The building survived.
The factory didn’t.

The night nothing happened.

At 12:13 a.m. Sunday, the ground moved beneath southern Los Angeles County. The earthquake was small: magnitude 3.2, centered south-southeast of Compton at a depth of roughly 8 miles. Residents reported feeling it across the region. No significant damage or injuries were reported.

The event mattered less for what it did than for where it happened. Reporting placed the epicenter over the Newport–Inglewood Fault Zone, the urban fault system associated with the destructive M6.4 Long Beach earthquake of 1933.

There is no evidence in the event itself that a larger earthquake is imminent. Small earthquakes should not be treated as countdown clocks. But the geography creates a useful counterfactual: what if materially stronger shaking crossed the same industrial economy?

The factories beneath the faults.

The industrial geography above Southern California’s fault systems has changed dramatically since 1933. Los Angeles County reported roughly 45,200 workers in aerospace product and parts manufacturing as of December 2024. Long Beach reports more than 7 million square feet occupied by its space and aerospace industry and more than 6,500 workers active in space and aerospace design, engineering and advanced manufacturing.

Those figures represent only part of the regional system. The wider Southern California economy includes precision machine shops, electronics manufacturers, robotics companies, space-hardware producers, optical systems firms, specialty processors, test laboratories and advanced-material operations. Their real-estate footprints can look ordinary. Their operational tolerances often are not.

That concentration changes the unit of earthquake analysis. A damaging event is not only a property-loss problem. It can become a manufacturing interruption, then a supplier interruption, and—where a difficult-to-replace node supports a strategic program—a mission interruption.

The wrong question is: “Did the building collapse?”

Commercial seismic diligence naturally starts with the shell: construction type, code era, retrofit status, probable maximum loss and the likelihood of serious structural damage. Those questions matter. But for advanced manufacturing they can understate what is economically at risk inside.

FEMA’s Hazus earthquake methodology explicitly separates structural damage, nonstructural damage, contents and equipment, monetary loss and business interruption. It also treats manufacturing/process machinery and computer/communications equipment among acceleration-sensitive exposures. That means an earthquake does not need to seriously damage columns or walls before it can damage the tenant’s economics.

A steel or tilt-up building may remain structurally usable while a machining center shifts, a rack overturns, process piping fails, a control cabinet is damaged, or precision equipment must be inspected and recalibrated.

For hard tech, there are at least three meanings of “safe”: safe for life, safe to occupy, and safe to produce. The third threshold may be the hardest to reach. A facility can therefore enter what this analysis calls the standing-building / dead-factory condition: physically recoverable real estate with commercially meaningful production unavailable.

The damage you cannot see from the street.

Consider a hypothetical precision-aerospace plant in the South Bay. The shaking stops. Employees evacuate. Engineers inspect the structure. There is no collapse, no dramatic façade failure, no image likely to lead an evening broadcast.

Inside, the operational story can be very different. A machining cell has tripped. Fixtures require verification. A coordinate-measuring machine must be checked. Sensitive equipment may require re-leveling. Work in process may need reinspection. A sprinkler leak can contaminate production space. None of these conditions requires a visibly destroyed building. Any one can stop a shipment.

For a conventional warehouse, a modest disturbance may be inconvenient. For an operation producing tight-tolerance aerospace components, optical systems, microelectronics or flight hardware, whether a machine remains upright is not enough. The relevant question is whether it can still produce output that passes inspection.

The earthquake lasts seconds. The economic event can last months.

Seismic intensity dominates the first minutes of a disaster. Downtime dominates many of the weeks that follow. A manufacturing facility can have several recovery clocks running at once.

Functional recovery exists as a distinct resilience objective because reoccupancy and restored function are not the same. For a mission-sensitive manufacturer, even “basic function” may not be sufficient: qualified production may require stable power, cooling, vacuum, specialty gases, clean environments, calibrated metrology and approved processes.

A building might recover on Day 5 while production recovers on Day 60 or Day 90. The gap is where inspection cost, labor inefficiency, lost throughput, expediting, replacement lead times and business interruption accumulate.

Liquefaction becomes a manufacturing event.

California’s seismic-hazard program describes liquefaction as the failure of loose, water-saturated sediment during strong shaking as pore-water pressure rises. State hazard zones identify areas requiring investigation; they do not mean every parcel inside a mapped zone will liquefy in a given event.

Nor does liquefaction “start” at one universal earthquake magnitude or PGA. Triggering depends on the intensity and duration of shaking, soil resistance, groundwater and site-specific conditions. But once permanent ground deformation occurs, the industrial chain can be direct: soil strength falls, settlement or lateral spreading develops, foundations or slabs move, buried services deform, machinery loses its reference plane, inspection begins, production stops.

A warehouse may tolerate a condition that a metrology lab, optical operation or precision machining cell cannot. The geotechnical question therefore becomes an economic one: how much differential movement can the process tolerate before the facility is functionally offline?

A factory does not need to be damaged to stop.

The failure can occur miles away. A water transmission line breaks. A substation trips. Telecommunications fail. A bridge or freeway closure delays employees and service technicians. A specialty processor on the other side of the basin goes offline.

For advanced manufacturing, “power restored” may itself be too crude a metric. Production may require stable voltage, phase quality, process cooling and supporting controls. Backup generation may sustain safety systems while leaving major production equipment unavailable.

The larger bill came after the shaking.

The historic USGS M7.8 ShakeOut scenario was built as a regional planning exercise, not a prediction. Its economic results are striking because they force earthquake analysis beyond the damaged structure.

The scenario’s economic work estimated $113 billion in property damage, $68 billion in direct and indirect business interruption, plus $11 billion in additional related costs. Business interruption alone was roughly 60% of modeled property damage.

The broader scenario modeled approximately 1,600 fires. It attributed roughly $87 billion in property and business-interruption losses to fire and approximately $24 billion in BI to water outages.

The lesson is not that those exact totals describe the next earthquake. It is that the loss pathways are connected: shaking triggers physical damage; infrastructure and fire can multiply it; downtime carries it into the economy.

Fire is not an afterthought.

Industrial earthquake loss can compound rapidly. Shaking damages electrical systems, gas lines, batteries, flammable-liquid systems or process equipment. At the same time, sprinklers, water pressure, road access and emergency-response capacity can be impaired.

For hard-tech occupancies, the aftermath can extend beyond burned equipment: smoke contamination, suppression-water damage, chemical cleanup, environmental investigation, replacement lead times and process requalification can convert a recoverable seismic event into something approaching an economic total loss.

The small supplier may matter more than the expensive building.

The economically critical node in a defense or aerospace supply chain is not necessarily the company occupying the largest property. A specialty heat-treatment shop, plating operation, nondestructive-testing lab, optical coater or precision machine shop can occupy comparatively modest real estate and still sit on a production path with few substitutes.

If such a node is unavailable for 90 days, its direct property damage may be small compared with a large prime contractor’s campus. The downstream schedule effect may not be. That is mission-interruption risk: the consequence of losing the production function, not merely the building.

This is also where apparent portfolio diversification can fail. Ten addresses can share one fault footprint, one liquefaction corridor, one power network, one water system, one freeway or one specialty supplier. To the insurer they are separate policies. To the catastrophe they can behave like one network.

Everyone may need the same repair crew at the same time.

Regional earthquakes produce another form of correlation: demand surge. Structural engineers, inspectors, electricians, sprinkler contractors, machine technicians, calibration specialists, switchgear and replacement equipment are all demanded simultaneously.

A facility’s restart time therefore cannot always be modeled independently from every other damaged facility. A machine that could ordinarily be inspected in three days may wait weeks if the OEM’s regional service queue is overwhelmed. A transformer available in normal conditions may become a long-lead item after hundreds of simultaneous failures.

The severe tail is built from combinations: strong shaking plus shallow groundwater, fragile equipment, prolonged utility interruption, limited repair capacity and a supplier outage. The individual components may each be manageable. Their correlation is what turns them into catastrophe.

The financial earthquake keeps moving.

Eventually the physical event reaches the capital structure. The tenant loses revenue while continuing to absorb payroll, rent, debt service and other fixed costs. Insurance may cover part of the loss, but deductibles, waiting periods, sublimits and claim-adjustment timelines affect liquidity.

The landlord then faces repair capital, tenant-credit deterioration and rent interruption. The lender sees weaker NOI and potentially lower debt-service coverage. The insurer sees multiple claims. The reinsurer sees regional accumulation.

By this point, the amount of concrete that cracked may no longer be the best measure of severity. The more decision-useful metric may be days until qualified production resumes—and how many other companies depend on that date.

The underwriting questions change.

For hard-tech real estate, seismic diligence should not stop after asking whether the structure is adequate. It should ask what shaking the equipment can tolerate; whether critical machinery is anchored; whether slab movement can interrupt production; how long inspection and recalibration take; what power quality the process needs; how long operations can tolerate utility outages; whether critical tools are duplicated; which machines have long replacement lead times; and whether suppliers with few substitutes sit in the same hazard footprint.

Those questions convert seismic engineering into operational underwriting. They also expose why a conventional structural PML can miss the asset that matters most: the production system inside the building.