VOLATILISATION PROCESSES IN WASTEWATER TREATMENT PLANTS AS A SOURCE OF POTENTIAL EXPOSURE TO VOCs

VOLATILISATION PROCESSES IN WASTEWATER
TREATMENT PLANTS AS A SOURCE OF POTENTIAL
EXPOSURE TO VOCs
Alexander P. Bianchi*f and Mark S. Varneyf
•Industrial Hygiene Department, Exxon Chemical Co., Cadland Rd, Hythe, Southampton SO45 6NP,
U.K., and tDepartment of Oceanographic Sciences, Oceanography Centre, University of Southampton,
Empress Dock, Southampton SO40 4WH, U.K.
(Received 17 October 1996)
Abstract—The results of a survey estimating volatilisation rates (as gas exchange constants and
flux rates) of a range of hazardous alkane, aromatic, organohalogen, organosulphide, ketone and
alcohol volatile organic compounds (VOCs) from a water bay are presented. More than 73
experimental test runs were earned out over a year under varying seasonal conditions to measure
the simultaneous concentrations of VOCs in air and water phases. The data were processed
employing new advances in ‘surface renewal’ volatilisation models. Based on the experimental
data, estimates of theoretical equilibrium constants, gas exchange constants and flux rates for each
VOC were made. Within the ranges of concentration which may reasonably be encountered by
process workers, the fluxes for a broad range of VOCs ranged from approximately 0.04x 10~8 to
9.0×10″‘ g cm”2 h~’. The results were used to make predictions about the relationship between
volatilisation processes and their implications for occupational hygiene nsk assessment. © 1997
British Occupational Hygiene Society. Published by Elsevier Science Ltd
INTRODUCTION
Emissions of volatile organic compounds (VOCs) from wastewaters in municipal
sewage treatment plants, surface impoundments, waste lagoons, industrial wastewaters
and drainage systems are often overlooked as sources of exposure to
hazardous substances. In many cases, the toxic effects of a broad range of VOCs
arising from such sources may have significant adverse consequences for public
health (Pellizzari, 1982; Namkung and Rittman, 1987; Ciccioli, 1993; Wallace, 1993)
and for wastewater plant workers within an industrial setting (de Mik, 1993). The
range of volatile chemical substances passing through wastewater treatment plants
are often unknown, which creates problems when attempting to measure the VOC
contribution to airborne environments or assess exposure implications for operating
personnel (Deacon, 1977; Dix, 1981; Berrafato and Wadden, 1986; Maugh, 1987;
GESAMP, 1991; Haz. Sub., 1993; de Mik, 1993). Up to 40% of the organic loading
within an industrial or municipal wastewater plant may be volatile (for example
VOCs with boiling point: <350°C; vapour pressure, 0.1-500 mm Hg at 20°C)
depending on the nature of materials flushed to sewer (Knap et al., 1979; Clark,
1986). Ecotoxicological impacts within the immediate receiving environment are
also closely associated with the effects of toxic vapour and liquid-phase VOCs in
effluent (Verscheuren, 1983; Stagg, 1986).
^Author to whom correspondence should be addressed.
437
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
438 A. P. Bianchi and M. S. Varney
Volatilisation has long been recognised as a mechanism whereby organic
compounds with appropriate physicochemical characteristics (that is aqueous
solubility and concentration, vapour pressure and Henry's Law constant) transfer
across the water/air interface from the 'aqueous' compartment to the 'atmosphere'
compartment (Kotzias and Sparta, 1993). Within wastewater and treatment plants,
volatilisation processes have, in the past, been viewed as a legitimate means for
reducing the total organic load at the discharge point. Moreover, these so-called
'evaporative losses' were incorporated into basic models employed to perform crude
estimates of volatile hydrocarbons emissions from oil/water separator bays
(Litchfield, 1971) and of volatile organohalogens from general wastewaters (Dilling
et al., 1975; Dilling, 1977).
Within the last 10-15 years, the limitations of engineering controls over
volatilisation processes have been recognised as an undesirable feature of public and
industrial sector wastewater management (Dix, 1981). In the U.S.A. uncontrolled
emissions were identified as an ongoing problem in municipal water treatment
plants, with unknown consequences for exposure (Pellizzari, 1982). Similar
problems have arisen in British municipal plants with respect to the reduction of
VOC emissions, particularly concerning odour control. For example, despite the
usage of a variety of 'chemical' controls on malodorous sewage streams (including
scrubbing, oxidation and ozonolysis), the release of volatile compounds, including
sulphur-containing organic thiols, polysulphides and hydrogen sulphide persisted as
a source of concern and complaint (Slater and Harling-Brown, 1986).
Selective improvements in industrial emission control have, however, been made
within the last 10-20 years. In the U.S.A., the petrochemical industry has steadily
developed new technology to reduce VOC emissions. In the early design and use of
primary wastewater treatment bays for industrial oil-water separation (for example
the API Separator; Litchfield, 1971), volatile hydrocarbon emissions arising from
wastewater treatment processes were not subject to control until the use of fixedroof
'vapour-encapsulating' structures and closed-loop vents were mandated by the
USEPA (Vincent, 1979). Within the last decade or so, VOC losses to atmosphere
from large-scale wastewater treatment plants have been the subject of heightened
concern with respect to environmental control and public health, especially in the
U.S.A. Today, the uncontrolled release of VOCs (termed 'secondary fugitive
emissions') are increasingly subject to legislative controls by USEPA and other
regulatory bodies (Springer et al., 1986).
Perhaps one of the most challenging problems encountered by air quality
scientists and occupational hygienists is the lack of suitable, consistent and
scientifically enduring models for estimating emission rates of VOCs from
wastewaters to the local airborne environment. Within the field of environmental
control, appropriate models are needed for estimating transfer of VOCs from water
to airborne compartments as part of emission loss assessments. Reliable models are
also needed for performing residence time calculations on anthropogenically-derived
VOCs as part of ecotoxicological risk assessments (Rogers et al., 1992). Accounting
for VOC losses from water-based processes in the manufacturing industry (for
example paint and dyestuffs) is also required by pollution control legislation.
From an occupational health perspective, predicting volatilisation behaviour
allows evaluation of personal exposure to hazardous VOCs, thus enabling adequate
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 439
controls to be established (for example engineering design, use of respiratory
protective equipment). Inevitably, because programmes of continuous air and water
VOCs monitoring in wastewater plants will be limited in practical terms by
constraints on time and resources, it is clearly desirable to have reliable models for
predicting emission rates of toxic VOCs.
Air/water volatilisation processes
In general, volatilisation models describing air/water interface transfer processes
have not enjoyed widespread use by environmental scientists nor occupational
hygienists in the study of environment and related health effects. Various
commentators have suggested that this may be because such models and the
implied processes are considered too theoretical, mathematically complex or
requiring significant amounts of input data which are not readily available (Ciccioli,
1993; Bianchi, 1994).
As with many predictive models which utilise differing input parameters and
incur major assumptions about natural processes, they may yield variable and often
conflicting results (Wadden and Berrafato-Triemer, 1989). Importantly, the status of
volatilisation models has undergone much change since the early 1990s following
major reappraisals of our understanding of physicochemical and meteorological
parameters which control air-water exchange. In the late 1980s the principal models
in common use to describe volatilisation processes from water bodies were of the (1)
'stagnant film', (2) 'surface renewal' and (3) 'turbulent boundary layer' type. Many
of the commoner so-called 'box' variants are in fact variants of stagnant film and
earlier surface renewal models. Although it is beyond the scope of this paper to go
into these in greater detail, the interested reader can find a summary of each in Liss
and Merlivat (1986).
In particular, the stagnant film model, developed and refined by Broecker and
Peng (1974), was widely applied in the U.S.A. Used throughout the 1970s and 1980s
it also formed the basis of many commercial computer packages for VOC flux
estimating. However, in the early 1990s further development and complementary
fieldwork on surface renewal and turbulent boundary layer volatilisation processes
significantly advanced the validity and effectiveness of models describing
volatilisation processes from water (Upstill-Goddard et al., 1990; Watson et al.,
1991). Simultaneously this new work also highlighted fundamental difficulties in the
stagnant film models. Crucially, the stagnant film model did not accurately predict
the correct dependence of gas transfer on molecular diffusivity nor account for
significant non-linear response in the effects of wind-induced turbulence on
volatilisation rates. The extent to which the model deviated from real situations
varied according to meteorological conditions and the physicochemical properties of
the organic compounds under examination. The literature also indicated that
previous calculations of emission rates based on the stagnant film model may
significantly underestimate actual flux levels, creating major implications for the
reliability of environmental databases founded upon it (Nightingale, 1991).
Some of the key findings of Upstill-Goddard et al. (1990) and Watson et al.
(1991) were later examined and validated through extensive field measurements
(Wanninkhof, 1992). Throughout 1992-93, in one of the first field studies intended
to estimate fluxes of low-level VOCs between environmental (that is estuarine and
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
440 A. P. Bianchi and M. S. Varncy
riverine) waters and air under typical environmental conditions, contemporary VOC
emission databases were reprocessed using the new surface renewal models and
found to be up to 10 times higher than predicted by the stagnant film model (Bianchi
and Varney, 1993; Bianchi, 1994).
Given that concentrations of VOCs in wastewater plants are usually much higher
than in marine environments, the consequences for emission rate estimates and their
interpretation in terms of environmental and occupational health are potentially of
greater relevance. Moreover, in this paper we present what we believe to be one of
the first attempts at applying advances in the development of surface renewal and,
where relevant, turbulent boundary layer models to a wastewater treatment bay
under simulated conditions, in the context of an environmental hygiene study.
SURFACE RENEWAL VOLATILISATION PROCESSES—KEY PRINCIPLES
In the study of gas exchange between air and water, the interface between the
two phases is considered as a two-layer (film) system; the main resistance to gas
transport arises from the gas and liquid phase interfacial layers across which
exchanging phases transfer by molecular processes (Liss and Slater, 1974). Since
transfer through the layer system is by molecular diffusion, Fick's first law in the
one-dimensional form (with z as the vertical direction) is applicable, that is
F = —D dc/dz (where F is the flux of gas through the layer, D is the coefficient of
molecular diffusion of gas in the layer material, and c is the gas concentration.
The flux of an organic compound across the water-air interface is a product of
the overall transfer velocity, k, and the extent of the disequilibrium between air and
surface water concentrations (Preston, 1992). The transfer velocity for any given
compound is dependent inter alia on factors such as Henry's Law constant, the
Schmidt number, windspeed and water temperature (Upstill-Goddard et al., 1990;
Watson et al., 1991). Additional parameters which influence the rate of volatilisation
of VOCs across the water-air boundary include aqueous solubility, vapour pressure,
diffusivity, wave action and bubble penetration (Mackay and Yeun, 1983; Liss and
Merlivat, 1986; Wadden and Berrafato-Triemer, 1989).
Some of the most important relationships governing exchange processes can be
summarised as:
F=A:(T)wAC (1)
where AC is the concentration difference driving the flux (F) and &(T)W is the total
transfer velocity (that is the gas exchange constant). The concentration difference is
the difference between the observed aqueous concentration and the calculated
concentration assuming the gas is in equilibrium with the atmosphere and obeys
Henry's law. It can be specifically expressed as:
AC = Ca/T' – Cw (2)
where Ca and Cw are the gas concentrations in air and water, respectively, and H is
the dimensionless and temperature-dependent Henry's Law constant. The total
transfer velocity can therefore be described as:
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatibsation processes in wastewater treatment plants 441
where ka and &w are the individual transfer velocities for chemically unreactive gases
in air and water phases, respectively, and a is a factor which quantifies any
enhancement of gas transfer in the water due to chemical reaction. Equation (3) can
be expressed in terms of resistances as:
^(T)w = rvl + r& (4)
where -R(T)w( = 1/^(T)W) is the total resistance, with rw(= \/akw) and ra(= \/Hka) the
resistances of water and air phases respectively.
By substitution of appropriate values for kw, ka and a in Equation (3) it can be
demonstrated that for many gases either rw or ra is dominant. Gases for which rw
is the dominant resistance to transfer mostly have high Henry's Law constants
(that is low solubility) and a is roughly equal to 1.0 [for example CH4, CO2,
CH3I, (CH3)2S)]. This category includes VOCs such as methane and
dimethylsulphide which are often found in industrial and municipal wastewaters,
and for these compounds, kw is the transfer velocity which controls their air/water
exchange.
In this study, most attention was focused on examining the effects of continuous
air movement over the surface of a body of water releasing VOCs to its local
environment using surface renewal concepts, typical of the environment in which
wastewater plants operate. Models of these transfer processes indicate that kw is
proportional to friction velocities in air (£/*) and also to windspeed (£/)• Similarly,
fcw is also proportional to the ratio of the transfer coefficients for momentum
(kinematic viscosity, v) and mass (molecular diffusivity, D) to the power — 2/3 (that
is fcwtfSc"2'3 at U= <5 m s~' or U*= <0.3 m s"1 (that is, a 'smooth surface'
regime) according to the definition of Liss and Merlivat (1986)). Sc is the Schmidt
number, v/D, a dimensionless ratio which is typically in the range of 0.5-2.0 for
gases and 500-2000 for liquids. A useful feature of the Schmidt number is that it
also expresses temperature dependence, notably for liquids in which case Sc
decreases rapidly with increasing temperature, as diffusivity rises and viscosity falls.
To estimate kw it is assumed the surface is smooth and that continuity of stress
across the interface is attained in order to convert the velocity profile in air to an
equivalent profile in water, that is:
kw = 0.082Sc-2/3(ra/rw)1/2£/* (5)
where ra and rw are the densities of air and water. For this model, A:w is proportional
to D ' . Wind tunnel experiments have shown that the relationship between kw and
Sc is not constant, and that under unsteady-state penetration mass-transfer
conditions, a lower dependence is indicated (that is kw is proportional to Sc"1''2)
for a 'rough surface regime' where the actual value of £/«4-13 m s~', and U* «0.3-
0.7 m s " ' (which represents a considerable increase in the slope of kw versus
windspeed). Here, the Schmidt number gives the transfer velocity (that is gas
exchange constant) as proportional to approximately D05. At windspeeds much
above t / = 1 0 m s ~ ' (that is which corresponds to the breaking-wave 'bubble'
regime associated with high winds over the water surface) gas transfer rates are
considerably enhanced, as confirmed by the application of dual-tracer experiments
in rough water environments (Watson et al., 1991).
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
442 A. P. Bianchi and M. S. Varney
Assuming the gas exchange constant (k(jyw) is strongly and non-linearly
dependent on the wind speed, three relationships which describe such variation
using carbon dioxide as the model gas, are:
A;(T)W = 0.17(7 (where £/< 3.6 ms"1) (6)
&(T)w = 2.85£/-9.65 (where 3.6 < U 13 ms”1) (8)
where fc(T>w is in cm h”1, and U is in m s~\ This model has been successfully used
to describe air-sea exchange fluxes ranging from low energy water bodies to large
scale open-ocean air-sea exchange associated with high wind speed and significant
surface turbulence (Wallace and Wirick, 1992).
EXPERIMENTAL DETAIL AND MONITORING SURVEYS
Volatilisation measurements were carried out over a 14-month period using a
redundant wastewater holding bay in the eastern Southampton dockland area (that
is an outdoor setting) access to which was made available by a commercial marine
engineering company. Meteorological variables and water temperature were
monitored continuously in addition to the aqueous and airborne concentrations
of selected VOCs, representative of those compounds frequently reported in
municipal and industrial waste streams (Aggazzotti and Predieri, 1986; Lawrence
and Foster, 1987; Thomas et al, 1987; Hazard et al, 1991; Rogers et al, 1992).
Samples were collected over 8-h intervals under a variety of weather conditions
spanning spring through to winter. Under circumstances intended to model severe
water contamination, samples were taken over 15-min periods to estimate the
possible consequences for short-term exposure. Average air temperatures ranged
through the periods of sampling from -21° to 32.2°C.
Experimental methods
The wastewater holding bay was an embedded, fined concrete structure of
approximate dimensions 18×10 m with depth 6.5 m, of which a ‘lip’ 1.4 m high
protruded above a ground-level concrete apron. The bay was gated at either end so
as to allow small vessels (and estuarine water from the Itchen sub-estuary) to enter
at one end and leave at the other. Contaminated water (post-experiment) was
rerouted to a dirty water holding bay. The basic layout shared many similarities in
structure and physical dimensions with municipal or industrial wastewater holdingbays
or oil/water separators. Water depth was recorded during each sampling period
allowing total water volume (and where necessary, residence time) to be calculated.
Water temperature was measured using mercury-in-glass thermometers. Meterological
factors such as relative humidity (%) and air temperature (dry bulb) were
measured using a portable thermohygrograph (Casella Ltd, Bedford, U.K.). Wind
velocity profiles were measured at various locations across the bay using a rotatingvane
anemometer (Casella). Meteorological data were checked daily with the
Southampton Weather Centre. Sampling was avoided during times of prolonged
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 443
rainfall due to the difficulties foreseen in quantitatively accounting for its potential
effects (for example temporary dilution of VOCs in surface water, VOCs ‘washout’
effects of wet precipitation).
Water sampling. Replicate water samples (1 1.) were drawn from the top 0.5 m
depth for analysis at 15-min and hourly intervals for storage in sealed, insulated
boxes cooled to 4°C. The majority of water samples were taken from the mid-point
of the bay (approximately 1 m towards the centre) where natural mixing conditions
were found to be relatively stable and free from internal turbulence (as previously
determined by observation of fluoroscein (BDH-Merck, Eastleigh, U.K.) dye
spiking experiments and water flow measurements using a current meter (Vector
Instruments, Oxford, U.K.). Water samples were taken to the laboratory in sealed
glass vessels (with no headspace) and analysed within 12 h of sampling. Analysis for
VOC content was conducted using a dynamic headspace open-loop multi-sorbent
bed ‘purge-and-trap’ method with automated thermal desorption (Perkin-Elmer
ATD-50, Beaconsfield, U.K.) and BP-1 capillary column-GC with simultaneous
FID/mass-spectral detection. A fuller description of the analytical details and
conditions employed, including calibration procedures, method performance and
quality assurance details, are published in Bianchi et al. (1989), Varney and Bianchi
(1990) and Bianchi et al. (1991) and will not be repeated.
Air sampling. Air samples were collected at a height of 2.0±0.5 m above the water
surface using sampling pumps (low-flow Accuhaler 808 model, MDA, Lincolnshire,
Illinois, U.S.A.; and Flo-Lite pumps, MSA, Pittsburgh, PA, U.S.A.) connected to
Perkin-Elmer ATD-50 sampling tubes packed to the Supelco ‘Carbotrap 300
specification’ [that is Carbotrap C (250 mg) 20/40 mesh; Carbotrap B (175 mg)
20140 mesh; Carbosieve S-III (105 mg) 60180 mesh] supplied by Supelco Inc
(Supelco, Bellefonte, PA, U.S.A.). Air sampling was carried out by suspending
sampling pumps on tripods to which short aluminium ‘poles’ were attached so as to
ensure that the sampling tubes were correctly positioned within the body of air
immediately above the water surface. Air samples were taken to coincide with water
sampling events at 15-min (Flo-Lite pump; pump flow rate = 500 ml min ~’) and 8-hr
intervals (MDA Accuhaler pump; pump flow rate = 50 ml min”1). Air sampling
tubes were capped with Swagelok® end-caps and sealed in glass-jars which were
then stored in separate boxes at 4°C for analysis within 12 h of sampling. Analysis
was carried out on the sampling tubes using Perkin-Elmer ATD-50 thermal
desorption and cGC-FID/MS techniques, very similar to methods used for water
samples. Fuller details of the sampling and analytical methods employed here,
including calibration and quality assurance steps used for airborne VOC samples
were based on the outline protocols given in CONCAWE (1986) and HSE (1989) and
further developed for environmental air sampling by Bianchi and Varney (1993).
Spiking experiments
Sampling measurements were carried out to determine the concentrations of
‘background’ VOCs in estuarine water from which the bay was filled, and the
airspace above it. Aqueous solutions of the VOC compounds of interest were
prepared by dissolving the pure compounds (AnalaR and HPLC spectrophoto-
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
444 A. P Bianchi and M. S. Varney
metric grade, Aldrich Ltd, Blandford Forum, U.K.) in water and injecting them into
the bay at a depth of 1 m using a motorised pump and braided steel-mesh hose. For
compounds that were slightly soluble and denser than water, saturated solutions
were prepared, diluted to an appropriate concentration and then injected into the
bay. In further experiments, drums of heavily contaminated water (for example
marine diesel, fuel oil and gasoline-contaminated wastewater) were added in order
to study the variation in volatilisation fluxes arising from the increased levels of
contamination.
RESULTS AND DISCUSSION
A summary of the air and water concentration data for the VOC substances of
interest is shown in Table 1. The data presented in Table 1 are representative of the
‘low concentration’ spiking experiments carried out by adding relatively low-tomoderate
concentrations of each of the VOCs of interest where the concentrated
aqueous solutions added to the bay were typically 5 1. of spike solution
(1000mgl.~’ of each compound), yielding final aqueous concentrations in the
approximate range of 5—15 000 ng l.~’ per component over the experimental
duration (that is JV=73, over 14 months). These concentrations were intended to
represent relatively low levels of contamination found in many typical influent
waters to wastewater bays and their discharge points.
Table 1. Typical air and water concentration ranges of selected VOCs in air (CJ and water (Cw) derived
from wastewater bay spiking experiments
VOC
n-Hexane
n-Dccane
Methylcyclohexane
2,2,4-Tnmethylpentane
Benzene
Toluene
Ethylbenzene
o-Xylene
1,2,4-Tnmethylbenzene
Naphthalene
Carbon tetrachloride
Chloroform
Trichlorofluoromethane
1,1,1 -Trichloroethane
Trichloroethylene
Methyl mercaptan
Dimethyl sulphide
Dimethyl disulphide
Butanone-2
Butanol-2
(Min-Max)*
0.05-33.2
0.07-13.6
0.02-20.2
0.05-24.3
O.KM8.5
0.34-65.7
0.29-60.3
0.36-58.3
0.40-50.0
0.44-28.3
0.23-27 3
0.77-99 3
0.01-20 2
0.75-^7.5
1.21-59.4
0.06-95.3
0.05-99.2
0.10-93.3
0.35-8.5
0.28-7.4
c.
(Mean)f
12.3
8.4
7.6
11.0
25.3
33.7
30.3
28.4
26.7
13.4
15.6
47.0
9.3
25.3
30.3
44.5
55.3
50.3
4.5
3.5
GMJ
6.0
3.3
3.0
5.6
11.9
15.9
14.1
13.3
12.3
6.5
7.3
23.5
4.5
12.2
14.3
20.4
27.5
25.6
2.3
1.8
(Min-Max)*
8.8^400
2.7-2203
3.3-1920
2.1-2504
10 4-8320
17.4-10122
15.2-10001
12.8-11030
10.2-8684
5.3-3686
12 3-6478
93 4-21300
15 6-3647
10.2-8500
20 3-9403
27 3-11 020
18.2-16304
29 3-13000
15.6-6900
18.8-7530
Cw
(Mean)f
1020
964
873
1285
4672
5602
4975
5204
4664
1405
3692
10203
1502
4633
4304
6422
8377
7002
3500
3790
GMJ
945
467
420
640
2202
2829
2704
2474
2645
693
1578
4365
745
2403
2102
3404
4355
3680
1200
1540
The data shown are based on a pool of A’= 73 ‘low concentration’ experiments and relate to 8-h timeweighted
average concentrations. Concentration data are expressed as ng l.~’
‘Arithmetic range shown as Min-Max values obtained.
tAnthmetic mean
JGeometnc mean.
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 445
The highest recorded concentration was for chloroform (Ca = 99.3 ng l.~’,
corresponding to CW = 213OO ng l.~’), a major contribution of which came from the
influent estuarine water used to flood the bay. This is consistent with the results of
Aggazzotti and Predieri (1986) who identified comparable levels of chloroform in
municipal water streams where it also represented the most ubiquitous compound
found in the highest concentration. Elevated levels of chloroform had previously
been reported by Bianchi (1994) and Dawes and Waldock (1994) in the Itchen subestuary
to Southampton Water, mainly as a by-product of water chlorination and
industrial processes. Organosulphide levels in ‘background’ water (and in air
samples) were also higher than anticipated. Up to 25% of the total amount of
methyl mercaptan, dimethylsulphide (DMS) and dimethyldisulphide (DMDS)
measured in the bay were also derived from surface water in the Itchen. Water
quality varies significantly within the estuary (Soulsby et al., 1985), highlighting the
need to account for extranneous sources of VOCs. The sources of the
organosulphides were traced to a nearby sewage outfall and unrelated biological
decay processes of plankton such as the photosynthetic ciliate {Mesodinium rubrum).
Among the alkylbenzenes, for example, toluene was also found in the highest
concentrations, particularly in air. The lowest airborne concentrations recorded
were for butanone-2 and butanol-2, which may be expected given their hydrophilic,
polar nature.
Theoretical equilibrium concentrations
Given the analytical data set made available by the air and water concentration
surveys, it was possible to estimate concentration differences (AC) across the airwater
interface. By using the theoretical Henry’s Law constants for each compound
(Equation 2), theoretical equilibrium concentrations were calculated (that is the
aqueous concentration in theoretical equilibrium with the concentration measured
in air) using the Henry’s constant protocol described by Nightingale (1991). A
summary of the theoretical equilibrium concentration values are shown in Table 2.
For all compounds studied, the actual aqueous concentration significantly exceeded
the theoretical air-water equilibrium concentration predicted by achievement of
interphase equilibrium, showing that, despite the relatively low range of aqueous
concentrations, the water was supersaturated with respect to atmospheric transfer.
Thus, in terms of Equation (2), Cw is mainly >>CaH~l. At these aqueous
concentrations, VOC movement would therefore be highly unidirectional (that is
from water to air), with the atmospheric ‘compartment’ representing the major sink.
This is an important finding, since under different environmental conditions the
direction of transfer may be reversed (that is from air to water). For example it was
previously shown by Bianchi and Varney (1993) that under certain conditions
following episodes of elevated airborne pollution, vapour-phase toluene may cross
the ‘air-water’ interface, representing a source of toluene to surface waters (that is
where CJJ~’ > Cw.
VOC flux calculations
Fluxes were derived by calculating the respective Co and Cw data for each
component, allowing for maximum and minimum recorded wind speed across the
surface of the water body. Using this data, gas exchange constants kmv/) were
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
446 A. P Bianchi and M. S. Vamey
Table 2. Equilibrium concentration values for representative air and water VOC concentration data
voc
n-Hexane
n-Decane
Methylcyclohexane
2,2,4-Trimethylpentane
Benzene
Toluene
Ethylbenzene
o-Xylene
1,2,4-Trimethylbenzene
Naphthalene
Carbon tetrachloride
Chloroform
Trichlorofluoromethane
1,1,1 -Trichloroethane
Trichloroethylenc
Methyl mercaptan
Dimethylsulphide
DimethyWisulphide
Butanone-2
Butanol-2
Estimated equilibrium
concentration
0.26
0.03
0.66
0.12
210.8
187.2
121.2
118.3
133.5
1595.2
14.4
522.2
1.86
105.4
101.0
111.3
1843.3
251.5
225.0
3500.0
Measured aqueous
concentration*
1020
964
873
1285
4672
5602
4975
5204
4664
1405
3692
10232
1502
4633
4304
6422
8377
7002
350
379
Calculated Henry’s
Law constantf
47
252
11.4
91.2
0.12
0.18
0.25
0.24
0.20
0.0084
1.08
0.09
5
0.24
0.3
0.4
0.3
0.2
0.02
0.01
Values are estimated from arithmetic mean data for air ( O and water (Cw) for the data set ( # = 73).
Concentration data are expressed as ng I.”1.
‘Measured aqueous concentration ng I.”1.
fHenry’s Law constants calculated according to Liss and Slater (1974), Sauer (1978) and Nightingale
(1991) corrected for temperature and pressure.
calculated for each component under the prevailing conditions. As these values
relate to the model gas CO2, a correction must be applied for other volatile
compounds to compensate for differences in molecular diffusivities. There are two
ways of accomplishing this. Firstly, by using the ratio of the square roots of the
molecular weights of CO2 and the organic compound, as detailed by Liss and Slater
(1974) and Nightingale (1991); or secondly, by examining the variation of the gas
exchange constant with both the compound of interest and temperature, which can
be described by a power dependence on the Schmidt number as Ki/K2= (Sc^ScO”
(Watson et al. (1991). For a windspeed of more than 3.6 m s ~ ‘ , the power
dependence (n) is approximately 1/2. As there are few predetermined values
available for molecular diffusivities, the molecular mass correction was applied to
values of A^-^ in this body of work.
Flux transfer rates were significantly a function of wind speed and VOC
concentration differences between air and water phases. In the Southampton Dock
area, wind direction is usually from the south-west for more than 75% of the year,
and normally exceeds 2.3 m s~’ for more than 80% of the time (Bianchi, 1994).
During winter months (that is which we classed as December through to March),
wind speeds during sampling exercises ranged from 3.6-13.9 m s~’. In summer
months (that is classed as May to September) wind speeds were much lower, from
1.9-4.8 m s”1.
Within the ranges of concentration tabulated in Table 1, fluxes for a broad range
of VOCs ranged from approximately 0.04xl0~8 to 9.0xl0~8 g cm”2 h~~’. A range
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 447
Table 3. Flux estimates for typical winter (5/2/95) and summer (23/7/95) weather conditions
voc
n-Hexane
n-Decane
Methylcyclohexane
2,2,4-Trimethylpentane
Benzene
Toluene
Ethylbenzene
o-Xylene
1,2,4-Trimethylbenzene
Naphthalene
Carbon tetrachlonde
Chloroform
Trichlorofluoromethane
1,1,1 -Trichloroethane
TricbJoroethylene
Methyl mercaptan
Dimethybulphide
DimethyWisulphide
Butanone-2
Butanol-2
c.
18.9
11.4
17.1
20.1
41.9
60.7
58.3
53.7
45.6
20.3
15.1
80.0
15.1
36.9
445
58.3
69.7
67.4
4.2
4.9
(5/2/95)’
Qv
3295
2105
1763
2209
7095
8654
7920
7562
6827
3057
4929
15300
3005
6250
8133
8745
14321
11450
5880
5956
10.21
7.95
9.58
8.87
10.34
9.87
8.79
9.19
8.64
8.36
7.64
8.63
8.08
8.20
8.27
13.66
12.03
9.76
11.16
11.01
FluxJ
336.4
167.3
168.8
195.9
692.8
820.9
675.7
674.4
570.2
76.2
375.4
1243.8
242.6
499.9
660.4
1181.5
1694.9
1084.6
632.8
116.3
c.
9.8
4.7
5.0
6.1
15.0
22.3
23.5
20 4
19.3
111
120
39.3
8.1
20.0
24 3
30.1
50.2
42.3
2.1
1.8
(23/7/95)t
560
420
440
620
2150
2720
2000
1950
1920
1700
1495
5500
655
2350
2090
3605
4107
3595
1190
1904
K 0.3 m s”1)
and U* = (6.1 + O.63t/iO)o5t/,o.
(where U\Q is the wind speed at a nominal height of 10 m).
Benzene, for example has a ScL= 1021 (at 20cC). By employing a typical value of
0.42 (of [/*) for winter, a flux rate of 16.5 cm h”1 is obtained. The ScL value
increases with increasing alkylation of the benzene ring (for example ScL
toluene = 1155) and hence flux rates are progressively lower for C3- and C4-
alkylbenzenes. ScL increases with decreasing water temperature and hence broad
estimates can be made of the variation of flux rate with water temperature. The
precise quantitative relationship between Schmidt number, temperature and related
environmental variables is not yet fully understood but it can be shown, for example
that if ScL increases by 2.8% per degree Centigrade decrease, the Schmidt number
for benzene at typical winter temperatures of about 70°C is predicted to be about
1650. At this theoretical value, the magnitude of Ktjy* (benzene) decreases from
16.5 cm h”1 (at 20°C) to 13.0 cm h”1.
Individual flux rates are also highly dependent on the absolute concentration of
contaminant in the water phase. This may be illustrated by comparing net fluxes
from water bodies which contain moderate concentrations of VOCs with those from
relatively pristine water bodies such as open sea water. For example, Nightingale
(1991) performed identical estimates for volatile organohalogens in the southern
North Sea, and estimated a net daily flux of 0.1 lx 10~10 g cm”2 h~’ for
chloroform. By way of comparison, the net daily fluxes obtained for chloroform
in our study were up to 1244xlO~10 g cm”2 h~\ between 150 and 12000 times
higher for summer and winter months respectively, highlighting the significant
differences in rate and total mass of transfer which is theoretically possible for the
same compound.
In terms of occupational health and environmental issues, these results also
suggest that volatilisation processes within wastewater bays may, under conditions
determined by AC=Ca//~’-Cw, their respective concentrations, ^mw, and air and
water temperatures, represent important factors which may determine personal
exposure to individual (or total) VOCs. For example, crude estimates based on the
experimental data from this study suggest that where ECW>5 mg I.”1, (as total
VOCs) in water (that is levels commonly encountered in wastewater plants), £Ca
may reasonably be expected to reach or exceed 50 mg m~3 (as total VOCs) in air
within the activity zone of workers, under moderate wind speed conditions (for
example 3-6 m s~’). More precise estimates of exposure to individual compounds
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 449
should be readily available when performing finer estimates. In theory, it should also
be possible for occupational hygiene personnel to make sufficiently useful
predictions of likely airborne concentrations based on measurements (or good
estimates) of aqueous VOC concentration and windspeed alone.
Contamination (spill) events
Importantly, increased loadings of contaminant material (including hydrocarbon
or solvent-based fluids) are relatively common occurrences in municipal and
industrial wastewater plants and can take place at any time, particularly after a spill
event. Such circumstances may arise after a process plant upset or following a civil,
traffic-related or industrial incident (for example fire, emergency release). Depending
on the nature of the VOCs in the influent water and prevailing weather conditions,
these factors will have a major impact on volatilisation rates. Further experiments
were carried out by injecting large volumes (200-300 1.) of contaminated effluent
water from local shipping activities (for example marine fuel oil, gasoline and dieselcontaining
ballast water) into the bay from large steel drums using drum pumps.
Sampling was carried out within 15-min time ‘windows’ within 5 min of charging
the bay, and for longer periods of time extending up to 8 h. Representative results of
a sample ‘injection’ of a mixture of ballast water severely contaminated by kerosine
and raw gasoline are shown in Table 4.
The results indicate that at relatively moderate values of K^-^ (that is 2.3-
3.7 cm h~’) proportionally high flux rates can be achieved ‘driven’ mainly to the
significant gradient (high AC) between air and water-phase concentrations (in
AC=CaH~i — Cw), especially in the first 15 min. Moreover, as the results suggest,
personnel working within the immediate zone would probably require respiratory
protection or limited work periods to prevent or reduce the risk of high single- and
Table 4. Flux estimates for a simulated contamination event following discharge of wastewater contaminated
by petroleum-fluids
VOC
n-Hexane
n-Decane
Benzene
Toluene
Ethylbenzene
o-Xylene
1,2,4-Trimethylbenzene
Naphthalene
c.t
111
52
156
242
219
215
127
7
(15-min
cw§
124
183
137
106
105
104
84
48
period)*
%)«
2.86
2.26
2.91
2 78
2.50
2.58
2.42
2.36
Fluxi
35
41
40
43
26
27
26
11
c.t
2.0
0.9
3.9
4.7
3.5
4.1
0.9
0.1
(8-h
cw§
3.5
4.7
5.1
6.7
1.5
2.1
0.7
0.3
period)t
^(T)w
3.67
2.90
3.72
3.57
3.21
3.31
3.10
3.03
Fluxi
1.3
1.4
1.9
2.4
0.5
0.7
0.5
0.1
Air and water concentration ranges of principal VOC are shown for air (C.) and water (Cw)
concentrations derived from sampling and analysis. Experimental data are based on 15-min and 8-h
sampling periods. K<x)w are expressed a s c m h " 1 corrected according to mean wind speed. Concentration
data are expressed as ng 1."'.
*Mean wind speed (15-min period) = 4.8 m s ~ \ mean air temp 15-min= 17.5°C, mean water
temperature = 10.2°C.
fMean wind speed (8-h) = 5.2 m s~', mean air temp (8-h) = 18.2°C, mean water temperature = 10.7°C.
JC« expressed as x 103 ng I."1.
§CW expressed as xlO6 ngl."1 .
> CaH~l). For all alkylbenzenes, values of
^(T)w reduce with increasing alkylation of the benzene ring. Given the relatively
greater chronic health risks associated with benzene compared to its alkylbenzene
counterparts, this feature would need to be adequately addressed during exposure
risk assessment.
Among the volatile organosulphides, values of K<j)w are highest for most low
molecular compounds with high vapour pressures (for example methyl mercaptan,
DMS). In particular, these factors may help to explain the labile nature of highly
malodorous methyl mercaptan (including ethyl and propyl mercaptans) within
municipal sewage works and waste transfer and recycling stations where these
compounds are frequently found, enhancing the probability of public odour
nuisance complaints (Slater and Harling-Brown, 1986; Bianchi, 1994).
Taken together, the foregoing discussion points suggest that a valid exposure
risk assessment would require the occupational hygienist to obtain a sound database
of aqueous concentration data in addition to airborne data, the latter activity being
the 'normal' premise of the practising hygienist. From our experience in assessing
volatilisation behaviour in open waters, wastewater lagoons, waste treatment plants
and industrial oil-water separators, it seems a valid and achievable precaution to
develop local predictive models which describe volatilisation fluxes and anticipated
airborne concentrations of commonly encountered VOCs under a range of
operating conditions regarded as 'routine' for the plant in question. Furthermore,
these results reinforce the necessity in considering the use of controlled
encapsulation, venting and filtration systems as a management strategy for VOCs
in wastewaters, as opposed to a dilute-and-disperse approach which calculations
indicate may exacerbate an already existing inhalation risk or enhance a nuisance
odour risk for certain types of VOCs.
Research carried out as part of this and earlier studies indicates that these
volatilisation models should also usefully apply to indoor water bays, tanks or
receptacles as much as they do to outdoor water bodies. New work on indoor paint
spraying tasks (unpublished at the time of writing) indicates that the surface renewal
model predicts airborne concentrations of VOCs (at levels of approximately 20-
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 451
500 mg m~3 +25%) associated with 'wet collection' drainage sluices used in
manual spraying activities adjacent to spray booths.
Perhaps one of the main factors which detracts from the validity of the model
lies in the existence of competing 'sinks' for VOCs. For example, the presence of
high concentrations of waterborne suspended solids in treatment plant bays
(inorganic flocculant, sand particles, organic sewage particles) may act as selective
adsorbent sites for hydrophobic VOCs, reducing the solute fugacity or partial
pressure, and so reducing the volatilisation rate. Adsorption to sediments results in
removal from the liquid phase and incorporation into a solid phase. Exchange may
also take place with and between resuspended particles at varying rates, depending
upon a variety of parameters. However, in a series of field measurements and
modelling experiments examining the extent of adsorption within water treatment
plant effluents, Bianchi (1994) demonstrated that even under conditions of high
suspended solids (that is 1200 mg I."1) with correspondingly high fractional organic
contents (that is approaching 100%), the maximum amount of VOCs removed by
adsorption was 10% (of total mass) for chloroform, and 5% (of total mass) for most
volatile aromatics (for example including benzene, toluene and C2-alkylbenzenes)
and organohalogens (for example including chloroform and carbon tetrachloride),
where actual aqueous concentrations were between 1 and 10 //g I."1. It is therefore
unlikely that adsorption would represent an important competing sink nor
significantly reduce volatilisation rates under conditions in which inhalation
exposure represented a potential risk.
Photo-oxidation processes have also been considered as a potential 'sink' for
VOCs. It is, however, unlikely that photo-oxidation reactions would occur at fast
enough rates to remove VOCs. Many of the final photo-oxidised products of
anthropogenic volatile hydrocarbons are aldehydes (Grosjean et al., 1978; Howard
et al., 1991). In particular some alkanes and alkenes oxidise to saturated aldehydes
(Cox et al., 1980), none of which were identified during this study.
SUMMARY AND CONCLUSIONS
This study has provided new and quantitative experimental data describing the
variation in gas exchange constants and water-to-air fluxes for a broad range of
hazardous VOCs commonly encountered in municipal and industrial wastewaters
under varying environmental conditions. The data were derived using a relatively
new volatilisation model which representatively accounts for the effect of
physicochemical characteristics and meteorological factors exerting major controlling
influences on volatilisation behaviour. With hindsight, the value of obtaining a
broad experimental data set over a year under changing climate conditions and in an
environment which closely resembled a real wastewater bay (as opposed to
laboratory simulations) cannot be underestimated. We believe that, despite potential
gaps in the experimental protocol, the data provide a useful platform from which to
examine the relationship between gas transfer velocities and wind speed as predicted
by Upstill-Goddard et al. (1990), Watson et al. (1991) and Wanninkhof (1992).
Further confirmation or comparison of the experimental findings was made difficult
by the apparent lack of data in the literature concerning similar studies of this
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
452 A. P. Bianchi and M. S. Varney
nature, which may indicate that this topic remains a new and still emerging field in
the discipline of occupational hygiene.
For all compounds we studied, the aqueous concentrations used to model typical
wastewater bay conditions significantly exceeded the theoretical air-water
equilibrium concentration predicted by achievement of interphase equilibrium.
Despite what we considered to be quite low absolute aqueous concentrations of
VOCs, it seems likely that most wastewaters will be supersaturated with respect to
atmospheric transfer. In most cases, aqueous concentrations of VOCs in wastewater
bays would probably exceed their theoretical air-water equilibrium concentrations.
The total set of VOCs data, including concentration differences (AC) between air
(Ca) and water (Cw) phases, gas exchange constants (kmv/), and fluxes were
calculated using the most up-to-date 'surface renewal' equations. To the best of our
knowledge, these models have been proven to be reliable although new work is
needed to examine their applicability to occupational environments and in particular
to a wide range of waterborne compounds used by industry. Within the ranges of
concentration commonly encountered in this study, fluxes for most VOCs ranged
from approximately 0.04 x 10~8 to 9.0×10~8 gem"2 h~', the highest values usually
being observed with alkylbenzene and organosulphide compounds. Furthermore,
given the higher concentrations of VOCs in wastewaters during winter associated
with greater windspeeds immediately above the air-water interphase, values of Kmw
were 20—30 times higher than in summer. These findings reinforce general
conclusions that programmes designed to minimise VOC emissions to air must
account for mechanisms which enhance volatilisation and incorporate precautionary
strategies around organic vapour control.
The results of our basic experiments intended to represent 'contamination spill
events' indicated that at relatively moderate values of Kfjy,, (that is 2.3-3.7 cm h~')
high flux rates are observed due to the significant gradient (high AC) between air
and water-phase concentrations (that is where AC=CaH~1 — Cw), especially in the
first 15min. Moreover, as our results highlight, personnel working within the
immediate vicinity may be at risk of higher levels of exposure than anticipated
during such occurrences.
We believe that these data should serve as a useful platform from which to
encourage the use of volatilisation modelling by occupational hygienists carrying
out risk assessments in a variety of wastewater treatment operations. Through
appropriate application of the theoretical models described, reasonable predictions
and estimates can be made of potential VOC exposure scenarios to workplace
personnel employed in wastewater storage and treatment activities.
Acknowledgements—The authors would like to thank Mr J. Leach and Mr A. Chidwick (Condive Marine
Co., Southampton) for access to and use of the water bay utilised for the field experiments, for assistance
in operating the equipment and arranging supply of the contaminated wastewater materials. The authors
would also like to thank Prof. P. Liss and Dr P Nightingale (School of Environmental Sciences,
University of East Anglia) for helpful comments on the use of the volatilisation models.
REFERENCES
Aggazzotti, G. and Predien, G. (1986) Survey of volatile organics (VHO) in Italy: levels of VHO in
drinking waters, surface waters and swimming pools. Water Research 20, 959-963.
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
Volatilisation processes in wastewater treatment plants 453
Berrafato, L. R. and Wadden, R. A. (1986) Predicted versus measured air emissions of volatile organics
from a simulated hazardous waste lagoon. In Toxic and Hazardous Wastes, ed. G. D. Boardman,
Proceedings of the 18th Mid-Atlantic Industrial Waste Conference, Lancaster, Pennsylvania, U.S.A.
Bianchi, A. P. (1994) Studies of volatile organic compounds in the water and sediments of a British
estuary—Southampton Water. PhD thesis, University of Southampton.
Bianchi, A. P. and Varney, M. S. (1993) Sampling and analysis of volatile organic compounds in estuarine
air by gas chromatography and mass spectrometry. Journal of Chromatography 643, 11-23.
Bianchi, A. P., Varney, M. S. and Phillips, A. J. (1989) A modified analytical technique for the
determination of trace organics in water using dynamic headspace and GC-MS. Journal of
Chromatography 467, 111-128.
Bianchi, A. P., Varney, M. S. and Phillips, A. J. (1991) Analysis of industrial solvent mixtures in
wastewater using a miniature purge-and-trap stripping method. Journal of Chromatography 557II,
429-439.
Broeclcer, W. S. and Peng, T. H. (1974) Gas exchange rates between air and sea. Tellus 26, 2135.
Ciccioli, P. (1993). VOCs and air pollution. In Chemistry and Analysis of Volatile Organic Compounds in
the Environment, eds H J. Th. Bloemen and J. Burn, pp. 92-165. Blackie Academic Press, London.
Clark, R. B. (1986) Marine Pollution. Clarendon Press, London.
CONCAWE (1986) Analysis of gasoline hydrocarbons (Report No. 8/86), Oil Companies Study Group
for the Conservation of Clean Air and Water, The Hague.
Cox, R. A., Derwent, R. G. and Williams, M. R. (1980) Atmospheric photooxidation reactions Rates,
rectivity and mechanisms for reaction of organic compounds with hydroxyl radicals Environmental
Science and Technology 14, 57—61
Dawes, V. J. and Waldock, M. J. (1994) Measurement of volatile organic compounds at UK National
Monitoring Plan Stations. Marine Pollution Bulletin 28(5), 291-298.
Deacon, E. L. (1977) Gas transfer to and across an air-water interface. Tellus 29, 363—374
Dilling, W. L. (1977) Interphase transfer processes. II: Evaporation rates of chloromethanes, ethanes,
ethylenes, propanes and propylenes from dilute aqueous solutions—comparisons with theoretical
predictions. Environmental Science and Technology 11, 408-409.
Dilling, W. L., Tefertiller, N. B. and Kallos, G. J. (1975) Evaporation rates and reactivities of methylene
chloride chloroform, 1,1,1-trichloroethane, trichloroethylene, tetrachloroethylene and other chlorinated
compounds in dilute solutions. Environmental Science and Technology 9, 833—838.
Dix, H. M. (1981) Waste disposal and water pollution. In Environmental Pollution. Wiley and Sons,
Chichester, U.K.
GESAMP (1991) Global changes and the air-sea exchange of chemicals. GESAMP reports and studies
No. 48.
Grosjean, D., Van Cauwenberghe, K., Fntts, D. R. and Pitts, J. N. (1978) Photo-oxidation products of
toluene—NO mixtures under simulated atmospheric conditions. Preprint of the 175th National
Meeting, Div. Environ. Chem., ACS, 12-17 March 1978. Am. Chem. Soc. Div. Environ. Chem. Prep
18, 354-356.
Gschwend, P. M., Zafiriou, O. C, Mantoura, R. F. C , Schwarzenbach, R. P. and Gagosian, R. B (1982)
Volatile organic compounds at a coastal site 1. Seasonal variations.. Environmental Science and
Technology 16, 31-38.
Hazard, S. A., Brown, J. L. and Betz, W. R. (1991) Extraction and analysis of hydrocarbons associated
with leaking underground storage tanks. Lc-Gc International 4, 29-35.
Haz. Sub. (Hazardous Substances) (1993) Sources of VOCs. 4, Issue 1, Monitor Press, Sudbury, U.K.
HSE (1989) Mixed hydrocarbons (C5 to C10) in air. Laboratory method using porous polymer diffusion
samplers, thermal desorption and gas chromatography. Methods for the Determination of Hazardous
Substances, MDHS66 Occupational Medicine and Hygiene Laboratory, Health and safety executive,
London.
Howard, P. H., Boethling, R S., Jams, W. F., Meylan, W. M. and Michalenko, E. M. (1991) Handbook
of Environmental Degradation Rates, Vols I, II and III. Lewis Publishers. Michigan, U.S.A.
Kotzias, D. and Sparta, C. (1993) VOCs and water pollution. In Chemistry and Analysis of Volatile
Organic Compounds in the Environment, eds H. J. M. Bloemen and J. Burn, pp. 175-199. Blackie
Academic Press, London.
Knap, A. H., LeB Williams, P. J. and Tyler, I. (1979) Contributions of volatile petroleum hydrocarbons
to the organic carbon budget of an estuary. Nature 279, 517-519.
Lawrence, A. R. and Foster, S. S. D (1987) The pollution threat from agricultural pesticides and
industrial solvents. Hydrogeological report of the British Geological Survey, No. 8712. Walhngford,
Oxfordshire, pp. 1-30.
Leaderer, B. P. (1982) Air pollutant emissions from kerosine space heaters. Science 218, 1113-1115.
Liss, P. S. and Slater, P. G. (1974) Flux of gases across the air-sea interface. Nature 247, 181-184.
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012
454 A. P. Bianchi and M. S. Varney
Liss, P. S. and Merlivat, L. (1986) Air-sea exchange rates:introduction and synthesis. In The Role of Airsea
Exchange in Geochemical Cycling, ed. P. Buat-Menard, pp. 113-127. Reidel, Dordrecht.
Litchfield, D. K. (1971) Controlling odours and vapours from API Separators. Oil and Gas Journal 69,
60-62.
Maugh, T. H. (1987) Chemicals—how many are there? Science 199, 162.
Mackay, D. and Yeun, A. T. K. (1983) Mass transfer coefficient correlations for volatilisation of organic
solutes from water. Environmental Science and Technology 17(4), 211-217.
de Mik, G. (1993) VOCs and occupational health. In Chemistry and Analysis of Volatile Organic
Compounds in the Environment, eds H. J. Th. Bloemen and J. Bum, pp. 268-279 Blackie Academic
PTCSS, London.
Namkung, E. and Rittman, B. E. (1987) Estimating volatile organic compounds from publicly owned
treatment works. Journal of the Water Pollution Control Association 59, 670-678.
Nightingale, P. D. (1991) Low molecular weight halocarbons in sea water. Ph.D thesis, University of East
Anglia.
Pellizzari, E. D. (1982) Volatile organics in aeration gases at municipal treatment plants. USEPA
Washington DC. EPA Report-600/2-82-056. PB82-227760.
Preston, M. R. (1992) The interchange of pollutants between the atmosphere and the oceans. Marine
Pollution Bulletin 24(10), 477^t83.
Rogers, H. R., Crathorne, B. and Watts, C. D. (1992) Sources and fifties of organic contaminants in the
Mersey estuary. Marine Pollution Bulletin 24(2), 82-91.
Sauer, T. C. (1978) Volatile liquid hydrocarbons in the marine environment. Ph.D. thesis, Texas A&M
University.
Slater, P. G. and Harling-Brown, L. (1986) Sampling and GC analysis of volatile sulphur compounds and
gases at sub v.p.m. levels in the presence of ozone. The Analyst (London) 111, 1059-1064.
Soulsby, P. G., Lowthion, D. and Houston, M., (1985) The identification and evaluation of
environmental quality in Southampton Water, UK, using limited resources. In Esluarine Management
and Quality Assessment, eds W. Halcrow and J. G Wilson. Plenum Press, New York.
Springer, C, Thibodeaux, L. J., Lunney, P. D. and Valsaraj, K. T. (1986) Emission of hazardous
chemicals from surface and near surface impoundments to air. Final Project Report Proj. No: 808161-
02: USEPA Solid and Hazardous Waste Research Div., Cincinnatti, OH, U.S.A.
Stagg, R. M. (1986) Physiological effects of toxicants on aquatic animals. In Environmental Toxicology
and Ecotoxicology. Proceedings of the 3rd International Course (W.H.O./I.L.O./U.N.E.P.),
Copenhagen.
Thomas, W. M., Lewis, W. K., Barron, R. J., Cooke, J. E. and Hair, D. (1987) Groundwater
contamination by trace organics—recent experiences in the UK.. Paper presented at the
Hydrogeological Group of the Brit. Geological Soc. (Geological Survey Reports, BGS, Keyworth,
Nottingham, U.K.).
Upstill-Goddard, R. C, Watson, A. J., Liss, P. S. and Liddicoat, M. I. (1990) Gas transfer velocities in
lakes measured with SF6. Tellus 42B, 364-377.
Varney, M. S. and Bianchi, A. P. (1990) Analysis of volatile organic micropollutants in the marine
environment using purge-and-trap methods with automated thermal desorption and capillary column
gas-chromatography/mass-spectroscopy. Paper presented at the International Environment and Safety
Conference, 10-11 January 1990. New Connaught Rooms, London, U.K.
Verscheuren, K. (1983) Handbook of Environmental Data on Organic Chemicals, 2nd Edn. Van Nostrand
Reinhold, New York.
Vincent, R. (1979) Control of organic gas emissions from refinery oil-water separators. California Air
Resource Board Report, Sacramento, U.S.A.
Wadden, R. A. and Berrafato-Triemer, L. R. (1989) Mechanisms of atmospheric release of chlorobenzene
and toluene from a storage lagoon under ambient conditions. Environmental Science and Technology
231, 1492-1498.
Wallace, L. A. (1993) VOCs and the environment and public health-exposure. In Chemistry and Analysis
of Volatile Organic Compounds in the Environment, eds H. J. Th. Bloemen and J. Burn), pp. 1-17.
Blackie Academic Press, London.
Wallace, D. W. R. and Wirick, C. D. (1992) Large air-sea gas fluxes associated with breaking waves.
Nature 356, 694-696.
Wanninkhof, R. (1992) Relationship between wind speed and gas exchange over the ocean. Journal of
Geophysical Research 97C5, 7373-7382.
Watson, A. J., Upstill-Goddard, R. C. and Liss, P. S. (1991) Air-sea gas exchange in rough and stormy
seas measured by a dual tracer technique. Nature 349, 145-147.
Downloaded from http://annhyg.oxfordjournals.org/ by guest on May 11, 2012

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: