St. Mark's Campanile

Why did St Mark's Campanile collapse in 1902?

The warning ran for a week. A fissure was noticed at the start of July 1902 at the north-east corner, at the height of the loggetta roof. It widened quickly, reached the belfry on 13 July, and the next morning the tower fell in on itself, into a square cleared in expectation of the worst (Consorzio Venezia Nuova, 2008).

The architect who directed the rebuilding published his own account of the causes the next year, and he located them inside the structure. The stair ramps did not rest on continuous walls. They rested on eight internal piers whose resisting section was little more than one square metre, carried from ground level to a height of 54 metres, and the springing of the arches tying the piers together cut into those sections and weakened them (Beltrami, 1903, pp. 83, 111). Two stretches of pier had been held together by thin iron bands for at least forty years, in plain view of everyone who climbed to the bells (Beltrami, 1903, p. 109).

He set out the fall in five stages: a stretch of internal pier broke up, the mass above it collapsed and dragged down the piers linked to it by arches, the rubble heaping up inside pushed outwards, the outer shaft split along its weakest line, the vertical series of ramp windows near the north corner, and the belfry and the crown followed (Beltrami, 1903, pp. 110 to 111).

How old was the tower that fell?

Little is documented about the campanile's origins and its first centuries. It stands where an old watchtower may have stood, and it appears to have been completed early in the twelfth century. The earliest pictures of it, from the fifteenth century, already show a structure broadly like the one on the square today, with a belfry carrying a pyramidal spire (Consorzio Venezia Nuova, 2008).

Lightning, fire and earth tremors then kept rebuilding it. A strike damaged the spire in 1489 and an earthquake shook the tower in March 1511. A major programme followed at the start of the sixteenth century, directed by Pietro Bon, which reset the belfry and the spire completely and put a wooden statue of the Archangel Gabriel, sheathed in copper, on a platform that turned with the wind. That is when the tower reached its height of about 100 metres. The statue was replaced in 1820 (Consorzio Venezia Nuova, 2008).

Between 1537 and 1549 the east side of the base was given the loggetta by Jacopo Sansovino, three wide arches with niches holding his statues of Minerva, Apollo, Mercury and Peace. Lightning strikes went on through the seventeenth and eighteenth centuries. In 1745 one tore open the north-east corner and three people were killed by falling stone, the same corner where the fatal crack appeared in 1902 (Consorzio Venezia Nuova, 2008).

What the collapse destroyed, and what survived

The tower fell on itself rather than across the square, so the damage around it was limited, though not slight. The north corner of Sansovino's library was destroyed, and the loggetta at the foot of the campanile was crushed. The loggetta was later rebuilt from the original fragments recovered on the site, which is why the building at the base is largely the sixteenth-century one (Consorzio Venezia Nuova, 2008).

Four of the five bells were damaged beyond repair. Only the Marangona, the largest and the one that rang the working day, came through practically intact (Consorzio Venezia Nuova, 2008).

The evidence of how the tower had failed was handled less carefully. Beltrami argued that the official inquiry chased immediate blame instead of the causes inside the building, and that the clearing of the rubble should itself have been run as an investigation, to record how each part of the structure had behaved during the fall (Beltrami, 1903, pp. 106 to 108). One find did support his sequence: a banded stretch of pier from about 20 metres above the ground was recovered in the rubble at base level, which is where it would have landed if it went first (Beltrami, 1903, p. 111). His conclusion was that the tower could have come down at any moment in the previous forty years (Beltrami, 1903, p. 110).

What holds the campanile up

The old tower stood on a stone block 222 square metres in area, resting on a raft of oak boards over a dense field of short wooden piles. Beltrami had two of those piles pulled out at the north-west corner in March 1903. They were about 25 centimetres across and 1.50 metres long, half of that length cut to a point, and they had been driven touching one another to compact the ground rather than to reach anything solid. The target was the shelly clay layer, which borings put between 3 and 5 metres below mean sea level (Beltrami, 1903, p. 88). Ten samples of the foundation timber were examined under the microscope in Rome: alder for the piles, oak for a piece of the raft, elm for the sheeting. The alder was judged to be in good condition (Beltrami, 1903, p. 87).

The weight above was the problem. The structure was estimated at 10,000,000 kg in 1885 and recalculated after the collapse at 14,400,000 kg, which over 222 square metres works out at 6.4 kg per square centimetre, rising to 8.64 under wind load (Beltrami, 1903, pp. 80 to 81). So the rebuilding widened the base, adding a perimeter strip at least 3 metres wide to bring the figure down to about 4 kg per square centimetre (Beltrami, 1903, pp. 82 to 83).

What the 1912 tower changed

Three things were done differently, on the project's own account: new bricks were used, the shaft was tapered to make it slimmer, and the belfry was built in reinforced concrete (Trevi, 2011). Over 3,000 piles about 4 metres long and about 20 centimetres across were driven around the old block, the gaps filled with concrete and crushed stone, with three layers of oak boards above them and then blocks of Istria stone and trachyte. The foundation came out at 410 square metres against the 222 of the old block, with the average pressure on the ground cut from about 9 to about 4 kg per square centimetre (Consorzio Venezia Nuova, 2008; Trevi, 2011).

The bigger change is the part nobody sees. Beltrami kept the 37 ramps at their old dimensions, course and inclination, and changed what carries them. The inclined barrel vaults were given up, since their springing cut into the walls that mattered most. Wall sections were made to decrease with height instead of running at one thickness all the way up, and the corners were given stronger pilasters tied in with courses of resistant stone instead of iron (Beltrami, 1903, pp. 111 to 112). He also stopped the inner shaft at the belfry floor, so the timber bell frame bears on continuous masonry and stands clear of the belfry walls, and the swing of the bells is absorbed there rather than carried down the piers (Beltrami, 1903, p. 112).

The settlement that never stopped

The widened foundation never quite acted as one piece with the old block, and small cracks were noticed in the trachyte steps at the base as early as 1914, two years after the inauguration. The consequence is an uneven load: compression concentrates at the four corners of the base section, because the base deforms while the box section of the shaft above it stays rigid (Trevi, 2011).

Manual measurement of the deep cracks ran from 1955, stopped in 1975, and started again with a diagnostic survey in 1993. Those readings showed the cracks widening almost linearly across 44 years, so the settlement was not slowing down. The same survey found cracks over a centimetre wide in the vault of the chapel of St Peter, which adjoins the bell tower inside the basilica (Trevi, 2011).

Surveys recorded a differential settlement of 60 centimetres in the foundation block, while the shaft itself is out of plumb by about 7 centimetres over its whole height, which is imperceptible on a tower of nearly 100 metres. An automatic structural monitoring system was installed in 1995 for the Magistrato alle Acque (Consorzio Venezia Nuova, 2008; Trevi, 2011). Beltrami's 1903 plan had put the widened foundation at about 450 square metres rather than the 410 reported as built, so the figure moved during construction (Beltrami, 1903, p. 82).

The titanium belt, and the concrete of 1912

The work that followed, carried out from 2008 and completed in 2011, did two things. The soil around and under the foundation was treated by wet soil mixing, the first use of the technique in Italy: about 100 cubic metres of cement grout went in through 92 steel pipes 12 metres long, and the treatment was reported to raise the foundation's resistance to stress tenfold. Then a belt of titanium bars 60 millimetres thick was tensioned around the buried stone block at two levels, each level a pair of bars 20 centimetres apart inside a protective sleeve, at 250 kN per bar, with the option of raising the tension later as the monitoring dictates. Titanium was chosen for durability in brackish water, and the bars sit outside the stone block so they can be removed and replaced. The project document puts the two levels at about 40 centimetres and about 2.3 metres below the pavement; the contractor's technical note gives 1.5 and 3.5 metres (Consorzio Venezia Nuova, 2008; Trevi, 2011). The tower stayed open to visitors throughout.

The reinforced concrete of 1912 is now the part under watch. In April 2024 small cracks and flaking concrete were reported on the internal tie rods of the spire, with fragments falling inside the tower and nothing reaching the square. The Procuratoria di San Marco called it wear on one of the oldest reinforced concrete structures in Venice and commissioned inspections (ANSA, 2024).