Implementing machine learning to optimize the cost-benefit of urban water clarifier geometrics

Li, Haochen; Sansalone, John J.

doi:10.1016/j.watres.2022.118685

Cited by 11 publications

(10 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For a design scenario within a typical five-device product range (1.2 m, 1.8 m, 2.4 m, 3 m and 3.7 m diameters (Hydro International & NJCAT 2015)), an increase in design from 1.2 m 2 to 2.5 m 2 surface area would represent a transition from a 1.2 m diameter device to a 1.8 m diameter device. As proprietary devices cost between €17,000 and €69,000 (2022 inflation adjusted) (USEPA 1999), 100% oversizing of hydrodynamic separators is unlikely to translate into significant additional costs, in contrast to the up to millions of euros achievable in more centralised clarification applications (Li & Sansalone 2022a). As such, while Pe sizing yields the Figure 8 | Estimated connectable area for hydrodynamic (Pe) and non-hydrodynamic (Ha) devices employing annualised η (50% or 83%) for the NJCAT sediment (Figure 2) under German annual rainfall intensities and flow weightings: 0.9 mm h À1 (50%), 2.2 mm h À1 (33%) and 9 mm h À1 (17%).…”

Section: Application Of Certification Results To Device Sizingmentioning

confidence: 99%

Evaluation and sizing of proprietary sedimentation devices for decentralised stormwater treatment

Houlker¹,

Pasing²,

Gesterding³

2022

Water Science &Amp; Technology

View full text Add to dashboard Cite

Suspended solids removal is a key performance measure for proprietary stormwater treatment devices. Various technologies are available, with manufacturers claiming hydrodynamic separators offer performance advantages. However, it is important to assess manufacturers' claims. Accordingly, this study seeks to compare the performance of proprietary devices, by applying dimensional analysis to third party certification data and experimental data from uncertified devices, and determine the accuracy of a single parameter estimation (Hazen or Péclet number) of removal efficiency. Statistical analysis indicates that device performance is well described by a single parameter estimation, transitioning from Hazen (NSE = 0.81 and RMSE = 5.1%) at low surface loading rates (SLR) in all technology types (high removal efficiency), to Péclet (NSE = 0.5 to 0.61 and RMSE = 5.9% to 4.3%) at higher SLR (low removal efficiency) for hydrodynamic separators. This indicates that performance at low SLR is well explained by gravity separation in all technology types, whilst in hydrodynamic separators removal at high SLR is better explained by gravity separation plus advection. Consequently, when high (>80%) removal efficiency is required there is no performance advantage between technology types. However, when low (<50%) removal efficiency is required hydrodynamic separators offer a 33% increase in treatment area.

show abstract

Section: Application Of Certification Results To Device Sizingmentioning

confidence: 99%

Evaluation and sizing of proprietary sedimentation devices for decentralised stormwater treatment

Houlker¹,

Pasing²,

Gesterding³

2022

Water Science &Amp; Technology

View full text Add to dashboard Cite

show abstract

“…3. Compared with existing lower-fidelity methods, performing design optimization with higherfidelity CFD simulations demands higher level of user's knowledge and skill (5,7). For example, in existing lower-fidelity methods such as RT, as this method is empirical and formulated as a closed-form algebraic expression, basin design is primarily a ''plug-and-chug'' and minimal mathematical skills and experience are required.…”

mentioning

confidence: 99%

“…With the transformed input design features (i.e., a nine-dimension vector) and the CFD simulated PM removal, a supervised ML task can be formulated. The data layout of such a supervised ML task is illustrated in Figure1d 5…”

mentioning

confidence: 99%

mentioning

confidence: 99%

“…Stakeholders can significantly benefit from optimizing basin design ( 3 – 5 ). Recent studies demonstrate that using computational fluid dynamics (CFD) simulation to optimize and guide basin design can reduce common basin design costs by 4.7X to 8X ( 4 , 5 ) for a given level of treatment. However, there exist many fundamental barriers leading to the stakeholders’ hesitation in adopting such an approach in engineering practice.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Extensibility of a Machine Learning Model for Stormwater Basin Design and Retrofit Optimization Through a User-Friendly Web Application

Spelman

Sansalone

2023

Transportation Research Record: Journal of the Transportation R

Self Cite

View full text Add to dashboard Cite

To sequester particulate matter (PM) and chemical loads, stormwater basins are a common infrastructure component of transportation land use systems in a planning/design or a retrofit phase. Basin design relies on residence time (RT), capacity–inflow ratio (CIR), and continuous stirred-tank reactor (CSTR) concepts to provide presumptive guidance for load reduction. For example, 14-day (or 21-day) RT is presumed to provide load reduction for total suspended solids (TSS) (e.g., commonly 80%), and for total phosphorus (TP) (e.g., commonly 60%). Despite such guidance, most existing basins are impaired—not meeting presumptive guidance, whether for TSS, nutrients, or chemicals (e.g., metals). RT guidance results in high initial basin design costs (land and construction). For impaired basin retrofit, RT guidance generates a significantly higher cost (area/volume increase), if such a retrofit is even feasible considering proximate infrastructure constraints. This study presents a computational fluid dynamics (CFD)-machine learning (ML) augmented web application, DeepXtorm, for cost–benefit optimization of basin design and retrofit. DeepXtorm is examined against basin costs from as-built data, RT, CIR, and CSTR (deployed in the Storm Water Management Model [SWMM]) over a range of load reduction levels of TSS (60%–90%) and TP (40%–60%). For a given load reduction requirement, DeepXtorm demonstrates up to an order of magnitude (or more) lower basin cost compared with RT, CIR, and CSTR models, and greater than an order of magnitude compared with as-built cost data. DeepXtorm represents a more robust and cost-effective tool for stormwater basin design and retrofit.

show abstract

Empowering Smart Cities: AI-Driven Solutions for Urban Computing

Haldorai,

Murugan

et al. 2024

EAI/Springer Innovations in Communication and Computing

View full text Add to dashboard Cite

Implementing machine learning to optimize the cost-benefit of urban water clarifier geometrics

Cited by 11 publications

References 54 publications

Evaluation and sizing of proprietary sedimentation devices for decentralised stormwater treatment

Evaluation and sizing of proprietary sedimentation devices for decentralised stormwater treatment

Extensibility of a Machine Learning Model for Stormwater Basin Design and Retrofit Optimization Through a User-Friendly Web Application

Empowering Smart Cities: AI-Driven Solutions for Urban Computing

Contact Info

Product

Resources

About