Subscriber access provided by Arizona State University Libraries
Article
Velocity dependent passive sampling for monitoring of micropollutants in dynamic stormwater discharges Heidi Birch, Anitha Kumari Sharma, Luca Vezzaro, Hans-Christian Holten Lützhøft, and Peter Steen Mikkelsen Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/es403129j • Publication Date (Web): 15 Oct 2013 Downloaded from http://pubs.acs.org on October 20, 2013
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 22
Environmental Science & Technology
1
Velocity dependent passive sampling for monitoring of
2
micropollutants in dynamic stormwater discharges
3
Heidi Birch, Anitha K. Sharma, Luca Vezzaro, Hans-Christian H. Lützhøft and Peter S. Mikkelsen*
4
Department of Environmental Engineering (DTU Environment), Technical University of Denmark
5
(DTU), Miljoevej, Building 113, 2800 Lyngby, Denmark
6
*Corresponding author, e-mail
[email protected], +45 4525 1600.
7
Micropollutant monitoring in stormwater discharges is challenging because of the diversity of sources
8
and thus large number of pollutants found in stormwater. This is further complicated by the dynamics in
9
runoff flows and the large number of discharge points. Most passive samplers are non-ideal for
10
sampling such systems because they sample in a time-integrative manner. This paper reports test of a
11
flow-through passive sampler, deployed in stormwater runoff at the outlet of a residential-industrial
12
catchment. Momentum from the water velocity during runoff events created flow through the sampler
13
resulting in velocity dependent sampling. This approach enables the integrative sampling of stormwater
14
runoff during periods of weeks to months while weighting actual runoff events higher than no flow
15
periods. Results were comparable to results from volume-proportional samples and results obtained
16
from using a dynamic stormwater quality model (DSQM). The paper illustrates how velocity-dependent
17
flow-through passive sampling may revolutionize the way stormwater discharges are monitored. It also
18
opens the possibility to monitor a larger range of discharge sites over longer time periods instead of
19
focusing on single sites and single events, and it shows how this may be combined with DSQMs to
20
interpret results and estimate loads over extended time periods.
ACS Paragon Plus Environment
1
Environmental Science & Technology
21
Page 2 of 22
For Table of Contents Only
22 23
INTRODUCTION
24
Stormwater runoff from residential and industrial areas contains a wide range of pollutants,1-3 which
25
can cause adverse effects in receiving waters and prevent these from obtaining a good chemical status as
26
e.g. required by the European Water Framework Directive4 as well as by the US Clean Water Act.5
27
However, the concentrations of pollutants in stormwater are variable from site to site and from event to
28
event.6 In order to justify the need for treatment at a particular site and to evaluate the selected
29
treatment, monitoring of stormwater discharges is important.7 Traditionally, this is achieved by using
30
autosamplers in a volume-proportional manner (a fixed volume of sample is collected at a defined
31
runoff volume interval) or flow-proportional manner (a flow weighted volume is collected at a defined
32
time interval). These campaigns are usually very costly and involve many practical difficulties.8
33
Stormwater sampling campaigns are thus often non-ideal, for example by omitting periods with extreme
34
weather (where sampling is difficult) or by focusing on spot-sampling or sampling of only the first part
35
of events.
36
Passive sampling devices (PSDs) have been developed for monitoring of water quality in surface
37
waters such as lakes and streams. Most PSDs are based on diffusion of solutes through a membrane or
38
diffusion layer to a collecting phase.9 In this way mainly the dissolved fraction (labile or bioavailable)
39
will be subject to sampling by PSDs, unless the solute’s affinity for the PSD-polymer is stronger than
40
for particles suspended in the water. Such samplers have been used for example for poly aromatic
ACS Paragon Plus Environment
2
Page 3 of 22
Environmental Science & Technology
41
hydrocarbons (PAHs) in stormwater drainage wells,10 heavy metals in the inlet and outlet from a
42
stormwater detention pond,11 and as a semi-quantitative approach to source tracing of metals in sewer
43
systems.12 The advantages of PSDs are lower costs for equipment and that the measurements are
44
integrated over time-periods of weeks to months, increasing the period represented by each sample
45
compared with e.g. grab sampling or flow proportional sampling. Most PSDs however collect solutes in
46
a time-integrative manner. When deploying time-integrative PSDs over periods of e.g. one month in a
47
highly dynamic storm drainage system with long periods without flow, the resulting sample will be
48
dominated by the concentration in the adjacent stagnant water during dry weather and not by the
49
concentrations occurring during runoff. The optimal PSD for stormwater runoff would therefore sample
50
proportionally to the flow in the system.
51
In recent years research has been conducted on a new passive sampling technique, which is based on
52
advective flow of water through the PSD (called SorbiCell).13 Successful application of this sampler has
53
been reported for measuring nitrate and phosphorous in surface waters and drainage water with a time-
54
integrative installation method,14 and sorbents are available for sampling a range of heavy metals,
55
volatile organic compounds, PAHs, pesticides and nutrients (www.sorbisense.com). The construction of
56
this flow-through sampler allows installing it in a new way, which produces velocity dependent
57
measurements that are appropriate when evaluating loads or flow weighted mean concentrations related
58
to dynamic discharge of pollutants. This velocity dependent passive sampling method may thus
59
potentially revolutionize monitoring of stormwater systems.
60
The aim of the work presented here is to test the new velocity dependent installation method for the
61
flow-through PSD in stormwater runoff. We focus on the heavy metals Cu, Zn and Pb at realistically
62
low concentrations, but the results are in principle valid for any dissolved micropollutant or
63
micropollutants sorbed to small particles for which appropriate sorbents are available. In order to
64
evaluate the precision of the PSD measurements, replicate PSDs were deployed. In order to evaluate the
65
accuracy of the PSD measurements, volume proportional samples were taken during parts of the
66
deployment period of the PSDs. However, we recognize that the ‘true value’ cannot be found for such ACS Paragon Plus Environment
3
Environmental Science & Technology
Page 4 of 22
67
complicated systems irrespective of the chosen sampling approach (there are always uncertainties and
68
biases connected with sampling methods in stormwater).8,15 Furthermore, an uncertainty-calibrated16
69
dynamic stormwater quality model,17 was used to evaluate loads and concentrations during the whole
70
deployment period including periods where no volume proportional samples were taken, in order to
71
validate the long-term average concentrations measured with velocity dependent passive sampling.
72 73
THEORY
74
Flow-through passive samplers
75
The flow-through PSD used here (SorbiCell), consists of a cartridge (6.5 cm long and 1 cm in outer
76
diameter) containing a macro-porous chelating resin sorbent (mesh particle size of 16-50) suitable for
77
sampling of e.g. Hg, Pb, Cu, Cd and a tracer salt (calcium-citrate tetrahydrate) held in the cartridge by a
78
spheriglass filter.13 When water passes through the cartridge, target analytes accumulate on the sorbent
79
and tracer salt is washed out proportionally to the passing water volume. The water volume that has
80
passed through the sampler during deployment (V) is then given by the equation: V=(Mt,0-Mt)/Ct,max
81
where Mt,0 is the initial tracer mass, Mt is the final remaining tracer mass after installation, and Ct,max is
82
the concentration of the tracer ion in solution.13 The concentration of analyte, Ca, is given by: Ca=Ma/V,
83
where Ma is the mass of analyte exracted from the PSD.13 A spheriglass filter in front of the sorbent
84
(100-160 µm) prevents larger particles from entering the sampler.
85
Assumptions behind installation of passive samplers for velocity dependent sampling
86
The principle behind the installation of PSDs for velocity dependent sampling is that the PSD is
87
placed directly in the stormwater flow. During runoff events water passes through the cartridge because
88
of the velocity head created by the flowing water. This method would, if the sampling was truly flow-
89
proportional and if the event runoff volume was known, result in an average measurement, which
90
represents the flow-weighted average concentration during runoff or the load of a pollutant passing the
91
sampling point during the deployment period. The assumptions which are necessary in order to regard
92
this sampler as flow-proportional are (a) that there is no uptake in the sampler when there is no flow in ACS Paragon Plus Environment
4
Page 5 of 22
Environmental Science & Technology
93
the system, (b) that the water velocity is linearly proportional to water flow at the site, (c) that the water
94
velocity at the sampling point is representative of the velocity over the whole flow cross-section, (d) that
95
the water velocity at the site is proportional to water flow through the sampler and the relationship is not
96
changing over time, and (e) that the total concentration is captured in the samplers (not only a fraction
97
of the pollutant such as the dissolved fraction). Further theory to elucidate these assumptions is given
98
below.
99
(a) Uptake in the sampler during no flow
100
The main uptake of analytes in the sampler is through advective flow through the sampler. Diffusion
101
will however result in uptake when there is no flow in the system. This can potentially bias the sampling
102
because analytes accumulate on the sorbent without wash-out of the tracer salt. It is well known that for
103
PSDs based on diffusion, the sampling rate varies linearly with surface area of the sampler.18-20 In order
104
to limit the diffusive uptake of analytes in the sampler during dry weather periods without flow,
105
diffusion to the sorbent is therefore limited through a small opening of the sampler. The maximum
106
sampling rate Rs,max during no flow can be estimated by assuming that the water boundary layer (WBL)
107
is rate limiting for the diffusion of analytes to the sorbent:18
, ∙
108
(1)
109
where Rs is the sampling rate of the sampler (volume of water per time e.g. L/d), A is the cross-section
110
area of the sampler, Dw is the aqueous diffusion coefficient of the pollutant, and δw is the thickness of
111
the WBL
112
(b) Velocity and flow relationship
113
The relationship between velocity and flow depends on the geometry and conditions of the monitored
114
system. If both flow and velocity is measured, the impact of weighting concentrations according to
115
velocity rather than according to flow can be estimated using a dynamic stormwater runoff quality
116
simulation model. This can be done by considering the measured flow and velocity and for each time-
117
step during a runoff event and weight the concentrations simulated by the simulation model (for the
ACS Paragon Plus Environment
5
Environmental Science & Technology
Page 6 of 22
118
same runoff event) in order to find an estimated velocity weighted event mean concentration (EMCv)
119
and an estimated flow weighted event mean concentration (EMCq):
120
∑ ∑
121
∑ ∑
∙
(2)
∙
(3)
122
where vi and qi are the ith measurements of velocity and flow, respectively, Ci is the pollutant
123
concentration in the ith measurement, and n is the number of velocity and flow measurements during the
124
runoff event.
125
(c) Cross-sectional velocity profile
126
The velocity of the water at the sampling point depends on the relative depth of installation as well as
127
the currents at the site. For relative depths of 0.1-1, the relative velocity in open channels is within
128
approximately 0.9-1.1 for turbulent flows, which can safely be assumed in stormwater runoff.21
129
(d) Relationship between runoff velocity and flow through the sampler
130
The relationship between water velocity and flow through the samplers has been tested in a flow-
131
channel by Kronvang et al.22 who found R2 = 0.77 for 1 week installations and R2 = 0.57 for 2 week
132
installations at 0.05-0.3 m/s. Change in this relationship over time can be caused by either instant
133
clogging if leaves or other ‘large’ objects get stuck on the sampler or slow clogging where small
134
particles captured in the sampler change the conductivity of the sampler over time. If slow clogging
135
occurs, it will lead to higher weighting of runoff during the first part of the sampling period than during
136
the last part. Also changes in the hydraulics around the sampler may have influence on the relationship
137
between runoff velocity and flow through the sampler. In both cases it is important to notice, that if a
138
peak concentration occurs either when the sampler is partially clogged or when the hydraulics around
139
the sampler disfavor sampling, this may have a negative effect on how much analyte is captured in the
140
sampler. But if intermittent clogging occurs and the water has average concentrations, it will not have an
141
influence on the evaluated concentrations.
142
(e) Fractions captured in the samplers ACS Paragon Plus Environment
6
Page 7 of 22
Environmental Science & Technology
143
Another important issue is the fraction of the analytes which is sampled by the PSD. The freely
144
dissolved and labile species are retained by the sorbent, which by design has a high affinity for the
145
analytes (here the heavy metals Cu, Zn and Pb). However, particles (including pollutants sorbed to the
146
particles) which are small enough to enter the sampler can also be captured in the sampler and therefore
147
included in the analysis. According to the manufacturer it is often seen that coloring of the polymer
148
material from e.g. humic material is more pronounced in the top of the cartridge, but after extreme flow
149
events coloring in the bottom of the cartridge is also experienced, and when sampling phosphorus the
150
PSD results correlate much better with traditional measurements of total phosphorous than with the
151
dissolved species orthophosphate.23 Other studies have found that heavy metals in stormwater were
152
either dissolved or mainly sorbed to particles smaller than 150 µm,24 that the main sources of traffic
153
related heavy metals were associated with particle sizes
