Delivery and Frequency Distributions of Combined Wastewater Collection System Wet and Dry Weather Loads Patrizia Piro1, Marco Carbone1, John Sansalone2
ABSTRACT: Combined wastewater collection systems continue to serve as a common urban conveyance method in urban areas of Europe and older urban areas of the United States. This study uses combined wastewater collection system monitoring data from the urbanizing Liguori catchment and channel in Cosenza (Italy) to illustrate eventbased delivery and distribution of conveyed pollutant indices. Motivated by recent European Union (EU) discharge control legislation, this study specifically differentiates the event-based delivery of these indices between dry and wet-weather flows. Although the relatively steady to diurnal-variable delivery phenomena in dry weather flows are known, transport limiting phenomena for wet-weather hydrology and mass delivery typically are not known for the same catchment. Limiting categories of transport for a pollutant phase are generated by variables such as flow volume and duration, stream power, hydrograph parameters, and previous dry hours (PDH). Transport limitations of wet and dry weather events from the 414-ha catchment were analyzed and characterized as limited by mass indices (first-order, first flush transport) or limited by flow (zero-order transport). Results indicated significant concentration differences between mass- and flow-limited events. Higher concentrations were associated with mass-limited events. Frequency distributions of flow, total suspended solids (TSS), chemical oxygen demand (COD), and five-day biochemical oxygen demand (BOD5) were consistently exponential for wet-weather and mass-limited events. In contrast, flow, TSS, and BOD5 concentrations were distributed normally for flow-limited events. Results indicated a reasonable linear relationship between discharged TSS, COD, and BOD5 (biochemical oxygen demand) for Liguori Channel discharges into the Crati River. Wet-weather event transport was predominately mass-limited for TSS, COD, and BOD5. Water Environ. Res., 84, 65 (2012). KEYWORDS: wastewater collection system, pollutant transport, pollutant distribution, combined sewer overflows (CSOs), particulate matter, rainfall-runoff. doi:10.2175/106143011X13188633467111
Introduction Since the late 1970s, combined sewer overflows (CSOs) have been permitted in Italy provided that a minimum dilution rate (discharge of wastewater divided by discharge of stormwater) was achieved (Italian Law n. 319, 1976). This dilution ratio 1
University of Calabria (Italy), Department of Soil Conservation, Arcavacata di Rende (CS), Italy.
2 University of Florida, Environmental Engineering Sciences Department, P.O. Box 116540, Gainesville, Florida, 32611-6450; e-mail:
[email protected].
January 2012
ranged from 1:2.5 to 1:6, depending on regional regulations; for example, the Lazio region ranged from 1:3 to 1:6 (Artina et al., 2004). Dilution ratios ranged from 1:4 to 1:6 for small urban watersheds characterized by smaller water supplies and lower pollutant concentration to 1:2.5 to 1:3 for large urban watershed characterized by distributed water supplies and lower pollutant concentrations (Artina et al.., 2004). Specifically, for the Calabria region, the location of this study, regulations required a 1:3 dilution rate for CSOs to the Crati River. The Crati River is the primary receiving water for Cosenza, Italy. Figure 1 illustrates a plan view of the urbanizing catchment discharging wet and dry weather flows to the Crati River. Figure 2 summarizes the monthly flows of an average water year for the Crati River in the mixing zone of the primary CSO from Cosenza, the Liguori Channel. Table 1 summarizes the interaction between the Liguori Channel total suspended solids (TSS) and five-day biochemical oxygen demand (BOD5) in the mixing zone of the Crati River (Calomino et al. 2004). Figure 2 presents the mean monthly flow value and the standard deviation for flow data measured from 1926 through 2008. The water chemistry indices presented in this study were evaluated between June 17, 1999, and February 1, 2001. Table 1 and Figure 2 illustrate that for a given CSO there are periods of the year, specifically July through September, during which the Crati River is effluentdominated with a decline in baseflow. The differences for inflows and concentrations in the upper Liguori Channel are a result of intense rainfall that results in CSO peaks higher than those of the river, as in the case for the events recorded on July 15 and September 5, 2000. Similar to other regions and countries, efforts to reduce CSOs in Southern Italy and Mediterranean regions of similar climates with combined wastewater collection systems have historically have been driven by human health, economic, aesthetic, natural resource, and ecological values (Viviani, 2008). Southern Italy has the additional challenge that local wastewater treatment in small communities frequently generates effluent treated with only primary clarification followed by disinfection, which results in higher differential loading to receiving waters (Viviani, 2008). Although the spatially dense historic urban center of Cosenza dates to the Roman Empire, recent modern urbanization since the end of World War II has led to a city with increased impervious area and a population of 200,000 that has expanded into the steep slopes along the Crati River valley. The urbanization of Cosenza is representative of many urban communities served by combined wastewater collection systems 65
Figure 2—Monthly flow variability in the Crati River at the location in the mixing zone.
Figure 1—Combined wastewater collection system and urbanizing Liguori catchment of Cosenza, Italy. discharging to local rivers; whether located in Europe or the United States (Sansalone and Glenn, 2007). The modern integration of impervious surfaces, hydraulically efficient conveyance systems, and increased population and anthropogenic pollutant-generating activities (e.g., dry weather wastewater flows or vehicular traffic-generated loads) has resulted in a complex matrix of pollutant types and phase interactions. Transport of these pollutant loads is driven by hydrologic parameters modified by urban system. These parameters include increased volume, peak flow, and altered temporal distribution of flow. Because of the complexities of pollutant type, phases, interactions, and event hetero-dispersivity within a phase, many examinations of pollutant transport (e.g., the classic first-flush transport behavior) rely on aggregate pollutant indices (as opposed to properties). Examples include particulate matter measured as TSS in comparison to particle size distributions (PSDs); total solids that combine soluble and particulate phases; or COD as an index for dissolved, particulate, or total (dissolved plus particulate) ultimate oxygen demand without identification of the biochemical fraction. Furthermore, many studies have considered that the predominance of the pollutant load can be associated with the particulate phase (Appel and Hudak, 2001; Farm, 2002; Lee and Bang, 2000; Maidment, 1993; Wanielista and Yousef, 1993; Wu et al., 1998). Although indices such as TSS, particulate-based COD (CODp), or dissolved COD (CODd) do not provide a constitutive relationship to granulometric indices such as PSD, combined wastewater collection system wet-weather flows typically are examined as aggregate measurements of oxygen demand or gravimetric particulate matter indices of the dissolved and particulate phase (Chocat, 1997; Piro et al., 2007; Sakrabani et al., 2005; Ying and Sansalone, 2008). For this study, aggregate indices are used. Whether for combined or separate wastewater collection systems (MS4s), it is recognized that concepts such as hydrologic restoration (embodied through low-impact development or sustainable urban drainage concepts) can reduce wet-weather flow effects of volumetric, hydraulic, and pollutant loads to wastewater treatment plants (WWTP) and CSOs to receiving waters. The historic alternative to hydrologic restoration has been optimization of 66
centralized and decentralized storage and equalization and commensurate capacity improvement, whether through maintenance for interception/inflow (I/I), sediment removal, optimized pumping schemes, or the construction of parallel collectors (O’Connor and Field, 2002a). With respect to combined wastewater collection system flow treatment, one recommended solution is to provide equalization and storage capacity near the head of the WWTP, thereby optimizing use of existing treatment unit operations and processes (UOPs) for treatment of detained wet-weather flows (O’Connor and Field, 2002b). In urban areas with combined wastewater collection systems, there is a dichotomy between more recent efforts in hydrologic restoration and pollutant source control compared to historical practices of providing conveyance capacity coupled with centralized WWTP of wet-weather flows. Since the 1990s, practitioners in the United States have tenuously suggested bridging this gap through promotion of ‘‘best management practices’’ (BMP) to control wet-weather effects. Such BMPs typically have included structural volumetric controls such as detention and retention basins, water ‘‘quality’’ basins, or sedimentation tanks. Commensurate with such volume-based controls is particulate matter, PSDs, COD, and BOD5 load attenuation. Best management practices also can include nonstructural measures. The U.S. Environmental Protection Agency (U.S. EPA, 1993) defined BMPs as ‘‘schedules of activities, prohibitions of practices, maintenance procedures, and other management practices to prevent or reduce the pollution of waters of the United States’’. Objectives This study hypothesized that the flow volume-based delivery and distribution characteristics of wet and dry weather flow pollutant indices are distinct. This hypothesis was tested for the Liguori catchment, an urbanizing catchment in Cosenza, Italy. These characteristics serve as inputs for management strategies to minimize CSO effects. Therefore, the first objective was to examine wet and dry weather pollutant load characteristics from monitored event-based data. The second objective was to categorize the delivery of pollutant indices as either flow- or mass-limited with the hypothesis that combined wastewater collection system delivery illustrates both limiting phenomena. The final objective was to investigate the frequency distributions for each pollutant index monitored in dry and wet-weather conditions. Water Environment Research, Volume 84, Number 1
Table 1—Characteristics of Liguori Channel, Cosenza Italy, total suspended solids (TSS) combined sewer overflows (CSOs) and Crati River five-day biochemical oxygen demand (BOD5) concentrations in the mixing zone. Crati River parameters Event
3
Liguori Channel overflow parameters
QU (m /s)
QD (m /s)
BODU (mg/L)
BODD (mg/L)
dOF (min)
QOF (m3/s)
TSS (mg/L)
1.08 0.82 2.16 2.00 1.43 2.66 1.96 3.21 2.90 2.03 2.13 2.50 2.34 2.20 2.69 2.13 2.69 2.70 2.26
1.67 0.99 2.57 4.92 3.35 3.42 2.31 3.46 3.29 2.14 3.29 3.25 3.02 2.30 3.61 2.55 2.97 3.26 2.45
2.8 2.9 2.3 2.4 2.7 2.1 2.4 1.7 1.9 2.4 2.4 2.2 2.2 2.3 2.0 2.4 2.0 2.0 2.3
36.6 218.7 98.7 84.5 452.8 9.9 175.3 9.0 14.2 8.4 113.5 26.0 19.4 10.3 11.4 117.8 17.1 17.9 7.6
43 38 74 198 24 375 18 90 49 18 47 26 86 72 276 25 208 276 172
0.59 0.17 0.41 2.92 1.92 0.76 0.35 0.26 0.39 0.11 1.16 0.75 0.68 0.10 0.92 0.42 0.28 0.56 0.19
197 3433 1600 314 2112 26 3095 201 215 256 806 215 139 440 30 1881 371 185 118
17/06/1999 14/08/1999 25/04/2000 15/07/2000 05/09/2000 09/09/2000 02/10/2000 07/10/2000 08/10/2000 27/10/2000 01/11/2000 19/11/2000 22/11/2000 25/11/2000 26/11/2000 03/12/2000 26/12/2000 05/01/2001 01/02/2001
3
QU 5 Flow rate measured in upstream of mixing zone. QD 5 Flow rate measured in downstream of mixing zone. BODU 5 Biochemical oxygen demand measured in upstream of mixing zone. BODD 5 Biochemical oxygen demand measured in downstream of mixing zone. dOF 5 Overflow duration time. QOF 5 Mean measured overflow rate to Liguori Channel. TSSOF 5 Total suspended solids measured in overflow volume by Liguori Channel.
Table 2—Hydrologic parameters for Liguori catchment, Cosenza, Italy.
Event
Rainfall duration (min)
Runoff duration (min)
Maximum Rainfall rainfall rate depth (mm) (mm/h)
tp (min)
tr (min)
Qp (m3/s)
Vt (m3)
Vb (m3)
Qbm (m3/s)
182 77 49 22 58 234 38 97
47 24 13 11 58 120 21 79
1.51 3.20 0.81 4.84 1.17 1.51 5.04 5.24
7503 10125 632 5258 3844 7406 21646 17630
3826 602 2316 1079 7383 6029 3556 3357
0.11 0.05 0.23 0.18 0.16 0.13 0.22 0.22
0 0 0
0 0 0
0.30 0.25 0.26
0 0 0
4230 13248 12816
0.26 0.17 0.16
Wet-weather flow events 24 March 1998 28 April 1998 2 October 2000 1 November 2000 18 December 2000 24 February 2002 29 January 2003 30 January 2003
595 159 222 103 627 605 156 207
462 151 131 92 750 684 255 241
0 0 0
0 0 0
1.1 4.1 1.5 6.2 2.2 10.6 9.4 9.2
6.0 12.0 2.4 60.0 12.0 3.0 24.0 24.0
0 0 0
0 0 0
Dry-weather flow events 8 April 1998 9 May 1998 17 June 1998
tp 5 Time to Qp measured from start of rainfall, [T]. tr 5 Time to Qp measured from start of runoff, [T]. Qp 5 Volumetric peak flow rate, [L3T21]. Vt 5 Total flow volume, [L3]. Vb 5 total base volume, [L3]. Qbm 5 mean volumetric dry weather flow rate, [L3T21]. January 2012
67
Table 3—Hydrologic data and selected indices for monitored wet-weather events from Liguori catchment, Cosenza, Italy. Gamma distribution parameters Event
Vr (m3)
g
W
Pu (W/m2)
24 March 1998 28 April 1998 2 October 2000 1 November 2000 18 December 2000 24 February 2002 29 January 2003 30 January 2003
6783.6 10064.9 232 5204.9 2960.9 7319.2 20927.5 12729.9
125.374 36.502 17.162 1.652 23.444 121.446 32.532 41.623
0.529 1.001 1.923 9.354 1.058 0.676 1.278 0.631
2.68 6.91 0.24 8.70 4.13 2.12 20.36 14.73
Vr 5 Total runoff volume, [L3]; g and W 5 Shape and scale factor parameters of the gamma distribution were fit to the dimensionless cumulative hydrograph for each event. Pu 5 Stream power [ML2T23].
Methodology The urbanizing Liguori catchment is located in Cosenza, Italy, in the southern region of Calabria. The catchment topographic elevations range from 205 to 431 meters above sea level, with a catchment surface area of 413.6 ha consisting of buildings (10.2%), pavement (37.6%) and vegetated open areas (51.4%). The urbanized catchment contributes wet-weather flow from 200 ha with a population of 50,000 people. Originally, the Liguori Channel was a natural tributary to the Crati River. The portion of the Liguori Channel in the urbanizing area is a structural combined wastewater conveyance. Wet-weather flows discharged as rainfall-runoff from the natural areas, above the urbanizing areas more proximate to the Crati River, are collected in a trapezoidal open-channel. At elevation 234 m a.s.l., the channel transitions to an enclosed polycentric concrete section. A wastewater collector transport flows by gravity to the Montalto WWTP upstream of the channel’s CSO discharge point to the Crati River at an overflow weir. Further details are provided elsewhere (Piro, 2007; Piro et al., 2007). Monitoring and Sample Analysis. Eight wet-weather and three dry weather events were sampled and analyzed in the Liguori channel during a monitoring campaign in 1998. Monitoring and sampling was conducted upstream of the CSO
Figure 3—Correlation between Liguori total suspended solid (TSS) overflows and Crati five-day biochemical oxygen demand (BOD5) and chemical oxygen demand (COD) in the mixing zone. discharge point in the Crati River. Continuous flow and rainfall measurements were recorded at intervals of one minute. Both wet and dry weather flow conveys loads of particulate matter, COD, or BOD5. Figure 4 illustrates the relationship between rainfall and runoff for wet-weather flows and compares the results to the temporal patterns of dry weather flows monitored. Figure 4 represents the signature of the urban/undeveloped watershed and the dry weather response of the wastewater collection system. Flow and rainfall intensity were normalized to peak event values so that a consistent scale could be used. The tmax represents the duration of the monitoring observation in each plot and was used to normalize the time axis. Rainfall duration indicates the difference in time between the first rainfall pulse measured and the last rainfall; runoff duration indicates the duration obtained from the difference between the hydrograph observed and baseflow hydrograph. After each event, hydrologic information was used to determine rainfall and runoff duration (td); rainfall depth; rainfall intensity (i); peak flow rate (Qp); runoff flow volume (V); baseflow volume (Vb); mean flow base (Qbm); previous dry hour (PDH); time to peak (tp); and peak flow (Qp). Table 2 summarizes the hydrologic data. For recorded wet-weather
Table 4—Monitoring data for wet- and dry-weather events (TSS = total suspended solids; COD = chemical oxygen demand; BOD5 = five-day biochemical oxygen demand). Range of parameters Event 24 March 1998 28 April 1998 2 October 2000 1 November 2000 18 December 2000 24 February 2002 29 January 2003 30 January 2003 8 April 1998 9 May 1998 17 June 1998
68
Flow type
Number of samples
TSS (mg/L)
COD (mg/L)
BOD5 (mg/L)
WET WET WET WET WET WET WET WET DRY DRY DRY
8 7 26 28 34 22 29 29 9 9 9
27–344 114–342 118–1870 141–1333 50–193 32–620 83–4487 102–1732 128–295 19–279 10–176
144–157 354–760 157–2607 160–1387 101–310 46–684 54–311 56–192 216–426 32–392 21–286
15–140 N/A 96–655 152–544 45–122 10–201 N/A N/A 105–207 25–185 13–132
Water Environment Research, Volume 84, Number 1
Figure 4—Event-based hydrology for urbanized Liguori catchment.
events, hydrologic indices were analyzed and summarized in Table 3. Pollutant indices were particulate matter (as TSS), oxygen demand (as COD and BOD5), and PSDs. To examine the differences between inorganic and organic loads under both dry and wet-weather conditions, sampling was conducted during two respective sets of conditions. Samples were collected at each time interval in 500-mL wide mouth polypropylene bottles from the start of observable rainfall at the site to the cessation of runoff for each particular event. Samples were collected at consistent 15-minute intervals throughout each event. Suspended solids concentration was analyzed according to the 2540D protocol (APHA et al., 1998). Chemical oxygen demand determinations were conducted using Standards Methods 5220B (APHA, 1998). The BOD5 was determined using Standards Methods 5210B (APHA, 1998). Table 4 illustrates the ranges of water chemistry indices measured from each event. January 2012
Runoff Event Transport To classify the transport of mass, events are categorized as mass- or flow-limited following the methodology of Sheng et al. (2008). A transport event is defined ‘‘flow-limited’’ when there is sufficient constituent mass available for transport throughout the event duration and the flowrate is the critical factor limiting the transport process. For a flow-limited process, mass delivery is proportional to flow volume. In contrast, a transport event is defined as ‘‘mass-limited’’ when there is not sufficient mass available; thereby limiting the mass-transport process. The various first-flush definitions in the literature have been shown to follow a mass-limited transport process (Sansalone and Cristina, 2004). A mass-limited transport process results in a disproportionate transport of mass with respect to runoff volume, a first-flush of mass. The different behavior between mass- and flow-limited has implications for volumetric control of wet or dry weather flows because these limiting transport 69
Figure 5—Transport of total suspended solids (TSS) from Liguori catchment.
categories are a function of volume transported and, therefore, affect volumetric design for treatment. Volumetric control based on this first-flush volume can provide a more economical design for volume-based controls. For mass-limited events where the predominance of mass is disproportionally transported with volume early in the event, control based on this volume is possible only if the transport behavior is known a priori. If masslimited transport of mass is first-order, the common transport assumption of a first-flush, the transport expression is as follows (Sheng et al. 2008): dM ~{KM dt
ð1Þ
Where,
SM~M0 (1{e{k1 SV )
ð3Þ
Where, SM SV M0 k1
5 5 5 5
Cumulative mass delivered, [M]; Cumulative volume, [L3]; Constituent mass at the beginning of the event, [M]; first-order coefficient, [L23].
For a flow-limited event, the transport is uncorrelated to the constituent mass available for transport, and the wash-off model has the following form (Sheng et al., 2008): ð4Þ
Where,
The transport coefficient K indicates the entrainment and delivery capability of the flow with respect to a particular constituent and is assumed to be linear related to the flow rate (Sheng et al. 2008).
70
Where, Q 5 Flowrate [L3T21] and k1 5 First-order coefficient [L23]. Combining these expressions yields the first-order masslimited washoff expression.
dM ~{K dt
M 5 Constituent mass remaining on the watershed; K 5 Wash-off coefficient, [T21]; and t 5 time.
K ~k1 Q
Figure 6—Transport of chemical oxygen demand (COD) from Liguori catchment.
ð2Þ
M 5 constituent mass available for transport [M] and K 5 washoff coefficient [MT21]. For flow-limited events, K was assumed to be linearly related to the flow rate (Sheng et al., 2006). K ~k 0 Q
ð5Þ
Water Environment Research, Volume 84, Number 1
Table 5—Mass- and flow-limited classifications for total suspended solids (TSS), chemical oxygen demand (COD), and five-day biochemical oxygen demand (BOD5) for each event. Basis Event
TSS
COD
BOD5
FL FL FL
FL FL FL
FL FL FL
ML FL ML ML FL ML ML FL
FL FL ML ML FL ML ML FL
ML N/A ML ML FL ML N/A N/A
Dry weather events 08 April 1998 09 May 1998 17 June 1998 Wet-weather events 24 28 02 01 18 24 29 30
March 1998 April 1998 October 2000 November 2000 December 2000 February 2002 January 2003 January 2003
SM~k0 SV
ð6Þ
Where, SM 5 cumulative mass delivered, M; SV is the cumulative volume, [L3]; and k0 is the zero-order coefficient, [ML23].
Figure 7—Transport of five-day biochemical oxygen demand (BOD5) from Liguori catchment.
Where, Q 5 flowrate [L3T21] and k0 5 zero-order coefficient [ML23]. Therefore, combining eqs. 4 and 5 yields the following expression:
In this study, cumulative volume (or elapsed time) was normalized to the total volume (or duration of flow), with the cumulative volume (or elapsed time) at the cessation of flow normalized to 1. Frequency Distribution of Constituents To analyze and highlight the variability of the different CSO indices and to correlate this with the conditions of flow (dryweather or wet-weather, mass-limited or flow-limited event) it was decided to statistically analyze the measured values for each event. For all indicators studied, and in particular for TSS, COD,
Figure 8—Concentration levels of each constituent on a wet and dry weather basis (TSS = total suspended solids; COD = chemical oxygen demand; BOD5 = five-day biochemical oxygen demand). January 2012
71
Figure 9—Concentration levels of each constituent on a mass and flow limited basis (TSS = total suspended solids; COD = chemical oxygen demand; BOD5 = five-day biochemical oxygen demand).
and BOD5, datasets were examined from the monitoring campaign. For each event these are:
N N N N
All All All All
the data measured in dry conditions; the data measured in wet weather; measured data for events classified mass limited; and measured data for events classified flow limited.
Each of the above set was statistically analyzed considering the main characteristics of early data and the random distribution of values. The frequency distributions were then compared with known probability distributions such as the uniform, normal, and log-normal distribution. Results and Discussion The relationships between TSS and COD and TSS and BOD5 were investigated in Figure 3, which illustrates that these relationships are both linear. The fit for both distributions is greater than 85%. This result suggests that pollutant transport in Liguori Channel is largely particulate-matter driven. This suggests that treatment should focus first a physical unit operation to attenuate flow volume and reduce particulate matter loads. If such primary treatment is insufficient for receiving water discharges, then secondary chemical or biological oxidation can be facilitated for the aqueous-phase pollutants and oxygen demand. Combined wastewater collection systems transport widely divergent flowrates and volumes when comparing wet and dry weather flows. Results in Table 2 and 3 illustrate the variability of the hydrologic data monitored in this study. For example, the mean and standard deviation for rainfall depth was 5.5 63.9 mm; maximum rainfall intensity was 17.9 6 19.0 mm/h; and peak flowrate of wet-weather flow was 2.9 6 1.9 m3/s. Furthermore, runoff volume on an event-basis varied from a maximum of 21 646 m3, on January 29, 2003, to a minimum value of 632 m3 on October 2, 2000. The variability of rainfall and wet- and dry-weather flow is illustrated for each event in Figure 4. 72
Event-Based Transport: Mass- or Flow-Limited To analyze transport for each dry- and wet-weather event, the dimensionless cumulative mass of TSS, COD, and BOD5 were plotted as function of the dimensionless cumulative flow volume. Figures 5 through 7 summarize these results for each event. These plots separately represent the intra-event transport (wet or dry weather) as either mass- or flow-limited. The diagonal line within each plot represents a zero-order (flowlimited) transport. Dry-weather flow events consistently exhibited flow limited transport of TSS, COD, and BOD5. Although most wet-weather events were mass limited for TSS and BOD5, three of the wet-weather events consistently exhibited flow-limited behavior.
Figure 10—Distribution of pH and conductivity on a wet- and dryweather basis. Water Environment Research, Volume 84, Number 1
Figure 11—Distribution for each constituent for wet- and dry-weather flows (TSS = total suspended solids; COD = chemical oxygen demand; BOD5 = five-day biochemical oxygen demand). The categorization results for event-based transport are summarized in Table 5. Results indicate that transport was mass limited for TSS in 55%, COD was 45%, and BOD5 in 63% of events observed. This variability is consistent with previous studies (Deletic, 1998; Diaz-Fierros, 2002; Gupta and Saul, 1996; Sansalone and Buchberger, 1997; Skipworth et al., 2000; Thornton and Saul, 1987; U.S. EPA, 1993). Identification of this transport variability is important for treatment designs based on volumetric capture and treatment of selected fractions of an event. A thorough study of the relationships between environmental parameters and characteristics of transport of pollutants has been made with particular reference to the parameters given in January 2012
Table 3. To understand possible links between modes of transport and the hydrological characteristics logistic regression has been applied, specifically with reference to rainfall duration; total runoff volume (Vr); hydrographs characteristics (defined with the parameters of the cumulative gamma distribution); stream power (Pu); and previous dry hour (PDH). The analysis showed that the parameters that most influence the type of transport are, in order of importance: total runoff volume (Vr); stream power (Pu); and PDH. Further investigation is needed to generalize the link between hydrological characteristics and mass- or flow-limited. Gravimetric concentrations associated with TSS, COD, and BOD5 are examined by comparing wet and dry weather 73
Figure 13—Flow and particle size distribution (PSD) distribution in wet- and dry-weather flows. a normal distribution for wet-weather flows, dry-weather flow pH is relatively constant. Frequency distributions of concentration for TSS, COD, and BOD5 summarized in Figure 11 are consistently exponential for wet-weather flows; distributions for dry-weather flow is consistently uniform for each. The variability of concentrations during wet-weather flows is clearly illustrated in Figure 11. When examined based on mass- and flow-limited events, TSS, COD, and BOD5 consistently illustrate exponential frequency distributions as summarized in Figure 12. In contrast, for flow-limited events, although TSS and BOD5 illustrate normal distributions, COD distribution remained exponential. The frequency distribution of wet-weather flows was exponential in comparison to a normal distribution for dry-weather flows as shown in Figure 13. Wet and dry weather PSDs were log-normally distributed across the range of PM diameters.
Figure 12—Wet-weather distributions for each constituent for mass- and flow-limited transport behavior (TSS = total suspended solids; COD = chemical oxygen demand; BOD5 = five-day biochemical oxygen demand). categories in Figure 8. The concentration results are illustrated non-parametrically. When examining median concentrations, there is a significant difference (a 5 0.05) for TSS. Concentration results are compared as mass- and flow-limited categories in Figure 9, also non-parametrically. Based on comparison, the median concentrations are significantly different (a 5 0.05) for TSS, BOD5, and COD. Concentration results also were examined based on frequency of occurrence, comparing wet and dry weather flows. Figure 10 illustrates that although conductivity is represented by a lognormal distribution for wet-weather flow, dry-weather flow conductivity is relatively uniform. Similar results also were reported for small urban-source area catchments (Sansalone et al., 2005). Results also illustrate that while pH is represented by 74
Conclusions This study examined and classified wet- and dry-weather transport in a 414 ha developing urban catchment with a combined trunkline wastewater collection system (Liguori Channel) as mass- or flow-limited. Although dry-weather flow produced flow limited transport for TSS, COD, and BOD5, wetweather transport was predominately but not exclusively masslimited for these indices. Wet-weather flows produced highly variable concentrations and flows that consistently exhibited exponential distributions. Mass-limited events produced significantly higher (a 5 0.05) concentrations than flow-limited events for TSS, COD, and BOD5. Similarly, wet-weather events produced significantly higher (a 5 0.05) concentrations than dry-weather events. In many cases, flows and concentrations were represented by a mean; for example, an event mean concentration that will reproduce event load. Additionally, the assumption of a normal distribution of flows or concentrations also was made. Results in this study indicate that, in contrast to such assumptions, distribution can be log-normal or exponential. Particularly for wet-weather flows, distributions were not Water Environment Research, Volume 84, Number 1
gaussian. Characterization, therefore, may be better represented by additional statistics that include the median. Results indicate that the variability of transport of TSS, COD, and BOD5 is driven largely by flow phenomena. Despite this variability, two limiting behavior emerged: transport based on mass limitations (a classic first-order exponential first flush); and flow limitations. These results are important when designing volumetric capture and treatment unit operations and processes for combined wastewater collection system flows. Submitted for publication February 2, 2011; revised manuscript submitted May 4, 2011; manuscript accepted for publication June 16, 2011. References American Public Health Association; American Water Works Association; Water Environment Federation (1998) Standard Methods for the Examination of Water and Wastewater, 20th ed.; American Public Health Association: Washington, D.C.; American Water Works Association: Denver, Colorado; Water Environment Federation: Alexandria, Virginia. Artina, S.; Calenda, G.; Calomino, F.; La Loggia, G.; Modica, C.; Paoletti, A.; Papiri, S.; Rasulo, G.; Veltri, P. (2004) Sistemi di Fognatura. Manuale di Progettazione (Sewerage System Design Manual); CSDU–HOEPLI: Milano, Italy. Appel, P. L.; Hudak, P. F. (2001) Automated Sampling of Stormwater Runoff in an Urban Watershed, North-Central Texas. J. Environ. Sci. Health, A36 (6), 897–907. Calomino, F.; Piro, P.; Veltri, P.; De Filpo, M.; Palma, G. (2004). Impatto Dello Sversamento di Acque Miste Nel Fiume Crati. (Impact Of The Combined Sewer Overflows In The Crati River.) Proceedings of the I Workshop Modeci - Modelli Matematici per la simulazione di Catastrofi Idrogeologiche (I Workshop Modeci - Mathematical Models for the Simulation of Hydrogeological Disasters); March, 30–31, 2004; Rende (CS), Italy. Chocat, B. (1997) Encyclope´die de l’Hydrologie Urbaine (Encyclopedia of Urban Hydrology); Lavoisier TEC&DOC: Lavoisier, Paris, France. Deletic, A. (1998) The First Flush Load of Urban Surface Runoff. Water Res., 32 (8), 2462–2470. Diaz-Fierros, T. F.; Puerta, J.; Suarez, J.; Diaz-Fierros, V. F. (2002) Contaminant Loads of CSOs at the Wastewater Treatment Plant of a City in NW Spain. Urban Water, 4 (3), 291–299. Farm, C. (2002) Evaluation of the Accumulation of Sediment and Heavy Metals in a Storm-Water Detention Pond. Water Sci. Technol., 45 (7), 105–112. Gupta, K.; Saul, A. J. (1996) Specific Relationships for the First Flush Load in Combined Sewer Flows. Water Res., 30 (5), 1244–1252. Italian Law n. 319 (1976) Norme per la Tutela delle Acque dall’Inquinamento (Standards for the waters protection against pollution). Gazzetta Ufficiale Italiana, 141. Lee, J. H.; Bang, K.W. (2000) Characterization of Urban Stormwater Runoff. Water Res., 34 (6), 1773–1780. Maidment, D.R. (1993) Handbook of Hydrology. McGraw-Hill: New York. O’ Connor, T. P.; Field, R. (2002a) U.S. EPA CSO Capstone Report: Control System Optimization. Proceedings of the 9th International Conference on Urban Drainage, September 8–13, 2002, Portland, Oregon. O’ Connor, T. P.; Field, R. (2002b) Control Strategy for Storm-Generated Sanitary-Sewer Overflows. Proceedings of the 9th International Conference on Urban Drainage, September 8–13, 2002, Portland, Oregon.
January 2012
Piro, P. (2007) Il Bacino Sperimentale Urbano del Canale Liguori nella citta` di Cosenza. Osservazioni Sperimentali Quali-Quantitative nel Periodo 1995–2003 (An Urban Experimental Catchment Canale Liguiori near the City of Cosenza. Experimental Observations of Water Quantity and Quality during the Period of 1995–2003); Nuova Bios: Cosenza, Italy. Piro, P.; Carbone, M.; Garofalo, G.; Sansalone J. (2007) CSO Treatment Strategy based on Constituent Index Relationships in a Highly Urbanized Catchment. Water Sci. Technol., 56 (12), 85–91. Sakrabani, R.; Ashley, R. M.; Vollertsen, J. (2005) The Influence of Biodegradability of Sewer Solids for the Management of CSOs. Water Sci. Technol., 51 (2), 89–97. Sansalone, J. J.; Cristina, C. M. (2004) First Flush Concepts for Suspended and Dissolved Solids in Small Impervious Watersheds. J. Environ. Eng., 130 (11), 1301–1314. Sansalone, J.; Glenn, D. (2007) Metal Distributions in Soils Receiving Urban Pavement Runoff and Snowmelt. Water Environ. Res., 79 (7), 736–752. Sansalone J. J.; Hird, J. P.; Cartledge, F. K.; Tittlebaum, M. E. (2005) Event-Based Storm Water Quality and Quantity Loadings from Elevated Urban Infrastructure Impacted by Transportation. Water Environ. Res., 77 (4), 348–365. Sansalone, J. J.; Buchberger, S. G. (1997) Partitioning and First Flush of Metals and Solids in Urban Highway Runoff. J. Environ. Eng., 123 (2),134–143. Sheng Y.; Ying, G.; Sansalone J. (2008) Differentiation of Transport for Particulate and Dissolved Water Chemistry Load Indices in RainfallRunoff from Urban Source Area Watersheds. J. Hydrol., 361 (1–2), 144–158. Sheng, Y.; Sansalone, J.; Calomino, F. (2006) Estimation of Solids Loadings to Rainfall-Runoff Unit Operations Using a Unit Pollutograph Concept for Source Area Watersheds. Water Sci. Technol., 54 (6–7), 363–369. Skipworth, P. J.; Tait, S. J.; Saul, A. J. (2000) The First Foul Flush in Combined Sewers: An Investigation of the Causes. Urban Water, 2 (4), 317–325. Thornton, R. C.; Saul, A. J. (1987) Temporal Variation of Pollutants in Two Combined Sewer Systems. Proceedings of the 4th International Conference on Urban Storm Drainage, Lausanne, Switzerland. U.S. Environmental Protection Agency (U.S. EPA) (1993) Combined Sewer Overflows Control Manual, Report EPA-625R-93-007; U.S. Environmental Protection Agency: Washington, D.C. Viviani, G. (2008). I Contributi Inquinanti Dei Sistemi di Drenaggio Urbano: Considerazioni Procedurali e Tecnologiche. (Contributions of Pollutants from Urban Drainage Systems: Procedural and Technological Considerations.) Proceedings of the 29u Corso d’Aggiornamento in Tecniche per la Difesa Dall’inquinamento (Refresher Course in Techniques for Protection Against Pollution), June 18–21, 2008, Guardia Terme (CS), Italy. Wanielista, M. P.; Yousef, Y. A. (1993) Stormwater Management. John Wiley and Sons Inc.: New York. Wu, J. S.; Allan, C. J.; Saunders, W. L.; Evett, J. B. (1998) Characterization and Pollutant Loading Estimation for Highway Runoff. J. Environ. Eng., 124 (7), 584–592. Ying, G.; Sansalone, J. (2008) Partitioning and Granulometric Distribution of Metal Leachate from Urban Traffic Dry Deposition Particulate Matter Subject to Acidic Rainfall and Runoff Retention. Water Res., 42 (15), 4146–4162.
75
