Twenty-four-hour whole-air samples were collected in evacuated stainless steel canisters and analyzed for volatile organic compounds (VOC) at selected sites in southeast Kansas from March 1999 to October 2000. The purpose was to assess the influence on air quality of four industrial facilities that burn hazardous waste located in the communities of Coffeyville, Chanute, Independence and Fredonia. Fifteen of the VOC analytes were found at concentrations above the detection limit and above levels observed in the blanks. Data were analyzed to investigate whether sampling site and date had a significant effect on VOC concentration. Results indicate that site and/or date were significant factors for many of the VOCs. To further investigate the temporal factor, sampling days were divided into four classifications based on wind direction: predominantly north winds, predominantly south winds, calm/variable winds and other winds. Results from statistical analyses show that wind direction was a significant factor for benzene, toluene, o-xylene, naphthalene, and carbon tetrachloride. Data from upwind and downwind samples were analyzed for the four cities of interest in the study area, to investigate the effect of the four targeted sources on VOC concentrations. Results from Fredonia showed higher concentrations of toluene, ethyl benzene, styrene, methyl chloride, and trichloroethylene in the upwind samples, although none of the results were statistically significant. Chanute also showed higher concentrations of the same compounds and m,p-xylene in the upwind samples; results were significant at the 0.05 level for toluene, ethylbenzene, and xylene. These results indicate that sources other than those targeted in the sampling network may be contributing to the VOC levels. Results from Independence showed higher concentrations of ethylebenzene and styrene in the downwind samples; results were statistically significant. These results indicate that the source targeted in the sampling network may be contributing to the VOC levels at those sampling sites.

1. Effects of Wind Direction on VOC
Concentrations in Southeast Kansas
Sergio A. Guerra, Dennis D. Lane, Glen A. Marotz,
Ray E. Carter, Carrie M. Hohl, Richard W. Baldauf
Department of Civil, Environmental, and Architectural
Engineering, University of Kansas

2. Introduction

Southeast Kansas supports the highest concentration of
hazardous waste burners in the country



U.S. EPA sponsored the Southeast Kansas Health Study
to investigate air quality and potential health effects
from ambient air in the area


3 cement kilns
1 commercial hazardous waste incinerator
(report available at http://www2.kumc.edu/ceoh/skhs/)
This study was a joint effort between the Department of
Civil and Environmental Engineering at KU and the KU
Medical Center

3. Scope



The environmental sampling element of
the project included the collection of 24-hr
VOC samples at selected sites in
Southeast Kansas
Effects of spatial and temporal factors on
these concentrations were investigated
Wind direction effects are of particular
interest

4. Air Quality Monitoring




Duration of Monitoring - March, 1999
through October, 2000
Sampling Sites in the cities of Chanute,
Coffeyville, Fredonia, and Independence
Additional Sites in Sedan, Tyro, Labette
County
Sampling Site Design, Sampling and
Analysis Protocols According to EPA
Guidelines

5. Sampling Area in Southeast Kansas

6. Sampling Area in Southeast Kansas
Coffeyville
Chanute
Fredonia
Independence
Background
CC Ag..
House
North Ash
Grove
East
Kennel
East of
River
Labette Co.Big Hill Lake
KDHE
Site
South Ash
Grove
Lincoln
School
Eisenhower
School
Labette Co.Junction160/169
Longfellow
School
Elderly
care
South
Farm
Radio
Tower
Sedan- Power
Pole
Admin.
Building
KDHE
office
South
Mound
Washington
School
Tyro- Water
Tower
Airport

7. Source-specific sites

Wichita windrose
(1984-1992)

8. Fredonia
Site
Sampling Methods
LaF-2
PM10 MiniVol*
LaF-3**
PM10 MiniVol
LaF-5
PM2.5 MiniVol
PM10 MiniVol
**Site was discontinued after March, 2000.

9. Data Collection- Canister with flow
controller and programmable
open/close valve

10. Sample analysis by GC/MS

VOCs analyzed
included:







Benzene
Toluene
Xylene isomers
Trichloroethene
Tetracholoethene
Chloroform
1,1,1trichloroethene

11. Data Analysis


Descriptive statistics were calculated for
the VOC concentrations
Data was analyzed for


Possible effects of sampling site
Possible effects of sampling date

12. Data Analysis

The temporal factor was further investigated to
determine the effect of wind direction on VOC
concentrations

Data was divided in four wind direction categories (from NCDC);





South
North
Calm/variable
Other
The effect of the targeted sources was also analyzed by
using the Wilcoxon signed rank test for the cities of
Chanute, Independence and Fredonia

13. Ambient Air VOC Concentrations
Producing Increased Cancer Risks
Increased Cancer Risks VOC
Concentration (µg/m3) producing risk levels
1 in 10,000
1 in 100,000
1 in 1,000,000
benzene
13
1.3
0.13
Bromoform*
90
9
0.9
carbon tetrachloride
7
0.7
0.07
Chloroform
4
0.4
0.04
1,2-dibromoethanea
0.5
0.05
0.005
1,2-dichloroethanea
4
0.4
0.04
1,1-dichloroethylenea
2
0.2
0.02
Hexachlorobutadienea
5
0.5
0.05
methylene chloride
200
20
2
1,1,1,2-tetrachloroethanea
10
1
0.1
1,1,2,2-tetrachloroethanea
2
0.2
0.02
170
17
1.7
1,1,2-trichloroethanea
6
0.6
0.06
Trichloroethylene
60
6
0.6
Tetrachloroethylene
* - not detected during this study

14. Ambient Air VOC Concentrations
Producing Other Health Effects
Other Health Effects
VOC
Minimum concentration producing health effects (µg/m3)
benzene
60
carbon tetrachloride
40
chlorobenzene
20
chloroethanea
10,000
chloroform
300
chloromethanea
90
dichlorobenzene isomersa
800
1,1-dichloroethanea
500
ethylbenzene
1000
methylene chloride
3000
naphthalene
3
styrene
1000
tetrachloroethylene
300
toluene
400
1,1,1-trichloroethane
1000
trichloroethylene
500
xylene isomers
400

15. Results

16. Distribution of VOC Concentrations,
24-hour samples
VOC
Number of Samples with Concentration in Each Range
0.1-1.0g/L
1.1-10.0g/L
10.1-100.0g/L
>100g/L
Health Effectsa
Benzene
54
6
0
0
58
Toluene
40
113
36
6
5
Ethylbenzene
16
56
20
10
2
m,p-xylene
32
42
4
0
0
o-xylene
36
32
0
0
0
Styrene
9
16
17
2
0
Naphthalene
76
46
1
0
14
Chlorobenzene
5
4
0
0
0
methylene chloride
3
26
8
2
31
Chloroform
12
6
0
0
18
carbon tetrachloride
28
2
0
0
30
1,1,1-TCA
7
1
1
0
0
Trichloroethylene
2
12
19
12
44
Tetrachloroethylene
3
16
2
0
14
Isooctane
1
19
7
0
na

17. VOC observations




Although there are 58 samples with benzene concentrations
exceeding the 1-in-1,000,000 level of increased cancer risk, only 5
of those exceeded the 1-in-100,000 level; none exceeded the 1-in10,000 level.
Other non-halogenated aromatic compounds were detected in a
large number of samples, especially toluene, ethylbenzene, and
naphthalene. However, only in the case of naphthalene were there
more than ten samples with concentrations which could produce
health effects.
There were 31 samples with methylene chloride concentrations
exceeding the 1-in-1,000,000 level of increased cancer risk. Only
eight of those exceeded the 1-in-100,000 level, and only one
exceeded the 1-in-10,000 level.
There were 18 samples with chloroform concentrations exceeding
the 1-in-1,000,000 level of increased cancer risk, thirteen of which
exceeded the 1-in-100,000 level, and two of which exceeded the 1in-10,000 level.

18. VOC observations



There were 30 samples with carbon tetrachloride
concentrations exceeding the 1-in-1,000,000 level of
increased cancer risk, fourteen of which exceeded the 1in-100,000 level; none exceeded the 1-in-10,000 level of
increased risk.
There were 44 samples with trichloroethylene
concentrations exceeding the 1-in-1,000,000 level of
increased cancer risk. Of those, 38 exceeded the 1-in100,000 level and 21 exceeded the 1-in-10,000 level.
There were fourteen samples with tetrachloroethylene
concentrations exceeding the 1-in-1,000,000 level of
increased cancer risk, only one of which the 1-in100,000 level.

19. Effects of sampling site



Sampling site was a statistically significant factor for 8 of
the 15 VOCs.
Site factor was further analyzed using upwind/downwind
sample pairs.
In Chanute and Fredonia, higher concentrations of six
VOCs were found in upwind samples.



Differences were not statistically significant in Fredonia, but
were in Chanute for toluene, ethylbenzene, and xylene.
Results indicate contributions from other sources.
In Independence, significantly higher concentrations of
ethylbenzene, styrene, methylene chloride, and
trichloroethylene were found in downwind samples.

Results indicate contribution from targeted source.

20. Effects of sampling date


Sampling date was a significant factor for 12 of the 15
VOCs
Wind direction accounted for much of the variation
among sampling dates.


Northerly winds frequently produced higher concentrations of
benzene and xylene, and occasionally produced higher
concentrations of carbon tetrachloride.
Southerly winds frequently produced higher concentrations of
toluene, but lower concentrations of naphthalene.

21. Conclusions



Fifteen VOCs were detected above detection limit.
Several VOCs were found at concentrations above published risk
levels, although infrequently.
Trichloroethylene concentrations were of most concern





44 values exceeded the 1 in 1,000,000 risk level
21 of those exceeded the 1 in 10,000 risk level
Statistical analysis showed that sampling site had a significant effect
on the concentrations of many compounds.
Larger than expected highest values were also found at the
Coffeyville-KDHE and Chanute-S. Ash Grove sites.
Sampling date was found to be significant for many of the
compounds.


Wind direction was shown to be a significant factor but not consistently.
For example, southerly winds typically produced higher than expected
conc. of toluene but lower than expected conc. of naphthalene.

22. Conclusions




For four of the compounds, concentrations at sites
downwind from the targeted source in Independence
were significantly higher than concentrations at upwind
sites.
However, in Chanute and Fredonia concentrations were
significantly higher at upwind sites.
During the present study several VOCs were found
above concentrations that could potentially affect human
health, though these levels were infrequent.
It could not be shown conclusively that targeted sources
contributed significantly to these concentrations.