[Note that this article is a transcript of the video embedded above.]

Long ago, maybe upwards of 1-2 million years ago, a river in the central part of what’s now Italy, emptied into what’s now the Ligurian Sea. It still does, by the way, but it did back then too. As the sea rose and fell from the tides and the river moved sediment downstream, silt and soil were deposited across the landscape. In one little spot, in what is now the city of Pisa, that sea and that river deposited a little bit more sand to the north and a little bit more clay to the south. And no one knew or cared until around the year 1173, when construction of a bell tower, or campanile (camp-uh-NEE-lee) for the nearby cathedral began. You know the rest of this story. For whatever reason, we humans love stuff like the Leaning Tower. There’s just something special about a massive structure that looks like it’s about to fall over. But you might not know that it almost did. Over the roughly six centuries from when it was built to modern times, that iconic tilt continued to increase to a point in 1990 when the tower was closed to the public for fear that it was near collapse. The Italian Government appointed a committee of engineers, architects, and experts in historical restoration to decide how to fix the structure once and for all (or at least for the next several centuries, we hope). And the way they did it is really pretty cool, if you’re into recreational geology and heavy construction. And, who isn’t!? I’m Grady, and this is Practical Engineering. Today we’re talking about the Leaning Tower of Pisa.

Five-and-a-half degrees. That was the average tilt of the tower in 1990 when all this got started. I have to say average because the tilt isn’t the same all the way up. And actually, that fact makes it possible to track the history of the lean back before it was being monitored. The tower started construction in 1173 and reached about a third of its total height by 1178 when work was interrupted by medieval battles with neighboring states. When work started back up nearly a century later, the tower was already tilting. But the masons didn’t tear it down and start over; they just made one side taller than the other to bring the structure back into plumb. By 1278, the tower had reached the seventh cornice, the top of the main structure minus the belfry, when work was interrupted again. One short century later, the belfry was finally built, and again with a relative tilt to the rest of the structure to correct for the continued lean. On the south side of the belfry, there are 6 stairs down to the main tower; on the north side, only four. The result of all this compensation by the builders is that the Leaning Tower of Pisa is actually curved. Knowing the timeline of construction and how the tilt varies over the height of the structure allowed historians to estimate how much sinking and settling the foundation underwent over time. By 1817, when the first recorded measurement was taken, the inclination of the tower was about 4.9 degrees, and it just kept going.

The new committee charged with investigating the issue first spent a lot of their time simply characterizing the situation. They drilled boreholes and tested the soil. They estimated stability using simple hand calculations. They built a scale model of the tower and tested how far it could lean before it toppled. They developed computer models of the tower and its foundation to see how different soil characteristics would affect its stability. All of the analysis and various engineering investigations all pointed toward the same result: the tower was very near to collapse. In 1993, one researcher estimated the factor of safety to be 1.07, meaning (generally) that the underlying soil could withstand a mere 7 percent more weight than the tower was imposing on it. There was basically no margin left to let the tower continue its lean. A similar tower in Pavia had collapsed in 1989, and the committee knew they needed to act quickly.

To start, they installed a modern monitoring system that could better track any movement over time, including surveying benchmarks and inclinometers. I have a video all about this type of instrumentation if you want to learn more after this. The committee also opted to take immediate temporary measures to stabilize the tower with something that could eventually be removed before developing a permanent fix. They built a concrete ring around the base of the tower and gradually placed lead ingots, about 600 tons in total, on the north side to act as a counterweight to the overhanging structure. As they added each layer of counterweights, they monitored the tilt of the tower. It was ugly, but it worked. For the first time in history, the tower was moving in the right direction. A few months after they finished the project, the tower settled into a tilt that was about 48 arcseconds or a hundredth of a degree less than before.

In fact, it worked so well, the committee decided to take it one step further. To reduce the visual impact of all those lead weights, they proposed to replace them with ten deep anchors that would pull the northern side of the tower downward to the ground like huge rubber bands. This fix didn’t go quite so smoothly. The engineers had assumed that the walkway around the base of the tower, called the Catino, was structurally separate from the tower. But what they found during construction of the anchor solution was that some of the tower was resting on the Catino. The project required removal of part of the catino to make room for a concrete block, and when they did, the tower started tilting again, this time in the wrong direction, and fast (about 4 arc seconds per day, enough for serious concern that the tower might collapse). They quickly abandoned the anchoring plan and added 350 more tonnes of lead weights to stop the movement and focus on a permanent solution.

Engineering ANY solution to a structure of this scale with such a severe tilt is a challenge in the best circumstances. But adding on the fact that the solution had to maintain the historical appearance of the building (including leaving the right amount of lean!) made it even tougher. And after the near disaster of the temporary fix, the committee knew they would have to be extremely diligent. They ultimately came up with three ideas to save the tower. The first one was to pump out groundwater from the sand below the north side of the tower, but they didn’t feel confident that they could predict how the structure would respond over the long term. Another idea was electroosmosis.

If you’ve seen some of my other videos about settlement, you know that it’s hard to get water out of clay, and there are quite a few clever ways engineers use to make it happen faster. One of those ways involves inserting electrodes into the soil and passing electric current through it. Clay particles have a negative surface charge, so the majority of the ions in the water between the particles are positively charged. Electro-osmotic consolidation takes advantage of this by applying a voltage across the soil, causing the water to migrate toward the cathode where it can be pumped to the surface. The idea seemed promising because, by carefully choosing the location of electrodes, engineers hoped they could selectively consolidate the clay below the north side of the tower, reducing its overall tilt. They even performed a large-scale field test near the tower to shake out some of the kinks and gather data on the effectiveness of the technique. But, it didn’t work at all. Turns out the soil was too conductive, so things like electrolysis, corrosion, heat, and all the other effects of mixing electricity and saturated soil made the process pretty much useless for this particular case.

So, the committee was down to one last idea: underexcavation. If they couldn’t get the soil below the tower to consolidate, they could just take some out. And again, they would need to test it out first. So, in 1995, they built a large concrete footing on the Piazza grounds not far from the Tower. Then, they used inclined drills to bore underneath the footing and gradually remove some of the underlying soil. Guide tubes kept the boring in the right direction, and a hollow stem auger inside two casings was advanced below the footing. The outer casing stayed in place while the inner casing moved with the auger. The auger and the inner casing were advanced past the outer casing to create a void, and when they were retracted, the cavity would gently close. At first, it wasn’t looking good. After an initial tilt in the right direction, the test footing started leaning the wrong way. But the crew continued refining the process and eventually got it to work, even finding it was possible to steer the movements by changing the sequence of underexcavation. It was finally time to try it on the real thing.

Knowing the risks and uncertainties involved, the engineers first designed a safeguard system for the tower if things started to go awry. Cable stays were attached between the tower and anchoring frames. The cables could each be tightened individually, giving the engineers opportunity to stop movement in any undesirable direction if the drilling didn’t go as planned. In 1999, they started a preliminary trial with 12 holes. And the plan went perfectly. Over the course of 5 months, the underexcavation brought the tilt up by 90 arcseconds, and after a few more months, it settled in at 130 arcseconds, about four hundredths of a degree. This gave the committee confidence to move on to the final plan.

Starting in 2000, 41 holes were drilled to slowly tilt the tower upright. Over the course of a year, 38 cubic meters of soil were removed from below the tower, roughly 70 tonnes. The lead counterweights were removed. A drainage system was installed to control the fluctuating groundwater levels that exacerbated the tilt. And, the tower was structurally attached to the Catino, increasing the effective area of the foundation. In the end, the project had reduced the tilt of the tower by about half a degree, in effect reversing time to the early 1800s when its likelihood of toppling was much lower. Of course, they didn’t straighten it all the way. The lean isn’t just a fascinating oddity; it is integral to the historical character of the tower. It’s a big part of why we care. Tilting is in the Campanile’s DNA, and in that way, the stabilization project was just a continuation of an 850-year-old process. Unlike the millions of photos with tourists pretending to hold the tower up, the contractors, restoration experts, and engineers actually did it (for the next few centuries, at least).