.R.: Living room中国煤化工CNMHGNATURAL GAS AND INDOOR AIR POLLUTIONAs shown in Table 4, CO2 concentrations were lowest either in living rooms or inkitchens before the stoves were turned on, but they were steadily rising above the standardalong with stove operation(0. 1% in public places). They were the highest in the NG grand declined gradually after the stoves were turned off.TABLE 4COz Concentrations in Indoor Air(%)Sites ofLPGRangeK.0.09130.072-0.1200.05520.04200690.138500700.210Before Fire Lr.07150064008300010.0600.16640085-0.240K.0.12530.0550.2200.13020.1190.1420.25440.1200.420K.0.228500900.2300.18230.1600.1990431702500.72030-45 minK.0.33650.2000.5000.19890.1800.240060790.2500.79030-45 minLR.0.13630.9-0.170007840.0580.1015-15 min.20740.1700.2800.14530.035070.1300.700K.0.12330.1200.1280.118300.33350.80.590K.0.13720.1030.1700.10430.10140.1060.21890077-0460Fire Off200680.0800.67200540.0800.14580.0560.250Note. K: kitchen: L R: living rooms.Method of AnalysisCO2 concentrations were determined by portable infrared analyzer GXH-3010C. In theightly closed kitchens the operator analyzed CO2 concentrations (%)during the combustionof NG, CG and LPG and at different periods of time after turning off the stoves, and livingrooms were used as controls. The CO2 standard in public places is 0. 1% or 0. 15%Organic Compounds in Combustion Products of NG, CG and LPGFindings in the studies are illustrated in Tables 5-8TABLE SContents of Organic Compounds on Particles as Combustion Products of NG CG and LPGTypes of GasContents of Organic Compounds(%)0.210.88中国煤化工CNMHGZHANG ET'ALMolecular FormulaMolecular Weight Chemical NameCuH3.3-dimethyl Hendecane2. 6,8-trimethyl Decane2682,4, 10, 14-tetramethyl Pentadecane5CushawC2 hu2, 6, 10, 14-tetramethyl HeptadecaneCishan(1-methyl dodecyle)-benzeneTABLE 7on Products of LPG Identified by GCMS- Molecular FormulaMolecular Weight ChemicalCubes3, 7, 7-trimethyl Decane2, 6, 8-trimethyl DecaneHeptadecaneC18H3OctadecaneCahe2, 6, 10, 14-tetramethyl HeptadecaneNonadecaneLarnel1C6H1013C2H4DocosaneCMhsTetracosane16CaUseCnhseptacosaneTable 8Organic Compounds as Combustion Products of CG Identified by GS/MSNo. Molecular Formula Molecular Weight Chemical NamesCutS2, 6, 8-trimethyl DecaneC12H8AcenaphtherC13H283,3-dimethyl Undecane5-propyl Tridecane(to be continued on the next page)中国煤化工CNMHGNATURAL GAS AND INDOOR AIR POLLUTIONMolecular FormulaHeptadecanePhenanthren567890u2345678g0Cushy(1-methyl dodecyl)-benzeneCcG282Eicosane96Chas310DocosaneCashesPentacosaneCaHCashesOctacosaneNonacosaneCir H3SON2819-octadecene AmideRadon Concentrations in Natural Gas Fields in BeijingRadon concentrations were determined in natural gas fields in Beijing, and the findingsare illustrated in Tables 9-10TABLE 9Daily Changes of Rn Concentrations in Natural Gas in Beijing(Ba/m)Times ofSampling.Morning104198494106100lll94101Note, 1992TABLE 10Monthly Changes of Rn Concentrations in Natural Gas in Beijing( Ba/m)0I157120103137106128921219770112Note.:1993As shown in Tables 9-10, Rn concentrations in the noon and evening, i.e. the peak timeof daily gas use, were basically similar, being 106 and 104 Bq/m respectively, while theaverage Rn concentrations in the morning were 92 Bq/m, lower than that in the noon and in中国煤化工CNMHGZHANG ETALthe evening. This might be explained by the fact that after staying in the NG pipe over nightRn went partially decayed in the morning. In the measurement for 9 successive days inMarch, the daily Rn concentrations in NG ranged from 84 to 1ll Bq/m, or lower thanrelative changes of RN concentrations within one day. The Rn concentrations ranged from70 to 157 Bq/m within a year, with an average value of 112 Bq/m and no evident seasonalchange was observed.The average Rn concentrations in the Renqiu gas field and the final users determinedfrom March 1992 to February 1993 was 112 Bq/m, which was higher than theconcentration of 73 Bq/m in the gas fields of north China as reported in 1986, but waslower than that in the coal gas used in Beijing and higher than that in the locallyliquefied petroleum gas". Nevertheless, this figure proved much lower than mataverages of Rn concentrations in 15 gas fields in Sichuan ProvinceConcentrations of 1-hydroxy Pyrene in Urine of the Study SubjectsAs shown in Table 11, 1- hydroxy poyrene concentrations in subjects’ urine werepositively correlated to B(a)P concentrations in the air of kitchens. The highest B(a)Pconcentrations were found either in subjects urine or in the air exposed to LPGi, the secondhigh concentrations were found in those exposed to CG and the lowest in those exposed toTABLE 11Concentrations of 1-hydroxy Pyrene in Subjects" Urine and in the Indoor AirB(a)P in Indoor Air(ng/m)Hydroxy Pyrene in Urine(umol/molNumber339±0930.277±0.237128±0.520.147±0.173DISCUSSIONComparison of Combustion Efects of NG CG and LPGAmong the three gases, NG showed the best combustion effects, which was followed inorder by LPG and CG. This is closely associated with their respective chemical composition(89%CH, for instance, in NG, 50% C3 Hg and 50% C4Hlo in LPG and 27% CH4 in CG). Thecomponents of the combustion products were directly related to composition of the gasesthemselves,as well as to conditions of combustion. the organic matters contained on theparticles as the combustion products of the 3 gases all included three normal alkanes, threebranched-chain alkanes and one alkyl substitute of benzene, suggesting that despite somedifferences in their original composition the products of combustion in them are partiallyidenticalComparison of Organic Components of NG CG and LPGBoth LPG and CG had rather complicated organic components, and the particles as theirombustion products all contained polycyclic aromatic hydrocarbons in addition to several中国煤化工CNMHGNATURAL GAS AND INDOOR AIR POLLUTIONnormal alkanes. The CG particles differed from the LPG ones by containing one morenormal alkanes, two additional polycyclic aromatic hydrocarbons(namely, acenaphtheneand fluoranthene)and also octadecene amide.Particles produced in the combustion process of the three gases all contained b(a)P as aresult of indoor air monitoring (Table 2). However, B(a)P was absent in their combustionproducts according to the findings of GS/MS, and pyrene was not detected in NG combustionproducts by GS/MS (Tables 6, 7 and 8). For B(a)P, HPLC with a sensitivity of 10 g wasapplied while for pyrene GS/MS with a sensitivity of 10"g was used. It is important to notethat though containing lower hydrocarbons as its chief components, NG was still able topolymerize into polycyclic aromatic hydrocarbons under a special combustion processEstimation of Internal Radiation Doses of Rn in Beijing InhabitantsBoth Rn and its daughter elements of short shelf life will produce internal radiationloses once they are inhaled. According to the model of dose estimation described in thedoICKP report(No 30), the aerodynamic diameter equals to 0. 2 um, suppose Rn's daughterelements of short shelf life have fp equivalent to 0.025. Assuming that F is exaggerated as Ior the equilibrium ratio between Rn and its daughter clements equals to 1, and that theinhabitants stay in indoor conditions all a year round, then the equivalent effective doses ofrespectively, and 1.88 HSv. a"in total. These doses only represent 1.8% of the yearlyequivalent effective doses to which Beijing inhabitants are exposed in indoor and outdoorconditions, or 0.8%o of the yearly equivalent effective doses(2. 43 HSv.a")to which they areexposed from all natural radiation sources in the cityCONCLUSIONSIn general, the indoor air pollution caused by NG, CG and LPG will be more severe inwinter than in summer and more evident in kitchens than in other rooms of the house. Theconcentrations of traditional pollutants such as particles, NO2, CO, SO2 are all withinnormal range, but they are lower in NG than in CG and LPGCombustion products of all the three gases contained lower organic components, nemore than 1% and of them NG was at the lowest levelThe NG from the Renqiu Oil Field which is currently used in Beijing has low Rn and itsdaughter elements, and their yearly effective radiation dose to which Beijing inhabitants arexposed is less than 1%0; nevertheless, in case the NG from Shaanxi Oil Field is used in thefuture, further monitoring of Rn will be necessaryTo reflect biological effect of B(a)P, the concentration of l-hydroxy pyrene wasdetected in urine samples of the subjects and the findings indicate that it is a fairly sensitiveindicatorConsidering the difficulty in sampling combustion products of the gases and a largequantity of samples required to test bioactivity of these products, micro Ames test was notestablished in the present study, but according to the findings in organic matters andavailable reports in the literature, it could be estimated that combustion products of NG willnot display high bioactivity, or their bioactivity will be lower than that of CG and LPGSample size in the present study was relatively small due to physical, technical andmanpower limitations and further study will be needed to explore the health impact ofnatural gas, an ever widely used domestic energy source in China中国煤化工CNMHGZHANG ETALREFERENCES1. Cao, Shouren, Cui, Jiusi, Zhao, Bingcheng, and Li, Baocheng (1984). Methods of monitoring air pollutants.Beijing Chemical Industry Publishing House, Edition 1, pp 114-117 and(In Chinese)2. Chen, Baosheng, Li, Shuming, and Zhang, Po(1988). Identificationdrocarbon in rural household by GCMS. Acta of Mass Spectorgraphy 9(54)(Suppl.3. Zhao, Zhenhua(1993). Chemistry of polycyclic aromatic hydrocarbons in relation to environmental health,China Science and Technology Publishing House, Beijing, pp. 98-121 (In Chinese4. Lin, Lianqing(1990). Study of source of domestic Rn in Beijing areas. Chinese Journal of Radiation Medicined Protection 10(1), 10-14(In Chinese)5. Lin, Lianqing and Chen, Zhipeng(1990). Natural Baseline radiation level in Beijing area. Chinese Journal ofRadiation Medicine and Protection 10(2), 73-77.(In Chinese)(Received February 21, 2003 Accepted June 2. 2003)中国煤化工CNMHG