I'm considering the mechanics of a "shock and awe" scene in my novel which could see an old dam being brought down. The era is equivalent to that of the middle Roman empire, so introducing explosives would be far too convenient and would require altering previous work for the simple logic of conflicts.

What was available, or has been used in the past, to undermine the foundations of large structures in minimal time?

It's undecided at present how deep the water is on the other side of the dam, and the construction materials. I'm most likely to use the Roman style of architecture (reference).

  • 13
    $\begingroup$ catapult, it can be a catapult was used (not in Rome empire) to demolish walls of castles. $\endgroup$ – MolbOrg Apr 23 '17 at 0:42
  • 6
    $\begingroup$ If you dig a trench around the edge of the dam, the water which begins to flow through the trench, will erode that space until the water runs out. It wouldn't necessarily destroy the dam, but compromise it's purpose. And that is assuming that the dam is built in a canyon and the walls of the canyon can be quickly eroded by flowing water. $\endgroup$ – Nolo Apr 23 '17 at 4:08
  • 5
    $\begingroup$ The first two articles mention mining under the wall. The people that do that are called sappers. en.wikipedia.org/wiki/Sapper $\endgroup$ – Martin York Apr 23 '17 at 4:20
  • 45
    $\begingroup$ You need a group of enraged Ents $\endgroup$ – RedSonja Apr 23 '17 at 17:50
  • 7
    $\begingroup$ youtube.com/watch?v=VipVo8zPH0U is how we do it today without explosives. $\endgroup$ – Bald Bantha Apr 23 '17 at 18:22

20 Answers 20


Forts using stone, rock and earth walls as fortifications (and in some cases, still standing !) were often attacked during sieges.

Such a wall has much in common with a dam.

To breach such a wall you tunnel underneath. It's a well established technique.

You dig a tunnel, using normal techniques to prop and seam your tunnel.

Then when you've dug enough, you set fire to the supporting structures (or otherwise destroy them) and, the attacker hopes, the subsidence will collapse the wall or (for a siege) damage or weaken it. I gather the tunnels were often (always ?) lined or filled with materials that would burn for an extended period to further weaken the wall above using heat.

I see no reason the same principles would not apply to a dam.

  • 93
    $\begingroup$ From this technique comes the word "to undermine", originally meaning "to make a passage or mine under (a wall, etc.), esp. as a military operation", and now more commonly "to weaken, injure, destroy or ruin, surreptitiously or insidiously" (definitions from the OED). $\endgroup$ – AlexP Apr 23 '17 at 1:13
  • 37
    $\begingroup$ That kind of approach is extra effective, because once you start the water flowing, it will very quickly erode the dam and result in a complete failure. Have a look at the Teton Dam, in which a small leak quickly resulted in a complete and catastrophic failure and many deaths. $\endgroup$ – Fake Name Apr 23 '17 at 3:10
  • 9
    $\begingroup$ Would the presence of the dammed water make digging the tunnel either impossible or exceedingly dangerous? $\endgroup$ – Rich Holton Apr 24 '17 at 1:51
  • 2
    $\begingroup$ Danger? no problem - send down enemy POWs to do the digging... $\endgroup$ – Baldrickk Apr 24 '17 at 9:41
  • 3
    $\begingroup$ @rich-holton : Digging a tunnel is always dangerous, but a dam, by it's nature, tends to be built with a water-proof or water-resistant layer that would make the danger of water (during undermining) no more than with normal groundwater, I'd guesstimate. I'd have thought that in these ancient techniques ventilation (i.e. the lack of) was the biggest problem. $\endgroup$ – StephenG Apr 24 '17 at 10:06

A more devious scheme to bring a dam down:

A secret second dam above the main target dam.

You simply install a second dam above the main dam (as high as possible ideally). You reduce the flow inconspicously so that the main dam manning do not see anything unusual, but slowly the reservoir of your second dam is filling.

Now you wait for the perfect time when the second dam reservoir is filled and the main target dam is nearly filled. If the main dam needs only some filling, you can increase the flow from the second dam.

Your attack commences with destroying the second dam which is purposefully built for being brought down (you have some pivot support columns which are disengaged). The water rushes down and gets faster and faster, converting the stored potential energy from a higher point to kinetic energy. Friction and obstacles will slow down the water masses to a point, but it will still be very fast.

When the water enters the main dam, an effect called the water hammer comes in effect. The main dam does not allow the moving water masses to continue running, so the moving water causes a sudden pressure increase. The incoming water not only causes water to slosh over, it literally pushes the dam crest apart. Result: catastrophic failure.

ADDITION: The original question states that we are on the technological level of the early Roman era, so we should neither expect to have a hydro dam like the Hoover nor a reservoir for a city of million people. It will be more like a dam with the height of metres and the reservoir like a big lake.

Still we can compare dynamic with static pressure. The dam need to withstand static pressure, so we can assume we need approximately a pressure with the same order of magnitude to break the dam.

\begin{eqnarray*} \rho & = & density(kg \, m^{-3}) \\ g & = & gravitational \; acceleration = 9.81 \approx 10 \; m \, s^{-2} \\ h & = & height \; m \\ v & = & velocity \; m \, s^{-1} \\ mean \; static \; pressure & = & \frac{1}{2} \, \rho \; g \; h (The \; dam \; holds \; this \; pressure) \\ dynamic \; pressure & = & \frac{1}{2} \; \rho \; v^2 \Rightarrow v \approx \sqrt{10*h} \end{eqnarray*}

Moderate flash flood velocity is 2.6 m/s and a very fast flashflood is in the range of 26 m/s. A moderate flashflood will be held by a 0.6m dam, a worst case scenario of 26 m/s would give an impressive height of 70 m. But the flash flood water will merge with the still water in the reservoir, so an inelastic collision will occur and the water slows considerably down. So the final velocity of the water will be the ratio $$ r = \frac{flash \; flood \; mass}{reservoir \; mass + flash \; flood \; mass}$$ of the flash flood speed (I also neglected friction and energy dissipation by waves).

Result: If the dam is something like 10 m high and the reservoir is big (10-100 times), even the ugliest flashflood will have no pressure effect. Moderate flash floods can be contained even with small dams. On the other hand, if the dam is only a few meters high and the reservoir has not a much bigger capacity (10 times) than the incoming water, an incoming massive flash flood is able to forcibly remove the dam.

  • 40
    $\begingroup$ This is so silly it's good. "Build a secret dam." $\endgroup$ – Nit Apr 23 '17 at 21:55
  • 21
    $\begingroup$ @Nit Ha, ha, a secret dam. You mean something which is quite massive and could be there undiscovered for 30 years? How ridiculous, ha ha ha... $\endgroup$ – Thorsten S. Apr 23 '17 at 22:22
  • 6
    $\begingroup$ You're not going to get a water hammer effect: the main dam's reservoir will dissipate the force of the incoming water. What will happen is that the surge of excess water will overtop and erode the dam, causing it to fail. $\endgroup$ – Mark Apr 23 '17 at 22:55
  • 3
    $\begingroup$ Several problems with this. First: the beaver dam example is nice but you can't rely on it. That dam's some 200 km inside an uninhabited forest; yours must be relatively close to the first dam, which would have people (for maintenance, fishing, bathing) going there day in day out. Second: it's not just "a secret dam". Its construction would need roads, carriages, a lot of workers, and probably a lot of noise (hammer and chisel is a very distinct sound) to also be secret. Third: water hammer effect only applies to closed systems. Open dams like the Romans' would just overflow. $\endgroup$ – walen Apr 24 '17 at 7:13
  • 6
    $\begingroup$ The fluid hammer effect does not apply to rivers and dams. However.... the plan will still work because flooding and uncontrolled overflow is the greatest threat to an otherwise "healthy" and well built dam. $\endgroup$ – MichaelK Apr 24 '17 at 11:27

Roman type dams were usually buttressed rather than relying purely on their weight to hold the water back, break away the supports and they will fail.

Earth dams just need a single point of failure to be induced and the weight of the water will do the rest in short order. Artillery targeting the non water side of either type of dam should be enough to make either type fail fairly quickly, they're not designed to withstand that sort of stress.

The Romans were excellent engineers, they would quickly recognise the weakness of a dam and either pull away supports or dig out making a point of failure.

  • 25
    $\begingroup$ The highest Roman dam at Subiaco (50 meters tall, built under Nero in the 1st century) is said to have been destroyed accidentally in 1305, when "two monks took stones from the wall, because they wanted to lower the water level, presumably in order to make the water further from their fields; the wall no longer withstood the weight of the water; apparently, a breach appeared and grew ever larger, until the wall finally gave way" (Wikipedia). $\endgroup$ – AlexP Apr 23 '17 at 1:28
  • $\begingroup$ @AlexP Interestingly the Subiaco wikipedia article linked above starts by telling these were gravity dams. This raises a point: the engineers would not have wanted their dam to collapse in a "shock and awe" fashion. So any dam holding enough water to create the effect would be a gravity dam that is torn apart by water gradually and any dam that might collapse catastrophically would hold too little water. Of course, for dramatic effect you can assume the engineers were not up to Roman standards, but people building large dams do tend to take the job seriously. $\endgroup$ – Ville Niemi Apr 23 '17 at 13:02
  • 8
    $\begingroup$ @VilleNiemi: Well, Wikipedia says that it was made of "masonry", and we must remember that it had stood for more than 12 centuries, of which at least eight without any maintenance... $\endgroup$ – AlexP Apr 23 '17 at 16:12
  • $\begingroup$ @AlexP Not sure about your point. Mine was that a large dam like Subiaco was designed in a way that made it resistant to catastrophic collapse and indeed even after centuries of neglect it failed gradually, just like people designing it wanted it to. Which is not "shock and awe" inducing. $\endgroup$ – Ville Niemi Apr 24 '17 at 17:50

You can of course excavate under the dam and have it crash for lack of support ("undermining"). This was the routine siege attack against walls and the crash would be quit abrupt. Its feasibility depends on the strength of the foundations and the ease of tunneling through rocks. Romans were quite proficient at that.

Depending on the situation, you could perhaps use a malvoisin - a high structure built near the dam. Get a heavy stone or iron ball secured to a chain hinged on the dam, raise it with pulleys on the malvoisin, let it fall down and impact the dam. Repeat as needed. Essentially you have built yourself a wrecking ball.


Make a hole in it.

Once there is water flow, the hole will expand. If the hole is low in the dam, the structure above it will collapse. But even if at the top of the dam, it will work its way down over time.

You might ask if dams don't already let water through. Sure, this is called a spillway. But spillways receive extra reinforcement so as not to compromise the overall structure. In other words, they are deliberately designed not to wear when water runs through them. If you choose a random portion of the dam (or the ground beside the dam), it won't have that reinforcement. The hole will expand.

Making the hole can be as simple as using a pick and shovel. Dams were often simply big earthworks then. You could dig through them. A dam made of mortared stone would be more difficult. You might find it easier to dig near the dam. That would often still be regular earth. The dam might also be buttressed. Then you could remove the buttress by digging around it and removing its support. A buttress near the center of the dam will likely cause the most strain when removed.

Note that the water is part of the dam's support. As you remove the water and relieve the pressure, the dam may collapse into the water. This is the shock and awe moment, when a small breech turns into total collapse. This is most likely with a buttressed dam.

The reason to do this rather than undermine the dam is that undermining requires more digging. Not only do you have to dig through the thickest part of the dam, but you also have to dig down to get there. And with a dam, it's unlikely that you'd be dodging defenders. Unlike a wall, where defenders stand on top and throw things at you. So there's less utility in starting a tunnel away from the wall.

Most dam construction in Roman times will be such that it is as easy to dig through the dam as under it. If a mortared stone dam, look to removing its support. This may be easier on the sides than underneath. Let the water do the hard work. You just need to give it a chance to start. I would only see undermining as necessary with a buttress, which is smaller.

  • $\begingroup$ Re "... spillways receive extra reinforcement so as not to compromise the overall structure." Oroville. And dams are often just big earthworks now. Oroville again. $\endgroup$ – jamesqf Apr 23 '17 at 4:09
  • $\begingroup$ I believe this is how dams often fail in real life and that this is known as piping failure community.dur.ac.uk/~des0www4/cal/dams/foun/seep4.htm $\endgroup$ – K-Feldspar Apr 26 '17 at 9:27

Speaking of tunnels... https://en.wikipedia.org/wiki/Ruina_montium

an ancient Roman mining technique that draws on the principle of Pascal's barrel. Miners would excavate narrow cavities down into a mountain, whereby filling the cavities with water would cause pressures large enough to fragment thick rock walls.

enter image description here

Mine tunnel to the center of the damm, as long as possible so that it contains a big quantity of air. Once the tunnel is ready, flood it as fast as possible. The trapped air builds up pressure as the water pushes in, and once it exceeds the tunnel walls resistance it escapes breaking the tunnel walls, effectively acting as the compressed air blast of an explosive's shockwave. Only no explosives needed, just water and air... and look, there's a big damm with water just behind you! How convenient.

Additionally, effects from hydrostatic pressure (the higher the water falls, the bigger the force it will do against the walls) add up for extra effect. What the air doesn't blast away, the water force will.


According to tradition, when Hannibal needed to clear boulders off Alpine paths, he used fire-setting; build a large fire against the rockface, and when it was properly heated, throw a large quantity of cold water (or vinegar, for the acid) on it to cause cracks through thermal shock.

This would work well against stone or masonry dams; it's not itself very destructive, but a thin crack all the way through would very quickly widen into a breach. An added bonus is that acetic acid would presumably react with, and weaken, the lime mortar.

  • $\begingroup$ Plus, you'd get that refreshing effervescent flow. I could see a great 80s-era tv spot out of this approach. $\endgroup$ – The Nate Apr 24 '17 at 15:02
  • $\begingroup$ Depending on the thickness of the damn, this is hindered by the heat transference to the rather large mass of water on the other side. Hannibal didn't have to boil a reservoir just to crack a single rock. $\endgroup$ – Ruscal Apr 25 '17 at 16:37

If water could penetrate cavities of the structure and freeze there that would generate large pressure (because water expands when it freezes) that can break rocks. This process is called ice wedging and it occurs in nature.


Roman concrete used lime mortar, strongly alkali. I don't know if they used it in dams, but it's plausible they might so it's not too far a guess.

Alkalis react with acids, which is why modern concrete and cement also come in sulphate (i.e., SO4 ion) resisting versions. Carbonic acid is also an issue. Acids were well within Roman technology, although I'm not sure of the practicalities. But if you could weaken a critical point with acid, or even by just gradually acidifying the water in some static part of the reservoir in contact with the critical mortar, perhaps it would gradually weaken the dam.

As a twist, maybe a tunnel would allow the sub-surface part of the dam foundation to be attacked, removing critical support invisibly, and without having to do it slowly to avoid detection, until suddenly......?

  • $\begingroup$ fondriest.com/environmental-measurements/parameters/… Acid can come from pretty much anything. Probably the quickest available would be diverting a source of water through a coal deposit, or for a slower method, you just need a source of carbon (i.e. trees) the link above recommends pine fir needles. $\endgroup$ – Scott Apr 24 '17 at 3:30
  • $\begingroup$ Like Hannibal crossing the Alps. $\endgroup$ – fectin Apr 24 '17 at 5:11
  • 1
    $\begingroup$ Romans used concrete and water-resistant mortars quite extensively - notably in the reservoirs, some of which are still functioning today. If memory serves me there is evidence they could even flood the Colosseum - they were the experts in controlling water in that period. $\endgroup$ – StephenG Apr 24 '17 at 10:11

@LSerni is probably right about how the Romans would really do this.

If you want some alternate technology within the grasp of the Romans you could destroy it by shaking it at its resonant frequency. Like buildings, dams have resonant frequencies and these are of interest mostly to prevent destruction by earthquakes.


Here is an interesting article about a skyscraper that was evacuated because it was shaking. The shaking was caused by 17 people exercising in unison. http://news.blogs.cnn.com/2011/07/19/scientist-tae-bo-workout-sent-skyscraper-shaking/

Additional reading: Tesla's oscillator, or "earthquake machine" which he claimed could bring down the Empire State Building. https://en.wikipedia.org/wiki/Tesla%27s_oscillator

I could not find a report of a dam which actually collapsed because of this phenomenon. But as far as the shocking awesomeness, using vibrations to destroy a structure should qualify. You could have people atop it stomping in unison faster or slower as the maestro directs. Could one affix a big piece of metal to a rigid structure and cause the structure to resonate by drawing a bow across the metal?

  • $\begingroup$ The walls of Jericho may have come down this way. $\endgroup$ – KalleMP Apr 25 '17 at 7:45

My vote goes towards piercing/hammering: Either with a stated wrecking ball (Malvoisin) or with flotsam (in autumn/spring when flow is high, throw a large, somewhat pointed trunk in the river, upstream of the dam...). If it has butresses, undermine these (possibly in a similar way: Chop down the largest tree nearby to fall away from the buttress, but after tying a rope from tree top to buttress -- rope slack on water, to be inconspicuous, so that it suddenly tenses as the falling tree reaches maximum velocity).

If you can build your secret dam upstream, even better --- use it to give the flotsam additional speed. This can be part of the plot: The dam(ned) engineers know of the 'secret' dam, and have calculated it to be ineffectual; so it's monitored to an extent and they're very prepared for 'attack day' --- however, they're WRONGLY prepared as in their overconfidence they've overlooked the other legs of the plan: The quickly-tied rope-to-buttress-on-falling-tree, and/or the flotsam-missile. (The flotsam-tree and tree-pulling both have the advantages of being normal jobs as timber is main building material --hence inconspicuous-- and quick versus the secret-dam-building --hence the full attack isn't understood until too late to defend, and it may be difficult to get the people prepared to counterattack (=heavily armed) to adjust to civil-defence jobs (plus, maybe they'll sink in the quagmire from their armour etc).

Tunneling under a river doesn't work (you drown, using roman tech), so castle-demolition-analogues won't work I think.

  • $\begingroup$ Welcome to worldbuilding, @user3445853 $\endgroup$ – kingledion Apr 24 '17 at 16:22

One possibility - but not sure how to execute it - would be to use steam. Dig a hole, put a sealed metal cylinder with water in it and fire it up. Steam has tremendous power and if there will be enough pressure in the cylinder it would be much like explosion. But someone would need to run some pretty neat equations on cylinder size and water content and I'm not sure you can build one with only blacksmith-level metal craft.

Any other way in a short form: not bloody likely.

A bit longer answer: unless it's a really small dam, good luck!

Long answer: there is a reason why most of the Roman-built dams are still around, with a lot of them still in use. You'd think that engineers who built roads and aqueducts that are around today and are definitely pieces of finest engineering would build a dam that is a simple wall? Yes, there are a few - like subiaco dam, which are like that, but they are exception to the rule. Subiaco especially, as it's said to be built so that Nero had a lake next to his villa.

But most of them were lime or concrete (yes, concrete) core, with compacted earth and masonry to protect it against erosion.

First of all you can forget catapult or trebuchet. Yes, they would crack the "outer layer", but will be completely ineffective against earth underneath.

With limestone or concrete core you can forget burrowing.

The only way to break solid Roman dam is with good-old earthworks: spades, picks and shovels.


I'm not sure if the physics of this hold up, but could you destroy a dam by movement of the water it holds back? I'm thinking you climb up a convenient high, rocky mountain next to the lake above a dam. You carve out a large, round boulder and let it roll down the mountain and into the water. The bow wave from the resulting impact stresses the dam sufficiently to breach it.

I guess if you were capable of all that, you might argue you'd be better off rolling the boulder onto or against the dam and have the same effect, although that would require a very conveniently located mountain.

  • $\begingroup$ Methinks you'd need more than a boulder, but if you could trigger a large and fast enough landslide, that would do it. $\endgroup$ – nigel222 Apr 25 '17 at 18:59

I think another way to do this would be to block the outflow of the dam, When the dam fills normally it would push water down the overflow.

Issues with the overflow can be dangerous.

Please see Oroville Dam had an issue with its overflow.

Very dangerous if Dam's have the overflow blocked.

  • $\begingroup$ A little trench along the side to focus the overflow to adjacent to the dam could help in such an approach. $\endgroup$ – The Nate Apr 24 '17 at 15:05

How to knock down a dam without explosives? Two words: Miley Cyrus

Miley Cyrus


How about a 'solar death ray'? Arguably a whole load of soldiers could spend a day or two polishing the insides of their shields and then hold them in a carefully choreographed way such that they each reflect a small amount of (bright?) sunshine to a single point on the dam. I'm not sure how you'd be able to ensure that every soldier's reflection was on the hot spot, but I guess you might be able to solve that by grouping them and having groups bring their reflections together at the end of the procedure. Even if it didn't work, you'd freak out the dam owners pretty badly.

It's been suggested Archimedes did this, although I'm not sure if it's true. Here's one attempt to recreate: http://web.mit.edu/2.009/www/experiments/deathray/10_ArchimedesResult.html


Heat it up, cool it off, heat it up, cool it off, heat it up, cool it off, heat it up, cool it off, heat it up, cool it off.

This seems stupid, but I assure you as someone who has some background in engineering if you stress out the material enough like this it will break. you just need to get it really hot and then get it cooled off quickly. I'm pretty sure this happened with a wall either in Egypt or Constantinople - I'm not really sure it was called that when it happened though. This method alone could break a damn by weakening its structure so much that it can no longer hold the load it was supposed to hold, but it would make more sense to use this method to try to weaken the damn before you hit it with a projectile battery.

Alternatively if this is too boring, you could just dig a bunch of holes. It's a more cartoonish solution, but it would work in theory. You could also just break all the support structures. The episode of Avatar the Last Airbender where team boomarang breaks the fire nation's drill to stop it from boring into the walled city displays this concept really well actually.


If you can go near the dam (meaning there is no secrecy involved, no guards, the dam is far from a village, etc. etc.), and if you want a cheap, quick (a few hours total), 100% reliable, 100% risk free method, then it's extremely easy to do.

Just use dry wood wedges:

  1. You need a boat, a bunch of dry wood wedges, hammer and chisel, and a bucket
  2. row to the middle of the dam
  3. use hammer and chisel to do a few holes -two or three should be enough, and no need to go through the entire dam, just go half the depth of the wedges.
  4. Plant the wedges inside the holes
  5. Fill the bucket with water and wet the wedges. Keep doing it till you start hearing crack noises coming from the stone
  6. Row away
  7. Just watch the dam getting teared in pieces.

Keep in mind that if you have time or resources you can skip the bucket part and either have a tiny bit of water flowing outside the holes where you put the wedges, or use a bronze tube to move water from the top of the dam to the wedges.

But in the end, whatever solution you choose to wet the wedges, you'll get your dam destroyed. And this is not hard science nor something unthinkable in that age, as we (humans being, I mean) have used this exact method since the dawn of humanity to tear apart mountains...


Try a ballista - https://en.wikipedia.org/wiki/Ballista

when it hits a tree or a rock, it pierces it easily. Such is the engine which bears this name, being so called because it shoots with very great force


Tesla supposedly brought down a building with a device about the size of a shoebox. Supposedly it used resonant vibration to compromise the structure over time - it was also attached to a metal portion of said structure.

As you are using the science-based tag, this might be enough to go by.

Also, considering the technological feats already accomplished in China, India, and the Americas around that time, it is not too far a stretch. Not to mention that the local neighbors, the Greeks, were building metal clockwork computers around that time. Just bring in an expert (foreigner with a briefcase... or, given the time period, a satchel or similar).

Even without using an electrical based device such as Tesla's, anything which produces rhythmic vibrations at the correct frequency would do the trick. A clockwork hammer mechanism might be able to pull it off, and has the advantage of the tech already having been worked out by the neighbors (again, the Greeks).

  • 7
    $\begingroup$ The middle Roman Empire didn't have electrical motor technology or use metal for construction, the last time I checked. $\endgroup$ – Nij Apr 23 '17 at 2:53
  • $\begingroup$ @Nij Might want to look into things a little further East. More locally, the Greeks were building computers around that time, so.... Shrug. $\endgroup$ – nijineko Apr 23 '17 at 19:43
  • 1
    $\begingroup$ There's a hell of a long way between custom-built handcrafted mechanical analogue orrery and a high speed electric motor with durability to establish structural resonance. Affirming a -1 for not making sense when the question is science-based. $\endgroup$ – Nij Apr 23 '17 at 21:00
  • 1
    $\begingroup$ ninjas in japan were already using self-driving cars by 1AD so it doesn't surprise me $\endgroup$ – Weyland Yutani Apr 25 '17 at 10:59
  • 1
    $\begingroup$ not to mention that while there is no evidence of smart phone use, archaeologists have found many ancient greek tablets $\endgroup$ – Weyland Yutani Apr 25 '17 at 11:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.