The slosh heard around the world: Solving the Brazilian penthouse pool problem
by John Kilpatrick and Trevor Haskett
Recently, an occupant of a luxury apartment in Brazil’s prestigious 117 m (582 ft) Millennium Palace experienced an alarming consequence of life in a tall building. A private soaking pool lived up to its name—by soaking the entire wood floor of the unit as the building swayed in a high wind. A video of the episode went viral on Facebook.
The Brazilian newspaper Gazeta do Povo reported the event happened during a storm that had wind gusts up to 90 km/h. The paper described the scene as being like a “ship at sea.”
This spectacular event highlights one of the nuances RWDI considers when we evaluate the wind stability of a tall building: Things respond differently to vibration. Below some threshold, a person can be relatively insensitive to sway: It takes a certain amount to make someone stumble. Water is a different matter.
Design for motion
In fact, one of the ways we can make a tall building more comfortable is to intentionally make water do exactly what it did in this video: slosh around. We can’t completely stop tall buildings from swaying. But if we put the right mass of water in the right place at the top of a tall building, in a tank of correct geometry with engineered obstructions to limit the wave heights, the sloshing water partly counteracts the sway. We use the instability of water to make buildings more stable.
This strategy is called tuned mass damping or TMD, and there are many ways to achieve the same result. (In fact, when water is the substance moving around, we use the more specific name tuned sloshing damper, or TSD.) Further, tuned mass damping is only one of the strategies we can use to improve how a building or structure—whether tall or typical—responds to the wind.
Design for probabilities
Besides thinking about what is likely to move and how, we also think about how often undesired motion might happen and what occupants would consider too often. For the event in Brazil, most people would probably consider once as too often.
We don’t have the full details of that event, but we can make an educated guess about whether the residents of that building need to get used to splashing pools. Given historical weather patterns in the region, residents of this building might have to deal with an event like this every 5 to 10 years. That is a broad estimate; a lot goes into correctly calculating the risk of nuisance events. (See the discussion at the end of this page for how we got to that number.) So, this particular nuisance is relatively infrequent, but we suspect the owner of the flat won’t like replacing floors every 10 years. And we can see from the worldwide media exposure that most people think such effects shouldn’t happen at all.
Understanding building motion
RWDI has a team of specialists who help clients understand building motion and advise designers on ways to prevent this sort of event. For example, they could modify the pool design parameters to ensure the sloshing of water is out of tune with the building’s vibration mode. They could also help the building developers understand how likely this scenario is. Even if the building owner is only willing to permit slight changes, a combination of such changes, chosen with an accurate understanding of risk, could be enough to make this event a rare occurrence, rather than a noteworthy and frequent one.
Our remark that the Millennium Palace might experience overflowing private pools about once a decade is a broad estimate. To give an accurate answer, we would consider the specific duration and direction of the wind gust, the surrounding terrain and where the gust is measured. But we conservatively assumed that the reported value was a 3-second gust (rather than 1 second or 1 minute) and that the gust was measured at the standard reporting height of 10 m at the nearest airport, Navegantes–Ministro Victor Konder International Airport, 8 miles from Balneário Camboriú. From work we’ve done at that airport, a gust like that would be associated with a mean hourly wind speed of 55–60 km/hr, which might happen once every 5 to 10 years. Whether this return period would be considered a problem would depend greatly on the effects it caused, to whom, and in what circumstances. These are all elements we help clients understand and quantify as they design for wind stability.