Blast, heat, debris and dust: how do pyroclastic surges damage buildings?
P J Baxter, R Boyle, P Cole, A Neri, R Spence and G Zuccaro, The impact of pyroclastic surges on buildings at the eruption of the Soufrière Hills volcano, Montserrat. Bulletin of Volcanology (2005) 67:292-313.
Montserrat residents are some of the last people in the world to need reminding of the types of damage that pyroclastic flows and surges can do to buildings, having seen so much at close quarters in recent years. Nevertheless, explosive volcanoes like the Soufrière Hills (SHV) continue to erupt at regular intervals around the world and therefore scientists have studied many aspects of the Montserrat eruption with an eye to learning things from it that can be applied to future volcanic crises elsewhere. Obviously such research may also be useful on Montserrat, if the eruption continues and moves into new patterns of behaviour.
One obvious point needs to be made at the beginning; the damage to buildings discussed here is inflicted by pyroclastic surges or “dilute pyroclastic density currents (PDCs)” as they are called in many recent research reports. Little or nothing needs to be said about the damage inflicted on a building by the dense glowing rock avalanche at the base of a pyroclastic flow. The result is obvious and inevitable. Likewise the lesson for both buildings and people in the path of a PF is simple: don’t be there (Fig. 1).
The situation with surges (dilute PDCs) is much more complicated and Baxter and his co-authors have used three notorious surge/PDC events during the Soufrière Hills volcano (SHV) eruption to work out how they behave when they hit buildings. These events were on 25 June, 21 September and 26 December 1997. The 25 June event is particularly important to understanding how surges behave because it happened in daylight and affected an area with a scattered population on that day. The accounts of survivors have been recorded both in video and in a detailed report by Loughlin, Baxter, Aspinall, Darroux, Harford and Miller in a 2002 book about the early years of the eruption published by the Geological Society of London.
What are volcanic surges?
Before we get down to details it would be useful to remember what pyroclastic density currents are all about. Either a lava dome collapse or a volcanic explosion generates a volume of gas (air plus gases from within the volcano) that is both hot and full of solid particles (dust and larger bits). The force of gravity then controls what this gas does next. If it is, as a whole (hot gas plus flying particles), less dense than the surrounding air, it rises as part of the eruption plume above the volcano. But if it is denser (heavier) than the surrounding air, gravity pulls it down and it flows downhill along the ground (Fig. 2).
On Montserrat this means that these density currents try to flow down to the sea. Then, if they reach it – as often happens below the Tar River Valley – the surge fractions flow onwards across the surface of the water because they are less dense than the sea below (Fig. 3). The heavier fraction below sinks into the sea.
Fig. 1 Near miss by a Montserrat block-and-ash pyroclastic flow. Note the tree fragments that form lethal missiles in surges (Maggie Mangan, USGS).
Fig. 2 A pyroclastic flow (hidden beneath dust clouds), and its attendant surges, on the Soufrière Hills volcano, Montserrat.
Fig. 3 Pyroclastic surge crossing sea below the Tar River Valley (M Stasiuk, Natural Resources Canada).
There is an excellent account of particle-laden density currents in general, written by Herbert Huppert for a school maths project, at this internet address: [http://plus.maths.org/issue20/features/huppert/index.html].
The final feature of surges (dilute PDCs) causes some of their strangest behaviour. A surge may be travelling down a hillside but all the time it is losing part of its suspended load of dust and larger particles, as they drop out of the cloud onto the ground. When enough of the particles have fallen, the remaining cloud of hot gas becomes lighter (less dense) than the air surrounding it and gravity again ensures that it simply rises at once to become a hot dusty cloud above the eruption.
Once surges are far enough from their volcanic sources not to be pushed by any original explosion, they are driven entirely by gravity and therefore move at speeds controlled by their density and the steepness of the slope. Eventually they slow and stop on level ground (or sea) but there are often places where they rush forward with enough momentum actually to climb uphill locally, as on Windy Hill on 25 June 1997.
Let’s put some figures to hot, fast and dust-laden (dense) surges:
How hot is hot?
Boiling pure water (at sea level) 100oC
Household oven at full power 200-250oC
Dry wood ignites about 250oC
25 June 1997 surge at Streatham Village about 400oC
26 December 1997 surge at St Patrick’s about 300oC
How fast is fast?
metres/second KPH MPH
Montserrat road speed limit 9 32 20
Usain Bolt (fastest human) 10.3 37 23
Tsunami nearing a coast 18 64 40
Hurricane Hugo, Montserrat 1989 71 260 160
Hurricane Katrina, New Orleans 2005 78 280 175
25 June 1997 surge in Mosquito Ghaut 55 198 123
25 June 1997 surge at Streatham 35 126 78
26 December 1997 surge at St Patrick’s 90 324 201
How dense is dense?
Montserrat morning air at sea level 1.2
(a little less dense when hotter and/or higher)
Pure alcohol 790
25 June 1997 surge at Streatham 1.6
26 December 1997 surge at St Patrick’s 100
Of course these estimates have been made in varied ways. Sources: Wikipedia; Loughlin et al.; Sparks et al.; Woods et al.; all in the Geological Society of London Memoir 21, 2002.
Turning to the damage that surges inflict on buildings, we can start for a comparison with the mayhem caused by being hit by a moving mass of relatively dense fluid. Nobody is surprised by a “wall of water” during a flood or tsunami hitting houses so hard that they are washed away. Surges are much less dense than water but they often move much faster, so that they too hit houses extremely hard. Scientists use a concept called “dynamic pressure” to measure the “pushing power” of a moving surge. Its formula is: multiply the density by the square of the speed of the surge and divide the result by two. The result is given in units called “kilopascals” (kPa). Rather than define a kilopascal here, you can look this point up in Wikipedia. This unit is conveniently imagined by knowing that a dynamic pressure in a volcanic surge (or a hurricane) of just under 1 kPa will begin to topple trees; 2 kPa is sufficient to flatten a forest (Montserrat/MVO International Scientific Advisory Committee main report 8; appendix 7, March 2007)
“Pushing power” is also important in studies of the damage done by explosions, and the damage caused by surges has been compared with the wreckage resulting from a nuclear explosion. Actually Baxter and his team noticed a subtle difference. Bombs, atomic or otherwise, send out a shock wave that hits objects in the way like a hammer. By contrast, surges push equally hard but do so less abruptly – more like a huge dinosaur leaning against the unfortunate building (Fig. 4).
Fig. 4 I feel like leaning on your house!
Dynamic pressures are obviously extremely hard to calculate accurately. You can hardly expect to wander around inside a surge with an anemometer and a sampling filter! Therefore the estimates have to be fairly imprecise. Nevertheless, some clever research efforts have focused on such things as the bending of metal fence posts and flagpoles, and on the toppling of trees. Baxter and his team also draw on previous work on buildings in the area around Vesuvius, Italy, where the fear of a future explosive eruption prompts plenty of research [http://evivo.pi.ingv.it]. They assemble a detailed table with the blast damage to buildings ranging from zero (dynamic pressure less than 1 kPa) to total devastation (dynamic pressure greater than 25 kPa).