WARMF: A Watershed Modeling Tool for Onsite Wastewater Systems

1
WARMF A Watershed Modeling Tool for Onsite
Wastewater Systems

• Laura Weintraub, MS, PE
• Systech Engineering, Inc.
• San Ramon, CA
• Co-authors
• Carl W. Chen, Systech Engineering
• Robert A. Goldstein, EPRI
• Robert L. Siegrist, Colorado School of Mines

AWRA Annual Conference San Diego, CA, November
3-6, 2003
2
Background and Motivation

• OWS account for 25 of domestic waste disposal
• OWS discharge important component of TMDL
  calculation
• Most models require external estimations of OWS
  loads using GIS
• Model that predicts OWS is more useful
• Calibrate to field data
• TMDL implementation plans
• Projecting trade-offs, OWS vs. centralized sewers
• Compare various OWS types

Onsite Wastewater System (OWS)
3
Project Overview

• Part of larger project sponsored by NDWRCDP
• Modeling, field data collection, laboratory
  experiments
• Watershed-scale model adapted to simulate OWS
  loading
• Algorithms developed based on OWS research and
  scientific principles
• Model calculates edge-of-drainfield pollution
  loads rather than requiring them as input
• Model predictions verified with observed data
• Case study of Dillon Reservoir watershed

4
WARMF Overview

• GIS-based watershed model and decision support
  system
• Five linked Modules including Consensus and TMDL
• Physically based, dynamic model
• Driven by meteorology, land use, point source,
  fertilizer, air quality data
• Simulates temperature, TSS, DO, nutrients, fecal
  Coli, Chl-a, etc.
• Peer reviewed by independent experts under EPA
  Guidelines
• Applied to watersheds throughout United States

Map of WARMF Applications
5
Catchment Hydrologic Simulation
Snow or Rain
Interception
Snow hydrology
Surface hydrology
Soil hydrology
6
Nonpoint Source Simulation
Wet deposition Dry deposition Fertilizer,
pesticide, animal dropping
Dissolution, advection, decay
Soil erosion wash off
Mineral weathering AMD Septic systems Organic
matter decay Nitrification Cation exchange Plant
uptake
7
Modifications for Modeling OWS

• Initial test add STE to each top soil layer
• Treatment based on original soil algorithms
• NH4-N retained through competitive cation
  exchange and nitrified
• PO4-P adsorbed according to linear isotherm
• Fecal Coli decay
• Required unrealistically high nitrification rate
  in soil

• Demonstrated need to incorporate biozone
  mechanism
• Biologically active soil layer develops in soil
  receiving STE
• Higher reaction rate than natural soil

8
Biozone Algorithm Hydrology

• Biomass bacteria grows on organic matter fed by
  STE
• Field capacity ? with bacteria accumulation,
  higher surface area, retains water like sponge
• Porosity ? with build up of plaque (dead bacteria
  and solid residue)

• Gap between field capacity and porosity reduces
  over time, infiltration rate ?, possible
  hydraulic failure of OWS

9
Biozone Algorithm Water Quality

• Biomass uptakes N (8) and P(2)
• Biomass decays, respires and sloughs off
• First order decay
• Nitrification, BOD decay and fecal coliform
• Decay rates a function of Biomass
• Temperature dependent rate constants
• Decay rates are scaled to biozone volume
• Phosphorus adsorption occurs in soil below biozone

10
Biozone Algorithm Testing

• Biozone module initially created independent of
  WARMF
• Tested with experimental data (Col. School of
  Mines)
• 16 columns tested
• 4 flow rates
• Accelerated loading for 138 days
• Loaded with actual STE
• Calibrated to establish decay rates in biozone

11
Study Area Dillon Reservoir Watershed

• 325 mi2 (840 km2)
• Water supply for City of Denver
• Approximately 1500 OWS servicing vacation and
  primary residences
• Rapidly developing watershed

12
3 Focus Areas

• Higher density of OWS
• Higher resolution modeling
• Surface water and groundwater field data
  collection

13
WARMF Input Data
14
OWS Characterization

• GIS spatial data
• number of OWS in each catchment
• assumed 2.5 people per household
• STE characteristics
• CFD Plots compiled from literature (Kirkland
  2001)
• STE effluent quality for various OWS types
  compiled

50th-percentile values for STE
15
Hydrology Calibration
Blue River above Dillon Reservoir relative error
0.070 cms (2.93) R2 0.84

• Blue River at Blue River, CO
• relative error 0.018 cms (2.33)
• R2 0.82

16
Water Quality CalibrationBlue River at Penn.
Creek (BR-3)
Nitrate
Ammonia
Phosphate
Fecal Coli
17
WARMF Consensus Module
18
Management ScenarioConvert Blue River Estates OWS

• Convert service for 906 residents (362 OWS)
• Flow and loading surface water discharge at S.
  Blue River WWTP
• Assume same treatment level from WWTP

19
Total Phosphorus Loading
Nonpoint ? Point ? Total Load ?
20
Total Nitrogen Loading
Nonpoint ? Point ? Total Load ?
OWS conversion only beneficial with a very high
level WWTP
21
Conclusions

• WARMF successfully enhanced for OWS
• Model accepts septic tank effluent
• Processes STE through biozone
• Discharges treated effluent to top soil layer of
  watershed model
• Provides linkage between OWS discharge and
  receiving water quality
• WARMF provides a tool for evaluating TMDLs or
  management alternatives related to OWS

22
Recommendations for Future Work

• Continue to improve biozone algorithm with
  additional monitoring and experimental data
• Incorporate OWS virus work into WARMF
• Share Dillon watershed application with Summit
  County stakeholders
• Apply WARMF to other watersheds with OWS issues

23
Acknowledgements

• Funding
• National Decentralized Water Resources Capacity
  Project (NDWRCP), USEPA
• Collaboration with
• Colorado School of Mines
• EPRI
• Summit County
• USGS

24
Summary

• Applied a watershed scale model (WARMF) to Dillon
  watershed
• Enhanced model resolution for 3 focus areas with
  high density OWS
• Incorporated additional parameter values from CSM
  field and laboratory studies
• Enhanced WARMF to model OWS using a biozone
  algorithm
• Developed hypothetical management scenarios and
  tested with WARMF