Roman aquaducts are famous, but didn't the water freeze during winter? Where did the cities get their water during winter?

EDIT: running water doesn't freeze as easily as motionless water, which leads to an important bit of info we need: how fast did the water run in Roman aquaducts? We might also need to know volumetric flow rate.

EDIT 2: Of course rivers don't freeze solid. I know it's only a surface or skin of ice. However, once the surface freezes, how did they access the liquid water underneath it? How did they pump that into their aquaducts?

  • 8
    Rome, and most of Italy, do not usually go below freezing. Also, I believe it was warmer back then anyway. In colder regions Romans did build aqueducts underground, which helped keep the water warmer.
    – Semaphore
    Commented Feb 22, 2016 at 8:19
  • 4
    The temperature in the Mediterranean area is not that low even in winter.
    – liftarn
    Commented Feb 22, 2016 at 8:20
  • 3
    In Mediterranean climate water does not usually freeze in winter.
    – Alex
    Commented Feb 22, 2016 at 13:01
  • 2
    I'm not sure why the assumption is that this question is about Italy-- The Romans obviously held a lot of territory where running water did freeze in the winter. @Semaphore's offhand comment about building aqueducts underground (if substantiated) is the answer to the question. The other comments are adding nothing to the discussion.
    – Era
    Commented Feb 22, 2016 at 20:40
  • 2
    Not long enough to be an answer: Water is a poor conductor of heat. Water also is one of the few substances that increase in density near the freezing point. The combination of the two plus a moderate flow rate means that rivers don't freeze solid, even as far north as Alaska, northern Canada, and Siberia. The cold climes of northern Germany are rather mild compared to that (and Rome never made it to northern Germany). Commented Feb 24, 2016 at 7:37

2 Answers 2



Aqueducts did, in some cases, freeze over. Aqueducts were built underground in the northern provinces to even out the temperatures over the day, and this helped a bit.

http://staff.civil.uq.edu.au/h.chanson/rom_aq.html http://www.eng-tips.com/viewthread.cfm?qid=262678

Flow rates are 5 to 70 L/s on modern models of Roman dropshafts. However, you can't prevent water from freezing over long distances just by increasing the flow rate. Presumably there was no running water if it became too cold.

  • 2
    Also consider that the whole length of a typical aquaduct wasn't the arched, above ground structure we see in pictures. For much of its length, it would have been a (possibly covered) ditch or tunnel. Thus the water, being in contact with the ground, would have been at some temperature above freezing. So when a parcel of water enters the arched section, it must cool to freezing, then give up its latent heat of freezing, before it becomes solid. But if the aquaduct's well-designed, it gets to the other side before that happens.
    – jamesqf
    Commented Feb 28, 2016 at 19:35

Roman aqueducts didn't freeze largely because they were in Italy. It's quite warm there, check climate data for Rome and other cities on wikipedia.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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