Why Hawaii Residents Should Not Wait for Grid Hardening

Grid hardening can prevent some power outages, but Hawaii residents should not expect the current program to protect their next outage: Tropical Storm Lala cut electricity to 60% of Hawaiian Electric customers—nearly 300,000 households—and the utility called invasive albizia trees the “No. 1 culprit.” Yet approved hazard-tree funding fell from $32.405 million to $11.052 million, while the transmission-hardening allocation dropped from $75.012 million to $41.186 million Civil Beat.

The Case for Waiting on Grid Hardening

The conventional view is reasonable. Stronger poles and transmission structures can survive conditions that would damage weaker equipment. Covered conductors can reduce some contact faults. Underground lines avoid direct exposure to wind, falling trees, ice, and airborne debris. Automated switches can isolate damage, while redundant paths and microgrids can preserve service for some customers.

The National Academies concluded that component hardening, physical security, and damage-prevention measures can reduce localized interruptions and mitigate large-area, long-duration outages. It also found that complete physical protection is impossible because the grid is extensive, interconnected, and exposed to many hazards National Academies.

HECO’s plan therefore is not meaningless. Its 2022 proposal called for $190 million over five years, with another $95 million in federal funding. Multi-year work is unavoidable on an island grid containing transmission corridors, substations, distribution circuits, vegetation zones, communications systems, and difficult terrain.

The consensus is also right that one major blackout does not prove every completed upgrade failed. A reinforced structure may survive while an unmodified line upstream fails. Switching may restore most of a circuit while homes beside the damage remain offline. Faster restoration is valuable even though it did not prevent the original interruption.

The problem is narrower: HECO’s post-Lala cause attribution points to vegetation, while the approved allocation reduced the spending most directly connected to trees and transmission exposure. That mismatch gives residents little evidence that their own outage probability will fall meaningfully on the plan’s timeline.

Test the Risk Around Your Address

This sketch does not calculate an actuarial outage probability; HECO has not published address-level figures that would support one. It sorts vegetation exposure and restoration difficulty using the failure mechanisms reported after Lala.

Choose your line, canopy, tree reach, route, and access conditions; the result identifies whether waiting or household backup wins for those inputs.

Lala Exposed the Plan’s Weakest Match

Lala affected Hawaii on August 15–16, 2026 after being downgraded from hurricane strength and passing about 75 miles offshore. Civil Beat reported that approximately 190,418 HECO customers on Oahu were affected and nearly 300,000 Hawaii households lost power statewide. Nine major Oahu transmission lines went down.

HECO Vice President Jim Kelly identified invasive albizia trees as “the No. 1 culprit,” adding, “All it takes is an errant branch.” Rural Ka‘u on Hawaii Island faced the prospect of weeks without power. Those facts point toward an immediate failure mechanism—trees contacting or bringing down lines—not grid age as an abstract category.

A separate snapshot reported more than 88,000 customers without electricity across Hawaii Island, Maui County, and Oahu at 7 p.m. HST on August 15. Gusts approached 90 mph in parts of Hawaii Island, and fallen trees damaged multiple poles and structures along a 42-mile transmission corridor on the Hamakua coast. Flooding, debris, downed lines, road closures, and unsafe conditions obstructed restoration Bernama.

The 88,000 figure should not be added to Civil Beat’s totals. It covered a particular time and multi-island snapshot, while the other figures describe broader storm effects without a directly comparable timestamp.

HECO’s explanation is also not a completed independent finding. Energy activist Henry Curtis petitioned the Public Utilities Commission for a formal inquiry into Lala and the March 2026 Kona storms. Without that inquiry or an equivalent asset-level review, the public does not know the hardening status of every failed line, pole, or structure.

That uncertainty cuts both ways. Critics cannot prove that every completed project had no effect. HECO also cannot establish, from the public information available, how many interruptions or outage-hours the work prevented.

The Spending Shift Does Not Match the Reported Cause

The original critical-transmission-hardening proposal allocated $75.012 million to that work. The approved amount was $41.186 million. Proposed hazard-tree funding of $32.405 million became $11.052 million.

At the same time, wildfire monitoring and mitigation increased from $14.102 million to $42.207 million as priorities shifted after the 2023 Lahaina fires. Wildfire mitigation is a legitimate safety obligation, not wasted spending. Weather monitoring, cameras, covered conductors, operational changes, and selective undergrounding can reduce ignition risk or make safety shutoffs more targeted.

But the investments are not interchangeable. A wildfire camera does not remove an albizia tree that can reach a transmission line. Monitoring can tell operators that hazardous conditions exist without preventing a branch from causing a fault. An operational shutoff may reduce fire risk while still leaving a household without electricity.

Only projects completed and operating before Lala could have affected Lala. Proposed, approved, funded, under-construction, and operational projects are different categories. A systemwide spending total does not reveal whether the specific assets exposed to the storm had been reinforced, cleared, automated, or given an alternate electrical path.

This is why the plan does not yet support a household-level promise. The utility reported vegetation as the leading cause, but the public record shows substantially less approved hazard-tree funding than proposed. Residents on overhead, tree-exposed feeders remain dependent on recurring inspection, trimming, removal, and access for repair crews.

Hardening Works Only When It Matches the Failure

“Grid hardening” combines measures that produce different outcomes. Physical reinforcement can prevent damage. Automated switching contains a fault after it occurs. A microgrid provides continuity inside a limited boundary. Sensors and remote controls can speed restoration.

Those outcomes should not be collapsed into “outages avoided.” A quickly restored interruption was still an interruption. A microgrid that islands successfully protects its own loads but does not prevent the wider utility outage.

Measure	Main Hazard	Primary Result	Limitation
Tree trimming and removal	Branch or tree contact	Prevention	Vegetation regrows
Stronger structures	Wind and structural load	Prevention or containment	Upstream assets can fail
Automated switching	Circuit fault	Containment and restoration	Damaged segment stays out
Household battery	Loss of grid supply	Local continuity	Runtime depends on load

When trees are the dominant cause, vegetation management is the most direct intervention. Covered conductor may reduce some branch-contact faults, and selective undergrounding removes a line from direct tree exposure, but neither makes the complete delivery chain outage-proof.

Undergrounding illustrates how strongly results depend on event and location. A Stanford research summary described a modeled counterfactual in which an additional 25% of overhead lines were buried. For the October 2017 New England nor’easter year, estimated annual outage time fell 10.8 hours, or 39.7%, from 27.3 hours per home. For California’s 2019 wildfire shutoffs, the modeled reduction was 2.4 hours, or 23%, from 10.5 hours to roughly eight hours. For the February 2021 Texas freeze, it was 2.5 hours, or 13.2%, from just over 19 hours Stanford.

These were modeled scenarios, not observations of additional lines actually buried and tested. Substantial outage time remained in all three. The variation is the point: a measure works when it addresses the event’s failure mechanism and covers the assets that matter.

There is no evidence-supported universal percentage for outages prevented by grid hardening. The result depends on the hazard, asset condition, network coverage, project completion, alternate paths, and whether generation is available to send through those paths.

Stronger Equipment Cannot Remove Network Weak Points

A power grid is a chain. Reinforcing one pole does not preserve service if a tree takes down the line feeding it. A protected substation can be stranded by failed transmission. An automated switch cannot reroute power if there is no viable alternate path or enough transfer capacity.

Severe storms also create compound failures. Wind, tree fall, flooding, debris, communications trouble, damaged roads, and unsafe working conditions can occur together. Surviving equipment may remain electrically isolated, and crews may be unable to reach failed equipment safely.

Automation has a last-mile limit. It can detect a fault, isolate the damaged section, and re-energize customers connected to an undamaged path. Customers beside the broken conductor, pole, or transformer generally remain without power until physical repairs are completed.

That does not mean automation failed. It means the technology delivered containment or partial restoration rather than prevention. Utilities should report those outcomes separately.

HECO Has Not Published the Data Needed to Prove Success

A raw outage count shows what happened with the existing system. It does not show what would have happened during the same storm without completed upgrades.

A credible Lala assessment would identify which lines, poles, substations, switches, and vegetation corridors had been upgraded and operational before the storm. It would report the wind, flooding, tree fall, or other exposure at each asset, followed by the precise failure mechanism.

Customer consequences also need more than one number. Useful measures include customers interrupted, interruption duration, total customer-hours lost, geographic spread, customers restored through switching, critical loads maintained, and time to major restoration milestones.

Comparable upgraded and unmodified assets would provide a stronger test than systemwide spending. Models can estimate a counterfactual, but assumptions about asset placement, failure rates, storm exposure, and remaining network capacity must be disclosed. Modeled avoided outages should not be presented as observed results.

Until HECO publishes that record, the defensible finding is limited: completed hardening may have prevented or contained some failures, but the available evidence does not show that residents on vegetation-exposed overhead feeders can expect reliable service soon.

Household Backup Solves a Different Problem

Grid reliability and household continuity are not the same objective. Utility work changes the probability or extent of an interruption. A properly configured battery, generator, or solar-and-storage system supplies selected household loads after the utility connection has already failed.

Civil Beat described a Lala resident whose solar-and-battery system maintained household electricity while utility service was down. That is an anecdote, not evidence of typical runtime or universal performance.

Actual capability depends on usable storage or fuel, household demand, prioritized loads, inverter and control design, available solar generation, equipment condition, and safe islanding. The draft evidence supplies no standard battery size or runtime suitable for every home, so no such figure can be promised.

The practical priority is to identify the loads that cannot wait for utility restoration: medically necessary equipment, refrigeration for essential food or medicine, communications, and safe temperature control. Backup equipment must be installed and operated according to manufacturer instructions, electrical requirements, utility rules, and local law. Generators must never be operated where exhaust can enter a home or occupied space.

Residents should still follow utility alerts and emergency instructions. Home backup does not make roads safe, eliminate storm hazards, or replace evacuation orders. It simply addresses the part of the problem HECO’s hardening timeline cannot guarantee: electricity at a particular address during the next interruption.

The Answer Depends on Which Failure You Mean

Grid hardening does prevent some equipment failures and customer interruptions. It can also contain damage, preserve selected loads, and shorten restoration. The National Academies’ evidence supports those narrower benefits, not an outage-free grid.

For Hawaii after Lala, the immediate evidence favors vegetation work over waiting for the broad hardening label to deliver reliability. HECO identified trees as the leading culprit, while approved hazard-tree and transmission allocations were reduced from their proposed levels. Until completed work matches that reported cause—and performance is documented asset by asset—household backup offers the more direct route to continuity.