6 Dependability Below Kensico Pressurization of Catskill Aqueduct from Kensico Reservoir to the UV Facility at EastviewKensico – City Tunnel (Previously known as City Tunnel No. 3 Stage 3)The Bronx – Queens Tunnel (Previously known as City Tunnel No. 3 Stage 4)City Tunnel No.3(Stage 2, Manhattan Leg)
7 2007 RWBT Identified as Primary Area of Concern RWBT conveys Delaware system water and is the primary source for Towns of Newburgh and Marlborough and approximately 50% of New York City’s supply
8 Project Background - Roseton Surface Expression Roseton LeakageProject Background - Roseton Surface ExpressionA dozen or so surface expressionsFlow observed on west bank of Hudson RiverTesting program and analysis estimates leakage of 15 to 35 mgd.Zhenqi Notes – improve quality of slide / image resolution? Call-out font sizes… enhance tunnel lines
9 How Long will the Tunnel be Out of Service? 2007 Shutdown ScheduleHow Long will the Tunnel be Out of Service?
11 Potential Solutions for Dependability Increase Aqueduct CapacitiesParallel TunnelsCroton Pump StationsAbandoned Sources (Westchester Co.)Hudson River and Harbor Surface Water, Hudson GroundwaterInterconnectionsDemand ReductionExpand Groundwater Use
12 Augmentation Project Screening Project Tiering/Prioritization – Focuses Effort on Best Projects and Combinations
13 Delaware Repair: Alternative water sources Project evaluation to identify a group of projects to meet NYC’s alternate water supply needs (2008)Factors:CostScheduleDEP Control26 ProjectsKey:Redundant tunnelsOptimization of existing systemGroundwaterDesalinization of Hudson River or Harbor waterInterconnections to New Jersey or Connecticut39 Projects
14 Delaware Repair: Alternative water sources Top 26 Projects falls to 17 with the removal of mutually exclusive projects.
16 Moving to DesignIn 2008 and 2009 DEP hired new consultant team to develop parallel tunnel and bypass tunnel concepts to address the RWBT leaks
17 Repair of the Leaks Roseton Repairs Wawarsing Repairs: Shaft 6 TunnelUnwatering ShaftShaft 2A17RondoutReservoirShaft 9Shaft 8EL 840Shaft 1Shaft 2Shawangunk MountainsShaft 4Shaft 5Shaft 5AShaft 7West BranchReservoirRondout CreekShaft 3Walkill RiverHudson RiverEl 503ElWawarsingRoseton RepairsNot possible from within tunnelAccess requires new shaftsBest Solution is BypassWawarsing Repairs:Possible from within tunnelAccess via Shaft 2AConfinement Relatively goodRoseton
19 Recent ModelingIn 2010 DEP shifted from evaluating a full repair of the RWBT to constructing a bypass around the leak in Roston, NYThis bypass included the use of inundation plugs to handle tunnel inundation, resulting in approximately a month shutdown period for the connection
20 OST Supply Curve ( )OSTv _Aug-SC_000Runs were conducted for these four supply levels
21 Threshold Approach - Objectives Objectives (during an outage):Provide advance notice of potential shortfall conditionsProvide DEP with enough time to take some preventive actionObjectives (now, during the planning process):Accurately simulate operations during an outageProvide a fuller picture of how various augmentation projects performProvide a framework to help support DEP’s risk analysisUncertainties in augmentation project capacityUncertainties in RWB repair duration
22 Likelihood of Emergency Actions for 15-Month Outage 22
23 Big Break Through!!! Inundation Plugs not necessary!!! This resulted in the following:A shorter shutdown period for the connection of approximately 10 monthsA phased connection of the bypass tunnel to the RWBTAllowed for an evaluation of shorter shutdown periods and bailout contingenciesWe evaluated four phased approaches:Fixed Staging4 month / 3 month / 3 month5/56/4Variable Stage (10 months total but can be broken and phased over 3 years)
24 Tunnel Outage Duration Evolution The design of the connection work always considered outage duration and risk mitigationSubsequent iterations improved risk pictureBase PlanPreemptive PlugReactive PlugDrainage Tunnel
26 Water Demand and Dry Weather Wastewater Flows Historic Flows and Future Projections 2012
27 Water Supply System Augmentation and Improvement Conservation / Demand ManagementUpper Catskill Aqueduct OptimizationQueens Groundwater Rehabilitation
28 Modeling Results End of 2012 As a result of three workshops, extensive modeling, and shutdown design updates:Variable Stage Shutdown allows for complete shutdown in one phaseIncludes a repair start on Oct 1 with augmentation starting June 1Provides for a singular phase repair based on modeling using forecasted 60 day look ahead for continuation of shutdown or bailout from repairProvides for allowance for remaining repair to be completed in subsequent years, if necessary58 mgd of augmentation (33 mgd Groundwater + 25 mgd Demand Management)Catskill Aqueduct maximum flow increased to 636 mgd via Catskill Rehabilitation ProjectCroton WFP flow at 250 mgd to account for diurnal demand pattern.
29 Outage Surface Plot: Variable Outage Index560Outage_PatternVarSourceNY_Q58New_Crot_Aq_max290CatAq_max636Reserve_Buffer10%NYC_Demand_level1070Temp_vary_DemandMost outages start in OctoberSubstantially longer Stage 1 outage duration with 10% reserve and CAT 636Year 1Year 2Year 3
30 Likelihood of Completion: Variable Staging Month #12345678910Likelihood of Completing x Outage DaysShows probability of successfully completing the variable stage shutdown repair during a three consecutive year periodAllows for three chances to complete the 10 month (300 day) repairCould complete entire 10 months in first yearOr could complete remaining repair period during subsequent one or two years if insufficient supply is available in first yearFixed stage shutdowns do not allow for this flexibility since entire three year period is required for each repair period4 lines for each color – these represent:NCA 250 / Demand PatternNCA 290 / Demand PatternNCA 250 / Demand RegressionNCA 290 / Demand RegressionWhy does this scenario appear to be more successful than the fixed outage scenarios here but appears to be less successful in the following runs...?Should discuss this...This plot shows the likelihood of completing 300 total outage days over a period of up to 3 years.
31 Likelihood of Completion: Summary Table 99% Probability of Success!
32 Next Steps Continue refining shutdown design and modeling Continue thinking about issues that could ariseDevelop operational plans for shutdown