Simple method to quantify extraneous water and organic matter degradation in sewer networks

Abstract

The quantity and quality of sewage in sanitary sewer systems can be heavily affected by extraneous water such as rainfall-runoff and groundwater, as well as the degradation of organic materials along sewer pipelines, which would lead to ultra-low influent concentrations of organic matter and much larger wastewater treatment volumes for wastewater treatment plants (WWTPs). Herein, we developed a simple method to quantify the amount of extraneous water and the organic matter degradation in sewer networks based on the balance model of water flow and pollutant loads—total phosphorus (TP) and chemical oxygen demand (COD)—between the entrance and exit of the sewer pipe network. Several important indicators reflecting the performance of the sewer systems, including collected sanitary sewage, groundwater infiltration, rainfall-runoff inflow, and the overall COD degradation rate, can be easily quantified. This method was used to estimate the function of two different sewer systems (system 1 and system 2) near Lake Erhai in Yunnan Province, China. The results successfully revealed the variations in the indicators, which were consistent with the theoretical analysis results and the observations. The results can provide guidance for the operation and management of sewer systems and WWTPs, as well as for the analysis of the environmental benefits of sewer systems. Moreover, the results can be used in designing new sewer systems in similar rural areas.

---

Water impact

Sanitary sewer systems are important infrastructures for sewage collection and subsequent treatment. However, most wastewater treatment plants (WWTPs) have suffered from ultra-low influent COD and much larger wastewater treatment volumes due to the collection of excessive undesired extraneous water, such as rainwater and groundwater, and the natural biodegradation of pollutants in sewer networks, which severely reduces the removal efficiency of WWTPs. To the best of our knowledge, as yet, it is difficult and usually very expensive to accurately evaluate the operating status of a sewer network and diagnose its problems. This paper therefore proposed a simple method to quantify the amount of extraneous water and the degradation of organic matter along sewer networks based on the balance model of water flow and pollutant loads by using the online monitoring data of influent from WWTPs, daily rainfall data from meteorological stations, sewage monitoring data at the sewer system entrance, and regional groundwater monitoring data.

1. Introduction

Sewage discharge is the main cause of environmental pollution and eutrophication in water bodies.^1–3 To reduce the negative impacts of sewage discharge on the environment, large-scale sewer systems and wastewater treatment plants (WWTPs) have been built. Sewer pipes and related facilities convey sewage to WWTPs where pollutants can be effectively removed. However, most WWTPs have suffered from ultra-low influent chemical oxygen demand (COD) and much larger wastewater treatment volumes. In China, the average influent COD of WWTPs is <250 mg L⁻¹ in 15 provinces.^4,5 The average influent COD of 356 WWTPs in Guangdong Province was 159 mg L⁻¹ from 2015 to 2017.⁶ In comparison, WWTPs in developed regions such as Europe and Singapore have relatively high influent COD values (470–612 mg L⁻¹).⁵ A measuring station of a catchment in Czechia showed a minimum COD concentration of 296 mg L⁻¹, whereas the concentration in the trunk sewer was only 80 mg L⁻¹.⁷ This difference is caused by the collection of excessive undesired extraneous water such as rainwater and groundwater in sewer pipes.^8–10 In particular in developing countries, the proportion of extraneous water in many WWTPs exceeds 50%. This also leads to a significant increase in the concentration of inorganic suspended solids (ISS) in sewage, which increases the sludge volume of the WWTPs.^4,5 Moreover, organic pollutants in sewer networks undergo biodegradation,^11,12 which may considerably decrease the influent COD concentration of the WWTPs. The ultra-low influent COD concentration severely reduces the biological removal efficiency of nitrogen and phosphorus due to the low carbon/nitrogen (C/N) and carbon/phosphorus (C/P) ratios.⁵ In order to meet the discharge standard, many WWTPs afflicted by ultra-low influent COD concentration need to use an external organic carbon source (such as glucose) for N removal and a large amount of chemical flocculants for P removal, and these measures significantly increase the operating costs.⁵

Given the importance and complexity of sewer systems, several studies have been conducted to assess the performance of sewer systems and diagnose their problems. These studies mainly fall into three categories: 1) investigation of the sewer pipes and related facilities using infrared cameras,¹³ closed-circuit television systems^14,15 or chemical tracers;¹⁶ 2) simulation of the effects of specific factors on sewer systems. These studies demonstrated that the rainfall-runoff and groundwater increase the sewage quantity^17–19 and pollutant transformation, sedimentation, corrosion, and flushing impact the sewage quality;^20–23 3) simulation of the sewage flow rate using a large and high-resolution sensor dataset, or prediction of pollutant exchange by long-term monitoring in a pilot sewer system.^24–27

Although these studies have made good progress, it is still not easy to accurately assess the function of a regional sewer system, which is usually distributed over a large area and has notable spatial differences. Digital image analysis can only be conducted under dry weather conditions and is expensive.^14,15 The combined ground penetrating radar technology and stable isotope approach are complex.^28–30 In addition, most models require many parameters that are difficult to obtain directly and may have a high degree of uncertainty,^31,32 which greatly limits their applications. Therefore, it is important to establish a simple method to quantify extraneous water and pollutant degradation in sewer systems.

In this study, we proposed for the first time a simple method to quantify extraneous water and COD degradation along sewer networks based on the balance model of water flow and pollutant loads—total phosphorus (TP) and COD—between the entrance and exit of the sewer network. With this method, several important indicators reflecting the performance of the sewer systems, including collected sanitary sewage (CSS), groundwater infiltration (GWI), rainfall-runoff inflow (RRI), and the overall COD degradation rate, can be easily quantified. We used this method to evaluate two different sewer systems near Lake Erhai in Yunnan Province, China. The results successfully revealed the variations in the indicators, which were consistent with the theoretical analysis results and the observations.

2. Methods

2.1. Study area

Lake Erhai is one of the major freshwater lakes in China located at high altitudes. It is located in Dali City in Yunnan Province and belongs to the Lancang River system upstream of the Mekong River. The surface area of Lake Erhai is 250 km², with a capacity of 2.88 billion m³ and an average water depth of 10.2 m. The land use along the river basin suggests that the sources of contaminants in Lake Erhai mainly originate from towns, villages, and farmland within the catchment area. Due to the recent development in tourism, the influx of tourists has increased the potential for pollution. The regional governments have undertaken environmental protection efforts in the lake basin. To minimize the negative impacts of sewage discharge from the towns and villages within the Erhai catchment area, the Dali City government has implemented a series of projects for centralized sewage collection and treatment for towns and villages since 2015. However, when the WWTPs and sewer networks were completed and put into operation, the problems of a large amount of influent wastewater and extremely low concentration of pollutants (particularly COD) occurred, especially during the rainy season. To ensure the stable operation of WWTPs and meet the discharge standards of N and P, it is required to add glucose as a carbon source to the wastewater for N removal and adopt chemical methods to enhance P removal.

In this study, two typical sewer systems and WWTPs near Lake Erhai, Yunnan Province, China were selected for a case study (Fig. 1). System 1 is located on the west coast of Lake Erhai and mainly collects sanitary sewage from rural residences and some of the livestock farm wastewater. The sewer main pipe has a total length of 58.61 km and a pipe diameter of DN400 to DN1000, with a service area of 12.74 km² covering 73 villages (including town centers). The sewage collected by system 1 enters WWTP 1, which has a designed capacity of 10 000 m³ d⁻¹. System 2 is located on the east coast of Lake Erhai and mainly collects sanitary sewage from tourist areas and parts of rural residences. The sewer main pipe has a total length of 14.98 km and a pipe diameter of DN400 to DN1000, with a service area of 3.03 km² covering eight villages (including town centers). System 2 enters WWTP 2, which has a designed capacity of 5000 m³ d⁻¹. Systems 1 and 2 collecting sewage adopt the design of separate sewer systems, but early-built sewer networks (sewage canals) in a small number of natural villages are combined systems. In addition, sewage is treated by septic tanks before entering the network system.

Fig. 1 Distribution diagram of the sewer network systems.

According to local hydrogeological data, Dali City has rich groundwater resources amounting to 262.8 million m³ and a resource modulus of 149 000 m³ a⁻¹ km⁻². The type of groundwater in the Lake Erhai plain area is the pore water of loose rock formation. The water depth is 0–5 m, and the groundwater bearing capacity ranges from strong (seepage volume from a single well is greater than 1000 m³ d⁻¹) to medium (seepage volume from a single well is 100–1000 m³ d⁻¹). The area has a subtropical monsoon climate and an annual average temperature of 15.1 °C, a maximum average temperature in July of 20.0 °C, and a minimum average temperature in January of 8.9 °C. The mean annual rainfall is 1000 mm, of which 90% falls in the wet season between May and October, and 10% falls in the remaining months that make up the dry season.

2.2. Water quality monitoring

The average concentration of COD and TP in the original sanitary sewage (OSS) was measured on site. In late November 2018, 19 and 4 monitoring points were set up in the service areas of systems 1 and 2, respectively (Fig. 1). The manual sampling site is at the end of the branch sewer pipe, where the sewage represents the confluent foul water discharged from all the households in a village (Fig. 2). Each point was sampled at 8:00, 12:00, 16:00, 20:00, and 24:00. The samples were kept at low temperature and mixed in equal proportion to analyze the average concentration. The COD was determined using the dichromate method (HJ828-2017),³³ and TP was determined using the ammonium molybdate spectrophotometric method (GB11893-1989).³⁴

Fig. 2 Schematic diagram of the sewage collection process.

2.3. Model and basic assumptions

The process of the wastewater collection and extraneous water inflow/infiltration in the sewer networks can be described as akin to a process of water mixing from different sources. As there may be an overflow or leakage loss during the transmission process of the wastewater in the network the water quantity and quality at the end of the network serve as the basis for analyzing the overall running status of the sewer network. When all of the collected wastewater enters the WWTP, the influent wastewater of the WWTP is then equivalent to the flow at the end of the sewer network. Therefore, the monitoring data of the influent wastewater of the WWTP can be used to analyze the overall running status of the sewer network.

For a selected sanitary sewer network system, during the continuous n-day monitoring period, the total influent wastewater (IWW) of the WWTP mainly consists of three parts: collected sanitary sewage (CSS), groundwater infiltration (GWI) and rainfall-runoff inflow (RRI). The process of sewage transport is also the process of pollutant load transport. After considering possible factors such as degradation, deposition in pipes and resuspension, the pollutant load in the pipeline also obeys a mass conservation relationship.

Under the effect of heterotrophic microorganisms, the organic matter in sewer systems decomposes into CO₂, CH₄ and other gases, resulting in the loss of the COD load in the sewage. Different from COD, under the effect of microorganisms, the proportion of gaseous phosphorus substances such as phosphine escaping from the sewage is very small, and phosphorus substances still remain in the sewage in a dissolved or solid state. Therefore, the total phosphorus (TP) load does not change significantly by degradation over a long period of time. In other words, the degradation loss of COD needs to be considered in the process of sewage transport in the pipeline, whereas the degradation loss of TP can be ignored. TN is also an important pollutant index in the sewage.³⁵ However, TN may have a significant degradation loss (for example, under hypoxic conditions with high nitrate concentrations). Conservative substances are used as indicators in some studies,^36–38 but the problem is that these substances are not routinely measured in the WWTP and their costs are often high. Therefore, taking into account the availability and the simplicity of the method, COD and TP are selected as indicators for quantification of extraneous water and organic matter degradation in sewer networks.

In addition to the degradation loss of pollutants, the deposition and resuspension of pollutants will occur in the pipeline.^39–41 However, during a long period of time when both deposition and resuspension of pollutants take place, the loss proportion induced by net deposition in the pipeline is relatively small as compared to the total amount of pollutants transported by the sewer network. Therefore, the effect of net deposition can be ignored normally when establishing the balance equations of pollutant loads.

Based on the above analysis, if the total amount of COD degradation loss in the study period is M_DCOD, then according to the principle of mass conservation, the following equations are obtained:

where V_IWW, M_ITP and M_ICOD are the total volume, TP mass and COD mass of WWTP influent wastewater during the n days of the study period, respectively. They can be obtained by the following equations , and , where V_IWWi, C_ICODi and C_ITPi are the volume, average TP concentration and COD concentration of WWTP influent on the ith day during the study period, respectively. V_CSS, V_GWI and V_RRI are the total CSS volume, GWI volume and RRI volume during the n days of the study period, respectively. C_OSCOD and C_OSTP are the average concentrations of COD and TP in the original sanitary sewage during the study period, respectively; C_GCOD and C_GTP are the average concentrations of COD and TP in the groundwater in the distribution area of the network, respectively; C_RCOD and C_RTP are the average concentrations of COD and TP in the rainfall-runoff, respectively.

As compared with sanitary sewage, the concentration of COD in the groundwater is generally negligible. Eqn (1) can be simplified as:

For the sewer networks dominated by separate sewers, the total amount of COD load from the rainfall-runoff is relatively small compared to that from the sanitary sewage. Thus, it can be assumed that the degradation loss of COD is all from the sanitary sewage. In this case, eqn (2) can be expressed as:

In eqn (3), α_COD is the overall degradation correction coefficient of the COD load in the process of sanitary sewage transport during the study period.

The degradation of COD in sewer pipes can be simplified as a first-order kinetic mode,^42,43 and the degradation loss of COD varies with the length of the pipes without considering the difference in flow velocity. If the overall COD degradation rate in 1 km length of the pipeline is K_COD, then α_COD can be expressed as follows:

where M_CODij is the COD load of sewage discharge unit j on the ith day; L_j is the length of the sewage pipelines from unit j to the WWTP; s is the total number of sewage discharge units.

To simplify the calculation, we assumed that the per capita COD load discharge for each sewage discharge unit (a natural village in this case) is the same. Then, the population equivalent can be used to replace the COD load.

where P_j is the population equivalent of sewage discharge unit j during the study period.

In fact, rainfall-runoff inflow in sewer systems only occurs when there is a relatively intense rainfall event. During the continuous n-day monitoring period, there are m days of rainfall generating surface runoff. After the rainfall event, the influence of surface runoff on the pipe network system will continue for a certain period. Afterwards, there will be a maximum of k days of rainfall-runoff that will flow into the sewer system (i.e., maximum rainfall-runoff impact period is m + k days). The total volume of RRI in the influent wastewater of the WWTP can be expressed as follows:

where V_RRIi is the volume of RRI in the influent wastewater of the WWTP on the ith day of the study period (n days); V_RRIj is the volume of RRI in the influent wastewater of the WWTP on the jth day of the maximum rainfall-runoff impact period (m + k days).

Although we cannot directly quantify the RRI on the jth day, we can estimate V_RRIj by comparing the differences in the quantity and quality of the influent wastewater of the WWTP between the jth day and the non-rainfall-runoff impacted days before and after the jth day because the impact of rainfall-runoff is abrupt and shorter than that in the entire monitoring period. The impact of rainfall-runoff on the influent wastewater of the WWTP is twofold: an increase in the influent wastewater and a decrease in the concentration of pollutants. Under special circumstances, the influent wastewater of the WWTP on rainy days may decrease, whereas the pollutant concentrations may increase. However, these are rare occurrences, and the RRI in the influent wastewater on these special days can be ignored. Therefore, we can calculate the RRI in two cases. In case 1, when the rainfall totals are low and the sewer system does not overflow, the increased water volume in the influent wastewater corresponds to the RRI amount. In this case, the amount of influent wastewater on the day before and after the rainfall-runoff impact days can be used to estimate the RRI (i.e., V_RRIj (case 1)) by using the linear interpolation method.

In case 2, assuming a large rainfall amount causing an overflow in the sewer system, to avoid underestimating the RRI in the influent wastewater, the changes in pollutant concentrations should be considered. When the wastewater flow in the rainy season does not exceed the delivery capacity of the main sewers and the overflow of rainwater and sewage mainly occurs in the branch sewer pipes, we can estimate the GWI on the rainfall-runoff impact days by using the linear interpolation method. Using eqn (1) and (3), the RRI (i.e., V_RRIj (case 2)) can also be estimated.

As we cannot determine the overflow on each rainy day in the study area based on the rainfall monitoring data, we take the larger value in the two cases as V_RRIj. If the results of V_RRIj in the two cases are zero or negative, the RRI in the influent wastewater can be ignored.

V_RRIj = max(V_RRIj(case 1), V_RRIj(case 2), 0) (7)

Considering the underlying surface conditions within the service scopes of systems 1 and 2 and the ground evaporation conditions of Dali City, the impact of rainfall-runoff on the influent wastewater of the WWTP should only be considered when the rainfall on that day is ≥5 mm. As the catchment areas of systems 1 and 2 are small and the distance of the sewage transport is short, when the rain stops, the impact of rainfall-runoff on the influent wastewater does not last for more than one day (but the rainwater can still have a continuous impact on the influent wastewater by infiltration). Therefore, all the days with daily rainfall ≥5 mm and the following day can be taken as the maximum rainfall-runoff impact period. Accordingly, the total volume of RRI during the study period can be calculated.

Using eqn (1) and (3), the total amount of the collected sanitary sewage V_CSS and the total amount of groundwater infiltration V_GWI during the study period can be obtained. Furthermore, the overall COD degradation correction coefficient (α_COD) and the overall COD degradation rate per kilometer length of the pipeline (K_COD) can be calculated using equations eqn (5) and (6).

2.4. Source of data and determination of parameters

The influent wastewater flow rates of WWTP 1 and WWTP 2 and the rainfall in the same period are presented in Tables S1 and S2 in the ESI.† The average concentrations of COD and TP in the OSS within the service areas of sewer networks, C_OSCOD and C_OSTP, can be obtained using the statistical methods after on-site sampling and monitoring. The related data are presented in Table S3 in the ESI.† The average concentration of TP in groundwater (C_GTP) can be obtained using statistical methods on groundwater monitoring data. The average concentrations of COD and TP in rainfall-runoff, C_RCOD and C_RTP, can be obtained via on-site monitoring, simulations, or comparisons with similar projects. The data types are listed in Table 1.

Table 1 The collected data

Type	Origin	Time series	Data
The influent volume and quality of the WWTPs	Online monitoring system of WWTPs	July 1, 2018–June 30, 2019	See Table S1†
Rainfall data	Meteorological monitoring station of Dali City	July 1, 2018–June 30, 2019	See Table S2†
C OSCOD and COSTP	Measured on-site	In late November, 2018	See Table S3†
C GTP	Historic data	2017	The average concentration of TP in the groundwater around Lake Erhai is 0.122 mg L^−1.
C RCOD and CRTP	Reference	2016	The average COD and TP concentrations of rainfall-runoff in the study areas are estimated to be 100 mg L^−1 and 0.5 mg L^−1, respectively^44,45

---

3. Results and discussion

3.1. Volume of rainfall-runoff collected by sewer systems

During the one-year study period, the influent wastewater flow rate of WWTP 1 (WWTP 2) was 3594–16 768 m³ d⁻¹ (1047–9438 m³ d⁻¹), and the average flow rate was 10 065 m³ d⁻¹ (3170 m³ d⁻¹). There were 111 rainy days in Dali City, with a total rainfall of 782.5 mm. July to October 2018 and June 2019 were typical rainy months in Dali City, with a total of 77 rainy days, and a cumulative rainfall of 640.8 mm, accounting for 81.9% of the total rainfall. November 2018 to May 2019 were typical dry months, with a cumulative rainfall of 141.7 mm, accounting for 18.1% of total rainfall. Although the rainfall frequency and volume in the dry season were smaller in the study period, there were some dates with high rainfall intensity. In all rainfall days, the cumulative rainfall with daily rainfall ≥5 mm was 656 mm, accounting for 83.8% of the total rainfall.

Therefore, considering the days when the daily rainfall is ≥5 mm with the next day as the filtering criteria and using eqn (7), the RRI in the influent wastewater of WWTPs 1 and 2 during the maximum rainfall-runoff impact days can be estimated (Fig. 3). The estimated RRI values have a good response to the occurrence of ≥5 mm rainfall, especially short-term rainfall. Therefore, the method can effectively estimate the RRI in the influent wastewater of the WWTP during the study period, considering the potential differences in the rainfall amounts among sampling sites, as well as the possible human interventions in sewer systems.

Fig. 3 Estimations of the rainfall-runoff inflow in the influent wastewater of WWTP 1 (a) and WWTP 2 (b) during the maximum rainfall-runoff impact days.

The total volume of RRI in the influent wastewater of WWTP 1 (WWTP 2) during the study period is 79 519 m³ (42 085 m³) with an average flow rate of 218 m³ d⁻¹ (115 m³ d⁻¹), accounting for 2.2% (3.6%) of mixed wastewater. This result indicates that the average flow rate of RRI for the two WWTPs is not large, which is consistent with the fact that systems 1 and 2 adopted separate systems.

On the one hand, the two selected systems (systems 1 and 2) adopted a drainage system dominated by separate systems; on the other hand, the possible overflow of rainwater and sewage in the pipe network have been deducted from the influent monitoring data of the WWTP; thus, the calculated rainfall-runoff inflow is very small compared to the annual wastewater volume.

3.2. Average flow rate of collected sanitary sewage and groundwater infiltration

The monitoring data for the influent wastewater volume and quality are obtained from the online monitoring system (Fig. 2), and the monitoring period is from July 1, 2018 to June 30, 2019. The data are presented in Table S1 in the ESI.† When the four parameters V_RRI, C_RTP, C_OSTP, and C_GTP are known, by using the monitoring data for the flow rate and TP concentrations at the inlet of the WWTPs (Fig. 4), the total volume (or average flow rate) of CSS and GWI during the study period can be calculated using eqn (1) and (3). As a result, the proportion of RRI, GWI, and CSS in the mixed wastewater collected by the sewer systems can be obtained (Fig. 4).

Fig. 4 Composition of the mixed wastewater collected by the sewer systems during the one-year study period.

In the mixed wastewater collected by system 1 (system 2), the average flow rate of the collected sanitary sewage is 5005 m³ d⁻¹ (1697 m³ d⁻¹), accounting for 49.7% (53.5%) of the mixed wastewater, and the average flow rate of groundwater infiltration is 4842 m³ d⁻¹ (1358 m³ d⁻¹), accounting for 48.1% (42.8%). Therefore, the amount of groundwater infiltration is essentially the same as that of sewage during the sewage transport. The average flow rate of groundwater infiltration per unit pipe length for systems 1 and 2 is 82.6 and 90.7 m³ d⁻¹ km⁻¹, and the average flow rate of groundwater infiltration per unit service area for systems 1 and 2 is 380 and 448 m³ d⁻¹ km⁻², respectively.

The quantification of extraneous water is important and useful for both the pipeline network rehabilitation and the evaluation of real environmental benefits of sewage collection and treatment systems. Therefore, it is a key issue for both engineering practice and water environment protection. Because the pipe network system is buried under the ground with many spots and quite wide areas, it is difficult to measure the amount of extraneous water directly in practice. Relevant researchers often use the concentration of monitoring pollutants to determine the amount of extraneous water. Compared with previous short-term or local studies,^35–38 the method established in this study realized the calculation and analysis of the annual amount of extraneous water mixed in the system based on the online monitoring data of the WWTP and the appropriate calculation parameters. Because the monitoring data are easily updated, this method is very suitable for the annual follow-up evaluation of the effectiveness of repairs on the pipe network system. According to the results of this study, extraneous water accounts for about (or nearly) half of the mixed sewage collected by system 1 and system 2. This indicated that although the sources of pollution in the service area of the two systems are different, that is, the calculation parameters are quite different, the defect degrees of the pipe network systems are similar. This can be explained by the fact that the sewage main pipe system of the two areas was newly built in the same project and distributed in the near shore area of Lake Erhai. Thus, this indicates that our calculation results are in good agreement with the construction situation of the project.

The proportion of extraneous water is only one aspect, and the more important index is the groundwater infiltration flow per unit pipe length. According to the Chinese “Code for Design of Outdoor Wastewater Engineering” (2006), the groundwater infiltration should be considered in the area with high groundwater levels. It can be measured using the infiltration amount per unit pipe length and per unit pipe diameter, or using 10–15% of the total amount of average daily total sanitary sewage and industrial wastewater. In the urban sewage pipe network of Guangzhou, China, the GWI values are 15.20, 145.15, and 710.13 m³ d⁻¹ km⁻¹, when the diameters of the reinforced concrete pipes are 300–600, 700–1000, and 1200–2000 mm, respectively.⁴⁶ The weighted average value of the GWI of the newly constructed concrete sewage pipeline in the Guangzhou sewage system is 210 m³ d⁻¹ km⁻¹. The GWI of the representative concrete pipeline with a diameter of 450–2460 mm in Shanghai is 50–166 m³ d⁻¹ km⁻¹. According to “Sewer Facilities Design Guide and Interpretation” (Japan Sewage Works Association, 2001), the GWI should be 10–20% of the daily maximum total sewage volume, based on the previous data. The British “Wastewater Treatment Plant” BSEN12255 recommends the estimation of GWI by observing the nighttime flow of the existing pipelines. According to the German Wastewater Engineering Association (2000), the infiltration amount should not exceed 0.15 L s⁻¹ hm⁻² (1296 m³ d⁻¹ km⁻²). The infiltration water volume in the United States is specified to be within the range of 0.01–1.0 m³ d⁻¹ mm⁻¹ km⁻¹ or 20–2800 m³ d⁻¹ km⁻².

In both systems 1 and 2, the amounts of the groundwater infiltration are generally within the range of the design standards or the relevant study results. However, as the service areas of systems 1 and 2 are both rural areas, the population density within the coverage of the pipe networks is smaller than in urban areas, so is the total amount of sewage discharge. This is the main reason for the significantly higher proportion of GWI.

In order to analyze the seasonal trends in GWI and the sewage volume, we shortened the calculation time to 30 d (n = 30) and obtained the average daily GWI amount and sewage flow before and after a certain date (Fig. 5). Fig. 5 also shows the average daily rainfall in 30 d. From July 2018 to June 2019, the average groundwater infiltration of systems 1 and 2 shows significant downward trends with large fluctuations during the decline. Compared with the average rainfall amounts in the same period, the trends in the average groundwater infiltration of systems 1 and 2 are similar to the regional average rainfall patterns, but with a time lag.

Fig. 5 Variation trends of 30 days average groundwater infiltration, collected sanitary sewage and rainfall of system 1 (a) and system 2 (b).

The Pearson correlation coefficient between the average groundwater infiltration of system 1 and the average rainfall 32 d before reaches a maximum of 0.851 (P < 0.001). The Pearson correlation coefficient between the average groundwater infiltration of system 2 and the average rainfall of 20 d before reaches a maximum of 0.768 (P < 0.001). Therefore, the response time lags between the groundwater infiltration and rainfall are 32 d and 20 d for systems 1 and 2, respectively. This result is consistent with the previous monitoring results of groundwater levels near Lake Erhai,⁴⁷ which indicates that the variations in groundwater levels caused by the seasonal rainfall may be the most important factor affecting the amount of GWI in systems 1 and 2. Moreover, the 30 d average of the influent wastewater amounts and pollution loads of the WWTP can better reveal the seasonal variations in groundwater infiltration in the area, while minimizing the impact of the short-term monitoring data fluctuations on the results.

The rainfall is a very important factor in determining the amount of GWI. For the two systems we studied, occasional rainfall events can still have a significant and long-lasting impact on GWI even in the dry season. This indicates that it is necessary to cover a complete one-year cycle in the research period if the sewer rehabilitation activities are evaluated by GWI. In this way, the impact of the uncertainty of rainfall events can be effectively reduced.

During the entire monitoring period, the average CSS of systems 1 and 2 showed an initially increasing trend followed by fluctuations. In fact, the CSS of the sewer systems is affected by many factors such as the seasonal changes in sewage discharge, sewage leakage, overflow during the rainy season, and the pipe network repair activities. In July and August, the amount of CSS was low, which indicates that the sewer networks may have significant sewage overflow in the rainy season. Based on the field survey of the sewer systems during the rainy season, the overflow of rainwater and sewage mainly occurred in the villages due to the low delivery capacity and blockage of branch sewer pipes, whereas the overflow of the main sewer pipes was not observed even on the days with maximum rainfall.

The results sufficiently reflected the impact of holidays on the amount of sanitary sewage collected by the sewer network system. For example, the peak value in early October 2018 corresponds to the China's National Day holiday, and the peak value in February 2019 corresponds to the Spring Festival. Therefore, the results in Fig. 6 can be used to further analyze the main factors affecting the sewage collection volume. When the sewage discharge is obtained based on the regional population data or daily water supply data, some important parameters such as the sewage collection rate and overflow/leakage rate can also be calculated. At present, it is still difficult to realize the real-time tourism statistics and accurately monitor the water supply volume of the tap water in the study area. These parameters are necessary to evaluate the pipe network maintenance projects and the operation and management of sewer network systems and WWTPs, and to protect the regional water quality.

Fig. 6 Analysis of the influence of degradation and extraneous water dilution on the average COD concentration of sewage.

3.3. COD degradation rate of the sewer systems

When the values of V_RRI, C_RCOD, C_OSCOD, and V_CSS are known, the overall degradation correction coefficient α_COD of the COD load can be obtained using eqn (5). The calculation results show that the COD degradation correction coefficients of systems 1 and 2 are 0.622 and 0.840, respectively. Combining the demographic data and the transmission distance of sewage, the approximate value of the COD degradation coefficient of sewage flowing through the unit length of the pipeline can also be calculated, and the results show that the overall COD degradation rates are 0.0731 km⁻¹ and 0.0534 km⁻¹, respectively.

The effect of extraneous water dilution and degradation on the average concentration of COD in wastewater was further discussed (Fig. 6). This result is valuable for the operation of WWTPs and the assessment of sewer network improvement, especially when it is difficult to accurately determine the amount of sewage discharge in the service area of a sewer network. Although the influence of sewage flow velocity in different areas is not considered in the discussion of COD degradation, the COD degradation correction coefficient and COD degradation rate per unit pipe length are still important for guiding the operation and management of the WWTPs, as well as the design of the pipe network system in similar areas. Compared with previous field studies,^43,48,49 the overall degradation rates of COD in systems 1 and 2 are at a relatively high level, and the overall degradation rate of COD in system 1 is higher than that in system 2. According to the monitoring results of the groundwater quality in 2017, the average concentration of nitrate in groundwater around Lake Erhai was 6.19 mg_N L⁻¹. Therefore, the high COD degradation rate may be related to the denitrification in groundwater.⁴⁹ In other words, groundwater does not only have a significant dilution effect on the concentration of organic matter, but it also further reduces the concentration of organic matter in the influent wastewater of the WWTPs through denitrification. This is highly unfavorable for the operation of WWTPs, especially in the rainy season.

In addition to the influence of sewage flow velocity, the difference in the COD degradation rate per unit pipe length is related to the degradability of COD and the aerobic/anaerobic conditions during sewage transportation.^50,51 In the service area of system 1, the villages are scattered and the sewage discharge amount per village is small. The foul water remains in the septic tank for a relatively long time, and the sewer pipes are mostly in an aerobic state due to numerous pipe joints. In system 2, the main sewage discharge source is located in the central market town where the tourist population is concentrated and the municipal facilities are adequate. Thus, the sewer pipes of system 2 have a better sealing property and are in a relatively anaerobic state.

The excessively low concentration of influent organic matter will have a significant negative impact on the biochemical treatment, and increase the cost of denitrification and phosphorus removal in the WWTP. Therefore, an important purpose of pipe network repair activities is to improve the concentration of influent organic matter of WWTPs. However, the degradation loss of organics in a pipe network system is often ignored by operation managers and engineers. The results showed that the degradation loss of organics (indicated by COD) is obvious in the pipe network system, and the degradation loss rate is mainly related to the layout of the system and the specific environmental conditions. Without considering the degradation loss of COD in rainfall runoff, the degradation loss rates of COD in sewage collected by systems 1 and system 2 were 37.8% and 16.0%, respectively. This suggested that the degradation loss of organic matter should be an important reference factor for future sewage collection and treatment engineering design, especially in rural areas where the pollution sources are relatively dispersed and the sewage transport distances are relatively long.

4. Conclusion

A method was established to quantify the RRI, GWI and organic matter degradation in sewer networks using rainfall data, TP and COD online monitoring data of WWTPs. Based on the method, the operating status of two different sewer systems near Lake Erhai was estimated, and the relationship between the GWI and the rainfall was further analyzed, as well as the influence of degradation and extraneous water dilution on the reduction of influent COD concentration of WWTPs. The results demonstrated that the extraneous water accounted for about half of the mixed sewage collected by system 1 and system 2. However, the GWI per unit length was still within the design specification, and the relatively high proportion of extraneous water was mainly related to the small total amount of sewage discharge in the study area. The amount of GWI varied with the rainfall, but there was an obvious lag. Organic matter indicated by COD exhibited obvious degradation in the network system, which is mainly related to the layout of the system and the specific environmental conditions. The results can provide guidance for the operation and management of sewer systems and WWTPs, as well as for the analysis of the environmental benefits of sewer systems. Moreover, the results can be used in designing new sewer systems in similar rural areas.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

The authors would like to acknowledge the Yunnan Key Research and Development Program (No. 2018BC001).

References

1. A. Montserrat, L. Bosch, M. A. Kiser, M. Pocha and L. Corominas, Using data from monitoring combined sewer overflows to assess, improve, and maintain combined sewer systems, Sci. Total Environ., 2015, 505, 1053–1061.
2. U. Rathnayake and A. H. M. Faisal Anwar, Dynamic control of urban sewer systems to reduce combined sewer overflows and their adverse impacts, J. Hydrol., 2019, 579, 124150–124163.
3. N. X. Cuong, T. C. Phuong, H. V. Hoan, N. T. Phuong, S. W. Chang, D. DucNguyencd, W. S. Guo, A. Kumar, D. Ducla and Q. V. Bach, Combined biochar vertical flow and free-water surface constructed wetland system for dormitory sewage treatment and reuse, Sci. Total Environ., 2020, 713, 136404–136413.
4. Y. S. Cao, J. G. Tang, M. Henze, X. P. Yang, Y. P. Gan, J. Li, H. Kroiss, M. C. M. van Loosdrecht, Y. Zhang and G. T. Daigger, The leakage of sewer systems and the impact on the ‘black and odorous water bodies’ and WWTPs in China, Water Sci. Technol., 2019, 79(2), 334–341.
5. Y. S. Cao, M. C. M. Loosdrecht and G. T. Daigger, The bottlenecks and causes, and potential solutions for municipal sewage treatment in China, Water Pract. Technol., 2020, 15(1), 160–169.
6. T. Zhu, G. Liu, S. Ma and S. Yuan, Analysis of influent water quality characteristics of urban wastewater treatment plants in Guangdong province, Ziyuan Zonghe Liyong, 2019, 37(5), 67–69 (In Chinese).
7. V. Bareš, D. Stránský and P. Sýkora, Evaluation of sewer infiltration/inflow using COD mass flux method: case study in Prague, Water Sci. Technol., 2012, 66(3), 673–680.
8. C. Karpf, S. Hoeft, C. Scheffer, L. Fuchs and P. Krebs, Groundwater infiltration, surface water inflow and sewerage exfiltration considering hydrodynamic conditions in sewer systems, Water Sci. Technol., 2011, 63(9), 1841–1848.
9. P. Staufer, A. Scheidegger and J. Rieckermann, Assessing the performance of sewer rehabilitation on the reduction of infiltration and inflow, Water Res., 2012, 46(16), 5185–5196.
10. S. Thorndahl, J. D. Balling and U. B. B. Larsen, Analysis and integrated modelling of groundwater infiltration to sewer networks, Hydrol. Processes, 2016, 30(18), 3228–3238.
11. P. Jin, B. Wang, D. Jiao, G. Sun, B. Wang and X. C. Wang, Characterization of microflora and transformation of organic matters in urban sewer system, Water Res., 2015, 84, 112–119.
12. T. Pai, S. Wang, H. Lo, L. Chen, T. Wan, M. Lin, C. Lin, P. Yang, W. Lai, Y. Wang and T. Lu, A simulation of sewer biodeterioration by analysis of different components with a model approach, Int. Biodeterior. Biodegrad., 2017, 125, 37–44.
13. M. Lepot, K. F. Makris and F. H. L. R. Clemens, Detection and quantification of lateral, illicit connections and infiltration in sewers with Infra-Red camera: Conclusions after a wide experimental plan, Water Res., 2017, 122, 678–691.
14. S. S. Kumar, D. M. Abraham, M. R. Jahanshahi, T. Iseley and J. Starr, Automated defect classification in sewer closed circuit television inspections using deep convolutional neural networks, Autom. Constr., 2018, 91, 273–283.
15. M. Beheshti and S. Saegrov, Detection of extraneous water ingress into the sewer system using tandem methods - a case study in Trondheim city, Water Sci. Technol., 2019, 79(2), 231–239.
16. Z. C. Zhao, H. L. Yin, Z. X. Xu, J. Peng and Z. W. Yu, Pin-pointing groundwater infiltration into urban sewers using chemical tracer in conjunction with physically based optimization model, Water Res., 2020, 175, 115689–115698.
17. J. Houhou, B. S. Lartiges, C. France-Lanord, C. Guilmette, S. Poix and C. Mustin, Isotopic tracing of clear water sources in an urban sewer: A combined water and dissolved sulfate stable isotope approach, Water Res., 2010, 44(1), 256–266.
18. C. Karpf and P. Krebs, Quantification of groundwater infiltration and surface water inflows in urban sewer networks based on a multiple model approach, Water Res., 2011, 45(10), 3129–3136.
19. G. Dirckx, A. Fenu, T. Wambecq, S. Kroll and M. Weemaes, Dilution of sewage: Is it, after all, really worth the bother?, J. Hydrol., 2019, 571, 437–447.
20. A. N. A. Schellart, S. J. Tait and R. M. Ashley, Estimation of uncertainty in long-term sewer sediment predictions using a response database, J. Hydraul. Eng., 2010, 136, 403–411.
21. G. Shahsavari, G. Arnaud-Fassetta and A. Campisano, A field experiment to evaluate the cleaning performance of sewer flushing on non-uniform sediment deposits, Water Res., 2017, 118, 59–69.
22. A. Campisano, C. Modica, E. Creaco and G. Shahsavari, A model for non-uniform sediment transport induced by flushing in sewer channels, Water Res., 2019, 163, 114903–114916.
23. G. Jiang, M. Zhou, T. H. Chiu, X. Sun, J. Keller and P. L. Bond, Wastewater-Enhanced Microbial Corrosion of Concrete Sewers, Environ. Sci. Technol., 2016, 50(15), 8084–8092.
24. L. García, J. Barreiro-Gomez, E. Escobar, D. Téllez, N. Quijano and C. Ocampo-Martinez, Modeling and real-time control of urban drainage systems: A review, Adv. Water Resour., 2015, 85, 120–132.
25. S. Kroll, T. Wambecq, M. Weemaes, J. Van Impe and P. Willems, Semi-automated buildup and calibration of conceptual sewer models, Environ. Model. Softw., 2017, 93, 344–355.
26. S. C. Troutman, N. Schambach, N. G. Love and B. Kerkez, An automated toolchain for the data-driven and dynamical modeling of combined sewer systems, Water Res., 2017, 126, 88–100.
27. Y. Hu, D. Scavia and B. Kerkez, Are all data useful? Inferring causality to predict flows across sewer and drainage systems using directed information and boosted regression trees, Water Res., 2018, 145, 697–706.
28. G. Liu, Y. Jia, H. Liu, H. Qiu, D. Qiu and H. Shan, A case study to detect the leakage of underground pressureless cement sewage water pipe using GPR, electrical, and chemical data, Environ. Sci. Technol., 2002, 36(5), 1077–1085.
29. O. Kracht, M. Gresch and W. Gujer, A stable isotope approach for the quantification of sewer infiltration, Environ. Sci. Technol., 2007, 41(16), 5839–5845.
30. O. Kracht, M. Gresch and W. Gujer, Innovative tracer methods for sewer infiltration monitoring, Urban Water J., 2008, 5(3), 173–185.
31. P. Willems, Quantification and relative comparison of different types of uncertainties in sewer water quality modeling, Water Res., 2008, 42(13), 3539–3551.
32. V. Gamerith, M. B. Neumann and D. Muschalla, Applying global sensitivity analysis to the modelling of flow and water quality in sewers, Water Res., 2013, 47(13), 4600–4611.
33. Ministry of Ecology and Environment the People's Republic of China, Water Quality-Determination of the chemical oxygen demand-dichromate method, 2017, HJ828-2017.
34. Ministry of Ecology and Environment the People's Republic of China, Water Quality-Determination of total phosphorus- ammonium molybdate spectrophotometric method, 1989, GB11893-198.
35. J. M. Shelton, L. Kim, J. Fang, C. Ray and T. Yan, Assessing the Severity of Rainfall-Derived Infiltration and Inflow and Sewer Deterioration Based on the Flux Stability of Sewage Markers, Environ. Sci. Technol., 2011, 45(4), 8683–8690.
36. Z. X. Xu, L. L. Wang, H. L. Yin, H. Z. Li and B. R. Schwegler, Source apportionment of non-storm water entries into storm drains using marker species: Modeling approach and verification, Ecol. Indic., 2016, 61(5), 546–557.
37. H. L. Yin, M. Xie, L. Y. Zhang, J. H. Huang, Z. X. Xu, H. Z. Li, R. J. Jiang, R. C. Wang and X. Zeng, Identification of sewage markers to indicate sources of contamination: Low cost options for misconnected non-stormwater source tracking in stormwater systems, Sci. Total Environ., 2019, 648(2), 125–134.
38. Z. C. Zhao, H. L. Yin, Z. X. Xu, J. Peng and Z. W. Yue, Pin-pointing groundwater infiltration into urban sewers using chemical tracer in conjunction with physically based optimization model, Water Res., 2020, 175(5), 115689–115698.
39. J. L. Bertrand-Krajewski, G. Chebbo and A. Saget, Distribution of pollutant mass vs volume in stormwater discharges and the first flush phenomenon, Water Res., 1998, 32(8), 2341–2356.
40. R. L. Lange and M. Wichern, Sedimentation dynamics in combined sewer systems, Water Sci. Technol., 2013, 68(4), 756–762.
41. I. Carnacina, F. Larrarte and N. Leonardi, Acoustic measurement and morphological features of organic sediment deposits in combined sewer networks, Water Res., 2017, 112, 279–290.
42. Y. Cao and G. J. Alaerts, A model for oxygen consumption in aerobic heterotrophic biodegradation in dual-phase drainage systems, Water Res., 1996, 30(4), 1010–1022.
43. S. Shahvi, E. Orsi, R. Canziani, E. Larcan and G. Becciu, Aerobic transformations of organic matters in sewer wastewater, a case study of Milan, Italy, Urban Water J., 2017, 14(10), 1058–1064.
44. C. Q. Wang, R. J. Chen, L. Wu, H. L. Song, Z. B. Fan, Q. J. Liu and X. N. Li, Release of pollutants with rainfall-triggered runoff from different underlying surfaces in villages typical of Yixing, Shengtai Yu Nongcun Huanjing Xuebao, 2016, 32(4), 632–638 (In Chinese).
45. Y. Gu and N. Zhang, Characterization of urban surface runoff in main districts of Kunming, Huanjing Gongcheng Xuebao, 2013, 7(7), 2587–2595 (In Chinese).
46. J. S. Lin, A study of infiltration Capacity of underground water in urban sewers, Shizheng Gongcheng, 2004, 4, 38–41 (In Chinese).
47. Y. Yang, D. Zhang, B. Lei, H. Liu, L. Zhai, Y. Mao, G. Feng and A. Chen, Variation characteristics of shallow groundwater level and its influencing factors in the nearshore vegetable field of Erhai Lake, J. Irrig. Drain. Eng., 2017, 36(12), 101–109.
48. K. Raunkjaer, T. Hvitved-Jacobsen and P. H. Nielsen, Transformation of organic matter in a gravity sewer, Water Environ. Res., 1995, 67(2), 181–188.
49. J. L. Huisman, T. Gasser, C. Gienal, M. Kuhni, P. Krebs and W. Gujer, Quantification of oxygen fluxes in a long gravity sewer, Water Res., 2004, 38(5), 1237–1247.
50. A. Guisasola, D. de Haas, J. Keller and Z. Yuan, Methane formation in sewer systems, Water Res., 2008, 42(6–7), 1421–1430.
51. W. Li, T. Zheng, Y. Ma and J. Liu, Current status and future prospects of sewer biofilms: Their structure, influencing factors, and substance transformations, Sci. Total Environ., 2019, 695, 133815–133826.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0ew00735h

---