Air Toxics in El Paso, Texas: Implications for Cross-Border

TransportL.-W. Antony Chen, Richard J. Tropp, Dongzi Zhu

Division of Atmospheric Sciences, Desert Research Institute, Reno, NV 89512

Wen-Whai LiDept. of Civil Engineering, University of Texas at El Paso, El Paso, TX 79968

Elda Rodriguez-HefnerEnvironmental Services Department, City of El Paso, El Paso, TX 79901

Presented at

EPA’s National Ambient Air Monitoring Conference

November 2-5, 2009

Nashville, TN

Introduction•

The Paso del Norte (PdN) Region–

Formed by mountains that surround El Paso, TX, and Sunland Park, NM, in the US, and Ciudád Juarez, Chihuahua, Mexico

–

One of the largest metropolitan areas along U.S.-Mexico border (2 Million residents)

•

Emissions–

Cross border traffic is 18 million vehicles yearly–

Other sources•

Cooking•

Heating•

Brick manufacturing•

Petrochemical•

Waste burning–

Different standards between US and Mexico•

Has some of the highest acetaldehyde and formaldehyde concentrations in the US

Introduction -

2•

Previous Studies–

GIS epidemiological•

Strong correlations between mobile source emissions and children’s pulmonary functions

•

Chemical composition of and community exposure to these emissions have not been quantified

•

How air toxics levels could be assessed using existing indicators (e.g., criteria pollutants) is unclear

–

Studies focusing on VOC and PM•

Limited to 1-2 receptors in downtown El Paso•

Did not look at acute short-term exposure near sources or spatial gradients

–

Currently available information on air toxics and emission sources insufficient to adequately apportion community exposure to sources in region

Introduction -

3

•

Pilot Study Objectives–

Characterize levels of selected air toxics•

Urban residential areas influence by cross-border pollution•

Specific local sources–

Cross-border traffic–

Waste burning–

Associate air toxics levels to major environmental indicators and/or criteria pollutant concentrations

–

Determine and document air toxics emission factors•

Cooperative Effort–

City of El Paso –

administration and logistics–

UTEP –

logistics, sampling locations and personnel–

DRI –

equipment, sampling, and laboratory analysis

El Paso Monitoring Sites

Border-Crossing to Highway Ramp

Bowie H.S.

UTEP

Trash Burning at DRI

Ciudad Juárez, Mexico

Population: 1,400,891

Pop. Density: 7452/km2

El Paso, U.S.

Population: 606,913

Pop. Density: 944.7/km2

12/10-14/08

12/15-18/08

12/19-22/08

12/23/08, 1/7-9/09

El Paso Monitoring Sites -

2•

Ambient Monitoring Sites–

UTEP –

next to TCEQ CAMS–

Bowie HS -

~1 block from TCEQ Chamizal CAMS/PAMS

•

Source Oriented Monitoring Sites–

Cross-Border Mobile Sources•

Behind El Paso Zoo•

Location where both cars and trucks coming from Mexico pass by to get to other parts of El Paso or IH-10 and other parts of the US

–

Waste Burning•

Originally to be at dump in Juarez•

Had to be abandoned due to safety concerns•

Eventually used household garbage from El Paso landfill–

Added PVC and tire scraps–

Burned at DRI facility in Reno, NV

El Paso Monitoring Sites -

3

Views at the El Paso Zoo site monitoring border-crossing vehicle emissions

Measurements•

Spectrum of Measurements–

Continuous (≤5 min)•

Gaseous–

CO, CO2

, NO-NO2

-NOx

, SO2–

Aldehydes, VOC,

•

Particulate –

PM10

& PM2.5–

PAH, BC–

PSD (10nm –

10µm)–

Integrated (~1-9 hrs)•

Gaseous–

Auto-GC -

VOCs–

DNPH/HPLC –

Aldehydes, carbonyls–

Canisters/GCMS –

VOCs•

Particulate–

Teflon filter –

mass & elements by XRF–

Quartz filter –

non-polar HCs by TD-GC/MS

Measurements -

2Method Target Pollutantsa Avg. Time Frequency

Fourier Transform Infrared Spectrometer (FTIR, Illuminator series, Midac, Costa Mesa, CA)

acetaldehyde, acrylonitrile, formaldehyde, hydrazine, methylene chloride, perchloroethylene, vinyl chloride, carbon monoxide, carbon dioxide, nitrogen oxides

2 sec Continuous throughout experiment

Auto-GC (TO-14 Air Sampling GC, SRI Instruments)

acrylonitrile, benzene, carbon tetrachloride, chloroform, 1, 3-dichloropropene, methylene chloride, perchloroethylene, propylene dichloride, 1, 1, 2, 2-tetrachloroethane, trichloroethylene, vinyl chloride, other VOC source markers

10 – 40 min

Hourly throughout experiment

DNPH/HPLC (2,4-dinitrophenyl-hydrazine derivatization cartridge sampling followed by High Performance Liquid Chromatography analysis)

acetaldehyde, acrolein, formaldehyde, other carbonyls 10 – 40 min

Overlap with auto-GC measurement; 10 – 15 per sampling location decided by the field operator for a total of 45 samples

Teflon Filter/XRF (Teflon membrane filter sampling followed by X-ray Florescence analysis)

arsenic, beryllium, cadmium, chromium, lead, manganese, mercury, nickel; and PM2.5 mass (by gravimetry)

10 – 40 min

Overlap with DNPH/HPLC measurement for a total of 45 samples

Quartz Filter/TD-GC/MS (Quartz-fiber filter sampling followed by Thermal Desorption- GC/MS analysis)

ethylene dibromide, ethylene dichloride, ethylene oxide, polychlorinated biphenyls, polycyclic organic matter*, quinoline*; and OC/EC (by Thermal/Optical Reflectance method)

10 – 40 min

Overlap with DNPH/HPLC measurement for a total of 45 samples

Canister/GCMS (Canister sampling followed by Gas Chromatography Mass Spectrometry analysis)

acrylonitrile, benzene, 1, 3-butadiene, carbon tetrachloride, chloroform, 1, 3-dichloropropene, methylene chloride, perchloroethylene, propylene dichloride, 1, 1, 2, 2-tetrachloroethane, trichloroethylene, vinyl chloride, other VOC source markers

10 – 40 min

Overlap with DNPH/HPLC measurement; 2 – 3 per sampling location decided by the field operator for a total of 10 samples

Photoelectric Aerosol Sensor (PAS, EcoChem Analytics PAS2000)

diesel particulate matter†, polycyclic aromatic hydrocarbons 1 min Continuous throughout experiment

CO2/H2O Gas Monitor (LiCor-840) carbon dioxide, water vapor 1 sec Continuous throughout experiment

DustTrakTM (Model 8520, TSI) PM10, PM2.5 by light scattering 1 sec Continuous throughout experimentElectric Low Pressure Impactor (Dekati ELPI) Particle size distribution (10 nm – 10 μm)‡ 5 sec Continuous throughout experiment

Aethalometer (Magee AE14U)b PM2.5 black carbon† by light absorption 5 min Continuous throughout experiment

a Detection limits vary (see SOPs); * Primarily for particle-phase fractions; † measure surrogate; ‡ aerodynamic diameter (12 size bins); b Not part of the In-Plume system and will be installed upon availability.

Measurements -

3

PTFE/Q

uartz Filters

Quartz/K

2 CO

3 Filters

Quartz/C

itric Acid Filters

Nucle

poreFilters

TSI 40241 M

assF

low Meters TS

I 40241 Mass

Flow Meters

Comtrol8 p

ort R

S232

to Ethernet

Gast

Pump

Gast

Pump

Gast Pum

pGast Pum

p

Gast

Pump

Gast

Pump

PM

2.5F

ilter Module

INLET FROM SOURCE

TSI DustTrak PM10

Grimm 1.108 Coarse

Grimm 1.108 Fine

TSI ELPI

MIDAC FTIR

VacuumPump

VacuumPump

Gast PumpGast Pump

Q = 10 lpm

Q = 2.5 lpm

Q = 2.5 lpm

Q = 1.7 lpm

Comtrol 8 port RS232 to E thernet

TSI 40241 M

assFlow

Meters F ie ld

Computer DAQ

EthernetHub

Q = 50 lpm

Com

trol8 port

RS23

2 to Ethernet

TSI DustTrak PM2.5 Q = 1.7 lpm

Bendex240

Cyclones

Mixing Plenum

Bendex240

Cyclones

Mixing Plenum

Real Time PM Module Gas and DAQModule

Li-Cor CO2Q = 1.0 lpm

Q = 113 L/m

LEGEND_____ Air Stream- - - -Data Stream

Li-Cor CO2

INLET FOR AMBIENT CO2

Q = 1.0 lpm

PhotoAcoust ic BCQ = 1.0 lpm

In-Plume Sampling System

Measurements -

4

DetectorlCA iii

Source

Radiation

H2

O and CO2

FTIR Spectra

Beer-Lambert law: log10

(1/T) = A absorbance

is absorption coefficient•

C is concentration•

L is the distance that the radiation travel through the sample l

1 =0 exp(-l)

0.5 cm-1 resolution, 10 m folded optic length, 2 L custom cell, 50 LPM flow, sample frequency once per 1.5 seconds.

Fourier Transform Infra-Red Spectrometer

Measurement -

5

ELPI

Electrical Low Pressure Impactor

Measures particle number concentration and aerodynamic size (between 7 nm and 10 m with filter stage)

Real-Time Continuous Particle Measurement

Measurements -

6

SRI Auto-GCDustTrak for PM by light scattering

EchoChem PAS for PAH, DPMAndersen Aethalometer for BC

Measurements -

7

Measurements -

8

PM2.5

y = 0.94x + 0.26R2 = 0.78

0

10

20

30

40

50

0 10 20 30 40 50

PM2.5 by TCEQ TEOM (g/m3)

PM2.

5 by

In-P

lum

e (

g/m

3 )

NOx

y = 1.18x + 0.02R2 = 0.98

0

0.05

0.1

0.15

0.2

0.25

0 0.05 0.1 0.15 0.2 0.25NOx by TCEQ (ppmv)

NO

x by

In-P

lum

e (p

pmv)

BC y = 1.24x + 0.60R2 = 0.66

0

3

6

9

12

15

0 3 6 9 12 15

EC by IMPROVE_A TOR Method (g/m3)

BC b

y Ae

thel

omet

er (

g/m

3 )

In-Plume System Measurement Comparisons

Results

0

5

10

15

20

25

30

35

40

12/9/08 12/11/08 12/13/08 12/15/08 12/17/08 12/19/08

Sampling Day

Wind Speed (m/sec)

0

90

180

270

360

450

540

Wind Direction (Deg)

RWSRWD

0

1

2

3

4

5

6

12/9/08 12/11/08 12/13/08 12/15/08 12/17/08 12/19/08

Sampling Day

CO, NOx, SO

2 Concentration (ppm

/ppb)

0

20

40

60

80

100

PM2.5 Concentration (g/m3 )

CONOxSO2 (ppb)PM25

UTEP Site (TCEQ C12) BOWIE Site (TCEQ C41)

0

10

20

12/9/08 12/11/08 12/13/08 12/15/08 12/17/08 12/19/08

Sampling Day

BC at 370/950 nm (g/m3 )

0

0.5

1

1.5

2

PAH, Toxics Concentration (g/m3 )BC 370 nm

BC 950 nmPAHformaldehydeacetaldehyde

UTEP Bowie HS

Ambient Monitoring Sites

Results -

2

Fleet average Emission Factor (g/kg fuel)

CO NO HC PM

Fleet Average 28.32 17.15 26.89 0.91

Standard Deviation 48.52 7.17 36.06 1.20

Sampled Vehicle Passes 971 971 971 296

On-Road Mobile Source Sampling

Results -

3

C O  E F  His tog ram

0

50

100

150

200

250

300

‐50 0 10 20 30 40 50 60 80 100 More

C O  E F  (g /kg  fuel)

Frequen

cy

Mobile Source CO Emission Factors

Results -

4Mobile Source NO Emission Factors

NO   His tog ram

0

50

100

150

200

250

300

‐10 0 5 10 15 20 25 30 40 50 More

NO  E F  (g /kg  fuel)

Frequen

cy

Results -

5

PM E F   His tog ram

0

10

20

30

40

50

60

70

80

0.05 0.1 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 4 More

PM E F  (g /kg  fuel)

Frequen

cyMobile Source PM Emission Factors

Results -

6

0.0001

0.0010

0.0100

0.1000

1.0000

PM

2.5

OC

EC As

Cd

Pb

form

alde

hyde

acet

alde

hyde

PA

Hs

Frac

tion

of P

M2.

5 Em

issi

on Motor VehicleGarbage Burn

Differences in Source Composition

Conclusions

•

Results are very preliminary•

Aerosol and air toxics levels highly influenced by wind direction, signifying cross-border transport

•

Good correlation between PM2.5

and CO and NOx Suggests on-road traffic as the dominant source of PM and polycyclic aromatic hydrocarbons (PAHs)

•

Emission profiles indicate strong emissions–

Formaldehyde from motor vehicles–

Black Carbon from unregulated domestic garbage burning•

Missing Auto-GC data making task of finding surrogates difficult

Future Work

•

Analysis for specific compounds•

Adjustments for time lags

•

Adjustments for missing data•

Source-receptor modeling

ReferencesChen, L.-W.A.; Moosmüller, H.; Arnott, W.P.; Watson, J.G. (2006). Novel approaches to

measure diesel emissions. J. Mine Ventilation Soc. South Africa, 59(April/June): 40-45.

Fujita, E.M. (2001). Hydrocarbon source apportionment for the 1996 Paso del Norte Ozone Study. Sci. Total Environ., 276(1-3):171-184.

Li, W.W.; Cardenas, N.; Walton, J.; Trujillo, D.; Morales, H.; and Arimoto, R. (2005). PM source identification at Sunland Park, New Mexico, using a simple heuristic meteorological and chemical analysis. J. Air Waste Manage. Assoc., 55(3):352-

364.

McCarthy, M.C.; Hafner, H.R.; Raffuse, S.M.; and Stiefer, P.S. (2004). Spatial variability in air toxics -

Supplementary technical memorandum. Report No. STI-903553-2591-STM. Prepared for Lake Michigan Air Directors Consortium, Des Plaines, IL, by Sonoma Technology, Inc., Petaluma, CA.

Mukerjee, S.; Norris, G.A.; Smith, L.A.; Noble, C.A.; Neas, L.M.; Haluk

Ozkaynak, A.; and Gonzales, M. (2004). Receptor model comparisons and wind direction analyses of volatile organic compounds and submicrometer

particles in an arid, binational, urban air shed. Environ. Sci. Technol., 38(8):2317-2327.

Neas, L.; Gonzales, M.; Mukerjee, S.; Smith, L.; and Svendsen, E. (2004). GIS-modeled indicators of mobile source emissions and adverse health effects among children in El Paso, Texas, USA. Epidemiology, 15:S66.

Nussbaum, N.J.; Zhu, D.; Kuhns, H.D.; Mazzoleni, C.; Chang, M.-C.O.; Moosmüller, H.; Chow, J.C.; Watson, J.G. (2009). The In-Plume Emissions Test-Stand: A novel instrument platform for the real-time characterization of combustion emissions. J. Air Waste Manage. Assoc., accepted.

Smith, L.; Mukerjee, S.; Gonzales, M.; Stallings, C.; Neas, L.; Norris, G.; and Ozkaynak, H. (2006). Use of GIS and ancillary variables compound and nitrogen dioxide to predict volatile organic levels at unmonitored locations. Atmos. Environ., 40(20):3773-3787.

Acknowledgements•

Barbara Zielinska and DRI’s Organic Analytical Laboratory for the analysis of carbonyl and canister samples

•

Judith Chow and John Watson for their advice and DRI’s Environmental Analysis Facility for the analysis of filter samples

•

Hamden Kuhn and DRI’s Particulate Emissions Measurement Laboratory for the use of the sampling van and in-plume monitoring system

•

This work was supported by a grant from the U.S. EPA Region 6 under its U.S./Mexico Border Air Quality Program

Thanks