How a structure ages and when ageing becomes visible

Design working life is a calculation parameter

Eurocode EN 1990 defines design working life as the period for which a structure is to be used for its intended purpose, with anticipated maintenance and without major repair being necessary. It sets five categories: 10 years for temporary structures, 10 to 25 years for replaceable structural parts, 15 to 30 years for agricultural structures, 50 years for common buildings, 100 years for monumental structures, bridges and other major civil engineering works.

The Italian technical construction standards adopt three categories, 10, 50 and 100 years, and the explanatory circular specifies that this is a conventional parameter related to the definition of design actions.

Exceeding that value describes a shift of the burden onto verification of actual condition, rather than a physical expiry date. A structure that is inspected and maintained exceeds it, and a neglected one shows problems sooner.

One figure clarifies the distance between design and reality: motorway bridges built in the 1960s and 1970s were designed to last 120 years, and the first signs of degradation appeared after 20 to 40 years.

At European regulatory level one anchor exists, but it concerns the trans-European networks alone: the revised TEN-T Regulation introduces an obligation to keep works in a safe and usable condition throughout their life, without setting any automatic verification on reaching the design working life.

In Italy 60 per cent of residential buildings were built before 1980, and 42.5 per cent have passed 50 years.

The mechanisms at work, and their timescales

The dominant mechanism is reinforcement corrosion, described in two phases. In the initiation phase aggressive substances penetrate the concrete cover without corrosion being active. In the propagation phase the depassivated steel oxidises.

Carbonation. Atmospheric carbon dioxide penetrates concrete following the square root of time law. A statistical review of 1,999 cases indicates an average coefficient of around 4 millimetres per square root of a year, implying an average depth below 29 millimetres after 50 years. With adequate concrete cover, carbonation is therefore a process of decades. Rising atmospheric carbon dioxide concentration and clinker reduction in cements may accelerate it in future.

Chloride attack. Typical of marine environments and of roads treated with de-icing salts, it is faster and produces localised corrosion with cell formation. Under severe conditions the local rate of section loss reaches 0.1 to 1 millimetre per year.

Alkali-silica reaction. It generally becomes visible after 20 to 30 years in temperate climates, earlier in hot climates. The finest cracks already reduce strength while remaining invisible to the naked eye, and they open the way to freeze-thaw action and to reinforcement corrosion.

Freeze-thaw. Progressive mechanical damage from freezing of water in the pores, aggravated by water ingress through pre-existing cracks.

Structural steel corrosion. Classified by environmental category, with first-year rates ranging from around 1.3 micrometres per year in the least aggressive environments to over 200 micrometres per year in extreme marine and industrial environments. Zinc from hot-dip galvanising corrodes 10 to 30 times more slowly than bare steel in the same environment.

When degradation becomes visible

This is the point that bears directly on anyone planning maintenance, and the evidence is clear.

The initiation phase, the one preceding the onset of corrosion, generally lasts decades with adequate concrete cover. The phase that follows leads to visible signs, and it is shorter.

A steel section loss of between 0.05 and 0.10 millimetres is sufficient to cause spalling of the concrete cover. A study on 40 millimetre cover indicates that under natural conditions the propagation phase lasts at least 7 years before reaching a crack width of 0.3 millimetres, the value the Eurocode takes as the serviceability limit state, and that even that value may underestimate the point at which delamination begins.

A study of 25 buildings finds that the corrosion rate in elements with unacceptable section loss was around 8 times higher than in cracked elements without section loss, because the process accelerates once the cover has cracked.

What can be observed earlier, and within what limits

Instrumental techniques detect degradation before it manifests visually. The literature documents that variation in dynamic characteristics allows corrosion damage to be identified even in the absence of surface cracking, and methods measuring corrosion rate, carbonation depth and chloride content return the state of the process while it is under way.

A quantitative limit has to be stated here, because it is what separates a usable measurement from an ambiguous one.

It follows that a series of dynamic measurements becomes interpretable with sufficiently long time series and with documented models correcting for environmental effects. A frequency shift of 1 or 2 per cent, read in isolation, describes a change of season with equal plausibility.

On residual life prediction. The literature organises diagnosis into levels: detection, localisation, quantification and prognosis. The first two levels are operationally available today. Quantified and reliable prognosis of the residual life of a real civil structure remains a research objective, as the reference literature states, and the mature methods in this field concern domains with controlled load histories and established fatigue models, different from building stock.

The case of precast industrial buildings

Industrial stock presents mechanisms of its own, and in Italy it spread from the 1950s and 1960s with clear spans of up to around 30 metres.

Corrosion of prestressing strands in roof elements and beams. Corrosion of prestressing steel is particularly insidious, because the material works at high stress and the process can advance without surface signs.

Friction and neoprene bearings. In statically determinate structures the beam-to-column bearings were often made by friction alone in construction predating the seismic codes, and the elastomer degrades over time.

Connections. These are the most vulnerable zone. Studies of systems widespread in Italy show that the connection between roof element and beam can lose up to 60 per cent of its strength beyond the capacity limit, with brittle failure.

Roof coverings. Thermal cycling, ultraviolet radiation, ponding from blocked drainage, and damage from subsequent installations which, without periodic checks, remain invisible until water infiltrates.

The maintenance backlog, and the cost of deferring

The total value of the European bridge stock is estimated at around €2 trillion, with around 10 per cent classified as significantly deficient and around 30 per cent of those in good condition lacking regular inspection.

On individual countries: the replacement backlog of the German rail network is estimated at between €44.5 and €49 billion, and around 5,000 motorway bridges require repair or reconstruction. The Netherlands Court of Audit quantifies a backlog of €54.5 billion, with a structural annual shortfall of €1.8 billion. In the United Kingdom the backlog for bridges managed by local authorities was estimated at £6.7 billion in the 2017-2018 survey, with 3,177 structures out of 71,652 classified as inadequate to carry the heaviest vehicles.

The cost of deferring has a classic formulation. The so-called law of fives, formulated by de Sitter in 1984 in the proceedings of a European workshop on durability, indicates that cost multiplies by around five at each stage of delay: quality in design and construction, preventive maintenance before corrosion begins, repair after it begins, rehabilitation or replacement after generalised corrosion. The same author warns that the figures serve to indicate where to concentrate attention rather than being taken literally.

When visual inspection is structurally insufficient

There is a category of elements for which direct observation returns little, and the most documented case in Europe shows it clearly.

The collapse of the Polcevera viaduct in Genoa on 14 August 2018, with 43 fatalities, concerned a structure 51 years old. The ministerial inspection commission attributed the cause to corrosion of the upper part of a stay, with section reduction of between 50 and 100 per cent on the majority of the strands examined, in a process that began in the first years of the structure's life.

Two elements are relevant to the subject of this page. The strands were embedded in concrete and therefore not visually inspectable. And a monitoring sensor installed on the same pier operated from 1992 to 1996 and was not subsequently replaced.

Where the material carrying the load lies beyond the reach of direct observation, instrumental measurement is the only route to knowledge.

What follows

The timescales of degradation are measurable, and the signs arrive late. The initiation phase lasts decades, the propagation phase years, and the appearance of visible signs comes at the end of both.

A single inspection describes an instant. Knowledge of the condition of a structure is built through repeated observation, and comparison between two surveys separated in time says what an isolated survey leaves out.

Instrumental measurement requires a series and a correction. The signal of initial damage has the same amplitude as environmental noise, so it becomes readable with long time series and with models compensating for the effects of temperature and operating conditions. The literature converges on this, and it serves as a criterion for evaluating any measurement system.