Events  // Event Case Study Series

Winter Storm Uri
February 2021
Texas · The Grid Collapse

A displaced polar vortex drove unprecedented cold across Texas, triggering simultaneous failures across every fuel type in ERCOT's generation fleet, cascading through the natural gas supply chain, and bringing the entire Texas power grid within four minutes of a total uncontrolled collapse — an outcome that would have left 26 million people without power for weeks or months. The catastrophe that actually occurred — 246 deaths, 4.5 million homes without power, USD 80–130 billion in economic losses — was a governance failure as much as a meteorological one, repeating with greater severity a failure mode that had been explicitly warned about in 2011.

Duration
February 10–21, 2021 (peak Feb 14–18)
Geography
All 254 Texas counties; parts of 14 other states
Peak Outages
4.5 million premises (~10 million people)
Economic Losses
USD 80–130 billion (economic); ~USD 20B insured
Fatalities
246 official; excess mortality significantly higher
Grid Near-Miss
~4 minutes from total uncontrolled collapse
4 minEstimated margin from total uncontrolled grid collapse on the morning of Feb 15 — ERCOT operators testified the grid was "seconds to minutes" from black start
30 GWGeneration capacity lost at peak — equivalent to the entire electricity needs of Florida, on the coldest night of the year in Texas
203 hrsConsecutive hours below freezing at Waco — 53 hours more than the previous record; DFW recorded 139 consecutive hours below freezing
500,000Insurance claims filed — 85% residential property; primarily frozen and burst pipes across all 254 Texas counties
10 yrsSince the 2011 Texas freeze — when FERC issued identical weatherization recommendations that went unimplemented
#1Largest manually controlled load shedding event in U.S. history — 20,000 MW deliberately cut to prevent total grid collapse

Executive Summary

Winter Storm Uri is the most consequential infrastructure failure in modern U.S. history, and the most analytically important winter storm event for the cat modelling industry. It is not primarily a meteorological extreme — the temperatures reached in Texas during February 2021, while severe, did not set records in most locations. It is primarily a system failure event: the collapse of an electricity generation system that had been designed for summer peak demand, never adequately weatherised for winter extremes, warned of exactly this failure mode ten years earlier, and left structurally isolated from neighbouring grids in a way that prevented the import of emergency power when the failure occurred.

Texas failed to sufficiently winterise its electricity and gas systems after 2011. Feedback between failures in the two systems made the situation worse. Overall, the state faced outages of 30 GW of electricity as demand reached unprecedented highs. The gap between production and demand forced the non-profit grid manager ERCOT to cut off supply to millions of customers or face a systems collapse that by some accounts was minutes away.

For cat modellers, Uri represents an event type that sits awkwardly between categories. It is not a wind event, not a flood event, not an earthquake. It is a demand-surge coupled with supply-collapse driven by extreme cold — a failure mode that standard winter storm cat models, calibrated to ice and snow accumulation damage, were not designed to capture. The losses were dominated not by physical damage from precipitation but by the catastrophic secondary consequences of power outage: frozen and burst pipes, water damage from thawed pipes, deaths from hypothermia and carbon monoxide poisoning, and business interruption across an economy that had never built resilience against multi-day power loss in winter conditions.

Why Uri Is a Defining Event for Cat Modellers
Uri exposed a failure mode — cascading infrastructure collapse from winter cold — that was absent from standard cat model frameworks but capable of generating losses comparable to major hurricane events. The USD 10–20 billion in insured losses came almost entirely from a single damage mechanism (frozen and burst pipes) triggered not by the storm's physical meteorological impact but by the secondary consequence of power outage. Cat models that estimate winter storm losses from ice accumulation and snow loading systematically missed the dominant loss driver entirely.

Meteorological Analysis — A Displaced Polar Vortex

The Polar Vortex and Its Disruption

The polar vortex is a large area of low pressure and cold air surrounding the Earth's polar regions. In its normal configuration, it is contained in the stratosphere above the Arctic by the jet stream — a band of strong upper-level winds that encircle the pole and act as a barrier keeping Arctic air in the north. When the polar vortex is strong and well-organised, it keeps cold Arctic air locked near the poles. When it is disrupted — weakened and destabilised by sudden stratospheric warming events — it can displace southward, carrying Arctic air to latitudes far south of its normal extent.

A disrupted stratospheric polar vortex was linked to the weather pattern that triggered the extreme cold that hit Texas in February. The warmer overall weather diminished the polar vortex's typical course enough that it was pushed further south rather than remaining in polar areas, resulting in the February 2021 storms. This displacement of the polar vortex brought a sustained intrusion of genuine Arctic air to the southern United States — not the ordinary cold fronts that Texas experiences several times each winter, but a prolonged residence of air masses originating directly over the Arctic ice cap.

Why Texas Was Uniquely Vulnerable to This Pattern

The same polar vortex disruption that brought extreme cold to Texas simultaneously affected states across the central United States. But Texas suffered disproportionately for reasons rooted in its geography, energy infrastructure, and regulatory structure:

The Grid Failure — A Cascading System Collapse

The Energy Mix and Simultaneous Multi-Fuel Failure

One of the most important — and most widely misunderstood — aspects of the ERCOT failure is that it was not caused by any single energy source. Every major generation fuel type in ERCOT's fleet failed simultaneously.

Natural Gas — 51% of ERCOT capacity
~50%
Largest contributor to outages

Frozen wellheads, frozen gathering lines, frozen pressure sensors and instruments at power plants. Gas production fell ~50% as upstream infrastructure froze. Even plants with contracted gas supplies couldn't receive deliveries as pipelines lost pressure. The feedback loop between gas production loss and power plant failure was the core cascading mechanism.

Wind — 24.8% of ERCOT capacity
~10%
Smaller contributor than widely reported

Wind turbines froze — blades iced, pitch mechanisms seized, nacelles unable to orient. However, wind capacity had already been low in ERCOT's winter planning assumptions. Politically prominent but proportionally smaller contributor to the outage than natural gas failure.

Coal — 13.4% of ERCOT capacity
~20%
Significant contribution

Coal stockpiles froze solid. Coal conveyors and handling equipment seized. Several coal plants lost fuel delivery capacity entirely. Water lines serving steam generators froze, causing forced outages. Coal contributed proportionally more to outages than its generation share would suggest.

Nuclear — 4.9% of ERCOT capacity
~1 unit
Coolant sensor failure

One nuclear unit (South Texas Project Unit 1) shut down when a coolant water sensor froze, triggering an automatic safety shutdown. While a smaller contributor to total outages, the nuclear failure illustrated that no fuel type was immune from the consequences of insufficient weatherisation.

Overall, the state faced outages of 30 GW of electricity as demand reached unprecedented highs. Gas production declined by nearly 50%, which lowered pressure in the pipelines, making it harder for power plants fuelled by natural gas to operate. The interaction between the power outage and the gas supply failure created a catastrophic feedback loop: power plants needed gas to generate electricity; gas wells and processing facilities needed electricity to operate pumps, heaters, and instrumentation; as power failed, gas production fell further; as gas production fell, more power plants lost fuel and shut down. Each failure made the next failure more likely.

The Cascade — From Cold to Near-Collapse

// The ERCOT Cascade — How Cold Almost Collapsed the Grid

Sustained Arctic cold penetrates Texas — unprecedented duration

Polar vortex displacement brings Arctic air to Texas for 7+ days. Temperatures across the state average 12°F on Feb 15 — 6°F colder than Alaska that same day. Infrastructure designed for mild winters begins failing progressively.

Natural gas production fails — upstream freeze-off

Wellheads, gathering lines, and processing equipment freeze across the Permian Basin, Eagle Ford, and other Texas production areas. Gas production falls ~50% from peak. Pipelines lose pressure as injection volumes drop.

Gas-fired power plants lose fuel — first generation outages

Natural gas plants begin losing fuel supply as pipeline pressure falls. Simultaneously, instruments, sensors, and cooling water systems at plants freeze, forcing additional outages from equipment failure. 51% of ERCOT's generation fleet begins to derate or trip offline.

Demand surges — heating load hits record 69,000 MW

As temperatures crash, electricity demand for heating surges to 69,000 MW — a winter record never previously approached in ERCOT planning scenarios, which had been calibrated to the 2011 event as the planning extreme.

Grid frequency begins to fall — approaching automatic cascade threshold

Grid frequency sagged below 59.4 Hz. Operators later testified the system was roughly four minutes from automatic protective cascade: relays across the fleet tripping generators offline to save the machines themselves — an uncontrolled collapse to black. A full black start of an islanded 90-GW grid was estimated at weeks to months of restoration time.

ERCOT initiates emergency load shedding — 20,000 MW deliberately cut

ERCOT ordered a total of 20,000 MW of rolling blackouts — the largest manually controlled load shedding event in U.S. history. This deliberate cut stabilised grid frequency above the cascade threshold, preventing total collapse — but left 4.5 million premises without power, some for four days, in sub-freezing temperatures.

Power outage causes gas production to fall further — amplifying feedback

As electricity was cut to rotating sectors, natural gas production facilities that needed power for pumps, heaters, and instrumentation lost electricity too, further reducing gas supply to the remaining operational power plants. The outage that was meant to stabilise the grid simultaneously worsened the fuel supply crisis that caused the outage.

Water infrastructure fails — pipe burst losses compound human toll

Without heat, building temperatures fall below freezing. Pipes freeze and burst across all 254 Texas counties. Water treatment plants lose power. Boil-water notices are issued across the state. The secondary water damage from burst pipes — which constitutes the majority of insured losses — begins accumulating across millions of residential and commercial properties.

The Four-Minute Margin — How Close Was Total Collapse?

The most haunting detail of the Uri crisis is how close the entire Texas power grid came to a total, uncontrolled collapse. At the depth of Uri, grid frequency sagged below 59.4 Hz, and operators later testified the system was roughly four minutes from automatic protective cascade: relays across the fleet tripping generators offline to save the machines themselves — an uncontrolled collapse to black. A black start of an islanded 90-GW grid was estimated in weeks to months.

To understand why this matters so profoundly: the difference between the blackout that actually occurred and the blackout that was narrowly avoided is the difference between days and months. The rolling outages that left 4.5 million people without power for up to four days were catastrophic. A total black start — where every generator on the grid trips offline and must be individually restarted in sequence, with no external power source available because ERCOT is isolated — would have been an order-of-magnitude larger catastrophe. The operators who initiated the deliberate load shedding made the decision that prevented this outcome, at the cost of choosing who would and would not have power during one of the coldest weeks in Texas history.

"At the depth of Uri, grid frequency sagged below 59.4 Hz, and operators later testified the system was roughly four minutes from automatic protective cascade — an uncontrolled collapse to black. A black start of an islanded 90-GW grid was estimated in weeks to months, an outcome so far beyond any storm that operators shed millions of customers deliberately to prevent it."

— Analysis of ERCOT grid frequency data and operator testimony, Power Enterprises 2025

The 2011 Warning — A Governance Failure Repeated

Uri was not the first time that Texas's power infrastructure had failed during a winter cold event. A nearly identical — if less severe — failure had occurred ten years earlier, during a cold weather event in February 2011. The 2011 event caused rolling blackouts across Texas, disrupted natural gas production, and left millions without power for shorter periods than 2021 but through essentially the same failure mechanisms. And critically, it generated the same recommendations that went unheeded.

// The Weatherisation Warning That Was Ignored — 2011 to 2021

2011 February 2011 freeze — similar but less severe failure. Gas plants and wind turbines freeze. Rolling blackouts across Texas. FERC and NERC conduct joint investigation.
2011 FERC/NERC report issued — explicitly recommends mandatory weatherisation standards for power plants and natural gas infrastructure in cold-weather-vulnerable regions including Texas. Recommendations are voluntary, not mandated.
2011–21 Recommendations largely unimplemented. Texas failed to sufficiently weatherise its electricity and gas systems after 2011. Weatherisation remained voluntary. Market structure provided no financial incentive for generators to invest in cold-weather protection that would only be needed in rare events.
2021 Winter Storm Uri — identical failure mechanisms, 10× the severity. 30 GW lost. 4.5 million without power. 246 deaths. USD 80–130 billion in economic losses. The 2011 recommendations were issued again — this time as mandatory requirements.
Post-2021 Texas SB 3 and ERCOT weatherisation mandate — Texas legislature mandates weatherisation for power generators, with inspections. Critical gas infrastructure added to load-shed priority list. New winter reserve products created. Post-Uri reform was real — weatherisation is now mandated and inspected, critical gas infrastructure is protected in load-shed plans. Tail risk reduced but structural isolation unchanged.

The 2011-to-2021 failure to implement weatherisation recommendations is the most striking governance parallel in the entire case study series. It mirrors the Turkey earthquake story — where identical building code enforcement failures produced identical disaster outcomes in 1999 and 2023. The lesson in both cases is the same: identifying a vulnerability and issuing recommendations without creating regulatory or market incentives for compliance does not reduce risk. It documents the risk for the post-event inquiry.

The Insurance Loss — A Different Kind of Winter Storm Claim

Loss Mechanism — Pipes, Not Precipitation

The insurance industry's experience with winter storms had historically been dominated by claims arising from the direct physical effects of precipitation: ice accumulation causing structural collapse (roofs, trees, utility lines), snow loading on buildings, and vehicle accidents on icy roads. Uri produced a fundamentally different claim pattern. Overall, there were 500,000 claims, and 85% were property claims — burst pipes, roof damage. The dominant loss mechanism was frozen and burst pipes — an interior building failure caused not by direct weather impact on the structure but by the secondary consequence of power outage in buildings designed for a warmer climate.

This distinction matters profoundly for insurance coverage purposes. Property policies that cover "damage caused by freezing" respond to frozen pipe claims. But policies vary significantly in how they handle the specific scenarios that Uri created:

Loss Quantum — Insured vs. Economic

Compared to Hurricane Harvey, the state's most expensive storm that resulted in an estimated ultimate loss of around USD 20 billion, Winter Storm Uri left Texas with total damages of around half that at USD 10.346 billion in insured losses. The USD 10–20 billion insured loss range against economic losses of USD 80–130 billion represents a protection gap of approximately 85–90% — driven by several structural factors:

ERCOT Market Dislocation — The Uninsured Financial Catastrophe

The single largest financial consequence of Uri was not property damage — it was the electricity market dislocation that occurred when ERCOT's energy-only market was left at its price cap of USD 9,000/MWh for an extended period. This cap — set for emergency conditions — produced electricity bills for some Texas commercial customers in the hundreds of thousands of dollars for a single week. Tens of thousands of residents and small businesses sought damages from power generators after losing electricity during the storm that resulted in billions in losses.

Retail electricity providers who had contracted to supply customers at fixed prices but were forced to purchase replacement power at USD 9,000/MWh — versus their contracted supply prices of USD 50–100/MWh — faced immediate financial ruin. Several Texas retail electricity providers filed for bankruptcy in the weeks following Uri. The USD 4.4 billion in overcharges that courts subsequently found were applied during the storm period represents an entirely uninsured category of financial loss with no cat model equivalent.

Cat Model Performance — What Winter Storm Models Missed

The Infrastructure-Failure Loss Driver

Standard winter storm cat models estimate losses from direct physical effects of winter weather: ice accumulation on structures and trees, snow loading on roofs, wind damage from blizzard conditions, vehicle accidents on icy roads, and flooding from snowmelt. These models are calibrated to historical winter storm loss patterns from the northeastern United States, where severe winter weather is common and building stock and infrastructure are designed for it.

Uri's dominant loss mechanism — interior building damage from frozen and burst pipes resulting from extended power outage — sits outside the scope of most winter storm cat model frameworks. The model would need to:

  1. Estimate the probability and duration of power outages as a function of cold weather severity
  2. Estimate the proportion of structures in the affected area that will experience interior temperatures below 0°C if power is lost for a given duration
  3. Estimate the pipe burst probability and expected water damage for those structures
  4. Apply location-specific vulnerability adjustments for building stock designed for warmer climates

None of these steps were standard in pre-Uri winter storm cat models for Texas. The result was substantial underestimation of Texas's exposure to catastrophic winter loss — a gap that became apparent only when the failure scenario actually occurred.

The Interdependency Problem

The cascading failure pattern of Uri — cold weather → power grid failure → gas production failure → more power grid failure → building heating failure → pipe burst → water damage — represents a chain of interdependencies across multiple infrastructure systems that no standard cat model was designed to trace. This is the same compound risk problem identified in the Sandy case study: multiple systems interacting in ways that amplify the total loss far beyond what any single-system model would predict. Power system engineers do not model insurance losses. Insurance cat modellers do not model power grid cascades. The gap between these modelling disciplines is exactly where Uri's losses resided.

Chronological Record

Feb 10–11

First wave — ice and freezing rain across North and Central Texas

The first of three successive winter storms begins affecting Texas, with ice accumulation up to half an inch in some locations. Gov. Abbott declares a state of emergency in all Texas counties. ERCOT issues an advisory about predicted extreme weather and asks for voluntary conservation. Power plants begin experiencing initial equipment issues.

Feb 13

Generation outages begin — ERCOT plea for conservation

Power generation units begin experiencing outages. ERCOT issues a public plea for customers to reduce energy usage after generation cannot meet demand. Natural gas production begins declining as upstream infrastructure starts to freeze.

Feb 14–15
1:25 AM

ERCOT initiates emergency load shedding — grid four minutes from collapse

With demand surging to 69,000 MW and generation failing rapidly, ERCOT initiates emergency load shedding at 1:25 AM on February 15. Grid frequency has fallen below 59.4 Hz — minutes from the automatic cascade threshold. ERCOT later testifies the grid was approximately four minutes from an uncontrolled total collapse requiring weeks-to-months of restoration.

Feb 15–18

Peak outage — 4.5 million premises; average Texas temp 12°F

The average temperature across Texas on February 15 was just 12°F — 6°F colder than the average temperature in Alaska that same day. Four and a half million premises are without power. Some will remain without power for four days. Texans burn furniture for heat, shelter in cars running with windows cracked (leading to carbon monoxide deaths), and queue for hours at warming centres. Water pipes across all 254 counties begin freezing and bursting.

Feb 15–21

Water system failures — boil-water notices statewide

Without power, water treatment plants fail. Pressure drops across municipal water systems as frozen and burst pipes create massive simultaneous leaks throughout distribution networks. Nearly half of Texans — 49% — experience water disruptions. Statewide boil-water notices are issued. The secondary water damage from thawing burst pipes begins accumulating across homes and businesses that the homeowners may not even be able to access due to road conditions.

Feb 18–21

Temperatures recover — the thaw reveals the damage

As temperatures rise above freezing, burst pipes begin thawing and flowing. Buildings that appeared structurally intact are discovered to have water damage ranging from soaked ceilings and walls to catastrophic flooding of entire floors. The insurance claims process begins. Plumbers, restoration contractors, and building suppliers face extraordinary demand backlogs as supply chain constraints from the pandemic compound the recovery challenge.

2021–22

500,000 claims filed; SB 3 mandates weatherisation

The Texas legislature passes Senate Bill 3, mandating weatherisation inspections for power plants and adding critical natural gas facilities to load-shed protection lists. FERC and NERC issue their final report — essentially reissuing the recommendations of their 2011 report with mandatory rather than voluntary status. Over 500,000 insurance claims are filed across the state, with the last complex commercial claims taking years to resolve.

Comparison — Uri vs. Katrina: Two Infrastructure Failure Events

Factor Winter Storm Uri 2021 (Texas) Hurricane Katrina 2005 (Louisiana)
Primary failure mechanism Power grid collapse from inadequate weatherisation; cascading gas supply failure Levee system structural failure from inadequate engineering and maintenance
Prior warnings ignored 2011 FERC/NERC recommendations — identical failure mode, unimplemented Multiple engineering warnings about I-wall designs and levee inadequacy
Primary insured loss mechanism Frozen and burst pipes from extended power outage — interior water damage Storm surge flooding — typically excluded from standard property policies
Cat model failure type Infrastructure cascade not modelled; interior pipe-burst loss absent from winter storm models Levee reliability assumed binary; surge undermodelled; CBI underestimated
Protection gap ~85–90% — electricity market losses, CBI, agricultural losses largely uninsured ~60–70% — flood exclusions from standard policies; NFIP undersubscribed
Governance failure Voluntary weatherisation recommendations after 2011 — not implemented Levee design inadequacy tolerated; protection level misrepresented to public
Key regulatory response Texas SB 3 — mandatory weatherisation; FERC/NERC rule changes USD 14.6B HSDRRS reconstruction; NFIP reform discussions; post-Katrina Emergency Management Reform Act

The Broader Context — Texas's Structural Vulnerability

The ERCOT Isolation — Energy Independence as Risk

Texas's decision to maintain its own isolated electricity grid — operated by ERCOT and deliberately kept separate from the national Eastern and Western Interconnections to avoid federal regulation by FERC — is both a point of political pride and a structural risk factor that Uri exposed dramatically. When ERCOT lost 30 GW of generation, it could import only 3–4 GW from neighbouring grids through its limited DC tie connections. A Texas integrated into the national grid could theoretically have imported power from states that were not simultaneously cold-stressed, buffering the most acute phase of the supply-demand imbalance.

The tradeoff is real: full integration with the national grid would subject Texas's electricity market to FERC regulation, ending the deregulated energy-only market structure that Texas has maintained for decades. The political economy of maintaining grid isolation — even after Uri — has meant that this structural vulnerability remains essentially unaddressed despite the recognition that it amplified the 2021 crisis.

The Energy-Only Market and Weatherisation Incentives

ERCOT operates an energy-only market — generators are compensated only for electricity they actually produce, rather than receiving a capacity payment for being available. This market structure has economic advantages: it encourages efficient dispatch and avoids paying generators for capacity that is never needed. But it creates a systematic underinvestment in reliability measures — like winterisation — that provide value only in rare, tail-risk scenarios. A generator weighing the cost of winterising its plant against the probability of needing it in any given year faces an unfavourable economic calculation when the events are rare — even if the expected value of weatherisation is clearly positive when the full probability-weighted consequences are considered. Texas failed to sufficiently weatherise its electricity and gas systems after 2011. The 2021 freeze suggests a need to rethink the state's regulatory approach to energy.

Legacy — What Uri Changed

// LEGACY 01

Winter Storm Cat Models — Pipe Burst Module

Uri exposed the absence of infrastructure-failure-driven interior pipe burst losses from standard winter storm cat models. Post-Uri, vendors began developing explicit power-outage-to-pipe-burst loss modules — estimating outage probability and duration as a function of cold weather severity, then modelling pipe freeze and burst probabilities for buildings with interrupted heating, calibrated to climatic design standards of the affected building stock.

// LEGACY 02

Infrastructure Cascade Modelling

The gas-power feedback loop — where electricity outages reduced gas production, which reduced electricity generation, which further reduced gas production — is a type of coupled infrastructure failure that no standard cat model framework was designed to represent. Uri has accelerated development of coupled infrastructure failure models that trace loss propagation across interdependent utility systems simultaneously.

// LEGACY 03

Texas SB 3 and Mandatory Weatherisation

The Texas legislature mandated weatherisation standards for power generators and critical natural gas infrastructure. FERC and NERC established new cold-weather reliability standards. The reforms are real and have improved winter resilience — subsequent cold events have not repeated the 2021 failure at comparable scale, validating the effectiveness of the weatherisation requirement when actually enforced.

// LEGACY 04

Energy Market Design and Reliability

Uri prompted deep re-evaluation of whether energy-only electricity market designs adequately incentivise reliability investments in tail-risk scenarios. ERCOT introduced new winter reserve products and capacity constructs to compensate generators for availability during extreme weather — a structural change to the market design that attempts to correct the underinvestment incentive that the energy-only structure created.

// LEGACY 05

Climate-Zone Vulnerability in Cat Modelling

Uri demonstrated that structures designed for warmer climates are categorically more vulnerable to extended cold events than similar structures in northern climates — not because they are less well-constructed, but because their design philosophy assumes temperatures that were not maintained during Uri. Cat models must explicitly incorporate climate-zone vulnerability adjustments when estimating losses from out-of-design-envelope cold events.

// LEGACY 06

Governance Failure as a Cat Model Input

The failure to implement the 2011 recommendations — despite full knowledge of the vulnerability — is structurally identical to Turkey's building code compliance failure and California's cat model prohibition. In all three cases, a known risk management gap was documented, not remedied, and produced predictable catastrophic losses. Cat models that assume governance structures function as designed systematically underestimate losses in environments where they demonstrably do not.

Summary — Key Analytical Takeaways

  1. Infrastructure interdependency creates loss mechanisms that single-peril models cannot capture: The gas-power feedback loop that amplified Uri's impact — where power outages reduced gas production, which reduced power generation, which deepened power outages — is a coupled infrastructure failure that no standard winter storm cat model was designed to model. Events that cascade across multiple interdependent infrastructure systems require multi-system models, not peril-specific single-system frameworks.
  2. The dominant loss mechanism was secondary, not primary: Uri's insured losses were not primarily from ice, snow, or wind — the conventional winter storm damage mechanisms. They were from frozen and burst pipes — an interior building failure triggered by the secondary consequence of power outage. Models calibrated to primary meteorological damage systematically underestimate loss when the dominant mechanism is infrastructure-failure-driven.
  3. Climate-zone design assumptions are a first-order vulnerability variable: Texas buildings and infrastructure were designed for a climate in which prolonged sub-freezing temperatures are rare. When that design assumption was violated for seven days, the entire built environment of the state's most populated regions failed simultaneously. Cat models must incorporate climate-zone design assumptions as a vulnerability input — not just construction type and age.
  4. Voluntary compliance with reliability recommendations is not risk mitigation: The 2011 FERC/NERC recommendations, if implemented, would have materially reduced the 2021 outcome. They were not implemented because compliance was voluntary. The same event, at three times the scale, produced mandatory requirements a decade later. This ten-year lag between documented vulnerability and enforced remediation is a governance failure that should be modelled as a systematic underestimation of risk in jurisdictions with voluntary rather than mandatory reliability standards.
  5. Grid isolation amplifies tail risk: ERCOT's limited interconnection with neighbouring grids converted a large but potentially manageable supply shortfall into a near-total-collapse scenario. Cat models for electricity-outage-driven losses must incorporate grid topology — specifically, the degree of interconnection available for emergency imports — as a modifier of the maximum credible outage duration and severity.
← All Events All Events →