IOP Conference Series: Earth and Environmental Science

PAPER • OPEN ACCESS

Effectiveness evaluation of LIDs through SWMM: A case study of typical

urban unit in Handan, China
To cite this article: Y C Huo et al 2019 IOP Conf. Ser.: Earth Environ. Sci. 344 012056

View the article online for updates and enhancements.

This content was downloaded from IP address 170.84.163.186 on 19/11/2019 at 03:52

The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

Effectiveness evaluation of LIDs through SWMM: A case

study of typical urban unit in Handan, China

Y C Huo1,2, X Wu1,3,6, Q H Luan1,4, J H Liu2 and D Ma5

1
School of Water Conservancy and Hydropower Power, Hebei University of
Engineering, Handan Hebei, 056038, China
2
Department of Water Resources, China Institute of Water Resources and Hydropower
Research, Beijing, 100038, China
3
Hydrology and Water Resources Investigation Bureau of Handan, 056006, China
4
Research Centre for Water Ecological Civilization & Social Governance of Hebei
Province, Handan, 056038, China
5
Hebei Province Hydrology and Water Resources Investigation Bureau, 050031,
China

E-mail: [email protected]

Abstract. LIDs are frequently utilized measures in urban waterlogging management in the
world and the evaluation of their effectiveness through model gets a lot of attention in urban
hydrological research. In this study, a typical urban unit of Handan, China was selected, and
regional SWMM model was calibrated and validated through measured maximum
waterlogging depths (MWD) of typical positions; the LIDs effectiveness on runoff process in
different return period events were evaluated through the SWMM. The results show that LIDs
have an effect on decreasing total runoff quantity and peak discharge and the effectiveness is
more significant when the return period is small. The research provides a technical support to
the scientific management of local flood and waterlogging and supplies a reference to local
Sponge City planning.

1. Introduction
Recently, the frequency of extreme rainstorms has increased due to global climate change [1]. In
addition, the accelerated urbanization has changed urban underlying surface, and the impervious
proportion areas have increased, resulting in lots of serious urban water problems [2]. Urban
waterlogging not only affects social and economic development, but also brings great inconvenience
to residents living, and even injures the lives [3]. In the late 1990s, low impact development (LID) as a
new conception was proposed in order to solve stormwater problem [4]. LIDs could control runoff
from the source through infiltration, filtration, evapotranspiration and other natural hydrological
processes, aiming to the reduction of runoff quantity and pollution, and the protection of the receiving
water, which is different from traditional stormwater management [5].
In recent years, with the implementation of the sponge city construction policy in China [6], LIDs
researches have been the focus in China. Many experts in the world had studied the effects of LIDs.
The effectiveness of concave greenbelt and permeable pavement were evaluated [7], and these related
results propose that the proportion of concave greenbelt area in the community area is at least 30%,
and the proportion of permeable pavement area in whole road area is at least 20%. The research

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

indicates that rain barrel cistern and porous pavement are effectiveness on the total runoff reduction,
and 50 percent of rain barrel/cistern, 50 percent of porous pavement and 25 percent of rain
barrel/cistern combined with 25 percent of porous pavement are more effective in the research
watersheds of 70 km2 and 40 km2 near Indianapolis, Indiana [8]. The comprehensive effectiveness and
benefits of LIDs (i.e. bio-retention, concave greenbelt, permeable pavement, vegetative swales, and
green roofs) also had been studied [9],[10].
Model simulation is an important means to study and analyze the effect of LIDs, such as SWAT
[11], MIKE SHE [12], MIKE URBAN [13] and SWMM [14],[15]. SWMM is a dynamic precipitation
- runoff simulation model developed by the Environmental Protection Agency (EPA), which is mainly
used for single event or long-term (continuous) simulation of runoff quantity and quality from
primarily urban areas. Through SWMM, the generated runoff quantity or quality in each sub-
catchment can be tracked, and the discharge, water depth and water quality in each pipe and channel in
any time step during any period can be simulated [16]. Due to these above properties, SWMM was
selected in this study.
Recently, the application frequency of LIDs increase greatly for waterlogging management in
China, however the cities in humid urban areas were more selected in the case studies [17] and few
cases were be focus on urban areas of semi-humid and semi-arid areas. In this research, a closed urban
unit in Handan was selected and different two scenarios (present with no LIDs and LIDs) in series of
rainfall events of different return periods were designed. Furthermore, different runoff processes in
different scenarios were simulated and compared through SWMM. This study can improve the
applicability of SWMM, also supply the reference to LIDs application in semi-humid and semi-arid
regions.

2. Study area and data collection

2.1. Study area

North Ring Road

Shangbi
Gauge

Hanlin
Chnanel

Fuyang
River

Zhizhang
River
Legend
Gauge
Outlet
Node
Pipeline
Subcatchment
River 0 0.2 0.4 0.8 1.2
kilometers
Study area
Handan city
Mubi
Province Gauge

Figure 1. Study area location and the distribution of calibration position.

2
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

Handan city is located in the south of Hebei province and the east of Tai-hang Mountains (figure 1),
China, of which the climate is semi-humid and semi-arid continental with warm temperate. The
average annual rainfall is 502.7 mm, and it is mostly concentrated on the period from June to August.
The study area is located in the main metropolis of Handan City and bounded by North Ring road in
the north, by Zhizhang River in the south, by Hanlin Channel in the east and by Fuyang River in the
west, with an area of 29.67 km2. The underlying surface consists of residential and commercial land,
road and green land, of which the area is 24.65 km2, 4.38 km2 and 0.66 km2 respectively and the
corresponding percentage is 83.1%, 14.7% and 2.2% respectively.

2.2. Data collection

The data in this study includes information of regional drainage pipe network, information of typical
rainfall processes and the corresponding measured MWDs of typical positions.
The information of drainage pipeline network includes 273 nodes, 272 pipelines and 196 sub-
catchments and the corresponding spatial distributions, which provided by the drainage department of
Handan city Administration Bureau. As the study area is bounded by river, total twenty one outlets
marked red in the figure 1 exist, of which twenty are utilized for discharging runoff into the river
directly, and one is utilized for discharging the sewage into the treatment plant (figure 1).
Because the study area is small, there is no gauge in the area and only two rainfall gauges (Mubi
and Shangbi) are nearby. Furthermore, the rainfall difference between these two gauges is little, so the
arithmetic mean method i. e. the average of these two gauges, was used to calculate the regional
rainfall in this study. Two typical processes rained on June 21, 2017("6.21") and May 19, 2018
("5.19") were selected, of which the time step is five-minute.
Due to the absence of observed discharge data, the MWDs on different waterlogging positions of
two typical storm-runoff processes was selected, investigated and measured. The relative error
between the simulated MWD value and measured MWD value was calculated for model calibration
and validation. Total four typical waterlogging positions were selected in this case, according to the
waterlogging situations of several storms.

3. Methodology

3.1. Model construction

3.1.1. Setting of parameter values. Subcatchment width, N-Imperv, N-Perv, Dstore-Imperv, Dstore-
Perv and others are the main parameters required for the simulation in this research. The Horton
infiltration model was adopted to simulate the infiltration process in the study area, and Max. Infil.
Rate, Min. Infil. Rate, Drying time, Decay constant was the parameters involved.
The above parameters were based on the values in the SWMM user manual [16] and then
calibrated. The subcatchment width is related to the subcatchment area. Because the subcatchment
area of the study area has a more regular shape, and refers to the SWMM user manual [16], the
method was selected to calculate the subcatchment width following equation (1). Therefore, the
empirical coefficient K value needed to be determined.

W  K a (1)

Where W is subcatchment width; K is empirical coefficient; a is sub-catchment area.

3.1.2. Parameter calibration and model validation. Due to the limited measured data, only the “5.19”
rainfall event was selected to calibration and MWD was used as indicator to determine the parameters.
Firstly, "5.19" process was simulated through the SWMM. Then, the nodes corresponding to the four
measured points were found in the model, and the maximum depths of the corresponding nodes were
derived, and finally checked with MWD.
After adjusting the parameters several times, the error range of all verification points (table 1) were

3
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

controlled within -20%~20%, which meets the requirements of the corresponding standard [18]. The
parameter calibration results are shown in table 2, and rainfall and discharge process of this event is
shown in figure 2(a).

Table 1. MWD comparison between measured and simulated

value during the calibration process.
Position No. Measured Value Simulated value relative error
(m) (m) (%)
(1) 0.06 0.06 0
(2) 0.07 0.07 0
(3) 0.09 0.08 11.1
(4) 0.07 0.07 0

Table 2. Initial and calibrated parameters.

Parameter Initial value Calibrated value
K / 2.1
N-Imperv 0.01 0.012
N-Perv 0.1 0.2
Dstore-Imperv 1.27mm 5mm
Dstore-Perv 1.27mm 1.27mm
Conduit roughness 0.01 0.01
Max. Infil. Rate 76mm/h 76mm/h
Min. Infil. Rate 12.7mm/h 12.7mm/h
Drying time 7h 7h
Decay constant 4 4

200 0 250 0
Rainfall Rainfall
0.5 200 3
160
Discharge
Discharge(L/s)
Discharge(L/s)

Discharge

Rainfall(mm)
Rainfall(mm)

1
120 150 6
1.5
80 100 9
2

40 50 12
2.5

0 3 0 15
0 100 200 300 400 500 600 0 100 200 300 400 500 600 700 800 900 1000
Time(min) Time(min)

(a) (b)
Figure 2. Rainfall and discharge process: (a) parameter calibration, (b) verification.

After the end of parameter calibration, “6.21” rainfall event was used to verify the reliability of the
model. The verify method was the same as the method of parameter calibration. The rainfall and
discharge process of this event is shown in figure 2(b), and the simulation results are shown in table 3.
According to the simulation results in table 3, the error of the four typical points were -16.7%, -12.5%,
0 and -16.7%, respectively. The error of (1) (2) (4) verification point were relatively high, mainly
because the model derived data was only accurate to two decimal places, and more accurate data could
not be given. According to the corresponding standard [18], the relative error range of the simulation
results is -20%~20%, and all verification points meet the requirements. Therefore, it can be considered
that the model has certain credibility of rain flood simulation in Handan area.

4
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

Table 3. Comparison on measured MWD and simulated MWD in validation.

Position No. Measured Value Simulated value Relative error
(m) (m) (%)
(1) 0.06 0.07 -16.7
(2) 0.08 0.09 -12.5
(3) 0.1 0.1 0
(4) 0.06 0.07 -16.7

3.2. Designed scenarios

Two scenarios were designed in this study. One was the current scenario, and the type of land use was
described in 2.1. The other was to add LIDs. For these two scenarios, this research simulates the
rainfall and discharge process at different return periods (2years, 5years, 10years and 20 years).

3.2.1. Designed precipitation. Based on rainstorm intensity formula of Handan [19] (equation (2)),
Chicago rain type generator was used to construct four designed precipitation with the return period of
2 years, 5 years, 10 years and 20 years, respectively. The rainfall lasted for 120 min and the peak
coefficient was 0.4. The rainfall process of different design frequencies is shown in figure 3.
7.802  7.5lg P (2)
i
(t  7.76)0.602
Where i is storm intensity, mm/min; P is rainstorm return period, year; t is duration of rainfall, min.

6
20 yrs
Rainfall intensity(mm/min)

5
10 yrs

4 5 yrs

3 2 yrs

0
0 20 40 60 80 100 120
Time(min)

Figure 3. Rainfall processes of different return periods.

3.2.2. LIDs selection and parameter setting. The LIDs include bio-retention cell, rain garden, green
roof, infiltration trench, rain barrel and others of which green roof and infiltration trench were selected
in this study. The main types of land use of study area are residential and commercial land, road and
green land. The infiltration capacity of green land is large, and waterlogging is not easy to occur, so
LIDs are not added in green land. Most of the impermeable land is residential and commercial land, so
green roof were designed on the roof of residential and commercial buildings, about 30 percent of the
total area. At present, all the surface of roads are impervious and the middle part is higher, thus, runoff
mainly concentrates on the pavement on both sides of the roads. Due to this, the artificial roads were
added with infiltration trench on both sides of the road, about 20 percent of the total area. The spatial
distribution of the LIDs is shown in figure 4.
In this study, green roof is composed of drainage layer, filter layer, soil matrix layer and vegetation
layer [20]. Refer to the results of others [21], combined with the regional situation, the parameters of
the green roof are set as shown in table 4. In this study, the infiltration trench is a gravel shallow
channel. According to the technical guide for sponge city construction [22], the parameters of

5
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

Infiltration trench are set in table 5. 06/21/2017 18

Subcatch Subcatch
LID Usage LID Usage
10.00 10.00
20.00 20.00
30.00 30.00
40.00 40.00
% %

(a) (b)
Figure 4. LID setting map: (a) present scenario without LIDs, (b) LID
scenario with two LIDs.

Table 4. Parameters of green roof.

Parameter Value Parameter Value
Berm height 150mm Soil thickness 150mm
Surface roughness 0.1 Soil porosity 0.5
Drainage Thickness 3mm Field capacity 0.2
Void fraction 0.5 Wilting point 0.1
Drainage roughness 0.1 Conductivity 12.7mm/h

Table 5. Parameters of infiltration trench.

Parameter Value Parameter Value
Storage thickness 300mm Surface roughness 0.1
Void ratio 0.75 Flow exponent 0.5
Seepage rate 0.5mm/h Offset height 6mm
Clogging factor 0

4. Results and analysis

Analyse of discharge process throughout study area is the focus of this study. Due to the study area is
bounded by the river, outlets are evenly distributed around, and the conflux time is not much different.
Therefore, the conflux time of a certain outlet is not considered separately. The changes in peak
discharge and runoff quantity of the overall system are analyzed at the two scenarios under different
return periods (2 years, 5 years, 10 years and 20 years).

4.1. Comparison of simulation results under different designed precipitation conditions

Due to the large return rainfall, the end of the discharge time is longer, and the trend of dewatering is
relatively stable. Therefore, only the first 6h of discharge process at different return periods (2, 5, 10,
and 20 years) were analyzed (figure 5). It is illustrated that the trend of different return periods were
similar, both gradually increasing and then gradually decreasing, and the runoff peaks were all
concentrated in 1.9 h~2.1 h after the start of rainfall. The four designed rainfalls began to runoff yield

6
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

was 0.86 h, 0.81 h, 0.76 h and 0.73 h after the start of rainfall, respectively. The time to start runoff
yield is continuously advanced as the return period increases. The peak discharge was 18.1 L/s, 66.9
L/s, 164 L/s and 347.9 L/s, respectively. As the return period increases, the peak value increased
rapidly. It can also be seen from figure 5 that the runoff quantity varies greatly under the four
scenarios. Under the 2 years (“2 yrs” for short) return period rainfall, the rainwater can be emptied
within 4 hours. In the case of a 5 yrs return period rainfall, rainwater evacuation requires 5.8 hours,
and in 10 and 20 years it takes more than 6 hours to drain the rainwater. It can be seen that the greater
the return period of the designed rainfall, the earlier the start of the flow, and the peak discharge and
total runoff quantity are also significantly increased.

2 yrs 5 yrs 10 yrs 20 yrs

350
Discharge(L/s)

300
250
200
150
100
50
0
0 1 2 3 4 5 6

Time(h)
Figure 5. The first 6 hours of discharge processes in different return periods.

4.2. Comparison of simulation results under different scenarios

Present 70
20 Present
60
Discharge(L/s)
Discharge(L/s)

LID LID
15 50

40
10
30

20
5
10

0 0
0 1 2 3 4 0 1 2 3 4 5 6

Time(h) Time(h)
(a) (b)
160 Present 350 Present
140 300
Discharge(L/s)

Discharge(L/s)

LID LID
120
250
100
200
80
150
60

40 100

20 50

0 0
0 1 2 3 4 5 6 7 8 0 2 4 6 8 10
Time(h) Time(h)
(c) (d)
Figure 6. Simulated rainfall runoff process in different return periods in two scenarios: (a) two
years, (b) five years, (c) ten years, (4) twenty years.

7
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

It can be seen from figure 6 that under different return periods, the LIDs group complete the discharge
process first than present group. With the increase of the return period, the closer the two groups of
discharge processes before the peak discharge occurs. It can be seen that the effect of LIDs weakens as
the return period increases. The specific values are shown in table 6.

Table 6. Runoff data under different scenarios.

Return Peak discharge(L/s) Total runoff quantity(m3)
period Present LID Ratio Present LID Ratio
2 yrs 18.1 11 39.23% 71.7 50.4 29.71%
5 yrs 66.9 62.1 7.17% 553 399 27.85%
10 yrs 164 152.2 7.20% 1940 1410 27.32%
20 yrs 347.9 329.7 5.23% 5180 3760 27.41%

Based on the analysis of data in table 6, it is concluded that the peak discharge of the LID scenario
under the design period of 2, 5, 10, and 20 years are relatively reduced by 39.23%, 7.17%, 7.20%, and
5.23%, respectively; the total runoff quantity are relatively reduced by 29.71%, 27.85%, 27.32% and
27.41%. It can be seen that the LIDs have a certain reduction effect on both peak discharge and total
runoff quantity, and the effect is most obvious under the 2 yrs return period rainfall, and the reduction
effect is continuously weakened as the return period increases.
The peak discharge reduction effect decreases rapidly under the 5 yrs return period. It can be seen
that the peak clipping effect of the combined LIDs is weaker in the face of large return period rainfall.
However, the weak of the total runoff quantity reduction effect is not obvious with the increase of the
return period.

5. Conclusion and discussion

In this study, The SWMM were calibrated and validated through measured MWDs of "5.19" rainstorm
and "6.21" rainstorm, respectively. Through SWMM, LIDs effectiveness in a closed urban unit of
designed runoff process in different return period events were evaluated. The peak discharge and total
runoff quantity in present scenario and LIDs scenario were analyzed and compared. The results show
that the related values in LIDs scenario were both lower than that in present, and the reduction effects
of LIDs decrease with the increase of the return period. The LIDs have best effect on two-year return
period event: the corresponding decrease of the peak discharge and total runoff quantity is 39.23% and
29.71% respectively. This study indicates that LIDs can effectively release the waterlogging pressure.
At the same time, the research provides a technical support to the scientific management of local flood
and waterlogging and supplies a reference to local Sponge City planning.
Due to the absence of data, only two measured rainfall were used for calibration and verification. In
the future, data collection of measured runoff and pipe flux should be increased to improve the
reliability of the model. In addition, in this study only green roof and infiltration trench are involved,
however different LIDs have different advantages. The method to combine and optimize sorts of LIDs
needs to be carried out in further research.

Acknowledgment
The researchers would like to extend their thanks to the National Natural Science Foundation of China
(grant No. 51739011) and the Natural Science Foundation of Hebei Water Authority, China (No.
E2017-57).

References
[1] Dietz M E 2015 Low impact development practices: A review of current research and
recommendations for future directions J. Water. Air. Soil. Poll. 22 4
[2] Zhang C, Wang Y and Ding W 2017 Vulnerability analysis of urban drainage systems: Tree vs.

8
The 5th International Conference on Water Resource and Environment (WRE 2019) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 344 (2019) 012056 doi:10.1088/1755-1315/344/1/012056

loop networks J. Sustainability 9 397

[3] Li Y X and Kang J H 2008 Optimization of settling tank design to remove particles and metals
J. Environ. Eng. 134 885-94
[4] Cheng M S, Coffman L, Riverson J, Shen J, Ouyang J, Lahlou M and Shoemaker L 2002 Low-
impact development management practices evaluation computer module J. WEFTEC. 2002
1522-34
[5] Dietz M E 2007 Low impact development practices: A review of current research and
recommendations for future directions J. Ecol. Chem. Eng. S. 186 351-63
[6] Yu K J and Li D H 2015 “Sponge city”: Theory and practice J. CCPR. 39 26-36
[7] Zhang C and Ding Z B 2014 Simulation study of concave greenbelt and permeable pavement
based on storm flood management model J. Water. Resour. Water. Eng. 25 185-9
[8] Ahiablame L M, Engel B A and Chaubey I 2013 Effectiveness of low impact development
practices in two urbanized watersheds: Retrofitting with rain barrel/cistern and porous
pavement J. Environ. Manage. 119 151-61
[9] Luan Q H and Fu X R 2017 Runoff effect evaluation of LID through SWMM in typical
mountainous, low-lying urban areas: A case study in China J. Water. 9 439-60
[10] Li J K and Deng C 2017 Comprehensive benefit evaluation system for low-impact development
of urban stormwater management measures J. Water. Resour. Manag. 31 4745-58
[11] Schuol J and Abbaspour K C 2007 Estimation of freshwater availability in the West African
sub-continent using the SWAT hydrologic model J. Hydrol. 352 1
[12] Demetriou C and Punthakey J F 1998 Evaluating sustainable groundwater management options
using the MIKE SHE integrated hydrogeological modelling package J. Environ. Modell.
Softw. 14 2
[13] Luan Q H and Zhang K 2018 The application of Mike Urban model in drainage and
waterlogging in Lincheng county, China C. IAHS. 379 381-6
[14] Luan Q H and Wang D 2017 Analysis and comparison on the flood simulation in typical hilly &
semi-mountainous region C. IGRED 100 012206
[15] Marleni N and Park K 2015 A methodology for simulating hydrogen sulphide generation in
sewer network using EPA SWMM J. Desalin. Water. Treat. 54 4-5
[16] Gironás J and Roesner L 2010 A new applications manual for the storm water management
model (SWMM) J. Environ. Modell. Softw. 25 813-4
[17] Liao Z L, Zhang G Q, Wu Z H, He Y and Chen H 2015 Combined sewer overflow control with
LID based on SWMM: An example in Shanghai, China J. Water Sci Technol. 71 1136
[18] Sun J C and Liang J Z 2001 Implementing the "code for hydrological information forecasting"
and making great efforts to modernize hydrological information forecasting J. Chin. Hydrol.
21 60-3
[19] Chen Q and Li M 2017 Analysis of short duration rainstorm pattern based on drainage and
waterlogging prevention planning J. Expr. Water. Resour. Hydrop. Inf. 38 38-40
[20] Justyna Czemiel Berndtsson J G 2010 Green roof performance towards management of runoff
water quantity and quality: A review J. Ecol. Eng. 36 351-60
[21] Chen X P and Huang P 2015 Research progress on regulation of green roof runoff J. Chin. J.
Appl. Ecol. 26 2581-90
[22] Bian H Y and Song J A 2016 The technical guide of sponge city construction J. Henan. Buil.
Mat. 2 138-40