Professional Documents
Culture Documents
SWMM Simulation of The Storm Water Volume Control Performance of Permeable Pavement Systems
SWMM Simulation of The Storm Water Volume Control Performance of Permeable Pavement Systems
SWMM Simulation of The Storm Water Volume Control Performance of Permeable Pavement Systems
net/publication/267213954
CITATIONS READS
54 4,070
2 authors:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Shouhong Zhang on 23 October 2014.
Abstract: The reliability of the low-impact development (LID) module of the widely used Storm Water Management Model (SWMM) for
Downloaded from ascelibrary.org by MCMASTER UNIVERSITY on 10/22/14. Copyright ASCE. For personal use only; all rights reserved.
modeling the runoff reduction performance of permeable pavement systems was evaluated through example applications with rainfall data
from Atlanta. The method of calculating infiltration through the pavement layers of permeable pavement systems of the LID module was
found to be inadequate when the depths of the pavement layer are less than 120 mm and the computational time steps are longer than 30 min.
An alternative method of representing permeable pavement systems as equivalent regular subcatchments is proposed. Using this method, the
hydrologic operation of permeable pavement systems can be modeled by SWMM or other hydrologic models. DOI: 10.1061/(ASCE)HE
.1943-5584.0001092. © 2014 American Society of Civil Engineers.
Author keywords: Permeable pavement; Storm water management; Runoff reduction; Storm Water Management Model (SWMM);
Low-impact development (LID).
Introduction Imran et al. 2013). The runoff reduction rate of a permeable pave-
ment system is defined as the fraction or percentage of runoff
A permeable pavement system (Fig. 1) is a structural low-impact volume reduced by the system over the long term; it varies signifi-
development (LID) practice that generally consists of a permeable cantly due to differences in design, climatic, and operating condi-
pavement layer underlain by a stone reservoir [U.S. Environmental tions (Brattebo and Booth 2003; Gilbert and Clausen 2006; Collins
Protection Agency (USEPA) 1999; Pennsylvania Department of et al. 2008; Drake et al. 2012). Accurate and reliable methods are
Environmental Protection (PDEP) 2006; North Carolina Division needed to estimate the long-term average runoff reduction rates of
of Water Quality (NCDWQ) 2007; Credit Valley Conservation permeable pavement systems to ensure that optimum systems can
Authority and Toronto and Region Conservation Authority (CVC be designed and constructed.
and TRCA) 2010]. The surface pavement layer may be composed The Storm Water Management Model Version 5 (referred to as
of pervious concrete, porous asphalt, or different types of structural SWMM) developed by the USEPA is widely used for single event
pavers. These pavement layers are usually highly permeable with and continuous simulation of runoff quantity and quality from
permeabilities ranging from tens to thousands of millimeters per urban catchments (Rossman 2010). The newly added LID module
hour (Bean et al. 2007; Kuang et al. 2011). Washed coarse aggre- (SWMM Version 5.1.006) is expected to have the capability of sim-
gate is used to form the stone reservoir to provide temporary ulating the storm water management performance of various types
storage for peak flow and storm water volume control purposes of LID practices including permeable pavements. In this paper, the
(USEPA 1999; CVC and TRCA 2010). A perforated pipe (known reliability of the LID module of SWMM for simulating the runoff
as an underdrain) may be horizontally placed somewhere below the reduction performance of permeable pavements is examined and
pavement layer but above the bottom of the stone reservoir. For its unstable behavior is demonstrated. An alternative method based
permeable pavement systems without underdrains, the subsoil on the basic SWMM algorithms is proposed for evaluating the
needs to be highly permeable and with low clay contents [e.g., less long-term average runoff reduction rates of permeable pavement
than 30% (USEPA 1999)]. However, systems located in low- systems.
permeability soils with an infiltration rate of less than 15 mm=h
are usually required to be incorporated with underdrains (CVC and
TRCA 2010). Methodology
Permeable pavements have emerged as a widely used technol-
ogy for on-site storm water control (Pratt et al. 1989; Booth and
Leavitt 1999; USEPA 1999; Brattebo and Booth 2003; Sansalone Hydrological Processes Involved
et al. 2012). Reduction of runoff volume is one of the main The hydrological processes occurring in a permeable pavement
storm water management roles that permeable pavements play during nonwinter seasons are depicted schematically in Fig. 1.
(Dreelin et al. 2006; Collins et al. 2008; Ball and Rankin 2010; As rain falls onto a permeable pavement system, part of the rain-
water is trapped by small depressions on the surface or adsorbed by
1
Ph.D. Candidate, Dept. of Civil Engineering, McMaster Univ., the pavement layer. The rest of the rainwater may either move
Hamilton, ON, Canada L8S 4L7. E-mail: zhangswater@gmail.com downward through the pavement layer into the stone reservoir
2
Professor, Dept. of Civil Engineering, McMaster Univ., Hamilton, ON, or flow away from the site as surface runoff. Due to the extremely
Canada L8S 4L7 (corresponding author). E-mail: guoy@mcmaster.ca
Note. This manuscript was submitted on September 10, 2013; approved
high permeability of the pavement layers, surface runoff seldom
on August 27, 2014; published online on September 22, 2014. Discussion occurs (Brattebo and Booth 2003; Collins et al. 2008). As rainwater
period open until February 22, 2015; separate discussions must be submitted moves downward into the stone reservoir, a minor portion of the
for individual papers. This technical note is part of the Journal of Hydrologic rainwater is adsorbed by the aggregates in the stone reservoir and
Engineering, © ASCE, ISSN 1084-0699/06014010(5)/$25.00. the rest will percolate through the bottom of the reservoir into the
J. Hydrol. Eng.
Table 1. Parameters of Permeable Pavement Systems Used in the
LID-SWMM Simulations
System component Parameter Value
Surface layer Storage depth 1.5 mm
Vegetative cover fraction 0
Surface roughness 0.015
Surface slope 1%
Pavement layer Thickness 1–200 mm
Void ratio 0.16
Impervious surface fraction 0
Permeability 254 mm=h
Clogging factor 0
Storage layer Height 450 mm
Downloaded from ascelibrary.org by MCMASTER UNIVERSITY on 10/22/14. Copyright ASCE. For personal use only; all rights reserved.
J. Hydrol. Eng.
reservoir is added to water held in the depression storage, and drain V rain − V r
Rrbasic−SWMM ¼ ð4Þ
outflow is treated as a part of surface runoff. V rain
To ensure that a permeable pavement system is equivalent to the
pervious subarea of the subcatchment in a basic SWMM model, the
depth of depression storage on the pervious subareas (denoted as Simulation Runs
Sd , in millimeters) should be set to equal the maximum storm water
retention capacity of the permeable pavement system. The maxi- The values of parameters required in the LID-SWMM simulations
mum storm water retention capacity of a permeable pavement sys- are all listed in Table 1. The values of parameters used in the basic
tem (denoted as Rc max , in millimeters) without underdrains can be SWMM simulations are all the same as the values of the corre-
calculated as sponding parameters required in the LID-SWMM simulations ex-
cept Sd , which can be determined using Eq. (2) or (3). Permeable
h p ep he pavement systems without vegetative covers are studied in the
Rc max ¼ Sds þ þ s s ð2Þ simulations. Therefore, the vegetative cover coefficient and the im-
1 þ ep 1 þ es
Downloaded from ascelibrary.org by MCMASTER UNIVERSITY on 10/22/14. Copyright ASCE. For personal use only; all rights reserved.
J. Hydrol. Eng.
Downloaded from ascelibrary.org by MCMASTER UNIVERSITY on 10/22/14. Copyright ASCE. For personal use only; all rights reserved.
Fig. 2. Comparison of (a) LID-SWMM; (b) basic SWMM simulation Fig. 3. Comparison of (a) LID-SWMM; (b) basic SWMM simulation
results (hp ¼ 50 mm) results (hd ¼ 50 mm)
step, the runoff reduction rate increases sharply from 0.50 to 0.78 as capacity of permeable pavement systems with underdrains. There-
hd increases from 0 to 40 mm; further increases in hd beyond fore, the runoff reduction rates determined from the LID-SWMM
40 mm do not result in any increases in runoff reduction rates. This and basic SWMM simulations should remain unchanged when hp
is physically unexplainable because increases in hd translate changes from 0 to 200 mm. The basic SWMM determined runoff
directly into increases in the maximum retention capacity of the reduction rate does remain at a constant level when different sim-
system [as expressed in Eq. (3)], which should definitely result ulation time steps are used and when hp changes from 0 to 200 mm
in some increases in runoff reduction rates. Further investigation [Fig. 3(b)]. However, the LID-SWMM determined runoff reduction
of the LID-SWMM simulation results showed that when 1 h is used rate is highly affected by simulation time steps and increases sig-
as the simulation time step, surface outflow V sr stays at approxi- nificantly from approximately 0.17 to approximately 0.86 when hp
mately 16,399 mm and drain outflow V dr remains at 0 mm with hd increases from 0 to 120 mm [Fig. 3(a)]. The LID-SWMM results
increases beyond 40 mm. This is physically incorrect because are again physically incorrect.
(1) the highest input rainfall intensity is 90.9 mm=h, which is The findings that LID-SWMM can provide reasonable results
far below the permeability of the pavement layers (254 mm=h), for some cases with very short time steps but cannot provide rea-
surface outflow should therefore never occur as a result of infiltra- sonable results for some other cases even with very short time steps
tion excess at the surface of the pavement layer; and (2) surface seem to suggest that the problem is perhaps not a simple bug in
outflow as a result of saturation of the maximum retention capacity programming. A preliminary examination of the source code of the
of the pavement system is unexpected when there is no drain LID-SWMM (http://www2.epa.gov/water-research/storm-water-
outflow from the system. The basic SWMM determined runoff management-model-swmm?#downloads) indicates that the actual
reduction rate does not change much with simulation time steps infiltration rate through the surface of the pavement layer is con-
[Fig. 2(b)]. Regardless of the different time steps used, results de- trolled by the following four factors: (1) the available rainwater on
termined from basic SWMM simulations are close to those deter- the surface of the pavement layer, (2) the permeability of the pave-
mined from the LID-SWMM simulations with a 5-min time step ment layer, (3) the available void space of the pavement layer, and
(with root-mean-squared error ranging from 0.068 to 0.097). (4) the simulation time step. The storage capacity of the stone res-
Fig. 3 shows the comparison of results for permeable pavement ervoir and the drainage capacity of the underdrain (if there is an
systems with hd fixed at 50 mm and hp changing from 1 to underdrain) are not considered in the determination of the actual
200 mm. Because the storage provided by the field capacity of infiltration rate through the pavement layer. This infiltration calcu-
the pavement layer is ignored (i.e., θfp ¼ 0) in both the LID- lation method is not complete because, in reality, both the storage
SWMM and basic SWMM simulations, the change of hp is not capacity of the stone reservoir and the drainage capacity of the
expected to result in any variations in the storm water retention underdrain affect the actual infiltration rate because they affect
J. Hydrol. Eng.
the possibility and extent of system saturation and the associated Booth, D. B., and Leavitt, J. (1999). “Field evaluation of permeable pave-
generation of saturation–excess runoff. This incomplete infiltration ment systems for improved storm water management.” J. Am. Plann.
calculation method may be one of the causes of the previously Assoc., 65(3), 314–325.
reported problems. Brattebo, B. O., and Booth, D. B. (2003). “Long-term storm water quantity
Cases without underdrains were also modeled. Similar unrea- and quality performance of permeable pavement systems.” Water Res.,
sonable results from LID-SWMM simulations were observed. 37(18), 4369–4376.
Without the separate calculation of drain outflows, the possible Collins, K. A., Hunt, W. F., and Hathaway, J. M. (2008). “Hydrologic
comparison of four types of permeable pavement and standard asphalt
causes of the unreasonable LID-SWMM results could not be as
in eastern North Carolina.” J. Hydrol. Eng., 10.1061/(ASCE)1084-0699
clearly identified as for cases with underdrains. The basic SWMM
(2008)13:12(1146), 1146–1157.
model, however, can provide more accurate representation of
Credit Valley Conservation Authority (CVC) and Toronto and Region
permeable pavements without underdrains using regular subcatch- Conservation Authority (TRCA). (2010). Low impact development
ments. That is why detailed results from cases without underdrains storm water management planning and design guide, Mississauga and
are not presented here.
Downloaded from ascelibrary.org by MCMASTER UNIVERSITY on 10/22/14. Copyright ASCE. For personal use only; all rights reserved.