On the kinetics of pollutants in storm water runoff

11th International Conference on Urban Drainage, Edinburgh, Scotland, UK, 2008

Gnecco et al. 1

On the kinetics of pollutants in storm water runoff

I. Gnecco1*, J.J. Sansalone2, and L.G. Lanza1

1 DICAT, Dept. of Civil, Environmental and Architectural Engineering, University of Genoa,

Via Montallegro 1, 16145 Genoa, Italy 2 Department of Environmental Engineering Science, University of Florida, 217 Black Hall

Gainesville, Florida 32605 USA

*Corresponding author, e-mail [email protected] ABSTRACT Transport, bioavailability and fate of metals in storm water runoff are investigated by analyzing data collected at three commercial land-use sites in the town of Genoa (Italy): a landside site within the airport area and a tourism and a dry-bulk terminal sites within the port area. The monitoring campaign, carried out from November 2005 to December 2006, provided hydrologic and water chemistry data. Results clearly indicate the difference between airport and port terminal sites in terms of total suspended solids concentration values ranging from 30 mg/l at the airport-landside site to 1 g/l at the dry-bulk terminal site. Findings of metals partitioning investigation indicate that particulate lead seems to dominate the dissolved fraction irrespective of the water quality characteristics such as TSS and pH, while copper and zinc confirm the tendency for a predominant dissolved fraction and such trend seems strongly related to the TSS mass delivery. Metals speciation was simulated using the thermodynamic equilibrium model MINTEQ. Results of this metal complexation model indicate that zinc speciation is strongly dominated by the ionic form, while copper and lead exhibit a higher tendency to form complexes, thus their behavior is related to the ligands concentration associated with the specific land use. KEYWORDS Runoff quality; heavy metals; partitioning; speciation INTRODUCTION Pollutant wash-off during a rainfall-runoff event is a complex process that is influenced by several factors ranging from hydrologic, land use and watershed characteristics to the availability, transport, partitioning and dissolution mechanisms of particulate matter and pollutants (Berretta et al., 2007). Since the seventies the experimental studies (Cordey, 1977; Wanielista et al., 1977) carried out to investigate non-point source pollution and its effect on receiving water bodies pointed out the relevance of the pollutant load associated with storm water runoff, especially for metals whose concentration have become a major concern due to their impact in terms of either acute or chronic toxic effects onto the aquatic environment. In storm water runoff, metals partition between the dissolved and particulate-bound fractions through surface complexation processes, ionic exchange and precipitation phenomena; as for the dissolved fraction, metals reveal different dominant aqueous species including the ionic form and organic or inorganic complexes.


2 On the kinetics of pollutants in storm water runoff

The present paper aims at illustrating data collected at three different experimental sites located within airport areas and port terminals, thus assessing runoff pollutant constituents related to specific port and aviation surface land uses. The main objective of this paper is to examine the partitioning of metals in storm runoff between the dissolved and particulate fractions, and the speciation of their dissolved fraction, in order to provide relevant information about their toxicity, bioavailability and mobility. The present study may also have direct implications on the operational choice between storm water treatment BMPs for particle separation or the treatment of the soluble fraction in order to achieve improved protection of the receiving aquatic ecosystem. METHODS Experimental sites Since 2005, a monitoring programme was carried out to characterize storm water runoff from commercial areas with different land uses in the territory of the Province of Genoa, Italy. In this paper, data collected at three experimental sites within airport and port terminals are examined. Each gauge station is equipped with an automatic sampler (12 glass bottles with a capacity of 950 ml. each) to collect runoff water samples directly from the drainage system coupled with level/velocity gauges for continuous flow measurements. Flow rate data are calculated after application of specific stage-discharge curves. A brief description of the instrumented catchments is as follows:

• the airport-landside catchment site is located within a parking area connected to the airport terminal. The monitored area is an asphalt pavement with partial zones of concrete pavement for an extension of about 1.4 ha.

• the tourism terminal site is located within the car ferries terminal and includes the access road for private and commercial vehicles to/from the car ferries embarkation point and the parking lot for vehicles and trucks. The monitored area is a concrete-paved surface with an area of about 5 ha.

• the dry-bulk terminal site is located within a terminal employed mainly for (animal) food handling and storage operations. The area includes the rail access and is equipped with the overhead-travelling crane and the conveyer system to transfer products from vessels to the storage facility or directly on trucks and train; the monitored area has an extension of 0.7 ha.

The monitoring campaign was carried out from October 2005 to December 2006, providing one-minute runoff flow rate data and discrete runoff samples collected at five minute frequency for the assessment of water chemistry data; one minute rainfall records are also available at the tourism terminal site. As for water chemistry characterization, laboratory tests of runoff samples were performed by the Department of Chemical and Process Engineering (University of Genoa). In particular the following water chemistry parameters were examined: pH, electrical conductivity (EC), redox potential (ORP), total suspended solids (TSS), total organic carbon (TOC), and total hydrocarbons. Dissolved and particulate metals (copper, zinc, lead, mercury and cadmium) were also determined. Transport of metals In order to examine the transport and fate of heavy metals in storm runoff, which are directly discharged into the receiving water body, the metals delivery behavior was analyzed. In an aqueous environment, metals partition between the dissolved and particulate-bound fractions depending on several hydrologic and water chemistry parameters such as rainfall pH, alkalinity, runoff residence time, solids characteristics (Sansalone and Buchberger, 1997).


Gnecco et al. 3

To assess the dominant phase of each metal, the dissolved metal fraction, fd, was calculated as:

dp

dd cc

cf

+= (1)

with cd and cp being respectively the dissolved and particulate-bound concentrations. Values of fd greater than 0.5 indicate a dominant dissolved fraction of metals. Focusing on the dissolved fraction, the distribution of aqueous species was investigated for zinc, copper and lead in order to identify the dominant species (free ions or complexes) of each specific metal and their corresponding evolution during the runoff event. Metals speciation was simulated by using the thermodynamic equilibrium model MINTEQ (Allison et al., 1991). The model is able to simulate acid-base reactions, while chemical precipitation reactions and oversaturation were not included, assuming storm water runoff as a dilute aqueous solution. Data collected during the monitoring campaign provide the main input model parameters: dissolved metal concentrations, pH, redox potential and electrical conductivity (translated to ionic strength). RESULTS AND DISCUSSION Water Quality Data At the airport-landside site ten rainfall-runoff events were monitored; as for the port area, ten and five rainfall-runoff events were collected respectively at the tourism and dry-bulk terminal sites. In order to illustrate the variability of the pollutant load associated with storm runoff, water quality data are presented as box plots representing statistical results on a sample basis. The lower and upper boundary of each box indicate respectively the 25th and 75th percentiles, while the line within the box marks the median. Whiskers above and below each box indicate the 90th and 10th percentiles; individual crosses/circles showed in the plot represent the 5th and 95th percentiles. In particular, Figures 1 to 3 provide box plots for the concentration of zinc, copper, lead and total suspended solids, illustrating the most significant concentration values recorded across the whole monitoring campaign. As for general water quality data, it can be noticed that the TSS concentration values strongly differ across the three monitoring sites: the median concentration values range from 20 mg/l at the airport-landside site to 150 mg/l at the tourism terminal, and to 0.6 g/l at the dry-bulk terminal site. Indeed, as for the latter site, cargo operations determine a significant products build-up on the impervious surface, thus causing relevant wash-off of particulate matter during rainfall events. In spite of the wide range of TSS concentration values measured at the three sites of concern, heavy metals associated with storm runoff reveal significant concentration throughout the different sites. Note that due to the high delivery of total suspended solids, samples collected at the dry bulk terminal site cannot be fractionated by filtration, thus laboratory tests have been performed only for the dissolved fraction of metals. As for the dissolved metal fraction, lead and mainly copper show comparable concentration distributions (as illustrated in Figures 2 and 3) while zinc reveals a wide range of concentration values with a median between 1 µg/l at the dry-bulk terminal and 100 µg/l at the tourism terminal (Figure 1), as already observed in previous studies (Berretta et al., 2006; Gnecco et al., 2005). As regards the particulate metal fraction, concentration values generally increase with increasing TSS mass delivery; in addition, zinc shows a more disperse distribution (varying across several orders of



magnitude) when compared with copper and lead, and such behavior pointed out the variability in terms of zinc partitioning, as illustrated in the next section. Finally, mercury and cadmium were only present in traces across the whole monitoring campaign.

Airport-Landside Tourism Terminal Dry-bulk Terminal

Zn [µ

g/l]

0.1

1

10

100

1000

TSS

[mg/

l]

1

10

100

1000

10000

Znd

Znp

TSS

Figure 1. Concentration values of suspended solids (TSS) and zinc for both the dissolved (Znd) and particulate (Znp) fractions, measured at the landside site within the Airport of Genoa and the tourism terminal and dry-bulk terminal sites within the Port of Genoa.


Cu

[µg/

l]

0.01

0.1

1

10

100

1000

TSS

[mg/

l]

1

10

100

1000

10000

Cud

Cup

TSS

Figure 2. Concentration values of suspended solids (TSS) and copper for both the dissolved (Cud) and particulate (Cup) fractions, measured at the landside site within the Airport of Genoa and the tourism terminal and dry-bulk terminal sites within the Port of Genoa.


Pb [µ

g/l]

0.01

0.1

1

10

100

1000

TSS

[mg/

l]

1

10

100

1000

10000

Pbd

Pbp

TSS

Figure 3. Concentration values of suspended solids (TSS) and lead for both the dissolved (Pbd) and particulate (Pbp) fractions, measured at the landside site within the Airport of Genoa and the tourism terminal and dry-bulk terminal sites within the Port of Genoa.


Gnecco et al. 5

Metals Partitioning Figures 4 and 5 illustrate the mass delivery of TSS and metals (both dissolved and particulate fractions) as a function of the elapsed time for two rainfall-runoff events monitored at the airport-landside site and tourism terminal site, respectively. For each metal, the mean value of the dissolved fraction fd and its standard deviation, σ, are indicated, together with the event mean concentration (EMC) values of the TSS and pH.

18 February 2006

Cum

ulat

ive

Mas

s (g

)

010203040506070

Znp

Znd

Cum

ulat

ive

Mas

s (g

)02468

101214

Cup

Cud

Elapsed time (min)

0 10 20 30 40 50 60 70

Cum

ulat

ive

Mas

s (g

)

05

101520253035

Pbp

Pbd

Elapsed time (min)

0 10 20 30 40 50 60 70

Cum

ulat

ive

Mas

s (k

g)

0.0

0.5

1.0

1.5

2.0TSS

fd=0.02σ=0.02

fd=0.05σ=0.01

fd=0.01σ=0.001

EMCTSS=65 mg/lEMCpH=7.4

21 October 2006

Cum

ulat

ive

Mas

s (m

g)

0

50

100

150

200Znp

Znd

Cum

ulat

ive

Mas

s (m

g)

0

50

100

150

200

250Cup

Cud

Elapsed time (min)

0 20 40 60 80 100 120

Cum

ulat

ive

Mas

s (m

g)

0

100

200

300

400

500Pbp

Pbd

Elapsed time (min)

0 20 40 60 80 100 120

Cum

ulat

ive

TSS

(g)

0

100

200

300

400

500TSS

fd=0.53σ=0.09

fd=0.87σ=0.09

fd=0.06σ=0.10


Figure 4. Cumulative mass delivery of dissolved (Med) and particulate (Mep) metals fractions for two events monitored at the airport-landside site; the corresponding mean value and the



standard deviation of the dissolved fraction, fd, are reported for each metal. The TSS mass delivery is also illustrated and the ECM values of TSS and pH are indicated.

18 February 2006C

umul

ativ

e M

ass

(g)

0

50

100

150

200

250Znp

Znd

Cum

ulat

ive

Mas

s (g

)

0

10

20

30

40

50Cup

Cud

Elapsed time (min)

0 20 40 60 80

Cum

ulat

ive

Mas

s (g

)

0255075

100125150

Pbp

Pbd

Elapsed time (min)

0 20 40 60 80

Cum

ulat

ive

Mas

s (k

g)

0

10

20

30

40

50TSS

fd=0.04σ=0.01

fd=0.06σ=0.02

fd=0.02σ=0.004 EMCTSS=329 mg/l

EMCpH=7.6

16 November 2006

Cum

ulat

ive

Mas

s (g

)

0.0

0.5

1.0

1.5Znp

Znd

Cum

ulat

ive

Mas

s (g

)

0

10

20

30

40

50Cup

Cud

Elapsed time (min)

0 20 40 60 80 100

Cum

ulat

ive

Mas

s (g

)

0

5

10

15

20Pbp

Pbd

Elapsed time (min)

0 20 40 60 80 100

Cum

ulat

ive

Mas

s (k

g)

0

10

20

30

40

50TSS

fd=0.03σ=0.05

fd=0.26σ=0.01

fd=0.04σ=0.04


Figure 5. Cumulative mass delivery of dissolved (Med) and particulate (Mep) metals fractions for two events monitored at the tourism terminal site; the corresponding mean value and the standard deviation of the dissolved fraction, fd, are reported for each metal. The TSS mass delivery is also illustrated and the ECM values of TSS and pH are indicated. By comparing the 18 February and the 21 October events monitored at the airport-landside site, the different behavior of metals clearly emerges from Figure 4. Copper and mainly zinc


Gnecco et al. 7

reveal their tendency for a predominant dissolved fraction as a function of the TSS mass delivery: fd values of Zn and Cu are respectively equal to 0.87 and 0.53 for the 21 October event, characterized by an EMC of TSS equal to 13 mg/l, while for the 18 February event (EMCTSS = 65 mg/l) the dissolved fraction of Zn and Cu show values below 0.1. Note that, in the latter event, the predominance of the particulate fraction is enhanced by the pH value, equal to 7.4. In addition, the zinc affinity for the dissolved fraction clearly emerges at the airport-landside site due to the limited TSS values measured: the mean fd value of zinc across the whole monitoring campaign is equal to 0.64. On the contrary, lead shows the greatest affinity for the particulate phase, irrespective of the TSS concentration and pH values. Note that for the 18 February event, the mass delivery of particulate fraction of metals strongly mimics the TSS pattern while, for the 21 October event, dissolved zinc tends to increase more rapidly than copper and mainly lead, whose delivery is clearly related to the particulate matter. Results obtained from data collected at the tourism terminal site indicate the predominance of the particulate-bound fraction for all metals, even if metals reveal the same sequence in terms of fd values (Zn > Cu > Pb) generally observed at the airport-landside site (Figure 5). Indeed, for the 16 November event, when the event mean concentration of TSS is lower than the one observed for the 18 February event, the zinc affinity for the dissolved fraction emerges: the fd value for Zn is one order of magnitude higher than those observed for copper and lead. Metals Speciation Metals speciation was modelled by assuming semi-equilibrium conditions for each discrete sample, thus performing the corresponding charge balance. Modelling results indicate a charge balance error generally lower than 10%. Figure 6, 7 and 8 illustrate results of metals speciation obtained for the airport-landside site, the tourism terminal and the dry-bulk terminal sites during one sample monitored rainfall-runoff event. In particular, the dominant aqueous species for each metal are plotted as a function of the elapsed time, and the temporal profile of the TSS and TOC concentrations are also illustrated together with the pH. The corresponding hydrograph is also plotted on all graphs. Note that the metal species presented in each plot sum to greater than 90% of the aqueous species for that metal; specific species representing only a limited fraction, or characterized by concentration values lower than 10-10 mol/l, are neglected. Findings of the metals complexation model indicate that zinc speciation is strongly dominated by the ionic form, ranging from 75 to 90% and 70 to 80% of the total aqueous concentration respectively at the airport-landsite and tourism terminal sites. Water quality characteristics such as the pH and TOC concentration reveal fairly low influence in determining the zinc speciation, which was generally constant with respect to the runoff hydrograph (Gnecco et al., in press). As for the hydrologic characteristics, the ionic form of zinc tends to increase at the hydrograph peak due to the dilution effect of runoff volume in terms of ligands concentration. The next species are more influenced by the specific site characteristics: ZnSO4 represents the second species at the airport-landside site, while the tourism terminal shows ZnDOM as the next species due to the high concentration of dissolved organic matter. As for the speciation of zinc, the exception occurs for the dry bulk terminal due to the particular water quality characteristics with respect to the previous sites: the magnitude of TSS mass delivery, together with the significant chloride concentrations (measured as electrical conductivity values); the latter are due to the salt layer applied on the terminal surface area for maintenance operations. Therefore, ZnDOM and ZnCl+ play a dominant role in zinc speciation mainly at the beginning of the rainfall-runoff event. In spite of such specific conditions, results confirm the tendency of zinc for the ionic form, as illustrated in Figure 8: due to the dilution effect of



increasing runoff volume, Zn+2 is predominant at the hydrograph peak; furthermore ZnCO3 and ZnOH+ rapidly increase with increasing pH (from 7.4 to 8.7).

Met

al s

peci

es [m

ol/l]

1e-9

1e-8

1e-7

1e-6

CuCO3 Cu DOM Cu+2

CuOH+

Q (l

/s)

0

5

10

15

20

1e-8

1e-7

1e-6

1e-5Zn+2 ZnSO4 Zn DOM ZnCO3 ZnHCO3

+

Time (min)

20 40 60 80 100 120 140

Met

al s

peci

es [m

ol/l]

1e-9

1e-8

1e-7PbCO3

Pb DOM Pb+2 PbOH+ PbHCO3

+ PbSO4

Time (min)

20 40 60 80 100 120 140

Q (l

/s)

0

5

10

15

20

SST,

TO

C*1

0 [m

g/L]

20

40

60

80

100

120

pH

6

7

8

TSS pH TOC

Figure 6. Dissolved metal species for Cu, Zn and Pb, calculated using the MINTEQ model, and the corresponding hydrograph for the 18 February 2006 event at the airport-landside site. The pollutograph of total suspended solids, total organic carbon and pH values are also reported.

Met

al s

peci

es [m

ol/L

]

1e-9

1e-8

1e-7

1e-6

CuCO3

CuDOM Cu+2

Q (L

/s)

0

20

40

60

1e-8

1e-7

1e-6

1e-5

Zn+2 ZnDOM ZnSO4

Time (min)

10 20 30 40 50 60 70 80

Met

al s

peci

es [m

ol/L

]

1e-9

1e-8

1e-7

PbDOM PbCO3

Pb+2 PbHCO3

+

PbOH+ PbSO4 PbCl+

Time (min)

10 20 30 40 50 60 70 80

Q (L

/s)

0

20

40

60

TSS,

TO

C*1

0 [m

g/L]

1e+1

1e+2

1e+3

1e+4

pH

6

7

8

TSS pH TOC

Figure 7. Dissolved metal species for Cu, Zn and Pb, calculated using the MINTEQ model, and the corresponding hydrograph for the 16 March 2006 event at the tourism terminal site.


Gnecco et al. 9

The pollutograph of total suspended solids, total organic carbon and pH values are also reported.

On the contrary, speciation model results for copper and lead prove their affinity to form complexes, thus complying with the covalent theory for divalent metal ions according to the predicted order of bonding preference, which is higher for Pb and Cu when compared with Zn (Glenn et al., 2001). Thus the ligands concentration at each experimental site strongly affects their species distribution. At the airport-landside site, PbDOM is the dominant species for lead, while copper shows CuCO3 as the dominant complex. The limited TOC concentration and pH values generally higher than 7 determine the predominance of carbonate copper: in terms of equilibrium constant (log K) the affinity to form CO3 complexes follows the order Cu > Pb > Zn while the equilibrium constant of Me-DOM species reveals this order of bonding preference Pb > Cu > Zn. In addition, a significant relevance of the ionic form emerges for copper and lead (generally representing between 15 and 20% of the total aqueous concentration) at the airport-landside site when compared with results reported in the literature (Dean et al., 2005), likely due the limited availability of ligands. At both terminal sites, due to the significant increase in terms of TSS and TOC at the airport site, a more pronounced predominance of PbDOM with respect to the next species PbCO3 and Pb+2 is observed. In addition, due to the high chloride concentration, the PbCl+ species was observed at the dry bulk terminal. As for copper, results indicate a more competitive behaviour between CuCO3 and CuDOM.

Met

al S

peci

es [m

ol/L

]

1e-10

1e-9

1e-8

1e-7

1e-6CuDOM CuCO3

Cu+2 CuOH+

Cu(OH)2

Cu(CO3)2-2

Q (L

/s)

0

5

10

15

1e-9

1e-8

1e-7

1e-6

1e-5

Zn DOM ZnCl+ Zn+2 ZnCO3

Zn(OH)+ Zn(OH)2

Time (min)

0 20 40 60 80 100

Met

al S

peci

es [m

ol/L

]

1e-10

1e-9

1e-8

1e-7

1e-6Pb DOM PbCl+ PbCO3

Pb+2

PbOH+ Pb(CO3)2

-2

Pb(OH)2

Time (min)

0 20 40 60 80 100

Q (L

/s)

0

5

10

15

TSS

, TO

C [g

/L]

0.01

0.1

1

10

pH

6

7

8

9

TSS pH TOC

Figure 8. Dissolved metal species for Cu, Zn and Pb, calculated using the MINTEQ model, and the corresponding hydrograph for the 18 February 2006 event at the dry-bulk terminal site. The pollutograph of total suspended solids, total organic carbon and pH values are also reported.

CONCLUSIONS The investigation on transport of metals associated with storm runoff points out:



• the relevance of zinc, copper and lead concentration values at the airport–landside site and the tourism and dry-bulk terminal sites, and their variability across the different rainfall-runoff events;

• the predominance of the particulate-bound fraction of lead, irrespective of the general water quality characteristics and specific site land-uses;

• the affinity of copper and mainly zinc for the dissolved fraction as a function of the concentration of total suspend solids and pH values;

• the ionic form reveals the dominant aqueous species of zinc across the whole monitoring campaign;

• copper and lead show their affinity to form carbonate species and complexes with the dissolved organic matter according to the water quality characteristics and thus depending on the specific land-uses.

Findings of the partitioning analysis and the metals complexation model allow to assess the impact of metals associated with storm water discharges on the receiving water bodies: the bioavailability of metal species decreases moving from the ionic form to weakly organic/inorganic species and to strongly bound complexes. Therefore the dominant species for copper and lead can be identified as the strongly bound metals, while the ionic form, whose predominance strongly emerges for zinc, is the most toxic among the dissolved metal species, thus posing a threat for the aquatic ecosystem. In light of the results presented in this paper, further investigations are needed in order to clearly relate the metals behavior to both the hydrological aspects of rainfall-runoff events and the specific land-use characteristics of each site. REFERENCES Allison, J.D., Brown, D.S. & Novo-Gradac, K.J. (1991). MINTEQA2/PRODEFA2, A geochemical assessment

model for environmental systems: Version 3.0 user’s manual. EPA/600/3-91. Berretta, C., Gnecco, I., Lanza, L.G. & La Barbera, P. (2007). Hydrologic control on storm water pollution at

two urban monitoring sites. Urban Water, 4(2), 107-117. Berretta, C., Gnecco, I., Lanza, L.G. & La Barbera, P. (2006). Quality of stormwater runoff from paved surfaces

of two production sites. Wat. Sci. Tech., 54(6-7), 177–184. Cordey, I. (1977). Quality characteristics of urban storm water in Sydney, Australia. Water Resour. Res., 13(1),

709-713 Dean, C.M., Sansalone, J.J, Cartledge, F.K. & Pardue, J.H. (2005). Influence of Hydrology on Rainfall-Runoff

Metal Element Speciation. J. Env. Eng., 131(4), 632-642. Glenn III, D.W., Liu, D. & Sansalone, J.J. (2001). Influence of Highway Runoff Chemistry, Hydrology and

Residence Time on Non-Equilibrium Partitioning of Heavy Metals – Implication for Treatment at the Highway Shoulder. Transp. Res. Rec., 1755, 129-140, Transportation Research Board, Washington D.C.

Gnecco, I., Sansalone, J.J. & Lanza, L.G. (2008). Speciation of zinc and copper in stormwater pavement runoff from airside and landside aviation land uses. Water, Air & Soil Poll., 192(1-4), 321-336.

Gnecco, I., Berretta, C., Lanza, L.G. & La Barbera, P. (2005). Storm water pollution in the urban environment of Genoa, Italy. Atm. Res., 77(1-4), 60-73

Sansalone, J.J. & Buchberger, S.G. (1997). Partitioning and First Flush of Metals in Urban Roadway Storm Water. J. Env. Eng., 123(2), 134-143.

Wanielista, M.P., Yousel, Y.A. & McLellon, W.M. (1977). Non point source effects on water quality. Journal Water Pollution Control Federation, 49(3), 441-451.

Documents

On the kinetics of pollutants in storm water runoff