modeling urban watersheds impacted by csos and ssos

MODELING URBAN WATERSHEDS

IMPACTED BY CSOS AND SSOS

Fifty Years of Watershed Management –

Past, Present and Future

September 24-26, 2012

Ted Burgess

Presentation Agenda

• Current state of modeling software

• Developing dry-weather flow rates

• Rainfall data used in model calibration

• Single-event and continuous model calibration

• Flow data analysis for model calibration

• Modeling green stormwater infrastructure

• Integration of (increasingly digital) field data and

SCADA with collection system models

Urban Watershed / Collection Systems Models

• Aqualyze – h3O

• Bentley – SewerGEMS, SewerCAD

• BOSS – StormNET

• CHI (Canada) – PCSWMM

• DHI (Denmark) – MIKE URBAN

• Delft (Netherlands) – SOBEK

• Innovyze (formerly MWH Soft) – InfoSWMM, InfoSewer, InfoWorks (formerly Wallingford Software, UK)

• U.S.ACE – HEC-HMS / HEC-RAS

• U.S.EPA – SWMM 5

• XP Software (Australia) – xpswmm

DHI MIKE URBANCHI PCSWMM

Innovyze InfoSWMM

U.S.EPA SWMM 5 and Commercial “Spin-offs”

Aqualyze h3Oxpswmm

Bentley SewerGEMS

BOSS StormNET

Gal/day/capita Water/Sewer Billing Analysis

• Data source requirements:

– Water (or sewer) billing records for winter periods

– Geo-reference property addresses to model catchments

Sewer Billing Data – Catchment Level

Sewer Billing Results

Radar-Rainfall Analysis:

2031 grid cells vs. 49 rain gauges

= 29.7%* + 1.5%* + 32.5% * + 36.4%*

Example: Area Weighted Radar-Rainfall

� Basin BW-AL-14 (352.618 ac or 1.427 km2)

Area Weighted Radar Rainfall for the basin is calculated as:

Area covered by radar pixels (1 * 1 km2) Area weights for each pixel

Selecting Calibration Storms

Selecting Calibration Storms (continued)

Unit Hydrograph Methodology in SWMM 5: Continuous

Simulation of Rainfall-Dependent Inflow/Infiltration

Rainfall Data Analysis for Continuous Simulation

Seasonal Wet Weather Response

Growth vs. Dormant Variability in Mean Dmax

U.S.EPA’s Sanitary Sewer Overflow Analysis and

Planning (SSOAP) Toolbox

SSOAP Toolbox - Data Flow

Sewer SystemGIS Database

Flow Monitoring Data

Rainfall

Data

Sewer System

Time Series

FlowVelocity

Depth

Time SeriesRainfall

Hydraulic Analysis

Data

SSO Volume

Capture Flow Volume

Overflow Frequency

Flooding LocationsPipe Capacity

External Data Sources

Database

ManagementTool

RDII AnalysisTool

DWF analysis resultsWet-weather selection results

WWF analysis results

RDII results

Event based RTK parametersRTK predictive analysis results

Internal Data Sources

Sewer System

Flow DataRainfall Data

RDII HydrographGeneration

Tool

RTK parameters

Rainfall DataSewer System

SSOAP-SWMM5Interfacing

Tool

RDII

Hydrograph

SWMM 5 Input File

SWMM 5

SWMM 5 Input File with RDII Hydrograph

SSOAP

SystemDatabase

in MS-ACCESS

0.2

0.1

0.0

0.4

0.6

0.8

1

2

3

4

2

3

4

1 SunJun 2008

8 Sun 15 Sun 22 Sun 1 Tue

OL-UP-10

Ra

inF

all (

in)

De

pth

(ft)

Flo

w (

mg

d)

Ve

locity (

ft/s

)

Date/Time

STA_OL-UP-10 0742S0009 0742S0009 (obs)0742S0010:0742S0009 0742S0010:0742S0009 (obs)

0.15

0.10

0.05

0.00

0.4

0.6

0.8

1

2

3

2

3

4

8 MonDec 2008

15 Mon 22 Mon 1 Thu

OL-UP-10

Ra

inF

all (

in)

De

pth

(ft)

Flo

w (

mg

d)

Ve

locity (

ft/s

)

Date/Time

STA_OL-UP-10 0742S0009 0742S0009 (obs)0742S0010:0742S0009 0742S0010:0742S0009 (obs)

0.3

0.2

0.1

0.0

0.4

0.6

0.8

1

2

3

2

3

4

1 MonSep 2008

8 Mon 15 Mon 22 Mon

OL-UP-10

Ra

inF

all (

in)

De

pth

(ft)

Flo

w (

mg

d)

Ve

locity (

ft/s

)

Date/Time

STA_OL-UP-10 0742S0009 0742S0009 (obs)0742S0010:0742S0009 0742S0010:0742S0009 (obs)

0.075

0.050

0.025

0.000

0.00

0.25

0.50

0.75

0

1

2

3

0

2

4

1 ThuJan 2009

8 Thu 15 Thu 22 Thu 1 Sun 8 Sun 15 Sun 22 Sun 1 Sun

OL-UP-10

Ra

inF

all (

in)

De

pth

(ft)

Flo

w (

mg

d)

Ve

locity (

ft/s

)

Date/Time

STA_OL-UP-10 0742S0009 0742S0009 (obs)0742S0010:0742S0009 0742S0010:0742S0009 (obs)

Example Calibration Results: Continuous RDI/I in SWMM5

Auto-calibration approaches

using Genetic Algorithm-based

techniques and graphical tools

can facilitate this process.

Continuous simulation runtimes for large

networks still require skeletonization

Modeling the relationship between sanitary/combined

sewers and storm sewers for SSO / CSO control

Precipitation

Initial Abstraction

Infiltration

Groundwater Recharge

(Deep Infiltration)

Infiltration to Collection Systems

Infiltration to Stormwater or

Combined Collection System

Infiltration to Sanitary Sewer System

Runoff

Runoff to Stormwater Collection System

Direct Inflow to Sanitary or Combined

Sewer System

19Session II – Modeling Urban Watersheds Impacted

by CSOs and SSOs

Green Stormwater

Infrastructure

Changes in SWMM5 facilitate green stormwater

infrastructure modeling

EffectiveImpervious

(DCIA)Pervious

Inlet

Impervious Pervious

ReceivingCatchment

Transfer

Transfer

Catchment routing –

traditional approach

Catchment routing –

new SWMM 5 options

SWMM5 LID Control Editors

Modeling21

Bio-Retention

Infiltration Trench

Porous Pavement

Rain Barrel

Vegetative Swale

Digital advancements in other sewer system technologies

allows integration of modeling with field information

3D Viewer Tool: Linking system condition and GIS data with sewer network data

SCADA integration example: Key Flow Control Structure

Whittier Street Regulator Gates

To Berliner Park

From/to DSR83

Regulator Gates

N

Whittier Street Storm Tanks

Modeler’s View of Collection System

Operator’s Views of Collection System

SCADA display of selected system conditions

SCADA to SWMM5 Control Rules Editor

Conclusions

• As computers get faster, models get bigger and more detailed (so we still live with long runtimes)

– CSO modeling: Typical year sufficient – no problem

– SSO modeling: Design targets ~ 2-10 year return periods – require much longer simulation periods and impractical runtimes

• Current focus on green stormwater infrastructure imposes new demands on established watershed modeling tools

• Convergence of field data tools (inspection databases, SCADA) and modeling tools is opening up new capabilities

Session II – Modeling Urban Watersheds Impacted

by CSOs and SSOs

EXTRAS

DWF Calibration Approach

DWF = GWI + Diurnal Pattern* Billing Data

– Basin level diurnal pattern and GWI developed from ADWF

flow monitoring data

– Catchment level billing data developed from winter season

water or sewer billing records

– Three parameters used to calibrate DWF:

• GWI Adjustment k1

• Pattern (or Billing) Multiplier k2

• Pattern Time Shift k3

DWF =(GWI + k1)+ (Diurnal Pattern time shift × k2) × k3 × Billing

End

of

Event

Start

of

Event

Rainfall

Dry

Weather

Flow

RDII

Flow

Metered

Flow

RDII Analysis Tool – Determination of RDII Parameters

Catchment Delineation Workflow

Data Process Results

MDC Sewer

Network

MDC Topographic

Data (2 foot contours)

MDC Storm

Network

MDC Property &

Building Polygons

ESRI ArcHydro

ArcGIS Spatial

Analysis

& VBA macro

Manual

Verification

Storm

Catchments

Sanitary/Combined

CatchmentsMDC Orthophotos

Modelers Views of Collection SystemRULE WSST_RG1A IF NODE 0006C0263

DEPTH <= 6.5’ (696’ ele.)THEN ORIFICE

WSST_RegulatorGate1 setting = 1

RULE WSST_RG1B IF NODE 0006C0263

DEPTH > 6.5’ (696’ ele.) AND < 7’ (696.5’ ele.)

THEN ORIFICE WSST_RegulatorGate1

setting = 0.6

RULE WSST_RG1C IF NODE 0006C0263

DEPTH >= 7.2’ (696.7’ ele.) AND <= 8’ (697.5’

ele.) THEN ORIFICE WSST_RegulatorGate1

setting = 0.3

RULE WSST_RG1B IF NODE 0006C0263

DEPTH > 8 (697.5’ ele.) THEN ORIFICE

WSST_RegulatorGate1 setting = 0

RULE WSST_RG2A IF NODE

WSSTControlHouse DEPTH <= 8.65 (700’ ele.)

THEN ORIFICE WSST_RegulatorGate2

setting = 1

RULE WSST_RG2B IF NODE 0006C0263

DEPTH > 6.5 (696’ ele.) THEN ORIFICE

WSST_RegulatorGate2 setting = 0

Columbus Sewer System Model Sanitary sewers: >12” Ø; combined sewers: >18” Ø

Columbus Sewer System Model Summary

• SWMM 5.00.22 engine / PCSWMM 2011 (CHI) interface

• Detailed model: 22,600 nodes / 3820 catchments

• RPM model: 4880 nodes / 1851 catchments

• Continuous simulation of RDII (RTK with IA) and surface

runoff (Green-Ampt)

• Diurnally-varied base flow and seasonally-varied GWI

• Calibration (2008-2009): 212 flow meter sites

• Validation (2010-2011): 60 permanent meter sites

• Radar-rainfall (1 sq km grid): calibration period

• Rain gauge data (49 sites): validation period

Columbus Sewer System ModelKey trunk sewers and flow control structures

Heavy Cleaning vs. Low Velocities

HGL vs Basement Elevations