There’s a a decent sized kingdom that is situated between the ocean and a gigantic dried out seabed. The Priest-King and his lackeys want to flood the useless desert because who wouldn’t want more water access and bragging rights.

The only issue is that they live in a world that’s in the Bronze Age. There’s no excavators or dynamite like they had to dig the Suez or Panama Canal: all they have is human muscles.

Thus, how long would it take to dig a 10 kilometer canal across rocky soil with just hand tools and animal traction? The canal's width is 100 meters and the depth is 10 meters. There is no significant change in elevation.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – Monty Wild
    Commented Jan 30, 2020 at 9:55

5 Answers 5


Fun trivia: the word canal, and related words like cane, are most likely the only words in modern English derived from the ancient Sumerian language (from 𒄀𒈾, "qi.na") - spoken by a culture that has not just been extinct for four thousand years, but it was completely forgotten until they dug them up again in the late 1800's. Then it took another half century to figure out they were not just the later Akkadians, who are also extinct but at least referenced in the Bible.

Sumerian city-states were very much in the Bronze Age, and yes, they built canals. Lots of them in fact! Luckily, this culture also had a writing system (in fact, they invented writing, effectively ending the prehistory), and they wrote down, directly or indirectly, how much time the various activities took, as a way to record their economy. This is the resource as close to the source as you can get, as no other early Bronze Age cultures had writing systems. You might want to ask a question like this on History.SE as well.

Anyway, this page provides some incredibly detailed resource and time estimations for a canal. It also shows that there's a lot more to canal-building than just digging. The sides were lined with bricks and bitumen to prevent leakage, clay for those bricks had to be dug, reeds harvested to fire those bricks, etcetera. It's quite a process.

For excavation only: the canals were dug in three "levels". The first, digging up to 0.75 metre deep, allowed for one labourer to move 6 cubic metres in a working day of twelve hours. For the second level, 0.75 to 1.5 metre depth, only 3 m³/day was possible. And for even deeper, 2 m³/day was the max. That can be because of either soil hardness or because it takes more effort to move up the dirt.

So, assume your canal is 6 metres wide, which is on the small side for Sumerian canals but doable, especially if this canal is not meant to be immediately navigable. It would take 7500 man-days for the first level, 15000 man-days for the second level, and if you make your canal 2 metres deep, another 15000 man-days for the third level. That's 37500 man-days in total, or 150 days if you've got a labour force of 250 men who do nothing but digging.

But, if you want this canal to be any more permanent, you have bake bricks, boil bitumen, and so on. See the website for more info about that.

Alternatively, if all you care about is getting water from A to B, without caring about navigability, then the width and depth of @Harper's answer (2m wide, 0.5 m deep) would be appropriate. Those dimensions would require only 1667 man-days to excavate, which 10 people can do in half a year.

On the other hand, the humongous 100 metre wide, 10 metre deep you now specified - well, that's 4.6 million man-days. Lagash, one of the biggest Sumerian cities, had a population of around 50 thousand. You would need to recruit one tenth of the population to do this within five years. And I am unconvinced the numbers scale well for something so many times larger than anything made at the time. If you insist on these measurements, count on this being the work of a generation.

  • 8
    $\begingroup$ Another example of canals that date to antiquity (though I'm not sure how to date China's Bronxe Age exactly) is the Chinese Grand Canal, parts of which date to the 5th century BCE: en.wikipedia.org/wiki/Grand_Canal_(China) And if one wants to flood a below sea level desert, lining the canal is counterproductive. You just dig a narrow canal, and let erosion widen it. See e.g. Salton Sea: en.wikipedia.org/wiki/Salton_Sea $\endgroup$
    – jamesqf
    Commented Jan 25, 2020 at 17:49
  • 1
    $\begingroup$ @jamesqf Looks like that was at the very tail end of the Chinese Bronze Age, at the cusp of the Iron Age. And fair point, though i had assumed that you would not want to spend so much effort on making a new lake, without also giving it a navigable access so boats can make use of it. $\endgroup$
    – KeizerHarm
    Commented Jan 25, 2020 at 18:04
  • 3
    $\begingroup$ Your historical background is good, but you are missing some geology. You don't need bricks or water tightening for a canal that has siginificant flow and an unlimited water supply. You also don't really dig deep, the only thing you need is to initiate the flow, then erosion will do the rest. $\endgroup$
    – Nobody
    Commented Jan 25, 2020 at 21:53
  • 2
    $\begingroup$ @Nobody You are correct. I added the answer for Harper's dimensions for a canal that would purely function to carry water from A to B, without regards for navigability. $\endgroup$
    – KeizerHarm
    Commented Jan 27, 2020 at 9:45
  • 9
    $\begingroup$ As a side note, how not to be amazed by today's web, where someone can type 𒄀𒈾 in Sumerian and I see this correctly and not as �� like on the ol' good times... $\endgroup$
    – WoJ
    Commented Jan 27, 2020 at 15:16

You don't need a huge canal, just a little one. Once the water starts moving, it's going to dig the canal it wants!

So you'll want to build a full-design-size system of headgates at the ocean side, so you can shut the darn thing off when it starts misbehaving.

After that, you only need to cut a channel wide and deep enough that 100% of the water doesn't evaporate while flowing the 10km. So 10cmx2cm (4"x1") won't do, but 200cmx50cm (7'x2') will be more than adequate. We're talking British canal lock.

At that point it's about directing and managing erosion. Getting it to scour where you want it to, stopping it and rock-walling or cementing where it should not, etc.

  • $\begingroup$ Is there a citation regarding canals eroding? As far as I know (eg. canals local to me), Canals are dredged pretty regularly, so this answer is completely incorrect, at least in many cases. $\endgroup$
    – pjp
    Commented Jan 30, 2020 at 16:20
  • 1
    $\begingroup$ @pjp I'm not saying those kinds of canals erode. Canals are human-engineered specifically not to, as that would be a grievous failure in canal design for a canal to erode after completion. I'm telling OP to use the considerable difference in water height (head) which exists naturally, to use water force to carve the canal in the first place. Remember OP is not after a navigation canal, but a water-level-equalization canal. $\endgroup$ Commented Jan 30, 2020 at 16:31

After his wife died due to poor access to medical care in his village, Dashrath Manjhi spent 22 years single-handedly carving a road through a hill 110m long, 9.1m wide, and up to 7.7m deep, using only a hammer, chisel, and other simple tools, such as fires and cold water to crack the stone via thermal shock.

So, one man, 110m x 9m x 3.5m*, 22 years.

Now just adjust those numbers to account for the length, width, and depth required for the canal and the amount of manpower that the Priest-King can bring to bear on the task.

You'll probably then want to bring the estimated time down some, since a government with a large workforce on the project could leverage much more effective tools -- they wouldn't all be out there working with individual hammers and chisels.

(* - since the hill he was cutting through rose to a peak, I'm just averaging the height)

  • 9
    $\begingroup$ There's also a considerable difference in the amount of effort needed to dig through "rocky soil", as described in the question, and actually cutting through solid rock for the entire distance. I'd expect an estimate based on the numbers you've provided to be dramatically higher than what would actually be required, even if there was no difference in the available tools. However, what that difference would be will depend on what the OP is actually visualizing wrt. the soil in the area which is being excavated. $\endgroup$
    – Makyen
    Commented Jan 25, 2020 at 15:38
  • 5
    $\begingroup$ That's a little less than 1 m^3/day, which is definitely in the ballpark provided KeizerHarm, and considering rock instead of soil. $\endgroup$ Commented Jan 25, 2020 at 22:22
  • 1
    $\begingroup$ @LawnmowerMan Plus there's the fact that Sumerians worked 12 hours a day. I assume Mr Manjhi had more to spend his time on. $\endgroup$
    – KeizerHarm
    Commented Jan 27, 2020 at 9:46

I would depend to a great extent on the nature of the rocky soil. If it’s fragments of rock that can be shovelled then that’s one thing but if it’s solid granite that’s something else entirely. It would also depend on the required depth and width not only the length.

Assuming that its loose material and not solid rock one technique that might be used would be to dig a narrow channel and let the sea wash through it and expand it by erosion. But ball park figure assume the initial channel is 10m deep, 10m wide and 10,000m long. That’s 1 million cubic metres. If five men can move 1 cubic metre in a day (a big if, but perhaps, the spoil needs to be moved and may be hard to excavate) that’s five million man hours or ten thousand men working 8 hours a day for a couple of months.

However if there’s solid rock it could easily take several orders of magnitude longer.

  • 1
    $\begingroup$ This. The historical data doesn't really apply, they didn't have an unrealistic empty seabed that would allow erosion to do most of the work. I suspect there would need to be some solid rock from a geological perspective, because otherwise water would have already found a way on its own. But you could still let erosion do a lot of the work. $\endgroup$
    – Nobody
    Commented Jan 25, 2020 at 21:47
  • $\begingroup$ It shouldn't be "too loose" (like sand) either, as this will complicate the canal walls actually remaining in place. This would require either construction of walls (with deep enough foundations), or a very shallow slope, making the amount of material to shovel out much higher for the same useful width/depth. $\endgroup$
    – jcaron
    Commented Jan 27, 2020 at 11:53
  • $\begingroup$ Very good answer, the difficulty in building canals isn't primarily the length, it's the material you are going through along with the environmental conditions of the site. That was exactly the mistake the suez canal designer made with panama. "However if there’s solid rock it could easily take several orders of magnitude longer." Or be completely impossible. I doubt very much you could build the panama canal regardless of how much time and people you threw at it. The current design would be flat out impossible with BA tech. $\endgroup$
    – eps
    Commented Jan 27, 2020 at 16:47

It depends on the size of your labour force how hard you want to work them. Using the figures from https://en.m.wikipedia.org/wiki/White_Sea%E2%80%93Baltic_Canal I reckon that you could build a 10km canal using hand tools in about a month with 100000 labourers, if you don't mind killing about 1000 of them.

  • 1
    $\begingroup$ Or hire Kiewit, and you'll have zero injuries. Because Kiewit. $\endgroup$ Commented Jan 25, 2020 at 22:42
  • $\begingroup$ It's going to depend entirely on what you need to go through and how deep it needs to be dug. Quite a few types of challenges encountered when building canals would probably render the project impossible with bronze age tech. $\endgroup$
    – eps
    Commented Jan 27, 2020 at 16:52

You must log in to answer this question.