Study of inundation and eco drainage system approach in Cicayur Kampong area

. Cicayur Kampong, Pagedangan District, Tangerang Regency is a residential area adjacent to the urban area. This area has experienced inundation since 2018. This study aims to evaluate existing drainage systems, assess the performance of channels in serving surface runoff to their catchment areas and provide alternative solutions by eco drainage approach using rainwater harvesting and recharge well. This study was carried out by processing daily rainfall data over a time span of 2011 to 2021. Hydrology analysis was performed manually, and the hydraulic simulation was conducted using the Storm Water Management Model (SWMM) 5.1. The rainfall intensity applied for modeling is 37.29 mm/hour, 46.96 mm/hour, and 53.41 mm/hour for 2, 5, and 10 years return periods, respectively. The results of the simulation of the existing drainage system show that there are 2 (two) channels that experience overflow during the 2 and 5-year return periods and increase to 5 (five) channels during the 10-year periods. Alternative solution with rainwater harvesting and recharge well shows a large reduction in surface runoff flowing into drainage channels by 45%.


Introduction
In general, flooding or inundation is a condition where drainage channels are no longer able to accommodate existing water and lead the water to overflow the surrounding area.The presence of flooding or inundation phenomenon can occur due to very high rainfall and the inability of a drainage system to accommodate rainwater.Sometimes external factors such as clogging drainage due to the presence of debris cause the water to overflow the surface.A serious rainwater increase in urban drainage can be a threat to the life and safety of urban residents' property [1].According to previous research, changes in land use increase the impervious value of an area [2][3].A significant increase in rainfall causes the channels to be unable to accommodate water at its existing capacity.Hence, with a higher impervious value of the area, it prevents the rainwater from infiltrating the soil.This land use changes because of a growth in the population, which increases the conversion of land use from open space to residential areas.The development of residential areas that are not supported by the planning of drainage systems and good drainage infrastructure may cause inundation or flooding [4].
The settlement area of Cicayur Kampong is located in pagedangan district which is the second most populous district in Tangerang Regency after Medang district (Fig. 1).It has a total population of 13,494 residents.This study focuses on the results of the evaluation of existing drainage in the Cicayur residential area, especially at the point of flooding in early 2020 on *Corresponding author: yureana.wijayanti@binus.ac.idJanuary 2 to 4 in Cicayur Village.The development of land use is related to the availability of water catchment areas, changes in use in land use on a large scale can then cause rainwater that should have entered the infiltration system to turn into surface runoff [5].In its control, the drainage system, which is an effort mechanism in forming water jetting for surface water and groundwater which will then be flowed a large excess from an area or area so that it does not cause inundation that can interfere with surrounding residential activities.This research will be supported by the use of Storm Water Management Model (SWMM) software.SWMM is software intended to simulate surface runoff in quantity and quality.
SWMM was developed by the United States Environmental Protection Agency (UESPA) in 1971 [6].SWMM uses modeling with input parameters that have been recorded in field conditions [7][8][9][10][11].Modeling using the SWMM can be built using five main environmental components: rainfall, temperature, and evaporation data; surface runoff components; subsurface groundwater components; pipe conveyance systems, channels, flow regulators, contaminant storage, and accumulation units [12][13][14][15].Not all of these components need to be included in the analysis process, most only include rainfall, surface runoff, and the hydraulics of the conveyance system.One of the popular applications in the implementation of drainage system analysis is EPA -SWMM.EPA -SWMM is a program developed by the Environmental Protection Agency or EPA from the United States in 1971.EPA -SWMM can be used to run a simulation of water amount and quality for urban drainage systems [16][17].1.In doing both planning and analysis of the drainage system can be done by this application because of its function to estimate the performance of channels in accommodating runoff in a drainage system.Runoff on EPA -SWMM modeling can be reviewed on open channels [18], closed channels, detention ponds, and pumps [19][20].EPA -SWMM delivers the quality and quantity of runoff affected by the catchment area, average flow, flow depth, water quality in each pipe, and open channel simulation time included in the addition of time.There are several components in modeling with EPA -SWMM applications such as rain gage, time series, subcatchment, junction nodes, outfall nodes, flow divider nodes, storage units, conduits, and orifices.
The objective of this study are to evaluate the performance of the existing drainage system against the maximum annual return rainfall, and to determine solutions or mitigation in an effort to overcome inundation.

Existing drainage system
The initial step is done by arranging the background and identifying the drainage system problems in the Cicayur residential area.The data was collected in the form of rainfall data, channel detail data, a topographic map of the area, a site plan, and photos of the site.Furthermore, the research was conducted by checking the safety of the channel slope by comparing the result of flow velocity calculations with the existing speed standards.With the data and the results of hydrological analysis, the simulation of existing drainage systems in residential cluster areas with SWMM can be done [3,[21][22].Fig. 2 shows the flowchart of this research.
Primary data is data obtained through direct observation of the object of study.The form of primary data in this study is the size of the dimensions of the existing drainage channel at the flood study site, flood duration information, and rainfall data, which were observed at the Budiarto Meteorological Station Post, Soekarno-Hatta Meteorological Station, Tangerang Geophysical Station and South Tangerang Climatology Station.The Serpong Rain Station post is considered accurate for the primary data collection of rainfall because it is located in the lower reaches of the Cisadane watershed, which is also the location of the Cicayur Sub-Watershed.

Fig. 2. Research Methodology, modified from [23].
The calculation of regional rainfall using the Polygon Thiessen method is a calculation produced through the polygon form of a perpendicular intersection of the connecting line between stations.The combination of the line meetings will then form an area that will then become the area represented by a rain post or station located on a polygon [24].The Polygon Thiessen method is used when there are stations used with a minimum number of three observation stations.Modeling Polygon Thiessen with software requires an administrative map of the location and coordinates of the observation station.Regional rainfall is then obtained by processing the polygon area value with recapitulated rainfall data into maximum daily rainfall according to the year of observation (Fig. 3 and Table 1).
The calculation of the selected type of distribution and its suitability testing can then be recapitulated to find out which type of distribution to use for further modeling.In choosing the type of distribution to be used, you will look at the results of statistical analysis, statistical analysis is represented by the values of Coefficient of Variant (Cv), Coefficient of Skewness (Cs), and Coefficient of Kurtosis (Ck).Based on the results of statistical analysis, the distribution that has no requirements that meet that the distribution used is the Pearson Type III Log distribution because it does not have specific requirements in ensuring that the distribution is correct, it is continued testing with the Chi-Square test and the Smirnov -Kolmogorov test.Fig. 3. Polygon Thiessen method of regional rainfall in study area.Distribution testing is generally aimed at determining the selected distribution equations that can represent the statistical distribution of the data under review.Here are the test results of both topics.
The distribution chosen is a distribution with test results X 2 or DMAXIMUM.The value describes the smallest outlier or aberration (outlier).In this test, the corresponding distribution type was Pearson Log Type III.Calculation of rain intensity to determine the depth of rain in units of time (Table 2).The calculation of rain intensity can be solved by several types of formulas.In this study, two methods will be used, namely the IDF Mononobe and the Alternative Block Method (ABM).The catchment area of the drainage system is 24612 m 2 .

Rainwater harvesting discharge and recharge well depth
The calculation of the discharge is used by the rational method (Equation 1).

QA=0.002778 × C × I × A (1)
where: QA = discharge (m 3 /s) C = landcover coefficient I = rainfall intensity A = surface runoff area Based on Sunjoto [25], a calculation of water level height in the recharge well was performed using the following Equation 2.
where 3 Result and discussion

Existing drainage system
In the results of modeling the number of channels that are not able to accommodate the most capacity, occurs in the simulation of the 3rd hour and the 45th minute.In the modeling, there are 5 nodes whose storage exceeds the capacity to produce backwater occurrence at junction 19 to outfall 32.The following is a summary report to see the flooding nodes from the modeling results.In the modeling of existing conditions, there is a channel that does not work, it is marked by channels J33 -J34 and J34 -J3 that connect to the secondary channel I (table 3).The capacity of conduits (Qc) that accommodate the surface runoff discharge (Qsr) from area I, sub-area I and sub-area III are insufficient.This result is in alignment with the field survey result where the location of the inundation took place.The drainage capacity is small which might be due to its location on small roads around settlements.The results of the simulation of existing conditions with the SWMM 5.1 program can be validated with the results of the comparison of the existing channel discharge with the design flood discharge.The result of comparison between surface runoff in 2, 5, and 10 years of return period and the existing capacity of the drainage system using the SWMM modeling represents the actual condition in the study area.Therefore, this model would further be utilized to simulate the scenarios to overcome the inundation in the future (Fig. 4).

Fig. 4. Existing inundation simulation result.
In the results of modeling the number of channels that are not able to accommodate the most capacity, occurs in the simulation of the 3rd hour and the 45th minute.In the modeling, there are 5 nodes whose storage exceeds the capacity to produce backwater events at junction 19 to outfall 32 (Fig. 5).The following is a summary report to see the flooding nodes from the modeling results.In the modeling of existing conditions, there is a channel that does not work, it is marked by channels J33 -J34 and J34 -J3 that connect to the secondary channel I.The channel is declared malfunctioning because according to surveys the location of the channel cannot serve or flow the flow and its dimensions are relatively small due to its location on small roads around settlements.So that the flow that should enter the channel is modeled to directly enter the J3 -J4 channel.Table 4 shows that the maximum inundation occurred in duration between 3 to 4 hours.This result is in alignment with the actual condition in the study area.

Implementation of recharge well and rainwater harvesting
In modeling, this scenario will also reactivate channels that were previously not working, at the existing channels of J33 -J34, J34 -J3, and J6-J14 as shown in Fig. 6.
With the Low Impact Development Control modeling on SWMM 5.1, infiltration wells will be modeled with infiltration trench modeling which will be applied every 100m 2 residential areas [26].The use of infiltration trenches as infiltration wells is considered the closest approach to the concept of infiltration wells [27].The calculation of the need for infiltration wells will be applied to conditions of 10 years return period.The infiltration well that will be applied is an infiltration well that is 5 m away from the septic tank of a house and 1 m away from the foundation of the house building (the sketch of the well location is shown in Fig. 7) [28].In this study, the scenario of implementing infiltration wells that will adjust to the Government Regulation of Public Works of the Republic of Indonesia No. 11 of 2014 [24] concerning Rainwater Management in Buildings.
The number of the recharge well implemented was calculated by assuming the size of the area between the two buildings, the area of 100 m 2 as the service area of each infiltration well will represent both types of buildings.
The rainwater discharge from the roof is 0.00126 m 3 /s.Thus the dimension of the groundwater well required is 2.55 m deep and a diameter of 1.5 m.
The surface runoff discharge that flows to the drainage system without rainwater harvesting and recharges well is 0.0079306 m 3 /s.While the flow discharge with the implementation of rainwater harvesting and recharge well is 0.0043815 m 3 /s.It shows a 45% reduction in surface runoff.

Conclusion
The distribution used in data processing is the Pearson Log Type III distribution method.The calculation of rain intensity was carried out at the 2, 5, and 10-year return periods with a magnitude of 37. The results of modeling the existing conditions for the 2nd anniversary and year 5 showed that there were 2 nodes that experienced overflow, namely nodes 30 and nodes 31, and four nodes that experienced almost full capacity, namely nodes 4 (Secondary 1), nodes 14, 15 and 16 (Tertiary 1).
The assumption of land use change for vacant land to built-up land in the 10th year resulted in four nodes experiencing almost full capacity.The application of infiltration wells on every 100 m 2 of impermeable land (built-up land) requires 156 wells with a diameter of 1.5 m and a depth of 2.55 m.After the application of the infiltration well, the amount of reduction of discharge flowed into the drainage is 45%, with 0.0035491 m 3 /s of 0.0079306 m 3 /s will enter the infiltration well in every 100 m 2 (each house) of the catchment area.
29 mm/hour, 46.96 mm/hour, and 53.14 mm/hour, respectively.Based on the calculation of channel capacity, overflow occurs on secondary channels I, Tertiary I, and Tertiary III.The results of the hour-to-hour rainfall modeling show that the highest flood volume occurred at the time of the highest rain intensity in the 3 rd -45 th minute rain hour in the 6-hour rain modeling.

Table 1 .
Maximum average rainfall using Polygon Thiesen.

Table 2 .
Rainfall intensity with various return periods.

Table 3 .
Comparison of surface runoff discharge (Qsr) and existing channel capacity (Qc).

Table 4 .
The summary of the flooded node.