13th IWA Specialized Conference on Small Water and Wastewater Systems 14 - 16 September 2016, Athens, Greece Session: Small Scale and Decentralized Wastewater Treatment and Management A Monte-Carlo based method for the identification of potential sewer mining locations Ι. K. Tsoukalas*, C. K. Makropoulos* and S. N. Michas ** * Department of Water Resources and Environmental Engineering, National Technical University of Athens, Heroon Polytechneiou 5, GR-15780, Zographou, Greece (E-mail: itsoukal@mail.ntua.gr; cmakro@mail.ntua.gr) ** Hydroexigiantiki Consultants Engineers, 3 Evias str, GR-15125, Marousi, Greece (E-mail: smichas@hydroex.gr)
2 Who is who… EYDAP Athens Water Supply and Sewerage Company, NTUA, Greece (CASE) National Technical University of Athens (RTD) Sewer Mining CHEMiTEC Pilot (Athens) Water & Environmental Technologies, Greece (SME) TELINT RTD Consultancy Services, UK (SME)
3 Context Athens has suffered rapid urbanisation resulting in few urban • green spaces Reuse , but at what scale ? • • Need for innovative management options and technologies for reuse needed to irrigate (primarily) green urban areas Current status (incl. devastated peri-urban forests). • Main WWTP in an island DEMO SITE (Psytalleia) • Increased energy costs for transp. • Peri-urban forests devastated by WWTP fires • Water scarcity
4 Enter Sewer Mining… The Athens Pilot brings together two emerging technologies: Fully automated packaged treatment low energy sensor Distributed plants networks coupled with distributed featuring membrane based, ICT intelligence (e.g. Advanced Metering small footprint, sewer mining technologies that allow direct mining and Monitoring Infrastructure, AMIs) of sewage from the network, close to innovative in terms of data fusion (b) point-of-use the with minimum data communication (c) interoperability and (d) mobile solutions for remotely infrastructure required and low controlling and operating transportation costs for the effluent the distributed infrastructure (against stringent performance criteria, incl. health and water quality standards)
Main concept The following general concept was developed as a basis of applications of the proposed solution: ICT controls Wastewater Unrestricted Sewer reuse irrigation network options Compact treatment Urban use system Pumping arrangement Industrial use
7 Benefits to be explored Benefits Upscale opportunities Case Athens is an opportunity to: Deployment in the east coast of • Increase reuse efficiency with Attica for: treatment at the point of use • Decrease transaction costs • Urban green compared to “centralised” reuse • Reduced water treatment cost (licensing / footprint / local • communities) Reused water withdrawal to • Increase % of reused water within avoid saltwater intrusion the highly constrained urban environment • Improve urban quality of life East coast through improved ecosystem services and; • Create new market for SMEs that can provide this service to, e.g. local municipalities A win-win scenario → SMEs will sell raw sewage using the existing centralised sewerage network of the water company and water companies will be able to sell untreated sewage, Map of Attica, Greece while also minimise the load to their centralised treatment facilities.
Identification of potential locations for sewer mining units: A Monte-Carlo approach Step 1: Spatial data pre-processing Step 2: Monte-Carlo Simulation Inputs : Uncertain parameters Χ (e.g., Variation Identification of: coefficients of wastewater discharge , BOD 5 loading). 1. Sewage network topology and assets (e.g., manholes, pipes) P( x ) 2. Hydraulic characteristics (e.g., P( x ) P( x ) pipe diameter, slope) X1 X2 Xn 3. Land uses (areas that will benefit from sewer mining – Sewer network Model e.g., parks) simulation Parameters Locate neighborhood model θ sewer network (e.g., SWMM5) components (e.g., nodes) Step 3: Results post-processing Outputs : Quantities of interest (e.g., concentration of BOD 5 at each pipe). Calculate metrics (e.g., utility functions, risk functions) & P( x ) P( x ) P( x ) perform multi-criteria analysis Location(s) selection Y1 Y2 Yn
Step 1 (a): Spatial data pre-processing Import: 1. Sewage network topology and assets (e.g., manholes, pipes) 2. Hydraulic characteristics (e.g., pipe diameter, slope) 3. Land uses (areas that will benefit from sewer mining – e.g., parks) 4. Other spatial data (e.g., aerial photo)
Step 1 (b): Spatial data pre-processing Pre-process: Why? Identify land uses (areas Offset green that will benefit from areas (10m) sewer mining – e.g., green Locate nested nodes areas, parks) Locate neighborhood sewer network components (e.g., nodes) How? Add offset to green areas (e.g., 10m). Locate the nodes that are inside each offset area. WWTP Find the path from each “selected” node Exit (e.g., WWTP). This path is unique for each node due to the “collective nature” of sewer networks.
Step 2: Monte-Carlo Simulation Why? The purpose of Monte-Carlo Step 1: Spatial data pre-processing simulation is to propagate the uncertainties of input parameters to the outputs. Also, allow the use of probabilistic Step 2: Monte-Carlo Simulation objective functions (metrics). How? Select number Sample uncertain For parameters X from Identify uncertain parameters Χ of Monte-Carlo i=1:N simulations N their distribution • Daily and hourly variation coefficients of wastewater discharge No • BOD 5 loading Quantities of interest Identify output of interest Sewer network i=N? (e.g., concentration • BOD 5 concentration of each pipe Simulation model of BOD 5 at each Alternatives? pipe). Yes Similar, a scenario-based approach End (instead or in conjunction with Monte- Carlo) could be employed (e.g., worst, middle, high conditions). Next step? Step 3: Results post-processing Define probabilistic objective functions (metrics). Post-process the results
Step 3: Results post-processing Why? Step 1: Spatial data pre-processing The purpose of this step is to use metrics (e.g., utility Step 2: Monte-Carlo simulation functions, risk functions) that quantify the output of interest (in our case H 2 S build-up) for a Step 3 (a) : Results post-processing chain of pipes (node exit Metric Z originally proposed by von Bielecki & Schremmer, (1987) and Pomeroy, node). (1990) for a single pipe I in order to quantify the probability of H 2 S build-up: How? Employ a modified version of 0.3 × 1.07 𝑈−20 × 𝐶𝑃𝐸 5 𝑗 × 𝑄 𝑗 the “quasi - quantitative” 𝑎 𝑗 = 0.5 × 𝑅 𝑗 1 3 indicator Z. 𝐶 𝑗 𝐾 𝑗 Calculate the E[ Z ] for given Where, i is the pipe index, T is the sewage temperature ( o C), J is the pipe slope, Q reliability level (R>75%) for is the discharge (m 3 /s), P is the wetted perimeter of the pipe wall (m) and B the each path for each green area surface width (m) of the stream. using the N simulation runs Modified Index Z of Pomeroy for a “ chain ” of pipes n : For each green area select the path with minimum E[ Z ]. 𝑜 Alternatives? 𝑎 𝑑 = 𝑏 𝑗 × 𝑎 𝑗 Similar, other metrics can be 𝑗=1 used that quantify the exact Where, a i = L i / L tot , L i is the length of pipe i , and L tot is the total length of pipes of amount of H 2 S in terms of mg/l. chain n . Next step? According to Pomeroy, (1990) if a pipe has Z i > 7500 then there are high chances Multi-criteria analysis and of H 2 S formation which could lead to odour and corrosion problems. selection of potential locations for sewer mining units.
Why? Step 1: Spatial data pre-processing The purpose of this step is to use multi- criteria analysis in order to identify Step 2: Monte-Carlo simulation potential locations for sewer mining unit placement. How? Step 3 (a): Results post-processing We have already calculated E[ Z ] for each node thus we can combine this Step 3 (b): Identify the Pareto set (Max{Area}, Min{Z}) information with: Based on the analysis: Information regarding the water For each green area the optimal node for the SM demand in the areas of interest (green placement is already found (step 3a). areas) Fuse the information regarding H 2 S build-up and green • We select as rough indicator for area water demand. water demand the area (m 2 ) of Green area 1 and 2 are suitable for SM placement. Green area 1 and 2 were selected based on a desired the park. reliability level Alternatives? Acceptance Similar, the actual water demand of each Max {Area} threshold area can be calculated if relevant G3 information is available. G4 Also other metrics can be employed in Next step? G2 Selection of potential locations for sewer mining units. Further analysis and modelling of G1 sewer mining unit for the selected locations. 7500 Min{E[Z]}
Study area: Kalyvia Thorikou, Greece
Study area results: Kalyvia Thorikou ID 12 ID 3 ID 22 Optimum path of green area (ID 3)
The GUI of Sewer Mining Placement Tool Step 2: Setup Monte- Step 3: Results post- Step 1: Load spatial data Carlo Simulation processing and pre-processing
Recommend
More recommend
