Who says looks do not matter?
Kevin Spence of Sheffield Hallam University explains the objectives and results of a research project concerned with ways of tackling aesthetic pollution from combined sewer overflowsThe larger municipal areas of the UK will continue to convey stormwater and wastewater in the same pipe networks for the foreseeable future. The hydraulic design of this infrastructure is well established, for example, combined sewer overflows (CSOs) are strategically located on these networks to prevent hydraulic overloading of both the network and downstream treatment works.
A project was formulated to address these needs and funds awarded by the Engineering and Physical Sciences Research Council and UK Water Industry Research (UKWIR) to Imperial College, Sheffield Hallam University, the University of Sheffield and Coventry University. Academic research personnel included David Butler, Adrian Saul, John Davis, Kevin Spence, Kim Littlewood, Manfred Schuetze, Chris Digman and James Houldsworth. David Balmforth was a principal driver during the inception of the project whilst at Sheffield Hallam University and remained an active and positive participant whilst employed by Montgomery Watson Harza. Committed and regular input into the project was supplied by UKWIR, Yorkshire Water and Scottish Water. The Environment Agency was also represented.
The ability to quantify aesthetic pollution production, as modified by the influence of such variables as total population number, gender, social, economic and ethnic grouping was investigated. Low and high income and ethnic areas were identified using a social deprivation map of the Sheffield area compiled from census data by the Geography Department, Sheffield University. This information was used with combined sewer asset data to identify discrete networks suitable to continuously collect both flow data and aesthetic pollutants during dry and wet weather conditions. Three distinct catchments were identified and consisted of a high income site with a white population, a low income site with 70% white and 30% Pakistani ethnicity, and a low income site with an ageing white population. Due to the configuration of sewer networks, the low income sites with ageing and ethnic populations were split into four and two subcatchments respectively. The total areas of the subcatchments ranged from 6.8-31.6ha with impermeabilities of 28-45%. Populations ranged from 359-1,810. Due to the topography of Sheffield, the average gradients varied from 1:15-1:47 and can therefore be judged as being relatively steep.
A postal questionnaire was devised and distributed to a random sample of 250 people in each of the three catchments. Information was gathered on age, sex, income and sanitary product usage. Results obtained from field sewer sampling showed an expected diurnal variation in both faeces and toilet tissue, but little temporal variation in other sanitary products. Both independent techniques of data acquisition allowed solid production rates per capita/day and the influence of socio-economic factors to be determined. For example, the sewer survey showed that faecal material typically constituted 60% of the sampled mass and for this solid 24g/person/d was produced at the high income and the low income and ethnic minority catchments, but 39g/person/d was produced at the low income catchment with an aging population. By dividing a catchment production value by the mean value from all catchments, a social, economic and ethnic day factor (SEED) was obtained for each solid. Social class and ethnicity made a difference to the number and type of solids that were input into the sewer system. For example, more wipes were used in the low income catchment with an aging population (SEED factor = 1.86) when compared with the catchments of high income (0.74) and low income with an ethnic minority (0.40). Good agreement was obtained between the two survey types. This data allows the quantity of solids generated by a catchment area to be evaluated. In practice, census data can be used to establish which SEED values should be used to adjust solid production rates.
These daily production rates are then modified to diurnal temporal profiles, based on previous work at Imperial College, and used as an input to the gross solids simulator (GROSSim) computer model that has been developed to predict the movement of gross solids in a combined sewerage system. It utilises drainage system data and continuous velocity/depth files from HydroWorks. The software simulates real-time series events, e.g. five dry days followed by a defined or measured rain event, followed by a further dry period. In dry weather flows, a proportion of solids entering the system is deposited in the upstream sections of the pipe network (typically 5%), which are not normally represented in a hydraulic model of the network. Storage equations modify the diurnal solids profile to account for these solids. These algorithms have been derived from field survey results which indicated that storage of material increases with antecedent dry weather period and housing density. The next rainfall event of sufficient magnitude will then mobilise the stored solids, for example when rainfall drives the discharge above three-times the dry weather flow. These, together with solids already being transported will arrive at a downstream location in the form of a ‘first foul flush’. A non-linear reservoir model of the upstream pipes predicts the rising limb, time-to-peak, peak magnitude and total quantity of solids flushed. This forms an input to the modelled network during a storm event. In the main sewer network where the pipes are greater than 300mm, solids are tracked individually because their advection velocity differs from the mean fluid velocity.
Deposition and subsequent re-erosion of the solids driven by the output from the HyroWorks model and governed by velocity and depth criteria developed from laboratory experiments and separate field experiments. A further component of the model is a CSO element which incorporates a family of performance characteristics for each solid associated with a particular CSO structure. These characteristics have been generated for different flow conditions using computational fluid dynamics. The GROSSim model has a user-friendly front end for easy data entry and model experimentation and produces output in a format that is easy to manipulate and plot. Default information is built into the programme, based on data obtained in this study. The model has been run to produce time-varying plots (aesthetographs) of solids at particular locations and accurately replicates dry weather flow and the first flush effect that occurs in some networks and catchments.
As efficiency characteristics for any particular CSO chamber (and performance data for screen) can be incorporated into GROSSim, spill and continuation flow aesthetographs can be calculated. This may be of use to screen manufacturers and designers in allowing realistic assessment of the solids load experienced during storms. It may also help explain some of the causes of screen failure. GROSSim is also able to track the position of solids throughout the system. This then enables a picture to be built up of not only the distribution of retention times of solids in the system but also those links where solids deposit for extended periods of time. This could be used as a diagnosis tool for sewers prone to blockage, in particular when analysing whether the cause of the problem is associated with low depths, low flows or to assess the potential increase in screenings’ loadings at STWs as a result of a CSO screening programme in the upstream catchment.
At present the model is calibrated and validated on the catchments studied during the course of this research project and it is intended to make the model more widely available in the near future. The general usefulness of the model beyond the restrictions of the current data set would be enhanced by application and verification on larger catchments. This is being explored at present in conjunction with companies that have expressed an interest. In conclusion, the main deliverable is a user-friendly, holistic modelling tool that will assist the water industry in decision-making during the control of aesthetic pollutants.