Islam et al. (2025) Variations of urban water balances considering subsurface sewer fluxes: a hydrologic modeling study

Identification

• Journal: Journal of Hydrology
• Year: 2025
• Date: 2025-11-14
• Authors: Imranul Islam, Amit Kumar, Kun Zhang
• DOI: 10.1016/j.jhydrol.2025.134599

Research Groups

• Civil & Environmental Engineering Department, University of Minnesota Duluth, MN, USA

Short Summary

This study developed an integrated hydrologic model to evaluate how sewer-mediated subsurface fluxes respond to variations in environmental and structural factors in urban water balances. Findings reveal that sewer-mediated fluxes are primarily sensitive to water table depth and pipe defect size, while natural fluxes (surface runoff, evapotranspiration) are more sensitive to native soil and trench backfill types; backfilling trenches with native soils reduces sewer fluxes and enhances natural outflows.

Objective

• To evaluate how sewer-mediated subsurface fluxes (inflow and infiltration, exfiltration) respond to variations in environmental and structural factors such as native soil type, water table depth, and pipe defect size.
• To assess the impact of trench backfilling with native soils on sewer-mediated subsurface fluxes and overall water balances.
• To quantify the relative contribution of rainfall and groundwater on sewer-mediated fluxes during and after storm events.

Study Configuration

• Spatial Scale: Trench-scale domain, 15 m × 6.5 m rectangular shaped domain, with a 0.50 m diameter circular pipe.
• Temporal Scale: Each simulation ran for 30 days, including a 24-hour rainfall event followed by a 29-day dry period. Warm-up simulations were performed for one year to establish initial steady-state conditions.

Methodology and Data

• Models used: HYDRUS-2D (for solving Richards' equation in 2D), Principal Component Analysis (PCA), and Least Square Boosting (LSBoost) for feature importance analysis.
• Data sources:
  • Climate data: 10-year 24-hour rainfall depth (98 mm) from NOAA ATLAS 14 for Duluth Harbor station, NRCS Type-II synthetic rainfall distribution, annual precipitation normal from GLISA (2025), and mean potential evapotranspiration (PET) derived from a climate ratio.
  • Soil parameters: Van Genuchten-Mualem model parameters for various soil types (sand, sandy clay loam, clay loam, medium sand/gravel, colmation layer) curated from Carsel & Parrish (1988), Karpf & Krebs (2013), and Peche et al. (2017).
  • Vegetation parameters: Standard turfgrass parameters for root water uptake from ˇSimůnek et al. (2012).
  • Pipe and trench design: Minnesota Department of Transportation Standard Specification for Construction (MnDOT, 2020).
  • Model validation data: Analytical solutions (Tang et al., 2017; Yu et al., 2024), experimental analysis (Tang et al., 2017), and numerical analysis (Tang et al., 2017).

Main Results

• Rainwater partitioning and water balances exhibited notable variations across different native soil types and water table depths.
• Sewer-mediated fluxes (inflow and infiltration (I&I) and exfiltration) are more sensitive to water table depth and pipe defect sizes.
• Natural fluxes of surface runoff and evapotranspiration (ET) are more sensitive to surrounding native soil and trench backfill soil types.
• Backfilling trenches with native soils (particularly clayey soils) instead of granular imported soils reduced sewer fluxes and enhanced natural outflows.
• A larger sewer defect (increasing from 10 mm to 30 mm) increased I&I flux from 0.28 m² to 0.76 m² at a water table depth of 2.0 m, and exfiltration flux from 0.07 m² to 0.17 m² at a water table depth of 3.5 m.
• Most sewer-mediated flux originated from groundwater; rainfall contribution was relatively limited, especially with a shallow water table (e.g., 5.6% for I&I during the initial 0–5 days at a water table depth of 2.0 m).
• In a shallow water table environment, I&I accounted for 26% of the water balance, with surface runoff (21%), ET (29%), and subsurface outflow (24%) constituting the remaining. In a deeper water table environment, exfiltration occurred (8%), and runoff (27%), ET (37%), and subsurface outflow (44%) constituted the outflows without I&I.

Contributions

• Developed and validated an integrated hydrologic model (HYDRUS-2D) to simulate detailed trench-scale urban water balances, including sewer-mediated subsurface fluxes.
• Systematically quantified the combined and relative influence of interacting environmental (native soil type, water table depth) and structural (pipe defect size, trench backfill materials) factors on urban water balance components.
• Provided physically based parameterizations for sewer-groundwater exchange processes, which can inform and improve catchment-scale urban hydrologic models.
• Offered actionable insights into the hydrological impact of trench backfilling configurations, suggesting that using granular embedment around the pipe with native soil backfill above can balance structural stability, reduce sewer-mediated fluxes, and optimize construction costs.
• Utilized advanced statistical methods (PCA and LSBoost) to identify the relative importance of various factors influencing different water balance components, enhancing understanding beyond single-factor analyses.

Funding

• National Science Foundation (NSF) Engineering Research Initiation (ERI) grant #2347541 ("ERI: Effects of urban water infrastructure and proximate soil profiles on coupled surface–subsurface hydrology").
• University of Minnesota McKnight Postdoctoral Teaching Fellowship (for Amit Kumar).
• Swenson College of Science and Engineering’s Dr. Howard Higholt professorship (for Kun Zhang).

Citation

@article{Islam2025Variations,
  author = {Islam, Imranul and Kumar, Amit and Zhang, Kun},
  title = {Variations of urban water balances considering subsurface sewer fluxes: a hydrologic modeling study},
  journal = {Journal of Hydrology},
  year = {2025},
  doi = {10.1016/j.jhydrol.2025.134599},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134599}
}