Cities rarely fail one infrastructure system at a time.
A power outage can stop pumps.
A failed telecommunications network can disable payment, dispatch and remote control.
A flooded substation can close a rail line, interrupt mobile service and disable a wastewater plant whose backup generator has limited fuel.
A damaged bridge can prevent repair crews from reaching the very utilities that depend on restoration.
This is the interdependency problem: infrastructure networks are individually planned, financed and regulated, but essential services are delivered by combinations of them.
The issue is now explicit in current policy. At its May 2026 Roundtable on Governing for Infrastructure Resilience, the OECD argued for a systems approach to manage interdependencies across networks, jurisdictions and supply chains and framed continuity of water, transport, energy, telecommunications, health and education as a core resilience objective. The OECD and World Bank’s March 2026 Compendium of Good Practices on Quality Infrastructure similarly emphasises risk-informed, lifecycle resilience rather than restoring damaged assets exactly as they were. In July 2026, the European Commission issued guidance for strengthening critical-infrastructure resilience against natural and human-made hazards and cross-sector risks. UN-Habitat’s 2026 work on intermediary cities and WUF13 urban systems also treats housing and essential services as interconnected systems rather than isolated sectors.
The reader job is therefore:
How should planners map which essential services depend on which other systems, identify nodes whose failure can cascade across the city, protect critical co-locations and access routes, set restoration priorities, test backup assumptions, and make capital plans reduce shared failure rather than accidentally concentrating it?
This article owns the spatial interdependency-and-cascade job.
It does not replace the Plan Integration Scorecard, individual power/water/transport/telecom owners, emergency management, cybersecurity, public finance, government or civilisation pages. Those own their systems. The Critical Infrastructure Interdependency Map asks one narrower advanced question: what happens between those systems when one stops working?
1. Start with essential services, not asset inventories
A city can catalogue substations, water plants, hospitals and bridges without understanding interdependency. Begin with services people need: drinking water, wastewater, electricity, communications, mobility, healthcare, food logistics and emergency response. Then trace the assets and external systems required to deliver each service. Service-first mapping reveals dependencies that asset-by-asset inventories miss.
2. Distinguish dependency from interdependency
A water plant may depend on electricity. If the power utility also depends on that water system for cooling or fire protection, the relationship becomes interdependent. Direction matters. Maps should use arrows, not just overlapping coloured layers. A one-way dependency creates a different failure path from a feedback loop.
3. Map direct dependencies first
Examples include pumps needing electricity, telecom towers needing power, hospitals needing water, and electric rail needing substations. These relationships are often known operationally but scattered across agencies. A shared map should capture them in a consistent form before attempting complex modelling.
4. Then map second-order dependencies
A hospital may have backup power, but the fuel supplier depends on roads and payment systems. A water utility may have generators, but operators need telecommunications to coordinate. Second-order dependencies explain why a system that appears redundant can still fail during a prolonged event.
5. Time-to-failure is as important as dependency itself
Some systems fail instantly when power disappears. Others have batteries, reservoirs or fuel for hours or days. Record the buffer. A dependency map that says ‘telecom depends on power’ is incomplete unless it also shows how long critical sites can operate without grid supply.
6. Recovery time matters too
A service may fail quickly but recover rapidly, or remain unavailable for weeks because specialised equipment must be replaced. Planning should distinguish immediate continuity from long-duration reconstruction. The OECD’s current infrastructure-resilience work emphasises the full lifecycle from prevention through recovery.
7. Identify nodes that serve several systems
A bridge may carry traffic, water mains, fibre and electricity. A tunnel may contain multiple utilities. A single flood at that node can create simultaneous failures. Multi-system nodes deserve higher visibility in capital and hazard planning because their importance exceeds any one owner’s asset rating.
8. Co-location can improve efficiency and increase common-mode risk
Shared corridors reduce land consumption and construction disruption. They also allow one landslide, fire or excavation incident to damage several networks. The correct response is not to ban co-location. It is to identify common-mode hazards and design separation, protection or alternative routes where the consequences justify it.
9. Power is a dependency multiplier
Electricity supports pumps, controls, lighting, communications, refrigeration, charging and many building systems. Yet treating power as the only critical layer is too simple. Grid restoration itself depends on communications, transport access, fuel, workers and functioning control systems. The map should show that reciprocity.
10. Water systems have hidden energy dependencies
Extraction, treatment, pumping and wastewater processes consume power. Gravity systems may provide some resilience; high-lift systems may fail quickly. Water utilities should identify which facilities have backup power, how long fuel lasts and which pressure zones are most sensitive to loss.
11. Wastewater failure can become a public-health and environmental cascade
If lift stations or treatment plants lose power, sewage can overflow, waterways can become contaminated and hospitals or dense districts may face sanitation problems. The Sewer Capacity owner remains canonical for normal capacity. The interdependency map focuses on what happens when supporting systems fail.
12. Telecommunications is an operational dependency across almost every sector
Utilities use fibre, mobile networks, radio and control links. Emergency agencies dispatch through communications. Residents receive warnings digitally. Payment and logistics systems depend on connectivity. Telecom should therefore be mapped not merely as a consumer service but as enabling infrastructure for other critical functions.
13. Telecom backup should be site-specific
A mobile network may remain available overall while key towers or exchanges fail because batteries are depleted or generators cannot be refuelled. The map should identify critical facilities, power arrangements and restoration access. General claims of ‘network redundancy’ are too broad for local resilience planning.
14. Digital dependence should be separated from cybersecurity
A control system can fail because electricity or fibre is lost even when no cyberattack occurs. Cybersecurity is a distinct specialist domain. The interdependency map records which services rely on digital communications and control so that physical and cyber resilience plans can read the same dependency structure.
15. Transport is both a service and a repair enabler
Roads, bridges, rail and ports move people and goods. They also move utility crews, replacement transformers, fuel, chemicals and emergency equipment. A damaged transport link can prolong failures in otherwise undamaged infrastructure. Restoration-route planning should therefore be part of infrastructure resilience.
16. Bridge criticality should include utilities it carries
Traditional transport analysis may rank a bridge by traffic volume. If it also carries a trunk water main and fibre corridor, its system value is much greater. Capital prioritisation should recognise bundled dependencies rather than evaluating only vehicle delay.
17. Fuel supply is a hidden dependency of backup power
Generators create confidence until a disruption lasts longer than on-site fuel. Refuelling depends on roads, depots, contracts, drivers and sometimes electric pumps at the fuel terminal. A resilience audit should ask how backup fuel reaches priority facilities during the same event that caused the outage.
18. Hospitals are infrastructure hubs as well as buildings
A hospital depends on power, water, telecoms, medical gases, waste removal, staff access and supply chains. The health-facility owner remains canonical for healthcare planning. The interdependency map identifies external systems whose failure can reduce hospital capability even when the building itself is intact.
19. Emergency services depend on ordinary infrastructure too
Fire stations need water, fuel, communications and passable roads. Ambulances need hospitals and transport routes. Police need communications and power. Emergency response should not be modelled as an unlimited external rescue resource; it is itself a dependent network.
20. Schools can become emergency facilities without being designed as critical infrastructure
Schools are sometimes used as shelters, distribution sites or cooling centres. If a resilience plan assigns that role, the building’s power, water, sanitation, communications and accessibility should be tested for emergency operation. The school allocation owner remains distinct; interdependency planning evaluates the emergency function actually assigned.
21. Shelters need service continuity, not only floor area
A hall that can hold 500 people may have toilets, cooling or backup power for far fewer during an outage. Capacity should be based on the infrastructure condition under which the shelter will operate. Emergency occupancy changes service demand.
22. Food systems depend on power, transport and telecoms
Cold storage needs electricity. Supermarkets need payment and logistics systems. Warehouses need roads and staff. A prolonged outage can become a food-access problem even if farms and national inventories are unaffected. The Food Map remains canonical; interdependency planning adds the failure pathways.
23. Payment systems can influence physical recovery
Fuel, food and repair transactions can be disrupted when digital payment or banking systems fail. Cash contingency and priority procurement are institutional responses. The planning map does not regulate finance systems; it identifies where physical service continuity assumes electronic commerce will remain available.
24. Critical facilities should have dependency profiles
For each hospital, water plant, data exchange, emergency centre or major station, record required external services, backup duration, access routes and restoration priority. A concise dependency profile is more actionable than a citywide map alone. It also creates a consistent basis for comparing facilities.
25. Dependencies should be ranked by consequence and replaceability
Some dependencies are severe but have substitutes. Others are modest yet irreplaceable. A hospital can sometimes use tanked water briefly; a rail signalling system may have no manual alternative. Rank dependencies by consequence, buffer, alternative and restoration time rather than by technical sophistication.
26. Common suppliers can create hidden concentration
Several utilities may rely on the same fuel depot, telecom provider, contractor or chemical supplier. That shared upstream dependency may not appear on infrastructure maps. Procurement and supply-chain information should therefore be included at a level appropriate for public planning and security.
27. Chemical supply can constrain water treatment
Drinking-water and wastewater plants may depend on chlorine, coagulants or other treatment chemicals delivered by road. A transport disruption can become a water-service problem days later. Inventory duration and alternate suppliers belong in the dependency profile.
28. Specialised replacement equipment can dominate recovery time
A damaged transformer, pump or control component may have a long manufacturing lead time. Utilities should identify high-consequence equipment and spares strategies. The Maintenance Ledger owner governs asset management; the interdependency map shows which external services remain unavailable while that component is missing.
29. Workforce availability is an infrastructure dependency
A system can survive physically but fail operationally if staff cannot reach it because of flood, evacuation or transport shutdown. Shift coverage, remote operation and staff accommodation can matter. People are part of infrastructure resilience.
30. Housing location can influence recovery workforce availability
If utility and emergency workers live across one vulnerable bridge or floodplain, access risk can be concentrated. Planning should not micromanage where employees live. It can identify transport chokepoints and provide alternative access or staging for priority workers.
31. Critical routes should be mapped for restoration crews
Priority road-clearing and bridge inspection plans should include routes to substations, water plants, telecom exchanges, hospitals and fuel depots. Restoring general traffic alone may not restore essential services fastest. A small local road can have high system importance if it is the only access to a critical node.
32. Debris management can affect restoration speed
Storm or earthquake debris can block utility sites and roads. Temporary debris storage also needs land. The Emergency and waste owners retain their systems. Interdependency planning identifies where debris clearance unlocks several services at once.
33. Flood maps should be overlaid with dependency nodes
A flooded substation has one consequence; a flooded substation feeding the water plant, emergency centre and telecom exchange has another. Hazard planning should prioritise protection or redundancy based on system consequence, not asset exposure alone.
34. Heat can create simultaneous demand and failure
Extreme heat raises electricity demand, stresses transformers, increases water use and drives people toward cooling centres. If power fails, health demand rises at the same time as cooling disappears. Interdependency analysis should include compound demand shocks, not only physical damage.
35. Drought can connect water and energy stress
Low water availability can constrain hydropower or thermal generation in some regions while high temperatures increase electricity demand. Water restrictions can affect industry. The Drought Capacity and energy owners remain canonical. The interdependency map shows cross-sector stress pathways.
36. Wildfire can isolate corridors without destroying the city core
A fire may close transmission lines, roads or telecom routes serving an urban area outside the burn zone. Resilience plans should include remote infrastructure that the city depends on. Municipal boundaries are poor proxies for service geography.
37. Earthquakes can create correlated failures across buried networks
Water, gas, power and telecom lines may cross the same fault or liquefaction zone. The Seismic Ground Map remains the hazard owner. Interdependency analysis identifies where several essential services share the same physical vulnerability and whether alternate routes exist.
38. Cyber incidents can become physical-service cascades
A cyberattack on operational technology can stop pumps, dispatch or grid functions. CISA’s 2026 guidance on isolating vital operational technology reflects the importance of knowing connections and separation points. Planning should not design cyber architecture, but resilience teams can use the same asset-and-dependency map to understand physical consequences.
39. Manual fallback should be tested, not assumed
Plans often state that a process can be operated manually. That may require staff, local controls, paper procedures and safe access that no longer exist. Exercises should verify actual fallback capability. A dependency should only be marked ‘mitigated’ when the alternate mode is credible.
40. Backup generators should be tested under full operating demand
A generator sized for life safety may not run all process equipment. A water plant may maintain basic pumping but not full treatment. Critical facilities should state what service level backup power actually supports. Partial continuity can be enough, but the map should not label it as full redundancy.
41. Batteries change the time profile of failure
Battery backup can bridge short outages and support graceful shutdown. It may not sustain multi-day operations. The dependency profile should record duration under realistic load. This allows emergency managers to see which services will fail first if restoration is delayed.
42. Microgrids can reduce some dependencies and create new ones
A critical district with local generation and storage may island from the grid. It can still depend on fuel, controls, maintenance or weather. The Microgrid District owner remains canonical. Interdependency mapping evaluates whether the claimed resilience survives the same hazard that interrupts the main grid.
43. Water storage can act as infrastructure battery
Reservoirs and tanks provide time between power loss and service failure. Their value depends on level at the start of an event, pressure zones and demand. Utilities should translate storage into hours of critical service under emergency assumptions. This makes a physical buffer operationally legible.
44. Redundancy should avoid the same hazard footprint
Two power feeders are not true geographic redundancy if both cross the same floodplain or tunnel. Two telecom carriers may share one fibre duct. Backup pathways should be tested for common-mode exposure. Apparent duplication can conceal one physical point of failure.
45. Diversity can be more valuable than duplication
A second identical system may share the same vulnerabilities. A different technology or route can provide stronger resilience: radio plus fibre, gravity plus pumping, road plus rail, local storage plus just-in-time supply. The appropriate mix depends on service needs and cost.
46. Critical co-locations should have explicit resilience design
Some cities deliberately cluster emergency centres, hospitals and utility control facilities. Co-location can improve coordination but concentrate risk. Site selection should consider independent power, communications, access and hazard exposure. A civic campus should not become a single disaster target.
47. Dispersal has operating costs too
Spreading critical functions across the region can reduce common-mode failure but increase staffing, maintenance and coordination. The objective is not maximum dispersal. It is a deliberate architecture of concentration, redundancy and alternate sites.
48. Capital plans should score interdependency benefit
A project that protects one substation may also protect the pumps, telecom nodes and hospital circuits it serves. Its resilience value is larger than the utility’s own asset score. Capital prioritisation can include cross-sector service benefits while public-finance owners retain budget decisions.
49. Projects can accidentally increase interdependency risk
A new smart-control system may centralise operations. A shared utility tunnel may concentrate networks. Electrifying backup systems may reduce fuel dependence but increase reliance on the grid. Every major capital project should ask whether it creates a new single point of failure.
50. New development should report critical-service dependency where scale justifies it
A giant data centre, hospital, industrial campus or high-rise cluster can create new power, water and telecom dependencies. The Concurrency Test remains canonical for normal capacity. Interdependency review asks whether the development also changes emergency-restoration priorities or creates concentrated service risk.
51. Regional growth should avoid building over restoration corridors
Utility access roads, helicopter landing areas, staging sites and temporary repair compounds can be lost to development. Strategic plans should protect practical access to major infrastructure. Resilience requires space for repair as well as space for the operating asset.
52. Temporary staging land should be pre-identified
After a major event, utilities may need places for crews, transformers, pipe, fuel and debris. Pre-identifying sites reduces improvisation. Locations should have transport access and avoid hazard zones. They can serve normal uses until activated through emergency agreements.
53. Mutual-aid routes are part of the dependency map
Outside crews and equipment often support restoration. Their arrival depends on regional roads, airports, ports and accommodation. A city can be internally well prepared yet isolated from mutual aid. Regional connectivity should therefore be included in recovery planning.
54. Restoration priorities should be service-based
The first asset repaired should not simply be the one easiest to fix or politically visible. Restoration sequencing should ask which repair restores the most essential services and which downstream systems are ready to receive service. A power feeder to a water plant may unlock more city function than a larger residential circuit initially.
55. Restoration sequence must avoid sending service into unsafe systems
Power should not be re-energised blindly where flooded equipment remains unsafe. Water pressure should not be restored before broken mains are isolated. Interdependency planning needs coordination between sectors so one utility’s recovery does not create another incident.
56. Exercises should test cascading scenarios
A tabletop exercise that assumes only power failure misses how cities actually behave. Test power plus telecom loss, flood plus blocked roads, heat plus water stress, or cyber disruption plus manual operations. Scenario design should reflect realistic compound hazards and known dependencies.
57. Exercises should include private operators
Telecommunications, fuel, logistics, hospitals and utilities may be privately operated. Public plans that exclude them will miss key dependencies. Information-sharing arrangements should respect commercial and security constraints while still supporting coordinated restoration.
58. Maps should protect security-sensitive detail
A public resilience map should not expose vulnerabilities that create security risk. Maintain appropriate access levels: public service-level information, restricted operational maps and specialist technical data. Transparency and security are design questions, not binary choices.
59. Data standards should allow agencies to exchange dependency information
One utility may store assets by engineering ID, another by address, another by GIS geometry. A shared resilience layer needs consistent location, service and dependency fields. The goal is not one giant database replacing agency systems; it is enough interoperability to understand cross-sector failure.
60. Dependency data should have owners and update cycles
Infrastructure changes constantly. A map created for one resilience plan becomes obsolete unless agencies update critical nodes, backup durations and routes. Assign data stewardship. Stale dependency data can be more dangerous than an acknowledged gap because it creates false confidence.
61. Unknown dependencies should be recorded explicitly
Not every relationship will be known. Mark uncertainty and investigate high-consequence gaps. A mature system distinguishes ‘no dependency’ from ‘dependency not yet assessed.’ This prevents absence of data from being interpreted as resilience.
62. A citywide map should be complemented by district playbooks
Downtown, port, hospital district and remote suburbs may have different failure sequences. District playbooks can identify local critical nodes, shelters, water points and access routes while remaining connected to the citywide system. Scale matters.
63. Interdependency crosses municipal borders
Power, water, telecom and transport systems often serve metropolitan regions. A city may depend on a reservoir, substation or data exchange in another jurisdiction. Regional governance is necessary because local land-use decisions can affect services far beyond local boundaries.
64. Cross-border national infrastructure requires higher-level coordination
Pipelines, electricity interconnectors, ports and telecom cables can cross countries. Local planners should know their regional importance without attempting to govern international systems. The dependency map can show external reliance and escalation routes to higher authorities.
65. Insurance and finance can influence restoration capacity
Operators need resources to repair damaged systems. Financial resilience belongs to its own governance domain. Interdependency planning can identify where prolonged restoration is likely because a critical provider lacks spare equipment, contractors or financing, then elevate that risk to the appropriate authority.
66. Equity should be measured through service interruption, not only asset location
A neighbourhood may contain no critical facility yet suffer the longest outage because it sits at the end of a network or lacks alternatives. The EJ Zoning Disparity Test remains a land-use diagnostic. Infrastructure equity should also track outage duration, access to backup services and restoration sequence.
67. Critical-service islands can support vulnerable communities
Libraries, community centres or schools with resilient power, water and communications can provide charging, cooling, information and basic services. Site selection should consider walking access and neighbourhood vulnerability. These hubs complement rather than replace systemwide resilience.
68. Public information systems need offline alternatives
Websites and apps may fail when telecom or power is disrupted. Plans should include radio, physical notice points, staffed centres or other channels. The communication system itself belongs inside the dependency analysis.
69. A worked example: flooded substation cascade
A river flood disables one substation. That circuit powers a wastewater lift station, mobile tower cluster and rail signalling. The interdependency map had identified the node, so flood protection and alternate feeds were prioritised before the event. During the flood, the water utility shifts pumping, telecom batteries bridge the outage and rail follows a predefined degraded-service plan. One asset failure does not become a citywide surprise.
70. A worked example: bridge closure blocks restoration
An earthquake leaves utilities intact on one side of a river but damages the only heavy-vehicle bridge. Replacement transformers and fuel trucks cannot reach a critical district. The resilience plan had identified a secondary route and staging yard, allowing crews to transfer equipment. Transport redundancy becomes utility resilience.
71. A worked example: prolonged heat outage
Extreme heat drives peak electricity demand and a grid failure. Cooling centres open, but several rely on the same failed feeder. A pre-event dependency review had designated some centres on separate circuits with batteries and backup cooling. The city directs vulnerable residents toward facilities that actually remain functional rather than shelters that exist only on a paper map.
72. The Critical Infrastructure Interdependency workflow
Step 1 — define essential services. Step 2 — map direct dependencies. Step 3 — map second-order and shared suppliers. Step 4 — record buffer and time-to-failure. Step 5 — identify multi-system nodes and common-mode hazards. Step 6 — map repair routes and workforce dependencies. Step 7 — test backups and alternate modes. Step 8 — rank cascade consequence. Step 9 — align capital plans. Step 10 — set restoration sequences. Step 11 — exercise compound scenarios. Step 12 — update the map after system changes.
73. An Interdependency audit
Ask whether essential services are defined; dependency arrows have direction; second-order dependencies are mapped; backup duration is known; power, water and telecom reciprocity is visible; bridges and corridors include carried utilities; fuel and chemical supply are included; hospitals and emergency services have dependency profiles; single points of failure are identified; redundant routes avoid common hazards; restoration access is protected; staff and mutual aid can reach sites; capital projects are screened for new concentration; cross-border dependencies are recognised; equity includes outage duration; data ownership and security levels are defined; and compound exercises validate assumptions.
74. The deepest test is whether a failure is surprising only once
No city can eliminate every cascade. The purpose of an interdependency map is to turn hidden relationships into planned ones. When an outage reveals that a water plant depended on one telecom route, that lesson should change the map, capital plan and exercise. The system succeeds when each disruption produces institutional learning: fewer unknown dependencies, more deliberate redundancy, clearer restoration priorities and infrastructure investments that strengthen the city as a connected organism rather than protecting one asset while leaving the next link exposed.
75. Restoration priorities should be agreed before the emergency
Utilities and emergency managers can identify priority service areas, critical facilities and sequencing principles in advance. During a crisis, exact conditions will change, but a pre-agreed framework reduces competition driven by whichever organisation calls loudest. Priorities should be based on life safety, public health, cascade prevention and restoration leverage, then adapted to the real event.
76. Priority lists should not become rigid scripts
A hospital may be the highest priority in one event but already operating normally on resilient power. A water plant may become the new bottleneck. The restoration framework should combine standing priorities with current system status. Resilience requires disciplined adaptation rather than automatic adherence to yesterday’s ranking.
77. Critical loads should be defined within facilities
A hospital, shelter or water plant may not need every circuit during an emergency. Identifying critical internal loads can extend backup duration and reduce generator or battery requirements. Building and utility engineers own the technical design. The interdependency map records the service level that remains available under degraded operation.
78. Fuel contracts should be tested against regional demand spikes
Many critical facilities may hold priority-fuel contracts with the same supplier. During a widespread outage, all of them can call simultaneously. A citywide audit should identify concentration and whether suppliers have enough trucks, depot access and fuel to honour priority commitments. Paper priority is not the same as physical delivery capacity.
79. Portable generation needs connection points before the outage
A utility may own mobile generators but lack safe, standardised ways to connect them at critical sites. Pre-installed transfer equipment and staging plans can make temporary power useful. Planning and capital programmes should identify where a modest pre-event investment dramatically shortens recovery.
80. Water tanker plans need loading and distribution sites
If a water network fails, emergency tankers require functioning source points, routes, traffic control and public distribution locations. A tanker fleet alone is not a water-continuity plan. The interdependency map can identify which distribution sites remain accessible under the hazards likely to break the network.
81. Telecom priority power sites should be coordinated with fuel and access plans
Some mobile sites and exchanges are far more important to emergency communications than others. Operators can identify priority restoration nodes under appropriate information-sharing arrangements. Road clearing and fuel delivery can then include those sites without publishing sensitive technical detail.
82. Data centres that support public systems should be distinguished from commercial load generally
A city may depend on particular cloud, colocation or municipal computing facilities for dispatch, permitting, records or utility control. The Data Center Zone owner remains the land-use route. Interdependency planning asks which public services fail if those specific digital facilities or their network connections are unavailable.
83. Cloud redundancy can still share physical dependencies
Two digital services may run in different logical environments but depend on the same regional fibre route, power grid or cooling-water system. Resilience claims should examine physical geography as well as software architecture. Logical redundancy can conceal local common-mode risk.
84. Satellite communications can provide a distinct fallback
Emergency agencies and some utilities can use satellite links when terrestrial networks fail. Capacity, equipment, line-of-sight and trained users still matter. An alternate technology is only useful if equipment is deployed, tested and included in exercises rather than stored unopened for the first crisis.
85. Public charging infrastructure becomes a resilience issue as fleets electrify
Electric buses, service vehicles and emergency-support fleets may depend on chargers that fail with the grid. The Charging Map remains canonical for ordinary deployment. Interdependency planning examines emergency fleet requirements, backup charging, depot resilience and whether electrification has replaced fuel dependence with an unrecognised electrical single point.
86. Electrification can improve resilience when paired with distributed energy
Electric systems can use local solar, batteries or microgrids and avoid disrupted fuel deliveries. The effect depends on design. The interdependency map should not treat electrification as inherently more or less resilient; it should compare actual dependencies under the relevant hazard.
87. Gas networks can be both dependency and hazard
Buildings, industry and power generation may depend on gas while emergency response may require isolation after earthquakes or fires. The relevant utility owner manages network safety. Interdependency planning records which essential services depend on gas and what happens when isolation is necessary.
88. District energy can concentrate thermal service
A district heating or cooling network can efficiently serve hospitals, housing and institutions. A central plant failure may also affect many buildings at once. The Thermal Network owner remains canonical. Resilience planning should test backup plants, loop configuration, power dependency and restoration access.
89. Sewage pumping priorities should reflect public-health consequence
Not every lift station has equal system importance. Some serve hospitals, dense districts or low-lying areas where overflow quickly reaches homes or waterways. Utilities can rank stations by consequence so temporary generation and repair crews are deployed where they prevent the largest cascade.
90. Stormwater systems can depend on electricity at the worst possible moment
Pumped drainage and movable flood barriers may require power precisely during storms that threaten the grid. Backup duration, protected electrical equipment and manual alternatives should be included in flood-resilience plans. The hazard owner defines flood conditions; the interdependency map reveals the power link.
91. Reservoir and dam operations depend on communications and control
Remote water infrastructure may rely on telemetry, power and road access. A communications loss can force conservative manual operation or reduce situational awareness. Operators should identify local fallback procedures and critical access. Regional planning should recognise remote assets whose failure has urban consequences.
92. Waste collection affects recovery and public health
After disasters, ordinary waste and debris can accumulate while roads are blocked and fuel is constrained. Collection failure can obstruct streets, attract pests and complicate shelter operations. The waste owner remains canonical. Interdependency mapping connects waste operations to fuel, transport, workforce and disposal-site access.
93. Ports and airports can be mutual-aid gateways
Regional recovery may depend on receiving generators, transformers, medical supplies or crews through ports and airports. Those gateways themselves rely on power, fuel, access roads and communications. Their dependency profiles should be included even when they sit outside the municipal boundary.
94. Supply-chain maps should focus on critical items, not every purchase
Attempting to map every supplier is unmanageable. Identify items whose absence stops essential service: treatment chemicals, transformer components, generator fuel, medical oxygen, communications equipment and specialised repair parts. For each, record inventory duration, alternate suppliers and transport dependency.
95. Interdependency scoring should remain interpretable
A sophisticated model can produce one composite ‘criticality’ number, but decision makers need to know why a node matters. Keep component scores visible: population served, dependent services, backup duration, alternate routes, hazard exposure and recovery time. Transparency helps capital decisions and prevents false precision.
96. Network models should be validated against real incidents
After an outage, flood or cyber event, compare predicted failure sequence with what actually happened. Did batteries last as expected? Were roads passable? Did mutual aid arrive? Update the model. A resilience map becomes more valuable when it is calibrated against lived operations rather than remaining a theoretical GIS product.
97. Near misses are valuable data
An outage narrowly avoided because one generator started, one valve held or one alternate route remained open can reveal a fragile dependency before catastrophe. Agencies should capture near misses and incorporate them into capital and maintenance priorities. Resilience learning should not wait for full failure.
98. The mature city treats interdependency as a permanent planning layer
Infrastructure relationships change as buildings electrify, communications move to new platforms, water systems add reuse, fleets change fuel and new hazards emerge. The interdependency map should therefore sit beside land-use, transport and hazard maps as a maintained planning layer. Its purpose is not to predict every cascade. It is to keep the city aware of which connections now matter enough to protect, diversify or rehearse.
99. Planning departments should know which land-use decisions change dependency topology
A rezoning can create a new hospital district, logistics hub, high-rise cluster or industrial campus whose service needs alter the network. Infrastructure agencies should receive those growth signals early. Interdependency analysis does not become a new permit for ordinary projects; it identifies large decisions that materially change restoration priorities, critical loads or common-mode exposure.
100. Utility projects should report whether they remove or move a single point of failure
A new main, feeder or fibre route may be advertised as resilience. The capital record should show which dependency it diversifies and whether the alternate path truly avoids the original hazard footprint. This makes resilience spending auditable and prevents duplicate infrastructure that follows the same vulnerable corridor from being counted as genuine redundancy.
101. Land for temporary service hubs should be protected from last-minute competition
During prolonged outages, cities may need water points, charging, cooling, communications, fuel staging and public information in accessible locations. Pre-identifying municipal or partner sites means those functions do not compete chaotically for parking lots during the emergency. Agreements can preserve ordinary use while specifying activation conditions.
102. Neighbouring jurisdictions should share restoration assumptions
One municipality may plan to receive water, emergency shelter or detour traffic from another during a crisis while the neighbour makes the same assumption in reverse. Regional exercises can expose these contradictions. Mutual support works only when capacity is not double-counted and cross-border routes remain operable.
103. Private continuity plans should connect to public priorities without becoming public property
Major operators hold commercially sensitive resilience information. Public agencies do not need every internal detail. They need enough to know service dependencies, backup duration, external support required and restoration interfaces. Structured information-sharing can protect security and commercial confidentiality while still allowing citywide cascade planning.
104. Performance targets should focus on service restoration time
Asset protection is important, but residents experience service. Useful targets can state how quickly critical water pressure, emergency communications, priority transport or hospital power should be restored under defined scenarios. Agencies can then choose the best mix of hardening, redundancy, spares and operational planning to meet the service objective.
105. The final map should make the city’s hidden contracts visible
Modern urban life rests on implicit contracts: electricity will power pumps, roads will deliver fuel, telecoms will coordinate repair, water will support hospitals, and workers will reach critical sites. The interdependency map turns those assumptions into explicit relationships that can be tested. Once visible, they can be diversified, protected, exercised or accepted consciously instead of being discovered under pressure.
Sources and further reading
- OECD, Roundtable on Governing for Infrastructure Resilience, 6 May 2026: https://www.oecd.org/en/events/2026/05/roundtable-on-governing-for-infrastructure-resilience.html
- OECD and World Bank, Compendium of Good Practices on Quality Infrastructure 2026: Rebuilding for the Future, 17 March 2026: https://www.oecd.org/en/publications/compendium-of-good-practices-on-quality-infrastructure-2026_6981eda5-en.html
- OECD, Policy Toolkit on Governance of Critical Infrastructure Resilience: https://www.oecd.org/en/publications/good-governance-for-critical-infrastructure-resilience_02f0e5a0-en/full-report/component-9.html
- European Commission, Guidance to strengthen resilience of critical infrastructure, 10 July 2026: https://home-affairs.ec.europa.eu/news/commission-issues-guidance-strengthen-resilience-critical-infrastructure-2026-07-10_en
- Ireland Department of Defence / Office of Emergency Planning, National Strategy on the Resilience of Critical Entities 2026–2029, 19 March 2026: https://www.gov.ie/en/department-of-defence/publications/national-strategy-on-the-resilience-of-critical-entities-20262029/
- UN-Habitat, Building resilience through disaster risk management in intermediary cities, 2026: https://unhabitat.org/building-resilience-through-disaster-risk-management-in-intermediary-cities
- WUF13 / UN system, Housing Resilience as an Urban Systems Challenge, 20 May 2026: https://www.unep.org/events/unep-event/housing-resilience-urban-systems-challenge
- CISA and international partners, CI Fortify — advice for isolating vital systems, 28 July 2026: https://www.cisa.gov/
Continue reading: Plan integration · Infrastructure resilience · Emergency planning · Full Town Planning Series Index.