Assessment of Runoff Control Effect with Improved Stepped Bioretention System (ISBS) under Various Rainwater Conditions

Abstract

Stepped bioretention systems have been increasingly used for rainwater treatment in hillside areas. However, the depth of aquifer and soil permeability coefficient limit the treatment effect of runoff rainwater, resulting in a large amount of overflow water, particularly during extreme rainfall events. Here, in contrast to the ordinary stepped bioretention system (OSBS), an improved stepped bioretention system (ISBS) was developed by changing the overflow channel and the inflow and overflow were analyzed under various rainwater conditions. ISBS has high stability and the ability to control runoff rainwater. The runoff rainwater volume reduction rate reached 51.5–100% and the removal rate of suspended solid, chemical oxygen demand, total phosphorus and total nitrogen were 31.2–47.9%, 27.1–51.7%, 26.5–59.0% and 26.7–46.9%, respectively. According to the working principle of the continuous stirred tank reactor (CSTR), the permeable water concentration of other rainfall events can be predicted by using the parameters obtained from extreme rainfall events. In general, ISBS is a very promising runoff rainwater treatment technology, which can reduce the overflow quantity and recharge groundwater under various rainwater conditions.

1. Introduction

As an unconventional water sources, rainwater carries a large amount of nitrogen, phosphorus and other pollutants directly into rivers, which not only waste water resources, but also cause serious water eutrophication [1,2]. As a low-impact development technology, the stepped bioretention facility is used for rainwater treatment in hillside areas to reduce soil erosion and protect vegetation growth, drawing on the advantages of the strong erosion resistance of terraced soil [3,4,5].

The existing research shows that although the use of water-absorbing materials and stabilizers can reduce soil erosion, it will reduce the soil permeability coefficient, thereby reducing the total amount of runoff pollutants trapped by soil. Yang et al. mentioned that under the mass ratio of 0.86% stabilizer, the shear strength of root soil composite is about 60% higher than that of plain soil, and the permeability coefficient can be reduced by two orders of magnitude. A field engineering application has proved that the soil and water conservation benefits are obvious by using the solidification system to solidify loess and planting grass for ecological restoration test [6]. Misiewicz et al. selected a super absorbent polymer (SAP) to measure the change value of the permeability coefficient of different SAP soil mixtures with time, and found that the addition of SAP would significantly hinder the flow rate of water in soil, thus reducing the permeability coefficient by several orders of magnitude [7]. Some studies have shown that the growth of plants can delay the process of medium plugging, and appropriately alleviate the above problems caused by the reduction of soil permeability coefficient. Fu et al. used a small experimental constructed wetland system to study the accumulation dynamics of various organic components in the plugging process, as well as their correlation with the medium permeability coefficient. The results showed that the growth of plants delayed the medium clogging process of wetlands [8]—the role of plants in delaying clogging is limited, after all. How to reduce the total amount of runoff pollutants under the premise of maintaining normal operations of biological detention facilities is an urgent problem that needs to be solved [9,10].

At present, although bioretention facilities have been widely used for rainwater treatment, some objective phenomena cannot be avoided. For example, natural rainfall is unrepeatable, and the correlation of various pollutants in artificial water distribution experiments is very low. Furthermore, global warming and other factors aggravate the unpredictability of rainfall. These phenomena make it impossible to analyze the rainwater treatment effect of bioretention facilities under all rainfall conditions [11,12]. Therefore, many scholars have proposed rainwater quality models to describe the response of watershed water quality to specific rainfall events, evaluate the pollutant concentration and load in the catchment area and evaluate the treatment effect of rainwater control measures. However, there are some limitations in the application of various rainwater quality models. For example, the Storm Water Management Model (SWMM), System for Urban Stormwater Treatment and Analysis Integration (SUSTAIN) and Model for Urban Stormwater Improvement Conceptualization (MUSIC) are mainly used to simulate the whole urban rainwater drainage system. The above models all use the first-order attenuation response model to calculate the pollutant degradation, ignoring the water quality change after runoff infiltration [13,14,15,16,17]. For example, the micro pollutant quality model (MPiRe), the micro pollutant rainwater treatment unit model (STUMP) and the HYDRUS-1D model in rainwater use the continuous stirred tank reactor (CSTR) and the convection dispersion equation (CDE) to simulate the hydrology and pollutant migration and transformation processes of the single rainwater treatment facility unit, without considering the systematization of the rainwater treatment facility [18,19,20].

In this study, the following specific objectives have been achieved by establishing an improved stepped bioretention system (ISBS), consisting of three rainwater treatment units and comparing it with the ordinary stepped bioretention system (OSBS): (1) ability to explore the advantages of the ISBS in reducing the total amount of runoff pollutants; and (2) the CSTR principle, which is used to predict the advantages of ISBS in improving the quality of permeated water.

2. Material and Methods

2.1. Bioretention System Set Up

This experiment explored the advantages of ISBS in purifying runoff rainwater by constructing two different types of stepped bioretention facilities, as shown in Figure 1.

Both OSBS and ISBS are composed of three stages of rainwater treatment units, each of which is made of acrylic acid (100 cm length × 50 cm width × 100 cm height), and the height difference between two adjacent treatment units is 30 cm. Each rainwater treatment unit is composed of an aquifer, soil layer, fine sand layer, filling layer and pebble layer from top to bottom.

In OSBS, the heights of the aquifer, soil layer, fine sand layer, filling layer and pebble layer are 10 cm, 40 cm, 5 cm, 30 cm and 5 cm, respectively. When the aquifer height reaches 10 cm, the runoff rainwater from the upper treatment unit can overflow into the top of soil layer of the lower treatment unit. At the same time, the runoff rainwater from each treatment unit can infiltrate into the underground soil through filtration of the soil layer, fine sand layer, filling layer and pebble layer.

In ISBS, the heights of the soil layer, fine sand layer, filling layer and pebble layer are 40 cm, 5 cm, 30 cm and 5 cm, respectively. The height of the aquifer is adjustable, and was 5 cm and 10 cm, respectively. When the aquifer height reaches 5 cm, the runoff rainwater from the upper treatment unit can overflow into the top of the filler layer of the lower treatment unit. When the aquifer height reaches 10 cm, the runoff rainwater from the upper treatment unit can overflow into the top of the soil layer of the lower treatment unit. At the same time, the runoff rainwater from each treatment unit can infiltrate into the underground soil through the filtration of the soil layer, fine sand layer, filling layer and pebble layer.

The soil used in this study is the alluvial soil of Xinxiang City, Henan Province, with a permeability coefficient of 5 × 10⁻⁶ m/s. Chinese pennisetum herb planted in the soil layer has a good removal effect on dissolved inorganic nitrogen and dissolved total phosphorus [21]. Over the filling zone is a transition layer made of well-graded sand to prevent soil from washing into the filling zone. The filler is composed of activated alumina with a good phosphorus removal effect. The main physicochemical properties of activated alumina are shown in

Table 1. Main physicochemical properties of activated alumina.

Physical indicators	Bulk density (g/mL)	Specific surface area (m2/g)	Pore volume (mL/g)	Loss on ignition (%)
	0.75	≥260	≥0.35	≤7.0
Chemical composition	Al2O3 (%)	SiO2 (%)	Fe2O3 (%)	Na2O (%)
	≥92	≤0.10	≤0.04	≤0.40

[22]. The upper sampling ports are set up on the side of each treatment unit to control the depth of the aquifer.

2.2. Experiment Design

In this experiment, the influent water of the ordinary and improved stepped bioretention devices came from the runoff rainwater of the same roof of Henan Normal University. The longitude and latitude of the roof in this experiment are 113°55′15″ and 35°19′59″. Due to the large area of the existing roof, the ratio of the area of bioretention device to the catchment area is limited to a certain extent. Therefore, some bricks are used to separate the roof to ensure that the catchment area of each stepped bioretention device is 30 m² [23].

In order to reduce the impact of pollutants generated during the construction of the stepped bioretention device on the purification of runoff rainwater, tap water shall be used to flush the device before the test until the overflow water is produced by the last stage treatment unit.

After the start of the experiment, the runoff rainwater is collected from the water intake of the first stage treatment unit at fixed time points (5 min, 10 min, 15 min, 20 min, 25 min, 30 min, 40 min, 50 min, 60 min, 80 min, 100 min and 120 min). After the overflow outlet of the last stage treatment unit discharged water, we took the overflow water every five minutes. When collecting water samples, we recorded the sampling date, start time and end time, and pretreat the samples for experimental analysis.

2.3. Data Analysis

Water samples were analyzed within 24 h. Suspended solid (SS) was measured according to the standard methods for the Examination of Water and Wastewater. Chemical oxygen demand (COD), total phosphorus (TP) and total nitrogen (TN) were measured according to standard methods by a UV–vis spectrophotometer (DR6000, HACH, Loveland, CO, USA) [24]. Event mean concentration was calculated and used to determine the different pollutants’ removal efficiency. In this study, the effluent concentration needs to meet Chinese “Environmental quality standards for surface water”, and the different water quality standard limits are presented in

Table 2. Quality of Chinese environmental quality standards for surface water.

Parameter	Concentration (mg/L)
	Category Ⅰ	Category Ⅱ	Category Ⅲ	Category Ⅳ	Category Ⅴ
TN	0.2	0.5	1.0	1.5	2.0
TP	0.02	0.1	0.2	0.3	0.4
COD	15	15	20	30	40

.

In this study, the working principle of CSTR is used to simulate the water quality control effect of the stepped bioretention system, composed of three-stage treatment units, and their governing equations are shown below [25,26]

$$\frac{d\left(MC\right)}{dt}=\frac{MdC}{dt}+\frac{CdM}{dt}=\left(Q_i·C_i\right)-\left(Q_0·C_0\right)-K·C·M\pm L$$

(1)

where M represents water volume, C represents concentration in the mixed volume, $Q_i$ represents inflow rate, $C_i$ represents concentration of the influent, $Q_0$ represents outflow rate, $C_0$ represents concentration of the effluent, K represents decay constant and L represents source (or sink).

3. Results and Discussion

3.1. Effect of Different Stepped Bioretention Facilities on Runoff Rainwater

3.1.1. Runoff Rainwater Volume Control

In order to analyze the runoff rainwater control effect of two different stepped bioretention systems under different rainfall conditions, ten rainfall events were monitored in this study. The rainfall data and runoff rainwater volume of different sampling points were shown in

Table 3. Rainfall data and runoff volume of different sampling points.

Rainfall Date	Rainfall Amount (mm)	Rainfall Duration (min)	Antecedent Dry Days (d)	Runoff Volume (L)				Runoff Reduction Rate (%)
				Water Inlet		Water Outlet		A	B
				A	B	A	B
5.10	10.5	98	15	267	280	0	0	100	100
5.23	17.5	85	12	458	461	228	122	50.2	73.5
6.1	22.8	51	8	597	602	397	292	33.5	51.5
6.14	5.1	35	12	131	127	0	0	100	100
6.17	2.3	46	2	58	60	0	0	100	100
6.29	15.7	159	11	398	404	128	26	67.8	93.6
7.25	8.9	226	25	235	231	0	0	100	100
7.31	25.3	198	5	654	649	354	254	45.9	60.9
8.8	21.3	157	7	556	561	258	166	53.6	70.4
8.17	8.2	178	8	211	216	0	0	100	100

Note: A represents ordinary stepped bioretention system (OSBR) and B represents improved stepped bioretention system (ISBR).

.

According to the data in Table 3, when the rainfall is less than or equal to 10.5 mm, neither of the two stepped bioretention facilities produces overflow water. The research of Yang et al. and Donjadee et al. show that with the increase of rainfall intensity, the runoff initiation time durations decreased and the runoff rate increased. At the same time, with the increase of rainfall amount, runoff volume increases, and thus increases soil loss [27,28]. By comparing the runoff reduction volume of two stepped bioretention facilities, the runoff control effect of ISBS is better than that of OSBS.

By comparing the rainfall data on May 23 and June 29, as well as the rainfall data on June 1 and August 8, it could be seen that with the increase of rainfall duration, soil pores are gradually filled with rainwater. This would lead to the gradual decrease of the soil permeability coefficient and finally reach a stable value, so the soil layer is no longer the main factor affecting infiltration, and thus reducing the advantages of ISBS in runoff control [29]. The above experimental phenomena are similar to the research of Pan et al., which shows that with the increase of the saturated permeability coefficient of in-situ soil, the underdrain outflow weakens, the exfiltration volume increases and the runoff control effects improve [30]. By comparing the rainfall data on June 29 and August 8, with the increase of rainfall, the runoff rainwater stored in the aquifer cannot penetrate in time, resulting in a large amount of overflow water. Compared with OSBS, the runoff rainwater stored in the ISBS aquifer can directly overflow to the top of the filling layer, which can effectively avoid the adverse impact of the permeability coefficient of the soil layer. Therefore, the runoff reduction volume of ISBS reached the optimal value in the short-term heavy rainfall event (51 min, 22.8 mm), and its runoff reduction volume is 110 L more than that of OSBS.

3.1.2. Runoff Rainwater Quality Control

Among the ten rainfall events mentioned above, only five rainfall events generated measurable runoff from the last-stage treatment unit. This study selected the five typical rainfall events for runoff rainwater quality analysis. The concentration changes curve of SS, COD, TP and TN was shown in Figure 2.

Through the comparative analysis of the data in Figure 2, it can be seen that the influent concentration of the two stepped bioretention facilities is somewhat different. The reason for this is that although the catchment area of the two stepped bioretention facilities is the same, the runoff coefficient and contamination degree of different catchment surfaces cannot be exactly the same.

It can be seen from the data in Figure 2 that compared with OSBS, ISBS did not reduce the removal of pollutants from runoff rainwater. Except for the rainfall events on May 23 and June 29, the effluent concentrations of COD, TP and TN in other rainfall events are lower than 40 mg/L, 0.4 mg/L and 0.8 mg/L, respectively, which meet the Category V value of Chinese “Environmental quality standards for surface water”. According to the analysis, the reason for this may be that after the construction of two bioretention facilities, the rainfall event on May 10 did not completely flush the residual pollutants on the catchment surface, resulting in the high effluent concentration on May 23. After the rainfall on June 1, although there were two rainfall events, the pollutants attached to the roof were not discharged in time, due to the small rainfall amount and its short duration. The above experimental phenomena are like those of Zhang et al. and Hou et al.; Zhang et al. established a transport model of dissolved pollutants based on two-dimensional physics and analyzed that the intensity and duration of rainfall events have significant effects on the transport rate of pollutants [31], while Huo et al. systematically analyzed the relative effects of rainfall characteristics and environmental factors on soil nutrient loss using a structural equation model (SEM). The results showed that maximum rainfall intensity over a 30-min period and rainfall depth had a positive effect on soil nutrient loss [32]. Therefore, after the roof was scoured by heavy rainfall on June 29, all the pollutants attached in the earlier stage were discharged, resulting in high effluent quality on June 29.

According to the analysis results of water quantity and water quality of the above 10 rainfall events, compared with OSBS, ISBS can not only maintain a better water quality control effect, but also increase the control amount of runoff rainwater volume. Particularly for short-duration heavy rainfall events that easily lead to soil erosion, ISBS can intercept a large amount of runoff rainwater through its own unique overflow path, reduce the discharge of pollutants into the river, and thus reduce the frequency of eutrophication. Wang et al. used the chemical properties of rainwater and the isotope analysis method to determine the transportation and source of nitrate, and concluded that reducing runoff volume should be the primary problem to control road runoff pollution [33].

3.2. Runoff Simulation Analysis of Stepped Bioretention Facility

In the above two-stepped bioretention facilities, the permeated water recharges the groundwater, which can increase water resources and prevent the decline of the groundwater level. However, groundwater recharge will cause a series of environmental problems, especially groundwater pollution. The existing research shows that if the drainage board is set at the bottom of the bioretention facility, although all the permeated water can be collected, the groundwater cannot be recharged. If perforated drain pipes are set at the bottom of the bioretention facility, although they can collect part of the permeated water, the geotextile covering the surface of the perforated drain pipe will affect the quality of permeated water [34].

In this study, the permeated water from stepped bioretention facility slowly infiltrates into the groundwater through the natural soil layer, so the quality of permeated water cannot be analyzed. The working principle of CSTR is used to infer whether the quality of permeated water meets the Chinese “reuse of urban recycling water-water quality standard for groundwater recharge”. At the same time, since the catchment area of the stepped bioretention facility is far larger than its surface area, the impact of natural rainfall can usually be ignored in the analysis. Therefore, Formula (1) can be converted into the following form

$$Q_{js}{C}_{js}-Q_{ys}{C}_{ys}-Q_{st}{C}_{st}=K{C}_{st}{M}_{kx}\pm L$$

(2)

where Q_js represents influent flow, C_js represents influent concentration, Q_ys represents overflow rate, C_ys represents overflow concentration, Q_st represents permeability discharge, C_st represents permeate concentration and M_kx represents internal pore volume of stepped bioretention facility.

When there is no overflow in the stepped bioretention facility, Formula (2) can be converted into the following form:

$$Q_{js}{C}_{js}-Q_{js}{C}_{st}=K{C}_{st}{M}_{kx}\pm L$$

(3)

$$C_{st}=\frac{Q_{js}{C}_{js}\pm L}{Q_{js}+K{M}_{kx}}$$

(4)

When there is overflow in the stepped bioretention facility, Formula (2) can be converted into the following form:

$$Q_{js}{C}_{js}-Q_{ys}{C}_{ys}-\left(Q_{js}-Q_{ys}\right)C_{st}=K{C}_{st}{M}_{kx}\pm L$$

(5)

$$C_{st}=\frac{Q_{js}{C}_{js}-Q_{ys}{C}_{ys}\pm L}{Q_{js}-Q_{ys}+K{M}_{kx}}$$

(6)

According to the data analysis in Table 3, for the rainfall with an extreme short duration and large amount and the rainfall with an extremely long duration, the overflow water of stepped bioretention facilities may be greater than the permeated water. Considering that in the existing research, the concentration of various pollutants in the permeated water is generally lower than that of the overflow water, on the premise of ignoring the permeated water, Formula (2) can be converted into the following form:

$$Q_{js}{C}_{js}-Q_{ys}{C}_{ys}\approx K{C}_{ys}{M}_{kx}\pm L$$

(7)

$$K\approx \frac{1}{n}{{\displaystyle \sum }}_{i=1}^n\left[\frac{\left(Q_{js}^{i+1}{C}_{js}^{i+1}-Q_{js}^i{C}_{js}^i\right)-\left(Q_{ys}^{i+1}{C}_{ys}^{i+1}-Q_{ys}^i{C}_{ys}^i\right)}{\left(C_{ys}^{i+1}-C_{ys}^i\right)M_{kx}}\right]$$

(8)

$$L\approx \pm \frac{1}{n}{{\displaystyle \sum }}_{i=1}^n\left(Q_{js}^i{C}_{js}^i-Q_{ys}^i{C}_{ys}^i-K{C}_{ys}^i{M}_{kx}\right)$$

(9)

The approximate values of K and L obtained from the above extreme rainfall events can provide a theoretical basis for solving the concentration of permeated water from other rainfall events.

4. Conclusions

In this study, the runoff control effect of ISBS was proposed and comprehensively evaluated. The specific conclusions are as follows:

(1) ISBS can effectively control runoff rainwater under different rainfall amounts, rainfall duration and antecedent dry days. The runoff rainwater volume reduction rate reached 51.5–100%, and the removal rate of suspended solid, chemical oxygen demand, total phosphorus and total nitrogen were 31.2–47.9%, 27.1–51.7%, 26.5–59.0% and 26.7–46.9%, respectively.

(2) The runoff control effect of ISBS is better than that of OSBS. Particularly for short- duration heavy rainfall events that can easily lead to water and soil loss, the special overflow path of ISBS intercepts a large amount of runoff rainwater, and reduces the discharge of pollutants into the river.

(3) The working principle of the CSTR is applied to the stepped bioretention system, which is composed of three treatment units. Using the parameters obtained from extreme heavy rainfall events and extremely long-duration events, the permeate water concentration of other rainfall events can be predicted, which provides a theoretical basis for groundwater recharge.