Gurugram flood attenuation: diagnosis and pilot design

Citywide hydraulic evidence and a field-testable Subhash Chowk intervention
Municipal Corporation of Gurugram · Technical engineering briefing · 31 August 2026

Design-event water balance: 31.48 Mm³ enters the model

Audited 107.3 mm / 3 h design-event diagnostic
31.48 Mm³total modelled source water during the six-hour simulation
Rainfall input30.98 Mm³ from a 107.3 mm, three-hour design hyetograph; peak intensity 73 mm/h.
Boundary inflow0.50 Mm³ enters from mapped upstream perimeter receivers.
InterpretationThis is total source water over six hours, not simultaneous ponded volume. The next slide separates rainfall forcing from hydraulic recovery.
Source: audited FloodAstra event package `20260810T021439Z...twmoderate`. 31.48 Mm³ = 30.98 Mm³ rainfall + 0.50 Mm³ external boundary inflow; it is total source water, not simultaneously ponded floodwater.

Design forcing ends at 3 h; hydraulic recovery continues to 6 h

Reviewed forcing and simulation window
0–3 h · rainfall forcing107.3 mm total; peak hourly intensity 73 mm/h. This drives the inflow hydrograph.
3–6 h · no additional rainfallThe model continues so recession, retained water and delayed discharge can be evaluated.
Source: reviewed event `20260810T021439Z_july2025_critical_3h_107mm_twmoderate`. Bar labels are rounded; the model forcing integrates to 107.3 mm.

Net floodwater transfer across wards and sectors

Audited design-event diagnostic
117 sectors · 257 transfersEvery reviewed sector and net boundary connection is retained.
Ward view remains availableThe live HTML switches between full sector and ward networks. Ward and sector totals are alternative aggregation views.
Largest net sector transfers2.940 Mm³ gross crossings
FromToNet m³
Sector 1Sector 11144,237
Sector 55Sector 5632,449
Sector 36ASector 36B28,124
Sector 75ASector 3525,597
FloodAstra 107.3 mm/3 h design event, six-hour simulation, 25 snapshots, 2 cm transfer threshold. The headline totals integrate gross crossings; each table row is net A→B minus B→A. Ward and sector views must not be added.

Tailwater and surcharge reduce usable drainage capacity

Hydraulic screening diagnostic

Backwater

Flow can remain downstream while the receiving level removes hydraulic head.

Surcharge

Water then returns to the road through an inlet or manhole.

Consequence

Increasing an upstream drain alone cannot remove a downstream head constraint. The entering hydrograph must also be reduced or delayed.

Source: signed FloodAstra drainage-flow screening for Rajiv Chowk, Subhash Chowk and Sheetla Mata Road. Direction and stress are event-specific; inferred inverts and provisional capacity priors require survey confirmation.

Required storage is the cumulative inflow–safe-outflow deficit

Governing storage relationship
Storage is required only while intercepted inflow exceeds the release rate that the receiving system can safely accept.
1 · Intercepted inflowQintercept(t) comes from the coupled surface model, then is corrected by surveyed road levels, inlet positions and capture efficiency.
2 · Safe outlet rateQsafe,out(t) is the minimum of spare drain capacity, treatment throughput, tested well acceptance and the permitted operating rate.
3 · Tank volumeThe maximum cumulative difference sets usable storage. Freeboard, maintenance isolation and recovery before the next storm are added during design.
Limiting case
No detention volume is required when the receiving system safely accepts the complete intercepted hydrograph.
Engineering relationship for preliminary sizing. The 14,000 m³ RechargeTwin case is a scenario input used to test response; it is not the derived construction volume for Subhash Chowk.

Subhash Chowk screening: candidate interception and storage areas

Catchment and network screening
18.41 km²runoff-screening area
90.8 kmreceiver-linked drain trace

1 · Intercept

Survey road levels and inlets on both sides of Netaji Subhash Marg.

2 · Store offline

Use suitable parking and green belts; retain mature trees.

3 · Recover safely

Use gravity; add pressure only after field tests.

Screening evidence from assimilated terrain, urban form and mapped drainage. The 90.8 km trace is network-linked, not proof that every segment contributes runoff. Catchment boundary and intercept points remain provisional until surveyed.

Modular storage follows surveyed geometry and safe isolation

Preliminary layout logic—not construction geometry
Parking or green-belt footprint confirmed by survey
Capture sump
screen + grit trap
Cell 1
Cell 2
Cell 3
Isolation manifold
controlled outlets
Road and inlet collection line
Usable storageDimensions are established from the surveyed footprint, structural depth, freeboard and maintenance requirements.
Capture conditionSurveyed inlet levels and conveyance must deliver the intended Qintercept(t) to the cells.
Isolation conditionEach cell needs independent shut-off, access, sediment removal and return-to-service capability.
Selection ruleChoose the layout meeting Vrequired with acceptable utility conflict, transfer head, tree retention and maintenance access.
Concept geometry only. A single tank may remain preferable where one surveyed footprint and inlet system satisfy the same hydraulic, structural and operational constraints.

Pretreatment is a hydraulic and water-quality control train

Mandatory safety architecture
1 · Road inletIntercept the surveyed flood hydrograph without obstructing ordinary traffic drainage.
2 · Coarse screenRemove litter, leaves and floating debris before water enters storage.
3 · Grit chamberSettle sand and coarse sediment; provide access for mechanical removal.
4 · Oil separationRetain free oil and fuel before baffled settling and filtration.
5 · FiltrationMedia and fine filtration are selected from measured source-water chemistry.
6 · Quality gateOnly water within approved limits may reach a controlled outlet.
Separated wasteTrash, grit, sediment and oil remain isolated for removal and disposal.
Continuous measurementsFlow, tank level, turbidity, conductivity, pH, pressure and aquifer head are logged.
Automatic stopAn out-of-limit quality, pressure, head or acceptance response closes the recharge path.
Delayed surface dischargeRelease only when monitored downstream capacity becomes available.
Recharge pilotOperate only after chemistry, borehole, pressure, acceptance and permit checks pass.
Design requirement: no open hydraulic path from raw floodwater to an injection well. Final limits and operation require laboratory analysis, CGWA/CGWB and applicable pollution-control review, and monitored pilot testing.

Injection capacity must be measured, not inferred from depth

Groundwater screening—not approval

Regional data can screen

Groundwater depth, broad aquifer context and locations where field testing should begin.

The pilot must establish

Bore log, lithology, confining-layer continuity, source-water compatibility, acceptance rate, recovery, safe pressure and neighbouring-well response.

Allowable rate

Capacity is assigned only after monitored tests define every limit.

Source: CGWB/NAQUIM groundwater screening and local GIS synthesis. Confined target, transmissivity, confining-layer continuity, water quality, safe pressure and sustainable acceptance require borehole and pilot tests.

3D terrain–groundwater context prioritizes field investigation

Interpolated screening product

One location, two surfaces

Terrain elevation and groundwater elevation are inspected together.

Spatial context

Ward, coordinates, groundwater depth, parks and observation points remain visible.

Purpose

Prioritize boreholes and monitoring. The 3D view does not certify an aquifer or a well.

Source: Gurugram groundwater 3D screening product. Surfaces are interpolated from available observations and terrain; uncertainty increases away from measurements.

Coupled two-park scenario: attenuation and recovery

Coupled scenario screening
10.6%peak road-p95 depth reduction25.2%peak pocket-volume reduction0 haoutside area worsened
Coupled FloodAstra two-park screen, 107.3 mm/3 h event: 14,000 m³ detention, assumed 800 L/s transfer capacity and assumed 96 L/s aggregate well acceptance. The six-hour run did not recover below 15 cm. Assumptions require field testing; this is not construction design.

Global precedents validate components, not local design

Independent precedents

Copenhagen

Surface cloudburst routes and underground tunnels are combined where urban space permits.

Saitama, Japan

A dedicated underground floodway conveys river floodwater to a receiving river; it is not aquifer recharge.

Metro Colombo

Wetlands, drainage and urban works operate as a connected green–grey flood system.

Salisbury / OCWD

Managed recharge follows pretreatment, monitored wells and defined operating limits—not direct raw-road-runoff injection.

Global precedents validate the components. Gurugram still requires its own hydrograph, geology, treatment specification and monitored pilot.
Primary references: Copenhagen Cloudburst Management Plan; Japan MLIT Metropolitan Area Outer Underground Discharge Channel; World Bank Metro Colombo programme; South Australia Salisbury MAR review; OCWD GWRS; CGWB artificial-recharge manual.

Screening volumes do not equal construction quantities

Modelled water-availability screen
Rajiv Chowk
8.42 crore Lmodelled peak surface water within 500 m; use only as water-availability scale.
Subhash Chowk
7.68 crore Lmodelled peak surface water within 500 m; final intercept volume requires catchment hydrograph.
Sheetla Mata Mandir Road · 250 m
5.26 crore Lpeak modelled surface water around the confirmed coordinate; local maximum is 0.55 m within 50 m.
Construction quantity
Not yet sizedStorage follows the intercepted hydrograph, usable footprint, treatment train and tested recovery rate.
Final storage and well counts follow surveyed inflow, usable land, treatment capacity and field-tested recovery.
Source: priority-location screening atlas. Volumes are model-derived screening quantities and require survey and design reconciliation.

Matched runs quantify benefit and incomplete recovery

Coupled two-park screening · Palam Vihar
Matched FloodAstra screen with identical rainfall, terrain, June 5 assets and network physics. The intervention uses assumed 14,000 m³ detention, 800 L/s transfer and 96 L/s well acceptance; it is not a measured or completed pilot result.

Evidence required before scale-up

Stage-gated investigation
01 · SurveyLevels, inlets, utilities, trees and usable footprint.
02 · InvestigateBore log, lithology, chemistry and confining layer.
03 · Pilot-testStep-rate, pressure, recovery, compatibility and neighbouring-well response.
04 · SimulatePaired before/after FloodAstra run with citywide transfer guardrails.
05 · Approve & designAuthorities set limits; detailed design resolves storage, treatment, structures and controls.
06 · Build & verifyInstrument the pilot; track flood response, quality, head, drawdown and maintenance.
Scale only when flood reduction, aquifer safety and operational recovery are demonstrated together.
Proposed gated pilot sequence; detailed approvals and responsibilities to be established by MCG and relevant groundwater/environmental authorities.

Decision gate: authorize survey and pilot design

Decision gate
Field surveyConfirm catchment pathways, levels, inlets, utilities, trees and buildable footprints.
Aquifer investigationComplete test bore, chemistry, treatment design and monitored injection testing.
Hydraulic comparisonCompare before/after peak, persistence, recovery and citywide transfer guardrails.
Return with a go/no-go construction gate. Expand only when measured performance agrees with modelled benefit.
Decision framing: authorize bounded field investigation, matched simulation and pilot design—not construction, final storage volume, well count or injection rate.

Backup · Terrain, roads and buildings are represented hydraulically

Source: Gurugram terrain and urban-form evidence panels for Rajiv Chowk, Subhash Chowk and Sheetla Mata Road.

Backup · Signed flow distinguishes stress and reversal

Source: FloodAstra signed drainage-flow diagnostics. Teal/magenta direction and stress interpretation is event-specific.

Backup · Controlled detention and treated recharge

Conceptual section. No construction dimensions or injection rate are implied.

Backup · The hydrograph determines storage and recovery

Illustrative 107 mm/3 h paired screening curves. Design requires the site-specific intercepted hydrograph.

[Sources] - FloodAstra Gurugram reviewed-event city map: local project asset `gurugram_extreme_combined_keyframe.png`.

[Sources] - `event_mass_balance.json`, `event_summary.json` and `exchange_report/fig_hyetograph.png`, run `20260810T021439Z_july2025_critical_3h_107mm_twmoderate`. - Modelled rainfall input 30,976,362 m3; external boundary inflow 503,704 m3; total sources 31,480,066 m3.

[Sources] - Reviewed forcing manifest and hyetograph for event `20260810T021439Z_july2025_critical_3h_107mm_twmoderate`.

[Sources] - `event_package/ward_flow_matrix.json` and `sector_flow_matrix.json`, run `20260810T021439Z_july2025_critical_3h_107mm_twmoderate`.

[Sources] - Complete D8 natural-slope partition at 10 m: 125 subcatchments covering 298.23 km2 of the reviewed Gurugram model domain. - Reviewed V2 terrain is used within Gurugram. FABDEM extends the terrain beyond the municipal mask so the city boundary is not treated as an artificial hydraulic outlet. - Receiving areas: Leg 1 23.0 km2; Leg 2 45.0 km2; Leg 3 / Badshahpur Drain 146.6 km2; Leg 4 35.3 km2; direct natural or external paths 48.4 km2. - Colours identify the first mapped trunk-leg corridor reached downstream. The classification is terrain connectivity screening, not surveyed inlet capture or event-specific discharge.

[Sources] - D8 natural-slope screening at 10 m on reviewed V2 terrain, FABDEM, Cartosat and SRTM. - Every terrain cell whose flow path reaches a 30 m corridor around the complete mapped Leg 1, Leg 2 or Leg 3 geometry is included. - `trunk_drain_legs.geojson` declares geographic coordinates although its vertices are UTM Zone 43N metres. This diagnostic corrects the CRS label in memory; it does not silently alter the production network source. - Catchments indicate terrain connectivity, not surveyed inlet capture, event discharge or the fraction hydraulically entering each leg. Overlap is possible and the three areas must not be added.

[Sources] - User-confirmed Sheetla Mata Mandir Road coordinate: 28.480884423197487, 77.04529521447296. - Reviewed V2 field-reportable result from event `20260810T021439Z_july2025_critical_3h_107mm_twmoderate`. - Nearest reviewed model cell: 7.75 m; point peak depth 0.319 m; duration above 5 cm 250.45 min. - Within 50 m: maximum peak depth 0.551 m and peak surface-water volume 2,353 m3. - Within 250 m: peak surface-water volume 52,608 m3 (5.26 crore L). - Within 500 m: peak surface-water volume 107,980 m3 (10.80 crore L).

[Sources] - https://international.kk.dk/sites/default/files/2021-09/Cloudburst%20Management%20plan%202010.pdf - https://www.ktr.mlit.go.jp/edogawa/edogawa00576.html - https://www.worldbank.org/en/results/2026/01/26/strengthening-flood-resilience-in-rapidly-growing-cities - https://www.waterconnect.sa.gov.au/Content/Publications/DEW/Managed%20Aquifer%20Recharge%20Schemes%20in%20Adelaide_Final.pdf - https://www.ocwd.com/gwrs/about-gwrs/ - https://cgwb.gov.in/cgwbpnm/publication-detail/362