Analysis of the Proportion of Incidental Water in Annual Wastewater Inflows to a Selected Treatment Plant During a 15-year Observation Period

The main aim of the study was to determine whether the amount and share of incidental water in wastewater inflows to the wastewater treatment plant increased with the age and development of the sewage system. The object of the study is a small sewerage system located in southern Poland, in the Lesser Poland Province, 15 km northeast of the Kraków city. The analysis of a 15-year observation series of daily wastewater flows, from 01.01.1994 to 31.12.2018, allowed a reliable assessment of the actions taken by the sewage system operator to reduce the sources of incidental water inflow to the sewage network. The study showed that the percentage share of incidental waters in the annual wastewater inflow to the analysed treatment plant decreased from 35.9% in 2004 to 8.4% in 2018. This reduction should be treated as a success of the remedial actions of the sewage system operator. However, the study showed at the same time that in 2017 alone, 56 709 m3 of incidental water flowed into the analysed sewerage system, compared to 166 075 m3 of wastewater. A similar case was repeated in 2013, when 53 060 m3 of incidental water and 132 860 m3 of real wastewater flowed into the treatment plant. These cases show that all remedial measures by the network operator are still insufficient when heavy precipitation occurs in a given year for a long period of time.


INTRODUCTION
The most important principles of the European environmental policy until 2020 and in the perspective until 2050 are protection, preservation and improvement of the natural capital of the EU as well as protection of its inhabitants from environment-related pressures and threats to health and well-being [https://europa.eu/european-union/topics/environment_pl]. In terms of these tasks, adequate and effective protection of surface and groundwater quality is particularly important.
The greatest threat to these waters are anthropogenic factors, in particular point source pollution, such as municipal and industrial wastewater discharges [Chmielowski et   Long-term monitoring of daily wastewater flows makes it possible to determine the volume of influent incidental water and to assess the effectiveness of any measures taken to seal the network and eliminate sources of inflow of incidental water to the sanitary sewer system [De Bénédittis 2004].
The study analysed the wastewater and incidental water inflows to a selected small rural wastewater treatment plant during a 15-year observation period. Such a long study period allows a full and detailed analysis and assessment of the issue in question, taking into account all the relevant factors.
The main aim of the study was to determine whether the amount of incidental water increases with the age and development of the sewer system, and whether the actions taken by network operators contribute to reducing the inflow of these waters into the analysed sewer system during the 15-year observation period.

Object of reserches
The object of the study is a small sewerage system located in southern Poland, in the Lesser Poland Province, 15 km north-east of the Kraków city. In 2004, when the sewerage system and treatment plant were commissioned, the length of the sanitary gravity sewerage network was 4.6 km, in the year of completion of the study (2018) it had a length of 36.1 km ( Table 1). The sewerage network is made of PVC pipes with diameters: 200, 250, 300, 350 and 400 mm, laid at a depth of 1.6 to 3.5 m below ground level. The number of households and inhabitants using the sewerage system is shown in Table 1. Wastewater, mainly domestic sewage, is discharged from the sewerage system to the mechanical-biological wastewater treatment plant with a capacity of 563 m 3. d -1 .
The wastewater treatment process consists of a Huber screen, Imhoff primary settling tank, flowthrough biological reactor with separate aerobic and anoxic zones and a vertical secondary settling tank. The wastewater in the aeration reactor is treated with fine bubbling system. The reactor allows for biological removal of nitrogen, while phosphorus is removed from wastewater in the secondary settling tank, after application of PIX. Currently, the average daily inflow of wastewater to the plant is 522 m 3 ·d -1 during rainless weather.

RESEARCH METHODOLOGY
The measurement data of daily wastewater flows were used in this study. This parameter was measured at the treated wastewater outflow from the treatment plant to the receiver using a Prosonic S FDU 90 Endress+Hauser ultrasonic flow meter. The measurement error of the effluent level was +/-2 mm. The continuous testing period was from 01.01.1994 to 31.12.2018. The measurement of wastewater flows was carried out daily at 7:00 a.m. and the measurement value was automatically recorded on a computer server archiving the operation parameters of the treatment plant. An OP2 tipping bucket rain gauge was also installed at the treatment plant; it was connected to a Mini Log B Endress+Hauser pulse data logger. The information on the occurrence and amount of rainfall was helpful in establishing the rain-free period. A dry period was defined as any day during which the total rainfall did not exceed 5 mm and the last rainfall greater than 5 mm occurred not earlier than 2 days before. The latter condition was intended to exclude increased sewage flows resulting from heavy recent rainfall and sewer retention from the dry weather period. After separating, for a given calendar year, the nonrainy periods during which only domestic sewage (without external waters) entered the sewage system, the value of average daily inflow of specific sewage q ść was calculated together with the value of standard deviation s ść . The daily inflow of incidental water to the sewerage system, during wet weather, was calculated using Eq. (1): where: q p -daily inflow of incidental water to the treatment plant during wet weather (m 3 ·d -1 ),  q d -daily inflow of sewage and incidental water to the treatment plant during wet weather (m 3 ·d -1 ), q ść -daily inflow of domestic sewage to the treatment plant during dry weather (m 3 ·d -1 ), s ść -standard deviation of daily sewage flow during dry weather (m 3 ·d -1 ).
The value q ść + s ść was taken as the cut-off level separating proper sewage from incidental waters. Figure 1 shows an example of daily inflows of proper sewage and incidental waters to the studied wastewater treatment plant in 2004. In addition, the diagram shows the time periods of rainless weather and daily precipitation amounts.
For each year of the study, the summed daily flows of proper sewage, incidental water and sewage and incidental water were calculated separately.
In the papers of many authors [Pecher 1999 Table 2 presents -separately for each calendar year -total inflows of wastewater, incidental water, wastewater and incidental water together. In addition, the table includes the values for the proportion of accidental waters in the annual sewage inflow to the analysed treatment plant, calculated using Eq. (2). The last column of the table shows unit amount of the sewage q j , per sewerage user (PE).

RESULTS AND DISCUSSION
The data obtained show that each year the volume of incidental water increased by 1942 m 3 on average. In 2004, 17 087 m 3 of incidental water flowed into the analysed sewerage system, in 2017 this volume increased by 332% to a volume of 56 709 m 3 . Thus, it can be considered that the annual volume of incidental water increased with the utilisation of the sewerage system and could be an indication of increasing leakage in the sewerage system. However, at the same time it should be noted that in 15 years the length of the sewer network increased from 4.6 km up to 36.1 km and the number of households connected to the sewer increased from 138 to 1077. The increase in the length of the sewer network and connections to buildings generally generates a marked increase in the sources of incidental water (more sewage wells and more buildings with the possibility of connecting roof gutters to house sewers). In order to correctly assess the load of the analysed treatment plant with incidental water it is necessary to analyse the course of the trend line of the share of incidental water on the time axis. This relationship will show the real state of sensitivity of the sewage network to incidental water inflows. In order to use statistical tools in the analysis, the statistical distribution of the SIW values must be examined. In the Shapiro-Wilk test, the W statistic was 0.926 and the p-value = 0.267 for an accepted level of p=0.05. Thus, there were no grounds to reject the hypothesis H 0 stating the normality of the analysed data. The F-value of the Fisher-Snedecor statistic, for the regression equation, was 33.177, while the p-value = 0.00013 for an assumed significance level of p=0.05. Thus, there were no grounds to reject hypothesis H 0 stating the significance of a linear relationship between the analysed variables. The Student's t-test also confirmed the significance of the directional coefficient (p-value = 0.000127) and the intercept (p-value = 0.0000001).
The obtained parameters of linear regression equation indicate that the share of incidental water in the inflow to the treatment plant decreased every year by 1.544%. During the analysed 15 years, the value of SIW decreased on average from 32.8 to 11.2% (Figure 2). This confirms that the remedial action taken by the operators of the analysed sewerage system have had the expected positive effect. In order to confirm this thesis, the graph presented in Figure 3 was created. It shows a simulation calculated on the basis of the data from the period 2005-2008, when the share of incidental water was subject to an average increase of 2.28% per year. This increase was related to the length of the network and the number of buildings connected to the sewerage system in 2005-2008, and then proportionally related to the length of the network and the number of buildings in 2018. The presented relationship shows that if no actions were taken to improve the tightness of the sewerage network against incidental water inflows, this share could be close to 64.7% (red line).
The presented results of calculations show that the SIW indicator gives a better reflect the incidental water load in sewer networks and treatment plants than the incidental water value itself, given in m 3 ·year -1 or m 3 ·d -1 .
An interview with the operator of the analysed sewerage system revealed that the greatest impact on the elimination of potential sources of incidental water was: • smoke testing and video inspections of the interior of sewerage collectors to detect and eliminate illegal or mistaken connections of roof gutters to sewers; • replacing manholes and improving their place to the road surface by using rings to distance the manhole tops; • placing greater emphasis on acceptance testing of new sewerage network sections and new connections to buildings; • environmental education of network users and introduction of charges with reduction of natural land retention.
A significant reduction in the inflow of incidental water to the analysed sewer network is also indicated by the value of daily sewage volume q j , calculated per 1 user of the network ( Table 2). In 2004, during the start-up of the sewage treatment plant, this value was 267.5 dm 3 ·PE -1 ·d -1 . It indicated a very careless technical acceptance after the network construction, as it is overestimated probably by more than 100 dm 3 ·PE -1 ·d -1 . Since 2013, this value has decreased from 211.8 to 161.4 dm 3 ·PE -1 ·d -1 in 2018. This indicator shows that some incidental water still flows into the network in 2018, but it already represents only 11.4 dm 3 ·PE -1 ·d -1 , because during dry weather this indicator was on average 150 dm 3 ·PE -1 ·d -1 .
The results obtained in this study should be referred to the results obtained by other authors. Weiss et al. [2002] investigating the share of incidental water in the sewerage systems of Baden-Wurttemberg, obtained a value of SIW equal to 35%, i.e. similar to that in the analysed sewerage system in 2004. A study by Krachta and Gujera [2005], carried out in Switzerland, found some incidental water in the inflows to wastewater treatment plants ranging from 35 to 65%. These results accurately reflect the extent of the negative prediction and confirm the predicted increase in the SIW without remedial action. A study in Norway, carried out by ØDEGARD [Hey et al. 2016] in 14 sewerage systems, showed that the proportion of incidental water was on average 67% (this value correlates with the negative forecast results obtained in this study). In contrast, in the Netherlands, the value of the SIW in the study by Schilperoort [2004] was 38%, while in Austria -from 25 to 50% [Ertl et al. 2008]. On the basis of the study by Kaczor [2002], conducted in 5 sewer systems in Poland, in the Małopolska Province, the SIW value was between 15.8 and 57.7%. In the study by Bugajski et al. [2017], also performed for a site in Poland, the SIW value averaged 36.2%.
Despite the noticeable improvement in the average value of the SIW factor, the annual inflow of incidental water to the analysed sewerage system in 2017 is alarming at the same time (Table 2). May and June the total rainfall was 271.2 mm, representing 42% of the annual total. Heavy rainfall, during the months mentioned, caused hydraulic overloading of the treatment plant and inflow of increased incidental water. In 2017, the total rainfall in July, August and September was 331.5 mm, corresponding to 49.3% of the annual total. These cases show that any remedial action by the network operator is not sufficient when heavy and long-duration precipitation occurs in a given year.

CONCLUSIONS
Summarizing the obtained results, it may be stated that the share of incidental waters in the inflow to the analysed sewerage system in 2004 (at the beginning of the study period) was at an average level in comparison with other objects in Poland and worldwide.
On the basis of the forecast carried out, it was shown that the absence of the actions to improve the tightness of the sewerage system could lead to a situation such as that observed at some sites in Switzerland and Norway, where incidental water accounted for more than 60% of the annual inflow to the treatment plant.
Against the background of the results obtained by other authors, the activities and corrective actions of the operator of the examined sewer network, which led to a decrease in the value of SIW from 35.9% in 2004 to 8.4% in 2018, should be positively evaluated. Such low values of SIW were not found in any research object described in the literature.
In addition, studies have shown that heavy precipitation with over long period of time, occurring periodically in particular years, despite the remedial actions taken by the network operator, still causes hydraulic overloading of the sewerage and wastewater treatment plants.
In summary, it should be stated that the longterm continuous observation of daily sewage flows allows for a very detailed analysis of the functioning of the sewage network and wastewater treatment plants. They allow irregularities in the operation of these facilities to be detected and the effects of remedial actions to be effectively assessed.