TEST SYSTEMThe platform is currently under development.

Technical article · bolting technology

Why regular inspections could not prevent the decisive damage

What the collapse of the Morandi Bridge teaches about concealed prestressing steel, corrosion and structural monitoring

The Morandi Bridge was inspected regularly. Nevertheless, the actual condition of the load-bearing prestressing steel in the stay that later failed remained largely concealed. The case shows that inspection can create safety only when the chosen method can detect the governing damage and the remaining uncertainty is assessed structurally.

On 14 August 2018, part of the Polcevera Viaduct in Genoa, commonly known as the Morandi Bridge, collapsed. The structural system around Pier 9 was affected. Approximately 240 metres of the deck fell into the Polcevera valley and 43 people died.

The collapse began with the failure of an essential load-bearing member. The system belonging to Pier 9 then collapsed within a short time. There was no adequate alternative load path.

The unusual structural system

The viaduct, built between 1963 and 1967, had three largely independent main structural systems at Piers 9, 10 and 11. The deck was suspended from tall reinforced-concrete pylons by inclined stays.

These stays were not exposed steel cables. Their prestressing wires were enclosed in concrete and additionally prestressed. The concrete was intended to protect the steel from corrosion, increase stay stiffness and reduce part of the stress range.

The arrangement had a major disadvantage: the condition of the internal prestressing steel could only be assessed to a limited extent from outside. An apparently intact concrete surface can conceal voids, incomplete grouting or internal corrosion.

Further image source

Wikimedia Commons provides a case-specific image collection on the collapsed Polcevera Viaduct. Connect does not reproduce the images.

Where the failure probably began

Forensic investigations and numerical back-analyses place the beginning of the collapse in the south-eastern stay of Pier 9, close to the upper anchorage.

Only simulations in which the stay failed in this region produced a collapse sequence consistent with the video evidence and debris pattern. These simulations assumed extensive corrosion and a correspondingly reduced effective steel area.

The public summary of the first-instance judgment published by the Genoa court on 17 July 2026 also identifies damage in the upper part of this stay as the starting point of the collapse mechanism. The full written reasons were not yet available when this article was published, and the judgment is not final.

A highly stressed stay that was indispensable to the whole structural system had probably lost a substantial part of its effective steel area in the critical region.

Why the damage was difficult to determine

The critical steel was inside a concrete enclosure. Ordinary visual inspection could mainly assess the external concrete surface, not how much load-bearing steel remained effective inside.

Different inspection and measurement methods were used during the bridge’s service life, including local openings, reflectometry, dynamic investigations and measurements of remaining prestress. They did not reliably answer the decisive question: what effective steel area was actually present near the upper anchorage?

Direct openings were made mainly in the early 1990s and mostly on other prestressing elements. The upper region of the Pier 9 stay that was later considered critical was not specifically opened.

Reflectometry produced partly contradictory results. Dynamic investigations could reveal changes in the global system but were not able to quantify local corrosion of individual wires. Deviations from expected prestress also did not uniquely establish the location and extent of damage.

Inspection without a conclusive result

If a non-destructive method cannot reliably detect the damage being sought, other measures remain necessary: targeted openings, complementary investigations, conservative structural assessment, strengthening or operational restrictions.

A condition that cannot be determined conclusively is not an unremarkable condition.

The findings at Pier 11

In the early 1990s, voids, incomplete grouting and damaged prestressing steel had already been found in the stays of Pier 11. That stay system was subsequently comprehensively strengthened.

It was therefore known that the original protection concept had not worked reliably everywhere. The technical question should have been whether the same fabrication, detailing and exposure could have produced similar damage in the comparable stays of Piers 9 and 10.

The planned strengthening

From 2015 onward, strengthening of the stays at Piers 9 and 10 was being planned. The work had not been completed before the collapse.

Planned strengthening does not increase the capacity of the existing structure. Until implementation, it must be decided explicitly whether additional investigations, closer monitoring, traffic restrictions or temporary safeguards are required.

An inspection report is not yet a condition assessment

Three steps must be distinguished for existing structures. First, observations and measurements are recorded. Second, their meaning for capacity, durability and robustness must be assessed. Only then can decisions on further investigation, strengthening or use restrictions be made.

An unremarkable surface does not prove the condition of concealed prestressing steel. A measurement is only as reliable as the method used. Repeating an insensitive method does not create additional safety.

What the case teaches today

Critical members must be inspectable

Members whose failure can trigger progressive collapse should be accessible, replaceable or reliably monitorable with suitable methods.

The inspection method must match the damage

Before inspection, the expected damage mechanism and the method’s detection capability must be defined. Inspection depth, spatial resolution and detection threshold are crucial for concealed prestressing steel.

Contradictory results are themselves a finding

If measurements do not agree, they must not be interpreted as reassurance. The uncertainty has to be resolved.

Damage in comparable members must be assessed system-wide

Voids, grouting defects or corrosion found in one similarly constructed member require assessment of the others.

A measure requires a binding conclusion

A known or seriously suspected capacity deficit needs a documented decision, a responsible owner and a deadline. Residual risk must be controlled until the permanent measure is complete.

Current German context

DIN 1076 governs the inspection and monitoring of German road structures. The January 2026 edition addresses structural safety, traffic safety and durability and includes inspection manuals, digital records, monitoring and imaging methods.

A structure has not been adequately inspected if the condition of its decisive load-bearing member remains unknown.

Limits of the conclusions

Forensic investigations and back-analyses strongly narrow down the probable initiation region, but not every detail of the damage sequence was directly observed. The legal assessment is also not final. This article therefore does not derive a final personal attribution of blame from the technical findings.

How the damage could have been prevented

A present-day risk-based condition assessment would have treated the severe damage found in the comparable stays of Pier 11 and the still-unknown condition of the concealed prestressing steel at Piers 9 and 10 as a system-wide safety case. Targeted openings in the critical upper anchorage region, a combination of inspection methods validated for the expected damage, a conservative residual-capacity assessment and independent review would have exposed the remaining uncertainty. Unrestricted continued use should not have been justified by repeated measurements or monitoring alone while the effective steel area remained unknown.

Critical decision point
Transfer of the Pier 11 findings to the comparable stays at Piers 9 and 10 and the decision on continued operation until planned strengthening.
Suitable present-day method
Targeted direct openings and sampling in critical areas, validated non-destructive methods, conservative residual-section and capacity assessment and independent risk-based overall evaluation.
Finding that would become visible
The actual load-bearing steel area in the critical stay was not adequately known; voids, incomplete grouting and corrosion had to be treated as a possible systematic weakness.
Required action
Restrict operation or close the structure until the condition is directly established or controlled by temporary safeguarding, strengthening or replacement.
Limit
Local concealed loss of individual prestressing wires can remain undetected despite monitoring and non-destructive inspection. If the method cannot reliably quantify the expected damage, direct openings or conservative safeguards are required.

Sources