comparison of modeling solutions in icmfor rdiisimulation · 2018-07-27 · outline introduction of...
Post on 09-Jul-2020
1 Views
Preview:
TRANSCRIPT
Comparison of Modeling Solutions in ICM for RDII SimulationPreliminary Pilot Study Using DeKalb Sewershed Model
HELEN LU, NANCY SCHULTZ, REGGIE ROWE
JACOBS
1
Outline Introduction of Rainfall Derived Inflow/Infiltration (RDII)
RDII simulation challenges
ICM model solutions for RDII simulation
Pilot study by using one sewershed model in DeKalb System
Summary and discussion
2
Definition of Rainfall Derived Infiltration and Inflow (RDII)
Source: EPA SSOAP Tool Box
Extraneous flows, both inflow and infiltration, derived from rainfall. RDII typically includes the fast response component (inflow) and the slow response component (infiltration).
RDII Simulation Challenges
4
ChallengesUncertainties:Source of the RDIIRDII response to various storm eventsOther factors
Complexity
5
RDII Sources
EPA Sanitary Sewer Analysis and Planning Toolbox
Antecedent Moisture Conditions
7
Other Uncertainties System pipe connectivities
Real time system performance and operation: moving solids, pump operation….
Flow data: Data Qualities
8
Tools for RDII Analysis SSOAP ToolboxUSEPA DevelopedCRADA with CDM
9
Important to know your flow data!
Complicated Process:How RDII flow gets to the sewer…..
Routing
Volume
Source: “Infowork ICM user help, Innovyze”
Sewer Pipe
10
Overland
Below Ground
ICM Model Solutions for RDIISimulation
11
RDIIModelingWEFFact Sheet notes the challenges and highlights the key considerations related to RDIIsimulation
12
ICM Model Solutions for RDIITriangular Unit Hydrograph: RTK
Hydrologic (surface runoff) models: volume and routing
Ground/Groundwater Infiltration Model (GIM)
13
Triangular Unit Hydrograph: RTK
14 Source: EPA Sanitary Sewer Analysis and Planning Toolbox
R-Values: % of runoff volume that
enters sewer system
Initial Abstractions and Monthly RTK
Since we want to use 1 set of RTKs, other parameters for those events need to be adjusted Initial abstraction parameters: storage
available for the runoff/infiltration from the rainfall before it reaches the sanitary sewer system. This storage can be due to soil conditions, topography or ponds etc. which would store the flow until its capacity is reached.
Monthly varied RTK to represent the seasonal variation
Rain.5”
15
Good News…
16
Runoff Models: Volume +Routing
17
LOST
GIM focus on the limitations of existing overland runoff models on slow response simulation
18
Source: http://blog.innovyze.com/2015/10/27/groundwater-infiltration-module-in-infoworks-icm-and-legacy-cs/
Runoff Models
GIM model
GIM Flow Schematic
LOSS TO BASEFLOW
(LOSSES FROM GROUNDWATER STOREDO NOT ENTER DRAINAGE NETWORK)
GROUNDWATER INFLOW TO SEWER NETWORKGROUND STORE INFLOW
SOIL STORE INFLOW
SOIL STORE DEPTH
GROUND STORE LEVEL
INFILTRATION TO GROUND STORE
INFILTRATION TO SOIL STORE Below Ground
GWI Model SetupGood article from Innovyze online Blog:
//blog.innovyze.com/2015/10/27/groundwater-infiltration-module-in-infoworks-icm-and-legacy-cs/
Over 12 parameters, but can be simplified by focusing on the major parameters
Require both GIM subcatchment parameters and Groundwater infiltration event
Need runoff model for direct inflow (fast responses)
20
GIM Subcatchment Parameters
LOSS TO BASEFLOW
(LOSSES FROM GROUNDWATER STOREDO NOT ENTER DRAINAGE NETWORK)
GROUNDWATER INFLOW TO SEWER NETWORKGROUND STORE INFLOW
SOIL STORE INFLOW
SOIL STORE DEPTH
GROUND STORE LEVEL
INFILTRATION TO GROUND STORE
INFILTRATION TO SOIL STORE
Percolation Parameters
Baseflow Parameters
Infiltration Parameters
A Long Journey……
22
A Quick SummaryTriangular Unit Hydrograph: RTK Widely used Empirical approach (forced curve)
Hydrologic (surface runoff) models: volume and routing Conceptualize the RDII responses as surface runoff inputs Limitation to represent slow response
Ground/Groundwater Infiltration Model (GIM) “complicated”, 12+ parameters, but can be simplified Very effective to simulate inter-event impacts due to antecedent moisture
conditions or a highly attenuated response to rainfall (long tail) Need to combine with a runoff model for direct inflow (fast response)
23
DeKalb Sewershed Model Pilot Study
24
DeKalb Pilot Study : Cobb Fowler Creek Sewershed Model
25
ApproachesCompare three model solutions for RDII simulation Fixed percentage runoff + New UK PR, proposed per model protocol
developed in earlier model studies by others RTK GIM + fixed percentage runoff model
Storm Events: Historical event: April 6, 2014, large storm (close to 1 year 24 hour) during
dry period Flow monitoring from January to June of 2015 (wet Period) January 4, 2015, largest storm (still less than 1 year 24 hour) captured during the flow
monitoring April 6 to 21, 2015, multiple small storm events (less than 1 year 24 hour) in the wettest month
26
On-going study, preliminary results for discussion only
Meter Locations
27
CBF8 CBF6
CBF3
CBF11
April 6, 2014
28
RainfallDepth (in)
RainObservedRTKFixed+New UK PRFixed PR+GIM
Peak (in/hr)RainObservedRTKFixed+New UK PRFixed PR+GIM
Average (in/hr)RainObservedRTKFixed+New UK PRFixed PR+GIM
FlowMin (MGD)
RainObservedRTKFixed+New UK PRFixed PR+GIM
Max (MGD)RainObservedRTKFixed+New UK PRFixed PR+GIM
Volume (US Mgal)RainObservedRTKFixed+New UK PRFixed PR+GIM
2.440 0.720 0.0170.390 3.070 6.1040.403 2.420 4.8500.403 2.471 4.4300.403 2.462 6.346
Observed / Predicted Report Produced by hlu3 (7/11/2018 8:33:43 PM) Page 1 of 1 Flow survey: >DynamicModelPilot>Pilot>FlowSurvey_graph>Flow survey_2013-2014 (4/3/2018 10:28:29 AM) Sim: >DynamicModelPilot>Pilot>Model group_Run>RTK_april2014>Rainfall event_2013_2014 (7/11/2018 2:36:39 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>1-yr-24-hr!_MWH_april2014!>Rainfall event_2013_2014 (7/9/2018 4:00:23 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>GWI4_CBF8_calb7_test9rev_April2014!>Rainfall event_2013_2014 (7/11/2018 7:44:23 PM)
January 4, 2015
29
RainfallDepth (in)
RainObservedRTKFixed + New UK PRFixed PR+GIM
Peak (in/hr)RainObservedRTKFixed + New UK PRFixed PR+GIM
Average (in/hr)RainObservedRTKFixed + New UK PRFixed PR+GIM
FlowMin (MGD)
RainObservedRTKFixed + New UK PRFixed PR+GIM
Max (MGD)RainObservedRTKFixed + New UK PRFixed PR+GIM
Volume (US Mgal)RainObservedRTKFixed + New UK PRFixed PR+GIM
2.440 1.680 0.0170.506 2.984 7.7410.436 1.850 4.7110.412 2.516 4.3710.436 2.435 6.054
Observed / Predicted Report Produced by hlu3 (7/11/2018 8:32:11 PM) Page 2 of 11 Flow survey: >DynamicModelPilot>Pilot>FlowSurvey_graph>Flow Survey Group>SF_2015Flow Survey Group_CBF (5/4/2016 3:02:19 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>RTK_Jan2015!>Rainfall_2015_v2_CBFonly_profileID (7/11/2018 3:17:11 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>1-yr-24-hr!_MWH!_Jan2015>Rainfall_2015_v2_CBFonly (7/10/2018 4:34:51 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>2015April_GWI4_CBF8_calb7_test9rev_Jan1>Rainfall_2015_v2_CBFonly_profileID (7/11/2018 7:03:09 PM)
April 6 to 21, 2015
30
RainfallDepth (in)
RainObservedRTKFixed + New UK PRFixed PR+GIM
Peak (in/hr)RainObservedRTKFixed + New UK PRFixed PR+GIM
Average (in/hr)RainObservedRTKFixed + New UK PRFixed PR+GIM
FlowMin (MGD)
RainObservedRTKFixed + New UK PRFixed PR+GIM
Max (MGD)RainObservedRTKFixed + New UK PRFixed PR+GIM
Volume (US Mgal)RainObservedRTKFixed + New UK PRFixed PR+GIM
5.140 2.040 0.0130.478 2.403 17.3040.403 1.737 12.7350.403 2.472 10.7810.436 2.439 16.822
Observed / Predicted Report Produced by hlu3 (7/11/2018 8:57:52 PM) Page 2 of 11 Flow survey: >DynamicModelPilot>Pilot>FlowSurvey_graph>Flow Survey Group>SF_2015Flow Survey Group_CBF (5/4/2016 3:02:19 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>RTK_Apr6-21_2015>Rainfall_2015_v2_CBFonly_profileID (7/11/2018 3:17:40 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>1-yr-24-hr!_MWH_Apr6-21>Rainfall_2015_v2_CBFonly (7/10/2018 4:38:27 PM) Sim: >DynamicModelPilot>Pilot>Model group_Run>2015April_GWI4_CBF8_calb7_test9rev_Apr6-21>Rainfall_2015_v2_CBFonly_profileID (7/11/2018 11:23:40 AM)
Summary & Discussion
31
Summary and DiscussionsGIM shows the best results to simulate inter-event impacts due to antecedent moisture conditions or a highly attenuated response to rainfall (long tail)
Selection and application of proper RDII simulation methods need to be tailored to system conditions and modeling objectives
No single RDII prediction method could be universally applicable due to the wide variety of site-specific RDII characterizations.
RDII model can be a highly effective tool in the decision-making process for I/I reduction and remediation programs, but also limited by inherent assumptions and conceptual representations of complex process
All methods require monitored data, important to know your data: limitations, quality etc.
The “state-of-the-art” in RDII modeling practice continues to evolve
32
Acknowledge Support fromDWM:Darren Eastall, Consent Decree
AdministratorLinda Li, Hydraulic Modeling Lead and her
modeling teamOthers involved in the Consent Decree
hydraulic modeling programConsent Decree Program Management Team (CDPMT)
33
Questions?
Technology Hype Cycle
34
Hope you are
here…..
top related