Sources of PM10 and PM2.5 in Cairo's ambient air - Hashemite ...

Environ Monit Assess (2007) 133:417–425 DOI 10.1007/s10661-006-9596-8

Sources of PM10 and PM2.5 in Cairo’s ambient air M. Abu-Allaban & D. H. Lowenthal & A. W. Gertler & M. Labib

Received: 12 September 2006 / Accepted: 30 November 2006 / Published online: 1 February 2007 # Springer Science + Business Media B.V. 2007

Abstract A source attribution study was performed to assess the contributions of specific pollutant source types to the observed particulate matter (PM) levels in the greater Cairo Area using the chemical mass balance (CMB) receptor model. Three intensive ambient monitoring studies were carried out during the period of February 21–March 3, 1999, October 27–November 27, 1999, and June 8–June 26, 2002. PM10, PM2.5, and polycyclic aromatic hydrocarbons (PAHs) were measured on a 24-h basis at six sampling stations during each of the intensive periods. The six intensive measurement sites represented background levels, mobile source impacts, industrial impacts, and residential exposure. Major contributors to PM10 included geological

M. Abu-Allaban Department of Water Management & Environment, Faculty of Natural Resources & Environment, The Hashemite University, Zarqa, Jordan D. H. Lowenthal (*) : A. W. Gertler Desert Research Institute, 2215 Raggio Parkway, Reno, NV 89512, USA e-mail: [email protected] M. Labib Egyptian Environmental Affairs Agency, Misr Helwan Road Bldg. 30, Maadi, Cairo, Egypt

material, mobile source emissions, and open burning. PM2.5 tended to be dominated by mobile source emissions, open burning, and secondary species. This paper presents the results of the PM10 and PM2.5, source contribution estimates. Keywords Urban air pollution . Source apportionment . Particulate matter . Ambient lead

Introduction Cairo, Egypt suffers from high ambient concentrations of atmospheric pollutants (Nasralla 1994; Sturchio et al. 1997), including particulate matter (PM), carbon monoxide (CO), oxides of nitrogen (NOx), ozone (O3), and sulfur dioxide (SO2). Nasralla (1994) reported particulate lead concentrations ranging from 0.5 μg/m3 in a residential area to 3.0 μg/m3 at the city center. Sturchio et al. (1997) measured total suspended particulate (TSP) and lead concentrations using stable isotopic ratios ( 207 Pb/ 204 Pb and 208 Pb/204Pb) at eleven sites in Cairo. Lead and TSP concentrations ranged from 0.08 and 25 μg/m3, respectively, at Helwan to over 3 and 1,100 μg/m3, respectively, at the city center. Rodes et al. (1996) measured fine (PM2.5) and coarse (PM10–PM2.5) concentrations as a part of a source apportionment study in Cairo from December 1994 through November 1995. The annual average

418

Environ Monit Assess (2007) 133:417–425

PM10 (sum of PM2.5 and coarse) concentrations exceeded the 24-h average US standard of 150 μg/ m3 at all sites except Ma’adi and the background site. An attempt was made to attribute the high PM levels to specific sources using the Chemical Mass Balance (CMB) source apportionment model (Watson et al. 1990). PM10 mass was dominated by the coarse fraction, suggesting a strong influence of fugitive dust sources. Emissions from mobile sources, oil combustion, and open/trash burning dominated the PM2.5 apportionments. In order to develop and implement a pollutioncontrol strategy and to reduce the health impact of air pollution in Cairo the Cairo Air Improvement Project (CAIP) was established. As part of the CAIP, source attribution studies were performed to assess the impact of various sources (e.g., lead smelters, motor vehicles, oil combustion, open burning, geological material, etc.) to ambient pollutant levels. In this paper we report the PM source attribution results of the ambient monitoring study performed during the periods of February 21–March 3, 1999, October 27–November 27, 1999, and June 8–June 26, 2002.

Experimental methods Sampling sites Six sites were selected from CAIP network (Gertler et al. 1999). Sites included: Background: Kaha, a Nile delta site with significant agricultural activity. During most of the year, the prevailing winds come from this direction. Industrial/Residential: Shobra El-Khaima and ElMaa’sara were chosen to represent residential areas that exhibit potential industrial activities. The Shobra site is located in a heavily industrialized area and is downwind from numerous Pb smelters and other industrial sources. This is one of the most highly polluted areas in the city. The El-Maa’sara site is near a number of cement plants and other industrial sources. Traffic: El-Qualaly Square, a site located downtown. The site is close to the road and has high light- and heavy-duty (bus) traffic.

Residential: Helwan and El-Zamalek were chosen. Helwan is impacted by emissions from nearby cement plants and has higher PM levels than some of the other residential areas. ElZamalek is located on one of the islands in the Nile and represents a residential area with limited nearby sources. Ambient measurements Ambient PM2.5 and PM10 samples were collected at the six sites using the sampling protocol described by Watson et al. (1994). All samples were of 24-h duration. During the February/March, 1999 study, samples were collected daily, while in the October/ November, 1999 and June, 2002 studies samples were collected every other day. Two medium-volume samplers designed to collect samples for chemical analyses were utilized. This type of sampler employs a Sierra-Andersen 254 PM10 inlet or Bendix PM2.5 cyclone to determine the size fractions collected. The ambient air is transmitted through the size-selective inlet and into a plenum. The inlet is located at the top of the plenum. Maintaining a constant pressure across a valve with a differential pressure regulator controls the flow rate in the sampler. For the size-selective inlet to work properly, a flow rate of 113 lpm must be maintained through the sampler. Flow rates of 20 lpm through each Savillex filter holder were used to collect adequate samples for gravimetric and chemical analyses. This flow rate was drawn simultaneously through two parallel filter packs, one with a ringed 47 mm Teflon-membrane filter (Gelman Scientific, Ann Arbor, MI) and one with a 47 mm quartz-fiber filter (Pallflex Corp., Putnam, CT). The remaining 73 lpm was drawn through a makeup air port. The flow rates were each set with a calibrated rotometer and were monitored with the same rotometer at each sample change. In addition to those samples recommended by Watson et al. (1994), we measured polycyclic aromatic hydrocarbons (PAHs). The PAHs were critical to apportion the carbon components of the PM based on the uniqueness of PAH compounds associated with light-duty spark ignition (LDSI) and heavy-duty diesel (HDD) vehicles (Fujita et al. 1998). PAH samples were collected on Teflonimpregnated glass fiber filters followed by an adsor-


bent cartridge of polyurethane foam and XAD-4 resin (TIGF/PUF/XAD-4) using the Sequential Fine Particulate/Semi-Volatile Organic Compounds Sampler (PSVOC sampler). The PSVOC sampler inlet is selected for PM2.5 with a Bendix 240 cyclone operating at 113 lpm. Prior to sampling, all sampling media were cleaned in the laboratory. The PUF plugs and XAD-4 resins were assembled into glass cartridges (10 g of XAD between two PUF plugs), wrapped in aluminum foil and stored in a clean freezer prior to shipment to the field. All filter packs and cartridges were stored cold and were shipped to and from the field by overnight carrier in coolers with blue ice packets. Source measurements Source emissions samples were collected using methods similar to those used in the ambient sampling program. Bulk soil and road dust samples were collected at each of the ambient-sampling sites. Emissions from various sources including brick manufacturing, cast iron foundry, copper foundry, lead smelting, refuse burning, Mazot oil combustion, refuse burning, and restaurants were sampled. Individual motor vehicle emissions were sampled from heavy- and light-duty diesel vehicles, spark ignition automobiles, and motorcycles.

419

Atomic Absorption (AA) Spectrometer. Organic and elemental carbon are measured by thermal/optical reflectance (TOR) on 0.5 cm2 punches taken from the remaining half of the quartz-fiber filter (Chow et al. 1993). The PAHs samples were analyzed for PAHs following the protocol described by Fujita et al. (1998). Receptor modeling methods The Chemical Mass Balance (CMB) receptor model was used to apportion PM and its chemical constituents to their sources (Watson et al. 1990). The CMB solution is based on effective variance weighting, which assigns more importance to species with lower relative uncertainties (Watson et al. 1984). The CMB procedure requires several steps. First, the contributing sources must be identified and their chemical profiles must be entered. Then the chemical species to be included in the model must be selected. The next step is the estimation of the fractions of each chemical species contained in each source type and the estimation of the uncertainties in both the ambient concentrations and source contributions. The final step is the solution of the set of chemical mass balance equations. These procedures are described in detail in an application and validation protocol (Pace and Watson 1987).

Analytical methods

Results

Teflon-membrane filters were weighed on a Cahn 31 Electro-microbalance before and after sampling to determine mass concentrations. Chemical analyses were performed on Teflon-membrane and quartz-fiber filters following the methodology described by Watson and Chow (1993). Briefly, the Teflon-membrane filters are analyzed for elements (Mg, Al, Si, P, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Br, Rb, Sr, Ba, U and Pb) by X-ray fluorescence (XRF) using a Kevex 700/800 analyzer. One-half of the quartz filter is extracted with distilled-deionized water. The extract was analyzed for chloride, nitrate, and sulfate ions by ion chromatography (IC) using a Dionex 4000I ion chromatograph, for ammonium by automated colorimetry (AC) using a TRAACS 800 Technicon auto analyzer, and for sodium and potassium by a Perkin-Elmer Model 2380 Double Beam

Mass and inorganic chemical species Upon analyzing the filters, 80 chemical species were detected. Average PM10, PM2.5 concentrations and major chemical species are presented in Tables 1 and 2. Other species are not shown because they were found in small amounts, but they were used in the CMB modeling. Most of the major species concentrations were orders of magnitude above their minimum detectable limits (MDL’s). As noted above, the effective variance solution in the CMB assigns more importance to species with low relative uncertainties. Mass concentrations in both size fractions were higher at all sites (except for PM2.5 at Shobra) during the fall 1999 sampling period. Shobra exhibited the highest average PM10 and PM2.5 mass concentrations. The lowest values in both size fractions were

420


Table 1 Statistical summary of PM10 mass and selected chemical species for the six intensive sites (average ± standard deviation, μg/m3) Species

Size

Sampling period

El-Zamalek

El-Qualaly

Helwan

Kaha

El-Maa’sara

Shobra

Mass

PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10 PM10

Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002

127.2±6.5 248.5±13.4 99.2±5.0 18.3±1.2 51.2±4.0 2.1±0.2 4.5±0.3 5.5±0.4 4.1±0.2 8.7±0.5 17.2±0.9 9.8±0.6 9.3±0.6 8.6±0.5 2.6±0.2 16.1±1.4 63.0±4.3 21.2±1.6 16.7±1.9 14.2±1.0 7.8±0.9 1.8±0.6 5.6±1.9 2.5±0.8 5.5±1.8 16.0±5.6 9.2±3.0 7.0±1.2 13.0±2.8 6.2±1.1 2.1±0.1 4.4±0.2 2.4±0.1 1.6±0.1 1.0±0.1 0.2±0.0

219.9±11.1 251.6±13.5 136.4±7.1 19.6±1.3 132.4±8.6 3.2±0.3 5.6±0.4 4.6±0.4 5.1±0.3 13.0±0.8 18.4±1.8 10.8±0.7 7.7±0.5 9.6±0.6 2.4±0.2 48.5±4.0 73.1±4.9 34.3±2.5 20.3±2.4 18.2±1.3 17.2±1.9 3.0±0.9 4.7±1.6 2.8±1.0 9.6±3.1 13.5±4.7 10.3±3.6 18.7±3.2 13.4±2.9 8.7±1.5 4.3±0.2 4.2±0.2 3.4±0.2 4.8±0.3 1.8±0.1 0.5±0.0

88.1±4.6 146.3±9.4 141.9±7.4 5.1±0.4 69.4±5.3 3.7±0.3 3.6±0.3 2.9±0.3 6.9±0.4 6.1±0.4 13.8±1.5 9.8±0.6 1.8±0.1 3.1±0.3 1.8±0.2 15.0±1.3 39.9±3.0 28.6±2.2 6.9±0.9 7.8±0.6 7.8±0.9 1.5±0.5 3.4±1.1 3.2±1.1 4.9±1.6 10.4±3.6 12.0±4.2 9.8±1.8 15.8±3.2 15.8±2.8 1.6±0.1 3.6±0.2 3.0±0.2 0.2±0.0 0.3±0.0 0.2±0.0

93.0±4.8 204.7±11.1 100.0±5.3 15.1±1.0 43.7±3.5 2.1±0.2 4.9±0.3 5.5±0.4 3.9±0.2 5.9±0.4 13.0±1.1 8.2±0.5 9.4±0.6 9.2±0.6 1.8±0.2 14.6±1.3 55.5±4.0 21.4±1.7 7.9±0.9 9.6±0.7 5.3±0.6 1.6±0.5 5.8±1.9 3.4±1.1 4.2±1.4 16.8±5.8 11.0±3.7 2.5±0.5 6.4±1.3 4.1±0.7 1.2±0.1 4.0±0.2 2.9±0.2 0.1±0.0 0.1±0.0 0.0±0.0

186.1±9.4 317.4±17.4 175.3±9.1 10.7±0.7 70.8±5.9 4.9±0.3 4.9±0.3 4.0±0.3 7.2±0.4 10.4±0.6 20.5±1.4 11.6±0.7 1.4±0.1 4.0±0.3 2.0±0.2 22.4±2.0 68.7±4.8 28.1±2.2 7.5±0.9 8.3±0.6 5.8±0.7 2.8±0.9 6.7±2.2 3.3±1.1 9.2±3.0 21.2±7.3 13.7±4.6 30.1±5.2 41.5±7.4 27.8±4.9 3.0±0.2 6.6±0.4 3.6±0.2 0.7±0.0 0.4±0.0 0.1±0.0

265.1±13.6 360.3±19.2 153.7±8.0 26.4±1.7 118.1±8.3 4.6±0.3 5.4±0.4 4.9±0.4 4.5±0.2 10.6±0.6 21.6±1.9 12.4±0.8 7.7±0.5 7.9±0.5 2.4±0.2 42.2±3.5 86.5±5.8 30.2±2.3 10.0±1.2 13.4±1.0 8.7±1.0 3.3±1.0 7.7±2.5 3.7±1.2 13.6±4.4 26.3±9.0 15.7±5.3 10.4±1.9 21.4±4.4 8.8±1.5 6.0±0.3 8.3±0.4 4.6±0.2 33.7±1.9 12.7±1.0 7.2±0.6

Chlorine

Nitrate

Sulfate

Ammonium

Organic Carbon

Elemental Carbon

Aluminum

Silicon

Calcium

Iron

Lead

observed at Helwan, a residential location. The background site, Kaha, generally had the second lowest mass levels. The correlations between measured and reconstructed (sum of species) were very high, which indicates that the data quality for the CAIP particulate measurements was quite good. The sum of the species consistently accounted for 70–80% of the measured mass. The difference is accounted for by the fact that the sum of species does not contain oxygen associated with geological species (e.g., Al, Fe, Si) or hydrogen, oxygen, nitrogen, and sulfur associated with organic carbon.

The PM2.5/PM10 ratio varied from 0.3 at ElMaa’sara to 0.8 at Shobra. El-Maa’sara is an industrial location impacted by emissions from nearby cement plants. The ratio of 0.3 at El-Maa’sara is consistent with coarse particle emissions from those activities. Shobra is a highly industrialized site with a number of lead smelters in the vicinity. The ratio of 0.8 is likely due to the impact of fresh combustion emissions, although it is still unusually high. One might have also expected very high ratios at El-Qualaly, the mobile source site; however, the observed ratio was 0.4. This site also had high levels of crustal species in


421

Table 2 Statistical summary of PM2.5 mass and selected chemical species for the six intensive sites (average ± standard deviation, μg/m3) Species

Size

Sampling period

El-Zamalek

El-Qualaly

Helwan

Kaha

El-Maa’sara

Shobra

Mass

PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5 PM2.5

Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002 Winter 1999 Fall 1999 Summer 2002

61.9±3.2 131.8±7.2 39.6±2.1 12.5±0.8 35.7±3.3 0.3±0.1 2.7±0.2 5.5±0.4 0.5±0.1 6.0±0.4 12.9±0.7 7.7±0.5 8.7±0.5 10.7±0.6 2.6±0.2 11.1±1.0 44.3±3.2 16.3±1.3 7.9±1.0 14.7±1.1 7.3±0.8 0.1±0.0 0.4±0.0 0.1±0.0 0.3±0.0 1.2±0.1 0.5±0.0 0.2±0.0 1.0±0.1 0.3±0.0 0.2±0.0 0.7±0.0 0.2±0.0 0.7±0.0 0.7±0.1 0.1±0.0

84.6±4.3 135.1±7.3 59.3±3.1 9.2±0.6 44.4±3.4 0.3±0.1 3.0±0.2 4.6±0.4 0.6±0.1 6.7±0.4 11.7±1.1 8.1±0.5 6.8±0.4 9.4±0.6 2.6±0.2 23.2±2.0 44.3±3.1 26.5±2.0 13.0±1.5 22.1±1.6 16.3±1.8 0.1±0.0 0.3±0.0 0.2±0.0 0.5±0.0 1.1±0.1 0.6±0.0 0.8±0.0 0.9±0.1 0.4±0.0 0.5±0.0 0.6±0.0 0.4±0.0 1.6±0.1 1.4±0.1 0.3±0.0

29.4±1.7 99.9±6.0 47.9±2.6 1.4±0.2 14.6±1.2 0.6±0.1 1.7±0.2 2.9±0.3 1.7±0.1 4.0±0.3 8.8±0.5 6.7±0.4 2.1±0.1 6.1±0.5 2.0±0.2 7.3±0.7 22.4±1.9 19.3±1.5 5.9±0.7 5.6±0.4 7.4±0.9 0.0±0.0 0.2±0.0 0.3±0.0 0.2±0.0 0.7±0.0 1.3±0.1 0.3±0.0 1.0±0.1 1.4±0.1 0.1±0.0 0.6±0.0 0.5±0.0 0.1±0.0 0.2±0.0 0.1±0.0

49.7±2.7 111.4±6.2 34.7±1.9 10.3±0.7 44.9±3.4 0.3±0.1 3.2±0.3 5.5±0.4 0.4±0.1 4.8±0.3 8.6±0.5 5.5±0.4 7.9±0.5 9.1±0.6 1.8±0.2 10.2±1.0 45.6±3.5 14.9±1.2 5.9±0.7 8.5±0.7 4.5±0.6 0.1±0.0 0.3±0.0 0.2±0.0 0.1±0.0 1.2±0.1 0.7±0.0 0.1±0.0 0.5±0.0 0.3±0.0 0.1±0.0 0.5±0.0 0.3±0.0 0.0±0.0 0.1±0.0 0.0±0.0

60.9±3.3 107.5±6.4 48.3±2.6 5.4±0.4 131.6±14.7 0.8±0.1 2.8±0.2 4.0±0.3 1.8±0.1 5.9±0.4 10.7±0.6 7.7±0.5 3.0±0.2 8.9±0.6 2.2±0.2 9.4±0.9 37.2±2.9 18.2±1.4 7.6±0.9 6.7±0.5 5.8±0.7 0.2±0.0 0.5±0.0 0.2±0.0 0.9±0.1 1.6±0.1 0.9±0.1 3.9±0.4 3.6±0.2 2.0±0.1 0.5±0.0 0.9±0.1 0.4±0.0 0.4±0.0 0.4±0.0 0.1±0.0

216.1±11.0 173.5±9.3 60.7±3.2 22.4±1.4 37.9±2.8 1.6±0.2 4.5±0.3 4.9±0.4 0.6±0.1 9.1±0.6 14.0±0.8 9.6±0.6 7.6±0.5 9.8±0.6 2.7±0.2 32.7±2.7 55.6±3.9 22.2±1.7 12.4±1.5 16.1±1.3 7.4±0.8 0.9±0.1 0.5±0.0 0.1±0.2 5.4±0.3 3.3±0.2 1.9±0.1 5.5±0.5 1.5±0.1 0.4±0.0 4.2±0.2 1.6±0.1 0.7±0.0 26.8±1.4 9.2±0.8 5.1±0.4

Chlorine

Nitrate

Sulfate

Ammonium

Organic Carbon

Elemental Carbon

Aluminum

Silicon

Calcium

Iron

Lead

the PM10 fraction, likely due to resuspended road dust, leading to the reduced ratio. Crustal components (Si, Ca, Fe, and Al) were significant at all sites. The majority of crustal material was in the coarse (PM10–PM2.5) fraction. The highest concentrations of PM10 crustal species, e.g., Si, were found at Shobra and El-Maa’sara, probably as a result of fugitive dust emissions from industrial operations at these sites. Organic carbon (OC), and elemental carbon (EC) were major components of PM at all sites. Potential

sources include mobile emissions, open burning, and fossil fuel combustion. The highest average PM10 OC levels were observed at El-Qualaly and Shobra. Shobra exhibited the highest lead (Pb) concentrations. The average PM10 and PM2.5 ambient lead concentrations were 7–34 and 5–27 μg/m3, respectively. Note that average winter 1999 Pb concentrations at Shobra were five times higher than those in summer 2002. Because other mass and species concentrations were generally higher during 1999,

422

we suspect that industrial lead operations may have been shut down or diminished at Shobra during the summer 2002 period. Lead concentrations at Shobra and Cairo in general are in excess of PM10 and PM2.5 mass concentrations observed in many cities in the US. It should be noted that leaded fuel was phased out in 1996 and so most of this Pb probably comes from other sources. Source profiles Fine and PM10 particulate source profiles used in the CMB modeling are presented in Table 3. Some profiles were developed as part of this study (CAIP profiles), while other profiles were compiled from previous studies (Chow and Watson 1999; Kuykendal 1990; and Rodes et al. 1996). Tests were done to determine which profiles best explained the ambient data. Geological CAIP profiles were used in all cases. These included soil (SOIL), unpaved road dust (UPRD), and paved road dust (PVRD) collected at the six sites. We relied on the AUTO and DIESEL profiles determined during the Rodes et al. (1996) study. These profiles fit the data better than did the CAIP motor vehicle profiles, probably because the OC/total carbon ratio in most of the CAIP motor vehicle profiles was greater than 90% while this ratio in the ambient fine samples averaged only 60%. The CAIP copper foundry and lead smelter profiles were needed to account for Zn and Pb, respectively. Although we used the Rodes et al. (1996) cement profile, we found that samples in both size fractions at El-Maa’sara were enriched in calcium, which could not be explained by the geological nor by cement profiles. To account for the excess calcium in Cairo, we used a pure CaCO3 (limestone) profile. We assumed that this was associated with cement production and report it as such. However, it is possible that the limestone contribution represents construction activities or other geological source contributions. The restaurant and gas power plant profiles were essentially pure OC. The Mazut power plant profile was enriched in vanadium (1.8%) and nickel (1.2%), which distinguishes oil combustion emissions. It was also enriched in sulfate (42%) and OC (15%), which made it collinear with vehicle and open burning as well as secondary sulfate profiles. The addition of PAHs to the CMB may help resolve this profile from other combustion profiles.


Soluble potassium (K+) accounted for over 90%, on average, of the fine potassium in the ambient samples. This is consistent with emissions from refuse and open burning, which are prevalent in Cairo. While soluble K+ is enriched in open burning emissions, it was essentially absent in all of the CAIP refuse burning profiles. For this reason, we used a open burning profile (VEGB1) from a US study (Chow and Watson 1999). Despite the utility of soluble K+ for distinguishing open burning contributions, there was still collinearity between the open burning and mobile source (DIESEL) profiles. Secondary ammonium sulfate (AMSUL), ammonium bisulfate (AMBSUL), sulfuric acid (H2SO4), and ammonium chloride (AMCL) profiles were needed to account for particulate sulfate, ammonium, and chloride. Pure sea salt profiles, needed to account for soluble sodium in the ambient samples, were used to account for marine aerosols. The ammonium chloride profile was used to account for high chlorine concentrations in excess of the marine contribution. PM CMB results The CMB results are presented in Table 4. The CMB model diagnostics for PM are quite good. For all average PM samples R2 was between 0.94 and 1.00. R2 greater than 0.8 is considered acceptable (Pace and Watson 1987). The calculated mass was within 10% of the measured mass. Chi-square, another measure of the goodness of fit, varied from 0.19 to 1.59, with the lowest values indicating the best fit. A chi-square less than two is considered acceptable (Pace and Watson 1987). A summary, by site, of the CMB results is presented below: Al El-Zamalek: The major sources of PM10 were geological material, Mazut oil, mobile sources, and open burning. Most of the secondary sulfate, ammonium chloride, and combustion (motor vehicle, open burning, and oil) were in the fine fraction. Lead and copper smelter contributions were small at this site. El-Qualaly: This site was dominated by mobile source emissions. PM10 was dominated by geological, mobile source, and open burning. About half of the mobile and open burning emissions was in the PM2.5 fraction. The lead and copper smelter contributions to PM10 were significant at this site.


423

Table 3 Source profiles used in the CMB source apportionment Number

Source ID

Description

Reference

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

BSSOIL HESOIL KASOIL ELMAUPRD QUAPVRD SHOBUPRD KAHAUPRD BRICK CASTFE CEMENT1 CUFOUNDR GASPP HDDBUS LDDMBUS LDSI1–3 LDSI4–5 LDSILATE MAZUTPP MOTO PBSMELT REFUSEBA REFUSED1 REFUSED2 REST AMSUL AMBSUL H2SO4 AMNIT NH4CL MAR100 27454 18323 37705 36586 AUTO CEMENT DIESEL OIL LEAD VEGB1 OFPP LIME

British School Soil Helwan soil Kaha soil El-Maa’sara unpaved road El-Qualaly paved road Shobra unpaved road Kaha unpaved road Brick foundry Cast iron foundry Cement plant Copper foundry Gas power plant Heavy duty diesel bus Light duty diesel microbus Light duty spark ignition Light duty spark ignition Late model light duty spark ignition Mazut oil power plant Motorcycle Lead smelter Refuse burning basatin Refuse burning desert Refuse burning desert Restaurant Secondary ammonium sulfate Secondary ammonium bisulfate Secondary sulfuric acid Secondary ammonium nitrate Secondary ammonium chloride Pure sea salt Sea salt 0.25 nitrate replacement Sea salt 0.5 nitrate replacement Sea salt 0.75 nitrate replacement Sea salt 1 nitrate replacement Light duty spark ignition Cement plant Heavy duty diesel Mazut oil Lead smelter Vegetative burning Oil fired power plant Limestone

CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP CAIP DRI PURE COMPOUND DRI PURE COMPOUND DRI PURE COMPOUND DRI PURE COMPOUND DRI PURE COMPOUND DRI – Watson et al. (1994) DRI – Watson et al. (1994) DRI – Watson et al. (1994) DRI – Watson et al. (1994) DRI – Watson et al. (1994) Rodes et al. (1996) Rodes et al. (1996) Rodes et al. (1996) Rodes et al. (1996) Rodes et al. (1996) Chow and Watson (1999) Kuykendal (1990) PURE CACO3

Helwan: PM10 at this site was dominated by geological material. While the marine contribution was almost entirely in the coarse (PM10– PM2.5) fraction, mobile, Mazut, and open burning contributions at this site were almost entirely in the fine fraction. Copper smelter contributions

were detected at low levels (1–2%) in both size fractions. Kaha: The largest contributors to PM10 were open burning and geological dust. Mobile source and Mazut oil contributions were small compared with other sites. PM2.5 was dominated by open burning.

Year

Winter 1999 Winter 1999 Fall 1999 Fall 1999 Summer 2002 Summer 2002 Winter 1999 Winter 1999 Fall 1999 Fall 1999 Summer 2002 Summer 2002 Winter 1999 Winter 1999 Fall 1999 Fall 1999 Summer 2002 Summer 2002 Winter 1999 Winter 1999 Fall 1999 Fall 1999 Summer 2002 Summer 2002 Winter 1999 Winter 1999 Fall 1999 Fall 1999 Summer 2002 Summer 2002 Winter 1999 Winter 1999 Fall 1999 Fall 1999 Summer 2002 Summer 2002

Location

El-Maa’sra El-Maa’sra El-Maa’sra El-Maa’sra El-Maa’sra El-Maa’sra El-Qualaly El-Qualaly El-Qualaly El-Qualaly El-Qualaly El-Qualaly El-Zamalik El-Zamalik El-Zamalik El-Zamalik El-Zamalik El-Zamalik Helwan Helwan Helwan Helwan Helwan Helwan Kaha Kaha Kaha Kaha Kaha Kaha Shobra Shobra Shobra Shobra Shobra Shobra

PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 PM2.5

Size

186±9 61±3 317±17 107±6 175±9 48±3 220±11 85±4 252±13 135±7 136±7 59±3 127±6 62±3 249±13 132±7 99±5 40±2 88±5 29±2 146±9 100±6 142±7 48±3 93±5 50±3 205±11 111±6 100±5 35±2 265±14 216±11 360±19 174±9 154±8 61±3

187±18 67±3 348±33 116±6 202±17 61±3 191±11 72±3 269±17 149±7 137±8 56±3 112±7 58±3 262±17 143±7 98±7 39±2 85±8 30±2 190±15 84±5 142±9 54±3 97±6 51±2 237±14 130±8 97±5 33±2 219±10 171±7 349±25 203±11 143±7 60±3

65±6 7±1 145±15 12±1 86±8 6±1 77±9 3±0 70±10 3±1 68±6 1±0 29±4 1±0 80±9 5±1 41±5 2±0 27±4 1±0 78±11 5±1 73±7 10±1 12±3 0±0 52±6 5±1 39±3 3±0 51±5 36±3 90±10 20±2 52±5 2±0

1±0 1±0 0±0 0±0 0±0 0±0 7±1 2±0 3±0 2±0 1±0 0±0 2±0 1±0 1±0 1±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 53±5 43±4 17±3 12±2 11±2 7±1

1±0 0±0 1±0 1±0 0±0 0±0 3±0 2±0 3±0 3±0 0±0 0±0 1±0 1±0 2±0 2±0 0±0 0±0 1±0 0±0 2±0 2±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 10±1 9±1 7±1 6±1 3±0 2±0

1±1 1±0 4±1 2±0 1±1 0±0 5±1 1±0 4±1 1±0 2±0 0±0 3±0 1±0 5±1 1±0 2±0 0±0 2±0 0±0 3±1 2±0 3±1 1±0 1±0 0±0 1±1 0±0 1±0 0±0 11±1 14±1 14±2 2±0 9±1 4±0

2±0 3±0 1±0 4±1 1±0 1±0 3±0 6±1 2±0 5±2 1±0 2±1 3±0 6±0 2±0 5±2 1±0 2±0 1±0 1±0 1±0 2±1 0±0 1±0 2±0 3±0 1±0 3±1 1±0 1±0 4±0 9±3 4±0 9±3 3±0 6±1

12±3 13±2 11±3 12±2 8±2 14±2 33±4 21±2 43±6 46±4 24±2 29±2 16±2 9±1 24±4 29±4 19±3 14±2 13±3 8±1 13±3 9±2 22±3 15±2 13±3 4±1 14±3 11±3 9±1 9±1 21±4 26±4 23±5 26±5 15±2 13±2

44±4 11±1 19±2 2±0 103±9 5±1 48±5 0±0 54±5 6±1 25±2 1±0 34±4 3±1 14±1 0±0 104±10 2±1 56±5 0±0 25±3 3±1 11±1 1±0 22±2 4±1 12±1 0±0 108±9 6±1 60±5 1±0 19±2 4±1 10±1 1±0 18±2 5±1 10±1 0±0 63±7 6±5 39±4 1±0 25±4 5±1 16±2 1±0 35±3 3±1 19±1 0±0 123±10 9±4 82±8 0±0 34±3 4±1 12±1 0±0 40±5 6±1 5±1 3±1 147±13 4±7 89±9 0±0 34±4 5±1 15±2 1±0

3±7 4±1 13±5 9±1 8±2 8±1 1±3 5±1 8±8 9±2 6±2 9±1 1±2 5±1 7±9 11±2 8±2 9±1 2±3 4±1 8±4 8±1 5±2 6±1 2±2 5±1 6±6 8±1 5±2 7±1 0±0 2±2 5±11 12±2 5±2 10±1

0±0 3±0 13±2 5±1 8±1 2±0 6±1 4±0 11±1 6±1 6±1 1±0 6±1 3±0 11±1 7±1 5±1 1±0 4±0 2±0 7±1 4±1 8±1 2±0 6±1 4±1 11±1 7±1 5±1 0±0 6±1 5±1 12±2 7±1 6±1 1±0

3±0 5±1 16±13 17±1 0±0 0±0 18±2 13±1 20±2 19±1 0±0 0±0 23±2 18±1 15±4 20±1 0±0 0±0 2±1 3±0 5±1 10±1 0±0 0±0 23±2 15±1 18±2 14±1 0±0 0±0 18±2 16±1 14±2 20±2 0±0 0±0

44±14 7±1 36±23 5±1 29±15 3±0 0±0 0±0 0±0 0±0 0±0 0±0 3±4 0±0 0±0 1±0 0±0 0±0 11±5 0±0 4±5 1±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 2±1 11±11 0±0 0±0 0±0

Measured Predicted Geological Lead Copper Steel Heavy Motor Open Marine Ammonium Ammonium Ammonium Cement mass mass material smelters smelters industry oil vehicles burning sulfate nitrate chloride plants

Table 4 Summary of PM2.5 and PM10 source attribution results for the six intensive sites (average±standard deviation, μg/m3)

424 Environ Monit Assess (2007) 133:417–425


El-Maa’sara: The PM10 fraction was dominated by geological material, cement, mobile sources, and open burning. The fine fraction was dominated by open burning and mobile emissions. The lead and copper smelter contributions were found almost entirely in the fine fraction. Shobra: The most unusual aspect of this location is the high PM Pb level. Eighty percent of the lead contribution was in the PM2.5 fraction. Most of this Pb is in the form of fresh emissions from secondary Pb smelters in the vicinity. Contributors to PM10 included geological material and mobile source emissions. The PM2.5 apportionment shows a similar distribution of source contributions, which is consistent with dominance of fine particles at this site.

Summary An intensive PM10 and PM2.5 sampling program was carried out at six sites in the greater Cairo area during a winter period from February 18 to March 4, 1999, during a fall period from October 29 to November 27, 1999, and during summer period from June 8 to June 26, 2002. Medium volume samplers were used to collect PM2.5, PM10, and PAH samples for subsequent chemical analysis and source apportionment modeling. The CMB receptor model coupled with source profiles measured during the CAIP and from previous studies was used to estimate source contributions to PM2.5 and PM10 mass. Depending on the sites, major contributors to PM10 included geological material, mobile source emissions, and open burning. PM2.5 tended to be dominated by mobile source emissions, open burning, and secondary species. Aside from the extremely high mass levels, two unusual features emerged. First, most sites had high levels of ammonium chloride during the two 1999 sampling periods. Second, lead concentrations were very high during winter 1999 at Shobra.

Acknowledgments The authors wish to acknowledge the support of USAID and the Egyptian Environmental Affairs Agency under contract no. 263-C-00-97-00090-00 for providing the funding for this work.

425

References Chow, J. C., & Watson, J. G. (1999). Cross border transport and spatial variability in Mexicali and California’s Imperial Valley. Atmospheric Environment, 34, 1833–1843. Chow, J. C., Watson, J. G., Pritchett, L., Pierson, W., Frazier, C., & Purcell, R. (1993). The DRI thermal/optical carbon analysis system: Description, evaluation and applications in U.S. air quality studies. Atmospheric Environment,27A, 1185–1202. Fujita, E., Watson, J. G., Chow, J. C., Robinson, N. F., Richards, L. W., & Kumar, N. (1998). Northern front range air quality study volume C: Source apportionment and simulation methods and evaluation. Final Report Prepared for Colorado State University. Gertler, A. W., Lowenthal, D. H., Howes, J. E., Jr., Sagebiel, J. C., Labib, M., Abu-Allaban, M., et al. (1999, June). Cairo source attribution study. (Paper presented at the AWMA 92nd National Meeting, St. Louis). Kuykendal, W. B. (1990). Air Emissions Special Manual. Volume II: Particulate Matter Species Profiles, U.S. Environmental Protection Agency (EPA), Research Triangle Park, N.C., EPA-450/2-90-0001b. Nasralla, M. M. (1994, October). Air pollution in greater Cairo. (Proceeding of the Italian–Egyptian study-days on the environment). Pace, T. G. & Watson, J. G. (1987). Protocol for Applying and Validating the CMB Model. EPA 450/4-87-010, U.S. Environmental Protection Agency, Research Triangle Park, NC. Rodes, C. E., Nasralla, M. M., & Lawless, P.A. (1996). An assessment and source apportionment of airborne particulate matter in Cairo, Egypt. Activity report no. 22, prepared for the USAID Mission to Egypt under EHP activity no. 133-RCm delivery order no. 7. Sturchio, N., Sultan, M., Sharkaway, M. E., Maghraby, A. E., & Taher, A. (1997). Concentration and isotopic composition of lead in urban particulate air, Cairo, Egypt, 1966. Argonne National laboratory, Argonne, IL, and Center for Environmental Hazard Mitigation, Cairo University, Cairo, Egypt. Watson, J. G., & Chow, J. C. (1993). Ambient air sampling. In K. Willeke, & P. Baron (Eds.), Aerosol measurement: Principles, techniques and applications. New York: Van Nostrand Reinhold, pp. 622–639. Watson, J. G., Chow, J. C., Lu, Z., Fujita, E. M., Lowenthal, D. H., Lawson, D. R., et al. (1994). Chemical mass balance source apportionment of PM10 during the Southern California Air Quality Study. Aerosol Science and Technology, 21, 1–36. Watson, J. G., Cooper, J. A., & Huntzicker, J. J. (1984). The effective variance weighting for least squares calculations applied to the mass balance receptor model. Atmospheric Environment,18, 1347–1355. Watson, J. G., Robinson, N. F., Chow, J. C., Henry, R. C., Kim, B. M., Pace, T. G., et al. (1990). The USEPA/DRI chemical mass balance receptor model, CMB 7.0. Environmental Software, 5, 38–49.